(Cambridge Manuals in Archaeology) R Lee Lyman-Vertebrate Taphonomy-Cambridge University Press (1994)

CAMBRIDGE MANUALS IN ARCHAEOLOGY
vertebrate
Taphonomy
T ap h o n o m y studies the tran sitio n o f organic m atte r from the biosphere into the
geological record. It is p articularly relevant to zooarchaeologists an d paleobiolo-
gists, w ho analyze o rganic rem ains in the archaeological record in an attem p t to
reconstruct hom inid subsistence p attern s an d paleoecological conditions. In this
user-friendly, encyclopedic reference volum e for stud en ts and professionals, R. Lee
Lym an, a leading researcher in tap h o n o m y , reviews the wide range o f analytical
techniques used to solve p articu lar zooarchaeological problem s, illustrating these in
m ost cases w ith ap p ro p ria te exam ples. He also covers the history o f taph o n o m ic
research an d its philosophical underpinnings. Logically organized an d clearly
w ritten, the book is an im p o rta n t u p d ate o n all previous publications on arch aeo
logical faunal rem ains.
VERTEBRATE TAPHONOMY
C A M B R I D G E M A N U A L S IN A R C H A E O L O G Y
Series editors
D on B rothw ell, University o f London

G raem e B arker, University o f Leicester
D ena D incauze, U niversity o f M assachusetts, A m herst
A nn Stahl, State University o f N ew York, Binghamton
A lready published
J.D . R ichards an d N .S. R yan, D ata processing in archaeology

Sim on H illson, Teeth
Alwyne W heeler and A ndrew K .G . Jones, Fishes
Peter G . D orrell, Photography in archaeology and conservation
Lesley A dkins and R oy A dkins, Archaeological illustration
M arie-A gnes C o urty , Paul G o ldberg and R ichard M acP hail, Soils and m icrom orphology it
archaeology
Clive O rto n , Paul Tyers an d A lan Vince, P ottery in Archaeology
Cambridge M anuals in Archaeology are reference h an d b o o k s designed for an in tern atio n al

audience o f professional archaeologists an d archaeological scientists in universities, museu
research laboratories, field units, an d the public service. E ach book includes a survey o f
cu rrent archaeological practice alongside essential reference m aterial on co n tem p o rary
techniques an d m ethodology.
VERTEBRATE TAP H O NO M Y
R. Lee Lyman
Departm ent o f Anthropology
University o f Missouri-Columbia
\ C a m b rid g e
|f |p U N IV E R S I T Y P R E S S
Published by the Press Syndicate o f the U niversity o f C am bridge
The Pitt Building. T ru m p in g to n Street, C am bridge CB2 1RP
40 W est 20th Street, N ew Y o rk , N Y 10011-4211, USA
10 S tam ford R oad, O akleigh, V ictoria 3166, A ustralia
© C am bridge U niversity Press 1994
First published 1994
P rinted in G reat B ritain at the U niversity Press, C am bridge
A catalogue record fo r this book is available fr o m the British Library
Library o f Congress cataloguing in publication data
L ym an, R. Lee.
V ertebrate tap h o n o m y / R. Lee Lym an.
p. cm. - (C am bridge m anuals in archaeology)
Includes bibliographical references an d index.
ISB N 0 521 45215 5 (hard). - ISBN 0 521 45840 4 (pbk.)
1. A nim al rem ains (A rchaeology). 2. T ap h o n o m y . 3. V ertebrates.
I. Title. II. Series.
C C 79.5.A 5L96 1994 93-28675
930.1'0285-dc20 C IP
ISBN 0 521 452155 h ard b ack

ISBN 0 521 458404 paperback
SE
To Barbara, John, and Michael
CONTENTS
page
L ist o f fig u res xiii
L ist o f tables xx
Preface xxiii
A cknow ledgem ents xxvi
1 W H A T IS T A P H O N O M Y ? 1
s * In tro d u c tio n 1
O n the analysis o f arch aeo lo g ical fa u n al rem ains 2
Basic concepts 3
G o als o f ta p h o n o m ic analysis in zo o arch aeo lo g y 5
T he challenge o f ta p h o n o m y 6
T a p h o n o m y ’s c o n trib u tio n to zo o arch aeo lo g y 7
T erm in o lo g y used in this b o o k 8
W h a t this b o o k is a n d w h a t it is n o t 9
2 TH E HISTORY A N D S T R U C T U R E OF
TAPHONOMY 12
A b rie f h isto ry o f ta p h o n o m ic research 12
O n th e stru ctu re o f tap h o n o m y : a p erso n al view 34
S um m ary an d con clu sio n 39
3 T A P H O N O M Y IN P R A C T I C E A N D T H E O R Y 41
In tro d u c tio n 41
E xam ples o f tap h o n o m ic analysis 41
U n ifo rm itaria n ism a n d actualism 46
A ctu alism in arch aeo lo g y a n d ta p h o n o m y 52
A n alogy 64
S u m m ary 68
4 STR U C TU R E AND Q U A N TIFIC A TIO N OF

VERTEBRATE SKELETONS 70
O n to g en y a n d allo m etry 70
Skeletal tissues 72
IX
x C ontents
P ro p erties o f skeletal tissues a n d ta p h o n o m y 82

V e rteb ra te skeletons 87
M od ificatio n o f skeletal tissues a n d tim e o f d ea th 95
Q u an tificatio n 97
S u m m ary 112
5 VERTEBRATE MORTALITY,
SKELETONIZATION, DISARTICULATION,
AND SCATTERING 114
-a* M odes o f d e a th 115
^ T he d em o g ra p h y o f m o rta lity 115
T he seasons o f m o rta lity 132
S k eleto n izatio n a n d d isarticu latio n 135
A nalysis o f d isarticu latio n a n d scatterin g 150
S u m m ary 160
6 A C C U M U L A T I O N AN D DISPERSAL OF
VERTEBRATE REMAINS 161
D ispersal, scatterin g , a n d a c cu m u latio n 161
A naly zin g dispersal 168
A naly zin g ac cu m u latio n 189
A c cu m u latio n a n d dispersal as m irro r im ages 219
S um m ary 220
7 FREQ UENCIES OF SKELETAL PARTS 223

H u m an u tilizatio n a n d tra n sp o rt o f carcass p a rts 223
S tru ctu ral density o f b ones 234
D ifferential tra n sp o rt versus differential su rv iv o rsh ip 258
W ith in -b o n e n u trie n ts 281
R eco n stru c tio n o f rav ag ed assem blages 283
O th e r sources o f v a ria tio n in bo n e stru c tu ra l density 288
A final co m m en t 289
S u m m ary 292
8 BUTCHERING. BONE FRA CTU RING, AND

BONE TOOLS 294
B utchering 294
F ra c tu rin g o f b ones 315
B one artifa cts 338
C ontents XI
B utchering, break ag e, a n d b o n e to o ls 350

S u m m ary 352
9 OTHER BIOSTRATINOMIC FACTORS 354

W e ath erin g 354
R o o t etching 375
T ra m p lin g 377
A b ra sio n 381
B u rning 384
O th e r biological ag en ts o f b o n e m o d ificatio n 392
P rese rv atio n a n d size biasing 397
C o m p a ra tiv e an aly tic techniques 398
S u m m ary 402
10 B U R I A L AS A T A P H O N O M I C P R O C E S S 404
D e p o sitio n a n d b u rial 406
S ed im en tatio n 406
B urial processes 413
S patial d istrib u tio n o f fau n al rem ains 415
S um m ary 416
11 DIAGENESIS 417
M in eralizatio n , leaching, en rich m en t 419
—> A nalysis o f chem ically altered b o n e 423
S edim ent o v erb u rd en w eight 423
P o st-b u rial m o vem ent 432
S um m ary 433
12 T A P H O N O M Y OF FISH, BIRDS, REPTILES,

AND AMPHIBIANS 434
Fish ta p h o n o m y 434
A vian ta p h o n o m y 446
R ep tilia n a n d a m p h ib ia n ta p h o n o m y 450
S u m m ary 450
13 DISCUSSION AND CONCLUSIONS 452

M u lti-v ariate ta p h o n o m ic analysis 453
A general th eo ry o f tap h o n o m y ? 463
C ontents
Bibliography
G lossary o f taphonom y term inology
In d ex
FIGURES
F igure
2 . 1. G e n eral re la tio n s o f the subdisciplines o f ta p h o n o m y re la page
tive to a n a n im a l’s life, d ea th , a n d scientific recovery. 17
2 .2 . M odeled tap h o n o m ic history o f a biotic co m m u n ity o r life
assem blage. 19
2.3. M ed lo ck ’s (1975) m odel o f the ta p h o n o m ic h isto ry o f a
/a u n a l assem blage. 24
2.4. F req u encies o f titles o f ta p h o n o m ic lite ratu re p er decade. 25
2.5. M e a d o w ’s (1981) m odel o f th e ta p h o n o m ic h isto ry o f a
fa u n al assem blage. 28
2 .6 . H esse an d W a p n ish ’s (1985) m odel o f a ta p h o n o m ic history
o f a zo o a rch aeo lo g ical assem blage o f fa u n a l rem ains. 29
2.7. B ehrensm eyer a n d K idw ell’s (1985) m odel o f a tap h o n o m ic
h isto ry w ith relatio n s o f subdisciplines o f ta p h o n o m y in d i
cated. 30
2 .8 . A n d rew s’ a n d C o o k ’s (1985) m odel o f a ta p h o n o m ic h isto ry
show ing stages o f m odification. 31
3.1. In tersectio n o f different kinds a n d intensities o f historic
(tap h o n o m ic) processes defining u n ifo rm itarian ism , a c tu a l
ism, a n d c a ta stro p h ism as p arad ig m s fo r explaining the
past. 50
3.2. S chem atic re p resen ta tio n o f the tra n sfo rm a tio n o f an a n i
m al fro m being a living o rg an ism to being a fossil show ing
w here p a rtic u la r bodies o f th eo ry are relevant, an d general
categ o ries o f tran sfo rm s a n d contexts. 65
3.3. A m odel o f (relatio n al) an alogical reasoning. 66
4.1. S chem atic illu stra tio n o f ossification a n d g ro w th o f e n d o
c h o n d ra l long b o n e (tibia) o f a m am m al. 71
4.2. S tru ctu re o f m am m alian b o n e a t different scales a n d levels
o f o rg an izatio n . 75
4.3. M ic ro stru c tu re o f m a m m alian b o n e show ing H a v ersian an d
lam ellar bone. 76
4.4. A p p e aran c e a n d d istrib u tio n o f tra b e c u la r a n d co m p act
bon e in a m am m alian long bone. 77
4.5. C ro ss section o f a typical m am m alian to o th show ing m ajo r
co m p o n en ts a n d regions. 80
xm
L ist o f figures
M o deled re latio n o f stress a n d strain, Y o u n g ’s m o d u lu s o f

elasticity a n d p o in t o f failure. 84
N o rth A m erican bison (Bison bison) skeleton. 88
G en eralized leporid o r ra b b it skeleton. 88
G en eralized teleost fish skeleton. 89
G en eralized frog (a m p h ib ian ) skeleton. 90
G en eralized tu rtle (reptile) skeleton. 91
G en eralized sn ake (reptile) skeleton. 92
G en eralized bird skeleton. 93
D irectio n al term s fo r v e rte b ra te skeletons. 94
C h ro n o lo g ical re la tio n s o f b o n e on to g en y , b o n e rem o d e l
ing, d ea th , a n d tap h o n o m y . 95
N o rm e d M N I p er skeletal p o rtio n frequencies an d n o rm ed
M A U p er skeletal p o rtio n frequencies fo r p ro n g h o rn a n te
lope rem ains fro m 39FA 83. 107
B ivariate sca tte rp lo t o f M N I p er skeletal p o rtio n freq u en
cies a n d M A U p er skeletal p o rtio n frequencies fo r p ro n g
h o rn an telo p e rem ains fro m 39FA 83. 108
B ivariate sc a tte rp lo t o f M N I p er skeletal p o rtio n freq u en
cies fo r left a n d right skeletal p o rtio n s o f p ro n g h o rn a n te
lope from 39FA 83. 109
T w o basic types o f age (m o rtality ) profiles. 119
A ge (m o rtality ) profiles fo r a p o p u la tio n w ith high m o rtality
a n d recru itm en t. 122
M o rtality profile fo r fossil horses. 123
M o rta lity profile fo r fossil antelope. 123
M o rta lity profile fo r arch aeo lo g ical deer rem ains. 124
M o rta lity profile fo r arch aeo lo g ical p ro n g h o rn an telo p e
rem ains. 125
E xpected a n d observed m o rta lity profile fo r w apiti killed by
the volcanic e ru p tio n o f M o u n t St. H elens. 126
T hree-po le g ra p h in g tech n iq u e fo r assessing d em o g ra p h ic
(m o rtality ) d a ta . 129
M o rta lity profiles fo r A frican bovid rem ains from K lasies
R iver M o u th a n d E lan d sfo n tein . 130
T hree-po le g ra p h o f m o rta lity d a ta from K lasies R iver
M o u th a n d E lan d sfo n tein . 131
S easonality a n d m o rta lity profiles fo r deer (Odocoileus spp.)
rem ains from archaeological site 4 5 D 0 1 8 9 . 133
S easonality a n d m o rta lity profiles fo r deer (O docoileus spp.)
rem ains from arch aeo lo g ical site 4 5 D 0 1 7 6 . 134
A p a rtia l, artic u la te d w apiti sk eleto n in situ. 136
B lum enschine’s (1986a, 1986b) c o n su m p tio n sequence p lo t
ted ag a in st flesh w eight. 149
L ist o ffig u re s xv
5.15. O rd e r o f jo in t d isarticu latio n at C asp er a n d H o rn e r II as

d eterm in ed by H ill’s (1979a, 1979b) m eth o d . 152
5.16. P ro p o rtio n o f artic u la te d jo in ts at C asp er a n d H o rn e r II as
d eterm in ed by T o d d ’s (1987b) m eth o d . 153
5.17. B ivariate sca tte rp lo t o f index o f skeletal d isju n ctio n an d
index o f frag m ent d isju n ctio n ag ain st stan d ard iz ed m eat
w eight yield for H o rn e r II bison. 158
5.18. B ivariate sca tte rp lo t o f index o f skeletal d isju n ctio n an d
index o f frag m en t d isju n ctio n ag a in st stan d ard iz ed m arro w
yield fo r H o rn e r II bison. 159
5.19. B ivariate sca tte rp lo t o f index o f skeletal d isju n ctio n against
index o f fragm ent d isju n ctio n fo r H o rn e r II bison. 159
6 . 1. T ypes o f b o n e occurrence based on m o rtality type (indivi
d u al, m ass), b o n e ac cu m u latio n agencies, tra n sp o rt, a n d
d u ra tio n o f a c cu m u latio n . 164
6 .2 . Classes o f bon e occurrence defined by d im ensions o f v a ria
bility in a c cu m u latio n agent (physical, biological), m o rtality
(single in d ivid u al, m ultiple individuals), ac cu m u latio n
ac tio n (passive, active), a n d d u ra tio n o f ac cu m u latio n
(sh o rt, long). 167
6.3. E q u id m o rtality profiles fo r M ag d a le n ian an d G ra v e ttia n
levels at S olutre, F ran ce. 169
6.4. F requ encies o f eq uid skeletal p a rts in the A u rig n acian level
o f S o lu tre, F ran ce. 170
6.5. C lassification o f b o n e dispersal g ro u p s acco rd in g to c u rre n t
velocity a n d p ro x im ity to the site w here bones begin tra n s
p o rt by fluvial actio n . 173
6.6. C lassification o f b o n e shape b ased on axial ratios. 177
6.7. A m irro r-im ag e rose d iag ra m show ing azim u th s o f long axis
o f long bones. 179
6 .8 . Idealized stereo g rap h ic p ro jectio n s o f fo u r possible d istri
b u tio n s o f long bo n e o rie n ta tio n a n d plunge o r dip. 182
6.9. A stereo g rap h ic p ro jec tio n o f the h o riz o n ta l a n d vertical
o rie n ta tio n o f five bones. 183
6 . 10. D istrib u tio n a n d o rie n ta tio n o f w apiti carcasses killed by
th e volcanic e ru p tio n o f M o u n t St. H elens. 184
6 . 11. A zim u th o f w apiti carcasses killed by the volcanic eru p tio n
o f M o u n t St. H elens. 185
6 . 12. B lum enschine’s (1986a) c o n su m p tio n sequence. 188
6.13. F req u en cies o f skeletal elem ents from carn iv o re kills an d
from a ca rn iv o re den p lo tte d ag ain st B lum enschine’s
(1986a) c o n su m p tio n sequence. 189
6.14. R elative frequencies o f skeletal p o rtio n s in different types o f
b o n e ac cu m u latio n s. 191
XVI /./.st o f figures
6.15. R o d e n t gnaw ed bones. 196-7

6.16. R a tio o f p o st-cran ial to cran ial skeletal p a rts accum ulated
an d d ep o sited by 19 species o f ra p to rs a n d m am m als. 202
6.17. R a tio o f distal to p ro x im al lim b elem ents ac cu m u lated an d
d ep o sited by 19 species o f ra p to rs a n d m am m als. 203
6.18. P ro p o rtio n o f co m p lete lim b elem ents in assem blages accu
m u lated by selected ra p to rs a n d m a m m alian carnivores. 204
6.19. R agged a n d c ren u lated edges resulting fro m m am m alian
carn iv o res gnaw ing m o d ern w apiti bones. 207
6 .2 0 . P ittin g a n d pu n ctures. 208
6 .21. P u n ctu res. 209
6 . 22 . F u rro w o n a m o d ern w apiti p ro x im al fem ur. 210
6.23. S cooping o u t on tw o distal fem ora. 211
6.24. D igestive co rro sio n o f first ph alan g es o f dom estic sheep. 211
6.25. C o m p a riso n o f diam eters o f p u n c tu re m ark s on sm all
m am m al b ones collected from a ro ckshelter, a n d the range
o f canin e d iam eters o f m o d ern carnivores. 214
6.26. A ttrib u tes o f m o d ificatio n to prey bones created by various
A frican carniv o res. 215
6.27. B ivariate sca tte rp lo ts o f relative frequencies o f b ones from
sm all m am m als a n d from large m am m als on the A frican
lan d scap e ag ain st b o n e frequencies in a h o m in id settlem ent. 221
7.1. A fam ily o f strategies fo r utilizing a n d /o r tra n sp o rtin g
an im al carcass p arts. 228-9
7.2. S catterp lo t o f ca rib o u % M A U values fro m A n av ik against
ca rib o u % M G U I values. 231
7.3. S catterp lo t o f B rain’s (1969) g o at b o n e stru c tu ra l density
values ag ain st the n u m b e r o f recovered g o a t b o n e specim ens
from a H o tte n to t village. 236
7.4. A n ato m ical lo catio n s o f scan sites w here p h o to n a b so rp tio
m etry m easu rem en ts have been ta k e n on u n g u late bones. 240-1
7.5. A n a to m ica l lo catio n s o f scan sites w here p h o to n a b s o rp tio
m etry m easu rem en ts have been ta k e n on m a rm o t bones. 242-3
7.6. A n a to m ica l lo catio n s o f scan sites w here p h o to n a b s o rp tio
m etry m easu rem en ts have been ta k e n on seal bones. 244-5
7.7. S catterp lo t o f % su rv iv o rsh ip o f deer skeletal p a rts from
4 5 0 K 4 ag ain st bo n e m ineral density values fo r deer. 249
7.8. S catterp lo t o f frequency o f individual scan sites in one
skeleton ag ain st bo n e m ineral density values fo r deer. 251
7.9. S c atterp lo t o f M A U frequencies o f m a rm o t skeletal p a rts
from the W hite M o u n ta in s ag a in st bo n e m in eral density
values fo r m arm o ts. 254
7.10. S catterp lo t o f M A U frequencies o f m a rm o t skeletal p a rts
L ist o f fig u res xvii
fro m the S alishan M esa site ag a in st bo n e m in eral density

values fo r m arm o ts. 255
7.11. S catterp lo ts o f g u an a co u tility indices ag a in st g u an aco bone
density. 259
7.12. S catterp lo ts o f % M A U frequencies o f deer-size an im al
rem ain s ag a in st the s tru c tu ra l density o f deer bones, the
% M G U I fo r sheep, a n d the % M G U I fo r caribou. 262-3
7.13. All possible c o m b in a tio n s (classes) o f c o rrelatio n coeffi
cients betw een the % M A U o f a b o n e assem blage, a n d b o th
b o n e density an d % M G U I. 264
7.14. S catterp lo ts o f % su rv iv o rsh ip o f skeletal p a rts afte r ra v a g
ing by h yenas ag ain st sheep bo n e s tru c tu ra l density an d deer
b o n e s tru c tu ra l density. 268
7.15. V a riatio n in sca tte rp lo ts o f % su rv iv o rsh ip o f skeletal p arts
afte r ravaging by hyenas against sheep % M G U I fo r long
b o n e ends a n d fo r long bo n e sh aft ends. 269
7.16. S catterp lo ts o f M N E frequencies from fo r F L K Z injanthro-
p us assem blage. 272
7.17. B ar g ra p h o f % w eig h t loss o f cow b ones over time. 279
7.18. N IS P -to -M N E ra tio s p lo tte d ag ain st w ith in -b o n e n u trien t
index fo r tw o taxa. 283
7.19. S catterp lo ts o f ca rib o u b o n e observed a n d re co n stru cted
frequencies ag ain st the c a rib o u % M G U I. 287
7.20. S ta n d ard iz ed fo od utility index fo r com plete bones p lo tte d
ag a in st th e % M A U o f surviving sheep bones. 292
8.1. E xam ples o f cut m arks. 305
8.2. D istal m etap o d ials show ing lo catio n s o f variously d o c u
m en ted cu t-m ark s. 310
8.3. P ro p o rtio n a l frequencies o f cu t-m a rk e d specim ens in
selected a n a to m ica l categories. 311
8.4. F ra c tu re types d escribed by S hipm an et a!. (1981), w ith
m o d ifications by M arsh a ll (1989). 319
8.5. F ra c tu re edge m o rp h o lo g y o f a b ro k e n m etac arp a l illus
tra te d using B iddick a n d T o m e n c h u k ’s (1975) system o f
p o la r co o rd in a tes a n d vertical planes. 322
8.6. F eatu re s o f fractu re surfaces show n o n a b o v id p ro x im al
m etacarp al. 323
8.7. L o ad in g points. 327
8.8. B ar g ra p h s o f th ree bo n e fra g m e n ta tio n a ttrib u te s fo r three
assem blages. 330-1
8.9. V a riatio n in th e p ro p o rtio n o f com plete skeletal elem ents
betw een tw o tax a o f owls. 334
8.10. P ro p o rtio n a l frequencies o f 1 cm size classes o f long bone
xviii L ist o f figures
diaphysis frag m en ts fo r tw o assem blages o f deer bones. 335

8 . 11. A m odel o f th e re latio n betw een N IS P a n d M N E in an
assem blage o f bones. 336
8 . 12 . P reh isto ric scap u la aw ls fro m eastern W ash in g to n . 341
8.13. P seu d otoo ls. 342
8.14. S catterp lo t o f M N E frequencies o f selected bison bones
ag ain st the bison food utility index. 349
8.15. D e m o g rap h y o f m o rta lity o f m a sto d o n carcasses re p o rted
by F ish er (1987). 350
8.16. Season o f m o rta lity o f m a sto d o n carcasses re p o rted by
F isher (1987). 351
9.1. Bone w eath erin g stages described by B ehrensm eyer (1978). 356-7
9.2. W e ath erin g profiles fo r carcasses d ead 0.5 to 1 yr, carcasses
dead 2.5 to 3 yr, carcasses d ead 4 to 10 yr, a n d carcasses dead
10 to 15 yr. 365
9.3. W e ath erin g profiles fo r tw o assem blages o f bones. 368
9.4. F req u en cy d istrib u tio n o f percentages o f b ones p er w e ath er
ing stage in th ree assem blages. 370
9.5. T h ree-p o le g ra p h o f bo n e w eath erin g d a ta fo r six assem
blages fro m O lduvai G o rg e a n d co n tro l assem blages o f
carcasses d ead fo r k n o w n n u m b ers o f years. 372
9.6. C u m u lativ e percen t frequency d istrib u tio n s fo r w eathering
stages o f b ones in sum m ed assem blages o f O lduvai G o rg e
thin d ep o sit sites a n d sum m ed assem blages o f O lduvai
G o rg e th ick d ep o sit sites. 373
9.7. R o o t etching o n a sheep m andible. 376
9.8. V ertical frequency d istrib u tio n o f tram p led artifacts. 378
9.9. S um m ary o f changes to bo n e subjected to heating. 386
9.10. C u m u lativ e p ercent o f w eight loss o f fresh an d b u rn e d bones
placed in acid. 390
9.11. R egression o f log o f live w eight ag ain st log o f the ra tio o f
n u m b er o f in dividuals expected to n u m b er o f individuals
observed. 397
9.12. S catterp lo t o f % differences in frequencies o f p ro x im al an d
distal h u m eri ag a in st % differences in frequencies o f p ro x i
m al a n d distal tibiae. 400
9.13. B one d e stru c tio n graphs. 401
1 1 . 1. B ivariate sca tte rp lo t o f N IS P :M N I ratio s p er skeletal p a rt
fo r tw o b o n e assem blages. 428
11.2 . B ar g ra p h show ing v a ria tio n in com pleteness index values
across seven sm all, co m p ac t bones from tw o sites. 430
11.3. B ivariate sca tte rp lo t o f com pleteness index values fo r six
sm all, co m p ac t bones. 431
L ist o f figures xix
12.1. P ro p o rtio n a l frequencies o f salm o n id c ran ial an d post-

cran ial rem ains. 439
13.1. E xam ple o f grap h ic tech n iq u e fo r su m m arizing an d c o m
p arin g ta p h o n o m ic d a ta for m ultiple assem blages. 459
TABLES
K inds o f ta p h o n o m ic d a ta th a t should be recorded for page

v erte b rate fossil rem ains. 22
F requ encies o f m a jo r kinds o f skeletal elem ents in different
m am m alian taxa. 98
F L K Z injanthropus bovid lim b b o n e d a ta . 103
F requen cies o f p ro n g h o rn an telo p e skeletal p o rtio n s from
site 39FA 83. 106
O bserved a n d expected M N I frequencies o f p ro n g h o rn
an telo p e skeletal p o rtio n s from site 39FA 83. 109
Life tab le fo r fem ale H im alay a n th ar. 117
Life tables fo r tw o h y p o th etical p o p u la tio n s o f m am m als. 120
O bserved a n d expected frequencies o f w apiti from c a ta
stro p h ic m o rta lity resulting from volcanic e ru p tio n o f
M o u n t St. Helens. 125
M o rta lity d a ta fo r tw o fossil assem blages. 130
R a n k o rd e r o f jo in t d isa rtic u la tio n in five m am m alian taxa. 145
Sequence o f d am ag e to bones o f un g u lates exploited by
N o rth A m erican wolves. 148
R an k ed general c o n su m p tio n sequence. 149
J o in t artic u la tio n d a ta for bison b ones fro m the C asp er site
an d th e H o rn e r II site. 151
Index o f skeletal d isju n ctio n a n d index o f frag m en t d isju n c
tio n fo r the H o rn e r II bison rem ains. 157
D im ensions o f variability in the process o f bo n e ac cu m u
lation. 165
C lasses o f v aria tio n in bone ac cu m u latio n . 166
A lig nm ent o f types o f b o n e occu rrence w ith bone ac cu m u
latio n classes. 167
C riteria p ro p o sed by W heat (1979) for distinguishing kill
sites, processing sites, a n d c o n su m p tio n sites. 171
M am m alian skeletal elem ents g ro u p ed by th eir susceptibi
lity to fluvial tra n sp o rt. 172
K o rth ’s (1979) settling g ro u p s aligned w ith V o o rh ies' (1969)
groups. 174
xx
L ist o f tables xxi
6.7. F luvial tra n s p o rt index values a n d sa tu ra te d w eight index

values fo r vario u s taxa. 175
6 .8 . O bserved an d expected frequencies o f 1084 b o n e specim ens
per 10° o rie n ta tio n class a t L u b b o ck Lake. 180
6.9. T h ree-d im en sio n al o rie n ta tio n d a ta fo r five fictional long
bones. 183
6 . 10. G n aw in g d am ag e to b ones typical o f fo u r tax o n o m ic g ro u p s
o f m am m alian carnivores. 213
6 . 11 . F req u encies o f skeletal p a rts o f tw o sizes o f m am m als from
the lan d scap e a n d fro m a h o m in id settlem ent. 220
7.1. B in fo rd ’s (1978) no rm ed utility indices fo r dom estic sheep
a n d carib o u . 226
7.2. M N E a n d M A U frequencies o f ca rib o u bones fo r tw o
eth n o arch ae o lo g ic al sites. 230
7.3. U tility a n d tra n s p o rt indices fo r v ario u s taxa. 232
7.4. U tility indices fo r b o n e p a rts o f vario u s m am m alian taxa. 233
7.5. F req u encies a n d stru c tu ra l density o f g o a t bones, an d
m easures o f sheep b o n e density. 236
7.6. A verage bo n e m ineral densities fo r deer, p ro n g h o rn a n te
lope, dom estic sheep, bison, g u an a co , a n d vicuna. 246-7
7.7. A verage bo ne m in eral densities fo r m a rm o ts a n d ph o cid
seals. 248
7.8. F requ encies o f re p resen ta tio n o f scan sites o f deer bo n e
from arch aeo log ical site 4 5 0 K 4 . 250
7.9. M A U values fo r the W hite M o u n ta in s m a rm o ts a n d the
S alishan M esa m a rm o ts, a n d co rresp o n d in g scan sites fo r
stru c tu ra l den sity values. 253
7.10. T ra d itio n a l density scan sites a n d m axim um density scan
sites typically co rrelated w ith M A U values. 257
7.11. % M A U frequencies fo r deer-sized an im als fo r site
45C H 302. 260
7.12. M N E a n d % M A U frequencies fo r h y en a-rav ag ed dom estic
sheep bones. 266
7.13. F req u encies o f skeletal p a rts a t F L K Z injanthropus an d
co m plete bo ne u tility index values. 271
7.14. C o rre la tio n coefficients betw een percen t w eight loss o f
skeletal p a rts due to ca rn iv o re gnaw ing over tim e, a n d bone
stru c tu ra l density. 278
7.15. N IS P to M N E ra tio s fo r selected parts. 282
7.16. R eco n stru c tin g ca rib o u bo n e assem blages fro m N u n a m iu t
sites. 286
7.17. E x perim ental d a ta fo r b o n e tra n s p o rt a n d survivorship, an d
how th o se d a ta w ould be trea ted in an archaeological
con text. 290
L ist o f tables
C arcass resources ex p lo itab le by a fa u n al p ro cesso r o r

h u m a n butcher. 295
Selected carcass-processing activities directed to w ard s
ex tra ctin g co n su m ab le carcass resources. 295
F a c to rs th a t influence utilized b u tch erin g techniques. 296
N IS P a n d frequencies o f cu t-m a rk ed specim ens in the F L K
Z injanthropus assem blage. 308
F req uencies o f cu t-m a rk ed m eaty lim b specim ens a n d m eta-
p o d ial specim ens in the F L K Z injanthropus collection. 312
F req uencies o f cu t-m a rk ed specim ens in jo in t, a n d m eaty
lim b sh aft lo cations. 313
F ra c tu re classification system o f D avis (1985). 321
F req uencies o f fractu re a ttrib u te s in th ree assem blages o f
h u m a n bones. 329
F req uencies o f skeletal p a rts in ra p to r pellets. 332
M N E frequencies o f bison b ones recovered fro m the Phillips
R an ch site. 348
W e ath erin g stages in large a n d sm all m am m als. 355
K o lm o g o ro v -S m irn o v D statistics betw een all possible p airs
o f carcass assem blages fro m m a jo r h ab itats. 362
F req uencies o f w eathered b ones in six assem blages from
O ld uv ai G o rge. 371
F req uencies o f b o n e p a rts fro m selected sites. 399
S ta n d a rd sed im ent size classes. 407
D ep o sitio n al settings an d a ttrib u te s o f sedim ents an d sedi
m e n ta ry units. 408
R atio s o f N IS P :M N I p er skeletal p a rt in tw o assem blages. 427
T u rb a tio n processes influencing b u rial, exposure, an d
m o v em en t o f fossils. 432
A verage skeletal com pleteness ratio s fo r v ario u s sized h o ri
zo n tal units a n d sites. 440
S tru c tu ra l density o f co h o salm o n skeletal elem ents. 442
S u m m ary o f crite ria for d istinguishing cu ltu ra lly from
n a tu ra lly dep o sited assem blages o f fish rem ains a ro u n d
large lakes. 445
D im en sion s a n d a ttrib u te states fo r ta p h o n o m ic analysis. 457
D efinition o f v ariables a n d listing o f values p er plo tted
variable. 460
P RE F A C E
W hen I sta rte d m y studies o f v e rte b ra te fa u n al rem ains recovered from

arch aeo lo g ical sites over tw enty years ago, I h ad no idea w h a t ta p h o n o m y w as
n o r w as I p a rtic u la rly co n cerned a b o u t w h a t are to d ay typically asked
q uestio n s co n cerning th e p re serv atio n a n d fo rm a tio n o f the a rch ae o fau n a l
record. B ut as I read the zo o a rch aeo lo g ical lite ratu re w hile co m p letin g m y
d o c to ra l d issertatio n in the m id -1970s, I fo u n d an increasing n u m b er o f pap ers
dealing w ith ta p h o n o m ic issues. T h e fact th a t since then it h as becom e
increasingly difficult to keep u p w ith the ever grow ing lite ra tu re o n ta p h o n o m y
is so m eth in g o f a m ixed blessing. It is a m ixed blessing because (a) we are
co n stan tly realigning the re la tio n betw een w h a t we w ant to learn a n d w h a t we
think we can learn fro m the v e rte b ra te fa u n al rem ains we recover fro m
arch aeo lo gical sites, a n d th u s o u r conclusions tend to be m uch m o re strongly
fou n d ed th a n even a d ecade ago (this is good), a n d (b) it is nearly im possible for
any one a n aly st to conceive o f all o f the logically possible ta p h o n o m ic p roblem s
th a t a single re aso n ab ly sized assem blage o f v e rte b ra te rem ains m ight present.
T he la tte r is n o t b ad; it ju s t m ean s a ta p h o n o m is t’s an d zo o a rch a eo lo g ist’s (and
th u s m y) jo b is m u ch m o re difficult now th a n it w as a m ere decade ago. Sim ply
p u t, the analysis o f zo o a rch aeo lo g ical rem ains is no longer the sim ple,
s traig h tfo rw ard task th a t it w as in the 1960s o r 1970s. T a p h o n o m ic research
has fo u n d a h om e in zo o arch aeo lo g y , a n d it is here to stay.
T o d ay , th e n u m b e r o f zo o a rch a eo lo g ists w ho sim ply identify the b ones, tally
them up, an d w rite a re p o ri a b o u t w h at p re h isto ric hom inids w ere eating, is
dim inishing. M o st re p o rts on zo o a rch aeo lo g ical rem ains w ritten in the p a st ten
years c o n ta in a m o re o r less d etailed co n sid eratio n o f a t least a few tap h o n o m ic
issues. T his b o o k is a b o u t how ta p h o n o m ic qu estio n s m ight be analytically
ad d ressed an d , som etim es, answ ered. It is a b o o k th a t I w an ted to w rite ten
years from now . H ow ever, w hen A nn S tahl talk ed to me in the S pring o f 1991
a b o u t the possibility o f w riting it, I realized, u p o n reflection, th a t now (from
M ay 1991 u ntil Ja n u a ry 1993) w as ju s t as g o o d a tim e as later. In fact, th e m ore
I th o u g h t a b o u t it, th e b etter the idea o f w riting it now becam e. M an y o f my
friends a n d professional colleagues were w ork in g h a rd on im p o rta n t ta p h o n o
m ic pro b lem s, a n d virtu ally all o f them w ere eager to tell m e w h a t they w ere
w o rk in g o n a n d w h a t they w ere learning. W ritin g the b o o k w ould, I decided, be
easy because o f all o f these w onderfully know ledgeable people, a n d there
w eren ’t m o re o f th em th a n I could keep tra c k o f w ith a little effort. A ny value
XXlll
xxiv Preface
this b o o k has is a trib u te to all o f those people w ho know ingly a n d u n k n o w

ingly helped m e w ith p u ttin g it to g eth er. F o r being a friend an d tap h o s
colleague as 1 w ro te this b o o k , I th a n k D ian e G iffo rd -G o n zalez, D o n a ld K.
G ra y so n , S tep han ie D. L ivingston, F io n a M arsh all, D ave N . S chm itt, and
M ary C. Stiner. I especially th a n k Lee A nn K re u tzer fo r finding a n d sending me
a co up le o f re p rin ts at the last m inute, a n d keeping m e in fo rm ed a b o u t her
studies o f b o n e density. M a n y o th e r people have helped m e over the years by
review ing som e o f m y m an u scrip ts a n d by alw ays being read y to sh are ideas
a n d rep rin ts. F o r help in m an y w ays ta p h o n o m ic a n d zo o arch aeo lo g ic, I th a n k
A n n a K . B ehrensm eyer, R o b e rt L. B lum enschine, R o b so n B onnichsen, Luis
A. B o rrero , V irginia L. B utler, G a ry H aynes, Jean H u d so n , Eileen Jo h n so n ,
R ich ard G . K lein, C u rtis W. M arean , D u n c a n M etcalfe, R ich ard M o rlan ,
Jam es F. O 'C o n n e ll, Paul W. P arm alee. Jam es Savelle, P at S hipm an, G e n try
Steele, an d L aw rence C. T o d d . T here are, to be sure, m an y o th ers w hose talks I
have h ea rd a n d w hose p ap e rs I have read; they have, no d o u b t, influenced m y
th o u g h ts m o re th a n I realize.
P erm ission to re p rin t som e o f the illu stra tio n s th a t are critical to the volum e
w as pro v id ed by several individuals d o in g im p o rta n t ta p h o n o m ic research. T o
these ind iv iduals I can only say “ I ow e you o n e:” P eter A ndrew s, A n n a K.
B ehrensm eyer, Lewis R. B inford, J. D. C u rrey , D iane G iffo rd -G o n zalez. B rian
Hesse, Eileen Jo h n so n , Lee A n n K reu tzer, L arry G . M arsh all, R ichard H.
M ead o w , S tanley O lsen, T. B. P arsons, R ich ard P o tts, a n d M ary C. Stiner.
I have been given m an y o p p o rtu n itie s to analyze an d study a rch ae o fau n a l
rem ain s over th e years. W ith o u t th a t b re a d th a n d d e p th o f experience, this
b o o k w ould be m u ch less th a n it is, an d , I p ro b a b ly w ould n o t have w ritten it.
F ra n k C. L eo n h a rd y an d C arl E. G u sta fso n in itiated m y interest in bones, and
F ra n k gave m e th e assem blage o n w hich I cu t m y teeth. I am deeply saddened
th a t his untim ely d ea th prev en ted his being here to see w h at he helped create.
M y early interests w ere fine tu n ed by D o n a ld K. G ra y so n , w ho provided me
access to several u n iq u e collections (including the M o u n t St. H elens crispy elk)
an d w ho knew w hen to let m e figure o u t I w as h ead ed in the w ro n g direction
an d w hen to n o t w aste tim e a n d tell m e I was w rong. O th e r friends w ho
pro v id ed boxes o f bones fo r m e to study include K e n n eth M . A m es, D avid R.
B rauner. R ich ard L. B ryant, T erry Del Bene. D avid T. K irk p a tric k , D ennis E.
Lew arch, M ichael J. O ’Brien, K e n n eth C. Reid, an d R ich ard E. R oss. In
p a rtic u la r, Jerry R. G alm has, over the p ast decade, seen to it th a t I d id n 't go
m o re th a n six m o n th s w ith o u t receiving a box o f bones in the m ail; th an k s,
Jerry , for en su rin g th a t I d id n ’t have to suffer w ithdraw al.
M an y people helped in sm all b u t im p o rta n t ways. G ail L aw rence, A m y
K o ch , a n d R o b D u n n helped w ith som e o f the early w o rd processing. E ugene
M a rin o a n d Paul Picha helped w ith co rresp o n d en ce via the fax m achine.
V irginia L. B utler, Jam es C ogsw ell, D o lo res C. E lkin, D o n a ld K . G ra y so n , an d
P au l P icha v ariou sly helped m e o b ta in several h ard -to -fin d articles a n d b o oks,
Preface xxv
a n d k ep t m e u p -to -d a te w ith w h at w as com ing off the presses. M ichael B.

Schiffer w as alw ays read y to visit a n d offer en co u rag em en t; th a n k s, M ike, fo r a
copy o f the to m a to bo ok. G re g o ry L. F o x m ad e sure I w ent fishing occasionally
d u rin g th e early w riting p hases, a n d th e stu d en ts in m y z o o a rch a eo lo g y class
m ad e sure I identified th e big e rro rs in som e o f m y early reasoning. A n n S tahl
an d A n n a K. B ehrensm eyer pro v id ed helpful co m m en ts on an early d ra ft o f
several c h a p te rs a n d th ereb y m ad e sure I w as o n the rig h t trac k . A n n
sub seq u en tly read th e entire m an u scrip t, d o in g all o f us a m a jo r service by
en su rin g th a t the m o re cu m b erso m e sentences w ere revised. L in d en Steele
p rin ted som e o f th e p h o to g ra p h s. T he U niversity o f M isso u ri-C o lu m b ia,
D e p a rtm e n t o f A n th ro p o lo g y helped get som e figures re p ro d u ced .
F inally, w ith o u t the faith in m y ability show n by m y wife, B arb a ra, a n d the
d istractio n s p ro v id ed by m y sons J o h n a n d M ike, it never w o u ld have been
finished.
Ja n u a ry 26, 1993
ACKNOWLEDGEMENTS
F o r perm ission to re p ro d u ce figures, I th a n k the follow ing:
F igu re 2.5, R ich ard M eadow ; F igures 2.6, 4.4, a n d 4.14, B rian H esse an d
T arax ac u m Press; F ig ure 2.7, A n n a K. B ehrensm eyer a n d T he P aleontological
Society; F ig u re 2.8, P eter A ndrew s a n d T he R oyal A n th ro p o lo g ica l Institute;
F ig u re 3.2, D ian e G iffo rd -G o n zalez a n d A cadem ic Press; F ig u re 3.3, D ian e
G iffo rd -G o n zalez a n d T he C en ter fo r the S tudy o f the F irst A m ericans;
Figures 4.1 a n d 4.3, T. S. P arso n s a n d W . B. S aunders C o m p an y ; F igure 4.2, J.
D. C u rrey a n d E d w ard A rn o ld L td.; F igures 4.9, 4.10, 4.11, 4.12, an d 4.13,
S tanley J. O lsen a n d th e P resid en t a n d Fellow s o f H a rv a rd C ollege, P eab o d y
M useum ; F ig ure 6.1, A n n a K. B ehrensm eyer a n d P len u m Press; F igure 6.5,
A n n a K. B ehrensm eyer a n d H a rv a rd U niversity, M u seu m o f C o m p arativ e
Z oology; F igu res 6.10, 6.21a, a n d 6.22, C en ter for the S tudy o f the F irst
A m ericans; F ig ures 7.4 a n d 8.2, A cadem ic Press; F ig u re 7.5, A cadem ic Press,
L td.; F ig ures 8.1a, 8.13a a n d 8.13b, Society fo r A m erican A rchaeology; F igure
8.4, L arry G . M arsh all a n d T he C en ter for the S tudy o f the F irst A m ericans;
F igure 8.6, Eileen Jo h n so n a n d A cadem ic Press; F ig u re 9.4. R ich ard P o tts and
A ldine de G ru y te r; F ig u re 9.7, Lewis R. B inford a n d A cadem ic Press.
xxvi
1
WH A T IS T A P H O N O M Y ?
O nly a sm all p a rt o f w hat once existed w as buried in the ground; only a p a rt o f w hat
was buried has escaped the destroying h an d o f time; o f this p a rt all has n o t yet com e
to light again; and we all know only to o well how little o f w hat has com e to light has
been o f service for o u r science.
(O. M ontelius 1888:5)
Introduction
T a p h o n o m y is the science o f the law s o f em b ed d in g o r burial. M o re com pletely,
it is th e study o f the tra n sitio n , in all details, o f organics from the b iosphere into
the lith o sp h ere o r geological record. T hese definitions were given by the
R ussian p aleo n to lo g ist I. A. E frem ov (1940) w ho coined the term from the
G reek w o rd s taphos (b u rial) a n d nom os (laws). T a p h o n o m y is, how ever,
im p o rta n t n o t only to p aleo n to lo g ists, b u t to archaeologists, especially zoo-
arch aeolo gists a n d p a le o e th n o b o ta n ists, w ho stu d y the org an ic rem ains
m ak in g u p p a rt o f th e archaeological record. T h a t im p o rtan ce h as com e to be
widely recognized in the p a st 20 o r 30 years. T a p h o n o m y is now seen as
im p o rta n t because it is o ften ta k e n to co n n o te th a t the zo o a rch aeo lo g ical an d
e th n o b o ta n ic a l reco rd s are biased if som e n o n -h u m a n -re la te d processes have
affected th e c o n d itio n o r frequencies o f biological rem ains. W hile th a t p ercep
tio n is o ften co rrect, I will show th a t this p ercep tio n is frequently in correct.
T h e re aso n arch aeo lo g ists sh o u ld be co n cerned w ith ta p h o n o m y is th a t it
involves th e fo rm a tio n o f w h a t is o ften a m a jo r p a rt o f the archaeological
record. If the arch aeo log ical reco rd is th o se m o d ern traces o f p ast h u m an or
ho m in id b eh av iors, then the d iscard ed rem ains o f m eals such as m am m al bones
a n d p la n t p a rts c o n stitu te a p o rtio n o f th e arch aeo lo g ical record. T hus,
ta p h o n o m ic research involving the zoological a n d b o tan ica l p o rtio n s o f the
archaeolo gical reco rd involves “ the study o f processes o f p re serv atio n a n d how
they affect in fo rm a tio n " co n tain e d w ithin these p a rts o f the reco rd (B ehrens
m eyer a n d K idw ell 1985:105).
G ra n tin g th e preceding, the reaso n fo r this b o o k ’s existence should be self-
evident. W h a t is p erh ap s n o t so evident, how ever, is the reaso n such a b o o k is
ap p e arin g now given th a t arch aeo lo g y has been p ra cticed w ithin a scientific
p arad ig m fo r ov er 100 years (e.g.. T rig g er 1989). In o rd e r to assess w hy
ta p h o n o m y is now seen as im p o rta n t, a n d to help explore w hy tap h o n o m ic
research o f th e late tw en tieth ce n tu ry ap p e a rs the w ay it does, the first p a rt o f
1
2 Vertebrate taphonom y
C h a p te r 2 review s th e h isto ry o f ta p h o n o m ic research an d assesses its cu rren t

statu s. T he second p a rt o f C h a p te r 2 presents a p erso n al view o f the stru ctu re o f
ta p h o n o m ic inquiry. T he h isto ry an d cu rre n t statu s o f tap h o n o m ic research
allow m e to tak e u p the to p ic o f C h a p te r 3, w h a t is variously called actualistic
research, eth n o arch ae o lo g y , m iddle-range research, o r n e o ta p h o n o m y , and
how this relates to identifying the fo rm a tio n a l dynam ics o f a zo o arch aeo lo g ical
reco rd from its m o d e rn static traces. C h a p te r 4 is devoted to a review o f
v erte b rate skeletal tissues a n d skeletons, a n d a discussion o f how to q u an tify
fau n al rem ains. C h a p te rs 5 th ro u g h 12 are d evoted to describing m an y o f the
co m m o n ly em ployed ta p h o n o m ic an aly tic techniques. In C h a p te r 13 m aterials
fro m earlier c h a p te rs are in teg ra te d a n d synthesized to pro v id e a fram ew o rk
for p erfo rm in g intensive a n d extensive ta p h o n o m ic analyses.
It is im p o rta n t here to in tro d u c e som e basic term s a n d co ncepts th a t are used
th ro u g h o u t this volum e. T he next few sections o f this first ch a p te r, then, are
d ev o ted to review ing som e o f the basics o f zo o arch aeo lo g ical analyses and how
tap h o n o m ic research c o n trib u te s to those analyses. T h a t b ac k g ro u n d leads to a
co n sid eratio n o f th e kinds o f c o n trib u tio n s ta p h o n o m ic research can m ake to
arch aeo lo g y in general. T he final topic o f C h a p te r 1 is a discussion o f w h at this
b o o k is m ean t to be, w h at it is n o t, a n d why.
On the analysis o f archaeological faunal remains
We m ust first elim inate causes o f erro r, an d discover w h at N atu re can do to bones
subm itted to her action.
(H. A. Breuil 1938:58)
A nalyses o f arch aeo log ical fau n al rem ains have been u n d e rta k e n at least since
th e late n in eteen th cen tu ry in N o rth A m erica ( R ob iso n 1978), an d p ro b a b ly for
at least 50 years p rio r to th a t tim e in E u ro p e (M o rlo t 1861). W hile once scarcely
m o re th a n a su b sid iary en d eav o r, arch aeo lo g ical site rep o rts now regularly
c o n ta in a section o n recovered faunal rem ains, often w ritten by a specialist, a n d
m an y m o re in d ep en d en tly published a n d in -d ep th studies o f faunal rem ains are
being p re p are d by specialists in zoology a n d archaeologists w ith zoological
train in g (L y m an 1979a). T his reflects the holistic a p p ro a c h o f archaeologists
trying to u n d ersta n d a n d explain the to ta lity o f h u m an history.
T h ere are tw o basic goals to analyzing p re h isto ric fau n al rem ains: reco n
stru ctio n o f ho m in id subsistence p a tte rn s, an d re co n stru ctio n o f paleoecologi-
cal co n d itio n s (H esse a n d W apnish 1985; K lein a n d C ru z-U rib e 1984). T he
fo rm er has been ch aracterized as an a tte m p t “ to explain, in the form o f
predictive m odels, the interface th a t existed betw een p reh isto ric h u m an
p o p u la tio n s an d the faunal section o f the biotic co m m u n ity ” (S m ith 1976:284).
T his goal is a n th ro p o lo g ica l in o rie n ta tio n as it addresses topics such as h u m an
diet, anim al reso urce p ro c u rem en t strategies, o r p re d a to r-p re y relationships.
A naly tic goals are atta in e d using a n th ro p o lo g ica l a n d ecological principles in
W hat is taphonom y? 3
analysis a n d in te rp re ta tio n (R a c k h a m 1983; e.g., L ym an 1992b). A nalyses o f

paleoecological co n d itio n s use zoological a n d ecological d a ta , m eth o d s, an d
th eo ry (D o d d a n d S ta n to n 1981; K ing a n d G ra h a m 1981) to re co n stru ct faunal
tu rn o v e r an d succession, p a leo e n v iro n m en tal h isto ry , a n d zoo g eo g rap h ic
histo ry (e.g., G ra y so n 1987).
T he tw o d istingu ished goals are n o t m u tu ally exclusive. B oth require
tax o n o m ic identification o f fa u n al rem ains, w hich necessitates adh eren ce to
zoological m eth o d a n d th eo ry . D a ta in te rp re ta tio n requires use o f ecological
principles w h eth er th o se co ncern h a b ita t preferences o f tax a o r d eterm in in g
av ailab le b iom ass o r m eat. In te rp re ta tio n o f a single assem blage o f fau n al
rem ains recovered fro m an arch aeo lo g ical site m ay accom plish eith er o r b o th
goals (K in g a n d G ra h a m 1981) because, in p a rt, an aly tic techniques overlap.
D istin ctio n o f the tw o goals is useful fo r discussion p u rp o ses, b u t is n o t
m a n d a to ry to ac tu a l analysis.
Basic concepts
In this vo lum e I focus o n v erte b rate rem ains. R esearch o n in v erte b rate
ta p h o n o m y is largely, b u t n o t entirely, fo u n d in the p aleo n to lo g ical literatu re.
M y re m a rk s are ap p licab le to the rem ains o f v irtu ally all anim al tax a, a n d
m an y are also ap p licab le to p la n t rem ains. I re stric t discussion a n d exam ples in
this volum e largely to m am m al rem ains fo r the sim ple reason th a t m ore
ta p h o n o m ic research has co ncerned m am m als th a n an y o th e r v erte b rate
tax o n o m ic gro u p; n o n -m a m m a lia n v erte b rates are covered in som e d etail in
C h a p te r 12.
T a p h o n o m y is generally co n stru e d as focusing on the p o stm o rtem , pre-, an d
p o st-b u rial histo ries o f fau n al rem ains. B urial is considered to be a stage
in term ed iate to pre- a n d p o st-b u ria l histories d u e to the p o ten tially destructive
an d d isru p tiv e n a tu re o f b u rial processes (e.g., D ixon 1984; K ra n z 1974a,
1974b). V ario u s arran g e m e n ts o f ta p h o n o m ic agents a n d processes have been
p o sited in the fo rm o f m odels depicting a general ta p h o n o m ic h isto ry (see
C h a p te r 2). G enerally, a bo n e m ay be gnaw ed, buried, exposed, reburied, re
exposed, b ro k en , tra n s p o rte d , a n d re b u ried p rio r to recovery (see the G lo ss
ary). R ealistic sequences o f ta p h o n o m ic facto rs m ay th erefo re require the
inclu sio n o f loops. A general ch ro n o lo g y o f ta p h o n o m ic agents a n d processes
affecting an im al rem ain s is called a taphonom ic history o r taphonom ic pathw ay.
A taphonom ic agent is the source o f force applied to bones, th e “ im m ediate
physical cau se” o f m o d ificatio n to an im al carcasses a n d skeletal tissues
(G iffo rd -G o n zalez 1991:228), such as gravity, a hyena, o r a hom inid. A
taphonom ic process is the d y n am ic ac tio n o f an ag en t o n an im al carcasses an d
skeletal tissues, such as dow nslope m o v em en t, gnaw ing, o r fractu rin g (relative
to the ag en ts listed in the preceding sentence). A taphonom ic effect o r trace is the
static result o f a ta p h o n o m ic process actin g o n carcasses a n d skeletal tissues,
the physical a n d /o r chem ical m o d ificatio n o f a bone. As we will see, ta p h o n o

m ic analysis involves identifying a n d /o r m easu rin g ta p h o n o m ic effects, a n d on
th a t basis id entifying a n d /o r m easu rin g the m ag n itu d e o f effects o f ta p h o n o m ic
processes a n d agents.
A ta p h o n o m ic h isto ry begins w hen one o r m ore m em bers o f a biotic
c o m m u n ity die. P o stm o rte m processes th a t affect carcasses a n d skeletal
tissues c o n stitu te the m a jo r p a rt o f a ta p h o n o m ic history. R ecovery o f faunal
rem ains is a p o ten tially biasing fa c to r because it affects the collected assem
blage th ro u g h differentially m oving a n d sam pling it. W h a t the collector
perceives as p ertin e n t o b serv atio n s m ay significantly influence w hich fossils are
collected an d w hich d a ta are recorded, a n d co n seq u en tly influence the final
an aly tic results. A large lite ra tu re alread y exists on this crucial topic (G am b le
1978). Z o o arch ae o lo g ists have becom e m u ch m ore aw are o f the stratig ra p h ic
a n d sed im en tary co n tex ts o f an im al rem ains an d the p o ten tial tap h o n o m ic
significance o f such geological an d c o n tex tu a l d a ta (e.g., R ap so n 1990). A s a
result, m o re care is ta k e n in the recovery o f an im al rem ains to d a y th a n in the
past.
A fa u n a is som e specified set o f a n im al tax a fo u n d in a geographic area o f
som e specified size, kind, a n d lo catio n at som e specified tim e (O d u m 1971:366-
367). F o r exam ple, one can specify a m o d ern in tertid al fa u n a o f the Pacific
R im , a p reh isto ric terrestrial fa u n a o f E u ro p e , a n d a P leistocene m am m alian
fau n a o f C o lo ra d o . Z oo logists stu d y fa u n as by observing living anim als.
P aleo n to lo g ists a n d zo o a rch a eo lo g ists stu d y fa u n as by analyzing fossils. I have
h ad several zo o a rch a eo lo g ists tell m e “ fossils are m ineralized an im al re m a in s,”
o r “ fossils are o ld er th a n 10,000 y ea rs.” I find n eith er o f these criteria in
d efinitions p u b lish ed by p aleo n to lo g ists (see the G lossary). I use the term fo s s il
in this volum e to d en o te an y trac e o r rem ain o f an an im al th a t died at som e tim e
in th e p a st (ascertain in g the age o f an im al d e a th is a sep a rate problem ).
A fo s s il record is som e set o f rem ains o f organism s, eith er o r b o th p lan ts and
an im als, hav ing a geological m o d e o f occurrence in som e defined geographic
space an d geological co n tex t (i.e., w ith a delineated spatial d istrib u tio n )
(m odified fro m L ym an 1982a, 1987e). A fossil reco rd consists o f those
o b servab le p h en o m en a such as the p a rtic u la r bones in a p a rtic u la r stratu m . A
fo s s il fa u n a consists o f th o se tax a rep resen ted by the fossil reco rd a t a specific
locality. T he term fossil fa u n a serves to em phasize the ta p h o n o m ic distin ctio n
betw een a living fa u n a a n d a fa u n a rep resen ted by fossils. W hile the term fossil
fa u n a as defined here is v irtu ally sy n o n y m o u s w ith the term s local fa u n a an d
p erh ap s fa u n u le (T ed fo rd 1970), the first term em an ates from the tap h o n o m ic
p erspective o f this volum e w hile the la tte r tw o em an a te from a paleoecological
perspective.
I d istin gu ish tw o kind s o f fossil faunas: those w ith o u t, a n d those w ith,
spatially asso ciated cu ltu ra l m aterials, o r paleontological fa u n a s a n d archaeo-
fa u n a s, respectively. T he d istin ctio n is n o t m e a n t to im ply w h eth er o r n o t
h u m an s h ad a role in the ta p h o n o m ic h isto ry o f a p a rtic u la r fossil assem blage.

A nalytically categ o rizin g a p a rtic u la r fossil reco rd as c o n stitu tin g a n atu ra lly
o r cu ltu rally d ep o sited set o f fa u n al rem ains is a m a jo r p a rt o f ta p h o n o m ic
analysis in zo o a rch a eo lo g y (e.g., A very 1984; B inford 1981b; P o tts 1984;
T u rn e r 1984). W h a t is m ean t by these tw o term s is sim ply w h eth er artifa cts are
o r are n o t sp atially associated w ith the fa u n al rem ains.
Goals o f taphonomic analysis in zooarchaeology

S ubsistence studies, by the n a tu re o f th eir research q u estio n s, req u ire k n o w l
edge o f the fo rm a tio n o f the arc h a e o fa u n a l reco rd (L y m an 1982a; M altb y
1985b; M ed lo ck 1975; R a c k h a m 1983). S im ilar know ledge is im p o rta n t to
p aleoecological research b u t fo r different reaso n s (B ehrensm eyer a n d H ill
1980; G iffo rd 1981; G ra y so n 1981). S ubsistence studies require th a t the fossils
co n stitu tin g th e a rc h a e o fa u n a be so rted in to a t least tw o categories: those
d ep o sited as a resu lt o f h u m a n (subsistence a n d o th er) behaviors, a n d those
n atu ra lly d ep o sited (B inford 1981b; T h o m as 1971). C u ltu rally deposited
fossils m u st be q u alitativ ely a n d q u a n titativ ely rep resen tativ e o f the fau n a
exploited, an d q u an tifica tio n techniques m u st p ro d u c e accu ra te relative a b u n
dances o f econom ically im p o rta n t tax a (G ray so n 1984; L ym an 1979b). P aleoe
cological studies need n o t have rep resen tativ e sam ples o f h u m an ly exploited
tax a, b u t m ay req u ire re p resen tativ e sam ples o f p reh isto rically e x ta n t faunas.
S am ple req u irem en ts are flexible in the sense th a t they have c e rtain to leran ce
lim its F o r exam ple, a bison kill site p ro b a b ly does n o t include all tax a
ex plo ited by a g ro u p o f peo p le yet it ca n be studied a n d analyzed. Sim ilarly, a
zo o a rch a eo lo g ist m ay focus only o n the m ic ro fa u n a a n d ig n o re larg er tax a in
an arch ae o fau n a . B o th dep en d o n the research qu estio n s being asked. S am ple
represen tativen ess is relative to som e p o p u la tio n w hich in tu rn is d ictated by
th e research goal. T h e representativeness o f a sam ple o f fau n al rem ains is
co n tro lled by the sam p le’s ta p h o n o m ic h isto ry , the sam pling tech n iq u es used
to collect the sam ple, and the research qu estio n s being asked o f the sam ple.
G iffo rd (1981) distinguishes tw o basic goals o f ta p h o n o m ic research: (1)
“ strip p in g aw ay ” th e ta p h o n o m ic o v e rp rin t fro m the fossil re co rd to o b ta in
a ccu rate re so lu tio n o f th e p re h isto ric b io tic co m m u n ity , a n d (2) d eterm in in g
the n a tu re o f th e ta p h o n o m ic o v erp rin t in o rd e r to be able to list the precise
ta p h o n o m ic m echan ism s responsible fo r a given fossil assem blage, en abling
the w ritin g o f ta p h o n o m ic histories. T he la tte r goal is sim ilar to studying the
fo rm a tio n o f the arch aeo lo g ical reco rd (Schiffer 1987), an d m ay lead to
con clusions re g ard in g th e h u m a n beh av io rs th a t c reated the fossil record. T he
fo rm er goal is seen as a necessary step to w a rd s paleoecological analysis because
the ta rg e t o f analysis req uires know ledge o f the p re h isto ric biotic com m unity.
F o r exam ple, as G ra h a m a n d K ay (1988:227) note, “ the definition o f ta p h o n o
mic path w ay s is n o t the u ltim ate goal o f paleobiological o r a n th ro p o lo g ica l
6 V ertebrate taphonom y
studies, b u t [it is] a m eans to fa c to r o u t biases a n d fu rth e r o u r fu n d a m e n tal

u n d ersta n d in g o f p ast ecosystem s a n d h u m an cu ltu re s.”
D e te rm in a tio n o f the exact ta p h o n o m ic h isto ry o f a p a rtic u la r fossil
assem blage is freq u en tly atte m p te d by zo o a rch a eo lo g ists w ho w ish to know
w hich tax a w ere exp loited by ho m in id s a n d the relative p ro p o rtio n s in w hich
they w ere exploited. M an y in te rp re ta tio n s th erefo re involve outlines o f the
suspected h u m a n b eh av io rs th a t resulted in th e fossil assem blage. F o r exam ple,
W h e a t’s (1972) d escrip tio n o f the h u n tin g an d b u tch ery process evidenced at
the O lsen -C h u b b u ck bison kill site is sim ply a n arrativ e m odel o f the suspected
ta p h o n o m ic h isto ry o f th a t site’s fossil record.
T he tw o goals o f ta p h o n o m ic research are n o t m u tu ally exclusive. S tripping
aw ay th e ta p h o n o m ic o v erp rin t requires th a t the o v erp rin t be know n. O nce the
ta p h o n o m ic o v erp rin t is kno w n , the p re h isto ric biotic co m m u n ity can, th e o re
tically. be d eterm in ed by analytically reversing the effects o f the tap h o n o m ic
processes. O f co urse, this p ro c ed u re requires the a ssu m p tio n th a t the sam ple o f
fossils is rep resen tativ e o f the biotic co m m u n ity , o r th a t it can be analytically
m ad e rep resen tativ e o f th a t co m m u n ity . T his assu m p tio n has been analytically
co n tro lled in cases w here an a rc h a e o fa u n a is directly co m p ared w ith a
p aleo n to lo g ical fa u n a in close g eo graphic an d tem p o ra l p ro x im ity to one
a n o th e r (e.g., B riuer 1977), a n d in cases w here tw o o r m o re geographically and
tem p o rally ad jacen t a rc h a e o fa u n a s are c o m p ared (e.g., G ra y so n 1983; G uil-
d ay et al. 1978). T h e covert a ssu m p tio n o f such co m p arativ e analyses is th a t
because each fossil assem blage has u n d erg o n e a m o re o r less u n iq u e ta p h o n o
mic history, sim ilar b u t in d ep en d a n t in terp retiv e results derived from the
assem blages are th o u g h t to represent p reh isto ric reality. T h a t is, the analyst
ca n h av e g re ater confidence th a t ta p h o n o m ic processes have n o t to tally
obscu red all in d icatio n s o f a p reh isto ric biotic co m m u n ity w hen all exam ined
fossil assem blages in d icate the sam e com m unity.
As we shall see, th e goals o f ta p h o n o m ic analysis ca n be m u ch m ore finely
distinguished. H ere I have show n th a t the tw o general goals o f tap h o n o m ic
analysis are easily aligned w ith the tw o basic goals o f zooarch aeo lo g ical
analysis, a n d th a t the tw o kinds o f analysis are, a n d in fact m u st be, synergistic.
T h a t is, d eterm in in g w hich an im al tax a w ere eaten by p re h isto ric peoples, a n d
how m uch o f each an im al tax o n w as eaten , is surely a ta p h o n o m ic problem .
T a p h o n o m y th u s p resents a challenge to zo o arch aeo lo g ists th a t can be sim ply
p h rased as “ W h a t are these b ones do in g in this site?”
The challenge o f taphonomy
U n d e rsta n d in g how ta p h o n o m ic processes affect q u a n tita tiv e m easures o f

fa u n al rem ains is a m ajo r challenge facing zooarch aeo lo g ical research (G ilbert
a n d Singer 1982; H o ltzm an 1979; A. T u rn e r 1983). Q u a n tita tiv e m easures such
as tax o n o m ic ab u n d a n ces, m eat w eights, a n d frequencies o f p a rtic u la r skeletal
elem ents are all affected by ta p h o n o m ic processes (B adgley 1986a; G ra y so n

1984). N o t only are q u a n tita tiv e d a ta im p o rta n t in m an y analyses, b u t so are
the d istrib u tio n s o f b ones a n d tax a w ithin a site (G ray so n 1983; L y m an 1980;
W h eat 1972). T a p h o n o m ic processes m ay obscure d istrib u tio n a l contexts,
u n related elem ents m ay becom e spatially associated, o r related elem ents m ay
lose th eir sp atial asso ciatio n (H ill 1979b). T he second m ajo r challenge in
zo o arch aeo lo g ical research is, then, ascertain in g the m eaning o f the d istrib u
tio n s o f fa u n a l rem ains. T he th ird m a jo r challenge is d eterm in in g h ow a n d w hy
the recovered fa u n a l rem ains differ fro m th e biotic co m m u n ity in w hich they
orig in ated . T his q u estio n is p a rtic u la rly im p o rta n t to paleoecologists as well as
zo o arch aeo lo g ists. O th er, m o re finely d istinguished challenges are easily
conceived, b u t these m a jo r ones tend to u n d e rp in virtu ally all o f th o se n a rro w e r
challenges.
Taphonomv’s contribution to zooarchaeology

E xam ples m ak e clear th e n a tu re o f the c o n trib u tio n o f ta p h o n o m y to z o o a r
chaeology. F o r in stance, arch aeo lo g ists have, beginning in the late 1970s,
ad o p te d o p tim al fo rag in g th eo ry from ecology as an ex p lan a to ry device (see
B ettinger 1991 fo r a review). P a rt o f th a t th eo ry d em an d s m easuring the
b re a d th o f the niche exploited by the subject fo rag er, in this case hum ans.
T ypically, niche b re a d th is m easu red by tallying the n u m b er o f p la n t and
anim al tax a ex plo ited. T h u s, the rem ains o f p lan ts a n d anim als fo u n d in
archaeo lo gical sites m u st m inim ally be so rted in to tw o categories: those
rep resen tin g tax a th a t w ere exploited, a n d th o se rep resen tin g tax a th a t were not
exp loited by people. T his entails exam ining the rem ains fo r in d icatio n s th a t
h u m an s ac cu m u lated a n d d ep o sited c e rtain o f the rem ains a n d in d icatio n s th a t
the o th e r rem ain s w ere n a tu ra lly deposited. I f som e o f the b ones have
b u tch erin g m a rk s on them , th en it is re aso n ab le to su p p o se th a t the tax o n o r
tax a rep resen ted w ere ac cu m u lated a n d b u tch ered by people. In this case, the
b u tch ery m a rk s are the in d icatio n o f a h u m a n agent in the tap h o n o m ic history
o f th o se rem ain s. I f the n early com plete a n d p artia lly a rtic u la te d skeleton o f a
b u rro w in g an im al such as a g o p h er is fo u n d in a k ro to v in a in a site, it is likely
th a t this in div id u al w as n atu ra lly d ep o sited a n d did n o t form a p a rt o f the
h u m a n o c c u p a n ts’ diet. H ere, the degrees o f skeletal com pleteness an d
artic u la tio n , th e b eh a v io r o f the tax o n , a n d th e co n tex t o f the rem ains are the
relevan t ta p h o n o m ic traces leading to the inference th a t these a n im al rem ains
were n atu ra lly dep osited. T hese are, as we shall see, sim plistic exam ples.
M o d ern ta p h o n o m ic analysis is seldom so straig h tfo rw ard .
A n arch ae o lo g ist in terested in m easu rin g the d ietary b re a d th o f som e
p re h isto ric h u m a n g ro u p m u st d istin g u ish betw een cu ltu rally a n d n atu ra lly
d ep o sited an im al rem ains. I f this is n o t do n e, th e n the an a ly st has sim ply
m easu red th e tax o n o m ic richness (n u m b er o f species) rep resen ted by the
sam p le o f an im al rem ain s, a n d n o t d ietary b re a d th . Inferences th a t h u m an s

b ro a d en ed th eir d ietary niche, o r n arro w ed it, o r did n o t a lte r it th ro u g h tim e,
will surely be in acc u rate w ith o u t such ta p h o n o m ic analyses.
In p aleoecological analysis, an im al rem ains ca n g ra n t insights to the
clim ato lo gical a n d floral e n v iro n m e n t to w hich a h u m a n g ro u p ad a p te d . If the
rem ain s o f a n an im al ta x o n th a t to d a y prefers co o l-m o ist en v iro n m en ts is
fo u n d in a site lo cated in a w arm -d ry h a b ita t, those rem ains p o ten tially indicate
th a t th e site m ay h ave been occupied d u rin g a tim e o f co o ler a n d m o ister
co n d itio n s (assum ing th e rem ains w ere d ep o sited a t the tim e o f h u m a n
o cc u p atio n ). T h a t in d icatio n can, how ever, be false if the an im al rem ains w ere
n o t locally derived. D id , p erh ap s, a far-ran g in g p re d a to r (h u m a n o r n o t) collect
th o se rem ain s som e significant d istan ce aw ay from the site a n d th e n tra n s p o rt
th o se rem ain s a n d d ep o sit them there? O r, did the tax o n in q u estio n actually
live o n o r very near, say, < 1 km , fro m th e site? A re the rem ains co rro d ed from
the digestive acids o f som e carnivore? I f so, th en p erh ap s, b u t n o t necessarily,
the rem ain s cam e from a significant distan ce aw ay. D o the rem ains represent
com p lete skeletons, o r selected p o rtio n s o f skeletons? Is the ta x o n represented
so large th a t it co u ld n o t have been tra n sp o rte d w hole to the site? P ro d u cin g
answ ers to these a n d related q u estio n s involves ta p h o n o m ic analysis a n d
pro v id es th e d a ta necessary to m ake inferences a b o u t the local o r d ista n t origin
o f th e rem ains in question.
O ne exam ple o f how ta p h o n o m y has co n trib u te d to o u r u n d e rsta n d in g o f
h u m a n p re h isto ry involves th e d eb a te over w h eth er o u r P lio-P leistocene
an cesto rs w ere h u n te rs o r scavengers. T he lite ratu re o n this to p ic h as grow n to
im m ense p ro p o rtio n s in the last decade (e.g., B inford 1981b, 1984b, 1988b;
B lum enschine 1986a, 1986b. 1987; B unn a n d K roll 1986, 1988; P o tts 1984,
1988; S h ip m an 1986a, 1986b). All o f th a t lite ratu re involves detailed ta p h o n o
m ic analysis. A s a result o f th a t research, the 1950s consensus th a t early
h o m in id s w ere h u n ters has been ch an g ed to one o f perceiving them as
ind ividuals w ho p ro b a b ly h u n ted sm all gam e a n d scavenged large gam e,
a lth o u g h o p in io n s differ on the precise m ag n itu d e o f the d ietary co n trib u tio n
o f b o th h u n tin g a n d scavenging. T his topic rem ains one o f the m o st discussed
in the lite ratu re o f th e 1990s.
Terminology used in this book

I follow several co n v e n tio n s in describing d a ta used in exam ples in this volum e.
T hese are n o t form alized b u t ra th e r are ad hoc con v en tio n s. “ P ” stan d s for
p ro x im al, “ D ” fo r distal. A b b re v iatio n s fo r p a rtic u la r skeletal elem ents vary
betw een in v estigators, a n d I have n o t a ttem p ted to be system atic in my use o f
these; ra th e r I have in m an y cases applied the original in v estig ato r’s a b b re v ia
tio n s w ith a p p ro p ria te definitions. A n assem blage o f fossils is som e analytically
defined set o f fa u n al rem ains usually, b u t n o t alw ays, from a p a rtic u la r spatio-
tem p o ra l co n text. I use the term “ b o n es” freq u en tly as a generic term fo r bones,
teeth, h o rn s, an tlers, etc. F inally, I use th e term s “ h o m in id s” a n d “ h u m a n s ”
interch ang eab ly .
In m an y cases I em ploy som e very basic statistical tests. T h ro u g h o u t, the
significance levels are d en o ted by P, P e a rs o n ’s p a ra m e tric c o rre la tio n coeffi
cient is d en o ted by r, S p e a rm a n 's ra n k o rd e r co rre la tio n coefficient is d enoted
by rs. D iscussion a n d d escrip tio n s o f these ca n be fo u n d in any in tro d u c to ry
text on statistics. In this age o f p erso n al c o m p u ters, I suspect we will see an
increasing n u m b e r o f statistical analyses o f ta p h o n o m ic d a ta . T he q u a n tita tiv e
aspects o f zo o a rch aeo lo g ical m aterials are a subject th a t could readily fill
a n o th e r volum e; m an y o f them are discussed in detail by G ra y so n (1984). A n
in tro d u c tio n to th e basic q u an tifica tio n un its o f zo o a rch aeo lo g ical research is
p ro v id ed in C h a p te r 4 o f this volum e, a n d ad d itio n a l co m m ents are p ro v id ed in
C h a p te r 8.
What this book is and what it is not

The objectives o f tap h o n o m ic analysis are very varied and its m eth o d s are eclectic,
being governed in p a rt by d isp arate ch aracteristics o f different types o f fossil
assem blages, an d in p a rt by the n atu re o f [research] problem s it is called u p o n to
address.
(R. D. K. T hom as 1986:206)
This b o o k is m ean t to review m an y o f the p o ten tially useful a n d in fo rm ativ e

an aly tic techniques ta p h o n o m ists have developed to help solve p a rtic u la r
zo oarch aeo lo g ical p roblem s. It is n o t m e a n t to pro v id e a set o f alg o rith m s for
solving conclusively all o r any p o te n tia l in terp retiv e problem s; it is n o t a
c o o k b o o k in th e sense th a t follow ing a p a rtic u la r recipe will p ro d u c e a tasteful
o r even edible p ro d u c t. T he h e a rt o f the volum e lies in the follow ing ch ap ters,
w hich are m e a n t to in tro d u ce th e novice’s m ind to, a n d refresh the e x p e rt’s
m em o ry concernin g , the d iversity o f v ariables th a t m u st be considered a n d the
p le th o ra o f an aly tic tech niques th a t m ig h t m ake u p a detailed ta p h o n o m ic
analysis. A s th e v o lu m e’s title in dicates, only v e rte b ra te ta p h o n o m y is c o n
sidered. F o r recent synopses o f in v e rte b ra te ta p h o n o m y , see the papers
in tro d u ce d by T h o m a s (1986) a n d the volum es edited by A llison a n d Briggs
(1991b) a n d D o n o v a n (1990). D u e to m y ow n lim itatio n s, m y review o f the
av ailab le ta p h o n o m ic lite ratu re is largely restricted to th a t p o rtio n o f it
pu b lish ed in E nglish. I to o k a zo o a rch aeo lo g ical perspective in w riting this
volum e, a n d th u s zo o a rch a eo lo g ists will, I hope, find m u ch o f value in its pages.
I ho p e as well th a t v erte b rate p aleo n to lo g ists m ay find som e o f the m aterial
useful.
N o w h ere in this volum e will a d etailed ta p h o n o m ic analysis o f a p a rtic u la r
zo o a rch aeo lo g ical co llection be fo u n d , a lth o u g h v ario u s collections are des
cribed to exem plify th e results o f p a rtic u la r ta p h o n o m ic processes o r are used
to illu stra te h ow a p a rtic u la r an aly tic to o l w orks. Sim ply, this b o o k is a review
o f m an y o f the an aly tic techniques used in the 1980s a n d early 1990s to help
determ ine th e ta p h o n o m ic h isto ry o f b o n e collections. B ecause archaeological
ta p h o n o m y is a ra p id ly d eveloping field, there is n o d o u b t th a t som e o f the
tech n iq u es review ed here will n o t be in use ten years fro m now , a n d techniques
n o t yet d eveloped a n d th u s n o t described here will be developed in the future.
As I n eared finishing w h at I th o u g h t w ould be a re aso n ab ly com plete first d ra ft,
I c o n tin u ed to en c o u n te r new ly published articles a n d to find references to
articles p u b lish ed years ago th a t I h a d n o t previously been aw are of. B ecause it
w as necessary to the co m p letio n o f this volum e, I sto p p ed review ing an d
in c o rp o ra tin g new d a ta a n d ideas in D ecem b er o f 1992. T h u s, w ith few
exceptions, the references cited herein w ere pu b lish ed p rio r to th a t date. T he
volum e is in som e w ays, then, incom plete a n d in o th e r w ays it is o u t o f d ate. 1
tak e these facts alo ne to in d icate th a t ta p h o n o m ic research is reaching, a n d
p erh ap s will c o n tin u e to enjoy fo r som e tim e, a p erio d o f florescence. T his
volum e is sim ply one m a rk o f this period.
T here is n o clearly stated , explicit p arad ig m for ta p h o n o m ic research a n d no
rules fo r how to do it (T h o m as 1986), except p erh ap s those u n d e r the um brella
o f u n ifo rm itarian ism . T a p h o n o m ic research in p re h isto ric co n tex ts has few
c riteria for assessing the validity o f a so lu tio n to a ta p h o n o m ic pro b lem . It is
n o t alw ays clear ho w to d eterm in e if a p a rtic u la r an aly tic tech n iq u e w as the
a p p ro p ria te one fo r a p a rtic u la r p roblem . In stea d , th e results o f ta p h o n o m ic
research are o ften ev alu ated fro m th e perspective th a t th o se results sh o u ld be
replicable if a n o th e r an a ly st uses the sam e d a ta a n d an aly tic procedures.
T his is n o t a volum e o n techniques o f zo o a rch a eo lo g ical analysis, a lth o u g h it
sho u ld be clear th a t m uch o f w h a t is described here often does (an d should
alw ays) m ak e u p m a jo r p o rtio n s o f m o d e rn zo o a rch aeo lo g ical research.
Several g o o d volum es o n zo o a rch a eo lo g ical analysis exist (e.g., D avis 1987;
H esse a n d W a p n ish 1985; K lein a n d C ru z-U rib e 1984), a n d these can be
c o n su lted in c o n ju n ctio n w ith use o f this volum e.
T his vo lum e is n o t a p icto ra l o r descriptive essay. T h ere are illu stra tio n s here,
b u t I have k ep t th e n u m b er o f p h o to g ra p h s to a m inim um because there are
n o w av ailab le so m an y excellent descriptive volum es co n tain in g n u m ero u s
p h o to g ra p h s o f v ario u sly m odified b o n es th a t I co u ld n o t h o p e to replicate the
ex ten t o r q u ality o f th eir coverage. T he re ad er is en co u rag ed to stu d y closely
th e volum es a n d p h o to g ra p h s fo u n d in A ndrew s (1990), B inford (1981b), B rain
(1981), H ay n es (1991), an d W hite (1992). I have chosen in this volum e to focus
o n w h a t an an a ly st m ig h t d o once the m odifications have been recognized in a
fossil assem blage. T h erefo re the m ajo rity o f the illu stra tio n s I have chosen to
in clude here are exam ples o f g ra p h s a n d ch a rts in ten d ed to show how vario u s
tap h o n o m ically m odified fossil assem blages m ay a p p e a r w hen g ra p h ed in a
p a rtic u la r w ay. I find it m u ch m o re m entally stim u latin g to m an ip u late
tap h o n o m ically m odified bones analytically th a n sim ply to describe som e as
carn iv o re chew ed a n d som e as b u rn e d , a n d it is such analytical m a n ip u la tio n

th a t in fact allow s us to m eet the challenges o f ta p h o n o m y head-on.
F inally, I have certain ly n o t covered all possible topics th a t m ight be covered
in a volum e o f this n atu re . W ith m inim al exceptions I d o n o t delve deeply into
the ch em istry o f fossilization n o r do I discuss the effects o f sam pling an d
d ifferential recovery. T he b o o k is longer th a n I in ten d ed as it is; to have
included ad d itio n a l topics w ould o f course have resulted in a longer book.
T h ere are tw o excellent w ays to learn a subject. O ne is to teach it, the o th e r is
to w rite a b o o k a b o u t it. H aving said th a t, how can I expect an y o n e to read this
book? I can have th a t ex p ectatio n (an d hope) because the next best w ay to learn
a subject is to re ad a b o u t it; n o t ju s t one article o r one b o o k , b u t as m an y as you
can lay y o u r h an d s on. 1 learned a lot because I read a lot in o rd e r to w rite this
b o o k. O th er w riters have said m an y o f the sam e things said here, o ften tim es
b etter. I offer this b o o k as a sta rtin g p o in t, then, to those w ith interests in
v erte b rate tap h o n o m y .
2
THE H I S T O R Y A N D S T R U C T U R E OF
TAPHONOMY
T ap h o n o m y is now in its salad days.

(P. D odso n 1980:8)
A brief history o f taphonomic research

O ne h allm ark o f a scientific discipline’s m a tu rity is the p u b licatio n o f a history
o f it. Several b rief histo ries have been w ritten ab o u t ta p h o n o m y (C adee 1990;
D o d so n 1980; O lson 1980). A lth o u g h inform ative, these have been largely
lim ited to the relatio n o f ta p h o n o m y to paleo n to lo g y . P erh a p s because
ta p h o n o m y d eveloped first in p aleo n to lo g y , w hich has as its focus the study o f
biological ev o lu tio n an d paleoecology, w hen arch aeo lo g ists b o rro w ed the
co n cep t they also b o rro w ed the c o n n o ta tio n th a t the fossil reco rd is p ro b a b ly
biased. T h a t is, in fo rm a tio n o n the ecology a n d m o rp h o lo g y o f anim als is lost
o r altered betw een th e tim e o f an o rg a n ism ’s d e a th a n d the tim e its rem ains are
recovered a n d stu died (D o d so n 1980; L aw rence 1968, 1971). N o rth A m erican
p aleo b io lo gists initially a d o p te d E frem o v ’s version o f ta p h o n o m y - th a t the
fossil reco rd is incom p lete a n d th ere fo re biased - ra th e r th a n the G e rm an
version w hich focused on re co n stru ctio n o f p ast en v iro n m en ts via detailed
paleob io lo gical analysis (C adee 1990:9-13). It is n o t surprising, then, th a t
N o rth A m erican zo o a rch a eo lo g ists see the zo o a rch aeo lo g ical re co rd as biased.
P aleo b io lo g ists to d ay o ften w orry a b o u t the “ fidelity” (K idw ell an d B osence
1991) an d com pleteness (e.g., M cK in n ey 1990) o f the fossil record; lack o f
eith er d en o tes a biased fossil reco rd w ith reg ard to how accu rately p aleo b io tas
are reflected. But because b ias is relative, n o t all assem blages o f an im al rem ains
are biased in all ways.
A n o th e r h allm ark o f the m a tu rity o f a discipline o r research topic is the
p u b licatio n o f b o o k s d ev o ted to it. Several b o o k s have a p p e are d (A llison an d
Briggs 1991b; B ehrensm eyer a n d Hill 1980; D o n o v a n 1990; L eM oine an d
M acE ach ern 1983; S hip m an 1981b), the first ones at the sam e tim e th a t the
histories o f ta p h o n o m ic research ap p eared . Im p o rta n tly , tw o m a jo r review
articles (B ehrensm eyer 1984; G ifford 1981) an d tw o m a jo r b o o k s (B ehrens
m eyer an d Hill 1980; B inford 1981b) a p p e are d a b o u t a decade ago an d , w ith
th e o th e r b o o k s pu b lish ed at this tim e, presented the ta p h o n o m ic perspective
explicitly to an arch aeo lo g ical audience. Several years ag o a new sletter devoted
to ta p h o n o m ic research w as in itiated by tw o p aleobiologists (P lotnick an d
Speyer 1989, 1990; P lo tnick a n d W alk er 1991). Special sym posia (a n o th e r
12
The history and structure o f taphonom y 13
h isto rical la n d m a rk o f a discipline’s d ev elopm ent) w ith zo o arch aeo lo g ical

ta p h o n o m y as th eir foci h av e been convened a n d the proceedings fro m them
p u b lish ed in an effort to synthesize th e c u rre n t state o f o u r know ledge
(B onnichsen a n d Sorg 1989; D ixon a n d T h o rso n 1984; H u d so n 1993; S olom on
et al. 1990; see also N itecki a n d N itecki 1987). T his b rie f synopsis gives one
cause to w o n d e r “ W hy is ta p h o n o m y so im p o rta n t in arch aeo lo g y to d a y ? ” T he
q u estio n becom es p artic u la rly cogent w hen the n in etee n th -cen tu ry a rc h a e o
logical lite ra tu re is exam ined.
N ineteenth-century zooarchaeological taphonom y

K now ing the history o f o n e’s discipline can save one a good deal o f unnecessary
originality. It can also give one a g reat m any good ideas, for the p ast never says
things quite the way the present needs them said.
(P. B ohannan and M. G lazer 1988:xv)
G ra y so n (1986) d o cu m en te d th a t research n o t unlike ta p h o n o m y has a deep

h isto ry in arch aeo log y . T he n in etee n th -cen tu ry d eb a te over the m eaning o f
b ro k e n stones th o u g h t to re p resen t a T ertiary -ag ed “ eo lith ic” cu ltu re resulted
in m uch research directed to w ard s identifying physical a ttrib u te s o f the stones
th a t u n am b ig u o u sly signified h u m an m o dification. G ra y so n (1986) notes th a t
W illiam B u cklan d early in the n in etee n th century, a n d C h arles Lyell, E d o u a rd
L arte t, a n d o th ers d u rin g the second h a lf o f th a t ce n tu ry w ere co n cern ed w ith
the ag en ts a n d forces w hich h ad created m ark s o n b ones recovered from
v ario u s p re h isto ric contexts. In re m a rk ab le an tic ip a tio n o f m uch m o d ern
research. B u ckland (1823:37-38), for exam ple, w rote:
I have had an o p p o rtu n ity o f seeing a C ap e hyaena at O xford . . . I w as enabled also
to observe the anim al’s m ode o f proceeding in the d estruction o f bones: the shin
bone o f an ox being presented to this hyaena, he began to bite off w ith his m o lar
teeth large fragm ents from its u p p er extrem ity, and sw allowed them w hole as fast as
they were broken off. O n his reaching the m edullary cavity, the b one split into
ang u lar fragm ents . . . he w ent on cracking it till he h ad ex tracted all the m arro w . . .
this done, he left u n to u ch ed the low er condyle, w hich contains no m arro w , and is
very hard . T he state an d form o f this residuary fragm ent are precisely like those o f
sim ilar bones a t K irkdale; the m ark s o f teeth on it are very few . . . these few,
how ever, entirely resem ble the im pressions we find on the bones o f K irkdale; the
sm all splinters also in form an d size, an d m an n er o f fracture, are n o t distinguishable
from the fossil ones . . . there is absolutely no difference betw een them , except in
p o in t o f age.
T he significance o f this o b serv atio n w as n o t lost o n m an y p re h isto ria n s

w o rk in g in the n in eteen th century. F o r exam ple, 50 years afte r its original
p u b licatio n the ab o v e c itatio n w as q u o te d a t g re ater length by D aw kins
(1874:281-283).
T o u rn a l (1833), citing B u ck lan d ’s research, to o k the idea o f hyenas as agents
o f b o n e ac cu m u latio n in caves one step fu rth e r. H e extended the an alo g y from
traces o f gn aw in g on b ones signifying hyenas in p a rtic u la r to signifying

carn iv o res in general.
T he m an n er in w hich the b ro k en an d gnaw ed bones o f different species are
accum ulated is easy to conceive. O ne has, so to speak, surprised n a tu re in the act,
w hen one has observed in o u r tim e th e charnel houses o f hyenas and oth er
carnivorou s anim als w hich carry th eir prey into g ro tto es in o rd e r to e at them , and
w hich a t length accum ulate im m ense quan tities o f gnaw ed bones, belonging to all
sorts o f anim als. N ow , in the case w hich concerns us, the identity is perfect, since
w ith the hyenas are fo u n d the bones which they have gnaw ed, an d even their
coprolites. A t the sam e tim e it has been observed th a t in the caverns the bones
accum ulated in the m ost rem ote passages. W h at has ju s t been said fo r the hyena
bone caverns applies equally to caverns w hich co n tain less ferocious o r sm aller-
sized carnivores.
(T ou rn al 1833 [1969:87])
N o te th a t T o u rn a l ex ten ded the an alo g y by lo o k in g n o t ju st at the gnaw ing

d am ag e on th e bo n es o f prey, b u t also a t the lo catio n s o f the b ones an d th eir
asso ciatio n s w ith th e rem ains a n d traces o f th eir p re d ato rs.
T he fact th a t som e o f the early ta p h o n o m ic lite ratu re w as pu b lish ed in
E u ro p e a n jo u rn a ls in n o n -E n g lish languages m ight have m ade som e o f th a t
in fo rm a tio n inaccessible to A m ericans, b u t tw o facts m ak e it clear th a t this did
n o t h in d er ta p h o n o m ic research in the N ew W orld. F irst, M o rlo t's (1861)
“ G en eral views on a rc h ae o lo g y ” w as published in E nglish by the S m ith so n ian
In stitu tio n . H e su m m arized som e E u ro p e a n research on the d e stru c tio n o f bird
b ones by canids. As well, M o rlo t (1861:300) n o ted " n e a rly all the cartila g in o u s
an d m o re o r less soft p a rts o f the [deer] b ones have been irregularly su b
tra c te d ,” an o b serv atio n suggested as well by th e avian ta p h o n o m y research he
cited. T h a t th e stru c tu re o r density o f skeletal elem ents w ould influence
w h e th e r o r n o t they survived a ttritio n a l processes w as also n o ted by D aw kins
(1869:207) w ho w rote th a t “ the stone-like m o lars o f the M a m m o th w ould
survive th e d estru c tio n o f all traces o f the b ones o f the sm aller anim als, a n d
rem ain in m an y in stances as th e sole evidence th a t P ostglacial m am m als ever
dw elt in the area w here they w ere fo u n d .” Such o b serv atio n s o n the in h ere n t
p ro p e rtie s o f fa u n al rem ains m ed iatin g ta p h o n o m ic processes an ticip ated
research th a t w as n o t to be ta k e n u p w ith rig o r u n til 100 years later.
T h e second re aso n I suspect there w as tran so c ean ic exchange o f ideas on
ta p h o n o m y involves explicit reference to the avian ta p h o n o m y cited by M o rlo t
by A m erican arch aeo lo g ists such as W y m an (1868), w ho w as concerned w ith
u n d ersta n d in g the fo rm a tio n o f w h a t are to d ay term ed shell m iddens. F u rth e r,
W y m an (1868:578) ex p an d ed the range o f ta p h o n o m ic a ttrib u te s one m ight
exam ine by n o tin g “ on loo k in g over the specim ens o f o u r collections, m ark s o f
teeth o f an im als w ere frequently noticed, som e o f them o f such size as m ight be
m ad e by dogs, b u t o th ers by a m uch sm aller an im al, as a cat o r m in k .”
U niquely. W ym an (1868) suggests th a t the size o f bo n e fragm ents w as a result
o f h u m an s b re ak in g the b ones so th a t the fragm ents co u ld be co o k ed in ceram ic
vessels w ith orifices o f sm all size o r lim ited capacities.
B u ck lan d ’s a p p ro a c h , involving ex a m in a tio n o f the a ttrib u te s o f m o d ern

bones th a t h ad been m odified by k n o w n ta p h o n o m ic processes in o rd e r to have
an an a lo g fo r in terp re tin g p re h isto ric specim ens, w as extended to include the
stu d y o f h u m a n ag en ts as well. L a rte t (1860 [1969:122]), fo r exam ple, believed
th a t th e “ very deep incisions” he observed on vario u s b o n es w ere th e result o f
them hav in g been c u t w ith “ an in stru m e n t having an edge w ith slight transverse
inflections, so as to p ro d u ce, by cu ttin g obliquely th ro u g h the bone, a p lan e o f
section so m ew h at u n d u la te d .” L a rte t (1860 [1969:123]) “ satisfied [himself], by
co m p arativ e trials on h o m o lo g o u s p o rtio n s o f existing anim als, th a t incisions
presen tin g such a p p e ara n ces co u ld only be m ad e in fresh b o n es,” an d he
o b tain ed “ a n a lo g o u s results by em ploying as a saw flint knives, o r [stone]
splinters w ith a sh a rp chisel-edge.” Such ex perim ental w o rk to derive analogs
for in terp re tin g the rem ains o f p a s t h u m a n activities w as to becom e a h allm ark
o f m o d e rn arch aeo lo g y (C h a p te r 3).
Peale (1871:390) re p o rte d th a t bones o f prey o f N o rth A m erican In d ian s are
" ro a ste d o n the coals o r b efore the fire, th en split w ith a stone h atch e t, a n d in
som e cases w ith a w edge d riven betw een the condyles w hen the bo n e has these
term in atio n s. [Som etim es bones] are b ro k e n in to sm all frag m en ts a n d boiled in
w ater u n til all the m arro w w hich they c o n ta in a n d the grease w hich ad h eres to
them are sep a rated , a n d rise to th e surface, w hen they are skim m ed o f f ’ a n d
stored. L o rd A v eb u ry (1913:321; b e tte r k n o w n as Sir J o h n L u b b o ck ) used
e th n o g ra p h ic in fo rm a tio n fro m D e n m a rk , A frica, a n d the A rctic, as an
an alog y to explain th e co n d itio n o f som e p re h isto ric m am m al rem ains. H e
w rote, “ som e o f the an cien t sto n e h am m ers a n d m o rta rs w ere no d o u b t used for
this p u rp o se , a n d th e p ro p o rtio n s o f the different bones afford us, I think,
indirect evidence th a t a sim ilar cu sto m [of b re ak in g bones fo r m arro w
ex traction ] p revailed am o n g the an cien t in h a b ita n ts o f so u th e rn F ra n c e .” L o rd
A v eb ury a p p a re n tly believed fra g m e n ta tio n destro y ed som e bones, a them e
picked up in later w o rk . H e also discussed (a) fra g m e n ta tio n p a tte rn s d istin c
tive o f a h u m a n fractu re ag en t, (b) th e influence o f fra g m e n ta tio n on the
tax o n o m ic id en tifiability o f v erte b rate rem ains, (c) the differential d estru c tio n
o f long b o n e epiphyses a n d diaphyses, a n d d e stru c tio n o f v erteb rae a n d ribs by
canids, (d) the c o rre la tio n o f o n to g en ic age o f an u n g u late skeleton a n d the
stru c tu ra l density o f bones, a n d (e) the c o rrelatio n o f s tru c tu ra l density o f bone
p a rts w ith th e n u tritio n a l value o f th o se p a rts fo r can id s (i.e., less dense bones
co n tain e d m ore b lo o d , m arro w , a n d grease an d th u s w ere b o th m ore n u tritio u s
an d m o re p ro n e to d estru c tio n by b o n e-g n aw in g carnivores). T his list o f
ta p h o n o m ic top ics is easily fo u n d in the lite ra tu re o f the 1980s.
T he range o f ta p h o n o m ic topics covered by n in etee n th -cen tu ry p re h is to r
ians, w hile n o t com pletely d o cu m en te d in this section, is sufficiently b ro a d th a t
I h ave to w o n d e r w h a t h ap p e n ed to this line o f in q u iry in the first h a lf o f the
tw en tieth cen tury; it seem s largely to have d isap p ea red from the published
reco rd d u rin g th a t perio d. It p ro b a b ly d isap p ea red d u e to the shift fro m the
cal reco rd to th e tw en tieth -c en tu ry focus o n establishing the te m p o ra l relations

o f arch aeo logical collections (D aniel 1981; T rigger 1989; W illey a n d S abloff
1980). W h atev er the reaso n , it is difficult to find co m m ents on ta p h o n o m ic
issues in th e arch aeo log ical lite ra tu re published d u rin g the first six decades o f
the tw en tieth ce n tu ry th a t echo the concerns o f the n in etee n th century. In the
early tw en tieth cen tu ry the tendency w as to conclude th a t people were the sole
ta p h o n o m ic ag en t one h a d to ac co u n t fo r d u rin g analysis a n d in te rp re ta tio n o f
a rc h a e o fa u n a l collections. T a p h o n o m ic research in the early a n d m iddle
tw en tieth ce n tu ry is largely fo u n d in the paleo n to lo g ical literature.
T aphonom y in the early and m iddle tw entieth century
T here is g reat need fo r b etter u n d erstan d in g o f the factors th a t act betw een the living
fauna and the preserv atio n o f p a rt o f it in the fossil state, as well as factors involved
in the fo rm atio n o f fossil deposits in general.
(G. G . Sim pson 1961:1683)
W eigelt (1927; E nglish tra n s la tio n 1989) suggested the term biostratinom y
(originally b io stra to n o m y ) w hich now is tak en to d en o te “ the study o f the
en v iro n m e n ta l facto rs th a t affect o rg an ic rem ains betw een a n o rg a n ism ’s d e a th
a n d th e final b u rial o f the re m a in s” (L aw rence 1979a:99). W eigelt’s (1927/
1989) volum e has been called " th e first m a jo r w ork on v erte b rate ta p h o n o m y ”
w ritten by “ th e first n a tu ra list to m o u n t a full-scale research effort to d o cu m en t
processes o f v erte b rate d ea th , decay, d isarticu latio n , tra n s p o rt, a n d b u rial, a n d
to d eterm ine th eir relevance to fossil p re se rv a tio n ” (B ehrensm eyer an d B adg-
ley 1989:vii, viii). L ater, M u ller (1963) used the term fossildiagenese (now
sim ply diagenesis) to d en o te “ fossilization events th a t tak e place afte r the final
b u rial o f org anic re m a in s” (L aw rence 1979b:245). T he re la tio n sh ip o f ta p h o
n om y, b io stratin o m y , a n d diagenesis w as often show n as in F ig u re 2.1 d u rin g
th e m iddle o f the tw en tieth cen tu ry (e.g., L aw rence 1968; N o e -N y g a a rd 1977).
R ich ter (1928) defined aktuo-paldntologie as the science o f the origin an d
p resen t-d ay m o d e o f fo rm a tio n o f fu tu re fossils (W arm e a n d H an tzsch el 1979).
A ctu alistic p aleo n to lo g y is “ the a p p lic a tio n o f the u n ifo rm ita ria n principle to
p aleo n to lo g ical p ro b le m s” (W arm e an d H antzschel 1979:4; see C h a p te r 3).
R ich ter (1928) also d istinguished the im p o rta n c e o f th e causes o f d e a th an d
th eir d irect consequences, b u ria l m ode, a n d alte ra tio n s o f the an im al carcasses
p rio r to diagenesis (W arm e a n d H antzschel 1979:5-6). It was E frem ov (1940),
th en , w ho sub su m ed all o f these stages betw een th e d e a th o f an org an ism and
the recovery o f its rem ains by a p aleo b io lo g ist in to th e field we to d a y label
taphonom y. P erh a p s because E frem o v pu b lish ed his new term a n d its definition
in a geological jo u rn a l, a n d m o st p revious w o rk o n the subject h a d been also
pu b lish ed in such ou tlets, th ere w ere, d u rin g th e first h a lf o f the tw entieth
cen tu ry , w h at K o ch (1989) has called tw o parallel tra d itio n s fo r the dev elo p
m en t o f tap h o n o m y : a p aleo n to lo g ical tra d itio n , a n d a n arch aeo lo g ical one.
The history an d structure o f taphonom y 17
Figure 2.1. G eneral relations o f the subdisciplines o f tap h o n o m y relative to an

an im al’s life, death, an d scientific recovery (after L aw rence 1968:1316, Figure 1).
T h eir coalescence in th e 1970s w as a m a jo r historical event in the developm ent

o f ta p h o n o m y as a research field in its ow n right, b u t first I co n sid er each
tra d itio n in tu rn .
P aleontological tradition
T his tra d itio n o f ta p h o n o m ic research is “ as old as p a le o n to lo g y ” (C adee
1990:17). C adee (1990:4-5) argues th a t “ ta p h o n o m ic reaso n in g in the tru est
sense” w as p erh ap s used first by L e o n a rd o d a V inci in the late fiftee n th -ea rly
sixteenth centuries, a lth o u g h his n o tes w ere n o t w idely available u n til after his
d eath . C adee also no tes th a t N iels S tensen (often referred to as N . S teno) m ade
m a jo r c o n trib u tio n s to ta p h o n o m ic re aso n in g d u rin g the seventeenth century.
T he estab lish m en t o f p aleo n to lo g y as a scientific discipline in the early
n in eteen th ce n tu ry resulted in the p u b licatio n o f n u m ero u s ta p h o n o m ic
o b serv atio n s, m an y o f w hich w ere fo u n d ed o n actu alistic research (see below ).
By th e end o f the n in etee n th century, G e rm a n y w as the cen ter o f p a le o n to lo g i
cal research (R udw ick 1976), a n d it w as here th a t m uch ta p h o n o m ic research
was p erfo rm ed at this tim e, a lth o u g h the G e rm a n school found little su p p o rt
o u tsid e o f G e rm a n y in th e early tw en tieth ce n tu ry (C adee 1990:9-13). T his is
im p o rta n t because the w ork o f a u th o rs such as W eigelt (1927/1989) an d
S chafer (1962/1972) is included in the G e rm a n school.
M u ller (1951) p resen ted a m a jo r synthesis o f m uch o f the p aleontologically
related ta p h o n o m ic research to th a t p o in t in tim e, especially the b io stratin o m ic
aspects. S chafer (1962/1972) su m m arized a n extensive b o d y o f research
co n cern in g th e d isin te g ratio n o f m arin e organism s, especially in v erteb rates.
T h e influence o f these a n d o th e r w o rk s o n E nglish-speaking p aleo n to lo g ists
m ight n o t have been significant due to language b arriers (B ehrensm eyer and
B adgley 1989; G iffo rd 1981:370), b u t O lson (1952, 1958), C lark et al. (1967),
L aw rence (1968), a n d V oorhies (1969) w ere influenced by the E u ro p e a n
research a n d th eir E nglish lan g u ag e p u b licatio n s helped b reach the language

b arrier. As well, increasin g in terest in the A m ericas in ta p h o n o m ic research and
issues d u rin g th e 1970s is a ttrib u te d by B ehrensm eyer a n d K idw ell (1985) to the
1972 p u b licatio n o f the E nglish tra n sla tio n o f S chafer's (1962/1972) volum e on
actu o p a leo n to lo g y . B ecause m o st o f these early w ritings w ere w ritten by
p aleo n to lo g ists in terested in p aleoecological issues, the tendency in this early
lite ra tu re was to focus on the fossil record as p o ten tially biased in term s o f how
well it reflected th e ac tu a l paleoecology o f the biotic co m m u n ity (F ig u re 2.2).
R eco g n itio n th a t th e fossil reco rd is p o ten tially biased resulted in m an y
an aly sts a tte m p tin g to d etect a n d co m p en sa te fo r th o se biases. S hotw ell’s
(1955) q u a n tita tiv e tech n iq u e o f m easu rin g how com plete the skeletons o f
in d iv id u al tax a w ere in o rd e r to distinguish tax a th a t lived n ea r th e site o f fossil
recovery (tax a m ak in g u p this p ro x im a l com m unity should have the m o st
co m p lete skeletons) from tax a th a t lived fa rth e r aw ay is a classic exam ple o f
developing an an aly tic tech n iq u e to co m p en sate for the ta p h o n o m ic processes
o f b o n e a c cu m u latio n a n d c o n c e n tra tio n . S hotw ell’s tech n iq u e w as a d a p te d to
the im p o rta n t zo o a rch aeo lo g ical q u estio n o f d istinguishing ta x a th a t ow ed
th eir presence in sites to h u m an activities (cultural bone) from tax a w hose
rem ains h ad been d ep o sited by n a tu ra l processes (T h o m as 1971; Z iegler 1973).
W hile there are serious p ro b lem s w ith b o th the a p p ro a c h ad v o cated by
S hotw ell fo r p aleoecological analysis an d the w ay the a p p ro a c h is applied to
arch aeo lo g ical assem blages (G ray so n 1978b; H o ltzm an 1979), recognition o f
th a t fact has only com e w ith increases in o u r know ledge o f the influence o f
ta p h o n o m ic facto rs on such q u an titativ ely based an aly tic techniques.
T o o ts (1965c) exam ined the m o d e rn d isa rtic u la tio n o f m am m al skeletons in
a ra re actu alistic stu d y o f the m iddle tw en tieth century. H e also discussed an
an aly tic techn iq ue fo r describing a n d in terp re tin g the o rie n ta tio n o f fossils
(T o o ts 1965a, 1965b), p o in tin g o u t th a t o rie n ta tio n processes w ere n o t
necessarily biasin g a n d th a t o rie n ta tio n d a ta w ere ra th e r v alu ab le to p aleo eco
logical studies. B oth th e stu d y o f d isarticu latio n o f skeletons (e.g.. Hill 1979a,
1979b; Hill a n d B ehrensm eyer 1984, 1985) an d o f fossil o rie n ta tio n (e.g.,
K re u tzer 1988) have been follow ed up recently by arch aeo lo g ists interested in
zo o a rch a eo lo g ical a n d ta p h o n o m ic p ro b lem s, a n d they o ften stress the value o f
such d a ta to research ra th e r th a n as signs o f p o te n tia l bias.
Explicit reco g n itio n o f the effects o f sam pling on p aleo n to lo g ical research
cam e in th e m iddle o f the tw entieth ce n tu ry (K ru m b ein 1965; M cK en n a 1962;
V oorhies 1970), ju s t as it did in arch aeo lo g y (B inford 1964; R ag ir 1967). G iven
th e p o te n tia l effects o f sam pling a n d sam ple size on m easures o f relative
tax o n o m ic ab u n d a n ces, a critical variab le in paleoecology, an aly tic techniques
to co n ten d w ith such effects w ere developed (e.g., S anders 1968). Som e o f these
w ere ad o p te d by zo o arch aeo lo g ists (e.g.. Styles 1981). B ut like S hotw ell’s
tech n iq u e, these tech niqu es fo r co n tro llin g sam pling effects have subsequently
been show n to be fau lty in som e situ atio n s (e.g., G ra y so n 1984; T ip p er 1979).
Figure 2.2. M odeled tap h o n o m ic h istory o f a biotic com m unity o r life

assem blage (after C lark and K ietzke 1967:117, Figure 53).
A gain, w h a t is u n d ersco red here is not the in accu racy o f techniques designed to
deal w ith com plex ta p h o n o m ic issues, b u t ra th e r th e fact th a t we co n tin u e to
learn in the 1990s a b o u t th a t com plexity a n d the lim its o f o u r analytical
ap p ro ach es.
T o d a y ta p h o n o m y is very m u ch a p a rt o f paleo n to lo g y . In a recent synthesis
o f p aleob io lo g y (Briggs a n d C ro w th e r 1990) a p p ro x im a te ly 100 pages are
d ev o ted to ta p h o n o m y . In fact, p aleo b io lo g ists b o rro w e d a term fro m m ining
geology, fo ssil-L a g erstd tten , to label s tra ta th a t are sufficiently rich in fossils o r
in w hich fossils are preserved in co n d itio n s sufficient “ to w a rra n t ex p lo itatio n ,
if only fo r scientific p u rp o se s” (S eilacher 1990:266). Sim ilarly, “ taphofacies
co n sist o f suites o f ro c k ch aracterized by p a rtic u la r co m b in atio n s o f preserva-
tio n al featu res o f th e co n tain e d fossils” (B rett an d Speyer 1990:258). O brution
deposits are “ co n se rv atio n L a g e rs ta tte n ” th a t co n sist o f ra p id ly bu ried , a n d
th u s ex ceptio nally well preserved, a rtic u la te d fossils (B rett 1990b:239), an d
“concentration L a g ersta tte n ” are dense c o n c e n tra tio n s o f fossils th a t m ay
co n sist o f m u ltip le d ep o sitio n al events (S eilacher 1990:267). T he florescence o f
ta p h o n o m ic term in o lo g y in d icated by these an d o th e r labels, a n d the recent
p u b licatio n o f tw o edited volum es dev o ted to ta p h o n o m y fo r p aleo n to lo g ists
(A llison a n d Briggs 1991b; D o n o v a n 1990) is a n effective in d icatio n o f the
co n tin u ally g row ing im p o rtan ce o f ta p h o n o m y in paleontology.
Archaeological tradition
W h a t w o u ld to d a y be co n sid ered ta p h o n o m ic research fo r archaeological
p u rp o ses w as ra re in th e early p a rt o f the tw en tieth century, b u t som e
im p o rta n t studies w ere u n d e rta k e n , such as D u c k w o rth ’s (1904) o b serv atio n s
th a t sm all stream s co u ld disperse the b o n es o f a h o rse skeleton. M a rtin (1910)
experim en tally b ro k e m am m al bones in an a tte m p t to discern c riteria d en o tin g
a h u m a n ag en t o f fractu re. B reuil(1932. 1938, 1939) a n d Pei (1932, 1938) were
also co n cern ed w ith how b ones m ig h t be m odified (b ro k en , incised) by h o m in id
an d o th e r ta p h o n o m ic agents. P erh ap s because som e o f this research was
pu b lish ed in n o n -E n g lish languages in E u ro p e a n jo u rn a ls, it seem s to have h ad
little im p act on A m erican archaeology. A lso, the research focus in the
A m ericas a t th is tim e o n cu ltu re h isto ry issues (W illey a n d S abloff 1980) m ade
these early ta p h o n o m ic studies irre le v an t to the concerns o f m o st practicin g
archaeo lo g ists.
T he im p o rta n c e o f ta p h o n o m ic research for z o o a rch a eo lo g y w as b ro u g h t
o u t clearly w ith D a rt’s (e.g., 1949, 1956b. 1957, 1960) p u b licatio n s o n his
p ro p o se d o ste o d o n to k e ra tic cu ltu re o f the au stralo p ith ecin es based o n collec
tio n s o f m o stly bov id rem ains recovered fro m lim estone caves in S o u th A frica.
D a rt w as co nvinced th a t the w ay som e bones w ere b ro k e n a n d the relative
a b u n d a n ces o f skeletal p a rts in d icated ho m in id s w ere th e ta p h o n o m ic agent
largely responsible fo r the fa u n al rem ains. D a rt (1956a; H ughes 1954) utilized
actu alistic d a ta in his arg u m e n ts a n d actively c o n trib u te d to discussions a b o u t
the artificial statu s o f m o difications to bones (e.g., D a rt 1958). R egardless o f

the eventual o u tco m e o f the d e b a te over w h eth er o r n o t the S o u th A frican
specim ens w ere v ario u sly m odified, ac cu m u lated , a n d d ep o sited by hom inids
(the c o n tro v ersy is n o t dead; e.g., B rain 1989; W olberg 1970), D a rt's ideas w ere
p ro v o cativ e a n d served as a m a jo r ca ta ly st for the d evelopm ent o f explicit
ta p h o n o m ic research in the service o f arch aeo lo g y (e.g., B onnichsen 1975;
B rain 1967a, 1967b, 1969, 1974, 1981; Hill 1976; Isaac 1967; R ead 1971; R ead-
M a rtin a n d R ead 1975; S hip m an a n d Phillips 1976; S hip m an a n d Phillips-
C o n ro y 1977).
T his aw ak en in g interest in arch aeo lo g ical ta p h o n o m y w as reinforced in the
A m ericas w ith th e re p o rt o f an excessively old b o n e to o l fro m n ear O ld C row in
the Y u k o n o f n o rth w e ste rn C a n a d a (Irving a n d H a rin g to n 1973). T h a t the
specim en in q u estio n was a to o l was never in d o u b t; its age w as, how ever, the
subject o f m u ch discussion a n d it has since been show n to be o f m iddle to late
H olocene age (N elson et al. 1986). B ut the original re p o rt, like D a r t’s, resulted
in a flurry o f p aleo n to lo g ical, arch aeo lo g ical, an d actualistic research (e.g.,
B onnichsen 1973, 1979; Jo h n so n 1985; L ym an 1984b; M o rla n 1980), som e o f
w hich w as re m a rk ab ly sim ilar to th a t o f Peale (1871) 100 years earlier (e.g.,
Z ie rh u t 1967). Interesting ly, as w ith the un settled statu s o f a n early P leistocene
A frican o ste o d o n to k e ra tic cu ltu re, the reality o f a m iddle to late Pleistocene
A m erican cu ltu re fo u n d e d on sim ilar to o l m ateria ls is n o t yet settled (e.g.,
Irving et al. 1989; M o rla n 1988).
A new level o f aw areness o f the im p o rtan ce o f ta p h o n o m ic processes was
atta in e d by the late 1960s a n d early 1970s. M ichael V o o rh ies' (1969) sem inal
m o n o g ra p h on v erte b rate ta p h o n o m y h ad ju s t been published, alo n g w ith the
im p o rta n t w ork o f C lark et al. (1967). T h e term ta p h o n o m y now ap p e are d in
th e title o f som e p aleo n to lo g ical studies (e.g., B oyd a n d N ew ell 1972; D o d so n
1971) w hereas it w as n o t even listed in th e index o f K u m m el a n d R a u p ’s 1965
H andbook o f Paleontological Techniques. G iven the new level o f aw areness
b o rn in th e late 1960s, it is n o t su rp risin g to find synopses o f p ro ced u res for
reco rd in g ta p h o n o m ic d a ta published d u rin g the early 1970s. O ne w as a u th
ored by tw o p ale o a n th ro p o lo g ists in terested in h o m in id ev o lu tio n (H ill an d
W alk er 1972); a second w as a u th o re d by tw o p aleo n to lo g ists (M u n th e an d
M cL eo d 1975). B o th articles focus on v e rte b ra te rem ains a n d are q u ite sim ilar
in th eir re co m m en d atio n s o f w h a t co n stitu te s im p o rta n t ta p h o n o m ic d a ta
(T able 2.1). In m an y respects the list o f d a ta is an ac cu ra te reflection o f w h a t is
typically reco rd ed to d ay , a lth o u g h m any o f the variables are now m ore
explicitly defined.
Som e zo o a rch a eo lo g ists p ro d u c ed explicit statem en ts a b o u t relev an t ta p h o
nom ic d a ta even th o u g h they did n o t use the term “ ta p h o n o m y ” in th eir
w ritings. G u ild ay et al. (1962:65), for exam ple, re p o rt th a t m an y w hite-tailed
deer (Odocoileus virginianus) bones fro m a n arch aeo lo g ical site they studied
h ad been gnaw ed by ca rn iv o res, a n d suggest th e less “ d u ra b le ” b ones o f
T ab le 2.1 Kinds o f taphonom ic data that should be recorded f o r vertebrate

fo s s il rem ains recom m ended by H ill and W alker (1972) and M u n th e and
M c L e o d (1 9 7 5 ). D istinction o f fie ld and laboratory data by L ym a n
F IE L D D A T A
1. G eographic locality o f fossil collection site (m apped, p h o to g rap h ed )
2. C ond itio ns an d tim e o f collection (visibility, w eather, date)
3. C ollection m ethod s and possible biases
4. S tratigraphic positio n o f fossils, including general geology an d to p o g rap h y o f collection
site and lithology and sedim entology o f stratu m (strata) co n tain in g fossils
5. H o rizo n tal an d vertical d istrib u tio n o f fossils - three dim ensional p o in t provenience o f
each individual specimen
6. A ssociated p la n t an d invertebrate fossils
7. D egree o f d isarticu lation an d to o th loss
8. O rien tatio n o f individual bones a n d /o r articu lated skeletons - azim uth o f long axis -
declination o f long axis
9. F lexion o r extension o f articu lated specimens
LA BO RA TO R Y DATA
10. T axonom ic identification
11. List o f elem ents present
12. A ttribu tes o f individual specimens
a. cracks an d flaking (w eathering)
b. fractures
c. crushing
d. abrasion
e. color
f. ro o t etching
g. d istortio n an d d efo rm ation
h. gnaw ing m arks
i. weight
j. shape (m easured as a m axim um length:m axim um w idth ratio)
k. area (m easured as the p ro d u c t o f m axim um length x m axim um w idth)
13. T axonom ic abundances
14. Size range o f taxa
15. O ntogenic age o f individual anim als
on to g en ically y o u n g deer are p ro b a b ly relatively ra re because o f th a t gnaw ing.

B onnichsen a n d S anger (1977) em phasize th a t b ones are n o t ra n d o m ly
d istrib u te d in sites (c o n tra Z iegler 1973) an d urge g re ater a tte n tio n to the
co n tex t an d asso ciatio n o f fau n al rem ains. T he published proceedings o f the
1974 A rch aeo zo o lo g ical C onference includes 47 articles c o n trib u te d by 49
a u th o rs from a ro u n d th e w orld, yet the term s ta p h o n o m y , b io stratin o m y , an d
diagenesis are n o t listed in the index (C lason 1975). Im p o rta n t tap h o n o m ic
to pics are non etheless discussed in th o se articles, including p a tte rn s o f bone
fractu re, traces o f in jury to anim als reco rd ed in th eir bones, bo n e frag m en t
sizes, a n d gnaw ing a n d ac cu m u latio n o f b ones by carnivores.
T a p h o n o m ic facto rs h ad seldom been discussed by zo o arch aeo lo g ists p rio r
to 1970. C o rn w all (1956) focused o n the effect fra g m e n ta tio n w ould have on
the identifiability o f an im al rem ains. H is only ta p h o n o m ic o b serv atio n s w ere
th a t n atu ra lly d ep o sited b ones “ will have suffered the ravages o f scavengers
an d o f decay, so th a t only th e m o re m assive a n d d u ra b le p a rts are likely to have
been p reserv ed ,” cu ltu rally d ep o sited an im al rem ains are likely to have been
“ sm ashed fo r th eir m arro w o r to o b ta in in d u strial m a te ria l,” a n d “ in fav o rab le
circu m stances the m aterial is fo u n d m u ch as it w as d iscard ed by m an , b u t very
o ften n a tu ra l agencies have fu rth e r affected it, so th a t, again, only the m ore
re sistan t frag m en ts are a v a ila b le” (C o rn w all 1956:184). C h ap lin (1971)
dev o ted tw o a n d a h a lf pages to “ the lim iting facto rs o f the archaeological
evidence.” H e em ph asized the need to d eterm in e how bones h ad been
accu m u lated a n d d ep o sited , statin g " th e tim e has surely com e to ap p ly som e
m ore conclusive tests to the p ro b le m ” (C h ap lin 1971:121). H e w as also w orried
a b o u t how rep resen tativ e o f the com plete site d ep o sit the collected faunal
rem ain s were, b u t offered only the suggestion th a t the an aly st assum e the
availab le sam ple w as a “ fa ir cross sectio n ” o f w h a t w as actually in th e deposit.
T h e term ta p h o n o m y w as n o t used in eith er C o rn w a ll’s o r C h a p lin ’s b o o k ,
alth o u g h b o th clearly were co n cerned a b o u t ta p h o n o m ic histories an d the
p o te n tia l biases created by such histories. T he sam e ca n be said fo r m o st o f the
jo u rn a l articles p u blish ed betw een 1900 a n d 1970. B ecause these articles seem
to be directed m o re to w a rd ed u c atin g arch aeo lo g ists a b o u t the value o f faunal
rem ains fo r ad dressin g a n th ro p o lo g ic a l an d zoological questions, it is perh ap s
n o t su rp risin g th a t the a u th o rs did n o t m en tio n th a t ta p h o n o m ic facto rs m ight
decrease the value o f th ose rem ains by som ehow biasing them (e.g., C h ap lin
1965; D aly 1969; G ilm o re 1949; W hite 1953c, 1956; W in tem b erg 1919; Ziegler
1965).
M ed lo ck 's (1975) review o f “ faunal an aly sis” in N o rth A m erica co n tain e d a
th o ro u g h review o f th e significance o f processes influencing w h a t a zo o ar-
ch aeo lo g ist m ight recover fro m a site. H e p o in ted o u t th a t, as o f th e early 1970s,
zo o arch aeo lo g ists h a d v ariously in terp re ted v aria tio n s in fa u n al assem blages
as resu ltin g fro m “ tem p o ra l o r cu ltu ra l differences, differential preserv atio n ,
b u tch erin g differences, o r fu n c tio n a l differences a m o n g sites” (M edlock
1975:224). M edlock em phasized the need fo r developing an aly tic techniques
for d isting uishin g th e effects o f these v ariables, techniques th a t them selves
sh o u ld be verified w ith ex p erim en tal w o rk . H e presen ted a m odel fo r the
fo rm a tio n o f a zo o a rch aeo lo g ical record th a t included virtually all variables in
m odels later form ally p ro p o se d as depicting a generalized ta p h o n o m ic history
(F ig u re 2.3).
M o re accessible to arch aeo lo g ists w as th e series o f p ap e rs on the ta p h o n o m y
a n d p aleoecology o f P lio-P leistocene h o m in id sites in A frica by B ehrensm eyer
( 1975a, 1975b, 1979; B oaz a n d B ehrensm eyer 1976). She focused n o t only on
the biasin g facto rs b u t also o n the paleoecological d a ta th a t could be derived
from d etailed ta p h o n o m ic analyses, a n d she w o rk ed to w ard s build in g stro n g
Figure 2.3. M ed lo ck ’s (1975) m odel o f the tap h o n o m ic history o f a faunal

assem blage (after M edlock 1975:225, Figure 2).
an alo g ical arg u m e n ts by p erfo rm in g v ario u s actu alistic a n d experim ental w ork
(B ehrensm eyer 1978, 1981; B ehrensm eyer a n d B oaz 1980; B ehrensm eyer et al.
1979; G iffo rd a n d B ehrensm eyer 1977). N o less im p o rta n t c o n trib u tio n s were
m ad e by several in d ivid uals w ork in g o n the ta p h o n o m y o f P lio-Pleistocene and
o th e r sites in A frica (e.g., B rain 1969, 1974, 1976; G ifford 1977; H ill 1978,
1979a, 1979b; Isaac 1967; S hip m an an d Phillips 1976; S hipm an a n d Phillips-
600
530
E Koch
g 500 Lyman
o
Q.
400
c
o
w
o 300
ro
o
5
Q. 200
05 100
n -
E
3 6 5 11
2 2 2 7
1 1 0 0 3 2 4
1
o o o o o o o o o o o o o o o o o
C\J CO in CD r^ CO O) o T— eg CO in id s co
CO CO CO CO 00 CO CO CO O) O) O) O) O) 0 0) 0) 0)
T— T-- T-- i— i— i— T— T— T— T— I— 1— T—
Figure 2.4. F requencies o f titles o f tap h o n o m ic literatu re p er decade (d ata from

K och 1989, supplem ented by Lym an).
C o n ro y 1977; S h ip m an a n d W a lk e r 1980). M u ch o f this research w as p resented

at the B urg W arten stein C onference in 1976 (B oaz 1980), the proceedings o f
w hich w ere pu b lish ed in 1980, u n d e r B ehrensm eyer’s co -ed ito rsh ip (B ehrens
m eyer an d Hill 1980).
R esearch ers w o rk in g in the N e ar E ast (e.g., G ilb ert 1979), E u ro p e (N oe-
N y g a a rd 1977; P ayne 1972a, 1972b), a n d the A m ericas (e.g., B inford an d
B ertram 1977; B riuer 1977; R ead 1971) also published ta p h o n o m ic analyses.
P aleo n to lo g ists pu b lish ed m u ch ta p h o n o m ic research in the 1970s a n d 1980s;
B ehrensm eyer a n d K idw ell (1985:109) re p o rt an “ average p u b licatio n rate in
ta p h o n o m y o f n early 50 articles p er y e a r” betw een 1975 a n d 1984, inclusively.
K o c h ’s (1989) b ib lio g rap h y o f ta p h o n o m y p rovides a g o o d sam ple o f refer
ences th ro u g h early 1987, w hich w hen su p p lem en ted w ith th e n ineteenth-
cen tu ry titles cited abov e a n d titles fo r 1987-1989, illu strates well the increase
in ta p h o n o m ic research (b o th th a t related to p aleo n to lo g y , a n d th a t related to
arch aeo lo g y ) th a t cam e w ith the 1970s (F ig u re 2.4). F ro m the 1940s to the
1950s th e n u m b e r o f titles n early q u a d ru p le d ; p erh ap s th a t increase is d ue to the
p o st-W o rld W a r II increase in all scientific endeavors. T he n u m b er o f titles then
do u bles from the 1950s to the 1960s, a n d d oubles ag ain from the 1960s to the
1970s. T here is only a 25.8% increase from the 1970s to the 1980s; if the
p u b licatio n ra te fo r th e 1970s (42.1 titles p er year) held in the 1980s, a b o u t 675
to 700 titles sh o u ld be recorded fo r the 1980s ra th e r th a n the 530 p lo tte d in
F ig u re 2.4. I suspect m an y titles are n o t included as th ere are 27 titles on
ta p h o n o m y p ublished in th e first fo u r issues o f the jo u rn a l Palaios w hich are

n o t tallied in F igure 2.4, a n d o th ers (e.g., N o e -N y g aa rd 1989; P lotnick and
Speyer 1989; W ilson 1988) w ere discovered afte r the figure w as com pleted.
F u rth e r, I have ten ded to exclude p u b licatio n s in n on-E nglish languages. N o
titles m o re recent th a n 1989 are included in F ig u re 2.4, b u t a p eru sal o f the
b ib lio g rap h y for this volum e will show the rate o f p u b licatio n o f ta p h o n o m y
articles has n o t significantly a b a te d in the early 1990s. D a ta presented by
Russell (1992) indicate the ra te o f p u b licatio n o f titles on ta p h o n o m y , as
com piled in the G E O R E F d a ta base (i.e., titles pu b lish ed in geological an d
p aleo n to lo g ical jo u rn a ls), seem s to have increased betw een 1982 a n d 1988, b u t
then leveled o ff from 1989 th ro u g h 1991. W h eth er o r n o t the p u b licatio n o f
articles a n d b o o k s on ta p h o n o m y co n tin u es at its p resen t ra te in to the late
1990s rem ains to be seen.
In N o rth A m erica, the 1960s w itnessed a shift from studying c u ltu re history
to a focus o n h u m a n b eh a v io r as it w as reflected in the arch aeo lo g ical record
(B in fo rd 1968; W illey a n d S abloff 1980). In terest in ta p h o n o m ic issues in the
service o f arch aeo lo g y cam e w ith the aw ak en in g re alizatio n in the late 1960s
a n d early 1970s th a t th e arch aeo lo g ical reco rd w as n o t a perfect reflection o f
h u m a n b eh av io r (A scher 1961a, 1961b, 1968; B inford 1964. 1977; Isaac 1967;
Schiffer 1972, 1976). E x trac tin g the dynam ics o f h u m a n b eh av io r from a static
arch aeo lo g ical reco rd w ould n o t be a stra ig h tfo w a rd process. In co n ju n ctio n
w ith a grow ing aw areness o f w h at zo o arch aeo lo g ical research could co n trib u te
to o u r know ledge o f h u m an p re h isto ry in general (e.g., C h ap lin 1965; D aly
1969), th e discipline o f arch aeo lo g y w as ready to a d o p t ta p h o n o m ic research as
an im p o rta n t a n d crucial p a rt o f its analy tical to o lk it (see also B inford 1981a;
Schiffer 1983, 1985).
R ecent taphonom ic research: the 1980s and 1990s

R esearch rep orts in archaeozoology, p aleo an th ro po lo g y , paleontology, and paleo
b o tan y are considered incom plete w ith o u t som e discussion o f the taph o n o m ic
history o f the collection.
(C. P. K och 1989:1)
In h er overview o f ta p h o n o m y a n d its relevance fo r arch aeo lo g ical research

G iffo rd (1981) p o in ts o u t th a t ta p h o n o m ic d a ta are im p o rta n t fo r detecting
p o te n tia l biases in th e zo o a rch aeo lo g ical reco rd , a n d no tes th a t ta p h o n o m ic
d a ta can p ro v id e significant paleoecological in fo rm atio n . N o t all a u th o rs
w ritin g a t th a t tim e w ere well a ttu n e d to the la tte r fact. F o r exam ple, Hill
(1978:88) w rites, “ ta p h o n o m y ultim ately deals w ith th e differences th a t exist
betw een an assem blage [of fossils] a n d the co m m u n ity o f anim als fro m w hich it
cam e. These differences co n stitu te the bias in m an y fossil assem blages a n d are
due to vario u s ta p h o n o m ic causes.” H ill’s d escrip tio n o f ta p h o n o m y is w ritten
fro m a p aleo n to lo g ical perspective a n d m ay n o t have been re ad o r com pletely
ap p reciate d by m an y zo o arch aeo lo g ists. In a p a p e r p erh ap s read by m ore

zo o arch aeo lo g ists M ead o w (1981) focuses on p o te n tia l biases th a t can
influence arch aeo lo g ical in te rp re ta tio n s based on q u a n tita tiv e d a ta a n d p re
sents a m odel o f th e fo rm a tio n o f the zo o a rch aeo lo g ical record rem iniscent o f
m odels p u b lish ed by p aleobiologists (F ig u re 2.5). T he decreasing size o f the
sym bols used by M eadow is sym bolic o f the loss o f in fo rm a tio n th a t occurs
d u rin g a ta p h o n o m ic history, as it is fo r C lark et al. (1967; see F ig u re 2.2).
T h e sym bolic loss o f in fo rm a tio n in d icated in F ig u re 2.5 eventually cam e to
d en o te th a t arch aeo log ical fa u n al rem ains are p o ten tially (and p ro b ab ly )
biased. H esse a n d W a p n ish (1985:19) are explicit a b o u t such sym bolism in th eir
dep ictio n o f a ta p h o n o m ic h isto ry (F ig u re 2.6) a n d say, “ the sizes o f the boxes
arra n g e d alo n g the vertical [tem poral] axis reflect the q u ality o f in fo rm atio n
availab le a t v ario u s p o in ts in tim e.” T hey e la b o ra te th a t sym bolism by dividing
each bo x in to tw o d istin ct sets o f in fo rm atio n : C o r cu ltu ra l in fo rm a tio n , a n d N
o r n a tu ra l in fo rm atio n (F ig u re 2.6). Im p o rta n tly , th eir d iag ra m im plies n a tu ra l
in fo rm atio n will, ov er tim e, increasingly m ask o r rem ove cu ltu ra l in fo rm atio n
th a t m ig h t otherw ise be derived from a bo n e assem blage. W hile p ro b a b ly true
at least som e o f th e tim e, this m u st be assessed fo r each assem blage. F inally, it is
un u su al to find cu ltu ra l processes listed as “ biases,” yet H esse a n d W ap n ish
(1985) are co rrect in d o in g so for the sim ple reason th a t bias, as n o ted earlier, is
relative.
In the 1980s p aleo n to lo g ists w ere very aw are o f w h a t co u ld be learned fro m
d etailed stu d y o f the traces o f ta p h o n o m ic processes. In fact, a sym posium
titled “ T h e P ositive A spects o f T a p h o n o m y ” w as held in 1984 as p a rt o f the
an n u a l m eeting o f the G eological Society o f A m erica (T h o m as 1986). B ehrens
m eyer a n d K idw ell (1985:105), tw o paleobiologists, p ro p o se d a “ new w orking
definition for the field as the stu d y o f processes o f preservation and how they
affect inform ation in the fo s s il record" because such a definition encom passes
n o t only the m o re tra d itio n a l foci o f ta p h o n o m ic research - loss a n d bias o f
in fo rm atio n - b u t w h at they call the “ positive c o n trib u tio n s ” m ade by
ta p h o n o m ic processes. T h eir m odel o f a ta p h o n o m ic h isto ry did not, for
in stance, utilize sym bols depicting in fo rm a tio n loss; ra th e r, it sim ply listed
catego ries o f ta p h o n o m ic processes in an im plied tem p o ra l o rd e r (F igure 2.7).
A stru ctu ra lly different dep ictio n o f a ta p h o n o m ic h isto ry was published th a t
sam e y ear by a p aleo n to lo g ist a n d a zo o a rch a eo lo g ist (A ndrew s a n d C o o k
1985). T h a t illu stra tio n (F ig u re 2.8) also did n o t sym bolically indicate in fo r
m atio n loss o r bias. Interestingly, n eith er o f these last tw o illu stra tio n s include
sym bolism fo r the positive c o n trib u tio n s o f a ta p h o n o m ic history. Positive
c o n trib u tio n s include the a d d itio n o f in fo rm a tio n to an im al rem ains, such as
traces o f the p re d a to r th a t exploited the an im al (e.g., gnaw ing m ark s) and the
processes th a t m oved (e.g., ab ra sio n a n d ro u n d in g due to fluvial actio n ), sorted
(e.g., selective rem o val o f p a rtic u la r skeletal elem ents d u e to fluvial action), an d
orien ted the b ones (e.g., fluvial ac tio n causing the long axis o f som e skeletal
Figure 2.5. M eadow ’s (1981) m odel o f the tap h o n o m ic h istory o f a faunal

assem blage. T he decreasing size o f the sym bol (from to p to b o tto m ) denotes the
loss o f in fo rm atio n th ro u g h tap h o n o m ic tim e (after M eadow 1981:Figure 1,
courtesy o f the autho r).
SAMPLE INFORMATION B I AS
PAST
PRESENT
Figure 2.6. Hesse an d W ap n ish ’s (1985) m odel o f a tap h o no m ic history o f a

zooarchaeological assem blage o f faunal rem ains. N o te how the n atu ra l factors
(N ) effectively rem ove the cu ltu ral in fo rm atio n (C) displayed by the assem blage.
C om p are to Figure 2.5. (after H esse an d W apnish 1985:19, Figure 9, courtesy o f
the au th o rs and T arax acu m Press).
SEQUENCE OF P R O C E S S E S SUB-DISCIPLINE
AFFE C TI NG P R E S E R V A T I O N OF TAPHONOMY
ORGANISM DIES OR B I OS PHER E

SH EDS BODY PARTS
NECROLOGY
— ORGANIC (SOFT) PARTS DECAY
SEDIMEN TARY PROCESSES

CO c
INTERACT WITH REMAINING PARTS
1 1 ” - ” 1
B I O S TR A T I NO M Y
‘ 3 8 I $ 8 '
— BURIAL
I I I
CHEMICAL ALTERATION DIAGENESIS I 1
AND LITHIFICATION
J ___L_
L I THOS PHE RE
— COLLECTION
I
F igure 2.7. B ehrensm eyer an d K idw ell’s (1985) m odel o f a tap h o n o m ic history
w ith relations o f subdisciplines o f tap h o n o m y indicated (after B ehrensm eyer and
K idw ell 1985:107, F ig u re 2; courtesy o f the au th o rs and T he Paleontological
Society). N o te the lack o f sym bolic loss o f inform ation.
elem ents to display p a tte rn e d o rien ta tio n s). T he positive aspects o f ta p h o n o

mic processes are ap p reciate d by p aleo n to lo g ists (e.g., A llison 1991; W ilson
1988), b u t w h a t a b o u t zoo arch aeo lo g ists?
A m o n g arch aeo lo g ists, B onnichsen (1989:2) th in k s B ehrensm eyer a n d K id-
w ell’s (1985) d efinitio n is a g o o d m o d ificatio n to E frem o v ’s (1940) original one,
b u t p ro p o se s th a t th eir new definition “ be explicitly extended to include the
goal o f learn in g a b o u t h u m a n ad a p tiv e system s” because he believes the new
definition is lim ited to “ the en h a n ced u n d e rsta n d in g o f p a st en v iro n m en ts th a t
one gains th ro u g h ta p h o n o m y .” T his is sim ilar to K o c h ’s (1989:2) suggestion
th a t “ arch aeo lo g ists a n d p ale o a n th ro p o lo g ists m ay find it m o re useful to
re sta te B ehrensm eyer a n d K id w ell’s definition as the stu d y o f the processes o f
p re serv atio n a n d m o d ificatio n , a n d how they affect geological, biological, an d
cu ltu ra l in fo rm a tio n in the geological re c o rd .”
T he significance o f an ac cu ra te definition o f ta p h o n o m y resides in its
specification o f the field o f inq u iry . W hat do ta p h o n o m ists study, a n d w hy do
they stu d y th a t m aterial? E frem ov (1940:85) originally defined ta p h o n o m y as
“ th e stu d y o f th e tra n s itio n (in all its d etails) o f an im al rem ains fro m the
b io sp h ere in to th e lith o sp h e re.” W hile ex p an sio n o f this definition has c o r
rectly com e to en com p ass the tra n s itio n o f an y org an ism fro m the bio sp h ere to
the lith o sp h ere, it is n o t a t all clear how the original definition differs
Figure 2.8. A ndrew s an d C o o k ’s (1985) m odel o f a tap h o n o m ic h istory show ing

stages o f m odification (after A ndrew s an d C ook 1985:689, Figure 7, courtesy o f
the au th o rs an d The R oyal A nthro p o lo g ical Institute).
significantly fro m B ehrensm eyer a n d K idw ell’s (1985) new definition o r

B o nn ich sen’s m o d ificatio n o f th a t new definition when the definitions alone are
considered w ith o u t th e ap p e n d ed e x p lan a tio n s o f th o se definitions. T h a t is
p a rtic u la rly so w hen one read s a little fu rth e r in E frem o v ’s sem inal p ap e r an d
finds th e sta te m e n t th a t “ the p assage fro m the b iosphere in to the lithosphere
occurs as a resu lt o f m an y interlaced geological a n d biological p h en o m en a.
T h a t is w hy, w hen this process is analyzed, th e geological p h e n o m e n a m u st be
stu d ied in th e sam e m easu re as the biological ones” (E frem ov 1940:85). T hus,
ta p h o n o m y is “ th e science o f the laws o f em b ed d in g [and unites paleontology]
b o th w ith geology a n d biology in to one general geo-biological historical
m eth o d o f stu d y ” (E frem ov 1940:93).
T he re aso n fo r th e suggestions th a t E frem o v ’s original definition be m odified
is th a t he, being a p aleo n to lo g ist, w as co ncerned w ith the paleobiologically
related biasing aspects o f ta p h o n o m ic processes. H e n o ted , fo r exam ple, th a t
fossil assem blages m ay b e a r little resem blance to th e com m u n ities o f living
o rg an ism s from w hich they w ere derived because fossil accu m u latio n s were
“ accid en tal selections” c o n d itio n ed in som e cases by the selective ac tio n o f
fluvial processes fo r b ones o f larger tax a a n d ag ain st the bones o f sm aller tax a.
A n d he w o n dered w hy fossils w ere som etim es fo u n d in dense c o n c e n tra tio n s
a n d a t o th e r tim es as iso lated a n d scattered single individuals o r bones. O nly by

c o n tro llin g for such bo n e a c cu m u latio n a n d dispersal facto rs o r u n d ersta n d in g
th eir causes, E frem o v reaso n ed , co u ld we p ro d u c e ac cu ra te re co n stru ctio n s o f
paleo co m m u n ities a n d e v o lu tio n ary histories. T h a t is, E frem ov clearly c o n
ceived o f his “ new b ra n c h o f p ale o n to lo g y ” as the stu d y o f the tra n sitio n o f
organ ism s from being alive to being the fossils (w hat a ta p h o n o m ist studies)
recovered by a p aleo n to lo g ist. Such study co u ld in fo rm the research er w hen a
fossil assem blage p ro v ides a clear reflection o f p re h isto ric biotic com m unities
a n d w hen a fossil assem blage p rovides a d isto rte d o r biased reflection o f a
p reh isto ric biotic co m m u n ity (w hy a researcher does ta p h o n o m ic research,
acco rd in g to E frem ov).
B ut bias is a p h e n o m e n o n th a t c a n n o t be isolated; th a t is, it m u st be given a
co n tex t in o rd e r th a t its n a tu re can be defined a n d th u s recognized. T o say
som eone o r so m eth in g is “ b iased ” tells us little a b o u t th a t p erso n o r object. T o
d eterm in e how an ind iv id u al is biased d em an d s th a t we kn o w the referential
context: “ biased w ith reg ard s to w h at?” A com plete, unm odified b o n e is n o t a
biased in d ic a to r o f the tax o n it represents, n o r is a com plete to o th . Sim ilarly, a
b ro k e n bo n e rep resen ted only by its distal en d is n o t a biased in d ic a to r o f the
skeletal elem ent rep resen ted n o r o f the ta x o n represented. H ow ever, it m ig h t be
im possible to d eterm in e the sex o f the ind iv id u al rep resen ted by th a t d istal h a lf
o f a skeletal elem ent, a n d in th a t sense the bo n e is “ b iased ” in term s o f the
sexual in fo rm a tio n th a t can be ex tracted fro m it; in this case the specim en is
biased ag ain st sexing due to th e fact th a t it displays n o n e o f the sexually
d iag n o stic m o rp h o m e tric featu res biologists have fo u n d to d en o te reliably the
sex o f in dividual an im als. In the sam e w ay, b u t fro m a purely tap h o n o m ic
perspective, the p u b is o f the in n o m in a te o f som e artio d a cty ls displays n o tab le
differences in th e ro b u stn ess a n d p o sitio n o f th e ilio-pectineal em inence. A
p re h isto ric a rtio d a cty l p u b is can be sexed by ex a m in a tio n o f th e ro b u stn ess an d
p o sitio n o f th a t em inence i f that p a rt o f the pubis is well preserved; if the
specim en has been w eath ered a n d bone m ateria l has exfoliated fro m th e bone
surface, o r if ro d e n t o r c a rn iv o re gnaw ing (o r any o f a m y riad o f o th er
processes) has rem o v ed the em inence o r o bscured its p o sitio n , th e n the
specim en is biased ag ain st sexing.
T hese exam ples m ak e it clear th a t bias in a ta p h o n o m ic sense is relative to the
q u estio n being asked o f the fossils (W ilson 1988). A p a rtic u la r m am m alian
fossil assem blage m ay accu rately reflect th e ta x a living in the area a t the tim e
the fossils were d ep o sited (be tax o n o m ically unbiased; this w as in p a rt the kind
o f bias E frem o v w as m o st co n cern ed a b o u t), b u t consist o f a d isp ro p o rtio n a te
n u m b er o f teeth (be biased in term s o f the relative a b u n d a n ces o f skeletal parts).
T h u s th e co n tex t o r referen t m u st be specified w hen it is said a fossil assem blage
is o r is n o t biased; n o t specifying the co n tex t results in the statem en t th a t a fossil
assem blage is o r is n o t biased hav in g an am b ig u o u s m ean in g a t best, a n d no
m eanin g at w orst.
T he n o tio n s th a t fossil assem blages are p o ten tially biased an d ta p h o n o m y is

co n cern ed solely w ith such biases has h a d a n influence o n how som e a rc h a e o
logists conceive ta p h o n o m y . G iven E frem o v ’s original definition, ta p h o n o m ic
processes are th o se energy-w ielding agents, w h eth er geological, clim atological,
o r biological, th a t affect an im al (an d p la n t) rem ains. H u m an s a n d their
ho m in id an cesto rs are an im als th a t can be considered biological ta p h o n o m ic
ag en ts actin g at least in the realm o f b io stra tin o m y a n d p erh ap s in the realm o f
necrology as well. T h u s H isco ck ’s (1990:44) suggestion th a t ta p h o n o m y be
defined as “ the stu d y o f th e tra n s fo rm a tio n o f objects, a n d th eir spatial
relatio n sh ips, afte r they leave a living system a n d before they are recovered by
the scien tist,” seem s to o n a rro w as it ap p e a rs to ignore necrology. H iscock
(1990:35) seem s to believe th a t w hile ta p h o n o m ic processes rem ove in fo r
m a tio n fro m th e fossil record, the in fo rm a tio n th o se processes a d d concerns
“ the [taphonom ic] processes them selves” an d n o t generally the paleoecology,
for instan ce, o f th e an im al rep resen ted . W ilson (1988), fo r one, w ould argue the
last is in acc u rate (see also B ehrensm eyer a n d K idw ell 1985). A nd Colley
(1990a:59), fo r a n o th e r, w ould c o u n te r H isco c k ’s suggestion by p o in tin g o u t
th a t “ the stud y o f w h a t people do w ith b o n es is often o f g reat in tere st in its ow n
rig h t” a n d such studies are the focus o f zo o arch aeo lo g ically o rien ted ta p h o
nom ic research.
B ased o n the preceding, it should be clear th a t statem en ts like, it is difficult to
d e m o n stra te th a t “ ta p h o n o m ic ra th e r th a n hom in id agencies are responsible
fo r [a bone] assem b lag e” (S peth 1991:37), o r th a t d isarticu latio n o f som e
an im al carcasses “ is p ro b a b ly a ttrib u ta b le to ta p h o n o m ic processes ra th e r
th a n b u tch erin g [by h u m an s]” (L a n d als 1990:139), o r sim ilar statem en ts
im plying th a t h u m a n a n d /o r h o m in id in tera ctio n s w ith an im al carcasses are
no t tap h o n o m ic, are incorrect. Such statem en ts are rare, an d I suspect they
reflect a discipline-centric view o f zo o a rch a eo lo g ists prim arily in terested in
h u m a n b eh av ior; th u s any n o n -h u m a n process is biasing a n d th erefo re
tap h o n o m ic. (R eed [1963:210-211] tu rn s this n o tio n a ro u n d a n d argues the
“ cu ltu ra l filter” biases th e fa u n al reco rd because h u m an ly accu m u lated fau n al
rem ain s will n o t accu rately reflect the p aleo fau n a.) T h e im plicit reaso n in g
seem s to be th a t because h u m a n activities are the subject o f interest, they
c a n n o t be biasing a n d th u s are n o t ta p h o n o m ic because ta p h o n o m y concerns
how th e fossil reco rd is biased. T h e n o tio n th a t ta p h o n o m y concerns only the
biases to a fossil reco rd com es fro m ta p h o n o m y ’s co n c ep tu al ro o ts in p a le o n to
logy w here it w as recognized th a t fossil assem blages seldom rep resen t co m p le
tely u n d isto rte d p ictu res o f p a st biological com m unities. In archaeology, the
co n tex t o f co n cern is generally h u m a n beh av io r, a n d w hile archaeological
assem blages o f an im al rem ains m ay well be biased w ith regards to th a t
beh av io r, this does n o t m ean h u m an s are n o t ta p h o n o m ic agents. T his, in the
final analysis, is th e re aso n in g u n d erp in n in g B o nnichsen’s an d K o c h ’s m odifi
catio n s o f B ehrensm eyer a n d K idw ell’s definition o f tap h o n o m y .
On the structure o f taphonomy: a personal view
E ach [taphonom ic] event has unique characteristics th a t require a p articu lar
ap p ro ach. [D ue to historical variability] useful analyses m u st select carefully from
am ong variables an d apply only those th a t are p ertin en t to the u n ique situation.
(E. C. O lson 1980:9-10)
Taphonom ic histories
T h e p aleo n to lo g ical fossil re co rd h as been fo rm ed to ta lly by n a tu ra l processes

in clud in g geological a n d biological processes. T hese n a tu ra l processes act u p o n
the av ailab le o rg anism s (w hich are in tu rn c o n d itio n ed by such n a tu ra l facto rs
as to p o g ra p h y , su b strate, veg etatio n , a n d clim ate) a n d affect the a d d itio n to,
m ain ten an c e in, a n d s u b tra c tio n fro m the p aleo n to lo g ical fossil reco rd o f
org an ism s a n d th eir rem ains.
T h e arch aeo lo g ical fossil reco rd is p o ten tially fo rm ed n o t only by th e sam e
n a tu ra l processes as th e p aleo n to lo g ical fossil reco rd , b u t also by h u m a n
processes. A n arch aeo lo g ical site consists o f c u ltu ra l a n d n a tu ra l objects th a t
are a d d e d , spatially a rran g e d , a n d preserved a n d /o r d estro y ed by various
h u m an a n d n a tu ra l processes. H u m a n b eh av io rs th a t result in the fo rm a tio n o f
a fossil reco rd h ave been labeled the cultural filte r (R eed 1963; D a ly 1969).
H u m an processes th a t affect p o te n tia l a d d itio n s to the fossil reco rd include
selective h u n tin g (S m ith 1979; W ilkinson 1976) an d bu tch ery p ractices (B in
fo rd 1978; N o e -N y g a a rd 1977, 1987). V a riatio n in b o th the archaeological and
the p aleo n to lo g ical fossil records is created by v arying the ad d itio n s, the m eans
o f a d d itio n , th e m ech anism s th a t d istrib u te fa u n al rem ains, a n d the m eans o f
m ain ten an c e a n d s u b tra c tio n o f an im al rem ains fro m th e fossil record.
T h e d istin c tio n b etw een p aleo n to lo g ical fa u n as a n d a rc h a e o fa u n a s is based
on c h a ra c te riz a tio n s o f th eir respective lack o f associated artifa cts o r presence
o f asso ciated artifa cts (C h a p te r 1). It m u st be em phasized, how ever, th a t these
ch a rac te rizatio n s are sim plistic. Som e fossil reco rd s m ay have no spatially
asso ciated cu ltu ra l m ateria ls even th o u g h h u m an s h a d an active role in th eir
fo rm a tio n . In th e absence o f associated artifacts, a ttrib u te s o f bo n e m odifica
tio n a ttrib u ta b le to h u m a n activities are cited as evidence o f h u m a n in terv e n
tio n . O ne exam ple o f this involves m a sto d o n (M a m m u t am ericanum ) b ones in
N o rth A m erica (F ish er 1984a. 1984b; G ilb o w 1981; G u sta fso n et al. 1979). In
these cases, m o d ificatio ns to b ones w hich are inexplicable given n a tu ra l
processes are cited as evidence o f h u m a n ta p h o n o m ic agents even th o u g h no
(or a few possible) artifa cts are asso ciated w ith the fa u n a l rem ains (see H aynes
a n d S ta n fo rd [1984] fo r sim ilar a rg u m e n ts re g ard in g N o rth A m erican late
P leistocene C am elops sp.). H ow ever, it is n o t clear as yet w hich o f these cases
actu ally rep resen t h u m an ly m odified carcasses (e.g., G ra h a m et al. 1983; see
H ay n es 1991 fo r a review o f p ro b o sc id e an tap h o n o m y ). T his in tu rn has
resulted in d etailed co m p arativ e analyses o f m a sto d o n rem ains clearly asso

ciated w ith artifa cts a n d sim ilarly aged m a sto d o n rem ains w ith no associated
artifacts in atte m p ts to derive bo n e m o d ificatio n crite ria th a t u n am b ig u o u sly
signify a h o m in id ta p h o n o m ic a g en t (e.g., F ish er 1987; G ra h a m a n d K ay 1988).
T he lite ratu re on a ttrib u te s o f bo n e m o d ificatio n is ex p an d in g rapidly
(B onnichsen a n d Sorg 1989; H u d so n 1993; a n d references therein) yet d eb ate
ab o u n d s over th e precise m ean in g o f m an y a ttrib u te s (e.g., Jo h n so n 1982, 1985
versus L ym an 1984b; a n d S hip m an 1981a; S hip m an an d R ose 1983a, 1983b;
O lsen a n d S hip m an 1988 versus E ick h o ff a n d H e rrm a n n 1985; B ehrensm eyer
et al. 1986, 1989; F io rillo 1989; O liver 1989). M any o f the a ttrib u te s o f bone
m o d ification used to distinguish cu ltu ra lly d ep o sited fro m n atu ra lly deposited
fau n al rem ain s are, w hen used alone, equivocal. M u ltip le kinds o f d a ta are
necessarily used reg ularly to help to d istin g u ish analytically the tw o types o f
fossil records.
T a p h o n o m ic h istories are re co n stru cted from the ab u n d a n ce, d istrib u tio n ,
a n d m o d ificatio n a ttrib u te s o f fossils. T he objects in a site, th eir frequencies,
physical attrib u te s, sp atial loci a n d asso ciatio n s, a n d geological a n d cu ltu ra l
asso ciatio n s are all th a t are o b serv ab le in the fossil record. A scientific
a p p ro a c h to ta p h o n o m y m u st realize w h a t the em pirical p h en o m en a o f the
fossil re co rd are, a n d p ro d u c e a m odel th a t p erm its ex p ectatio n s to be p h ra sed
co n cern in g fossil assem blage c o n te n t a n d d istrib u tio n ; i.e., the archaeologi-
cally visible fossil record. Such a m odel w ould ideally be universally applicable
an d yet specific en o u g h to g ra n t insights to p a rtic u la r ta p h o n o m ic p ath w ay s. A
first step to m odel b uild ing involves u n d e rsta n d in g the basic stru c tu re o f
ta p h o n o m ic processes an d effects. A s defined in C h a p te r 1, a taphonom ic
process is the d y nam ic actio n o f som e source o f force o r the physical cause o f
m od ificatio n to an im al carcasses a n d skeletal tissues. A taphonom ic effect is the
static result o r trac e o f a ta p h o n o m ic process th a t acted on a n d m odified
carcasses o r skeletal p a rts an d tissues. 1 begin by co n sid erin g som e general
characteristics o f ta p h o n o m ic processes, a n d then tu rn to som e general
categories o f ta p h o n o m ic effects.
Processes th a t fo rm the fossil reco rd v ary alo n g three dim ensions. T he
o b j e c t dim en sion consists o f three categories: a d d itio n , su b tra c tio n , an d
m ain ten an ce o f objects. A n im al rem ains m ay be a d d e d to, rem oved from , o r

m ain tain ed w ith in a p a rtic u la r sp atially defined assem blage o f rem ains. T he
s p a t i a l dim en sio n consists o f m ovem ent, a n d no n m o v em en t. F a u n a l rem ains
m ay be m oved w ith in (an d p e rh a p s rem oved from ) o r n o t m oved w ithin a

p a rtic u la r sp atial u n it (the b o u n d arie s o f w hich define the assem blage). T he
m o d i f i c a t i o n d im en sio n consists o f tw o possibilities: a b o n e m ay be m odified
from its n a tu ra l, living state (e.g., be b ro k e n , b u rn e d , m ineralized), o r it m ay

n o t be so m odified (o th er th a n p erh ap s drying). A com plete skeletal elem ent
th a t retain s its a n a to m ica l in teg rity a n d features, if n o t fossilized, ca n be
co n sid ered to be n o t m odified. T a p h o n o m ic processes create variability
betw een a n d w ithin p a rtic u la r fossil records. B ecause the m a jo r expressions o f

ta p h o n o m ic effects in archaeological co n tex ts are assum ed to be largely
a ttrib u ta b le to h u m an processes, ra n d o m v aria tio n in fossil assem blage
c o n ten t, d istrib u tio n , a n d m o d ificatio n a ttrib u te s is n o t a n ticip ated , an an tici
p a tio n b o rn e o u t by studies o f archaeological fa u n as (e.g., L ym an 1978;
M ead o w 1978; N o e -N y g a a rd 1977; P ozorski 1979; R ap so n 1990; Stiner
1990a). H ow ever, n a tu ra l processes also create n o n -ra n d o m p a tte rn s in the
arch aeo lo gical reco rd (B inford 1981a, 1981b; B rain 1969; H aynes 1980a,
1980b; Hill an d B ehrensm eyer 1984, 1985). T h e first step in analysis th en , is to
identify p a tte rn s asso ciated w ith n a tu ra l processes. T his m ay be accom plished
by co m p arin g the fossil reco rd u n d er stu d y to m odels o f the variable(s) u n d er
scrutiny, such as the relative frequencies o f skeletal elem ents in com plete
skeletons. S u b seq u en t to this co m p ariso n , the analyst can begin to assess the
m ean in g o f an y d etected p a tte rn s by co m p arin g the fossil reco rd to p a tte rn s
displayed by n eo ta p h o n o m ic records created by v ario u s kn o w n tap h o n o m ic
processes. B rain's (1967b. 1969) co m p ariso n o f b o n e p a rt frequencies in
e th n o arch ae o lo g ic al co n tex ts w ith frequencies o f b o n e p a rts fro m S outh
A frican caves is a classic exam ple o f this co m p arativ e ap p ro a c h .
T a p h o n o m ic histo ries are in itiate d w hen an an im al dies (the m o d e o f
m o rta lity can be a significant ta p h o n o m ic variable; see C h a p te r 5). S oft tissues
m ay th en be rem oved, b ones m ay becom e d isarticu lated , scattered , buried,
fossilized, m ay ero de o r chem ically d e te rio ra te aw ay, a n d m ay eventually be
recovered. O f course, v ario u s ta p h o n o m ic processes m ay o r m ay n o t act
sim u ltan eo u sly on a bo n e o r carcass, a n d m ay o r m ay n o t affect p a rtic u la r
carcasses o r bones. T h e set o f p o te n tia l effects o f ta p h o n o m ic processes m ay be
arran g e d in to fo u r categories: d isarticu latio n , d ispersal, fossilization, and
m echanical m od ificatio n . All o f these are m o n ito red from the startin g p o in t o f
a co m plete, unm odified skeleton. T h a t is, we kn o w w h a t a living skeleton looks
like, an d w h a t ta p h o n o m ists are in terested in is how th a t skeleton, o r w h a t is
left o f it, ap p e a rs w hen it is recovered from a geological context.
D isarticulation refers to th e a n a to m ica l disasso ciatio n o f skeletal elem ents; it
o ften is the first step in the loss o f a n a to m ica l in teg rity o f a skeleton.
D isa rtic u la tio n is related to so ft tissue th a t fu n ctio n s to ho ld jo in ts to g eth er
(D o d so n 1973; H ill 1980; S chafer 1962/1972; T o o ts 1965c). C hem ical o r
m echanical b re ak d o w n an d rem oval o f soft tissue u ltim ately results in d isa rti
c u latio n (e.g., C oe 1978; M icozzi 1986; P ayne 1965). B ecause soft tissue
a n a to m y varies fro m jo in t to jo in t, the process o f d isarticu latio n is extrem ely
com plex u n d e r n a tu ra l co n d itio n s, b u t is n o t so com plex as to preclude
c o n stru c tio n o f m odels o f n a tu ra l d isa rtic u la tio n (H ill 1979b, 1980).
D ispersal o f skeletal p a rts m ean s the increase o r the decrease o f d istance
betw een bones. D ispersal o f skeletal elem ents m ay precede, o r be sim u ltan eo u s
w ith o r su b seq u en t to d isarticu latio n , a n d is related to d isa rtic u la tio n because
it co n cern s the sp atial lo catio n o f fossils. W hile d isarticu latio n requires only a
few centim eters o f sp atial d isasso ciatio n o f p a rts to d estro y an a to m ica l

integrity, d ispersal entails centim eters to kilom eters (H ill 1979b). M odels o f
d ispersal have been co n stru c te d for fluvial tra n s p o rt (e.g., B ehrensm eyer
1975b), h u m an tra n s p o rt (e.g., B inford 1978), ra p to r tra n s p o rt (P lug 1978),
tra n s p o rt by p o rc u p in e s (B rain 1980), ca rn iv o re tra n s p o rt (B inford 1981b),
an d ra n d o m processes (H ill 1979b). M o st analyses o f fossil assem blages th a t
co n sid er ta p h o n o m ic issues assess w h eth er o r n o t th e fossil assem blage has
been tra n s p o rte d to its recovery lo catio n fro m a n e a r o r a fa r source location.
Shotw ell (1955) developed an an aly tic tech n iq u e to assess w h eth er o r n o t a
fossil assem blage h a d been tra n s p o rte d , a n d to assess w hich ta x a in the
assem blage were locally derived a n d w hich w ere p ro b a b ly intrusive o r n o n
local. W hile S hotw ell’s (1955) tech n iq u e w as la te r ad a p te d to distinguishing
n atu ra lly fro m cu ltu ra lly d ep o sited tax a in a rc h a e o fa u n a s (T h o m as 1971), it
has since been show n to c o n ta in serious flaws (G ray so n 1978a, 1978b; W olff
1973). A n aly sts still ad d ress this issue, using techniques such as assessing the
degree o f ab ra sio n o f b o n es (B ehrensm eyer 1975b) to d eterm in e w h eth er the
assem blage o r p o rtio n s th e re o f h av e been fluvially tra n sp o rte d to the collection
locality, a n d ex p erim ental d a ta to infer the agent o f tra n s p o rt (B inford 1981b;
L ym an 1985a).
Fossilization d en o tes the a lte ra tio n o f b o n e chem istry (C o o k et al. 1961;
R olfe a n d B rett 1969). T he type o f sed im en tary m atrix in w hich th e bo n e is
d ep o sited largely determ ines the p a rtic u la r types o f fossilization processes.
S eco n d ary d e te rm in a n ts include e n v iro n m e n ta l c o n d itio n s such as soil m o is
ture regim es as d eterm in ed by p re c ip ita tio n a n d te m p eratu re. Som e fossiliza
tio n processes, especially w eath erin g (B ehrensm eyer 1978), m ay result in
fra g m e n ta tio n o f bones.
M echanical alteration den o tes the s tru c tu ra l a n d /o r m o rp h o lo g ical a lte
ra tio n o f the o rig in al living b o n e by m echanical o r physical processes.
E xam ples in clude fra g m e n ta tio n a n d ab rasio n . F o r instance, each bo n e in an
an im al is a com plete, discrete object. T h e cause o f the an im a l’s d e a th a n d /o r
p o stm o rte m fa cto rs m ay resu lt in b ro k e n bo n e (L y m an 1984b, 1989b).
F ra g m e n ta tio n , th en , is th e d e stru c tio n o f th e original an a to m ica l discreteness
o f a b o n e by g en eratin g m u ltip le discrete objects fro m the original by
m ech an ical o r physical m eans, in this case by the lo ad in g o f force on the bone.
A b ra sio n is the m odification o f original bo n e m o rp h o lo g y by the a p p lica tio n o f
frictio n al forces to bo n e surfaces o r edges (O lsen a n d S hip m an 1988; S hipm an
an d R ose 1988).
Discussion
T a p h o n o m y is co n cern ed w ith differences a n d sim ilarities betw een fossils an d
o rg an ism s, a n d betw een a fossil reco rd a n d th e p re h isto ric fa u n a from w hich it
derived. C o n cern in g the latter, obvious differences include the presence o f
living o rganism s. E cological a n d ethological studies o f extinct tax a are, o f

course, im possible. E ven th o se ta x a w ith m o d ern , living c o u n te rp a rts are n o t so
easily d ealt w ith w hen rep resen ted by fossils because stu d y in g living tax a
p resents certain difficulties (C oe 1980). D ifferences a n d sim ilarities betw een
living a n d fossil fa u n as p re sen t ta p h o n o m ic challenges to paleoecological
research because the ecological principles used to stu d y living fa u n as are
co m m o n ly used in p aleoecological research (V an C o u v erin g 1980; W estern
1980). C o n seq u en tly , the fossil record m u st be analytically re co n stitu ted into a
fossil fa u n a o r the original biotic co m m u n ity to answ er m an y research
qu estio n s. A s n o ted in C h a p te r 1, subsistence studies using a rc h a e o fa u n a s face
sim ilar an aly tic challenges (K ing a n d G ra h a m 1981).
C lues fo r d eveloping analy tical tech n iq u es for m eeting these an aly tic ch a l
lenges can be fo u n d by considering the fo u r categories o f ta p h o n o m ic effects
(d isartic u la tio n , d ispersal, fossilization, m echanical alte ra tio n ) in light o f the
v ariab ility in ta p h o n o m ic processes. F o r exam ple, new o r ad d itio n a l fossils
c a n n o t be ad d ed to an assem blage th a t is in situ w ith o u t m o v em en t o f the
“ n ew ” fossils. T he only conceivable w ay this m ay h ap p e n is if the assem blage
m oves to a new lo catio n a n d is d ep o sited a ro u n d the “ new ” fossils w ith o u t the
la tte r’s m o vem en t, as in som e fluvial settings (B oaz 1982). A co m m o n an d
readily conceivable w ay fo r new fossils to be ad d ed to a n assem blage w ith o u t
m o v em en t o f the fo rm er is fo r the sam pling universe to be enlarged such th a t
ad d itio n a l fossils are collected fro m new areas o f a site.
T h e second im p o rta n t p o in t deriving fro m co m p ariso n o f processes an d
effects is th a t different processes can have sim ilar effects; equifinality is a very
real pro blem . D isarticu la tio n , d ispersal, a n d m echanical alte ra tio n all involve
m o v em ent o f the fossils w hile fossilization does n o t req u ire m ovem ent. F inally,
all effects an d process categories co n cern frequencies (ad d , m a in tain , su b trac t)
a n d d istrib u tio n s (m ove, n o n -m o v e) o f fossil categories. It is th ere fo re p e rti
n en t to discuss possible techniques fo r m easu rin g each ta p h o n o m ic effect in the
fossil record . I do this briefly here, a n d in m o re detail in later ch ap ters.
Hill (1979b:744) co ncludes th a t " th e d eterm in in g c o n tro ls o f the [d isarticu la
tion] p a tte rn are in h ere n t in the an a to m y o f the d ead anim al itself a n d th u s
in d ep en d e n t o f the ag ents w hereby it is realized.” D isa rtic u la tio n m ight be
m odeled by a ran k o rd erin g o f the cro ss-sectional area o f soft tissue s u rro u n d
ing jo in ts. T he basic an aly tic a ssu m p tio n m ig h t be p h ra sed as: the g re ater the
cro ss-sectional area o f soft tissue associated w ith a jo in t the longer th e jo in t will
rem ain in tact su b seq u en t to the a n im a l’s d eath . T his assu m p tio n o f course
presum es th a t soft tissues associated w ith each jo in t are q u alitativ ely identical,
w hich is unlikely (H ild eb ran d 1974; R o m er an d P arso n s 1977). T h u s the
am o u n t a n d type o f connective tissue seem to be variables w ith significant
influence on the d isarticu latio n process. S tudy o f d isa rtic u la tio n in the fossil
reco rd requires d etailed d a ta on the lo catio n s a n d sp atial asso ciatio n s o f fau n al
rem ain s (C h a p te r 5).
D isp ersal is a com plex process m inim ally co n tro lled by d isarticu latio n , type
a n d stren g th o f d ispersal m echanism , su b strate, to p o g ra p h y , an d b o n e density,

size, a n d m o rp h o lo g y . Hill (1979b:269-270) hypothesizes th a t scatterin g is
caused by processes th a t act ra n d o m ly . D e p a rtu re s from the ra n d o m p a tte rn
suggest n o n -ra n d o m ly actin g processes w hose id en tity m u st be determ ined.
H ill's (1979b) h yp o th esis co u ld be used as the first null hypothesis to be tested
w ith fossil d ata. T h en , intrinsic p ro p e rtie s o f bones can be used to generate
ex p ectatio n s re g ard in g d istrib u tio n a l p a tte rn s o f fossils (e.g., F ro stic k an d
R eid 1983; K o rth 1979). C learly, d a ta on b o n e lo catio n , o rie n ta tio n , a n d angle
o f d ip sh ould be reco rd ed d u rin g field recovery, as well as sedim entological
d a ta in d icatin g m o d e o f d ep o sitio n an d tu rb a tio n processes (C h ap ters 5-9).
F o ssilizatio n m echan ism s are m inim ally d e p e n d a n t o n clim ate, d ep o sitio n al
m atrix , a n d bon e po ro sity . T here ap p a re n tly is no detailed m odel o f fossiliza
tion c o m p arab le to H ill’s (1979b) m odels o f d isarticu latio n a n d dispersal.
D o cu m en ted processes o f fossilization (e.g., R olfe a n d B rett 1969; S chopf
1975; W h itm er et al. 1989) in d icate, how ever, th a t in o rd e r to study fossiliza
tio n, d a ta re q u ired include m atrix chem istry a n d m ineralogy, chem istry o f the
fossils a n d original chem istry o f the bones, clim atic (p ast an d present)
in fo rm a tio n such as te m p e ra tu re , p re cip ita tio n , a n d g ro u n d w a te r regim es, an d
a k n ow ledge o f th e geologic a n d pedogenic processes th a t c o n trib u te d to the
fo rm a tio n o f p a rtic u la r s tra ta (C h a p te r 11).
M ech an ical a lte ra tio n seems to be largely co n tro lled by bo n e stru c tu re an d
m o rp h o lo g y , at b o th m icroscopic a n d m acroscopic levels, an d bone p o ro sity
a n d density. In o rd e r to m easu re m echanical a lte ra tio n in the fossil reco rd , the
m in im al req u isite d a ta are frequencies o f frag m en t types (e.g., B unn 1989;
T o d d a n d R ap so n 1988) an d w h eth er o r n o t fragm ents o f a bone are associated,
in situ, p o lished, a b ra d e d , o r display o th e r featu res (C h ap ters 8-10). F o r
exam ple, K lein a n d C ru z-U rib e (1984) suggest sedim ent o v erb u rd en m ay crush
m o re deeply b u ried bones; all else being equal, the an a ly st co u ld m easure
frag m en t sizes to d eterm in e if frag m en ts d ecreased in size w ith increasing d ep th
(e.g., L ym an a n d O ’Brien 1987).
D esp ite the pleas o f several a u th o rs over tw o decades ago (H ill 1978; Hill and
W alk er 1972; M u n th e a n d M cL eo d 1975) the kinds o f d a ta m en tio n ed above
(T able 2.1) w ere seldom pu b lish ed by an aly sts p rio r to the m id-1980s. B ecause
a b u n d a n ces o f fossil categories are im p o rta n t to m any tra d itio n a l analyses an d
in te rp re ta tio n s, b o n e frequency d a ta are n early alw ays published, a n d m ore is
k n o w n a b o u t the ta p h o n o m y o f frequencies o f bone p a rts th a n virtually any
o th er v ariab le o f the fossil record.
Summary and conclusion

T aph o nom y has com e o f age!
(P. A. A llison 1991:345)
T a p h o n o m ic histories are usually com plex. T a p h o n o m ic processes vary along

three dim ensio ns (objects [added, su b trac te d , m ain tain ed ], sp atial [m oved, n o t
m oved], m o d ificatio n ) w hile ta p h o n o m ic effects m ay be a rra n g e d in fo u r

general catego ries (d isartic u la tio n , d ispersal, fossilization, m echanical alte r
atio n ). C o n sid e ra tio n o f the process a n d effect categories reveals th e types o f
d a ta re q u ired fo r ta p h o n o m ic analyses. I have d istinguished tw o basic goals o f
z o o a rch aeo lo g ical fa u n al analysis (d e te rm in atio n o f h u m a n subsistence p a t
terns a n d p re h isto ric ecological co n d itio n s) a n d tw o types o f fossil records
(a rch a eo fau n a s a n d p aleo n to lo g ical faunas). All o f these issues are fo u n d in
one form o r a n o th e r in m o st ta p h o n o m ic studies.
It is im p o rta n t to keep in m ind th ro u g h o u t the re m a in d er o f this volum e th a t
ta p h o n o m ic processes are historical a n d cum ulative. T h a t is, they have a
d irectio n th ro u g h tim e in the explicit sense th a t effects o f ta p h o n o m ic processes
w hich o cc u r late in th e h isto ry m ay d ep en d o n the effects o f tap h o n o m ic
processes w hich o ccu rred early in the ta p h o n o m ic history. A ta p h o n o m ic
h isto ry results in a fossil assem blage w hich m ay p o o rly reflect the q u a n tita tiv e
p ro p e rtie s o f the biotic co m m u n ity from w hich the fossils derived. T ap h o n o m ic
processes som etim es m im ic a n d o th e r tim es ob fu scate th eir respective effects,
th ereb y re n d erin g the w ritings o f ta p h o n o m ic histories difficult. T he h isto ry o f
ta p h o n o m ic research, especially in arch aeo lo g y , illu strates w hy ta p h o n o m ic
research a t the end o f th e tw en tieth ce n tu ry ap p e ars th e w ay it does. T he
com plexities o f ta p h o n o m ic processes can be d escribed by a sm all set o f general
processes a n d th eir respective general effects. T his sim ple fram ew o rk guides us
to w ard s re co g n itio n o f d a ta requisite to ta p h o n o m ic analyses. M y p u rpose
here has been to p ro v id e a general fram ew o rk th a t allow s us now to tu rn to a
general co n sid eratio n o f the p ractice a n d th eo ry o f ta p h o n o m ic research.
3
T A P H O N O M Y IN P R A C T I C E AND
THEORY
H istory suggests th a t the ro ad to a firm research consensus is extraordinarily

ard u o u s . . . In the absence o f a paradigm o r som e can d id a te for paradigm , all o f the
facts th a t could possibly p ertain to the developm ent o f a given science are likely to
seem equally relevant. A s a result, early fact-gathering is a far m ore nearly ran d o m
activity th an the one th a t subsequent scientific developm ent m akes fam iliar.
(T. S. K u h n 1970:15)
Introduction
T he fo u n d a tio n s fo r ta p h o n o m ic research w ere laid in the n in etee n th a n d early
tw en tieth cen turies w ith a focus o n o b serv atio n s o f m o d e rn processes th a t
resu lted in d ep o sits c o n ta in in g b ones w ith certain m odifications (B ehrens
m eyer a n d K idw ell 1985). E arly ta p h o n o m ists follow ed the u n ifo rm ita ria n ist
a p p ro a c h used by geologists o f the n in etee n th a n d tw en tieth centuries. T h a t
a p p ro a c h a n d its p resen t stru c tu re in the service o f zo o a rch a eo lo g ical ta p h o
n o m y is review ed in th e second p a rt o f this c h a p te r. P rio r to th a t I review
several exam ples o f w h a t I co n sid er to be g o o d ta p h o n o m ic analyses. T hese
illu stra te w h a t m ak es fo r stro n g co nclusions an d lead to a co n sid eratio n o f
u n ifo rm itarian ism a n d actu alism as m eth o d o lo g ies fo r studying the p ast. T his
in tu rn leads to a c o n sid e ra tio n o f e th n o arch ae o lo g y an d m iddle-range
research. F in ally , because actu alism a n d m iddle-range research u ltim ately lead
to an alo g ical arg u m e n ts, th e stru c tu re o f such arg u m e n ts is described.
Examples o f taphonomic analysis
T he crite ria I used to select th e exam ples review ed w ere sim ple. T h e analysis
m u st be p u b lished in a generally available form so th a t the original can be
co n su lted by in terested readers. T he analysis m u st have explicit hypotheses
th a t w ere being tested, a n d explicit a ssu m p tio n s a n d m eth o d s. As well, the d a ta
m u st be av ailab le (generally in the published articles) fo r e v a lu a tio n a n d
a d d itio n al analysis. In keeping w ith the d istin ctio n o f p aleo n to lo g ical an d
arch aeolo g ical tra d itio n s o f ta p h o n o m ic research, I begin w ith tw o exam ples o f
ta p h o n o m ic analysis in p aleo n to lo g ical co n tex ts, a n d th e n tu rn to exam ples
from arch aeo lo g ical contexts.
41
The extin ctio n o f Irish elk

E x tin ctio n o f th e larg e-an tlered Irish elk (M egaloceros giganteus) n ear the end
o f the P leistocene h as o ften been explained as resu ltin g from th eir having
becom e m ired in bo g m u d a n d /o r being d ro w n ed in m arshes o r bogs in p a rt
because th eir an tlers w ere large, heavy, a n d cum bersom e. B arn o sk y ’s (1985,
1986) analysis o f rem ains o f this large cervine is instructive. H e excavated a p a rt
o f B allybetagh Bog n ea r D u b lin , Ireland, a n d stu d ied specim ens previously
collected from it. H is ex cav atio n s covered 21 m 2 a n d ran g ed fro m 1.5 to 2.5 m
deep. H e exam ined 35 skulls, all m ales, from this bog, a n d c o m p ared them to
v ario u s sam ples o f Irish elk skulls fro m o th e r localities. Som e o f the c o m p a ra
tive specim ens w ere o n display in m useum s (large a n d com plete specim ens) an d
o th ers w ere n o t on display (sm all a n d /o r broken).
B arn o sk y (1985) lists six test im p licatio n s o f th e m irin g -d ro w n in g due to
large antlers h ypothesis. O nly one is m et w ith available d ata: all individuals in
the bog he sam pled were m ales (fem ales d o n o t carry antlers). But the o th er five
im p licatio n s are n o t m et. A n tlers are sm aller th a n average in B arn o sk y ’s
sam ple; sk eletons are n o t a rtic u la te d n o r com plete; the b ones are em bedded in
clay d ep o sits to o th in for the an im als to have becom e m ired; the deposits are
n o t d istu rb ed by tra m p lin g o r struggling o f these an im als as they should have
been by m ired anim als; a n d the bog w aters w ere a p p a re n tly shallow enough (as
in ferred from geologic d a ta ) to preclude d ro w n in g o f u p rig h t anim als. N o te
th a t the six test im plications referred to require detailed m o rp h o m etric,
co n tex tu al, asso ciatio n al, a n d stra tig ra p h ic d ata.
F ailin g to confirm the m irin g -d ro w n in g hypothesis, B arnosky (1985, 1986)
p ro p o se s a n d tests tw o altern ativ e hypotheses. P leistocene overkill by h u m a n
h u n ters (M a rtin a n d K lein 1984) is quickly d iscard ed because no artifa cts have
been fou nd asso ciated w ith rem ains o f Irish elk, a n d " th e few exam ples o f
m od ificatio n to Irish elk bones re p u ted to have been inflicted by h u m an s
c a n n o t be distin guished from o th er n atu ra lly created kinds o f break s, a b r a
sions, g n aw -m ark s, o r scratch es” (B arn o sk y 1986:132). F u rth e r, the oldest
arch aeo log ical evidence o f h u m a n s in Irelan d d ates betw een 9000 a n d 8500 BP,
w hile the Irish elk was extinct there ca. 10,600 BP. T ig h t chro n o lo g ical an d
stra tig ra p h ic c o n tro l o f the Irish elk rem ains described by B arnosky m akes the
arg u m e n t fo r absence o f a h u m a n ta p h o n o m ic agent convincing, a n d th u s the
overkill hy poth esis c a n n o t be sustained.
T h e second altern ativ e hyp o th esis is th a t “ m ale Irish elk visited bogs m o re
o ften th a n fem ales did d u rin g w inters, w hen unfit anim als died a n d d eco m
posed n e a r the w a te r’s edge, in som e cases o n the ice, a n d w ere scavenged an d
tra m p le d ” (B arn o sk y 1985:340). E vidence bearin g on the six test im plications
fo r th e m irin g -d ro w n in g h y pothesis are c o n sisten t w ith the w interkill h y p o th e
sis, as are fo u r o th e r test im plications specific to the la tte r hypothesis. All elk
died w ith an tlers a tta c h e d , suggesting an a u tu m n -w in te r d e a th season based on
Taphonom y in practice and theory 43
an alo g y w ith m o d ern cervines. M o rta lity w as d em o g rap h ically a ttritio n a l, as it

sh o uld have been if th e Irish elk m o st susceptible to m a ln u tritio n -re la te d d ea th
(the y o u n g an d the very old) w ere dying. B arn o sk y ’s (1985:341) Irish elk tend to
be sm all in b o d y a n d a n tler size, suggesting “ som e c o m b in a tio n o f lim ited
resources, m a ln u tritio n , o r disease d u rin g fetal o r p o s tn a ta l g ro w th .” F inally,
in m o d e rn cervines, “ m ale m o rta lity is g re ater th a n fem ale m o rta lity d u rin g
w inter, a p p a re n tly because m ales, unlike fem ales, eat little d u rin g th e fall ru t
a n d en ter th e w in ter in p o o r c o n d itio n ,” a n d m ales m o re o ften seek w inter
shelter in valley b o tto m s (n e ar bogs) a n d th u s m an y die n e a r lake (bog) shores
“ because they need w ater a n d because they are easy prey on ice” (B arn o sk y
1985:343). T his explains the o v erab u n d an c e o f m ales relative to fem ales in
collections o f Irish elk fossils. T he test im plications fo r the w interkill h y p o th e
sis u n d ersco re the necessity o f ag e-sex d em o g ra p h ic d a ta , m o rp h o m e tric d a ta ,
an d th e use o f m o d e rn analogs as co m p arativ e bases in ta p h o n o m ic analyses.
B arn o sk y n o t only p ro vides m an y (b u t n o t all; e.g., th e “ scratch es” o n Irish
elk b o n es are n o t described) relev an t d a ta , he considers th ree sep arate
hypotheses. T he w interkill h y p o th esis succeeds because its test im p licatio n s are
m et w hereas test im p licatio n s o f o th er hypotheses are not. F u rth e r, B arnosky
(1985,1986) co n sid ered a d o zen test im p licatio n s, som e w ith n egative evidence
(p aleo n to lo g ically /arch aeo lo g ically invisible test im plications) a n d som e w ith
positive evidence (p aleo n to lo g ically /arch aeo lo g ically visible test im plications)
to confirm his hypothesis.
Crocodilian scatology
F ish e r (1981:263) atte m p te d to establish “ u n a m b ig u o u s sig n a tu re ” crite ria o f

cro co d ilian c o n su m p tio n a n d d igestion o f m icro v erte b rates via the co n tro lled ,
lo n g -term feeding o f eight in d iv id u al crocodiles a n d the collection a n d stu d y o f
all scats a n d re g u rg ita ted m aterials. H e suggests e x tra p o la tio n o f his actualistic
d a ta “ to fossil cro co d ilian s seem s ju stifiab le as a w o rk in g h y p o th esis” (F isher
1981:270). H e justifies th e e x tra p o la tio n by review ing the digestion o f bones by
o th er v e rte b ra te p re d a to rs a n d n o tin g the u n iq u e a ttrib u te s o f bones subjected
to cro co d ilian c o n su m p tio n a n d digestion; “ only cro co d ilian s are kn o w n to
decalcify calcified tissues, w hile leaving th eir org an ic m atrices in ta c t” (F ish er
1981:270). H e also describes suspected cau sal facto rs fo r the observed effects on
bones con sum ed by crocodiles, such as v aria tio n in re te n tio n tim e in the
sto m ach a n d bon e ro b u sticity influencing th e ex ten t o f decalcification. H e then
exam ines a P aleocene fa u n a a n d , by carefully elim in atin g o th e r h y p o th etical
causes o f decalcification such as ab rasio n a n d su b aerial w eathering, he p resents
a co n v incing arg u m e n t th a t p re h isto ric crocodiles w ere the ta p h o n o m ic agent
th a t created the fossil assem blage.
F ish e r’s (1981) clearly stated a p p ro a c h to ta p h o n o m ic analysis via the
actu alistic m eth o d (see below ) is lau d ab le a n d ap p ro x im a te s th a t o f others. H e
concisely d o cu m en ts th e distinctiveness o f a set o f a ttrib u te s displayed by a set

o f an im al rem ains. H e believes these a ttrib u te s ca n be used as diagnostic
criteria given suspected causal re la tio n s betw een the a ttrib u te s o f bone
m od ification a n d p a rtic u la r ta p h o n o m ic processes, a n d he elim inates o th er
possible causes. F ish er (1981) clearly describes the origin o f his e x p lan a to ry
prem ises an d th eir a tte n d a n t w eaknesses by in d icatin g several tim es th a t they
are a t p resen t th e best ap p ro x im a tio n , a n d fu rth e r research m ay require their
revision. F ish e r’s stu dy is one o f the best ta p h o n o m ic studies to a p p e a r in recent
years.
A rchaeological exam ples

It is reaso n ab le to w o n d er if zo o arch aeo lo g ists w ith interests in ta p h o n o m ic
p ro b lem s p erfo rm ta p h o n o m ic analyses in a m an n er sim ilar to the tw o
p aleo n to lo g ical exam ples ju s t review ed. T h a t is so because it has been
p ro claim ed th a t h u m an s a n d th eir h o m in id an cesto rs are sufficiently distinct
from o th e r b o ne-processing o rganism s th a t special tre a tm e n t is req u ired w hen
they are p erh ap s p a rt o f the ta p h o n o m ic h isto ry o f a bo n e assem blage (e.g.,
B inford 1981b). In m y view, such arg u m e n ts are tax o n -cen tric a n d u n w a r
ra n te d . F o r exam ple, in an analysis o f an assem blage o f cow (Bos taurus) bones
recovered fro m th e g ro u n d surface inside a cave in N e v ad a, I tested the
hy po thesis p ro p o se d by the ex c av ato r th a t the cattle bones rep resen ted the
rem ain s o f a w in ter m eal left by N ativ e A m ericans early in the tw entieth
cen tu ry (L y m an 1988a). I derived test im plications from the h y pothesis and
searched the cow b o n es fo r evidence o f them . I failed to find convincing
evidence th a t p eo p le h ad an y th in g to do w ith the cattle rem ains. O f the test
im p licatio ns derived, tw o clearly are n o t m et (skulls were n o t fractu red for
b ra in e x tra ctio n b u t ra th e r w ere d isarticu lated along sutures; no b u tchering
m ark s w ere fo und), th ree d o n o t s u p p o rt the hypothesis o f h u m an in terv en tio n
b u t n eith er do they refute it (som e bones m ay have been co o k ed b u t m any h ad
been b u rn e d by n a tu ra l fires; there was no evidence o f differential tra n s p o rt or
u tilizatio n based on econom ic u tility o f skeletal p arts; to o th e ru p tio n an d w ear
suggest a co arse d iet a n d possible w in ter d e a th due to m aln u tritio n ), a n d tw o
w ere n o t m et b u t are am b ig u o u s in term s o f ind icatin g a h u m an tap h o n o m ic
agent.
W hile it m ig ht be suggested th a t the h y p o th esis-testin g p ro c ed u re I used for
th e N e v a d a cattle rem ain s w as reaso n ab le because the b o n e assem blage seems
to re p resen t a n a tu ra lly dep o sited one, th a t does n o t im ply the p ro c ed u re is
in a p p ro p ria te fo r assem blages th a t w ere d ep o sited by h om inids. F u rth e r, while
m y co n clu sio n th a t h u m an s h ad little to d o w ith the cow bones is in direct
c o n tra st to the e x c a v a to r’s, th a t does n o t m ean I believe the ex c av ato r was
stu p id . I believe he w as tap h o n o m ically naive, as we all w ere in the late 1960s
w hen he did his analysis a n d offered his conclusion, a n d th e difference in o u r
T aphonom y in practice and theory 45
conclu sion s sim ply “ m easures the increase in o u r u n d e rsta n d in g o f ta p h o n o

mic pro cesses” over th e in tervening years (L y m an 1988a: 104).
B inford ( 1984b: 10—14) argues th a t we m u st m ain tain a healthy scepticism
a b o u t know ledge claim s p resen ted to d a y because those claim s are to som e
degree th e p ro d u c t o f o u r “ know ledge o f the m o m e n t;” thus, w hen we can
d e m o n stra te th a t a suspected cause-effect relatio n (fo r instance) is in accu rate
“ it is alw ays because we have gained know ledge th a t w as no t available to
[earlier w orkers].” T his echoes F ish e r's (1981) suggestion th a t the cause-effect
relatio n s he p o stu lates are “ w o rk in g h y p o th eses” subject to revision on the
basis o f fu rth e r research.
B inford (1984b) presents several suspected causal links betw een the lo catio n
a n d o rie n ta tio n o f b u tch erin g m ark s on v ario u s bones a n d jo in ts (the effects o f
bu tch erin g ), an d the process o f b u tchering. T he relevant linkages o r causal
relatio n s are fo u n d ed on w h eth er the an im al carcass being b u tch ered is fresh
a n d supple o r p artia lly desiccated an d rigid. Based o n his actualistic research,
B inford (19 8 4 b :7 1) w rites "w h en a carcass is stiff, the jo in ts are generally b o u n d
- th e tissue has sh ru n k an d locked the artic u la tio n in to a fixed p o sitio n , m aking
m an ip u latio n o f th e jo in t im possible. T his m eans th a t the o rie n ta tio n o f cuts
relative to th e sh ap e o f bones will generally be in re g u lar an d determ in ed places,
ra th e r th a n th e [fresh carcass situation] in w hich the o rie n ta tio n o f the c u t shifts
as the jo in t is flexed d u rin g d ism em b erm en t.” B inford (1984b) uses this
suspected causal re la tio n to explain b u tch ery m ark s he observed o n b ones from
K lasies R iver M o u th C ave in S outh A frica site as in d icatin g m an y o f the bovid
rem ains a p p e a r to h ave com e fro m rigid, p artia lly desiccated carcasses
collected by the h o m in id o cc u p an ts o f the site.
In a n o th e r exem plary analysis, S tiner (1990a, 1990b, 1991a) considers the
d em o g ra p h y o f the p o p u la tio n o f large m am m als rep resen ted by fossil
collections from M iddle P aleolithic sites in Italy. She re p o rts th a t she “ was
b o th su rprised a n d fru stra te d by the lack o f clear logical co n n ectio n s betw een
[a m o rtality ] p a tte rn a n d [its] ca u se” a n d th a t “ a g reat d isp arity c u rren tly lies
betw een th e technically so p h isticated m eans fo r co n stru c tin g m o rta lity profiles
a n d the know ledge av ailable to in te rp re t these p a tte rn s ” (S tiner 1991 d:2).
Stiner reviews m uch o f the ecological lite ratu re o n m o rtality , a n d describes how
the b eh a v io ral v ariab ility o f ca rn iv o ro u s p re d a to rs influences the m o rtality
p a tte rn s o f their prey. She builds a m odel fro m available d a ta w hich in c o rp o r
ates cau sal links betw een p re d a to r an d prey beh av io rs as causes, a n d their
effects in term s o f the kind o f prey m o rta lity th a t will resu lt fro m certain
in tera ctio n s o f p re d a to r an d prey taxa. T he m odel is actu alistic because it is
based on m o d ern cases, a n d it becom es an a n a lo g w hen S tiner (1990a, 1990b,
1991a) uses it to help in te rp re t the p re h isto ric reco rd . B ut S tiner is a p p ro p ria
tely cau tio u s. She no tes th a t the m odel is n o t an alg o rith m fo r deciphering
dem o g rap h ic d a ta derived from zo o arch aeo lo g ical rem ains. She p o in ts ou t its
w eaknesses, such as it being based o n d iach ro n ic assem blages a n d th u s it m ay
n o t be ap p licab le to relatively sy nchronic assem blages. T h e m odel c a n n o t, by

itself, “ estab lish the cause o f d e a th o r the agency responsible fo r creatin g a bone
assem b lag e,” a n d in d ep en d e n t ta p h o n o m ic research m u st help establish those
causes a n d agencies (S tin er 1990b:341). T he m odel is suggestive o f p re d a to ry
b eh av io rs due to th e suspected causal linkages o f p re d a to r beh av io rs a n d prey
m o rta lity type. But S tiner (1990b:343) recognizes th a t ad d itio n a l n eo ta p h o n o -
m ic d a ta “ will u n d o u b te d ly im p ro v e a n d s tre n g th e n ” th e m odel.
S u m m a ry
T h e exam ples review ed sh are several featu res o f the fo u n d a tio n s o f ta p h o n o

m ic research. T hese include reference to m o d e rn cases w herein the re la tio n s o f
causes a n d effects are k n o w n o r suspected, testing o f h y p o th etical tap h o n o m ic
histories, a n d th e realizatio n th a t all conclusions reg ard in g tap h o n o m ic
h istories are the best acco u n ts presently available. C on clu sio n s m ay be m o d i
fied as new d a ta a n d /o r co n cep ts in eith er m o d e rn o r p re h isto ric settings com e
to light. Sim ply, ta p h o n o m ic research is a p a rtic u la r kind o f scientific research.
A ta p h o n o m ist is n eith er o m n ip o te n t n o r infallible. T heories an d concepts are
assessed in the em pirical w orld, m odified, a n d assessed again. T o delve fu rth e r
in to th e en terp rise o f ta p h o n o m ic research, we n ow tu rn to its epistem ological
un derp in n in g s.
Lniformitarianism and actualism
T here are o perations p ro p e r to the surface o f this globe by w hich the form o f the
habitab le e a rth m ay be affected; o p eratio n s o f w hich we un d erstan d b o th the causes
and effects, and, therefore, o f w hich we m ay form principles fo rju d g in g o f the past.
(J. H u tto n 1795; cited in S charnberger et al. 1983:312)
G iffo rd (1981) o utlin es the co m m o n p ercep tio n o f the th eo retical a n d m e th o d

ological basis o f ta p h o n o m ic analysis. She ad vocates th a t we p h ra se o u r
research goals a n d m eth o d s in term s o f the “ u n ifo rm ita ria n m eth o d o lo g y an d
a ssu m p tio n s” w hich u n derlie o u r discipline (G ifford 1981:397). T he basis o f
this ad v o cacy can be fo u n d in the G e rm a n a k tiio -p alae o n to lo g ie p ro p o se d by
R ich ter (1928) a n d defined by him as the science o f the origin an d p resen t-d ay
m ode o f fo rm a tio n o f fu tu re fossils, a n d the ap p lica tio n o f th a t know ledge to
paleo n to lo g ical p ro b lem s via the m eth o d o lo g y o f u n ifo rm itarian ism (W arm e
an d H an tzsch el 1979). U n ifo rm itaria n ism , in one form o r a n o th e r, perm eates
all aspects o f geology (H o o y k a as 1970; S chum m 1985; W a tso n 1969), p a le o n
to log y (S. J. G o u ld 1965, 1967, 1979; S im pson 1970), paleoecology (L aw rence
1968, 1971; N a irn 1965; S cott 1963), arch aeo lo g y (A scher 1961a, 1961b;
B inford 1981b; R .A . G o u ld 1980; Stiles 1977), arc h a e o fa u n a l analysis (L ym an
1982a; M edlock 1975; R ack h a m 1983) a n d ta p h o n o m y (B rain 1981; H ill 1978,
1988; S hip m an 1981b). It is th ere fo re a p p ro p ria te to co n sid er first the concept
o f u n ifo rm itarian ism a n d the related co n cep t o f actualism .
W hat is uniform itarianism ?

M o dern [m ethodological] un ifo rm itarian ism has no substantive c o n te n t-th a t is, it
asserts n o th in g w hatever ab o u t n atu re. U n ifo rm itarian ism m ust be viewed as telling
us how to behave as scientists an d n o t as telling n atu re how it m ust behave.
(J. H. Shea 1982:458)
W hile the co n cep t o f u n ifo rm itarian ism is usually a ttrib u te d to Jam es H u tto n ,
it received g re ater success as a guiding scientific co n c ep t in the w ritings o f
C h arles Lyell (G o u ld 1979, 1982, 1984; H a n eb erg 1983; H o o y k a a s 1970;
R u dw ick 1971; W ilson 1967). T he detailed m ean in g a n d c o n n o ta tio n s o f the
co n cep t have ch an g ed over the 160 o r so years since its in tro d u c tio n (W.
W hewell [1832] suggested the term in a review o f Lyell's Principles o f G eology),
b u t its basic m eanin g h as rem ained m o re o r less the sam e. It is p erh ap s because
o f the ev o lu tio n o f th e co n c ep t's m ean in g th a t m uch m isu n d erstan d in g exists
a n d v ario u s m ean in gs are a ttrib u te d to it. H ere I review w h a t u n ifo rm ita ria
nism stan d s fo r to d ay , an d w h a t som e scientists (m ostly geologists) th in k ab o u t
the co n cept.
U niform itarianism , as p h ra sed by Lyell, consists o f tw o m a jo r p arts: a
testab le th eo ry , a n d an an alytic p ro c ed u re o r assu m p tio n (G o u ld 1965, 1979;
R udw ick 1971). T h e th eo ry entails tw o hypotheses. T he first is labeled
“ g ra d u a lism ” a n d suggests th a t rates o f ch ange have been u n ifo rm th ro u g h o u t
tim e a n d th a t large results are n o t the p ro d u c t o f su dden c a sta stro p h ic causes,
b u t th e ac cu m u lated effect o f in n u m era b le m in u te changes (G o u ld 1979:126—
127). T h e second h y p o th esis is labeled “ n o n p ro g re ssio n ” a n d suggests th a t the
co n fig u ratio n o f th e e a rth is in a d ynam ic steady-state; change is incessant b u t
cyclic (G o u ld 1979:126-127). L abelled “ su b stan tiv e u n ifo rm ita ria n ism ”
(G o u ld 1965), the th eo retical p a r t o f u n ifo rm itarian ism now seem s false, a n d if
“ to o rigidly held [becom es] an a p rio ri assu m p tio n , stifling to the fo rm u latio n
o f new hy p o th eses w hich m ay b e tte r explain certain d a ta ” (G o u ld 1965:226).
T he an aly tic p ro c ed u re , labelled “ m eth o d o lo g ic al u n ifo rm ita ria n ism ” (G o u ld
1965), m ak es the p a st am en ab le to purely scientific ex p lan a tio n , a n d also
consists o f tw o p arts. F irst, it assum es th a t n a tu ra l law s are in v a ria n t in tim e
an d space. S econd, it assum es th a t processes have been in v a ria n t in tim e a n d
space, th ere fo re p a st results m ay be p ro p e rly ascribed to causes now in
o p e ra tio n (G o u ld 1979:123, 126). T he tw o p a rts o f m eth o d o lo g ical u n ifo rm i
taria n ism are req u isite to in ferrin g p a st dynam ics; th e a n aly tic process involves
the asso ciatio n o f m o d ern results w ith p a rtic u la r m o d e rn processes. W hen
sim ilar results, som e form ed in an cien t tim es a n d o th ers fo rm ed in m o d ern
tim es, are fo u n d , the inference is m ad e th a t th e processes w ere the sam e o r at
least sim ilar in b o th the p a st a n d p resen t cases.
S im pson (1963:24-25, 1970) suggests “ im m a n en t p ro p e rtie s” are the
“ u n ch a n g in g p ro p e rtie s o f m a tte r a n d energy a n d p rinciples arising th e re fro m ”
(including processes such as gravity) a n d “ co n fig u ratio n al p ro p e rtie s” are the
arran g e m e n ts a n d o rg a n iz atio n s o f the m a tte r m ak in g u p the w orld. T he
fo rm er is su b su m ab le u n d er m eth o d o lo g ical u n ifo rm itarian ism , the la tte r is

m ore o r less sy n ony m o us w ith su b stan tiv e u n ifo rm itarian ism . It is, o f course,
the fo rm e r th a t is o u r co ncern here.
G o u ld (1965:223) defines m eth o d o lo g ical u n ifo rm itarian ism as “ a p ro c e
d u ral p rinciple assertin g sp atial a n d tem p o ra l invariance o f n a tu ra l law s.” But
S im pson (1970:59) suggests “ the in v ariab ility o f n a tu ra l laws is indecisive
a b o u t such basic p ro b lem s as th eir sufficiency o r as to w h eth er the actio n s o f all
h istorically relev an t law s are c u rren tly (‘ac tu a lly ’) o b serv ab le.” A s W a tso n
(1966) m ak es clear, this is a p ro b lem o f scale given, for instance, th a t hu m an s
live only a b o u t 70 years an d th u s one individual c a n n o t observe the fo rm a tio n
o f a n ero sio n al fe atu re th e m ag n itu d e o f th e G ra n d C an y o n . W a tso n (1966)
p o in ts o u t th a t if the law s o f ero sio n w ritte n by scientists are co rrect, th en we
can infer h o w th e G ra n d C an y o n a n d sim ilar features w ere form ed by reference
to th o se laws.
K itts (1977:63-64) discusses tw o prin cip al aspects o f m eth o d o lo g ical u n ifo r
m itarian ism . F irst, th e g ro u n d s u p o n w hich the assertio n o f u n ifo rm ity rests
are n o t testable. If the g ro u n d s fo r the assertio n are th a t the p ast w as like the
p resen t (th e n a tu ra l law s we have w ritten are tem p o ra lly in v aria n t), we ca n n o t
test th em because we c a n n o t observe th e past. Second, m eth o d o lo g ical
u n ifo rm itarian ism im poses restrictio n s o n statem en ts m ad e a b o u t the p ast, i.e.,
ap p eal to s u p e rn a tu ra l o r u n k n o w n forces are precluded. C learly, w ith o u t this
re stric tio n one co u ld m ake an y statem en t a t all a b o u t the p a st, including
statem en ts th a t c a n n o t be em pirically tested. Y et, K itts (1977) co n tin u es, one
m ust subscribe to m eth o d o lo g ical u n ifo rm itarian ism in o rd e r to m ak e h isto ri
cal inferences. "B ecause it is im possible fo r us to observe an y th in g except the
present, o u r in te rp re ta tio n s o f p rio r events m ust necessarily consist o f in fer
ences b ased u p o n p resen t o b se rv a tio n s” (H u b b e rt 1967:29-30). O u r assu m p
tions m u st th ere fo re be ( 1) n a tu ra l law s are in v a ria n t w ith tim e, a n d (2 )
v io latio n o f n a tu ra l law s by an y n o n -n a tu ra l m echanism is excluded from
co n sid eratio n .
G o o d m a n (1967) a n d S hea (1982) arg u e th a t u n ifo rm itarian ism is a term for
how we d o science. It involves the c o n fro n ta tio n o f em pirical d a ta w ith theory,
an d m od ification o f the la tte r if necessary, in o rd e r to derive an ex p lan a tio n o f
the em pirical item s u n d e r stu d y (see also G o u ld 1965, 1967; W a tso n 1966). This
p ro c ed u re is c o m m o n to all sciences as it presum es n a tu ra l law s are tem p o rally
in v aria n t. All sciences involve in d u ctio n a n d sim plicity, tw o key aspects o f
m eth o d o lo g ical u n ifo rm itarian ism (G o o d m a n 1967; G o u ld 1967). Sim plicity is
the denial o f processes o r forces unique to the p a st (G o u ld 1965), w hich is to say
th a t m eth o d o lo g ical u n ifo rm itarian ism assum es n a tu ra l law s are atem p o ral.
S im plicity recognizes th e fact th a t scientists w rite the laws, w hich are d escrip
tions o f how we th in k things w ork to d a y (G o o d m a n 1967). If we d o n o t have an
ex p lan a tio n fo r a p h e n o m e n o n th a t w as created in th e p ast, th e n we have not
y e t observed th e processes in o p e ra tio n to d a y w hich create th a t kin d o f
p h en o m en o n (S co tt 1963).
S alm o n (1953) co ncludes th a t because we c a n n o t d e m o n stra te n a tu ra l laws

to be in v a ria n t in the p ast, o u r conclusions are inductively derived. Such
co nclusions c a n n o t be show n to be tru e b u t ra th e r “ we ca n ju stify them by
show ing th a t they are useful in pred ictin g an d ac tin g ” (S alm on 1953:47). W e
can, S alm o n (1953) suggests, ju stify inductive a rg u m e n ts w ith o u t assum ing any
unifo rm ity o f n atu re . T h u s, I believe, because o u r inductive conclusions are
based on the a ssu m p tio n o f u n ifo rm ity o f n atu re , o u r conclusions are
accep tab le in a scientific sense. I agree w ith K itts (1977:13) th a t m eth o d o lo g ical
u n ifo rm itarian ism “ m ay be view ed as a m eth o d o lo g ical device o r co n v en tio n
th a t lim its the gen eralizatio n s used in p rim ary h isto rical geological ex p lan a
tions to statem en ts th a t m eet the em pirical req u irem en ts set for an y valid
g en eralization in science, th a t is, th a t they m ay have been verified a n d have n o t
been falsified here a n d n o w .”
W hat is actualism?
. . . the anim als died durin g a few weeks o f d ro u g h t, and their rem ains were
accum ulated a ro u n d the few rem aining pools o f w a te r-a theory th a t from an actuo-
geological p o in t o f view is very well founded.
(B. K urten 1953:69)
A ctualism asserts sp atial a n d tem p o ra l in v arian ce o f n a tu ra l laws (H an eb erg

1983), p artic u la rly those co n cerned w ith m echanical, chem ical, a n d physical
processes (b u t n o t b eh av io ral ones). It th u s is eq u iv alen t to m eth o d o lo g ical
u n ifo rm itarian ism . “ It refers to u n ifo rm ity [of im m a n en t properties] in all fo u r
dim ensions o f space a n d tim e” (S im pson 1970:63). A ctualism “ d en o tes the
m eth o d o lo g y o f in ferrin g the n a tu re o f p a st events by an alogy w ith processes
o b serv able in a c tio n a t th e p re se n t” (R u d w ick 1976:110). A c tu o p a le o n to lo g y is
th e expression o f actu alism p e rta in in g to p aleo n to lo g ical issues. N e o ta p h o -
n o m y (H ill 1978) a n d eth n o arch ae o lo g y (Stiles 1977) are sim ilar expressions o f
actu alism relevan t to th eir respective fields o f study. All o f these term s m ay be
su bsu m ed u n d e r the co n cep t o f actualism .
N o n -a ctu alistic m eth o d s assum e th a t changes m ay have been w ro u g h t by
agents o f different k in d th a n th o se o bservable to d ay , a n d include c a ta stro p h ism
in its classic sense (agents o f different kin d a n d intensity). A ctualistic m eth o d s
assum e ag en ts w ere o f the sam e kin d a n d o f the sam e o r different intensities
(H o o y k a as 1970). W hile including classic m eth o d o lo g ical u n ifo rm itarian ism
(agents o f sam e kin d a n d intensity), actualistic m eth o d s also allow varian ce in
energy o r in tensity across tim e w hile denying tem p o ra lly un iq u e kinds o f agents
(H o o y k a a s 1970; F ig u re 3.1).
A re actu alistic m eth o d s necessary to the stu d y o f h isto rical p h enom ena?
M an y w ould an sw er “ Y es” (B inford 1981b; G ifford 1981; G o u ld 1965;
S im pson 1970; W a tso n 1969, 1976), w hile a m in o rity w ould answ er “ N o ”
(S cott 1963). T he d isag reem en t results from perceived sh o rtco m in g s o f the
actu alistic m eth o d , th e degree to w hich actualistic m eth o d s are req u ired a n d /o r
Vertebrate taphonom y
Inte nsit y of P r o c e s s e s
Same as Today D i f f e r e n t f ro m Today
ZJ) classic actualism

<D
TD uniform itarianism,
O
H actualism
c/>
CD
a)
E im m an e n t p r o p e r tie s
CD
CO
methodological u n i f o r m i t a r i a n i s m
3ub3tantive
Z7> u niform ita rianism

CD
TD
O nonactualistic nonactualistic
methods, methods,
.£ E
o
n o n u n i fo rm i ta r i a n n o n u n i fo rm i t a r i a n
methods metho ds
c
a>
cu
c a t a stro p h is m
Figure 3.1. Intersection o f different kinds an d intensities o f historic (taphonom ic)

processes defining u n iform itarianism , actualism , an d cata stro p h ism as p aradigm s
for explaining the past. Substantive un ifo rm itarian ism can encom pass all four
categories; m ethodological u n ifo rm itarian ism an d im m an en t p roperties assum e
processes o f the sam e kind an d o f either the sam e o r different intensity as
observed today.
used to stu d y p a st p h e n o m en a , a n d co n fu sio n o f su b stan tiv e a n d m e th o d o lo g i

cal u n ifo rm itarian ism . A ctualism is req u ired a n d largely u n q u estio n ed w hen a
p a rtic u la r fossil m u st be identified as a fem ur (G ifford 1981). T he identification
is u n q u estio n ed because o f th e necessary a n d sufficient causal re latio n betw een
the genetic co n tro ls a n d onto g en ic processes (im m an en t p ro p erties) resulting in
the fo rm a tio n o f th e fem ur, a n d th e fem u r itself. T o arg u e th a t the n a tu ra l
h isto ry a n d b eh a v io r o f living h y aen as are th e sam e as th o se o f extinct g enera o f
h y aen as is, how ever, q u estio n ab le because this tra n sfe r o f m o d e rn ecological
p aram eters to th e p ast (L aw rence 1971) is based on sufficient causal relatio n s o r
co n fig u ratio n al p ro p e rtie s a n d represents su b stan tiv e u n ifo rm itarian ism .
T h ere are tw o m a jo r criticism s typically leveled against actualistic m eth o d s
th a t are sim ilar to th o se o f m eth o d o lo g ic al u n ifo rm itarian ism . T he first is th a t
w hen th e actu alistic m eth o d is used, the an a ly st m u st assum e th a t observable

processes a n d th eir results are the sam e to d a y as in the p a st (n a tu ra l law s are
tem p o rally in v arian t), an d this is an u n p ro v a b le an d th erefo re p o ten tially false
assu m p tio n (B info rd 1981 b:27; S im pson 1970:62-63). A dvocates o f the a c tu a
listic m eth o d in d icate th a t the a ssu m p tio n m u st be w a rra n te d in o rd e r to use the
m eth o d (B inford 1981b; G o u ld 1965). A s we have seen, the a ssu m p tio n o f
tem p o ra l in v arian ce o f n a tu ra l law s c a n n o t be w a rra n te d because it ca n n o t be
tested; we c a n n o t observe the p ast. S im pson (1970:62) argues th a t geological
a n d p aleo n to lo g ical o b serv atio n s on p h en o m en a created over the en tire history
o f th e w o rld are ex p lain ab le using actualistic m eth o d s, a n d “ th a t is the source
a n d p rinciple s u p p o rt o f the c a n o n o f ac tu a lism ” in th o se a n d o th e r sciences. O f
course, sim ply because it w orks does n o t m ake it so. W e c a n n o t present
h isto rical events as “ instances o f co n firm a tio n fo r [the assum ption] we w ish to
te st” if th a t assu m p tio n “ has been p resu p p o se d in in ferrin g the ev en ts” (K itts
1977:79). N o netheless, geologists, p aleo n to lo g ists, an d arch aeo lo g ists m ust,
a n d do, g ra n t th e assu m p tio n , largely because it w orks, a n d there is n o stro n g
altern ativ e as yet.
T he second criticism suggests th a t exclusive use o f the actu alistic m eth o d as a
basis o f in te rp re ta tio n m ay conceal im p o rta n t insights to the p a st (S cott 1963;
G o u ld 1980; B inford 1981b; B ehrensm eyer 1988a). W hile this criticism m ight
be co n fu sin g su b stan tiv e w ith m eth o d o lo g ic al u n ifo rm itarian ism , a n d thus
co n fig u ratio n al w ith im m a n en t p ro p e rtie s, ad v o cates o f the actu alistic m eth o d
argue th a t w hen h isto rical p h en o m en a c a n n o t be explained using actualistic
m eth o d s, “ a m in im al inference is th a t know ledge o f present processes is
in co m p lete” (S im pson 1970:84). W hen actu alistic a n d p re h isto ric d a ta c a n n o t
be m atch ed an d the la tte r con seq u en tly n o t explained, “ a m axim al inference is
th a t there have been p ast processes n o t now o p erativ e” (S im pson 1970:84). T he
u n ex p lain ab le “ resid u e” (S cott 1963:516) o f p h en o m en a seems to be sm all
co m p ared to th e n u m b e r an d d iversity o f p h en o m en a th a t are explainable using
actu alistic m eth o d s, b u t has served as the basis for a rejection o f u n ifo rm ita ria
nism (G o u ld 1980).
T he second criticism can be p h ra sed a n o th e r way: W hen an u n explainable
residue o f p h en o m en a rem ains th e an a ly st m u st invoke ad hoc arg u m en ts to
explain the residue a n d this “ w eakens the assu m p tio n ” o f tem p o ra l invariance
o f n a tu ra l laws (S co tt 1963:513). A d v o cates o f the actu alistic m eth o d argue
th a t the occasio n al necessity o f inv o k in g a d h o c arg u m e n ts is d u e largely to
incom plete know ledge o f p resen t processes, n o t som e in te rn a l w eakness o f the
m eth o d . E ven th o se reco m m en d in g th a t actu alism be a b a n d o n e d realize th a t
“ th ere is n o such th in g as the final o r u ltim ate in te rp re ta tio n - only b e tte r an d
b e tte r ap p ro x im a tio n s o f p ast reality ” (G o u ld 1980:46).
It is one th in g to criticize a m eth o d o f analysis a n d entirely a n o th e r to offer an
equally (o r m o re) p ro d u c tiv e altern ativ e. W hile n o t a d irect criticism o f the
actu alistic m eth o d , a stro n g arg u m e n t fo r a b a n d o n m e n t o f it w ould be the
d ev elo pm en t o f a m e th o d n o t subject to eith er o r b o th o f th e tw o criticism s and

yet cap ab le o f p erm ittin g e x p lan a tio n s c o m p arab le to , b e tte r th a n , o r m o re
com plete th a n th o se p erm itted by actualistic m eth o d s. T h u s far. no altern ativ es
have been outlined.
Actualism in archaeology and taphonomy
S olom on (1990:29) suggests th a t w hen ta p h o n o m ic research is im plem ented

the w ay E frem o v in ten d ed , it involves “ the seeking o f laws th a t describe the
processes by w hich th e rem ains o f an im als becom e in c o rp o ra te d in to the
lith o sp h ere, w ith equal em phasis given to the geological a n d biological
fo rm a tio n o f the site.” She (S olom on 1990:29) argues th a t “ w h at we d o . in
arch aeo lo g y specifically, is a c tu o p a le o n to lo g y ” as defined by R ich ter (1928).
S o lo m o n (1990:30-31) d raw s a tte n tio n to perceived differences betw een E fre
m o v ’s ta p h o n o m y , w hich she describes as law seeking a n d thus n o m o th etic,
an d actu o p a leo n to lo g y , w hich she describes as n o n -n o m o th e tic a n d em ploying
u n ifo rm ita ria n m eth o d s “ to solve p ro b lem s fo u n d at individual sites.” T h ere
fore. w h at ta p h o n o m ists d o is a c tu o p a le o n to lo g y an d n o t tap h o n o m y , a n d “ to
a tte m p t to invoke a law w hich will explain w h a t h ap p e n ed at a site is fo o lh ard y .
T here is no such law ” (S o lo m o n 1990:31).
S o lo m o n (1990:26) defines a law as a “ co rrec t sta te m e n t o f invariable
sequence betw een specified co n d itio n s an d specified p h en o m en a; reg u larity o f
n a tu re .” T h u s S o lom on has identified a ta p h o n o m ic (a n d sem antic) red
h erring. T o suggest th a t because ta p h o n o m y is a h isto rical science, a n d th a t
h isto rical sciences d o n o t o r c a n n o t use law s is false. It ignores the fact th a t
biological evolu tio n an d the stu d y o f it via p aleo n to lo g y is an h isto rical science
involving th e use o f law s o f n a tu ra l selection o r w hat G o u ld (1986:64) term s
“ n o m o th etic u n d e rto n e s.” F u rth e r, one o f the m a jo r p ro p o n e n ts o f studying
the fo rm a tio n a l (including ta p h o n o m ic) processes o f the arch aeo lo g ical record
has w ritten th a t “ fo rm a tio n a l processes exhibit regularities th a t can be
expressed as law s” (Schiffer 1987:11). C ad ee (1990:13) suggests th a t “ the
m eth o d s th a t E frem ov p ro p o se d fo r ta p h o n o m ic a l studies are no t significantly
different from those o f W eigelt a n d R ic h te r.” E frem ov (e.g.. 1940:92) w as n o t
only aw are o f his p redecessors (e.g., E frem ov 1940:92), he no ted th a t he was
after " p rin c ip le s” ra th e r th a n laws (E frem ov 1958; afte r C adee 1990), w hich I
tak e to m ean he w as well aw are o f the difference betw een su b stan tiv e an d
m eth o d o lo g ical u n ifo rm itarian ism .
I suspect w h at S olom on (1990) has d o n e is confused su b stan tiv e an d
m eth o d o lo g ical u n ifo rm itarian ism , a n d co nfused im m an en t a n d co n fig u ratio
nal p ro p erties. Im m an e n t p ro p e rtie s include those im m u tab le physical and
chem ical reactio n s th a t o ccu r w ith p red ictab le results regardless o f spatio-
tem p o ra l co ntex t. C o n fig u ratio n al p ro p e rtie s, because they are co n tex t speci
fic, are h isto rical a n d m u tab le; because they co n cern h u m a n b eh av io rs in
arch aeolo g y , an d b eh a v io r changes th ro u g h tim e, we m u st test any su b stan tiv e

(co nfig u ratio n al) law s o f h u m an beh av io r. W h a t is im p o rta n t is the fact th a t
while tap h o n o m ists speak o f “ ta p h o n o m ic h isto ries,” ta p h o n o m ic research is
fo u n d ed in m eth o d o lo g ical u n ifo rm ita ria n ism a n d im m a n en t p ro p e rtie s, o r
actualism .
U se o f the actualistic m eth o d is im plicit th ro u g h o u t m uch o r all o f
arch aeo lo gical m eth o d an d theory. O ne clear statem en t o f its use is p rovided by
W a tso n (1976:621):
A rchaeologists w ho p ropose laws o f cu ltu ral evolution [and] o f h u m an behavior
assum e - as do o th er scientists - th a t the p ast was like the present, an d although
com binations an d rates m ay have been different th en from w hat they are now , the
basic b ehavioral characteristics o f m en an d m aterial were n o t different then from
w hat they are now. T his m ay in fact be w rong. B ut if h u m an n atu re and the
environm ent were radically different in the p ast from w hat they are now , we assum e
th a t there has been lawlike change from the past to the present th a t can be derived
and u nd erstoo d from its physical rem ains. This general u n iform itarianism is the
prim ary pro ced u ral o r m ethodological assum ption o f archaeology, as it is o f all
o th er sciences.
T reatises on arch aeo log ical m eth o d a n d th eo ry are quick to p o in t o u t th a t if

p reh isto ric d y nam ics are a goal o f analysis, inferences re g ard in g p ast dynam ics
m u st be fo u n d ed o n how the m o d ern w orld w orks (B inford 1981b; G o u ld
1980). T a p h o n o m ists, because they o f necessity deal w ith p a st d ynam ics, m ake
sim ilar arg u m en ts (G ifford 1977, 1981; Hill 1978, 1980). T he arg u m e n ts are
straig h tfo rw ard : p re h isto ric dynam ics are n o t o bservable, only th eir static
results are, alo n g w ith m o d ern dynam ics a n d th eir re su lta n t static effects.
T he p ro b lem th u s becom es one o f d eterm in in g how the actu alistic m eth o d is
im plem ented in arch aeo lo g y an d tap h o n o m y , an d if it is im plem ented properly.
T he actu alistic m eth o d o f stu d y in g the p ast is superficially sim plistic (e.g.,
F ish er 1981). O bserve present processes in actio n , estab lish o r p o stu late causal
relatio n s betw een p a rtic u la r processes o r dynam ics a n d p a rtic u la r static effects,
an d m atch the m o d ern record to the p ast, inferring sim ilar processes fo r sim ilar
effects on the basis o f the causal relations. Is this how arch aeo lo g ists and
tap h o n o m ists perceive an d use the actu alistic m ethod? I co n sid er each in tu rn .
A rchaeology and actualism

T he m o st c o m m o n fo rm th a t the actu alistic m eth o d has ta k e n in a rch aeo lo g y is
e th n o g ra p h ic analo g y . T his is reflected in statem en ts like “ in its m o st general
sense in te rp re tin g by an alo g y is assaying any belief a b o u t no n o b serv ed
b eh a v io r by referral to observed b eh a v io r w hich is th o u g h t to be re le v an t”
(A scher 1961 a :3 17). It is generally ca u tio n e d th a t an alo g s derived from
e th n o g ra p h ic d a ta be used m erely as clues to general co n d itio n s, n o t p a rtic u
lars, because th ere is the p o te n tia l o f m u ltip le m o d e rn an alo g s fo r p a rtic u la r
archaeo lo gical p h en o m en a (A scher 1961 a, 1961 b). Pleas fo r e th n o g ra p h ic d a ta

w ith specific ap p licab ility to arch aeo lo g ical research have been m et w ith a
v o lu m in o u s b o d y o f eth n o arch ae o lo g ic al d a ta g enerated by arch aeo lo g ists
(e.g., G o u ld 1978; K ra m e r 1979; an d B onnichsen a n d Sorg 1989; H u d so n
1993).
A classic exam ple o f e th n o g ra p h ic analo g y is fo u n d in B inford (1967).
B inford (1967) p o in ts o u t th a t analogic re aso n in g is inductive, a n d e th n o
g rap h ic an alo g y involves m atch in g the observable arch aeo lo g ical re co rd w ith
the observed eth n o g ra p h ic reco rd follow ed by the p ro p o sa l th a t, w hen sim ilar,
b o th static reco rd s can be explained as resulting fro m the sam e processes. Such
arg u m e n ts can o nly be ev alu ated by m o n ito rin g c o n c o m ita n t v aria tio n s o f
a ttrib u te s in b o th static records, a n d if th e v aria tio n s are sim ilar in b o th
reco rd s, th e inference is stren g th en ed (B inford 1967). T he im p licatio n here is
th a t this co n c o m ita n t v a ria tio n indicates a causal re la tio n betw een the m o n i
to red variables. B in fo rd 's (1967) inference re g ard in g a p a rtic u la r a rch ae o lo g i
cal p h en o m en o n is q u estio n ed by M u n so n (1969), w ho provides an equally
viable b u t altern ativ e inference based on a different a n alo g . T his illu strates th a t
the a ttrib u te s b o th B inford a n d M u n so n have used are n o t, p erh ap s, causally
related to d y n am ic processes w hich created the ph en o m en a.
A different kind o f analogic re aso n in g is pro v id ed by Schiffer (1976:162),
w ho estim ates the a m o u n t o f lithic m ateria l in a site o n the basis o f w h a t was
recovered (m ultip lyin g the a m o u n t in the recovered sam ple by the reciprocal o f
the sam p lin g fractio n ). Schiffer here assum es a u n ifo rm ity betw een th e density
o f lithic m aterial in the sam pled p o rtio n o f the site, a n d the density o f lithics in
the u n sam p led p o rtio n , w hich in tu rn co n stru es the density o f lithic m ateria l as
causally related to th e a m o u n t o f sam pled space. W h e th e r o r n o t the sam ple
p o p u la tio n a n d th e sam p led space are each rep resen tativ e o f th eir respective
sam ple universe in a statistical sense is n o t k n o w n . A n d w hile it is tru e th a t as we
dig m ore we tend to find m o re, it is n o t a t all clear th a t lithics are ra n d o m ly
d istrib u te d th ro u g h o u t the site, th erefo re, it is n o t k n o w n if the variables
involved are co rrelated . O f relevance to ta p h o n o m ic studies is the sim ilar
p ractice o f estim atin g the n u m b e r o f an im als o r bones in a kill site from the
n u m b er recovered (e.g., D ibble a n d L o rra in 1968; R eher a n d F riso n 1980).
S eldom are these estim ates ju stifiab le because they en tail co n fig u ratio n al
p ro p e rtie s a n d su b stan tiv e u n ifo rm itarian ism .
A tte m p ts to b ase an alo g s o n causal re la tio n s are reflected in A sch er’s (1961a)
c h a ra c te riz a tio n o f th e “ new an a lo g y ” as including b o u n d a ry co n d itio n s (Stiles
1977; W a tso n 1979). B o th the a n a lo g an d the p h e n o m e n o n u n d e r scrutiny m u st
o ccu r in sim ilar e n v iro n m e n ta l settings, o r the technological levels o f the tw o
involved cultures m u st be sim ilar. A nalogies are even stro n g er, it is arg u ed , if
the m o d e rn a n a lo g is from a h u m a n g ro u p k n o w n to have descended fro m the
archaeolo gical g ro u p (e.g., A n d e rso n 1969; C h a rlto n 1981; Lange 1980); this
im plies th a t an an a lo g o u s re la tio n will be stro n g er a n d a causal re la tio n m ore
certain if the m o d e rn an a lo g a n d the p re h isto ric subject are also h o m o lo g o u s

(i.e., share an an cesto r). B o u n d ary c o n d itio n s are sim ultaneously co n stru e d as
p ro v id in g a m ean s o f co n tro llin g the choice o f m o d e rn analogs in o rd e r to
increase confidence in re su lta n t inferences by im p ro v in g the distinctiveness o f
the ch osen a ttrib u te s (A scher 1961a), a n d as a device th a t results in lim itatio n s
on possible inferences (F re em an 1968). B oth o f these facto rs are believed to
im pose a kind o f p ro xy causal relatio n on the variables in the analog. If a
p reh isto ric an d an existing society occupy sim ilar en v iro n m en ts an d utilize
sim ilar technologies, then beh av io rs m u st be sim ilar due to the assum ed direct
re la tio n o f a p a rtic u la r tech n o lo g y fo r exploiting a p a rtic u la r en v iro n m en t.
T his re aso n in g rings w ith en v iro n m e n ta l determ inism a n d is surely fo u n d ed in
su b stan tiv e u n ifo rm itarian ism a n d co n fig u ratio n al pro p erties.
T o use actu alism pro p erly , B inford (1981 b:26—27) argues th a t a “ necessary
causal re la tio n ” betw een a p a rtic u la r process an d its result(s) m u st be e sta b
lished, a p o in t B inford (1972) recognized in the d eb a te w ith M u n so n (1969).
B inford (1981 b:26) states th a t “ if we can isolate causal re la tio n s betw een
things, a n d if we can u n d e rsta n d such relatio n s in term s o f m o re general
principles o f necessity, such as the theories o f m echanics, then we have a stro n g
w a rra n t fo r the inference o f the cause fro m the observed effect.” F o r B inford
(1981 b:26-29), a necessary causal re la tio n is one th a t is “ c o n s ta n t an d u n iq u e,”
an d estab lish m en t o f such relatio n s will allow us to specify archaeologically
visible “ sig n atu re p a tte rn s ” th a t allow d iscrim in atio n o f one agent o r process
from all oth ers. E stab lish in g these kinds o f relatio n s faces tw o challenges: (1) is
the re latio n in fact causal a n d no t ju s t c o rre la tio n a l (a p o in t I re tu rn to in the
follow ing section), a n d (2) was the process ch a rac te ristic o f the past? T he latter
clearly involves im m an en t p ro p e rtie s, o r w hat B inford (1981 b:29) term s
“ etern al o b jects.” T he search fo r im m a n en t p ro p e rtie s o r tem p o rally in v aria n t
n a tu ra l law s has been term ed m iddle-range research by B inford (1977).
T h e c o n cern fo r d iag n o stic c riteria is a n im p o rta n t one w hich h as p erm eated
m any treatises on archaeological reasoning. F ritz (1972:137) im plies th a t
d iag no stic criteria in d icate the presence o f p a rtic u la r p ast p h en o m en a , b u t m ay
be difficult to establish. H e notes th e actu alistic m eth o d is requisite to
explaining th e archaeolo gical re co rd b u t there is a n absence o f explicit
a rg u m e n ts atte m p tin g to d e m o n stra te th a t a p a rtic u la r clu ster o f crite ria are
d iag no stic o f a p a rtic u la r p ast d ynam ic (F ritz 1972:138, 143). H e labels these
“ arg u m e n ts o f relevance” the req u irem en ts o f w hich involve u n d erta k in g
actu alistic research a n d establishing diag n o stic criteria. S m ith ’s (1977:607, 611)
“ p lausib ility c o n sid e ra tio n s” are id entical to arg u m e n ts o f relevance, an d
n arro w the rang e o f possible ex p lan a to ry h ypotheses. T he ex p lan a to ry h y p o th
eses w ith the low est p rio r p ro b a b ilities o r w ith the least distinctive criteria are
n o t considered. I f necessary, the final selection o f one o f several com peting
h y p o th eses is accom plished by exam ining co n c o m ita n t v a ria tio n o f different
variab les in th e arch aeo lo g ical a n d actu alistic records. O f course, co n c o m ita n t
v a ria tio n suggests th a t criteria w ere in fact distinctive b u t w ould n o t necessarily

in dicate th a t th e re la tio n s betw een variables w ere cau sal o r o f B in fo rd ’s
“ n ecessary” kind.
G o u ld (1980) recom m ends the disposal o f u n ifo rm ita ria n arg u m en ts. H e
no tes th a t w hile arch aeo lo g ists generally assum e th a t th e processes w hich
stru c tu re th e e th n o g ra p h ic re co rd have also stru ctu re d the archaeological
reco rd , ( 1) this n o tio n is self-lim iting because we assum e the very things we are
trying to find o u t, ( 2 ) th e n o tio n is subject to the fallacy o f affirm ing the
c o n seq u en t, (3) we have no g o o d an alo g s fo r m an y a d a p ta tio n a l form s o f
cu ltu re, a n d (4) u n ifo rm itarian ism , w hich he does n o t define, is a “ seductive
n o tio n ” w hich p resu p p o ses th a t the m o d e rn w orld a n d the p a s t are sim ilar
(G o u ld 1980:29-32). H e argues th a t “ while som e processes m ay be subject to
th e prin cip le o f u n ifo rm itarian ism , o th ers m ay n o t,” th ere fo re analogic
reaso n in g sh o uld only be used to lo o k fo r co n tra sts o r anom alies betw een the
m o d ern w orld a n d the p ast (G o u ld 1980:33-35). T his “ co n tra stiv e a p p ro a c h ”
is “ an a lo g y ’s last h u rra h ;” “ analo g u es are b e tte r at in fo rm in g archaeologists
a b o u t w h at they d o know o r can expect to k n o w ” (G o u ld 1980:37). T he
ch aracteristics G o u ld (1980) a ttrib u te s to u n ifo rm itarian ist arg u m e n ts
describe su b stan tiv e u n ifo rm ita ria n ism a n d co n fig u ratio n al p ro p e rtie s, an d
G o u ld ’s an o m alies are ex plained using m eth o d o lo g ic al u n ifo rm ita ria n ism and
im m a n en t p ro p e rtie s (see also C raw fo rd 1982).
G o u ld (1980:37) ou tlines a n altern ativ e a p p ro a c h , beginning w ith the
statem en t th a t “ all scientific law s have irreducible p ro p e rtie s o f statin g [causal]
relatio n s th a t are in v ariab le in tim e an d space. These law s are derived from
observ in g regularities in tim e a n d sp ace.” G o u ld (1980:42) w ants to develop
general p ro p o sitio n s “ a b o u t h u m a n b eh a v io r th a t po sit relatio n s th a t are
in v ariab le in tim e a n d space a n d are susceptible to testing [because] th e p a st can
be perceived only in term s o f o u r p re sen t-d ay ideas a b o u t it.” G o u ld (1980:251)
explicitly states tw o requisite steps a n d im plies a th ird step to e x p lan a tio n o f the
arch aeo lo g ical reco rd . F irst, observe a n d m o d el ad ap tiv e b eh a v io r in c o n te m
p o ra ry societies. S econd, establish convincing linkages betw een p a rtic u la r
kinds o f ad a p tiv e b eh a v io r a n d distinctive “ archaeological sig n atu res” in
h u m an residues th a t identify these kinds o f behaviors; in o th e r w ords, establish
“ necessary causal re la tio n s” in B in fo rd ’s (1981b) term s, betw een processes and
effects. A n d th ird , c o m p are a ttrib u te s the fo rm a tio n o f w hich have been
observed to m o d ern static p h en o m en a created in the p a st an d , w hen sim ilar,
infer, o n th e basis o f the causal re la tio n , sim ilar processes created b o th . W hile
G o u ld (1980) has p erh ap s co rrectly ch a rac te rized e th n o g ra p h ic an alo g y as an
an aly tic p ro c ed u re fo r th e m a tc h in g o f fo rm s a n d in ferrin g sim ilar causes, his
suggested p ro c ed u re fo r explaining the archaeological reco rd is an accu rate
ren d itio n o f the actualistic m eth o d , m eth o d o lo g ical u n ifo rm itarian ism , and
the role o f im m a n en t p ro p e rtie s (W ylie 1982a).
M o re recently, G o u ld (1990:48) softens his stance on actu alism w hen he
w rites th a t (m eth o dolo g ical) u n ifo rm itarian ism o p erates a t tw o levels in

arch aeo lo g ical inference:
the level o f site-transfo rm atio n [particularly natural] processes; and the level o f real-
w orld co n strain ts o n h u m an behavior, as these are observed to o p erate in the
present an d are assum ed to have o p erated in sim ilar ways u n d er sim ilar circum
stances in the p ast . . . A t b o th o f these levels, the use o f general unifo rm itarian
assum ptions to build a bridge betw een the present and the p ast is co m p arab le to th a t
o f the o th er historical sciences.
But, G o u ld (1990:50) is still h esitan t to accept actu alism because w hile “ general
u n ifo rm ita ria n assu m p tio n s play an essential role in the stepw ise process o f
arch aeo log ical inference, they do n o t necessarily (o r even usually) provide
co m p lete ex p la n a tio n s.” As we have seen, ad v o cates o f the actualistic
a p p ro a c h w ould resp o n d by suggesting th a t an incom plete e x p lan a tio n results
from insufficient in fo rm a tio n on relevant m o d ern processes (e.g.. C h a rlto n
1981; L ange 1980), o r in ad eq u a tely o r in accu rately w ritten n a tu ra l laws.
Som e arch aeo lo g ists still call u p o n b o u n d a ry co n d itio n s a n d historical
relatedness to en h a n ce the stren g th o f an alogical arg u m e n ts (e.g.. C h a rlto n
1981). O th ers tak e a different ap p ro a c h . F o r instance, faced w ith the p o ten tial
th a t th e a ssu m p tio n th a t n a tu ra l law s are tem p o ra lly in v aria n t m ight be false,
Bailey (1983:3) suggests stu d y o f m ultiple sets o f “ in d ep en d en tly verifiable
d a ta , [all o f w hich are arch aeologically visible] ra th e r th a n to explain one set o f
[archaeological] d a ta in term s o f a n o th e r set w hich is archaeologically invisible
an d can only be derived by e x tra p o la tio n from a n o n -a rch a eo lo g ic al c o n te x t.”
T his so u n d s sim ilar to B in fo rd ’s (1967) suggested m o n ito rin g o f c o n c o m ita n t
v aria tio n o f a ttrib u te s in an a tte m p t to ensure causal re la tio n s are being utilized
in the analogy.
M u rra y a n d W a lk e r (1988) ad v o c ate a re fu tatio n strateg y w hen using
analogies. T hey arg u e th a t o th ers use a co n firm atio n ist, “ v erificationist,” o r
“ self-fulfilling p ro p h e c y ” strateg y (M u rra y an d W alk er 1988:266-267). A
re fu tatio n ist strateg y atte m p ts to show an analogically based conclusion is false
“ such th a t in ab ility to refute them helps pro v id e fu rth e r ju stific atio n for their
ac cep tan c e.” Such a strateg y w ould, they argue, lead to significant b re a k
th ro u g h s in research because it w ould open up " h ith e rto unim agined areas o f
p o ten tial k n o w led g e” (M u rra y a n d W alk er 1988:261, 283). T his is an im p o rt
a n t p o in t w a rra n tin g co n sid eratio n in an y analogically based arg u m e n t, b u t it
is also a n o th e r c h a ra c te riz a tio n o f good scientific research w ithin the co n tex t o f
the actu alistic m eth o d (G o o d m a n 1967; G o u ld 1965; Shea 1982).
M an y arch aeo lo g ists now realize th a t an alo g ical a rg u m e n ts d o n o t specify
all a n d only id entities betw een p h en o m en a. F ollow ing W ylie (1985), M u rra y
a n d W alk er (1988:262) correctly n ote th a t analogies are n o t "equivalences o f
identity; if they w ere so, the w o rk ed an alogy w ould be su p erflu o u s.” H ow ever,
th eir m isu n d erstan d in g o f the d istin c tio n betw een su b stan tiv e a n d m e th o d o lo
gical u n ifo rm itarian ism (F ig u re 3.1) lead M u rra y a n d W alker (1988:279) to
argue n o n u n ifo rm ita ria n su b stan tiv e theories m ay be necessary. In c o n tra st,
H o d d e r (1982:14) deals w ith his related su b stan tiv e co n cern th a t “ if we
in terp re t the p ast by an alo g y to the presen t, we can never find o u t a b o u t form s
o f society a n d cu ltu re w hich d o n o t exist to d a y ” by suggesting we m u st build
stro n g er an alo g ic al/a ctu alistic arg u m en ts.
H o d d e r (1982:14) states “ because sim ilarities in som e aspects d o n o t
necessarily, certain ly o r logically im ply sim ilarities in o th ers, we can never
prove [analogically-based] in te rp re ta tio n s .” H o d d e r (1982:16) distinguishes
fo r m a l analogies fro m relational analogies, n o tin g th a t the fo rm er pro d u ce
co n clusions based on sim ple sim ilarities betw een tw o objects, one o f w hich is
b etter o r m o re fully k n o w n o r u n d e rsto o d th a n the oth er. H ere, the conclusion
tak es th e fo rm th a t because the tw o objects sh are som e p ro p e rtie s visible o r
k n o w n fo r b o th , they also share o th e r p ro p e rtie s only k n o w n o r visible fo r one.
“ Such analogies are w e ak ,” H o d d e r (1982:16) argues, because “ the observed
a sso ciatio n o f [shared] ch aracteristics o f the objects o r situ atio n s m ay be
fo rtu ito u s o r accid en tal.” F o r exam ple, because one object is o f bo n e an d is
also a frag m en t o f a h u m eru s does n o t necessarily m ean the next bo n e object
e n c o u n tered will also be a hum erus; to arg u e th a t second object is also a
h u m eru s w ould co n stitu te a fo rm a l analogy. In c o n tra st to form al analogies,
th en , H o d d e r (1982) suggests arch aeo lo g ists should use relatio n al analogies
w herein associated a ttrib u te s are in te rd e p e n d e n t o r causally related. F o r
exam ple, to determ ine th e fu n ctio n o f a stone tool, archaeologists once sim ply
ex am in ed th e sh ape o f artifacts. B ut because shape m a y not be directly related
to to o l fu n ctio n (a screw driver sh ap e den o tes a screw driver fu n ctio n u n til th a t
screw driver is used to pry open a can o f p a in t, o r a soup ladel w ould n o t w ork
well fo r op ening a can o f p ain t), arch aeo lo g ists h ave tu rn e d to form al a ttrib u te s
o f artifa cts in terd e p en d en t w ith a n d causally related to to o l fu n ctio n , such as
use w ear (e.g., S alm on 1981).
H o d d e r (1 9 8 2:16 ,18,19) suggests stro n g fo rm al analogies m ay be b u ilt by (a)
n o tin g th a t the m o re sim ilarities tw o p h en o m en a share, “ the m ore likely are
o th er sim ilarities to be expected;” (b) using h o m o lo g o u s p h en o m en a (the direct
h isto rical a p p ro a c h ) in building an alogical arg u m en ts; (c) d o cu m en tin g m u lti
ple cases across m an y different instances w here relev an t a ttrib u te s an d
processes are associated; an d (d) the an aly st co u ld lim it th eir conclusions to low
levels, av o id in g b ro a d o r general sim ilarities. R elatio n al analogies are stro n g er
still because they explicitly involve “ som e necessary re la tio n betw een the
v ario u s aspects o f the a n a lo g y ;” th a t is, th e associated a ttrib u te s are th o u g h t to
be “ re le v an t” o r causally related to the inferred p ro p e rtie s (H o d d e r 1982:19-
20). T o im p rove such analogies, H o d d e r (1982:20-21) suggests we co n sid er no t
only the “ relevant causal links betw een the different p a rts o f the an a lo g y ," bu t
also the “ c o n te x ts” o f the p a rts o f the analogy; th a t is, the cu ltu ra l and
arch aeo lo g ical co n tex t o f the p h en o m en a o f in terest m u st be co n sid ered as
p o ten tially causally related to the observed a n d inferred p ro p e rtie s (H o d d e r

1982:24—25; see also S alm on 1981). R elevant linkages are theoretically
in fo rm ed ju d g em en ts a b o u t w hich a ttrib u te s should be causally related.
W hile o th e r discussions o f how to build analogies can be fo u n d in the
arch aeo log ical lite ratu re , th e ones I have review ed are som e o f the m a jo r ones
to have been p u blish ed in the p a st tw o decades. W h a t these discussions indicate
is a set o f differing id eas a b o u t h ow to use the actu alistic m eth o d . M a n y o f them
also discuss to one degree o r a n o th e r the p ro p e r use o f eth n o arch ae o lo g ic al
d a ta . Such d a ta m ay serve as a d etectio n device fo r recognizing significant
p a tte rn s o r relatio n s o f m ateria ls in the arch aeo lo g ical record n o t otherw ise
im m ediately a p p a re n t (B inford 1981b; G o u ld 1980, 1990; K ra m e r 1979). These
d a ta m ay also serve as a device by w hich significant, o r causal, relations
betw een m aterials in th e archaeological reco rd an d the processes (h u m a n and
n a tu ra l) w hich resulted in th a t record can be inferred (B inford 1981b; C h a rlto n
1981; G o u ld 1980, 1990; K ra m e r 1979; Stiles 1977; T rin g h am 1978). F inally,
these d a ta m ay serve as a source fro m w hich general e x p la n a to ry m odels m ay
be derived, the m odels sub seq u en tly being used to explain the archaeological
reco rd (B inford 1968; C h a rlto n 1981; G o u ld 1980, 1990; G o u ld a n d W atso n
1982; L ange 1980; Schiffer 1978; Stiles 1977; T rin g h am 1978). T he first use can
h ard ly be faulted. T h e second a n d th ird uses are sim ply statem en ts o f how the
actu alistic m eth o d is p ro p erly em ployed.
A rch aeo lo g ists have long acknow ledged the actu alistic m eth o d as the m eans
to ex plain h u m a n k in d 's p ast (e.g., C h a rlto n 1981; G ra y so n 1986). M uch o f the
polem ic, while co u ch ed in the ja rg o n pecu liar to archaeology, co n tain s all the
basic elem ents o f the m ethod. T he inescapable conclusion is, how ever, a rc h a e o
logical an alo g y o ften lacks the req u ired causal relatio n s betw een processes a n d
effects (does n o t consist o f re la tio n a l analogies), an d also lacks the diagnostic
criteria som e believe are requisite to use o f the actualistic m eth o d . A s M u rra y
an d W a lk e r (1988) im ply, d iag n o stic criteria in the sense o f B inford (1981 b) are
n o t req uired ; if such were av ailab le o u r arg u m e n ts w ould n o t be analogical but
ra th e r id en tification al. W ith o u t the estab lish m en t o f causal relations betw een
processes a n d effects (relevant linkages), a ttrib u te s o f p h en o m en a used as
sig n atu re criteria are inductively derived em pirical gen eralizatio n s (B inford
1977, 1981b). A nalogical ex p lan a tio n is only as so u n d as the em pirical
g eneralizatio n , a n d d a ta in d icatin g th a t the em pirical g eneralization is in acc u r
ate are sure to be fo u n d (W ilier a n d W ilier 1973). If causal relatio n s betw een
processes a n d effects are kn o w n o r suspected, we m ay consider inferences m ore
highly p ro b a b le (B inford 1981 b:26-27), a n d exceptions should be m uch h ard er
to find (W ilier a n d W ilier 1973). It is in the la tte r w here F ish e r’s (1981) analysis
o f cro co d ilia n ta p h o n o m y finds its stren g th , w hich brings us to a c o n sid eratio n
o f actu alism in ta p h o n o m y .
T aphonom y and actualism
H ill (1978:88) suggests th a t actu alistic ta p h o n o m y (w h at he labels neotapho-

nom y) “ involves relevant e x p e rim en tatio n o r o b serv atio n s on the co n d itio n o f
m o d ern v erte b rate rem ains in v ario u s closely defined en v iro n m e n ts” a n d is
“ designed to te s t” h y p o th etical assertio n s a b o u t causes resulting in the effects
observed in a fossil assem blage. Hill (1978:89) suggests th a t while actualistic
o b serv atio n s m u st be m ad e to a c c o u n t fo r differences betw een fossil assem
blages a n d living co m m u nities, it is p re m a tu re to a ttrib u te observed effects to o
rigidly to p a rtic u la r processes because the details are often p o o rly know n;
specifically, cau sal relatio n s betw een processes a n d effects are u n k n o w n , and
d iag n o stic criteria are u n av ailab le. D ifferences betw een a fossil assem blage and
the living co m m u n ity fro m w hich it derived can be subsum ed u n d e r tw o basic
categories: ( 1) differences in state (chem ical, asso ciatio n al, lo catio n al, etc.) o f
fossil a n d living o rganics, a n d ( 2 ) differences in how ecological p aram eters are
reflected by the fossil assem blage a n d the once living com m unity. K now ledge
o f the fo rm e r is requ isite to d iscernm ent o f the latter, a n d know ledge o f the
fo rm er is o b ta in e d by actu alistic research leading to the e stab lish m en t o f causal
relatio n s a n d the defin itio n o f d iagnostic criteria (Hill 1978:98).
S h ip m an (1981 b: 12) w rites th a t if the “ first law o f ta p h o n o m y ” is u n ifo rm i
taria n ism o f cause a n d effect, th en the “ second law o f ta p h o n o m y ” m u st be th a t
only u n iq u e a n d “ d istin ctiv e” aspects o f an effect o r result can be considered
“ d iag n o stic” o f a p a rtic u la r cause o r process. She correctly argues th a t “ the
o ccurrence o f a p a st event can be deduced only by d e m o n stra tin g th a t its effects
differed from tho se o f o th e r sim ilar ev en ts” (S hipm an 198 lb : 12). S hip m an has
here set u p th e estab lish m en t o f b ico n d itio n a l statem en ts as the qualification
fo r in ferrin g p a st events. T h a t is, h er “ d iag n o stic” sig n atu re criteria are o f the
“ if X, a n d only if X, th en Y ” type, w here X is the event o f interest a n d Y is the
distinctive sig n atu re o f events o f kind X. As M u rra y a n d W alker (1988) and
W ylie (e.g., 1982a, 1982b, 1985) m ak e clear, arg u m e n ts o f this so rt are n o t
analogies, they are identifications. B ut as we have seen, we seem to lack
d iag n o stic sig n atu re criteria in S h ip m a n 's b ico n d itio n a l sense. I suspect th a t
S h ip m an w ould agree th a t im m a n en t p ro p e rtie s are, u ltim ately, w h a t we are
after, a n d th a t o u r ta p h o n o m ic conclusions are, in fact, an alo g y based.
K lein a n d C ru z-U rib e (1984) im ply ta p h o n o m ic analyses involve strip p in g
aw ay biases in th e fossil reco rd , a n d w hile valuable, can be co n tra ste d w ith
w h a t they label th e com parative approach. T he la tte r involves co n tro lled
co m p ariso n s o f different p re h isto ric assem blages, a n d w hen tw o assem blages
differ in term s o f one o r m o re a ttrib u te s o f th eir respective fau n al rem ains, a
difference betw een the tw o assem blages in one o r m o re n o n -fau n al a ttrib u te s
such as asso ciated artifa cts o r sedim entological co n tex ts serves as a signal o f
th e source o f p o te n tia l differences in the ta p h o n o m ic histories. T he c o m p a ra
tive a p p ro a c h is a v aluable one fo r p o in tin g o u t w ays to en hance o u r
u n d ersta n d in g , b u t in m y view it rests so u n d ly on the ta p h o n o m ic a p p ro a c h

w ith w hich it is c o n tra ste d by K lein an d C ru z-U rib e (1984).
In o ne o f the few trea tm e n ts o f the to p ic o f w hich I am aw are, K lein and
C ru z-U rib e (1984:9) list several “ lim ita tio n s” to the actualistic m eth o d on
w hich the ta p h o n o m ic a p p ro a c h is fo u n d ed (see also C ru z-U rib e 1991). T he
first is th a t o b serv atio n a l co n d itio n s m ay affect results; fo r exam ple, carnivores
in zo o s m ay process b ones differently th a n carn iv o res in the wild. Second, som e
biological b o n e-m o d ify ing agents, including h u m an s, n o lo n g er exist in
co n tex ts sim ilar to p reh isto ric ones. T h ird , biological b o n e-m odifying agents
th a t are presently extinct c a n n o t be observed. T hese three " lim ita tio n s”
co n cern su b stan tiv e u n ifo rm itarian ism a n d co n fig u ratio n al p ro p e rtie s, a n d
n o t th e m eth o d o lo g ical u n ifo rm itarian ism a n d im m a n en t p ro p e rtie s o f a c tu a
listic research. T h u s, listing these lim itatio n s helps us focus o n the la tte r ra th e r
th a n the form er. T he fact th a t m any m o d ern studies are tem p o rally brief, th at
is, sp an only m o n th s o r a few years, w hile m o st p re h isto ric fa u n al assem blages
rep resen t m uch lo nger tim e spans o f ac cu m u latio n is an im p o rta n t p o ten tial
lim itatio n , b u t seem s to be a m a tte r o f the scale o f o b serv atio n like th a t alluded
to earlier co n cern ing the fo rm a tio n o f the G ra n d C an y o n . F inally, p reh isto ric
assem blages h av e b o th p o st-d ep o sitio n al an d p o st-b u rial histories w hereas
m o d ern sets o f an im al rem ains typically do n o t have either, p artic u la rly the
latter. Som e m ig h t d isp u te the lack o f a p o st-d ep o sitio n al h isto ry as m any
m o d ern bon e assem blages have been a n d are being studied afte r th eir dep o si
tion (e.g., B oaz 1982; O lsen a n d S hipm an 1988). T he m o st serious lim itation
involves stu d y in g p o st-b u rial processes; because they are bu ried th e bones
ca n n o t be d irectly observed as they u n d erg o diagenetic processes. H ere, o f
course, is w here experim ental research co u ld fill a m ajo r gap in actualistic
research.
G iffo rd (1981:388-389) p o in ts o u t th a t actualistic o b serv atio n s will indicate
the range o f v aria tio n in fossil assem blages p ro d u ced by sim ilar agents u n d er
sim ilar co n d itio n s, w hich in tu rn will suggest the m ost distinctive criteria to be
used in the m atch in g process. N o te th a t this is n o t the sam e as the b ico n d itio n al
d iag n o stics o f S h ip m a n ’s “ if, a n d only i f ' sort. C o m p a ra tiv e stu d y o f m any
fossil assem blages will. G ifford (1981) suggests, aid in the search for distinctive
criteria by indirectly co n tro llin g fo r sto ch astic v aria tio n ; sp u rio u s c o rrelatio n s
betw een causes a n d effects can be m ore readily detected a n d discarded. In o rd e r
to establish th a t visible effects are d iag n o stic o f p a rtic u la r causes, G ifford
(1981:393—394) suggests the follow ing pro ced u re: (1) observe d ynam ic in te ra c
tions betw een p o stm o rtem o rg an ic rem ains a n d processes th a t o p erate o n them
at the scale o f th e ind iv idual skeletal elem ent; ( 2 ) establish the n a tu re o f
cause-effect relatio ns; (3) elim inate possible causes u ntil each criterio n has only
one possible cause, o r each crite rio n is diag n o stic o f a p a rtic u la r tap h o n o m ic
process; (4) estab lish the expected ran g e o f v a ria tio n in a diag n o stic criterion;
(5) test possible cause-effect re la tio n s a n d suspected diag n o stic crite ria w ith
fu rth e r o b serv atio n s by chan g in g th e scale o f in v estig atio n to the level o f fossil
assem blage a n d “ p red ictin g the stru c tu re o f assem blages p ro d u c ed by the
actio n o f specified processes.” G ifford (1981:394) argues th a t “ only if p red ic
tions pass this test o f actu alistic ev a lu a tio n sh o u ld they be em ployed in analysis
o f fossil m a te ria ls.” T h e testing p ro c ed u re is im plicitly perceived as p a r t o f
establish in g causal relatio ns an d echoes B in fo rd ’s (1967) re co m m en d atio n s for
testing e th n o g ra p h ic an alo g s by exam ining co n c o m ita n t v aria tio n o f a ttrib u te s
suspected to be causally related in actualistic settings. B oth display a p p a re n t
co n fu sio n betw een c o rrelatio n a n d cau satio n . It is p erh ap s because exhaustive
testing is literally im possible th a t G ifford (1981) suggests elim in atio n o f o th er
p ossible causes fo r p a rtic u la r effects as an altern ativ e to testing. T his is sim ilar
to B in fo rd ’s (1981b) reco m m ended p ro c ed u re o f arg u m en t by elim ination. In
b o th cases, the re treat to a p o sitio n o f elim inating som e causes suggests “ if, and
only i f ' d iagno stic criteria will be difficult to establish. But th a t is n o t a
d am n in g o r fatal criticism o f the actu alistic m ethod.
G iffo rd (1981:394) believes th a t “ the gravest p ro b lem in actualistic research
is assum in g th a t a given process is a necessary a n d sufficient cause o f an
observ ab le a ttrib u te w hen no such re latio n has actu ally been estab lish ed .” A
sufficient cause is o ne th a t is cap ab le o f creatin g a p a rtic u la r result, b u t it is no t
th e only o ne cap ab le o f p ro d u c in g th a t result. F o r exam ple, a house m ay b u rn
d ow n fo r several reasons: a careless sm o k er w ho falls asleep in the house, an
electrical sh o rt circuit, arso n , o r a lightning strike. All are sufficient to cause the
h ouse to b u rn d o w n , b u t n o n e o f these is necessary to th e h o u se b u rn in g dow n
as a sp ark from the fireplace in the house (a n o th e r sufficient cause) m ight
p ro d u ce the sam e result. It is necessary, in o u r b u rn in g h ouse exam ple, th a t the
sm o k er n o t aw ak en in tim e to p u t the blaze o u t before it is o u t o f co n tro l, and
th a t th e ashes from w h atev er he is sm oking c o n ta c t som e flam m able m aterial.
Sim ilarly, it is necessary fo r the arso n ist to light flam m able m aterial, an d fo r the
blaze n o t to be discovered p rio r to its becom ing u n co n tro llab le fo r a b urned-
d ow n h o use to result (see S alm on 1984 for fu rth e r discussion o f necessary and
sufficient causes).
K lein a n d C ru z -U rib e (1984:9) suggest th a t “ ca u satio n m ay be observed; it
does n o t have to be in ferred ” w hen w o rk in g in actu alistic contexts. W hile
superficially true, som e p h ilo so p h ers w ould disagree (e.g., S alm on 1984).
R eq u irem en ts fo r estab lish in g causal relatio n s include (1) the cause a n d the
effect m u st be co n tig u o u s in tim e a n d space, b u t sole use o f this re q u irem en t can
result in the " p o st hoc fallacy” (S alm on 1963:74) th a t te m p o ra l-sp a tia l
coincidence d en o tes a causal re latio n ; a n d ( 2 ) there is som e co n n ectio n betw een
a cause a n d an effect such th a t the tw o are regularly coincident, an d this m ay
involve th e specification o f necessary an d sufficient co n d itio n s (S alm on
1984:211). E stab lish in g necessary a n d sufficient causal re la tio n s is sim ilarly a
difficult m a tte r at best. In sh o rt, actualistic research alo n g the lines p ro p o se d by
G iffo rd (1981) a n d exem plified by F ish er (1981) is used to propose such
re la tio n s a n d linkages. “ C au sal statem en ts are h y p o th eses” (S alm on 1963:88),

an d th u s are cap ab le o f ev a lu a tio n a n d re fu tatio n . N u m ero u s ta p h o n o m ic
studies have, in fact, n o t established o r even explicitly p ro p o se d causal
linkages. F o r exam ple, M ellet (1974) sim ply co m p ared m o d ern a n d fossil
bones a n d inferred a sim ilar ta p h o n o m ic process on the basis o f form al
sim ilarities o f the bones. D o d so n a n d W exlar (1979), S hipm an a n d W alker
(1980). H offm an (1988), an d K u sm er (1990) go fa rth e r to w ard s describing the
effects o f certain ta p h o n o m ic processes including th a t discussed by M ellet
(1974), b u t they offer only b rie f discussions o f w hy the causes they m o n ito r
p ro d u c e th e effects they describe (see also A ndrew s 1990). V oorhies (1969) an d
K o rth (1979) also sum m arize actu alistic o b serv atio n s, b u t suggest causal
reason s w hy som e o f the effects they observe result fro m vario u s processes;
th eir causal linkages include intrinsic p ro p e rtie s o f the bones such as stru ctu ra l
density, size, a n d shape, a n d how th o se p ro p e rtie s in terre late w ith the
ta p h o n o m ic processes th ey m o n ito r. W hile I re tu rn to causal linkages below , it
is im p o rta n t at this ju n c tio n to n ote th a t as long as em pirical generalizations
(inductively d o cu m en te d instances o f processes a n d effects w ith o u t explicit
p ro p o se d causal linkages) c o n tin u e to serve as in terp retiv e bases, d eb a te over
the precise ta p h o n o m ic m eaning o f v ario u s a ttrib u te s o f bo n e assem blages will
con tin u e.
T h e fossil reco rd is an effect o r result o f ta p h o n o m ic causes o r processes. T o
explain the effects a n d w rite ta p h o n o m ic histories, tap h o n o m ists a tte m p t to
d eterm in e th e causes. T h e m eth o d o f d e te rm in a tio n involves th e actualistic
m eth od . A n tec ed an t statem en ts (B inford 1981b; G ifford 1981; Hill 1978;
S h ip m an 1981b) to this effect are th erefo re co rrec t in in ten t. E quifinality, the
p ro p e rty o f allow ing o r having the sam e effect o r result fro m different events, is
p o o rly co n tro lled in m an y cases, how ever, because o f a lack o f causal linkages
betw een the a ttrib u te s o f interest. A s recent reviews (e.g., B onnichsen a n d Sorg
1989; H u d so n 1993) show , w hile ac tu alistic d a ta are being collected a t ever
in creasing rates, ta p h o n o m y still regularly lacks such linkages. T he result is
th a t m an y analyses em ploy a set o f criteria w hich are suspected to be diagnostic
o f p a rtic u la r ta p h o n o m ic processes b u t w hich are in fact em pirical g en eraliza
tions. In te rp re ta tio n s based o n these em pirical gen eralizatio n s are published,
only to have a “ c a u tio n a ry ta le " published a few years later th a t q uestions the
em ployed crite ria's distinctiveness based on o th e r actualistic d a ta th a t also lack
any explicit statem en t a b o u t causal relations. A ctualistic research ca n n o t
sim ply reco rd in p u t (kinds a n d n u m b ers o f bones) to a n d o u tp u t (m odifications
to kin d s a n d n u m b ers o f bones in p u t) o f ta p h o n o m ic processes. It m u st also be
designed to in v estigate ca u satio n . It m u st seek answ ers to why, given p a rtic u la r
in p u ts a n d processes, a p a rtic u la r o u tp u t results.
B ecause estab lish in g causal relatio n s a n d linkages is difficult (S alm on 1984),
they are o ften sim ply p ro p o se d as “ w o rk in g h y p o th eses” (F ish er 1981).
R elatio n al p ro p o sitio n s o f causes a n d effects a b stra c te d fro m actualistic
o b serv atio n s can th en be p h ra sed to m ak e purely logical m odels (B inford

1981b), m uch as the exam ple by S tiner (1990a, 1990b, 1991b) review ed earlier
illustrates.
A b stractio n is a m a tte r o f establishing an isom orphism betw een theoretical
nonobservables and em pirical observables. T he m ore sim ilar the m odel an d the
em pirical case are in all respects covered by th e theoretical statem ent, the b etter the
theory will explain o r predict. T he p o in t o f indifference is reached w hen the erro r in
prediction is w ithin the lim its o f m easurem ent error.
(W ilier and W ilier 1973:26).
E x p lan a to ry m odels derived by a b stra c tio n are logically co rrect, b u t they are
n o t ab so lu te n o r are they b eyond a lte ra tio n o r disp lacem en t by o th e r m odels
w ith different stru ctu re s th at pro v id e m ore sufficient o r efficient ex p lan atio n s.
T h u s few er c a u tio n a ry tales sh o u ld result. W hile discussed here as possible
altern ativ es, arg u m e n t by elim in atio n a n d use o f th eo retical causal relatio n s o r
logical m odel b u ild in g o p erate m o st effectively w hen used to g eth er, as in the
exam ples review ed above.
G ifford (1981) im plies suspected causal linkages can com e fro m a tte n d a n t
bodies o f th eo ry for each stage in a ta p h o n o m ic h isto ry (F igure 3.2). G ifford-
G o n zalez (1989b:46) em phasizes th e necessity o f causal linkages in analogically
based arg u m e n ts w hen she notes “ the goal o f actualistic research should be to
d istin g u ish ca u sal/fu n ctio n al re la tio n s” betw een ta p h o n o m ic processes o r
causes, a n d th eir a tte n d a n t effects (F ig u re 3.3). B ut she also argues th a t we m ust
begin to ex p an d o u r an alogical h o rizo n s fro m one line o f evidence a n d analogy
to em b race “ clu sters” o f relatio n al analogies; this is sim ply a re sta te m en t o f the
n o tio n th a t m ultiple lines o f evidence are b e tte r th a n single lines (see also
G iffo rd -G o n zalez 1991). T his is especially so due to the generally h y p o th etical
n a tu re o f causal relatio n s betw een ta p h o n o m ic processes a n d th eir effects. If all
lines o f evidence a n d analogically fo u n d ed results p o in t to w a rd the sam e
co nclu sion reg ard in g th e ta p h o n o m ic h isto ry o f a bo n e assem blage, we can
h ave g re ater confidence in the tru th fu ln ess o f th a t p a rtic u la r conclusion th a n in
one fo u n d ed on a single line o f evidence o r single analog.
Analogy
N eo tap h o n o m ic analogs are based on direct o bservation o f cause and effect,

throu gh actualistic o r experim ental d ata.
(E. Jo hnso n 1985:159)
T he actu alistic m eth o d involves arg u m en t by analogy, so it is a p p ro p ria te to

review th e stru ctu re o f stro n g an alogical arg u m en ts. In arch aeo lo g y such
things as m iddle-range th eo ry build in g (B inford 1977, 1981b) an d the identifi
catio n o f site fo rm a tio n processes (e.g., Schiffer 1987) include m a jo r aspects o f
actu alistic research a n d th u s often tak e the form o f analogical arg u m en ts. Such
arg u m en ts have been described in detail by a p h ilo so p h er w ith interests in
arch aeo log ical science.
relevan t
ELEMENT T R AN S F O R M A T I O N SEQOENCE
theory
biologic ANATOMI CAL CONTEXT
DEATH T R A N S F O R M S
J-2
S UR F AC E CONTEXT
biostrati no mic
BURIAL TRANS FORMS
S E D I M E N T A R Y CONTEXT
DI AGENETI C T R A N S F O R M S
diagenetic
FOSSI L S E D I M E N T A R Y CONTEXT
EXPOSURE/SAMPLING TRANSFORMS
recovery
S A M P L E A S S E M B L A G E CONTEXT
D E S C R I PT I ON ( S Y S T E M A T I C S / T Y P O L O G Y )
anal ytic T RANSFORMS
OBSERVATI ONAL DATA
Figure 3.2. Schem atic rep resen tatio n o f the tran sfo rm atio n o f an anim al from
being a living organism to being a fossil show ing w here p articu lar bodies o f
theory are relevant, an d general categories o f tran sfo rm s an d contexts (after
G ifford 1981:387, Figure 8.1; courtesy o f the a u th o r an d A cadem ic Press).
Figure 3.3. A m odel o f (relational) analogical reasoning. Shaded area is w here

o bservations are m ade in the present o r w here actualistic research takes place
(after G ifford-G onzalez 1989b:44, Figure 1; courtesy o f the a u th o r an d The
C enter for the Study o f the F irst A m ericans).
W ylie (1982a, 1982b, 1985, 1988, 1989a, 1989b) explores the stru c tu re o f
an alo g ical arg u m en ts, a n d no tes first such a rg u m e n ts are inductive, a n d they
are am pliative; they resu lt in conclusions th a t generally co n ta in m ore in fo r
m a tio n th a n the initial d a ta a n d prem ises. T his som etim es leads to scepticism
th a t an y reliable know ledge o f the p ast will be derived, b u t th a t scepticism is
“ m isp laced " (G iffo rd -G o n zalez 1989b) a n d m ain tain s a m isconception o f
science. W ylie (1982b:42) ch aracterizes a realist view o f science as one th a t
acknow ledges a n d em phasizes the am pliative aspect o f scientific know ledge
an d seeks to im p rove it. A nalogical a rg u m e n ts are n o t in h eren tly faulty, bu t
they can be catego rized as involving eith er w eakly o r stro n g ly a rg u ed fo rm al o r
relatio n al analogies (e.g., H o d d e r 1982, above). W eakly arg u ed analogies are
th o se w hich fail to “ specify the (usually lim ited) p o in ts on w hich an analogy
holds (i.e., to specify the positive, negative a n d n eu tra l aspects o f an analogical
co m p ariso n o f item s o r contexts) a n d an in d iscrim in ate ca rry in g over o f all
featu res o f the [actualistic] source to the [prehistoric] su b ject" (W ylie
1982b:43). F o rm a l analogies are th o se w hich only specify p o in ts o f sim ilarity
(and less often, p o in ts o f dissim ilarity) betw een the m o d ern source an d the
p reh isto ric subject p h en o m en a , a n d co nclude sim ilar processes created b o th
p h en o m en a w ith o u t co n sid eratio n o f the possible causal a n d /o r stru c tu ra l
linkages betw een the observed p h en o m en a a n d the processes (a fter H o d d e r
1982; W ylie 1982a, 1988). “ A nalogical arg u m e n t is form ally valid; the diffi
culty is ju s t th a t th e relevant m a jo r prem ises are in secu re,” b u t it is n o t ju st
fo rm al sim ilarities o f the source a n d subject b u t ra th e r th a t the observed
(effects) p ro p e rtie s are som ehow n o n accid en tally related to, o r m o re th an
sim ply co rrelated w ith, th e in ferred (causal) p ro p e rtie s th a t serves as the basis
fo r an alo g ical arg u m e n ts (W ylie 1988:136).
A s K elley a n d H a n e n (1988:170) note, “ does saying th a t A caused B m ean

an y th in g m o re th a n saying th a t w henever A occurs B occurs?” T h a t is, “ the
co n cep t o f cause involves at least c o rre la tio n ” (K elley an d H an en 1988:249),
bu t does it involve m o re th a n this em pirically? T hey suggest being able to
predict successfully th e resu lt o f a process indicates m o re th a n a sim ple
c o rrelatio n (K elley a n d H a n en 1988:170), b u t conclude that:
relevance is the chief d esideratum , an d in o rd er to determ ine relevance, it is

necessary to explore the negative an d neu tral aspects o f analogies as well as the
positive ones. C ausal, functional, an d stru ctu ral analogies are m ore likely to be
theoretically relevant th a n “ m ere” sim ilarities.
(Kelley and H anen 1988:269)
T h e re la tio n a l p ro p e rtie s (causal a n d /o r stru c tu ra l linkages) n o t only w a rra n t

th e an alo g ical con clu sion , b u t guide in q u iry directed to w ard im p ro v in g them
(W ylie 1988:144). M u rra y a n d W a lk e r's (1988) re fu ta tio n ist strateg y is sim ply
p a rt o f b u ild in g stro n g inductive o r co n firm a tio n ist analogical arg u m en ts.
T h u s W ylie (1988, 1989a) ca n re p o rt th a t arch aeo lo g ists w o rk in g in the G re a t
P lains o f N o rth A m erica w ere able to refute the analogically based n o tio n th a t
late p re h isto ric people th ere w ere n o m ad ic horsem en a n d conclude th a t this
a d a p ta tio n ap p e are d very late in tim e w hile previous cultures involved m ore
sed en tary h o rticu ltu ralists. D etailed w o rk on the m o d ern o r source side o f the
analo g ical e q u a tio n " c a n serve n o t ju s t to w iden the ran g e o f an alo g s on which
in te rp re ta tio n ca n draw , b u t also to sh arp ly lim it them , p ro v id in g an im p o rta n t
basis fo r critical assessm ent o f th eir cred ib ility ” (W ylie 1989a: 13). O ne m an n er
o f w o rk in g on th e subject o r p re h isto ric side o f the e q u a tio n involves calling
u p o n m ultiple, in d ep en d en t lines o f evidence (W ylie 1989a: 15). Sim ply put,
w o rk in g b ack a n d fo rth o r h o rizo n ta lly betw een subject a n d source sides o f the
eq u a tio n , a n d u p a n d dow n o r vertically w ith in each side o f the e q u a tio n by
stu dy in g m ultiple attrib u te s, sh o u ld ultim ately p ro d u c e strong, analogically
based inferences. T o use W ylie’s (1989a) m e ta p h o r, such tack in g b ack an d
fo rth , an d b uildin g o f suspension cables, respectively, will sim ultaneously
reinforce an d c o n stra in o u r inferences. B uilding stro n g , relatio n al analogies,
th en, involves clearly specifying the “ n u m b e r a n d extent o f sim ilarities betw een
source a n d subject, the n u m b e r a n d diversity o f sources cited in the prem ises in
w hich kn o w n an d in ferred sim ilarities co -o ccu r as p o stu la te d for the subject,
a n d th e expansiveness o f the co nclusions relative to the prem ises” (W ylie
1985:98).
G iffo rd -G o n zalez (1991:226) argues th a t as zo o arch aeo lo g ists first and
ta p h o n o m ists second, we generally w ish to study “ the life re latio n sh ip s -
ecological, social, a n d cu ltu ral - o f p reh isto ric h o m in id species,” a n d such
stu d y is an alog ical in n atu re . She suggests th a t d u rin g the 1980s ta p h o n o m ists
becam e very a d e p t a t identifying ta p h o n o m ic agents; th a t is, we now know
w h a t kin ds o f visible traces a h u n g ry ca rn iv o re will leave o n a bo n e a n d how
th o se differ from th e visible traces p ro d u c ed on a b o n e th a t was b u tch ered by a
sto ne too l-w ieldin g ho m in id . T he relatio n al analogies used to p ro d u c e such

id en tificatio ns are stro n g a n d we tend to have m u ch confidence in the
identifications.
T o b uild confidence in the inferences b u ilt u p o n o u r identifications o f the
creato rs o f traces, G iffo rd -G o n zalez (1991:228-229) suggests use o f a hierarchy
o r "n ested system o f an aly tical ca teg o ries” described by the follow ing set o f
term s:
trace: visible a ttrib u te displayed by a bone th at has undergone a taph o n o m ic
process (e.g, a scratch o n a bone);
c a u s a l a g e n c y : the im m ediate physical causes pro d u cin g a trace; the im m ediate
interaction o f m aterials producing a trace (e.g., the g rating o f a sand grain

against a bone);
e f f e c t o r : the item o r m aterial th a t effects the m odification o f a b one (e.g., a sand
grain grating against a bone);

a c t o r : the source o f force o r energy th a t creates traces (e.g., an ungulate w ith sand
grains adhering to its h o o f tram p lin g a bone);

b e h a v i o r a l c o n t e x t : the prehistoric systemic en vironm ent in w hich the ta p h o n o
mic process to o k place (e.g., a herd o f ungulates m illing a b o u t a w aterhole);

e c o l o g i c a l c o n t e x t : the type o f ecosystem an d enviro n m en t in w hich the actors
lived (e.g., an A frican savanna).
Id en tificatio n o f traces involves d escrip tio n o f the fo rm al a ttrib u te s (size,

shape, lo catio n , o rie n ta tio n , frequency, etc.) o f th o se traces, a n d the use o f
relatio n al an alo g y to identify the causal agency th a t created them . U sing
aggregates o f different traces in c o n ju n ctio n w ith aggregates o f identifications,
we derive an alo g y -b ased identifications o f effectors an d actors. B ecause o f the
com plexities o f such m ulti-faceted reasoning a n d decreasing ce rtain ty o f the
cause-effect linkages betw een ad jacen t levels w ithin the h ierarchy, there is a
“ d ilu tio n o f c e rta in ty ” (G iffo rd -G o n zalez 1991:2 3 1) as we m ove from d escrip
tio n o f the traces th ro u g h the id en tificatio n o f ac to rs to w ard s in ferrin g the
b eh av io ral con tex t. W hile som e o f this decrease in confidence can be overcom e
as we p erfo rm m o re actu alistic research focused on the kinds o f aggregates o f
traces created in different b eh a v io ral a n d ecological co n tex ts, we m ay never
h ave stro n g confidence in inferences o f such co n tex ts based o n p re h isto ric
m aterials given th e com plex n a tu re o f the re aso n in g involved. A s one p ro
gresses fro m casual agency to co n tex t, the n u m b e r o f necessary relatio n al
analogies increases, a n d b o tan ica l, sedim entological, artifa c tu a l, site stru c
tu ral, a n d o th e r kinds o f d a ta com e increasingly into play. T his is an im p o rta n t
to p ic th a t will re a p p e a r in la te r ch a p te rs, a n d to w hich I re tu rn fo r detailed
co n sid eratio n in th e final ch ap ter.
Summary
It seems th a t archaeology m ust rely o n analogy in the fo rm u latio n o f hypotheses

and theories to a greater degree th a n m ust o th e r social sciences.
(J. H. Kelley and M. P. H an en 1988:378)
A ctualistic research is presently perceived as the basis fo r m o st ta p h o n o m ic an d

archaeolo gical analysis a n d in te rp re ta tio n . D a ta derived from actualistic
research are, how ever, co m m o n ly used as a source o f em pirical generalizations
o r fo rm al analog ies ra th e r th a n to build re la tio n a l analogies a n d p o stu late
diag n o stic criteria. W hile the d eriv a tio n o f d iag n o stic sig n atu re p a tte rn s o r
criteria o f the b ico n d itio n a l “ if, a n d only i f ’ so rt are som etim es suggested to be
the desired end p ro d u c t o f actu alistic research, because o f the difficulties in
estab lish in g such relatio n s a n d criteria, the result is usually a series o f em pirical
gen eralization s o r fo rm a l analogies because the relev ant linkages (necessary
a n d sufficient c a u sa l/stru c tu ra l relations) are n o t alw ays clearly specified, even
in h y p o th etical form . F o rtu n a te ly , this situ a tio n seem s to be im p ro v in g as
tap h o n o m ists gain a clearer u n d ersta n d in g o f the actu alistic m eth o d o f
learning a b o u t the past.
In follow ing ch a p te rs I describe vario u s kinds o f d a ta derived from a c tu a lis
tic research a n d an aly tic techniques fo r co n ten d in g w ith th o se kinds o f d a ta in
p reh isto ric contex ts. T he explicit re aso n in g beh in d the relevance o f these kinds
o f d a ta a n d an aly tic techn iques to ta p h o n o m ic analysis is epitom ized in F igure
3.3. T ypically, the desire o f zo o a rch a eo lo g ists p erfo rm in g ta p h o n o m ic
analyses is to identify the agents, im plem ents, a n d processes w hich resulted in
the a ttrib u te s displayed by a b o n e assem blage, w here agents a n d processes,
loosely defined, are the sources o f the forces th a t result in m odifications to
carcasses an d b ones, a n d im plem ents are the objects th a t serve as the
m ech an ism o f force tran sm issio n to carcasses a n d bones. T h e d istin ctio n o f
agents, processes, a n d im plem ents is fo r h euristic p u rp o ses a n d m ay be
u n n ecessary to som e form s o f analysis. But now it is tim e to tu rn to the
m aterials o f ta p h o n o m ic analysis p rio r to c o n sid e ra tio n o f its techniques.
4
STRUCTURE AND QUANTIFICATION

OF V E R T E B R A T E S K E L E T O N S
T he fossil record is com posed alm ost entirely o f the preserved h ard p arts o f
organism s.
(D. K. M einke 1979:122)
Introduction
I presum e th e re ad er possesses som e basic know ledge o f archaeological,

p aleo n to lo g ical, biological, ecological, a n d an a to m ica l principles a n d c o n
cepts, b u t it is im p o rta n t to review som e basics o f v erte b rate skeletal an ato m y .
In this c h a p te r I p re sen t a general discussion o f w h a t a p p e a r to be ta p h o n o m i
cally significant p ro p e rtie s o f bones, teeth, a n d related m aterials. B o th m icro
scopic a n d m acro sco p ic features are review ed, as well as the principles o f
o n to g en y a n d allom etry. W hile only superficially covered here, these topics all
w a rra n t careful c o n sid e ra tio n in m an y ta p h o n o m ic analyses, a n d the serious
stu d e n t will find th e references cited a good place to s ta rt learning m ore a b o u t
them .
I also co n sid er som e basic issues re g ard in g the q u an tifica tio n o f v erte b rate
rem ains. In this c h a p te r I review the q u a n tita tiv e u n its co m m only used in
v erte b rate ta p h o n o m y ; ad d itio n a l details are pro v id ed in o th e r ch ap ters. As
w ith th e stru c tu re o f v erte b rate skeletons a n d skeletal tissues, an extensive
lite ratu re con cern in g the q u an tifica tio n o f v erte b rate rem ains exists, a n d the
in terested re ad er is en co u rag ed to inspect th a t literatu re.
Ontogeny and allometry

O n to g en y a n d allo m etry are tw o in terre lated p h en o m en a th a t often play an
influential role in co n tro llin g the kind o f skeletal tissue u p o n w hich tap h o n o m ic
processes m ig h t o p e ra te a n d u p o n the possible effects o f those processes.
O ntogeny involves the g ro w th a n d d evelopm ent o f an org an ism fro m its
c o n cep tio n to its d e a th (F ig u re 4.1). E m b ry o n ic developm ent, fetal dev elo p
m ent, a n d p o stn a ta l d evelopm ent a n d g row th are all included in v erteb rate
o n to g en y (Shea 19 8 8 b :4 0 1). A llom etry is the study o f relatio n s betw een the
sizes o f v ario u s b o d y p arts, a n d it m ay involve the scaling o f individual body
co m p o n en ts relative to body size (A lexander 1985; Shea 1988a:20). F o r
exam ple, large m am m als have n o t only absolutely larger skeletons th a n sm all
m am m als, b u t due to the larger b o d y m ass (w eight) o f the fo rm er they also have
70
S tructure an d quantification o f vertebrate skeletons 71
F igure 4.1. Schem atic illu stratio n o f ossification an d grow th o f en d o ch o n d ral

long bone (tibia) o f a m am m al. W hite, cartilage; heavy stipling, en d o ch o n d ral
bone; black, perich o n d ral bone; light stipling, m edulary cavity. A , cartilaginous
stage. B, C, initial ossification o f diaphysis. D , grow th o f diaphysis and
ossification o f epiphyses. E, n ear fusion o f epiphyses an d diaphysis, an d initial
form atio n o f m edullary cavity. F , bone grow th (ontogeny) com plete.
R eproduced by perm ission from : R om er, A. S. an d P arsons, T. S. The vertebrate
body, Figure 108. Philadelphia: W. B. S aunders C om pany. C opyright 1977 W. B.
S aunders C om pany.
relatively larger skeletons. T h a t is, to build a m ouse the size o f a n elep h an t

w ould be possible b u t w ould involve m ak in g the m ouse 100 tim es longer, 100
tim es w ider, 100 tim es h igher a n d th u s increasing its volum e by 1 m illion, bu t
the cross-section al areas o f th e b ones w ould be increased by only 10,000
(A lexan d er 1985:26; see also L aB arb era 1989; Shea 1992; S m ith 1984).
T h ere are a t least th ree kinds o f allom etric relations. “ A n o rg a n o r stru ctu re
ca n grow m o re quickly th a n the b o d y as a w hole (positive allom etry), m ore
slow ly (negative allo m etry), o r w ith the sam e speed (iso m etry )” (R ensch
1960:133). O ne o b viou s negative allo m etric re la tio n is the d ecreasing size o f the
h u m an h ead relative to the body d u rin g g ro w th (Shea 1988a:21). T h e relative
g ro w th rates, o r allo m etric relatio n s, o f tw o o rg an s o r stru ctu re s m ay vary (e.g.,
fro m p ositive to negative) d u rin g the o n to g en y o f an organism . G ro w th rates
can be m easu red fo r an o rg a n o r stru ctu re , a p a rt o f a n o rg a n o r stru c tu re , o r
the com p lete org an ism (R ensch 1960).
A llo m etric relatio n s can be determ in ed fo r a tra it across m ultiple tax a, as
w ith th e large m ouse end o w ed w ith a n elep h a n t skeleton, w ithin a ta x o n as a
fu n ctio n o f the o n to g en ic age o f individuals, o r as a fu n c tio n o f different-sized
ad ults. O f in terest to a tap h o n o m ist is the process o f bone fo rm a tio n in a
grow ing v erte b rate, as w hen th e ca rtila g in o u s m odel o f a lim b b o n e is replaced
by b o n e tissue (F ig u re 4.1). T he relevance o f such allo m etric re la tio n s is well
illu stra te d by the n u m ero u s zo o a rch a eo lo g ists w ho have suggested the reason
th a t th e rem ain s o f on to g en ically y o u n g m am m als are ra re in m an y a rc h a e o lo
gical collections is th a t th eir b o n es are generally o f low s tru c tu ra l density. These

b ones are n o t fully ossified a n d th u s are readily co n su m ed by scavenging
carn iv o res o r rem oved by o th e r ta p h o n o m ic ag en ts such as fluvial a c tio n due to
th eir relatively low b u lk density a n d high p o ro sity .
Skeletal tissues
Because skeletons resist decay after d eath , they have becom e the aw e-inspiring
epitom e o f death itself in every h u m an culture.
(H. Francillon-V ieillot et al. 1990:473)
V erteb rate bodies co nsist o f v ario u s kinds o f soft tissues, an d v ario u s kinds o f
h a rd tissues. T he fo rm er consist o f m uscles, ligam ents, ten d o n s, a n d hide; these
m ay be fo u n d in arch aeo lo g ical co n tex ts w here p re serv atio n al c o n d itio n s are
ex cep tio n al, b u t typically they do n o t preserve. H a rd tissues are th o se th a t
m ak e u p the skeleton, a n d include bone, to o th , cartilage, a n d h o rn a n d antler.
It is th e h a rd tissues th a t typically preserve in arch aeo lo g ical co n tex ts a n d thus
it is th e h a rd tissues, the skeletons, w ith w hich n early all ta p h o n o m ic analyses
are con cern ed . T he re a d e r sh o u ld u n d e rsta n d th a t in th e follow ing, unless
otherw ise n o ted , the w o rds bo n e, cartilage, to o th , an tler, a n d h o rn refer to the
tissue, a n d n o t som e specific skeletal elem ent.
Bone
Bone is a tissue th a t has evolved as a s tru c tu ra l m ateria l (M einke 1979). It thus
displays ra th e r re m a rk ab le m echanical p ro p e rtie s in a n engineering sense
(C u rrey 1984; M a c G re g o r 1985). As well, because it is a living tissue, it can n o t
only re sp o n d m echanically to stresses, it can also re sp o n d by alterin g its
p ro p erties. A n y o n e w ho has suffered an accident resulting in one o r m ore
b ro k e n bones readily ap p reciate s these facts. Bone is o ften referred to as a
“ tw o -p h a se” o r “ c o m p o u n d ” m ateria l (M einke 1979). T his is so because bone
consists o f a b o u t 70% in o rg an ic m ateria l, usually hyd ro xya p a tite, a calcium
p h o sp h a te m ineral w ith the general chem ical co m p o sitio n C a 10(P O 4) 6.2O H ,
a n d 30% o rg an ic m a tte r, m o stly the s tru c tu ra l p ro te in collagen, w hich ca n also
vary in chem ical co m p o sitio n (F ran cillo n -V ieillo t et al. 1990:515). T he fo rm er
m aterial is resistan t to co m p ressio n forces a n d the la tte r to ten sio n forces
(C u rrey 1984; Jo h n so n 1985). T he h y d ro x y a p a tite crystals are su p p o rte d in an
extensive system o f collagen fibers, being fo u n d b o th w ithin an d a ro u n d those
fibers; “ one o f the long axes o f the m ineral plates is alw ays fairly well aligned
w ith th e collagen fibrils” (C u rrey 1984:26). C ollagen fibers are infinitely long
c o m p ared to the discrete a p a tite crystals. T he m inerals co n fer rigidity and
h ard n ess, an d the o rg an ic m a tte r confers tou g h n ess, resiliency, an d elasticity to
bones (H ild e b ra n d 1974:96; R o m e r an d P arso n s 1977:150).
T he term “ a p a tite ” refers to a diverse g ro u p o f calcium p h o sp h a te m inerals,
Structure and quantification o f vertebrate skeletons 73
w ith o th e r m in o r in clu ded elem ents vary in g fro m tissue to tissue (the w ord
“ a p a tite ” derives fro m th e G re ek w o rd apati w hich m ean s “ to deceive” ); th u s
there is a diversity o f a p a tite m inerals (C arlso n 1990:531). “ Studies o f in organic
a p a tite have d em o n stra te d th a t a high fluorine co n te n t increases the stability
(decreases solubility) o f the m in eral . . . In creased c a rb o n a te c o n te n t raises
a p a tite so lu b ility ” (C arlso n 1990:531). T he chem ical co m p o sitio n o f skeletal
tissue, th u s, can influence diagenesis (see C h a p te r 11).
B one tissue is alive w hile th e a n im al is alive. It serves as a m etab o lic reservoir
fo r vario u s m inerals, especially calcium a n d p h o sp h a te s (de R o u sseau
1988:95). B one tissue is “ in a state o f c o n s ta n t m etab o lic exchange w ith th e rest
o f th e b o d y ” (M a c G re g o r 1985:7) a n d th u s if the b o d y is n o t tak in g in
a p p ro p ria te n u trie n ts in the p ro p e r a m o u n ts, b ones will re act accordingly. T his
creates som e skeletally visible sig n atu res o f d ietary deficiencies an d disease th a t
are o f g reat use to zo o a rch a eo lo g ists (e.g., B aker a n d B rothw ell 1980) an d
p erh ap s to ta p h o n o m ists as well.
T h ere are tw o basic m odes o f b o n e tissue fo rm a tio n . D erm al bone form s
directly in th e m esenchym e o r n e a r b o d y surface (skin) cells. B ones o f the
h u m a n c ran ial v au lt are d erm al bones. Endochondral bone is “ p re fo rm e d ” by a
cartila g in o u s m odel w hich is replaced by b o n e tissue (F ig u re 4.1). L o n g bones
in h u m an s are en d o c h o n d ral. D u rin g the o n to g en y o f en d o c h o n d ra l bone, a
long b o n e consists o f three d istin ct p arts. T he diaphysis is the sh aft p o rtio n , and
the epiphyses (pi.; sin g u lar fo rm is epiphysis) are th e tw o ends. E ach is a center
o f ossification d u rin g onto g en ic dev elo p m en t (F ig u re 4.1). L o n g -b o n e grow th
occurs a t th e ends o f th e d iaphysis in th e zone k n o w n as the m etaphysis w here
th e cartilag e is replaced by bone. B efore the epiphyses a n d diaphysis begin to
g row to g e th e r th ey are sep a rated by a disc o f cartilag e k n o w n as th e epiphyseal
p la te an d , u p o n its d isap p earan ce, the epiphyses a n d diaphysis grow to g eth er
a n d fuse in to one discrete object, th e a d u lt bone. L o n g b ones grow in
circum ference o r g irth as they grow in length, w ith successive layers o f co m p ac t
b o n e (som etim es referred to as periosteal bone) d ep o sited a ro u n d the ou tsid e o f
th e diaphysis.
Cells th a t d ep o sit b o n e m a te ria l are called osteoblasts, a n d as the o steo b lasts
b ecom e em b ed d ed in the bone m atrix they becom e osteocytes or, literally, bone
cells. B one g ro w th o ccurs until ad u lt size is reached, b u t bo n e rem odeling
o ccurs th ro u g h o u t th e life o f a n organism . R em odelling o f bone is ac co m
plished by osteoclasts w hich re so rb bone, creatin g in p a rt incom plete H a v e r
sian system s (see below a n d “ in terstitia l system s” in F ig u re 4.3). H aversian
system s are series o f sm all canals co n tain in g b lo o d vessels a n d nerves w hich
b ra n c h th ro u g h the bone; bones receive n u trim e n t fro m these. T he periosteum
is the m e m b ra n o u s sh eath fo u n d on the ex tern al surface o f a b o n e th a t can be
stim u lated to p ro d u c e new bone, as w hen an org an ism break s a bone. T he
a rtic u la r surfaces o f b ones are n o t covered w ith p erio steu m , b u t a th in layer o f
cartilage. In te rio r (m arro w o r m edullary) cavities o f som e e n d o c h o n d rial
bones, especially th e long bones o f the lim b, are lined w ith a m em b ra n e know n
as th e endosteum (bo n e here is som etim es referred to as endosteal bone).
T h e m a jo r s tru c tu ra l elem ent o f b o n e tissue is th e osteon. A n o steo n is sim ply
a ro u g h ly cylindrical stru c tu re o f successive lam ellae su rro u n d in g a centrally
lo cated H a v ersian canal. T he alig n m en t o r o rie n ta tio n o f collagen fibers an d
a p a tite crystals varies fro m one lam ella to a n o th e r w ithin the osteon, w hereas
the long axis o f the o steo n ten d s to be p arallel to the long axis o f the bone
(F ig u re 4.20- In d iv id u al H av ersian canals are linked to one a n o th e r betw een
o steo n s by ra d ia tin g b lo o d vessels th a t occupy V o lk m a n ’s canals. In d iv id u al
o steo n s are jo in ed to one a n o th e r by a su b stan ce called cement.
T h ere are several form s o f bo n e th a t can be distinguished, basically at
different scales. A t a fine scale, fo r m am m als, w oven bone, lam ellar bo n e, an d
parallel-fibered bo n e can be distinguished (C urrey 1984:26-27). W oven bone
form s quickly a n d h as ra n d o m ly orien ted fine collagen fibers. M ineral crystals
in w oven b o n e are also ra n d o m ly o rien ted (M a c G re g o r 1985:4). W oven bone
has irre g u la r trab e cu lae a n d is tran sien t, being the initial kind o f bone
d ep o sited in th e fetus a n d a ro u n d b o n e fractures. Spaces a ro u n d b lo o d vessels
in w oven bo n e are m o re extensive th a n th o se fo u n d in lam ellar bone. L am ellar
o r lam in ated b o n e fo rm s m ore slow ly a n d has a n organized stru c tu re w ith the
collagen a n d b o n e fibers a rra n g e d in layers called lam ellae (C u rrey 1984;
M einke 1979). W ith in each lam ella the collagen fibers form g ro u p s; the
in d iv idu al g ro u p s o f fibers m ay display d istin ct o rien ta tio n s, an d these ca n vary
betw een ad jac en t lam ellae (F ig u re 4.2). “ Parallel-fibered bone is stru ctu ra lly
in term ed iate betw een w oven bo n e a n d lam ellar b o n e ” (C u rrey 1984:27), an d
seems to be m ore ra re th a n lam ellar a n d w oven bone. Sm all cavities o r lacunae
p erm eate all three kinds o f bo n e, a n d b lo o d vessels in all types are fo u n d in
canaliculi.
A t a m o re general stru c tu ra l scale fo u r kinds o f b o n e can be distinguished:
w oven bon e, lam ellar bone, H av ersian system lam ellar bone, a n d fibrolam ellar
b o n e (C u rrey 1984:28-29). W oven bo n e a t this scale occurs in areas several
m illim eters in all d irectio n s in y o u n g b o n e a n d fractu re calluses. L am ellar bone
at this scale also extends o ver relatively large areas, such as a ro u n d the outside
surface o f m am m alian long bones (“ circum ferential lam ellae” [C urrey
1984:28]) (F ig u re 4.3). H aversian system lam ellar bone is form ed w hen the bone
m ateria l a ro u n d a b lo o d vessel is ero d ed by o steoclasts, an d new b o n e is
d ep o sited in co n cen tric layers on the (inner) surface o f the resulting cavity
(F ig u re 4.3). T he b lo o d vessel(s) rem ains at the cen ter o f the concentric
lam ellae. T hese H a v ersian system s can vary co n sid erab ly in how they are laid
out. T h eir o u te r lim it consists o f a “ cem ent s h e a th ” a n d very few canaliculi
cross it so th a t cells o u tsid e the sh ea th are “ cut off m etabolically fro m the blood
vessel in the m iddle o f the H av ersian system ” (C u rrey 1984:29), alth o u g h
individual H av ersian system s are con n ected by V o lk m a n n 's canals (M a c G re
g o r 1985:5). Fibrolam ellar (o r lam inar) b o n e is “ fo u n d p a rtic u la rly in large
Structure a nd quantification o f vertebrate skeletons 75
Figure 4.2. S tructure o f m am m alian bone a t different scales an d levels o f

organ izatio n, a, collagen fibril w ith associated m ineral crystals; b, w oven bone,
collagen fibrils random ly arranged; c, lam ellar bone, show ing separate lam ellae
w ith collagen fibrils oriented in p articu lar dom ains; d, w oven bone, w ith blood
channels as black spots; e, p rim ary lam ellar bone w ith lam ellae show n as light
dashes; f, H aversian bone, each H aversian system w ith concentric lam ellae
aro u n d a central blood channel; g, lam in ar bone w ith altern atin g layers o f woven
and lam ellar bone; h, co m p act b one types from low er levels; i, cancellous
(trabecular) bone. R eproduced by perm ission from : W ainw right, S. A. et al.
1976, Figure 5.14. L ondon: E dw ard A rn o ld Ltd. C o pyright 1976 by S. A.
W ainw right, W. D. Biggs, J. D . C urrey an d J. M . G osline.
Interstitial system
Outer basic lamellae
Canal of Volkmann
Figure 4.3. M icrostru ctu re o f m am m alian bone show ing H aversian an d lam ellar
bone. R eproduced by perm ission from : R om er, A. S. an d P arsons, T. S. The
vertebrate body, F igure 105. Philadelphia: W . B. S aunders C om pany. C opyright
1977 by W. B. Saunders C om pany.
m am m als” the b o n e o f w hich m u st grow quickly in d iam eter (C urrey 1984:29).

T his is a c o m b in a tio n o f w oven o r parallel-fibered a n d lam ellar bone, w ith
w oven bo n e first quickly fo rm in g a kind o f scaffolding, a n d lam ellar bone
form in g la te r on th e o u tsid e o f the stru c tu re an d eventually filling in the cavities
o f th e originally form ed w oven bone. T his creates a lte rn a tin g layers o f w oven
or parallel-fibered a n d lam ellar bo n e (C u rrey 1984:29; M a c G re g o r 1985:5).
A t th e final an d highest scale, com pact (dense) bone is relatively solid w hereas
cancellous (trabeculated, spongy) bone h as large spaces. P o ro sity o f th e fo rm er
is lim ited largely to the H a v ersian canals a n d lacunae; cancellous b o n e is quite
p o ro u s given its stru c tu re o f plates a n d stru ts. T he a m o u n t a n d d istrib u tio n o f
co m p ac t a n d spongy bone tissues is d ep e n d en t on each individual b o n e ’s
p a rtic u la r stru ctu re a n d fu n ctio n (R o m er a n d P arso n s 1977:151) (F ig u re 4.4).
L o n g b o n es typically consist o f m o stly cancellous bo n e a t th e ir epiphyseal ends
an d co m p ac t (lam ellar) bo ne m ak es u p the diaphysis. C ancellous b o n e can tak e
several form s, b u t basically consists o f bone stru ts an d plates th a t are variously
in terc o n n ected a n d o f v ario u s o rie n ta tio n s, largely d ep en d in g o n the fu n c tio n
o f th e p a rtic u la r skeletal elem ent. T hese stru ts a n d p lates are called trabeculae
S tructure and quantification o f vertebrate skeletons 77
Figure 4.4. T he ap pearan ce an d d istrib u tio n o f trab ecu lar o r cancellous and
com pact or cortical bone in a typical m am m alian long bone, a proxim al
hum erus. R eprod u ced by perm ission from : Hesse, B. an d W apnish, P. A nim al
bone archeology, Figure 25. W ashington, D .C .: T arax acu m Press. C o p yright
1985 by B. Hesse and P. W apnish.
(tra b ecu la m ean s stru t). In m am m als, the spaces o r pores o f cancellous bone
are usually filled w ith m arro w , a n d m o st cancellous b o n e occurs in the ends o f
long b ones (C urrey 1984). L am in ated o r lam ellar an d H av ersian bo n e m ake up
co m p ac t bone.
T he bo n e tissue o f birds is sim ilar to th a t o f m am m als, b u t the thickness o f
the w alls o f bird b ones tends to be less, relative to a b o n e ’s d iam eter, th a n in
m am m als. B ird bon es are n o t m a rro w filled, b u t they fill w ith calcium as a
reserv o ir fo r egg-shell p ro d u c tio n (M a c G re g o r 1985:8). R eptilian bo n e is
cellular a n d “ in all b u t one m a jo r re p tilian g ro u p the bone tissues are v ascu lar
[but] in virtually all ad u lt reptiles, som e localized areas can be n o n -v a scu lar in
s tru c tu re ” (E nlow 1969:45, 47). T he bo n e o f am p h ib ian s is avascular. H a v e r
sian system s are lacking in th e b o n e o f m an y reptiles including lizards an d
snakes (E nlow 1969:47). C o m p a c t b o n e o f lizards a n d snakes is “ v irtu ally n on-
v asc u la r” a n d lizard long bones have a “ relatively lim ited extent o f cancellous
trab e cu lae in the m id -d iap h y sis” (E nlow 1969:62-63). B one o f teleost fishes has
n o b o n e cells a n d lacks osteocytes (C u rrey 1984; E nlow 1969).
T u rtle shells are p e rh a p s the m o st in trig u in g v e rte b ra te skeletal p arts. These
shells consist o f the d ish -sh ap ed d o rsa l carapace a n d the ven tral flat plastron;
th e tw o are co n nected alo n g the sides by a n a rea called the bridge. T h e shell
consists o f an o u te r epid erm al, h o rn y surface cover a n d an in n er derm al, bony
arm o r (Z an g erl 1969:312). In ad u lt turtles, the o u ter a n d in n er layers are
sep a rated by a “ spongy m iddle region co n tain in g large n u m b ers o f spherical
cavities o f different sizes. O n b o th sides o f this there are zones o f co m p act
lam ellar bo n e, co n ta in in g m o d era tely n u m ero u s ra d ia l v asc u la r c a n a ls” (Z a n
gerl 1969:313). T h e in n er layer o f lam ellar b o n e is m o re v asc u la r th a n th e o u ter
layer. T h e shell does n o t serve as a m in eral reservoir w hereas the lim b bones do
(Z an g erl 1969:313).
C u rre y (1984:36) considers individual b ones to consist o f three basic shapes:
tu b u la r, ta b u la r, a n d s h o rt bones. T he fo rm e r are elo n g ated in one direction
an d in cross-section are a p p ro x im a te ly circular. T hey are ex p an d ed a t the ends,
an d in clude the long b ones o f the lim bs a n d the ribs. T a b u la r bones are those
th a t are p a rtia lly flatten ed, such as th e pelvis, scapula, a n d som e bones o f the
b ra in case. S h o rt b o n es are ro u g h ly the sam e dim en sio n in all d irections, an d
include carp als, som e tarsals, a n d som e phalanges. D avis (1987:47) dis
ting u ish es cylindrically sh aped b o n es (e.g., long bones o f th e lim bs), flat bones
(e.g., skull, scap ula, in n o m in a te , rib), a n d irreg u larly shaped bones (e.g.,
verteb rae). M icozzi (1991:54) distinguishes fo u r “ m o rp h o lo g ical types o f
[hum an] b o n es.” L o n g bones include the m andible, clavicle, h u m eru s, radius,
ulna, fem ur, tib ia, a n d fibula. S h o rt bones include carpals, m etac arp a ls, tarsals,
m etatarsals, a n d phalanges; p h alan g es are the “ tru e sh o rt b o n e s” (M icozzi
1991:54). F la t b ones include the fro n ta l, p arie tal, occipital, tem p o ra l, sternum ,
scap u la, sacrum , ilium , ischium , p u b is, a n d ribs. Irre g u la r b ones include the
v erteb rae, p atella, h yoid , sp h en o id , m axilla, n asal, eth m o id , lacrim al, p alate,
an d vom er. All th ree a rra n g e m e n ts are sim ple m o rp h o lo g ical typologies m ean t
to u n d ersco re som e o f th e sim ilarities a n d differences in b o n e shapes. A m ore
g eom etric a n d less an a to m ic a l a p p ro a c h to categorizing bones acco rd in g to
th eir sh ap e is d escribed in C h a p te r 6 (F ig u re 6 .6). As in d icated there, th a t
system is quite relev ant to som e ta p h o n o m ic problem s.
C artilage
C artilag e is som etim es fo u n d in arch aeo lo g ical contexts. T here are several
basic types o f cartilag e, b u t all are so m ew h at elastic a n d are h a rd e r th a n m o st
S tru ctu re and quantification o f vertebrate skeletons 79
soft tissues such as m uscle b u t so fter th a n b o n e (H ild eb ran d 1974:126; R o m er

a n d P a rso n s 1977:150). C artilag e m ay tak e the fo rm o f a firm gel th ro u g h w hich
is sp read a n etw o rk o f connective tissue fibers. C artilag e cells are called
chondrocytes, a n d these are iso lated w ithin the m atrix (gel) they h av e secreted.
C artilag e typically does n o t c o n ta in b lo o d vessels. C artilag e m ay becom e
calcified via th e d ep o sitio n o f calcium salts in the m atrix , w hich creates a h ard ,
b rittle m aterial sim ilar to bone. Such calcification precedes the replacem ent o f
cartilage by b o n e d u rin g the ossification o f b ones th a t a re g row ing (F ran cillo n -
V ieillot et al. 1990:520-523; R o m er a n d P arso n s 1977:149-150).
C artilag e m ain ly occupies in tern a l areas o f th e b o d y (R o m er a n d P arso n s
1977:150). T h a t is, w ith few exceptions (such as the m am m alian ea r a n d nose)
cartilage is never in th e skin a n d seldom n e a r th e surface o f the body. C artilag e
is m o re a b u n d a n t in y o u n g anim als, a n d less a b u n d a n t in ad u lts due to the
rep lacem en t o f m u ch cartilage w ith bo n e d u rin g ontogeny. B ecause cartilage
generally h as a low er stru c tu ra l density th a n bone, it ten d s to m ed iate the
effects o f ta p h o n o m ic processes less well th a n bone.
Tooth
T he m a jo r fu n c tio n o f teeth is m asticatio n (K ay 1988). M am m alian teeth are

largely co m p o sed o f enam el a n d dentine, w ith a m in o r a m o u n t o f cem entum
(F ig u re 4.5). E nam el is the h a rd e st skeletal tissue, an d is co m p o sed o f elongate
crystals o f h y d ro x y a p atite ; the a p a tite crystals in enam el are larger th a n they
are in d en tin e o r in b o n e (C arlso n 1990:535). O nly a b o u t 1 -3 % o f enam el tissue
is o rg an ic (a n o n -co llag en o u s form ); th e re m a in d e r is m in eral w hich is
org an ized in to p rism s (H ild eb ran d 1974:95; R o m er a n d P arso n s 1977:301).
E nam el is h a rd e r, denser, a n d less soluble th a n d en tin e because o f its low
p o ro sity a n d low o rg an ic c o n ten t. It is a stiff, nonelastic m aterial th a t is quite
re sistan t to diagen etic chem ical change. E nam el is, how ever, “ m o re b rittle a n d
h as less com pressive stren g th th a n d e n tin e ” (C arlso n 1990:537) because it is so
highly m ineralized a n d so n o n -p o ro u s. D entine is h a rd e r th a n co m p ac t bone
b u t so fter th a n enam el. It is also co m p o sed o f h y d ro x y a p atite b u t the tissue
itself is a b o u t 30% o rg an ic a n d 70% m in eral by w eight (C arlso n 1990:533).
D en tin e is “ p erm eated by sin u o u s su b -p arallel canals called d en tin al tu b u les,”
a n d it is th u s a “ fairly p o ro u s m a te ria l” (C arlso n 1990:533, 534). B ut like bone,
“ th e s tru c tu ra l in term ix tu re o f collagen fibers a n d a p a tite crystals in dentine
creates a co m p o site m ateria l w ith a hig h degree o f elasticity a n d stre n g th ”
(C arlso n 1990:534). C em entum is a type o f b o n e b u t is in p a r t acellular (M einke
1979). It is a m in eralized connective tissue th a t “ is sim ilar to b o n e b o th
u ltra stru c tu ra lly a n d b io m ech an ically ” (C arlso n 1990:534). It is 70% ap a tite,
25% o rg an ic, an d 5% w ater by w eight (C arlso n 1990:534).
M am m al teeth have a crow n, neck, a n d ro o t (F ig u re 4.5). O ntogenically, the
enam el cro w n form s first, th e n the b o d y o f th e to o th (dentine) form s, an d
finally as th e to o th eru p ts th e ro o t form s. T h e to o th is set in an alveolus or
Figure 4.5. C ross section o f a typical m am m al to o th show ing m ajo r com ponents
and regions.
socket in th e jaw . M am m alian d en titio n s are heterodont; th a t is, different kinds

o f teeth are fo u n d in each jaw . T hese typically are called, from fro n t to back,
incisor, canine, p re m o la r, a n d m o la r teeth. T he n u m b er a n d m o rp h o m e try o f
teeth is tax o n o m ically d istinctive in m o st anim als. T h e tusks o f elephants,
w alrus (O dobenus rosm arus), a n d n arw h al (M onodon m onoceros) are specia
lized teeth (see C arlso n 1990:546-553 fo r a su m m ary o f the m icro stru c tu ral
in tertax o n o m ic v aria tio n in teeth). E le p h a n t tusks are, in fact, a specialized
version o f d en tin e (F ran cillo n -V ieillo t et al. 1990:474).
G iven th a t the chem ical, m a c ro -stru c tu ra l, a n d shape p ro p e rtie s o f teeth are
different fro m bo nes, teeth tend to resp o n d differently to ta p h o n o m ic processes
th a n bones. W e ath erin g (C h a p te r 11) o f teeth, fo r instance, is q u ite different
from w eath ering o f bones. W hile b o n e ten ds to split a p a rt a n d exfoliate in w hat
m igh t be th o u g h t o f as a tim e-transgressive process leading eventually to bone
du st, teeth sim ply fall to pieces.
A n tler an d horn
H o rn s a n d an tlers are co m p o sed o f q u ite different m aterials. H o rn is largely
restricted to bovines, a n d is m ainly k eratin , a p ro tein th a t also m ak es u p hair,
h o o f-sh eath s, a n d feath ers. It is, in effect, fu n ctio n ally p a r t o f the exoskeleton
(M a c G re g o r 1985:19). As a tissue h o rn tends to grow slow ly in a n epiderm al
layer a ro u n d a v ascu lar bony core referred to as the " h o rn co re ” a n d arising
from the fro n ta l bo n e o f the skull. T he horn core, o r “ os cornu, orig in ates in the
su b cu tan e o u s conn ective tissue a n d fuses secondarily w ith the u nderlying skull
. . . cartilag e is n o t k n o w n to be p resen t in the d evelopm ent stages o f h o rn bone
[which] develops as an in d ep en d e n t cen ter o f ossificatio n ” (G oss 1983:67).
“ H o rn y lam ellae are ad ded [at the base] intern ally , in the n a tu re o f a cone
w ithin a cone. T his results in the distal displacem ent o f m ateria l p ro d u ced
e a rlier” (G oss 1983:57), a n d som etim es results in th e fo rm a tio n o f g row th
increm ents. G en erally if rem oved, h o rn is n o t reg en erated , alth o u g h the
p ro n g h o rn (A ntilocapra am ericana) o f N o rth A m erica a n n u a lly casts th e old
h o rn a n d grow s a new h o rn (O 'G a ra an d M a tso n 1975; R o m er a n d P arso n s
1977:133-134; S olo unias 1988).
A n tler resem bles bo n e, a n d grow s at the tip w ithin a b lo o d -rich skin called
velvet from a p ro tru d in g p o rtio n o f the fro n ta l b o n e called a pedicle (G oss
1983:57). A n tler grow s seasonally, ossifies, a n d is shed a n n u a lly (M odell 1969).
In d iv id u al an tlers o f m o d ern cervids are lo n g cylinders th a t v ario u sly b en d an d
b ra n ch , a n d w hich are filled w ith cancellous b o n y tissue (G oss 1983). It is
generally th o u g h t th a t antlers grow by ossification at the grow ing tips w ith
ad d itio n a l an tler m aterial being a d d e d to the surface. B lood vessels are in tern al
a n d ex tern al (in th e velvet) to th e grow ing an tler. T hese b lo o d vessels shrivel
a n d die w hen the a n tle r com pletes g ro w th , a n d th e velvet is shed. A n tle r tissue
“ consists p rim arily o f co arsely -b u n d led w oven b o n e ” (M a c G re g o r 1985:12).
In d iv id u al an tlers h ave a cancellous core encased by c o m p a c t tissue; th e fo rm er
occupies progressively less cross-sectional a rea as one progresses o u t individual
a n tle r tines such th a t the tips o f the tines are co m posed only o f the co m p ac t
tissue. T h e ra tio o f m in eral to org an ic m a tte r o f a n tler is sim ilar to th a t o f bone,
a n d a n tle r p ro d u c tio n m ay result in re d u ctio n o f bone m ineral co n ten t
(M a c G re g o r 1985:13). A n tle r g ro w th a n d develo p m en t d epends n o t only on
the age o f the an im al b u t on the n u tritio n a l statu s a n d h ea lth o f the anim al as
well (B row n 1983).
O ther tissues
T h u s fa r I have co n sid ered the skeletal tissues w ith w hich m o st v erte b rate
ta p h o n o m ists deal. T h ere are several o th e r tissues, how ever, th a t w a rra n t
m en tio n. A m p h ib ian s, like fish, have b o n y scales, b u t the scales o f reptiles are
k eratin ized stru ctu re s (F ran cillo n -V ieillo t et al. 1990:483). F ish scales are
com plex, p o ly m o rp h ic stru ctu re s th a t have a varied b u t basically derm al
origin. “ T he scales o f fishes fo rm a m ore o r less c o n tin u o u s derm al skeleton on
the bo dy , w hich m o d u lates in to the specialized d erm a l elem ents o f the fins (fin
rays) on one h a n d , a n d the m o u th a n d p h a ry n x (teeth) o n the o th e r”
(F ran cillo n -V ieillo t et al. 1990:477). F in rays a n d scales are h o m o lo g o u s w ith

m ineralized skeletal elem ents (F ran cillo n -V ieillo t et al. 1990:474). R elated
stru ctu re s fo u n d in som e fishes are the b o n y plates kn o w n as scutes a n d spines,
b o th o f w hich are m odified scales (F rancillon-V ieillot et al. 1990:486). O to lith s
are “ acellu lar m ineralized co n cretio n s w hich occu r in th e in n er e a r” o f m an y
fishes (F ran cillo n -V ieillo t et al. 1990:524). T hey are o rg an s o f equ ilib riu m and
are “ generally co m p o sed o f calcium c a rb o n a te in the fo rm o f p u re a ra g o n ite ”
(C asteel 1976:20). T he shell o f the arm a d illo (D a sy p o d id ae) is co m p o sed o f
scales, each o f w hich consists o f a d erm a l b o n y plate, an ep id erm al k eratin o u s
covering, a n d co n nectiv e tissue betw een the tw o.
W hile it is som etim es fo u n d in arch aeo lo g ical co n tex ts (e.g., K eepax 1981),
av ian egg shell is, ap p a ren tly , ra rely fo u n d as it is seldom re p o rte d in the
literatu re. A v ian egg shell is m ain ly calcite b u t h as a n o rg an ic co m p o n e n t as
well. It is so m ew h at p o ro u s, a n d a p p a re n tly preserves well in alk alin e sedi
m ents (K eep ax 1981:317). In fo rm a tio n on the m ic ro stru c tu re o f avian egg shell
is c o m p a ra b le to th a t o u tlin ed above fo r b ones an d teeth (e.g., B ecking 1975;
T yler 1969).
Properties o f skeletal tissues and taphonomy

[The knee] is a stru ctu re o f such m echanical im plausibility th a t it it m ust be designed
to do som ething very well. U n fo rtu n ately , I do n o t know w h at th a t som ething is.
(J. D. C urrey 1984:184)
W h a t is th e ta p h o n o m ic significance o f w h eth er b o n e tissue is co m p o sed o f

w oven o r H a v ersian bone? W h y sh o u ld we w o rry th a t bo n e tissue consists o f
b o th a m ineral co m p o n e n t a n d a n o rg an ic co m p o n en t? T hese a n d sim ilar
co n cerns are im p o rta n t because they influence the u ltim ate effect a p a rtic u la r
ta p h o n o m ic ag en t o r pro cess will h av e o n a b o n e specim en. C u rre y (1984:3-4)
n otes th a t “ b ones fu n c tio n m ainly by n o t d efo rm in g ap p reciab ly u n d e r load . . .
T he stiffness o f a b o n e a n d its stren g th dep en d on tw o factors: the stiffness an d
stren g th o f the b o n e m ateria l itself a n d also th e build [m o rp h o m etry , o r size an d
shape] o f th e w hole b o n e .” T h u s, bones have a fu n c tio n in th e b o d y o f an
o rg an ism , a n d th a t fu n c tio n d ictates th e taphonom ic strength o f th e bone; th a t
is, a b o n e ’s a n a to m ica l fu n ctio n d eterm ines the b o n e ’s m echanical pro p erties,
a n d the m echan ical p ro p e rtie s o f a bone in tu rn m ed iate the effects o f
ta p h o n o m ic processes on the bone. W e begin w ith a review o f som e o f the
basics o f bio m ech anics before we tu rn to m echanical p ro p e rtie s o f skeletal
tissues. M o st research on the m echanical p ro p e rtie s o f skeletal tissues has
focused on bone, a n d th a t skeletal tissue is th ere fo re the one I focus on in the
follow ing, a lth o u g h co m m en ts a b o u t the o th e r basic kinds o f skeletal tissues
are offered w here possible.
Basic biom echanics

Stress is a m easu re o f the intensity o f force, an d can be m easured as the force
p er u n it o f cross-sectio n al area. Strain involves the ch ange in shape o f a b o d y
u n d er stress an d is d im ensionless (C u rrey 1990:11). Stress a n d strain are related
a n d are initially p ro p o rtio n a l to one a n o th e r. T he steep p o rtio n o f the curve in
F ig u re 4.6a show s this initial re la tio n sh ip betw een the origin o f the curve an d
the y ie ld point. B eyond the yield p o in t o f the curve, “ irreversible changes have
ta k e n place in th e m a te ria l” a n d strain increases a t a m u ch m o re ra p id ra te th a n
stress (C u rrey 1990:11). T h e initial p a r t o f the curve is called Y oung’s m odulus o f
elasticity a n d m easures the stiffness o f the m aterial. T he m ineral co m p o n e n t o f
skeletal tissue h as a hig her Y o u n g ’s m o d u lu s a n d is m o re b rittle th a n the
org an ic co m p o n en t. E v entually the specim en b reak s, at w hich p o in t the
ultim ate stress a n d ultim ate strain ca n be m easured. T he to ta l area u n d e r the
curve is a m easu re o f the to u g h n ess o f the m aterial. A n tler tissue, fo r exam ple,
ten d s to be to u g h e r th a n bo n e (F ig u re 4.6b). A to u g h m aterial resists cracks
travelin g th ro u g h itself (C urrey 1984, 1990), a n d the co m posite n a tu re o f
skeletal tissue, being p a rt m ineral a n d p a rt organic, m akes it very tough,
to u g h e r th a n eith er the m ineral o r o rg an ic fractio n alone. N o te th a t the ra te o f
lo ad in g can influence th e shape o f the stre ss-stra in curve in F igure 4.6a (C urrey
1990:12; see C h a p te r 8).
T eeth, because they are m ostly m ineral, have a high Y o u n g ’s m o d u lu s
relative to co m p act bone. B ut enam el is b rittle a n d n o t very to u g h (C urrey
1990:19). Y o u n g ’s m o d u lu s is low er in cancellous bone, w hich is b e tte r able to
ben d d u e to its relatively p o ro u s stru ctu re , th a n in co m p ac t b o n e (C urrey
1990:23). T he g re ater the m in era liza tio n o f co m p ac t bone, the g re ater the
stiffness a n d th e less the toughness. A n tlers, w hich have relatively low m ineral
co n ten t, h ave a low Y o u n g ’s m o d u lu s b u t are q u ite to u g h relative to co m p act
b o n e w ith high m ineral co n ten t.
Biom echanics and skeleta l tissues

C u rre y (1984:58) observ ed th a t a p a tite is the m ineral o f choice fo r b ones ra th e r
th a n , say, calcium c a rb o n a te , because “ a p a tite is very re lu c ta n t to fo rm large
cry stals” a n d “ large cry stals are b rittle a n d in general to be avoided in m aterials
th a t need to be to u g h .” [M artill (1990:272) no tes th a t the “ u ltra-m icro sco p ic
size o f th e b o n e m in eral has m ad e it difficult to study, especially w ith re g ard to
diagenetic processes th a t m ay have affected the original s tru c tu re .” ] T he
inclusion o f o rg an ic m a tte r in bo n e increases the tou g h n ess a n d stren g th o f the
tissue because “ th ere is a lim it to g re ater stren g th p ro v id ed by h y d ro x y a p atite
ap p o sitio n alone, fo r at a ce rtain level increased in o rg an ic m aterial will m ake
the co m p o site m ateria l b rittle ” (B u rr 1980:113). T h o se studying the m echani-
a.
strain
b.
Figure 4.6. a, M odeled relatio n o f stress an d strain, show ing Y o u n g ’s m od u lu s o f

elasticity an d p o in t o f failure (fracture), b, C o m p ariso n o f stress-strain relation
betw een b one an d an tler (after C urrey 1990).
cal p ro p e rtie s o f b o nes have, nonetheless, fo u n d th a t “ an u n d e rsta n d in g o f the

related effects o f p o ro sity a n d m in era liza tio n p rovides the m o st a d e q u ate
pictu re o f the m ech anical b eh a v io r o f b o n e ” (B urr 1980:120).
B u rr (1980:110) no tes th a t a b o n e ’s m echanical in teg rity is d eterm in ed in
p a rt by the m ineral m ass (density) o f the b o n e tissue. " D e n sity is a fu n c tio n o f
m in eralizatio n o f the o steo n a n d the p o ro sity p er u n it volum e o f bone. M ineral
salts ac co u n t fo r a b o u t 90% o f the density o f b o n e ” (B u rr 1980:110). B ecause
bo n e m in eral c o n te n t varies across a bone, w h a t I term the structural density o f
the b o n e (C h a p te r 7) will vary across the bone. F o r exam ple, increased w eight
b earin g increases the den sity o f a bone, a n d th e p a rts o f a b o n e w ith the g reatest
density are th o se p a rts th a t tend to u n d erg o the g re atest com pressive and
tensile stresses d u rin g the life o f the o rg an ism (B urr 1980:111). B ecause the
stru c tu ra l density o f bon e is determ in ed as a w eight:volum e ra tio , the p o ro sity
o f a b o n e specim en influences the stren g th o f th a t specim en: “ increasing
p o ro sity w ith age reduces the stren g th o f the b o n e ” (B u rr 1980:113), and
“ in creased m in eralizatio n intensifies the deleterious effects o f p o ro sity in older
b o n e by d ecreasing the ability o f b o n e to a b so rb energy by plastic a n d elastic
d e fo rm a tio n ” (B u rr 1980:117).
T he re la tio n betw een b o n e stren g th a n d m in eral c o n te n t h o ld s fo r b o th
cortical a n d tra b e c u la r bone (B urr 1980:117), as well as o th e r skeletal tissues.
B ut d u e to the g re ater m acro sco p ic p o ro sity o f tra b e c u la r o r cancellous bone, it
is less stiff a n d w eak er th a n co m p ac t b o n e (C urrey 1984:80). T he arran g e m e n t
an d co n tig u ity o f trab e cu lae influence the stiffness o f tra b e c u la r b o n e (B urr
1980:118). P o ro sity is also fo u n d in bo n e tissue at a m icroscopic level; “ the
p o ro sity o f co m p ac t b o n e is rep resen ted by the p ercen tag e (volum e) o f cavities
o r spaces fo rm ed by th e H a v ersian c a n a ls” (Jo h n so n 1985:166). T h u s, B u rr
(1980:120) n o tes th a t “ increasing re co n stru ctio n o f bo n e - i.e., a g re ater
p ercentage o f seco n d ary H av ersian system s - reduces the m echanical stren g th
o f bo n e by (1) in creasin g p o ro sity , (2) decreasing m in eralizatio n , since y o u n g er
b o n e is less calcified th a n o lder bo n e, a n d (3) increasing the n u m b e r o f osteons
p er u n it a re a .” F u rth e r, B u rr (1980:120) no tes th a t because the p erio steal bone
fo u n d on the o u tsid e o f a m a m m alian long bo n e is lam ellar ra th e r th a n
H av ersian in stru ctu re , it is denser (because it is m o re highly m ineralized) an d
less p o ro u s th a n th e H a v ersian bo n e w hich is fo u n d o n the in te rn a l surface o f
the skeletal elem ent fo rm in g the w alls o f the m a rro w cavity (F ig u re 4.1). C u rrey
(1984:86) con clu des th a t “ in no respect does H av ersian bone seem to have
m ech an ical p ro p e rtie s su p erio r to those o f fib ro lam ellar b o n e .” Y o u n g bones
have a h igh er im p act stren g th (are m ore able to w ith stan d im p acts w ith o u t
break in g ) th a n o lder bones (C u rrey 1984:93). T his is no d o u b t because the
fo rm er are less m in eralized, a n d th u s less b rittle, th a n the latter. A n d , as n oted
in th e prev iou s section, w oven bo n e is m o re p o ro u s th a n , say, lam ellar o r
H av ersian bone.
T h e n u m b er, size, a n d o rie n ta tio n o f collagen fibers, a n d the degree o f their
crosslin k in g all influence the m echanical p ro p e rtie s o f b o n e (B urr 1980:120-

121). A n exam ple w ith ta p h o n o m ic im p licatio n s is pro v id ed by T ap p e n (1969;
T ap p e n an d Peske 1970) w ho show s th a t the o rie n ta tio n o f split-lines and
w eath erin g cracks, w hich are induced d u rin g the drying a n d resulting sh rin
kage o f bon e (the initial stages o f subaerial w eathering; C h a p te r 9), is related to
the o rie n ta tio n o f th e m ajo rity o f the collagen fibers in the bone. T hus
w eath erin g crack s a n d split-lines on long bones tend to be p arallel to the long
axis o f the bone. B ones w ith a spiral o rie n ta tio n to th eir collagen fibers tend to
break spirally (H ill 1976). T he im p licatio n here is th a t individual bones o r
skeletal elem ents can be conceived o f as having a grain; “ m o st bone, except
w oven bo ne, has a definite grain , p ro d u c ed by the co -o rien ted cem enting
to g e th e r o f collagen fibrils a n d th e ir m ineral. T his gives b o n e a m icro stru c tu re
equ iv alen t to th a t o f a fibrous co m p o site w ith a very high volum e fractio n o f
fibers” (C u rrey 1984:86).
Jo h n so n (1985:167) n o tes th a t th e fractu re o f co m p ac t b o n e o n a m icro
scopic level is “ directly related to the a m o u n t a n d d istrib u tio n o f o steons, the
d istrib u tio n a n d o rie n ta tio n o f collagen fibers, a n d th e co m b in ed response to
force o f o steo n s a n d collagen fibers.” She (Jo h n so n 1985:167) w rites th a t
“ o steo n s w hose collagen fibers follow a steeply spiraled course a ro u n d the
lo n g itu d in al axis o f th e o steo n exhibit g re ater tensile b u t less com pressive
stren g th s th a n those o steons w hose fibers have a low angle o f spiral (low er
tensile b u t g re ater com pressive stre n g th s).” T he p o in t here is th a t the fractu rin g
o f bo nes is in p a rt related to the g rain o f the b o n e tissue (see C h a p te r 8).
W e k n o w very little a b o u t how the biom echanical p ro p e rtie s o f skeletal
tissues relate to forces such as ab ra sio n (C u rrey 1990:11). G eniesse (1982:38)
no tes th a t because force ap p lica tio n s at different angles to the grain o f bone
tissue h ave different d e fo rm a tio n a n d fra g m e n ta tio n effects, ab ra sio n forces
will p ro b a b ly have different effects dep en d in g on the d irectio n o f th o se forces
relative to b o n e tissue grain. She artificially a b ra d e d co m p act bovid bone u n d er
co n tro lled co n d itio n s, a n d collected a n d w eighed the bone tissue rem oved by
several instan ces o f b o th 300 a b ra sio n stro k es a n d a 30 second p erio d o f
ab rasio n . W hile statistically significant results w ere n o t o b tain ed , in every case
m o re b o n e tissue w as rem oved by ab rasiv e force applied p erp en d icu lar to the
g rain o f th e bo n e th a n by ab rasiv e force applied parallel to the grain. G eniesse
(1982:40) suggests th a t one m a jo r fa c to r co n tro llin g this difference w as the
g re ater ch an ce o f tearin g off larg er pieces o f bo n e tissue w hen w o rk in g against
the g ra in w hereas ab ra sio n forces w o rk in g p arallel to the g rain tend to w ear
d o w n th e b o n e in sm aller pieces.
S u m m a ry
T he preced in g co m m en ts have focused on tw o p ro p e rtie s o f bo n e tissue: its

stru c tu ra l density (an d p o ro sity ), a n d its grain. B o th p ro p e rtie s ow e th eir
m an ife sta tio n s w ithin a p a rtic u la r skeletal elem ent to the m icro stru c tu re o f
bo ne tissue a n d its v a ria tio n across the skeletal elem ent. T he m o rp h o m e try o f a
skeletal elem ent h as been discussed in only g eneral term s in this section, b u t it is
clear (C h a p te r 8) th a t this v ariab le to o is im p o rta n t in term s o f w h eth er o r n o t a
p a rtic u la r skeletal elem ent ca n be m ade in to a usable tool. A s well, the
m o rp h o m e try o f a b o n e will influence how readily it is fluvially tra n s p o rte d and
w h eth er o r n o t the bo n e will m ove d o w n slo p e d u e to g ravity (C h a p te r 6). B one
den sity in p a rtic u la r is considered in som e d etail in C h a p te r 7, a n d the influence
o f b o n e m ic ro stru c tu re on how a b o n e fractu re s is considered in C h a p te r 8. T he
influence o f bone m ic ro stru c tu re o n diagenetic processes is considered in
C h a p te r 11. T he discussion in this section h as been superficial, b u t th a t is in p a rt
because, w ith the excep tion o f fractu re m echanics, the influence o f bo n e tissue
m icro stru c tu re in p a rtic u la r o n ta p h o n o m ic processes h a s n o t been intensively
explored. I hav e in d icated w hy the m ic ro stru c tu re o f all skeletal tissues should
be so explored.
Vertebrate skeletons
T he skeleton’s obvious m echanical function is to serve as the b o d y ’s scaffolding.

(H. F rancillon-V ieillot et al. 1990:473)
A vertebrate is, sim ply, a n an im al w ith a b ack b o n e. T he tax o n o m ic phylum

C h o rd a ta includes anim als th a t have, eith er th ro u g h o u t th eir life o r a t som e
stage in th eir dev elo pm ent, an in te rn a l su p p o rtin g rod. T h a t ro d is c o n tin u o u s
d ow n m uch o f the length o f the body along the d o rsal m idline. If it consists o f
in divid ual v erteb ral elem ents, th e an im al is classified as a m em b er o f the
v erte b rate su b p h y lu m (H ild e b ra n d 1974; R o m er an d P arso n s 1977).
T here is a g reat deal o f v a ria tio n in the skeletons o f verteb rates. T h a t
v aria tio n is fo u n d n o t only in the kinds o f tissues m ak in g up the skeletons, but
in th e a c tu a l stru c tu re o f the skeleton. A bird skeleton is n o t nearly the sam e as,
for in stance, the skeleto n o f a reptile a lth o u g h the tw o are phylogenetically
related. F igu res 4.7 th ro u g h 4.13 p resen t generalized sketches o f various
v erte b rate skeletons. In follow ing c h a p te rs I presum e m u ch o f w h a t is show n in
these d raw in gs is well k n o w n to the reader. T he general d irectio n al term s used
to describe skeletal p a rts are illu stra te d in F ig u re 4.14.
A skeleta l elem ent (o r bone o r to o th ) is an “ a n a to m ica l o rg a n ” (F ran cillo n -
V ieillot et al. 1990:480). S keletal elem ents m ak in g up v e rte b ra te skeletons can
be divided in to tw o basic categories: axial a n d ap p e n d ic u lar. A x ia l elem ents are
fo u n d n ea r the m idline o f the to rso , they m ay be single o r p aired, a n d m o st tend
to be b ilaterally sym m etrical. T hese include the c ran iu m (single), m andibles
(paired), v erteb rae (single), ribs (paired), a n d stern u m (single). Appendicular
elem ents are o ften said to be paired; th a t is, th ere are discrete left a n d right,
b ilaterally sym m etrical instances o f each. T hese include the lim b bones, the
TH OR AC IC
LUMBAR
SACRUM CERVICAL
CAUDAL £
SCAPULA
FEMUR
SKULL
PATELLA
M A N D IBLE
STERNUM HUMERUS
ULNA
R AD IU S
TARSAL CARPAL
.ME TA T AR S A L
METACARPAL
1st P H A L A N X -----
-2nd P H A L A N X —
— 3 rd P H A L A N X
Figure 4.7. N o rth A m erican bison (Bison bison) skeleton, show ing locations o f
m ajo r skeletal elem ents.
Figure 4.8. G eneralized leporid o r ra b b it skeleton, show ing locations o f m ajor

skeletal elem ents.
S tructu re and quantification o f vertebrate skeletons 89
N om enclature O f G en era lized Teleost F ish S keleto n
second dorsal f i n f i r s t dorsal f i n

anterior
f t f i n rays dorsal
pterygiophores
caudal f i n
neural
ultim a te vert series
caudal
vertebra -fr o n ta l
lacrym al
prem axilla
-maxilla
' haem al spine'
p leu ra l
branched soft-rays so ft f i n rays ribs
f ®T quadrate
a rtic u la r
penultim ate vert. o p e rcu la r\p re o p e rcu la r
pterygiophores pectoral f i n j j
anal f in subopercular
tru n k vert. cleith ru m /
a n terio r an al spine radials
Figure 4.9. G eneralized teleost fish skeleton, show ing locations o f m ajo r skeletal
elem ents. R eproduced by perm ission from : O lsen, S. J. Fish, amphibian and
reptile remains fr o m archaeological sites. P eabody M useum Papers, vol. 56, no. 2,
Figure 3; C o pyright 1968 by the President an d Fellow s o f H arv ard College.
b ones o f th e p ec to ral (shoulder) girdle (clavicle if p resent, scapula, hum erus),

a n d the b ones o f th e pelvic girdle (in n o m in ate, co nsisting o f the pubis, ilium ,
ischium ). In m am m als, “ b lo o d -fo rm in g red m arro w [is] fo u n d w ithin the flat
bones o f th e skull, stern u m a n d ribs an d in the cancellous extrem ities o f the long
bones. Y ellow m arro w , m ostly fat, occupies the in terio rs [m edullary cavity] o f
the long b o n e d iap h y ses” (M a c G re g o r 1985:9; see also C u rre y 1984:104-105).
V e rteb ra te skeletons in general have fo u r lim bs (F ig u res 4.7 a n d 4.8), w ith
the ex ceptio n o f fish w hich have fins instead o f lim bs (F ig u re 4.9). S nakes have
no lim bs (F ig u re 4.12) b u t som e species have skeletal vestiges o f the lim bs their
rep tilian an cesto rs once sp o rted . T u rtles a n d to rto ises are unique fo r the shells
they ca rry (F ig ure 4.11). In birds the forelim bs are m odified fo r flight, bones are
th in-w alled to decrease w eight, a n d the stern u m is enlarged to su p p o rt the flight
m uscles (F ig u re 4.13).
M o st ta p h o n o m ic research has focused on w ith in -sk eleto n v aria tio n in the
effects o f ta p h o n o m ic agents a n d processes. T h a t is, an aly sts have stu d ied how
bovid fem ora re sp o n d to p a rtic u la r ta p h o n o m ic processes a n d c o m p are these
responses to th o se o f bovid scapulae to the sam e processes. M u ch less w o rk has
been d ev o ted to co m p ariso n s of, say, the responses o f a b ird h u m eru s a n d a dog
h u m eru s to a p a rtic u la r ta p h o n o m ic process. Som e research along these lines is
beginning to a p p e a r (e.g., C h am b ers 1992; K re u tzer 1992; L ym an eta l. 1992a),
parasphenoid ,premaxilla nasal terminal phalanx
— maxilla — pTootic phalanges

—ethmoid—
netacarpal
J r onto- parietal-
squamosal— radiale-
quadrate cartilage I qltadraiojugal columella centrale

•ulnare
exoccipital atlas-
P, . transverse processvertebrae
mento-meckelian bones 1
sacrum \humerus
dentary
.urosi
remur
tibio-Jibula
astragalus
D calcaneum.
cent rale
episternum
tarsalia- -prehallux
scapula \ ,n
r \ clavicle
■phalanges
mstatarsai
coracoid
ternum
suprascapula
terminal phalanx
xiphisternum
Figure 4.10. G eneralized frog (am phibian) skeleton, show ing locations o f m ajor
skeletal elem ents. R eproduced by perm ission from : Olsen, S. J. Fish, amphibian
and reptile remains fr o m archaeological sites. P eabody M useum Papers, vol. 56,
no. 2, Figure 10; C o p yright 1968 by the P resident and Fellows o f H arv ard
College.
prem axilla
sym physis palatine
-prefrontal
quadratojugal
splenial postjrontal
* dentary prootic
squamosal
articular
phalanges paroccipital
metacarpals supraoccipital
radiale plus centrale
j,Jb& fci H carpalia
intermedium y j \ ( \ ln a r e
Figure 4.11. G eneralized tu rtle (reptile) skeleton, show ing locations o f m ajor
skeletal elem ents. R eproduced by perm ission from: Olsen, S. J. Fish, amphibian
no. 2, Figure 11; C op yright 1968 by the President an d Fellows o f H arv ard
College.
prem axilla rootic splenial coronoid

septomaz. & vomer palatine
prefrontals, dentary D angular

fr o n ta l -
4 - basisphenoid ex occipital
supraorbital' articular
basioccipital
p o stfro n ta l squamosal ^
transversum quadrate
supraoccipital
a tla s '
epistropheus
pa rieta l pterygoid
cervical vertebrae
prefrontal^ | squamosal j quadrate
n asa ls
surangular
& articular
ectopterygoid
/ poison fa n g s
dentary
dorsal vertebrae -
caudal vert.
spinous process
prezygopophy:
centrum \/
inferior lam ella \ / postzygopophysis
haem al process
G H I
Figure 4.12. G eneralized snake (reptile) skeleton, show ing locations o f m ajor
skeletal elem ents. R eproduced by perm ission from : O lsen, S. J. Fish, amphibian
no. 2, Figure 12; C o p yright 1968 by the P resident an d Fellow s o f H arv ard
College.
S tru cture and quantification o f vertebrate skeletons 93
a tla s
Figure 4.13. G eneralized bird skeleton, show ing locations o f m ajo r skeletal
elem ents. R eproduced by perm ission from : Olsen, S. J. Fish, amphibian and
reptile remains fr o m archaeological sites. P eabody M useum Papers, vol. 56, no. 2,
A ppendix Figure 3; C o p y rig h t 1968 by the P resident an d Fellow s o f H arv ard
College.
V ertebrate taphonom y
Anterior
Cranial
T ransverse
Figure 4.14. D irectional term s for v ertebrate skeletons. R eproduced by

perm ission from : Hesse, B. an d W apnish, P. A nim al bone archeology, Figure 30.
W ashington, D .C .: T arax acu m Press. C o p yright 1985 by B. H esse an d P.
W apnish.
bone r e s o r p t i o n
bone g r o w t h bone r e g e n e r a t io n
>
allom etry
RECOVERY
BIRTH Onto geny Matu re DEATH TAPHONOnV FROM
FOSSIL
ANTEMORTEM P E R IM O R T E M POSTMORTEM RECORD
Figure 4.15. C hronological relatio n s o f bone ontogeny, bone rem odeling, death,
and taph o n o m y . Tim e passes from left to right.
b u t is th u s fa r restricted to different m a m m alian taxa. G iven th e m ed iatin g

influences o f m icro stru c tu ra l p ro p e rtie s o f skeletal tissues on ta p h o n o m ic
processes, an d th e m a jo r m a c ro stru c tu ra l a n d m ic ro stru c tu ra l v aria tio n
betw een fish, b ird , reptile, a m p h ib ian , a n d m am m al skeletons a n d skeletal
tissues, such research seem s necessary to the c o n tin u ed g ro w th o f o u r ta p h o
n o m ic know ledge.
Modification o f skeletal tissues and time o f death

T he w o rd skeleton is ta k e n fro m the G reek skeletos w hich m eans “ d ried u p ” or
“ w ith ere d .” T a p h o n o m ists are co n cerned w ith the c o n d itio n s of, a n d the
m o difications to, skeletons, skeletal elem ents, a n d skeletal tissues. T h a t is, they
id entify a n d stu d y such co n d itio n s a n d m o d ificatio n s based on m odels o f the
skeletons o f living org anism s, individual skeletal elem ents o f living organism s,
a n d skeletal tissues o f living organism s. D o c u m en tin g a n d explaining differ
ences betw een th e living a n d the fossil ap p e ara n ce o f skeletons, elem ents, an d
tissues is the subject o f ta p h o n o m ic inquiry. B ut in using a m odel o f a specific
living skeleto n as th e co m p arativ e baseline, the analyst m ust co n sid er th a t the
m odel is based on a n o rm o r average. S keletal elem ents a n d skeletal tissues in
p a rtic u la r can be m odified fro m the n o rm d u rin g the lifetim e o f a n organism ;
such m od ification s occu r before d e a th a n d are k n o w n as antem ortem m odifica
tions. G iven th e general consensus th a t ta p h o n o m ic histories begin w ith the
d ea th o f o rg an ism s, it m ust be realized th a t tap h o n o m ists tend to be interested
n o t in a n te m o rte m m o d ifications b u t ra th e r m odifications th a t o ccur at the
tim e o f d e a th o r p erim ortem m odifications, a n d m odifications th a t o cc u r after
d e a th o r p o stm o rtem m o d ifications (F igure 4.15).
D ep en ding on the research qu estio n s being asked, arch aeo lo g ists m ay be
interested in a n tem o rte m dam ages, such as w hen congenital defects o r injuries
to an o rg an ism are stud ied (e.g., B aker a n d B rothw ell 1980:82-95). If h u m an
w arfare o r h u m an h u n tin g practices are the subject o f interest, th en p e rim o r
tem m od ification s m ay be o f in terest (e.g., W ilson 1901, a n d N o e -N y g aa rd
1974, 1975a, 1975b, respectively). Sim ilarly, stu d y o f arch aeo lo g ical traces o f
h u m a n can n ib alism tend s to focus on p erim o rte m a n d p o stm o rte m m odifica

tions o f h u m an rem ains (e.g., F lin n et al. 1976; C. G . T u rn e r 1983; W hite 1992),
alth o u g h here in p a rtic u la r the line sep a ratin g the tim e o f p erim o rtem from
p o stm o rtem m odifications m ay be unclear. M odifications to skeletons th a t
result fro m processing o f freshly tak en prey fo r im m ed iate c o n su m p tio n m ay
be difficult to disting u ish from m odifications th a t occur, say, a m o n th o r m ore
a fte r d e a th , such as w hen prey are cached o r sto red fo r later co n su m p tio n .
D istin ctio n o f p erim o rte m fro m p o stm o rte m m o d ificatio n is p erh ap s less
critical th a n disting u ish in g these tw o fro m a n te m o rte m m odifications. T h a t is
so because ta p h o n o m ic histories begin w ith an im al d eath . O f course the cause
o f d e a th m ay be an im p o rta n t ta p h o n o m ic v ariab le (e.g.. W eigelt 1927/1989),
an d we re tu rn to it in C h a p te r 5. H ere, we m u st co n sid er how the analyst
identifies a n te m o rte m m odifications.
A n tem o rte m m odifications include readily identified p h e n o m e n a such as
healed fractures. O rtn e r a n d P u tsc h a r (1981:61) indicate th a t an tem o rte m
“ tra u m a tic , fatigue a n d p ath o lo g ical fractu res will vary in the healing p rocess.”
A tra u m a tic fractu re results in the ru p tu re o f b lo o d vessels in H a v ersian canals,
the perio steu m , an d the m arro w ; released blo o d form s a h e m ato m a a ro u n d the
fractu re. P erio steu m is strip p ed aw ay from the b o n e surface fo r a few
m illim eters ad jacen t to the fractu re site, an d this ap p e a rs to in itiate the
fo rm a tio n o f callus. T he h e m a to m a (b lo o d clot) is p erm e ated by fibrous
connective tissue w hich m ay be the source o f the fibrous callus th a t unites the
fractu re surfaces. T he u n m ineralized fibrous callus p rovides the m atrix fo r the
fo rm a tio n o f w oven bo n e a n d th e p rim ary bo n e callus. T he w oven bone is
ev en tu ally replaced w ith lam ellar bone. H ealed fractu res m ay be m ark e d by
excessive b o n y tissue a t the fra c tu re site, a n d n o n -n o rm a l alig n m en t o f the tw o
pieces (O rtn e r a n d P u tsc h a r 1981:61-64).
O th e r a n te m o rte m m odifications, such as a b n o rm a l skeletal developm ent
a n d disease o r m a ln u tritio n in d u ced m odifications, are, like p erim o rte m an d
p o stm o rte m m odifications, identified on the basis o f m odels o f n o rm al or
average skeletons, elem ents, a n d tissues. M an y a n tem o rte m m odifications are
identified on the basis o f the presence o f a b n o rm a l bo n e g row th o r re so rp tio n
(B aker a n d B rothw ell 1980; O rtn e r a n d P u tsc h a r 1981). In o rd e r fo r either to
occur, th e skeletal tissue m u st be alive, in d icatin g the o rg an ism w as alive. T here
is m o re re so rp tio n a n d less re g en eratio n im m ediately a fte r a w o u n d to a bone is
inflicted, a n d as the w o u n d heals th ere is less re so rp tio n a n d m ore regeneration.
Resorption o ften creates pits in bone, bo n e regeneration (initial d ep o sitio n o f
w oven b o n e th a t is ev entually replaced by lam ellar bone) first occurs as tiny
no d ules o f new b o n e m aterial. H ealed w o u n d s to b ones display a relatively
sm o o th surface a n d ex tra bone tissue relative to the n o rm (after M orse 1983).
G a rla n d (1988:324) argues th a t m acro sco p ic in spection o f b ones “ gives little
insight in to the in tera ctio n s w hich have tak en place betw een bones a n d the
v ario u s chem ical a n d physical facto rs a n d biological ag en ts w ithin th e burial
e n v iro n m e n t.” H e suggests th a t m icroscopic in spection o f skeletal tissues,
especially h istological studies, can be im p o rta n t fo r d istinguishing p ath o lo g ical

co n d itio n s fro m ta p h o n o m ic effects (e.g., G a rla n d et al. 1988). T h a t is so
because o f the m icro sco pically distin ct responses o f skeletal tissues to disease
a n d to , in p a rtic u la r, p o st-d ep o sitio n al ta p h o n o m ic processes (C h a p te r 9). I
suspect G a rla n d is co rrect, a n d the research o f o th ers tends to confirm this
suspicion (e.g., H a c k e tt 1981; H a n so n a n d B uik stra 1987; P iepenbrink 1986).
T here is n o d o u b t an area o f m icroscopic ta p h o n o m ic research co n cerning
p aleo h isto lo g y a n d diagenetic processes th a t is, as yet, little explored.
T he precedin g overview focuses largely o n m am m alian tissues as th o se are
the best k n o w n a n d th e m o st typical in the fossil record. O th e r h a rd tissues th a t
m ight be o f co n cern to a v erte b rate ta p h o n o m is t such as the baleen o f w hales (a
k eratin o u s m aterial) have n o t been covered. C o n su lta tio n o f the references
cited a n d th eir respective references sh o u ld pro v id e a good sta rt to w ard s
p ro v id in g in fo rm a tio n on the m aterials n o t covered as well as fu rth e r details
a b o u t th e m ateria ls th a t have been reviewed.
Quantification
[In ratio n al o r w ell-reasoned quantification] the in vestigator adm its to his graphs,
so to speak, only item s o f evidence th a t are relevant to the p articu lar m a tte r under
investigation, an d th a t are as accurate as practicable, w ith the p ro b ab le lim its o f
sam pling and experim ental erro r expressed graphically.
(J. H. M ackin 1963:139)
A n an a ly st interested in ta p h o n o m ic issues h as an a d v a n ta g e over, say, a lithic

tech n o lo g ist o r som eone in terested in p re h isto ric ceram ics. T he ta p h o n o m ist
h as a m odel fro m w hich to w ork. T he ta p h o n o m ist know s, fo r instance,
m am m als have fo u r legs, the n u m b e r o f b o nes a n d teeth m ak in g up a single
sk eleto n o f a p a rtic u la r species (T able 4.1), a n d w h at each com plete bo n e and
to o th o f th a t skeleton lo o k ed like w hen th a t skeleton w as endow ed w ith
m uscles, a h e a rt, lungs, etc., a n d w as w alking a ro u n d . D o c u m e n ta tio n o f how
a n d w hy th e pile o f b ones a n d teeth lying o n th e la b o ra to ry table differ from
th a t m odel o f a living o rg an ism is the w o rk o f the ta p h o n o m ist. O ne o f the first
im p o rta n t steps in such d o c u m e n ta tio n involves m easu rin g how a b u n d a n t
p a rtic u la r skeletal p a rts a n d p a rtic u la r tax a are in a collection o f fau n al
rem ain s (G ray so n 1984). In this section, I define several key co n cep ts an d
term s, describe several o f the regularly used q u a n tita tiv e un its, a n d discuss
several basic co n cep ts o f q u an tifica tio n a n d m easu rem ent.
C oncepts
T he process o f m easu rem en t involves “ the ap p lica tio n o f a set o f p ro c ed u ra l
rules fo r co m p arin g sense im pressions w ith a scale a n d fo r assigning sym bols to
o b se rv a tio n s” (G ib b o n 1984:40). M easurem ents result fro m co m p arin g o b ser
v atio n s m ad e o n p h e n o m e n a w ith a scale acco rd in g to a set o f rules an d
T ab le 4.1 Frequencies o f m ajor kinds o f skeleta l elem ents in different

m am m alian taxa
Skeletal elem ent H omo B ovid/C ervid Equid Suid C anid Felid Castor Pinniped
cranium 1 1 1 1 1 1 1 1
m andible 1 2 2 2 2 2 2 2
atlas 1 1 1 1 1 1 1 1
axis 1 1 1 1 1 1 1 1
cervical (3-7) 5 5 5 5 5 5 5 5
thoracic 12 13 18 14-15 13 13 14 15
lum bar 5 6-7 6 6 -7 7 7 5 5
sacrum 3 1(4-5) 1(4-5) 1(5) 1(4) 1(3) 1(3) 1-4 1(3)
innom inate 2 2 2 2 2 2 2 2
rib 24 26 36 28-30 26 26 28 30
sternum 3 1(6) 1(6) 1(6-8) 1(6) 1(8) 1(8) 1(5) 1(8-9)
scapula 2 2 2 2 2 2 2 2
clavicle 2 0 0 0 0 2 2 0
hum erus 2 2 2 2 2 2 2 2
radius 2 2 2 2 2 2 2 2
ulna 2 2 2 2 2 2 2 2
carpal 16 12 14-16 16 14 14 14 14
m etacarpal 10 2 2 8 10 10 10 10
fem ur 2 2 2 2 2 2 2 2
patella 2 2 2 2 2 2 2 2
tibia 2 2 2 2 2 2 2 2
fibula 2 distal only proxim al 2 2 2 proxim al 2
only only
astragalus 2 2 2 2 2 2 2 2
calcaneum 2 2 2 2 2 2 2 2
o th er tarsals 8 6 8 10 10 12 12 10
m etatarsal 10 2 2 8 10 8 10 10
first phalanx 20 8 4 16 20 18 20 20
second phalanx 16 8 4 16 16 16 16 16
third phalanx 20 8 4 16 20 18 20 20
Note:
3 In m atu re individuals there is one sternum and one sacrum , m ade up o f the n u m b er o f
individual stern ab ra an d sacral vertebra indicated in parentheses.
assigning o ne o r m o re sym bols o r values to each o b serv atio n p er p h en o m en o n .

Q u a n titativ e units m ak e up the scale a n d are o f different levels o f m ath em atical
p o w er (S h en nan 1988:11-12; Stevens 1946), a n d o f different kinds (G ib b o n
1984:55). T h e level o f a q u a n tita tiv e u n it is d eterm in ed by rules o f o rd e rin g and
distance, an d is critical to d eterm in in g a p p ro p ria te statistical tests (see G ra y so n
1984 fo r relevant discussion reg ard in g zo o arch aeo lo g ical d ata). N om inal-level
m easures reco rd differences in kin d o r categ o ry w ith no o rd erin g o r d istance
betw een categories. O rdinal-level m easures reco rd ra n k o rd e rs o f m agnitude;
g re a te r-th a n a n d less-than relatio n s are specified, b u t n o t how m u ch g re ater or
less th a n . T here is a n o rd e r to b u t no specification o f d istance betw een

m easu rem en ts. Interval-level m easures reco rd an o rd e r a n d how m uch g reater-
th a n a n d how m uch less-than in term s o f fixed a n d equal units o f m easurem ent.
B oth o rd e r an d d istan ce betw een m easu rem en ts are know n. R atio-level m ea
sures are th e sam e as interv al scale m easures, b u t also have a n a tu ra l zero point.
N o m inal-level m easu res are som etim es described as q u alitativ e, discrete, a n d /
o r d isco n tin u o u s; o rd in al, in terv al, a n d ratio-level m easures are som etim es
ch aracterized as q u a n tita tiv e a n d /o r c o n tin u o u s.
Q u a n tita tiv e u n its can be o b serv atio n a l o r analytical. O bservational units are
em pirical m an ife sta tio n s th a t are easily observed p ro p e rtie s o f phen o m en a;
they are easily experienced w ith o n e ’s senses a n d ca n be directly m easured.
M easu rin g o b serv atio n a l units by tallying the frequency o f specim ens or
d eterm in in g the length o f ind iv id u al specim ens p ro d u c es fu n d a m en ta l m easure
m ents o f p ro p e rtie s o f specim ens. A n a lytica l units consist o f o b serv atio n a l u nits
th a t h ave been m odified, o ften m ath em atically , to reflect som e com plex,
indirectly ob servable p ro p e rty o f the p h en o m en a u n d er study. In archaeology
the p ro p e rty o f in tere st is indirectly o b serv ab le because the archaeological
reco rd is static, a n d th e p ro p e rty o f in tere st ten d s to include a d ynam ic process
th a t is believed to be som ehow related to th a t record.
A n aly tical u nits m ay tak e eith er o f tw o form s: derived, o r interpretive.
D erived units are m o re com plex th a n o b serv atio n a l u n its because they are
defined by som e specified m ath em atica l re la tio n betw een fu n d a m e n ta l m e a su r
em ents. D erived units in clude such things as ra tio s o f fu n d a m e n ta l m easu re
m ents, a n d req u ire an aly tical decisions ab o v e a n d b ey o n d the choices o f a scale
o f co m p ariso n a n d a rule set fo r assigning sym bols. M easu rem en ts o f derived
units p ro d u c e derived m easurem ents (G ib b o n 1984:55). D erived units, in the
follow ing, tend to have non-explicit, unclear, o r only w eakly established
relatio n s to th eo retical o r in terp retiv e co ncepts, a lth o u g h they m ay play a
significant role in co m p arativ e analyses. Interpretive units are very com plex
b ecause they are stru ctu re d to m easure som e a b s tra c t o r th eo retical concept,
a n d th e term s ap p lied to such u n its o ften include a n am e fo r them . M ea su re
m ents o f in terp retiv e u n its have been called fia t o r p r o x y m easurem ents
(G ib b o n 1984:55).
D erived m easu rem en ts, in m y view, are m ath em atica lly gen erated in the
h opes th a t som e h id d en p a tte rn w ithin the un its m easu red will be revealed; th a t
p a tte rn m ay o r m ay n o t be causally related to the p ro p e rty we w ish to m easure.
Sim ilarly, p ro x y m easu rem en ts are m ath em atica lly gen erated b u t there tends
to be som e re aso n to su p p o se the m easu red in terp retiv e u n its are causally
related to the p ro p e rty we seek to m easure. F o r exam ple, m easures o f an
in d iv id u a l’s so cioeconom ic sta tu s o ften include fu n d a m e n ta l m easures o f
a n n u a l incom e, o cc u p atio n , a n d the level o f ed u c atio n attain ed . O n one h an d ,
the typically stro n g co rre la tio n betw een incom e a n d statu s serves as an
em pirical g en eralizatio n w a rra n tin g th e co n clu sio n th a t m easures o f socio
eco n o m ic statu s ca n be o b tain ed by m ath em atica lly m a n ip u la tin g fu n d a m e n tal

m easu rem en ts o f incom e, o cc u p atio n , a n d the like, to p ro d u c e p roxy m easures
o f an in terp re tiv e u n it we m ight term a “ socioeconom ic statu s in d ex .” D erived
units, o n th e o th e r h a n d , are less closely allied w ith th eo retical concepts, an d
m ay sim ply be used to p ro d u c e derived m easures o f different sets o f p h en o m en a
one w ishes to co m p are. D erived u n its m ay a tta in the statu s o f in terp re tiv e units
as we learn m o re a b o u t th e ir th eo retical c o n n o ta tio n s, an d conversely,
in terp retiv e u n its m ay ch ange th eir statu s a n d becom e analy tical un its if
research suggests th eir suspected causal re la tio n to a d ynam ic process is less
stro n g th a n originally believed.
N I S P and M N I
N IS P a n d M N I are th e q u a n tita tiv e un its m o st co m m o n ly e n c o u n te red in the

z o o a rch aeo lo g ical lite ratu re , a n d they ten d to have generally agreed u p o n
m eanings; th a t is, the units they seek to m easure are clear (e.g., G ra y so n 1984;
K lein a n d C ru z-U rib e 1984; L ym an 1985a). N I S P is defined as the n u m b e r o f
identified specim ens p er tax o n ; it is an o b serv atio n a l unit. T he ta x o n can be a
subspecies, species, genus, fam ily, o r h ig h er tax o n o m ic category. M N I is
defined as th e m in im u m n u m b e r o f ind iv id u al anim als necessary to a c c o u n t for
som e an aly tically specified set o f identified fa u n al specim ens; it is a derived unit
because it m ay o r m ay n o t tak e inter-specim en v a ria tio n such as age, sex, o r size
in to acco u n t. It is im p o rta n t to u n d e rsta n d th a t M N I trad itio n ally m eans the
m in im um n u m b e r o f individual anim als necessary to ac co u n t for all the kinds
o f skeletal elem ents fo u n d in the skeleton o f a tax o n . the hum eri, the scapula,
the cervical v erteb rae, etc. A s we will see, M N I ca n m ean so m eth in g else if the
analytically specified set o f identified fau n al rem ains does n o t include all the
kind s o f skeletal elem ents in a skeleton.
In the definitions o f N IS P a n d M N I, the w o rd "id e n tifie d " m ust be clear.
T ypically, “ iden tified” m eans “ identified to ta x o n .” It can also m ean “ id en ti
fied to skeletal elem en t,” such as a h u m eru s, a tibia, a th o rac ic v erte b ra, o r a
v erteb ra. B ecause it is typically necessary to identify the skeletal elem ent p rio r
to iden tifyin g th e ta x o n rep resen ted by a specim en (L ym an 1979c), the m eaning
o f identified as “ identified to ta x o n ” usually also (im plicitly) entails the
m ean in g “ identified to skeletal elem en t.” T h a t is, the la tte r is o ften necessary to
th e fo rm er. T he fo rm e r is n o t, how ever, alw ays necessary to the latter.
It is critical to define explicitly w h a t is m e a n t by “ specim ens” a n d “ skeletal
elem en ts.” G ra y so n (1984:16) defined a specim en as “ a bone o r to o th , or
frag m en t th ereo f, w hile a n elem ent is a single com plete bo n e o r to o th in the
skeleton o f an a n im a l.” A specim en is an archaeologically discrete p h e n o m e n o
logical u n it, such as a com plete h u m eru s, a d istal h a lf o f a tibia, o r a m andible
w ith teeth in it. A skeletal elem ent is a discrete, n a tu ra l an a to m ica l u n it o f a
skeleton, such as a h u m eru s, a tibia, o r a to o th . Specim ens can, b u t need n o t be
skeletal elem ents a n d are o b serv atio n a l un its. S keletal elem ents are a n a to m ica l
un its th a t m ay be rep resen ted by frag m en ts o r w hole b ones a n d are re p re
sented, p artia lly o r com pletely, respectively, by specim ens. A com plete fem ur
recovered fro m a site is a specim en, an o b serv atio n a l u n it, a n d a skeletal
elem ent. A frag m en t o f a fem u r such as the distal en d is a specim en, an
o b serv atio n a l unit, a n d represents b u t is p hen o m en o lo g ically n o t, technically, a
skeletal elem ent.
T ap h o n o m ists h ave a skeletal m odel to w hich the o b serv atio n a l units
(specim ens) they stu d y can be related. T h a t m odel consists o f the individual,
an a to m ica lly discrete a n d v ario u sly a rtic u la te d skeletal elem ents m ak in g up
the skeleton. F o r exam ple, it is n o t u n u su al to find a tab le listing the frequencies
o f fo relim bs, fem o ra, p ro x im al tibiae, o r th o rac ic sections o f the vertebral
co lu m n in a zo o a rch a eo lo g ical rep o rt. T hese a n a to m ic a l categories are
fo u n d ed on the m odel o f a skeleton consisting o f discrete skeletal elem ents, and
the categ ories are o f varying scales o f inclusiveness o f the skeletal elem ents. T he
an a to m ica l categ o ry “ distal tib ia ” is less inclusive th a n the categ o ry “ h u m e r
u s” w hich in tu rn is less inclusive th a n the category “ th o racic section o f the
verteb ral c o lu m n .” O nly the second categ o ry is directly co m p arab le to an
an a to m ica lly discrete skeletal elem ent. T he first categ o ry includes som e
analytically specified p o rtio n o f an an a to m ica lly discrete skeletal elem ent
w hereas th e last includes several analytically specified a rtic u la te d b u t a n a to m i
cally discrete skeletal elem ents. T hese a n a to m ica l categories o f varying scales
o f inclusiveness can a n d o ften d o serve as the q u a n tita tiv e un its w ithin w hich
o b serv atio n a l u n its (specim ens) are tallied.
Som e a u th o rs use the term “ bo n e frag m en ts” o r ju st “ frag m en ts” w hen in
fact they m ean “ specim ens” as defined earlier. T h u s, w hen eith er o f the form er
tw o term s is used, w h a t o ften are included in the tallies are b o th frag m en ts o f
skeletal elem ents a n d com plete skeletal elem ents. “ S pecim ens” is a m ore
satisfacto ry term as it has n o c o n n o ta tio n a b o u t the kin d o f the p a rt o f the
skeleton being tallied (bone, to o th , h o rn ) o r a b o u t the an a to m ica l co m p lete
ness o f th e p a rt. T h e explicit d istin c tio n o f elem ents a n d specim ens is critical to
tap h o n o m ists co n cerned w ith m easu rin g the extent a n d in ten sity o f bone
frag m en tatio n because such m easures include N IS P :M N E ratio s, N IS P :M N I
ratio s, a n d the like (C h a p te r 8). If it is n o t clear w h a t a specim en is an d how it
m ight differ from a skeletal elem ent, ra tio s like these will n o t be replicable an d
th eir ta p h o n o m ic significance will be obscure.
B ecause specim ens, as defined above, are the fu n d a m e n ta l o b serv atio n a l
units o f zo o arch aeo lo g y , it sh o u ld be clear th a t the ten acity a n d identification
skills o f th e an aly st m ay influence N IS P m easures (W hite 1992). T he specim ens
I ca n identify to skeletal elem ent a n d ta x o n m ay be different from those
so m eone else can identify. W hile in ter-a n aly st v aria tio n in w h at is identifiable
(an d th u s co u n tab le ) h as n o t been stu d ied in d etail, I suspect th a t this source o f
v a ria tio n betw een an aly sts m ay be m in im al in a g re at m an y cases.
W hile M N I (as defined above) a n d N IS P are q u a n tita tiv e units th a t are basic
to m u ch o f zo o arch aeo lo g ical analysis, there are o th e r q u a n tita tiv e un its th a t
p lay m a jo r roles in m o d e rn ta p h o n o m ic analysis. W e tu rn to several o f these
now.
MNE
T he recen t focus in zo o a rch a eo lo g y on ta p h o n o m ic issues h as b ro u g h t w ith it a

shift fro m m easu rin g frequencies o f tax a (usually accom plished w ith N IS P an d
M N I) to m easu rin g, a m o n g o th e r things, frequencies o f p o rtio n s o f skeletons
o f ind iv id u al tax a. M an y analyses w ith such a focus use the q u a n tita tiv e unit
M N E , o r som e d eriv a tio n thereof. T he term M N E signifies the m inim um
n u m b e r o f a p a rtic u la r skeletal elem ent o r p o rtio n o f a tax o n , such as the
m in im u m n u m b e r o f bovid p ro x im a l h um eri, o r the m in im u m n u m b e r o f
caprin e th o rac ic sections o f th e v erte b ral colum n. A nalysts typically d ep en d on
the form o f d a ta p re sen tatio n to m ak e it clear th a t the M N E values they publish
are n o t necessarily the m inim um n u m b e r o f an a to m ica lly com plete skeletal
elem ents (a n alo g o u s to M N I), b u t ra th e r are o ften o f som e p o rtio n o f a skeletal
elem ent o r som e m u lti-skeletal elem ent p o rtio n o f a skeleton.
Because M N E is th e m in im u m n u m b e r o f skeletal p o rtio n s necessary to
ac co u n t fo r the specim ens rep resen tin g th a t p o rtio n , the sam e p ro b lem s plague
th e d eriv a tio n o f M N E values as plague the d eriv a tio n o f M N I values (see
G ra y so n 1984 for a discussion o f the latter). O ne m ay (e.g., B unn an d K roll
1988; H esse a n d W ap n ish 1985; P o tts 1988), o r m ay n o t (e.g.. B inford 1984b),
fo r in stance, tak e in to ac co u n t age, sex, size, o r even tax o n o m ic differences
betw een th e specim ens fo r w hich a m in im u m n u m b e r is desired. M N E is
th erefo re an analytical u n it ra th e r th a n an o b serv atio n a l unit. T he an a ly st uses
som e set o f criteria by w hich specim ens are considered to be in d ep en d e n t (each
o f tw o o r m o re specim ens represents a sep a rate case) o r in terd e p en d en t (tw o or
m ore specim ens rep resen t the sam e case). T hese criteria sh o u ld be, b u t seldom
are, explicit.
M N E h as becom e an im p o rta n t q u a n tita tiv e unit, reflecting the fact th a t
tap h o n o m ists are co n cerned w ith h ow a n d w hy archaeological fau n al rem ains
differ from the set o f skeletal elem ents m ak in g up a com plete skeleton. Each
m am m al, fo r exam ple, has tw o hum eri, tw o tib ia, a n d seven cervical vertebrae.
T h e relative a b u n d a n ces o f different kinds o f skeletal elem ents one observes in
an arch aeo log ical co llection can be c o m p ared to the relative ab u n d a n ces o f
skeletal elem ents in a com plete skeleton. E xplaining w hy archaeologically
observed relative frequencies o f skeletal elem ents differ fro m o r are sim ilar to
th o se in a com plete skeleton h as p ro v e n to be an im p o rta n t a n d fruitful
an aly tical step (C h a p te r 7). F u rth e r, N IS P :M N E ra tio s are useful fo r m easu r
ing th e degree o f fra g m e n ta tio n o f different skeletal elem ents (see d iscussion o f
F ig u re 8.11). It is critical, then, to co n sid er how M N E values are derived fro m a
set o f specim ens.
S tru cture a nd quantification o f vertebrate skeletons 103
T able 4.2 F L K Z in ja n th ro p u s bovid limb bone data (fro m Bunn 1986; Bunn
and K roll 1986)
Skeletal elem ent N IS P en dsa N IS P sh afts M N E ends M N E shafts M N E co m p
hum erus 30 58 19 20
radius 28 57 14 22
m etacarpal 21 32 15 16
fem ur 14 58 6(8b) 17b 22
tibia 20 128 11(15b) 21b 31
m etatarsal 24 28 15 16
Notes:
a n um ber o f identified specimens w ith one o r b o th ends; the latter are com plete
b from B unn an d K roll (1988:142); taxonom ic/size differences are accounted for in values in
parentheses.
M a re a n a n d S pencer (1991:649-650) m en tio n tw o w ays to derive M N E

values. O ne involves m easu rin g the p ercen tag e o f the com plete circum ference
rep resen ted by a lo n g -b o n e sh aft frag m en t, a n d th en sum m ing th o se p ercen t
ages fo r each m easu red p o rtio n o f a skeletal elem ent. T his is sim ilar to a m eth o d
described by K lein a n d C ru z-U rib e (1984:108) as re co rd in g the “ fractio n by
w hich an identifiable b o n e is rep resen ted [using] co m m o n a n d intuitively
obv io u s fractio n s (e.g., 0.25, 0.33, 0.5, 0.67) a n d n o t a tte m p tin g g reat preci
sio n .” T h e fractio n s are sum m ed to p ro d u c e a n M N E value fo r each skeletal
p o rtio n . F o r instance, if the an a ly st records one com plete p roxim al fem ur, a
frag m en t rep resen tin g one h a lf o f a p ro x im al fem ur, a n d a frag m en t re p resen t
ing o ne th ird o f a p ro x im al fem ur, th e sum o f th e fractio n s w ould be
(1.0 + 0.5 + 0.33 = ) 1.83, for a n M N E o f tw o p ro x im al fem ora. O n one h an d ,
M a re a n a n d S pen cer’s (1991) m eth o d seems to be relatively accu rate, b u t can
result in slight (an d p ro b a b ly statistically insignificant) overestim ates. C are
m u st be ta k e n w ith K lein a n d C ru z -U rib e ’s (1984) m eth o d . T h a t is so because if
th e th ree p ro x im al fe m o ra pieces n o te d ab o v e all include th e g re ater tro c h a n te r,
th en th e M N E is n o t tw o, it is in fact three. K lein a n d C ru z -U rib e ’s (1984)
m eth o d can , o f course, easily be m odified to a c co u n t fo r such o verlap o f
specim ens. In fact, th e second m eth o d m en tio n ed by M a re a n a n d Spencer
(1991:652) “ involves using the co m p u te r to co u n t the n u m b er o f p o rtio n s w ith
overlapping sections” (em phasis ad d ed ), b u t they do n o t describe this m ethod.
B unn a n d K ro ll (1986, 1988) describe three w ays to derive M N E values. T he
analy st m ay d eterm in e (1) the m inim um n u m b e r o f com plete lim b b o n e skeletal
elem ents necessary to ac co u n t fo r only the specim ens w ith one o r b o th a rtic u la r
ends, (2) th e m in im u m n u m b er o f com plete lim b bo n e skeletal elem ents
necessary to ac co u n t fo r only the specim ens o f lim b bo n e shafts (w ith o u t an
a rtic u la r end), an d , (3) the m in im u m n u m b e r o f com plete skeletal elem ents
necessary to a c co u n t fo r b o th the specim ens w ith one o r b o th artic u la r ends an d
the sh aft specim ens. T hese are labeled, respectively, the M N E e n d s, the
M N E sh a fts, an d the M N E c o m p in T ab le 4.2 an d , as B unn (1986, 1991)
em phasizes, these values can be different. W h a t is p erh ap s co nfusing is th a t all

th ree are labeled M N E in the published record, yet th ere are significant
differences betw een th e three sets o f values due to v a ria tio n in how they are
derived. T h e co n fu sio n is ex acerb ated by the fact th a t in all b u t one o f B u n n ’s
(1986, 1991; B unn a n d K ro ll 1986, 1988) published rep o rts on the F L K
Z injanthropus fau n al assem blage the M N E e n d s fo r fem ora is listed as 6 a n d for
tibiae is listed as 11 while in B unn an d K roll (1988) those values are easily
derived as 8 an d 15, respectively, from th eir d a ta tables. T he difference here
resides in the h igh er values resulting from m y d istinguishing the size class o f
tax a rep resen ted by the specim ens, a n d th u s m y in tro d u c in g yet a n o th e r kin d o f
M N E value, one th a t tak es in to ac co u n t the size o f the o rg an ism rep resen ted by
the bones.
M N E is a derived m easure. As w ith M N I, th ere are several w ays to derive
M N E values. T he im p o rta n t p o in t here is th a t, like M a re a n a n d S pencer (1991)
an d B unn a n d K ro ll (1986,1988), the analyst m u st be explicit a b o u t the criteria
used to derive M N E values w hen they are presented. T his will ensure th a t
co m p ariso n s o f o ne a n a ly st’s M N E values w ith a n o th e r’s M N E values is n o t a
co m p ariso n o f app les a n d oranges.
M N I an d M A U per skeleta l portion

B inford (e.g., 1984b) d em o n stra te d a n d p o p u lariz ed the use o f M A U as a
q u a n tita tiv e unit. T he h isto ry o f this q u a n tita tiv e unit is interesting. B inford
beg an his z o o a rch aeo lo g ical studies w ith a c h a p te r in a b o o k p ublished in 1977.
In th a t c h a p te r he p resen ted q u a n tita tiv e d a ta as “ M N I” values, o r “ the
m in im u m n u m b e r o f in d iv id u al anim als rep resen ted by each a n a to m ica l p a r t”
(B inford a n d B ertram 1977:79). W hile one m ight presum e, given this d escrip
tio n a n d B in fo rd ’s term in ology, th a t his M N I values are calculated like W h ite's
(see th e discussion o f M N I above, a n d especially the discussion in follow ing
p a ra g ra p h s), they are no t. B inford (and B ertram 1977:146) p o in ted o u t he was
no t in terested in th e q u a n tita tiv e unit signified by the tra d itio n a l (e.g.,
W h itean ) m eanin g o f M N I. b u t ra th e r in the survivorship o f different skeletal
p a rts (B inford a n d B ertram 1977) an d how h u m an s differentially dism em ber
an d tra n s p o rt carcass p o rtio n s (B inford 1978). T h erefo re, B inford (and
B ertram 1977:146; B inford 1978:70) divided the observed bone co u n t (M N E )
fo r each a n a to m ica l u n it (such as p ro x im al fem ur) by the n u m b e r o f tim es th a t
an a to m ica l u n it o ccurs in one com plete skeleton. H e w as, in effect, s ta n d a rd iz
ing th e observed frequencies o f all “ a n a to m ica l u n its” acco rd in g to th eir
frequency in one an im al in o rd e r to m o n ito r how m an y o f each o f the various
p o rtio n s o f carcasses w ere represented. B inford (1984b:50) m ade it clear th a t
his “ M N I” values p er skeletal p o rtio n w ere n o t the sam e as W h ite ’s M N I
values w hen he “ decided to reduce the am b ig u ity o f language by no longer
referrin g to an a to m ica l frequency co u n ts as M N Is ” a n d in tro d u ce d the term
M A U fo r his sta n d a rd iz e d frequencies o f skeletal p arts. M A U stan d s fo r the

m in im u m n u m b e r o f an im al u n its necessary to a c co u n t fo r th e specim ens in a
collection.
W h ite (e.g., 1953a) can be credited w ith p o p u lariz in g the tra d itio n a l
tech n iq u e o f calcu latin g M N I values in N o rth A m erica. H e w rote, “ the m eth o d
I have used is to sep a rate the m o st a b u n d a n t elem ent o f the species fo u n d into
rig h t a n d left co m p o n en ts an d use the g re ater n u m b e r as the u n it o f calcu la
tio n ” (W hite 1953a:397). T h a t is, how m an y ind iv id u al anim als are necessary
to a c c o u n t fo r th e com bined h u m eri, scapulae, m andibles, cervical vertebrae,
etc. in a collection o f rem ains rep resen tin g a tax o n . T he analytically specified
set o f fa u n al rem ains consists o f all o f the kinds o f skeletal elem ents included in
th e skeleton o f a tax o n . W h ite is, how ever, also describing a m eth o d for
deriving a n M N I value fo r each p aired skeletal p a rt; th a t is, how m an y
in d iv id ual an im als are necessary to a c c o u n t fo r the h um eri, fo r the scapulae, fo r
the m an d ib les, etc. T he M N I determ in ed on the basis o f the hum eri, then, m ay
n o t be th e sam e as the M N I determ in ed on the basis o f the scapulae. In this case
the specified set o f fa u n al rem ains consists o f som e analytically defined set o f
an ato m ically lim ited categories o f faunal rem ains, such as p ro x im al fem ora
an d cervical v erteb rae. T he la tte r M N I values, w h a t I call M N I per skeletal
portion values, are w h a t are o f im p o rtan ce here because o f th eir sim ilarity to
MAU.
W h ite som etim es listed b o th M N I values p er skeletal p o rtio n a n d the M N I
frequencies o f b o th lefts a n d rights o f p aired bones (W hite 1952b, 1953c, 1955,
1956), alth o u g h he did n o t consistently d o so (W hite 1952c, 1953b, 1954). H e
suggested th a t to “ divide [the to ta l M N E o f paired elem ents] by tw o w ould
in tro d u ce g reat e rro r because o f the possible differential d istrib u tio n o f the
kill” (W hite 1953a:397). T h a t is, W hite suspected som e significant w ithin-site
d istrib u tio n a l d a ta m ight be m ask ed by calcu latin g w h a t B inford calls M A U
values. F o r exam ple, W hite (1953c:59) w ro te “ in m o st o f the featu res in the sites
from w hich I h ave identified the b o n e the d iscrep an cy betw een the rig h t a n d left
elem ents o f th e lim b b ones w as to o g re at to be ac co u n ted for by accident o f
p re serv atio n o r sam pling. T his leads one to believe th a t studies o n the
d istrib u tio n o f th e kill m ig h t be p ro fitab le .” W hen co m p arin g tw o assem blages,
the an a ly st sho u ld lo o k “ fo r large discrepancies betw een the [frequencies of]
rig h t a n d left elem ents. Sm all discrepancies are n o t necessarily significant
because th ey m ig h t be d ue to the accidents o f p re serv atio n o r sam p lin g ” (W hite
1953c:61). W hite did n o t suggest how the an aly st m ig h t distinguish betw een
“ g re a t” a n d “ sm all” discrepancies, n o r did he stu d y the sp atial d istrib u tio n o f
left a n d rig h t elem ents in a site. In identifying such an a ly tical avenues, how ever,
W h ite w as clearly offering an arg u m e n t to ju stify h ow he calcu lated M N I per
skeletal p o rtio n values.
W hite believed h u n ters, b u tch ers, a n d co nsum ers m ight d istin g u ish betw een
the left a n d rig h t sides o f an an im al, a n d b u tch er, tra n s p o rt, a n d d istrib u te the
T ab le 4.3 Frequencies o f pronghorn antelope skeleta l portions fr o m site

39FA 83. C olum ns B and C fr o m W hite ( 1952b). C olumn D is the m a xim u m
o f either B or C . E = ( B + C ) Jr 2
Skeletal p art M N I Left M N I Right MNI MAU

(A) (B) (C) (D) (E)
m andible 18 19 19 18.5
pelvis 13 19 19 16
scapula 24 24 24 24
P hum erus 3 0 3 1.5
D hum erus 26 30 30 28
P radius 28 25 28 26.5
D radius 23 22 23 22.5
P ulna 23 22 23 22.5
P m etacarpal 27 11 27 19
P fem ur 11 6 11 8.5
D fem ur 6 10 10 8
P tibia 9 9 9 9
D tibia 19 31 31 25
P m etatarsal 22 15 22 18.5
tw o sides o f a large m am m al differentially, yet he did n o t a tte m p t to find

evidence fo r this in any o f the bo n e assem blages he studied. B in fo rd ’s (1978:70)
experience w ith th e N u n a m iu t suggested to him th a t “ h u n ters m ak e no such
d isc rim in a tio n ” betw een left a n d right sides o f large m am m al carcasses. This
su b sta n tia te d his belief th a t investigating how people differentially b utcher,
tra n s p o rt, a n d d istrib u te p o rtio n s o f prey carcasses d em an d e d a co u n tin g u n it
w hich ig n o red the d istin c tio n o f left a n d right elem ents, a n d focused on, say, the
n u m b er o f forelim bs versus the n u m b e r o f hindlim bs versus the n u m b er o f rib-
cages th a t w ere tra n sp o rte d . B inford th u s p ro p o se d the M A U q u a n tita tiv e unit
to fulfill this analytical fu nction. T he difference betw een how W hite chose to
co u n t carcass p o rtio n s a n d how B inford chose to co u n t carcass p o rtio n s is n o t a
trivial d istinctio n. If W hite is co rrec t in his suspicion th a t carcasses m ay have
been differentially d istrib u te d based on the side o f the carcass, then calculating
M N I values o f b o th left an d rig h t elem ents a n d n o t dividing th eir sum by tw o
w ould be m o re a p p ro p ria te th a n deriving M A U values. M A U m asks such
v aria tio n , a n d M A U values can be easily derived from frequencies o f left and
right elem ents; sim ply sum the lefts a n d rights, an d divide by tw o. N o te th a t I
am n o t saying M A U is a p o o r q u a n tita tiv e unit; it is not.
B inford (1978, 1981b, 1984b; B inford an d B ertram 1977) typically n o rm s his
M A U per skeletal p o rtio n values to w h at are called % M A U values by dividing
all M A U values by the greatest M A U value in the assem blage. W hite (1952b,
1952c, 1953b) did n o t n o rm the M N I p er skeletal p o rtio n values in his early
p u b licatio n s, b u t did in his later ones (1953c, 1954, 1955, 1956) using a
CL CL
F igure 4.16. N o rm ed M N I p er skeletal p o rtio n frequencies an d no rm ed M A U

per skeletal p o rtio n frequencies fo r p ro n g h o rn antelo p e rem ains from 39FA 83
(from T able 4.3).
p ro c ed u re sim ilar to B in ford's. I suspect W hite n o rm ed frequencies to rem ove

the effects o f sam ple size w hen co m p arin g different sized assem blages. It was
n o t necessary to n o rm the values w hen he w as d escribing different assem blages,
b u t in his later p u b licatio n s c o m p ariso n o f assem blages is a m ajo r p a rt o f his
research. W h ite’s n o rm in g tech n iq u e w as used by o th e r w o rk ers fo r c o m p a ra
tive p u rp o ses as well (G ilb ert 1969;K ehoe a n d K ehoe 1960; W o o d 1962,1968).
W o o d (1962) in fact p ro d u c ed g ra p h s o f n o rm ed W h itea n M N I values m uch
like the g ra p h s used by B inford (1978, 1981b, 1984b; B inford a n d B ertram
1977), except B inford g rap h ed M A U values. Such a g ra p h , using b o th W hitean
M N I values an d B in fo rd ian M A U values, is presented in F ig u re 4.16. D a ta for
th a t g ra p h are based on the p ro n g h o rn an telo p e (A ntilocapra americana)
rem ains recovered fro m site 39FA 83 a n d re p o rted by W hite (1952b) (T able
4.3).
T h e g ra p h in F ig u re 4.16 is o f h isto rical in tere st an d , m o re im p o rta n tly , it
un d ersco res the difference betw een the M N I a n d M A U q u a n tita tiv e units
w hen th ey are used to m easu re frequencies o f skeletal p o rtio n s. F o r exam ple,
using the d a ta from 39F A 83, F igure 4.17 show s th a t the M A U values for
skeletal p o rtio n s ten d to be less th a n the M N I values fo r skeletal p o rtio n s (all
MNI (x)
Figure 4.17. B ivariate scatterp lo t o f M N I p er skeletal p o rtio n frequencies and
M A U per skeletal p o rtio n frequencies for p ro n g h o rn an telope rem ains from
39FA 83 (from T able 4.3). Low er line is the simple, best-fit regression line; upper
line is diagonal (origin o f 0, slope o f 1).
p lo tte d p o in ts fall below the d iag o n al line). T his is the p red ictab le resu lt w hen
the frequencies o f left a n d right elem ents differ. Interestingly, the sim ple, best-
fit regression line th ro u g h the p o in t sca tte r ( y = —0.2931 + 0.9018x; r = 0.96,
P = 0.0001) h as a slope < 1 a n d suggests th a t as frequencies o f skeletal p o rtio n s
increase the difference betw een M N I a n d M A U values increases.
F ig u re 4.18 illu strates the differences betw een frequencies o f left (to tal = 252)
an d right (to tal = 243) elem ents o f p ro n g h o rn an telo p e in the 39FA 83 collec
tion. T he sim ple, best-fit regression line (y = 2.7837 + 0.8096x; r = 0.72, P
< 0 .0 1 ) has a slope < 1, suggesting increasingly g reater differences betw een the
frequencies o f left a n d rig ht elem ents as frequencies o f elem ents increase. I f left
an d rig h t elem ents w ere consistently o f equal o r near-eq u al frequencies, the
best-fit line w ould be a d iag o n al line (y = 0 + 1x). H erein lies one w ay to search
an aly tically for w hat W hite c h aracterized as “ discrepancies” in the frequencies
o f left a n d righ t elem ents. If p o in ts ab o v e the d iag o n al line in F igure 4.18
represen t b ones fro m one arch aeo lo g ical co n tex t, an d p o in ts below the
d iag o n al line rep resen t b ones fro m a n o th e r co n tex t, th e n in tra-site differential
d istrib u tio n o f th e kill possibly occurred. But. are the differences betw een the
frequencies o f left a n d right elem ents significant, an d if so, are such differences
fou n d fo r all p aired elem ents?
T o ad d ress the preced ing q u estio n , I calculated ad ju sted residuals fo r each
categ o ry o f skeletal p a rt (T able 4.4; see E veritt 1977 fo r a d escrip tio n o f the
p ro ced u re; ad ju sted residuals are read as s ta n d a rd n o rm al deviates). T hey
suggest tw o, o r p erh ap s three o f the skeletal p a rts occur in a b u n d a n ces
T ab le 4.4 O bserved a nd e x p ected ( in parentheses) M N I frequencies o f

pronghorn antelope skeleta l portions fr o m site 39F A 83, and adjusted residuals
and probability values fo r each skeleta l portion
A djusted A djusted
Skeletal p art M N I Left residual P M N I Right residual P
m andible 18(18.8) - 0 .2 6 0.397 19(18.2) 0.28 0.390

pelvis 13(16.3) - 1.21 0.113 19(15.7) 1.22 0.111
scapula 24(24.4) - 0 .1 2 0.452 24(23.6) 0.12 0.452
P hum erus 3(1.5) 1.76 0.039 0(1.5) - 1 .7 6 0.039
D hum erus 26(28.5) - 0 .7 1 0.239 30(27.5) 0.73 0.233
P radius 28(27) 0.29 0.386 25(26) - 0 .3 0 0.382
D radius 23(22.9) 0.03 0.492 22(22.1) - 0 .0 3 0.488
P ulna 23(22.9) 0.03 0.492 22(22.1) - 0 .0 3 0.488
P m etacarpal 27(19.3) 2.61 0.004 11(18.7) - 2 .6 2 0.004
P fem ur 11(8.7) 1.13 0.129 6(8.3) - 1 .1 6 0.123
D fem ur 6(8.1) - 1.07 0.142 10(7.9) 1.09 0.138
P tibia 9(9.2) - 0 .1 0 0.460 9(8.8) 0.10 0.460
D tibia 19(25.5) - 1.95 0.026 31(24.5) 1.98 0.024
P m etatarsal 22(18.8) 1.09 0.138 15(18.2) - 1.10 0.136
LEFT
Figure 4.18. B ivariate scatterp lo t o f M N I per skeletal p o rtio n frequencies for left
and right skeletal p o rtio n s o f p ro n g h o rn an telope from 39FA 83 (from T able 4.3).
D iagonal line has a 0.0 origin on the y axis; simple, best-fit regression line has an
origin o f 2.78.
significantly different fro m th a t expected given ra n d o m chance. T here are m ore

left p ro x im al m etac arp a ls a n d m o re rig h t distal tib iae (and few er right
p ro x im al m e ta c a rp a ls a n d few er left d istal tibiae) th a n chance alo n e allows.
T here m ay also be m o re left p ro x im al hum eri (and few er right p roxim al
h u m eri) th a n chance alon e allow s, b u t th ere are so few o f this skeletal p o rtio n
th a t I w o rry th a t sam ple size m ay be influencing th e result. R egardless, it is clear
th a t in th is sam ple the an a ly st m ay w a n t to exam ine the asso c ia tio n a l co n tex ts
o f p ro x im al m etac arp a ls an d distal tibiae to determ ine if, as W hite suggested,
th ere is evidence fo r differential d istrib u tio n o f the kill. A n d , analysis o f
ad ju sted residuals p ro v id es a n o th e r w ay to search analytically fo r W h ite’s
“ large d iscrep an cies” in frequencies o f left a n d rig h t elem ents.
T h e im p o rtan ce o f co m p arin g frequencies o f skeletal p o rtio n s as m easu red
by B in fo rd ’s M A U a n d W h ite’s M N I is th a t the tw o units m easure different
p ro p e rtie s o f a b o n e collection. T h u s, c o rrelatio n s o f skeletal p a rt frequencies
w ith eith er B in fo rd ’s (1978) econom ic u tility indices o r a m easu re o f the
survival p o ten tial o f skeletal p a rts (e.g., L ym an 1984a) m ay be significantly
influenced by the co u n tin g u n it used (see C h a p te r 7). A gain tu rn in g to the
39FA 83 d a ta , we find the follow ing S p e a rm a n ’s rh o co rre la tio n coefficients:
Q uantitativ e unit C arib o u M G U I Bone density
M N I rs = 0.73 -0 .5 1
P= 0.008 0.06
M A U rs = 0.64 - 0 .5 3
P= 0.02 0.053
This set o f coefficients follow s the tra d itio n o f using o rd in a l scale statistics
w hen co m p arin g b o n e frequencies w ith a utility index o r th e stru c tu ra l density
o f a b o n e p a r t (see C h a p te r 7 fo r m o re com plete discussion). W hile the
coefficients an d p ro b a b ility values d o n o t differ greatly w h eth er one uses the
M N I values o r the M A U values, v a ria tio n betw een the coefficients u n d ersco res
the fact th a t the q u a n tita tiv e unit used can influence these kinds o f statistical
results. C learly, M A U is the unit o f choice fo r b o th c o rrelatio n s as it m ore
accu rately m easures the relative frequencies o f skeletal p a rts th a n M N I,
especially w hen differences betw een frequencies o f left a n d right elem ents are
great. T w o left hum eri rep resen t an M A U o f 1, b u t an M N I o f 2. W hen one is
in terested in d eterm in in g if frequencies o f skeletal p a rts are the result o f
differential tra n s p o rt o r differential d estru c tio n , the n u m b e r o f individual
an im als is irrelev an t; w h eth er m o re h u m eri o r m o re tib iae are rep resen ted is
p a ra m o u n t. A n d , because n o t all b ones o f the sk eleto n are p aire d , the
frequencies o f skeletal p a rts m u st be w eighted in o rd e r to assess accurately
w hich skeletal p a rts are a b u n d a n t a n d w hich are ra re, c o m p ared to their
relative a b u n d a n ces w ithin a com plete skeleton. W eighting is accom plished by
dividing th e observed freq uency o f each skeletal p a rt by the expected frequency
(the m ax im u m possible frequency if all w ere present).
S tructure and quantification o f vertebrate skeletons
D iscussion
B oth how q u a n tita tiv e units are defined a n d how they are o p eratio n aliz ed m ust
be explicit in o rd e r to ensure c o n c o rd an c e betw een the c o u n tin g u n its used an d
the research q u estio n add ressed w ith th o se units. T h e u ltim ate co n cern is th a t
the an aly st m ak es clear w h a t is being co u n ted , how it is being co u n ted , a n d w hy
specim ens are c o u n ted th a t w ay. P a rt o f the key to p ro d u c in g reliable a n d valid
q u a n tita tiv e m easures resides in the ac cu ra te definition o f ta rg e t p o p u la tio n s
a n d sam ple p o p u latio n s. Z o o arch ae o lo g ists in terested in d eterm in in g paleo-
en v iro n m en tal co n d itio n s fro m fa u n al rem ains require a m easure o f the fa u n a
th a t w as e x tan t a t th e tim e o f site o cc u p atio n w hereas zo o arch aeo lo g ists
in terested in d eterm in in g p re h isto ric h u m a n subsistence need to m easure the
fa u n a th a t w as killed o r h arv ested by h u m a n h u n ters (L ym an 1982a:337). A.
T u rn e r (1983:312-313) m ade this p o in t w hen he d istin g u ish ed betw een the
excav ated sam ple, the killed p o p u la tio n , a n d the living p o p u la tio n o f anim als.
T he first is the set o f faunal rem ains recovered by the arch aeo lo g ist from a site;
the second is the set o f anim als p ro c u red by the p re h isto ric o cc u p an ts o f the
site; the th ird is the fa u n a e x tan t at the tim e the site w as occupied. T u rn e r was
concern ed w ith estim atin g tax o n o m ic ab u n d a n ces w ithin e x tan t fa u n as (targ et
p o p u la tio n ) on the basis o f excavated sam ples (sam ple p o p u la tio n ), an d he
m ad e it a b u n d a n tly clear th a t the tw o w ere n o t necessarily c o rrelated , in p a rt
because the killed p o p u la tio n (archaeologically sam pled p o p u la tio n ) need n o t
be a ra n d o m sam ple o f the e x ta n t fauna.
B rew er (1992:207) d istinguishes a target population , the g ro u p o f things the
an aly st w ishes to m ak e inferences a b o u t, fro m a sam ple population, w h a t the
an aly st w o rk s w ith a n d w h a t serves as the basis o f o n e ’s inferences; the “ sam ple
p o p u la tio n m u st be relev ant to the ta rg e t p o p u la tio n , w hich in tu rn m u st be
defined by th e qu estio n s being ask e d ." T h a t is, th ere m u st be co n co rd an ce
betw een th e h y p o th esis being ev alu ated , the an aly tic techniques used, a n d the
co u n tin g u n its w hich are analyzed. It is, fo r exam ple, the lack o f co n c o rd an c e
betw een M N I m easures o f fossil tax o n o m ic ab u n d a n ces a n d tax o n o m ic
ab u n d a n ces in p reh isto rically e x ta n t fa u n as th a t has c o n trib u te d to this
q u a n tita tiv e u n it’s fall fro m analy tical favor. Likew ise, it is th e lack o f
co n c o rd an c e betw een N IS P m easures o f skeletal p a rt frequencies resulting
from differential fra g m e n ta tio n o f skeletal elem ents a n d actu al frequencies o f
skeletal p a rts th a t resu lted in th e in tro d u c tio n o f M N E . W e m u st be explicit
a b o u t w hy we have ch osen th e q u a n tita tiv e u n it we have used in o u r analyses
(L y m an 1982a:361). In o th e r w ords, we m ust specify how the quantitative units
we use to m easure the sam ple population relate to the quantitative properties o f
the target population we wish to infer. Such specification sh o u ld help us
d eterm in e if th e q u a n tita tiv e u n it we h av e chosen is the a p p ro p ria te one.
W h ich c o u n tin g u n it sh o u ld one use, a n d how should th a t u n it be o p e ra tio
nalized, in a p a rtic u la r situ atio n ? S om etim es it is clear (e.g., B adgley 1986a),

som etim es it is n o t. H ere is w here a m ajo r, tw o-step research effort should be
d irected, as it m ay n o t be clear w hich o f several u n its is the a p p ro p ria te one to
use. T h e first step is to specify explicitly the targ e t p o p u la tio n one is trying to
m easure. T h a t is, w h a t is the q u a n tita tiv e p ro p e rty o f interest? Such specifica
tion sh o u ld pro v ide clues to an a p p ro p ria te sam ple p o p u la tio n an d q u a n tita
tive u n its th a t give ac cu ra te m easu rem en t o f the targ e t p o p u la tio n ’s properties.
T his first step m ay be all th a t is necessary to a p a rtic u la r analytical problem . If it
is n o t, th e second step involves actualistic studies geared to w ard d eterm in in g
w hich o f th e analytical units (an d how it is actually co u n ted ) m o st clearly,
co n sistently , an d u n am b ig u o u sly reveals archaeologically detectab le p a tte rn s
th a t reflect th e p ro p e rty o f interest. Such research will allow som e analytical
units to be utilized as in terp retiv e units. B rain (1969), fo r exam ple, detected the
im p o rta n t u n it o f m easu rem en t now called % survival (or, % su rv iv o rsh ip )
d u rin g eth n o arch ae o lo g ic al research, a n d th a t u n it is regularly used to d a y to
help explain frequencies o f skeletal p o rtio n s in archaeological assem blages
(C h a p te r 7). S tiner (1991 b, 1991 e) likewise described several un its o f m easu re
m en t th a t allow th e id entification o f ta p h o n o m ic processes th a t have
influenced arch aeo lo g ical assem blages. B oth B rain a n d S tiner h ad p a rtic u la r
research q u estio n s in m ind, q u estio n s th a t suggested the kinds o f co u n tin g
units th a t logically sh o u ld be used. N e ith er o p erated in a strictly inductive or
blindly em pirical m o de, and, im p o rta n tly , they b o th w ere sufficiently clear
a b o u t h o w they defined th eir co u n tin g u n its a n d how they o p eratio n aliz ed
th em th a t o th ers can replicate a n d test th eir results, as well as ev alu ate the
a p p ro p riaten ess o f th o se un its fo r n .y s u r i n g specific pro p erties.
Summary
[An] ap p ro a ch to the p ractice o f tap h o n o m y is to en um erate an d then explain the
differences betw een fossil collections an d living com m unities o f anim als.
(A. Hill 1988:563)
A v e rte b ra te sk eleto n is a com plex entity. Its stru ctu re , fro m m icroscopic to
m acro sco pic levels, can have significant influences o n the effects tap h o n o m ic
processes h ave on its co n stitu e n t p a rts (the skeletal elem ents). F u rth e r, how
th ose c o n stitu en t p a rts are co u n ted d u rin g analysis is a com plex m atter. If
A n d rew H ill’s sta te m e n t q u o te d ab o v e is co rrec t, then clearly q u an tifica tio n is
im p o rta n t to ta p h o n o m ic analysis. H o w m an y a n d w hich skeletal elem ents o f
on to gen ically y o u n g ind iv id u als are p resen t in an assem blage? A re th o se values
d ifferent fro m the values observed fo r o ntogenically old individuals? H ave
an tlers a n d teeth been m odified by ta p h o n o m ic processes in such a m a n n e r as to
alter th eir a b u n d a n ces relative to the a b u n d a n ces o f lim b bones? W ere
verteb rae fluvially tra n s p o rte d b u t ca rp als n o t so tra n sp o rte d ? A re fem ora
b ro k e n b u t p h alan g es n o t bro k en ? A nsw ers to these a n d sim ilar q uestions
consist o f a large p a rt o f the ta p h o n o m ic d a ta researchers reco rd a n d analyze as

they try to u n ravel, u n d e rsta n d , a n d w rite ta p h o n o m ic h istories o f assem blages
o f v e rte b ra te rem ains. A nd, they all d ep en d o n q u a n tita tiv e d a ta o f one fo rm or
an o th er. W ith this c h a p te r as b a c k g ro u n d to the basic issues, it is, then, to
d escrib in g how one p ro d u ces answ ers to such qu estio n s th a t we tu rn in the
re m a in d er o f this volum e.
5
V E R T E B R A T E MO R T A L I T Y ,
SKELETONIZATION,
D I S A R T I C U L A T I O N , AND
SCATTERING
K lahn has m ade a sh arp distinction betw een the tw o m ain g roups o f causes o f death;
dying and being killed. By dying he m eans n o rm al d eath due to old age o r sickness.
By being killed he refers to vigorous individuals th a t becom e victim s o f accident,
enemies, o r the forces o f n ature.
(J. W eigelt 1927/1989:21)
Introduction
T a p h o n o m y is co n cern ed w ith the differences betw een w h a t the p aleo n to lo g ist
o r zo o a rch a eo lo g ist lays ou t in the la b o ra to ry fo r study, an d , variously, the
biotic co m m u n ity a n d /o r ind iv id u al anim als represented by th a t laid-out
m aterial. In a way, ta p h o n o m ic histories begin w ith th e d e a th o f an organism .
T his is n o t exactly true, a lth o u g h it is precise given m o st definitions o f
ta p h o n o m y (see C h a p te rs 1-3). It is n o t exactly tru e because the b ehavioral
p a tte rn s, ecological predilections, a n d life h isto ry o f an o rg an ism m ay influence
the m o d e o f m o rtality an d the ta p h o n o m y o f th a t o rg a n ism ’s carcass. As a
sim ple exam ple, terrestrial v erte b rates have different ta p h o n o m ic histories
th a n a q u a tic v erte b rates sim ply due to the different m edium in w hich they
n o rm ally die. K n o w in g so m eth in g a b o u t the behaviors, ecology, a n d lives o f
th e o rg an ism s w hose rem ains are being studied ca n th u s be a g re at benefit to the
tap h o n o m ist.
In this c h a p te r we explore the vario u s w ays anim als die a n d are killed, how
th o se m o des o f d e a th m ig h t influence su b seq u en t episodes in the ta p h o n o m ic
h isto ry o f a carcass, a n d som e a n aly tic techniques used to d eterm in e p re h isto ric
m odes o f d eath . W e also explore how m o rta lity influences age a n d sex
d em o g ra p h ic p a ra m e te rs in d icated by fossils. T here are m an y w ays to die.
In tim ately related to these m odes o f d e a th are w h eth er the rem ains being
stud ied re p resen t an active o r a passive ac cu m u latio n (C h a p te r 6), a synchronic
o r d iach ro n ic a c cu m u latio n (C h a p te r 6), the n u m b e r o f organism s represented,
a n d the d em o g rap h ic p ro p e rtie s o f th o se d ead organism s. In this c h a p te r I
in tro d u ce the subjects o f sk eleto n izatio n o f carcasses a n d scatterin g o f bones,
variab les th a t are in tim ately related to th e m o d e o f death.
114
M o rta lity, skeletonization, disarticulation, scattering 115
M odes o f death
A n individual anim al m ay die accidentally, o f old age, as a victim o f parasites or
oth er enemies, from lack o f food o r as a consequence o f external forces.
(W. Schafer 1962/1972:9)
C auses o f d e a th o f o rg anism s are m an y a n d varied, especially w hen co m p ared

to th e lim ited n u m b e r o f w ays the life o f a n org an ism m ig h t be initiated. T his is
because o f the accid en tal fa cto r in m o rtality . W eigelt (1927/1989) provides the
follow ing list o f acciden tal m odes o f death:
1. death due to volcanic activity;
2. death due to poisonous gases;
3. death due to fire;
4. death by drow ning;
5. death by becom ing m ired in m ud, quicksand, oil, o r asphalt;
6. death due to flooding;
7. death due to fluctuation in salinity o f w ater;
8. death due to drought;
9. death due to o v erpo p u latio n (m alnutrition);
10. death due to hunting (predation);
11. death due to freezing;
12. death due to falling th ro u g h ice.
O th er accidents include falling from high places an d intraspecific activities such

as co m p etitio n betw een m ales fo r breeding privileges leading to the d e a th o f
one o r b o th o f the c o m b a ta n ts o r th e d e a th o f a juvenile w ho go t in the w ay. All
o f these h ave a source ex tern al to th e o rg an ism th a t dies in c o n tra s t to disease o r
old age th e source o f w hich is relatively in tern a l to the organism .
B erger (1991:127) suggests th a t “ anim als w ith ‘n o rm a l’ lives rarely m ake
fossils [and thus] p re serv atio n o f the un lu ck y is the n o rm in p a le o n to lo g y .”
T ho se u n fo rtu n a te indiv iduals th a t die d u e to an accid en t are the ones th a t
typically experience a ta p h o n o m ic h isto ry conducive to p re serv atio n in c o n
tra s t to tho se th a t die o f old age. T his “ survival o f the u n lu ck y ” (B erger
1991:127) m ay be true, b u t it m ay also be trivial because it is u n u su al indeed
w hen all m em bers o f an age g ro u p die o f senility. In the exam ples described
below , typically less th a n h a lf o f a n age g ro u p die o f old age; m ore th a n h a lf die
o f o th e r causes. T h u s we have a b e tte r chance o f finding the rem ains o f an
o rg an ism th a t died accidentally th a n the rem ains o f a senile individual sim ply
because m o re o f th e fo rm er th a n o f the la tte r e n ter the fossil record.
The demography o f mortality

A m ong the tools developed by neoecologists, th a t o f the life table analysis m ay at
least confer u pon paleoecology a w elcom e ad d itio n to exact m ethodology.
(B. K u rten 1953:47)
M o des o f d e a th ca n be categorized as density d ep en d en t o r in d ep en d en t, age

d ep en d en t o r in d ep en d en t, a n d occasionally as sex d ep e n d en t o r in dependent.
T h e la tte r tw o o ften d ep en d on a n d h ave th e ir g re atest effect w hen different age

c o h o rts o r sex g ro u p s display different behaviors. M o rta lity facto rs such as
p re d a tio n , disease tran sm issio n , a n d co m p etitio n (p e rh ap s leading to m a ln u tri
tion) are d ep e n d en t on the density o f the p o p u la tio n being affected; th eir effects
on m o rta lity increase in m ag n itu d e as p o p u la tio n density increases. T he y oung
are th e ex p en dab le p a rt o f the p o p u la tio n , a n d the rates o f th eir survival an d
recru itm en t influence p o p u la tio n density. Such p o p u la tio n p a ra m e te rs o f
m o rtality lead us to co n sid er first the d em o g rap h ics o f m o rtality .
The basics
S tudy o f p o p u la tio n p a ra m e te rs o f fossil tax a has been u n d e rta k e n in detailed,
system atic fashio n fo r at least fo u r decades (e.g., K u rte n 1953, 1958; V oorhies
1969). T h eo retical aspects o f the d em o g ra p h ic characteristics o f verteb rates,
especially m am m als, are now ra th e r refined (B arlow 1984; C au g h ley 1966,
1977; C raig an d O ertel 1966; C zaplew ski et al. 1983; D eevey 1947; P olacheck
1985). As well, tech niq u es o f m o rta lity analysis are being applied to m o re an d
m ore v erte b rate tax a in p aleo n to lo g ical a n d arch aeo lo g ical settings, an d
m o rtality in e x tan t p o p u la tio n s c o n tin u es to be studied (B erger 1983; C oe et al.
1980; C rib b 1985, 1987; K o r th a n d E v an d e r 1986; L ym an 1989a, 1991a, 1991b;
S tiner 1990b, 1991a). In fact, a recent edited volum e is dev o ted specifically to
analy zing an d in terp re tin g v e rte b ra te m o rta lity p a tte rn s as evidenced by
a rch ae o fau n a s (S tiner 1991c).
A cohort is a g ro u p o f ind iv id u al o rganism s th a t w ere b o rn sim ultaneously
(C au g h ley 1977:85). C o h o rts can include all individuals b o rn d u rin g a p a rtic u
lar d ay, week, m o n th , season, or, typically, a so lar year. B ecause d e a th is
c o n tin u o u s th ro u g h o u t the existence o f a c o h o rt, regardless o f how old the
in divid u als are, th ere are alw ays few er indiv id u als in each succeedingly older
age class th a n in the im m ediately preceding age class; th a t is, using the typical
tim e u n it o f one year, w ith the passing o f each year there are few er individuals
still alive in a c o h o rt. R eal c o h o rts (all individuals b o rn in one year) are seldom
stu died by biologists o r paleo n to lo g ists; ra th e r, techniques o f d em o g rap h ic
analysis described here are usually applied to a p o p u la tio n o f indiv id u als o f all
ages as if th a t p o p u la tio n was m ade up o f the d em o g ra p h ic h isto ry o f a c o h o rt
from th e b irth o f th a t c o h o rt’s first ind iv id u al to the d e a th o f its last individual.
In th e follow ing, it is presum ed th a t a relatively ac cu ra te tech n iq u e for
assessing th e o n togenic age o f m o d ern an d fossil specim ens has been used; th a t
is, a tech n iq u e allow ing d e te rm in a tio n o f a n in d iv id u a l’s age a t d e a th w ith in a
m ax im u m ± 2 m o n th p erio d has been em ployed. A geing techniques are n o t
described here, b u t m an y such techniques exist a n d are d escribed in m an y o f the
references cited in this section.
A c o h o rt’s m o rta lity p a tte rn is typically a n d form ally p resen ted in th e fo rm
o f a “ life ta b le .” A n exam ple o f a life tab le fo r the H im alay a n th a r (H em itragus
T ab le 5.1 L ife table f o r fe m a le H im alayan thar (fro m C aughley 1966)
Age Frequency Survival M o rtality M o rtality Survival

rate rate
X f, I, dx 9* Px
0 205 1.000 0.533 0.533 0.467

1 96 0.467 0.006 0.013 0.987
2 94 0.461 0.028 0.061 0.939
3 89 0.433 0.046 0.106 0.894
4 79 0.387 0.056 0.145 0.855
5 68 0.331 0.062 0.187 0.813
6 55 0.269 0.060 0.223 0.777
7 43 0.209 0.054 0.258 0.742
8 32 0.155 0.046 0.297 0.703
9 22 0.109 0.036 0.330 0.670
10 15 0.073 0.026 0.356 0.644
11 10 0.047 0.018 0.382 0.618
12 6 0.029
jem la h icu s) is given in T ab le 5.1 (from C aughley 1966). T h e tab le is co n stru c te d

as if the in d iv idu al anim als w ere b o rn o n the sam e day a n d the n u m bers
surviving to each su b seq u en t b irth d a y w ere recorded. T ab le 5.1 follow s
biological tra d itio n a n d lists only the fem ales, as they are the source o f new
in div id u als in the p o p u latio n ; p aleo n to lo g ists typically list all ageable indivi
d u als regardless o f sex. By colum n, x is the age o f the c o h o rt by one-year
intervals. T he n u m b er surviving in each successive y ear is listed u n d e r f x. T he
p ro b a b ility th a t an ind ividual o f a p a rtic u la r age will survive to its next
b irth d a y is given u n d e r lx; this value is calcu lated by dividing the respective f x
value fo r an age g ro u p by the original size o f the c o h o rt, in this case 205. T he
fo u rth co lu m n, dx, lists the p ro b a b ility o f dying d u rin g the age interval x to
x + 1. T h u s the p ro b a b ility th a t a n ew b o rn th a r in this c o h o rt will die before
reach in g its first b irth d a y is calculated as the p ro b a b ility th a t it will survive
w hen b o rn (1.000) m inus the p ro b a b ility o f survival w hen it is one year old
(0.467), o r 1.000 —0.467 = 0.533. By co n v en tio n dx values are p lo tte d on the
row fo r th e b eg inning o f the age interval. T he ra te o f m o rtality , qx, is calculated
as dx/lx a n d is the p ro p o rtio n o f anim als alive at age x th a t die before reaching
age x + 1. T he final co lu m n lists survival rates, sym bolized as p x. T hese are the
p ro p o rtio n o f an im als alive at age x th a t survive to age x + 1 a n d are the
co m p lim en ts o f th eir respective qx values, o r 1 —qx.
Life tables have seldom been pu b lish ed fo r a rc h a e o fa u n a l assem blages,
p erh ap s because th eir value to ta p h o n o m ic analysis is n o t evident. K o ik e an d
O h taish i (1985) a n d L y m an (1987c) in d ep en d en tly show ed th a t the d ata
c o n ta in e d in a life tab le ca n be useful fo r detecting p re d a to ry pressure on a
p o p u la tio n a n d re su lta n t changes in its d em o g ra p h ic stru ctu re . K n o w in g th a t a
ta x o n ’s b eh a v io r is age specific, such as juveniles fo rm in g sep a rate g ro u p s aw ay

from ad u lts, th e d em o g rap h ic stru c tu re o f a fossil p o p u la tio n o f th a t tax o n
m ig h t be ac co u n ted fo r by referral to th a t behavior.
T ypically, tap h o n o m ists lo o k a t a g e -fre q u en cy d istrib u tio n s. In such cases,
the n u m b e r o f d ead indiv iduals p er age class is tallied (usually as M N I values)
a n d all are p lo tte d in a h isto g ram , each vertical b a r rep resen tin g a n age class
a n d th e height o f a b a r being scaled to the frequency o f individuals p er age class.
T w o basic types o f a g e -fre q u en cy d istrib u tio n s, o r w h a t are o ften called
m o rta lity profiles, are recognized by p aleo n to lo g ists (H u lb e rt 1982; K u rte n
1983; V o orhies 1969). Z o o arch ae o lo g ists have fu rth e r d istinguished several
v aria n ts o f the tw o basic p a tte rn s to aid in th eir in te rp re ta tio n o f fo o d -g ettin g
p ractices o f p re h isto ric people (K lein 1982a; Levine 1983; S tiner 1990b, 1991a).
O ne basic m o rta lity type is referred to as “ a ttritio n a l” o r “ n o rm a l” m ortality.
It is m o d eled as a frequency d istrib u tio n o f age classes in w hich very y o u n g an d
very old ind iv idu als are o v errep resen ted relative to th eir live ab u n d a n ces, an d
rep ro d u ctiv ely active ad u lts are u n d errep re sen ted because o f varying m o rtality
rates acro ss age classes (C raig a n d O ertel 1966). A ttritio n a l m o rta lity is
selective. T h o se age classes m o st susceptible to n a tu ra l ecological m o rtality ,
such as juveniles to o y o u n g to flee p re d a to rs o r escape tra p s a n d senile
ind iv id u als to o w eak to escape p re d a tio n o r survive less th a n o p tim a l e n v iro n
m en tal co n d itio n s, are m ore p ro n e to die w hereas healthy ad u lts in th eir prim e
re p ro d u ctiv e years are less p ro n e to die. T his susceptibility to m o rta lity results
in a b im o d al frequency d istrib u tio n , w ith one m ode to the fa r left a n d the o th er
to th e rig h t o f cen ter (F ig u re 5.1b); the frequency d istrib u tio n is o ften referred
to as “ U -sh a p e d ” (e.g., K lein 1982b:53). A ttritio n a l m o rtality results from
n o rm al o r ro u tin e ecologically related (accidental) d e a th s o f p o p u la tio n
m em bers. T he m od el o f a ttritio n a l m o rta lity given in F ig u re 5.1b assum es th a t
m o rta lity is slow a n d reflects the ra te o f biom ass tu rn o v e r, a n d is said to
illu stra te “ a b alan ced p ictu re o f a fa u n a as it existed in n a tu re ” over tim e
(V o o rh ies 1969:23).
T h e second basic m o rta lity type is referred to as a “ c a ta stro p h ic ” o r “ m ass”
m o rta lity p a tte rn . It is m odeled as a frequency d istrib u tio n o f age classes in
w hich successively o lder age classes are rep resen ted by few er a n d fewer
in d ividuals; in o th e r w ords, the frequency d istrib u tio n is u n im o d al w ith
extrem e positive skew ing (F ig u re 5.1a) a n d is referred to as “ L -sh a p e d ” (e.g.,
K lein 1982b:53). T h e choice o f the label “ c a ta s tro p h ic ” fo r the L -shaped
m o rta lity profile is fo u n d e d in the fact th a t m ass, o r non-selective, m o rta lity
p ro d u ces a synchro nic “ s n a p s h o t” (V oorhies 1969) o f a p o p u la tio n ’s age
stru ctu re at the tim e o f d ea th , b u t as n o ted below , this label m ay be a p o o r
choice. T heoretically, because it is non-selective, c a ta stro p h ic m o rta lity will
result in p ro p o rtio n a lly m o re prim e-age ad u lts dying th a n a ttritio n a l, selective
m o rtality (V oorhies 1969:46-47). N a tu ra l c a ta stro p h ic m o rta lity events
include floods (B oaz 1982), d ro u g h ts (C o n y b eare a n d H aynes 1984; H aynes
1984. 1987, 1988a, 1988b), a n d volcanic e ru p tio n s (L y m an 1987c, 1989b);
M o rta lity , skeletonization, disarticulation, scattering 119
500
□ N Alive
_C/5 £1 N Dead
4 0 0
03
ZJ
■g
>
3 0 0
C
0) 200 -
XI
E
3
100 -
*r »
0 1 2 3 4 5 6 7 8 9
Age Class
□ N Dead
w
co 200 -
3
;o
TD
jZ
1--------- 1—
Number of
_i.
O
o
1...........-
n n n
0
1 2 3 4 5 6 7 8 9
Age Class
Figure 5.1. T w o basic types o f age (m ortality) profiles (from T able 5.2A). a,
b lank bars den ote the n u m b er o f individuals alive in a co h o rt each year during
the existence o f the co h o rt, and stippled bars den o te the n u m b er th a t m ust die
each year (in each age class); b, the n u m b er th a t die each year (in each age class).
arch aeolo gical eq u ivalents o f c a ta stro p h ic m o rta lity ca n be fo u n d in ju m p sites

a n d tra p s w here large n u m b ers o f anim als w ere killed sim u ltan eo u sly (e.g.,
Levine 1983; R eh er 1974).
W h a t are typically called c a ta stro p h ic a n d a ttritio n a l m o rta lity are, from a
d em o g rap h ic a n d a n ecological perspective, in terre lated . A n in sta n ta n e o u s
event th a t causes the sy n ch ro n o u s d e a th o f all m em bers o f a p o p u la tio n
p ro d u ces an a g e -fre q u en cy d istrib u tio n o f th o se indiv id u als w ho have, to th a t
T ab le 5.2 L ife tables f o r two hypothetical

populations o f m am m als. For A , 5 0 0 fe m a le s give
birth once a year; f o r B, 5 0 0 fe m a le s give birth twice
a yea r or all have twins (fro m Klein 1982b)
Age N u m b er N u m b er M o rtality
alive dead rate
X h dx ?-v
A. 0 500 250 0.50

1 250 25 0.10
2 225 22 0.10
3 203 20 0.10
4 183 37 0.20
5 146 44 0.30
6 102 41 0.40
7 61 37 0.60
8 24 24 1.00
9 0 0
B. 0 1000 700 0.70

1 300 150 0.50
2 150 75 0.50
3 75 30 0.40
4 45 9 0.20
5 36 0.20
6 29 6 0.20
" 23 12 0.50
8 11 11 1.00
9 0 0
in stan t, survived m o re ro u tin e o r a ttritio n a l m o rta lity (K lein 1982b:58). “ T he

perfect m ass-m o rtality p o p u la tio n is a ‘fro zen ’ living p o p u la tio n ” (C raig and
O ertel 1966:351). T h u s the tw o frequency d istrib u tio n s in F igure 5.1 could
derive from the sam e p o p u la tio n o f anim als. A n a ttritio n a l ag e -fre q u en cy
d istrib u tio n o r m o rtality p a tte rn can be derived from the ag e -fre q u en cy
d istrib u tio n fo r the living p o p u la tio n (o r the “ c a ta s tro p h ic ” m o rta lity profile)
“ sim ply by g ra p h in g only the loss o f indiv id u als fo r each age g ro u p ” (K o rth
an d E v an d e r 1986:228). T h u s the stippled bars in F igure 5.1 rep resen t d ea th s in
a p o p u la tio n over som e m u lti-y ear tim e p erio d while the b lan k b ars rep resen t
the age stru c tu re o f the living p o p u la tio n at any one p o in t in tim e d u rin g th at
tim e in terv al (presum ing, as we shall see, th a t th e p o p u la tio n w as o f stable size
an d m a k e u p d u rin g th e tim e interval).
But th ere are co m plicating facto rs in the relatio n betw een the frequency
d istrib u tio n s (b lan k versus stippled bars) in F igure 5.1. In p artic u la r, re c ru it
m en t a n d m o rta lity rates m u st be off-setting if a p o p u la tio n is to rem ain stable
in size th ro u g h m u lti-g e n era tio n a l (usually m ulti-year) tim e. A s well, if th e age
stru c tu re o f th e p o p u la tio n is to rem ain u n ch a n g ed th ro u g h tim e, m o rtality

m u st be d istrib u te d a m o n g the age classes in such a m a n n e r as to ensure a stable
age stru ctu re . K lein (1982b) illu strates h ow m o rta lity rates m u st ch ange in the
face o f ra p id recru itm en t rates (o r vice versa) in o rd e r to m a in ta in a stable
p o p u la tio n size a n d stru ctu re. In a p o p u la tio n having 500 b irth s p er y ear a n d
m axim al longevity o f 9 years, 500 d e a th s per year m ust o ccu r w ith the
d istrib u tio n show n in T able 5.2a an d F igure 5.1. If th a t p o p u la tio n o f 500
in div id uals h as 1000 b irth s p er year, such as all fem ales giving b irth to tw ins,
th en 1000 d ea th s m u st o ccur p er year w ith th e d istrib u tio n show n in T able 5.2b
a n d F ig u re 5.2. C o m p arin g F ig u re 5.1 a n d F ig u re 5.2 show s how increased
re cru itm en t suppresses the m o d e signifying the d e a th o f old indiv id u als p ast
th eir prim e, p ro d u c in g an a g e -fre q u en cy d istrib u tio n w ith m o re o f an “ L-
sh a p e ” (such an effect co uld be en h an ced by m ak in g the scale on the y-axis
larger). W hile th e d a ta show the frequencies o f d ea th s are “ U -sh a p e d ,” this
exercise u n d erscores th a t sim ple inspection o f the sh ap e o f an a g e -fre q u en cy
d istrib u tio n like F igure 5.2 can be m isleading.
G iven th e ab o v e co m p licatin g facto rs, it is n o t su rp risin g th a t p a le o n to lo
gists a n d zo o arch aeo lo g ists have used a set o f crite ria to help determ ine
w hether, for exam ple, an L -sh ap ed a g e -fre q u en cy d istrib u tio n derived from
fossil rem ains actu ally represents an instance o f c a ta stro p h ic (ag e-indepen
d en t) m o rtality , o r an instance o f a ttritio n a l m o rtality . T he sh ap e o f the a g e -
frequency d istrib u tio n is one o f th o se criteria, b u t it c a n n o t be th e only one.
O th er criteria used to identify th e k in d o f m o rta lity p a tte rn rep resen ted include
th e discreteness o f age classes, ta p h o n o m ic in fo rm a tio n , a n d paleoecological
d a ta (K lein 1982b:58-59). By discreteness o f age classes I m ean the observed
ran ge o f v a ria tio n in o n to g en ic ages a ro u n d age class m id -p o in ts. T he
discreteness o f age classes crite rio n applies only to th o se tax a w ith seasonally
restricted b irth in g seasons (C raig a n d O ertel 1966:351). T his is so because if
b irth s o ccu r th ro u g h o u t the year, then a c o h o rt w ould consist o f individuals o f
all ages (any m o n th ) w ithin a year. If b irth s are tem p o ra lly restricted to a m o n th
o r season, th en all in d ividuals in a c o h o rt will be w ithin a m o n th o r tw o o f
exactly th e sam e age. F o r a seasonally b irth in g tax o n a n d age classes th a t each
re p resen t one year, a ran g e o f v a ria tio n in tim e o f d e a th o f ± 5 m o n th s per age
class suggests m o rta lity w as a ttritio n a l as individuals w ere dying virtu ally all
year long. A ran ge o f v a ria tio n in tim e o f d e a th o f ± 1 m o n th suggests a ra th e r
lim ited tim e d u rin g the year w hen anim als died an d th a t m o rta lity was
cata stro p h ic.
A s w ith the sh ap e o f th e a g e -fre q u en cy d istrib u tio n , ag e-class discreteness
sh o u ld n o t be used alone. T h a t is so because individuals in a p o p u la tio n m ight
be dying th ro u g h o u t the year, o r attritio n a lly , yet the locality sam pled m ay
re p resen t a seasonally restricted a c cu m u latio n o f th o se rem ains, such as
a ro u n d w ater-h oles used only d u rin g th e d ry season. Som e exam ples will m ake
the im p o rtan ce o f this clear.
1000 -
□ N Alive
1! N Dead
800 -
600 -
400 -
200 -
0- II— U — v — U— I.*-"*,
0 1 2 3 4 5 6 7 8 9
Age Class
N Dead
600 -
iO
TO
•g
>
C 400 -
a>
_Q
E 200 -
1 2 3 4 5 6 7 8 9
Age Class
F igure 5.2. Age (m ortality) profiles fo r a p o p u latio n w ith high m o rtality and
recruitm ent (from T able 5.2B). a and b as in Figure 5.1.
H u lb e rt (1982) describes a set o f fossil th ree -to ed h o rse (N eohipparion cf.

leptode) rem ain s th e a g e -fre q u en cy d a ta fo r w hich in d icate m o rta lity was
a ttritio n a l (F ig u re 5.3) a n d the age classes fo r w hich are ± 0.35 years. T he lack
o f relatively n o n -d iscrete age classes a n d th e U -sh a p ed m o rta lity profile, once
th e a p p a re n t u n d e r-re p re se n ta tio n o f the y o u n g est age class is acco u n ted for,
suggest m o rta lity w as a ttritio n a l. K u rte n (1983) describes a set o f fossil
an telo p e (P achy tragus solignaci) rem ains from T unisia the ag e-freq u en cy
d istrib u tio n o f w hich suggests m o rta lity w as c a ta stro p h ic (F ig u re 5.4) yet the
60
252 □ N Obs. Dead
[1 N Est. Dead
J2
co
40 -
-g
>
TD
c
CD 20 -
jO
E
v* * r ■ ■nr '» V--- *Y t*---“r
innnnn Age Class

I
t-
I
CM
Figure 5.3. M o rtality profile for fossil horses (d ata from H u lb ert 1982). Age
classes are one year each. B lank bars are frequencies observed in fossil record;
stippled b ar is estim ated.
150
□ N Dead
125 -
Number of Individuals
100 -
75 -
50 -
25 -
0 V *i* *i* 'V

1 2 3 4 5 6 7 8 9 1 0
Age Class
Figure 5.4. M o rtality profile for fossil antelo p e (d ata from K u rten 1983). Age
classes are one year each.
2 0 -i----------------------------------------
□ N Dead
— 15-
-g
>
T3
z. 10 "
0
CD
-Q
1 s-
Z
0
1 2 3 4 5 6 7 8 9
Age Class
Figure 5.5. M ortality profile for archaeological deer rem ains (d ata from Sim pson
1984). A ge classes are one year each.
lack o f discrete age classes a n d geological d a ta indicate the sam ple resulted
from seaso n al sam p lin g o f a p o p u la tio n u n d erg o in g a ttritio n a l m o rtality .
Sim ilarly, in h er stu d y o f m ule d eer (Odocoileus hem ionus) rem ains from an
arch aeolo gical site, S im pson (1984) fo u n d an L -shaped a g e -fre q u en cy d istri
b u tio n (F ig u re 5.5) b u t arg u ed th a t because the age classes were n o t discrete
(w ere ± 3 m o n th s), the rem ains ac cu m u lated d u rin g one season over m ultiple
years. N im m o (1971) derived a U -sh a p ed a g e -fre q u en cy d istrib u tio n (F igure
5.6) fo r an assem blage o f p ro n g h o rn an telo p e (A ntilocapra am ericana)
rem ains, b u t arg u ed th a t because the age classes w ere relatively discrete ( ± 1
m o n th ), the tax o n h ad seasonally-restricted b irth in g , a n d the fossil rem ains
were recovered from an archaeological kill site, m o rtality was ca ta stro p h ic.
M o re th a n the sh ape o f the a g e -fre q u en cy d istrib u tio n derived for a set o f
fossils m ust be co nsid ered if one wishes to infer w h eth er m o rta lity was
“ a ttritio n a l” o r “ c a ta stro p h ic .” A n d the la tte r term m ay well be a m isn o m er if
it is allow ed to d en o te w h eth er m o rtality w as d iach ro n ic o r synchronic ra th e r
th a n ju s t the shap e o f the m o rta lity profile described by an ag e-freq u en cy
d istrib u tio n . B ut the sh ape o f such frequency d istrib u tio n s is a logical place to
s ta rt o n e ’s analysis. K lein (e.g., 1982a; K lein an d C ru z-U rib e 1984:57-60)
suggests using the K o lm o g o ro v -S m irn o v tw o-sam ple D statistic as a w ay to
d eterm in e w h eth er the shape o f an a g e -fre q u en cy d istrib u tio n ap p ro x im a te s an
L -sh ap ed o r a U -sh a p ed frequency d istrib u tio n . T he p ro c ed u re is to take an
ag e -fre q u en cy d istrib u tio n fo r a kn o w n m o rta lity event, w h eth er a ttritio n a l or
c a ta stro p h ic, a n d c o m p are the cu m ulative p ercentage d istrib u tio n o f age
classes in th a t d istrib u tio n w ith th a t o f the p o p u la tio n o f u n k n o w n m ortality.
A n exam ple will m ak e this clear.
T ab le 5.3 O bserved and ex p ected frequencies o f w apiti fr o m catastrophic

m o rta lity resulting fr o m volcanic eruption o f M o u n t S t. H elens (fro m L ym a n
1987c)
Age N u m b er of N um ber of Difference

Class D ead O bserved (C um . % ) D ead E xpected (C um . % ) (D)
0 13(0.157) 19 (0.275) 0.118*

1 20 (0.389) 14 (0.478) 0.089
2 11 (0.517) 10 (0.622) 0.105
3 14(0.68) 8 (0.738) 0.058
4 18 (0.889) 6 (0.825) 0.064
5 3 (0.924) 4(0 .8 8 3 ) 0.041
6 1 (0.934) 3 (0.926) 0.008
7 2 (0.957) 2 (0.955) 0.002
8 1 (0.967) 2 (0.984) 0.017
9 1 (0.977) 1 (1.00) 0.023
>10 2(1.00) 0 (1 .0 0 ) 0.000
T otal: 86 69
Note:
a g reatest D.
30
□ N Dead
cc
1 20
ID
o
<5
10
ZJ
0
1 2 3 4 5 6 7
Age Class
Figure 5.6. M ortality profile for archaeological p ro n g h o rn antelope rem ains
(data from N im m o 1971). Age classes are one year each.
T he m o rta lity o f w apiti (Cervus elaphus) resulting fro m the volcanic eru p tio n
o f M o u n t St. H elens in the state o f W a sh in g to n in M ay o f 1980 p ro d u c ed the
kn o w n ca ta stro p h ically -g en e rated a g e -fre q u en cy d istrib u tio n described in
T ab le 5.3 (L y m an 1984b, 1987c, 1989b). B ecause the sam ple is n o t re p resen ta
tive o f the ex tan t w apiti p o p u la tio n due to b eh av io ral v aria tio n betw een
different ag e-sex g ro u p s (L y m an 1987c, 1989b), the observed n u m b er o f
20
_w
05
3
■g
>
£.
o 10
CD
-Q
E
3
Figure 5.7. Expected an d observed m o rtality profile fo r w apiti killed by the

volcanic eru p tio n o f M o u n t St. H elens (from T able 5.3). A ge classes are one year
each.
in d iv id uals described p er age class w as sm o o th ed using p o ly n o m ial regression

(C aughley 1977) to p ro d u c e the expected n u m b e r o f indiv id u als p er age class
given in T ab le 5.3. T h e frequency d istrib u tio n o f the expected n u m b e r o f dead
in d iv id uals p er age class is clearly c a ta stro p h ic a n d differs visually fro m th a t fo r
the observ ed n u m b e r o f individuals. B ut is th a t difference significant? C u m u la
tive p ercen tages w ere ca lcu lated fo r each a g e -fre q u en cy d istrib u tio n a n d the
differences betw een th o se p ercentages by age class o r row w ere derived (T able
5.3). T he greatest difference, o r D statistic, is 0.118 an d falls in the youngest age
class. T h a t observed D is less th a n the tab led D o f 0.22 w hen P = 0.05. T hus, the
K o lm o g o ro v -S m irn o v test suggests there is no statistically significant differ
ence betw een the tw o frequency d istrib u tio n s in F igure 5.7. If, fo r exam ple,
these w ere b o th fossil assem blages, one w ould conclude, on the basis o f the
K o lm o g o ro v -S m irn o v tw o-sam ple D statistic, th a t they b o th represented
c a ta stro p h ic m o rta lity events (on the basis o f m o rta lity profile shape an d the
discrete age classes) even th o u g h the frequency d istrib u tio n s a p p e a r different.
Interpreting the dem ographics o f m ortality
T h u s fa r th e discussion has co n cern ed identifying tw o basic types o f m o rta lity

profiles o riginally identified in p aleo n to lo g y . A s n o ted above, z o o a rc h a e o lo
gists have ex p an d ed the n u m b e r o f m o rta lity profile types (e.g., Levine 1983), in
p a rt d u rin g atte m p ts to d istinguish wild prey anim als fro m dom estic prey
an im als (e.g., C ollier a n d W hite 1976; W ilkinson 1976). W e still have m uch left
to learn in this p a rtic u la r reg ard (e.g., C rib b 1985, 1987), yet the dem o g rap h ics
o f m o rta lity rem ain a n im p o rta n t ta p h o n o m ic clue. T his is in spite o f th e fact

th a t “ the cause o f d e a th does n o t necessarily in d icate w h eth er the p a tte rn o f
m o rtality is a ttritio n a l o r c a ta s tro p h ic ” (S h ip m an 1981 b: 18). Sim ilarly, as we
h ave seen in c o n ju n ctio n w ith the discussion o f F ig u res 5.3 th ro u g h 5.6,
w h eth er th e p a tte rn o f m o rta lity is a ttritio n a l o r c a ta stro p h ic does n o t
necessarily indicate the cause o f d e a th o r w h eth er m o rta lity w as d iach ro n ic o r
synchronic.
Stiner (1990b, 199Id) em phasizes, correctly I th in k , th a t until we know an d
u n d e rsta n d the m o d e rn ran g e o f v a ria tio n in m o rta lity p a tte rn s, we will have
difficulty u n d e rsta n d in g w h a t vario u s m o rta lity p a tte rn s signify w hen id en ti
fied in p re h isto ric contexts. She suggests th a t we th ere fo re m u st study th a t
m o d e rn v a ria tio n in c o n ju n ctio n w ith the ecologies a n d beh av io rs o f b o th
p re d a to r a n d prey species. F o r exam ple, L ym an (1989b, 1991a, 1991b) reviews
th e m o d e rn b ehav iors o f p in n ip ed tax a d u rin g the b reed in g a n d b irth in g season
an d how escape b eh avio rs, o r lack thereof, resulted, in the case o f one tax o n , in
the d e a th o f m an y ad u lt breeding-age m ales th a t defend th eir breeding
territo ries ag a in st all com ers a n d few d eath s o f o th e r ag e-sex classes. In the case
o f a n o th e r p in n ip ed tax o n , the a d u lt m ales a n d fem ales o f w hich ten d to flee at
signs o f d an g er, m o stly helpless a n d relatively naive a n d im m obile new b o rn s
were killed by h u m an h u n ters. H u d so n (1991) show s how p ro c u rem en t
tech no lo gy exerts sim ilar influences on the ag e-sex stru c tu re o f the killed
p o p u la tio n o f a sm all cervid. Based on n e o ta p h o n o m ic d a ta . H o c k ett (1991)
argues th a t assem blages o f leporid (ra b b its a n d hares) bones ac cu m u lated by
ra p to rs will be d o m in a te d by rem ains o f juveniles. H u m a n h u n ters, in c o n tra st,
create lepo rid b o n e assem blages w ith relatively high frequencies o f rem ains o f
ad u lts. T his difference exists because o f an u p p e r size lim it o f ind iv id u al prey
th a t ra p to rs can effectively prey u p o n (H o c k e tt 1991). S im ilar selection o f
y o u n g er age classes o f prey h as been d o cu m en te d fo r som e m am m alian
carn iv o res (e.g., A nd rew a n d E vans 1983; H aynes 1980b). In fact, in a unique
stu d y o f the d em o g ra p h y o f prey m o rtality , S m ith (1974) suggests wolves
(Canis lupus) focus th eir p re d a tio n o f deer (Odocoileus virginianus) on old
ad u lts a n d juveniles w hereas the p re h isto ric h u m a n g ro u p he w as stu d y in g to o k
m o stly n ea r-a d u lts a n d p rim e ad u lts. S m ith concludes th ere w as, th en , no direct
co m p etitio n fo r deer betw een the h u m a n s a n d w olves, a n d th a t the tw o
p re d a to rs k ep t th e d eer p o p u la tio n stable because the n et effect o f p re d a tio n by
the tw o p re d a to rs w as th a t all age classes w ere exploited.
S tu d y o f m o rta lity profile shape, ag e-class discreteness, a n d co n tex t o f the
fossils g ra n ts insights to w h eth er m o rta lity w as synchronic (m ass, o r literally
ca ta stro p h ic ) o r d iach ro n ic (a ttritio n al). O nce th a t is d eterm in ed , stu d y o f
o th e r variab les such as b eh a v io ral v a ria tio n betw een d istin c t ag e-sex gro u p s
a n d to p o g rap h ic al, geological, a n d arch aeo lo g ical c o n tex tu a l d a ta m ay reveal
w hy m o rta lity w as sy nchronic o r d iach ro n ic. W hy does a n instance o f
p re h isto ric m o rta lity h ave the d em o g ra p h ic expression it does? W e next tu rn to
how this q u estio n is answ ered.
A n a lyzin g the dem ography o f m ortality

K lein (1982a) suggests th a t in sp ectio n o f a m o rta lity profile in c o n ju n ctio n w ith
know ledge o f the b eh av io r o f the tax o n involved a n d stu d y o f the geological
co n tex t o f the fossil sam ple allow s the d istin ctio n o f an im al p o p u la tio n s th a t
were g en erated by h u n tin g fro m those th a t were gen erated by scavenging. In
arch aeo lo g ical contex ts, tax a th a t can be tak en in large n u m b ers sim u lta
neously, such as in drives o r tra p s, will p ro d u c e c a ta stro p h ic profiles (L -shaped
ag e -fre q u en cy d istrib u tio n s) w hereas tax a th a t are n o t beh av io rally am enable
to being tak en in large n u m b ers a t one tim e m u st be h u n te d individually and
will ten d to p ro d u ce a ttritio n a l m o rta lity profiles (U -sh ap ed ag e -fre q u en cy
d istrib u tio n s). In the latter, very y o u n g individuals are m ore ra re th a n in
p aleo n to lo g ical assem blages, suggesting th a t in the la tte r scavenging by
carn iv o res has selectively rem oved the rem ains o f the youngsters. T hus, in
arch aeo lo g ical co n tex ts low p ro p o rtio n s o f y o u n g sters in a p o p u la tio n disp lay
ing a ttritio n a l m o rta lity suggests scavenging by h u m an s o f prey anim als th at
h ad been killed a n d co n su m ed by n o n -h u m a n carnivores. H igh p ro p o rtio n s o f
y o u ng sters, how ever, suggest active h u n tin g by h u m an s. If a c a ta stro p h ic
m o rtality profile is fo u n d in an arch aeo lo g ical co n tex t, h u n tin g an d scavenging
are best d istin guished by geological d a ta . In these cases, scavenging by hu m an s
o f a p o p u la tio n o f an im als th a t died d u e to a n a tu ra l c a ta stro p h e is, K lein
(1982a: 153) supposes, a ra re event. T herefore, if a c a ta stro p h ic m o rtality
profile has been derived from fossils th a t clearly ac cu m u lated over a long tim e
span (h u n d red s o r th o u sa n d s o f years given stra tig ra p h ic a n d c o n tex tu a l d ata),
then it is likely th a t m o rta lity represents artificial c a ta stro p h e s an d h u n tin g
w hereas a relatively sh o rt-term ac cu m u latio n (e.g., a single event) m ay
rep resen t h u n tin g o r scavenging by h u m a n p re d a to rs. G eological d a ta m u st be
called u p o n to d eterm in e if the rem ains are in a sedim entary co n tex t suggestive
o f n a tu ra l m o rta lity (such as a fiash-fiood d ep o sit) o r artificial m o rta lity such as
at th e base o f a cliff.
S tiner (1990b, 1991a, 1991 d) em ploys a three-pole g ra p h in g tech n iq u e to
assess m o rta lity p attern s. E ach o f the three axes represents a different age class:
juveniles (ap p ro x im a tely the first 20% o f n a tu ra l ecological longevity, bu t
defined by S tiner [ 1990b:311] as from b irth to the age at w hich a p artic u la r,
tax o n -d e p en d en t decid u o u s to o th is shed), prim e-age ad u lts (breeding age),
a n d old ad u lts (p a st th eir prim e, ap p ro x im ately the last 30% o f n a tu ra l
ecological longevity) (F ig u re 5.8). B ased on p o p u la tio n s w ith k n o w n m o rtality
profiles, Stiner defines areas on the three-pole g ra p h th a t c o rresp o n d ed to the
tra d itio n a l U -sh ap ed an d L -shaped a g e -fre q u en cy d istrib u tio n s as well as
th ree ad d itio n a l types o f m o rta lity th a t she labels ju v en ile-d o m in ated , prim e-
d o m in ated , a n d o ld -d o m in ated . H e r g ra p h in g tech n iq u e gives the analyst
a n o th e r w ay to visualize v a ria tio n in m o rta lity across age classes. B ased on
d em o g rap h ic d a ta co m p iled fo r prey tax a, S tiner uses this g rap h in g technique
100X ol d
OX ol d
Figure 5.8. T hree-pole grap h in g technique for assessing dem ographic (m ortality)
d ata (after Stiner 1990b:318, Figure 6).
to distin g u ish, in a g eneral w ay, the kinds o f h u n tin g techniques th a t p ro d u ce a

p a rtic u la r kind o f m o rtality . C u rso rial p re d a to rs tend to “ engage in long chases
o f th eir q u a rry ” a n d generate U -sh ap ed m o rta lity profiles (S tiner 1990b:322).
P red a to rs w hich am b u sh th eir prey tend to p ro d u ce L -shaped m o rtality
profiles because the e n c o u n te r o f prey is d eterm in ed by chance. P re d a to rs th a t
dep en d on scavenging ten d to p ro d u c e o ld -d o m in ated a g e -fre q u en cy d istrib u
tions, w h eth er they are cu rso rial o r am b u sh p re d a to rs, p ro b a b ly because prim e
ad u lts are less susceptible to a ttritio n a l d e a th a n d the rem ains o f juveniles are
less likely to survive gnaw ing by carnivores.
A d a ta set described by K lein (1982a) can be used to illu strate K lein ’s and
S tin e r’s techniques. T h a t d a ta set concerns tw o collections (T able 5.4) o f
rem ain s o f th e ex tinct g ian t A frican buffalo Pelorovis antiquus, a large bovid
th a t K lein (1982a) ch aracterizes as ra th e r im m une to p re d a tio n due to its size
a n d suspected social co hesion in defensive situ atio n s. B oth collections a p p a r
ently re p resen t palim p sest accu m u latio n s; th a t is, m ultiple stra tig ra p h ic units
a n d /o r d ep o sitio n al events are rep resen ted by each collection. P lo ttin g the a g e -
freq uen cy d istrib u tio n s for the tw o assem blages, it is o b vious th a t the m o rtality
profile fo r K lasies R iver M o u th co n tain s m an y m o re juveniles th a n the profile
fo r E lan d sfo n te in (F ig u re 5.9). These tw o m o rta lity profiles are statistically
different (D = 0.557, P < 0 .0 1 ) . K lein (1982a) believes the K lasies R iver M o u th
d a ta rep resent active h u n tin g o f the anim als due to the significantly g reater
n u m b e r o f in div iduals in the first age class relative to the low frequency o f
y ou n g in dividu als in the E lan d sfo n te in assem blage, w hich seems to represent a
T ab le 5.4 M o rta lity data fo r two fo s s il assem blages

o /P e lo ro v is an tiq u u s (fro m K lein 1982a). Each age
class represents 10% o f the natural ecological
longevity o f the taxon
Age class K lasies R iver M o u th E landsfontein
1 41 8
? 1 12
3 1 2
4 4 11
5 4 10
6 7 12
3 19
8 3 15
9 0 6
10 0 0
40 □ Klasies
El Elandsfontein
ra
-g 30
>
T3
_C
o 20
CD
JQ
E
- 10
i—
0 r-n f r r r i
Figure 5.9. M o rtality profiles for A frican bovid rem ains from K lasies River
M o u th an d from E landsfontein (from T able 5.4). A ge classes are each 10% o f
n atu ra l ecological longevity.
n a tu ra l ac cu m u latio n . T he greatest difference in the cu m u lativ e frequency

d istrib u tio n in fact occurs in the y o u n g est age g roup.
F o llow ing S tin er’s (1990b) p ro c ed u re a n d lum ping the first tw o age classes
into the ju v enile age g ro u p , the th ird th ro u g h the seventh age classes in to the
prim e age g ro u p , a n d the eighth th ro u g h the ten th age classes in to the old age
g ro u p , th e p lo t o f p o in ts on the th ree-p o le g ra p h (F ig u re 5.10) indicates the tw o
assem blages are different, a n d th a t th e K lasies R iver M o u th assem blage
100* old
Figure 5.10. T hree-pole g rap h o f m o rtality d a ta from K lasies River M o u th and

E landsfontein (from T able 5.4; see Figure 5.8).
ap p ro x im a te s a U -sh ap ed profile w hereas the E lan d sfo n te in assem blage

ap p ro x im a te s an L -shaped m o rta lity profile. T he latter is p erh ap s a fun ctio n o f
the g ra p h as o th e r d a ta described by K lein (1982a) in d icate b o th profiles are
a ttritio n a l. C u m u lativ e frequencies o f the K lasies R iver M o u th profile d o n o t
differ significantly fro m an a ttritio n a l profile m odeled fro m the n u m b e r o f dead
in d iv idu als in T ab le 5.2a (Z) = 0.141, P > 0 .0 5 ) b u t cum ulative frequencies o f
the E lan d sfo n te in profile do differ fro m th a t m o deled a ttritio n a l profile
(D = 0.416, Z3< 0 .0 1 ). B oth the K lasies R iver M o u th an d E landsfontein
assem blages differ significantly from the c a ta stro p h ic a g e -fre q u en cy d istrib u
tion m odeled by the n u m b er o f live individuals in T ab le 5.2a ( / 5< 0.01 for
b o th ). K lasies R iver M o u th has p ro p o rtio n a te ly m ore juveniles th a n the
m odeled c a ta stro p h ic profile a n d E lan d sfo n te in has p ro p o rtio n a te ly m ore
prim e-ag ed in d ividuals. S tin er’s g ra p h (F igure 5.10) suggests ju s t such a
difference betw een th e tw o collections.
If K lein (1982a) is co rrec t th a t the E lan d sfo n te in assem blage represents a
n a tu ra l ac cu m u latio n , w h at does the w eakly p rim e-d o m in ated m o rtality
profile fo r th a t assem blage signify? P erh a p s this assem blage consists m ostly o f
prim e-age m ales th a t died d u e to m a ln u tritio n afte r the stress o f the breeding
season. K lein (1982a: 155) re p o rts th a t m an y o f th e rem ains occur “ n e a r w h at
[at the tim e th e rem ains accum ulated] w ere p ro b a b ly p eren n ial sources o f
w a te r.” D id this species o f bovid re sp o n d in a m a n n e r sim ilar to B arn o sk y ’s
(1986) Irish elk (see C h a p te r 3)? A lternatively, the E lan d sfo n te in rem ains are
w eakly o ld -d o m in ated (F ig u re 5.10), so p e rh a p s it w as individuals o f b o th sexes
th a t w ere n earin g th e end o f th eir prim e years th a t w ere dying; these age-sex
c o h o rts w ould h ave been m o re susceptible to the vicissitudes o f dry-season
en v iro nm ents. W h e th e r these bovids w ere m ostly m ales o r individuals o f bo th
sexes alm o st p ast th eir prim e, this analysis o f the d em o g ra p h y o f th eir m o rtality
directs us to w ard o th e r kinds o f d a ta th a t m ay co n trib u te to a ta p h o n o m ic
e x p lan a tio n for th eir d e a th a n d occurrence as fossils. A b im o d al d istrib u tio n o f
sizes o f rem ain s w o uld suggest b o th sexes w ere represented. I f season o f d ea th
co uld be d eterm in ed , th a t m ight p ro v id e c o rro b o ra tin g evidence fo r dry-season
stresses. R egardless o f the o u tco m e o f such analyses, this possibility leads us to
a n o th e r topic, th e seaso n ality o f m o rtality .
The seasons o f mortality

As sh o u ld be clear from the preceding, the season o f m o rta lity m ay be an
im p o rta n t ta p h o n o m ic v ariab le in e x p lan a tio n s o f w hy a bone assem blage
ap p e ars th e w ay it does, especially in term s o f the d em o g ra p h y o f m o rtality .
M eth o d s fo r d eterm in in g th e season o f an im al d e a th are m any. M o n k s (1981)
p ro vid es a d etailed review o f m o st o f them , an d I d o n o t re ite rate th a t
discussion here. It suffices to n o te th a t tw o p ro p e rtie s allow the season o f d eath
to be d eterm ined . F irst, m an y tax a have seasonally restricted breeding and
b irth in g seasons. S econd, due to p a tte rn e d o n to g en ic dev elo p m en t, it is
possible to assign a season o f d e a th to organism s w ith a restricted season o f
b irth based on th eir observed stage o f o n to g en ic developm ent. T he age o f an
ind iv idu al o rg an ism at d e a th can be assigned to it based on th e second
p ro p e rty , b u t season o f d e a th ca n only be d eterm in ed fo r indiv id u als o f those
tax a w ith th e first p ro p e rty .
F o r cervids such as deer (Odocoileus spp.) in the n o rth w e ste rn U n ited States,
the seasonally restricted b irth season occurs betw een M ay 1 a n d Ju n e 30. A n
in d iv id u al th a t dies w hen it w as exactly one y ear old died Ju n e 1 ± 30 days.
U sing th a t know ledge o f m o d ern deer, plus know ledge o f the on to g en y o f
m o d e rn deer, one ca n estim ate th e season o f d e a th o f d eer rep resen ted by
p reh isto ric rem ains. F igures 5.11a a n d 5.12a illu strate exam ples ta k e n from
tw o arch aeo lo g ical sites in ea ste rn W a sh in g to n state. In b o th cases, the
sea so n ality -freq u e n cy d istrib u tio n s, o r seasonality profiles, suggest m ost deer
w ere dy ing in late sum m er th ro u g h early w in ter m o n th s. B ecause th e presence
o f b u tch ery m ark s (see C h a p te r 8) on m an y o f the b ones indicates h u m an s
played a role in th eir a c cu m u latio n an d d ep o sitio n , the analyst m ight in terp re t
the season ality profiles as in d icatin g th a t d eer h u n tin g w as m o st intensive in the
fall, a n d call on evidence such as the co n d itio n o f un g u lates in tem p erate
latitu d es as in d icatin g the fall m o n th s are the tim e w hen deer are o f the m ost
n u tritio n a l value to h u m an s (e.g., Speth a n d S pielm ann 1983).
T he p receding in te rp re ta tio n o f F igures 5.11a a n d 5.12a suggests th a t the
m o rta lity profiles fo r th e tw o assem blages sh o u ld a p p ro x im a te a ca ta stro p h ic
12
10
_w
co
=3
8 -
>
TO
- 6 H
Cl)
_Q
E
c cn
-8 <
Q_ 13 D
<
Month of Death: ‘t
1
2
1 5
J/5
ro
■g
2> 10 -
TD
<D
-Q 5 -
E
CD C\J o CO CD cvj o
Age at Death: CO G) o eg CD 00
J. T—
JL J. JLCO LO
I
CO
T1 1
o>
1
r^
1
in
T
CO
a> o C\J CD
Figure 5.11. Seasonality (a) and m o rtality (b) profiles fo r deer (Odocoileus spp.)
rem ains from archaeological site 4 5 D 0 1 8 9 in eastern W ashington (from Lym an
1988b). A ge a t death in (b) is in m onths.
10
Vi
w 8 -
05
13
■g
>
TD 6 -
c.
CD 4 -
_Q
E
z
co CO C\J o CO co C\J o
CO in r^ O) o C\J CO 00
Age at Death:
i r^
co
in
LO
CO
r^
T
1
T
cr>
T1
r^
T
LO
T
CO
cn o C\J CO
Figure 5.12. Seasonality (a) an d m ortality (b) profiles for deer (Odocoileus spp.)
rem ains from archaeological site 4 5 D 0 1 7 6 in eastern W ashington (from Lym an
1985b). Age at death in (b) is in m onths.
o r L -sh ap ed a g e -fre q u en cy d istrib u tio n . T hey in fact d o (F igures 5.11b an d

5.12b). T h a t this is the case is due to the seasonally-restricted b irth in g season
fo r the tax o n involved, an d , like the m o rtality profiles o f K u rte n (1983; F igure
5.4) a n d S im pson (1984; F ig u re 5.5), the m o re o r less seasonally-restricted
ac cu m u latio n o f deer rem ains. In these cases, the seasonally restricted m o rta l
ity o f d eer helps ac co u n t fo r the m o rta lity profile, a n d the m o rta lity profile
helps ex plain th e seasonality profile.
A less freq uently used technique for assessing seasonality is th e presence o f
c h itin o u s exuvia o f fly p u pae. G ilb ert a n d Bass (1967) determ ine the season
w hen h u m a n s w ere b u ried in a site in S o u th D a k o ta by th e presence o f such
insect rem ain s a n d the lim ited season (d u rin g w arm m o n th s) w hen such
o rg an ism s are active to d ay . C h o m k o a n d G ilb ert (1991) infer a Ju n e th ro u g h
S ep tem ber season o f site fo rm a tio n a n d d ep o sitio n o f bo n e pieces in a pit based
o n th e asso ciatio n o f fly p u p ae w ith the bones. G u th rie (1990) sim ilarly infers
th e season o f d e a th fo r an in d iv id u al o f a late P leistocene bison (Bison sp.) in
A lask a based on the presence o f fly p u p ae. T hese insects te n d to atta c k
carcasses quickly, a n d th eir presence suggests som e delay betw een d e a th a n d
bu rial. In fact, G u th rie (1990:83) cites evidence from S o u th A frica ind icatin g
m am m al carcasses there m ay be com pletely skeletonized by flies w ithin five
days o f d e a th in the su m m er, a n d w ithin 14 days in the w inter. Sim ilarly, P ayne
(1965) re p o rts th a t im m a tu re pig (Sus scrofa) carcasses a b o u t 1.0 to 1.4 kg in
size placed outsid e in S o u th C a ro lin a w ere com pletely skeletonized (bits o f hide
rem ained ) by feeding insects a fte r eight days. T h u s the im p act a n d utility o f
insects fo r u n d e rsta n d in g ta p h o n o m ic processes seem s great. N o t only m ight
th eir rem ains, w hen associated w ith v e rte b ra te skeletons, suggest delayed
b urial a n d season o f d ea th , b u t th eir presence m ay help explain w hy som e
carcasses are fo u n d in a m o re o r less artic u la te d state w hile o th ers are
d isarticu lated . T h a t is, as we see in the next section, the ap p e ara n ce o f d ea th in
the fossil reco rd can be m ightily influenced by m ere insects as well as larger
ta p h o n o m ic agents.
Skeletonization and disarticulation

If air is cu t off, m alig n an t p u trefactio n sets in, causing the fo rm atio n o f noxious
substances.
(J. W eigelt 1927/1989:5)
N a tu ra l tra p s such as chim ney caves, bogs, a n d ta r pits are generally evident
from geological d a ta , an d the cause o f d e a th is ra th e r m ore evident in such
co n tex ts th a n in o th e r kinds. F o r the m o re typical assem blages o f bones
recovered from op en sites, geological d a ta m ay p ro v id e clues as to the m ode o f
d ea th , b u t the a p p e ara n ce o f the d ead an im als m ay also p ro v id e im p o rta n t d ata
on how they died. F o r exam ple, B arn o sk y ’s (1985, 1986) Irish elk fossils were
recovered fro m p ea t deposits suggesting they died in a bog, a n d th a t is the
Figure 5.13. A p artial, articu lated w apiti (Cervus elaphus) skeleton in situ.
source o f the m irin g h y p o thesis he tests. L y m an (1989b) a n d V o orhies (1981)

describe th e effects o f volcanic eru p tio n s on the ap p e ara n ce o f m o rtality .
In d iv id u al m am m als th a t w ere shielded fro m th e explosive force o f the e ru p tio n
sim ply collap sed a n d w ere quickly b u ried by volcanic ash fro m the e ru p tio n ,
th u s th eir skeletons w ere still largely a rtic u la te d w hen recovered (e.g., F igure
5.13). T h e fact th a t these skeletons w ere stratig ra p h ic ally w ithin a thick d ep o sit
o f volcanic ash suggested the in d iv id u als died o f suffocation. B ut th eir ra p id
b u rial resulted in m o stly co m p lete skeletons w ith in d iv id u al b ones in th eir
p ro p e r a n a to m ica l po sitio ns. W h a t I am to u ch in g o n here are th o se im m ediate
p o stm o rte m processes o f so ft tissue d ec o m p o sitio n , sk eleto n izatio n , a n d
d isarticu latio n , a n d it is these processes to w hich I now tu rn .
I w as eatin g m y lu n ch th e first tim e I read W eigelt’s (1927/1989) grap h ic
d escrip tio n s o f im m ediately p o stm o rte m a lteratio n s to an im al carcasses; it w as
all I co u ld do to finish m y p e a n u t-b u tte r a n d jelly sandw ich. B ut W eigelt’s a n d
o th e rs ’ d escrip tio n s o f w h a t h a p p e n s in th e first few days o r w eeks a fte r d ea th
are im p o rta n t because they illu stra te h ow the m o d e o f d e a th can influence the
ta p h o n o m ic h isto ry o f an an im al carcass. S chaefer (1962/1972:10) w as quite
aw are o f this w hen he w rote “ usually en o u g h tim e h as passed [betw een d ea th
an d burial] fo r th e skeletal p a rts to have lost th eir original [soft tissue]
co n n ectio n s o r fo r ex tern al ag en ts to h ave sep a rated them . H ence th ere are only
tw o w ays in w hich com p lete skeletons o f v erte b rates can be preserved: either
q uick b u rial in areas o f ra p id sed im en tatio n o r burial in places sheltered from
tra n s p o rta tio n fo rces.” W hile p aleo n to lo g ists a n d zo o arch aeo lo g ists need no t
be greatly co n cerned w ith im m ediately p o stm o rte m changes in o rganism s due
to th e tem p o ra l rem o teness o f th e m o rta lity event rep resen ted by a fossil (e.g.,
S h ip m an 1981b h a rd ly m en tio n s such changes), we d o need to h av e a basic
g rasp o f th e processes involved a n d th eir p o te n tia l effects on a carcass. T his is
especially so w hen one subscribes, as I do, to the p o sitio n th a t ta p h o n o m ic
effects are cum u lative, a n d w h a t h ap p e n s early in a ta p h o n o m ic h isto ry can
influence w h at h ap p en s later in th a t history.
In th e follow ing, it is im p o rta n t to realize th a t skeletonization (soft tissue
rem oval) occurs d u e to m icro -o rg an ism s such as b ac te ria a n d fungi, to sm all
org an ism s such as insects, a n d to m ed iu m a n d large o rganism s, such as
v ultures, h yenas, a n d o th e r scavenging carn iv o res. T he first tw o categories
often are discussed u n d er the term d eco m p o sitio n , even th o u g h the b acteria
an d insects are scavenging fo o d if th a t is ta k e n to m ean ea tin g the flesh o f dead
an im als w hich the co n su m er has n o t killed. B ut here I follow the literary
tra d itio n o f co n sid erin g b ac te ria a n d insects u n d e r th e general categ o ry o f
d ec o m p o sitio n ag ents an d processes. I begin w ith a d esc rip tio n o f the kinds o f
settings a n d processes resu ltin g in th o se ra re cases w hen soft tissues are
preserved before tu rn in g to the m o re co m m o n cases o f soft tissue d e c o m p o
sition, sk eleto n izatio n , a n d d isa rtic u la tio n (see A llison a n d Briggs 1991a for
m o re extensive trea tm e n t).
S o ft tissue preservation
P o stm o rtem tran sfo rm atio n by the action o f m icro-organism s has n o t been well
studied.
(M . S. M icozzi 1991:38)
Som e o f th e b est-k n o w n cases in w hich m am m alian soft tissues have preserved

over several cen turies o r m illennia co n cern the h u m an m um m ies fo u n d in
E gypt. T h ere are m u ltip le w ays fo r a m um m ified carcass to re su lt from
ta p h o n o m ic processes. T he term “ m um m ification refers to all n a tu ra l a n d
artificial processes th a t b rin g a b o u t p reserv atio n o f the body o r its p a r ts ”
(M icozzi 1991:17) T his definition o f the process m ay be to o inclusive as it can be
perceived to include th e p re serv atio n o f ju s t bones. It is m y im pression th a t the
term m um m y, w hen app lied to an entity, typically refers to a carcass w ith at
least som e preserv ed soft tissue (often a m a jo r p o rtio n o f the so ft tissue found
on the living o rg anism ) in ad d itio n to som e bones. It is in this last sense th a t I
use th e term m um m ification; to d en o te the processes th a t b rin g a b o u t the
p re serv atio n o f a t least som e soft tissues, as well as a t least som e o f the b ones, o f
an anim al. I d o n o t co n sid er the slippery q u estio n o f “ how m uch soft tissue
m u st be preserved fo r a p a rtia lly skeletonized carcass to be considered a
m u m m y ra th e r th a n sim ply a skeleton?”
M icozzi (1991:17) suggests th ere are th ree m a jo r categories o f m um m ifica
tio n processes. N a tu ra l m um m ification involves n a tu ra l processes, eith er singly
o r several, such as desiccation a n d freezing. Intentional m um m ification involves
h u m a n processes w hich deliberately exploit n a tu ra l processes, such as freezing
a carcass to preserve it. A rtificial m um m ification is also in ten tio n a l a n d involves
h u m a n s b u t the processes used involve ones n o t available n atu ra lly . T hese
in clude the use o f p reservative substances such as resins a n d oils, a n d processes
such as sm oke curing. B ecause the la tte r tw o kinds o f m u m m ification tend to
have been little ex p lored by ta p h o n o m ists, I do n o t co n sid er them fu rth e r here
(see M icozzi 1991 fo r a useful discussion o f such processes). In the re m a in d er o f
this section, th en , I review only n a tu ra l m um m ification.
It is th e extrem es o f en v iro n m e n ta l c o n d itio n s th a t ten d to fa v o r n a tu ra l
m u m m ificatio n . F o r exam ple, extrem e cold such as freezing te m p e ra tu re s tend
to fa v o r p re serv atio n o f soft tissues (see the review in G u th rie 1990). Low
tem p eratu res in h ib it b acterial activity w hich results in soft tissue d e c o m p o
sition. F reezin g also results in su b lim atio n , o r freeze-drying (M icozzi 1991:9),
w hich in h ib its p u tre fa c tio n (see below ), a process w hich requires m oisture.
H o t, d ry en v iro n m en tal c o n d itio n s are also conducive to the p re serv atio n o f
soft tissues d u e to desiccation w hich, as n o ted , in tu rn prevents p u trefactio n .
C arcasses dep o sited in m ineral salts som etim es have preserved soft tissues due
to th e d esiccatory effects o f the salts. A q u a tic en v iro n m en ts th a t are acidic an d
a n a e ro b ic also lead to soft tissue p re serv atio n d u e to in h ib itio n o f bacterial
activity (A llison 1990), as evidenced by D a n ish “ bog p e o p le ” preserved by the

tan n ic acids created in peat bogs (M icozzi 1991). M an y p la n t p ro d u c ts have
an ti-b a cterial a n d /o r insecticidal effects w hich m ight in h ib it d eco m p o sitio n o f
an im al soft tissues, b u t w h eth er these p ro p e rtie s accrue to carcasses o f anim als
d ep o sited n e a r p la n ts p ro d u c in g such p ro d u c ts is n o t clear.
T h e d u ra tio n o f so ft tissue p reserv atio n , even if it does n o t preserve dow n
th ro u g h the cen tu ries for the ta p h o n o m ist, is im p o rta n t. B ones can tak e on
d ifferent p o stm o rte m p o sitio n s relative to one a n o th e r given changes in
a tta c h e d soft tissues. W eigelt (1927/1989) d istinguishes w h a t he calls th e passive
position o f v erte b rate skeletons o f carcasses th a t d ecom posed in w a te r from
w h at he calls the contorted position o f skeletons o f carcasses th a t d ecom posed in
su b aerial contexts. H e w rites “ w hen a carcass dries o u t, sh rin k in g m echanism s
o ften b rin g a b o u t u n n a tu ra l, c o n to rte d po sitio n s n o t usually fo u n d in living
anim als. T he m o st co n sp icu o u s is th e o ften observed [dorsally concave]
c u rv atu re o f the cervical p a rt o f the spinal co lu m n ” (W eigelt 1989:103). W eigelt
(1989:88) describes the passive p o sitio n as “ sim ilar to the n a tu ra l stru c tu re o f
the b o d y .” I tak e “ th e n a tu ra l stru c tu re o f the b o d y ” to m ean the p o sitio n s o f
a n a to m ica l p a rts relative to one a n o th e r w hen the org an ism is alive. D o d so n
(1973:15) has n o ted th a t “ afte r rigor m ortis is lost, a carcass carried by flowing
w ater to its place o f b u rial will show a passive p o sitio n , ra th e r th a n th e rigid
p o sitio n o f an an im al b u ried in a state o f rigor o r afte r desiccation on la n d .”
T he o rie n ta tio n o f carcasses (see C h a p te r 6) m ay, th u s, reveal details a b o u t
dep o sition .
D istin ctio n o f the passive a n d c o n to rte d po sitio n s o f skeletons underscores
the im p o rtan ce o f detailed sp atial a n d c o n tex tu a l d a ta for tap h o n o m ic
analysis. T h a t kin d o f d a ta c a n n o t be derived fro m a p a p e r b ag full o f bones
even th o u g h all m ight have been recovered from the sam e ex cav atio n unit.
L ater in this discussion I describe p ro ced u res fo r the analy tical re artic u la tio n o f
skeletons th a t ca n be app lied w hen co n tex tu a l d a ta are lacking. B ut, even if the
skeleton o f an an im al has been a rra n g e d in a passive o r c o n to rte d (o r som e
o th er) p o sitio n , soft tissues seldom preserve; m o re o ften th a n n o t, they
decom pose.
S chafer (1962/1972:20) no tes th a t “ carcasses o f m arin e m am m als th a t die
from n a tu ra l causes d rift fo r w eeks on the surface o f th e sea,” a n d as soft tissues
d eterio rate, bones becom e d isarticu lated a n d “ fall to the [sea b o tto m ] one by
one as fro m a d riftin g sack [and th u s are] sp read over m iles o f the sea flo o r.”
C arcasses flo at d u e to gas p ro d u c tio n d u rin g decay, a n d the p h en o m en o n
S chafer describes has been term ed " b lo a t a n d flo at” (A llison a n d Briggs
1991 a:27). A llison an d Briggs (1991 a:27) indicate th a t the p ro p en sity for a
carcass to float d u rin g decay o f soft tissues is co n tro lled by the “ s tre n g th ” o f
so ft tissue, the ra te o f gas p ro d u c tio n , a n d h y d ro sta tic pressure. W hen the
tissues are sufficiently w eakened by decay the increasing volum e o f gas breaks
th e carcass a p a rt, the gas escapes, a n d the carcass sinks. R a th e r u n u su a l carcass

o rien ta tio n s, degrees o f skeletal com pleteness, a n d states o f artic u la tio n m ay
result.
S o ft tissue decom position

D ead organism s are a valuable food source in any environm ent. I f this food source
is utilized by m acro-organism s it is term ed scavenging; if it is utilized by m icrobes,
such as fungi and b acteria, it is term ed decay.
(P. A. A llison 1990:213)
T h e rem oval o f soft tissues fro m v e rte b ra te skeletons involves organism s o f

m an y sizes, from b acteria to large m am m alian carnivores. In the follow ing, I
begin w ith discussion o f the sm all o rganism s th a t rem ove soft tissues, the
process an d effect o f w hich are usually referred to as d ec o m p o sitio n o r decay,
a n d w o rk u p th ro u g h progressively larg er organism s.
M icro-organism s
S u b seq u en t to th e d ea th o f a v erte b rate, soft tissues generally d ecom pose due to
the actio n o f b ac te ria an d enzym es. F ollow ing the forensic lite ratu re , w here
m u ch research co n cern in g soft tissue d ec o m p o sitio n h as been re p o rte d , decay
involves th e d eco m p o sitio n o f p ro tein u n d e r aerobic co n d itio n s; autolysis
involves en zym atic b re ak d o w n o f tissue by the enzym es in the (once living)
o rg an ism , enzym es th a t assisted in m etab o lic functions; putrefaction involves
th e b acterial b re ak d o w n o f p ro te in u n d er a n a ero b ic co n d itio n s, w ith the source
o f th e b acteria being eith er in tern a l to the o rg an ism o r ex tern al to it (H aglund
1991:25). Soft tissue d ec o m p o sitio n usually proceeds successionally “ from
w ithin th e carcass d u e to the actio n o f enteric m icro -o rg an ism s [indigenous
b acterial m icro flora, som e o f w hich are anaerobic], a n d fro m [outside the
carcass] by co lo n izatio n [of the carcass] w ith soil m icro -o rg an ism s a n d decay
o rg a n ism s” (M icozzi 1991:37, 39, 42). P u trefa ctio n occurs only in the presence
o f m o istu re a n d in m o d era te tem p eratu res; desiccation a n d tem p eratu res less
th a n 4°C p ro h ib it p u trefactio n because these c o n d itio n s in h ib it bacterial
g ro w th (Jan aw ay 1990; M icozzi 1991:37, 38, 40). "D e c o m p o sitio n due to
b ac te rial ac tio n is ra p id in en v iro n m en ts ch a rac te rized by tem p eratu res
betw een 15°C a n d 37°C” (M icozzi 1991:41). T he ra te o f d ec o m p o sitio n slows
w hen a carcass is b u ried , in p a rt d u e to the low te m p e ra tu re o f the enveloping
sed im en t in h ib itin g b acterial g ro w th , b u t also d ue to the decreased access o f the
soft tissues to ca rrio n insects (M icozzi 1991:37). F ungi w hich m ay c o n trib u te to
the d ecay process ten d to be aero b ic a n d “ are restricted to the surface o f the
c a d a v e r” (Jan aw ay 1990:147).
H ag lu n d (1991) follow s P ayne (1965; see also C oe 1978, 1980), a n d lists six
c h ro n o lo g ical stages o f d ecom position: fresh, b lo ated , active decay, ad v anced
decay, dry, a n d skeletal rem ains. “ Sm all am o u n ts o f decaying tissue rem ain
th ro u g h o u t th e d ry stage [and] the skeletal stage m ay re ta in som e ligam entous

tissue, a rtic u la r cartilage, a n d o th e r ca rtila g e ” (H a g lu n d 1991:26). M icozzi
(1991:43) suggests th a t so ft tissue d ec o m p o sitio n w hich is free fro m the effects
o f c a rrio n insects is slow er th a n w hen such insects are p resen t, a n d is
c h aracterized by five stages: fresh, b lo ated a n d d ec o m p o sitio n , flaccidity an d
d eh y d ratio n , m um m ification, a n d desiccation a n d d isin teg ratio n . D isa rtic u la
tio n sets in by the final stage. Jo h n so n (1975) describes fo u r stages o f
d ec o m p o sitio n in sm all m am m als: fresh, b lo a t, decay, a n d dry. B loating results
fro m gases p ro d u c e d by a n a ero b ic p ro te in d ec o m p o sitio n caused by p u tre fa c
tion; th e b lo atin g stage is b rie f (tw o to five days) in w a rm en v iro n m en ts, long
(several w eeks) in cool en viro n m en ts. T he decay stage involves m o stly aerobic
p ro tein d ec o m p o sitio n , a n d c a rrio n insects a b a n d o n th e carcasses d u rin g the
dry stage (Jo h n so n 1975). D ifferent a u th o rs have b ro k e n the deco m p o sitio n
process in to different stages, p e rh a p s because differing e n v iro n m e n ta l c o n
d itio n s ex acerb ate o r in h ib it v ario u s d ec o m p o sitio n al processes, o r because
d eco m p o sitio n is a process, it is a c o n tin u u m , the stages o f w hich are artifa cts o f
o u r o b serv atio n s. W h atev er the case, the im p o rta n t p o in t is th a t d ec o m p o
sition is a process.
T he rates o f d eco m p o sitio n a n d sk eleto n izatio n are d ep e n d en t on the
en v iro n m e n t o f carcass d ep o sitio n , the cause o f d e a th , the co n d itio n o f the
carcass a t d ea th , a n d o th e r facto rs. E m aciated bodies d ecom pose m o re rapidly
th a n h ea lth y bodies. G iv en the rule o f th u m b th a t a chem ical re actio n doubles
in ra te fo r every 10°C rise in te m p e ra tu re (H a g lu n d 1991), carcasses decom pose
m o re ra p id ly in w a rm er en v iro n m en ts (e.g., C oe 1978). F ro ze n bodies preserve
qu ite well as chem ical reactio n s are negligible (e.g., G u th rie 1990). D e c o m p o
sition is m o st ra p id in carcasses lo cated o n the g ro u n d surface o r in air, o f
m o d era te ra te fo r carcasses in w ater, a n d slow est for b u ried carcasses (H aglund
1991:29). Saponification, a chem ical re actio n in w hich fa t is hyd ro ly zed an d
co n verted to ad ipocere, occurs in cool, m o ist en v iro n m en ts, a n d “ adipocere
fo rm a tio n serves to re ta rd o th e r form s o f d ec o m p o sitio n ” (H ag lu n d 1991:32).
Insects as agents o f so ft tissue rem oval

M o v in g u p the size scale o f agents a n d processes th a t rem ove soft tissues from
an im al carcasses, th e categ o ry ab o v e b ac te ria generally discussed by forensic
scientists involves insects. T hree kinds o f insects are typically observed o n soft
tissue rem ain s (M icozzi 1991:44). N ecrophagus insects are p rim ary consum ers
o f c a rrio n a n d m ainly include larv ae o f insects th a t are n o t n o rm ally n e c ro p h a
gus them selves. O m nivorous insects co n su m e c a rrio n a n d n e c ro p h ag u s insect
larvae. Predator a n d parasite insects prey o n n ec ro p h ag u s insects alone. O th er
categories o f insects include th o se sim ply living in o r o n a carcass o r w hich by
ch ance are fo u n d th ere (M icozzi 1991:44; see also Jo h n so n 1975; P ayne 1965).
O nly th e n ec ro p h ag u s an d o m n iv o ro u s insects are involved in soft tissue
rem oval.
A ctu alistic research by Jo h n so n (1975), P ayne (1965), a n d o th ers (e.g.,

R o d rig u ez a n d B ass 1983) indicates different insect tax a ten d to characterize
different stages o f soft tissue d eco m p o sitio n , alth o u g h insects are m o st
a b u n d a n t d u rin g the decay stage (betw een the b lo at a n d dry stages). T h u s, a
ta p h o n o m ist m ight gain insights in to w hen d u rin g d eco m p o sitio n a carcass was
buried if rem ains o f p a rtic u la r insect species are fo u n d associated w ith the
rem ains. T h e ta p h o n o m ist m ig h t also gain som e idea o f h ow long the carcass
was exposed as m o st insects tend n o t to exploit buried ca rrio n . E n v iro n m en tal
facto rs such as te m p e ra tu re a n d m o istu re influence insect activity, a n d th u s
different carcass d ep o sitio n al h a b ita ts m ay influence w h eth er an d , if so, w hich
insect ta x a exploit an im al ca rrio n .
F in ally , it is im p o rta n t to n o te th a t som e insects eat bone tissue (e.g.,
B ehrensm eyer 1978; see C h a p te r 9), an d som e insects m ove bones (S hipm an
an d W a lk e r 1980). T he fo rm er m ay n o t be m ed iated by the size o f the bone
relative to the size o f the insect, alth o u g h the la tte r p ro b a b ly is (S h ip m an an d
W alk er 1980). T hese tw o effects as gen erated by insects have, how ever, no t
been stu d ied in d etail in actu alistic settings, a n d p erh ap s fo r th a t re aso n these
effects h ave n o t been freq uently in ferred in p re h isto ric settings.
Joint an a to m y
B ecause large-scale ag en ts th a t rem ove soft tissues have tended to be discussed
in term s o f th eir influence o n the d isa rtic u la tio n o f v erte b rate skeletons, it is
difficult to se p a ra te sk eleto n izatio n an d d isa rtic u la tio n for discussion
p u rp o ses. T he difficulty arises as well because th e size o f large-scale tissue
rem o v ers allow s th em to rem ove soft tissue a n d d isarticu late the included bones
sim u ltan eo u sly. Sm all-scale tissue rem overs such as insects an d b acteria, due to
th eir sm all size, typically c a n n o t sep a rate a bone a n d soft tissue pack ag e from a
carcass a n d th en m ove th a t package. In the absence o f large-scale soft tissue
rem o v ers v e rte b ra te carcasses a p p e a r sim ply to fall a p a rt an d rem a in sem i
artic u la te d (C oe 1978, 1980; H aynes 1991; Jo h n so n 1975: P ayne 1965; P ayne et
al. 1968). F o r b o th sm all- an d large-scale soft tissue rem overs, one fa c to r seems
co n sisten tly to m ed iate d isarticu latio n , o r the a n a to m ica l d isasso ciatio n o f
skeletal p a rts, a n d th a t is jo in t an a to m y . T h u s, it is im p o rta n t to review jo in t
a n a to m y before discussing d isarticu latio n a n d large-scale soft tissue rem overs.
S yn d esm o lo g y, th e stud y o f ligam ents, jo in ts, an d artic u la tio n s, a n d arthrol-
ogy, th e stu d y o f jo in ts, have resulted in a w ealth o f in fo rm a tio n on jo in t
an a to m y . M u ch o f th a t in fo rm a tio n concerns the m echanics o f v ario u s jo in ts,
bu t th ere are also d a ta relevant to ta p h o n o m ic studies o f d isarticu latio n . In
p a rtic u la r, it is re aso n ab le to suppose th a t the kind o f artic u la tio n betw een tw o
skeletal elem ents, a n d the kind(s) o f soft tissues h o ld in g those tw o skeletal
elem ents to g eth er, influence the p o stm o rtem interval betw een d e a th and
d isa rtic u la tio n o f th e tw o bones.
Som e jo in ts o r artic u la tio n s result in tw o bones being relatively im m obile

relative to one a n o th e r, such as th e su tu ra l jo in ts betw een bones o f th e skull. In
o th e r jo in ts th e artic u la te d b ones are only slightly m o v ab le relative to one
a n o th e r, such as is fo u n d w ith the in terv e rteb ra l jo in ts a n d pubic sym physis o f
m am m als. F inally, the artic u la te d bones o f som e jo in ts are q u ite m obile
relative to o ne a n o th e r, such as in the elbow o r knee jo in ts o f h u m an s. These
types o f jo in ts v ary in term s o f th e degree to w hich the a rtic u la te d bones
in terlo ck w ith one a n o th e r a n d in term s o f the a m o u n ts a n d kinds o f connective
tissues h o ld in g them to g e th e r (H ild e b ra n d 1974:450—455; M icozzi
1991:49-50).
Jo in ts can be classified acco rd in g to the type o f m ateria l hold in g them
to geth er. T he follow ing is a b stra c te d from M o o re ’s (1985:33-36) discussion o f
such a classification, in w hich m a jo r jo in t categories are in d icated by n um bers
an d jo in t types w ithin a category are in d icated by letters.
1) F ib ro u s joints: bones are u nited by fibrous tissue;

A) Sutures: bones u nited by a thin layer o f fibrous tissue; little to no m ovem ent o f
bones; occur only in th e skull;
B) Syndesm osis: bones jo in ed by a sheet o f fibrous tissue, either a ligam ent o r
interosseous fibrous m em b ran e (as betw een radius and ulna); slight to consider
able m ovem ent o f bones possible;
2) C artilaginous joints: bones u nited by cartilage;
A) P rim ary cartilaginous joints: bones u nited by hyaline cartilage w hich perm its
slight bending (e.g., epiphyseal cartilage plate betw een epiphysis an d diaphysis
o f a long bone, rib to sternum connection);
B) Secondary cartilaginous joints: a rticu lar surfaces o f articu lated bones covered
w ith hyaline cartilage and cartilage surfaces jo in ed by fibrous tissue a n d /o r
fibrocartilage; stro n g a n d slightly m oveable (e.g., inter-vertebral centra, pubic
sym physis);
3) Synovial joints: norm ally provide free m ovem ent betw een jo in ed bones; have four
distinguishing features, (1) jo in t cavity, (2) a rticu lar cartilage, (3) synovial m em
brane, an d (4) fibrous capsule; accessory ligam ents strengthen the capsule an d lim it
m ovem ents o f the jo in t in undesirable directions;
A ) Plane joints: perm it sliding, as betw een tw o carpal bones;
B) H inge joints: perm it m ovem ent in one axis a t right angles to bones involved, as
h u m an elbow;
C) P ivot join ts: perm it ro ta tio n aro u n d a lo ngitudinal axis th ro u g h a bone, as the
atlas-ax is joint;
D) C ondyloid joints: allow m ovem ent in tw o directions, as the m e ta c a rp a l-
p halange jo in t o f hum ans;
E) Saddle joints: allow m ovem ent in tw o directions as a rticu lar surfaces are saddle
shaped, as the c a rp a l-m e ta ca rp a l jo in t o f the h u m an thum b;
F) Ball and socket joints: highly m ovable, m u lti-directional m ovem ent, as the
hu m an hip an d sho u ld er joints.
T his classification is so m ew h at tax o n o m ically stru c tu re d , as the type o f

connective tissues seem s m o st im p o rta n t, follow ed by finer-scale details o f the
a rtic u la tio n , such as the jo in t’s m o rp h o lo g y a n d m obility.
D isa rtic u la tio n involving only m icro -o rg an ism s (b acteria an d insects) seems
to begin w ith m obile jo in ts, an d progresses to the slightly m obile jo in ts m an y o f
w hich have in trao sseo u s ligam ents hold in g them to g eth er. Im m obile su tu re
jo in ts are the last to d isarticu late because they are held to g eth er by ligam ents
an d th e b o n es have, in a w ay, interw oven w ith each o th e r (M icozzi 1991). These
general n a tu ra l sequences m ay vary slightly d ep en d in g on the kinds o f
organ ism s involved in so ft tissue rem oval.
D isarticulation
F ro m a ta p h o n o m ic perspective, the im p o rtan ce o f soft tissue rem oval resides
in th e fu n ctio n o f such tissues to hold the bones to g eth er in the fo rm o f a
skeleton. As the so ft tissues are rem oved, a skeleton will fall a p a rt; th a t is, bones
th a t w ere artic u la te d in life will becom e d isarticu lated , a n d p erh ap s eventually
sp atially disassociated o r scattered (H ill 1979a, 1979b). T he access o f sca
vengers, w h eth er m am m als, insects, o r b acteria, to carcasses influences the ra te
o f sk eleto n izatio n . B urial effectively rem oves carcasses from m any large
scavengers, an d n o t only re ta rd s the ra te o f soft tissue d ec o m p o sitio n , b u t also
the ra te o f b o n e d isarticu latio n an d scattering. A b ler (1985), fo r exam ple,
rep o rts on a dom estic sheep ( Ovis aries) th a t becam e m ired in a m u d d y sw am p.
O nly the d o rsal p a rt o f the b ack a n d the h ead were n o t bu ried , a n d only these
exposed p o rtio n s w ere “ m o st vu ln erab le to scavenging, decay, a n d w e ath er
ing” (A bler 1985:250). L ym an (1989b) re p o rts sim ilar results fo r carcasses o f
N o rth A m erican w apiti (Cervus elaphus) killed by a volcanic e ru p tio n ; only
p o rtio n s o f carcasses n o t buried by volcanic ash were exposed to scavenging
an d w eatherin g , b u ried p o rtio n s o f carcasses w ere still a rtic u la te d an d
u n w eath ered a year a n d a h a lf afte r the eru p tio n . M icozzi (1991:4 9 ,51) suggests
“ d ec o m p o sitio n o f soft tissue occurs from the to p (head) d o w n w a rd " a n d thus
“ th e m an d ib le a n d skull generally have the first o p p o rtu n ity to becom e
d isarticu lated from the re m a in d er o f the sk eleto n .” A b le r’s (1985) suggestion
m ay th u s sim ply co n cern a p artic u la ristic fa c to r th a t exacerbates the generality
suggested by M icozzi (1991).
Hill (1979b) describes a d isa rtic u la tio n sequence for the topi (D am aliscus
ko rrig u m ), an d Hill a n d B ehrensm eyer (1984) describe sim ilar sequences for the
w ildebeest (C onnochaetes taurinus), dom estic cow (Bos sp.), B urchell's zebra
(E quus burchelli), a n d G ra n t's gazelle (G azella granti). All are sum m arized in
T ab le 5.5. Hill a n d B ehrensm eyer (1984) fo u n d g reat overall sim ilarity betw een
th e five d isa rtic u la tio n sequences. W hile they d o c u m e n t som e differences
betw een them , they fo u n d those differences co u ld n o t be explained by reference
to tax o n o m ic v aria tio n o r by v a ria tio n in carcass size. Im p o rta n tly , M icozzi
(1991:50) co rrectly re p o rts th a t H ill's studies did n o t identify a n d distinguish
th e activity o f d eco m p o sitio n o rganism s (b acteria a n d insects) fro m large-scale
scavengers (e.g., hyenas).
T ab le 5.5 R a n k order o fjo in t disarticulation in fiv e m am m alian taxa (fro m

H ill 1979b, H ill and Behrensm eyer 1984). N um bers denote order o f
disarticulation; 1 = fir s t jo in t to disarticulate, 30 = last jo in t to disarticulate.
Jo in t type (a fter M oo re 1985) given in parentheses; S, synovial jo in t; C,
cartilaginous jo in t; F, fib ro u s jo in t
D om estic B urchell’s G ra n t's

Joint Topi W ildebeest Cow Z ebra G azelle
c ran iu m -m an d ib le (S - hinge) 5 4.5 2 2 6

c ra n iu m -a tla s (S - hinge) 20 4.5 6 8.5 6
atlas-ax is (S - pivot) 7 20 13 22.5 6
a x is-th ird cervical (C - secondary) 29 30 24 26 30
cervical-cervical (C - secondary) 30 29 25 29 29
seventh cervical-first thoracic 27.5 25 22.5 22.5 23
(C - secondary)
th o racic-th o racic (C - secondary) 22 24 26 20 17
th o racic-rib (S - condyloid) 23 21 21 11 11
th irteenth th o racic-first lum bar 25.5 22.5 28 27 24.5
(C - secondary)
lu m b a r-lu m b a r (C - secondary) 25.5 25 29 28 27
seventh lu m b ar-sa cru m (C - secondary) 27.5 27 30 22.5 28
sacrum -first caudal (C - secondary) 2.5 9 20 8.5 2.5
c au d al-cau d al (C - secondary) 2.5 3 9 4 2.5
sacru m -in n o m in ate (S - plane) 24 28 22.5 22.5 24.5
in n o m in ate-fem u r (S - ball and socket) 11 9 8 6.5 13
fem u r-tib ia (S - condyloid) 19 13.5 15 12.5 20.5
tib ia -ta rsa ls (S - plane) 14.5 13.5 15 12.5 20.5
tarsals-m e tatarsal (S - plane 21 15.5 19 25 20.5
m etatarsal-first phalan x (S - hinge) 17 15.5 11 17 18.5
first phalan x -seco n d phalanx 17 17.5 10 17 15.5
(hind) (S - hinge)
second p h a la n x -th ird phalanx 17 19 4 5 15.5
(hind) (S - hinge)
fo relim b-body (none) 1 1 1 1 1
scap u la-h u m eru s (S - hinge) 4 2 5 3 4
h u m eru s-rad iu s, u lna (S - hinge) 11 17.5 16 17 18.5
rad iu s-u ln a (F - syndesm osis) 14.5 22.5 27 30 26
rad iu s-carp als (S - plane) 6 11 14 14 13
carp als-m etacarp al (S - plane) 8.5 12 18 17 13
m etacarp al-first p h alan x (S - hinge) 8.5 7 12 10 9
first phalan x -seco n d p halanx 13 9 7 17 9
(fore) (S - hinge)
second p h a la n x -th ird phalanx 11 6 3 6.5 9
(fore) (S - hinge)
T o d eterm in e ifjo in t a n a to m y , as ch a rac te rized by M o o re ’s (1985) typology,

influenced the o rd e r o f jo in t d isarticu latio n as determ in ed by Hill, I assigned
each jo in t in T able 5.5 to one o f M o o re ’s types. H ill (1979b) notes th a t m ost o f
the jo in ts he co n sid ered are synovial jo in ts. In fact, 19 o f the 30 jo in ts in T able
5.5 are synovial jo in ts (five kinds o f synovial jo in ts are represented), 10 are
seco n d ary ca rtila g in o u s jo in ts, a n d one is a fibrous syndesm osis jo in t. I
calcu lated th e averag e ra n k o f d isa rtic u la tio n fo r all jo in ts o f three types for the
to p i; o th e r jo in t types w ere only rep resen ted by one o r tw o jo in ts. T he average
ra n k o f d isa rtic u la tio n for the 10 seco n d ary ca rtila g in o u s jo in ts is 21.5 ( ± 10.3),
for the five synovial p lane jo in ts the average ra n k is 14.8 ( ± 7 .8 ) , a n d fo r the 10
synovial hinge jo in ts the average ra n k is 12.4 ( ± 5.4). T hese cru d e d a ta suggest
jo in t a n a to m y does influence the o rd e r o f jo in t d isarticu latio n d o cu m en ted by
Hill.
H ill’s (1979a, 1979b. 1980; H ill a n d B ehrensm eyer 1984. 1985) research
focused o n spatially isolated carcasses a n d jo in ts. A fte r studying several loci
w ith m u ltip le carcasses o f d om estic cow s in close p roxim ity, T o d d (1983a:40 -
41) re p o rts th a t “ as th e n u m b e r o f carcasses increases a n d as the d istance
betw een carcasses decreases, the g re ater the p o te n tia l fo r [a d isarticu latio n
sequence] different from th a t observed by H ill (1979b).” F u rth e r, if a carcass is
on th e p erip h ery o f a cluster o f carcasses, it will tend to d isarticu late m ore
ra p id ly th a n a carcass n e a r the ce n te r o f the clu ster d u e to the fo rm e r’s g re ater
accessibility to scavenging carn iv o res a n d tram p lin g agents (T o d d 1983a).
T o d d (1983b:54) suggests som e o f the differences betw een the d ec o m p o sitio n
a n d d isa rtic u la tio n o f a n isolated carcass fro m a g ro u p o f carcasses results from
the c reatio n by the la tte r o f a “ distinctly different ta p h o n o m ic m icro-clim ate
th a n th a t w hich w o u ld o p erate on a single carcass in the sam e settin g .” T his is
sim ilar to , b u t at a different (m ulti- o r betw een-carcass) scale th a n th a t (single-
o r w ith in -carcass) used by M icozzi (1991:42) w ho notes th a t the te m p e ra tu re o f
d eco m p o sin g so ft tissue m ay differ fro m the am b ien t sedim ent o r air te m p e ra
tu re “ d u e to the creatio n o f an in tern a l m icro e n v iro n m e n t by th e ac tio n o f
b a c te ria .”
G iv en th e im p o rta n c e o f m o istu re fo r skeleto n izatio n , the season o f d e a th in
tem p erate clim ates m ay exert stro n g influences o n rates o f sk eleto n izatio n an d
th u s on d isarticu latio n . In tem p erate N o rth A m erica, T o d d (19 8 3 b :7 1) sug
gests “ we co u ld expect a g re ater degree o f skeletal d isa rtic u la tio n a n d dispersal
w here an im als died in the late spring o r sum m er th a n in those w here d ea th
o ccu rred in the fall o r w in ter” (see also G a llo w ay et al. 1989). T he presence o f
snow d rifts ex acerb ates the effects o f m o istu re as the d rifts m elt (T o d d 1983a,
1983b:72). W ith the exception o f the effects o f snow d rifts, H aynes (1991) has
com e to conclusio ns sim ilar to T o d d ’s, even th o u g h H aynes studied carcasses
o f A frican elep h an ts, anim als several tim es larger th a n the d om estic cows
stu d ied by T o d d . F inally, M icozzi (1986) fo u n d th a t previously frozen c a r
casses ten d to d isarticu late m o re ra p id ly th a n fresh carcasses.
N u m e ro u s ac tu alistic studies in d icate th a t as the frequency o f p re d a to rs
increases relative to th e frequency o f prey, the intensity o f e x p lo itatio n o f a

given prey carcass by p re d a to rs will increase, a n d th u s the rate o f skeleto n iza
tio n a n d ex ten t o f b o n e d isp ersal will also b o th increase (D ’A n d re a an d
G o tth a rd t 1984; H ay nes 1980a, 1980b, 1982; H ew son a n d K olb 1986; T o d d
1983a). In his stu d y o f N o rth A m erican w olves (Canis lupus), H ay n es (1980b,
1982) fo u n d th a t as the n u m b er a n d h u n g er o f w olves increased, the intensity
a n d ex ten t o f carcass e x p lo itatio n increased, an d , the wolves ten d ed to m ore
fully exp lo it carcasses o f anim als they h ad killed th a n carcasses o f d ead anim als
they fo u n d . H e described the sequence o f prey carcass u tiliza tio n a n d the
sequence o f b o n e d e stru c tio n in o rd e r to a p p ro x im a te stages o f carcass
ex p lo itatio n (T ab le 5.6).
A sequence o f ex p lo itatio n o f prey carcasses by A frican carn iv o res was
developed by B lum enschine (e.g., 1986a, 1986b, 1987), w ho m o n ito re d n u m er
ous carcasses o f an im als b o th killed a n d co n su m ed by th e carn iv o res, a n d
scavenged carcasses th e original p re d a to r fo r w hich w as u n k n o w n . T h a t
sequence (T able 5.7) is sim ilar to the one H aynes d o cu m en ted fo r N o rth
A m erican w olves, a lth o u g h d etailed co m p ariso n betw een the tw o is n o t
possible because H a y n es’ sequence is n o t as detailed as B lum enschine’s.
B lum enschine’s general carcass c o n su m p tio n sequence indicates th a t tissue
ex tern al to bones, o r flesh, is ea te n in a p a rtic u la r sequence by m am m alian
scavengers first, a n d th en th a t sam e sequence is follow ed w hen scavengers
exploit tissues such as m arro w a n d “ p u lp ” th a t are inside bones (T able 5.7).
T he so ft tissue c o n su m p tio n sequence tends to begin w ith the skeletal p arts
w ith large a m o u n ts (by w eight) o f tissue (B lum enschine a n d C a ro 1986), an d
ends w ith the skeletal p a rts w ith low am o u n ts o f tissue (F ig u re 5.14; rs = 0.673,
/ 5< 0 .0 5 ). O m ittin g the rib-cage a n d cervical v erteb rae, w hich B lum enschine
(1986a:644) suggests are a n o m a lo u s because these p a rts are “ difficult to gain
com p lete access to w ith o u t first eatin g p a rts p o ste rio r o r (less frequently)
a n te rio r to th e m ,” significantly stren g th en s the re la tio n sh ip betw een the
c o n su m p tio n sequence a n d flesh w eight (rs — 0.833, P = 0.02).
H a y n es’ (T able 5.6) an d B lum enschine's (T able 5.7, F igure 5.14) d a ta are
im p o rta n t because they help explain H ill’s (T able 5.5) sequence o f d isa rtic u la
tion. H ill’s sequence w as ap p a re n tly influenced by m am m alian scavengers, bu t
to an u n k n o w n degree, an d th u s it is n o t su rp risin g th a t it tends to in d icate the
ap p e n d ic u lar skeletal elem ents becom e d isarticu lated p rio r to the axial skele
to n . T his is sim ilar to th e sequence o f soft tissue c o n su m p tio n fo u n d by H aynes
a n d B lum enschine. M o re d etailed c o m p ariso n s c a n n o t be p erfo rm ed w ith the
d a ta a t h an d .
Hill (1979a, 1979b, 1980; Hill a n d B ehrensm eyer 1984, 1985) suggests the
n a tu ra l d isarticu latio n sequences he describes can be c o m p ared to d isa rtic u la
tio n sequences p ro d u c ed by h u m an s b u tch erin g sim ilar tax a as a w ay to
m o n ito r how jo in t a n a to m y influences b u tch ery . H ill (1979a; H ill a n d B ehrens
m eyer 1985) finds g re at sim ilarity betw een his n a tu ra l sequences a n d one N o rth
A m erican P a leo in d ian bison (Bison sp.) kill site assem blage, a n d also finds
5.6 Sequence o f dam age to bones o f ungulates exp lo ited b y N orth A m erican wolves (fro m H aynes 1982; with
ications an d additions fr o m H aynes 1980a)
l elem ent Light to m oderate utilization Full utilization Heavy utilization
les still articu lated , an g u lar process one hyoid present, an g u lar disarticulated, ventral border
lightly gnaw ed, hyoids present process gnaw ed b roken off, medial (lingual) su
gnaw ed
no d am ag e to bones nasals to oth-scratched nasals ragged a t ends, an terio
prem axillaries m ay be broken
ae som e processes gnaw ed m ost vertebral processes gnaw
and rem oved
edges o f ilia and ischia gnaw ed, ilia and ischia partly gone only stum ps o f ilia and ischia
trab ecu lar bone exposed rem ain
a vertebral b o rd er gnaw ed and vertebral b o rd er splintered and blade crunched an d splintered
ragged jagged, d isarticulated from glenoid fossa p o rtio n may rem
hum erus
us greater tuberosities gone or tuberosities gone, to o th scoring proxim al end gone, ^ o f proxi
furrow ed on shaft shaft gone, distal condyles gn
tro ch an teric stu m p left, g reater medial condyle gouged, surface o f distal end gone, head nearly g
tro ch lear rim scored at right lateral condyle gone, trochlca shaft breaking up
angle to long axis, m inor dam age well opened, tro ch an teric stum p
to m edial condyle gone, to o th m arks u ndercut head
lateral proxim al end grooved or crest open o r gone, m edullary proxim al end gone, fracture e
beveled, som e furrow ing and cavity exposed a t lateral sh arp with localized rounding
gouging proxim al end, m edial edges tarsals still articu lated, shaft b
last
R A L O B SE R V A T IO N S
tw o o r three lim bs still three o r four lim bs d isarticulated few m etapodials and phalang
articu lated with body, three or from body present
four m atap o d ials an d all
phalanges present
T ab le 5.7 R a n ke d general consum ption sequence,

fro m fir s t to last eaten (fro m Blum enschine 1986a)
C arcass unit R ank C arcass u n it Rank
H in d q u a rter flesh H indlim b m arrow

pelvis 1 fem ur 13
lum bar 3 tibia 14
fem ur 2 m etatarsal 16
tibia phalanges 15
F o re q u a rte r flesh F orelim b m arrow
rib-cage 4 hum erus 17
scapula 6 radius 19
hum erus 5 m etacarpal 20
rad iu s-u ln a 9 phalanges 18
cervical 8 H ead contents
H ead flesh m axilla pulp 21
tongue 10 m andible pulp 23
m andible 11 brain 22
m axilla 12 fro n tal pulp 24
20000
consumption sequence
15000
Flesh Weight (gm)
TJ
CD
E
3
CO
10000 c *o
o
o ©
E
13
(/)
C
o
5000 o
CO
oj
0 m nzO
a> (/) CO
> (IS CD 3k_ 3 o
E _Q aJ Q_ "D
a) 0) E o a) CO >
k_ aJ
-Q E o
ZJ </) a>
o
Figure 5.14. B lum enschine’s (1986a, 1986b) con su m p tio n sequence p lotted
against flesh weight. See text for discussion.
co n sisten t differences betw een the taxon-specific n a tu ra l sequences a n d th a t

bison assem blage. H e th u s concludes “ d isa rtic u la tio n is extrem ely u n ifo rm in
different large h erb iv o re species, w h atev er the agent o f d ism em berm ent,
h u m a n o r n o n -h u m a n ” (H ill an d B ehrensm eyer 1985:144).
H u m a n s, like large q u a d ru p e d a l carn iv o res, have a size ad v a n ta g e relative to
m any p rey carcasses such th a t they ca n m ove carcass p a rts ra th e r th a n sim ply
cause a carcass to fall a p a rt as b acteria an d ca rrio n insects do. T h a t ad v a n ta g e
is ex acerb ated by the use o f to o ls by h u m an b u tch ers to aid in task s o f skinning,
defleshing, a n d dism em b erm ent. T o o l use, in tu rn , o ften creates traces in the
form o f b u tch erin g m ark s w hich allow the an aly st to identify h u m a n s as agents
o f d isarticu latio n (C h a p te r 8).
It is im p o rta n t to n o te th a t, th u s far, discussion o f d isa rtic u la tio n has been
lim ited to sep a ratio n o f bones at th eir p o in ts o f a rtic u la tio n , o r at jo in ts. M ost
a u th o rs suggest th a t jo in ts are th e easiest places fo r n o n -in d u stria l h u m a n
b u tch ers to d isarticu late b ones o f a carcass (e.g., B inford 1978; G ilb e rt 1979;
R ead 1971). B u t b o n es can also becom e an a to m ica lly d isasso ciated by b re a k
ing th em d u rin g b u tch ery (o r by co n su m p tio n by large q u a d ru p e d a l c a rn i
vores), as n o ted by B in fo rd (1978:63) w hen he w rites “ u n d e r c ertain co n d itio n s
(e.g., fro zen carcasses) bones m ay be b ro k e n d u rin g [dism em berm ent]” (see
C h a p te r 8).
Analysis o f disarticulation and scattering

It is well w orth while considering fragm ents in th eir context, as found. Several
undistinguished chips an d flakes o f bone, fo u n d together, m ay have jo in s preserved
w hich will enable som ething significant to be rebuilt.
(I. W. C ornw all 1956:197)
If articulation is defined as tw o o r m ore skeletal elem ents being in th eir p ro p e r

an a to m ica l po sitio n s relative to one a n o th e r, a n d w ithin a cen tim eter o f each
o th er if n o t in fact to u ch in g , a n d if scattering is defined as increasing the spatial
d istance betw een an ato m ically related bones (see the G lossary), th en clearly
b o th are m o st readily m easured in the field d u rin g the collection o f faunal
rem ains. R eco rd in g w hich b o n es w ere fo u n d in c o n ta c t w ith one a n o th e r a n d
reco rd in g th e d istan ce betw een b ones lying close to, b u t n o t in c o n ta c t w ith, one
a n o th e r are preferab le to sim ply n o tin g w hich bones cam e fro m th e sam e o r
ad jac en t ex cav atio n un its, a n d th e n tu rn in g the bones a n d th o se no tes over to
the ta p h o n o m ist fo r study. Y et the latter is com m onplace.
A b u n d a n t eth n o arch ae o lo g ic al d a ta in d icate T h o m as (1971:367) w as c o r
rect tw en ty years ago w hen he w ro te “ the d ietary practices o f m a n ten d to
d estro y a n d disperse th e b o n es o f his prey-species.” It is the ra re case w hen
b ones th a t w ere artic u la te d in life are fo u n d artic u la te d in archaeological (or
eth n o arch ae o lo g ic al) contexts. (A lth o u g h artic u la te d skeletal elem ents m ay be
m ore co m m o n in p aleo n to lo g ical co n tex ts th a n archaeological ones, there are
T ab le 5.8 Jo in t articulation data f o r bison bones fr o m the Casper site and the
H orner I I site. S ee te x t fo r discussion
C asper H o rn e r II
P o ten tial8 P otential2

Joint H ill's n C JF (% articulated) H ill’s n C JF (% articulated)
scap u la-h u m eru s 1 0.11 70(1.4) 17 0.70 77 (22.1)

h u m eru s-rad iu s/u ln a 38 4.28 65 (58.5) 45 1.84 122 (36.9)
rad iu s-carp als 22 2.48 37 1.52
carp als-m etacarp al 21 2.36 61 2.50
m e ta c a rp a l-1st phalanx 26 2.93 57 2.33
1st p h a lan x -2 n d phalanx 25 2.82 55 2.25
2nd p h a la n x -3 rd phalanx 25 2.82 52 2.13
fem u r-tib ia 15 1.69 53 (28.3) 30 1.23 95 (31.6)
tib ia-tarsals 19 2.14 45 (42.2) 50 2.05 110(45.4)
tarsals-m etatarsal 24 2.70 52 (46.2) 89 3.64 126 (70.6)
m e ta ta rs a l-1st phalanx 26 2.93 96 3.93
1st p h a lan x -2 n d phalan x 25 2.82 94 3.85
2nd p h a lan x -3 rd phalanx 25 2.82 92 3.77
H ill's N = 444 = 1221
Note:
a P otential num ber o f articu latio n s (% o f p o ten tial th a t are actually articulated).
no d a ta I am aw are o f to su b sta n tia te such a claim .) In the follow ing, I review

several tech niq ues fo r m easu rin g d isarticu latio n a n d several o th ers fo r m easu r
ing scattering. It is th u s im p o rta n t to n o te th a t the tw o processes a n d their
results are n o t m u tu ally exclusive. T he analy tical techniques described below
are g o o d ones, irrespective o f w h eth er they are used to m easure d isarticu latio n
o r dispersal. T h e sub -section h eadings are, th en , heuristic devices ra th e r th a n
logical ones.
A rticulation and disarticulation

Hill (1979a, 1979b) m easures the o rd e r o f jo in t d isa rtic u la tio n w ith the
follow ing eq u atio n :
100nH-NR = C JF [5.1]
w here N = the to ta l n u m b e r o f all in ta c t o r a rtic u la te d jo in ts in a bone

assem blage, n = th e to ta l n u m b e r o f a p a rtic u la r skeletal jo in t in a bone
assem blage, R = th e n u m b e r o f tim es a p a rtic u la r jo in t occurs in a single
skeleton, a n d C J F is th e c o rrected jo in t frequency. A s an exam ple, I calculated
C J F values for th e fro n t an d re a r lim bs o f N o rth A m erican bison skeletons as
d o cu m en te d at the C asp er Site (F riso n 1974) an d the H o rn e r II site (T odd
1987b). T h o se frequencies (T able 5.8) in d icate the s c a p u la -h u m e ru s (shoulder)
jo in t a n d th e fe m u r-tib ia (knee) jo in t w ere the “ first” to be d isarticu lated
Figure 5.15. O rder o f jo in t d isarticu latio n a t C asper an d H o rn er II sites as

determ ined by H ill’s (1979a, 1979b) m ethod. A bbreviations: scap, scapula; hum ,
hum erus; rad, radiu s/u ln a; carp, carpals; m tcrp , m etacarpal; F lp h , forelim b first
phalanx; F 2 p h, forelim b second p halanx; F 3 p h , forelim b th ird p halanx; fem,
fem ur; tib, tibia; tars, tarsals; m ttrs, m etatarsal; R lp h , rear lim b first phalanx;
R 2ph, rear lim b second p halanx; R 3ph, rear lim b th ird phalanx.
accordin g to H ill’s m odel, b u t then the “ o rd e r” o f d isarticu latio n differs

betw een th e tw o assem blages (F ig u re 5.15). S p e a rm a n ’s rh o betw een the tw o
sets o f C J F values is w eak (rs = 0.52, P = 0.07), suggesting th a t bison lim bs at
these tw o sites w ere d isarticu lated in ra th e r different orders.
H ill’s (1979a, 1979b) m eth o d is fo u n d ed on calcu latin g the p ro p o rtio n a l
frequency o f a rtic u la te d jo in ts fo r each type o f jo in t rep resen ted , w eighted by
how m an y tim es a jo in t type occurs in a skeleton a n d th e to ta l n u m b e r o f all
types o f a rtic u la te d jo in ts. T h u s the an aly st m ay find one distal h u m eru s an d
one p ro x im al ra d iu s th a t are n o t artic u la te d , o r the analyst m ay find only the
distal h u m eru s o r th e p ro x im al rad iu s, o r the analyst m ay find n eith er the distal
h u m eru s o r the p ro x im al rad iu s, a n d the C J F value fo r the elbow jo in t w ould be
the sam e in all th ree cases because it only tallies the artic u la te d jo in ts. T h a t is,
H ill’s m e th o d a n d e q u a tio n [5.1] m ay be useful fo r co m p a ra tiv e p u rp o se s (e.g.,
F ig u re 5.15), b u t this does n o t co n sid er how m an y jo in ts are possible given the
b ones recovered, a n d it does n o t co n sid er h ow m an y o f th o se possible jo in ts are
n o t artic u la te d .
T o d d (1987b: 142-146) suggests the an aly st co u n t the n u m b er o f artic u la te d
jo in ts th a t are possible fo r a b o n e assem blage, a n d th e n calculate th e p ercen t
age o f th o se possible jo in ts th a t are ac tu ally a rtic u la te d . B u rg ett (1990:163)
80
□ Casper
S 13 Horner I
| 60
S
iS
o
C 40
<
w
c
o
Z1 20
o
vP
O'
I
0
X>
03
E
0) i (/)
E k_
OJ
JZ
Figure 5.16. P ro p o rtio n o f articu lated jo in ts a t C asper an d H o rn e r II sites as

determ ined by T o d d ’s (1987b) m ethod. A bbreviation: scap, scapula; hum ,
hum erus; rad /u l, radius-ulna; fem, fem ur; tib, tibia; tars, tarsals; m ttrs,
m etatarsal.
labels this th e “ percen tag e o f p o te n tia l a rtic u la tio n s ,” o r P PA . F o r exam ple, an

assem blage o f 15 distal left h um eri a n d 20 p ro x im al left radii has the p o ten tial
to h ave 15 a rtic u la te d left elbow jo in ts. N o te th a t the p o ten tial n u m b e r is the
low er o f the tw o values because b o th bones are req u ired to m ake a n a rtic u la tin g
jo in t. T h e p ro c ed u re assum es th a t all specim ens are from individuals o f the
sam e sex a n d o n to g en ic age. If 10 o f th o se 15 p o te n tia l elbow jo in ts are
artic u la te d o r in tact, th en 67% (10 o f 15 = P P A ) are artic u la te d . Such calcu la
tio n s fo r selected jo in ts in the C asp er a n d H o rn e r II bison assem blages (T able
5.8) indicate, fo r exam ple, a b o u t the sam e p ro p o rtio n o f knee a n d an k le jo in ts
are artic u la te d (an d d isarticu lated ) at b o th sites, b u t m ore sh o u ld er an d ankle
jo in ts at H o rn e r II are artic u la te d th a n at C asper, a n d m o re elbow jo in ts are
artic u la te d at C asp er th a n at H o rn e r II (F ig u re 5.16).
B urg ett (1990:163) labels a n o th e r statistic the “ p ercentage o f surviving
a rtic u la tio n s,” o r PSA . T he PSA is “ the frequency o f artic u la tio n s recorded for
each jo in t observed, divided by the frequency in w hich each jo in t occurs in a
com p lete skeleton, [and m ultiplying the result by 100 to provide] a percentage
v alu e" (B u rg ett 1990; 155). T his statistic w as designed specifically fo r actualistic
research w herein the n u m b e r o f anim als (a n d th u s the frequency o f p o ten tial
jo in ts) is know n. In B u rg ett’s case, the n u m b er o f anim als w as alw ays one.
In arch aeo lo g ical cases, the an aly st could use the M N I as an a p p ro x im a tio n
o f th e n u m b er o f anim als in an assem blage o f fau n al rem ains, a n d m ultiply

the n u m b er o f jo in ts p er skeleton by the M N I to derive a d e n o m in a to r
for calcu latin g th e PSA . F inally, B urgett (1990) suggests co m p arin g PSA
values fo r the to ta l v erteb ral co lu m n (e.g., 27 jo in ts in one bison [Bison bison]
w ould be the d en o m in a to r) w ith the PSA values fo r the com plete forelim b
(d e n o m in a to r = 16 w hen M N I is 1) a n d PSA values fo r the com plete hindlim b
(d e n o m in a to r = 14). F o relim b jo in ts are: th o ra c ic -sc a p u la , sc a p u la -h u m e ru s,
h u m e ru s-ra d iu s-u ln a , ra d iu s-u ln a -c a rp a ls, c a rp a ls-m e ta c a rp a l, m e ta c a rp a l-
first p h alan g es, first p h alan g e s-se c o n d ph alan g es, second p h a la n g e s -th ird
p h alan g es, tim es tw o fo r b o th lim bs. H in d lim b jo in ts are: in n o m in a te -fe m u r,
fe m u r-tib ia , tib ia -ta rsa ls , ta rs a ls-m e ta ta rs a l, m e ta ta rsa l-first ph alan g es, first
p h a lan g e s-sec o n d phalan ges, second p h a la n g e s-th ird phalanges, tim es tw o for
b o th lim bs.
B urgett (1990) fo u n d little relatio n betw een d isarticu latio n o f bison an d
w apiti (Cervus elaphus) skeletons scavenged by coyotes (Canis latrans) an d the
live w eight o f th e anim als. Sim ilarly, th ere w as no re la tio n betw een the
on to g en ic age o f th e prey carcasses a n d the PSA . H e did find, as H ill (1979b)
did, th a t the axial skeleto n rem ain ed artic u la te d longer th a n bones o f the lim bs.
O nce th e state o f a rtic u la tio n a n d d isa rtic u la tio n are recorded, the an a ly st can
begin to stu d y o th e r p h en o m en a , such as frequencies o f b u tch ery m arks,
season al v a ria tio n in an im al d e a th , p o st-d e p o sitio n a l d istu rb an ce , b o n e w e a th
ering, a n d the sedim en tary co n tex t in seeking an ex p lan a tio n fo r v a ria tio n in
th e observed state o f a rtic u la tio n a n d d isarticu latio n .
D ispersal and scattering

T he u su al term s fo r analy tical re -a rtic u la tio n are refitting studies o r studies o f
conjoining pieces. Such studies h a d th eir o rigin in the arch aeo lo g ical analysis o f
sto n e artifacts. Schild (1976:96), fo r exam ple, re p o rts th a t th e “ d em an d in g
stu d y tech n iq u e involving m atch in g stray flint flakes a n d blades, a n d even
com p leted to o ls, w ith th e flint ‘co res’ fro m w hich the b lan k s used fo r tool
m a n u fa c tu re w ere stru ck , generates w h a t we call articulation nets, w hich define
the b o u n d arie s o f a single o cc u p atio n u n it.” C ah en a n d M o eyersons (1977:813)
re p o rt th a t th eir
reassem bly o f w orked stones (th a t is, jo in in g to g eth er the tools, flakes, core and
fragm ents struck off from the sam e block) [produced] fitting artefacts at several
different depths from w hich different rad io carb o n dates have been o b tain ed (the
vertical distance betw een jo in in g pieces som etim es exceeds 1 m ) . . . the vertical
d istribu tion o f the elem ents o f the reconstructed cores does n o t follow the o rd e r in
w hich they have been struck off. T his excludes the hypothesis o f a later reutilisation
o f older artifacts.
T hese o b serv atio n s led them to conclude th a t su b su rface m ovem ent o f the
sto n e artifa cts h ad resu lted in th eir differing vertical proveniences.
S tudy o f the h o riz o n ta l a n d vertical proveniences o f refitting o r conjoining

pieces has becom e a v aluable analy tical to o l fo r u n d e rsta n d in g site fo rm a tio n a l
processes (H o fm an 1981, 1986). Such m echanical refitting is n o t unlike
w o rk in g th ro u g h a com plex jigsaw puzzle w herein the sam ple o f pieces one is
try ing to fit to g e th e r actu ally consists o f several v ario u sly incom plete a n d
com p lete puzzles. It w as atte m p te d w ith fa u n al rem ains soon afte r in trig u in g
results b egan to em erge fro m the study o f refitting sto n e a rtifacts. F o r exam ple,
J o h n so n (e.g., 1982, 1987) refit bo n e pieces d istrib u te d across a single h o riz o n
tal pro v en ien ce to id entify “ b u tch erin g a rtic u la tio n n ets.” C o n jo in in g pieces o f
skeletal elem ents fro m different vertical proveniences w ere re p o rte d by V illa et
al. (1986:433) in th eir a tte m p t to identify clusters o f bones th a t h a d been
“ processed a n d d iscard ed a t the sam e tim e .” Such clusters, w hen d istu rb ed by
p o st-d ep o sitio n al (an d by im p licatio n p o st-b u tch erin g ) processes, w ere id en ti
fied by V illa et al. (1986:433) as clusters o f b o n e “ pieces th a t can be refitted an d
by a h ig h er (th an average) d en sity o f pieces w ithin a restricted area as show n by
h o riz o n ta l a n d vertical p lo ts o f th eir observed p o sitio n s.”
V illa et al. (1986) exam ined th e pieces o f p a rtic u la r skeletal elem ents th a t fit
to geth er. Jo h n so n (1982, 1987) exam ined b o th pieces o f skeletal elem ents th a t
m echanically fit to g eth er, a n d skeletal elem ents th a t an a to m ica lly seem ed to
rep resen t the sam e in d iv id u al anim al. T h a t is, J o h n so n p erfo rm ed b o th
m echanical refitting an d anatom ical refitting analyses. T he fo rm er is an a lo g o u s
to refitting sto n e a rtifa cts a n d involves the m echanical process o f assessing
w h eth er tw o pieces fit to g e th e r (just like w ith a jigsaw puzzle); these m ay be
frag m en ts o f th e sam e bone, such as a sh aft frag m en t a n d a n end frag m en t, o r it
m ay be tw o a rtic u la r ends th a t w ere artic u la te d in life. A n a to m ica l refitting
m ay involve th e use o f biological o b serv atio n s such as age a n d sex v aria tio n
betw een th e in d iv id u als rep resen ted in the fau n al collection; if only one ad u lt
a n d one s u b a d u lt o rg an ism are rep resen ted in the collection, th en it is a
relatively stra ig h tfo rw a rd m a tte r to assign specim ens th e o n to g en ic age o f
w hich can be d eterm in ed to one o r the o th e r individual. T his kind o f
an a to m ica l refitting becom es m ore difficult as the n u m b e r o f individuals w ithin
p a rtic u la r age, sex, a n d /o r size categories increases. F o r large sam ples w ith
m any ind iv iduals in each age/sex/size categ o ry T o d d (1987b) suggests the
principle o f b ilateral sym m etry m ay allow the an a to m ica l refitting o f paired
bones. Such b ilatera l refits require m uch co m p a ra tiv e d a ta to establish the
range o f v aria tio n betw een p aired bones. Sim ilarly, T o d d (1987b: 179) suggests
such co m p arativ e d a ta m ay also p erm it “ in term em b eral refittin g ” o f skeletal
elem ents because fo r m an y ta x a “ the length o f a fem u r can be pred icted from
the length o f the h u m eru s o f the sam e a n im a l.”
If th e assem blage consists o f only 10 specim ens, each specim en m u st be
checked ag a in st the o th e r nine to d eterm in e if the tw o refit. T h u s, for 10
specim ens there are 45 possible refits o f a n y tw o. T hese 45 possible p airs do no t,
o f course, ac co u n t fo r the fact th a t fo r each tw o possibly refitting specim ens
th ere are m u ltip le w ays they m ig h t fit to g eth er. F o r exam ple, if each specim en
h as tw o ends, th en th ere are fo u r possible w ays fo r an en d o f one specim en to
refit w ith an end o f th e o th e r specim en. T he 45 possible pairs, in the case o f o u r
10 specim ens, becom es 180 possible ways to refit an end o f one specim en to the
end o f a n o th e r specim en. T he m ag n itu d e o f the refitting task is ex acerb ated by
the fact th a t bo ne specim ens n o t only have ends, b u t they also often h ave tw o or
m ore sides o r edges, such as w hen a specim en consists o f only the lateral p o rtio n
o f a long b o n e diaphysis.
T h ere are, th en , several w ays to refit bo n e specim ens, o r to d etect w hich
specim ens co n join . T h e im p o rta n t q uestions, once th e refitting specim ens have
been identified (p e rh ap s regardless o f the m eth o d o f refitting, b u t see below ),
co n cern how to su m m arize th o se d a ta , a n d d eterm in in g w h at the kind and
degree o f refitting m ean s in term s o f ta p h o n o m ic processes. O ne m ight sim ply
c o u n t th e n u m b e r o f bone specim ens, including no n -id en tifiab le ones, in an
assem blage, a n d also co u n t the n u m b e r o f refit pieces to derive a p ro p o rtio n o f
specim ens w ith refitting pieces (if tw o specim ens refit, tally b o th as hav in g a
specim en th a t co n jo in s w ith it). G iven the lab o r-in ten siv e n a tu re o f refitting,
how ever, the u n clea r ta p h o n o m ic m ean in g o f a statistic such as the p ro p o rtio n
o f specim ens th a t refit w ith at least one o th e r specim en m ay be pointless to
calculate.
S tudies o f co n jo in in g bo n e specim ens focus n o t ju s t o n the fact th a t som e
specim ens can be refit w ith one a n o th e r. T hose studies also exam ine the
h o riz o n ta l a n d vertical p rovenience o f the refit specim ens. T h u s it is the relative
spatial lo catio n s o f refitting specim ens th a t are im p o rta n t to refitting studies.
V illa et al.'s (1986) m ap show s th a t the dense clusters o f bones th o u g h t on the
basis o f co n tex tu al a n d stra tig ra p h ic d a ta to represent sy n ch ro n o u s depositio-
nal events h ad refitting pieces som ew hat rem oved from them . J o h n s o n ’s (1982,
1987) “ b u tch erin g artic u la tio n n e ts” in d icate refitting pieces w ere found
dispersed betw een several distin ct clusters o f bones, suggesting the clusters
were c o n te m p o ra ry a n d p ro b a b ly represent subsets o f the sam e b u tch erin g
event. If it is rem em bered th a t disarticulation is the sim ple an a to m ica l
disasso ciatio n o f b o nes a n d involves th eir sp atial se p a ra tio n som e m inim al
distance, a n d th a t scattering is the fu rth e r sp atial disasso ciatio n o f the bones,
th en th e d istance betw een co n jo in in g specim ens can be used as a m easure o f
d isarticu latio n a n d scattering.
T o d d (1987b: 189, 193) presents tw o m eth o d s for m easu rin g the degree o f
scatterin g. F o r p a ire d b ones such as fem ora, m andibles, a n d the like, w hich can
be an a to m ica lly refit, he uses the index o f skeletal disjunction (ISD ). T his index
is in div id u ally calcu lated for each kind o f p aire d skeletal elem ent (for fem ora,
for m an d ib les, etc.) by first m easu rin g the m inim um d istance betw een each
an ato m ically refit pair. T hen, the average m inim um distance is calcu lated for
each kind o f p aired skeletal elem ent. T h e n u m b e r o f co njoined p airs is
m u ltip lied by tw o, a n d ad d ed to the n u m b e r o f u n p a ire d specim ens in the
T ab le 5.9 In d e x o f skeleta l disjunction ( I S D ) and index o ffr a g m e n t

disjunction ( I FD) f o r the H orner I I bison rem ains (from T odd 1987b)
Index o f skeletal disjunction

E lem ent N o f pairs M D T o tal N % Pairs ( M D t % Pairs) 100 S tandardized ISD
H um erus 9 0.99 43 41.86 2.365 48.58

R adius 10 1.10 43 46.51 2.365 48.58
F em ur 8 2.05 38 42.11 4.868 100.00
T ibia 11 0.85 44 50.00 1.700 34.92
Index o f fragm ent disjunction

M inim um N
N o f refit o f elem ents
E lem ent fragm ents w ith refits MD Col. 3 - Col. 2 Col. 4 -r C ol. 5 S tandardized IF D
(1) (2) (3) (4) (5) (6) (7)
H um erus 14 7 0.39 0.50 0.78 29.78
R adius 4 2 0.35 0.50 0.70 26.73
F em ur 31 13 1.10 0.42 2.62 100.00
T ibia 12 4 0.41 0.33 1.24 47.44
co llection to p ro d u c e the to ta l n u m b er o f b ones in a category. T he n u m b er o f

b ones w ith a co n joined m ate is divided by the to ta l n u m b er o f b ones in a
categ o ry to derive the percentage o f b ones w ith m ates. T he m ean m inim um
d istan ce betw een an a to m ica lly refit specim ens (in m eters) is then divided by the
p ercentage o f bones w ith m ates a n d m ultiplied by 100. T he resulting n u m bers
are th en scaled from 1 to 100 to derive the IS D . E q u a tio n [5.2] sum m arizes the
calcu latio n s o f the IS D , a n d the m eth o d fo r calcu latin g T o d d ’s (1987b) index o f
fra g m e n t disjunction (IF D ) is sum m arized in e q u a tio n [5.3]. In these eq u atio n s,
M D is the m ean m in im u m d istance betw een refit specim ens, a n d j is the skeletal
elem ent u n d er co n sid eratio n (e.g., fem ur, o r m an d ib le, o r hum erus). In [5.2]
relatively co m plete, an a to m ica lly refitting skeletal elem ents are considered,
an d in [5.3] m echanically refitting frag m en ts o f skeletal elem ents are
considered.
(M P . -j- % o f j th a t a n atom ically refit) 100__________________________________________ ^^ ^
m axim um value in the nu m e rato r for the assem blage
M D, h- (m inim um num ber o f , w ith refit fragm ents to tal nu m b er o f refitted fragm ents o f ,) ^
m axim um value in the nu m e rato r for the assem blage
A n exam ple will help b ring the indices in to focus. D a ta presented by T o d d

( 1987b: 189. 193) fo r th e H o rn e r II bison rem ains are given in T able 5.9. B oth
the IS D values a n d IF D values are g reatest fo r the fem ur, an d are relatively low
fo r th e o th e r lim b bones. T hese exam ples illu stra te how to calculate ISD an d
IF D values, b u t w h at is the ta p h o n o m ic significance o f such values? F o r one
thing, g re ater index values in d icate g re ater scatterin g (spatial disassociation),
100 -
□ ISD
♦ IFD
80
radius
- 60 tibia
"O
c
TO
Q 40 humerus
CO
20
20 40 60 80 1 oo
Standardized Meat Weight Yield
Figure 5.17. B ivariate scatterp lo t o f index o f skeletal disjunction (ISD ) an d index
o f fragm ent disjunction (IF D ) against standardized m eat w eight yield for H o rn er
II bison.
a n d th u s one m igh t calculate e ith er index to d eterm in e w hich kind(s) o f skeletal

elem ents are greatly scattered a n d w hich are n o t. T he indices m ight also be
calcu lated fo r different tax a to d eterm in e if th ere is tax o n o m ic v a ria tio n in
scatterin g . B ut again, the q u estio n will arise as to w hy som e categories o f
skeletal p o rtio n s are m o re scattered a n d o th e r categories are less scattered.
As T h o m as (1971) observed, h u m an s tend to sca tte r the rem ains o f anim als
they b u tch er. T h u s one m ig h t p re d ic t skeletal p o rtio n s w ith high econom ic
u tility will be m ore scattered th a n skeletal p o rtio n s w ith low econom ic utility
(see C h a p te r 7). I follow ed T o d d ’s (1987b) lead an d p lo tte d the IS D an d IF D
values in T ab le 5.9 ag a in st the average m eat yield a n d ag ain st the average
m arro w yield fo r each k ind o f b o n e as m easu red on three a d u lt bison (one m ale,
tw o fem ales) by E m erso n (1990). T he results suggest fem ora display the
g reatest IS D due to th eir high m eat yield as the o th e r skeletal elem ents are
n eith er as g reatly scattered (low er IS D ) o r as m eaty (low er m eat yield) (F igure
5.17). T h e scatterin g o f these com plete, n o n -fra c tu re d b ones m ay result from
d ism em b erm en t to en h ance tra n s p o rt (lighten the w eight) o f skeletal p o rtio n s
o r to en h an ce rem oval o f m e a t fro m the bones. T he IF D does n o t seem to be
stro n g ly related to the m a rro w yield per categ o ry o f skeletal p o rtio n (F igure
5.18). T his is p erh ap s so because b re ak in g bones fo r m a rro w e x tra ctio n m ay
n o t re su lt in the d isasso ciatio n o f frag m en ts o f the b ro k e n bo n e as once it is
b ro k e n a n d the m a rro w ex tracted there is little econom ic reaso n to m ove the
fragm ents. F o r exam ple, F igure 5.19 in dicates th a t the IS D a n d IF D are no t
greatly different fo r each kin d o f skeletal elem ent in this sam ple. T h a t is, the
100 femur *
□ ISD
' ♦ IFD
8 0
ISD and IFD
6 0
? *
d /■ humerus tibia
4 0
radius 1 a
♦
\
20 H
! ----- 1— I T | — '------- T .....T ' 1

2 0 4 0 6 0 8 0 100
Standardized Marrow Yield
Figure 5.18. B ivariate seatterp lo t o f index o f skeletal disjunction (ISD ) and index
o f fragm ent disjunction (IF D ) against stan d ard ized m arro w yield fo r H o rn e r II
bison.
100 -
femur
8 0
6 0
4 0
humerus
radius
20
ISD
Figure 5.19. B ivariate seatterp lo t o f index o f skeletal disjunction (ISD ) against
index o f fragm ent disjunction (IF D ) for H o rn er II bison.
an ato m ically refitted bones (m easured by the IS D ) w ere dispersed to a degree

sim ilar to th e m echanically refitted frag m en ts (m easured by the IF D ).
W h a t I have so u g h t to d o w ith the ab o v e exam ple is n o t to solve the
tap h o n o m ic p ro b lem o f w hy skeletal p a rts have the d istrib u tio n s they d o at the
H o rn e r II site. T a p h o n o m ic q u estio n s are seldom so easily answ ered. R a th e r, I
have sim ply illu stra te d how tw o m easures o f scatterin g m ight be ap p lied to a
bo n e co llection w hen the an aly st seeks to explain the d istrib u tio n o f skeletal
p a rts in a p a rtic u la r site.
Summary
T ap h o n o m ic processes, by definition, can only begin afte r a n an im al dies. T he

cause o f m o rta lity o ften has significant ta p h o n o m ic consequences. C a ta
stro p h ic o r non-selective m o rta lity o f m u ltip le individuals in a sm all geo
g rap h ic area can resu lt in a n a b ru p t a n d m a jo r in p u t to the fossil record.
N o rm a l a ttritio n a l m o rta lity provides a relatively c o n sta n t in p u t o f a few
carcasses each year, generally scattered over a large sp atial area. S tudy o f the
d em o g ra p h y o f th e d ead p o p u la tio n o f organism s rep resen ted by a fossil
collection helps the an aly st to ascertain the m ode, a n d p erh ap s the cause, o f
m o rtality . W hen analysis o f the d em o g ra p h y o f m o rtality is d o n e in co n ju n c
tion w ith d e te rm in a tio n o f the season o f m o rtality , a d d itio n al insights to causes
o f m o rta lity m ay be g ran ted .
Seldom are soft tissues o f a n im al carcasses preserved. W hen they are, such
p re serv atio n o ften involves w h a t m ight be th o u g h t o f as m u m m ification p ro
cesses. M o re typically, soft tissues are rem oved a n d the skeleton progressively
looses its a n a to m ica l integrity; th a t is, the b ones becom e d isarticu lated an d
ev entu ally spatially disassociated. T here are m any w ays for an anim al carcass
to be skeletonized, a n d these usually involve th e c o n su m p tio n o f soft tissues by
v ario u s m icroscopic a n d m acro sco p ic organism s. T he o rd e r o f jo in t d isa rtic u
latio n in a skeleton seem s to be a fu n c tio n o f the stru c tu re o f a jo in t a n d its
asso ciated soft tissue. A n alysts interested in sk eleto n izatio n a n d d isarticu latio n
typically calcu late the p ro p o rtio n o f artic u la te d jo in ts represented in a collec
tion o f b o nes, a n d m ay p erfo rm vario u s kinds o f analyses o f co n jo in in g bones
an d b o n e frag m en ts to estim ate the degree o f d isarticu latio n a n d scattering.
S om etim es com p lete carcasses a n d som etim es only p a rts o f carcasses are
d ispersed fro m the site o f an im al d e a th a n d original carcass dep o sitio n . These
carcasses a n d carcass p a rts m ay eventually be accu m u lated in a p a rtic u la r
g eo g raph ic place by any o f a m yriad o f bo n e ac cu m u latin g processes an d
agents. It is the dispersal an d ac cu m u latio n o f faunal rem ains to w hich we tu rn
in th e next ch ap ter.
6
A C C U M U L A T I O N A N D D I S P E R S A L OF
VERTEBRATE REMAINS
W hat are all these bones doing here?

(P. Shipm an 1979:42)
Introduction
P aleo n to lo g ists a n d zo o arch aeo lo g ists tend to be o p tim al fo rag ers w hen
ch o o sin g a place to collect fau n al rem ains. F o r the fo rm er, it saves tim e an d
m oney. F o r th e latter, it n o t only saves tim e a n d m oney, b u t the places chosen
are usually selected by an arch ae o lo g ist because they c o n ta in a dense co n c en t
ra tio n o f artifacts; if b ones are sp atially associated w ith the artifacts, then they
to o are usually collected (because they are th ere a n d ) regardless o f their
frequency p er u n it volum e o f sedim ent. T ypically, few arch aeo lo g ical sites are
excavated o r collected sim ply because they c o n ta in an im al rem ains. D e te rm in
ing how th e an im al rem ains cam e to be in the lo catio n s fro m w hich they are
collected, regardless o f th eir g eographical a n d geological p o sitio n w hen
collected, is one o f the m ost fu n d a m e n ta l aspects o f ta p h o n o m ic research.
W h y are b on es densely c o n c e n tra te d in a p a rtic u la r lo catio n , b u t n o t in
su rro u n d in g areas? W hy are som e kinds o f bones p resen t o r a b u n d a n t an d
o th e r k in d s ab se n t o r rare? W hy are som e skeletons com pletely artic u la te d ,
som e p artia lly artic u la te d , a n d o th ers to tally disarticu lated ? W hy are som e
bones spatially close an d o th ers spatially d ista n t from one a n o th er? W hy are
som e b o n es o rien ted one w ay a n d o th ers o rien ted a n o th e r way? W hy d o som e
assem blages h ave lots o f ca rn iv o re rem ains a n d o th ers have few relative to the
frequency o f h erb iv o re rem ains? T his sam pling o f q u estio n s im plies a n u m b er
o f variab les th a t m ig h t be m easu red in an analysis o f a c cu m u latio n . Som e o f
these are discussed a t length in o th e r ch ap ters. In this c h a p te r the variables th a t
seem to be m ost directly related to d eterm in in g how bones have been
ac cu m u lated a n d /o r dispersed are considered. W e begin w ith som e general
co n cep ts before delving in to details o f v ariab le m easu rem en t a n d analysis.
Dispersal, scattering, and accumulation

C o n cen tra tio n o f rem ains from m ultiple individuals occurs either by active tra n s
p o rt o f prey rem ains to a focal area o r by passive accum ulation at a site w here
pred atio n occurs repeatedly.
(C. Badgley 1986a:336)
A nim al carcasses begin th eir ta p h o n o m ic h isto ry as artic u la te d skeletons.

S k eleto n izatio n leads to d isa rtic u la tio n a n d scatterin g o f individual segm ents
161
o f carcasses a n d skeletons. S om etim es th o se segm ents are sets o f artic u la te d

skeletal elem ents held to g e th e r by connective tissue (such as a com plete lim b)
a n d o th e r tim es they are in d iv id u al, discrete skeletal elem ents o r frag m en ts
thereof. H ill (1979b:268-269) defines scattering as “ the increase in d ispersion o f
the p a rts o f a sk eleto n ,” a n d im plies th a t skeletal p a rts becom e m o re dispersed
as they becom e m ore spatially d ista n t fro m the skeleton th a t w as th eir origin;
“ sca tte rin g can be seen to h ave occu rred locally a ro u n d the p o in t o f d e a th [and]
leads to th e d isp ersio n o f rem ains over a co n siderable a re a .” D ispersal is
sy n o n y m o u s w ith scattering. T he degree o f dispersal o f skeletal p a rts o f
in d iv id u al carcasses d epends in p a rt on the length o f tim e betw een skeleto n iza
tio n a n d burial; b o nes o f carcasses th a t are b u ried p rio r to sk eleto n izatio n tend
to stay a rtic u la te d a n d n o t be dispersed.
T he degree o f scatterin g o f skeletal p a rts o f a carcass also d epends on the
ta p h o n o m ic processes an d agents o f scatterin g o r bone tra n s p o rt th a t act u p o n
the p a rts. Such processes a n d agents include scavengers, carnivores, hu m an s,
fluvial actio n , gravity, an d tram p lin g . T he first three focus on ex p lo itatio n o f
so ft tissues a n d bones (see C h a p te r 7), the la tte r three tend to act on the physical
a n d m o rp h o m e tric a ttrib u te s o f bones regardless o f th eir soft tissues o r
econom ic a n a to m y (see C h a p te r 7). T he k in d a n d degree o f scatterin g o f bones,
then, is o ften d ep e n d en t on sk eleto n izatio n processes (w hich p a rtic u la r indivi
d u al b ones o r subsets o f artic u la te d bones are initially detach ed from a
carcass), th e d ep o sitio n al e n v iro n m e n t o f the carcass, a n d the processes o f
scattering.
S catterin g has as its startin g p o in t the com plete set o f a rtic u la te d bones
m ak in g up a skeleton. B ehrensm eyer (1983:98) notes, dispersal o f bones by
p re d a to rs a n d scavengers results in “ lim bs a n d som etim es even the h ead being
drag g ed aw ay from th e v erteb ral colum n, w hich generally acts as a kind o f
‘a n c h o r’ fro m w hich o th e r p a rts are re m o v e d .” T hus, scattering, as a ta p h o n o
m ic process, is a t least im plicitly m o n ito re d o r m easu red fro m the original
d ep o sitio n al p o sitio n o f the com plete skeleton. A ccu m u latio n , as a ta p h o n o
mic process, is usually m o n ito red from a different p o sitio n on the landscape
relative to the lo catio n o f the originally d ep o sited co m p lete skeleton.
T w o basic types o f bo n e ac cu m u latin g processes ca n be d istinguished: active
an d passive (B adgley 1986a, 1986b). A ctive accum ulation processes are those
w hich, via tra n s p o rt o r m ovem ent o f skeletal p a rts (w h eth er o r n o t as com plete
carcasses/skeletons) significant distances fro m the lo catio n o f an im al d eath ,
result in relatively dense c o n c e n tra tio n s o f bones a n d teeth in a spatially lim ited
area. Such processes are labeled “ sp atially focused processes” by B ehrens
m eyer (1983:94). A ctive ac cu m u latio n involves forces a n d energy external to
the anim al(s) w hose b ones are accu m u lated . Passive accum ulation p rocesses are
th o se w hich do no t involve tra n s p o rt o f skeletal p a rts significant distances from
the lo catio n o f an im al d eath ; such processes are n o t spatially focused, an d have
been co n sid ered to re p resen t n o rm a l a ttritio n a l m o rta lity a n d d ep o sitio n o f
A ccum ulation and dispersal o f vertebrate rem ains 163
an im al rem ain s close to the place o f d e a th (B ehrensm eyer 1983). Passive

ac cu m u latio n involves forces a n d energy in tern a l to the anim al(s) - its b eh av io r
- w hose bones are accum ulated. A good exam ple o f an active a c cu m u latio n is
an an im al den (including h u m a n h a b ita tio n sites) to w hich ad u lt m eat eaters
b rin g p o rtio n s o f prey carcasses fo r pro v isio n in g them selves a n d /o r th eir
young. Passive accu m u latio n s can involve loci w here m u ltip le individual
an im als have died o r w ere killed, such as fluvially created paleo n to lo g ical
q u a rry sites (e.g., V o orhies 1969), volcanically created c a ta stro p h ic m o rtality
o f g reg ario us an im als (e.g., L ym an 1989b), o r h u m an ly g enerated bison kill
sites (e.g., F riso n 1974); in these instances a n im al rem ains h ave n o t been m oved
far from the lo catio n o f an im al d eath . O ften, b u t n o t alw ays, the form o f
m o rtality in these cases is a ttritio n a l (see C h a p te r 5), an d w hen m o rta lity is
a ttritio n a l, passive a c cu m u latio n across a large section o f landscape results in
the slow b u t co n tin u o u s ad d itio n o f an im al carcasses to the (fu tu re) fossil
record (e.g.. B ehrensm eyer 1982,1983; B ehrensm eyer an d B oaz 1980; B ehrens
m eyer et al. 1979).
U n lik e d isp ersal, ac cu m u latio n need n o t begin w ith an entire carcass or
skeleton. T h a t is so because w hereas m easures o f dispersal typically m o n ito r
the spatial disassociation o f p a rts o f individual skeletons, m easures o f ac cu m u
latio n typically m o n ito r the spatial concentration o f bones regardless o f
w h eth er th o se b o nes derive fro m the sam e in d iv id u al carcass o r m ultiple,
ind ep en d en t carcasses, a lth o u g h d e te rm in a tio n o f the n u m b er o f carcasses th a t
c o n trib u te d bones (e.g., M N I) m ay be im p o rta n t. Basically, w hen tap h o -
nom ists co n sid er ac cu m u latio n , they are seeking to identify the tap h o n o m ic
agents a n d processes th a t collected a n d d ep o sited the fa u n al rem ains. D ispersal
involves m o v em en t o f b ones aw ay from the lo catio n o f an im al d e a th an d
carcass d ep o sitio n , w hereas a c cu m u latio n involves m ovem ent a n d d ep o sitio n
o f b ones in a lo catio n aw ay fro m the lo catio n o f an im al d e a th a n d carcass
dep o sitio n . B oth involve m o v em en t o f an im al p arts, b u t for the fo rm er the
m o n ito rin g perspective o f the an aly st is lo cated at the site o f an im al d e a th an d
o riginal carcass d ep o sitio n w hereas fo r the la tte r the an a ly st’s m o n ito rin g
perspective o rig in ates at the site o f bone d ep o sitio n (typically spatially
rem oved from the site o f carcass deposition).
B ehrensm eyer (1987:430) outlines a typology o f the “ m ajo r types o f bone
occurence, for an im als larger th a n 15 kg, acco rd in g to causes [of m o rta lity an d
accu m u lation] a n d how these relate to biological a n d physical agencies,
tra n s p o rt a n d tim e” (F ig u re 6.1). T h a t typological system concerns w h a t can be
co n sid ered interpretive types, as the "ty p e s o f o ccurence” include such c a te g o r
ies as “ cach es” a n d “ p re d a tio n a re n a s.” B ehrensm eyer (1987:431) outlines
“ general ta p h o n o m ic c h a rac te ristic s” o f each type; th a t is, she p rovides a list o f
a ttrib u te s the an a ly st can stu d y to d eterm in e if a p a rtic u la r bone assem blage
rep resen ts a “ d e n ” o r a “ tr a p .” T he typological system is instructive because it
identifies fo u r v ariables th a t p lay a significant role in the ac cu m u latio n h isto ry
T Y P E OF OCCURRENCE
/
M A SS DEATHS
IN D IV ID U A L DEATHS
I N D IV ID U A L K IL L S U ntra nsport ed
PREDATIO N ARENAS
TRAPS T IM E
SPECIF IC
EATING AREAS
CACHES AND DENS Biological Agencies
CAVES Transp orted
HYDRAU LIC CONCENTRATIONS ■ Ph ysical Agencies
Favora ble B u r i a l Conditions TIM E

’a v e r a g e d
F igure 6.1. T ypes o f bone occurrence based on m o rtality type (individual, m ass),
bone accum ulation agencies, tran sp o rt, an d d u ra tio n o f accum ulation (after
B ehrensm eyer 1987:430, Figure 1, courtesy o f the a u th o r an d Plenum Press).
o f a fossil assem blage. F irst, th e n u m b e r o f individuals th a t die (m ass o r

m u ltip le versus single) establishes th e frequency o f rem ains th a t m ay eventually
be in c o rp o ra te d in to th e fossil record. Second, b o n e assem blages m ay be
u n tra n s p o rte d (passive ac cu m u latio n ) o r tra n sp o rte d (active accu m u latio n ),
an d th ird , ag ents o f tra n s p o rt m ay be biological (e.g., carnivores) o r physical
(e.g., fluvial action). F inally, assem blages m ay re p resen t “ tim e specific”
o ccurrences con sistin g o f fa u n al rem ains ac cu m u lated over a relatively sh o rt
tim e p erio d , o r “ tim e a v e rag e d ” occurrences th a t consist o f fa u n al rem ains
a c cu m u lated over a long tim e perio d .
E ach o f the fo u r variables - m o rtality , passive o r active ac cu m u latio n ,
a c cu m u latio n agent, a n d ac cu m u latio n d u ra tio n - in B ehrensm eyer’s (1987)
schem e h as tw o v ariab le states (T able 6.1). A rran g in g these fo u r dim ensions o f
v ariab ility so th a t they intersect one a n o th e r to fo rm all possible co m b in atio n s
o f th e fo u r variables a n d eight v ariab le states results in 16 possible classes o f
b o n e accu m u latio n s (T able 6.2, F ig u re 6.2). T hese classes are n o t interpretive,
b u t ra th e r sim ply describe all possible co m b in atio n s o f the variable states
identified by B ehrensm eyer (1987), so m eth in g her typ o lo g y (F ig u re 6.1) does
n o t do. F u rth e r, the classes do n o t differentially w eight the dim ensions o f
v ariab ility , as B ehrensm eyer’s typ o lo g y does, an d th u s each resulting class is
defined o n th e basis o f the sam e suite o f dim ensions. B ehrensm eyer’s (F igure
6.1) “ m ass d e a th s ” type is different th a n h er “ ea tin g a re a s” type due to
differential w eighting o f variables; the fo rm e r is d istinguished on the basis o f
T ab le 6.1 D im ensions o f variability in the process o f bone accum ulation
V A R IA B L E
V ariable State
com m ents
I. M O R T A L IT Y
A. Single Individual
1. a ttritio n al m o rtality (see C h ap ter 5)
B. M ultiple Individuals
1. m ass o r c atastro p h ic m o rtality (see C h ap te r 5)
II. A C C U M U L A T IO N A C T IO N
A. U n tran sp o rte d
1. passive accum ulation
B. T ran sp o rted
1. active accum ulation
III. A C C U M U L A T IO N T Y PE
A. Biological
B. Physical
IV. A C C U M U L A T IO N D U R A T IO N
A. Tim e Specific
1. short d u ratio n , fine grained
2. one m o rtality event (either one o r m ultiple individuals)
3. one accum ulation action
4. one accum ulation type
B. Tim e A veraged
1. long d u ratio n , coarse grained
2. m ore th an one m o rtality event (m ultiple instances o f either o r b o th one or
m ultiple individuals)
3. m ore th a n one accum ulation event
4. p erh ap s m ore th an one accum ulation action
5. perh ap s m ore th an one accum ulation type
the n u m b e r o f d ea d in div id u als w hereas the la tte r is d istinguished o n th e basis

o f one possible o u tco m e for one o r m o re d ead individuals (they w ere eaten).
T he classification system described by T ab le 6.2 a n d F igure 6.2 can subsum e
all o f the in terp retiv e types described by B ehrensm eyer (1987), an d can
ac co m m o d ate o th e r in terp retiv e types as well (T able 6.3). T he classification
system does n o t, how ever, result in the sam e se p a ra tio n o f som e o f the types as
B ehrensm eyer’s typology. P a rt o f the re aso n fo r th a t is the tax o n o m ic stru ctu re
o f th e ty p o lo g y in F ig ure 6.1 a n d the differential w eighting o f the variable states
in th a t typology. All v ariab le states are equally w eighted in the classification in
T ab le 6.2 a n d F ig u re 6.2, th u s all v ariab le states m ake an equal c o n trib u tio n to
defining th e classes o f bo ne ac cu m u latio n .
M y in ten t in p resen tin g the typology in F ig u re 6.1 a n d the classification in
F ig ure 6.2 is to highlight som e o f the im p o rta n t variables in b o th schem es. F o r
exam ple, w h eth er o r n o t a b o n e occurrence is the result o f passive (u n tra n s-
T ab le 6.2 Classes o f variation in bone accum ulation (fro m Table 6.1)
M ortality A ccum ulation A ccum ulation A ccum ulation Cell in

type type action d u ratio n Figure 6.2
single biological passive short 1

single biological active short 2
m ultiple biological passive short 3
m ultiple biological active short 4
single physical passive short 5
single physical active short 6a
m ultiple physical passive short 7
m ultiple physical active short 8
single biological passive long 9
single biological active long 10
m ultiple biological passive long 11
m ultiple biological active long 12
single physical passive long 13“
single physical active long 14
m ultiple physical passive long 15
m ultiple physical active long 16
Note:
a Cell n o t visible in Figure 6.2.
p o rted ) o r active (tra n sp o rte d ) a c cu m u latio n is significant because in the case

o f th e fo rm er th ere often, b u t n o t alw ays, ten d to be few er fossils p er u n it area
th a n in cases o f the la tte r (see below ). But even passive ac cu m u latio n , if it is
spatially focused a n d o f sufficient tem p o ra l d u ra tio n , can result in ra th e r dense
c o n c en tratio n s o f fossils. F u rth e r, as w as em phasized in C h a p te r 5, the
d u ra tio n o f ac cu m u latio n can resu lt in a set o f sp atially ju x ta p o se d b u t
tem p o rally sep a rate in d ividual d ea th s w hich m ay have the a p p e ara n ce, in som e
respects, o f a sh o rt-d u ra tio n m ass d e a th event. B iological a n d physical
a c cu m u latio n agen ts can (a) exploit individual o r m ass d ea th s, an d (b) op erate
over sh o rt o r long tim e periods. It is those fossil assem blages th a t have form ed
due to m u ltiple ac cu m u latio n events th a t are term ed “ tim e a v e rag e d ” by
B ehrensm eyer (e.g., 1982).
T a p h o n o m ists o ften begin th eir study o f a fossil assem blage w ith a q u estio n
like “ W h y are these b o n es hereT ’ T hey seek to identify th e processes w hich
ac cu m u lated the fossils in o rd e r to be able to m ake inferences a b o u t the fossil
assem blage a n d /o r th e fa u n a it represents, inferences th a t are n o t based on an
assem blage th a t is som ehow biased, w ith reg ard s to th o se inferences, by the
processes w hich resu lted in the ac cu m u latio n . K n o w in g the agents and
processes o f bon e ac cu m u latio n , one can allow , say, fo r the fact th a t a
h y d rau lic co n c e n tra tio n will n o t c o n ta in m an y bones w ith low bu lk densities
because such b ones tend to float m ore readily th a n bones w ith high bulk
T ab le 6.3 A lig n m ent o f types o f bone occurrence ( Figure 6.1) with bone
accum ulation classes (T a b le 6.2, Figure 6.2)
Class Type(s) o f bone occurrence (references w ith exam ples o f type)
1 individual deaths, individual kill sites, tra p s (B ehrensm eyer et al. 1979; B urgett 1990)
2 eating areas, caches, dens, caves (B inford 1981b; B rain 1981; Blum enschine 1986b)
3 m ass deaths, p red atio n arenas, trap s (F riso n 1974; H aynes 1991; Olsen 1989b)
4 eating areas, caches, dens, caves (B inford 1981b; B rain 1981; L am 1992)
5 individual deaths, trap s (L ym an 1989b; O liver 1989)
6 hydraulic co ncen tration s (B oaz 1982; Stew art 1989)
m ass deaths (B utler 1987; G rah a m and O liver 1986; L ym an 1989b)
8 hydraulic co ncen tration s (Boaz 1982; S tew art 1989)
P a ss iv e J Active JP a s s i v e j Active
Accu m u lation Action
S h o r t D u r a t io n JLong D uration
Accu m u lation D u ratio n
F igure 6.2 Classes o f b one occurrence defined by dim ensions o f variability in

accum u lation agent (physical, biological), m ortality (single individual, m ultiple
individuals), accu m u latio n action (passive, active), and d u ra tio n o f accum ulation
(short, long) (from T able 6.2). E ach n um bered block is a distinct class; classes 6
and 13 are n o t visible.
d ensities a n d th u s the fo rm er are w innow ed o u t w hereas the latter, w hile they

m ight be m oved by fluvial actio n , tend to rem ain as a lag dep o sit (e.g.,
B ehrensm eyer 1975b; K o rth 1979; V oorhies 1969). It is to the techniques used
for identifying b o th dispersal (w hich is w here ac cu m u latio n , in a w ay, begins)
an d ac cu m u latio n th a t we now tu rn .
Analyzing dispersal
T he prim ary assum ption for the archaeologist to evaluate is th a t the dietary
practices o f m an tend to destroy and disperse the bones o f his prey-species.
(D. H. T hom as 1971:367)
D isp ersal an d scatterin g o f skeletal p a rts has no t been as extensively o r

intensively stu died as the a c cu m u latio n o f skeletal p arts. D ispersal and
scatterin g begin w ith d isa rtic u la tio n (see C h a p te r 5) an d end at som e (spatial
an d tem p o ral) p o in t o f bo ne a c cu m u latio n a n d dep o sitio n . T he tra n sitio n from
d isarticu latio n to a c cu m u latio n is n o t clear, so distin g u ish in g a stage o f
dispersal betw een the tw o is n o t easy. H ow ever, as n o ted earlier, d ispersal, by
d efinition, can be (spatially) m easu red fro m the original d ep o sitio n al lo catio n
o f an in dividu al carcass. T h u s, the refitting techniques a n d analyses o f
d isarticu latio n described in C h a p te r 5 are b o th ap plicable to studying d isp e r
sal. T he indices o f skeletal d isju n ctio n (IS D ) a n d frag m en t d isju n ctio n (IF D )
(C h a p te r 5) can be used to m easure the degree o f dispersal o f the p a rts o f
ind iv id u al sk eleto ns if o th e r evidence suggests the lo catio n o f an im al d e a th is
included in o r n ear the sam pled space. L arge IS D o r IF D values d en o te
significant d ispersal, b u t d o n o t identify the agent(s) o r process(es) responsible
for th a t dispersal. If th e m o n ito rin g perspective o f the an a ly st m u st originate
from the lo catio n o f an im al d ea th a n d carcass d ep o sitio n in o rd e r to m easure
dispersal, how is th a t locus identified fro m the zo o a rch aeo lo g ical record? Such
an identificatio n w ould seem to be the first step in the analysis o f d ispersal, an d
m ay involve tech niq u es fo r assessing how b ones w ere accum ulated.
N o n -z o o arc h aeo lo g ica l as well as zo o a rch aeo lo g ical criteria in d icatin g th at
a site is a kill site seem to be the ones o ften used to identify a lo catio n as the place
o f an im al d e a th a n d original carcass d eposition. S im ilarly, n o n -fau n al
arch aeo lo g ical crite ria are o ften used in c o n ju n ctio n w ith zo o a rch aeo lo g ical
criteria to infer a site is a h a b ita tio n , co n su m p tio n , o r o c c u p atio n al site, th a t is,
so m eth in g o th e r th a n a kill site. In the case o f th e fo rm er, archaeological
evidence o f a tra p o r ju m p is often used, a lth o u g h o th e r lines o f evidence should
also play an im p o rta n t role. O lsen (1989b), fo r exam ple, show s th a t the
tra d itio n a l in te rp re ta tio n o f the U p p e r P aleolithic site o f S o lu tre in F ran c e as a
h o rse ju m p is w ro n g fo r a n u m b e r o f reasons. Z o o arch ae o lo g ic al a n d ta p h o n o
m ic evidence suggests the wild horse (E quus fe ru s) rem ains recovered from this
site rep resen t a c a ta stro p h ic m o rta lity event (F ig u re 6.3), b u t the to p o g ra p h ic
setting o f the rem ains, the inferred m ig ratio n al p a tte rn s o f the p reh isto ric
O CM -S’ CD 00
Figure 6.3. E quid m o rtality profiles for M agdalenian an d G rav ettian levels at
Solutre. F rance (after Levine 1983). Age classes are in years and are given as
class m id-points.
horses, a n d the b eh a v io r o f m o d ern wild horses all p o in t to p reh isto ric h u n ters
in tercep tin g an d d irecting m ig ratin g herd s in to a cul-de-sac at the b ase o f a cliff
w here h orses w ere co rralled a n d killed.
F a u n a l d a ta suggestive o f m ass kill sites include ca ta stro p h ic m o rtality
profiles, thick, dense c o n c e n tra tio n s o f b ones o f m ultiple individuals w ith som e
bones still artic u la te d , an d , the com plete o r n ear-co m p lete re p re se n ta tio n o f
an im al skeletons. L evine's (1983) o n to g en ic d a ta fo r the horse rem ains from
tw o cu ltu ra l h o rizo n s at S olutre are suggestive o f c a ta stro p h ic m o rtality
(F ig u re 6.3; see C h a p te r 5 fo r discussion o f the analysis o f o n to g en ic d a ta ), w ith
the ex cep tio n th a t y o u n g horses are ra re (b u t see also H u lb e rt 1982). T he la tte r
m ay be a ttrib u ta b le to p o o r p re serv atio n o f the rem ains o f y o u n g h orses due to
th eir low stru c tu ra l density (see C h a p te rs 7 a n d 9). O lsen (1989b) re p o rts 4,483
bo ne a n d to o th specim ens o f h o rse w ere recovered fro m 36.3 m 2 o f excavation,
fo r a n average o f 123.5 specim ens p er m 2. She refers to the stratig ra p h ic
occurrence o f the h o rse rem ains as the “ dense ‘h o rse m a g m a ’ ” (O lsen
1989b:298). O f the skeletal elem ents O lsen (1989b) describes, all m a jo r p o rtio n s
o f th e sk eleton are p resen t (she does n o t describe v erte b rae o r ribs), a lth o u g h in
v aryin g frequencies (F ig u re 6.4; see C h a p te r 7 fo r techniques o f dealing w ith
such frequencies), a n d m any individual anim als are represented. T h a t people
h ad so m eth in g to d o w ith this ac cu m u latio n o f b ones is in d icated by the fact
th a t o f the 2,484 bo n e specim ens rep resen ted in F ig u re 6.4, seven have
b u tch erin g m ark s o n th em (see C h a p te r 8 fo r d iscussion o f b u tch erin g m arks).
astragalus
2nd phalanx
mandible
carpals
navicular
1st phalanx
radius
calcaneum
metacarpal
metatarsal
patella
humerus
tibia
3rd phalanx
femur
maxilla
scapula
innominate
ulna
0 10 20 30 40
MNI per Skeletal Element
Figure 6.4. Frequencies o f equid skeletal p a rts in the A u rignacian level o f
Solutre, F ran ce (after Olsen 1989b).
W h e at (1979, see also W h e at 1972) suggests the set o f crite ria su m m arized in
T able 6.4 allow s th e d istin c tio n o f kill sites, processing sites, a n d h a b ita tio n or
c o n su m p tio n sites. T he values given th ere are n o t ab so lu tes b u t ra th e r serve to
illu stra te th e relative frequencies o f the listed v ariables fo r a given site type. F o r
exam ple, as one read s d o w n the list from kill to processing to c o n su m p tio n
sites, the p ro p o rtio n o f bones still artic u la te d w ith o th e r bones decreases, the
p ro p o rtio n o f bo nes n o t a rtic u la te d w ith o th e r bones (th a t are d isarticu lated )
increases, a n d the average n u m b e r o f bones m ak in g up a set o f artic u la te d
bones decreases. T his is only logical as it m easures d isa rtic u la tio n a n d dispersal
o f b o nes, tw o m a jo r effects o f b u tch erin g (see C h a p te r 8). B u t we m u st n o t lose
sight o f the fact th a t o th e r processes also result in the d isarticu latio n and
dispersal o f bones; th e a ttrib u te s in T able 6.4 do, nonetheless, pro v id e one clue
to w h eth er o r n o t the site represents the original lo catio n o f an im al d e a th an d
carcass d ep o sitio n .
E vidence o f o cc u p atio n al features such as houses in a site, a b u n d a n t
evidence o f sto n e to o l m a n u fa c tu rin g , a n d /o r n u m ero u s p la n t processing tools
m ay p ro m p t the in v estig ato r to explore the possibility th a t the site was a
h a b ita tio n , w here such activities are expected, ra th e r th a n a kill site, w here such
activities p ro b a b ly rarely to o k place. As T ab le 6.4 indicates, it seem s th a t few
b ones will be a rtic u la te d , a n d the n u m b e r o f bones in a set o f a rtic u la te d bones
will be sm all. B ut th e p ro p o rtio n s o f a rtic u la te d an d d isarticu lated b ones tell us
little a b o u t dispersal; ra th e r, they m easu re the degrees o f artic u la tio n and
d isarticu latio n a n d only indirectly m o n ito r dispersal. T o m easu re dispersal, as
T ab le 6.4 C riteria proposed by W heat (19 7 9 ) f o r distinguishing kill sites,

processing sites, an d consum ption sites. N o te that the frequencies p resented are
fo r several specific sites, and are m eant as guides to relative frequencies rather
than as absolutes
% o f bones % o f bones A verage n u m b er o f M ake-up o f

Site type articu lated n o t articu lated bones articu lated tool assem blage
Kill 49 51 10 m any stone an d few

bone b u tchering tools
Processing 28 72 6 few stone and m any
bone b u tchering tools
C am p o r consum p tio n 2.5 97.5 3 9
defined ab o v e, we need to k n o w the o rig in al d e a th lo catio n o f the anim als

rep resen ted by th e b ones in a h a b ita tio n site, b u t because th e b ones in a
h a b ita tio n site w ere ac cu m u lated in th a t site, we p ro b a b ly could n o t determ ine
the lo catio n w here the rep resen ted an im als w ere killed (see O ’C onnell et al.
1992 fo r an in terestin g discussion o f kill sites created by h um ans). W e w ould,
p erh ap s, be able, via an aly tical refitting, to d eterm in e how fa r the bones were
dispersed from one a n o th e r w ithin the c o n su m p tio n site once they h ad been
accu m u lated th ere, a n d th u s we co u ld m easu re the relative degree o f dispersal
o f b o nes w ithin a site. B ut we w ould n o t k n o w h ow o r how fa r th e b ones were
dispersed from the kill site. F o r exam ple, L ym an (1989b) used an a to m ica l
refitting to identify th e dispersal o f bones fro m k n o w n lo catio n s o f carcass
d ep o sitio n . Som e b ones h ad m o v ed dow nslope, a p p a re n tly as a resu lt o f fluvial
tra n s p o rt. B ones fo u n d u p slo p e o f the original carcass lo catio n h a d been
gnaw ed by carn iv o res, leading to th e inference th a t these specim ens h a d been
dispersed by carnivo res. B ut L ym an w as dealing w ith a lim ited n u m b er o f
carcasses rep resen tin g a kn o w n c a ta stro p h ic m o rta lity event, a n d he sam pled a
h o rizo n ta lly large a rea (ca. 100 m x 100 m ) in w hich m o st b ones w ere fo u n d on
the surface. W ith o u t these ad d itio n a l d a ta , a n d w hen bones have been
preh isto rically ac cu m u lated from kill sites a kilo m eter o r m o re from the
arch aeo log ically sam p led site o f b o n e d ep o sitio n , such inferences a n d m easures
o f d ispersal m ay n o t be possible.
T h e stu d y o f d isp ersal o f skeletal p a rts fro m th e o rig in al site o f an im al d ea th
a n d carcass d ep o sitio n m ay be to o dem an d in g ; th a t is, we m ay be u n ab le
analy tically to d eterm in e the o rig in al lo catio n o f carcass d ep o sitio n . D ispersal
o f b o n es can be, a n d in fact o ften is, stu d ied w ith o u t m eeting this d em an d . In
the follow ing I review several w ays dispersal has been a n d can be studied
w ith o u t clear kn ow ledge o f the lo catio n o f o riginal carcass d eposition.
Fluvial dispersal
T h u s far d ispersal by h u m an s h as been considered in the sense th a t the fossil
co n tex ts h ave been described as arch aeo lo g ical. W h a t, then, a b o u t n a tu ra l
T ab le 6.5 M a m m alian skeleta l elem ents grouped by their susceptibility to

fluvial transport (fro m Voorhies 1969:69). C olum ns labeled " I & I I ” a n d " II
& I I I ” are transitional betw een m ain groups
G ro u p I 1 & II G ro u p II II & III G ro u p III

im m ediately m oved, gradually rem oved, lag deposit
m ay float o r bounce stay in con tact
along b o tto m w ith b o tto m
rib scapula fem ur ram us o f skull

vertebra phalange tibia m andible m andible
sacrum ulna hum erus
sternum m ctapodial
pelvis
radius
agents o f bone dispersal? O ne o f the m o st-stu d ied m echanical dispersal

processes is fluvial tra n s p o rt (e.g.. B ehrensm eyer 1975b. 1982; B oaz 1982; Boaz
an d B ehrensm eyer 1976; D o d so n 1973; F ro stick an d R eid 1983; H a n so n 1980;
K o rth 1979; V oorh ies 1969; W olff 1973). V oorhies (1969) pioneered these
studies w hen he p erfo rm ed experim ents w ith d isarticu lated bones o f dom estic
sheep (O vis aries) a n d co y o te ( Canis latrans) in a flum e. H is results indicate
som e skeletal elem ents are m o re likely to be m oved by fluvial processes th an
o th ers (T able 6.5). B ehrensm eyer (1975b) su b sequently ela b o ra te d on V o o r
hies’ schem e, n o tin g th a t the stru c tu ra l density o f bones as well as th eir size and
shape influences the p ro b a b ility th a t a p a rtic u la r bo n e will be fluvially
tra n sp o rte d . In fact, w h at cam e to be called “ V oorhies G ro u p s ” o f bones (sets
o f b o n e d isplaying v arying p ro b a b ilities o f being m oved by fluvial processes)
are w eakly co rrelated w ith the stru c tu ra l density o f b ones as the la tte r was
m easu red by B ehrensm eyer (rs = 0.495, P = 0.06). F u rth e r, B ehrensm eyer’s
ex p an d ed V o o rh ies G ro u p s are strongly co rrelated w ith th e stru c tu ra l density
o f b o n es (rs = 0.775, P = 0.001) (see C h a p te r 7 fo r discussion o f the stru c tu ra l
density o f bones). W olff(1973) referred to these processes a n d results as sorting
to d en o te the b ones o f a skeleton, if d isarticu lated , w ould be so rted by
h y d ro d y n a m ic processes in to g ro u p s o f readily m oved a n d n o t readily m oved
skeletal elem ents.
B ehrensm eyer (1975b:489) discusses the "d isp e rsal p o te n tia l” o f b ones in a
fluvial m edium , a n d n o tes th a t "since V oorhies G ro u p I [Table 6.5] is the m ost
easily affected by fluvial tra n s p o rt, its presence o r absence in fossil assem blages
can p ro v id e specific in fo rm a tio n on the sed im en tary h isto ry o f bo n e assem
blages.” T he absence o f G ro u p I bones suggests the stu d ied assem blage is a lag
assem blage (G ro u p I b ones h av in g been w innow ed out); th e presence o f G ro u p
I b ones suggests a non-fluvially w innow ed assem blage. F u rth e r, presum ing
th a t the fluvial tra n s p o rt o f the bones began at the site o f an im al d ea th , then
“ th e p ro p o rtio n s o f different V oorhies G ro u p s in fossil assem blages should
t IT)
O
Beyond
lim its of
dispersal
Voorhies
Voorhies
Group 1
Voorhies
Voorhies
Group 1
Voorhies
Voorhies
Groups 1
and II
Voorhies
I—
>-
j—
Group 1 Group 1 Groups 1 Group II
and II
Voorhies Voorhies Voorhies Voorhies

cL J cLlJ Group 1 Groups 1 Group II Groups II
CO CO (and II) and II and III
<E ^
s s
Voorhies Voorhies Voorhies complete
Groups Groups II Group III removal
1, II, III and III (lag)
(undisturbed) (winnowed)
Site where
transport processes INCREASING CURRENT VELOCITY
begin to affect bone3
Figure 6.5. C lassification o f b one dispersal g roups according to cu rren t velocity
and proxim ity to the site w here bones begin tra n sp o rt by fluvial action; see T able
6.5 for skeletal elem ents in each V oorhies G ro u p (after B ehrensm eyer 1975b:491,
Figure 5; courtesy o f the a u th o r an d M useum o f C o m p arativ e Z oology, H arv ard
U niversity).
pro vid e evidence fo r the p roxim ity o f fossils to the original th an a to c o e n o se an d

the h a b ita ts o f th e living an im als” (B ehrensm eyer 1975b:490, 491). This
re la tio n sh ip is m odeled in F igure 6.5.
B ehrensm eyer (1975b) m easu red the influence o f the sh ap e o f b ones on their
p o te n tia l fo r fluvial tra n s p o rt by stu d y in g th eir settling velocities, a them e
picked up on by K o rth (1979) in his stu d y o f the ta p h o n o m y o f m icro verte
b rates (see also D o d so n 1973). B ehrensm eyer (1975b) a n d K o rth (1979) argue
th a t the settling velocity o f a b o n e is related to the p o ten tial th a t the b o n e will be
fluvially tra n sp o rte d . K o r th ’s (1979) ex p erim en tal results o n settling velocities
o f bo n es o f sm all m am m als a n d D o d s o n ’s (1973) experim entally derived
tra n s p o rt g ro u p s o f m ouse bones align fairly well w ith V o o rh ie s’ G ro u p s
(T able 6.6). K o rth (1979) notes th a t sedim entary analysis o f the m atrix
s u rro u n d in g the recovered fossils can pro v id e im p o rta n t c o rro b o ra tin g evi
dence th a t a p a rtic u la r bo n e assem blage w as, o r w as n o t, subjected to fluvial
tra n sp o rt. If, fo r exam ple, th e b o n es a n d th e sedim ent particles h ave eq u iv alen t
T ab le 6.6 K o rth's (19 7 9 ) settling groups aligned with V oorhies’ (1969)

Groups (T a b le 6 .5 ). A ll are m a m m a l bones
A nim al live V oorhies G ro u p s

w eight I I & II II II & III III
Settling group: T axa < 500 g

rib atlas calcaneum m o lar m andible
p halanx lum bar astrag alu s incisor tibia
m etap o d ial fem ur maxilla
radius hum erus skull
scapula
Settling group: T axa > 10 kg

rib atlas skull m andible m andible
scapula
m o lar
fem ur
astrag alu s
settling velocities, th e n it is likely the b o n es w ere fluvially tra n sp o rte d .

B ehrensm eyer (1975b:499) indicates “ th e m o st likely places fo r final b u rial [of
fluvially tra n s p o rte d bones] are in th e actively ag g rad in g p a rts o f a channel
such as p o in t bars a n d sand o r gravel b ars [where the co arse fractio n o f the
sedim en t lo ad is deposited] . . . B ones will m ove alo n g a ch an n el, suffering
progressive ab rasio n , u n til they en c o u n te r such a s itu a tio n .”
B oaz (1982) stu d ied m o d e rn sets o f bones in the M a ra R iver o f T an z an ia.
H e r stu d y o f a set o f w ildebeest (C onnochaetes taurinus) carcasses a n d bones
indicates th a t w hen this m ig ratin g species crosses large rivers, m an y individuals
d ro w n a n d th e carcasses th en flo at d o w n stre am w here they are ac cu m u lated in
areas o f shallow w ater. She no tes th a t “ herd size a n d crow ding m ay be m ore
im p o rta n t th a n flood levels a n d c u rre n t velocities as causes fo r m ass d row nings
d u rin g river crossin gs” (B oaz 1982:173). T he a b u n d a n c e o f carcasses in such
accu m u latio n s resu lts in low levels o f d am ag e by carn iv o res, an d , d ep e n d in g on
th e sedim ent lo ad o f th e fluvial system , the carcasses a n d bones m ay be quickly
b u ried only to be re-exposed a n d a b ra d e d by the sedim ent load. Som e bones
m ay be w innow ed o u t by fluvial ac tio n , b u t the h o riz o n ta l a n d vertical
d istrib u tio n o f b o n es seem s to p ro d u c e c o n c e n tra te d , dense accu m u latio n s o f
fau n al rem ains.
F riso n a n d T o d d (1986) m easu red the d istan ce each o f several individual
skeletal elem ents o f a n In d ian elep h a n t (E lephas m a x im u s) w as tra n sp o rte d by
fluvial ac tio n over m u ltip le experim ents. F ro m th o se m easu rem en ts they
derived a fluvial tra n s p o rt index (F T I) w ith the eq u atio n :
lo g I0[(M TDj) 100] [6.1]
in w hich M T D is th e m ean tra n s p o rt d istan ce o f skeletal e le m e n t, o ver m ultiple

tra n s p o rt episodes. T he values fro m e q u a tio n [6.1] are scaled fro m 1 to 100 (all
T ab le 6.7 Fluvial transport index ( F T I) values and saturated weight index

( S W I ) values f o r various taxa. Indian elephant after Frison and T odd (1 9 8 6 );
other ta x a derived fr o m B ehrensm eyer ( 1975b). B lanks indicate no data
Indian elephant SWI

Skeletal
elem ent SWI FT I D om estic sheep R eedbuck T opi Z ebra
cranium 100.00 100.0 100.0

m andible 69.35 34.56 19.1 23.4
atlas 2.76 41.97 10.4 11.1 21.4 20.6
axis 11.0 10.9 21.2 22.3
cervical 2.08 96.64 8.4 12.1 16.4 19.2
thoracic 1.66 76.43 5.0 5.1 7.1 8.2
lum bar 1.10 76.21 4.7 11.9 10.7 5.6
sacrum 6.97 80.11 11.3 17.8 31.9
caudal 0.37 92.43
rib 1.74 53.98 3.8 5.6 8.7 4.6
scapula 16.50 62.95 14.5 22.6 40.0
hum erus 26.40 57.77 27.7 54.7 81.0 63.5
rad iu s-u ln a 18.70 49.95 21.9 43.5 73.8 50.8
m etacarpal 1.05 68.83 9.4 36.4 40.7 20.3
pelvis 57.20 0.00 42.9 43.9
fem ur 26.40 24.26 31.9 92.4 96.9 100.0
patella 0.98 90.19 1.0 3.5 6.2 3.3
tibia 8.56 72.84 27.7 92.4 100.0 68.9
fibula 1.22 60.19
astragalus 1.08 96.83 2.3 7.7 8.1 8.2
calcaneum 1.66 100.00 3.0 11.6 12.9 10.2
m etatarsal 10.2 41.4 45.2 28.2
first phalanx 1.4 3.7 5.2 5.7
second phalanx 0.5 4.2 2.7 5.4
third phalanx 0.6 1.8 2.4
values divided by the m ax im u m value a n d m u ltip lied by 100) to p ro d u c e the

F T I values (T able 6.7). F riso n a n d T o d d (1986) fo u n d th a t skeletal elem ents
w ith F T I values > 7 5 (sacrum , p atella, astra g alu s, calcan eu m , cervical, th o ra
cic, lu m b ar, sacru m ) are basically the sam e as th o se fo u n d in V oorhies G ro u p I
(T able 6.5); skeletal elem ents w ith F T I values o f 50 to 74 (ribs, scapulae,
h u m eri, tibiae, a n d m etac arp a ls) are sim ilar to V o orhies G ro u p II; a n d skeletal
elem ents w ith low F T I values ( < 5 0 ; atlas, m an d ib le, pelvis, ra d iu s -u ln a ,
fem ur) are nearly th e sam e as th o se in V oorhies G ro u p III. In b iv ariate
sca tte rp lo ts o f the F T I values a g a in st th e frequencies o f C o lu m b ian m a m m o th
(M a m m u th u s colum bi) bones in tw o clusters a t th e C lovis-aged C o lb y site in
W yom ing, F riso n a n d T o d d (1986) fo u n d little clear relatio n betw een the tw o,
a n d co n clu d ed th a t fluvial tra n s p o rt h a d n o t created th e b o n e clusters.
F riso n a n d T o d d (1986:68) also re co rd e d th e w eight o f th e skeletal elem ents
o f th e In d ia n elep h a n t a fte r they w ere sa tu ra te d , a n d n o rm e d the w et w eights to
a scale o f 1 to 100 to p ro d u c e a s a tu ra te d w eight index (SW I). T he SW I is
inversely b u t im perfectly co rrelated w ith the F T I (r= —0.67, P = 0.001),

u n d ersco rin g “ the degree to w hich facto rs o th e r th a n w eight influence [fluvial]
tra n s p o rt p o te n tia l” (F riso n a n d T o d d 1986:68). F riso n a n d T o d d fo u n d little
clear relatio n betw een the tw o clusters o f m a m m o th b o n e at the C olby site and
the SW I, an d co n cluded th a t the w eight o f the bones h ad exerted little influence
on w h eth er o r n o t b o n es ended u p in the clusters. T hey did, how ever, find
significant relatio n s betw een w h at they called the “ dispersed b o n es” n o t in the
tw o bo n e clusters a t the site, an d b o th the F T I a n d the SW I, suggesting to them
th a t fluvial tra n s p o rt h a d m oved these “ dispersed b o n e s.”
W h eth er o r n o t F riso n an d T o d d ’s (1986) conclusions reg ard in g the C olby
site p ro v e to ho ld u p u n d er fu rth e r study, it is clear th a t the F T I a n d SW I
indices th ey developed are o f value in helping us u n d e rsta n d the dispersal o f
bones. S im ilar indices fo r o th e r tax a sh o u ld be developed. F riso n a n d T o d d ’s
(1986:66) F T I an d SW I values fo r In d ian elep h a n t are listed in T ab le 6.7. Also
listed in th a t table are SW I values I derived from wet w eight d a ta described by
B ehrensm eyer (1975b:572) fo r fo u r m a m m alian species (dom estic sheep, Ovis
aries; reed b u ck , Redunca sp.; topi, D am aliscus sp.; an d zebra, Equus sp.). Such
indices can be used in b iv ariate sc a tte rp lo ts a n d /o r w ith c o rrelatio n coefficients
to d eterm in e if b o n e frequencies a n d the indices are co rre la te d (see C h a p te r 7
fo r exam ples o f how such indices are used). Significant c o rrelatio n s w ould
suggest fluvial tra n s p o rt has played a role in d istrib u tin g (dispersing) the bones.
Shape o f fossils
M o st research ers stu d ying fluvial tra n s p o rt suggest th a t the shape o f a bone
exerts a stro n g influence on w h eth er it is tra n sp o rte d . F ro stic k a n d R eid (1983)
describe a sim ple w ay to m easure a n d classify the shape o f bones. T hey suggest
“ the th ree m u tu ally p e rp en d icu lar axes o f the fossils p ro v id e a m easu re o f b o th
p article sh ape a n d sp h e ric ity . . . A xial ra tio s (c/b, b /a) subdivide specim ens into
spheres, ro d s, discs, a n d b lad es” (F ro stick an d R eid 1983:159). T he axes they
refer to are: a, m axim u m dim ension; b, m id dim ension; a n d c, m inim um
dim ension. T hese are p o ten tially referred to as length, w idth, a n d thickness,
respectively, if it is realized th a t length is n o t necessarily related to som e
an a to m ica l o rie n ta tio n o f the m easured bone. P lo ttin g the axial ra tio s (c/b, b/a)
ag ain st one a n o th e r allow s specim ens to be classified acco rd in g to th eir basic
shape, as show n in F ig u re 6.6. N o te th a t as one m oves fro m the u p p er right o f
the g ra p h to th e low er left, sphericity decreases. S phericity is defined by the
e q u a tio n
(b c/a2)033 [6.2]
w here a, b, a n d c are defined as above. F ig u re 6.6 show s fo u r fictional bones, the

d im ensions o f w hich are given, p lo tte d on the grap h . F ro stic k a n d R eid's
(1983:163) ex p erim en tal d a ta indicate th a t bones classified on the g ra p h as
blades a n d discs are less likely to m ove do w n slo p e th a n bones w ith shapes
a p p ro x im a tin g ro d s a n d spheres, o r a t least the fo rm er will m ove m o re slowly
1U —
2
disc
o4 sphere
bonel: e = 1 0 , b = 1 , c = 1
bone 2: a = 1, b = 1, c = 1
0.5 - o3
bone 3: a = 1 0, b = 5 , c = 1
blade rod
bone 4: a = 5, b = 4, c = 1
0.1 - 1<
I I
0.1 0.5 1. 0
c/b
Figure 6.6. C lassification o f bone shape based on axial ratio s (after F ro stick and
Reid 1983). Bone dim ensions can be any unit o f linear m easurem ent (e.g., cm).
See text for discussion.
th a n the latter. D o d so n (1973:18), fo r exam ple, re p o rts th a t verteb rae (sphere

like shape) w ere m ore easily dispersed by fluvial ac tio n th a n m andibles (b lad e
like shape).
F aced w ith an assem blage o f variously com plete bones an d frag m en ts o f
bones, th e ta p h o n o m is t co u ld m easu re the three p erp en d icu lar axes o f each
specim en an d plo t them on the g ra p h in F ig u re 6.6. If the m ajo rity o f the
specim ens classified as rods a n d spheres occupy low er vertical p o sitio n s th a n
th o se classified as discs a n d blades, the an a ly st co u ld th e n suggest th a t fluvial
tra n s p o rt h ad dispersed the bones. T his presum es, o f course, th a t the relative
vertical p o sitio n s o f the b ones are n o t d u e to fau ltin g o r tectonic tilting o f the
stra tig ra p h ic u n it in w hich the fossils are fo u n d , a n d th u s th a t the ta p h o n o m ist
has d etailed proven ien ce d ata.
O rientation o f bones
F lu v ial ac tio n will n o t only preferen tially so rt a n d tra n s p o rt skeletal elem ents,
it will also result in p a tte rn e d o rie n ta tio n s o f long bones (T o o ts 1965a, 1965b;
V o o rh ies 1969). T h ere are tw o basic w ays in w hich o rie n ta tio n d a ta are
typically p resented: rose diagram s, a n d stereographicprojections (F io rillo 1988;
T o o ts 1965a). T he fo rm er in c o rp o ra te s tw o -d im en sio n al o rie n ta tio n d ata
based o n th e az im u th o f th e lo n g axis o f specim ens w ithin a h o riz o n ta l plane
w hereas th e la tte r in c o rp o ra te s b o th the az im u th a n d the vertical p lu n g e o r dip
o f th e long axis relative to a h o riz o n ta l plane. I first describe the fo rm er, a n d
th en tu rn to th e latter.
T o o ts (1965a:220) suggests “ tw o -d im en sio n al th in k in g ac co u n ts for the
ten d en cy to dep ict o rie n ta tio n d a ta grap h ically in a rose d ia g ra m .” H e argues

th a t a rose d iag ra m is “ perfectly co rrect w hen studying tw o -d im en sio n al
p h e n o m e n a ” (T o o ts 1965a:220), a n d I believe he w ould agree th a t a rose
d iag ram is a p p ro p ria te if there is m inim al dip o r plunge in the long axis o f
fossils w ith a definite long axis. P resum ing such is the case, one can m easure
w h at is typically called the “ p referred o rie n ta tio n ” (e.g., K re u tzer 1988) o r the
a z im u th o f a long bo ne w ith a co m p ass if the bo n e is in situ, o r w ith a p ro tra c to r
if m ap s o f bo n e d istrib u tio n s are sufficiently detailed. A long bone actu ally has
tw o o rien ta tio n s, such as 160° a n d 340°, if m easu red from (say) the p roxim al
end a n d from the distal end. T h a t is, ignoring w h eth er the p ro x im al to d istal o r
distal to p ro x im al o rie n ta tio n is m easu red , th en th e possible ran g e o f v a ria tio n
in o rie n ta tio n sp ans only 0° to 180°. If a rose d iag ra m o f o rie n ta tio n d a ta is
p ro d u c ed as a com plete 360°, then, fo r instance, the d a ta betw een 0° a n d 180°
are the ac tu a l m easu rem en ts, a n d the re m a in d er o f the g ra p h (from 181° to
360°) is a m irro r im age o f the form er. T he o rie n ta tio n o f a long bo n e m ay be
m easu red to the n ea rest degree, b u t o rie n ta tio n d a ta are typically g ra p h ed in a
rose d iag ra m by som e > 1° interval, such as by each 10° (e.g., F io rillo 1988;
K re u tzer 1988; S h ip m an 1981b).
“ R ose d iag ram s are useful as in d icato rs o f ra n d o m o r n o n ra n d o m p a tte r n
ing in th e o rie n ta tio n o f long axes o f b ones . . . A sym m etry o f a d iag ra m is
assu m ed to be an in d ic a to r o f n o n ra n d o m p o sitio n in g o f b o n e s” (F riso n and
T o d d 1986:53). T o in te rp re t rose d iag ram s the analyst m u st first assum e th at
th e skeletal elem ents w hose o rie n ta tio n s are p lo tte d w ere “ free to m ove in
resp on se to d irectio n al forces” a n d m ust fu rth e r assum e th a t the bones m oved
across an essentially flat surface th a t did n o t influence th eir final o rie n ta tio n
(F riso n a n d T o d d 1986:54). (R eg a rd in g the form er, D o d so n ’s [1973:17]
ex p erim ents w ith m o use bones suggest to him th a t th e o rig in al o rie n ta tio n o f
b ones p rio r to fluvial actio n “ has som e effect on the susceptibility o f b ones to
m o v em en t,” an o b serv atio n also n o ted by F riso n a n d T o d d [1986].) Sets o f
a rtic u la te d b ones will be orien ted differently th a n their d isarticu lated co n sti
tu en t p a rts, a n d if sufficient soft tissue is still a tta c h e d such sets m ay w eigh so
m uch th a t they c a n n o t be m oved by fluvial action. F riso n a n d T o d d ’s
ex p erim en ts w ith elep h a n t bones, a n d experim ents by o th ers (e.g., H a n so n
1980; V o o rhies 1969), in d icate th a t m icro to p o g ra p h ic features, ch an n el w idth
relative to bo ne length a n d shape, a n d o th e r featu res can result in bones having
o rie n ta tio n s th a t are n o t related to c u rre n t direction. P erh a p s fo r th a t reaso n
S hipm an (1981 b :7 1) reco m m ends th a t only sam ples w ith at least 72 b ones for
w hich o rie n ta tio n can be m easu red be p lo tte d in rose diag ram s. S m aller
sam ples m ay c o n ta in to o m an y bones the o rie n ta tio n s o f w hich are n o t a
fu n c tio n o f fluvial ac tio n , b u t ra th e r are the result o f the influence o f o th er
facto rs (e.g., tram plin g ); larger sam ples sh o u ld m ore clearly display o rie n ta tio n
p a tte rn s. A lso, if g eo m o rp h ic featu res such as a channel o r slope are a p p a re n t
in the stra tig ra p h ic reco rd , b o n e o rie n ta tio n should be related to th a t feature
a n d n o t ju s t the co m pass az im u th (e.g., F ro stic k a n d R eid 1983).
Figure 6.7. A m irror-im age rose d iag ram show ing azim uths o f long axis o f long
bone (after K reutzer 1988:226, Figure 4; courtesy o f the a u th o r an d A cadem ic
Press). B lack wedges are non-significant values; w hite wedges are significant
values (see T able 6.8).
T h e m irro r-im a g e rose d ia g ra m in F ig u re 6.7 p resen ts o rie n ta tio n d a ta for

1,084 b o nes com piled by K re u tzer (1988) fo r tw o late P leistocene/early
H o lo cen e clusters o f b iso n b o n es a t the L u b b o c k L ake site in T exas. K re u tzer
(1988) c o m p ared th e observed a n d expected frequencies o f specim ens p er 10°
o rie n ta tio n class by calcu latin g ad ju sted residuals fo r each class, a n d finding
the p ro b a b ility o f observ ing such ad ju sted residuals if o rie n ta tio n w as ra n d o m
(E v eritt 1977 describes the statistical technique). T he results (T able 6.8)
in d icate th a t th e black w edges (10° classes) in F ig u re 6.7 are n o t significantly
different fro m expected values w hereas the w hite w edges are significantly
different from expected values. K re u tzer (1988) concludes th a t fo u r classes
have m o re bon es th a n expected given ra n d o m chance, a n d seven classes have
few er b ones th a n expected given ra n d o m chance. In light o f experim ental d a ta
(V oorhies 1969) in d icatin g th a t long b ones w ith one heavy end will have th eir
long axes o rien ted p arallel to th e c u rre n t d irectio n a n d bones w ith ends o f
ap p ro x im a te ly eq u al w eight will have th eir long axes o rien ted p erp en d icu lar to
cu rren t directio n , K re u tzer (1988:227) concludes th a t the specim ens from
T ab le 6.8 O bserved and e xp ected frequencies o f 1084 bone specim ens per 10°
orientation class at L ubb o ck L a ke, T exas (fro m K reutzer 1988)
D egree Expected O bserved A djusted

class frequency frequency residual P robability C onclusion
0-10° 60.22 119 8.03 < 0 .0 5 m ore th an exp.

11-20° 60.22 50 - 1.35 > 0 .0 5
21-30° 60.22 61 0.10 > 0 .05
31-40° 60.22 69 1.17 > 0 .0 5
41-50° 60.22 83 3.05 < 0 .0 5 m ore th a n exp.
51-60° 60.22 70 1.30 > 0 .0 5
61-70° 60.22 45 - 2 .0 0 < 0 .0 5 fewer th a n exp.
71-80° 60.22 47 - 1.74 < 0 .0 5 fewer th a n exp.
81-90° 60.22 44 - 2 .1 3 < 0 .0 5 fewer th a n exp.
91-100° 60.22 112 7.05 < 0 .0 5 m ore th an exp.
101-110° 60.22 23 - 4 .8 5 < 0 .0 5 fewer th a n exp.
111-120° 60.22 34 - 3 .4 3 < 0 .0 5 fewer th a n exp.
121-130° 60.22 59 - 0 .1 6 > 0 .0 5
131-140° 60.22 80 2.65 < 0.05 m ore th a n exp.
141-150° 60.22 61 0.10 > 0 .0 5
151-160° 60.22 48 - 1 .6 1 > 0 .0 5
161-170° 60.22 42 - 2 .3 9 < 0 .0 5 fewer th a n exp.
171-180° 60.22 37 - 3 .0 4 < 0 .0 5 fewer th a n exp.
L u b b o ck L ake w ere fluvially o rien ted because tw o m a jo r g ro u p s o f o rien

tatio n s are detectab le, a n d they are a t rig h t angles to one an o th er.
B ehrensm eyer’s (1990:234) suggestion th a t “ in stro n g cu rren ts elongate
b ones generally o rien t p arallel to the flow directio n , w ith the heavier end
u p strea m [whereas] in shallow w a te r o r w eak cu rren ts elongate b ones m ay
o rien t p erp en d icu lar to the c u rre n t” does n o t in v alid ate K re u tz e r’s (1988)
co n clu sio n. It suggests o th e r lines o f evidence one m ight inspect. D o the
skeletal elem ents p lo tte d in F ig u re 6.7 have ends o f equal weight? Is there
sedim ento log ical evidence fo r tw o distinct fluvial events, one w ith a stro n g
cu rre n t (coarse sedim ent), one w ith a w eak cu rre n t (fine sedim ent)? K re u tzer
(1988) identified a significant p a tte rn in the b o n e o rie n ta tio n d a ta . O th er
analyses can refine a n d e la b o ra te h er conclusions.
V o o rh ie s’ (1969:66) ex perim ental d a ta in d icate th a t long bones tend to
o rien t p arallel to flow d irectio n if they are subm erged, b u t w hen th e b ones are
o nly p artia lly sub m erg ed they te n d to o rien t p e rp en d icu lar to flow direction.
S h ip m an (198 l b : 69) suggests the term p referred orientation d en o tes a p a rtic u
lar, d o m in a n t co m p ass o rien ta tio n ; th a t is, one o f the tw o ends o f a long bone is
typically u p strea m , o r d o w n stre am . T his h as been called p olarity, o r the
ten d en cy to have a p a rtic u la r o rie n ta tio n given in h eren t p ro p e rtie s in the
sym m etry o f the fossils (T o o ts 1965a:220). As n o ted , V oorhies (1969:66)
in d icates th a t th e “ h eav ier e n d ” o f lo n g b o n es ten d s to be u p stre a m once a long
b o n e h as been subjected to fluvial actio n . I f eith er en d is likely to be u p strea m

(o r d o w n stre am ), S hipm an (1981 b :7 1—72) suggests the term preferred axis o f
o rie n ta tio n be used in o rd e r to u n d ersco re the absence o f p o larity . M irro r-
im age rose d iag ra m s result w hen p o larity is ab se n t o r n o t recorded. R ose
d iag ram s m ay, o f course, be d ra w n w hen p o larity is n oted.
T o o ts (1965a:220) notes th a t “ in geology m o st p h en o m en a are three-
d im en sio n al [and] p ro jectin g th ree-d im en sio n al d a ta in to one p lan e . . . m ay
lead to the loss o f critical in fo rm a tio n .” T h u s, rose d iag ra m s m ay oversim plify
som e situ atio n s, a n d T o o ts (1965a, 1965b) a n d F io rillo (1988; see also R ap so n
1990) suggest using stere o g rap h ic p ro jectio n s to illu stra te the th ree -d im en
sional o rie n ta tio n o f fossils. Such p ro jectio n s g ra p h th ree-d im en sio n al o rien
tatio n s as “ lines w ith in a u n it sphere. T he p o in ts w here the lines intersect the
sphere, p ro jected o n to the h o riz o n ta l p lane, fo rm th e stereo g rap h ic p ro je c tio n ”
(F io rillo 1988:1). E ach p o in t a ro u n d th e p erip h ery o f th e p ro jec tio n (circle)
rep resen ts a specim en the long axis o f w hich is h o riz o n ta l w hereas each p o in t
n ear the cen ter o f the p ro jec tio n rep resen ts a specim en th e long axis o f w hich is
vertical. T he closer a p o in t lies to the o u tsid e edge o f the p ro jectio n , the n earer
the rep resen ted specim en is to being h o rizo n ta l; th e closer a p o in t lies to the
cen ter, th e n ea rer the rep resen ted specim en is to being vertical (90° from
h o rizo n ta l). F ig u re 6.8e show s h ow th e vertical plunge o r dip o f a bone is
p lo tte d on th e h o riz o n ta l p lan e o f a stere o g rap h ic p ro jectio n . E ach p o in t’s
p o sitio n a ro u n d the circle o f the p ro jec tio n indicates its o rie n ta tio n in the
h o riz o n ta l plane, p lo tte d to the n earest degree (sim ilar to a rose d iagram ).
F o u r m odels o f possible stereo g rap h ic p ro jectio n s are show n in F ig u re 6.8.
T hese m odels show b o th the plu n g e o r dip a n d the o rie n ta tio n o f each p lo tte d
bone. T he m odel in F ig u re 6.8a describes a case in w hich all specim ens are
v ertical o r n early so a n d th u s p referred o rie n ta tio n s are n o t detectable. F igure
6.8b illu strates a case in w hich all specim ens are nearly h o riz o n ta l a n d display
no p re ferred o rie n ta tio n . F ig u re 6.8c illu strates a case in w hich all specim ens
are essentially h o riz o n ta l a n d display p referred o rien ta tio n s. R esults like th a t
in F ig u re 6.8d in d icate no p referred o rie n ta tio n a n d ra n d o m plunge am o n g
specim ens. N o te th a t a fifth m odel, such as w ould be represented by K re u tz e r’s
(1988) d a ta show n in F ig u re 6.7, w ould p ro d u c e p o in t scatters every 90°, such
th a t F ig u re 6.8c w ou ld have tw o ad d itio n a l p o in t scatters, one on each side o f
the circle as well as a t the to p a n d b o tto m .
S tereo g rap h ic p ro jec tio n s are easy to com pile a n d g enerate. F o r exam ple,
T ab le 6.9 lists fictional d a ta fo r five long bones. T he plunge o r d ip o f each bone
is m easu red as th e degrees fro m h o riz o n ta l (0°) o f th e long axis o f each bone
(perfectly vertical is 90°). A s well, the a z im u th o f the lo n g axis o f each bone is
m easu red from th e p ro x im al to the distal end to a c co u n t fo r p o larity . T h a t is,
the az im u th is m easu red as the angle betw een the 0° line an d the line defined
from the p ro xim al end to the d istal end o f the bone (in F ig u re 6.9, the distal end
is to w a rd th e ce n te r o f the circle a n d the p ro x im al end is p o in ted to the ou tsid e
a b
stereo projection c e n t e r of
stereo projection
Figure 6.8. Idealized stereographic p rojections o f fo u r possible d istrib u tio n s o f

long bone o rien tatio n an d plunge o r dip. M odified from T o o ts (1965a) and
Fiorillo (1988). E ach individual bone is represented by a dot; b orders added to
em phasize groups, a, long axes o f bones have a n ear vertical o rien tatio n w ith no
p referred azim uth, b, long bones have an approxim ately h o rizon tal o rien tatio n
w ith no preferred azim uth, c, long bones have an appro x im ately h o rizo n tal
o rien tatio n w ith one p referred azim uth, d, long bones have a ran d o m o rien tatio n
horizontally and vertically, e, a stereographic projection fro m the side illustrating
how plunge o r dip is p lotted.
T ab le 6.9 Three-dim ensional orientation data fo r

fiv e fic tio n a l long bones ( see Figure 6.9 J
V ertical H o rizontal
Bone dip o r plunge azim uth
A 30 10
B 5 50
C 0 75
D 85 310
E 45 200
180
Figure 6.9. A stereographic projection o f th e h o rizo n tal an d vertical o rien tatio n

o f five bones (from T able 6.9).
o f th e circle). P o la rity need n o t be p lo tte d in stereo g rap h ic p ro jectio n s, in w hich

case a m irro r-im ag e stereo g rap h ic p ro jec tio n w ould result. T h e g ra p h in F igure
6.9 o f th e d a ta in T ab le 6.9 m ight be ta k e n to indicate a preferred o rie n ta tio n o f
m ost b o n es to have th eir long axes m ore o r less p arallel to a 45° o r 225° line, an d
th ree o f th e five bones are h o riz o n ta l o r nearly so.
O rientation o f carcasses
In d iv id u al a n im al carcasses m ay have a p a rtic u la r o rie n ta tio n ju s t as individual
b ones m ay. C arcass o rie n ta tio n d a ta m ay reveal aspects o f th eir tra n s p o rt an d
d ep o sitio n al h isto ry (D o d so n 1973; S chafer 1962/1972; W eigelt 1927/1989).
F o r exam ple, L y m an (1989b) m easu red the o rie n ta tio n o f carcasses o f w apiti
(Cervus elaphus) killed as a result o f a volcanic e ru p tio n . O rie n ta tio n was
Figure 6.10. D istribu tio n an d o rien tatio n o f w apiti carcasses killed by the
volcanic e ru p tio n o f M o u n t St. Helens. T hree-digit num bers den o te individual
carcass locations; arrow s p o in t from tail to head o f carcasses for w hich
o rien tatio n could be d eterm ined (from L ym an 1989b: 161, Figure 12; courtesy o f
T he C en ter for the Study o f the F irst A m ericans).
m easu red as th e az im u th o f the a rtic u la te d axial skeleton. P o la rity was

acco u n ted fo r by m easu rin g the az im u th o f the line defined from the tail to the
h ead o f each carcass (F ig u re 6.10). L ym an (1989b: 161) n o ted the com pass
b earin g fro m the carcasses to the vo lcan o a n d co n cluded th a t “ p o stm o rtem
o rie n ta tio n o f these cervids seems to be u n re la ted to the lo catio n o f the volcano
responsible fo r th eir d e a th s" based on F igure 6.10. A p lo t o f the carcass
o rie n ta tio n d a ta on a p o la r g ra p h (F igure 6.11), how ever, suggests 9 o f 13
w apiti tu rn e d aw ay fro m the clo u d o f volcanic ash (aw ay from the volcano) th a t
sw ept ov er th em a n d resulted in th eir suffocation (all carcasses a p p e a r to
rep resen t an im als th a t sim ply d ro p p e d d ead w here they stood). C h i2 analysis
in d icates, how ever, th a t this value (9 o f 13) is n o t statistically different fro m a
ra n d o m p a tte rn (ch i2 = 1.92, 0.25 > P > 0 .1 ) .
R egardless o f th e ta p h o n o m ic significance o f the w apiti carcass o rien ta tio n s,
such d a ta m ay be im p o rta n t. W h eat (1972:29), fo r instance, co n cluded th a t the
“ o rie n ta tio n s ” o f 14 co m plete bison skeletons at the O lsen -C h u b b u ck site in
C o lo ra d o w ere due to th eir being p acked in to the b o tto m o f a n arro w arro y o .
T w o o f th e carcasses h ad , in fact, “ nearly v ertic al" o rien ta tio n s. A n a d d itio n al
27 n early com plete skeletons displayed o rie n ta tio n s sim ilar to the 14 com plete
skeletons, a n d th eir o rie n ta tio n s also seem ed to be related to th eir p o sitio n in
Figure 6.11. A zim uth o f w apiti carcasses killed by the volcanic eru p tio n o f
M o un t St. H elens. E ach sm all circle represents a single carcass; p o larity is
indicated w ith tail at center o f large circle an d head aw ay from the center. N o te
th a t m any carcasses are facing aw ay from the crater (see text for fu rth er
discussion). C om p are w ith Figure 6.10.
the a rro y o . These few exam ples un d ersco re n o t only the im p o rtan ce o f carcass
o rie n ta tio n d a ta fo r ta p h o n o m ic concern s (w hy are the skeletons n o t in life-
positions?), b u t the significance o f the geological co n tex t o f the carcasses. T he
latter, o f course, applies w ith eq u al force to the discussion o f b o n e o rie n ta tio n
above. It sh o uld also be o b vious th a t if o ne ca n m easu re the o rie n ta tio n o f a
carcass, one has p ro b a b ly fo u n d the lo catio n w here the carcass o f the dead
an im al w as d ep o sited , if n o t w here the an im al in fact died, a n d th u s dispersal o f
in d iv idu al b ones from th a t lo catio n co u ld be m easu red if individual bones can
be an a to m ica lly o r m echanically refit to the skeleton (see C h a p te r 5 on
refitting).
Abrasion
B ehrensm eyer (1975b) an d K o rth (1979) suggest th a t fluvial tra n s p o rt m ight
resu lt in the a b ra sio n o f bones (see also B rom age 1984). A b ra sio n (see the
G lo ssary ) results fro m th e tu m b lin g o f b ones in a liquid th a t co n ta in s sedim ent.
T h e physical ero sio n o f bone results fro m its c o n sta n t b u t shifting c o n ta c t w ith
the sedim ent grains. F o r exam ple. B oaz (1982:142) re p o rted th a t in a sam ple o f
236 w ildebeest (C onnochaetes taurinus) b ones she collected from fluvial
en v iro n m en ts in A frica, 59 displayed no evidence o f any kind o f d am ag e, 48
h ad been b ro k e n by carnivores, 61 h ad been b ro k e n a n d th en a b ra d e d , a n d 67
show ed a b ra sio n d am ag e only. She n o ted th a t the effects o f ab ra sio n by the
sandy sedim ent w ere “ ch a rac te rized by a ‘w earin g -aw ay ’ o f the o u te r tab le o f
bo ne a n d expo su re o f the in n er cancellous p o rtio n ” (B oaz 1982:147). K o rth
(1979:263-265) exam in ed tw o m ouse skeletons (one each o f P erom yseus an d
M icro tu s) in tu m b lin g b arrels c o n ta in in g w a te r an d q u a rtz grains averaging 2 -
4 m m in diam eter. T he b arrels w ere ro ta te d so as to a p p ro x im a te a linear
velocity o f 24 cm /sec, a n d the b ones w ere exam ined every 10 h o u rs fo r 80 h ours.
T he m ouse skeleto n s first d isarticu lated , a n d the h y p so d o n t (rootless) teeth fell
from the m axillae a n d m andibles, b u t the ro o ted teeth stayed in th eir alveoli.
B ones o f th e skull th en d isarticu lated , a n d th eir edges becam e ro u n d e d . S harp
edges o f all b ones show ed evidence o f ro u n d in g early in the process, an d bone
surfaces becam e progressively th in n er. E ventually, som e b ones b ro k e due to
the th in n in g .
S h ipm an an d R ose (1983a:79, see also S hip m an an d R ose 1988) c a u tio n th a t
“ b ones in a tu m b lin g b arrel are exposed to the im p act o f sed im en tary particles
m o re co n tin u o u sly th a n are b ones in n a tu ra l stream c o n d itio n s [and] tu m b led
bones are also subjected to m o re c o n sta n t velocity th a n is likely to be realistic.”
T h eir tu m b lin g b arrel ex perim ents w ith b ones o f dom estic sheep (O vis aries)
in d icate th a t m icrosco pic features on bo n e surfaces are quickly a n d often
com pletely rem oved by ab rasio n , bu t “ grossly a p p a re n t changes in bone
surfaces o ccur afte r a b o u t 35 h o u rs o f a b ra s io n ” (S hipm an a n d R ose
1983a:79). In th eir exp erim ents “ sed im en tary ab rasio n rarely p ro d u c ed
scratches o r o th e r elo n g ate grooves, regardless o f the sedim ent size, the
in clusion o r exclusion o f w a te r [in the tu m b lin g barrels], the co n d itio n o f the
b ones (fresh, w eath ered, fossilized, w hole, o r b ro k en ), a n d the d u ra tio n o f
tu m b lin g ” (S h ip m an a n d R ose 1983a:79).
S hip m an an d R ose (1988:317) list six facto rs th a t influence the rate and
n a tu re o f a b ra sio n o f b o n e by sedim ents:
1. the grain size o f the sedim entary particles w ith w hich the bones are tran sp o rted ;
2. the com position o f the sedim entary particles;
3. the presence o r absence o f soft tissue on the bone;
4. the co n d itio n o f the b one a t the onset o f tra n sp o rt (fresh o r w eathered; b ro k en or
whole; m ineralized or unm ineralized);
5. the presence o r absence o f w ater in the sedim entary system; and
6. the d u ra tio n o r distance involved in tran sp o rt.
T hey observed th a t only a few bones show ed m acroscopically o r grossly visible

ab rasio n features, such as exposed cancelli (as re p o rted by B oaz 1982), a fte r 20
to 35 h o u rs o f tu m b ling . S hape changes in b ones w ere extrem ely subtle early in
th e tu m b lin g process. “ T he sm aller the g rain size, the faster the ra te o f
ab rasio n , all o th e r thin gs being e q u a l” (S h ip m an a n d R ose 1988:323). M ore

a n g u la r (“ s h a rp e r” ) sedim ent a b ra d e d bones faster th a n m ore ro u n d e d sedi
m ent particles. B one edges a b ra d e d faster th a n surfaces, p ro b a b ly because o f
g re ater exposed surface a rea relative to volum e (see also A ndrew s 1990).
F in ally , convex surfaces o f b ones ten d ed to a b ra d e fa ste r a n d p rio r to concave
surfaces.
R ecognizing traces o f ab ra sio n is only one aspect o f th e pro b lem , as several
processes can a b ra d e bone, processes such as tra m p lin g (B rain 1967a), eolian
activity (S h ip m an a n d R ose 1988), a n d fluvial tra n s p o rt. T he fo rm er tw o
processes are discussed in C h a p te r 9, w here it is n o ted th a t the d istrib u tio n o f
ab rasio n d am ag e across ind iv id u al bo n e specim ens o ften can help the analyst
d eterm in e w hich process w as responsible fo r the ab rasio n . It suffices here to
n o te th a t fluvial tra n s p o rt o f b ones ab ra d e s the entire surface o f the specim en
w hereas ab ra sio n by eolian activity a b ra d e s only the exposed o r to p surface(s)
o f specim ens. T ra m p lin g m ay a b ra d e bones, b u t also creates deep scratches in
bo ne surfaces, so m eth in g fluvial ab ra sio n does n o t pro d u ce. F inally, sedim en-
tological analysis a n d d e te rm in a tio n o f w h e th e r the m atrix enclosing the bones
is eolian o r fluvial in origin m ay help the an aly st so rt o u t the responsible
ta p h o n o m ic process. B ut, w hile it seem s th a t c o a rser a n d m ore an g u lar
sedim ents m ay a b ra d e bones m o re extensively a n d ra p id ly th a n fine tex tu red ,
su b a n g u la r o r ro u n d e d sedim ent particles, G ra h a m a n d K ay (1988:237)
c a u tio n th a t “ re d ep o sitio n a n d recycling o f bo n e specim ens m ay m ake
b reak ag e, polish, a n d ab ra sio n a p p e a r in co n g ru o u s w ith sed im en tary particle
size.” T hey fo u n d heavily a b ra d e d bo n e specim ens sp atially associated w ith
u n a b ra d e d b o n e specim ens, a n d all w ere in ra th e r fine sedim ent. T h u s w hile the
sed im en tary co n tex ts o f th e b ones are im p o rta n t ta p h o n o m ic variables (see
C h ap ters 10 a n d 11), the an a ly st sh o u ld n o t be m islead by th a t context.
D ispersal by hom inids and scavenging carnivores

O th er intensively studied dispersal m echanism s include an im als th a t m ove o r
collect bones. M o st h ave been studied fro m th e v an tag e o f the site to w hich
these dispersal m echanism s b rin g bones a n d dep o sit them , such as lairs, nests,
o r dens. T h u s I co n sid er th em in m o re d etail in the discussion o f ac cu m u latio n
below . H ow ever, th ere is one in terestin g w ay th a t the an aly st m ight study
d ispersal o f an im al bones by m a m m alian ca rn iv o res a n d scavengers fro m a kill
site. B lu m enschine’s (1986a) general sequence o f c o n su m p tio n o f carcass p arts
(T able 5.7) in dicates w hich carcass p a rts generally are the first to be rem oved
(consum ed o r sp atially m oved aw ay) from the site o f an im al d eath . B ones o f the
h in d q u a rte r a n d lu m b a r region are the first to go, b ones o f the fo re q u a rte r are
second, a n d b ones o f th e h ead are the last. S im ilar sequences o f tra n s p o rt o f
skeletal p a rts aw ay fro m a kill site by h u m a n s have also been developed (see
C h a p te r 7). T h e p o in t here is, th e bones p re sen t offer a clue as to w h eth er the
12
10
8
je:
ro 6
DC
Figure 6.12. B lum enschine’s (1986a) co n su m p tio n sequence. T he ran k is the

o rd er in which flesh associated w ith the indicated skeletal elem ent is consum ed
(from T able 5.7). C.S., co n su m p tio n sequence; I.C .S., inverse con su m p tio n
sequence.
assem blage rep resents a kill site from w hich b ones w ere rem oved o r a site to
w hich b ones w ere b ro u g h t.
B lum enschine’s (1986a) c o n su m p tio n sequence can be m odeled as a b ar
g ra p h , th e h eig h t o f th e b ars being based on the ra n k o rd e r o f co n su m p tio n o f
each skeletal p a rt (F ig u re 6.12; see B lum enschine an d C av allo [1992:93] for a
less sch em atic illu stratio n ). T he b a r height is eq u iv alen t to the relative
frequency o f each p a rtic u la r skeletal p a rt one w ould expect to find at a kill site
w here carn iv o res v ariou sly co n su m ed a n d rem oved b ones follow ing the
c o n su m p tio n sequence. T he inverse c o n su m p tio n sequence (B lum enschine
1986a) can sim ilarly be m odeled (F ig u re 6.12), an d the bars rep resen t the
relative frequencies o f skeletal p a rts one w ould expect at a den o r lair to w hich
bones h a d been tra n sp o rte d . F o r exam ple, B inford (1981b:214—216) su m m a
rized the frequencies o f bones from ca rn iv o re kills o f w a te rb u ck (Kobus
ellipsiprym nus) in E ast A frica (as re p o rted by A. P. H ill) a n d frequencies o f
reed b u ck ( R edunca sp.) bones fro m a ca rn iv o re den (as re p o rte d by R . G.
K lein). P lo ttin g th o se tw o sets o f frequencies ag a in st B lum enschine’s c o n su m p
tio n sequence suggests th ere is little re la tio n sh ip betw een the tw o variables
(F ig u re 6.13). In fact, th ere is n o significant c o rre la tio n betw een the co n su m p
tion sequence an d eith er bone assem blage (for the kills rs = 0.06, P = 0.82; for
th e den r s = 0.30, P = 0.34). O th e r evidence suggests the frequencies o f bones
have been significantly influenced by the d e stru c tio n o f m an y o f them due to
carn iv o re c o n su m p tio n . T hus, in this case, th e an aly st w ould w a n t to p ursue
100 -
80 -
I* 60 -
0
cr
1 40-
20 -
Figure 6.13. Frequencies o f skeletal elem ents from carnivore kills a n d from a
carnivore den plo tted against B lum enschine’s (1986a) con su m p tio n sequence (see
Figure 6.12).
o th e r lines o f evidence w hen studying the dispersal o f bones, including all

facto rs th a t m ight influence b o n e frequencies (see C h a p te r 7 fo r extensive
discussion). N o netheless. B lum enschine’s (1986a) c o n su m p tio n sequence is a
v aluable an alytical to o l th a t can be used as p a rt o f initial analyses exam ining
the d ispersal o f b ones from the lo catio n o f carcass d ep o sitio n . F inally, the
inverse c o n su m p tio n sequence can be used to help stu d y b o n e ac cu m u latio n ,
o u r next top ic o f discussion.
Analyzing accumulation
1 arb itrarily define cultural b one as those fragm ents o f n o n -h u m an to o th and
osseous m aterial deposited as the result o f h u m an activity. Bone deposited by o th er
m echanism s is term ed natural bone.
(D. H. T h o m as 1971:366)
B ackground scatters
Z o o arch ae o lo g ists a n d ta p h o n o m ists typically deal w ith b o n e assem blages
th a t w ere collected from sp atial loci displaying higher th a n average co n c en t
ra tio n s o f anim al rem ains. A n o b vious ta p h o n o m ic p ro b lem th u s concerns
id entifying the agent(s) o r process(es) responsible for creatin g th o se dense
c o n c e n tra tio n s o f bones. O ne w ay to begin to a p p ro a c h this p ro b lem is to
exam ine w h a t the average o r n o rm a l b o n e ac cu m u latio n m echanism s are on
the lan d scap e, a n d /o r to d eterm in e w h a t has been referred to as th e “ b a c k

g ro u n d ” sca tte r o r n o rm al density o f faunal rem ains across the lan d scap e (e.g.,
B ehrensm eyer 1983).
B ehrensm eyer (1982, 1983, 1987; B ehrensm eyer a n d B oaz 1980; B ehrens
m eyer et al. 1979) has gone fa r to w a rd s identifying a n d describing w h a t typical
b a c k g ro u n d scatters o f bones lo o k like in eastern A frica. In stu d y in g the
passive ac cu m u latio n o f bones across the lan d scap e, researchers begin w ith the
o b serv atio n th a t “ th e n u m b e r o f b ones o f a species w hich ac cu m u late in an
e n v iro n m e n t dep end s initially on the p o p u la tio n size a n d an n u a l d e a th ra te o f
th a t species” (B ehrensm eyer a n d B oaz 1980:75). T hey also p o in t o u t th a t the
rem ain s o f sm aller ta x a are less likely to (a) preserve a n d (b) be recovered
(B ehrensm eyer et al. 1979; see C h a p te r 9 fo r detailed discussion). B ehrens
m eyer (1983) notes th a t th ere are differences in b a c k g ro u n d scatters th a t tend to
c o rrelate w ith the d ep o sitio n al m icro en v iro n m en t. F o r exam ple, the average
density o f fa u n al rem ain s (tax o n o m ically identifiable specim ens only; ribs,
ste rn a b ra e , a n d long bo n e d iaphysis frag m en ts n o t included) in the sw am p
h a b ita t (0.0083 sp ecim en s/m 2) is fo u r tim es hig h er th a n it is in the b ush (0.002
sp ecim ens/m 2). T hese densities re p resen t an ac cu m u latio n p erio d o f 10-20
years, a n d p ro jectin g these frequencies across g re ater tim e sp an s a n d co rrectin g
fo r the fact th a t n o t all o f the specim ens will be bu ried a n d th u s preserved for
fu tu re recovery, B ehrensm eyer (1983) suggests th a t only a b o u t 4.4 specim ens/
m 2 will be d ep o sited a n d preserved over 10,000 years.
F req u en cies o f skeletal p a rts differ n o t only w ith d ep o sitio n al h a b ita t, they
also v ary w ith w h eth er a bo n e assem blage represents a passive ac cu m u latio n
across th e lan d scap e o r an active, spatially focused ac cu m u latio n . D a ta (from
B ehrensm eyer 1983; B ehrensm eyer a n d B oaz 1980) in d icatin g this v aria tio n
are su m m arized g rap h ically in F ig u re 6.14. T here, it is clear th a t the assem blage
labeled “ dispersed b o n e s” has a p p ro x im a te ly equal n u m b ers o f skulls, verte
b rae, fo relim b bones, a n d h in d lim b bones; this assem blage represents the
passively ac cu m u lated a ttritio n a l b a c k g ro u n d sca tte r o f bones across the
landscape. T he assem blage labeled " p re d a tio n p a tc h ” is also a passive ac cu m u
latio n , b u t it is one rep resen tin g a geographic lo catio n w here m ultiple anim als
were killed, one a t a tim e, over tim e by p re d a to rs. T he p re d a tio n p atch
assem blage ten ds to h av e relative bo n e frequencies o p p o site th o se o f the
“ hyen a d e n " assem blage, reflecting the fact th a t hyenas have selectively
rem oved b on es fro m an im al d e a th loci to th eir dens. T he “ b u ried b o n es”
assem blage ten d s to reflect fairly accu rately the “ average sk eleto n ” assem
blage, suggesting th a t b o nes o f the fo u r m a jo r categories o f skeletal elem ents
p lo tte d all have a b o u t the sam e chance o f being buried a n d th u s in co rp o rated
in to the (fu tu re) fossil record. T h ere are n o tab le differences, how ever, betw een
th e “ averag e sk eleto n ” a n d o th e r assem blages, all p ro b a b ly reflecting differen
tial dispersal a n d d e stru c tio n o f b ones w ithin the fo u r categories o f skeletal
p arts.
Accum ulation and dispersal o f vertebrate rem ains 191
Skull
60 Vertebra
o Forelimb
c: Hindlimb
o
cr
a)
k_
40
LL
C ,
0
1 20
o
CL
0
A verage Dispersed Buried
Skeleton Bones Bones
Figure 6.14. Relative frequencies o f skeletal p o rtio n s in different types o f bone

accum ulations (after B ehrensm eyer 1983; B ehrensm eyer an d Boaz 1980).
H ay nes (1988b) presen ts d a ta on n a tu ra l b a c k g ro u n d scatters o f bones for

b o th A frica an d C a n a d a . H e c o m p ared five assem blages o f b ones fro m “ m ass
d e a th sites” w ith tw o “ cu m ulative assem blages” o f bones. A m ass d e a th site is
defined as a “ relatively circum scribed locus w here a h e rd o r g ro u p o f anim als
(o f one o r m o re tax a) died over a b rie f tim e sp an d u e to a single agency o f
d e a th .” an d a cu m u lative site is defined as a place “ w here bones o f large
m am m als h av e am assed o ver tim e d u e to n u m ero u s different m o rta lity events
such as serial p re d a tio n , the re g u lar an d h ab itu a l killing o f prey an im als in the
sam e loci” (H ay nes 19 8 8 b :2 19). H ay n es fo u n d th a t the ra tio o f bones per
in d iv idu al (M N E :M N I) did n o t differ betw een the tw o kinds o f assem blages,
b u t th e density o f b ones (frequency p er m 2) w as g re ater a t the m ass d e a th sites
(avg. = 0.14) th a n a t the cu m ulative sites (avg. = 0.09). H e notes th a t sites w ith
high densities o f b o nes b u t only one ta x o n rep resen ted “ are n o t alw ays the
result o f m ass d e a th s ” (H ay n es 1988b:230). H aynes also notes, as B ehrens
m eyer did. th a t b o nes o f sm all an im als have less chance o f survival a n d
in c o rp o ra tio n in to th e fossil reco rd th a n b ones o f large anim als (see C h a p te r 9).
F inally, he re p o rts th a t ca rn iv o res an d scavengers tend to rem ove bones from
carcasses in m ass d e a th sites b u t feed heavily on bones at d e a th sites when
carcasses are scarce.
Passive m ass accum ulations

T h e w o rld -fam o u s L a B rea ta r pits c o n ta in a w ell-know n passively ac cu m u
lated , m u lti-ta x o n , m u lti-in d iv id u al fossil assem blages (e.g., S tock 1956). T he
frequency o f b ones p er u n it a rea ten d s to be q u ite high in such settings. T here

are several w ays fo r such m ass accu m u latio n s to be gen erated passively.
H aynes (1988b. 1991), fo r exam ple, has fo u n d th a t loci such as w aterholes
w here m u ltip le indiv idu als die, even th o u g h it m ay be one individual dying at a
tim e, ten d to have large, dense a c cu m u latio n s o f bones. In d iv id u als m ay die
from th irst such as d u rin g tim es o f d ro u g h t, from sta rv a tio n , o r from m ultiple
events o f p re d a tio n by carnivores. In a w ay, the w aterholes serve as m agnets,
draw in g in n u tritio n a lly stressed individuals, o r p o ten tial prey.
N a tu ra l tra p s m ay o r m ay n o t serve as m ag n ets for an im als. F o r exam ple,
O liver (1989) describes a n a tu ra l pit-fall tra p . Shield T ra p C ave, in M o n ta n a .
H is analysis o f a p o rtio n o f th e excavated m aterial revealed over 5,000
identifiable bones rep resen tin g a t least 52 individual m am m als o f 10 species.
M o st o f these rem ains, O liver (1989) believes, rep resen t an im als th a t acciden
tally fell in to the cave a n d eith er died as a result o f the fall o r could n o t escape
from the cave a n d died fro m starv a tio n . O liver (1989:81) suggests th a t
“ carn iv o res w ere a ttra c te d to th e tra p by the presence o f d ead a n d /o r dying
anim als on th e cavern flo o r" b u t few w ere able to survive the 14 m fall in to the
cave; th o se th a t did soo n succum bed to injuries resulting from th e fall. T he
n o tio n th a t som e n a tu ra l tra p s served as m ag n ets for c a rn iv o res is em ployed as
an e x p lan a to ry device by W hite et al. (1984). T hey fo u n d th a t som e caves w hich
served as n a tu ra l tra p s c o n ta in e d m uch hig h er p ro p o rtio n s o f carn iv o res
relative to herbivores (average o f 2 caves = 32.8% o f the M N I) th a n (a) caves
w hich did n o t serve as n a tu ra l tra p s (average o f 18 caves = 4 .7 % o f the M N I),
a n d (b) a living m am m alian fa u n a (2% carnivores). T hey explained this
difference as resulting from “ the pecu liar co n fig u ratio n o f the tw o [carnivore-
trap ] caves an d th e b eh a v io r o f the ca rn iv o res” (W hite et al. 1984:250). In
p artic u la r, th e tw o caves they believe to re p resen t ca rn iv o re tra p s w ere difficult
to get o u t of, a n d the carn iv o re tax a rep resen ted by the faunal rem ains tend to
be o p p o rtu n istic feeders o f c a rrio n . A nim als th a t died in the carn iv o re tra p
caves (including carn ivores) w ould have served as “ b ait fo r the tr a p ” (W hite et
al. 1984:246).
In brief, passive m ass accu m u latio n s o f bones have th ree distinguishing
attrib u tes. F irst, th ere m u st be som e fa c to r w hich results in anim als being
a ttra c te d to a locus y ear a fte r year. Second, the p ro b a b ility th a t a t least som e o f
these anim als will die in th a t attra c tiv e (?) locus m u st be g re ater th a n in o th er
loci. T h ird , th e an im als effectively ac cu m u late them selves; th a t is, th ere is no
ag ent o f carcass o r b o n e a c cu m u latio n th a t is ex tern al to the accu m u lated
carcasses o r bones, b u t ra th e r the process o f a c cu m u latio n involves the
b eh av io rs o f the ac cu m u lated an im als them selves. In the preceding few
p a ra g ra p h s, several such loci have been identified. T hese include n a tu ra l trap s
w here th e a ttra ctiv e feature seems to be c a rrio n , the “ a ttra c tiv e ” featu re is th a t
th e tra p is hid d en , o r the attra c tiv e fe atu re m ay be a relatively iso lated w atering
hole d u rin g d ry seasons.
A ctive m ass accum ulations

In c o n tra s t to passive m ass ac cu m u latio n s, active m ass accu m u latio n s are
created by b o n e-accu m u latin g agents a n d processes th a t are ex tern al to the
ac cu m u lated carcasses an d bones. T hese ag en ts an d processes include geologi
cal processes such as fluvial ac tio n , a n d biological processes such as h u m an s
an d d en n in g carnivo res. I only briefly co n sid er geological processes in the
follow ing as they have been d ea lt w ith in earlier p o rtio n s o f this c h a p te r. I then
co n sid er som e o f the n a tu ra l biological processes, a n d then tu rn to h u m an s as
b o n e accu m u lato rs.
Fluvial accum ulations

F requ en cies o f skeletal p a rts can be a ttrib u te d to fluvial ac cu m u latio n by
reference to V o orhies G ro u p s o r the like (F ig u res 6.5 a n d 6.6; T ab le 6.5).
A ndrew s (1990:18-19) atte m p te d to sim ulate the effects o f fluvial tra n sp o rt on
sm all m am m al bones by placing a sam ple o f ro d e n t bones in a ro ta ry tu m b le r
alo n g w ith different sedim ent grades (from relatively fine [silt a n d sand] to
co arse [gravel], a n d several large clasts ad d ed to the p rev io u s tw o). H e found
th a t afte r a b o u t 200 h o u rs o f tu m b lin g there w as “ a slight degree o f ro u n d in g o f
the b o n es [and] afte r 300 h o u rs ro u n d in g w as p ro n o u n c e d " fo r all sedim ent
grades. B ones w ere b ro k e n a n d chipped m o re w hen large clasts w ere included.
H e con clud es th a t “ th e n a tu re o f the b re ak ag e is largely co n tro lled by the
s tru c tu ra l p ro p e rtie s o f the b o n e s” (A ndrew s 1990:19).
H arvester ants as bone accum ulators

In a u n iq u e stu d y S hipm an a n d W a lk e r (1980) d o cu m en t the bone-collecting
activities o f h arv ester an ts in A frica. T he assem blage o f fau n al rem ains created
by these insects is ch a rac te rized by dense co n c e n tra tio n s o f bones a n d teeth
a ro u n d th e e n tra n ce tu n n els o f the insects’ b u rro w s. T he tax a represented are
all sm all a n d o f local d eriv atio n . T here is a w ide diversity o f m am m alian taxa
represented , inclu d ing d iu rn a l a n d n o c tu rn a l, a n d fossorial a n d n o n -fo sso rial
form s; this m ay be different fro m a ttrib u te s o f a fa u n a rep resen ted by rem ains
collected by, say, a n o c tu rn a l m am m alian p re d a to r o r ra p to r (see below ). T he
an t-g e n era te d bo n e assem blage co n tain e d a b u n d a n t “ ro b u s t” skeletal ele
m ents (S h ip m an a n d W alker 1980:499), a them e we re tu rn to below .
R odents as bone accum ulators

H offm an an d H ays (1987) c o n d u c te d an experim ent in w hich they placed a
n u m b e r o f m am m al b o nes in a cave in N o rth A m erica occupied by the eastern
w o o d ra t (N eoto m a floridana). T hey fo u n d th a t this sm all ro d e n t m oved an d
re d istrib u ted m an y o f the bones. T h e w ood ra ts did n o t seem to select bones for
gnaw ing o r on the basis o f b o n e m o rp h o lo g y o r texture, b u t ra th e r m oved all
b ones < 100 g in w eight a n d > 0.3 g. T his species o f w ood ra t weighs betw een
170 a n d 340 g. V arying p ro p o rtio n s o f m am m al, fish, a n d tu rtle b ones were

m oved by th e w o o d rats; overall, 55% o f all bones w ere m oved. H o riz o n ta l
m ov em ent w as, on average, 1 to 2 m , w ith a m ax im u m distan ce o f 5 m . M an y
b ones w ere also m oved 1 to 2 m vertically. H offm an an d H ays (1987) d o not
in dicate h ow a ta p h o n o m ist m ig h t analytically d eterm in e if w o o d ra ts have
re d istrib u ted bones in a site.
H o c k e tt (1989b) indicates the frequencies o f skeletal p a rts accu m u lated by
w ood ra ts m ay be a fu n ctio n o f w h a t w as available for them to accum ulate.
“ V erteb rae a n d ribs m ay be co m m o n sim ply as a reflection o f the fact th a t they
are a b u n d a n t elem ents in an u n g u late ca rca ss” a n d ra re elem ents m ay be rare
because m an y o f them w ere co n su m ed o r rem oved fro m a carcass by carnivores
(H o c k ett 1989b:31). R eg ard in g the latter, H o c k ett (1989b) re p o rts th a t 51 % o f
the n o n -ro d e n t b o n es he collected fro m w o o d ra t nests display evidence o f
ca rn iv o re gnaw ing. Few o f them , how ever, displayed signs o f hav in g been
gnaw ed by the w o o d rats. H o c k e tt suggests th a t the size o f a bone m ay lim it
w h eth er it is tra n sp o rte d by a w ood ra t. H o c k e tt (1989b:33) suggests bones
> 1 .1 cm w ide, > 2 9 .5 cm long, a n d > 5 4 .5 g can n o t be tra n sp o rte d by the
bu sh y -tailed w ood ra t (N eotom a cinerea), a n d if “ no n e o f th e th ree individual
[sizes is exceeded by a fossil bone], th en the bo n e in q u estio n m ay h av e been
d ep o sited a t the site by w ood ra ts;” larg er b ones w ere p ro b a b ly d ep o sited by
o th e r agents. I w ould a d d th a t the presence o f w ood ra t rem ains in the bone
assem blage, evidence o f w o o d ra t nests in the site, a n d the presence o f ro d e n t
gnaw ing m a rk s on som e bones in the assem blage m ight be tak en as circ u m sta n
tial evidence th a t this sm all ro d e n t is p erh ap s responsible fo r the d istrib u tio n o f
som e b o n es (e.g., G ra y so n 1988; L ym an 1988a). F igure 6.15a show s three
do m estic sheep (Ovis aries) astrag ali collected fro m a w o o d ra t nest in N ew
M exico. A ll show typ ical extensive ro d e n t gnaw ing, a n d while the precise
id en tity o f th e gnaw ing ag en t is n o t kno w n , the co n tex t o f the bones suggests
w ood ra ts are responsible. F ig u re 6.15b show s tw o distal tibiae o f deer
(Odocoileus sp.). O ne o f them w as collected fro m a ro ck sh elter in M issouri an d
displays extensive ro d e n t gnaw ing. A gain, the precise id en tity o f the gnaw ing
agen t is n o t kno w n , b u t m a rm o t (M a rm o ta sp.) rem ains are co m m o n in the
ro ck sh elter dep o sits a n d this large sciurid lives in a n d a ro u n d the sh elter to d ay .
B rain (1980, 1981) re p o rts th a t the A frican p o rc u p in e (H y strix africaeaustra-
lis) tra n s p o rts a n d accu m u lates m am m al rem ains (bone a n d h o rn ) a n d to rto ise
shell. T his ro d e n t does n o t seem to d estro y b ones o f low stru c tu ra l density (see
C h a p te r 7). T he A frican p o rc u p in e collects a w ide ran g e o f sizes o f bone
specim ens; B ra in ’s (1980) sam ple fro m a lair occupied by this species co n tain e d
specim ens ran g in g betw een 1 an d 750 g, w ith the m ajo rity (70% ) w eighing
betw een 1 a n d 50 g. T he m axim um length o f the specim ens is 90 cm , b u t the
m ajo rity fall betw een 2 a n d 15 cm . O f the 1708 to ta l b o n e specim ens collected
from th e lair, 1,043 (61 % ) display evidence o f having been gnaw ed by
p orcu p ines. M o st o f the b ones collected by the p o rcu p in es are “ bleached,
d efatted b o n es” (B rain 1980:123; 1981:116). B rain (1981:116) concludes th a t

“ the m o st reliable in d icatio n th a t a b o n e a c cu m u latio n in a cave has been built
up by po rcu p in es is th e presence o f typical gnaw ing m ark s on d efatted an d
freq uen tly w eathered b o n es.” Im p o rta n tly , B rain (1981:117) no tes th a t “ the
incidence o f gnaw ed b o n es in the w hole collection fro m an y site will dep en d on
th e a b u n d a n c e o f b o n es available to the p o rc u p in e s a t the tim e;” if b ones are
readily available, few will be gnaw ed; if b ones are scarce, m an y will be gnaw ed.
M ag u ire et al. (1980:91) in d icate th a t the A frican p o rc u p in e creates tw o
types o f d am ag e to u n g u late bones. F irst, gnaw ing m ark s tak e th e form o f
“ b ro a d , co n tig u o u s shallow scrape m a rk s .” Second, cancellous bo n e is often
scooped o u t o f the ends o f long b ones, creatin g “ tu b u la r sh a fts” (M ag u ire et al.
1980:93). T he scoo p ed o r hollow ed o u t cancellous bo n e m ay resem ble c a rn i
vore d am age, b u t M ag u ire et al. (1980:91) re p o rt th a t distinctive p o rcu p in e
gnaw ing m ark s a re “ in v ariab ly also p re se n t.” T hey re p o rt th a t “ p o rc u p in e s are
in cap ab le o f splittin g o r cracking th e shafts o f lim b bones, so th a t b o n e flakes
are exceedingly ra re in porcupine-collected bo n e a c c u m u la tio n s” (M ag u ire et
al. 1980:93). T he gnaw ing they illu strate is sim ilar to th a t show n in F igure 6.15.
D ix o n (1984) suggests the N o rth A m erican p o rc u p in e (E rethizon dorsatum )
m ay accu m u late bones in caves a n d dens, b u t th u s fa r clear d o c u m e n ta tio n o f
such b eh a v io r is n o t available. It has been d o cu m en te d th a t this species gnaw s
b ones (D ixo n 1984). It seems b o th A frican a n d N o rth A m erican p o rcu p in es
collect b o n es fo r gn aw in g in o rd e r to keep th eir incisors in g o o d co n d itio n , th u s
it is n o t surprising th a t m ost o f the bones they collect are d ry an d at least slightly
w eath ered because d ry b o nes are easier to gnaw (are less slippery from grease;
see C h a p te r 9). A s well, this p ro b a b ly ac co u n ts for the fact th a t bones o f sm all
v erteb rates d o n o t seem to be collected by p o rc u p in e s (A ndrew s 1990:7).
Birds as bone accum ulators

L im ited research in dicates v ario u s species o f (A frican) v u ltu re (A ccipitridae)
accu m u late b ones a t th eir nesting sites (M u n d y a n d L edger 1976; P lug 1978).
T his b eh a v io r a p p a ren tly p rovides a source o f calcium for grow ing chicks
(M u n d y a n d L edger 1976; R ich ard so n et al. 1986). T h ere are no ro d e n t o r sm all
m am m al bones in som e o f these collections (P lug 1978). M an y o f the b ro k e n
b ones show evidence o f h av in g been b ro k e n by carn iv o res, a n d som e bones are
still artic u la te d , suggesting these specim ens m ay have been in tro d u ce d to the
nest as sources o f m e a t o r o th e r soft tissue fo r the chicks (M u n d y a n d L edger
1976; P lug 1978). B one specim ens o f b ovids are virtually all b ro k e n an d range
in length from 1 to 40 cm , w ith a m o d al value o f 3 -6 cm . Som e o f the bones th a t
w ere reg u rg ita ted by v u ltu re chicks h ad been p artia lly digested a n d have a
c o rro d ed a p p e ara n ce (R ich ard so n et al. 1986:38). “ A rtifa c ts” are also a c cu m u
lated by v ultures. T h u s w hen a nest d eterio rates a n d collapses, the b o n e a n d
a rtifa c t ‘fa ll-o u t’ could create a sm all ‘p seu d o -site’ o n the g ro u n d .
S o lo m o n et al. (1986) re p o rt th a t the A u stra lia n g re at bow er b ird (C hlam y-
Figure 6.15. R odent gnaw ed bones, a, distal view o f three dom estic sheep (Ovis
aries) astrag ali th a t have been extensively gnaw ed by w ood rats (N eotom a sp.)
recovered from a w ood ra t nest in N ew M exico; note the exposed cancellous
bone tissue in the tw o left specim ens an d the incisor grooves on the left m argin o f
all three; b, a long bone shaft fragm ent gnaw ed by m arm o ts (M arm ota sp.)
recovered from a rockshelter in M issouri; c, distal tibiae o f deer (Odocoileus sp.),
up p er specim en gnaw ed by m arm o ts an d recovered from a rockshelter in
M issouri, low er specim en is an ungnaw ed com p arativ e specim en placed in the
sam e o rien tatio n as the u p p er specimen.
dera nuchalis) collects bones as well as artifa cts. T hey suggest “ a n a b a n d o n e d ,

d ecom po sed a n d dispersed collection o f b ow er objects co u ld be m isin terp reted
as a n open arch aeo logical site” (S olom on et al. 1986:308). T hey suggest bow er
birds m ay rem ove bo n es fro m existing arch aeo lo g ical sites o r a decom posed
bow er n ear a site co uld c o n ta m in a te th a t site. T hey collected over 300
v erte b rate specim ens, all o f w hich rep resen t m am m als, from tw o bow ers. All
specim ens are < 3 0 g , an d 84% o f them are < 10 g. All specim ens are < 19.0 cm
long, a n d 81% are 0.2 to 0.9 cm in m ax im u m dim ension. A b o u t 80% o f the
bones have length:w eight ra tio s less th a n 30 m m /g. T he p ro p o rtio n o f
specim ens identifiable to skeletal elem ent a n d the p ro p o rtio n identifiable to
tax o n are b o th high relative to zo o a rch aeo lo g ical collections. N early 80% o f
the specim ens have a glossy sheen o r polish “ o n the ends o f long b ones an d in
the m iddle o f sh o rt, ch u n k y b o n es” (S olom on et al. 1986:314). A few specim ens
have p u n ctu re s a n d beak m ark s. F ra g m e n ta ry specim ens seem to have been
b ro k e n w hen dry, a n d som e a p p e a r to have been b ro k e n by th e b ow er birds.

F inally, m an y o f the specim ens have relatively high stru c tu ra l densities (see
C h a p te r 7).
M an y species o f owls tend to ro o st in caves a n d rockshelters. T hey cast
(reg u rg itate) pellets w hich are co m posed o f m a tte d fur, hair, bones, teeth, and
o th e r undigested m aterial. O ne o r tw o pellets are cast p er day a n d w hen a ro o st
is occupied for long tim e periods, owls have the p o te n tia l to create m ajo r
accu m u latio n s o f b on es o f th eir prey (K u sm e r 1990). A ctualistic research on a
n u m b e r o f species o f owls indicates th a t som e a ttrib u te s o f such ow l-generated
b o n e accu m u latio n s m ay be d iag n o stic o f this bo n e accu m u lato r. F o r exam ple,
bo n e assem blages resu ltin g fro m ac cu m u latio n by owls ten d to have a b u n d a n t
bones o f all elem ents o f the skeleton, b u t m andibles a n d fem o ra m ay be the
m o st a b u n d a n t elem ents (D o d so n a n d W exlar 1979; K u sm er 1990). T hese
bones m ay re p resen t a lim ited n u m b e r o f the species p resen t in the area (e.g.,
only th e rem ain s o f sm all m am m als [no large m am m al skeletal parts] are
p resen t in ow l pellets), m an y b ones m ay be fro m im m a tu re prey anim als, a n d
m an y o f th e rep resen ted species will p ro b a b ly be n o c tu rn a l a n d /o r crep u scu lar
(K u sm e r 1990). B rain (1981) sum m arizes d a ta in d icatin g a t least som e species
o f A frican ra p to rs a n d owls exploit m a m m alian prey ra th e r o p p o rtu n istically ,
a n d th u s th e created fossil assem blage m ay be a relatively a c cu ra te reflection o f
th e sm all m am m al p o p u la tio n e x ta n t in the p re d a to r’s fo rag in g area. It seems
th a t the frequencies o f different skeletal elem ents will n o t allow the an a ly st to
d istin g u ish th e p a rtic u la r ra p to r species responsible fo r a p a rtic u la r bone
ac cu m u latio n (H o ffm an 1988; see below).
Owls ten d to d am ag e the b o n es they accum ulate. T h ere is lim ited digestive
c o rro sio n (K u sm e r 1990), alth o u g h bones ac cu m u lated by haw ks m ay show
m o re such co rro sio n th a n th o se ac cu m u lated by owls. Skulls o f prey ac cu m u
lated by owls in su m m er m o n th s m ay be m ore frag m en ted th a n those
ac cu m u lated in w in ter m o n th s (L ow e 1980). H aw ks a n d the screech owl (O tus
asio) seem to b re ak th e bones o f th eir prey m o re th a n m o st owl tax a th u s far
stu died (H o ffm an 1988). M o st scapulae a n d in n o m in a te s will be b ro k e n by
m an y owl species, a n d fem ora, radii, m andibles, a n d hum eri will tend to be
com p lete a n d n o t b ro k e n (D o d so n a n d W exlar 1979; K u sm er 1990). H offm an
(1988) defined nine categ ories o f fra g m e n ta tio n , reco rd in g to the n earest 25%
the p o rtio n o f a skeletal elem ent rep resen ted by a specim en: com plete, 25% o f
th e p ro x im al p o rtio n p resen t, 50% o f the p ro x im al p o rtio n present (proxim al
h a lf present), 75% o f th e p ro x im al p o rtio n (distal 25% m issing), 25% o f the
shaft p resen t, 50% o f the sh aft present, 25% o f the distal p o rtio n present, 50%
o f the distal p o rtio n p resen t (distal h a lf present), an d 75% o f the distal p o rtio n
(p ro x im al 25% m issing). H e found th a t haw ks an d the screech owl tend to
create m ore fragm ented b ones a n d m ore kinds o f frag m en ted bones th a n owls
in general.
H o c k ett (1989a, 1991) fo u n d th a t N o rth A m erican ra p to rs accu m u late
bones in o p en sites as well as caves a n d rockshelters, a n d ten d to accu m u late
m o re juv enile lep o rid bones th a n b ones o f a d u lt leporids. In n o m in ates are

freq u en tly dam ag ed , th e tibia is o ften b ro k e n in to a diaphysis cylinder
(especially fo r sm all leporids o r Sylvilagus spp.), the tran sv erse processes o f
v erteb rae a n d th e g re ater tro c h a n te r o f th e fem ur are o ften dam ag ed , and
fo relim b b o n es ten d to o u tn u m b e r h in d lim b bones in ra p to r-a c c u m u la te d
assem blages. T he skull is usually disassociated (dispersed) from the rest o f the
carcass, th e occipital is b ro k e n , a n d th ere are b eak a n d /o r talo n p u n ctu re s
beh in d the eye sockets. T he ascending ra m u s o f the m an d ib le o ften has
p u n ctu re s o r is b ro k en . F o relim b s an d hin d lim b s are often dispersed as
a rtic u la te d units, a n d lo n g b ones m ay h ave one o r m o re p u n ctu re s o n one side.
H u m a n a c cu m u lato rs o f leporid b ones ten d to create m o re tib ia diaphysis
cylinders fo r large lep o rids (L epus spp.), ac cu m u late m o re bones o f m atu re
leporids, a n d vario usly b u rn a n d b u tch er the carcasses (H o c k e tt 1991).
In the m o st intensive a n d extensive study to d ate. A ndrew s (1990) describes
in som e d etail the effects a n u m b e r o f avian p re d a to rs have on m am m alian
bones. H e also co m p ares th e m odifications m ad e to bones by these avian
p re d a to rs to th o se m ad e by som e m am m alian carn iv o res (sum m arized in the
follow ing su b section). A ndrew s (1990) indicates there are fo u r basic categories
o f evidence the an a ly st can exam ine in atte m p ts to identify the p re d a to r
resp o n sible fo r a b o n e ac cu m u latio n . T he presence o f pellets o r scats, som e o f
w hich can be identified to the tax o n th a t created an d dep o sited them , is one
categ ory o f evidence, b u t these seldom preserve (A ndrew s 1990:28). Second,
the size ran g e o f prey tax a m ay pro v id e a n in d icatio n o f the p re d a to r’s identity,
b u t th e size o f the p re d a to r a n d the size o f th e prey are n o t tightly co rrelated .
T h ird , th e tax a o f prey m ay also pro v id e som e hints as to the id en tity o f the
p re d a to r, b u t again these tw o variables do n o t seem to be tightly co rrelated as
m an y p re d a to rs are fa r to o o p p o rtu n istic in th eir fo rag in g h ab its (A ndrew s
1990:29). It is the fo u rth categ o ry o f evidence, m o d ificatio n o f bones, th a t
A n d rew s (1990) finds to be th e m o st a c cu ra te sig n atu re o f p a rtic u la r p re d a to ry
taxa. H e d istinguishes three kinds o f bo n e m odification - skeletal p a rt
frequencies o r b o n e loss, bo n e break ag e, a n d digestive c o rro sio n - a n d reviews
each in som e detail fo r a n u m b e r o f p re d a to rs. I sum m arize th a t discussion in
the follow ing p a ra g ra p h s.
Bone loss: N o tin g th a t owls, haw ks, a n d m am m alian carn iv o res often
d estro y som e bones o f th eir prey, A ndrew s (1990:45) suggests th a t calcu latio n
o f th e relative p ro p o rtio n s o f skeletal elem ents rep resen ted in an ac cu m u latio n
m ay p ro v id e clues to th e identify o f the b o n e a c cu m u lato r. T hese p ro p o rtio n s
are calcu lated w ith th e e q u a tio n
Ri = Ni/(M NI)Ei [6.3]
w here R, is the relative p ro p o rtio n o f skeletal elem ent i, Nj is the observed

frequency o f elem ent i in the assem blage, M N I is the m inim um n u m b e r o f
ind iv id uals, a n d E ; is the frequency o f skeletal elem ent i in one prey skeleton.
T his e q u a tio n acco u n ts for the fact th a t th ere are, fo r exam ple, m o re hum eri
th a n skulls in a single m am m al skeleton, a n d is m ath em atica lly identical to
eq u a tio n s used by o th ers fo r sim ilar p u rp o ses (e.g., D o d so n a n d W exlar 1979;
K o rth 1979; K usm er 1990; H o ffm an 1988; S hipm an a n d W alk er 1980).
T he p u rp o se o f e q u a tio n [6.3] is to p ro d u c e a frequency d istrib u tio n o f
different skeletal elem ents so th a t different assem blages m ay be co m p ared . F o r
exam ple, if it can be d em o n stra te d using actu alistic d a ta th a t haw ks dep o sit
m an y fem o ra an d few m andibles w hereas owls d ep o sit few fem ora a n d m any
m an d ib les, th en we will h av e a v alu ab le an aly tical tech n iq u e fo r identifying the
agent o f b o n e accu m u latio n . In fact, p revious researchers have suggested such
d istin ctio n s do n o t a p p e a r to be possible. H offm an (1988:85) c o m p ared the
frequency d istrib u tio n s o f skeletal elem ents across seven species o f ra p to r (4
owl a n d 3 haw k species) a n d co ncluded “ differential elem ent re p resen ta tio n as
m easu red by o rd in al ra n k in g is insufficient for d istinguishing clearly betw een
ra p to r species.” Sim ilarly, afte r stu d y o f th e frequencies o f skeletal p a rts
created by 10 species o f owls, tw o species o f haw ks, a n d seven species o f
m am m alian carn iv o re, A ndrew s (1990:49) co n cluded th a t “ at this level o f
analysis it is n o t possible to distinguish a d e q u ately betw een a d iu rn al ra p to r,
som e o f the owls, o r m am m alian c a rn iv o res.”
T he p ro b lem is readily in d icated using the Rj frequencies o f skeletal p arts
re p o rte d by A n d rew s (1990). T h o se frequencies fo r the b a rn owl (T y to alba)
a n d th e kestrel (Falco tinnunculus) are co rrelated {r = 0.572, P = 0.02), those
frequencies fo r th e b a rn owl a n d the co y o te (Canis latrans) are co rrelated
(r = 0.63, P = 0.007), a n d th o se frequencies fo r the kestrel a n d co y o te are
c o rrelated (r = 0.726, P = 0.03). T hese statistics in d icate there is perh ap s
insufficient v a ria tio n in skeletal p a rt frequencies gen erated by these bone-
a c cu m u latin g agents to allow them to be distinguished o n this basis alone.
(P o ten tial in ter-a n aly st v a ria tio n in how b ones w ere qu an tified a n d /o r in
id en tificatio n skill preclu des co m p ariso n of, say, b a rn owl skeletal p a rt
frequencies re p o rted by D o d so n a n d W exlar [1979], H o ffm an [1988], K usm er
[1990], a n d A ndrew s [1990].) T h u s as A ndrew s (1990:49) notes, “ a ra th e r m ore
d etailed fo rm o f analysis becom es n ecessary” if we w ish to identify a n d
d istin g u ish these k in d s o f b o n e accu m u lato rs. W e tu rn to th a t o th e r fo rm o f
analysis in later sections o f this ch ap ter.
A ndrew s (1990:45) suggests one co u ld also “ express the d istrib u tio n o f
skeletal elem ents ag a in st the to ta l n u m b e r o f bones in th e sam p le,” a lth o u g h
this results in an “ ex ag g eratio n o f the a b u n d a n ces o f such elem ents as ribs and
v e rte b ra e .” T he a d v a n ta g e to this la tte r tech n iq u e, A n d rew s (1990:45) sug
gests, is th a t it does n o t dep en d on the calcu latio n o f M N I, “ at best an
u nreliab le estim ate o f th e tru e n u m b ers o f individuals in a sam ple” (see also
G ra y so n 1984). A n drew s (1990:45—46) uses ch i2 to c o m p are skeletal p a rt
frequencies calcu lated this w ay fro m different assem blages. W hile this a n a ly ti
cal p ro c ed u re m ay be a re aso n ab le one, it m u st be em phasized th a t the analyst
needs to be clear w h eth er th e frequency d istrib u tio n o f “ skeletal elem ents” is

fo u n d ed on N IS P o r M N E values (see F ig u re 8.11 an d associated discussion).
B ecause th e average R; value fo r an assem blage (calculated fro m R; fo r all
included skeletal elem ents) “ is co m m only used to ch aracterize assem blages and
it is o ften suggested th a t d ep o sitio n al agents m ay be identified th ro u g h average
Ri v alu es” (K u sm e r 1990:630; e.g., K o rth 1979), if average R ; values are at least
p artia lly a fu n c tio n o f sam ple size o r N IS P (as K u sm e r’s 1990 research
indicates) as well as a m easu re o f differences in the tre a tm e n t o f b ones by
d ep o sitio n al agents, th en yet a n o th e r p ro b lem exists w ith using these values as
an aly tical tools for identifying agents o f bo n e ac cu m u latio n . This is so because
we as yet d o n o t k n o w w h a t p o rtio n o f th e v a ria tio n in average R; values can be
a ttrib u te d to sam ple size differences a n d w h a t p o rtio n o f th a t v a ria tio n can be
a ttrib u te d to differences in the id en tity o f the d ep o sitio n al agents.
W hile one o f th e m o st intensively a n d extensively d o cu m en te d variables
available, th e relative frequencies o f skeletal p a rts g en erated by vario u s ra p to rs
a n d m am m alian ca rn iv o res do n o t seem to allow the an aly st to identify clearly
a b o n e -acc u m u la tin g ag ent. T his m ay also be tru e o f large m am m al bones
ac cu m u lated by large m am m alian p re d a to rs, including hom inids (see C h a p te r
7). It is in p a rt for this reaso n th a t m o st an aly sts have tu rn e d to o th e r kinds o f
d a ta in atte m p ts to identify b o n e-acc u m u la tin g agents.
O ne k in d o f d a ta involves calcu latin g th e ra tio o f c ran ial to p o st-cran ial
rem ains, a n d a n o th e r involves calcu latin g th e ra tio o f p ro x im al lim b elem ents
to distal lim b elem ents. A ndrew s (1990:49) calculates the ra tio o f cran ial to
p o st-cran ial rem ain s tw o ways. O ne tech n iq u e involves calcu latin g the ra tio o f
the sum o f the frequencies o f fem ora, tibiae, hum eri, radii, an d ulnae, a n d the
sum o f the frequencies o f m andibles (lefts a n d rights co u n ted separately),
m axillae (lefts a n d rights co u n ted separately), a n d isolated m olars, m ultiplied
by 100 to derive a p ercentage. B ecause in a single skeleton the ra tio o f these tw o
categ o ries o f skeletal p a rts is 10 p o st-cran ial to 16 cran ial p a rts, A ndrew s
m ultiples the observed ra tio by 5/8 to w eight the ra tio according to the
frequency o f these b on es in one skeleton. T he o th e r tech n iq u e o f calcu latin g a
ra tio o f p o st-cran ial skeletal p a rts to cran ial skeletal p a rts described by
A n drew s (1990:49) involves calcu latin g the ra tio o f the sum o f the frequencies
o f fem o ra a n d h um eri, an d the sum o f the frequencies o f m andibles an d
m axillae (fo r b o th , lefts a n d rights are tallied separately), a n d m ultiplying th a t
ra tio by 100 to p ro d u c e a percentage. B ecause the ra tio o f these skeletal p a rts is
4:4 in a single skeleton, no w eighting o f the ra tio is necessary. A g ra p h o f the
first described ra tio o f p o st-cran ial to c ran ial skeletal p a rts using A n d rew s’
(1990:49) d a ta suggests ow ls tend to p ro d u c e the highest ratio s, m am m alian
carn iv o res p ro d u ce m id-level ra tio s, an d haw ks p ro d u c e low ra tio s (F igure
6.16).
A n d rew s (1990:49-50) also suggests calcu latin g the ra tio o f m ajo r distal lim b
p a rts (tib iae plus radii) to m a jo r p ro x im al lim b p a rts (fem ora plus hum eri)
snowy owl
long-eared owl
tawny owl
barn owl
short-eared owl
E. eagle owl
great grey owl
coyote
little owl
red fox
bat-eared fox
V. eagle owl
S. eagle owl
genet
mongoose
pine marten
kestrel
arctic fox
hen harrier
0 100 200 300
Ratio of postcranial to cranial skeletal parts (X100)
Figure 6.16. R a tio o f po stcran ial to cranial skeletal p a rts accum ulated and
deposited by 19 species o f rap to rs and m am m als; see text for discussion (from
A ndrew s 1990). Snowy owl (N yctea scandiaca); long-eared owl (Asio otus); taw ny
owl (S trix aluco); b arn owl ( T yto alba)', short-eared owl (A sio flam m eus); E. eagle
owl = E u ro p ean eagle owl (Bubo bubo)\ g reat grey owl (S tr ix nebulosa); coyote
(Canis latrans); little owl (A thene noctua); red fox ( Vulpes vulpes); b at-eared fox
(Otocyon megalotis); V. eagle owl = V erreaux eagle owl (Bubo lacteus)\ S. eagle
owl = spotted eagle owl (Bubo africanus); genet = sm all-spotted genet (G enetta
genetta); m ongoose = w hite-tailed m ongoose (Ichneum ia albicauda); pine m arten
(M aries martes)', kestrel (Falco tinnunculus); arctic fox (A lo p ex lagopus)', hen
harrier (Circus cyaneus).
m u ltiplied by 100. A g ra p h o f these d a ta fo r 19 tax a o f bo n e ac cu m u lato rs

indicates owls d ep o sit a b o u t equal n u m b ers o f p ro x im al a n d distal p a rts o f
lim bs (F ig u re 6.17). H a w k s ten d to d ep o sit few er distal lim b p a rts th a n owls,
an d m am m alian carn iv o res dep o sit the fewest distal lim b p a rts relative to
p ro x im al lim b p arts.
Bone breakage: I have alre ad y m en tio n ed H o ffm an ’s (1988) tech n iq u e o f

tallying the p ro p o rtio n o f a bo n e fragm ent rep resen ted by a specim en, and
using th a t d a ta to calcu late the n u m b e r o f kinds o f frag m en ts a n d the diversity
(frequency o f re p resen ta tio n o f each) o f the kinds o f fragm ents. H e concludes
th a t h aw ks p ro d u c e m ore b ro k e n bones th a n m o st ow ls (H o ffm an 1988:87).
A n d rew s (1990:52) fo u n d sim ilar results, plus he n o ted th a t sm all m am m alian
barn owl
V. eagle owl
snowy owl
long-eared owl
tawny owl
great grey owl
short-eared owl
coyote
E. eagle owl
arctic fox
kestrel
little owl
hen harrier
S. eagle owl
red fox
genet
mongoose
bat-eared fox
pine marten
0 20 40 60 80 100
Ratio of distal to proximal limb elements (X100)
Figure 6.17. R atio o f distal to proxim al lim b elem ents accu m u lated and
deposited by 19 species o f ra p to rs an d m am m als; see text fo r discussion (after
A ndrew s 1990). Bone accum ulating tax a as in Figure 6.16.
carn iv o res p ro d u c e m o re b ro k e n bones (few er com plete bones) th a n haw ks,

an d som e owls p ro d u c e very few b ro k e n bones.
I calcu lated the p ro p o rtio n o f the M N E (m inim um n u m b e r o f elem ents) th a t
w ere com plete from vario u s assem blages re p o rted by A ndrew s a n d E vans
(1983). D o d so n a n d W exlar (1979), H o ffm an (1988), a n d K u sm er (1986). O nly
d a ta for th e h u m eru s, ulna, fem ur, an d tibia w ere com piled. As g rap h ed in
F ig u re 6.18, these d a ta c o rro b o ra te the o b serv atio n s o f others. M o st owls
b re ak few er th a n 25 % o f th e included skeletal elem ents. M a m m a lia n c a rn i
vores ten d to b re a k 50% o r m o re o f th e included lim b bones, a n d haw ks
essentially b re ak all o f th e in cluded skeletal elem ents (see C h a p te r 8 fo r fu rth e r
discussion o f the q u an tifica tio n o f frag m en ted bones).
F ra g m e n ta tio n o f cran ial elem ents show s sim ilar p a tte rn s. M o st owls
stu d ied by A ndrew s (1990:53-64) tend to leave skulls w ith m axillae still
attac h ed w hereas h aw k s a n d carn iv o res p ro d u c e m o re iso lated m axillae. Owl
pellets c o n ta in m o re incisors (b o th u p p e r a n d low er) th a n haw k pellets and
m am m alian carn iv o re scats. M o st owls b re a k few m an d ib les a n d virtu ally no
m o lars o r incisors w hereas h aw k s a n d m a m m alian carn iv o res b re a k m an y
m and ib les. H aw k s b re a k a few m o lars a n d incisors, b u t m a m m alian carn iv o res
seem to p ro d u ce th e m ost b ro k e n m o lars a n d incisors. T h u s, if the an a ly st can
GH owl 1
GH owl J
barn owl
short-eared owl
barn owl
GH owl
barn owl
barn owl
barred owl
GH owl
genet
screech owl
mongoose
bat-eared fox
screech owl
coyote
pine marten
sparrow hawk
RL hawk
red-tailed hawk
red fox
0 20 40 60 80 100
% complete limb elements
F igure 6.18. P ro p o rtio n o f com plete lim b elem ents (hum erus, ulna, fem ur, tibia)
in assem blages accum ulated by selected rap to rs and m am m alian carnivores. GFI
owl = great horned owl (Bubo virginianus); b a rn owl ( Tyto alba); sh ort-eared owl
(Asio flam m eus); b arred owl (S tr ix varia); screech owl (O tus asio); sp arro w haw k
(Falco sparverius)\ R L haw k = rough-legged haw k (Buteo lagopus); red-tailed
haw k (Buteo jamaicensis)', m am m als as in F igure 6.16.
m easure th e degree to w hich th e specim ens m ak in g u p a fossil collection are

frag m en ted , th ey will be closer to identifying the ag en t o f b o n e accu m u latio n .
Digestive corrosion: A ndrew s (1990:32) fo u n d th a t digestive co rro sio n was

greatest in m a m m alian ca rn iv o re accu m u latio n s; w hile such co rro sio n w as
present in ra p to r accu m u latio n s, it w as less p ro n o u n c ed . H e suggested th a t this
was th e case because th e fo rm er tax a digest food in the sto m ach a n d intestine
w hereas ra p to rs only digest in the stom ach, th u s food in m am m alian carnivores
is subjected to lo n g er digestion tim es. A ndrew s (1990:32) fo u n d th a t to o th
enam el m o st readily displayed the effects o f digestive co rro sio n , p ro b a b ly
because it is the m ost highly m ineralized skeletal tissue (C h a p te r 4); d en tin e and
bo n e also displayed digestive co rro sio n b u t less extensively th a n enam el (see
also F ish er 1981).
M o re m o lars w ere digested by m am m alian carn iv o res a n d haw ks th a n by
owls in A n d rew s’ (1990:65) sam ples. D igestion seem ed also to be m ore
intensive a n d extensive for the fo rm er th a n fo r the la tte r p re d a to rs. T he
occlusal co rners o f salient angles p ro g ress fro m slightly ro u n d e d , to flattened,
to com plete loss o f enam el a n d exposure o f dentine, as one shifts fro m owls to
d iu rn al ra p to rs (haw ks) to m a m m alian ca rn iv o res (A ndrew s 1990:67). Skeletal
elem ents o f th e lim b w ere also m o re intensively a n d extensively c o rro d e d fro m
digestion by haw ks a n d m am m alian ca rn iv o res a n d less c o rro d e d by owls
(A ndrew s 1990:79; F ern a n d ez-Ja lv o a n d A ndrew s 1992:410).
R en sb erg er a n d K re n tz (1988) re p o rt th a t c o rro sio n features on ro d en t
b ones a n d teeth c reated by th e digestive ac tio n o f c o y o te ( Canis latrans) an d
great h o rn e d owl (Bubo virginianus) tend to be sim ilar. M an y featu res visible
u n d er scan n in g electron m icroscopy w ere n o t evident u n d er sta n d a rd light
m icroscopy. These include deep so lu tio n fissures w ith ro u n d e d edges a n d pits.
T he fissures a n d pits seem to o rig in ate in canals b en e ath the bone surface.
F ern a n d ez-Ja lv o a n d A ndrew s (1992) pro v id e a useful synopsis o f digestive
c o rro sio n stages displayed by ro d e n t cheek teeth an d incisors first re p o rted by
A nd rew s (1990). L ig h t digestion is restricted to the occlusal co rn ers o f the
salient angles o f cheek teeth a n d does n o t p en e trate below the alv eo lar m argin;
the en tire enam el surface o f incisors is slightly p itte d a n d the distal tip o f the
enam el surface m ay be com pletely rem oved, in d icatin g digestion w hile the
incisor w as still in place in the jaw . M oderate digestion o f cheek teeth is signified
by the rem o v al o f en am el alo n g th e entire edge o f th e salient angle, p e rh a p s w ith
som e pittin g; the en tire enam el surface o f incisors is m o re com pletely affected
an d th e d en tin e has a w avy surface. C o rn ers a n d salient angles o f cheek teeth
are heavily ro u n d e d a n d m uch enam el is rem oved an d den tin e exposed in heavy
digestion o f cheek teeth; incisors have isolated islands o f enam el a n d w avy
d en tin e surfaces. E xtrem e digestion results in cheek teeth th a t are rarely
identifiable an d w ith qu ite d am ag e d dentine; incisors have very little o r no
enam el rem ain in g a n d d en tin e is quite d am aged.
M a m m alian carnivores as bone accum ulators

T h ere has p ro b a b ly been m o re research a n d m o re w ritten o n this categ o ry o f
b o n e-acc u m u la tin g ag ents th a n an y o th er, no d o u b t because m an y o f its
m em b ers occupy caves an d ro ck sh elters ju s t as h o m in id s do. T he z o o a rch a eo -
logist a n d ta p h o n o m ist is th u s regularly faced w ith d eterm in in g if the bones in
such a site are p resen t there d u e to the ac tio n o f h u m a n s o r the ac tio n o f
carn iv o res. M ost actualistic studies have focused on relatively large carnivores.
A n d rew s a n d E vans (1983), how ever, pro v id e an im p o rta n t stu d y o n the bone
collecting activities o f relatively sm all m a m m alian carnivores. T hey exam ine a
n u m b er o f a ttrib u te s o f bo n e assem blages accu m u lated by sm all carnivores,
som e o f w hich are c o m p ared ab o v e w ith bo n e assem blages ac cu m u lated by
owls a n d haw ks. W hile in this section we focus on large m am m alian p re d a to rs
an d b o n e accu m u lato rs, it is im p o rta n t to sum m arize o b serv atio n s m ade o n the
m o dification s to bones created by sm all, dog-sized o r fox-sized carnivores.
S tallib rass (1984, 1990:153-155) lists several a ttrib u te s o f bone m odification
p ro d u c ed by fox-sized canids. B ones in scats are highly fragm ented, even m ore
so th a n is fo u n d in som e owl pellets. F rag m en ts rarely exceed 1 cm in m axim um

dim ension, an d b on es o f dom estic sheep (Ovis aries)-sized prey are sm all and
unidentifiable. Iso lated teeth are co m m o n a n d m ost m andibles o f sm all
m am m als are b ro k en . Som e bones have p u n c tu re m ark s fro m gnaw ing an d
som e display digestive c o rro sio n . D iaphysis frag m en ts are splinters ra th e r th a n
tubes. As we will see, these are ra th e r sim ilar to the m odification attrib u te s
d o cu m en te d for larger p re d ato rs.
T he A frican leo p ard (P anthera pardus) carries prey carcasses into trees for
u n in te rru p te d c o n su m p tio n (B rain 1980, 1981; C avallo a n d B lum enschine
1989). B ones fall o u t o f the tree as the leo p ard consum es the carcass. If those
bones re m a in a ro u n d the base o f th e tree, a sm all co n c e n tra tio n o f b ones m ay
build u p. O n e o f th e m o st in terestin g a ttrib u te s o f the skeletal rem ains (o th er
th a n gn aw ing d am age) is the d estru c tio n o f the b ra in case an d p o stero -d o rsal
p o rtio n o f th e eye sockets (B rain 1981:93).
In th e O ld W o rld , th e m o st stu d ied large m am m alian ca rn iv o re is the hyena.
T h ere are three e x tan t species o f hyena: the bro w n hyena (H yaena brunnea), the
strip ed h y en a (H ya ena hyaena), a n d the sp o tted h y en a (C rocuta crocuta). T he
bo ne-co llectin g h ab its a n d b o n e-m o d ify in g beh av io rs a n d results o f those
b eh av io rs fo r each species have been studied in som e detail (B rain 1981;
H o rw itz a n d Sm ith 1988; K e rb is-P e terh a n s an d H o rw itz 1992; L am 1992;
R ich a rd so n et al. 1986; S kinner a n d v an A a rd e 1991; a n d references therein).
T he sp o tte d h y en a h as been ch a rac te rized as “ th e m o st effective e x tan t bone-
crack in g c a rn iv o re ” (M a rean a n d Spencer 1991:648). In the N ew W o rld , m uch
research h as been d o n e o n w olves (Canis latrans), coyotes (Canis latrans), an d
do m estic dogs ( Canis fa m ilia ris) (B inford 1981b; B urgett 1990; H aynes 1980,
1983a; K e n t 1981; K lippel et al. 1987). A n ecd o tal d a ta have also been collected
fo r th e A frican lion (P anthera leo) a n d N o rth A m erican bears ( Ursus spp.)
(H ay n es 1983a). M an y o f the a ttrib u te s o f b o n e m odification p ro d u c ed by
these large m a m m alian ca rn iv o res can be sum m arized as follow s.
Ragged-edged chewing (M ag u ire et al. 1980:79-80) is typically seen o n thick
bo ne such as the en ds o f lim b bo n e shafts (F ig u re 6.19). A lso k n o w n as
crenulated edges (B in fo rd 1981 b :5 1), this type o f d am ag e m ay occu r on very
th in b o n e w hen the to o th p en e trates an d rem oves p a rt o f the edge o f the bone.
Shallow p ittin g (M ag u ire et al. 1980:79-80) o r p ittin g (B inford 1981b:44-48)
often has a restricted d istrib u tio n a n d is p ro d u c ed w hen the bo n e is sufficiently
s tro n g o r dense to w ith sta n d th e p ressures o f teeth a n d n o t be p u n ctu re d
(F ig u re 6.20). P unctures (B inford 1981b:44-48), p u n cta te depressions, or
perforations (M ag u ire et al. 1980:79-80) result w hen the b o n e collapses u n d er
the p ressu res o f teeth , leaving a clear, m o re o r less oval d epression in the bone,
o ften w ith flakes o f th e o u te r w all o f the b o n e pressed in to th e p u n c tu re (F igure
6.21). S hip m an (e.g., 1981 a:366) suggests p u n ctu re s are p ro d u ced by the
p ressu re o f a single to o th cusp o r canine ap p lied at a n angle ap p ro x im ately
p erp en d icu lar to th e bo n e surface. P u n ctu res decrease in d iam eter as d ep th
fro m th e b o n e surface increases.
Figure 6.19. a, ragged an d cren u lated p roxim al end o f a m o d ern w apiti (Cervus
elaphus) hum erus shaft resulting from gnaw ing by carnivores, p ro b ab ly coyote
(■Canis latrans), collected from W ash in g to n state; b, detail o f the cren u lated edge,
note polishing o f points o f crenulations an d furrow s.
a
Figure 6.20. P itting and punctures, a, left, deer (Odocoileus sp.) second p halanx
collected from a rockshelter in M issouri show ing a large p u n ctu re (pitting is
evident on the reverse side) p ro b ab ly resulting from m ultiple bites, right,
com plete ungnaw ed com p arativ e specimen; b, right, an intensively chew ed deer
distal hum erus show ing pittin g , collected from a rockshelter in M issouri, left, an
ungnaw ed com parativ e specimen.
F igure 6.21. Punctures o n (a) a patella an d a v entral rib o f w apiti (Cervus
elaphus), collected from W ash in g to n state (from L ym an 1989b: 153, Figure 7;
courtesy o f T he C en ter for the Study o f the F irst A m ericans), and (b) a proxim al
fem ur o f deer (Odocoileus sp.) collected from a rockshelter in M issouri.
Figure 6.22. F u rro w o n a m odern w apiti (Cervus elaphus) proxim al fem ur

collected from W ashington state (from L ym an 1989b: 153, Figure 7; courtesy o f
T he C enter for the Study o f the F irst A m ericans).
Striations, gouge m a rks (M ag u ire et al. 1980:79-80), o r scoring (B inford

1981b:44-48) are usually sh o rt, p arallel, a n d linear o r stra ig h t m ark s th a t are
ro ug h ly p e rp en d icu lar o r tran sv erse to the long axis o f the bone. T hey m ay be
quite close to g eth er, are usually o n the sh afts o f long lim b bones, a n d tend to
follow the surface o f the bone. T hey ap p a re n tly result from d ragging the teeth
across th e surface o f the bone. S hip m an (e.g., 1981 a:365) labels such m ark s
tooth scratches a n d ch aracterizes them as “ elongate grooves th a t m ay vary
from V -shaped to U -sh ap ed in cross-section [and] the b o tto m o f the groove is
s m o o th .” T o o th scratch es m ay o ccu r singly, as sets o f parallel o r subparallel
m ark s, o r as clusters w ith different o rien ta tio n s. V arious co n tig u o u s o r close,
irre g u la r a n d ra n d o m ly -o rien ted grooves (M ag u ire et al. 1980:79-80) o r
furrow s ( B inford 198 lb :4 4 -4 8 ; H ay n es 1980, 1983a) are generally fo u n d on the
ends o f long lim b b on es w here b o n e tissue is cancellous (F ig u re 6.22). Scooping
out o r hollowing out (B in ford 198 lb :4 4 -4 8 ; H aynes 1980, 1983a; M ag u ire et al.
1980:79-80) is the resu lt o f extrem e furro w in g , a n d involves the rem oval o f
significant p o rtio n s o f the cancellous bo n e tissue fro m epiphyseal ends o f lim b
b ones (F ig u re 6.23). S coo ping o u t results in the p ro d u c tio n o f large, irreg u lar
holes in lo n g bon e ends.
B oth hyenas a n d N o rth A m erican canids p ro d u c e acid-etching a n d corrosion
o f bon e (F ig u re 6.24). Such b ones m u st have passed th ro u g h the digestive tra c t
Figure 6.23. Scooping o u t on tw o distal fem ora o f m od ern w apiti (Cervus
elaphus) collected from W ashington state.
Figure 6.24. Digestive co rro sio n o f first phalanges o f dom estic sheep (Ovis aries)
collected from M issouri. Left specim en is a w hole com p arativ e specimen; note
pitting on center specim en an d feathering o f fractu re edges o n right specimen.
o r been re g u rg ita ted a fte r som e tim e in the stom ach. B oth g ro u p s o f carn iv o res
also splinter and crack bones (see also C h a p te r 8). T his fractu rin g pro d u ces
lunate o r crescent-shaped fra c tu re scars, th o u g h t o f as “ h a lf o f a p u n c ta te bite
m ark w hen the stren g th o f the bite has been sufficient to split the b o n e so th a t
the fra c tu re line passes th ro u g h th e p u n c ta te p e rfo ra tio n ” by M ag u ire et al.
(1980:79-80). T here is som etim es a n associated b o n e flake p a rtia lly detach ed
from the concave surface (m edullary cavity side) o f the b o n e at the p o in t o f
to o th co n tac t. W h en m am m alian ca rn iv o res gnaw b ones they “ a tta c k the ends
o f long b o n es first” (B inford 1981 b :51). C hanneled bones are p ro d u c ed by
“ p u n ctu rin g th e b o n e b ack fro m the tran sv erse edge, leaving a ch a n n el ru n n in g
parallel to the lo n g itu d in al axis o f the b o n e ” (B inford 1981 b: 51). Chipping back
results fro m chew ing th e edge o f a b ro k e n long bone; the b o n e edge is
co n tin u o u sly ch ip p ed an d to o th scoring on the external surface o f the bo n e is
frequently associated. L icking ch ip p ed edges o r ends can p ro d u ce ro u n d e d and
p o lished edges a n d ends th a t have the a p p e a ra n c e o f “ use w e a r” (B inford
1981b; H ay n es 1980, 1983a). O ften, to o th d am ag e is b ip o la r due to the vise-like
ac tio n o f the u p p e r a n d low er to o th row s (e.g., F ig u re 6.20). L o n g -b o n e
d iaphyses th a t have h ad the ends rem oved can be collapsed by gnaw ing
carn iv o res, p ro d u c in g m any diaphysis splinters. In carn iv o re-g n aw ed assem
blages, “ evidence o f gnaw ing, pressure-flaked edges, incised scarrin g o n the
o u tsid e o f th e flake, p ittin g a n d ab ra sio n fro m re p eated vise-like m ash in g o f a
b o n e surface, a n d so o n will occu r on large splinters ( > 4 cm long); sm all
sp linters will exhibit no such m o dification. T he latter m ay show signs o f having
been co rro d ed by sto m ach acids an d be em bedded in feces” (B inford
1981 b:60).
T he m ere presence o f one o r m o re o f these to o th -m a rk in g a ttrib u te s w ould
sim ply in d icate th a t a large m a m m alian c a rn iv o re h a d access to the bones, an d
n o t d irectly in d icate th e ca rn iv o res ac cu m u lated the bones. T hese a ttrib u te s
m ay allow m o re precise id entification o f the ca rn iv o re ta x o n responsible if the
bones are n o t to o extensively gnaw ed acco rd in g to H aynes (1983a) (see T able
5.6 fo r a schem e o f stages o f ca rn iv o re d e stru c tio n o f bones). H is crite ria for
id en tify in g th e general tax o n o m ic fam ily o f gnaw ing c a rn iv o re are rep ro d u ce d
in T ab le 6.10. H aynes (1980, 1983a) a n d K en t (1981) c a u tio n th a t m am m alian
carn iv o res ca n gnaw a n d m ove b o n es a n d yet leave no traces o f to o th m ark s on
the bones.
O th er a u th o rs have arg u ed th a t the size o f a to o th p u n c tu re o r lu n ate flake
scar p ro v id es a g o o d in d icatio n o f the id en tity o f the c a rn iv o re tax o n . F o r
exam ple, M o rla n (1983:256) suggests th a t the lo ad in g p o in ts o f teeth on bones
“ will h ave d iam eters slightly larg er th a n th e c o n ta c t areas o f the teeth, a n d for
m o st carn iv o res such d iam eters will be relatively sm all, no m ore th a n a few
m illim eters. L arg er notches resem bling lo ad in g p o in ts m ay be preserved on a
carn iv o re-in d u ced fractu re, b u t they are co m p o sites o f m an y sm aller flake scars
an d do n o t represen t discrete p o in t lo a d in g .” T h a t is, m ultiple bites m ight be
d etected as m u ltip le discrete lo ad in g p o in ts. B ut w here d o we m easu re the
T ab le 6.10 G nawing dam age to bones typical o f fo u r taxonom ic groups o f

m am m alian carnivores (fro m H aynes 1983a)
D am age type C anids H yenids U rsids Felids
T o o th m arkin g on co m p act bone: 3 5 2-1 1-0

5 = m ost expected, 1 = least expected
G rinding off prom inences vs. biting through: 2 2-3 5 1
5 = m ostly grinding (leaves sm ooth stum ps);
1 = m ostly biting th ro u g h (leaves a rough stum p,
usually an irregular rim o f com p act bone)
T o o th im pression shape in trab ecu lar bone: 3 3 5 1
5 = square o r rectangular;
3 = cone o r tru n cated cone;
1 = “ axe-edge” o r elongated V -shape
“ c o n ta c t a re a ” o r d iam eter o f a to o th ? T his is a sticky p ro b lem because teeth

can p en e trate bo n y m aterial, th ereb y increasing the cross-sectional area o f
co n tac t betw een th e to o th an d the bone. T his is especially so for tra b e c u la r o r
cancellous bone w hich is easily a n d readily p e n e tra te d by teeth. A nd, w hat
a b o u t assem blages gnaw ed by m o re th a n one ta x o n o f carnivore? These
difficulties can be illu stra te d w ith an exam ple.
W hile an alyzin g a co llection o f sm all m am m al rem ains (a p p ro x im a tely 7,000
specim ens identified to genus o r species) recovered from a ro ck sh elter in the
state o f W ash in g to n , I m easured the m ax im u m a n d m inim um d iam eter o f
p u n ctu res. M easu rem en ts w ere tak en at the surface o f the bone. B ecause the
assem blage is late H o locene in age, I m easured the d iam eters o f a set o f m o d ern
m am m alian ca rn iv o re canines ap p ro x im a te ly 2 -3 m m from the distal tip o f the
canine. T h e 2 -3 m m value w as chosen because m o st o f the p u n ctu re s were
ab o u t th a t deep. P lo ttin g the frequencies o f to o th p u n c tu re d iam eters against
the ran g e o f m easured canine diam eters p ro d u ces an am b ig u o u s in d icatio n o f
the tax o n o m ic id en tity o f the b o n e-gnaw ing c a rn iv o re d espite the fact th a t the
m easu red carn iv o res are the m a jo r ones historically re co rd ed in the site area
an d rem ain s o f each o f them w ere fo u n d in the ro ck sh elter (F ig u re 6.25). I
suspect th a t these results w ere derived because the gnaw ing m a rk s were
p ro d u c ed by several ca rn iv o re ta x a w ith different-sized canine teeth.
B lum enschine a n d Selvaggio (1988, 1991) inspected b o th pits (F ig u re 6.20)
a n d h am m ersto n e -g en e rated m ark s; they term the la tte r percussion m arks.
T hey fo u n d th a t th e tw o can be d istinguished u n d e r low (10 x ) m agnification.
P ercussion m ark s alw ays have dense p atch es o f in icro striatio n s in o r e m a n a t
ing fro m the pit defining the m ark . T o o th -g en e rate d pits rarely have a few, less
densely o ccu rrin g stria tio n s th a t generally require m ore pow erful m agnifica
tio n to be visible. U n lik e to o th m ark s, percu ssio n m ark s should occu r in
ap p ro x im a te ly th e sam e lo catio n on specim en after specim en o f the sam e
skeletal elem ent.
H ay n es (1983a) p rovid es a descriptive guide for d istinguishing the gnaw ing
Range of comparative
canine diameters
— Yulpes vulpes
Lynx sp. —
— Taxidea taxus
— Nustela visorr
F — Mustela frertata
R
E 5 -
Q
U 3 -
E
N 1 -
C r — r
Y 1.7 2.3 2.9 3.5 4.1
Class midpoints for average diameter
of puncture marks (m m)
Figure 6.25. C om p ariso n o f diam eters o f p u n ctu re m ark s on small m am m al

bones collected from a rockshelter in W ash in g to n state, an d the range o f canine
diam eters o f m o dern carnivores.
d am ag e d o n e to u n g u late b ones by u rsids, can id s, a n d felids th a t is b ased on the

a p p e a ra n c e o f th e gnaw in g d am age. R ic h a rd so n (1980; R ich a rd so n et al. 1986)
p resen ts d a ta (sum m arized in F ig u re 6.26) in d icatin g hyenas d am ag e and
d estro y m o re b on es o f prey th a n lions (Panthera leo), leo p ard s (Panthera
pardus), feral dogs (Canis fa m ilia ris), a n d b lack -b ack e d ja c k a ls (Canis m esom e-
las). A s well, hy enas ten d to frag m en t b ones m o re o ften th a n these o th er
carnivo res, as in d icated by the N IS P :M N E ra tio s in F igure 6.26. Such
c o m p a ra tiv e analyses are useful fo r establishing ca rn iv o re taxon-specific
a ttrib u te s o f bo n e m o d ificatio n b u t have, to d ate, been rarely published.
P erh ap s th a t is because we are still far from d o cu m en tin g the full ran g e o f
v ariab ility in m odifications to an im al carcasses a n d prey bones th a t ca n be
p ro d u c ed by p a rtic u la r ca rn iv o re tax a (e.g., L am 1992).
W h a t is beco m in g a b u n d a n tly clear as research progresses is th a t it will often
be n ecessary to stu d y m ultiple a ttrib u te s o f the fossil assem blage in o rd e r to
id entify agents o f b o n e ac cu m u latio n (e.g., N o e -N y g a a rd 1989). T his is quite
clear in a recen t synopsis o f a ttrib u te s th a t m ig h t be used to d etect the
ac cu m u latio n o f b o nes by hyenas. C ru z-U rib e (1991; see also K lein 1975)
describes a ttrib u te s she believes to be distinctive o f a b o n e assem blage created
by h yenas, a n d n o tes h ow these a ttrib u te s differ fro m a bo n e assem blage
ac cu m u lated by hom in id s. It is im p o rta n t to n o te th a t all o f h er collections are
Prey Size NISP:MNE
jackal 100— 300
jackal 10— 100

m m
feral dog 10— 100
leopard 10— 100
lion 100— >300

i
spotted hyena W77777X >300
spotted hyena 100— 300

C l %undamaged
brown hyena □ %survival 10— 100
20 40 60 80 1 00
Proportion
Figure 6.26. A ttrib u tes o f m odification to prey bones created by various A frican
carnivores. Prey size is live w eight o f anim al in kg; N IS P :M N E ratio is
R ich ard so n ’s (1980) “ frag m en tatio n ra tio ” divided by 100 (after R ich ard so n
1980:114, Figure 4; an d R ich ard so n et al. 1986:29, F igure 3).
p re h isto ric ones; th a t is, she h as n o ex perim ental co n tro ls over the fossil
assem blages she co m p ares because the ac tu a l b o n e-acc u m u la tin g agents are
n o t kn o w n . M an y o f th e a ttrib u te s she lists do, how ever, tend to align ra th e r
well w ith a ttrib u te s d o cu m en ted in actu alistic contexts. In h er collections,
C ru z-U rib e finds th a t ca rn iv o res m ake up a t least 20% o f the rem ains (as
m easu red by M N I) ac cu m u lated by hyenas w hereas ca rn iv o res m ake up less
th a n 13% o f the rem ain s ac cu m u lated by hom inids. S econd, m any b u t n o t all
bones ac cu m u lated by hyenas display striatio n s, p ittin g , grooves, scooping out
o f can cellou s bon e, a n d digestive co rro sio n . T h ird , bo n e assem blages ac cu m u
lated by h yenas c o n ta in m an y long bo n e cylinders; th a t is, the m ore o r less
co m plete diaph y sis will be present b u t will lack the epiphyses. H om inid-
g en erated accu m u latio n s ten d to have m o re b ro k e n diaphyses a n d in tact
epiphyses. F o u rth , m o rtality profiles o f prey exploited by hyenas (see C h a p te r
5) ten d to be a ttritio n a l w hereas m o rta lity profiles o f prey exploited by
h o m in id s m ay be eith er a ttritio n a l o r c a ta stro p h ic . F ifth , in hyena a c cu m u
lations, sm all u n g u lates tend to be b e tte r rep resen ted by cran ial bones a n d large
ung u lates by p o st-cran ial b ones (as m easu red by M N I). In hom in id ac cu m u
latio n s, th ere is no clear re la tio n betw een frequencies o f cran ial a n d post-
cran ial bones, a n d prey size. F inally, sm all, h a rd , stru ctu ra lly dense (see
C h a p te r 7) b ones such as tarsals, carp als, sesam oids, a n d ph alan g es tend to be
a b u n d a n t in w ell-preserved hom in id ac cu m u latio n s, b u t are ra re in hyena-
ac cu m u lated assem blages because hyenas sw allow a n d digestively destroy
these elem ents.
T he crite ria o u tlin ed by C ru z-U rib e (1991) are sim ilar to ones o u tlin ed by
S tiner (1990a, 1991b, 1991e) a n d sum m arized in C h a p te r 7. B o th a u th o rs were
in terested in d istin g u ish in g carn iv o re-cre ated b o n e accu m u latio n s fro m those
created by h om inids. T his brings us, th en , to th e last m a jo r bo n e ac c u m u la to r
zo o a rch a eo lo g ists a n d arch aeo lo g ical ta p h o n o m ists ten d to consider. I have
fo u n d this to be a con v en ien t place to m ak e this to p ical shift, b u t the perceptive
re ad er will n o te som e o verlap betw een th e im m ediately preceding discussion
a n d w h a t follows.
H om inids as bone accum ulators

T he p ro b lem o f recognizing h o m in id s as b o n e-acc u m u la tin g agents is one o f
an alytically so rtin g o u t w h at T h o m as (1971) calls cultural bone from natural
bone (L y m an 1982a). T he fo rm e r ca te g o ry includes specim ens dep o sited as a
result o f h u m a n activity; the la tte r ca te g o ry includes specim ens (accu m u lated
an d ) d ep o sited as a result o f n a tu ra l processes. W hen th e an a ly st is in terested in
h u m a n subsistence practices, a logical second step in the d istin ctio n process is
th e so rtin g o f cu ltu ra lly d ep o sited an im al rem ains into th o se rep resen tin g food
w aste a n d th ose rep resen tin g n o n -fo o d rem ains. A b u ried pet o r beast o f
b u rd e n rep resen ts cu ltu ra lly d ep o sited b o n e b u t m ay n o t have been eaten.
Several kind s o f crite ria have been used to distinguish cu ltu ra l a n d n a tu ra l
bone, w ith the im p licatio n generally being th a t these sam e criteria also help to
d istin g u ish cu ltu rally d ep o sited fo o d bones from c u ltu rally d ep o sited n o n -fo o d
bones. I briefly review each o f these criteria in th e follow ing.
Burning or charring: B ecause m an y h u m an s c o o k m eat b efo re co n su m in g it,

there is o ften a chance th a t b o n e em bedded in the m eat will be b u rn e d (criteria
fo r recognizing b u rn e d b o n e are review ed in C h a p te r 9). W h e th e r a b o n e is
b u rn ed d u rin g co o k in g depends o f course on how m eat is coo k ed , a n d w hether,
say, b o n e in m eat b ro iled over a fire will be b u rn e d dep en d s on w h e th e r the
bo ne is exposed to th e heat. Balm e (1980) suggests th a t tax a w ith m u ch b u rn ed
b o n e p ro b a b ly re p resen t h u m a n fo o d resources w hereas tax a w ith little b u rn e d
b o n e in th e sam e site m ay n o t have been used as a fo o d resource. Evidence
described by G ra y so n (1988; see also L ym an 1988a) indicates this distin ctio n
m ay n o t be a valid one w hen used alone. F u rth e r, B alm e (1980) fo u n d b u rn ed
b o n e in a cave th a t co n tain e d clear evidence o f h u m a n o cc u p atio n , b u t also
fo u n d b u rn e d bo n e in a cave th a t a p p a re n tly h ad n o t been occupied by people
a lth o u g h in the la tte r site th ere w as m u ch less b u rn e d b o n e th a n in the fo rm er
(see B riu er 1977 fo r sim ilar results).
G ra y so n (1988:27-29) exam ined the p ro p o rtio n s o f individual b u rn e d
skeletal elem ents o f ja c k ra b b it (L epus sp.) recovered fro m a cave in U ta h . H e
fo u n d skull a n d u ln a specim ens w ere b u rn e d less often, a n d tib ia specim ens
were b u rn e d m ore o ften th a n can be ac co u n ted for by ra n d o m chance in the
assem blage as a w hole, b u t w hen stratig ra p h ic ally distin ct assem blages w ere
stu died , these results chan ged. T h ere w as no co n sisten t b etw e en -stra ta p a tte rn
in th e b u rn in g o f ja c k ra b b it specim ens. F o r this set o f rem ains, b u rn in g was
ra n d o m ly d istrib u te d across skeletal elem ents from s tra tu m to stra tu m because
the b u rn in g occu rred afte r the skeletons h ad been d isarticu lated a n d deposited,
an d the b u rn in g w as th e result o f in situ n a tu ra l b u rn in g o f the d ry organic
m a tte r in the cave. G ra y so n (1988:29) em phasized th a t the lack o f precise
p ro ven ien ce in fo rm a tio n fo r ind iv id u al specim ens precluded ad d ressin g the
reaso n s fo r th e differential b u rn in g o f the fa u n al rem ain s (w ere tib iae m ore
o ften d ep o sited in stra tig ra p h ic areas th a t w ere to becom e b u rn e d th a n skulls
a n d ulnae?). G ra y s o n ’s (1988) stu d y is in stru ctiv e because it illu strates how the
ta p h o n o m ist m ight begin to u n rav el the b u rn in g h isto ry o f a set o f anim al
rem ains.
C om m inution o f bone: Boiling bone to ex tract grease m ay result in co m m i

n u ted bon e o r sm all b o n e fragm ents. T he presence o f extrem ely sm all bone
frag m en ts in large q u an titie s m ig h t be in terp re ted as indicative o f bo n e grease
p ro d u c tio n an d th u s th a t the tax a represented by the bones were exploited as a
food resource. H ow ever, sm all bone fragm ents are often difficult to identify to
skeletal elem ent (L y m an a n d O 'B rien 1987) a n d th u s they are also difficult to
identify to tax on. V ehik (1977:172-173) suggests fo u r lines o f evidence can be
used to in d icate b o n e grease p ro d u c tio n : (1) the presence o f m any sm all pieces
o f u n b u rn e d bone; (2) low frequencies to absence o f bones w ith high grease
c o n te n t (see C h a p te r 7 fo r m easures o f grease c o n te n t o f p a rtic u la r skeletal
elem ents); (3) th e presence o f h am m ersto n e s, anvil stones, fire pits, a n d
th erm ally fractu re d rock; an d (4) the loss o f collagen fibers from the bone
frag m en ts. T he la tte r, o f course, can also resu lt fro m diagenetic processes (see
C h a p te r 11).
M ineralization, weathering, and staining: T he degree o f w eathering, m in erali

za tio n , a n d stain ing o f b ones occasionally varies betw een intrusive o r n atu ra lly
dep osited an d cu ltu ra lly d ep o sited bo n e (descriptions o f these m odifications
are given in C h a p te rs 9 a n d 11). C u ltu ra l bo n e sh o u ld be m ore heavily
w eath ered , stained , a n d /o r m ineralized th a n intrusive bo n e because the latter
was d ep o sited in site sedim ents som e tim e afte r the form er. H ow ever, sim ply
because a b o n e is stain ed in a m a n n e r suggesting it is n o t intrusive does n o t
m ean it w as d ep o sited by hom inids. A dog, fo r exam ple, co u ld in tro d u ce non-
cu ltu ra l b o n e to a site sim u ltan eo u sly w ith the d ep o sitio n o f cu ltu ra l fo o d bone.
Butchering and technology m arks: T he presence o f b u tch erin g m ark s o r

m odificatio ns to b o n es resulting fro m the p ro d u c tio n o f bone to o ls are good
in d icatio n s th a t fa u n al rem ains w ere dep o sited as the result o f h u m a n activities
(these a ttrib u te s are described in C h a p te r 8). H ow ever, n o t only is th ere no
d etailed stu d y d isting u ish ing these tw o k in d s o f m odifications (w h at we m ight
term b u tch ery m ark s a n d tech n o lo g y m ark s, respectively), all cultu rally
d ep o sited fau n al rem ain s will n o t necessarily display one o r the o th er. T he
usual p ro c ed u re , how ever, is to argue th a t if som e o f the rem ains o f a tax o n
display such m odifications, th en all rem ains o f th a t tax o n w ere p ro b a b ly
ac cu m u lated an d dep o sited by ho m in id s (e.g., N o e -N y g aa rd 1989).
Ethnographic analogy: E th n o g ra p h ic an alo g y m ay be helpful in suggesting

w hich tax a w ere eaten , bu t there are at least tw o p o te n tia l problem s. F irst,
eth n o zo o lo g ical d a ta m ay be am biguous. F o r exam ple, in w estern N o rth
A m erica th ere are tw o species o f L y n x , the b o b ca t (L. rufus) an d the lynx (L.
canadensis), w hich in som e areas are sym patric. Som e e th n o g ra p h ic accounts
for this region refer to the “ w ild cat” as an exploited tax o n b u t w ith o u t listing
the tax o n o m ic nam e. T his is the genus L y n x , b u t w hich species? T he second
p o te n tia l p ro b lem is th a t the e th n o g ra p h ic record, as has been p o in ted ou t
m any tim es, is only o ne fram e o f an as yet unfinished featu re film. C u ltu res
evolve a n d change, a n d w h at people w ere eating in, say, 1905 w hen e th n o
g rap h ic d a ta w ere reco rd ed m ay n o t be w h a t the an cesto rs o f those p eople were
eatin g a th o u sa n d years earlier.
S ke le ta l com pleteness: T h o m as (1971:367) suggested th a t because “ the die

tary practices o f m an tend to d estro y a n d disperse the bones o f his prey-
species,” th e set o f tax a w ith lesser relative skeletal com pleteness is the one th at
can be identified as consisting o f cu ltu rally deposited bone w hereas the set o f
tax a w ith high relativ e skeletal com pleteness is th e one th a t can be identified as
co nsisting o f n atu ra lly dep o sited bone. W hile use o f this tech n iq u e has been
ad v o cated by som e (e.g.. W ing a n d B row n 1979; Z iegler 1973), o th ers have
arg u ed th a t the co m p u ted statistic a ctu ally m easures sam ple size (e.g., G ra y so n
1978b). A s well, e th n o g ra p h ic research sub seq u en t to T h o m a s ’ (1971) sugges
tio n has show n th a t h u m an s d o n o t alw ays accu m u late a n d dep o sit fractio n s o f
an im al carcasses (see C h a p te r 7). A m easure o f relative skeletal com pleteness
ca n n o t d isting u ish , say, an incom plete ra b b it skeleton d ep o sited by n atu ra l
processes fro m a co m p lete ra b b it skeleton d ep o sited by cu ltu ra l processes. This
does n o t m ean m easures o f skeletal com pleteness are no t useful analytical
to ols, a n d we re tu rn to them in later ch ap ters.
C o n text an d associations: V ariously b u rn e d , co m m in u ted , m ineralized, an d

b u tch ered bon e is readily believed to have been m odified an d dep o sited by
cu ltu ra l processes w hen such b ones are associated w ith u n d isp u ted evidence o f
h o m in id s, such as artifacts. T h a t is so because, w hile a fossorial ro d e n t m ay die
in its b u rro w w ithin a site a n d th u s its rem ains m ay becom e associated w ith
artifacts, it is d o u b tfu l th a t th o se ro d e n t rem ains will also be b u rn e d , highly
fractu re d , a n d v ario u sly m ineralized o r stained, o r th a t they will display
b u tch erin g m ark s. T h e co m bined a ttrib u te s o f b u rn in g , frag m en tatio n , sim ilar
m in eralizatio n o r stain ing across m ultiple specim ens, b u tch ery m ark s, an d
asso ciatio n w ith artifa cts, all p o in t to the sam e ac cu m u latio n agent. A s m o re o f
these a ttrib u te s fail to be present, the inference th a t the rem ains represent
cu ltu ra lly ac cu m u lated bone progressively w eakens.
Accumulation and dispersal as mirror images
O ne possible effect o f ac cu m u latio n a n d dispersal th a t m ight be considered

d u rin g analysis is th a t, b eginning w ith a co m p lete carcass, som e b ones o f th a t
carcass m ay be dispersed fro m the lo catio n o f carcass d ep o sitio n a n d a c cu m u
lated in a n o th e r location . T h u s, dispersal a n d ac cu m u latio n o f b ones m ay have
w h a t m ig h t be referred to as m irror image effects on the c o m p o sitio n o f skeletal
p a rts in a carcass from w hich selected b ones have been dispersed a n d the
co m p o sitio n o f th e dispersed a n d ac cu m u lated skeletal p a rts. B inford
(1981 b:222) said it well w hen he w ro te “ a tra n s p o rte d [actively accum ulated]
assem blage sh o u ld look like the ‘o p p o site ’ o f the n o n tra n sp o rte d [carcasses
from w hich bones have been dispersed] assem blage; th a t is, it [the tra n sp o rte d
assem blage] sh ou ld be the p ro p o rtio n a l inverse o f the p a rts th a t w ere n o t
tra n s p o rte d .” W hile this is a logical su p p o sitio n re g ard in g the effects o f
ac cu m u latio n a n d dispersal o f skeletal p a rts on the relative frequencies o f those
p arts, B inford (1981 b:222—229) h ad som e difficulty finding em pirical evidence
th a t this p redicted re la tio n held in the real w orld.
B unn et al. (1991) describe an d c o m p are th e % M A U frequencies o f skeletal
p a rts they reco rd ed on the landscape su rro u n d in g som e settlem ents occupied
by A frican foragers, a n d the frequencies o f skeletal p a rts in one o f those
settlem ents. T hey com piled the d a ta w ith in tw o general size classes o f anim als;
sm all m am m als w eigh betw een 2.2 an d 115 kg, large m am m als w eigh betw een
115 a n d 905 kg (T able 6.11). If B in fo rd 's (1981b) susp icion is co rrect, then the
frequencies o f skeletal p a rts recorded on the A frican lan d scap e should be
inversely co rrelated w ith the frequencies o f skeletal p a rts in the settlem ent.
T h a t is so because a t least som e o f the bones o n the lan d scap e display
bu tch erin g m ark s a n d th u s p ro m p te d B unn et al. (1991) to conclude th a t these
p ro b a b ly derived from an im al kill sites created by h o m in id activities. S catter-
plots o f th e frequencies o f b ones o n th e landscape ag ain st the frequencies o f
bones in th e settlem ent for b o th size g ro u p s o f m am m als, an d th eir respective
statistics, d o n o t m eet B in fo rd ’s (1981b) ex p ectatio n (F ig u re 6.27). As B inford
(1981b) before them , B unn et al. (1991) suggest the expected re la tio n is n o t
fo u n d due, at least in p a rt, to v a ria tio n in th e survival o f different skeletal p arts.
F o r exam ple, they in te rp re t the bo n e frequencies as in d icatin g th a t b ones seem
to preserve b etter, a n d h ave a g re ater p ro b a b ility o f being in c o rp o ra te d into the
(future) fossil reco rd if they are d ep o sited in the settlem ent th a n if they are left
on th e lan d scap e regardless o f the size o f the rep resen ted an im al. F u rth e r, the
rem ain s o f larg er anim als ten d to be m o re a b u n d a n t o n the lan d scap e th a n the
T ab le 6.11 Frequencies ( % M A U ) o f skeletal p a rts o f two sizes o f m am m als

fr o m the landscape and fr o m a hom inid settlem ent. S ize classes 1 and 2, anim al
live-weight size = 2.25 to 115 k g (5 to 250 lb s); size classes 3 a n d 4 anim al
live-weight size = 115 to 905 k g (250 to 2000 lb s). S ize classes and data fr o m
Bunn et al. (1991)
Skeletal Size classes ]1 and 2 Size classes 3 and 4
p art landscape settlem ent landscape settlem ent
cranium 100.0 100.0 100.0 28.6

m andible 29.2 62.5 47.4 57.1
cervical 19.1 50.0 63.9 53.1
thoracic 1.2 21.5 20.2 33.4
lum bar 8.3 70.8 24.6 57.1
sacrum 8.3 25.0 26.3 57.1
rib 0.3 9.6 12.0 18.0
pelvis 25.0 37.5 44.7 14.3
scapula 29.2 100.0 36.8 42.9
hum erus 16.7 87.5 26.3 100.0
radius 20.8 25.0 21.1 85.7
fem ur 16.7 50.0 47.4 85.7
tibia 20.8 62.5 42.1 57.1
m etapodial 13.0 25.0 22.4 35.7
rem ains o f the sm aller anim als regardless o f the skeletal p a rt represented.
F inally, B unn et al. (1991:52) suggest th a t the frequencies o f lim b elem ents o f
large an im als are hig h er th a n th e frequencies o f lim b elem ents o f sm all anim als
in the settlem en t due to differential tra n sp o rt.
W hile it is d o u b tfu l th a t excavations will o ften be extensive en o u g h to allow
direct co m p ariso n o f skeletal p a rt frequencies in sites rep resen tin g lo catio n s
from w hich bones w ere dispersed w ith skeletal p a rt frequencies in sites
rep resen tin g lo catio ns in w hich bones w ere accu m u lated , B in fo rd 's (1981b)
ex p ectatio n reg ard in g the inverse relatio n o f these tw o kinds o f bo n e assem
blages is a re aso n ab le one. But as his research an d th a t o f B unn et al. (1991)
in d icate, even in eth n o arch ae o lo g ic al settings the ex p ectatio n m ay n o t be m et
due to differential p re serv atio n a n d /o r differences in the tra n s p o rt o f skeletal
p arts. W e re tu rn to these topics in C h a p te r 7.
Summary
The problem is to determ ine the identity o f the prim ary accu m u lato r when evidence
has been distorted by secondary o r even tertiary anim al activity o n bones . . .
In te rp re ta tio n o f m ulti-agency accum ulations is difficult, if n o t im possible, after
inadequate excavation o r collection.
(G. A very 1984:347)
I have in this c h a p te r discussed vario u s o f the characteristics o f the dispersal

a n d ac cu m u latio n o f an im al rem ains. I have also review ed m an y o f the
A ccum ulation and dispersal o f vertebrate remains 221
100
<
Q)
E
_a>
c
a>
C/5
Landscape %M AU
100 -
3
<
a
05
E
.2?
CD
CO
20 40 60 80 100
Landscape %M AU
Figure 6.27. Bivariate scatterplots o f relative frequencies o f bones from (a) small
m am m als and (b) large m am m als o n the A frican landscape against bone
frequencies in a hom inid settlem ent. If tap h o n o m ic processes o th e r th an
tran sp o rt by hum ans have n o t altered bone frequencies, the p o in t scatter should
define a d iagonal line w ith a negative slope.
a ttrib u te s o f fossil assem blages p ro p o se d as d iag n o stic o f p a rtic u la r dispersal

and ac cu m u latio n processes a n d agents. W h a t w a rra n ts em phasizing is th a t
a rc h a e o fa u n a l assem blages m ay be the result o f m ultiple dispersal an d
a c cu m u latio n agents; w hen th a t is the case, the analyst is in for som e difficult
w ork. In a w ay, as n o ted in the p rev io u s section, the pro b lem is reduced to a
q u estio n o f w h eth er an assem blage o f bones (o r w hich p o rtio n s o f it) was
cu ltu ra lly o r n a tu ra lly d eposited. T he a ttrib u te s described in this c h a p te r are
typically used to help answ er this question.
A c cu m u la tio n a n d d ispersal o f v erte b rate skeletons are n o t easily dis

tin g u ish ed co n cep tually . D ispersal tends to be conceived o f as tak in g place
relative to th e orig in al lo catio n o f carcass dep o sitio n . A c cu m u la tio n can be o f
carcasses, in w hich case the original locus o f carcass d ep o sitio n is exam ined
such as in th e case o f kill sites. A c cu m u la tio n is typically conceived o f relative to
w here carcasses a n d bo nes are finally d ep o sited . A c cu m u la tio n can be o f
in d iv idu al skeletal specim ens, in w hich case th e b ones have been tra n sp o rte d
fro m th e orig in al lo catio n s o f carcass d ep o sitio n to a n o th e r lo catio n such as a
c a rn iv o re ’s den site o r a h u m a n h a b ita tio n site, o r it can be o f carcasses such as
th o se o f an im als w hich are tra p p e d in a bog. A c cu m u la tio n is ch aracterized as
passive w hen the b eh av io rs o f the anim als w hose bones are being studied
d ictate how the b o nes are accu m u lated . A ccu m u latio n is active w hen it is the
actio n o f a ta p h o n o m ic process external to the anim als w hose b ones are being
stu d ied th a t results in th eir ac cu m u latio n . M u ch o f this distin ctio n , as well as
ad d itio n a l a ttrib u te s o f ac cu m u latio n , is su m m arized in F ig u re 6.2 a n d its
asso ciated discussion.
Iden tify in g th e m echan ism s a n d agents o f b o n e dispersal an d ac cu m u latio n
involves the study o f m an y a ttrib u te s o f an assem blage o f fa u n al rem ains.
K now ledge o f w hich fa u n al rem ains p a rtic u la r ag en ts ten d to disperse and
accu m u late, a n d know ledge o f how th o se dispersal a n d a c cu m u latio n agents
m odify b o nes, are the keys to w riting dispersal a n d ac cu m u latio n histories o f
assem blages o f fau n al rem ains. Such know ledge helps answ er the q u estio n
“ w h a t are all these b ones d o in g here?” A ttrib u te s o f fossil assem blages o th er
th a n th o se discussed in this c h a p te r also help answ er this q u estio n , b u t as we see
in the next ch a p te r, som etim es the q u estio n being asked is m o re specific th an
sim ply id entifying an ac cu m u latio n agent.
7
F R E Q U E N C I E S OF S K E L E T A L P ART S
A fter death, verteb rate carcasses are often subjected to the sam e m echanical laws o f
tra n sp o rt by geologic agents as any o th er co m p o n en t o f the sedim ent. T heir specific
gravity, w hich changes according to changing buoyancy, an d the relatio n o f m ass to
surface area specify the m echanical arrangem ent.
(J. W eigelt 1927/1989:160-161).
Introduction
O ne o f th e m o st obv ious a n d visible p ro p e rtie s o f a fa u n a l assem blage is the
frequencies o f each o f the p a rtic u la r skeletal elem ents th a t m ak e u p the
collection. A co m p lete m am m al skeleton, fo r exam ple, alw ays consists o f tw o
h u m eri, tw o scap ulae, tw o m andibles, one skull, etc. F ro m this m odel o f
relative frequencies o f skeletal p a rts in a n individual, one ca n p re d ic t w h at
sho u ld be fo u n d in a fossil assem blage th a t c o n tain s, fo r instance, 10 skulls;
here, 20 h u m eri, 20 scapulae, 20 m an d ib les, etc., sh o u ld be fo u n d if ta p h o n o m ic
processes h av e n o t resulted in the rem o v al o f certain kinds o f bones, an d
sam p ling a n d recovery processes have n o t failed to find c e rtain skeletal p arts.
A nalysis o f skeletal p a rt frequencies, o r w h a t are som etim es called skeletal p a rt
profiles, has, in the p a st 15 years, becom e a m ajo r p a rt o f ta p h o n o m ic research.
In fact, th e references cited in this c h a p te r show th a t th ere has been a m ajo r
b u rst o f p u b lic a tio n o n this to p ic in the late 1980s a n d early 1990s.
Human utilization and transport o f carcass parts

It is necessary no t only to give a reasonable explan atio n for the presence o f the
elem ents fou n d b u t also for the absence o f those n o t found. F ro m here o n o u t the
stud en t is on his own.
(T. E. W hite 1953c:61)
Z o o arch ae o lo g ists h av e long been in terested in explaining the v aria b ility w ith
w hich skeletons o f p a rtic u la r tax a are represented. F o r exam ple, nearly 40
years ago T h e o d o re W hite, a p aleo n to lo g ist w ho also did zooarch aeo lo g ical
research in N o rth A m erica, p ro p o se d in terp re tiv e assu m p tio n s fo r explaining
th e v a ria tio n in frequencies o f skeletal p a rts in zo o a rch a eo lo g ical collections.
T hese assu m p tio n s w ere fo u n d e d o n W h ite ’s know ledge o f m am m alian
a n a to m y a n d eth n o g ra p h ic d a ta co n cern in g h u m a n b u tch ery a n d subsistence
practices. W h ite (1953c:59, 1956:401) p h ra sed the q u estio n in the follow ing
223
way: “ W hich elem ents w ere b ro u g h t in to the cam p o r village a n d w hich

elem ents w ere left at th e kill?” T he in terp retiv e assu m p tio n s were p h ra sed as
follows: (1) “ W hile it c a n n o t be expected th a t all [skeletal] elem ents will [be
eq ually a b u n d a n t] it is difficult to escape the inference th a t som e p a rts w ere n o t
b ro u g h t in to c a m p ” (W hite 1952a:337), a n d (2) “ Since the low er lim b does n o t
ca rry an y usable m eat it is conceivable th a t it was ch o p p e d off a n d left a t the
place o f th e kill in o rd e r to reduce the lo a d ” (W hite 1953b: 162). Sim ply p u t,
th en , W h ite believed th a t skeletal elem ents w ith large am o u n ts o f usable m eat
w ere m o re likely to be tra n s p o rte d fro m th e kill to the c o n su m p tio n site th a n
elem ents w ith sm all am o u n ts o f usable m eat. Basically, elem ents w ith large
am o u n ts o f usable m eat w ere believed by W hite to consist o f the p ro x im al lim b
bones w hereas elem ents w ith sm all am o u n ts o f usable m eat included the distal
lim b bones.
Im p o rta n tly , W h ite also allow ed fo r th e fact th a t n o t all b o n e assem blages
recovered from suspected h a b ita tio n , n o n-kill sites co nsisted o f m ostly p ro x i
m al lim b elem ents. I f p ro x im al lim b bones w ere ab sen t from non-kill sites,
W h ite believed h u m a n b u tch ery practices h ad destro y ed the bones (e.g.,
red u ced them via fra g m e n ta tio n to u n identifiable fragm ents) a n d /o r th a t
scavenging carn iv o res such as dogs h ad co nsum ed them . F o r an y skeletal
elem ent th a t was rep resen ted in low ab u n d a n ce, W hite (1956:402) suggested
the an aly st co n sid er w h eth er the m issing elem ents w ere ab sen t due to an
“ accid en t o f p reserv atio n , o r an accident o f sam p lin g .” H e even suggested
som e skeletal elem ents rep resen ted in low a b u n d a n ces m ay have been m ade
in to to o ls a n d th e n been carried aw ay, o r if fo u n d , w ere m odified o r w orn
b ey o n d reco gnition.
In the O ld W o rld , p erh ap s the b est-k n o w n stu d y o f skeletal p a rt ab u n d a n ces
is th a t o f P erkins a n d D aly (1968), w ho p ro p o se d the concept o f the schlepp
effect to a c co u n t fo r v aria tio n s in skeletal p a rt ab u n d a n ces (see also K lein
1976). D a rt’s (e.g., 1957) study o f S outh A frican P lio-P leistocene bovid
rem ains p ro m p te d him to p ro p o se th a t differential use by early h o m in id s o f
skeletal elem ents as to o ls ac co u n ted fo r v a ria tio n in frequencies o f b ones an d
b o n e p a rts, b u t his p ro p o sa l w as im m ediately co n tro v e rsial (see C h a p te r 3).
P erkins an d D a ly ’s (1968) schlepp effect, on the o th e r h an d , did n o t fo ster the
sam e level o f co n tro v e rsy u ntil a decade afte r its p ro p o sa l (e.g., B inford 1978,
1981b). T h e schlepp effect, fro m the G e rm a n verb m ean in g “ to d ra g ,” as
o riginally described, “ co m bines a fa c to r related to the size o f the gam e anim al
w ith one related to th e distance betw een kill site an d hom e settlem en t” (P erkins
an d D aly 1968:104). T he schlepp effect can be defined as “ the larger the anim al
and the fa rth e r aw ay from the p o in t o f co n su m p tio n it is killed, the few er o f its
b ones will get 'sc h le p p ed ' b ack to the cam p, village, o r o th e r a re a ” (D aly
1969:149).
P erkins an d D aly (1968:104) p ro p o se d the schlepp effect as a w ay to account
fo r low relative a b u n d a n ces o f cattle lo n g b ones associated w ith high a b u n
Frequencies o f skeleta l parts 225
dances o f ca ttle fo o t bones in a p re h isto ric site occupied by h u n ters. T hey

suggested th a t " p e rh a p s because the feet m ad e convenient han d les fo r dragging
the m eat-filled hide [and because the fo o t bones] co n tain useful sinews a n d have
been called ‘the h u n te r’s sewing k it’,” the archaeologically observed relative
a b u n d a n ces o f fo o t bones a n d long bones could be ac co u n ted for by reference
to the schlepp effect. A ccording to P erkins a n d D aly a p re h isto ric village
occupied by p asto ra lists should p ro d u c e a p p ro x im a te ly equal frequencies o f
ca ttle long b ones a n d fo o t bones. P erkins a n d D aly (1968:104) refer to “ N ew
W o rld arch ae o lo g ists” w ho, they state, have show n th a t skeletal p a rts will be
rep resen ted in u n eq u a l a b u n d a n ces a t cam p sites to w hich selected p a rts o f
anim als b u tch ered elsew here w ere b ro u g h t. B ecause the only such a rc h a e o
logist they cite is W h ite (1953c), it is tem p tin g to suggest th a t th eir d evelopm ent
o f the co n cep t o f the “ schlepp effect” h ad its ro o ts in his w ork.
B in fo rd (1978:12) w ished to arg u e th a t v aria tio n s in the frequencies o f
skeletal p a rts “ resu lted from v ariab le strategies in th e h u m a n use o f food
sou rces.” T h a t is, he w an ted to derive the sam e kinds o f conclusions th a t W hite,
a n d P erk in s a n d D aly h ad derived. B inford (1978:11) th erefo re b egan w ith the
assu m p tio n th a t " a n y v ariability in the relative frequencies o f a n a to m ica l p arts
am o n g arch aeolo gical sites m u st derive from the dynam ics o f th eir use.” H e
p o in ted o u t th a t people b u tc h e r (skin, eviscerate, d ism em ber, disarticu late,
deflesh, etc.; see C h a p te r 8) a n d sca tte r segm ents o f ind iv id u al carcasses d u rin g
th eir ex p lo itatio n . B inford (1978:10) fo u n d th a t “ w h at is clearly lacking from
o u r cu rre n t u n d ersta n d in g [of how people b u tch er carcasses] is a specific
know ledge o f the p a rtic u la r effects th a t m ig h t be expected to resu lt from
activ ities” such as b u tch erin g a n d schlepping. T o p ro v id e th a t specific k n o w l
edge he m easu red th e a m o u n ts o f m eat (w eight o f fat a n d m uscle tissue),
m arro w (m arro w cavity volum e m ultiplied by the p ercentage o f fatty acids
present in the m arro w ), a n d grease (volum e o f skeletal p a rt for cancellous p a rts
m ultiplied by the p ercen tage o f fa tty acids p resen t in the m arro w ) associated
w ith skeletal p a rts o f tw o d om estic sheep (O vis aries) a n d one ca rib o u (R angifer
tarandus). T he m easured values w ere used to c o n stru c t indices o f the food
utility o f the in divid ual carcass p a rts for h u m an consum ers. T h en , he c o n
stru cted “ a c o m p o u n d index o f general utility th a t m o n ito rs quite closely the
actu al p ro p o rtio n s o f different usable fo o d co m p o n en ts o f the a n im a l” based
on the am o u n ts o f m eat, m arro w , an d grease associated w ith each skeletal p a rt
because it w as clear th a t the th ree fo o d substances w ould sim ultaneously
influence choices m ad e by people a b o u t how carcasses w ere to be b u tch ered ,
sto red , a n d /o r tra n sp o rte d (B inford 1978:72). F inally, B inford (1978:74)
altered th e values o f the co m p o u n d , general utility index (G U I) to create a
“ m odified general utility in d ex ” (M G U I) th a t reflected the fact th a t anim als are
n o t alw ays b u tch ered in to the discrete skeletal p a rts fo r w hich the G U I values
h ad been m easu red; ra th e r, som e p a rts w ith low G U I values m ay rem ain
a tta c h e d to (ride w ith d u rin g tra n s p o rt) skeletal p a rts th a t have higher G U I
le 7.1 B in fo rd ’s (1 978) norm ed utility indices f o r dom estic sheep and caribou (m ea t, marrow, grease, G U I, M
. Sheep C aribou
etal ________________________________________________________ _______________________________________________
ent M eat M arro w G rease GUI MGUI M eat M arro w G rease GUI
r/h o rn 1.0 1.03 1.0

12.86 1.0 25.74 12.87 9.05 1.0 — 17.49
dible
tongue 43.36 10.35 11.75 43.50 43.6 31.1 5.74 12.51 30.26
out tongue 14.12 10.35 11.75 11.65 11.65 11.4 5.74 12.51 13.89
18.65 1.0 7.19 18.68 18.68 10.1 1.0 13.11 9.79
18.65 1.0 9.47 18.68 18.68 10.1 1.0 12.93 9.79
cal 55.32 1.0 15.00 55.33 55.33 37.0 1.0 17.46 35.71
cic 46.47 1.0 9.82 46.49 46.49 47.2 1.0 12.26 45.53
ar 38.88 1.0 14.74 38.90 38.90 33.2 1.0 14.82 32.05
100.0 1.0 9.30 100.0 100.0 51.6 1.0 7.50 49.77
um 90.52 1.0 11.05 90.52 90.52 66.5 1.0 26.00 64.13
ula 44.89 6.23 3.85 45.06 45.06 44.7 6.40 7.69 43.47
m erus 28.24 28.26 56.67 29.50 37.28 28.9 29.69 75.46 30.23
m erus 28.24 41.21 9.38 29.31 32.79 28.9 28.33 27.84 29.58
dius 14.01 35.40 32.54 15.18 24.30 14.7 43.64 37.56 16.77
dius 14.01 68.98 18.77 15.81 20.06 14.7 66.11 32.70 17.82
als 4.74 1.0 22.98 5.00 13.43 5.2 1.0 36.47 5.51
tacarpal 4.74 62.93 13.24 6.37 10.11 5.2 61.68 16.71 8.24
etacarpal 4.74 71.85 33.59 6.79 8.45 5.2 67.08 42.47 8.83
s 81.30 9.57 34.65 81.51 81.50 49.3 7.85 29.26 47.89
m ur 78.24 38.62 23.68 79.38 80.58 100.0 33.51 26.90 98.32
m ur 78.24 56.05 100.0 80.58 80.58 100.0 49.41 100.00 100.0
ia 20.76 57.84 56.40 22.71 51.99 25.5 43.78 69.37 27.57
ia 20.76 100.0 26.84 23.41 37.70 25.5 92.90 26.05 29.46
galus 6.37 1.0 24.38 6.64 23.08 11.2 1.0 32.47 11.23
neum 6.37 23.11 34.38 7.27 23.08 11.2 21.19 46.96 12.40
tatarsal 6.37 64.16 12.37 8.02 15.77 11.2 81.74 17.88 15.03
etatarsal 6.37 73.52 33.33 8.46 12.11 11.2 100.0 43.13 16.24
phalanx 3.37 33.77 15.70 4.33 8.22 1.7 30.00 33.27 ■ 3.52
nd phalanx 3.37 25.11 13.33 4.10 8.22 1.7 22.15 24.77 3.03
phalanx 3.37 1.0 9.82 3.49 8.22 1.7 1.0 13.59 1.85
Frequencies o f skeleta l p a rts 227
values. B in fo rd (1978:74) reaso n ed th a t a skeletal p a r t w ith a low G U I an d

a tta c h e d to a skeletal p a rt w ith a high G U I should tak e on a utility value
eq u iv alen t to th e average o f the tw o values (see B inford 1978:15-34, 72-75;
M etcalfe a n d Jo n es 1988; Jo n es a n d M etcalfe 1988; fo r detailed discussion o f
how th e utility indices w ere derived). B ecause B inford (1978) w as interested in
differential tra n s p o rt o f skeletal p arts, his M G U I w as co n stru c te d to ac co u n t
fo r “ rid e rs,” o r skeletal p a rts w ith low G U I values th a t w ere tra n s p o rte d d u e to
th eir atta c h m e n t to p a rts w ith high G U I values. All utility values B inford
(1978) derived are given in T ab le 7.1. N o te th a t all o f these values have been
n o rm ed to a scale o f 1 to 100 by dividing all derived values by th e greatest
derived value in a colu m n. T h u s, the n o rm ed M G U I is typically referred to as
the % M G U I.
B inford (1978:81) m odeled how v a ria tio n in the h u m an tra n s p o rt and
u tilizatio n o f carcass p a rts w ould be reflected by c o n c o m ita n t v aria tio n in
skeletal p a rt frequencies an d in the % M G U I value fo r individual skeletal p arts.
T h a t m o d elin g to o k the form o f a fam ily o f curves, each rep resen tin g a
b iv ariate sca tte rp lo t o f po in ts derived by p lo ttin g the utility o f a p a rtic u la r
skeletal p a rt on th e x-axis ag a in st a m easu rem en t o f the frequency o f th a t
skeletal p a rt in an assem blage on the y-axis (F ig u re 7.1). T he unit fo r m easuring
th e frequency o f skeletal p a rts B inford (1978) chose to use is the m inim al anim al
unit (M A U ; o rigin ally called M N I values, B inford [1984b] changed the term to
M A U ). M A U values are calculated by first ascertain in g how m an y o f a
p a rtic u la r skeletal p a rt o r elem ent are present; th ree o v erlap p in g frag m en ts o f
distal h u m eri in d icate a m in im u m n u m b e r o f three distal h um eri, typically
called an M N E (m in im u m n u m b e r o f elem ents) value (see C h a p te r 4). A fter the
M N E p er skeletal p a rt is determ in ed , th o se values are then divided by the
n u m b e r o f tim es th a t p a rt occurs in one skeleton to p ro d u c e an M A U value.
T h u s an M N E o f th ree distal h um eri w ould be divided by tw o to derive an
M A U o f 1.5 distal h u m eri. I f 26 th o rac ic v erte b rae are present, a n d th e tax o n
u n d e r stu d y h as 13 th o rac ic v erte b rae p er in d iv id u al, an M A U o f tw o th o racic
verteb rae w ould be n oted. In B in fo rd 's p ro c ed u re , M A U values are n o rm ed on
a scale o f 1 to 100 (by dividing all M A U values fo r an assem blage by the
m ax im u m M A U value in th e assem blage) p rio r to p ro d u c in g the sca tte rp lo t
fro m w hich inferences o f differential tra n s p o rt an d u tility are derived. T he
n o rm ed values are typically referred to as % M A U values.
A s a n exam ple o f h ow this an aly tic tech n iq u e is im plem ented, the e th n o a r-
chaeologically d o cu m en te d site o f A n av ik (B inford 1978:78) can be used. This
site is a ca rib o u (R angifer tarandus) k ill-butchery site occupied in the spring by
the N u n a m iu t o f A laska. B oth M N E a n d M A U frequencies o f ca rib o u bones
recovered from this site are given in T ab le 7.2. N o te th a t B inford (1978) did no t
re p o rt M N E values fo r this site, b u t for sake o f illu stra tio n I derived them from
the M A U values he p resen ted in o rd e r to illu stra te the re la tio n sh ip betw een
M N E a n d M A U values. A b iv ariate sc a tte rp lo t (F ig u re 7.2) o f the % M A U
a 100
80
=> 60
<
° 40
20
0
0 20 40 60 80 100
%M G UI
b 100
80
60
%MAU
40
20
0
0 20 40 60 80 100
%M G UI
100
80
60
%MAU
40
20
0
40 60
%MGUI
d 100
80 -
60
=>
<
40
20
—r~ —i— — r~ —r~

20 40 60 80 1 00
% M G UI
%MGUI
Figure 7.1. A fam ily o f strategies for utilizing a n d /o r tra n sp o rtin g anim al carcass
parts based on the % M G U I (after B inford 1978; see also T h o m as and M ayer
1983). a, reverse (bulk) strategy; b, go u rm et strategy; c, b u lk strategy; d, unbiased
strategy; e, reverse g ourm et strategy.
values fo r ca rib o u rem ains fro m A n av ik ag a in st the ca rib o u % M G U I (T able

7.1) show s w h a t is readily inferred to be a reverse u tility curve based on the
m odel in F ig u re 7.1. T h a t the statistical re la tio n betw een the tw o sets o f
variab les is significant is in d icated by a co rre la tio n coefficient (rs) o f —0.85
(P < 0 .0 0 1 ). T h u s, th e an aly tic tech n iq u e is sim ply p erfo rm ed , a n d p roduces
w h a t a p p e a r to be readily in te rp re ta b le results. W e re tu rn to w h eth er the latter
is in fact the case la te r in this ch ap ter.
P erh a p s n o t surp risingly, given the logic beh in d B in fo rd 's (1978) m odel o f
u tility a n d tra n s p o rt curves (F ig u re 7.1). th e explicit know ledge o f the food
T ab le 7.2 M N E and M A U frequencies o f caribou bones f o r two

ethnoarchaeological sites created by the N unam iut and reported by B inford
(1 9 7 8 :7 8 ,3 2 3 )
Site: A navik Ingested

Skeletal p a rt MNE MAU %M AU MNE MAU %M AU
antler (2)a 106 53 100.0 1 0.5 7.69

skull (1) 44 44 83 1 1 15.38
m andible (2) 78 39 73.5 3 1.5 23.08
atlas (1) 40 40 75.4 1 1 15.38
axis (1) 44 44 83 0 0 0.0
oo
cervical (5) 210 42 79.2 4
O
12.31
thoracic (13) 260 20 37.7 13 i 15.38
lu m bar (6) 144 24 45.2 1 0.16 2.46
sacrum (1) 19 19 35.8 no d a ta
in nom inate (2) 45 22.5 42.4 1 0.5 7.69
rib (26) 364 14 26.4 26 1 15.38
sternum (7) 105 15 28.3 5 0.71 10.92
scapula (2) 36 18 33.9 13 6.50 100.0
P hum erus (2) 30 15 28.3 0 0 0.0
D hum erus (2) 33 16.5 31.1 7 3.5 53.84
P r a d iu s (2) 39 19.5 36.7 4 2 30.77
D radius (2) 49 24.5 46.2 3 1.5 23.08
carpals (12) 360 30 56.7 8 0.75 11.54
P m etacarpal (2) 63 31.5 59.4 3 1.5 23.08
D m etacarpal (2) 63 31.5 59.4 1 0.5 7.69
P fem ur (2) 18 9 16.9 5 2.5 38.46
D fem ur (2) 18 9 16.9 0 0 0.0
P tibia (2) 26 13 24.5 0 0 0
D tibia (2) 27 13.5 25.4 5 2.5 38.46
astragalus (2) 44 22 39.6 4 2 30.77
calcaneum (2) 46 23 38.6 3 1.5 23.08
P m etatarsal (2) 73 36.5 68.9 4 2 30.77
D m etatarsal (2) 46 23 43.3 2 1 15.38
first phalanges (8) 292 36.5 68.9 6 0.75 11.54
second phalanges(8) 288 36.0 67.9 11 1.37 21.08
th ird phalanges (8) 284 35.5 66.9 20 2.5 38.46
Note:
a n um ber o f tim es a skeletal p a rt occurs in one individual
u tility o f carcass p a rts (T able 7.1), a n d th e straig h tfo rw a rd an aly tic technique
o f g en eratin g a sea tte rp lo t th a t could be in terp re ted in term s o f the m odel, the
tech n iq u e w as quickly p u t to use by arch aeo lo g ists (e.g., L andals 1990; Speth
1983; T h o m as a n d M ay er 1983). H ere w as an alg o rith m th a t ap p a ren tly
g ra n te d significant insights to the h u m an beh av io ral p a rt o f the tap h o n o m ic
h isto ry o f a b o n e assem blage. T he m a n n e r in w hich the utility indices were
d erived cam e u n d e r close scru tin y (e.g., C hase 1985; Jo n es a n d M etcalfe 1988;
100
80
3 60
<
20
%MGUI
Figure 7.2. S catterplot o f carib o u % M A U values from A navik ag ain st caribou
% M G U 1 values.
M etcalfe a n d Jo nes 1988), b u t no one co u ld deny the utility o f the m odel o r

in v alid ate th e specific kn ow ledge B inford h ad presented. T h u s it sh o u ld com e
as no su rp rise th a t researchers began to gen erate utility indices fo r o th e r taxa
(B lum enschine a n d C a ro 1986; B o rrero 1990; B rink a n d D aw e 1989; E m erson
1990; K o o y m an 1984, 1990; L ym an et al. 1992b; Will 1985), m any o f w hich are
sum m arized in T ables 7.3 an d 7.4. V irtually all o f these indices w ere derived in a
m a n n e r sim ilar to th a t used by B inford (1978). T h a t is, the m uscle a n d fa t tissue
asso ciated w ith a skeletal p a rt w as w eighed an d , less often , the a m o u n ts o f
m arro w a n d grease w ere estim ated o n the basis o f the volum e o f each skeletal
p art. T h o se w eight values w ere th en m ath em atica lly co n v erted to the index
values. T h e indices h ave variously been called utility indices, econom ic utility
indices, o r food u tility indices (F U I). M o st o f them have been derived by
archaeolo g ists w ho bu tch ered carcasses a n d m easu red (typically w eighing) the
a m o u n t o f fo od tissue (m eat, m arro w , fat, grease) associated w ith a skeletal
p a rt such as a p ro x im al h u m eru s, a com plete skeletal elem ent such as the
scap u la, a n d sections o f the skeleton such as the ribs o r th o rac ic vertebrae.
F ollo w in g B inford (1978) m o st an aly sts n o rm th eir indices to a scale o f 1 to 100,
a n d it is these n o rm ed values th a t are given in T ables 7.3 a n d 7.4.
B ecause m o st o f these indices are fo u n d ed on m easu rem en ts fro m one o r a
few (alw ays < 4 ) in div idu als, I believe they are a t best o rd in al scale. T he indices
are based o n averag e m easu rem en ts o f m eat a n d fa t a m o u n ts associated w ith,
fo r exam ple, th e scapu lae (left a n d right) o f only one o r a few individuals. T hey
th u s (a) m ute ind ividu al v a ria tio n such as th a t displayed by individuals o f
different age, sex, a n d n u tritio n a l statu s (typically co rrelated w ith season), n o t
7.3 U tility and transport indices f o r various taxa
C arib ou FU I Phocid seal Im pala A lcelaphine Llam a F U

l (M etcalfe and FUI tran sp o rt tran sp o rt (M engoni
t Jones 1988) (L ym an et al. 1992b) (O ’C onnell et at. 1990) G on alo n s
9.1 27.4 13 3 14.75
ble w ith skull 13 3
tongue 31.1 9.95
out tongue 11.5 5.25
nd axis 10.2 w ith cervical with cervical w ith cervical 8.57
l 37.1 35.8 16 4 64.15
c 47.3 24.9 16 4 61.75
33.2 32.9 16 4 77.97
51.6 100.0 9.5 3.5 100.00
m 66.6 2.7 w ith rib w ith rib 99.35
a 44.7 19.8 12.5 3 41.66
us 36.8 10.7 11 2 36.68
ulna 25.8 4.8 10.5 1 23.00
rpal 5.2 10 1 6.53
lipper w ith carpals 2.3
phalanges
nate 49.3 44.5 15 4 40.18
100.0 4.5 10 2 75.94
bula 62.8 (w /tarsals) 16.5 10 (w /tarsals) 2 (w /tarsals) 43.04
rsal 37.0 9.5 2 11.46
pper w ith tarsals 7.7
phalanges
ges 19.4 15 2 4.78
7.4 U tility indices f o r bone p a rts o f various m am m alian taxa
G u an aco M odified Bison Bison m odified M uskox m o

l general utility gu an aco utility fat total products general utili
(B orrero 1990) (L ym an 1992a) (Brink and D aw e 1989) (E m erson 1990) (Will 1985)
10.0 10.0 no d a ta 14.2 11.23

ble included with skull
tongue 5.7 5.7 no d ata 36.55
out tongue no d ata no d a ta no d ata 24.16
8.8 8.8 no d ata 6.4 20.47
8.8 8.8 no d a ta 7.8 20.47
l 51.3 51.3 no d a ta 56.6 47.50
c 22.1 22.1 no d a ta 84.7 66.00
44.9 44.9 no d a ta 82.9 61.57
100.0 100.0 no d ata 100.0 55.57
m 8.5 8.5 no d a ta 52.9 83.54
a 38.4 38.4 no d ata 31.6 28.89
erus 23.8 38.4 40.5 31.6 55.58
erus 23.8 23.8 22.0 25.1 48.69
s 7.8 15.8 33.5 16.5 41.81
us 7.8 7.8 25.7 12.1 48.83
15.8 19.8 no d a ta 20.8
7.8 7.8 no d ata 12.1
1.3 4.5 no d ata 6.6 42.54
carpal 1.3 2.6 8.9 3.9 36.25
acarpal 1.3 2.4 15.2 2.6 49.82
inate 40.2 40.2 no d a ta 54.7 83.59
r 83.2 83.2 31.4 69.4 57.56
ur 83.2 83.2 35.2 69.4 100.00
21.3 52.2 33.5 40.8 80.62
21.3 21.3 14.1 25.5 61.25
lus 1.7 11.5 no d a ta 13.6 55.23
eum 1.7 11.5 no d a ta 13.6 55.23
lo-cuboid 1.7 11.5 no d a ta 13.6 55.23
tarsal 1.7 6.6 12.4 7.5 49.22
atarsal 1.7 4.4 22.7 4.5 64.27
alanx 2.1 2.1 no d ata 2.4 31.61
phalanx 2.1 2.1 no d ata 2.4 22.64
halanx 2.1 2.1 no d ata 2.4 7.45
to m en tio n in te r-p o p u la tio n v aria tio n , a n d (b) are n o t average values fo r the
com p lete ran g e o f v a ria tio n th a t different in dividuals o f a ta x o n m ay display
because few in dividuals h ave been m easured. E ven if we eventually o b ta in the
w eight o f m uscle tissue fo r b o th sexes o f each age class o f individuals o f a tax o n
an d th o se in d iv id u als died d u rin g v ario u s seasons o f th e year, th e fact th a t we
c a n n o t yet d eterm in e the sex, o n to g en ic age, a n d season o f d e a th o f a n anim al
rep resen ted by m ost isolated bones o r teeth we find, forces us to tre a t the
econ o m ic u tility indices as o rd in a l scale. T h u s w hile th e sca tte rp lo ts o f F igures
7.1 a n d 7.2 are in terv al scale a n d are useful heuristics, o rd in a l scale statistics are
called fo r w hen analyzin g the d a ta in the scatterp lo ts.
P ro b ab ly also as a direct result o f B in fo rd 's (1978) efforts, o th e r researchers
began to m o n ito r how m o d ern h u n te r-g a th e re rs differentially tra n s p o rt c a r
cass p a rts o f som e tax a (B unn et al. 1988; O 'C o n n e ll et al. 1988; 1990;
O 'C o n n ell an d M arsh all 1989) regardless o f th eir food utility. R a th e r th a n , for
exam ple, defleshing a carcass a n d w eighing the m eat, m arro w , a n d grease
asso ciated w ith p a rtic u la r bones to derive a utility index, these analysts
reco rd ed w hich b o nes w ere tra n s p o rte d fro m each in d iv id u al carcass o f
an im als th a t w ere killed o r scavenged by m o d e rn h u n ter-g ath ere rs. T hey then
o rd ered th e skeletal p a rts fro m those m o st often tra n s p o rte d to th o se least
o ften tra n sp o rte d . I (L y m an 1992a) term these “ tra n s p o rt indices” an d
su m m arize them in T ab le 7.4, w here the higher the n u m b er given fo r a skeletal
p a rt d en o tes a g re ater ch ance th a t a p a rt will be tra n sp o rte d ; th a t is, these
tra n s p o rt indices have n o t been no rm ed . As w ith the utility indices, I believe the
tra n s p o rt indices are b est trea ted as o rd in al scale.
T h e p u rp o se o f p resen tin g all o f th e available indices here should be evident:
th ey can be used to help explain varying frequencies o f skeletal p a rts in an
assem blage o f bones. As B inford (1987:453) notes, these indices “ can be
co nsid ered q u ite literally as a fram e o f reference, th a t fu n c tio n m u ch like a
screen u p o n w hich slides are p ro je c te d .” F o r exam ple, presu m in g the bones
were cu ltu rally accu m u lated (C h a p te r 6), these indices can serve as one fram e o f
reference o r one p a rt o f an ex p lan a tio n fo r observed ab u n d a n ces o f skeletal
p arts. Several such fram es o f reference (V oorhies G ro u p s, T ab le 6.5; fluvial
tra n s p o rt indices, T ab le 6.7) described in C h a p te r 6 p erta in to fluvial tra n sp o rt
processes. Below I co n sid er o th e r fram es o f reference a n d show h ow still o th er
ta p h o n o m ic v ariables m u st be co n sid ered to clarify w h eth er o r n o t differential
u tiliza tio n a n d tra n s p o rt o f carcass p a rts by h u m an s w as responsible fo r any
given set o f bones, b u t the utility a n d tra n s p o rt indices are a good place to sta rt
th e search fo r ex p lan a tio n s o f skeletal p a r t frequencies.
Structural density o f bones

T he investigator can begin w ith a concrete basis for com parison, since he know s the
exact ratio o f the different elem ents in th e living a n im a l. . . H e has a precise ratio o f
Frequencies o f skeletal parts 235
expectation w ith w hich to com pare the observed, and, if the sam ple size is large, the
differential rep resen tatio n o f the elem ents can be studied in detail.
(R. D. G u th rie 1967:243)
G iven a n in terest in ex plaining skeletal p a rt frequencies by reference to

tra n s p o rt o r utility indices, the zo o a rch a eo lo g ist m u st be sure th a t those
frequencies are indeed a reflection o f econom ic decisions a n d n o t som e o th er
fa cto r, such as differential preserv atio n . A s will becom e clear, the p ro b a b ility
th a t a skeletal p a rt will survive the rigors o f v ario u s ta p h o n o m ic processes is at
least p artia lly a fu n ctio n o f th a t p a r t’s s tru c tu ra l density (g /cm 3). It is to a
co n sid eratio n o f th a t p ro p e rty th a t we now tu rn .
W hile th e s tru c tu ra l density o f skeletal p a rts w as referred to in the n in eteen th
cen tu ry as p o ten tially m ed iatin g v ario u s ta p h o n o m ic processes, such as
carn iv o re gnaw ing (see C h a p te r 2), th a t p ro p e rty did n o t receive serious
a tte n tio n a g ain u n til G u th rie (1967) m en tio n ed it as a possibly im p o rta n t fa cto r
affecting the frequencies o f skeletal p a rts o f P leistocene m am m als in A laska.
G u th rie (1967) did n o t, how ever, m easu re the stru c tu ra l density o f bones.
B etter k n o w n to zo o a rch a eo lo g ists is the stu d y by B rain (1967b, 1969), w ho
m easu red th e s tru c tu ra l density o f g o a t (Capra hircus) bones in an a tte m p t to
ac co u n t fo r v aryin g frequencies o f skeletal p a rts in an assem blage kn o w n to
h ave been ravag ed by dogs (Canis fa m ilia ris) a n d people. A t the sam e tim e,
V o orhies (1969) n o ted th a t the s tru c tu ra l density o f bones seem ed to exert
s tro n g influences on the tra n s p o rta b ility o f skeletal elem ents by fluvial
processes, b u t he did n o t m easu re the stru c tu ra l density o f bones. S ub seq u en t
research by b o th th o se in terested in zo o arch aeo lo g ical (B inford a n d B ertram
1977) a n d p aleo n to lo g ical (B ehrensm eyer 1975b; B oaz a n d B ehrensm eyer
1976) p ro b lem s ex p an d ed g reatly o n the results o f the early studies, a n d several
p ro v id ed m easures o f the s tru c tu ra l density o f bones.
In m easu rin g th e stru c tu ra l density o f bones, ta p h o n o m ists have been intent
on p ro v id in g the explicit know ledge necessary to s u p p o rt statem en ts such as
those m ad e in the n in eteen th a n d tw en tieth centuries th a t “ h a rd ” skeletal p arts
ten d to survive c a rn iv o re gnaw ing w hereas “ so ft” p a rts do n o t (see C h a p te r 2).
T h u s B rain (1969) show ed how his m easu rem en ts o f the stru c tu ra l density o f
g o a t b ones co rrelated w ith the frequencies o f skeletal p a rts rem ain in g afte r the
assem blage o f b ones h ad been ravaged by dogs an d people. W hile B rain (1969)
only m easu red the stru c tu ra l density o f eight skeletal p a rts (T able 7.5), th a t w as
sufficient to show the stro n g m ed iatin g influence th a t s tru c tu ra l density has on
the survival o f b ones (F ig u re 7.3). T he co rre la tio n betw een B rain ’s (1969)
m easu rem en ts o f stru c tu ra l density an d the frequencies o f skeletal p a rts th a t
survived th e ra v ag in g is stro n g a n d significant (rs = 0.80, / >= 0.03).
Som e recent research has been fo u n d ed in p a r t o n the arg u m e n t th a t B rain ’s
(1969) research w as flaw ed because it “ lacked system atic co n tro l [of] the exact
frequencies o f b o n e p rio r to rav ag in g by d om estic d o g s” (M a re a n et al.
1992:102). W hile B rain (1969) did n o t in fact re p o rt (a n d a p p a ren tly did not
T ab le 7.5 Frequencies and structural density o f goat bones reported b y Brain

(1 9 6 9 ), and B ehrensm eyer’s ( 1975a) m easures o f sheep bone density
Skeletal p art G o a t bone density F requency Sheep bone density
P hum erus 0.58 33 1.26

D hum erus 0.97 336 1.75
P radius/ulna 1.10 279 1.64
D rad ius/u lna 0.97 114 1.59
P fem ur 0.75 28 1.47
D fem ur 0.72 56 1.42
P tibia 0.82 64 1.32
D tibia 1.17 119 1.64
D humerus
3 0 0
P radius/ulna
CO
c
Q)
E 200
o
Q)
Q.
CO D radius/ulna D tibia
o ■ ■
Z 100
D femur
P humerus ■ ■ P tibia
■ ■ P femur
0 — i— i— i— i— ■— i— ■— i— '— i— ■— i >
0 . 5 0 . 6 0 . 7 0 . 8 0 . 9 1 . 0 1 . 1 1 . 2
Structural Density
Figure 7.3. S catterp lo t o f B rain ’s (1969) g o a t bone stru ctu ral density values
against n u m b er o f recovered g o at bone specim ens from a H o tte n to t village.
kn ow ) th e p re-rav ag in g frequencies o f bones, tw o im p o rta n t p o in ts m u st be

m ade. F irst, the new research confirm s B rain ’s conclusions th a t b ones w ith low
stru c tu ra l densities will m ore often be destro y ed by c a rn iv o re rav ag in g th an
bones w ith high stru c tu ra l densities. S econd, B rain’s research began w ith the
im plicit assu m p tio n th a t the b ones he stu d ied rep resen ted once com plete
skeletons, the a ssu m p tio n all tap h o n o m ists m u st m ak e in studies o f skeletal
p a rt frequencies. T h a t a ssu m p tio n is necessary because ta p h o n o m ists studying
p reh isto ric collections never know w h a t w as originally p resen t “ p rio r to
ra v ag in g .” T he only place to sta rt in studying a skeletal p a rt profile from a
p reh isto ric co n tex t is w ith a m odel o f a set o f com plete skeletons, the original
n u m b er o f com p lete sk eletons typically being determ in ed by the M N I fo r an
assem blage. A s we will see, the a ssu m p tio n m ay well be w ro n g , b u t it is a m a tte r
Frequencies o f skeletal p arts 237
o f s ta rtin g w ith ju s t th is m odel, o r n o n e a t all. U ltim ately , th en , B rain ’s (1969)

p ion eerin g research estab lished a n im p o rta n t an aly tic technique a n d a signifi
c a n t ta p h o n o m ic relation: the s tru c tu ra l density o f skeletal p a rts can exert
stro n g influences o n bon e frequencies.
T h e term stru c tu ra l density m ay seem cum bersom e, b u t it is m a n d a to ry to
discussions o f ta p h o n o m ic issues because the term density can d en o te a
m ass:volum e ra tio such as g /cm 3, or, a freq u en cy :u n it area o r volum e ratio
such as th e n u m b er o f bones p er m 2 o r m 3. S tructural density is m ean t to den o te
the ra tio o f the m ass o f a substance to its volum e. If the m ateria l is
h om o g en eo u s, stru c tu ra l density is a c o n s ta n t o f the substance; if the m aterial
is h etero g en eo u s, s tru c tu ra l density is a n average ch a rac te ristic o f the sam ple
m easu red . T h u s S h ip m an (1981 b:23—25) suggests th a t m an y m easures o f bone
stru c tu ra l density are in fact m o re p ro p e rly labeled m easu rem en ts o f bone
“ c o m p o sitio n ” because bone as a m ateria l is m o re o r less h o m o g en eo u s, b u t its
co m p o sitio n in the sense o f the ra tio o f spongy to co m p ac t bo n e varies w ithin
an d betw een skeletal elem ents. In o th e r w ords, the p o ro sity o f bone varies, an d
hence L y m an (1984a:264), follow ing m ateria ls science, distinguished true
density fro m bulk density as follows:
D, = M /V s [7.1]
D b= M /V t [7.2]
H ere, D t = tru e density, D b = b u lk density, M = m ass, Vs = volum e o f the

su b stan ce exclusive o f p o re space volum e in the m easu red specim en, and
Vt = volum e o f th e sub stan ce including volum e o f the p o re space in the
specim en m easured. A m easure o f to ta l p o ro sity (fp) can be derived from
e q u a tio n s [7.1] a n d [7.2] w ith e q u a tio n [7.3]:
fp—(Dt Db)/Dt —1 (D b/D t) [7.3]
A synthesis an d ev a lu a tio n o f m uch o f the research o n the stru c tu ra l density o f

bones w as p resen ted by L ym an (1984a), w ho show ed th a t the m easu rem en ts o f
density th a t h ad been tak en to th a t tim e tended to be dissim ilar due to
differences in how density h ad been m easured. M easu res o f stru c tu ra l density
o f bo n e w ere vario usly m easures o f bu lk density, tru e density, o r som e hybrid
o f these because p o ro sity a n d o th er im p o rta n t variables th a t influenced
m easu rem en t results h ad been differentially co n tro lled (L y m an 1984a). T h u s it
c a n n o t be ascertain ed if, fo r exam ple, B rain ’s m easu rem en ts o f d om estic g o at
bo ne stru c tu ra l density differ from B ehrensm eyer’s (1975b) m easu rem en ts o f
do m estic sheep (O vis aries) b o n e s tru c tu ra l density (T able 7.5) because o f
differences betw een the tax a o r because o f differences betw een the m easu re
m en t techn iq ues used. T his is a critically im p o rta n t d istin ctio n . T he fact th a t
B ehrensm eyer’s (1975b) density values co rrelate w ith b o th B rain ’s (1969)
density m easures (rs = 0.82, 7^ < 0.03) a n d w ith the frequency o f g o at bones in
T ab le 7.5 (rs = 0.83, Z3 < 0.03) co u ld result fro m eith er the s tru c tu ra l density o f
bones, regardless o f tax o n , being a ro b u st ta p h o n o m ic fa cto r, o r dissim ilar

m easu rem en t tech niq u es p ro d u c in g tw o sets o f values (o f different physical an d
s tru c tu ra l p ro p e rtie s) eith er o r b o th o f w hich are spu rio u sly co rrelated w ith the
frequencies o f surviving bones.
L y m an (1984a) a n d o th er researchers (C h am b ers 1992; E lkin a n d Z an c h e tta
1991; K re u tzer 1992; L ym an et al. 1992a) have used a tech n iq u e called p h o to n
a b so rp tio m e try o r p h o to n d en sito m etry to derive w h a t are called “ b o n e
m ineral d en sities” fo r very specific p a rts o f skeletal elem ents. T he technique
involves th e m easu rem en t o f how w eak a p h o to n b eam o f k n o w n stren g th
becom es w hen it passes th ro u g h the selected p a rt o f a bone. A scan site “ is th a t
a rea o r p a r t o f the b o n e w hich w as ac tu ally m easu red [or scanned] by the
p h o to n b e a m ” (L y m an 1984a:272). T he higher the m ineral c o n te n t o f the
m easu red p a rt o f th e bo n e, the w eaker the b eam is th a t com es th ro u g h the bone
(the few er th e p h o to n s th a t pass th ro u g h it). A p h o to n d etec to r co u n ts the
p h o to n s th a t pass th ro u g h the bone, an d this is co n v erted to a m easu rem en t o f
bo n e m in eral density by a co m p u te r m o d u le atta c h e d to the d e te c to r (see
K re u tzer 1992 a n d L ym an 1984a fo r d escrip tio n s o f the basic m achine). L ym an
(1984a:272-273) chose scan sites (1) th a t w ould allow assessm ent o f k n o w n
s tru c tu ra l v a ria tio n w ithin each bone, (2) th a t w ere easy to locate a n d define on
the basis o f a n a to m ica l features in o rd e r to allow m easu rem en t o f m ultiple
specim ens, a n d (3) th a t tran sected skeletal p a rts co m m o n ly fo u n d in a rc h a e o lo
gical sites. S u b seq u en t researchers have ten d ed to select the sam e scan sites
(e.g., K re u tz e r 1992). W e m ay u ltim ately w ish to m easu re o th e r areas o r p a rts
o f a skeletal elem ent (see below ), b u t fo r n ow the m easu red scan sites are
sufficient fo r m an y analyses.
T h e m easu rem en ts o f bo n e m in eral d en sity are ap p ro x im a tio n s o f bulk
density, a n d are given as g /cm 3 fo r the scan site. U n lik e p rev io u s m easures o f
stru c tu ra l density, th e stru c tu ra l p ro p e rty being m easu red is identical betw een
an aly sts because the m easu rem en t tech n iq u e is identical. T his enhances the
validity o f e x p lan a tio n s fo r different p a tte rn s o f bo n e su rv iv o rsh ip across
different tax a. I f results like th o se in the exam ple ab o v e using B rain ’s density
m easu rem en ts fo r go ats a n d B ehrensm eyer’s m easu rem en ts fo r sheep (T able
7.5) are fo u n d using bone m ineral densities, the reaso n fo r the difference
b etw een th e co rre la tio n coefficients calcu lated betw een b o n e survivorship and
bo n e s tru c tu ra l density a n d th e less th a n perfect c o rrelatio n betw een th e tw o
sets o f den sity m easu res m u st reside in tax o n o m ic v a ria tio n in bo n e stru ctu ra l
d ensity (e.g., K re u tzer 1992).
T he p h o to n d en sito m etry m easu rem en t tech n iq u e is ra th e r inexpensive an d
n o n -d estru ctiv e; one h as b u t to h av e b ones to m easu re a n d a d en sito m eter w ith
w hich to m easure them . M an y such m achines are h o used in S chools o f
M edicine w here they are used to stu d y onto g en ic changes in the m in eral density
o f h u m a n bones; som e m ay be fo u n d in h o sp itals w here they are used as
d iag n o stic tools. T h u s far, bo n e m in eral densities have been m easu red fo r deer
(■Odocoileus spp.), p ro n g h o rn an telo p e (Antilocapra americana), dom estic

sheep (Ovis aries), N o rth A m erican bison (Bison bison), N o rth A m erican
m a rm o ts (M a rm o ta spp.), S o u th A m erican v icuna a n d g u an a co (L a m a spp.),
a n d seals (P hoca spp.). S can sites fo r all tax a th a t have been m easu red are
show n in F igures 7.4, 7.5, a n d 7.6, a n d den sity values are given in T ables 7.6
an d 7.7. B ecause o th e r research indicates th a t the bulk density o f b ones varies
betw een tax a d u e in p a rt to v a ria tio n in lo c o m o to r m odes (e.g., C u rre y 1984;
Stein 1989; W all 1983), ta p h o n o m ists sh o u ld enlarge this list o f m easu red taxa.
T he s tru c tu ra l density values in T ables 7.6 a n d 7.7 can serve m u ch like the
u tility indices d escribed in the preceding section: as a fram e o f reference fo r
assessing th e frequencies o f skeletal p a rts in an assem blage o f bones. T h a t is so
because experim ental d a ta (e.g., B ehrensm eyer 1975b; H aynes 1980b, 1983a;
M arean a n d Spencer 1991; V oorhies 1969) a n d eth n o arch ae o lo g ic al d a ta (e.g.,
B inford a n d B ertram 1977; B rain 1969; W alters 1984, 1985) in d icate th a t
several ta p h o n o m ic processes are m ed iated by the s tru c tu ra l density o f bones.
F o r in stance, th o se b ones w ith the low est bulk density also have the greatest
p o ro sity w hich in tu rn m eans they have the g re atest surface a rea to volum e
ra tio s (the larg er a n d /o r m o re freq u en t the pores p er u n it volum e, the g reater
the surface a rea p er u n it volum e). M ech an ical a n d chem ical a ttritio n should
have g re ater effects o n b ones w ith low b u lk densities (high p o ro sity ) sim ply
because th ere is g re ater surface area to w ork on. A nd, in so far as g reater
p o ro sity reduces w eight p er u n it volum e, w h e th e r o r n o t a b o n e is tra n sp o rte d
by such processes as fluvial ac tio n will be influenced by the bu lk density o f th a t
bone.
T h e an aly tic p ro c ed u re fo r using the bo n e stru c tu ra l density values is a
sim ple one an d is sim ilar to th a t used w ith the utility indices. T he m ajo r
difference is th a t, as originally c o n stru c te d , one tallies the frequency o f each
skeletal p a rt fo r w hich a s tru c tu ra l density value is available. T h u s fo r B rain ’s
(1969) density values, one tallies the M N E o f p ro x im al h um eri, distal hum eri,
etc. fo r th e eight skeletal p a rts listed in T ab le 7.5. F o r the scan sites fo r deer
listed in T ab le 7.6, one tallies how m an y o f each scan site is rep resen ted in the
collection. T ho se frequencies are th en co n v erted to % su rv iv o rsh ip values
based on th e m in im um n u m b e r o f ind iv id u al an im als (M N I) rep resen ted by the
bo n e collection. If, for exam ple, th ere is an M N I o f 8, then if all b o n e p arts
survived th ere sh o u ld be 16 o f each p aire d elem ent such as distal h um eri, 8
skulls, 8 atlas v erteb rae, 8 axis v erteb rae, 40 cervical v erteb rae (5 p er indivi
d u al), 208 ribs (26 p er individual), 64 first ph alan g es, etc. T he % su rv iv o rsh ip
values are deriv ed by dividing th e observed M N E frequency by th e frequency
expected given 100% su rv iv o rsh ip (see the discussion o f e q u a tio n [7.4] below ).
T h u s 4 observed distal hum eri w ould rep resen ted 25% su rv iv o rsh ip w ith an
M N I o f 8 (100% su rv iv o rsh ip = 16). N o te th a t such a n o rm in g p ro c e d u re is n o t
necessary fo r B rain ’s (1969) g o a t b o n e d a ta because all skeletal p a rts are p aired
(T able 7.5); th a t is, all values p lo tte d in F ig u re 7.3 w ould have the sam e relative
view A T 3 A11
O X IS A x (
do r s a l
view A X 2
AX3
cervical
U LI.
HUI UL2 M C I-
HU2 ■R Al MC 2
RA2
HU3 MC 3
R A3
left scapula MC4

lateral v i ew MC 5
MC6 '
posterior view
metacarpal
UL3- a n t e r i o r view
left radius
anterior view left ulna
PI 3 PI2 Pll lateral view
PA I
P2 3 p22
l eft patella first phal on*

anterior view
medial view
second phalanx third phal anx
medial view me d i a l view
FE I
CA4
C A ICA2 CA3
NCI left naviculo

TI 3 cuboid MR 3
F E4
MR 4
MR 5
ASI
TI4 l eft MR6
astragalus metatarsal
- T 15
anterior view
left f e m u r left tibia
anterior view anterior view
Figure 7.4. A natom ical locations o f scan sites w here p h o to n ab so rp tio m etry
m easurem ents have been taken on ungu late bones (from L ym an 1984a:274-275,
Figure 2; courtesy o f A cadem ic Press, Inc.).
axis
lateral view
ATI ATI
Cx
atlas
dorsal view
C LI
L UI
I l e f t da3' vicle
dorsal view
left rib
posterior view
lumbar
RI4
RI5
■MR2
metatarsal dorsal view sternabra
dorsal view ventral view
right calcaneum
dorsal view
right radius
posterior view
right scapula
lateral view right astragalus
right humerus
dorsal view
anterior view
patella
anterior view
B2
- f e f Fl
right ulna right fem ur right fibula right tibia
lateral view posterior view anterior view
anterior view
Figure 7.5. A natom ical locations o f scan sites w here p h o to n ab so rp tio m etry
m easurem ents have been tak en on m arm o t bones (from L ym an et al.
1992a:Figure 4; courtesy o f A cadem ic Press, Ltd.).
left mandi
lateral view lumbar
lateral view lateral view
atlas
cranial view
SC 2
sacrum
dorsal view
left rib
posterior view
axis
lateral view
left radius
medial view
Frequencies o f skeletal p a rts 245
left calcaneum
dorsal view
anterior
right humerus right femur

anterior view posterior view
left astragalus
dorsal view
anterior
right fibula-tibia
anterior view
right innominate
lateral view
Figure 7.6. A nato m ical locations o f scan sites w here p h o to n ab so rp tio m etry
m easurem ents have been tak en on seal bones (after C ham bers 1992).
T ab le 7.6 Average bone m ineral densities f o r deer, pronghorn antelope,
dom estic sheep ( L ym a n 1982b, 1 9 8 4 a ), bison ( K reutzer 1992), guanaco and
vicuna ( E lkin and Z a n ch etta 1991)
scan site bison deer pronghorn sheep guanaco vicun
2&3CP 0.50
5MC 0.62
AC1 0.53 0.27 0.14 0.26 0.22 0.18
AS1 0.72 0.47 0.39 0.54 0.65 0.55
AS2 0.62 0.59 0.48 0.63
AS3 0.60 0.61 0.57 0.60
ATI 0.52 0.13 0.12 0.07 0.17 0.18
AT2 0.91 0.15 0.13 0.11
AT3 0.34 0.26 0.32
AX1 0.65 0.16 0.13 0.13 0.17 0.16
AX2 0.38 0.10 0.11 0.14
AX3 0.97 0.16 0.17
CA1 0.46 0.41 0.29 0.43
CA2 0.80 0.64 0.55 0.58 0.66 0.49
CA3 0.49 0.57 0.50 0.56
CA4 0.66 0.33 0.20 0.43
CE1 0.37 0.19 0.12 0.12 0.24 0.23
CE2 0.62 0.15 0.12 0.13
C U N E IF 0.43 0.72 0.64
DNI 0.53 0.55
DN2 0.61 0.57
DN3 0.62 0.55
DN4 0.53 0.57 0.62
DN5 0.53 0.57
DN6 0.57 0.31
DN7 0.49 0.36
DN8 0.79 0.61
FE1 0.31 0.41 0.16 0.28
FE2 0.34 0.36 0.20 0.16 0.37 0.37
FE3 0.34 0.33 0.21 0.20
FE4 0.45 0.57 0.33 0.36
FE5 0.36 0.37 0.30 0.24
FE6 0.26 0.28 0.27 0.22 0.29 0.23
FE7 0.22
LATMAL 0.56 0.52 0.63
HU1 0.24 0.24 0.06 0.13 0.28 0.23
HU2 0.25 0.25 0.12 0.22
HU3 0.45 0.53 0.25 0.42
HU4 0.48 0.63 0.44 0.37
HU5 0.38 0.39 0.33 0.34 0.40 0.34
HYOID 0.36
IL1 0.22 0.20 0.16 0.23
IL2 0.52 0.49 0.33 0.47
IS1 0.50 0.41 0.28 0.49
IS2 0.19 0.16 0.32 0.11
LU1 0.31 0.29 0.15 0.26 0.26 0.19
LU2 0.11 0.30 0.11 0.22
LU3 0.39 0.29 0.10
LUNA R 0.35 0.83 0.66
MCI 0.59 0.56 0.33 0.40 0.60 0.54
MC2 0.63 0.69 0.41 0.55
T ab le 7.6 (con t.)
scan site bison deer pronghorn sheep guanaco vicuna
MC3 0.69 0.72 0.57 0.67

MC4 0.60 0.58 0.45 0.54
MC5 0.46 0.49 0.40 0.38 0.45 0.39
MC6 0.53 0.51 0.44 0.50
MR1 0.52 0.55 0.47 0.43 0.59 0.50
M R2 0.59 0.65 0.45 0.53
MR3 0.67 0.74 0.57 0.68
MR4 0.51 0.57 0.43 0.51
MR5 0.40 0.46 0.39 0.31 0.43 0.38
MR6 0.48 0.50 0.44 0.39
NCI 0.48 0.39 0.26 0.59 0.42
NC2 0.64 0.33 0.26
NC3 0.77 0.62
PA1 0.31 0.39 0.44
P ll 0.48 0.36 0.24 0.43
P12 0.46 0.42 0.38 0.40 0.65 0.53
P13 0.48 0.57 0.45 0.55
P21 0.41 0.28 0.23 0.34
P22 0.25 0.24 0.39 0.55 0.40
P23 0.46 0.35 0.30 0.42
P31 0.32 0.25 0.25 0.30 0.39 0.17
PUI 0.55 0.46 0.34 0.45
PU2 0.39 0.24 0.25
RA1 0.48 0.42 0.26 0.35 0.41 0.40
RA2 0.56 0.62 0.25 0.36
RA3 0.62 0.68 0.57 0.52
RA4 0.42 0.38 0.30 0.19
RA5 0.35 0.43 0.34 0.21 0.37 0.38
RI1 0.27 0.26
RI2 0.35 0.25
RI3 0.57 0.40 0.37 0.31
RI4 0.55 0.24
RI5 0.33 0.14
SCI 0.27 0.19 0.11 0.20 0.20 0.20
SC2 0.26 0.16 0.25 0.16
SCA PH OID 0.42 0.98 0.68
SP1 0.50 0.36 0.27 0.25 0.38 0.30
SP2 0.48 0.49 0.10 0.33
SP3 0.28 0.23 0.30 0.19
SP4 0.43 0.34 0.15 0.32
SP5 0.17 0.28 0.21
ST1 0.22
TH1 0.42 0.24 0.24 0.14 0.19
TH2 0.38 0.27 0.19
T il 0.41 0.30 0.18 0.16 0.33 0.26
TI2 0.58 0.32 0.26 0.20
TI3 0.76 0.74 0.48 0.59
TI4 0.44 0.51 0.40 0.36
TI5 0.41 0.50 0.29 0.28 0.51 0.42
T R A PM A G 0.52 0.74 0.65
UL1 0.34 0.30 0.28 0.18
UL2 0.69 0.45 0.26 0.26
UL3 0.44
U N C IF 0.44 0.78 0.70
T ab le 7.7 Average bone m ineral densities f o r m arm ots ( L ym a n et al. 1992a)
and ph o cid seals ( C ham bers 1992)
M A RM O T SEAL
scan site density scan site density scan site density scan site density
AC I 0.44 AC1 0.47
AS1 0.71 AS1 0.45 AS2 0.55 AS3 0.56
ATI 0.67 ATI 0.54 AT2 0.42
AX I 0.45 AX1 0.56 AX2 0.49
CA1 0.58 CA1 0.45
CA2 0.84 CA2 0.45 CE1 0.35
CL1 1.09 CU1 0.56
DN1 0.72 DN1 0.59
DN2 0.58 DN2 0.84
DN3 0.59 DN3 0.90
DN4 0.59 DN 4 0.84 DN6 0.89
DN5 0.71 DN5 0.64 DN7 1.11
FE1 0.56 FE1 0.50
FE2 0.73 FF.2 0.53
FE3 0.66 FE3 0.52
FE4 0.70 FE4 0.69
FE5 0.39 FE5 0.45
FE6 0.48 FE6 0.57
FIB1 0.46 FI1 0.39 FI3 0.90 FI4 0.88
FIB2 0.55 FI2 0.78 FI5 0.76
HU I 0.37 HU1 0.43
HU2 0.44 HU2 0.39
HU3 0.62 HU3 0.57
HU4 0.77 HU 4 0.67
HU5 0.62 HU5 0.60
IL1 0.46 IL1 0.60
IL2 0.83 IL2 0.63
IS1 0.58 IS1 0.67
IS2 0.96 IS2 0.75 1S3 0.55
LU1 0.34 LU1 0.38
MR1 0.81
MR2 0.85 NA1 0.57
PAT1 0.83
PI1 0.72
P12 0.79 PU1 0.70
PU1 1.04 PU2 0.71
RA1 0.79 RA1 0.63
RA2 0.97 RA2 0.69
RA3 0.95 RA3 0.71
RA4 0.70 RA4 0.39
RA5 0.51 RA5 0.45
RI1 0.64 RII 0.40
RI2 0.73 RI2 0.50
RI3 0.74 RI3 0.62
RI4 0.79 R14 0.63
RI5 0.54 RI5 0.29
SCI 0.33 SCI 0.43 SC2 0.34
SP1 0.58 SPI 0.49
SP2 0.51 SP2 0.48 SP4 0.63
SP3 0.46 SP3 0.61 SP5 0.41
ST1 0.26 TH1 0.34 TH2 0.37
T il 0.45 T il 0.39
TI2 0.53 TI2 0.47
TI3 0.87 TI3 0.86
TI4 0.74 TI4 0.56
TI5 0.56 TI5 0.48
UL1 0.66 UL1 0.44
UL2 0.99 UL2 0.66
UL3 0.40 UL3 0.35
UL4 0.95 UL4 0.79
Bone Mineral Density
Figure 7.7. S catterp lo t o f % surv iv o rsh ip o f deer skeletal p a rts from 4 5 0 K 4

against bone m ineral density values for deer.
values afte r d eriv a tio n o f % su rv iv o rsh ip because all w ould have been divided
by tw ice th e M N I.
T he % su rv iv o rsh ip values can be p lo tte d ag ain st the density values as in
F ig u re 7.3, a n d a co rre la tio n coefficient betw een those frequencies an d the bone
density values can be calculated. A s an exam ple, the frequencies o f each o f 95
scan sites rep resen ted in an assem blage o f d eer (Odocoileus sp.) bones recovered
from arch aeo lo g ical site 4 5 0 K 4 in eastern W a sh in g to n state are given in T able
7.8, alo n g w ith th eir respective 100% su rv iv o rsh ip values based o n a n M N I o f
15, a n d th eir respective % su rv iv o rsh ip values. T he sc a tte rp lo t o f % su rv iv o r-
ship frequencies fo r this assem blage ag ain st bo n e m in eral density for d eer scan
sites (T ab le 7.6, exclusive o f ca rp als a n d d istal fibula) suggests th a t as bone
m in eral den sity increases, the frequency o f skeletal p a rts increases (F ig u re 7.7).
T he co rre la tio n coefficient betw een the tw o variables is positive a n d significant
(rs = 0.65, P < 0.001), suggesting there is a stro n g re la tio n betw een the tw o
variables. A reaso n ab le co n clu sio n w ould be th a t som e ta p h o n o m ic process
th a t is m ed iated by the stru c tu ra l density o f the skeletal p a rts h ad a m ajo r
influence o n the frequencies o f skeletal p a rts. Such processes include, b u t are
n o t lim ited to, c o n su m p tio n by ca rn iv o res o f low -density skeletal p arts
(typically called c a rn iv o re a ttritio n o r rav aging), h u m a n c o n su m p tio n o f low
density skeletal p a rts, differential fra g m e n ta tio n leading to less dense bone
p a rts being cru sh ed in to u n identifiable p o w d e r in c o n tra s t to denser b o n e p arts
being sim ply b ro k e n in to sm all b u t recogn izable pieces d u rin g the e x tra ctio n o f
grease a n d /o r m arro w , p o st-d ep o sitio n al cru sh in g by sedim ent o v erb u rd en , o r
som e c o m b in a tio n o f these.
T able 7.8 Frequencies o f representation o f scan sites o f deer bones fr o m
archaeological site 4 5 0 K 4 (fro m L ym a n 1982b)
Scan N N % Scan N N 0/
/o
site obs. exp. survivorship site obs. exp. survivorship
DN1 10 30 33 DN2 10 30 33
DN3 12 30 40 DN4 13 30 43
DN5 10 30 33 D N6 13 30 43
DN7 14 30 47 DN8 5 30 17
A TI 0 15 0 AT2 2 15 13
AT3 3 15 20 AX 1 0 15 0
AX2 0 15 0 AX3 0 15 0
CE1 7 75 9 C E2 7 75 9
TH1 2 195 1 TH2 9 195 5
LU1 7 105 7 LU 2 12 105 11
LU3 3 105 3 SCI 2 15 13
SC2 0 15 0 SP1 13 30 43
SP2 11 30 37 SP3 6 30 20
SP4 5 30 17 SP5 2 30 7
HU1 2 30 7 HU2 12 30 40
H U3 13 30 43 HU4 26 30 87
H U5 5 30 17 UL1 4 30 13
U L2 6 30 20 UL3 0 30 0
RA1 4 30 13 RA2 9 30 30
RA3 19 30 63 RA4 3 30 10
RA5 4 30 13 M CI 6 30 20
M C2 6 30 20 M C3 20 30 67
M C4 7 30 23 M C5 3 30 10
M C6 16 60 27 R11 6 390 2
RI2 17 390 4 RI3 36 390 9
RI4 35 390 9 RI5 21 390 5
ST1 4 105 4 IL1 3 30 10
IL2 13 30 43 AC1 7 30 23
IS1 3 30 10 IS2 0 30 0
PU1 30 23 PU2 0 30 0
FE1 4 30 13 FE2 1 30 3
FE3 13 30 43 FE 4 20 30 67
FE5 6 30 20 FE 6 1 30 3
PA1 1 30 3 T il 0 30 0
TI2 14 30 47 TI3 15 30 50
TI4 9 30 30 TI5 6 30 20
AS1 18 30 60 AS2 24 30 80
AS3 26 30 87 NCI 3 30 10
NC2 2 30 7 N C3 5 30 17
CA1 2 30 7 CA2 1 30 3
CA3 4 30 13 CA 4 6 30 20
MR! 3 30 10 MR2 10 30 33
M R3 29 30 97 MR4 8 30 27
M R5 3 30 10 M R6 13 60 22
P ll 16 120 13 P12 7 120 6
P13 15 120 12 P21 3 120 2
P22 2 120 2 P23 3 120 2
P31 2 120 2
co
Bone Mineral Density

Figure 7.8. S eatterp lo t o f frequency o f individual scan sites in one skeleton
against bone m ineral density values fo r deer.
It is im p o rta n t to n o te th a t, w hile the M N E frequencies o f scan sites in the

4 5 0 K 4 assem blage co rrelate w ith the s tru c tu ra l density o f the scan sites
(rs = 0.48, P < 0.001), th e conversion o f th o se frequencies to % su rv iv o rsh ip is a
critical step in the analysis. It is critical because o u r in terest lies n o t in how
m an y b o n es th ere are, b u t in how m an y survived fro m th e original carcasses,
an d a m am m alian carcass has different M N E frequencies o f different skeletal
elem ents. A n o b v io us exam ple fo r artio d a cty ls such as deer is th a t they have 2
h um eri (one left, one right, b o th in the fro n t legs) b u t 8 first ph alan g es (2 per
fo ot, w ith 2 hindfeet a n d 2 forefeet). G iven an interest in the n u m b er o f bone
p a rts per carcass th a t survived, the observed frequencies m u st be co n v erted to
% su rv iv o rsh ip values so th a t, fo r instance, a g re ater a b u n d a n c e o f first
p h alan g es th a n h u m eri reflects differential su rv iv o rsh ip ra th e r th a n the differ
en t n u m b ers o f skeletal p a rts p er living individual. A se a tte rp lo t o f bone
m ineral density values a g ain st the M N E frequency o f scan sites in one com plete
skeleton (F ig ure 7.8) is suggestive o f a n inverse o r negative re la tio n betw een the
tw o variables. T h ere is no significant c o rrelatio n betw een these tw o variables
(rs= —0.06, P = 0.56), b u t the se a tte rp lo t show s th a t the tendency w ould be for
sca tte rp lo ts w ith M N E values o n the y-axis to a p p e a r as inverse relatio n s if
these values w ere n o t co n v erted to % su rv iv o rsh ip values. T h a t th e 4 5 0 K 4
M N E values are im perfectly co rrelated w ith the % su rv iv o rsh ip values
(rs = 0.76, P < 0.001) u n d ersco res the difference betw een M N E frequencies an d
% su rv iv orship values. A significant c o rre la tio n m ay be fo u n d betw een the
M N E values fo r an assem blage a n d b o n e density, a n d n o t betw een the
% su rv iv o rsh ip values a n d b o n e density, b u t the ta p h o n o m ic significance o f the

fo rm er w o u ld be unclear.
I believe th e stru c tu ra l density values listed in T ables 7.6 a n d 7.7 are at best
o rd in al scale. T he reaso n in g fo r this belief is identical to th a t used w ith the
tra n s p o rt a n d u tility indices. T h e s tru c tu ra l density values are averages o f
different in d ividuals, essentially all skeletally m atu re (e.g., all epiphyses are
fused) b u t o f different ages, sexes, n u tritio n a l statuses, a n d genetic p o p u latio n s.
It is well k n o w n th a t the stru c tu ra l density o f bones varies w ith age, sex,
n u tritio n a l statu s, a n d genetics, a n d different bones will vary differently. T h u s
the d istal h u m eru s m ay be 40% m o re dense th a n the p ro x im al h u m eru s in one
in d iv id u al b ut only 20% m ore dense in a n o th e r individual. A gain, because we
c a n n o t yet d eterm in e th e sex, onto g en ic age, a n d n u tritio n a l statu s o f the
in d iv id u al rep resen ted by m o st o f the b ones we find, it is best to tre a t the
stru c tu ra l density m easures as o rd in al scale a n d to use o rd in a l scale statistics
w hen searching for evidence o f d en sity -m ed iated attritio n .
A b o v e it w as suggested th a t, given the stro n g statistical re la tio n betw een the
% su rv iv o rsh ip o f scan sites in the 4 5 0 K4 assem blage a n d the s tru c tu ra l density
o f th e scan sites, it seem ed reaso n ab le to conclude th a t this assem blage had
u n d erg o n e one o r m o re d en sity -m ed iated a ttritio n a l processes. By density-
m ediated attrition is m ean t the loss o f skeletal p a rts due to th eir stru ctu ra l
density; typically skeletal p a rts w ith low s tru c tu ra l density are lost b ecause they
are m o re easily destro y ed (b u t see the discussion o f b o n e to o ls in C h a p te r 8).
N o specific a ttritio n a l agent is identified, b u t ra th e r only th o se ta p h o n o m ic
processes th a t are buffered o r m ed iated by the s tru c tu ra l density o f skeletal
p a rts are im plicated. T h u s, an y one o r m o re o f a n u m b e r o f d ensity-m ediated
ta p h o n o m ic processes m ight be responsible fo r the frequencies o f skeletal p a rts
in th e 4 5 0 K 4 deer b o n e assem blage, a lth o u g h those th a t are n o t density
m ed iated c a n n o t be com pletely elim inated (see below ). Several density-
m ed iated processes w ere n o ted in the discussion o f the 4 5 0 K 4 assem blage.
O thers include fluvial tra n s p o rt, w hich ten d s to so rt (rem ove) low -density
bones (the w innow ed o r rem oved p o rtio n o f the assem blage) from th o se o f high
density; th e la tte r m ay rem ain (a lth o u g h th eir lo catio n m ay be changed) as a lag
dep o sit (see T ab le 6.5 a n d F ig u re 6.5). G n a w in g by ro d e n ts ten d s to focus on
the densest skeletal p a rts (e.g.. B rain 1980; M o rla n 1980). H u m an selection o f
dense b on es fo r to o l m a n u fa c tu re (see C h a p te r 8), su b aerial w eathering, an d
diagenetic processes th a t m o re rapidly affect skeletal p a rts w ith low stru ctu ra l
densities th a n skeletal p a rts w ith high stru c tu ra l densities (e.g., L ym an a n d Fox
1989) are o th e r possibilities, alth o u g h no n e o f these has been studied actualisti-
cally. O bviously, d eterm in in g w h eth er a p a rtic u la r assem blage o f bones
u n d erw en t som e d en sity -m ed iated ta p h o n o m ic process involves m ak in g scat
terp lo ts a n d calcu latin g a c o rre la tio n coefficient betw een values o f skeletal p a rt
stru c tu ra l density a n d % su rv iv o rsh ip . B ut it does n o t end there, as a n exam ple
will m ak e clear.
L y m an et al. (1992a) arg u e th a t n eith er the utility indices n o r the stru c tu ra l
T ab le 7.9 M A U values fo r the W hite M ountains m arm ots (fro m Grayson

1989) and the Salishan M esa m arm ots (L y m a n n .d .b ), and corresponding
scan sites f o r structural density values; skeleta l p a rt abbreviations are used in
Figure 7.10
T rad itio n al M axim um

Skeletal part W hite M o u n tain s Salishan M esa scan site scan site
Skull (SK) 55.50 13.00 none none

M andible (M A N D ) 84.00 23.50 DN3 DN1
A tlas (AT) 7.00 4.00 A TI ATI
Axis (A X) 3.00 1.00 AX1 AX1
Inno m in ate (PELV ) 8.00 3.50 AC1 PU1
S capula (SCA P) 24.00 5.50 SP1 SP1
P hum erus (PH ) 12.50 3.00 HU1 HU2
D hum erus (D H ) 33.50 7.50 HU5 HU4
P radius (PR ) 52.00 9.00 RA1 RA2
D radius (D R ) 28.50 1.00 RA5 RA4
P ulna (PU ) 65.50 13.00 UL1 U L2
D ulna (D U ) 12.00 2.50 U L3 U L3
C arpals (C A R P) 7.57 n.d. none none
P m etacarp al (PM C ) 3.70 n.d. none none
D m etacarpal (D M C ) 0.50 n.d. none none
P fem ur (P F) 9.00 6.00 FE2 FE2
D fem ur (D F ) 4.50 2.00 FE6 FE 6
P tibia (PT) 3.50 1.00 T il TI2
D tibia (D T) 58.00 8.00 TI5 T I4
A stragalus (AST) 64.00 4.50 AS1 AS1
C alcaneum (C A LC ) 14.50 1.50 CA2 CA2
O ther T arsals (T A R S) 3.86 n.d. none none
P m etatarsal (PM T) 5.50 n.d. MR1 MR1
D m etatarsal (D M T ) 1.40 n.d. MR2 MR2
D m etapodial (D M P ) 1.25 n.d. none none
Phalanges (Ph) 1.98 n.d. none none
d ensity values o f skeletal p a rts sh o u ld be used alone to in te rp re t skeletal p a rt

frequencies. T o do so w ould result in ta c it acceptance o f a sim ple c o rrelatio n
d en o tin g a cau sal relatio n betw een tw o variables th a t m ight n o t in fact be
causally related. T h a t is, a c o rre la tio n betw een bo n e frequencies an d bone
stru c tu ra l density is a necessary c o n d itio n fo r inferring a causal relatio n
betw een th e tw o variab les, b u t it is n o t a sufficient c o n d itio n fo r such an
inference. L y m an et al. (1992a) exam ine tw o assem blages o f yellow -bellied
m a rm o t (M a rm o ta flaviventris) bones (T able 7.9) a n d a tte m p t to explain why
one o f th em is a n d th e o th e r is n o t co rrelated w ith the stru c tu ra l density values
for m a rm o t b ones (T able 7.7). T hey fo u n d th a t the M A U values fo r the
m a rm o t b o n e assem blage fro m the W hite M o u n ta in s o f C alifo rn ia (G ray so n
1989') are n o t co rrelated w ith m a rm o t bo n e densitv (rc = 0.3. P = 0.28). as
100 -
80 -
60 -
40 -
20 -
0-
0.3 0.4 0.5 0.6 0.7 0.8 0.9
Density
Figure 7.9. S catterp lo t o f M A U frequencies o f m arm o t skeletal p arts from the
W hite M o u n tain s against b one m ineral density values fo r m arm ots.
im plied in F ig u re 7.9, suggesting th a t d en sity -m ed iated a ttritio n w as no t

responsible for th e vary in g frequencies o f skeletal p a rts (note the vertical
sca tte r o f p o in ts in F ig u re 7.9). T he W hite M o u n ta in s m a rm o t bones were,
how ever, q u ite frag m en ted (G ray so n 1989). T h u s, L ym an et al. (1992a) suggest
th a t differential fra g m e n ta tio n influenced the skeletal p a r t profiles. T h ey call
u p o n th e prem ise th a t sm aller frag m en ts will be less identifiable because they
are less likely to have d iag n o stic la n d m a rk s (e.g., L ym an a n d O ’Brien 1987).
T he m a jo r test im p licatio n o f this hy p o th esis is th a t ra re k in d s o f skeletal p a rts
sh ou ld have sm aller average p roportional com pleteness values (% o f a com plete
specim en rep resen ted by a frag m en t) th a n the m o re co m m o n kinds o f skeletal
p arts. W hile frag m en t size d a ta are u n av a ila b le fo r the W hite M o u n ta in s
m aterials, a n d th u s the h y pothesis c a n n o t be tested, w h a t is im p o rta n t here is
th a t the in itial p ro jec tio n o f the m a rm o t bo n e frequencies on the fram e o f
reference o f b o n e s tru c tu ra l density directs us to o th e r kinds o f d a ta th a t m ay
help a c c o u n t fo r the vary ing frequencies o f skeletal p arts. L ym an et al. (1992a)
fo u n d a significant c o rre la tio n betw een s tru c tu ra l density a n d the M A U values
o f m a rm o t bones fro m the S alishan M esa site in eastern W a sh in g to n (rs = 0.46,
P = 0.06), as im plied in F ig u re 7.10. T h o se m a rm o t bones seem to be less
frag m en ted th a n the W hite M o u n ta in s bones. B etw een these tw o assem blages,
th en , a q u estio n is raised a b o u t how stru c tu ra l density m ed iates frag m en tatio n
a n d th e re su lta n t id en tifiability o f fa u n al rem ains, a to p ic to u ch ed o n in
C h a p te rs 6 a n d 8.
It m u st be p o in ted o u t th a t the % su rv iv o rsh ip value is the sam e v ariab le as
th e % M A U value discussed in th e preceding section o f th is ch ap ter. This
■ MAND
20
■ pu
10
■ PR
DT" ■ m
■ PF
PELV SCAP ■ ■ AST
PH ■ - " AT C ALC
DU .P T .C R ■ CF ■
0 t-------' ------ 1
------ >------ 1-------1------ 1
------ <------ r
0.3 0.4 0.5 0.6 0.7 0.8 0.9
D en sity
Figure 7.10. S catterp lo t o f M A U frequencies o f m arm o t skeletal p a rts from the

Salishan M esa site ag ain st bone m ineral density values for m arm ots.
equivalence can be show n as follow s. B rain (1969, 1976) originally calculated

% su rv iv o rsh ip in an assem blage as
(M N E l) 100________ _
M N I (nu m ber o f tim es ■
, occurs in one skeleton)
T he d e n o m in a to r in e q u a tio n [7.4] tells us how m an y o f skeletal p o rtio n , to

expect given 100% su rv ivorship, a n d it is equal to th e m ax im u m M N E value
observed in an assem blage. It can th u s be w ritten as
(M N E,) 100_________
m axim um M N E in the assem blage
E q u atio n [7.4] (a n d by im p licatio n eq u a tio n [7.5]) is my re co n stru ctio n o f

B rain 's (1967b. 1969, 1976) p ro c ed u re , as he did n o t use the term s M N E o r
M N I. B rain (1967b:4), for exam ple, describes the bo n e assem blage he studied
as vario u sly con sisting o f long b o n e “ distal en d s” a n d “ pieces” o f long bone
ends a n d o th e r skeletal elem ents a n d skeletal p o rtio n s (in the senses defined in
C h a p te r 4). H e gives a g ood d escrip tio n o f the d eriv a tio n a l p ro c ed u re he uses
w hen he w rites “ in the case o f ribs, 26 o f w hich are fo u n d in a single g o at
skeleton, th e original n u m b e r c o n trib u te d by 64 [ = M N I] go ats m u st h ave been
1664 [ = 6 4 x 2 6 ]. O nly 170 [ = M N E ] have been fo u n d , in d icatin g a 10.2%
survival” ( B rain 1969:18). I presum e B rain (1967b:3; 1969:17) used a q u a n tita
tive u n it eq u iv alen t to M N E in his p ro c ed u re fo r m easu rin g % su rv iv o rsh ip
b ecause he gives the n u m b e r o f rib “ p a rts ” as 174. But, (170 h- 1664)
1 0 0 = 10.2% . Sim ilarly, the N IS P fo r m andibles is given by B rain as 188, w h at I
believe to be the M N E is given by him as 117, a n d the % su rv iv o rsh ip is given as

91.4% ( = [1 1 7 -^ (2 x 6 4 )] 100).
As defined by B inford (1978, 1981b, 1984b), % M A U is derived w ith the
e q u a tio n
[M A U J 100_________ [76]
m axim um M A U in the assem blage
N o te th a t M A U , is derived w ith the e q u a tio n

__ ___ MNEj
[7.7]
num ber o f tim es j occurs in one skeleton
S u b stitu tin g e q u a tio n [7.7] in to e q u a tio n [7.6],

[MNE, -r num ber of times j occurs in one skeleton] 100
[7.8]
maximum M N E in the assemblage —num ber o f times that skeletal portion occurs in one skeleton
T he d e n o m in a to r in eq u a tio n s [7.6] and [7.8] tells us how m an y to expect given

100% survival, a n d is d ep e n d en t o n the assem blage u n d er study. T he im p o rt
a n t p o in t to n ote is th a t the expression “ - n u m b e r o f tim es j occurs in one
sk eleto n ” in the n u m e ra to r a n d d e n o m in a to r in e q u a tio n [7.8] cancel each
o th e r o u t, a n d w hen they d o , eq u a tio n s [7.5] a n d [7.8] are identical.
T h a t eq u a tio n s [7.4] an d [7.5], an d e q u a tio n s [7.6] a n d [7.8], are m easuring
the sam e p ro p e rty can be illu strated by using B rain ’s d a ta for the rib a n d fo r the
m an dible. F o r the rib, e q u a tio n [7.8] is solved as
f '7 0 - 2 6 ] 100 = [6.538]
1664 -f- 26 64 1' J
F o r the m an dible, e q u a tio n [7.8] is solved as

[117- 2 ]J 0 0 = [58.5] 100
1 2 8 -2 64 1' J
T hese are precisely the results fo u n d w hen using B ra in ’s p ro c ed u re , o r

e q u a tio n s [7.4] a n d [7.5]. T h u s it should be clear th a t % su rv iv o rsh ip is
m easu red precisely the sam e w ay as % M A U ; b o th variables m easure the sam e
p ro p e rty , a n d are one a n d the sam e.
Recall th a t the tra n s p o rt a n d utility indices described in the p receding section
are sim ply one “ fram e o f reference” (B inford 1987) th a t can be used as a single
step in co n stru c tin g ex p lan a tio n s for the frequencies o f skeletal p arts. T he
stru c tu ra l density values are a n o th e r fram e o f reference a n d th u s a n o th e r step
in such co n stru c tio n s. T h us, it is reaso n ab le to ask if the % M A U values for
ca rib o u rem ains at th e A navik site (T able 7.2) co rrelate w ith the stru c tu ra l
density o f bo n e p a rts. W e as yet lack stru c tu ra l density values fo r ca rib o u , bu t
we can use the density values for deer in T ab le 7.6 as th a t species is confam ilial
w ith ca rib o u , a n d th u s is the closest genetic relative fo r w hich we have density
values. T h en , we m u st decide w hich p a rtic u la r density values to use because the
Frequencies o f skeletal p arts 257
T ab le 7.10 T raditional density scan sites and m a xim u m density scan sites
typically correlated with M A U values
MAU T rad ition al M axim um MAU T rad itio n al M axim um

category scan site scan site category scan site scan site
m andible DN4 DN8 innom inate AC1 IL2

atlas A TI AT3 P fem ur FE2 FE1
axis AX1 AX1 D fem ur FE6 FE6
cervical CE1 CE1 patella PA1 PA1
thoracic TH1 TH 2 P tibia T il T il
lum bar LU1 LU2 D tibia TI5 TI5
sacrum SCI SCI astragalus AS1 AS3
rib R13 RI3 calcaneum CA2 C A2
sternum ST1 ST1 naviculo-cuboid NCI NC3
scapula SP1 SP2 P m etatarsal MR1 MR1
P hum erus HU1 HU2 D m etatarsal MR6 MR6
D hum erus HU5 HU5 F irst phalanx P12 P13
P radius RA1 RA2 Second phalanx P22 P23
D radius RA5 RA5 T h ird phalanx P31 P31
carpal none S C A P H O ID
% M A U values are fo r skeletal p a rts such as p ro x im al h um eri an d there are tw o

places on the p ro x im al h a lf o f the hum eri fo r w hich we have density m easures,
scan sites H U 1 a n d H U 2 (F ig u re 7.4). In m y original w ork w ith this pro b lem
(L y m an 1984a), I selected the scan site th a t seem ed best to typify the stru ctu ra l
density o f a skeletal p a rt. O th e r researchers have suggested using the m axim um
density value reco rd ed for a p a rtic u la r M A U skeletal categ o ry (G ifford-
G o n zalez a n d G a rg e tt n.d.; R ap so n 1990). I (L ym an 1992a) subsequently
referred to the fo rm er set o f values as the traditional density values, a n d the
latter as th e m axim u m density values, b o th o f w hich are listed in T able 7.10.
T h e A n av ik ca rib o u % M A U values are n o t co rrelated w ith the tra d itio n a l
deer b o n e density values (rs = - 0.003, P = 0.93) o r the m axim um density values
(/•s = 0.06, P = 0.75). U sing only these tw o fram es o f reference, utility indices
an d stru c tu ra l density values, th en , m ight p ro m p t us to conclude th a t the
frequencies o f ca rib o u skeletal p a rts a t A n a v ik are a ttrib u ta b le to the o p eratio n
o f a p a rtic u la r tra n s p o rt strateg y practiced by the h u m a n o ccu p an ts o f the site
a n d n o t to d en sity -m ed iated a ttritio n . O ne m ig h t suspect such a case w ould be
easily in terp re ted w hen fo u n d in an archaeological context. Sim ilarly, the
ca rib o u rem ains B inford (1978:323) recovered from the eth n o arch aeo lo g ical
In gsted site (T able 7.2) are co rrelated w ith the s tru c tu ra l density o f deer bones
(rs = 0.38, P = 0.05 fo r b o th tra d itio n a l an d m axim um density values) b u t n o t
w ith th e ca rib o u % M G U I (rs = 0.01, P = 0.9). T he frequencies o f ca rib o u bones
at the Ingsted site seem to be a ttrib u ta b le to d ensity-m ediated a ttritio n a l
facto rs ra th e r th a n to a p a rtic u la r tra n s p o rt strateg y follow ed by h u m an
h u n ters. A gain, one m ig h t suspect sim ilar statistically clear results fo u n d in

arch aeo lo g ical settings w ould be readily in terp reted . U n fo rtu n a te ly , as any
serious stu d e n t o f ta p h o n o m y know s, n o t all zo o a rch aeo lo g ical cases are so
tra n sp a re n t.
Differential transport versus differential survivorship
The problem
In te rp re ta tio n s derived from the tw o variables m easured by the tra n s p o rt or
utility indices a n d the stru c tu ra l density values require d a ta on bone freq u en
cies, typically % M A U o r % su rv iv o rsh ip values, w hich we have seen are in fact
the sam e. G iven how in te rp re ta tio n s are derived from , fo r exam ple, sca tte r
p lots like th o se in F ig u res 7.2 a n d 7.3 a n d th eir associated (o rd in a l scale)
co rre la tio n coefficients, the p ro b a b ility th a t a skeletal p a rt will be tra n sp o rte d
an d utilized by h u m an s m u st be in d ep en d e n t o f (m ust n o t co rrelate w ith) the
p ro b a b ility th a t this skeletal p a rt will survive d en sity -m ed iated a ttritio n a l
processes. I f these tw o v ariables are som ehow in te rd e p e n d e n t o r are sim ply
co rrelated (not necessarily causally related), then significant c o rrelatio n s
betw een b o n e frequencies a n d b o th the tra n sp o rt o r utility indices an d
stru c tu ra l density m ay result, a n d the an a ly st will be faced w ith a case o f
equifm ality. T h a t is, sim ply using the an aly tic pro ced u res described th u s far
will result in the research er n o t know ing w h eth er to call u p o n differential
tra n s p o rt an d utility o r differential d e stru c tio n as the ex p lan a tio n fo r the
frequencies o f skeletal p a rts observed in a bo n e assem blage.
Several years ago I show ed th a t th ere is a w eak negative c o rrelatio n betw een
th e b o n e m in eral density values fo r d eer a n d several o f the utility indices listed
in T ab le 7.1 (L y m an 1985a). I su b sequently show ed th a t m an y o f the indices
listed in T ables 7.3 a n d 7.4 are also negatively c o rrelated w ith the bone m ineral
density values fo r d eer (L ym an 1992a). T hese results p ro m p t several co n clu
sions. F irst, b ones th a t have low s tru c tu ra l densities tend to ran k high in utility
an d b o n es th a t h av e high stru c tu ra l densities ten d to ra n k low in utility. (N o te
th a t I use the term “ u tility ” in the re m a in d er o f this discussion to d en o te b o th
econom ic u tility a n d the p ro b a b ility th a t a skeletal p a rt will be tra n sp o rte d ;
utility is th u s a sh o rth a n d fo r b o th econom ic utility a n d tra n s p o rt indices.)
T h u s, b o th d en sity -m ed iated d estru c tio n a n d a reverse u tility strateg y (F igure
7.1a) will p ro d u c e a negative h yp erb o lic (L -sh ap ed ) curve w hen % M A U values
are p lo tte d ag ain st % M G U I values. S econd, n eith er the % M G U I n o r the
s tru c tu ra l density values, alone o r in c o m b in atio n , will pro v id e tru stw o rth y
bases fo r e x p lan a tio n s o f skeletal p a rt frequencies. O th e r kinds o f evidence are
necessary to help so rt o u t w hich process(es) (tra n s p o rt o r d estru c tio n ) w as
respo nsib le for a p a rtic u la r b o n e assem blage (L y m an 1985a, 1991c, 1992a).
T he p ro b lem is well illu strated by the g u an a co general utility index an d the
m odified g u an aco u tility index (T able 7.4). B oth indices are negatively c o rre
D en sity
100 -
ZD
ro
CD
c
03
0
T3
.2?
t5
o
0.1 0.2 0.3 0.4 0.5 0.6 0.7

D en sity
Figure 7.11. a, S catterp lo t o f the general g uanaco utility index against guanaco
bone density, b, S catterplot o f the m odified g u anaco utility index against
g uanaco b one density.
lated w ith the s tru c tu ra l density o f g u an a co skeletal p a rts (T able 7.6).

S catterp lo ts o f the g u an a co general utility index against b o n e density
(rs = —0.71, P = 0.0005; F ig u re 7.11a) a n d o f the g u an aco m odified utility index
ag ain st b o n e den sity (rs = —0.55, P = 0.004; F ig u re 7.11b) clearly show the
relatio n . T h e fact th a t g u an aco bones o f low density tend to ra n k high in utility
w hile g u an a co b on es o f high density tend to ra n k low in utility m eans, w hen a
significant negative co rre la tio n betw een % M A U values fo r an archaeological
co llection o f g u an a co rem ains a n d the utility index fo r g u an aco is fo u n d , the
analyst m ay also find a significant positive c o rrelatio n betw een th o se % M A U
T ab le 7.11 % M A U frequencies f o r deer-sized

anim als f o r site 45C H 302 (fro m L ym a n n.d.a)
Skeletal p art %M AU Skeletal p a rt %M AU
m andible 100.0 P ulna 25.0

atlas 37.5 P m etacarp al 43.8
axis 37.5 D m etacarp al 68.8
cervical 13.75 P fem ur 12.5
thoracic 9.6 D fem ur 6.2
lum bar 44.4 P tibia 3.1
innom inate 28.1 D tibia 78.1
sacrum 6.2 naviculo-cuboid 53.1
rib 27.5 astragalus 96.9
sternum 0.9 calcaneum 34.4
scapula 34.4 P m etatarsal 59.4
P hum erus 6.2 D m etatarsal 56.2
D hum erus 62.5 first phalan x 78.1
P radius 53.1 second phalan x 34.4
D radius 21.9 th ird phalan x 17.2
frequencies a n d th e s tru c tu ra l density o f g u an a co bones. A gain, in such cases

the in terp retiv e p ro b lem is deciding w h eth er the bone frequencies are the result
o f differential tra n s p o rt a n d utilizatio n by people, differential, density-
m ed iated d estru c tio n , o r a co m b in a tio n o f the tw o.
A n exam ple will m ak e the significance o f the preceding p ro b lem clear. T he
late H o lo cen e-ag ed site o f 45C H 302 in eastern W a sh in g to n p ro d u c ed a large
sam ple o f deer (O docoileus sp.), b ig h o rn sheep (Ovis canadensis), a n d p ro n g
h o rn an telo p e (A ntilocapra am ericana) rem ains (L ym an n.d.a). B ecause these
three ta x a overlap in size, som e o f th eir b ones c a n n o t be identified to tax o n
(such as ribs a n d m o st v erteb rae), a n d m an y o th e r bones c a n n o t be identified to
tax o n if the specim ens are incom plete o r b ro k en . I tallied the M N E o f all
rem ain s fo r these three tax a, plus all rem ains th a t could n o t be identified to
tax o n b u t th a t could be identified as a m em b er o f this size class o f m am m als if
th e skeletal p a rt rep resen ted by a specim en co u ld be d eterm in ed (T able 7.11).
T he % M A U frequencies fo r the 45C H 302 assem blage o f deer-sized b o n es n o t
only co rrelate stro n g ly w ith the stru c tu ra l density o f d eer bones (rs = 0.67,
P < 0.001), b u t also w ith b o th the ca rib o u % M G U I (rs= —0.44. P < 0 .0 5 ) and
the sheep % M G U I (rs= —0.55, P < 0.005). T hese relatio n s are show n in F igure
7.12, a n d the in terp retiv e p ro b lem should be a b u n d a n tly clear fro m these
g raphs: is th e assem blage o f deer-sized rem ains fro m 45C H 302 the result o f
differential u tilizatio n a n d tra n s p o rt by h u m an s, d en sity -m ed iated a ttritio n , o r
b oth? T h e co rre la tio n coefficients a n d the sc a tte rp lo ts d o n o t help answ er this
qu estion .
In atte m p tin g to p ro d u c e a n answ er to the preceding q u estio n , G ra y so n
(1988:70-71) suggested th a t w hen d en sity -m ed iated d e stru c tio n w as the
responsible process a n d differential tra n s p o rt w as n o t, th e statistical re latio n

betw een % M A U a n d bo n e density will be positive a n d significant (typically,
P < 0 .0 5 ) while the statistical relatio n betw een % M A U a n d % M G U I will be
insignificant (typically, P > 0 .0 5 ) . C onversely, G ra y so n arg u ed , differential
tra n s p o rt co uld be inferred if % M A U a n d % M G U I are significantly an d
positively co rrelated b u t % M A U a n d stru c tu ra l density are insignificantly
c o rrelated . T his p ro c ed u re o f calcu latin g a n d co m p arin g co rre la tio n coeffi
cients w as used by K lein (1989) in an effort to show th a t rem ains o f b ovids o f
different size classes recovered from the A frican K lasies R iver M o u th site were
vario u sly a ttrib u ta b le to d ensity-m ediated a ttritio n a n d n o t differential tra n s
p o rt. G ra y so n (1988:71) n o ted th a t o th er co m b in atio n s o f statistical c o rre la
tions “ will be m ore difficult to in te rp re t.”
All possible co m b in atio n s o f statistical c o rrelatio n s betw een % M A U values
an d b o n e density, a n d betw een % M A U values a n d % M G U I are show n in
F ig u re 7.13. I have given each possibility a nam e: class 1, class 2, class 3, etc.
(L ym an 1991c). G ra y so n 's (1988) so lu tio n s to the equifinality p ro b lem no ted
abo v e are in d icated (C lasses 2 a n d 4), a n d o th e r possible in te rp re ta tio n s o f
vario us o f the possible classes are also n o ted in F ig u re 7.13. W h a t is a
ta p h o n o m ist to d o w ith this ran g e o f possibilities? T o assess the significance
an d m ag n itu d e o f this q u estio n , I calculated c o rre la tio n coefficients betw een
bo n e density values in T able 7.6 a n d the % M A U values, a n d utility indices in
T ables 7.1 a n d 7.4 a n d % M A U values, fo r a sam ple o f 184 eth n o arch ae o lo g ic al
and arch aeo log ical assem blages o f a rtio d a c ty l bones. N inety-seven o f the
assem blages com e fro m N o rth A m erica, one from S o u th A m erica, 44 from
A frica, 30 from E u ro p e, 10 fro m the N e a r E ast, a n d one each from the
C arib b ea n a n d A sia. F o rty -o n e are eth n o arch ae o lo g ic al, 90 are H olocene-
aged, 52 d ate to the P leistocene a n d one is pre-P leistocene in age. W hen taxa
w ere sm all o r co nfam ilial w ith deer, I used the deer (Odocoileus sp.) density
values a n d the ca rib o u a n d sheep % M G U I values; w hen tax a w ere large o r
were b ison, I used the b iso n (Bison bison) density values a n d the bison to ta l
p ro d u c ts index. I th en identified each set o f coefficients p er assem blage as one o f
the classes, initially keeping eth n o arch ae o lo g ic al assem blages sep a rate from
arch aeo log ical ones (F ig u re 7.13).
C o n sid erin g only th o se c o rre la tio n coefficients for w hich P < 0 .0 5 as signifi
c an t, 84 o f the 184 (45.7% ) assem blages are positively co rrelated w ith bone
density, a n d o f th ose 84 only 9 (10.7% ) are eth n o arch ae o lo g ic al assem blages.
O nly th ree assem blages, all eth n o arch ae o lo g ic al, are negatively co rrelated with
bo n e density. T hese results suggest th a t p o st-d ep o sitio n al a n d p o st-b u rial
d estru c tio n , w hich th e arch aeo lo g ical assem blages have u n d erg o n e b u t the
eth n o arch ae o lo g ic al assem blages have n o t (o r a t least n o t to the sam e degree),
are m a jo r ta p h o n o m ic facto rs (see C h a p te r 11). F orty -sev en (25.5% ) o f the 184
assem blages are negatively co rrelated a n d 14 (7.6% ) are positively co rrelated
w ith a u tility index. A b o u t one th ird (67, o r 36.4% ) o f the 184 assem blages
Sheep %MGUI
co rrelate w ith n eith er bon e density n o r the utility index (fall in C lass 5). T h u s, a
th ird o f th e assem blages we m ight co rrelate w ith b o th b o n e density a n d utility
m ay be inexplicable using only the variables o f food utility a n d stru c tu ra l
density. F u rth e r, only 74 (40.2% ) o f the 184 assem blages fall in C lasses 2 an d 4;
this m eans th a t less th a n h a lf o f these assem blages m ight be explainable using
only G ra y s o n ’s (1988) statistical criteria. F inally, 16.8% (31 o f 184) o f the
assem blages co rrelate significantly w ith b o th bone density a n d utility. T h a t is,
if this sam ple o f 184 assem blages is rep resen tativ e o f the to ta l variability in
fau n al assem blages, th en the equifinality p ro b lem will be en c o u n te red in one o f
every six assem blages we analyze. W e have, then, reached a p o in t in the search
100 -
80 -
3 60 -
<
20 -
0-
0 20 40 60 80 100
Caribou %MGUI
Figure 7.12. a, S catterp lo t o f % M A U frequencies fo r deer-size anim al rem ains
from 45C H 302 against th e stru ctu ral density o f deer bones, b, S catterp lo t o f
% M A U frequencies fo r deer- size anim al rem ains from 45C H 302 ag ain st the
% M G U I for sheep, c, S catterp lo t o f % M A U frequencies for deer-size anim al
rem ains from 45C H 302 against the % M G U I for caribou.
for e x p lan a tio n s o f varying frequencies o f skeletal p a rts th a t d em an d s o th er

kinds o f d a ta a n d o th e r kinds o f analysis. B ut w h at kinds o f d a ta a n d w hat
kinds o f analysis?
Tow ard a solution 1: counting units

W e ca n begin to discern an answ er to the preceding q u estio n if we first exam ine
the co u n tin g u n its used in studies o f skeletal p a rt frequencies. In m o st general
term s, th e co u n tin g unit is som e analytically specified p o rtio n o r p a rt o f a
skeleton. H o w p a rtic u la r skeletal p a rts are defined is critical. T he necessary an d
sufficient c o n d itio n s fo r identifying a specim en as rep resen tin g a m em b er o f a
p a rtic u la r categ o ry o f skeletal p a rt o r p o rtio n m u st be clear. T he original utility
indices (T able 7.1) w ere d eterm in ed fo r proxim al a n d distal halves o f long
bones, a n d p o rtio n s o f the axial skeleton consisting, w ith the exceptions o f the
atlas a n d axis v erteb rae, o f m o re th a n one skeletal elem ent. E xam ples o f the
la tte r include th e co u n tin g unit labeled " rib " an d the co u n tin g unit labeled
" th o ra c ic v e rte b ra ;” the fo rm er consists o f 26 individual skeletal elem ents and
the la tte r consists o f 13 individual skeletal elem ents in a typical u n g u late (T able
4.1). G iven th e skeletal units (p o rtio n s o f the skeleton) fo r w hich utility index
values w ere derived, it w as only logical to d eterm in e the m inim um n u m b e r o f
each o f th o se skeletal p o rtio n s - each is a n an a to m ica lly defined c o u n tin g u n it -
% M A U :B O N E D E N S I T Y
N E G A T IV E P O S IT IV E
SI G NI F IC A N T IN SIGNIFICANT S IG NI F IC A N T
totals:
LU h-
> 2 C L A S S 3 C L A S S 2 C L A S S 1
reverse utility reverse u tility reverse utility

< 2
CD Li_ w in now lag o r ravaged
O E = 9
CO E = 2, A = 0 E = 7, A = 11 E = 0, A = 2 7 A = 38
— 1—
D z : C L A S S 6 C L A S S 5 C L A S S 4
0 < v in n o v lag o r ravaged

H <->
B? L±-
E = 24
3 1
I ’
O
LlJ
oo
c
■si
E 1 6, A 51 A 99
O
= = =
II
ll
m
II
O
ll
o
S 1
[
C L A S S 9 C L A S S 8 C L A S S 7
—
1
2 ! bulk or gourmet bulk or gourmet bulk o r gourmet
u tility u tility u tility
> s
w innow lag o r ravaged E = 8
E -
</> ^ E = 1, A = 0 E = 6, A = 6 E = 1, A = 0 A = 6
o y
C l CO
totals: E = 3, A = 0 E = 29, A = 68 E = 9, A = 75 E = 41
A = 143
Figure 7.13. All possible co m b in atio n s (classes) o f co rrelatio n coefficients
betw een the % M A U o f a bone assem blage, an d b o th bone density an d % M G U I.
Classes th a t w ould represent a p artic u lar utility o r tra n sp o rt strategy are
indicated, as are those classes th a t represent ravaged assem blages, and fluvially
sorted lag (no t tran sp o rted ) an d w innow (tran sp o rted ) assem blages. E = the
nu m ber o f ethnoarchaeologically docum ented assem blages th a t fall w ithin a
class; A = the n um b er o f archaeological assem blages th a t fall w ithin a class.
in an arch aeo lo g ical assem blage (often referred to as the m inim um n u m b er o f

elem ents, o r M N E values), a n d it w as only logical to use the stru c tu ra l density
value w hich seem ed to typify each p a rtic u la r skeletal unit (L ym an 1985a).
M u ch a rc h a e o fa u n a l a n d p aleo n to lo g ical fau n al d a ta w ere, at th a t tim e, being
pu blished in ju s t such a form (L y m an 1984a). T he suggestion to use the
m ax im um stru c tu ra l density v alue fo r a skeletal unit ra th e r th a n the tra d itio n a l
o ne in o rd e r to pro v id e a m ore co nservative test o f the density-m ediated
d estru c tio n hy p o th esis is re aso n ab le (G iffo rd -G o n zalez a n d G a rg e tt n.d.), b u t
these tw o sets o f s tru c tu ra l d en sity values fo r deer are tig h tly co rrelated at an
o rd in al scale (rs = 0.89, P < 0.001) as are the tw o sets o f values fo r bison
(rs = 0.75, P < 0.001). T h e su b stitu tio n o f one fo r the o th e r n eith er app reciab ly
changes results like th a t in F ig u re 7.13 (L y m an 1992a, 1993b) n o r does it help
us d eterm in e w h eth er skeletal p a rt frequencies are the result o f differential
tra n s p o rt o r differential d estru ctio n .
A ta p h o n o m ist m u st ask if the skeletal u n its used to c o u n t skeletal p a rts are
the a p p ro p ria te ones. As we have seen, % M A U (o r % survivorship) values are
a p p ro p ria te because they are typically in terp re ted in term s o f the % M G U I an d
like m easures w hich have as th e ir co u n tin g u n it p ro x im al a n d distal halves o f
long bones. T his does n o t m ean, how ever, th a t p ro x im al a n d distal halves o f
long b o nes are alw ays o r are the only a p p ro p ria te m easure. F o r exam ple,
M etcalfe a n d Jon es (1988) a n d L ym an et al. (1992b) describe utility indices for
com p lete long b o nes ra th e r th a n p ro x im al a n d distal halves. F o r these indices,
com p lete long b o n e skeletal units are the a p p ro p ria te co u n tin g u n it (again,
typically referred to as M N E values in arch aeo lo g ical cases). F o r the stru c tu ra l
density values, the a p p ro p ria te co u n tin g u n its are the scan sites (F igures 7.4,
7.5, 7.6); the fact th a t m o st bone frequency d a ta h ad been (an d still are)
pu b lish ed as co u n ts o f p ro x im al a n d d istal halves o r as com plete skeletal
elem ents p ro m p te d the selection o f p a rtic u la r scan sites (T able 7.10) w hen
co rrelatin g p ublished bone frequencies w ith bone density (L ym an 1984a,
1985a). T h u s, it seem s we need to c o u n t o u r b ones several different w ays, o r at
several different levels o f inclusiveness, in o rd e r to generate the frequency d a ta
a p p ro p ria te to th e fram e o f reference to w hich we w ish to relate th a t d a ta .
B ecause o f the com plexity a n d the significance o f counting skeletal p a rts, it is
critical a t this ju n c tu re to review briefly the basic co u n tin g units presently in
use. R ecall fro m C h a p te r 4 th a t a specim en is a discrete, identifiable skeletal
p a rt such as a to o th , a com plete ph alan g e, o r the p ro x im al end o f a tibia. A
skeletal elem ent is a com plete, an a to m ica lly discrete u n it, such as a first m o lar
o r a h um eru s. T hu s, three specim ens o f the distal h a lf o f the h u m eru s are three
specim ens th a t re p resen t th ree skeleta l elem ents, w hereas tw o d istal ends o f
h u m eri a n d a p ro x im al end o f a h um eri are th ree specim ens th a t m inim ally
(assum ing they are all fro m left elem ents) re p resen t tw o elem ents. C om plexity
arises w hen tallying skeletal p a rt frequencies to cast u p o n the utility o r
stru c tu ra l density fram es o f reference because we m u st co n v e rt specim en
co u n ts in to % M A U values in such a m a n n e r th a t the q u a n tita tiv e u n its for
b o th reference fram es a n d the b o n e assem blage are th e sam e (see the discussion
o f e q u a tio n s [7.4] a n d [7.6] above). T h a t sam eness com es fro m d eterm in in g the
m in im um n u m b e r o f each skeletal u n it rep resen ted by th e b o n e collection. T his
dem an d s th a t one d eterm in e w hich specim ens belonging to a kin d o f skeletal
u n it are in d ep en d e n t o f one a n o th e r; i.e., d o n o t re p resen t th e sam e individual
skeletal elem ent. T he skeletal un its are o f course d eterm in ed a n d defined by the
fram e o f reference (e.g., fo r som e u tility indices, the skeletal units fo r tallying
long b ones are th e p ro x im al a n d distal halves), a n d those units have tra d itio n -
T ab le 7.12 M N E an d % M A U frequencies o f hyena-ravaged dom estic sheep

bones (fro m M arean an d Spencer 1991). O riginal M N E is 50
Surviving Sheep bone D eer bone Sheep

Skeletal p a rt MNE % S urvivorship density density %MGUI
P fem ur end 9.95 19.9 0.28 0.41 80.58

P fem ur shaft 32.05 64.1 0.20 0.33
M id fem ur shaft 50.00 100.0 0.36 0.57
D fem ur shaft 32.85 65.7 0.24 0.37
D fem ur end 7.30 14.6 0.22 0.28 80.58
P tibia end 7.80 15.6 0.16 0.30 51.99
P tibia shaft 43.00 86.0 0.20 0.32
M id tibia shaft 50.00 100.0 0.59 0.74
D tibia shaft 44.05 88.1 0.36 0.51
D tibia end 25.20 50.4 0.28 0.50 37.70
P m etatarsal end 26.48 53.0 0.43 0.55 15.77
P m etatarsal shaft 30.65 61.3 0.53 0.65
M id m etatarsal shaft 45.89 91.8 0.68 0.74
D m etatarsal shaft 31.90 63.8 0.51 0.57
D m etatarsal end 15.45 30.9 0.31 0.46 12.11
ally been called m in im u m n u m b ers o f elem ents (M N E ). O bviously, the w ord

“ elem en t” in the term M N E is n o t necessarily the sam e kind o f skeletal unit as
the term "e le m e n t” as it is defined earlier in this p a ra g ra p h an d in C h a p te r 4.
B inford (1978:70-71) reaso n ed th a t because “ o u r in terest is in the ac tu a l use
m ade o f anim als as fo o d ,” M A U values will provide “ u n d isto rte d conversions
o f th e ac tu a l co u n t o f bones in to an im al units [that] accurately describe the
relative p ro p o rtio n s o f a n a to m ica l p a r ts ” a n d avoid o v erestim atin g the
am o u n t o f m eat presen t at a site (see L ym an 1979b fo r a sim ilar discussion).
R ecall th a t M N E is th e m inim um n u m b e r o f each skeletal p a rt necessary to
acco u n t fo r the specim ens o f each p a rt, a n d th a t M A U is the M N E p er skeletal
p a rt divided by the n u m b er o f tim es th a t skeletal p a rt occurs in a single
skeleton. T h ere are several w ays to tally skeletal frequencies o th e r th a n M A U
an d M N E , each w ith th e express p u rp o se o f studying v aria tio n in skeletal p art
frequencies in the fossil record. F o r exam ple, L ym an (1979b) suggests one can
c o u n t the m in im u m n u m b e r o f b u tch erin g units; his suggestion is fo u n d ed on
an analysis o f h isto ric fau n al rem ains th a t h a d been b u tch ered w ith saws, a n d it
has been follow ed by an aly sts studying h isto ric faunal rem ains (e.g., S chulz and
G u st 1983). L ym an (1979b) also suggests, follow ing R ead (1971), th a t one can
c o u n t th e m inim um n u m b er o f “ skeletal p o rtio n s ” to o b ta in a n o th e r m easure
o f v aria tio n in how skeletons are represented. H e distinguishs three skeletal
p o rtio n s: the fo re q u a rte rs (forelim bs), h in d q u a rte rs (hindlim bs including
pelvis), a n d rib -v e rte b ra e skeletal p o rtio n . T h ere have been o th e r v aria n ts on
this them e, such as S tin er's (1990a, 1991 b) suggestion th a t skeletal p a rts can be
Frequencies o f skeleta l p arts 267
tallied by w h a t she calls “ an a to m ic a l reg io n s.” She defines nine an a to m ica l

regions: h o rn /a n tle r, h ead, neck (cervical verteb rae), axial colum n below the
neck (th o racic v erteb rae, ribs, stern u m , lu m b a r v erteb rae, pelvis, sacrum ),
u p p er fro n t lim bs (scapula, h um erus), low er fro n t lim bs (rad iu s-u ln a, carpals,
m etac arp a l), u p p e r hind lim bs (fem ur), low er h in d lim bs (tibia, tarsals,
m etatarsal), a n d feet (ph alanges). A ll o f these suggestions can be subsum ed
u n d er the general ru b ric o f a p a tte rn reco g n itio n a p p ro ach . T h a t is, different
scales o f inclusiveness o f p o rtio n s o f skeletons have been p ro p o se d in o rd e r to
determ in e if detectable p a tte rn s in the frequencies o f those p o rtio n s occur
across m u ltip le assem blages.
M a re a n an d S pencer (1991) pro v id e d a ta th a t are very conducive to
illu stratin g the issue o f a p p ro p ria te co u n tin g units. In th eir study o f the im pact
on bo n e assem blages by the sp o tte d h y en a (C rocuta crocuta), they fo u n d a
stro n g co rre la tio n betw een the s tru c tu ra l density o f lo n g bo n e p a rts a n d the
frequencies o f lo n g b o n e p a rts o f dom estic sheep (Ovis aries) th a t survived
gnaw ing. T he c o rrelatio n coefficient they describe w as based only o n the
frequencies o f p ro x im al a n d d istal ends o f th ree long bones. T hey also re p o rt
the frequencies o f p a rts o f shafts th a t survived hyena a ttritio n fo r each o f those
long b on es (T able 7.12). T h e frequencies o f b o th sh aft a n d end p a rts are alm ost
significantly co rrelated w ith the sheep (rs = 0.48, P = 0.07) bo n e density values
an d are co rrelated w ith the d eer (rs = 0.60, P = 0.02) bo n e density values (F igure
7.14). W h a t is o f g re ater interest to the q u estio n o f how to disentangle those
cases w here % M A U values are positively co rrelated w ith b o n e density an d
negatively co rrelated w ith the % M G U I is illu stra te d in F ig u re 7.15. T here, the
difference betw een the sca tte rp lo ts resides in the high su rv iv o rsh ip o f proxim al
an d distal sh afts o f the three long bones a n d the low su rv iv o rsh ip o f th eir
p ro x im al a n d distal ends. T he sc a tte rp lo t o f the p ro x im al a n d distal ends
ag ain st th e sheep % M G U I (F ig u re 7.15a) has the ap p e ara n ce o f a reverse
utility curve (F ig u re 7.1a) w hereas th e sc a tte rp lo t o f the p ro x im al a n d distal
shafts ag a in st th e % M G U I (F ig u re 7.15b) has the ap p e ara n ce o f a bu lk utility
curve (F ig u re 7.1c).
G iv en th a t % su rv iv o rsh ip a n d % M A U should be calcu lated o n the basis o f
the m a xim u m observed M N E fo r the skeletal p a rt being tallied (recalling first
how % M A U values are calculated [see eq u a tio n s 7.4-7.10], a n d second, th a t an
M N E fo r a p ro x im al o r distal h a lf o f a long b o n e can be based on the m inim um
n u m b e r o f ends, th e m in im u m n u m b e r o f n ear-en d sh aft frag m en ts, o r som e
co m b in a tio n o f th e tw o such th a t specim en in d ependence is ensured), the
difference in the tw o sca tte rp lo ts in F ig u re 7.15 m ay seem irrelev an t because the
an aly st sh o u ld be p lo ttin g the values in F ig u re 7.15b in sca tte rp lo ts such as
th o se exem plified in F ig u re 7.2. T h a t is so because the p ro x im al a n d distal shaft
p a rts o u tn u m b e r the p ro x im al a n d d istal ends, so the fo rm e r sh o u ld be the basis
for deriving th e p lo tte d M N E values ra th e r th a n th e latter. T he difference
betw een th e sc a tte rp lo ts in F ig u re 7.15 is, how ever, q u ite relev an t to o u r
100
80
Q.
3 60
>
CO 40
>8
O'*
20
0
0. 1 0.2 0.3 0.4 0.5 0.6 0.7
Sheep Bone Structural Density
b 100
80
Q.
|>
60
£
CO 40
o'
20
0
0.2 0.3 0.4 0.5 0.6 0.7 0.8
Deer Bone Structural Density

Figure 7.14. S catterplots o f % su rvivorship o f skeletal p a rts after ravaging by
hyenas (from M arean an d Spencer 1991) against (a) sheep bone stru ctu ral
density and (b) deer b one stru ctu ral density.
discussion fo r tw o reaso ns. F irst, th a t difference u n d ersco res th e fact th a t

co rrelatio n s betw een bo n e frequencies a n d stru c tu ra l density sh o u ld be calcu
lated on the basis o f the categories o f skeletal p a rts th a t m o st closely
co rresp o n d to th e scan sites. F req u en cies o f p ro x im al a n d distal sh a ft p a rts
sho u ld n o t be c o m p ared to the stru c tu ra l densities o f p ro x im al a n d d istal ends.
T h e second thin g th a t is illu stra te d in F ig u re 7.15 is th a t d ensity-m ediated
d estru c tio n has influenced the frequencies o f som e skeletal p a rts in this
assem blage (long bo n e ends) b u t n o t o th ers (long b o n e n ear-en d shafts a n d
a 100 -I----------------------------------------
80 -
Q_
w 60 -
o
>
£
Z 4° -
o-
20 -
0 -
0 20 40 60 80 100
Sheep %MGUI
b 100
80
Q-
W 60
o
>
w 40
20
0
0 20 40 60 80 100
Sheep %MGUI
Figure 7.15. V ariation in scatterplots o f % survivorship o f skeletal p a rts after
ravaging by hyenas (from M arean and Spencer 1991; see T able 7.12) against
sheep % M G U I for (a) long bone ends an d (b) long bone shaft ends.
m id-shafts). T his is im p o rta n t because it seem s im p ro b ab le th a t h u m an

b u tch ers w ould tra n s p o rt, fo r instance, p ro x im al fem u r shafts m u ch m o re
freq u en tly th a n p ro x im al fem u r ends, a n d th e sam e goes fo r distal fem u r shafts
an d distal fem u r ends, p ro x im al tib ia shafts a n d p ro x im al tib ia ends, etc. If the
kin d o f p a tte rn in g in skeletal p a rt su rv iv o rsh ip show n in F ig u re 7.15 occurs
across all m a jo r lim b o r long bones, I suspect the an a ly st w ould be safe in
suggesting th a t th e differential tra n s p o rt hypothesis m ay ac co u n t fo r freq u en
cies o f ends o f shafts, b u t d en sity -m ed iated d estru c tio n acco u n ts for the
frequencies o f lo n g bo n e ends. I f b o th the lo n g bo n e ends (w ith low stru c tu ra l

densities) a n d the c o rre sp o n d in g sh aft ends (typically hav in g higher stru ctu ra l
densities th a n ad jac en t ends) are b o th co rrelated w ith bo n e density, then it
w ould be unw ise to co n clude th a t eith er m easu re o f b o n e frequencies (ends o r
sh aft ends) p ro v id e a valid m easu re o f differential tra n sp o rt because b o th the
bo ne ends an d the sh aft ends co u ld have been subjected to density-m ediated
a ttritio n .
W h e th e r o r n o t I am co rrect in these suggestions is largely irrelevant; their
co rrectn ess can be assessed by eth n o arch ae o lo g ic al research, w hich show s us
yet a n o th e r directio n fo r such research to proceed. W h a t is im p o rta n t in the
preceding exam ple is th a t the scale o f the c o u n tin g unit w as ch an g ed from the
M N E o f p ro x im al a n d distal halves o f long bones to the M N E o f scan sites
(F ig u res 7.4, 7.5, a n d 7.6). If archaeologically observed frequencies o f the scan
sites co rrelate w ith th eir stru c tu ra l density, then there is every reaso n to
a n ticip ate th a t th e frequencies o f the larg er co u n tin g units o f p ro x im al an d
distal halves o f b on es will also, because the la tte r are (o r should be) the
m ax im u m M N E o f a scan site fo u n d o n th e p ro x im al o r distal h a lf o f a long
bone. S im ilar arg u m e n ts apply to all o th e r skeletal p a rts w hen th o se p a rts are
defined by the utility o r tra n s p o rt indices in such a m a n n e r th a t they co n tain
m o re th a n one scan site.
R ap so n (1990) n o tes th a t long bo n e ends are rare b u t long bone shafts are
a b u n d a n t in the arch aeo lo g ical collection he studied, a n d he th erefo re suggests
th a t d en sity -m ed iated d e stru c tio n o f low -density long bo n e ends h ad occurred,
b u t th e h igh -den sity long bo n e sh afts h ad n o t been destroyed. F ifteen years
earlier K lein (1975:286) arg u ed , follow ing n e o ta p h o n o m ic o b serv atio n s
re p o rted by Sutcliffe (1970), th a t hyenas d estro y long b o n e ends by gnaw ing
an d c o n su m p tio n , b u t leave long bo n e shafts, including the p ro x im al an d distal
ends o f the shafts. K lein suggested th a t h o m in id b u tch ers seem no t to destroy
long b o n e ends because they b re a k a bo n e op en by im p actin g som e p a rt o f the
sh aft ra th e r th a n the end. K lein (1975:279) describes bovid rem ains from a late
Pleistocene hyena den. T h a t assem blage co n tain e d m ore long bone sh aft ends
(n = 290) th a n long b o n e ends ( n = 170; b o th values are fo r the sum o f the
h u m eru s, rad iu s, fem ur, a n d tibia). T his is a statistically significant difference
(chi2 = 31.3, P cO .O O l).
E x p erim en tal w o rk confirm s the significance o f K lein ’s (1975) a n d R a p s o n ’s
(1990) suggestion. O n the basis o f th eir actu alistic research, M a re a n an d
Spencer (1991:655) ind icate th a t “ if 100 percent o f the [bone] sam ple has been
rav ag ed by hyenas th en large differentials will exist betw een M N E s calculated
o n lim b-b on e ends a n d M N E s calcu lated on m iddle-shaft pieces.” T hey
calcu late a m id d le-sh aft M A U to lim b-bone end M A U ra tio , an d find th a t tw o
to five tim es as m an y shafts will survive rav ag in g by carn iv o res as long-bone
ends. (N o te: in this case, fo r all skeletal p a rts considered, M A U = M N E -r- 2, so
end to sh aft ra tio s are the sam e w h eth er M A U o r M N E values are used.) T his in
T ab le 7.13 Frequencies o f skeletal p a rts at F L K Z in ja n th ro p u s (fro m Bunn

and K roll 1986, 1988) and com plete bone utility index values (fro m M etca lfe
and Jones 1988)
Skeletal M N E based on ends M N E based on C om plete bone

elem ent only (proxim al/distal) ends an d shafts utility index
hum erus 19(5/19) 20 36.8

radius 14(14/5) 22 25.8
m etacarpal 15 (15/8) 16 5.2
fem ur 6 (6/6) 22 100.0
tibia 11 (10/11) 31 62.8
m etatarsal 15(15/10) 16 37.0
tu rn leads to the suggestion th a t such ratio s allow the an aly st to m easure “ the
in tensity o f carn iv o re ra v ag in g ” (M a re a n a n d S pencer 1991:655), higher ratio s
(m ore sh aft M A U s p er end M A U s) in d icatin g m o re intensive ravaging.
Setting aside fo r the m o m en t th e suggestion th a t sh aft to en d ra tio s m ight be
used to m easu re how ravaged a b o n e assem blage is, M a re a n a n d S pencer’s
(1991) d a ta in d icate a possible w ay to escape the equifinality p ro b lem posed
w hen skeletal p a rt frequencies co rrelate positively w ith bo n e density an d
negatively w ith utility indices. C arn iv o re d estru c tio n sh o u ld be evidenced by
carn iv o re gn aw ing m ark s (see C h a p te r 6 a n d below ), low lo n g b o n e end to long
bo ne sh aft ratio s, a n d a positive c o rrelatio n o f frequencies o f long bo n e ends
w ith b o n e stru c tu ra l density. W hen this co m b in a tio n o f a ttrib u te s is fo u n d , the
an aly st ca n d eterm in e th e M N E o f com plete long bones on the basis o f long
bo ne sh aft pieces (as well as o v erlap p in g long b o n e ends; see B unn 1991; B unn
an d K ro ll 1986, 1988), an d p lo t th o se values ag ain st a utility index such as th a t
o f M etcalfe an d Jo n es (1988) fo r co m p lete lim b bones. T h a t is, sim ply shift the
inclusiveness o f th e c o u n tin g unit, in this case fro m p ro x im al an d distal halves
to co m p lete long bones. Such a shift in scale fo r co u n tin g the b ones requires, o f
course, a sim ilar shift in the scale o f the a n a to m ica l u n its in the u tility index.
A s a n exam ple o f th e preceding, we can use the bo n e assem blage recovered
from th e P lio-P leistocene F L K Z injanthropus site (B unn 1986, 1991; B unn and
K ro ll 1986, 1988). R elev ant d a ta are su m m arized in T ab le 7.13. M a re a n an d
Spencer (1991:655) in te rp re t th e ra tio s o f hindlim b shafts to ends fo r this
assem blage as “ stro n g ly suggest[ive of] intensive, p e rh a p s 100 percen t, c a rn i
vore ra v ag in g .” T h e M N E o f lo n g bo n e ends co rrelates stro n g ly w ith the
m ax im u m s tru c tu ra l density p er skeletal p a rt (rs = 0.81, P = 0.007), also sug
gesting d en sity -m ed iated d e stru c tio n has influenced the b o n e frequencies in
this assem blage. T h e M N E values as d eterm in ed only fro m long b o n e ends are
negatively (if insignificantly) co rrelated w ith the ca rib o u M G U I (rs = —0.28,
P = 0.35). T h u s it is n o t su rp risin g th a t th e frequencies o f skeletal p a rts as
determ in ed from th e long bo n e ends describes a negative re la tio n w hen p lo tte d
Utility
Figure 7.16. S catterp lo ts o f M N E frequencies derived from long bone ends only,
an d from long bone ends an d long bone shafts, for F L K Zinjanthropus
assem blage (after Bunn 1986; Bunn an d K roll 1986, 1988), against food utility
index (from M etcalfe an d Jones 1988). Best fit sim ple regression lines are show n
to em phasize the differences in the tw o sets o f p lotted p oints (com pare with
Figure 7.1).
ag ain st M etcalfe an d Jo n e s’ (1988) com plete bo n e utility index (F ig u re 7.16).

But, in clu d ing the sh aft pieces alo n g w ith the ends to derive M N E values results
in skeletal p a rt frequencies th a t describe a positive re latio n w hen p lo tte d
again st th a t utility index (F ig u re 7.16). In this case, then, using com plete bone
M N E values ra th e r th a n M N E values derived from only long bo n e ends seems
to get us o u t o f the in terp retiv e d ead end o f equifinality. It ap p e ars th a t the
fau n al rem ain s a t the F L K Z injanthropus site owe th e ir presence th ere a t least
in p a rt to th eir econom ic utility because bones o f higher utility are m ore
freq u en t th a n bones o f low er utility.
R egardless o f the correctness o f the preceding in te rp re ta tio n , it is im p o rta n t
to re tu rn to the n o tio n th a t long bo n e sh aft to long bone end ratio s m o n ito r the
in ten sity o f carn iv o re gnaw ing. M a re a n a n d S pencer’s (1991) experim ental
d a ta are com pelling. F o r exam ple, the m ax im u m M N E fo r fem urs based on
ends is 9.95 bu t th e fem u r M N E based on m iddle shafts is 50.47; the m axim um
M N E fo r tibiae based on ends is 25.2 b u t based on m iddle shafts it is 50.22; the
m axim um M N E fo r m e ta ta rsa ls based on ends is 26.48 but based on m iddle
shafts it is 45.89. T he ratio s o f end M A U s to sh aft M A U s are 5.07, 1.99, an d
1.73, respectively fo r fem urs, tibiae, an d m etatarsals. T hese are im pressive
in d icatio n s o f the in tensity o f c a rn iv o re rav ag in g w hen it is realized th a t these
ra tio s w ou ld be 1 if the entire carcass (to tal, if frag m en ted , bones) w as p resen t
a n d u n rav ag ed . It is im p o rta n t to realize, how ever, th a t the significance o f the
ra tio s resides in the fact th a t en tire b ones (end to sh aft ra tio s = 1) w ere given to
th e h yenas by the researchers. T h a t is, the hyenas h ad the o p p o rtu n ity to ea t all
o f each bon e. T he q u estio n th en is, d o the ra tio s reflect w h a t w as tra n s p o rte d
a n d given to the hyenas (w hat the hyenas h a d access to), o r does it reflect w h at
they did n o t consum e? W o u ld the ratio s be as high if twice as m an y shafts as
ends (end to sh aft ra tio = 0.5) h a d been given to the hyenas? T his is the critical
issue in th e co n fla tio n o f the effects o f differential tra n s p o rt w ith differential
d estru c tio n on skeletal p a rt frequencies. D oes the presence o f a p a rtic u la r
skeletal p o rtio n influence how ca rn iv o res exploit o th e r kinds o f skeletal
p o rtio n s in an assem blage? W hile M a re a n e ta l.{ 1992) a p p a re n tly seem to th in k
th e an sw er to this q u estio n is negative (they m u st assum e, if im plicitly, th a t in
p re h isto ric cases, as w ith studies o f differential tra n s p o rt assum ing com plete
skeletons were initially p resen t, com plete b ones w ere initially present), there is
no so u n d basis to th in k they are co rrect, a n d an e cd o tal d a ta suggest they m ay
be in co rrect (e.g., H ayn es 1980b, 1982). A d d itio n a l ex perim ental w o rk is called
for here. A n d , recall th a t while I suggested earlier th a t it seem ed unlikely th a t
h u m an b u tch ers w ould tra n s p o rt p ro x im al ends, proxim al shafts, m id-shafts,
distal sh afts, a n d distal ends as sep a rate pieces a n d in different ab u n d a n ces, we
do n o t kn o w this fo r a fact.
M a re a n a n d Spencer (1991) fo u n d th a t o f the 50 hin d lim b s o f d om estic sheep
they fed to sp o tted hyenas, a n average o f 70% ( ± 18% ) o f the ends b u t only 7%
( ± 4 % ) o f th e sh afts w ere d estro y ed (averages o f 12 m easu rem en ts fo r bone
ends, a n d 4 m easu rem en ts fo r shafts). T hey are th u s certainly co rrec t th a t the
frequencies o f diaph y sis frag m en ts tallied as M N E values p er com plete lim b
b o n e will p ro v id e m o re ac cu ra te m easures o f the frequencies o f th o se skeletal
elem ents in an assem blage th a n M N E values based on frequencies o f long bone
ends. A s n o ted in c o n ju n ctio n w ith earlier discussion o f T ab le 7.12 a n d F igures
7.14 a n d 7.15, the % su rv iv o rsh ip values fo r M a re a n a n d S pencer’s experim en
tally g en erated assem blage a re n early significantly co rrelated w ith b o n e density
w hen all skeletal p a rts (ends, p ro x im al a n d distal shafts, m id shafts) are
included (rs> 0 .4 8 , / 5< 0 .0 7 ). O m ittin g the epiphyseal ends an d retain in g only
the p rox im al a n d distal sh aft, a n d m id -sh aft skeletal p a rts, the c o rrelatio n
d isap p ea rs (for sheep bo ne density rs = 0.24, P = 0.5; fo r deer b o n e density
rs = 0.31, P = 0.38). T hese results suggest the u tility g ra p h in F ig u re 7.15a is a
fu n ctio n o f d en sity -m ed iated d e stru c tio n w hereas the utility g ra p h in F igure
7.15b is n o t a fu n c tio n o f d en sity -m ed iated d estru c tio n (recall th a t in the
previo u s section it w as em phasized th a t such conclusions should be fo u n d ed on
m o re th a n ju s t the c o rre la tio n coefficient).
T h e exam ple ju s t review ed is a n im p o rta n t one. It indicates o ne w ay we m ight
begin to d isen tan g le the effects o f d en sity -m ed iated d e stru c tio n fro m differen
tial tra n s p o rt o f skeletal p arts. H ow ever, it is by no m ean s going to w o rk in all
cases. S tin er (1991b, 1991 e), fo r instance, claim s th a t M N E co u n ts based on
long b o n e shafts m ay n o t alw ays p ro d u c e h ig h er frequencies th a n M N E co u n ts
based o n lo n g b o n e ends. T h ere are tw o w ays to ev alu ate this claim indirectly,
w ith an exam ple, a n d w ith ex p erim en tal d ata.
T he skeletal p a rts in the 45C H 302 bo n e assem blage (F ig u re 7.12, T ab le 7.11)
w ere tallied by c o u n tin g the M N E per long b o n e half; th a t is, the m inim um
n u m b er o f p ro x im al halves o f h u m eri necessary to ac co u n t fo r p ro x im al ends,
p ro x im al shafts, a n d p ro x im al-m id shafts o f h um eri w ere ac co u n ted fo r in
deriving the M N E a n d th u s the % su rv iv o rsh ip o f p ro x im al hum eri. Y et, th a t
assem blage is significantly co rrelated w ith b o n e density (rs = 0.67, P < 0.001). If
only frequencies o f p ro x im al a n d distal halves o f h um eri, radii, m etacarp als,
fem ora, tib iae, a n d m e ta ta rsa ls are considered, a n d the m ax im u m density
values fo r th e p ro x im al a n d distal halves o f these bones are used, the c o rrelatio n
is even stro n g e r (rs = 0.81, P = 0.008). T h u s it is clear th a t the 45C H 302
assem blage has been subjected to d en sity -m ed iated a ttritio n a n d a utility g rap h
c a n n o t be derived even from th e frequencies o f shafts o f lim b b ones in th a t
assem blage. T his exam ple does n o t d irectly su b sta n tia te S tin er’s (1991 b, 1991 e)
claim , b u t it does in d icate M A U frequencies based o n specim ens o f long bone
sh afts m ay som etim es c o rrelate w ith bo n e density.
T he ex p erim en tal d a ta w ith w hich to ev alu ate S tin e r’s (1991b) claim th a t
M N E values based o n long bo n e shafts a n d long b o n e ends will n o t alw ays
o u tn u m b e r M N E values based only o n long bo n e ends consist o f sh aft
frag m en t to end fragm en t ratios. Bunn (1989) used a h am m ersto n e a n d an anvil
to b re ak cow (B os sp.) lim b bones, a n d derived ra tio s o f identifiable sh aft
frag m en ts (specim ens th a t co u ld be identified to skeletal elem ent represented)
to identifiable end fragm ents. B u n n ’s (1989) ra tio s are N IS P o f shafts to M N E
o f each elem ent, a n d are 2.6 fo r the h u m eru s, 2.6 fo r the rad iu s, 2.5 fo r the
fem ur, a n d 4.7 fo r the tibia. T h o se ra tio s are higher if the sh aft frag m en ts th a t
w ould be arch aeo lo g ically unidentifiable are included: 5.1 fo r the h u m eru s, 4.6
for th e ra d iu s, 4.5 fo r th e fem ur, a n d 6.8 fo r th e tibia. B u n n ’s d a ta ca n also be
used to calcu late N IS P o f shafts to N IS P o f ends ratio s. F o r identifiable shafts
th o se ra tio s are 1.3 fo r the h u m eru s, 0.9 fo r th e rad iu s, 1.2 fo r the fem ur, a n d 2.2
for th e tibia. F o r all sh aft frag m en ts, th o se ra tio s are 2.6 fo r the h u m eru s, 1.6
for th e rad iu s, 2.2 fo r the fem ur, an d 3.2 fo r the tibia. B inford (1978:155)
reco rd ed th e average n u m b e r o f “ splinters a n d ch ip s” o f long b o n e shafts per
artic u la r end p ro d u c ed by N u n a m iu t E skim os fractu rin g ca rib o u (R angifer
tarandus) long b o nes to e x tra ct m arro w . T he average N IS P o f shafts to M N E o f
ends ra tio s B inford (1978) re p o rts a re 17.5 fo r th e h u m eru s, 12.2 fo r the radius,
8.2 fo r th e m etac arp a l, 16.6 fo r th e fem ur, 19.7 fo r the tibia, a n d 6.6 fo r the
m etatarsal. B in fo rd ’s (1978) ra tio s are a b o u t three tim es g re ater th a n B u n n ’s
(1989) fo r th e sam e v ariables (N IS P o f shafts to M N E o f ends). T h a t difference
is p erh ap s a ttrib u ta b le to tax o n o m ic v ariatio n .
T he p o in t o f B in fo rd ’s (1978:156) ra tio s is th a t “ if we observe m o re splinters
a n d chips th a n we estim ate [should be fo u n d given a ce rtain n u m b e r o f ends],
th en we h ave re aso n to suspect th a t som e a rtic u la r ends w ere destro y ed b eyond
reco g n itio n o r som e a rtic u la r ends w ere rem oved fro m the site afte r the bones
h ad been b ro k e n for m a rro w ." T h a t is, B inford (1978) suggests we can use
th o se ra tio s to estim ate the n u m b er o f artic u la r ends th a t should be present, a
p ro c ed u re he p u ts to use in a later p a p e r (B inford 1988). T h e p o in t o f B u n n ’s
(1989) ra tio s is to help su b sta n tia te his c o n ten tio n th a t d eterm in in g the
m in im u m n u m b er o f bones represented by b o th long b o n e shafts a n d long bone
ends will usually result in g re ater an d th u s m ore ac cu ra te M N E values th a n
w hen ju s t long b o n e end pieces are used (e.g.. B unn 1986, 1989; B unn a n d K roll
1986, 1988).
B oth th e ra tio s described by B unn (1989) a n d those described by B inford
(1978) in dicate m ore shaft specim ens (N IS P ) th a n end specim ens (N IS P ) will
result solely fro m h am m ersto n e b reak ag e o f u n g u late lo n g bones. R em oval o f
end pieces by carn iv o res will increase the ra tio values. In the absence o f
ca rn iv o re rav ag in g o r o th e r ta p h o n o m ic processes w hich selectively rem ove
long b o n e ends, th e ra tio s o f pieces should be like th o se d o cu m en te d by B unn
an d B inford. an d , im p o rtan tly , the ra tio s o f M N E values based on long bone
ends to M N E values based on long bone shafts should be a p p ro x im a te ly one.
T h u s it seem s th a t each sam ple m u st be ev a lu a ted as to w h eth er h ig h er skeletal
co u n ts are in fact o b tain ed using diap h y sis-b ased M N E s o r epiphysis-based
M N E s.
I d o n o t m ean in the preceding to suggest th a t ra tio s o f long bo n e shafts to
long b o n e ends are n o t valuable analy tical tools. C ertain ly they are. W h a t I do
m ean to im ply, how ever, is the fact th a t even these ra tio s (or, m y frequencies o f
scan sites) m ay n o t be sufficient fo r analytically d isentangling the im pacts o f
differential tra n s p o rt an d differential, d en sity -m ed iated d estru c tio n on a set o f
p reh isto ric bones, a n d thus m ay be insufficient for gaining a tap h o n o m ic
u n d e rsta n d in g o f varying skeletal p a rt frequencies. O th e r kinds o f d a ta are
necessary to su p p lem en t the frequencies o f scan sites rep resen ted , a n d sh aft to
end ratio s.
T ow ard a solution 2: other attributes

M arean et a i (1992:117) suggest th a t so rtin g o u t the influences o f density-
m ed iated d e stru c tio n from the influences o f econom ically fo u n d ed decisions o f
skeletal p o rtio n tra n s p o rt requires d a ta o n “ the incidence o f h a m m ersto n e
percussion a n d to o th m ark in g o n bones in ta n d e m w ith analyses o f b o d y p a rt
re p re se n ta tio n .” T h a t suggestion is fo u n d e d o n a n u m b e r o f actualistic
ex perim en ts (B lum enschine 1988; B lum enschine a n d Selvaggio 1988, 1991;
M a re a n 1991; M a re a n an d S pencer 1991; M a re a n et al. 1992). W hile in m ost
cases these ex perim en ts focus on the effects o f sp o tted h y en a (C rocuta crocuta)
a n d co n cern th e im p acts o f ca rn iv o re rav ag in g on b ones o f d om estic sheep
(O vis aries), th eir results have b ro a d im plications fo r study o f skeletal p a rt
frequencies.
In an in n o v ativ e stu d y o f the effects o f carn iv o res o n skeletal p a rt freq u en

cies, B lum enschine (1988) created three kinds o f bone assem blage. O ne, called
“ carn iv o re o n ly ” assem blages, consisted o f bones o f bovid carcasses killed, o r
scavenged, a n d co n su m ed by large A frican carnivores. T he second kind, called
“ h am m ersto n e o n ly ” assem blages, involved fresh, defleshed bovid bones
b ro k e n w ith a h a m m ersto n e w hile th e bo n e lay o n a sto n e anvil. Som e o f these
assem blages w ere n o t fu rth e r subjected to ta p h o n o m ic processes, b u t o th ers o f
them w ere placed o n the lan d scap e im m ediately follow ing b reak ag e an d
rem oval o f m a rro w to p ro d u c e the th ird kin d o f assem blage, called “ sim ulated
site” assem blages. T h e sim u lated site assem blages w ere scavenged by c a rn i
vores, m ain ly sp o tted hyenas, a n d the bones later collected fo r study.
C o m p ariso n s betw een the three kinds o f assem blage reveal significant
differences in a ttrib u te s th a t m ight be used to help unravel the tap h o n o m ic
processes we have been discussing. F irst, skeletal p a rts in the sim ulated site
assem blages have significantly few er to o th m ark s (8% to 45% o f the specim ens
per assem blage display such m ark s) th a n the skeletal p a rts in the carn iv o re only
assem blages (66% to 100% ). S econd, m o st epiphyseal ends o f long bones in
b o th the sim ulated site assem blages an d in the ca rn iv o re only assem blages w ere
co n su m ed by carn iv o res, a n d th u s th e m ere absence o f epiphyses does n o t allow
th e an a ly st to d istin g u ish bo n e assem blages rav ag ed only by carn iv o res from
assem blages initially subjected to m a rro w e x tra ctio n a n d su b sequently to
ca rn iv o re ravaging. T h ird , p ro p o rtio n s o f long bo n e en d pieces w ith carn iv o re
to o th m ark s are high, lo n g -b o n e n ear-en d pieces are m id-level in frequency,
an d p ro p o rtio n s o f lo n g -b o n e m id -sh aft pieces are low in frequency fo r b o th
the sim ulated site a n d th e carn iv o re only assem blages (B lum enschine 1988).
B lum enschine’s (1988) experim ents in d icate the p ro p o rtio n o f carn iv o re
g n aw ed b o n e pieces is high in assem blages th a t did n o t have m a rro w rem oved
an d th e b on es w ere u n b ro k e n w hen carn iv o res en c o u n te red them (carnivore
only), a n d low in assem blages in w hich the m a rro w h ad been rem oved a n d the
b ones h a d been b ro k e n p rio r to c a rn iv o re action. T hus, the p ro p o rtio n a l
frequency o f carn iv o re-g n aw ed o r to o th -m a rk e d specim ens in a n assem blage
m ay help in d icate the p re-gnaw ing c o n d itio n o f bones processed by carnivores
an d , w h eth er o r n o t people h ad access to the bones p rio r to the carnivores.
F u rth e r. B lum enschine a n d Selvaggio (1991) indicate th a t long bones in the
sim ulated site assem blages te n d to have m an y h am m ersto n e percu ssio n m ark s
(see C h a p te r 8) lo cated in the m id -sh aft area, w hile to o th m a rk s ten d to occur
on the epiphyses a n d near-epiphysis p o rtio n s o f the long bones.
W hile th e presence o f h am m ersto n e o r p ercussion m ark s o n bones signifies a
h o m in id ta p h o n o m ic ag ent, the presence o f a carn iv o re to o th m a rk sim ilarly
signifies a (q u a d ru p e d a l) ca rn iv o ro u s ta p h o n o m ic agent. T o o th m ark s on
m an y o f the epiphyseal o r near-ep ip h y seal p o rtio n s o f long bones in co n ju n c
tio n w ith p ercu ssio n m ark s o n m id -sh aft p a rts w ould suggest com plete bones
h ad been p rocessed by h u m an s a n d sub seq u en tly rav ag ed by carnivores. As
w ith th e ra tio o f long b o n e shafts to long b o n e ends, th en , the frequency an d

d istrib u tio n o f to o th m ark s a n d p ercussion m ark s m ay help identify the
ta p h o n o m ic h isto ry o f a bone assem blage, an d th u s help unravel the ta p h o n o
m ic m ean in g o f v ary in g skeletal p a rt frequencies. C arn iv o re to o th m ark s,
how ever, m ay n o t alw ays be p resen t on bones even if carn iv o res h ad u n h in
dered access to them .
K en t (1981) show s th a t dogs (C anis fa m ilia r is), at least, can gnaw bones and
yet leave n o visible gnaw ing m ark s (im p o rtan tly , she also show s th a t dogs
re d istrib u te bones; th a t is, dogs o ften m ove bones from th eir o riginal d e p o sitio
nal loci in sites). H aynes (1983a: 171) show s th a t different ca rn iv o re tax a gnaw
b ones in d istin ct w ays, b u t notes th a t w h eth er an assem blage o f bones displays
gnaw ing d am ag e o r n o t d epends on several variables th a t “ affect an individual
p re d a to r’s b e h a v io r." F u rth e r, the presence o f gnaw ing m ark s on bones does
n o t necessarily m ean the observed frequencies o f bones are the result o f
carn iv o re-g en erated attritio n . T o illu stra te this, it is necessary to assum e th a t as
a bo n e specim en becom es sm aller, th ro u g h frag m en tatio n o r attritio n a l
processes th a t rem ove o r d estro y bo n e tissue, th a t specim en becom es less
identifiable as to the ta x o n rep resen ted a n d the skeletal elem ent represented.
Several studies in d icate this is a reaso n ab le a ssu m p tio n (L y m an a n d O ’Brien
1987; W a tso n 1972).
A s b o n e specim ens becom e sm aller a n d less identifiable, progressively m ore
o f them becom e “ an aly tically a b se n t” (L y m an a n d O 'B rien 1987); th a t is,
p ro x im al h um eri m ay be p resen t in a collection b u t som e o f the pieces o f
p ro x im al h u m eri are so sm all o r m odified by a ttritio n th a t they are un id en tifi
able, a n d th u s they are, fo r analy tical (q u an tificatio n ) p u rp o ses, ab sen t. T he
frequency o f such an aly tically ab sen t pieces sh o u ld increase th ro u g h tim e as
ta p h o n o m ic processes co n tin u e to affect a bo n e assem blage. T his u n d ersco res
th e tim e-tran sg ressiv e o r cu m u lativ e n a tu re o f ta p h o n o m ic processes; it also
im plies th a t there is a th resh o ld at w hich a p a rtic u la r skeletal p a rt will cease to
be an alytically p resen t (identifiable). T his is im p o rta n t in o u r co n sid eratio n o f
the presence o f carn iv o re gnaw ing m ark s o n bones a n d the frequency o f
gnaw ed b ones suggesting a bo n e assem blage h as u n d erg o n e d ensity-m ediated
a ttritio n because the extent o f ca rn iv o re gnaw ing d am ag e is cum ulative.
G a rv in (1987) fed dom estic cow (Bos taurus) bones to dogs, an d d o cu m en ted
the a m o u n t o f b o n e p a rt w eight loss over tim e d u e to gnaw ing. H e fed the dogs
defleshed, fresh p ro x im al a n d distal halves o f h um eri, radii, fem ora, a n d tibiae
(distal rad ii h ad ca rp als a tta c h e d , a n d d istal tib iae h a d tarsa ls a ttac h ed ). H e
record ed th e p ercen tag e o f to ta l w eight th a t w as lost p er specim en every tw o
days o v er a 22 d ay period . I presum e th a t an y w eight loss den o tes the loss o f
bo n e m ateria l a n d th u s the p ro d u c tio n o f identifiable gnaw ing m ark s. G iven
th a t a ttritio n o f b ones g en erated by carn iv o re gnaw ing seem s to be m ed iated by
the stru c tu ra l den sity o f th e bones, one ca n pred ict th a t skeletal p a rts w ith low
stru c tu ra l densities will lose w eight m ore rap id ly th a n skeletal p a rts w ith high
T ab le 7.14 Correlation coefficients ( r J betw een percent weight loss o f skeletal

p arts due to carnivore gnaw ing over tim e ( data fr o m Garvin 1987), and
different m easures o f bone structural density. See also Figure 7.17
N u m b er o f Bison M axim um D eer M axim um

days gnaw ed density bison density density deer density
2 rs= 0.02 0.06 - 0 .2 9 - 0 .4 6

P= 0.91 0.85 0.45 0.23
4 - 0 .1 3 - 0 .1 7 - 0 .3 3 - 0 .4 8
0.73 0.66 0.38 0.20
6 - 0 .2 0 - 0 .2 6 - 0 .4 6 - 0 .6 6
0.59 0.49 0.23 0.08
8, 10, 12, 14 - 0 .5 4 - 0 .5 2 - 0 .6 0 - 0 .7 4
0.15 0.16 0.11 0.05
16 - 0 .5 6 - 0 .5 7 - 0 .5 5 -0 .8 1
0.13 0.13 0.14 0.03
18, 20 - 0 .6 7 - 0 .6 2 - 0 .6 9 - 0 .8 8
0.07 0.10 0.06 0.02
22 - 0 .7 8 - 0 .7 4 - 0 .7 1 - 0 .8 6
0.04 0.05 0.06 0.02
stru c tu ra l densities. T h u s, as m o re w eight (m ore b o n e m ateria l) is lost th ro u g h

tim e (as dogs h ave progressively m o re days to gnaw o n b o n e p arts), the p ercen t
w eight loss recorded by G a rv in (1987) should be m ore strongly a n d negatively
co rrelated w ith b o n e stru c tu ra l density. T his is precisely w h a t is seen in the
c o rrelatio n coefficients betw een percen t w eight loss a n d the b o n e density values
(T able 7.14).
T he p ercent w eight loss over the first tw o w eeks is n o t co rrelated w ith bone
density (T able 7.14). T his indicates th a t the m ere presence o f gnaw ing m ark s on
bo n e p a rts does n o t necessarily m ean ca rn iv o re-g en erated d en sity -m ed iated
a ttritio n has p ro d u c ed the observed bo n e frequencies. T he coefficients indicate
th a t th ree w eeks’ w o rth o f gnaw ing is necessary before such a ttritio n will
p ro d u c e a statistically significant re la tio n betw een bo n e frequencies a n d the
stru c tu ra l density o f b o n e p arts. T his tim e-transgressive o r cu m u lativ e effect o f
carn iv o re gnaw ing is density m ed iated , as show n in F ig u re 7.17, w here it is clear
th a t b o n e m aterial w as lost m o re ra p id ly (a n d a p p a re n tly w as still being lost
afte r th ree weeks) fro m th e low -density p ro x im al h u m eru s th a n fro m th e high-
density distal h u m eru s o ver the three-w eek gnaw ing perio d . T h u s the m ere
frequency o f gn aw ing m ark s m ay be m isleading, by itself, as an in d ic a to r o f
ca rn iv o re-g en erated d en sity -m ed iated a ttritio n . T he an a ly st m ay w a n t to
d eterm in e if th e p ro p o rtio n a l (% ) frequency o f skeletal p a rts displaying
gnaw ing m ark s varies directly w ith the s tru c tu ra l density o f the p arts. T h a t is,
are m o re o f the skeletal p a rts w ith low density (a n d high m arro w a n d grease
80 -
□ P humerus
w HU D humerus
10
o
-j 60 -
x:
g>
'cd
^ 40 -
o'
o>
>
ro
| 20 -
3
o
0-
8 10 12 14 16 1 8 20 22
D ays Gnaw ed
Figure 7.17. B ar g rap h o f % w eight loss o f cow bones over time. N o te th a t the
low stru ctu ral density proxim al hum erus lost m ore w eight m ore quickly th an the
high stru ctu ral density distal hum erus over the first 10 days, and then continued
to lose w eight while the distal hum erus lost alm ost no w eight over the last 12
days.
co n ten t) gnaw ed th a n p a rts w ith high density (an d low m a rro w a n d grease
co ntent)? I f so, one w ould then have ad d itio n a l evidence suggesting th a t
carn iv o res h ad in fact destro y ed som e lo n g b o n e ends.
P eople b re ak an im al bones, som etim es intensively, in o rd e r to e x tra ct grease
a n d m a rro w (see the discussion o f “ W ith in -b o n e n u trie n ts” below ). T h u s, one
m ig h t ask , if high frequencies o f percussion m ark s in d icate intensive frag m en
ta tio n o f b ones by h u m a n s (e.g., B lum enschine 1988; M a re a n 1991), do they
also in d icate th e an aly tical absence o f ce rtain skeletal p a rts due to the sm all size
o f the fragm ents? W hite (1956) suggested fractu rin g o f b ones d u rin g b u tc h e r
ing m ay sm ash b o nes b ey o n d reco g n itio n , a n d B inford (1981b) agreed th a t this
co uld certain ly h ap p e n (how B inford suggests we deal w ith an assem blage o f
b ones th a t h as u n d erg o n e d en sity -m ed iated a ttritio n is described later). But
n eith er o f these a u th o rs n o r an y o th e r ta p h o n o m ist has a ttem p ted to find o u t
how the frequency o f percussion m ark s o n bones m ight covary w ith density-
m ed iated d estru ctio n . O ne m ight ask if th ere is a relatio n betw een how bones
b reak an d the stru c tu ra l density o f bones. A gain, w hile we seem to have learned
a lo t a b o u t how bo n es b reak , a n d w hy th ey b re ak th e w ay they d o , in the p ast
decade, n o one h as a tte m p te d to find a c o rrelatio n betw een these tw o variables
(b on e fra g m e n ta tio n is dealt w ith in C h a p te r 8). Som e h ave suggested selection
o f p a rtic u la r b ones fo r to o l m a te ria l m ig h t resu lt in the rem oval o f b ones w ith
high stru c tu ra l densities (e.g., M a c G re g o r 1985). B ut again, no detailed study

o f the c o v a ria tio n in utilized bo n e tool m aterial a n d the stru c tu ra l density o f
utilized b ones has been perform ed.
S tiner (1990a, 1991 b) su m m arizes a set o f a ttrib u te s w hich she believes can be
used to d eterm in e if the d o m in a te biological cause o f a b o n e assem blage was a
n o n -h u m a n carn iv o re (especially a hyena), o r a hom in id . F irst, assem blages o f
faun al rem ain s th a t are d o m in a te d by h ead p a rts (tallied w ith isolated teeth
excluded) are p ro d u c e d by ca rn iv o res w hereas h o m in id -cre ated b o n e assem
blages h ave a p p ro x im a te ly e q u al frequencies o f all skeletal p a rts (are n o t head-
d o m in ated ) a n d the an im al carcasses are all m o re o r less an ato m ically
com plete. S econd, carn iv o res ta k e m o re old ad u lt prey as these ten d to be
scavenged as well as h u n ted ; h o m in id s ten d to p ractice little selection w ith
reg ard to age o f prey (a m b u sh h u n tin g ) o r focus on prim e-aged individuals (see
C h a p te r 5). T h ird , b o n e assem blages resu ltin g fro m ca rn iv o re activity display
m o re evidence o f gn aw in g th a n assem blages resu ltin g fro m h u m a n activity.
F o u rth , “ carn iv o re la trin e s” are associated w ith bone assem blages pro d u ced
by ca rn iv o res b u t n o t w ith bo n e assem blages p ro d u c ed by hom inids.
T o th e last o f S tin e r’s (1991b) a ttrib u te s we m ig h t a d d such th in g s as
ca rn iv o re co p ro lites in asso c ia tio n w ith carn iv o re accu m u latio n s b u t n o t w ith
bo n e accu m u latio n s created by ho m in id s (e.g., K lein 1975), a n d m o re evidence
o f co rro sive d am ag e to bones th a t have passed th ro u g h ca rn iv o re digestive
trac ts in the fo rm er assem blages th a n the la tte r (see C h a p te r 6). M an y o f the
a ttrib u te s su m m arized by S tiner (1991b) are precisely th o se th a t have long been
in use by ta p h o n o m ists to discern the ta p h o n o m ic agent responsible fo r a bone
ac cu m u latio n (C h a p te r 6). W h a t is im p o rta n t a b o u t these a ttrib u te s in the
co n tex t o f ex p lainin g b o n e frequencies is th a t they b rin g to b e a r o th e r lines o f
evidence to help explain w hy a p a rtic u la r skeletal p a rt profile a p p e ars the w ay it
does, lines o f evidence th a t are to som e ex ten t in d ep en d e n t o f the stru c tu ra l
density a n d utility o f skeletal p a rts. T herefore, skeletal p a rt frequencies th a t fall
in C lass 1 o f F ig u re 7.13 m ay p ro v e to be inexplicable given only the c o rrelatio n
coefficients betw een b o n e frequencies, a n d stru c tu ra l density an d food utility,
b u t w ith reference to o th e r lines o f evidence, such as d em o g ra p h y o f m o rtality ,
freq uency o f h ead p a rts, a n d frequencies o f coprolites, gnaw ing dam ag e, an d
digestive co rro sio n , th o se frequencies m ay in som e in stances be explicable.
L y m an et al. (1992a) p e rh a p s said it well w hen they referred to th e cliche th a t
a statistical co rre la tio n does n o t necessarily d en o te a cau sal re la tio n betw een
tw o v ariables. A fte r finding a significant c o rre la tio n b etw een the frequencies o f
m a rm o t (M a rm o ta sp.) skeletal p a rts a n d th e s tru c tu ra l density o f th o se p arts,
L ym an et al. (1992a) w ent o n to argue th a t differential recovery o f sm all
skeletal p a rts did n o t seem to be c o n trib u tin g to th e co rre la tio n because sm all
skeletal p a rts w ere excluded fro m th e frequencies. F u rth e r, evidence o f
carn iv o re gnaw ing o n som e o f th e b ones suggested th a t th e p o te n tia l existed for
b o n e d e stru c tio n d u e to such a ta p h o n o m ic process. F inally, bo n e frag m en
ta tio n did n o t seem to be c o n trib u tin g to th e significance o f the c o rrelatio n

because som e o f th e m a rm o t bones w ere com plete. T h a t is, they, like Stiner,
used m u ltip le lines o f evidence to derive a conclusion. This re tu rn s us to
B in fo rd ’s (1987:453) statem en t q u o te d earlier: indices o f utility, tra n sp o rt, and
stru c tu ra l density o f b o n e p a rts are “ fram es o f reference,” a n d as such th eir use
as exp lan ato ry algorithm s fo r skeletal p a rt profiles is n o t advised; ra th e r, they
sh ou ld be used as one o f several steps in building a ta p h o n o m ic e x p lan a tio n fo r
th o se profiles.
T h u s far in this c h a p te r som e o f the sim pler initial steps in explaining the
frequencies o f skeletal p a rts have been considered. T here are, as th e preceding
tw o p a ra g ra p h s im ply, m ore com plex issues involved as well. In the next
section, several o f these are review ed.
Within-bone nutrients
M arsh all an d P ilgram (1991) suggest th a t h u m a n ex tra ctio n o f n u trie n ts w ithin
b ones, p artic u la rly m arro w , m ay have a significant influence o n skeletal p a rt
profiles. T h ey p ro p o se th a t bones th a t ra n k h igh in w ith in -b o n e n u trie n t utility
will tend to be m o re fragm ented th a n b ones th a t ra n k low in w ithin-bone
n u trie n t utility. T h e initial effect o f fra g m e n ta tio n d u rin g , say, m arro w
e x tractio n , will th u s be initially to raise the N IS P p er skeletal elem ent. As
fra g m e n ta tio n co n tin u es, how ever, som e frag m en ts will becom e so sm all as to
becom e unidentifiable, and th u s the later effect o f frag m en tatio n will be to
reduce N IS P values. G iven th a t M N I is o ften a statistical fu n c tio n o f N IS P
(G ray so n 1984), the u ltim ate effect o f frag m en tatio n will be to reduce M N I
values. Sim ilarly, intensive fra g m e n ta tio n will reduce the observed M N E (and
th u s M A U ) values for an assem blage (see C h a p te r 8).
R eferred to as “ an aly tical ab sen ce” by L ym an an d O 'B rie n (1987), it is n o t at
all clear how we m ight an alytically overcom e the effect o f fra g m e n ta tio n an d
reduced id entifiability o f bo n e specim ens (b u t see the discussion o f refitting in
C h a p te r 5). P ro b ab ly the m o st im p o rta n t issue here, from a tap h o n o m ic
perspective, is th e a tte m p t to identify such effects, an d to m easure their
m ag n itu d e. T his is precisely w h a t M arsh a ll a n d P ilgram (1991) d o w hen they
co m p are th e degree o f fra g m e n ta tio n o f rem ains o f tw o taxa. T hey m easure
frag m en tatio n as the p ercen tag e o f identified specim ens th a t are com plete for
each skeletal elem ent. A n o th e r w ay to m easu re fra g m e n ta tio n is to calculate
the N IS P :M N E ra tio per skeletal elem ent. Sim ilarly, one could calculate an
N IS P : M N I ra tio , to c o m p are across tax a, if the sam e skeletal elem ents are used
for all tax a (see C h a p te r 8 fo r m o re on frag m en tatio n ).
H o w w ould such ratio s be used to assess the extent o f o r in tertax o n o m ic
v aria tio n in the e x p lo itatio n o f w ith in -b o n e n u trien ts such as m arro w an d
grease? O ne co uld p lo t the N IS P :M N E ra tio s for m a rro w -c o n ta in in g bones
ag ain st a m arro w utility index, pred ictin g th a t skeletal elem ents w ith high
T ab le 7.15 N IS P : M N E ratios f o r selected skeletal

p arts (fro m M arshall and Pilgram 1991)
Skeletal p art C ap rin e ratio C attle ratio
P hum erus 2.7 4.3

D hum erus 2.9 3.5
P radius 3.5 3.3
D radius 2.2 4.8
P fem ur 2.2 2.2
D fem ur 3.9 6.3
P tibia 3.0 2.9
D tibia 2.7 6.7
m arro w indices w ould be m ore b ro k e n a n d th u s have h ig h er N IS P :M N E ra tio s

th a n skeletal elem ents w ith low m a rro w indices. F o r exam ple, the N IS P :M N E
ra tio s fo r cap rin e (sheep a n d goats) a n d cattle rem ains from an A frican
N eo lithic site are given in T ab le 7.15, a n d are p lo tte d ag a in st the sheep m arro w
u tility index (T able 7.1) a n d the fat utility index fo r bison (T able 7.4) in F igure
7.18, respectively. N e ith er p a ir o f variables is significantly co rrelated (for
caprin es, rs = —0.13, P = 0.72; fo r cattle, rs= —0.18, P = 0.64), bu t b o th p oint
scatters seem to define inverse relatio n s, an d th a t re la tio n ap p e a rs to be steeper
for th e cattle rem ains th a n fo r the cap rin e rem ains (recall th a t utility indices are
o rd in al scale, as are arch aeo lo g ical b o n e frequencies, so the ap p e ara n ce o f
steepness m ay be m o re a p p a re n t th a n real). B ut the cattle rem ains seem to be
m o re frag m en ted th a n the cap rin e rem ains; the average N IS P :M N E ra tio for
the fo rm er is 4.25 (SD = 1.6) a n d for the la tte r it is 2.9 (S D = 0.6). R ecall th a t
fra g m e n ta tio n has th e initial effect o f increasing N IS P values, b u t as pieces are
b ro k e n in to progressively sm aller pieces the second effect is to reduce N IS P
values. F ig u re 7.18 co u ld reflect m an y o f the ca ttle b ones w ith high fa t an d
m arro w c o n te n t being b ro k e n b ey o n d recognition. C ap rin e bones m ay have
been less intensively b ro k e n th a n cattle bones because the fo rm er have less
m arro w an d fat co n ten t. T his explains the steeper slope o f the best-fit
regression line fo r cattle an d the nearly h o riz o n ta l slope o f the best-fit line for
cap rin es (an d th u s th e steepness m ay be real as well as a p p a re n t).
R egardless o f w h eth er o r n o t the preceding is co rrect, it illu strates how the
ex p lo itatio n o f w ith in -b o n e n u trie n ts m ight influence b o n e frequencies. It
un dersco res the fact th a t fra g m e n ta tio n can p ro d u c e an analytical absence o f
skeletal p arts. W ith in -b o n e n u trien ts, such as m arro w , grease, a n d the fat
asso ciated w ith neu ro logical tissues o f the axial sk eleto n (S tiner 1991b),
rep resen t critical variables th a t have n o t been intensively studied. If co m p ared
to such variables as N IS P :M N E ratio s, the p ro p o rtio n a l frequency o f each
skeletal unit displaying gnaw ing m ark s a n d /o r percussion m ark s, a n d the like,
w ith in -b o n e n u trie n t indices m ay reveal details o f the ta p h o n o m ic h isto ry o f a
b o n e assem blage. U ltim ately , like the frequencies o f gnaw ing m ark s a n d long
Frequencies o f skeleta l p arts
cattle
□ Cattle
■ Caprine
Within-Bone Nutrient Index

Figure 7.18. N IS P :M N E ratio s p lo tted against w ithin-bone n u trien t index for
tw o taxa. Sim ple best-fit regression lines show n fo r reference. See text for
discussion.
bon e shaft M N E to long bone end M N E ratio s, they m ay help us explain why
som e skeletal p a rts are relatively ra re in som e assem blages.
Reconstruction o f ravaged assemblages
W e wish to recognize w hether we are dealing w ith an assem blage th a t has been
tran sp o rte d o r w ith a residual p o p u latio n , th a t is, w h at rem ains after o th e r p arts are
tran sp o rted . C om plicating the m a tte r m ay be the presence a n d /o r absence o f
destru ction coupled w ith tra n sp o rt as well as different levels o f destru ctio n in
different settings. H ow do we begin unraveling such a com plicated set o f possible
conditions?
(L. R. Binford 1981b:217)
T h u s far, I have covered several an aly tic techniques a n d fram es o f reference

developed to help explain vary in g skeletal p a r t frequencies. O nce we k n o w th a t
a p a rtic u la r bo n e assem blage has u n d erg o n e carn iv o re rav ag in g o r density-
m ed iated d estru c tio n , can we still discuss differential tra n s p o rt o f skeletal p a rts
by p reh isto ric hom inids? C an we p erh ap s reco n stru ct, fo r instance, the p re
rav ag in g assem blage? I f we can answ er this second q u estio n in the affirm ative,
then we can also an sw er the first q u estio n in the affirm ative. It is w ith the second
q u estio n th a t this section is concerned.
In a sem inal c o n trib u tio n to the m eth o d a n d th eo ry o f m o d ern zo o a rch a eo -
logy, U e rp m a n n (1973:319) urg ed zo o arch aeo lo g ists to “ develop m odels o f
bo n e d isin te g ratio n so th a t the m issing m aterial can be determ in ed an d
calcu lated scientifically.” H e suggested th a t th e a ttritio n a l processes actin g on

faun al rem ains “ c o n fo rm to certain laws w hich could be u n d ersto o d by the
ap p lica tio n o f statistical m eth o d s [and eventually the an aly st w ould] be able to
calcu late th e m issing q u a n tity ” o f bones th a t h ad been subjected to those
a ttritio n a l processes (U e rp m a n n 1973:319). O ne o f the principle criticism s o f
m easu rin g a b u n d a n ces o f fa u n al rem ains a t the tim e U e rp m a n n w ro te involved
the effects o f differential p re serv atio n an d the fact th a t such p re serv atio n m ight
be differentially d istrib u te d across tax a (see the review in G ra y so n 1984). If the
b ones o f ta x o n “ 1” w ere well preserved b u t th e bones o f ta x o n “ 2” w ere p o o rly
preserved, a n in acc u rate assessm ent o f the relative a b u n d a n ces o f the tw o
w ould resu lt from sim ple co u n ts o f b ones a n d p ro b a b ly as well from m inim um
nu m b ers o f individuals. I f a statistical m odel like th a t envisioned by U e rp m a n n
co uld be developed, zo o arch aeo lo g ists w ould have a pow erful analytic tool
allow ing th em to estim ate m o re accu rately relative tax o n o m ic ab u n d an ces.
V ario u s research ers h ad been searching fo r regularities th a t m ight co n stitu te
such a m odel p rio r to the p u b licatio n o f U e rp m a n n ’s p a p e r (e.g., B rain 1967b,
1969), an d m an y c o n tin u ed to exam ine th o se regularities o f bo n e a ttritio n after
1973 (see references in L ym an 1984a). W hen, h a lf a decade afte r U e rp m a n n ’s
(1973) plea, a form al m odel o f bone a ttritio n allow ing the “ m issing q u a n tity ”
o f b o n es to be estim ated w as co n stru c te d , th a t m odel w as geared to w ard
re co n stru ctin g the ab u n d a n ces o f p a rtic u la r skeletal elem ents o f a single taxon.
T he m odel is based o n a p o ly n o m ial eq u a tio n derived by B inford a n d B ertram
(1977:138) fro m an analysis o f the statistical re la tio n betw een the stru c tu ra l
density o f bo n e p a rts a n d the frequencies o f bo n e p a rts in an assem blage know n
to have been ravag ed by carn iv o res. T he m odel has been used to estim ate the
"m issin g q u a n tity ” o f bones by B inford (1978, 1981b, 1984b) several tim es. It
has also been used by B lum enschine (1986a) to re co n stru ct w hat is th o u g h t to
be a rav ag ed assem blage, a n d it h as been m en tio n ed w ith a p p ro v a l by o th er
research ers (e.g., M a re a n 1991; T u rn e r 1989).
G iven th e a p p a re n t logic o f U e rp m a n n ’s original p h ra sin g o f the problem ,
an d the av ailab ility o f a m odel in the fo rm o f a n easily solved eq u a tio n , w hy is
re co n stru ctio n o f the m issing q u a n tity o f b ones n o t co m m o n p lace in ta p h o n o
m ic analysis? P erh ap s, as T u rn e r (1989:6) notes, the p ro b lem is th a t one m ust
assum e p a rtic u la r b o n e p a rts are ra re d u e to destru ctiv e agents ra th e r th a n
faulty arch aeo lo g ical recovery. M o re generally, the re co n stru ctio n p ro ced u re
requires th e a ssu m p tio n th a t, w hen ab u n d a n ces o f skeletal p a rts are positively
co rrelated w ith the stru c tu ra l densities o f th o se p a rts, ra re skeletal p a rts are
ra re because they have low s tru c tu ra l densities a n d they w ere th u s originally
present b u t w ere destro y ed by d en sity -m ed iated a ttritio n a l agents. T h a t is, one
m ust assum e th a t b o nes w hich are ab se n t w ere in fact p resen t p rio r to
p erfo rm in g the re co n stru ctio n ; m o re plainly, one is assum ing precisely w h at
one is try in g to determ ine. In term s o f ascertain in g the econom ic utility or
tra n s p o rt p a tte rn (F ig u re 7.1), if we fo u n d a positive c o rre la tio n betw een bone
frequencies a n d stru c tu ra l density, we w ould be p ro m p te d to “ re c o n stru c t” the

frequencies o f th e b o n es destroyed by d en sity -m ed iated attritio n . B ut to do so
w ould dem and th a t we assum e those ra re, low -density (high utility) bones were
tra n sp o rte d , utilized, an d dep o sited on the site by people, only to be later
d estroyed. T his is precisely w h a t we w a n t to determ ine; did the people actually
tra n s p o rt a n d utilize th ose bones? I suggest this single requisite a ssu m p tio n is
why th e re co n stru ctio n technique developed by B inford a n d B ertram (1977) is
n o t used by m o st tap h o n o m ists; no one is w illing to g ra n t the assu m p tio n .
But, fo r sake o f discussion, let us g ra n t the assu m p tio n . W h a t h ap p e n s next?
U sing stru c tu ra l density values they m easu red , B inford a n d B ertram (1977)
derived a p o ly n o m ial e q u a tio n describing the statistical re la tio n betw een the
density values as th e in d ep en d e n t v ariab le a n d % M A U frequencies in an
assem blage k n o w n to have been ravaged by carnivores. T hey th en solved the
e q u a tio n fo r each skeletal p a rt using the stru c tu ra l density values to derive
w h a t they called “ S P ” values, o r the p ro p o rtio n a l frequency o f a p a rtic u la r
bo ne p a rt exp ected to survive d en sity -m ed iated a ttritio n . T o re co n stru ct the
assem blage o f b o n e p a rts originally presen t, one divides the observed M A U per
skeletal p a rt (w hich, as we have seen, is the sam e as the su rv iv o rsh ip value) by
the SP value for th a t skeletal p a rt to “ yield an estim ate o f the M A U s originally
p resen t before the a ttritio n a l ag en t destro y ed th e b o n e s” (B inford 1978:210).
T h a t is, B inford is suggesting th a t w hen the an aly st has an assem blage o f
p re h isto ric b o n e p a rts the frequencies o f w hich are positively co rrelated w ith
th eir s tru c tu ra l densities, n o t only does it a p p e a r th a t the b o n e assem blage has
u n d erg o n e d en sity -m ed iated a ttritio n , b u t the an aly st should re c o n stru c t the
assem blage to d eterm in e w h a t w as p resen t p rio r to d estru ctio n . So, the
q u estio n now is, w h a t is the effect o f such an analytic re co n stru ctio n on skeletal
p a rt frequencies? A n exam ple is the easiest w ay to answ er this question.
B inford (1978:210-211) re co n stru cted b o n e assem blages created by N u n a -
m iu t d ogs using th e bones recovered fro m “ dog y a rd s” a n d follow ing the
p ro ced u re o u tlin ed in the preceding p a ra g ra p h . T he observed frequencies o f
bo ne p a rts from tw o d og y ard s are given in T ab le 7.16, as are the reco n stru cted
frequencies. It is im p o rta n t to n o te th a t the observed a n d the re co n stru cted
frequencies are tig h tly co rrelated ; for the M o rry D o g Y a rd assem blage r s = 0.85
(P = 0.0001) a n d fo r the R u lla n d D og Y a rd assem blage rs = 0.94 ( /3 = 0.0001).
T hese coefficients in d icate the re c o n stru c tio n h as n o t significantly altered the
relative a b u n d a n ces o f skeletal p a rts in eith er assem blage. T he final q u estio n to
ask th en , is, has the process o f re co n stru ctio n resulted in frequencies o f bone
p a rts th a t w ould lead us to a co n clu sio n different th a n th a t we m ig h t derive
using th e observed frequencies o f skeletal parts? A s F ig u re 7.19 m ak es clear,
re co n stru ctio n w ould not lead us to a different conclusion. In th a t figure, b o th
the sca tte rp lo t o f observed frequencies a n d the sca tte rp lo t o f re co n stru cted
b o n e frequencies suggest a reverse utility strategy.
W hen U e rp m a n n (1973) called fo r th e develo p m en t o f a statistical m odel
T ab le 7.16 R econstructing caribou bone assem blages fr o m N u n a m iu t sites

(fro m B inford 1978)
M orry M orry R ulland R ulland

dog yard % M A U dog yard % M A U dog yard % M A U dog yard % M A U
Skeletal observed reconstructed observed reconstructed
skull 34.4 42.26 84.6 100.00

m andible 100.0 100.00 100.0 96.10
atlas 31.2 44.43 15.4 20.99
axis 31.2 56.88 15.4 26.92
cervical 14.7 35.32 15.4 36.34
thoracic 8.6 22.75 7.7 19.25
lum bar 15.5 30.88 20.0 38.09
pelvis 28.1 31.17 46.2 49.12
rib 5.2 18.94 1.6 5.65
sternum 0.0 0.00 0.0 0.00
scapula 0.0 0.00 38.4 62.31
P hum erus 3.1 18.96 0.0 0.00
D hum erus 12.5 20.30 23.1 35.93
P radius 18.7 39.64 23.1 46.83
D radius 25.0 48.40 30.8 57.20
carpal 7.5 22.75 7.7 22.34
P m etacarpal 3.1 7.68 7.7 18.17
D m etacarpal 3.1 8.36 15.4 39.57
P fem ur 3.1 7.51 0.0 0.00
D fem ur 6.3 21.05 0.0 0.00
P tibia 6.3 18.94 7.7 22.34
D tibia 9.4 12.91 15.4 20.45
tarsal 6.3 18.84 7.7 22.48
astragalus 9.4 23.04 15.4 36.34
calcaneum 6.3 15.35 15.4 36.34
P m etatarsal 9.4 19.79 7.7 15.61
D m etatarsal 6.3 18.94 7.7 22.48
first phalanx 2.5 12.63 1.5 7.53
second phalanx 2.5 17.52 1.5 10.36
third phalanx 2.5 25.26 1.5 14.94
th a t w o u ld allow the an aly tic re c o n stru c tio n o f frequencies o f b ones m issing

fro m an assem blage d u e to a ttritio n a l processes, his plea m ad e g o o d sense. It
still does. I f we can b u ild such a m odel, we ca n effectively erase m u ch o f the
tap h o n o m ic o v erp rin t o r bias th a t plagues the q u an tifica tio n o f tax o n o m ic
ab u n d an ces. Sim ilarly, such a m odel, if p ro p e rly c o n stru c te d , w ould allow us to
ascertain w hich skeletal p a rts w ere frequently tra n s p o rte d an d w hich p arts
w ere rarely tra n s p o rte d in th o se cases w hen frequencies o f skeletal p a rts seem
to be a t least in p a r t th e resu lt o f d en sity -m ed iated a ttritio n . T he single m odel
developed to d ate fo r the la tte r p u rp o se , how ever, d em an d s an a ssu m p tio n th a t
seems u n w a rra n te d , th e a ssu m p tio n th a t ra re bones o f low stru c tu ra l density
100 - ■
Morry Dog Yard
□ Observed
+ Reconstructed
Z) +
<
* \
■ + □+o+++++ s
++ + + +
* Jt □ g
' S? a □
i ■ 1 i ' i
1.0
- o
60 80 100
c\j
o
%MGUI
100 + □ Rulland Dog Yard

+
□ □ Observed
80 + Reconstructed
+
ZD 60 - +
<
+ fi
4 0 '- +
□
+♦ a
20 -. + +
ct +
+°h°□
-fa □
+
□lb am □+ □
T *—1
--------1
--------»
-------*-*
20 40 60 80 1 00
%MGUI
Figure 7.19. S catterplots o f carib o u bone observed an d reconstructed frequencies
against the carib ou % M G U I (d ata from B inford 1978).
w ere in fact once presen t in an assem blage. T o g ra n t the a ssu m p tio n results in
p resu m in g th a t w h a t we are trying to show m ig h t have h ap p e n ed did in fact
h ap p e n . E ven g ra n tin g the assu m p tio n in th o se cases w here a stro n g positive
c o rre la tio n is fo u n d betw een bo n e frequencies a n d the stru c tu ra l density o f
b o n e p a rts, w hich are the logical cases to m ak e such an assu m p tio n , th e analytic
process o f re co n stru ctio n does n o t p ro d u c e m ark ed ly different conclusions
because th e re co n stru cted frequencies are sim ply observed values increased
p ro p o rtio n a te ly to th eir stru c tu ra l density. T hus, to d ate, U e rp m a n n ’s (1973)
plea h as gone unfulfilled.
Other sources of variation in bone structural density

In his sem inal stud y on bo n e density B rain (1969) notes th a t a b o n e's stru ctu ra l
d en sity co rrelates w ith th e o n to g en ic age o f the bone; in p artic u la r, he notes
th a t skeletal p a rts w ith early fusing epiphyses tend to be denser th a n skeletal
p a rts w ith late fusing epiphyses. B inford a n d B ertram (1977:105-109) also note
a stro n g c o rrelatio n betw een the onto g en ic age o f a b o n e an d th a t b o n e ’s
stru c tu ra l density. T he significance o f th eir o b serv atio n is th a t they d o c u
m en ted w ith em pirical d a ta a p ro p e rty recognized at least 15 years earlier by
G u ild ay et al. (1962) th a t the b ones o f im m a tu re individuals preserve less well
th a n the b ones o f m a tu re individuals. It sh o u ld also be clear th a t in at least
som e tax a, the stru c tu ra l density o f b ones n o t only increases as the anim al
m atu res, b u t it decreases as the anim al ap p ro ach e s senility (e.g., P erzigian
1973). T h u s, rem ains o f ontogenically y oung a n d very old individuals m ay be
ra re because th eir b o n es did n o t preserve. A s n o te d in C h a p te r 5, ontogenically
y o u ng ind iv iduals tend to be a b u n d a n t a n d ontogenically old individuals tend
to be ra re in living p o p u la tio n s. T he precise n a tu re o f th e effects o f o n to g en y on
th e stru c tu ra l density o f b ones a n d the re su lta n t effects on bone p re serv atio n
h av e yet to be stu d ied in th e sam e d etail as the w ith in -sk eleto n v a ria tio n in bone
density described earlier in this ch a p te r. L ym an (1984a), fo r exam ple, selected
skeletally m a tu re (all epiphyses fused) skeletons to m easure, as did K reu tzer
(1992). W e need, then, d a ta o n the stru c tu ra l density o f b ones from skeletally
im m a tu re (e.g., epiphyses unfused) individuals.
T he d a ta in T ables 7.6 a n d 7.7 m ak e it clear th a t the stru c tu ra l density o f
skeletal elem ents varies w ithin a single skeleton. But all scan sites th u s far
d o cu m en te d are fo r com plete cross-sectional areas o f a p a rt o f a skeletal
elem ent. M a re a n (1991:688) notes th a t, fo r exam ple, the m edial h a lf o f the
p ro x im al rad iu s o f sm all to m edium -sized bovids is p ro b a b ly m ore dense th a n
the lateral side, “ p resu m ab ly in response to g re ater b o d y w eight load in g on the
m edial artic u la r facet.” T h u s, different scan sites m ay eventually be m easured
to help b e tte r to a c c o u n t for frag m en ts o f the b o n e p o rtio n s defined by the scan
sites (F ig ures 7.4 a n d 7.5). D o c u m e n tin g the v a ria tio n in density across scan
site RA1 (F ig u re 7.4), fo r exam ple, m ight help explain M a re a n ’s o b serv atio n .
B inford a n d B ertram (1977:117) also n o te th a t, given the m ineral reservoir
fu n ctio n o f living bone, seasonal n u tritio n a l stress m ight result in the skeletal
elem ents o f anim als th a t died w hen well n o u rish ed being o f hig h er density th a n
th o se sam e bones in an an im al th a t died w hen u n d er n u tritio n a l stress. This
sh ould be no surprise as it is clear th a t the n u tritio n a l q u ality of, say, a bovid in
a te m p erate en v iro n m e n t will v ary w ith th e seasons (e.g., S peth 1983), b u t to
d ate few d a ta on seasonal v a ria tio n in the stru c tu ra l density o f bones are
available. H o rw itz a n d S m ith (1990) re p o rt th a t w ell-nourished dom estic sheep
(Ovis aries) ewes show no seasonal v aria tio n in the co rtical thickness o f the
d iaphysis w alls o f m etap o d ials, b u t ewes u n d e r dietary stress display such
v aria tio n . R am s show n o such v a ria tio n w h eth er well n o u rish ed o r seasonally
u n d er d ietary stress, leading H o rw itz a n d S m ith (1990) to suggest th a t the
ad d ed stress o f g estatio n a n d la c ta tio n o n ewes living u n d e r c o n d itio n s o f
seasonally restricted foo d supplies can result in seasonal flu ctu atio n s in bone
m ass. T h a t tran slate s, o f course, in to seasonal flu ctu atio n s in the stru c tu ra l
density o f bones. T his in tu rn suggests th a t a fu rth e r step in research on the
stru c tu ra l den sity o f bones m ay require the g en eratio n o f density d a ta for
in d iv id u als th a t died at different seasons o f the y ea r a n d u n d e r different
n u tritio n a l regim es.
W e h ave learn ed m u ch in the last tw o decades a b o u t how the stru ctu ra l
density o f b ones m ed iates a n d buffers the effects o f ta p h o n o m ic processes an d
agents. B ut we still h av e m u ch m o re to learn. W hy, fo r exam ple, d o b ones o f
sm all m am m als a p p a ren tly n o t preserve well co m p ared to the bones o f large
m am m als (e.g., B ehrensm eyer et al. 1979; see C h a p te r 9)? Is it because the
surface to volu m e ra tio o f bones is g re ater in sm all m am m als? It does n o t seem
to be because o f v a ria tio n in th e s tru c tu ra l density o f bones, as L ym an et al.
(1992a) re p o rt th a t m o st scan sites fo r m a rm o ts are denser th a n their
h o m o lo g u es in deer, a n d th u s one m ig h t expect o n this basis alone th a t b ones o f
sm all m am m als will w ith stan d d en sity -m ed iated ta p h o n o m ic processes b etter
th a n b o n es o f large m am m als. T he s tru c tu ra l density o f skeletal p a rts is an
im p o rta n t fram e o f reference th a t w a rra n ts fu rth e r study, b u t it is n o t the only
one. C o v a ria tio n o f th e s tru c tu ra l density o f bones w ith o th er stru c tu ra l an d
m o rp h o m e tric p ro p e rtie s sh o u ld be stu d ied sim u ltan eo u sly w ith o u r co n tin u ed
learning a b o u t stru ctu ra l density.
A final comment
R egardless o f the q u a n tita tiv e u n it used to tally frequencies o f b o n e specim ens,
it is critical to rem em b er th a t the reason fo r tallying b o n e frequencies an d
co n stru c tin g skeletal p a rt profiles is to study, an d hopefully to explain, the
differences a n d sim ilarities betw een the archaeologically observed skeletal p a rt
frequencies a n d the frequencies o f skeletal p a rts in a set o f com plete skeletons.
T he success o f M a re a n et al.'s (1992) discussion o f the im pacts o f carnivore
rav ag in g o n b o n e su rv iv o rsh ip a n d th u s skeletal p a rt frequencies, fo r instance,
resides in the fact th a t they knew w h a t the original, p re-rav ag in g skeletal p a rt
profile lo o k ed like. T hu s, they could m easu re exactly the p ro p o rtio n o f bone
p a rts rem ain in g afte r ravaging. A rchaeologically, we never kn o w w h a t w as
originally presen t, o r w h at the p re-rav ag in g skeletal p a rt profile lo o k ed like,
a n d th u s we c a n n o t m easure the p ro p o rtio n lost due to ca rn iv o re ravaging. W e
are forced to sta rt w ith the m odel o f a com plete skeleton (o r set o f M N I
co m plete skeletons).
T he significance o f th e preceding can be m ad e clear by fu rth e r c o n sid eratio n
o f M a re a n et al.'s (1992) d a ta . A nalyzing th o se d a ta (T able 7.17) ju s t as one
7.17 E xp erim en tal data f o r bone transport and survivorship (fro m M arean et al. 1992), and how those data w
a ted in an archaeological context. S F U I fr o m M etca lfe and Jones (1 9 8 8 ). See te x t f o r discussion
M arean et a l.'s E xperim ental d a ta treated as if

experim ental d ata recovered from an archaeological context
l O riginal Surviving O riginal Surviving %M AU M axim um S tru

MNE MNE MAU MAU Surviving SFUI D ensity (deer)
ble 0.0 0.0 0.0 0.0 0.0 12 0.61

6.0 2.0 6.0 2.0 8.0 10 0.26
6.0 1.0 6.0 1.0 4.0 10 0.16
l 30.0 2.0 6.0 0.4 1.6 37 0.19
c 18.0 0.0 6.0 0.0 0.0 47 0.27
56.0 0.0 28.0 0.0 0.0 33 0.30
49.0 25.25 25.5 12.6 50.0 49 0.49
36.0 1.0 1.4 0.04 0.2 52 0.40
bra 0.0 0.0 0.0 0.0 0.0 67 0.22
140.0 5.0 28.0 1.0 4.0 49 0.19
a 0.0 0.0 0.0 0.0 0.0 45 0.49
us 0.0 0.0 0.0 0.0 0.0 37 0.53
0.0 0.0 0.0 0.0 0.0 26 0.68
rpal 0.0 0.0 0.0 0.0 0.0 5 0.72
50.0 50.0 25.0 25.0 100.0 100 0.57
50.0 50.0 25.0 25.0 100.0 63 0.74
lus 50.0 18.0 25.0 9.0 36.0 63 0.61
rsal 50.0 45.9 25.0 22.9 91.6 37 0.74
alanx 100.0 4.0 25.0 0.5 2.0 19 0.57
w ould if they h a d been derived fro m a n arch aeo lo g ical co n tex t u n d ersco res
som e o f the p ro b lem s discussed th u s far. F irst, M arean et al. (1992) did not
begin w ith a set o f com plete skeletons, as evidenced by the M A U values one can
derive from th eir M N E values fo r the pre-rav ag ed assem blage. T he p re
rav ag in g M A U values range from 0 to 28, a n d even d isreg ard in g the skeletal
p a rts k n o w n to have pre-rav ag in g values o f 0, the range is 1.4 to 28. In an
archaeolo gical case one m ight assum e th a t all o f the M A U values sta rte d as the
sam e, m axim u m observed (M N I-b ase d ) value, or, one m ight assum e th a t
skeletal p a rts n o t p resen t archaeologically (such as m an d ib les in this case) were
no t present to be rav ag ed a n d so should n o t be considered d u rin g analysis.
U n fo rtu n a te ly , the lite ra tu re is silent on w hich altern ativ e is the co rrec t one,
an d it is also ra th e r q u iet a b o u t w hich a ssu m p tio n is m ad e by p a rtic u la r
an aly sts in p a rtic u la r situ atio n s. (I have, in com piling the d a ta sum m arized in
F ig u re 7.13, o m itted fro m analysis those skeletal p a rts n o t tallied o r discussed
by th e o riginal inv estigators, a n d used only th o se zero values suggested to be
ac cu ra te by the original investigators.) S econd, if the original, pre-rav ag in g
M A U values are tre a te d as rep resen tin g w h a t w as tra n s p o rte d to a site, those
values d o n o t co rrelate w ith the stan d ard iz ed food utility index fo r com plete
bones (S F U I; M etcalfe a n d lo n e s 1988) (rs = 0.28, P = 0.23) n o r d o they
co rrelate w ith the m axim um stru c tu ra l density value recorded for each skeletal
elem ent (rs= —0.08, P = 0.72). Even o m ittin g the M A U values kn o w n to be
initially zero does n o t result in significant statistical re la tio n s betw een the p re
ravag in g M A U values a n d the S F U I o r bo n e s tru c tu ra l density (rs = 0.19,
P = 0.52; an d , rs = 0.28, P = 0.34, respectively). T his is g o o d because it in dicates
th a t th ere is little chance th a t the p o st-rav ag in g b o n e frequencies will be
co rrelated w ith b o n e density o r the S F U I sim ply because the p re-rav ag in g bone
frequencies w ere co rrelated w ith those variables.
T he th ird thing revealed by M a re a n et al.'s (1992) experim ental d a ta is th a t
th e p o st-ra v ag in g % M A U values, w hen tre a te d as a n arch aeo lo g ical collection
m ig h t be (zero values included in statistical analysis), d o not co rrelate w ith bone
stru ctu ra l density (rs = 0.16, P = 0.5), n o r d o those values co rrelate w ith the
S F U I (rs = 0 .3 6 , P = 0 . 13). If we o m it all skeletal p a rts fo r w hich the surviving
(p o st-rav agin g) % M A U is zero, th en a significant co rre la tio n is fo u n d betw een
% M A U an d bo n e stru ctu ra l density (rs = 0.65, P = 0.04), a n d a significant bu t
slightly w eak er c o rre la tio n is fo u n d w hen only th o se M A U values know n to be
zero in th e p re-rav ag in g assem blage are o m itted (rs = 0.58, P = 0.04). I f all o f the
p o st-rav ag in g % M A U values eq u al to zero are o m itted , as m ig h t h ap p e n
d u rin g analysis o f an archaeological collection, the rem aining bone frequencies
are n o t co rrelated w ith the S F U I (rs = 0.45, / 5= 0.15), alth o u g h the sca tte rp lo t
o f these values has th e ap p e ara n ce o f an unbiased utility g ra p h (F ig u re 7.20).
T he p o in t here sim ply is th a t if M a re a n et al.'s d a ta h ad been derived from an
arch aeo log ical co n tex t, they w ould n o t kn o w w hich (p o st-rav ag in g ) M A U
values eq u al to zero sh o uld be o m itted (th o se skeletal p a rts w ere never present
100
8 0
O)
c
■>
£ 6 0
3
CO
=3
< 4 0
0s
20
SFUI
Figure 7.20. S tandardized fo o d utility index (S F U I) fo r com plete bones (from
M etcalfe and Jones 1988) p lo tted against the % M A U o f surviving (post-
ravaging) sheep bones (from M arean and Spencer 1991). Filled sym bols are o f
elem ents originally in tro d u ced to carnivores; open sym bols are o f bones n o t
introd uced to carnivores.
to be ravaged) an d w hich should be re tain ed (those skeletal p a rts tra n s p o rte d to

the site b u t destroy ed by ca rn iv o re ravaging) in the statistical analysis. This
un d ersco res th e facts th a t (a) in an arch aeo lo g ical co n tex t one typically begins
w ith a m odel o f a co m plete (set of) skeleton(s) as a m odel, a n d (b) differential
tra n s p o rt a n d differential d e stru c tio n o f skeletal p a rts can p ro d u c e the sam e
k ind o f skeletal p a rt profile (L y m an 1985a). C learly, m o re th a n ju st the
econom ic utility, tra n s p o rt index, a n d s tru c tu ra l density fram es o f reference are
re q u ired to help us p ro d u c e e x p lan a tio n s o f skeletal p a r t frequencies.
Summary
T he m ajo r flaw in inferential argum ents based on excavated d a ta is the assum ption,
alw ays im plicit, th a t the absence o f evidence is evidence fo r absence.
(M . B. Schiffer 1987:356)
Since at least the m iddle o f the tw en tieth cen tu ry , arch aeo lo g ists have been
try in g to perfect m eth o d s to explain w hy som e p o rtio n s o f an im al carcasses are
a b u n d a n t a n d o th e r p o rtio n s a re ra re in sites. O n one h a n d , p e rh a p s because o f
the focus o n arch aeo logical collections, T h e o d o re W hite suggested th a t people
m ig h t h ave tra n sp o rte d carcass p a rts differentially based o n the econom ic
value o f th e p arts. T his w as follow ed by D ex ter P erkins an d P atricia D a ly ’s
p ro p o sa l th a t the d istance carcass p o rtio n s h a d to be tra n s p o rte d m ay have
influenced w hich carcass p a rts w ould be tra n sp o rte d . B oth o f these suggestions
su b seq u en tly fo u n d em pirical s u p p o rt in the eth n o arch ae o lo g ic al re co rd (e.g.,

B inford 1978; a n d O 'C o n n e ll et al. 1988,1990; respectively). O n the o th e r h an d ,
p erh ap s because p aleo n to lo g ists are no t co n cerned w ith the h u m an b eh av io ral
m ean in g b u t ra th e r the paleoecological m eaning o f skeletal p a rt frequencies,
m em bers o f this g ro u p o f researchers such as R. D ale G u th rie a n d C. K. B rain
so u g h t e x p lan a tio n s in a n o th e r aren a. P erh ap s, they suggested, differential
p re serv atio n o f skeletal p a rts w as in p a rt a fu n ctio n o f in h eren t stru c tu ra l
p ro p e rtie s o f the bones. T h e stru c tu ra l density o f bo n e p a rts w as one such
p ro p e rty , an d it w as easily m easured.
T o d ate, th ere seem s to h av e been a g re a t deal m o re research on the
econom ic, differential tra n s p o rt type o f ex p lan a tio n th a n on the differential
p re serv atio n ex p lan atio n . F o r exam ple, econom ic utility indices applicable to
h isto ric a rc h a e o fa u n a s have been developed a n d applied to various collections
(H u elsbeck 1989, 1991; L ym an 1987d; R eitz 1986; S chultz a n d G u st 1983;
S inger 1985). H ow ever, no one to the best o f m y know ledge has yet m easured
the s tru c tu ra l den sity o f cattle o r g o at bones, a lth o u g h I have m easu red the
s tru c tu ra l density o f a few d om estic sheep b ones (L ym an 1982b). If this c h a p te r
accom plishes an y th in g , I h o p e it illu strates the necessity o f exploring how we
m ig h t m easu re the influences o f n a tu ra l, n o n -h u m a n tap h o n o m ic processes on
bo n e frequencies, a n d th a t such ex p lo ra tio n m u st keep pace w ith o u r e th n o a r
ch aeo lo g ical research on the influences o f h u m a n ta p h o n o m ic processes.
N u m e ro u s utility indices a n d m easures o f th e s tru c tu ra l density o f bones
have now been p u blish ed. B ut these tw o ta p h o n o m ic variables are n o t
necessarily in d ep en d en t. T h u s, ad d itio n a l a ttrib u te s o f th e fa u n al rem ains are
n ow reg u larly stu d ied , a ttrib u te s such as the ra tio o f long bone shafts to long
bo ne ends, to pro v id e o th e r lines o f evidence th a t tend to su b sta n tia te eith er the
d ifferential tra n sp o rt o r the differential d estru c tio n in te rp re ta tio n . W hile som e
v ariab les such as th e h u m a n selection o f p a rtic u la r b ones fo r m ak in g to o ls have
n o t been exp lored in detail (see C h a p te r 8 fo r a co n sid eratio n o f this problem ),
sufficient d a ta exist to d e m o n stra te th a t a m ere c o rre la tio n betw een one fram e
o f reference an d th e frequencies o f skeletal p a rts in an assem blage is n o t a
sufficient w a rra n t fo r arg u in g the p a rtic u la r fram e o f reference is the reaso n for
th o se frequencies. I have, in this c h a p te r, o u tlin ed the ta p h o n o m ic variables
th a t m u st be co n sid ered, a n d the m ajo r an aly tic techniques fo r building
e x p lan a tio n s o f skeletal p a rt frequencies.
8
BUTCHERING, BONE FRACTURING,

AND BONE TOOLS
Introduction
T he m a n n e r in w hich an im al carcasses a n d skeletal elem ents com e a p a rt o r are

tak en a p a rt is an im p o rta n t ta p h o n o m ic variable. H u m a n s b u tch er anim als
a n d th a t b eh a v io r often, b u t n o t alw ays, v ario u sly m odifies bones. In fact, it
m ight be arg u ed th a t b u tch erin g an im al carcasses is the single greatest
ta p h o n o m ic (an d b io stratin o m ic ) fa c to r in th e fo rm a tio n o f h u m an ly created
fossil assem blages. H u m an s exploit anim als fo r a variety o f reasons, but
basically to e x tra ct resources, w h e th e r energy (food) o r m aterials fo r to o ls o r
clothing. D u rin g th a t ex p lo itatio n , skeletons are d isarticu lated an d bones are
b ro k e n a n d v ariou sly m odified. B ut as we have seen in previous ch ap ters
(especially C h a p te r 6), n o n -h u m a n ta p h o n o m ic processes ca n result in the
d isarticu latio n o f skeletons a n d fra g m e n ta tio n o f bones. In this c h a p te r, I
review these processes, focusing o n th e m o d ificatio n o f skeletal elem ents for
w hich h o m in id s in p a rtic u la r are responsible.
Butchering
T he fragm ents o f A urochs exhibiting very deep incisions, ap p aren tly m ade by an
in strum ent having a w aved edge . . . in w hich I th o u g h t I recognized significant
m arks o f utilization an d flaying o f a recently slain anim al, were o b tain ed from the
low est layer in th e cu ttin g o f the C anal de l’O urcq, n ear P aris . . . I have o btained
analogous results by em ploying as a saw those flint knives fo u n d in the sands o f
Abbeville.
(E. L artet 1860 [1969:122])
T h e term butchering ten d s to h o ld different c o n n o ta tio n s fo r different analysts.

P erh a p s th a t is because it has seldom been explicitly defined. L ym an
(1987a:252) defines butchering “ as th e h u m a n re d u ctio n a n d m o d ificatio n o f an
an im al carcass in to co n su m ab le p a rts .” In this definition, “ c o n su m a b le ” is
“ b ro a d ly co n stru e d to m ean all form s o f use o f carcass p ro d u c ts, including but
no t restricted to c o n su m p tio n o f p ro d u c ts as fo o d ” (L y m an 1987a:252). It is
im p o rta n t to n o te th e inclusion o f the w ord “ h u m a n ” in the definition because
L ym an (1987a:251-252) d istinguishes b u tch erin g fro m w h a t he calls fa u n a l
processing, defined as “ th e re d u ctio n a n d m o d ificatio n o f an an im al carcass
in to co n su m ab le p a r ts .” A s we have seen in earlier ch a p te rs, m an y organism s
294
Butchering, bone fra ctu rin g , and bone tools 295
T ab le 8.1 C arcass resources exploitable by a fa u n a l processor or hum an

butcher (fro m L ym a n 1987a:252)
hide m arro w fat/b lu b b er

hair grease m eat
sinew (tendon, ligam ent) blood juice
bone teeth brains
h o rn /an tler viscera (a n d /o r their contents) hooves
T ab le 8.2 S elected carcass-processing activities directed towards extracting

consum able carcass resources (see Table 8 .1 ) (fro m L ym a n 1987a)
evisceration skinning, hide rem oval

disarticulation, dism em berm ent defleshing, filleting, m eat extraction
bone extraction brain extraction
m arrow extraction blood extraction
bone grease extraction bone juice p ro d u ctio n
periosteum rem oval sinew o r ten d o n rem oval
o th e r th a n ho m in id s process carcasses in to co n su m ab le p a rts (e.g., ca rrio n

insects, b ac te ria, hyenas). O nly ho m in id s, how ever, butcher an an im al carcass.
R ussell (1987:386) states th a t the goal o f b u tch ery “ is rem oval o f m e a t,” and
th u s im plies th a t b u tch ery is solely the rem oval o f m eat from carcasses. This
p ercep tio n is, I th in k , to o n arro w . B inford (1978:63) w rites “ b u tch erin g is in
reality a ta sk o f d ism em b erm en t. T h ro u g h it th e anatom y*of a large an im al is
p a rtitio n e d in to sets o f b ones th a t m ay be ab a n d o n e d , tra n sp o rte d , o r allocated
to different uses.” T his aligns well w ith m y definition if the w ords “ sets o f
b o n es” are ta k e n to d en o te archaeologically visible results o f a process
involving n o t ju s t b ones b u t m uscle, fa t, a n d hide as well as bones. B inford
(1978:48) also states th a t “ b u tch erin g is n o t a single act b u t a series o f acts
beginning w hen th e an im al is killed an d co n tin u in g a t varying ju n c tu re s until
the an im al is to tally co n su m ed o r d iscard ed ." T h u s while tra n sp o rt o f skeletal
p a rts (see C h a p te r 7) m ay occu r betw een b u tch erin g acts, tra n s p o rt is not
technically a p a rt o f b u tchering. N o r does b u tch erin g include the h u m an
beh av io rs o f co o k in g a n d co n su m p tio n . O f p a rtic u la r interest, given the
definition, are (1) the carcass resources a b u tch er (or any fau n al processor)
seeks to exp lo it (T able 8.1), a n d (2) the processes o r h u m a n activities necessary
to re n d er a carcass in to those resources (T able 8.2). T h ere are m an y kinds o f
p o te n tia l resources to ex tract from , say, a 200 kg artio d a cty l, a n d m any
processes th a t can be used to ex tra ct them . E th n o arc h aeo lo g ica l research
ind icates there are m an y facto rs w hich influence th e p a rtic u la r resources
exp lo ited a n d th e m a n n e r in w hich they are exploited (T able 8.3).
B utchering con sists o f a set o r series o f sets o f h u m a n activities directed
T ab le 8.3 Factors tha t influence utilized butchering techniques (m odified and

expa n d ed fr o m L ym a n 1987a:253)
N atu ral factors

Prey anim al: taxon, size o f carcass, age an d sex o f anim al, h ealth statu s o f anim al
N atu re o f procurem ent:
Scavenged: con d itio n o f carcass (rancid?), com pleteness o f carcass
H unted: num b er o f anim als killed, n u m b er o f people present, type o f kill site (location,
accessibility, geological conditions, geographic conditions)
Spatial relationships o f kill site, h ab itatio n site, and processing areas
Tim e o f day: heat, a m o u n t o f light rem aining, w eather
Season o f the year: heat, precip itatio n (type an d am o u n t)
D ietary statu s o f people: im m ediate versus long-term n u tritio n al needs
C ultural factors
Technology: available versus used, cu rated versus expedient tools
G u stato ry preferences
P rep aratio n and consum ption: co oking vessel size, preserv atio n technology (if any), storage
capabilities and kinds
E thnic gro u p involved: first anim al rituals, kin present a t kill site versus kin present at
h ab itatio n site, selective hun tin g
to w ard s th e ex tra ctio n o f co n su m ab le resources fro m a carcass. It has a

te m p o ra l d u ra tio n , m ad e u p o f the set a n d o rd e r o f activities ca rried o u t to
e x tra ct resources fro m a carcass. T he actio n s a n d activities involved in
b u tch erin g h av e been term ed th e butchering process o r butchering techniques,
a n d th e results o f th e process have been term ed th e butchering pattern (L ym an
1987a:252). Z o o arch ae o lo g ic al research betw een 1950 a n d 1980 w as aim ed at
in ferrin g th e h u m a n b eh av io rs m ak in g u p a b u tch erin g process a n d evidenced
by th e b u tch erin g p a tte rn as d eterm in ed fro m an assem blage o f fa u n al rem ains
(e.g.. W h eat 1972). A rule o f th u m b in such en d eav o rs w as to presum e th at
p reh isto ric b u tch ers utilized the m o st p ra g m a tic b u tch erin g process, an d
evidence o f b u tch erin g co u ld be in terp re ted as reflecting efficient h u m an
activity (e.g., Spiess 1979; W hite 1956).
P rio r to th e ex ten d ed discussion o f ta p h o n o m ic issues in the late 1970s an d
early 1980s, analysts interested in studying p reh isto ric b u tch erin g techniques
assu m ed m an y a ttrib u te s o f b o n e m o d ificatio n could be a ttrib u te d to h u m a n
activity. T he a ssu m p tio n seems to have been fo u n d e d largely o n the basis o f the
asso c ia tio n o f th e an im al rem ains w ith artifa cts (B inford 1981b; L ym an
1987a). W ith the increasing ta p h o n o m ic aw areness o f the late 1970s, z o o a r
chaeo lo g ists m o re frequently studied m odification a ttrib u te s displayed by
b o n es in a tte m p ts to estab lish clearly th o se m odifications th a t co u ld be
u n am b ig u o u sly ascrib ed to h u m a n activities. W ays to establish th e id en tity o f
som e m ark s as b u tch erin g m ark s o r stone to o l cut m ark s w ere extensively
discussed in the early 1980s (B unn 1981, 1982; O lsen 1988; P o tts 1982; P o tts
an d S h ipm an 1981; S h ipm an 1981a, 1981b; W alker an d L ong 1977). D iscus
sions a n d ex perim ents co n cerning m ark s m ad e d u rin g the fractu rin g o f bones
by v ario u s ag ents a n d the kinds o f fractu res p ro d u c e d by them w ere presen ted
(e.g.. B onnichsen 1979; M o rla n 1980; S hip m an et al. 1981). Som e o f the latter
were co n cerned w ith identifying bo n e to o ls th a t h ad been only m inim ally
m odified. T he re m a in d er o f this c h a p te r presents an overview o f research
con cern in g these topics.
Butchering m arks
W hile n o t w ith o u t p reced en t (e.g., G u ild ay et al. 1962), it w as largely as a result
o f the w o rk o f P at S h ipm an (1981a, 1981b, 1983, 1986a, 1986b; S hip m an and
R ose 1983a, 1983b, 1984) th a t analysts began exam ining the m icroscopic
m o rp h o lo g y o f v ario u s scratches on bones. S h ip m a n ’s w ork in d icated th a t
m ark s m ad e by sto n e to o ls are m o rp h o lo g ically d istin ct fro m th o se m ade by,
fo r exam ple, ca rn iv o re a n d ro d e n t teeth (F ig u re 6.1). S hip m an arg u ed th a t cut
m arks m ad e by sto n e to o ls will (1) be V -shaped to U -sh a p ed in cross section,
b u t ten d to w a rd the fo rm er, (2) be elo n g ate, (3) have m ultiple, fine p arallel
striae o n th e w alls o f the m ark , a n d (4) som etim es display w h at she called
“ sh o u ld er effects” (sm all striae parallel to the m ain striatio n ) a n d /o r “ b a rb s ”
(w ould have a sm all b a rb o r h o o k at one end). T hese a ttrib u te s, p artic u la rly the
th ird one, h ave becom e the m a jo r crite ria fo r identifying cu t m ark s, b u t are
seldom used alo ne to identify scratches o n b o n es as cu t m ark s. T h a t is because
som e o f these a ttrib u te s can be created by n o n -h o m in id ta p h o n o m ic processes
such as the gnaw ing actio n o f ca rn iv o re teeth (E ick h o ff a n d H e rm a n n 1985)
an d tram p lin g o f bones on san d y su b strates (B ehrensm eyer et al. 1986, 1989;
F io rillo 1989; H aynes an d S ta n fo rd 1984). H aynes (1991:163) argues th a t the
b u tch ery m ark s S h ip m an p ro d u c ed experim entally w ere not m ad e d u rin g the
butchering o f a n an im al carcass, b u t ra th e r w ere deliberately m ade to be visible,
o ften on defleshed bones. T hus, the criteria O lsen a n d S h ip m an (1988) list, for
exam ple, as d istin guishing tram p lin g -g en e rated fro m b u tch erin g m ark s m ay be
invalid as the la tte r w ere n o t created d u rin g the process o f ex tra ctin g resources
from a carcass, especially by b u tch ers w ith a w o rk in g know ledge o f the
an a to m y o f the an im al they w ere b u tch erin g a n d w ith som e co n cern o f
p reserving a sh a rp to o l edge. N onetheless, the m icro -m o rp h o lo g ica l a ttrib u te s
S h ip m an p ro p o se d have com e to be the ones an aly sts exam ine first. A nalysts
th en typically exam ine ad d itio n a l a ttrib u te s o f the m ark s, p artic u la rly th eir
a n a to m ica l lo catio n a n d o rie n ta tio n (L y m an 1987a; e.g., G ib e rt a n d Jim enez
1991; N o e -N y g a a rd 1989). (Som e researchers suggest th e directio n in w hich the
cu t w as m ad e a n d the h an d ed n ess o f the b u tch er can be asc ertain ed by
m icro sco pic ex a m in a tio n o f the cu t m a rk [B rom age a n d Boyd 1984]. O th er
researchers suggest th a t co m p u te r-g e n e ra te d th ree-d im en sio n al m odels o f
surface irreg u larities o n bones such as b u tch erin g m ark s m ay aid in d e te rm in
ing w h eth er o r n o t a b o n e w as fresh w hen it w as cu t [D uring a n d N ilsso n 1991].)
G u ild ay et al. (1962:63) suggest th a t a c u t m a rk can be identified using tw o

criteria. F irst, a m a rk m u st occu r o n “ specim en afte r specim en at precisely the
sam e lo catio n o n the b o n e .” Second, they suggest th a t each categ o ry o f m ark
sh o u ld h av e a d etectab le a n a to m ic a l p u rp o se o r reaso n fo r o ccu rrin g w here it
does. T he second crite rio n is related to th e n o tio n th a t anim als will be
b u tch ered efficiently, a n d as a result each m ark has a fu n ctio n o r p u rp o se for
existing. T h u s we read statem en ts such as “ an im als w ere b u tch ered p rim arily at
the jo in ts since this is the easiest m e th o d ” (R ead 1971:53) a n d “ the sim plest way
for m en w ith o u t po w er saws to divide a carcass is to follow the n a tu ra l p ath s o f
m uscles a n d cu t them only w here they need cu ttin g , at their origins and
in sertio n s” (G ilb ert 1979:152).
T h u s th e ad d itio n a l crite ria used by zo o a rch a eo lo g ists to identify m ark s on
bones as cu t m ark s involve the a n a to m ic a l placem ent an d o rie n ta tio n o f a
m ark , a n d the fu n ctio n o f a p a rtic u la r categ o ry o f m ark suggested by its
lo catio n a n d o rie n ta tio n . B inford (1981 b:46^47) suggests c u t m ark s differ from
gnaw in g m ark s (see C h a p te r 6) in th a t th e fo rm e r “ rarely follow the ‘c o n to u r’
o f th e b o n e [surface]" w hereas the la tte r d o (see N o e -N y g a a rd 1989 fo r a recent
overview ). C u t m ark s, B inford (1981 b:47) argues, “ generally result from three
activities: (a) skin n ing , (b) d isarticu latio n , a n d (c) filleting.” S kinning cut
m ark s are fo u n d a ro u n d the sh aft o f low er legs a n d ph alan g es, a n d along the
low er m arg in s o f the m an d ib le o r o n the skull. D isa rtic u la tio n cut m ark s occur
o n th e “ edges, o r a rtic u la r surfaces o f the ends o f long bones, a n d o n the
surfaces o f v erteb rae o r pelvic p a r ts ” (B inford 1981 b:47). F illeting cu t m ark s
g enerally p arallel th e lo n g axis o f th e bone. H ere, B inford is suggesting som e
general rules for identifying the fu n ctio n o f a cu t m ark fro m a ttrib u te s o f its
lo catio n a n d o rien ta tio n . M o st an aly sts to d a y re co rd the lo catio n a n d o rien
ta tio n o f cu t m ark s by illu stra tin g each m a rk on line draw ings o f skeletal
elem ents, m u ch like th e draw ings in F ig u res 7.4 a n d 7.5. I have fo u n d th a t
several views o f each elem ent, m inim ally a n a n te rio r a n d a p o ste rio r view o f
long b ones, o ften are sufficient (e.g., L ym an 1991a).
T h a t people use h am m ersto n e s to b reak bones m u st also be considered in
any d iscussion o f b u tch erin g m a rk s (the k in d s o f fractu re s gen erated are
discussed later in this ch ap ter). M o rla n (1980:50) suggests th a t “ the u p p e r size
lim it o f carn iv o re to o th c o n ta c t a rea is sm aller th a n the u p p e r lim it o f a
h am m ersto n e c o n ta c t a re a .” B u n n (1982:44) no tes th a t “ a p o in ted h a m m e r
sto n e ca n p ro d u c e a sm all, to o th -sized in d e n ta tio n w here the bo n e b re ak s, a n d
a series o f o v erlap p in g to o th -in d u c e d in d e n ta tio n s ca n resem ble the b ro a d ,
a rc u a te in d e n ta tio n s w hich ro u n d e d h am m ersto n es m ore typically p ro d u c e .”
P o tts (1982:215) suggests th a t a h a m m ersto n e will p ro d u c e bone flakes th a t are
b ro a d e r th a n they are long a n d have p la tfo rm s th a t are b ro a d e r th a n they are
thick. H e believes th a t the resu ltin g flake scars will be b ro a d , arcu ate
in d en tatio n s, w hereas carn iv o re to o th -c re a te d flake scars will a p p e a r as sm all
notches. B ased o n a d d itio n a l ex p erim en tal w o rk , P o tts (1988:101) suggests
th ere is m a jo r o v erlap betw een the le n g th -w id th a ttrib u te s o f h y en a-p ro d u ced

b o n e flakes an d hu m an -w ield ed h a m m e rsto n e -g en e rated bo n e flakes. G iven
this case o f equifm ality (see also F igure 6.25 a n d associated discussion), it is
fo rtu n a te indeed th a t B lum enschine an d S elvaggio’s (1988, 1991) experim ental
w o rk w ith to o th -c re a te d pits a n d h am m ersto n e -g en e rated percussion m ark s
(see C h a p te r 6) p rovid es crite ria th a t seem to be distinctive o f b o th kinds o f
m ark s. T h e a n a to m ica l lo catio n s o f h am m ersto n e -g en e rated im p act scars m ay
be reco rd ed o n line d raw ings o f b ones ju s t like cu t m arks.
A conceptual fra m e w o rk f o r analyzing butchering

W ith in th e co n tex t o f an alyzing b u tch erin g p ractices, transport o f carcasses an d
carcass p a rts can be defined as the h u m an m ovem ent o f an im al p ro d u c ts from
one p o sitio n on th e lan d scap e to a n o th e r p o sitio n (see C h a p te r 7). T ra n s p o rt o f
anim al b o d y p a rts usually im plies ca rry in g such p a rts from the p ro c u rem en t
lo catio n to a resid en tial o r c o n su m p tio n lo catio n . B u tch ery a n d tra n s p o rt are
n o t in d ep en d e n t processes, n o r are they to tally a n d alw ays in terd e p en d en t.
E th n o arc h aeo lo g ica l d a ta in d icate th a t once a n an im al carcass is p ro c u red ,
h u m an s are faced w ith a m y riad o f decisions. T ra n s p o rt logistics m ay o r m ay
n o t d ic ta te if a n d how the carcass is red u ced o r b u tch ered in to tra n s p o rta b le
pieces (B inford 1978; B unn et al. 1988; O ’C onnell a n d M arsh a ll 1989;
O ’C on nell et al. 1988, 1990). Sm all carcasses m ay be tra n s p o rte d w hole,
w ith o u t b u tch ery (H u d so n 1990; Y ellen 1991 a:5), a n d th u s sm all carcasses m ay
“ be b ro u g h t to cam p skeletally com plete o r virtu ally so ” (B artram et al.
1991:102). L arge carcasses are typically b u tch ered in to sm all, tra n sp o rta b le
pieces only som e o f w hich m ay in fact be tra n s p o rte d (e.g., O ’C onnell et al.
1990, 1992). In som e cases n early all o f the skeleton o f p ro b o sc id e an s - the
largest o f terrestrial v erte b rates - is tra n sp o rte d ; in o th ers n early the entire
sk eleton is n o t tra n sp o rte d (C ra d e r 1983; F ish e r 1992).
T he zo o a rch aeo lo g ical lite ra tu re is, o n one h a n d , full o f discussions a b o u t
differential tra n s p o rt o f carcass p o rtio n s. T hese discussions largely concern
w hy ce rtain p o rtio n s are tra n s p o rte d a n d o th ers n o t, a n d h ow to recognize such
tra n s p o rt in arch aeo lo g ical co n tex ts (see C h a p te r 7). O n the o th e r h a n d the
lite ra tu re is relatively silent on th e in terp lay o f tra n s p o rt logistics a n d bu tch ery
(see B a rtra m et al. 1991; G iffo rd -G o n z alez 1989a; Y ellen 1991a, 1991b; for
recent a d d itio n s to the discussion). T h u s it is im p o rta n t to in teg ra te the
b u tch erin g process w ith the processes o f tra n sp o rt. B utchering typically
involves a set o f activities th a t, betw een the tim e o f carcass p ro c u re m e n t a n d
final d isp osal o f carcass p o rtio n s, occur in varying o rd e rs an d frequencies o r
intensities fo r different carcasses.
B in fo rd ’s (1978, 1981b) ca te g o riza tio n o f the v ario u s b u tch erin g activities as
skinning, d ism em b erm en t o r d isarticu latio n , a n d filleting o r rem oving m eat
fro m b o n es is a useful one th a t ca p tu res the essence o f the m ain activities o f a
h o m in id b u tch er. E x trac tio n o f viscera, b lo o d , b rain s, m arro w , grease, bone,

an d sinew , a n d p erio steu m rem oval are o th e r activities th a t m ight be c o n
sidered su b sid iary to , in p a rtic u la r, d ism em b erm en t a n d filleting o r food
e x tra ctio n (L y m an 1987a). A s im plied earlier, all o f these activities have the
p o te n tia l to d am ag e bones; th a t is, all can p ro d u c e w h a t are typically called
b u tch ery m ark s. B ecause b u tch ery o f an an im al carcass is a process, it can be
conceived o f as a set o f stages o f g re ater o r lesser discreteness. I have fo u n d the
follow ing stages to p ro v id e a useful co n c ep tu al fram ew ork: (1) kill-butchery
stage, follow ed by tra n sp o rt o f variously b u tch ered carcass p o rtio n s, (2)
secondary butchery stage, p erh ap s follow ed by a n o th e r tra n sp o rt episode (even
if only re d istrib u tio n w ithin the site o f c o n su m p tio n ), a n d (3) fin a l butchery-
consum ption stage. T h e im p o rtan ce o f these stages is th a t they im ply th a t
b u tch ery m ark s will be ad d ed to th e b ones o f a carcass as it passes th ro u g h the
b u tch erin g stages. O ne m ight expect, then, few er m ark s to occu r on bones
d isposed o f early in the b u tch ery process (p rio r to tra n s p o rt) w hereas bones
re ta in ed (a n d tra n sp o rte d ) have th e o p p o rtu n ity to u n d erg o fu rth e r b u tch ery
p rio r to co n su m p tio n an d m ay display a d d itio n al m ark s th a t w ere created
d u rin g final processing o r co n su m p tio n .
T h o m as an d M ay er (1983) inferred, in the case o f the G atecliff S helter
H o riz o n 2 b ig h o rn sheep (O vis canadensis), th a t they w ere dealing w ith w hat I
term a seco n dary b u tch ery site based on vary in g frequencies o f skeletal p arts.
T h a t inference in tu rn resulted in th eir co n c ep tio n o f the b u tch erin g m ark s they
observed as rep resen tin g early stages o f b u tch erin g ra th e r th a n final o r
in term ed iate stages. T h o m as a n d M ay er (1983) use the co n cep t o f archaeologi
cal m onitoring perspective in th eir discussion. Sim ply p u t, the m o n ito rin g
perspective is th e arch ae o lo g ist’s o b serv atio n a l p o sitio n w ith in a p re h isto ric
h u m a n b eh a v io ral co ntex t. T h o m a s a n d M ay e r w ere exam ining th e skeletal
p a rt frequencies resulting fro m a p a rtic u la r tra n s p o rt strategy. T hey w ere in the
geo g rap h ic place aw ay from w hich bones w ere tra n sp o rte d , a n d th u s p a rtic u la r
bones w ere ra re w h ereas o th ers w ere frequent. I f th eir m o n ito rin g perspective
h ad been in the lo catio n to w hich bones h ad been tra n sp o rte d , the kinds o f
bones th ey observed to be ra re w ould have been a b u n d a n t a n d th e kinds o f
bones th ey observed to be a b u n d a n t w ould h av e been rare. W ith in th e co n tex t
o f the b u tch ery stages, the a n a ly st’s m o n ito rin g perspective will p erh ap s
influence th e an a to m ica l p lacem en t a n d frequency o f b u tch erin g m ark s. Y ellen
(1977), B inford (1984a), an d G iffo rd -G o n zalez (1989b, 1989c) re p o rt e th n o a r-
chaeological d a ta fo r w h at I term final b u tch ery -co n su m p tio n sites. T h eir d ata
in d icate th e bones, a n d the b u tch ery -re la ted d am age they have su stain ed , in
such co n tex ts rep resen t the final stages o f processing (see also Y ellen 1991a).
T h u s it seem s th a t n o t only m ay the a n aly tical m o n ito rin g perspective influence
the kin d s o f skeletal p a rts an arch ae o lo g ist m ay find (see C h a p te r 7), it m ay also
influence th e k in ds a n d frequencies o f b u tch ery m ark s on th o se skeletal p arts.
H o w a carcass is b u tch ered a n d tra n s p o rte d are n o t only influenced by o n e’s
m o n ito rin g perspective o r w hich stage o f b u tch ery is rep resen ted by the studied
assem blage, they are influenced by a n u m b e r o f o th e r facto rs as well. G ifford-
G o n zalez (1989b) co rrectly observes th a t the a n a to m ica l d istrib u tio n an d
frequ ency o f cu t m a rk s re su lt fro m costs a n d benefits o f tra n s p o rt, the
techn o lo g y used to b u tch er anim als, an d co o k in g a n d c o n su m p tio n practices
(see also L y m an 1987b). H e r (G iffo rd -G o n zalez 1989a) eth n o arch aeo lo g ical
d a ta suggest the size an d a n a to m y o f a carcass m ay co n stra in b u tchery
practices, b u t fo r a n y carcass size a n d an a to m ic a l categ o ry there are a n u m b er
o f different d ism em b erm en t a n d defleshing strategies th a t m ay be used.
Sim ilarly, the tra n s p o rt logistics involved in m oving carcasses an d carcass p arts
fro m p rim a ry a n d seco n d ary kill-b u tch ery sites to the locus o f final b u tchery
an d co n su m p tio n will influence how an an im al is bu tch ered an d th u s the
bu tch ery -re la ted d am ag e inflicted on bones. B inford (1978, 1981b) labels those
facto rs w hich influence how an anim al is b u tch ered a n d tra n s p o rte d con
tingency fa c to rs (see T ab le 8.3 fo r a list o f these).
U nless one is d ealin g w ith a kill-b u tch ery site w here relatively m inim al
co n su m p tio n occu rred , the arc h a e o fa u n a l rem ains w ith w hich we typically deal
are p ro b a b ly c o n su m p tio n w aste, such as those rem ains fo u n d in village or
cam p sites. M u ch as th e lithic an aly st m ay focus on the debitage resulting from
p ro d u c in g a to o l, the zo o a rch a eo lo g ist focuses on the debitage resulting from
p ro d u cin g , typically, a m eal, p a rtic u la rly w hen th e site is one a t w hich
co n su m p tio n w as relatively m o re freq u en t th a n initial processing. T h e effects
o f c o n su m p tio n on an archaeological bone assem blage can be significant if, for
exam ple, processing includes bo n e frag m en tatio n , th u s reducing the identifia-
bility o f specim ens (L y m an a n d O 'B rien 1987).
If th e subject o f study is frequencies o f bu tch ery m ark s, then a m ajo r
a ssu m p tio n m u st be g ra n te d (L y m an 1992c, 1993a). T he assu m p tio n consists
o f the prem ise th a t, given som e set o f b ones X, som e subset X ' o f those bones
will be b u tch ered , a n d o f th o se b u tch ered bones som e subset X " will sustain
d am ag e in the fo rm o f b u tch ery m ark s. T he a ssu m p tio n is th a t som e p ro p o r
tio n o f each skeletal elem ent w as b u tch ered a n d som e lesser p ro p o rtio n will
display b u tch ery m ark s, a n d th o se p ro p o rtio n s will directly a n d positively
co v ary a t least a t a n o rd in a l (b u t p e rh a p s n o t an interval) scale. N o te th a t while
I h av e said “ b ones are b u tc h e re d ," this is a s h o rth a n d fo rm o f saying “ carcasses
a n d /o r carcass p a rts, in cluding hide, viscera, m uscle, a n d o th e r soft tissues are
bu tch ered ; strictly speaking, b ones are only b u tch ered w hen they are b ro k e n
fo r m a rro w o r grease ex tra c tio n .” I use th e s h o rth a n d fo rm because it is the
bones, n o t the so ft tissues, carcasses, o r lim bs th a t we stu d y in archaeological
co ntexts.
I f b u tch ery m a rk s are ep ip h en o m en a, th a t is, if they are in som e sense an
un in ten d ed , acciden tal, fo rtu ito u s, o r in cid en tal result o f b u tch ery activities,
then frequencies o f b u tch ered bones are p o ten tially am b ig u o u s in d icato rs o f
the q u a n tita tiv e aspects o f h u m a n behaviors, a n d th u s term s such as “ b u tch ery
p a tte r n ” w o u ld be in a p p ro p ria te given its h u m a n b eh av io ral im p licatio n s (see

L ym an 1993a fo r a d d itio n a l discussion). A fictitious exam ple m akes clear the
significance o f the assu m p tio n . L et us say th a t 10 fem o ra a n d 10 h u m eri were
av ailab le fo r b u tch ery (X = 20). O f those, six fem o ra a n d five h u m eri w ere
bu tch ered (X ' = 11). O f th ose b u tch ered elem ents, fo u r fem o ra a n d tw o hum eri
display b u tch ery m ark s (X" = 6). T he critical statistical relatio n here is th a t
m ore fem o ra th a n h u m eri w ere b u tch ered , a n d m ore fem ora th a n hum eri
actu ally display arch aeo logically visible b u tch ery m arks; th a t is, even th o u g h
60% o f th e fem o ra a n d 50% o f the h u m eri w ere b u tch ered , 67% o f the
b u tch ered fem o ra a n d 40% o f th e b u tch ered h u m eri display archaeologically
visible b u tch ery m ark s. I f this re la tio n sh ip fails to ho ld , such as a case in w hich
six o f th e ten fem o ra a n d five o f th e ten h u m eri w ere b u tch ered b u t one fem ur
(17% o f th o se actu ally b u tch ered ) a n d fo u r h um eri (60% o f th o se actually
b u tch ered ) disp lay b u tch ery m ark s, th en an aly sts atte m p tin g to discern
b u tch ery in ten sity a n d h u m a n b eh a v io ral p a tte rn s m ay be m isled by the
arch aeo lo gically visible reco rd o f b u tch ery practices. I co n sid er the issue o f
p re serv atio n o f b u tch ery m ark s below .
It is im p o rta n t to realize th a t som e u n k n o w n p ro p o rtio n o f b u tch ered bones
(X ') m ay be archaeologically invisible; only th a t p o rtio n o f X ' w hich consists o f
b u tch ery -m ark e d b ones (X ") is archaeologically visible. T herefore, the
a ssu m p tio n th a t X ' a n d X " are positively a n d directly related is generally
o p eratio n aliz ed by an aly sts w ho p resen t the p ro p o rtio n (percentage) o f all
b ones (X) th a t display b u tch ery m ark s (m akes u p X "). T h a t is, fo r an y one k ind
o f skeletal elem ent (i), th e an a ly st n o rm ally presents the ab so lu te frequency o f
b u tch ery -m ark e d b ones (X"i) a n d the p ro p o rtio n a l frequency o f butchery-
m ark e d b o n es = (X"i/Xi) 100. In te rp re ta tio n o f th o se frequencies usually
involves co m p arin g th o se values fo r different skeletal p a rts (i), o r ta x a o r bone
assem blages (j); X",j c o m p ared to X " i+ y , o r (X'^j/Xy) 100 c o m p ared to (X ";j+J
Xjj + j) 100, respectively. H ow ever, w h a t one actu ally is seeking to m easu re is the
h u m a n b eh a v io r in d icated by X'y, w hich m ay, o f course, be archaeologically
invisible. T h u s th e assu m p tio n is re q u ired th a t the frequency o f butchery-
m ark e d b ones in an assem blage (Xjj") is directly a n d positively related to the
n u m b e r o f b o nes in th a t assem blage th a t w ere in fact b u tch ered (Xjj') in o rd e r to
infer h u m a n beh avio rs.
If th e frequencies o f b u tch ery -m ark e d b ones are n o t positively co rrelated
w ith th e frequencies o f b u tch ered b ones th en frequencies o f b u tch ery -m ark e d
bones, eith er ab so lu te o r p ro p o rtio n a l, are p o ten tially am b ig u o u s in d icato rs o f
th e q u a n tita tiv e aspects o f h u m a n b u tch erin g behaviors. W hen, w h eth er, a n d
how o ften the assu m p tio n th a t the frequency o f b u tch ery -m ark e d bones is
positively co rrelated w ith th e frequency o f b u tch ered bones is w a rra n te d is
a n o th e r m a tte r. B ecause eth n o arch ae o lo g ic al research has n o t yet been p u b
lished on this issue, an aly sts in terested in analyzing a n d in te rp re tin g freq u en
cies o f b u tch ery m ark s m u st m ak e the a ssu m p tio n th a t th e tw o variables are
Butchering, hone fra ctu rin g , and bone tools 303
positively co rrelated . M ak in g the a ssu m p tio n then leads to co n sid eratio n o f

how to co u n t b u tch ery m arks.
Q uantification o f butchering m arks
T h a t th e frequency o f b u tch ery m ark s is in fact a significant v ariab le in the

analysis o f p reh isto ric b u tch ery is a p p a re n t in the literatu re. K o o y m an
(1984:54) re p o rts o n th e p reh isto ric b u tch ery o f m o a (a large flightless b ird o f
the tax o n o m ic o rd e r D in o rn ith ifo rm es) carcasses at the site o f O w ens F erry in
N ew Z ealan d . H e suggests cu t m ark s on m o a fem ora “ are to o few a n d scattered
to in d icate any g eneral processing strategy; they p ro b a b ly in d icate occasional
altern ativ e [butchery] p a tte rn s.” T h a t is, the in fre q u en t occurrence o f cuts on
the fem ora relative to frequencies o f cuts o n o th e r bones suggests to K o o y m an
th a t m o a legs w ere n o t all b u tch ered in like m an n ers. R ap so n (1990:287)
calcu lates the n u m b e r o f “ cu t m ark s p er u n it a r e a ” o f bo n e specim en surface to
pro v id e a “ m easu re o f cut m a rk in ten sity w hich c o n tro ls fo r differences in
[bone] specim en size” a n d size o f the b u tch ered tax o n . H e th en identifies
differences an d sim ilarities in b u tch ery based on v aria tio n in the density
(frequency per unit area) o f cut m ark s on bones o f anim als o f different sex an d
taxon.
B unn an d K roll (1986:432) state th a t “ frequencies o f cu t m ark s on different
skeletal p a rts can be directly linked to the skinning, d isarticu latio n , an d
defleshing o f carcasses.” T hey infer, fo r exam ple, th a t because 20% o f the
b ones m ak in g up the elbow jo in t in th eir sam ple display cu t m ark s, “ rep eated
d ism em b erm en t o f the elbow jo in t is d o c u m e n te d .” B inford (1986:446) sug
gests th a t high frequencies o f cu t m ark s on long bone shafts in d icate “ extrem e
difficulty in p rocessing alre ad y p artia lly desiccated lim b p a rts th a t h ad been
previously ravag ed by c a rn iv o res,” th u s there was little m eat for the b u tch ers to
ex tract. In co n tra st, B unn a n d K roll (1986:450, 449-450) arg u e th a t “ the
presence o f m an y slicing m a rk s o n once-m eaty lim b bones in d icates th a t
h o m in id s rem oved su b stan tial q u an titie s o f m eat from the b o n es” because such
high frequencies “ are m o st likely to o ccu r w hen it is difficult o r im possible to see
w here the b o n e is, as w hen a com plete, m eaty lim b b o n e is being defleshed.”
B inford (1988:127) suggests th a t “ the n u m b e r o f cu t m ark s, exclusive o f
d ism em b erm en t m ark s, is a fu n c tio n o f differential in v estm en t in m e a t o r tissue
re m o v a l.” T he g re ater th e effort invested, the m o re cu t m ark s will result. H e
fu rth e r suggests “ the n u m b ers o f cu t m ark s a n d th eir frequencies o n different
b ones m ay reflect very different processing o p e ra tio n s;” skinning m ay result in
m ore m a rk s on m etap o d ials a n d few er on fem o ra, all else being equal (B inford
1988:128).
R egardless o f w ho is co rrect in the above, it is clear th a t th e frequency o f cut
m ark s is felt by these an aly sts to be an im p o rta n t variable th a t in som e m an n er
reflects h u m a n b eh aviors. In all o f the preceding exam ples we find im plicit
a d o p tio n o f th e a ssu m p tio n th a t frequencies o f b u tch ery -m ark e d b ones (X")

will be d irectly a n d positively co rrelated w ith the frequencies o f bu tch ered
bones, m ark e d a n d u n m a rk e d (X '). G iven the in terp retiv e im p o rtan ce ascribed
to frequencies o f b u tch ery -m ark e d b ones a n d frequencies o f b u tch ery m ark s on
p a rtic u la r b ones, it is m a n d a to ry th a t som e explicit m eth o d be used to co u n t
th em (M a ltb y 1985a). T he n u m b e r o f b o n e specim ens displaying bu tch ery
m ark s is a ra th e r straig h tfo rw a rd q u a n tita tiv e unit: sim ply tally the N IS P
(n u m b er o f identified specim ens) th a t have o r display b u tch ery m ark s. But is
N IS P an a p p ro p ria te unit fo r such quan tificatio n ?
T he co u n tin g u n its fo r tally in g b u tch ery m a rk frequencies typically,
a lth o u g h n o t alw ays, define som e lim ited an a to m ica l space on a skeletal
elem ent. T ypically the to ta l frequency o f th a t space (X , above) an d the
freq u ency o f th a t space w ith b u tch ery m ark s (X ", above) are tallied, regardless
o f h o w frag m en ted o r com plete th e ind iv id u al specim ens rep resen tin g th a t
space are. T h a t is so because th e an a ly st is in terested n o t ju s t in the frequency o f
bu tch ery m ark s, b u t in th eir a n a to m ica l d istrib u tio n because th a t d istrib u tio n
relates to b u tch erin g fu n ctio n s (e.g., L ym an 1987a; M altb y 1985a, 1985b).
T h u s one tallies X a n d X " fo r un its such as lateral distal tib ia o r p ro x im al
g re ater tro c h a n te r o f th e fem ur. T he term s “ specim en” a n d “ N IS P ” th u s have
different c o n n o ta tio n s in b u tch ery analysis th a n in, say, the analysis o f
tax o n o m ic a b u n d a n ces (see C h a p te r 4 fo r discussion o f the latter). In the latter,
a specim en o r an N IS P o f o ne represents som e discrete archaeological object,
w h eth er it is a com plete fem ur o r ju s t th e d istal end o f a fem ur. In b u tch ery
analysis, the p o rtio n o f the specim en displaying b u tch ery m ark s is tallied. F o r
exam ple, tw o o f five recovered d istal h um eri m ig h t display b u tch ery m ark s,
resu lting in the con clu sion th a t 40% o f the recovered distal hum eri have
b u tch ery m ark s. T he fact th a t th ree o f th o se five h u m eri specim ens are
co m p lete b ones, one consists o f the distal end a n d shaft, a n d th e fifth one
consists o f ju s t th e distal condyle is irre le v an t to tallying frequencies o f
b u tch ery m ark s. If the sh aft o f the fo u rth specim en displays a cut m ark , it is
tallied in th e “ d istal h u m eru s s h a ft” category. T h u s the m ean in g o f the term s
“ sp ecim en ” a n d “ N IS P ” in the analysis o f b u tch ery resides in th e an a to m ica l
area sense ra th e r th a n in the archaeologically discrete object sense.
T he n u m b e r o f b u tch ery m ark s is a p o ten tially difficult co u n tin g unit to
op eratio n alize. I tally each discrete, n o n a d ja c e n t ( > 1 cm a p a rt) a n d n o n
o v erlap p in g m a rk as a n instance o f force a p p lic a tio n (L y m an 1987a, 1992c,
1993a). W hile so m ew h at subjective (e.g., a clu ster o f striae is tallied as one
in stan ce even th o u g h m ultiple instances o f force ap p lica tio n are clearly
rep resen ted ), this seem s to be th e p ractice generally follow ed. I p refer this
p ro c ed u re because w hen striae in p a rtic u la r, a n d flake scars to a lesser degree,
ov erlap , c o u n tin g th em individually m ay be im possible as one m a rk tends to
d estro y traces o f a m a rk m ad e previously (F ig u re 8.1a).
U se o f the p ro p o rtio n a l frequency o f an a to m ica l areas w hich display
Butchering, bone fra ctu rin g , and bone tools
Figure 8.1. Exam ples o f cut m arks, (a) Steller’s sea lion (Eum etopias jubatus)
hum erus w ith cu t m ark s o n proxim o-lateral surface, scale b ar is 5 cm (from
L ym an 1992c:251, Figure 1; courtesy o f the Society fo r A m erican A rchaeology);
(b) deer (Odocoileus sp.) distal m etap o d ial w ith cu t m ark s on lateral surface.
b u tch ery m ark s in analysis d em an d s th a t th o se frequencies be tallied in such a

m a n n e r as to tak e in to ac co u n t v ario u s ta p h o n o m ic facto rs w hich destroy
b u tch ery m arks. W e ath erin g (B ehrensm eyer 1978), gnaw ing by carn iv o res o r
ro d e n ts (G ray so n 1988), ro o t etching (C h a p te r 9), a n d o th e r processes w hich
alter the ex terio r surface o f bones can o b lite rate b u tch ery m ark s (M altb y
1985a, 1985b). T h us, the an aly st tallies (a) the n u m b er o f surfaces o f each
an a to m ica l area th a t have the p o ten tial to display b u tch ery m ark s, (b) the
n u m b er o f surfaces o f each a n a to m ica l a rea th a t have been m odified by non-
b u tch ery -related processes such th a t b u tch erin g m ark s th a t m ay have existed
have been o b lite rated , a n d (c) the n u m b er o f surfaces o f each a n a to m ica l area
th a t disp lay b u tch ery m arks. P ro p o rtio n a l frequencies o f b u tch ery -m ark ed
specim ens are derived by dividing the th ird value by the first; the second value is
ig n o red in b u tch ery analysis, b u t w hen sum m ed w ith the first value will provide
a to ta l co u n t fo r th e p a rtic u la r an a to m ica l a rea u n d e r co n sid eratio n (w hich
m ay be im p o rta n t to d eriv atio n o f M N E an d M N I values). T hus, fo r exam ple,
20 distal m edial tib ia frag m en ts m ay be in a sam ple, b u t three are heavily
w eath ered a n d tw o h ave been gnaw ed by ro d en ts, leaving only 15 w ith the
po ten tial to display b u tch ery m arks. If five display such m ark s, th en 33%
[(5 -t- 15) 100] is the p ro p o rtio n a l frequency o f specim ens w ith the p o ten tial to
display b u tch ery m ark s th a t in fact display them (the M N E o f distal tib ia w ould
be 20). U nless th ere is evidence to suggest otherw ise, it is p ro b a b ly reaso n ab le
to assum e, a n d in fact m u st be assum ed, th a t the d e stru c tio n o f b o n e surfaces
w hich m ig h t show b u tch ery m ark s is n o t related to the presence o r absence o f
b u tch ery m ark s on th o se d estro y ed surfaces.
A n a lyzin g butchering practices
T here h av e been several w ays th a t zo o arch aeo lo g ists have a p p ro a c h e d the

analysis o f b u tchering . O ne involves stu d y o f the frag m en ts o f bones, a topic I
reserve fo r th e next section. M o re recently, th e focus h as been on the b u tch erin g
m ark s observed o n bones. In the follow ing, I review several ap p ro ach e s to
analyzing an d in terp re tin g b u tch erin g by su m m arizing how various analysts
have stud ied b u tch erin g m arks.
M uch o f the m o st intensive stu d y o f b u tch ery m ark s has been u n d erta k en
w ith bovid fa u n al rem ains recovered fro m the P lio-P leistocene sites excavated
at O lduvai G o rg e, T an z an ia. H ere, the cataly st for such intensive study resides
in th e im p o rtan ce o f u n d ersta n d in g the early stage in h o m in id ev o lu tio n and
b eh av io r rep resen ted by the collections. S tudy o f b u tch erin g m ark s on bones
recovered from sites a t O lduvai began in earn est ju st over a decade ago (B unn
1981, 1982; P o tts an d S h ip m an 1981; S hip m an 1983), a n d co n tin u es today.
T w o researchers have been largely responsible fo r this research (o th ers are, o f
course, also involved), a n d th eir ap p ro ach e s to the stu d y o f b u tch erin g m ark s
differ som ew hat. I review each in tu rn .
Butchering, bone fracturing, and bone tools 307
S h ip m an (1981a, 1983. 1986a. 1986b. 1987, 1988b: P o tts an d S hipm an 1981)

used S E M techn iques to identify cu t m a rk s o n som e o f the bones recovered
from O lduvai Bed I sites. She th e n used ch i2 analysis to assess w h eth er those
m ark s w ere ra n d o m ly d istrib u te d across “ n e a r jo in t” a n d “ m id sh a ft” locations
on th e b ones, a n d w h eth er th o se m a rk s w ere ra n d o m ly d istrib u te d across ”
m e at-b earin g ” a n d “ n o n -m e a t-b e a rin g ” bones, relative to (a) a sam ple o f
carn iv o re gnaw ing m ark s o n 70 bones fro m O lduvai a n d (b) a sam ple o f cut
m ark s o n b o nes recovered fro m the N e o lith ic p a sto ra list site o f P ro lo n g ed
D rift (G ifford et al. 1980). She fo u n d th a t “ the O lduvai cut m ark s are
d istrib u te d significantly differently fro m the P ro lo n g ed D rift cut m ark s a n d do
n o t clu ster n e a r jo in ts [and that] the O lduvai cu t m a rk d istrib u tio n is
ind istin g u ish ab le fro m th a t o f th e 70 carn iv o re to o th m a rk s” (S hipm an
1986a:30). S hip m an (e.g., 1986b:698) concludes th a t the anim als represented
by th e bones fro m th e O lduvai sites w ere system atically b u tch ered , an d
p ro b a b ly w ere scavenged ra th e r th a n h u n ted .
S h ip m a n ’s (1986a, 1986b, 1988b) assu m p tio n s include the follow ing: (a) cut
m ark s in n ea r-jo in t lo catio n s signify d isarticu latio n , (b) c u t m ark s o n m eat-
b earin g b ones (w hich are never defined o r identified) signify defieshing o r
filleting, (c) cu t m ark s o n m etap o d ials re p resen t skinning, a n d (d) a n a p p ro p ri
ate an a lo g u e fo r a system atic b u tch erin g p a tte rn can be fo u n d in the frequency
d istrib u tio n o f cu t m ark s across n ea r-jo in t a n d m id sh aft loci o n m eat-b earin g
a n d n o n -m ea t-b earin g b ones in a sam ple o f cu t-m a rk ed bones fro m a N eo lith ic
site. These assu m p tio n s have been criticized as p o o rly fo u n d ed a n d in som e
cases refu ted by eth n o arch ae o lo g ic al d a ta (B unn a n d B lum enschine 1987;
G iffo rd -G o n zalez 1989c; L ym an 1987b). S h ip m a n ’s analytical m eth o d s are,
how ever, in trig uin g . T hey display a shift fro m th e m o re co m m on, to th a t p o in t
in tim e, p ro c ed u re o f illu stra tin g a skeleton w ith the general a n a to m ica l
lo catio n s o f cu t m ark s show n (e.g., G u ild ay et al. 1962), to use o f the frequency
d istrib u tio n o f c u t m ark s across p a rtic u la r an a to m ica l loci as an im p o rta n t
source o f in fo rm a tio n on h o m in id b ehavior. B unn a n d K ro ll’s (1986) analysis
is im p o rta n t because it illu strates the next step, bey o n d S h ip m a n ’s w ork,
to w ard s increasing the in tensity a n d ex ten t to w hich d a ta on b u tch ery m ark
p lacem en t a n d frequency d istrib u tio n s are studied.
B unn a n d K ro ll (1986:436; see also B unn 1981, 1982, 1983) begin th eir study
o f th e bo v id rem ains fro m the Z injanthropus site at O lduvai w ith the statem en t
“ th e lo catio n a n d frequency o f cu t m ark s o n different skeletal p a rts can be used
in c o n ju n ctio n w ith a know ledge o f an im al a n a to m y to identify p a tte rn in g in
th e b u tch erin g tech niqu es o f p re sen t a n d p a st h u m a n s .” O n th a t sam e page,
they go on to suggest th a t “ cu t m ark s on n o n m eaty skin-covered bone surfaces,
on o r n e a r epiphyses w here connective tissues bind artic u la tin g jo in ts, an d on
m eaty bones at p o in ts o f m uscle a tta c h m e n t ca n provide u n am b ig u o u s
d o c u m e n ta tio n o f carcass skinning, jo in t d isarticu latio n , a n d defieshing,
respectively.” T h ey p re sen t d a ta o n the frequency d istrib u tio n o f c u t m ark s
T ab le 8.4 N I S P a nd frequ encies o f cu t-m a rked specim ens in the F L K

Z in ja n th ro p u s assem blage (m o d ifie d fro m Bunn and K roll 1986)
Small bovid Large bovid

A natom ical
category N IS P N cut % cut N IS P N cut % cut
m axilla and cranium 49 0 0.0 46 2 4.3

m andible 20 2 10.0 160 11 6.9
vertebral centrum 29 1 3.4 57 1 1.8
vertebral process 28 0 0.0 59 3 5.1
rib 217 1 0.5 423 21 5.0
sternabra 1 0 0.0 1 0 0.0
scapula 7 0 0.0 22 2 9.1
P hum erus 3 1 33.3 4 0 0.0
S hum erus 13 3 23.1 45 8 17.8
D hum erus 6 4 66.7 17 6 35.3
P radius 4 2 50.0 15 3 20.0
S radius 12 3 25.0 45 10 22.2
D radius 3 0 0.0 6 1 16.7
P ulna 6 0 0.0 10 2 20.0
S ulna 4 0 0.0 11 1 9.1
D ulna 0 0 0.0 4 0 0.0
carpals 13 0 0.0 17 0 0.0
P m etacarpal 6 2 33.3 10 2 20.0
S m etacarpal 12 1 8.3 20 1 5.0
D m etacarp al 3 0 0.0 2 0 0.0
innom inate 14 2 14.3 26 7 26.9
P fem ur 1 0 0.0 6 1 16.7
S fem ur 17 2 11.8 41 7 17.1
D fem ur 2 1 50.0 5 0 0.0
patella 1 0 0.0 1 0 0.0
P tibia 9 2 22.2 1 0 0.0
S tibia 36 11 30.6 92 7 7.6
D tibia 3 2 66.7 7 2 28.6
tarsals and D fibula 24 2 8.3 28 1 3.4
P m etatarsal 12 2 16.7 6 0 0.0
S m etatarsal 15 1 6.7 13 2 15.4
D m etatarsal 2 0 0.0 4 1 25.0
phalange 27 0 0.0 13 0 0.0
sesam oid 16 0 0.0 21 0 0.0
S long bone 208 5 2.4 667 17 2.5
S m etapodial 17 0 0.0 42 3 7.1
totals: 840 50 6.0 1947 122 6.3
across v ario u s skeletal p a rts o f sm all bovids (10-110 kg live w eight) an d large
bovids ( > 110 kg live w eight), sum m arized in slightly m odified fo rm in T able
8.4.
B un n a n d K ro ll (1986) p re sen t a n u m b e r o f inferences o n the basis o f th eir
d a ta (T ab le 8.4). Som e o f these can be sum m arized as follow s: (1) A b u n d a n t cut
m ark s o n “ m eaty ” lim b bones (n o t defined by B unn a n d K roll) indicate large
q u an titie s o f m eat w ere cut fro m those bones. (2) P ro p o rtio n a te ly few er m eaty
lim b b ones o f large m am m als have cut m ark s th a n m eaty lim b b ones o f sm all
m am m als, b u t few er n o n -m eaty n o n lim b bones o f large m am m als have cut
m ark s th a n n o n -m ea ty n o n lim b b ones o f sm all m am m als; B unn a n d K roll
believe this is the case because th ere is m o re m eat o n the n o n -m eaty , non lim b
bones in large m am m als th a n in sm all m am m als. (3) C u t m ark s on m etap o d ials
rep resent “ skin n ing o p e ra tio n s.” (4) C u t m ark s in m id -sh aft lo catio n s on
m eaty lim b b ones rep resen t defleshing. B unn an d K ro ll’s (1986:436) assu m p
tio n a b o u t the lo catio n o f a cu t m a rk being indicative o f th a t m a rk ’s fu n ctio n
(skinning, d isarticu latio n , defleshing) m ay be co rrect in a general sense, given,
for exam ple, the eth n o arch ae o lo g ic al d o c u m e n ta tio n pro v id ed by B inford
(1981b) on the c o v a ria tio n o f m ark lo catio n an d m ark function. H ow ever,
L ym an (1987a:263-265, 1987b:711) notes th a t while B inford (1981b) provided
a fu n ctio n al typo lo gy for cu t m ark s based on 108 a n a to m ica l lo catio n s and
orientations o f such m ark s, the general n ea r-jo in t categ o ry o f m ark s, for
exam ple, includes n o t ju s t d isarticu latio n m ark s b u t defleshing a n d skinning
m ark s as well (F ig u re 8.2). P a rt o f the reaso n fo r such v a ria tio n m ay reside in
w h eth er th e carcass being b u tch ered is fresh a n d supple, frozen, o r som ew hat
desiccated a n d stiff, such as a several-day-old carcass th a t is scavenged (B inford
1984b: 110-112).
T he frequency d is trib u tio n o f p ro p o rtio n s o f cu t-m a rk ed specim ens across
different a n a to m ica l loci suggests to B unn a n d K roll (1986), a m o n g o th er
things, a th o ro u g h a n d system atic b u tch erin g process w as applied to b o th size
classes o f bovids (B u n n 1983), a n d m u ch m eat w as rem oved from the bones by
the butch ers. B inford (1988:127) c o u n ters the la tte r by n o tin g th a t " th e n u m b er
o f c u t m ark s, exclusive o f d ism em b erm en t m ark s, is a fu n c tio n o f differential
in v estm en t in m eat o r tissue rem oval. W hen a b u tch er w ho is filleting m eat
seeks to get all th e ad h e rin g tissue off the bones, there will be m any cut m arks; if
little effort is m ad e to clean th e bones, relatively few c u t m ark s re su lt.” B unn
an d K ro ll (1988:144) reply by suggesting th a t “ a sim ple c o m p a riso n o f cut-
m ark frequencies a m o n g skeletal elem ents a n d an im al size g ro u p s is ill-advised
as the sole basis fo r establishing w h eth er o r n o t a su b stan tial p o rtio n o f the
m eat w as p resen t w hen the carcass w as b u tc h e re d ,” a n d indicate the freq u en
cies o f cu t m ark s they reco rd ed on the b ones fro m O lduvai are sim ilar to the
frequencies d o cu m en te d at N eo lith ic sites w here, it is presum ed, cu t m ark s
w ere p ro d u c ed o n b ones w ith m u ch m e a t ad h e rin g to them (e.g., G iffo rd et al.
1980; M arsh all 1986).
MEDIAL VIEW
DISTAL VIEW
Figure 8.2. D istal m etap o d ials show ing locations o f variously docum ented cut-
m arks (from L ym an 1987a:264, Figure 5.1; courtesy o f A cadem ic Press). M 1 -M 4
from ethnoarchaeological contexts (B inford 1981b); A -G from prehistoric
contexts.
I p resen t th e preceding discussion w ith o u t the in ten t o f resolving the d eb ate

a b o u t the significance o f the cu t-m a rk e d bones from O lduvai, a n d w ith o u t the
in ten t o f tak in g sides. W h a t I h o p e to h ave d o n e in these few p a ra g ra p h s is offer
th e re a d e r so m eth in g o f the flavor o f the w ay an aly sts have ap p ro a c h e d
b u tch ery -m ark d a ta . T here is, o f course, m u ch I have n o t said th a t can be found
by re ad in g th e references cited. B ut there are also several o th e r item s here th a t
can be to u ch ed on, item s n o t to be fo u n d in the references cited. In p artic u la r,
there are several w ays to stu d y the b u tch ery d a ta in T ab le 8.4 n o t explored by
the a u th o rs w ho have previously discussed them . I review several o f them here.
F irst, th e p ro p o rtio n o f cu t-m a rk e d bones o f sm all bovids at O lduvai is n o t
significantly different from the p ro p o rtio n o f cu t-m a rk e d bones o f large bovids
(arcsine /s = 0.3, P > 0.5; see S okal an d R o h lf [1969:607-608] fo r discussion o f
the arcsine tra n sfo rm a tio n statistic). T his suggests th a t overall, the carcasses o f
sm all bovids w ere b u tch ered ju s t as intensively as carcasses o f large bovids.
Second, the p ro p o rtio n o f cu t-m a rk e d specim ens p er a n a to m ica l categ o ry for
sm all bovids, as listed in T ab le 8.4, is n o t co rrelated w ith the to ta l N IS P p er
an a to m ica l catego ry fo r sm all bovids (r = 0.19, P = 0.26); the sam e is tru e for
B utchering, bone fra ctu rin g , and bone tools 311
Small Bovid
60 HI Large Bovid
"O
Cl)
■£ 40
ro
O
20
</) </) </) 03 03 k_ k_ 03 03
~o
03
k_
_:d
5
“O
13
03
i_
_c
=3
CL
_c
13
CO
3
E
£
13
E
©
H—
.D
QL CO
Q_ CO Q CL
F igure 8.3. P ro p o rtio n al frequencies o f cut-m arked specim ens in selected

anatom ical categories (from T able 8.4).
large b o vids (r = 0.15, P = 0.40). T h u s the p ro p o rtio n s o f cu t-m a rk e d bones in

th e tw o an im al size categ ories do n o t seem to be a fu n c tio n o f sam ple size
m easu red as N IS P (n ote th a t the N IS P p er a n a to m ic a l categ o ry fo r sm all
bovids is strong ly co rrelated w ith the N IS P p er a n a to m ica l categ o ry fo r large
bovids; r = 0.95, P < 0.001). T h ird , the p ro p o rtio n o f cu t-m a rk ed specim ens p er
a n a to m ica l categ o ry fo r sm all b ovids is significantly b u t w eakly co rre la te d w ith
the p ro p o rtio n o f cu t-m a rk ed specim ens p er a n a to m ic a l categ o ry fo r large
bovids (r = 0.48, P = 0.003). T his suggests th a t there are, p erh ap s, differences
betw een how large a n d sm all bovids w ere b u tch ered , g ra n tin g , o f course, the
a ssu m p tio n to any such analysis n o ted earlier th a t th e frequency o f butchery-
m ark e d bones is directly co rrelated w ith th e frequency o f b u tch ered bones.
T h a t th ere m ay be differences in how these tw o size gro u p s o f bovids were
b u tch ered is suggested by F ig u re 8.3. T his brings us to B unn an d K ro ll’s (1986)
con clu sio n n o ted abo v e th a t p ro p o rtio n a te ly few er m eaty lim b bones o f large
bovids th a n sm all bovids h av e cu t m ark s, b u t few er n o n -m ea ty n o n lim b bones
o f large b ovids h av e cu t m ark s th a n n o n -m ea ty n o n lim b bones o f sm all bovids.
H o w m ig h t this be stu died analytically?
I p resu m e m eaty lim b b ones include th e h u m eru s, ra d iu s-u ln a , fem ur, a n d
tibia; th e m etap o d ials are n o n -m ea ty lim b bones. T h ere is no statistically
significant difference in the p ro p o rtio n a l frequencies o f b u tch ery -m ark e d
m etap o d ials fo r th e tw o size categories o f bovids (T able 8.5). T here are,
T ab le 8.5 Frequencies o f cu t-m a rked m e a ty limb specim ens and m etapodial

specim ens in the F L K Z in ja n th ro p u s collection (fro m Table 8.4)
Small bovid Large bovid Arcsine

A natom ical
category N IS P N cut % cut N IS P N cut % cut ts P
M eaty limbs 119 31 26.0 309 48 15.6 2.398 < 0 .0 2

M eaty lim bs + shafts 327 36 11.0 976 65 6.7 2.383 < 0 .0 2
M etapodials 65 6 9.2 97 9 9.3 0.022 > 0 .5
how ever, significant differences in the p ro p o rtio n s o f b u tch ery -m ark e d m eaty
lim bs, w h eth er o r n o t B unn a n d K ro ll’s unidentifiable long-bone shaft
frag m en ts are included. P ro p o rtio n a te ly m ore m eaty lim b specim ens o f sm all
bovids display cut m ark s th a n m eaty lim b specim ens o f large bovids. T o help
d eterm in e why such a difference exists, I so rted the m eaty lim b specim ens into
categories th a t c o n stitu te the m ajo r lim b jo in ts: the sh o u ld er includes all
scapula specim ens; the elbow includes the distal h um erus, proxim al rad iu s, and
p ro x im al ulna; th e w rist includes the distal rad iu s, d istal ulna, carp als, and
pro x im al m etacarp al; the hip includes the p roxim al fem ur; the knee includes
the distal fem ur, p atella, a n d p ro x im al tibia; the ankle includes the distal tibia,
tarsals a n d distal fibula, a n d p roxim al m e ta ta rsa l (a fter L ym an 1991a, 1992c,
1993a). R elevant d a ta are sum m arized in T ab le 8.6. A rcsine statistics indicate
the only significant differences are in the hind leg; the knee jo in t was a p p a re n tly
m o re freq u ently bu tch ered in sm all bovids th a n in large bovids, an d the tibia
sh aft o f sm all bovids w as m ore frequently bu tch ered th a n the tibia shaft in large
bovids.
T h e perceptive re ad er will note tw o things a b o u t the preceding. F irst, I have
n o t presu m ed o r inferred an y th in g a b o u t the fun ctio n o f p a rtic u la r categories
o f m ark s, such as n ea r-jo in t m ark s rep resen tin g d isarticu latio n . Second,
w ith o u t m ore precise d a ta on the lo catio n o f the m ark s on scap u la and
in n o m in a te specim ens, the statistical analyses in T able 8.6 m ay be faulty. T h at
is so because in th a t tab le I included all scapula specim ens even th o u g h som e o f
the b u tch erin g m ark s m ay have been on the b lade o f the scap u la in n o n -n e a r
jo in t loci, an d I excluded the in n o m in a te even th o u g h som e o f the cut m ark s
m ay h av e been n e a r th e acetab u lu m in n ea r-jo in t loci. N onetheless, m y p u rp o se
in p resen tin g this analysis has n o t been to d eterm in e precisely w h a t b u tch erin g
activities w ere p erfo rm ed by P lio-P leistocene hom inids at O lduvai. R ath er, I
have sim ply illu strated how one m ight p erfo rm such an analysis w ere th a t o n e’s
goal. T h ere are three o th er an aly tical avenues th a t can be described using the
O lduvai d a ta in T ab le 8.4.
F irst, are the p ro p o rtio n a l frequencies o f cu t-m a rk ed jo in ts a fun ctio n o f
jo in t tightness? O ne m ig ht expect the skeletal p a rts c o n stitu tin g a jo in t th a t is
tightly b o u n d by ligam ents, ten d o n s, a n d m uscles to display m o re cu t-m a rk s
T ab le 8.6 Frequencies o f cut-m arked specim ens in jo in t, and m ea ty limb shaft

locations ( com piled fr o m Table 8.4)
Small bovid L arge bovid Arcsine

A natom ical
category N IS P N cut % cut N IS P N cut % cut h P
shoulder 3 1 33.3 4 0 0.0 1.509 > 0 .2

elbow 16 6 37.5 42 11 26.2 0.828 > 0 .4
wrist 22 2 9.10 37 3 8.1 0.132 > 0 .5
hip 1 0 0.00 6 1 16.7 0.779 > 0 .4
knee 12 3 25.0 7 0 0.0 2.203 < 0 .0 5
ankle 39 6 15.4 41 3 7.3 1.160 > 0 .2
S hum erus 13 3 23.1 45 8 17.8 0.419 > 0 .5
S radius-ulna 16 3 18.8 56 11 19.6 0.071 > 0 .5
S fem ur 17 2 11.8 41 7 17.1 0.525 > 0 .5
S tibia 36 11 30.6 92 7 7.6 3.119 >0.01
th a n th e b ones o f a less tig h tly b o u n d jo in t. I calcu lated th e average ra n k o f

d isarticu latio n fo r th e top i, w ildebeest, a n d G ra n t's gazelle to pro v id e a general
m easure o f lim b jo in t tightness from T ab le 5.5. T h a t o rd e r is, from the first
(least tig ht) to th e last jo in t (tightest) to n atu ra lly d isarticu late: sh o u ld er, w rist,
hip, elbow , a n d the knee a n d ankle are tied for last. T h a t o rd e r o f n a tu ra l
d isarticu latio n does n o t co rrelate w ith eith er the p ro p o rtio n o f cu t-m a rk ed
bones per jo in t o f sm all bovids (rs = 0.06. P = 0.91) o r w ith the p ro p o rtio n o f
cu t-m a rk ed b ones p er jo in t o f large bovids (rs= —0.01, P = 0.98), as listed in
T able 8.6. If a significant c o rrelatio n h ad been fo u n d , one m ight arg u e th a t (a)
m any o f the m ark s are d isarticu latio n m ark s, an d (b) m ark frequencies are
related to th e la b o r inv estm ent re q u ired to d isarticu late jo in ts.
Second, the p ro p o rtio n s o f individual cu t-m a rk ed lim b bones o f b o th large
an d sm all bovids d o n o t co rrelate w ith the food utility index (T able 7.3). T h a t is
so using tw o different d a ta sets. In one, I calculated the p ro p o rtio n s o f cut-
m ark e d specim ens fo r ends a n d shafts o f the scapula, hum erus, rad iu s-u ln a,
m etac arp a l, in n o m in a te , fem ur, tibia, an d m etatarsal. In the o th er, I calculated
the p ro p o rtio n s o f cu t-m a rk ed specim ens fo r the scap u la a n d in n o m in a te , an d
the sh afts only o f th e h u m eru s, ra d iu s-u ln a , m etac arp a l, fem ur, tibia, an d
m etatarsal. O ne m ight be p ro m p te d , given such a result, to w o n d er if the cut
m ark s are n o t related to m eat e x tra ctio n based on the ex p e ctatio n th a t bones
w ith m o re m eat on them w ould p ro m p t m o re effort to rem ove th a t m eat,
resu ltin g in m o re cu t-m a rk e d specim ens. T his ex p e ctatio n is, in fact, one thing
th a t B inford (1988) a n d B unn a n d K ro ll (1986, 1988) agree on in at least a
general sense: m ore m ark s re p resen t m ore investm ent to rem ove soft tissue, but
w h eth er th a t m ean s th ere is m u ch o r little soft tissue p resen t is unclear.
F in ally , follow ing th e re aso n in g th a t the fo o d utility o f a skeletal p a rt m ight
influence the frequency o f specim ens o f th a t p a rt w hich are b u tch ered an d th u s
the frequency w ith w hich th a t p a rt will display b u tch erin g m ark s, the frequency
o f flake-scarred b ones in a n assem blage should, p erh ap s, co rrelate w ith the
m arro w u tility index, the grease u tility index (e.g., T ab le 7.1), o r som e
c o m b in a tio n th ereo f. T he frequency o f flake-scarred specim ens per an a to m ica l
categ o ry in the O ld uvai sam ples has n o t been re p o rted , preclu d in g the
an aly tical p u rsu it o f this line o f inquiry.
S u m m a ry
T he av ailab le lite ra tu re on b u tch erin g is extensive. L ym an (1987a:251)

re p o rted th a t over tw o d ozen b o o k s a n d articles on this to p ic alone a p p e are d in
eight m o n th s d u rin g 1984 a n d 1985. A su b stan tial n u m b e r o f articles ap p eared
in 1986. T he stu dy o f b u tch ery has co n tin u ed to be an im p o rta n t to p ic in the
lite ra tu re (e.g.. B artra m et al. 1991; Y ellen 1991a, 1991b). A nalysis o f w hat
a p p e a r to be b u tch erin g m ark s has recently been used to identify ritualistic
tre a tm e n t o f b o th an im al rem ains (N o e -N y g a ard an d R ichter 1990) and
h u m an rem ains (T o rb en so n et al. 1992), a n d h u m an cannibalism (W hite 1992).
T he analysis a n d in te rp re ta tio n o f b u tch ery -m ark e d bones is a research topic
th a t is perceived by m an y zo o arch aeo lo g ists as im p o rta n t. A n d , given the
definition o f ta p h o n o m y subscribed to in this volum e a n d the definition o f
b u tch ery used in this c h a p te r, b u tch erin g should be considered a n im p o rta n t
research top ic by ta p h o n o m ists if there is evidence th a t hom inids w ere involved
in th e ta p h o n o m ic h isto ry o f the bo n e assem blage u n d e r study.
In this section I have o u tlin ed an d exem plified analy tical pro ced u res and
assu m p tio n s th a t are necessary to the stu d y o f butchering. T he co n cep tu al
fram ew o rk fo r such stu dy u n d ersco res th e im p o rtan ce of, in p a rtic u la r, an
a ssu m p tio n th a t is critical to the analysis o f frequencies o f b u tch ery -m ark ed
bones: the frequencies o f b u tch ery -m ark e d bones m ust be directly co rrelated
w ith th e frequencies o f b u tch ered bones, som e o f w hich m ay n o t be butchery-
m ark ed . I f b u tch ery m ark s are to som e significant degree fo rtu ito u s epipheno-
m ena, w hich they m ay well be in som e cases (e.g., L ym an 1993a), then there is
little p o in t in d o in g m o re th a n sim ply reco rd in g th a t som e bones in an
assem blage are b u tch ery m ark ed . H ere, 1 th in k , resides an area d em an d in g
m an y m o re eth n o arch ae o lo g ic al d a ta th a n are presently available. W e need to
k now n o t o nly how h o m in id b u tch erin g reduces an im al carcasses in to co n su m
able p a rts, b u t how th o se b eh av io rs are reflected archaeologically. W e need
details o f h o w th a t reflection ap p e a rs in term s o f frequency d istrib u tio n s o f
m ark s across p a rtic u la r a n a to m ica l loci, a n d how a n d w hy v a ria tio n betw een
such frequencies can be created. W e need to know m o re a b o u t how to co u n t
b u tch ery m ark s (sh ould they be tallied by skeletal elem ent, by p ro x im al end,
distal end, a n d shaft). W ith o u t these a n d o th e r kinds o f in fo rm atio n , o u r
analyses o f b u tch erin g m ark s will rem ain m u ch like th o se described above:
inductive p a tte rn -re c o g n itio n studies th a t identify p a tte rn s an d v aria tio n s in
b u tch ery m a rk d a ta , the h om in id b eh av io ral significance o f w hich is deb atab le.
In this first section, I have focused on b u tch ery m ark s. B ut b u tch erin g often
also results in b ro k e n b ones (e.g., B inford 1978; L ym an 1978; N o e -N y g aa rd
1977, 1987; Y ellen 1977, 1991a, 1991b). A n d while this section also tends to
focus o n b u tch erin g to e x tra ct fo o d resources, it sh o u ld be clear th a t o th er
kinds o f resources are also typically ex tracted from an im al carcasses (T able
8.1). O ne o f these resources is bo n e used to m ak e tools. M ak in g b o n e tools can
result in b ro k e n bones. T h u s bo n e fra g m e n ta tio n seem s to be a reaso n ab le
bridge betw een discussion o f b u tch erin g to e x tra ct fo o d a n d discussion o f bone
tools. T h us, it is to the im p o rta n t topic o f fractu re d bones th a t we now tu rn .
F ractu rin g of bone

It can n o t be assum ed th a t all split an d fractu red bone on an archaeological site has
been b roken by man.
(J. D. C lark 1972:149)
P erh aps it w as R ay m o n d D a rt’s (e.g., 1957) claim th a t P lio-Pleistocene

hom in id s h ad b ro k e n som e o f the bones recovered from the M a k a p a n sg a t
lim ew orks in S o u th A frica in distinctive w ays th a t served as the ca ta ly st for
intensive an d extensive investigation o f how an d w hy bones are b ro k e n the way
they are. C ertain ly D a rt’s claim s w ere then, a n d are now , co n tro v e rsial (e.g..
Hill 1976; M ag uire et al. 1980; R ead 1971; R e a d -M a rtin an d R ead 1975;
S h ip m an an d Phillips 1976; S hip m an a n d P hillip s-C o n ro y 1977;W olberg 1970;
see B rain 1989 fo r a recent overview ), as are sim ilar claim s fo r som e b ro k e n
bones fo u n d in N o rth A m erica (e.g., B onnichsen 1973, 1979; H ay n es 1983b;
Irving a n d H a rin g to n 1973; Irving et al. 1989; Jo h n so n 1983, 1985; L ym an
1984a; M iller 1969, 1975; M o rla n 1980, 1983, 1984, 1988; S a d e k -K o o ro s 1972,
1975). T he co n tro v ersy has led to a great deal o f actu alistic research on bone
frag m en tatio n . T o u n d e rsta n d fully the significance o f b o n e fractu re for
ta p h o n o m ic studies, I begin w ith a discussion o f the fractu re m echanics o f
b ones before describing types o f fractu res a n d identifying agents o f bone
fractu re.
M echanics o f bone fra ctu re

F o rce m u st be ap p lied to a bo n e to b re a k it. Strain is the ch ange in linear
dim en sio n s o f a b o d y resu lting fro m the a p p lic a tio n o f force. Stress is the ratio
o f the a m o u n t o f force ap plied to the a rea over w hich the force is acting, a n d is
o ften used synon y m ou sly w ith strength. E lasticity is the p ro p e rty th a t allow s a
b o d y to re tu rn to its original sh ap e a n d size afte r an ap p lied force is rem oved.
T h e m odulus o f elasticity is the ra tio betw een u n it stress a n d u n it strain , a n d it
m easures the stiffness o f a m aterial, n o t its elasticity. P lo ttin g the stress
p ro d u c ed in a m ateria l against the strain , an d increasing the stress, pro d u ces a
curve in d icatin g the stiffness o f the m aterial. T he resulting curve is used to
d eterm in e the b re ak in g p o in t a n d th e a m o u n t o f energy a b so rb ed by the
m aterial p rio r to its fractu re (C u rrey 1984; D avis 1985:55-56, 62-63; Jo h n so n

1985).
A tensile force is one w hich tends to pull a body ap art, w hereas a compressive force is
one w hich tends to p ush a b o dy together. A shearing force is one w hich causes one
p a rt o f a body to slide in a direction opp o site to th a t o f an im m ediately adjacent
part. Torsion o r tw isting com bines tensile and shearing forces, while bending
involves a c o m b in atio n o f tensile an d com pressive forces.
(Evans 1961:110-111)
S ta tic loading involves the ap p lica tio n o f c o n sta n t com pressive pressure,
generally w ith an even d istrib u tio n o f force (Jo h n so n 1985:192). D ynam ic
loading involves focused su dden im p act (Jo h n so n 1985:170, 192). W h en either
k in d o f lo ad in g exceeds a b o n e ’s tensile stren g th the bo n e fractures. T he m o re
ra p id th e lo ad in g rate, the less the m ax im u m strain a b o n e can w ith sta n d an d
the less energy it can a b so rb before fractu re (D avis 1985:63). F ra c tu re begins in
the o u te r layer o f the b o n e a n d progresses in w ard . Stress w aves created by the
fractu re fro n t are v ariou sly ab so rb ed , reflected a n d diffused by tra b e c u la r bone
an d th u s fractu re fro n ts tend n o t to pass th ro u g h the epiphyseal ends o f long
bones; stress w aves d efo rm th e dense co rtical b o n e o f long b o n e diaphyses an d
m ay fractu re them . D yn am ic lo ad in g im p arts bending forces to a long bone,
a n d “ sh earin g is alo n g a helical course th a t is inclined a t a 45° angle to the
lo n g itu d in al axis o f the long b o n e ’’ (Jo h n so n 1985:171) to p ro d u ce a spiral o r
helical fractu re , w hich is a tensile failure. “ D ry an d m ineralized bo n e exhibits
h o riz o n ta l ten sio n failure in w hich the fractu re fro n t cuts across the diaphysis
an d p ro d u ces p erp en d icu lar, parallel, o r d iag o n al b re a k s” (Jo h n so n 1985:172).
T he co m p o site n a tu re o f b o n e tissue - being p a rt m in eral a n d p a r t org an ic -
results in it being m echanically stro n g , stro n g e r th a n eith er m ateria l alone
(C h a p te r 4). A b o n e ’s m ic ro stru c tu re governs bone fractu re an d the resu ltan t
k ind o f fractu re, w here fra c tu re is “ a localized m echanical fa ilu re” (Jo h n so n
1985:160). C u rre y (1984:49) states th a t “ b o n e deform s ra th e r little before
fractu rin g . . . it does n o t fra c tu re w ith a sm o o th surface, as a really b rittle
m aterial does, a n d so c a n n o t be tru ly b rittle .” Jo h n so n (1985:160) elab o rates
th a t fresh, green b o n e co n tain s m o istu re a n d m arro w in the m ed u llary cavity,
a n d will “ n o t beh ave in a b rittle (o r inflexible) m an n er, b u t ra th e r, is a
viscoelastic (i.e., flow able a n d d efo rm ab le), ductile m aterial cap ab le o f w ith
stan d in g g re at a m o u n ts o f p ressu re a n d d e fo rm a tio n before failure [fracture].
B one does behave in a b rittle m a n n e r w hen it is well d rie d .” A b o n e ’s stiffness,
tensile stren g th , com pressive stren g th , a n d h ard n ess are all increased by drying
(D avis 1985:66-67). D ry in g causes a b o n e to shrink, th ere b y increasing its bulk
density a n d reducing its p o ro sity . T hus d ry bone behaves m ore like an
in o rg anic m ateria l th a n w et bone, a n d “ a lth o u g h dry b o n e m ight be stro n g er in
static lo ad in g, it w o uld be m o re likely to fail [fracture] at sm aller forces d u rin g
dy n am ic lo ad in g ” (D avis 1985:67).
J o h n so n (1985:167) no tes th a t osteo n s tend to decrease tensile stren g th
(m ore o steo ns m ean m o re cem ent, w hich is m echanically w eak) w hereas

lam ellae te n d to increase tensile stren g th (see also D avis 1985:65). F ra c tu re
begins w ith w h a t are called m icrocracks, w hich ten d to follow cem ent lines
a ro u n d o steo n s (see C h a p te r 4). “ A s failure [fracture] occurs, the p a th s th a t the
p ro p a g a tin g fra c tu re fro n ts tak e are th e resu lt o f a d ynam ic in tera ctio n
betw een th e lo ad in g device, fractu re dynam ics, a n d b o n e stru c tu re ” (Jo h n so n
1985:170). H a v ersian an d V o lk m a n ’s canals reduce a b o n e ’s stiffness, a n d as
the p ro p o rtio n o f in terstitial lam ellae increases, the stress a bo n e can w ith stan d
w ith o u t fractu rin g is reduced (D avis 1985:65-66).
T his co n sid eratio n o f the biom echanics o f fractu re does n o t do ju stic e to the
extensive lite ra tu re o n the topic. B ut I h av e been b rie f because, as D avis
(1985:70) notes, m u ch o f th a t lite ra tu re co n cern s sam ples o f bone tissue ra th e r
th a n b on es o r skeletal elem ents. T a p h o n o m ic agents fra c tu re com plete skeletal
elem ents, a n d the lite ra tu re o n the b iom echanics o f w hole b ones is ra th e r
lim ited (see D av is 1985 an d M o rla n 1980 fo r in tro d u c tio n s to th a t literatu re).
D a v is’ (1985) m o st in terestin g conclusions re g ard in g the fractu re o f co m
plete skeletal elem ents can be sum m arized as follow s. (1) B one size as so rted by
the live w eight o f th e anim al co n trib u tin g the bone does n o t influence bone
fra c tu re p a tte rn s. (2) M a c ro stru c tu ra l differences betw een skeletal elem ents
ac co u n t fo r fractu re location. “ D ifferences in cross-sectional thickness o f the
co rtical b o n e betw een elem ents a n d the presence o f crests such as th e an terio rly
p o sitio n e d tibial crest are the facto rs m o st likely to d eterm in e th e lo catio n o f a
fra c tu re ” (D avis 1985:94). (3) M ic ro stru c tu ra l differences in skeletal elem ents
exert som e influence o n fra c tu re form . “ F o r exam ple, the collagen alig n m en t in
h um eri is m ore likely to p ro d u ce fractu re s w ith curved edges” (D avis 1985:97).
M id -sh aft cross-sectio nal shape a n d d iam eter relative to cortical bo n e th ic k
ness also seem to influence fra c tu re form . (4) L ightly w eathered bo n e will
fra c tu re obliquely, b u t th e p ro p o rtio n o f such fractu res decreases as th e extent
o f w eath erin g increases. T here is a tendency fo r m o re intensively w eathered
bones to req uire less force fo r fractu rin g . (5) “ In static load in g the b o n e fails in
ten sio n usually a t a p o in t o p p o site to the p o in t o f loading. In d ynam ic loading,
several fractu re fro n ts fo rm a n d ra d ia te aw ay fro m the p o in t o f im p a c t” (D avis
1985:108). (6) “ A high p ercentage o f o blique fractu res is expected in any
assem blage d ue to p ro p e rtie s in h e re n t in th e b o n es” (D avis 1985:129). (7)
S tatistically significant g re ater th a n expected frequencies o f o blique fractu res
(spiral fractu res, see below ) o cc u r u n d e r co n d itio n s o f static a n d to rsio n al
lo ad in g , w hereas d y n am ic lo ad in g p ro d u c es few er oblique fractu re s th a n
expected. T his is p ro b a b ly because u n d e r low strain rates such as static a n d
to rsio n a l lo adin g , fractu re is betw een lam ellae a n d obliquely o rien te d collagen
fibers w hereas th e h igh strain ra te s in d u ced by d ynam ic lo ad in g c reate fractu re
fro n ts th a t m ay cross lam ellae (D avis 1985:133). (8) D ynam ically lo ad ed bones
are m o re likely to display ro u n d e d fractu re ends th a n statically lo ad ed a n d
fractu re d bones.
D a v is’ (1985) research results reflect the variables an aly sts stu d y w hen
exam ining b ro k e n bones. T a p h o n o m ists focus on the m o rp h o lo g ical attrib u te s
o f fractu res in a tte m p ts to identify the ag en t o f b o n e fractu re, the kin d o f force
th a t resulted in a fractu re, a n d the co n d itio n o f a bone w hen it w as fractu red .
W e tu rn , then, first to types o f fractu res, a n d subsequently to identifying
fractu re agents.
Types o f fra c tu re
Jo h n so n (1985:172) uses the follow ing term s a n d definitions to discuss a ttri
b utes o f b ro k e n bones: fra ctu re location - the area w here failure occurred;
fra ctu re fr o n t - th e leading edge o f force a n d its directio n (can be determ ined
from featu res o n the fractu re surface); fra c tu re surface - created by failure, a
cross-section o f co m p ac t b o n e exposed by the passage o f force th ro u g h it;
fra ctu re shape - th e o utline co n fig u ratio n o f exposed co m p ac t b o n e th a t
reco rd s th e p ro p a g a tio n p a th tak en by the fractu re fro n t in planview . 1 utilize
this term in olo gy in th e follow ing discussion.
S h ip m an et al. (1981) describe a n d illu strate seven types o f fractu re. T heir
schem e w as m odified by M arsh a ll (1989:14) to include an eighth type (F igure
8.4). S h ip m an et al. (1981:260) p o in t o u t th a t “ only rarely will a p a rtic u la r type
o f b re ak identify the ag en t o f b re ak ag e u n am b ig u o u sly .” T hey n o te th a t the
lo catio n a n d frequency o f different types o f break s on different types o f bones
can be c o m p ared w ithin an d betw een assem blages. T hey suggest th a t c o m p a ri
son o f such d a ta fo r assem blages w ith k n o w n agents o f bo n e fractu re w ith
p reh isto ric assem blages m ay reveal the ag en t o f b reak ag e in the latter. M ost
im p o rtan tly , they argu e, co rrectly I believe, th a t the an aly st should “ co m p are
like w ith like” in such analyses (S h ip m an et al. 1981:259); th a t is, m am m al
b ones o f like-sized an im als should be co m p ared ra th e r th a n , say, the b ro k e n
b ones o f a ra b b it w ith the b ro k e n b ones o f a cow o r bird. T his m akes sense
b ecause different b ones will b reak differently sim ply because o f m icro stru c tu ral
differences.
T he types o f fractu re in F igure 8.4 are generalized. A s well, the necessary and
sufficient co n d itio n s fo r identifying a p a rtic u la r specim en as displaying one
k ind o f fractu re o r a n o th e r are n o t specified by S hip m an et al. (1981) o r by
M arsh all (1989). G iffo rd -G o n zalez ( 1989a: 188) defines several o f the fractu res
types: perpendicular o r transverse fractu re s are at a rig h t angle to the long axis
o f the bone; longitudinal fractu res a re p arallel to the long axis o f the bone; spiral
fractu res are curved in a helical, p artia lly helical, o r com pletely helical p a tte rn
a ro u n d th e circum ference o f the shaft. S hipm an (1981 a :3 7 1—372) a n d Jo h n so n
(1985:175), how ever, n o te th a t the type “ spiral fra c tu re ” consists o f at least tw o
subtypes, one hav ing a sm o o th fractu re surface (fractu re fro n t a p p a re n tly w ent
betw een collagen bu nd les) a n d the o th e r having a ro u g h fractu re surface
(fractu re fro n t ap p a re n tly w ent th ro u g h o r w as p erp en d icu lar to collagen
Stepped o r C o lu m n a r Sawtoothed
Irregular Sm ooth
P e r p e n d ic u la r P e r p e n d ic u la r
Figure 8.4. F ractu re types (after M arsh all 1989:14, F igure 1; courtesy o f the
a u th o r an d The C enter for the Study o f the F irst A m ericans). All are schem atic,
w ith fractures occurring on the distal end o f an artio d acty l fem ur show n in
posterio r view.
bundles). T he fo rm er is called a T ype I spiral fractu re by S hipm an (1981a) an d

is term ed a horizontal tension fa ilu re or fra c tu re by Jo h n so n (1985); it is believed
to result from the b reak ag e o f d ry bone. T he la tte r is called a T ype II spiral
fractu re by S hip m an (1981a) a n d a true spiral o r helical fra c tu re by Jo h n so n
(1985); it is believed to result fro m the b re ak ag e o f fresh, green bone. T he
fractu re planes o f T ype I/h o riz o n ta l ten sio n failures illu stra te d by Jo h n so n
(1985:174) are eith er p erp en d icu lar to the long axis o f the bone o r form an
ob liq u e angle relative to the long axis o f the bone. D av is’ (1985) term “ oblique
fra c tu re ” includes b o th T ype I a n d T ype II spiral fractures.
A m o re p ra ctical p ro b lem fro m a n an a ly tical s ta n d p o in t concerns how one
identifies the k ind o f fractu re displayed by a specim en. B unn (1982:43) p o in ts
o u t th a t “ on a specim en w ith m an y fractu re d edges, it is n o t o b vious how to
define th e scale at w hich descriptive term s such as spiral, stepped, jagged, etc.
sh o u ld ap p ly o r ho w to define a b o u n d a ry betw een the end o f one such fractu re
sh ap e a n d the beginning o f a n o th e r.” P o tts (1982:214), fo r exam ple, re p o rts

th a t in one assem blage o f b ones he studied (N IS P = 627) he fo u n d “ possible
segm ents o f spiral fractu res and lo n g itu d in al b re ak s o n a t least one side o f over
90% o f th e specim ens.” T hese o b serv atio n s u n d ersco re th e descriptive an d
ra th e r general n a tu re o f the fractu re types in F ig u re 8.4.
A n o th e r typ o logy o f fractu re sh ap e is described by Jo h n so n (1985:172, 177).
She illu strates six types o f fractu re shapes as seen in planview . T hese are:
“ tran sv erse (stra ig h t edge), w ide curved (ro u n d ed edge, sides far a p a rt), n a rro w
curv ed (ro u n d ed edge, sides close), p o in ted (converging apex edge), stepped
(in terru p ted edge), a n d scalloped (series o f sem icircles o r curves alo n g an
edge).” R espectively, J o h n s o n ’s types o f fractu re shapes are m o st closely
aligned w ith the sm o o th p erp en d icu lar, spiral, spiral (respectively), V -shaped,
step p ed o r co lu m n ar, a n d irre g u la r p erp en d icu lar types in F ig u re 8.4. H o w
ever, th ere is m in im al agreem ent in d etail betw een th e tw o typologies. This
ag ain u n d ersco res th e subjective a n d in tu itiv e n a tu re o f the typologies, and
indirectly illu strates how difficult they m ay be to use.
B o th typologies described ab o v e are m e a n t to offer insight to the c o n d itio n
o f th e b o n e w hen it w as b ro k e n . F o r exam ple, based on ex perim ental d ata,
J o h n so n (1985:176) m odifies a schem e originally p ro p o se d by M o rla n
(1980:48-49) a n d outlines a set o f criteria m an y analysts use to discrim inate
betw een fractu re d b o nes th a t w ere b ro k e n w hen fresh o r green a n d th o se th a t
w ere b ro k e n w hen the b o n e w as m ineralized. She w rites “ fresh b re ak fractu re
surfaces have the sam e co lo r as the o u te r co rtical surface, exhibit a sm o o th
tex tu re, a n d fo rm acute a n d o b tu se angles w ith the o u te r co rtical surface.
[M ineralized] b re ak fra c tu re surfaces have a co n tra stin g co lo r, exhibit a ro u g h
texture, a n d fo rm rig h t angles w ith the o u te r co rtical su rface” (Jo h n so n
1985:176). T he fra c tu re surface o f specim ens b ro k e n w hen d ry (n o t m in era l
ized) m ay have a ro u g h , b u m p y tex tu re a n d have angles th a t are a co m b in atio n
o f acu te, o b tu se, a n d rig h t relative to th e o u te r co rtical surface. B ones w ith
spiral o r helical (T ype II) fractu res th a t have a ro u g h fractu re surface seem to
rep resen t bones th a t w ere green o r fresh w hen b ro k en , as do flaked bones.
B ones w ith spiral o r helical fractu re s th a t h av e a sm o o th fractu re surface w ere
p ro b a b ly d ry a n d /o r slightly w eathered; these fractures m ay be stepped in p a rt
d ue to th e helical fractu re fro n t e n c o u n te rin g a split line crack. W eath ered a n d /
o r fossilized b on es te n d to b re ak in such m an n ers as to fall in th e stepped,
lo n g itu d in al, a n d sm o o th p erp en d icu lar types. F ra c tu re surfaces m ay pass
th ro u g h th e epiphyseal ends o f d ry bones.
D avis (1985) b ro k e tibiae, fem o ra, a n d h u m eri o f v ario u s A frican bovids.
T h e b ro k e n bo n es w ere v ario u sly fresh a n d dry, a n d lo ad in g w as variously
static a n d d y n am ic b u t was consistently ap p lied to m id -sh aft locations. Som e
b ones w ere also b ro k e n using to rsio n a l loading. F ollow ing re co m m en d atio n s
by Hill (1980), D avis (1985:78-79) developed a system fo r classifying bone
fractu res. T h e classification uses a term in o lo g y th a t is in ten d ed to be purely
T able 8.7 Fracture classification system o f D avis (1985)
A ttribu te
A ttrib u te states (letter designations after D avis 1985:82)
F ractu re orientatio n: relative to the long axis o f the specimen; for long bones, the long axis is
parallel to the long axis o f the com plete bone [N O TE: these a ttrib u te states are in co rp o rated
into the fracture m orp holo gy attrib u te below]
X: mixed
Y: parallel
Z: oblique
F ractu re surface location: based on the anatom ical p osition o f the area o f m axim um exposure
o f fracture surface w ith the specim en lying flat on a h o rizo n tal surface w ith the m arro w cavity
exposed (upw ard)
A: an terio r
B: posterior
C: medial
D: lateral
F ractu re m orphology: the to tal form o f the fracture by independent coding o f the lateral,
proxim al, and distal edges o f the fracture; the fractu re surface closest to the epiphysis is the
proxim al end, the fracture surface farth est from the epiphysis is the distal end, the two
rem aining sides o f the fracture surface are the lateral edges
A: parallel and sm ooth
B: oblique an d sm ooth
C: oblique and stepped
H: curved in one o r m ore planes
I: irregular fractures
J: V -shaped end m orphologies
K: ho rizontal to long axis
descriptive an d n o t im p licate a fractu re ag en t (T able 8.7). D avis (1985:80-81)

suggests th a t each fractu re can be described w ith a fo u r-letter code w hich
describes th e “ fo rm o f the fractu re. F o r exam ple, a fractu re w ith straig h t
m edial a n d lateral edges p arallel to the long axis o f th e bo n e an d p ro x im al an d
distal edges straig h t a n d p erp en d icu lar [to th e long axis o f the bone], w ould be
co ded as an A A A A fra c tu re .” A helical spiral fractu re illu strated by D avis
(1985:82) is coded as a JJJJ a n d a saw -to o th e d specim en (F igure 8.4) as an IIII.
D av is’ (1985) tre a tm e n t o f fractu re lo catio n circum vents the p ro b lem n o ted by
B unn (1982) a n d cited ab o v e a b o u t fractu res th a t display v aried m orphologies.
D a v is’ system allow s the an aly st to reco rd fo u r fractu re m o rp h o lo g ies in fo u r
ad jo in in g lo catio n s alo n g the fractu re surface. H er system does n o t allow
reco rd in g w h eth er the p roxim al edge o f the fractu re surface is n ea r an
epiphysis, in a p ro x im al sh aft lo catio n , o r in a m id sh aft location. I have in
T ab le 8.7 atte m p te d to clarify som e aspects o f D a v is’ classification system , b u t
several p ro b lem s rem ain. F o r exam ple, she m en tio n s th a t several a ttrib u te
states sho uld be reco rd ed relative to one o r m ore o f the follow ing planes: the
“ z ” axis, the “ h o riz o n ta l ax is,” a n d the “ vertical ax is.” D a v is’ (1985) a tte m p t
P r o x i m a l End
P o la r Coordinate Degrees
D i s t a l End
Figure 8.5. F ractu re edge m o rphology o f a b roken m etacarp al (show n in Figure

8.6) illustrated using Biddick and T o m en ch u k ’s (1975) system o f p o lar
coo rd in ates an d vertical planes. Shaded area is the p o rtio n o f the bone
represented.
to circum vent som e o f the in ter-a n aly st subjectivity she correctly perceives in
typological schem es like those described in preceding p a ra g ra p h s is effectively
th w arte d because she defines no n e o f these planes. I believe, how ever, th a t the
a ttrib u te s an d a ttrib u te states she uses are im p o rta n t ones, a n d w ith som e
clarification her classification could becom e a sta n d a rd p a rt o f the recording
an d analysis o f b o n e fractures.
P erh ap s the m o st rig o ro u s system for reco rd in g the m o rp h o lo g y and
lo catio n o f a fractu re is p ro p o se d by Biddick an d T o m en c h u k (1975). T hey
describe a system o f d ividing a long b o n e in to ten equal-sized length sections
w ith w h a t they term “ vertical p la n e s.” T his system utilizes p o la r co o rd in a tes to
the n earest 10° at each vertical p lan e to reco rd fractu re shape. T h e p o lar
co o rd in a tes are m ark e d on elastic b an d s w hich are placed on a specim en a t each
vertical plane, w ith 0° aligning w ith the a n te rio r sagittal p lan e an d 180° aligning
CONVEX
Figure 8.6. F eatures o f fractu re surfaces show n on a bovid proxim al m etacarpal

(after Jo h n so n 1985:177, Figure 5.5; courtesy o f the a u th o r and A cadem ic Press).
w ith the p o ste rio r sag ittal plane. A read in g is ta k e n a t each p o la r co o rd in a te

from the n earest vertical plane a n d re co rd ed o n a b iv aria te g ra p h w ith vertical
planes o n th e vertical axis a n d p o la r c o o rd in a te s o n th e h o riz o n ta l axis.
In te rp o la tio n betw een p lo tte d p o in ts p ro d u ces a tw o -d im en sio n al illu stra tio n
o f th e lo catio n a n d sh ap e o f the fractu re surface in the fo rm illu stra te d in F igure
8.5, w hich show s th e m o rp h o lo g y o f the ex terio r edge o f th e fractu re surface o f
the b o n e illu stra te d in F ig u re 8.6.
W hile B iddick a n d T o m en c h u k (1975:248) em phasize th a t th eir system
p ro vid es d a ta th a t are “ precise, rep eatab le, co m p u terizab le, a n d statistically
an a ly z ab le,” no one to the best o f m y know ledge has em ployed it, a lth o u g h a
d ecade la te r M iinzel (1986) described a sim ilar system fo r re co rd in g bone
frag m en ts using a th ree-d im en sio n al c o o rd in a te system . G iven the subjectivity

o f the typologies o f fractu re types described in preceding p a ra g ra p h s, I think
B iddick a n d T o m e n c h u k ’s (1975) system w a rra n ts careful co n sid eratio n by
ta p h o n o m ists in terested in studying details o f the m o rp h o lo g y o f bone
fractures.
A com m ent about spiral fra ctu res
A m ajo rity o f th e lite ra tu re o n b o n e fractu re concerns w h a t have been generally

labeled spiral fra ctu res. T h a t lite ra tu re h ad its significant beginnings w ith
claim s by D a rt (e.g., 1957) th a t early P leistocene ho m in id s utilized a “ crack-
an d -tw ist” m eth o d o f b re ak in g bones to e x tra ct m a rro w an d p ro d u c e bone
tools. T he b o d y o f lite ratu re th a t resulted, b o th in O ld W o rld an d N ew W orld
contexts, is tru ly im m ense. A sam pling o f th a t lite ratu re is cited in the first
p a ra g ra p h o f this section. W h a t has becom e clear in the years since D a rt’s
claim s w ere m ade is th a t any n u m b e r o f tap h o n o m ic processes can create spiral
fractu res (Type 11/true spiral fractures, above), including tram p lin g (H aynes
1983b, 1991; M y e rs eta l. 1 9 8 0 ),c a rc a sse sfa llin g so m e d ista n c e (L y m a n 1984b),
carn iv o re gnaw ing (B inford 1981b), an d hom in id activities (B onnichsen and
W ill 1980; Z ierh u t 1967). It is also now clear th a t at least tw o kinds o f fractures
have been term ed “ spiral fra c tu re s,” b o th h o riz o n ta l ten sio n failures an d true
spiral o r helical fractu res (Jo h n so n 1985; see above). M o st im p o rtan tly , it is
now k n o w n th a t because a b o n e displays a tru e spiral fractu re (T ype II) does
not signify the ag ent o f fractu re. Jo h n so n (1985:175) re p o rts th a t a tru e spiral o r
helical fractu re indicates only th a t the bone w as b ro k e n w hen fresh; " it does n o t
necessarily indicate the agency inv o lv ed .” She suggests th a t identifying the
agent o f b o n e fractu re requires study o f a ttrib u te s displayed by the bone
surface an d featu res o f the fractu re. D e te rm in atio n o f the agent o f fractu re thus
now involves the stu d y o f m ultiple attrib u te s, no t ju s t the kind o f fractu re
displayed by a specim en. It is identifying agents o f bone fractu re to w hich we
now tu rn .
Fracture agents
Id entifying the ta p h o n o m ic ag en t responsible for the fractu red bones in an

assem blage, w hen possible, can tell us m uch a b o u t the fo rm a tio n a l h isto ry o f a
b o n e assem blage. As listed by D avis (1985:7), m a jo r causes o f bo n e fractu re
include (b u t are n o t lim ited to) ho m in id s m ak in g tools, feeding h o m in id s (e.g.,
L ym an 1978; N o e -N y g a a rd 1977), feeding carn iv o res (e.g., B inford 1981b),
tram p lin g (C h a p te r 9), facto rs related to clim ate such as subaerial w eathering
(C h a p te r 9), a n d p o st-b u rial facto rs such as com pression forces induced by
o v erb u rd en w eight (C h a p te r 11). K u rte n (1953:70) suggests th a t in bone
assem blages w ith m u ltip le indiv id u als a n d a c a ta stro p h ic m o rta lity profile
(C h a p te r 5), b ro k e n b o nes m ay have “ resulted from an im als tu m b lin g do w n a

precipice w hen fleeing before a fire.” L y m an (1984b) describes b ones b ro k e n by
processes related to volcanic activity. C ertain ly the list could be exp an d ed to
include p ath o lo g ical o r a n te m o rte m fractu res as well as o th e r tap h o n o m ic
processes th a t b re a k bones.
T he p ro b lem w hen studying a n assem blage including b ro k e n skeletal
elem ents is to d eterm in e how a n d w hy they are b ro k e n (an d less frequently w hy
som e specim ens are com plete skeletal elem ents, see below ). T he analytical
p ro c ed u re generally involves the study o f m u ltip le a ttrib u te s, especially those
w hich a p p e a r on the basis o f actualistic d a ta to be d iagnostic o f tap h o n o m ic
agents k n o w n a t least occasionally to b re ak bones. I review several o f th e b etter
kn o w n a n d m o st intensively stu d ied agents o f bo n e fractu re , a n d describe the
d iagn o stic a ttrib u te s they produce.
Feeding carnivores as fra c tu re agents

T he d istin c tio n betw een m o d ificatio n a n d d estru c tio n o f bone on one h a n d ,
a n d sim ple fractu re o f bones o n the o th er, is n o t alw ays a clear one in the
lite ra tu re (e.g., B inford 1981b). I review the m a jo r kinds o f m odifications to
bones (C h a p te r 6) a n d a ttrib u te s o f d e stru c tio n o f bones (C h a p te r 7) elsew here.
T h ose discussions su p p lem en t the follow ing.
B one-gnaw ing carn iv o res tend to b re ak b ones tw o w ays, b o th o f w hich
involve static loadin g . O ne involves first rem oving, by chew ing a n d gnaw ing,
one o r b o th epiphyseal ends o f a long bone. T his significantly w eakens the
s tru c tu ra l stren g th o f the bone. T he d iaphysis is th en chew ed, a n d it “ collapses
in to long rectilin ear sp linters th a t generally follow the lo n g itu d in ally aligned
collagen b u n d les” (Jo h n so n 1985:192). G n a w in g m ark s in the fo rm o f furrow s
an d p u n ctu re s rem ain on the epiphyseal end, scoring a n d p ittin g o ccur on the
diaphysis, a n d the end o f the diaphysis m ay be scalloped (F ig u res 6.19, 6.20,
a n d 6.23). T he o th e r w ay ca rn iv o res fractu re b ones involves sim ply chew ing the
co m plete bon e, w ith force applied alo n g th e diaphysis u ntil the bo n e fractu res
(Jo h n so n 1985:192). S coring a n d p ittin g o f the diaphysis are co m m on. T he
k ind s o f fractu res th a t resu lt fro m these tw o w ays in w hich bones are gnaw ed
dep en d o n the skeletal elem ent (m ic ro stru c tu ra l a n d m a c ro stru c tu ra l features)
an d w h eth er th e b o n e is fresh a n d green o r som ew hat d ry (for m ore details, see
B inford 1981b; B onnichsen 1973, 1979; H aynes 1980b, 1982, 1983b; M iller
1969, 1975).
In the only case re p o rted in detail o f w hich I am aw are, M aguire et al. (1980)
describe som e sheep b on es k n o w n to have been b u tch ered a n d gnaw ed by
hu m an s. T h ey in d icate th a t in th eir lim ited sam ple, h u m a n gnaw ing only
cru sh ed an d cru n ch ed fairly soft bone, such as alo n g the edge o f the ischium .
T h u s, th e sim ple presence o f gnaw ing d am ag e o n b ro k e n bones does n o t m ean
carn iv o res are the fractu re agent. C h an n eled fractu res a n d ch ip p ed -b ack edges
on th e diap hy ses o f long bones, fo r exam ple (C h a p te r 6), do n o t seem to be
created by h o m in id gnaw ing, a lth o u g h th e la tte r m ay be occasionally p ro d u ced

by h o m in id s m ak in g b o n e to o ls (see below ). T h u s th e d istrib u tio n s as well as
th e kin d s o f gnaw ing d am ag e m u st be considered.
H om inids as agents o f bone fra c tu re

H o m in id s can b re ak b ones several ways. V irtually all o f th em utilize d ynam ic
loading. D irect percu ssion o ften creates “ p o in t lo ad in g ” (e.g., Jo h n so n
1985:192), percu ssio n pits (B lum enschine a n d Selvaggio 1988, 1991), a n d /o r
flake scars (L y m an 1987a). O th e r featu res dep en d in p a rt on such things as
w h eth er the bone, w hen im p acted , is su p p o rte d by a single anvil u n d er the
im p act p o in t, a n anvil a t each end o f the bone, o r an anvil a t one end a n d the
o th e r end rests on the g ro u n d (Jo h n so n 1985:209: M iller 1989:386).
F resh long b o nes b ro k e n via d ynam ic lo ad in g tend to p ro d u c e tru e spiral o r
helical (T ype II) fractu res. B ut such a fractu re type does n o t, by itself, indicate a
h o m in id fractu re agent. T he presence o f a lo ad in g o r im p act p o in t, in
c o n ju n ctio n w ith th e absence o f gnaw ing m ark s a n d a lo ad in g p o in t the
d iam eter o f w hich is g re ater th a n th a t p ro d u c ed by ca rn iv o re teeth (b u t see
F ig u re 6.25 a n d associated discussion) seem to be the a ttrib u te s used by m ost
analy sts to disting uish h u m an ly b ro k e n b ones fro m th o se b ro k e n by o th er
processes a n d agents.
T h e im pact o r loading p oint is a circu lar o r oval depressed area m ark e d by
incipien t ring crack s o r cru sh ed b o n e (refer to F ig u re 8.6 in this a n d the
follow ing p a ra g ra p h ). T h e o u te r cortical edge, w hen view ed from the exterior,
often will display a crescen t-sh ap ed n o tch a t the lo ad in g p o in t (F ig u re 8.7).
Im p ac t p o in ts are distin guished fro m rebound points w hich rep resen t force
red irected back in to th e b o n e fro m the anvil su p p o rtin g it. In lo ad in g points,
bo n e flakes an a lo g o u s to lithic flakes m ay still be atta c h e d w ithin the m edullary
cavity a t th eir d istal ends; alternatively, flake scars m ay be visible w ithin the
m ed u llary cavity a n d b en e ath the lo ad in g p o in t. A re b o u n d p o in t ten d s to be
sm aller th a n its related lo ad in g p o in t, is lo cated o p p o site its lo ad in g p o in t, an d
displays little to no cru sh in g (Jo h n so n 1985:194, 210).
F ra c tu re surfaces o n b ones b ro k e n w hile fresh ten d to be variously w eakly
concave a n d convex, have a relatively fine tex tu re m acroscopically b u t a rough
tex tu re m icroscopically, a n d fo rm acu te a n d o b tu se angles w ith the o u ter
co rtical surface o f th e bone. O th e r a ttrib u te s o f fractu re surfaces are variously
form ed as a result o f stress relief a n d im p ed im en ts to the fractu re fro n t as it
looses stren g th . H ackle m arks are d isco n tin u o u s curved grooves a n d ridges;
ribs are sem icircular o r a rc u a te ridges th a t are usually c o n tin u o u s a n d concave
relative to the origin o f the fractu re (Jo h n so n 1985). B oth hackle m ark s a n d ribs
are stress relief featu res th a t spread o u tw a rd from the p o in t o f im p act an d are
d iag n o stic o f d y n am ic loading. T he presence o f hackle m ark s a n d ribs indicates
d y n am ic lo ad in g o f green bone; th eir absence does n o t, how ever, in d icate static
loading. C hattering ap p e a rs as highly ac cen tu ate d , closely spaced, straig h t
Figure 8.7. L oadin g p oints, a, deer (Odocoileus sp.) hum erus shaft w ith flake still
in place; b, lower, deer rad iu s shaft w ith flake still in place; upper, flake rem oved
from a loading point.
peaks a n d valleys; stepping is created by split lines causing an in te rru p tio n to the
flow o f force a n d results in stepped o r jag g ed fractu re edges (F ig u re 8.4).
C h a tte rin g a n d step p in g are in terference features. W edge fla kes are bo n e flakes
th a t are rem oved from the ex terio r cortical surface o f the bo n e created by
b end in g failure w hen the bo n e flexes (Jo h n so n 1985:194, 197).
Fracture due to other agents

F ra c tu re d ue to tram p lin g is discussed in C h a p te r 6. C ru sh in g an d fractu rin g o f
b ones due to th e w eight o f sedim ent o v e rb u rd en is m en tio n ed in the follow ing
d iscussion an d in C h a p te r 11. S ubaerial w eathering induced fractu re is
discussed in C h a p te r 9. T he final topic o f b o n e fractu re to cover, then, is how
one analyzes a n assem blage co n tain in g som e b ro k e n skeletal elem ents. In the
follow ing, several different kinds o f analyses are sum m arized to illu strate
v ario u s kin d s o f d a ta a n d an aly tic techniques.
A nalysis o f fra c tu re d bones
Fracture types
In an intensive a n d extensive stu d y o f p a tte rn in g in bo n e fractu res, V illa a n d
M ah ieu (1991) describe a series o f a ttrib u te s o f fractu res, a n d c o m p are the
frequency d istrib u tio n s o f th o se a ttrib u te s across th ree p reh isto ric assem blages
o f h u m a n rem ains from F ran ce. T he L ate N eolithic S arrian s assem blage is
from a collective b u rial a n d is m ade up o f bones believed to have been b ro k e n in
situ via o v erb u rd en cru sh in g as the b ones dried. C o n jo in in g frag m en ts lay
ad jac en t to one a n o th e r, incom plete fractu re s o r crack s w ere n o te d in som e
specim ens, a n d b re ak ag e occurs in b o n es resting o n concave o r convex
surfaces. T h e E arly a n d M iddle N e o lith ic F o n tb re g o u a assem blage is th o u g h t
to re p resen t an in stan ce o f ca n n ib a lism in w hich bones w ere b ro k e n w hile fresh
by ho m in ids. A b o u t 20% o f th e specim ens have im p act notches, h a lf o f w hich
have m icroflakes ad h e rin g to th e im p act p o in t, a n d 30% o f the specim ens have
cu t m ark s. F ra c tu re surfaces have sh arp edges. C o n jo in in g specim ens are
sep a rated by as m u ch as 50 cm , a n d w hen ad jac en t they are n o t in co rrect
a n a to m ica l p o sitio n relative to one a n o th e r. S tratig rap h ic b o u n d arie s closely
define th e extent o f th e bone cluster. C a rn iv o re gnaw ing d am ag e is n o t present.
T he B ezouce assem blage is fro m a collective b u rial a n d consists o f bones
b ro k e n d u rin g excavation. T h e bones w ere d ry (n o t fresh) w hen bro k en .
V illa a n d M ah ieu (1991) c o m p are the p ro p o rtio n a l frequencies o f several
a ttrib u te s p er assem blage. T he fracture angle is the angle form ed by the fractu re
surface a n d th e co rtical surface o f the bone, a n d includes th ree a ttrib u te states:
(1) o b liqu e (o b tu se o r acute), (2) rig h t, a n d (3) o blique a n d rig h t (fractu res w ith
v ariab le angles). T h e fra c tu re outline o r shape includes th ree a ttrib u te states: (1)
tran sv erse (fractu re surface is stra ig h t a n d tran sv erse to the bo n e long axis), (2)
curved (spiral fractu re co m b in ed w ith V -shaped o r p o in te d fractures; com plex,
T ab le 8.8 Frequencies o ffr a c tu r e attributes in three assem blages o f hum an

bones. Values f o r a ttrib ute states are N IS P (% o f toted N I S P ) (fro m Villa
and M ahieu 1991)
Fractu re attrib u te Sarrians Bezouce F o n tb reg o u a
F ractu re angle, to tal N IS P 269 253 174

oblique 22 (8.2) 27(10.7) 114(65.5)
right 176 (65.4) 174 (68.8) 47 (27.0)
oblique and right 71 (26.4) 52 (20.5) 13(7.5)
F ractu re outline, to ta l N IS P 358 287 261
transverse 193 (53.9) 144(50.2) 92 (35.3)
curved/V -shaped 74/32 (29.6) 59/23 (28.6) 42/92 (51.3)
interm ediate 59(16.5) 61 (21.2) 35 (13.4)
F racture edge, to tal N ISP 358 287 261
sm ooth 111 (31.0) 158 (55.1) 163 (62.5)
jagged 247 (69.0) 129 (44.9) 98 (37.5)
Shaft circum ference, to ta l N IS P 226 93 151
< 1/2 16(7.1) 33 (35.5) 115(76.2)
> 1/2, < com plete 10(4.4) 0 (0.0) 23 (15.2)
com plete 200 (88.5) 60 (64.5) 13 (8.6)
m u ltid ire ctio n a l fra c tu re surface), a n d (3) in term ed iate (stra ig h t m o rp h o lo g y

b u t d iag o n al to b o n e long axis, a n d stepped). Fracture edge refers to the tex tu re
o f the fractu re surface an d includes tw o a ttrib u te states: (1) sm o o th a n d (2)
jag g ed . S h a ft circum ference excludes specim ens < 4 cm long a n d h as three
a ttrib u te states: (1) m ax im u m circum ference is less th a n h a lf o f the original
skeletal elem ent, (2) m ax im u m circum ference is m o re th a n h a lf in a t least a
p o rtio n o f the length, a n d (3) co m p lete circum ference in at least a p o rtio n o f the
length. S h a ft length concerns how m uch o f the diaphysis is represented,
excludes specim ens < 4 cm long, does n o t include co n sid eratio n o f the presence
o r absence o f the epiphyseal ends, a n d consists o f fo u r a ttrib u te states: (1) shaft
is < o n e fo u rth the length o f the com plete bone, (2) sh aft is > one fo u rth
b u t < one h a lf o f com p lete length, (3) sh aft is > one h a lf b u t < th ree fo u rth s o f
com plete length, a n d (4) shaft is > three fo u rth s o f com plete length. (T he shaft
circum ference a n d sh aft length variables are sim ilar to ones d eveloped by B unn
[1983].) S h a ft fra g m en ta tio n is d eterm in ed by p lo ttin g sh aft circum ference and
sh aft length in a b iv aria te g ra p h (Villa a n d M ah ieu 1991:42). V illa an d M ahieu
also exam ine th e b re a d th /le n g th ra tio s o f sh aft frag m en ts, including only
specim ens w ith a b re a d th less th a n the original com plete b o n e an d > 4 cm long
because sm aller specim ens are difficult to identify to skeletal elem ent.
D a ta fo r fractu re angle, fractu re outline, fractu re edge, sh aft circum ference,
a n d sh aft length fo r the th ree assem blages are given in T able 8.8. C h i2 analyses
o f th e p ro p o rtio n a l frequencies o f fractu re angles a n d fra c tu re outlines indicate
the S arria n s an d B ezouce assem blages are sim ilar to one a n o th e r a n d b o th
Fracture Angle
■ oblique
0 right
□ oblique & right
Sarrians Bezouce Fontbregoua

Assemblage
Fracture Outline
■ transverse
curved/V-shaped
□ intermediate

Assemblage
differ from the F o n tb re g o u a assem blage (F ig u re 8.8a a n d 8.8b). T h u s Villa an d

M ah ieu (1991) suggest th a t these tw o fractu re a ttrib u te s aid in distinguishing
bones fractu re d w hen fresh from those fractu re d w hen dry. C h i2 analysis o f the
fractu re edge v ariab le did n o t distinguish the assem blages in this m an n er, an d
Villa a n d M ah ieu (1991:40) co n clu d e th a t “ this a ttrib u te will n o t d iscrim in ate
betw een old a n d fresh b o n e b re a k a g e .” T hey suggest th a t the kind o f force
applied - static o r d yn am ic - m ay exert a stro n g er influence on this variable
th a n w h eth er a b o n e is old an d desiccated o r fresh w hen bro k en . T he
p ro p o rtio n s o f catego ries o f sh aft circum ference com pleteness also indicate
g reater sim ilarity betw een the S arrian s a n d Bezouce assem blages th a n betw een
eith er o f th em a n d the F o n tb re g o u a assem blage (F ig u re 8.8c). V illa an d
Shaft C ircum ference
■ < 1 /2
El > 1/2
□ complete

Assem blage
Figure 8.8. B ar grap h s o f three bone frag m en tatio n a ttrib u te s for three
assem blages, a, percentage o f to tal frequency p er k in d o f fractu re angle; b,
percentage o f to tal frequency per kind o f fracture outline; c, percentage o f total
frequency per size o f shaft circum ference (from T able 8.8).
M ah ieu (1991:41) co n clu d e th a t “ high frequencies o f com plete sh aft diam eters
a p p e a r to ch aracterize assem blages o f p o st-d ep o sitio n ally b ro k e n b o n e s” (see
also C h ap ters 10 a n d 11). T hey co n clu d e th a t p o st-d ep o sitio n al b re ak ag e tends
to create sh o rt tu b u la r (com plete circum ference) pieces o f shafts w hereas
fractu re o f green bo n e creates splinters o f vario u s lengths w ith variously
incom plete a n d co m p lete circum ferences. T h u s the b re a d th to length ra tio is
low est fo r th e F o n tb re g o u a assem blage a n d is higher fo r the S arrian s an d
B ezouce assem blages (the la tte r tw o are n o t statistically significantly different).
V illa a n d M a h ie u ’s (1991) analysis dep en d s on the frequencies o f various
m o rp h o lo g ical a ttrib u te s o f the fractures. T his raises the issue o f how to
q u an tify o r tally such attrib u tes. It seem s th a t V illa a n d M ah ieu (1991) tally the
frequency o f specim ens o r N IS P th a t display each attrib u te . T h u s a single
specim en d isp laying a spiral (o r transverse) fractu re w ould be tallied as an
N IS P o f 1. But tw o possibilities th a t m ight influence such tallies are evident.
F irst, given th a t som e specim ens could be refit (e.g., V illa et al. 1986), the
p ro b a b ility th a t a p a rtic u la r fra c tu re type w as tallied tw ice fo r the sam e skeletal
elem ent seem s great. T h a t is, a b o n e b ro k e n in to tw o pieces an d displaying an
ob liqu e fra c tu re will be tallied tw ice, once fo r each piece. Is this th e co rrec t w ay
to tally th is a n d o th e r fractu re a ttrib u tes? T he specim ens are an a to m ica lly
in terd e p en d en t, b u t sho uld they be tre a te d as in d ep en d en t units d u rin g
q u an tificatio n ? T h e significance o f this m ay be g reat if, say, o blique fractu res
are m o re easily refit th a n jag g ed o r stepped fractures. M o re study o f this aspect
o f the q u an tifica tio n o f fractu re a ttrib u te s is required.
T h e second issue co n cern in g the q u a n tific a tio n o f b ro k e n bones involves the
8.9 Frequencies o f skeleta l p a rts in raptor pellets (from H offm an 1988). N whole = num ber o f whole specim en
lete skeleta l elem ents. % whole = 100 ( N whole j N I S P ) . N I S P . M N E calculated w ithout whole specim ens
G reat horned owl Screech owl Diurnal hawk
l part NISP N whole MNE % whole NISP:M NE NISP N whole M NE % whole NISP :MNE NISP N whole MNE % whole NISP
le 92 57 84 62.0 1.30 24 4 18 16.7 1.43 26 0 18 0.0 1.44

81 25 78 30.9 1.06 19 0 17 0.0 1.12 3 0 3 0.0 1.06
us 92 80 90 97.6 1.20 25 9 22 36.0 1.23 12 0 7 0.0 1.46
69 53 66 76.8 1.23 17 11 16 64.7 1.20 4 0 4 0.0 1.10
81 64 80 79.0 1.06 10 0 6 0.0 1.67 5 0 4 0.0 1.46
nate 73 30 73 41.1 1.00 18 3 18 16.7 1.00 5 0 5 0.0 1.00
99 83 95 83.8 1.33 26 12 18 46.2 2.33 11 0 7 0.0 1.95
100 91 96 91.0 1.60 29 9 18 31.0 2.22 12 0 7 0.0 1.96
s: 687 483 662 70.3 168 48 133 28.6 78 0 55 0.0
= 48 11 9
fact th a t som e b ro k e n bones m ay display one kin d o f a ttrib u te on one p a rt o f

the fractu re surface a n d a n o th e r, d istin c t kin d o f a ttrib u te o n a different
p o rtio n o f th e fra c tu re surface. T he frequency w ith w hich this will occur
dep en d s, o f course, o n how a ttrib u te s o f fractu re are defined. T h u s, the
a ttrib u te state “ o bliqu e fra c tu re ” is d istin c t fro m th e a ttrib u te state “jagged
fra c tu re ,” b u t only by h av in g an a ttrib u te state “ o blique an d ja g g e d ” w ould
one be able to tally a specim en displaying b o th a ttrib u te states. T he la tte r m ay
o ccu r w hen a diap h ysis specim en is obliquely fractu re d o n one end b u t has a
jag g ed fractu re a t the o th e r end. A gain, th e influence o f this kind o f possibility
has n o t been co n sid ered in th e literatu re. It should be in the future.
P roportions o f broken bones

A p o ten tially im p o rta n t ta p h o n o m ic v ariab le th a t sh o u ld be considered in
studies o f b o n e fra g m e n ta tio n is the frequency o f com plete skeletal elem ents o r
u n b ro k e n w hole bones. T o ad d ress this topic, I distinguish the extent o f
fra g m e n ta tio n fro m the in ten sity o f frag m en tatio n . By e x te n t o f fragm entation
I m ean th e p ro p o rtio n o f the to ta l N IS P o f an assem blage th a t consists o f
b ro k e n a n d incom plete skeletal elem ents a n d /o r the p ro p o rtio n o f the to tal
N IS P o f a n assem blage th a t consists o f w hole, u n b ro k e n , com plete skeletal
elem ents. By intensity o f fra g m en ta tio n I m ean the size o f the fragm ents. A n
assem blage o f very intensively b ro k e n skeletal elem ents will have m an y sm all
frag m en ts w hereas an assem blage o f less intensively b ro k e n skeletal elem ents
will h ave m an y large fragm ents. O f course, w hen m easu rin g the intensity o f
fra g m e n ta tio n , o ne m u st be sure to co m p are frag m en ts th a t com e fro m bones
o f sim ilar size. In the follow ing I first co n sid er m easures o f the extent o f
frag m en tatio n . I th en describe tw o w ays to m easu re the in ten sity o f
frag m en tatio n .
E x te n t o f fra g m en ta tio n
T he p ro p o rtio n o f th e to ta l N IS P rep resen ted by w hole skeletal elem ents is a
ra th e r qu ick w ay to m easu re the degree to w hich the b ones in an assem blage are
b ro k e n (e.g., D o d s o n a n d W exlar 1979). F o r exam ple, the d a ta in T ab le 8.9
in d icate th a t bones recovered fro m g re at h o rn e d owl (Bubo virginianus) pellets
are m o re likely to be w hole (70.3% o f the to ta l N IS P are w hole) th a n bones
recovered fro m pellets d ep o sited by screech owls ( O tus asio) (28.6% o f the to ta l
N IS P w hole); pellets cast by d iu rn a l h aw ks c o n ta in n o w hole bones. T he
differences in p ro p o rtio n s o f w hole bones in the to ta l assem blages are
statistically significant: g reat h o rn e d owl to screech ow l arcsine ?s = 9.94,
P < 0.001; g reat h o rn e d owl to d iu rn a l h aw k /s = 16.61, P < 0.001; screech owl
to d iu rn al haw k /s = 8.23, / ) < 0 .0 0 1 .
T h e an a ly st co u ld now inspect individual skeletal elem ents to d eterm in e if
certain skeletal elem ents are m o re o ften com plete w hen dep o sited by great
h o rn e d owls th a n w hen dep o sited by screech owls. O n one h an d , the rad iu s
c.
F igure 8.9. V ariation in the p ro p o rtio n o f com plete skeletal elem ents (% W hole)
deposited by tw o tax a o f owls (from T able 8.9).
ten d s to have as m uch chance o f being w hole w hen d ep o sited by great h o rn ed

owls as w hen d ep o sited by screech owls (76.8% w hole versus 64.7% w hole,
respectively; arcsine ?s = 0.99, P > 0.2). O n the o th e r h a n d all o th e r skeletal
elem ents are m uch m ore likely to be b ro k e n by screech owls th a n they are by
great h o rn e d owls (F ig u re 8.9). A nd, the p ro p o rtio n s o f w hole specim ens per
skeletal elem ent are n o t statistically sim ilar betw een these tw o bone-accum u-
latin g a n d fractu rin g ag ents (r = 0.55, P = 0.15). W hile o th e r d a ta suggest these
differences sho u ld p ro b a b ly n o t be used as d iagnostic a ttrib u te s o f these bone-
accu m u latin g ag ents (e.g., D o d so n an d W exlar 1979; K u sm er 1990), I have
here show n how the p ro p o rtio n o f w hole bones can serve as an im p o rta n t
m easu re o f th e e x te n t o f bone frag m en tatio n (see also T o d d an d R ap so n 1988).
In tensi ty o ffra g m e n tat ion

T h ere are tw o o b v io us w ays to m easure the intensity o f frag m en tatio n . 1
co n sid er each in tu rn .
Sizes o f broken bones: As im plied by V illa an d M ah ieu (1991), the size o f

d iap h y sis frag m ents is a good m easure o f the intensity o f frag m en tatio n . T o
illu strate this, I co m p are tw o sam ples o f bone fragm ents recovered from late
H olocene arch aeo log ical sites in N o rth A m erica. O ne sam ple com es from a site
in eastern W a sh in g to n ( 4 5 0 K 11; L ym an, u n p ublished d ata); the o th e r sam ple
com es from a site in M issouri (2 3 L N 104; L ym an a n d O 'B rien 1987). F o r each.
o
c
CD
3
cr
a>
c
a>
oV_
O)
CL
C\J CO LO CD N
Size Class (1 cm)

Figure 8.10. P ro p o rtio n al frequencies o f 1 cm size classes o f long bone diaphysis
fragm ents in tw o assem blages o f deer bones.
the lengths o f long bo n e diaphysis frag m en ts o f deer (Odocoileus sp.) a n d deer

sized artio d a cty ls w ere m easu red to the n ea rest m illim eter. T he sam ple from
4 5 0 K 11 consists o f 306 specim ens; the sam ple from 2 3 L N 104 consists o f 972
specim ens. B oth sam ples a p p e a r to be largely a ttrib u ta b le to the b u tch erin g
activities o f h u m a n o cc u p an ts o f th e sites. T he frequency d istrib u tio n s o f the
percen tag e o f to ta l specim ens p er 1 cm size class a p p e a r to be ra th e r sim ilar
betw een th e tw o sites (F ig u re 8.10). H ow ever, th e tw o d istrib u tio n s are
significantly different statistically (K o lm o g o ro v -S m irn o v tw o-sam ple
D = 0.233, / >< 0 .0 1 ). T he g reatest differences occu r in the 2 .1 -3 .0 to 4 .1 -5 .0 cm
size classes; p ro p o rtio n a te ly m o re frag m en ts fro m 2 3 L N 104 occu r in these size
classes th a n in the 4 5 0 K 1 1 sam ple. As well, the 4 5 0 K 1 1 sam ple h as p ro
p o rtio n a te ly m ore specim ens 6.1 to 10.0 cm long. T he frequency d istrib u tio n o f
frag m en t size classes (F ig u re 8.10) suggests th e 23L N 104 assem blage is m o re
intensively frag m en ted th a n the 4 5 0 K 1 1 assem blage because th e form er
assem blage has m o re sm aller frag m en ts th a n the la tte r assem blage. W hy these
differences o ccu r th u s becom es the focus o f fu rth e r analysis. W hile I d o n o t
have th e d a ta necessary to p u rsu e such analyses, I w ould next inspect the
p ro p o rtio n o f specim ens th a t display carn iv o re gnaw ing m ark s, th e p ro p o rtio n
o f specim ens display ing flake scars a n d p ercussion m ark s, the w eath erin g
stages disp layed by in d iv id u al specim ens, a n d o th e r a ttrib u te s k n o w n to covary
an d be ta p h o n o m ic a lly in terre lated w ith fra g m e n ta tio n processes. A n o th e r
NISP
Figure 8.11. A m odel o f the relatio n betw een N IS P an d M N E in an assem blage
o f bones. P lo tted points will alw ays fall on the diagonal (N IS P = M N E ) o r in the
shaded area (N IS P ) M N E ). The d a rk shaded area represents the fact th at a
skeletal elem ent can only be b ro k en in to a finite nu m b er o f identifiable
fragm ents, here m odeled as 15 fragm ents p er one M N E .
a ttrib u te o f th e assem blages th a t m ig h t help u n rav el th e fra g m e n ta tio n h isto ry

o f these assem blages is N IS P :M N E ratios.
N I S P . M N E ratios: R ic h a rd so n (1980:111) w as one o f the first ta p h o n o m ists

to calcu late N IS P :M N E ra tio s, o r w hat he called a fragm entation ratio.
S u bseq uen t researchers have follow ed his lead (e.g., Schick et al. 1989). Recall
th a t M N E is the m in im u m n u m b e r o f com plete skeletal elem ents necessary to
a c c o u n t fo r th e specim ens identified, a n d th a t th o se specim ens m ay be
v ario u sly co m plete a n d in co m p lete bones. If, fo r exam ple, there are 2 p ro x im al
left h u m eri a n d 6 d istal left h u m eri in an assem blage, th e N IS P fo r left h um eri is
8 w h ereas the M N E fo r left h um eri is 6 i f n o n e o f the p ro x im al specim ens
o verlaps w ith a d istal specim en. I f one o f the p ro x im al specim ens h as a p o rtio n
o f d iap h y sis th a t is also rep resen ted on one o f the d istal specim ens, the tw o
specim ens o verlap a n d cam e fro m different skeletal elem ents. In this case the
N IS P w ould still be 8 b u t the M N E w ould be 7. G iven how N IS P a n d M N E are
tallied, N IS P will alw ays be g re ater th a n o r equal to the M N E fo r an
assem blage (F ig u re 8.11). If N IS P = M N E , then eith er all o f the specim ens are
com plete skeletal elem ents, o r all o f the specim ens o verlap o r re p resen t the
sam e p o rtio n o f a skeletal elem ent, such as w hen all are the distal end o f the
tibia. If N IS P > M N E , th en som e specim ens are frag m en tary a n d o verlap o f

specim ens is lim ited. B ecause any given skeletal elem ent can be b ro k e n into
som e finite n u m b e r o f specim ens th a t are identifiable to skeletal elem ent (e.g.,
L ym an an d O 'B rien 1987), M N E will be < N I S P to som e lim ited degree
(F ig u re 8.11). T h e in tensity o f fra g m e n ta tio n den o tes the sm allness o f the
frag m en ts. Sm all frag m en ts are less likely to o verlap w ith one a n o th e r th a n
large frag m en ts a n d be unlikely to be show n to be in d ep en d e n t o f one an o th er.
T h u s, a large frag m en t will c o n trib u te one to N IS P a n d p ro b a b ly will
c o n trib u te one to M N E tallies w hereas a sm all frag m en t will co n trib u te one to
tallies o f N IS P b u t will p ro b a b ly n o t c o n trib u te one to M N E tallies (the
sm allest frag m en ts will be unidentifiable a n d c o n trib u te to neither).
N 1 S P :M N E ra tio s should be calculated using tallies fo r N IS P a n d M N E
d erived only fro m fragm ents; w hole o r com plete skeletal elem ents should be
excluded. In clu d in g w hole bones in tallies o f N IS P a n d M N E reduces th eir
p ro p o rtio n a l difference because b o th are increased the sam e ab so lu te a m o u n t
by inclusion o f w hole bones. F o r exam ple, in a n assem blage o f bo n e fragm ents
only th a t has an N IS P o f 10 a n d an M N E o f 5, the N IS P :M N E ra tio is 2.0.
A d d in g 2 com p lete skeletal elem ents to the assem blage p ro d u ces an N IS P o f 12,
an M N E o f 7, a n d a N IS P :M N E ra tio o f 1.71. T h u s, th e m ore com plete bones
in an assem blage (the g re ater the p ro p o rtio n o f the to ta l N IS P th a t are
com p lete bones), th e less the p ro p o rtio n a l difference will be betw een N IS P and
M N E , a n d th e low er the N IS P :M N E ratio . In T able 8.9, the ra tio for tibiae
d ep o sited by th e g re at h o rn e d owl w ould change fro m 1.60 using only
frag m en ts to 1.04 if w hole b ones are included; the ra tio fo r scapulae w ould
change fro m 1.06 to 1.04 because there are few er co m p lete scapulae (30.9% )
th a n th ere are com p lete tibiae (91.0% ).
T h e N IS P :M N E ra tio s in T ab le 8.9 w ere calcu lated w ith o u t w hole bones
in clu d ed in the tallies. T h ose ra tio s in d icate th a t fem o ra a n d tibiae d ep o sited by
screech ow ls a n d d iu rn a l haw ks (ratio s all > 1 .9 5 ) are m o re intensively frag
m en ted th a n th o se skeletal elem ents are w hen dep o sited by g reat h o rn e d owls
(ratio s < 1.60). T he average N IS P :M N E ra tio across the eight skeletal elem ents
listed in T ab le 8.9 fo r g reat h o rn e d owl (avg. = 1.22) is m o re different from the
average ra tio fo r screech ow ls (avg. = 1.52; / = 1.58, / >= 0.14) a n d d iu rn a l
h aw k s (avg. = 1.43; t= 1.38, f >= 0.19) th a n the la tte r tw o are from each o th er
(/ = 0.43, P = 0.61), suggesting th a t screech owls a n d d iu rn al h aw k s m ore
intensively frag m en t bones (b reak them in to sm aller pieces) th a n g reat h o rn e d
owls.
F in ally , it is im p o rta n t to n o te th a t m o st researchers w ork in g w ith bone
assem blages d ep o sited by ra p to rs (A ndrew s 1990; D o d so n a n d W exlar 1979;
H o ffm an 1988; K u sm er 1990) re p o rt th a t these b o n e-acc u m u la tin g an d
dep o sitin g agents ten d m o st freq u en tly to b re ak the scap u la a n d the in n o m i
nate. T h e percen t w hole values fo r these tw o skeletal elem ents across all three
tax a o f ra p to rs tend to be sm all ( < 3 0 % ) . N IS P :M N E ra tio s fo r these skeletal
elem ents, w hen calcu lated using only frag m en ts, ten d n o t to reflect this fact (all
ra tio s < 1.12). H ow ever, the p ercen t w hole values in com bination with the
N 1 S P :M N E ra tio s fo r these tw o skeletal elem ents in tu rn suggest these ra p to rs
d ep o sit large frag m en ts o f scapulae an d in n o m in ates, an d those fragm ents
o v erlap one an o th er.
S u m m ary
Bones are m ade up o f a co m posite, viscoelastic m aterial th a t, w hen green an d

fresh, fractu res acco rd in g to b iom echanical p ro p e rtie s th a t have evolved
th ro u g h tim e. S tudies o f bone as a m aterial g ra n t insights to th eir fractu re
m echanics. U n d e rstan d in g those m echanics in tu rn allow s us to discern the
c o n d itio n o f th e b o n e w hen it w as b ro k en . T his is largely restricted to
d istinguishing fractu re s th a t occu rred w hen the bone w as fresh a n d green from
fractu res th a t to o k place w hen the bo n e w as d ry o r m ineralized. T he im p o rta n t
topic o f how bones b re ak w hen b u rn e d (see C h a p te r 9 fo r discussion o f
b u rn in g ) h as n o t been covered here because little research has been do n e on it.
T ypologies o f fractu res have been described by vario u s researchers, b u t m any
o f these suffer from having to o generally o r am b ig u o u sly defined a ttrib u te s and
no clear specification o f th e a ttrib u te s th a t are necessary a n d sufficient to assign
specim ens to a type. T his m akes the typologies difficult to use an d no t readily
replicable from an aly st to analyst. Biddick a n d T o m e n c h u k 's (1975) d escrip
tive tech niqu e m ay pro v id e d etailed d escrip tio n an d exact re co rd in g o f fractu re
edges a n d surfaces, b u t is less readily co nverted to a typological o r classification
system . C learly th ere is ro o m fo r im p ro v em en t here.
S tudy o f how bones are b ro k e n is geared generally to w ard s ascertain in g the
ta p h o n o m ic ag ent responsible fo r the frag m en tatio n . T he exam ples o f analyses
o utlin ed abo ve focus on how frag m en tatio n m ight be analytically m an ip u lated
d u rin g such ascertain m en t. A nd while several o f those exam ples utilize bones
dep osited by ra p to rs, there is no reaso n w hy th o se sam e analytical techniques
c a n n o t be ap plied to bones th o u g h t to have been b ro k e n a n d dep o sited by
hom inids.
Bone artifa cts
A rtifacts are defined as objects th a t have been m ade o r m odified by h u m an beings.

(J. Bower 1986:14)
T ool d eterm in ation com es from recognition o f a use-w ear p a tte rn after hum ans
have been established as the agency o f [bone] breakage.
(E. Joh n so n 1985:175)
A rtifa c ts an d tools
T he defin ition o f a rtifact given ab o v e is sim ilar to th a t fo u n d in virtu ally every

in tro d u c to ry arch aeo lo g y text. Its significance becom es clear w hen the im plied
d efinition o f a tool given ab o v e is considered. Tools are a rtifa cts th a t exhibit
B utchering, bone fracturing, and bone tools 339
use-w ear th a t w as created because th e a rtifa c t w as used by h u m an s to p erfo rm

som e fu n ction. Such a definition is no d o u b t far m ore n arro w th a n m ost
arch aeo lo g ists will accept because artifa cts m ay be used insufficiently to
g en erate use-w ear. A n d , o f course, one could w o n d e r if b ead s a n d o rn a m e n ts
m ad e from b o n e a n d to o th are tools; they are used, b u t p ro b a b ly will no t
display use-w ear. Sim ilarly, because h u m a n s as ta p h o n o m ic agents can m odify
bones d u rin g b u tch ery by eith er gen eratin g cut m ark s on them o r break in g
th em fo r m arro w ex tra ctio n o r b u rn in g th em as fuel in a fire, does th a t
m o d ificatio n m ake tho se b ones artifacts? By definition it does. H ow ever, m ost
arch aeo lo g ists ten d to so rt an im al rem ains - h u m an ly m odified o r n o t - from
artifa cts in to the sep a rate categ o ry eco fa cts: “ the m aterial residue o f the
en v iro n m e n t” (B ow er 1986:14). A gain, this is a typical definition. I believe th a t,
given p receding sections o f this c h a p te r, h u m an s d o m odify bones a n d th u s by
defin itio n the m odified b ones w ould be artifa cts. B ut I also believe th a t stone
to o l m ateria l a n d fine-grained sedim ent used to m ake projectile p o in ts an d
p o ttery , respectively, also are m aterial residues o f the en v iro n m en t. T he
d istin c tio n o f artifa cts, tools, a n d ecofacts, th en , is n o t so clear-cut as it m ight
first ap p e ar.
O n one h a n d the preceding p a ra g ra p h m ay sim ply c a p tu re the essence o f a
sem an tic q uibble. W e all ra th e r intuitively kn o w w h a t artifa cts are a n d how
they differ from ecofacts, a lth o u g h the d istin ctio n betw een artifa cts a n d tools
m ay be less clear. O n the o th e r h a n d , th e discussion hinges o n the term
“ m o d ificatio n ,” a n d p erh ap s th a t is w here we should start. F o r exam ple, as
n o ted earlier in this c h a p te r, the creatio n o f b u tch ery m ark s - a fo rm o f
m o d ificatio n - m ay well be an incidental b y p ro d u c t o f th e h u m a n b eh a v io r o f
m od ifying an im al carcasses in to co n su m ab le resources. A s such, bu tch ery
m ark s w o uld be unintentional epiphenom ena o f the in ten d ed m o d ificatio n (see
L y m an 1993a fo r ad d itio n a l discussion). F ra c tu re d bones resulting from
m a rro w e x tra ctio n w ould be an in ten d ed result o f th e activity, b u t th e in ten t is
to p ro d u c e a n edible resource, n o t a to o l o r a rtifa ct, a lth o u g h as we will see
b ones b ro k e n fo r the fo rm er m ay be used fo r the latter. M a n u fa c tu rin g an
a rtifa c t o r to o l results fro m the intentional m o d ificatio n o f a n a tu ra l substance,
w h eth er th a t su b stan ce is clay fo r a ceram ic, sto n e fo r a projectile p o in t, o r bone
fo r a n awl. U se-w ear is an u n in te n tio n a l b y p ro d u c t o f use, regardless o f the tool
m aterial.
G iven th a t highly m odified artifa cts are easily recognized an d identified as
artifa cts, th e q u estio n o f g reatest m o m en t fo r tap h o n o m ists is identifying those
objects th a t m ight be artifa cts given the lim ited m odification they display. T he
re m a in d er o f this section, th en , deals exclusively w ith this topic.
Id en tifyin g bone tools

T he p ro b lem o f distin g u ish in g n atu ra lly m odified objects fro m those m odified
by h o m in id s a n d th u s those rep resen tin g artifa cts, especially sto n e artifacts,
has a deep h isto ry (G ray so n 1986). T he parallels betw een techniques developed
for so rtin g sto n e artifa cts from a collection o f rocks, som e n atu ra lly a n d som e
artificially m odified, a n d bo n e artifa cts from a pile o f faunal rem ains, are
re m ark ab le. T he basic tech n iq u e involves study o f tw o kinds o f a ttrib u te s, w hat
I have term ed a ttrib u te s d en o tin g purposefulness a n d a ttrib u te s o f prim itiveness
(L y m an 1984b). As B inford (1981 b:4—8) d o cu m en ts, the fact th a t som e
m odificatio n s result in " re d u n d a n t p a tte rn in g p ro d u c in g a result to a design o r
p la n ” h as lo ng served arch aeo lo g ists searching fo r p u rp o sefu l m odification o f
objects. P urp osefulness reflects in ten tio n a l m odification in the sense no ted
earlier. P u rp o sefu l m odifications are “ repetitive, system atic, p lan n ed and
c o n tro lle d ” (A scher a n d A scher 1965:244—245). T he prim itiveness o f a m odifi
c a tio n “ involves b o th the so p h istic atio n a n d degree o f m o d ificatio n to the
object. T he less the degree o f so p h istic atio n [and degree] o f m o dification, the
m o re ten u o u s the identification o f an object as an a rtifa c t” (L ym an 1984b:325).
P u rp o sefu l sh ap in g o f a n object by h o m in id s to p ro d u ce a desired m o rp h o
m etry can be term ed m anufacturing m odification (L y m an 1984b:325). M a n u
factu rin g m o d ificatio n can ran g e in so p h istic atio n a n d degree from the
extensive m o d ificatio n necessary to p ro d u c e a S o lu trean biface o r a C lovis
p o in t to th e m inim al o r lack o f m an u fa c tu rin g m odification required o f w hat
h av e been called in sta n t, im p ro m p tu , o r expedient tools. O bviously, the m ore
extensive the m a n u fa c tu rin g m o d ificatio n to a bone specim en, the easier it will
be to identify th a t specim en as a n a rtifa c t sim ply because “ no reaso n ab le
co m b in a tio n o f conceivable ag en ts o th e r th a n people could have p ro d u c ed [the
m o d ificatio n ]" (L y m an 1984b:328). Sem enov (1964:143-195) still provides one
o f th e m ost extensive an d detailed co n sid eratio n s o f prim itive bo n e tool
m a n u fa c tu rin g m o d ificatio n available, a n d M a c G re g o r (1985) describes m ore
m o d ern tech n iq u es o f m ak e bo n e tools. U se-w ear, because it is an incidental
resu lt o f an object being used as a tool, m ay n o t be very extensive o r obvious. In
co n ju n ctio n w ith m in im al m a n u fa c tu rin g m odification, m inim al use-w ear can
result in a b o n e object th a t has been only slightly m odified, a n d th a t will th u s be
difficult to identify as an artifact.
Extensively m odified bones, such as the scapulae show n in F igure 8.12,
clearly could no t have been p ro d u c ed by n a tu ra l processes. T here w ould be
little arg u m e n t th a t h om inids w ere the m a jo r agent o f m odification, th a t the
intend ed result w as a too l (an aw l) to be used, a n d th a t the sam e kind o f tool w as
being p ro d u c ed by tw o distinct m a n u fa c tu rin g p ro ced u res. T he p ro b lem o f
id entifying bon e to o ls clearly resides w ith th o se bones th a t have been m inim ally
m odified. In these cases, the distribution o f the a ttrib u te s o f m o d ificatio n
becom es a critical v ariab le (see below ). W hile stu d y o f th a t variable m ay well
increase th e p ro b a b ility th a t m inim ally m odified bones are identified as tools,
F ig u re 8.13 illustrates specim ens w ith m inim al ro u n d in g a n d /o r ch ip p in g o f
fractu re p o in ts an d edges th a t w ere n o t p ro d u c ed by h om inids, b u t w hich m ight
be m isidentified as to o ls if fo u n d in asso ciatio n w ith in d isp u ta b le artifacts.
G iv en th a t highly m odified artifa cts are easily recognized an d identified as
a
Figure 8.12. Prehistoric deer (Odocoileus sp.) scapula awls from eastern
W ashington state, a, u p p er three specim ens are b lan k s form ed by pecking and
breaking po sterio r m argin from the scapula blade, low er specimen is finished
product; b, m edial surface o f a scapula show ing the engraved groove used to
rem ove the p o sterio r m argin from the blade.
Figure 8.13. Pseudotools, a, natu rally b ro k en an d flaked hum erus shaft; b.

naturally brok en, flaked an d p o in ted tibia shaft (from L ym an 1984b:321, Figure
4; and L ym an 1984b: 320, Figure 3, respectively; courtesy o f the Society for
A m erican A rchaeology).
artifacts, how do we d eterm in e if objects th a t display lim ited m odification are

artifacts? In the follow ing I review som e o f th e criteria frequently used to so rt
skeletal rem ains in to th ose m odified by h o m in id s fo r use as artifa cts an d tools
a n d th o se used as to o ls, from all o th e r specim ens, regardless o f w h eth er the
la tte r w ere m odified (such as those th a t are b u tch ery m ark e d ) by ho m in id s or
C o n text
T h e co n text o f a skeletal specim en den o tes its sp atial lo catio n a n d its spatial
asso ciatio n w ith o th e r objects, including cu ltu ra l item s such as u n d isp u ted
artifa cts a n d n a tu ra l item s such as sedim ents. A n archaeological co n text is
p ro d u c ed by a u n iq u e c o m b in a tio n o f n a tu ra l a n d cu ltu ra l processes o p eratin g
on n a tu ra l a n d cu ltu ra l m aterials in a lim ited geographic area over som e
tem p o ra l span. A n arch aeological co n tex t is usually specified by labelling the
spatial area an archaeological site. T he m o st typical in d icatio n th a t one has
fo u n d a site is the presence o f artifacts (for discussion here, these include
discrete p o rta b le objects as well as w h a t are o ften term ed features - sets o f
discrete objects w hich owe a t least p a rt o f th eir artificiality to th eir spatial
asso ciatio n s w ith one a n o th e r), o r objects th a t owe any o f th eir a ttrib u te s to
h u m an activity. W hen u n q u estio n ab le a rtifa cts are fo u n d in som e frequency,
usually in relatively dense (frequency p er u n it space) c o n c e n tra tio n s, the
arch aeo lo g ist declares th a t a site has been found.
T h u s, a p rim ary definitive crite rio n fo r an arch aeo lo g ical co n tex t is the
presence o f a site th e definitive criterio n in tu rn fo r w hich is the presence o f
artifacts, typically ceram ic o r sto n e artifacts. T his results in a n ea r-tau to lo g ical
line o f re aso n in g w hen the arch aeo lo g ical co n tex t o f b o n e item s is used to
assign them an artificial status. B inford (1981 b:4—8), fo r one, has show n th at
the sim ple presence o f bones in a site does n o t necessarily m ean th o se bones
were artificially deposited . T his n o tio n has a m uch d eeper h isto ry w hen the
search fo r w ays to so rt cu ltu rally deposited b o n e from n a tu ra lly dep o sited bone
is recalled (e.g., T h o m as 1971). T he arch aeo lo g ical co n tex t o f bones, then,
ca n n o t be used a lo n e to recognize eith er cu ltu ra lly d ep o sited bones o r h u m an ly
m odified bones. T ypically, co n tex t is only the first clue, o ften im plicitly called
u p o n , a n d suggests th e po ten tia l th a t a p a rtic u la r b o n e w as cu ltu ra lly deposited
an d m odified by hom inid s.
K ind o f bone
In N o rth A m erica in p artic u la r, analysts searching fo r w h at have been called
bone expediency tools h ave used criteria re g ard in g the kin d o f b o n e fro m w hich
a p u rp o rte d to o l is m ade to help identify these objects. O nly bones o f
a p p ro p ria te stru ctu re , w eight, a n d stren g th w ere em ployed as expediency tools.
Y et no form al list o f w hich skeletal elem ents are a p p ro p ria te h as been

p u blish ed . Review o f the skeletal elem ents inferred to have been used as
expediency tools in dicates v irtu ally an y b o n e o f the skeleton m ay be an
expediency too l. T h a t is so because, w hile only one p a rt o f a skeletal elem ent
m igh t p ro v id e a w o rk in g edge, w eight an d b alan ce req u irem en ts can be m et by
using a jo in t co nsistin g o f several a rtic u la te d bones. T h erefo re any b o n e m ay be
a p a rt o f a b o n e expediency to o l (see L ym an 1984b fo r a d d itio n al discussion).
T h e b o n e expediency to o l co n c ep t allow s fo r the possibility th a t p reh isto ric
h u n ters b u tch ered an an im al w ith its ow n bones. T hu s, the b u tch er need only
k n o w how to p ro d u c e a bo n e expediency to o l ra th e r th a n ca rry a b u tch erin g
to o l kit a ro u n d th e landscape. E x p erim en tal w ork suggests, how ever, th a t
while this is certain ly possible, a m ore efficient bone expediency tool w ould be
m ade from a b o n e o f a n an im al th a t died som etim e previous to the b u tch erin g
event (F riso n 1982). T h u s a bo n e expediency to o l m ay be m ade fro m the bones
o f a tax o n different fro m th o se being b u tch ered . A single m odified cam el
(■C am elops sp.) b o n e in a b ison (Bison sp.) kill site, for exam ple (F riso n 1982), is
a likely c a n d id a te fo r an expediency tool. In such cases, the exotic bo n e m ay be
m o re w eath ered th a n the c o m m o n bones. S hipm an an d R ose (1988:308) re p o rt
th a t bo n e flakes ten d to a b so rb grease a n d m o istu re fro m the flesh o f the
b u tch ered an im al, effectively d ulling the edge a n d “ m ak in g it im possible to
c o n tin u e to use th e sam e b o n e flake w ith o u t co n sid erab le sacrifice in
efficiency.” T his m ay in d icate th a t w hen such to o ls are fo u n d they signify a
scarcity o f o th er, m o re efficient to o l m aterials.
M odification attributes
T he kind a n d d istrib u tio n o f m o d ificatio n on skeletal p a rts seem to be the
m a jo r criteria an aly sts use to help identify b ones m ade in to a n d /o r used as tools
by h om inid s. D oes a b ra sio n o cc u r only on the fractu re surface, the distal end o f
the fractu re , o r th e en tire surface o f a specim en? Is p o lish fo u n d only on the
distal end o f the fractu re o r is it fo u n d over the entire fra c tu re surface plus the
ex terio r co rtical surface o f the bone? T he im p licatio n s o f answ ers to these an d
sim ilar q u estio n s fo r identifying bo n e to ols reside in the a ssu m p tio n th a t the
m o d ificatio n will have a d istrib u tio n restricted to the w ork in g edge if it is use-
related w ear a n d a less restricted d istrib u tio n if it is th e result o f n a tu ra l
processes.
Kinds o f m odification
A rchaeo lo gical evidence o f m a n u fa c tu rin g b o n e to o ls can tak e the fo rm o f
ch ip ped fractu re edges, g ro u n d fractu re edges (including striae), a n d the
creatio n o f d etritu s (b o n e flakes). U se-w ear m o d ificatio n is restricted to the
a ttritio n a l loss o f b o n e tissue a n d ca n consist o f polish, ro u n d in g , sm oothing,
a n d m icro flakin g o f fractu re edges a n d surfaces (Jo h n so n 1985:213-217;
B utchering, bone fracturing, and bone tools 345
L ym an 19 8 4 b :3 18). A b ra sio n , polish, a n d ro u n d in g are discussed in detail in

C h a p te r 9, a n d flaking is m en tio n ed in earlier sections o f this c h a p te r. These
m odification s m ay be m acro sco p ically visible (F ig u re 8.13), b u t som e tap h o -
nom ists have arg u ed th a t m icroscopic ex a m in a tio n o f them is im p o rta n t (e.g.,
d 'E rric o et al. 1984; S h ip m an 1988a).
O n th e basis o f th eir stu d y o f m icroscopic featu res o f experim entally used
bon e to o ls a n d e th n o g ra p h ically d o cu m en te d bone tools, S hipm an a n d R ose
(1988:312) re p o rt th a t all utilized edges develop a m acroscopically visible gloss
o r polish. U n d e r m icroscopic ex a m in a tio n , w ear is “ restricted to raised areas o f
the edge o r rugosities th a t actu ally com e in to c o n ta c t w ith the sub stan ce being
cut; these initially raised areas becom e ro u n d e d , glassy [in texture], a n d sm o o th
an d lose surface detail progressively w ith c o n tin u ed u tiliz a tio n ” (S h ip m an an d
R ose 1988:314). T hey suggest th a t the use-w ear o n a bo n e tool m ay show a
p referred axis o f m o tio n b ased o n m icro striae, a n d th a t use-w ear m icro p o lish
sho u ld n o t be unifacially d istrib u te d over the fu n c tio n a l w o rk in g edge o r tip
(S h ip m an a n d R ose 1988:329). N a tu ra l a b ra sio n does n o t seem to p ro d u c e the
fine, glassy polish seen in use-w orn specim ens, b u t to help distinguish artificial
w ear fro m n a tu ra l ab ra sio n the sed im en tary co n tex t o f the specim en sh o u ld be
exam ined (S h ip m an 1988a:282).
Som e n a tu ra l processes can p ro d u c e m odifications to bones th a t m im ic use-
w ear m od ificatio n . O liver (1989) d o cu m en ts w h a t m ay be a co m m o n case w hen
an an im al suffers a n a n te m o rte m bo n e fractu re. Sixty percent (35 o f 58) o f a
sam ple o f biso n (Bison sp.) bones displaying p erio steal reactive bo n e g ro w th o r
p rim ary callus also display polish a t the incom pletely healed fra c tu re edge.
N o tin g th a t th e p erio steal sh ea th shields m u ch o f the co rtical bo n e ad jac en t to
the fractu re from abrasive forces, O liver (1989:87) suggests “ the p olish on these
specim ens w as m ost certainly created by the m ovem ent o f the fractu re d pieces
across o ne a n o th e r a n d the su rro u n d in g tissue in the live an im a l.” Because
p rim ary callus is w oven bo n e th a t is easily ero d ed , b ones b ro k e n in live anim als
by n a tu ra l process m ay leave b ro k e n bones w ith polished fractu re edges in the
fossil record.
Villa (1991:199-200) illu strates w h at she describes as an “ e la b o ra te biface
m ad e o n elep h an t b o n e ” fro m a m iddle P leistocene site in Italy. T his is an 18.7
cm long a n d 8.2 cm wide, bifacially flaked piece th a t in o u tlin e has the
a p p e ara n ce o f an A cheu lian h an d ax e. It is m ore extensively flaked th a n a
specim en illu strated by A g e n b ro ad (1989:142) fro m the late Pleistocene H o t
S prings M a m m o th Site in S o u th D a k o ta th a t w as ap p a re n tly created by
tram p lin g . N o r is the extent a n d frequency o f flake scarring o f the Italian
specim en a p p a re n t in actualistically d o cu m en te d n atu ra lly b ro k e n A frican
elep h a n t (L o xo d o n ta africana) bones illu stra te d by H aynes (1991:148-149). In
th e case o f the Ita lia n specim en, the frequency a n d p a tte rn e d lo catio n - on
m o re th a n one edge a n d m ore th a n one surface - o f the flake scars seem to
in d icate h om in id in ten t o r m a n u fa c tu rin g m odification.
O lsen (1989a) discusses the differences betw een n a tu ra l w ear p a tte rn s on

cervid an tlers a n d cu ltu ra l d am ag e to these b o n y stru ctu re s. A n tler tine flakers
used by h o m in id s to w o rk sto n e tend to have use-w ear a t th e tip in the form o f
blu n tin g. M icroscopically the b lu n tin g results from heavy p ittin g due to
cru shin g o f the tissue; this som etim es pro d u ces a w ear facet th a t is p erp en d icu
lar to th e long axis o f the a n tle r tine. S triae are visible m icroscopically o n the
an tle r tips, an d o ccasionally a very sm all lithic flake is em b edded in the an tler
tip. S ections o f a n tle r beam used as soft h am m ers fo r percussion red u ctio n o f
sto n e to o ls display use-w ear in the form o f “ heavy p ittin g , sets o f fine parallel
striae, a n d slan ting V -shaped cu ts co m p actly d istrib u te d over the lo catio n o f
im p a c t” (O lsen 1989a: 134). N a tu ra l surface m odifications to a n tle r resulting
from cervid b eh av io rs include polish a n d ab rasio n o f sections o f the an tler
beam a n d tines. A b ra sio n a n d polish o n the tines ten d s to extend fu rth e r dow n
the tin e to w ard the m ain beam th a n artificial w ear. R an d o m ly o rien ted shallow
cuts w idely dispersed ov er the tines also m ay be p ro d u c ed by n a tu ra l b ehaviors
o f the cervids.
D istribution o f m odification attributes

A n aly sts typically exam ine the a n a to m ica l d istrib u tio n o f th e v ario u s kinds o f
m o d ificatio n (e.g.. M yers et al. 1980). But o th e r a ttrib u te s as well m ust be
con sid ered because it is becom ing clear th a t certain processes ca n result in
p a tte rn e d a n d restricted d istrib u tio n s o f w h a t a p p e a r to be use-related w ear
m odification s b u t w hich in fact are n o t use-related (e.g., L ym an 1984b; W hite
1992; see C h a p te r 9). F o r exam ple, Jo h n so n (1985:190) suggests th a t use-w orn
bo n e to o ls are d isting uished by ro u n d in g a n d polishing on a localized o r
restricted area o f th e fractu re edge a n d ad jacen t o u te r co rtical a n d in n er
m ed u llary surfaces, w ith no associated w eath erin g o r ca rn iv o re dam age.
F u rth e r, as im plied by S hip m an a n d R o se’s (1988) ex p erim en tal results cited
above. Jo h n so n (1985:191) notes th a t often the use-w ear polish is restricted to
the “ convex fractu re su rface.”
L ym an (1984b:328—329) suggests th a t the analyst exam ine the co n tex tu al
d istrib u tio n o f v ario u s m o d ificatio n a ttrib u te s o f p u rp o rte d b o n e expediency
tools. T his effectively shifts the focus o f analysis from one o n the technological
a n d use h istories o f th e to o ls to o n e o n th eir fu n ctio n al contexts. F o r exam ple,
are b o n e “ ch o p p e rs” used m o re freq u en tly in sites form ed d u rin g cold m o n th s
a n d bo ne “ fleshers” used m ore frequently in sites form ed d u rin g w arm m o n th s
(e.g., F riso n 1982)? T h eo retically fo u n d ed ex p ectatio n s, g ro u n d e d in eth n o a r-
chaeo log ical o b serv atio n s, co n cern in g the kinds o f use-w ear a n d m a n u fa c tu r
ing m o d ificatio ns th a t should be fo u n d given p a rtic u la r kinds o f sites a n d
p a rtic u la r circu m stances o f b u tch erin g (T able 8.3) co u ld help z o o a rc h a e o lo
gists k n o w w h a t kin ds o f m odifications a n d m odified bones to lo o k fo r in
p a rtic u la r cases.
A nalysis o f bone tools
O nce b o n e to ols have been so rted from an assem blage o f fau n al rem ains, their
analysis can proceed m u ch as w ith any ca te g o ry o f a rtifa ct, w h eth er the bone
to o ls are extensively o r only m inim ally m odified by m a n u fa c tu re a n d /o r use.
F o r exam ple, m o d ificatio n resulting from m a n u fa c tu re m ay reveal som ething
a b o u t th e tech n o lo g y used to p ro d u c e the tools. O bviously technological
analysis will be restricted to th o se specim ens th a t have been m ade a n d n o t ju st
used. T h e scap u la specim ens illu strated in F ig u re 8.12 are from a single site
d atin g to a b o u t 1500 B.P. a n d lo cated in eastern W a sh in g to n state. T hey
in dicate th a t o cc u p an ts o f this site used tw o basic technological strategies to
ex tra ct th e p o ste rio r b o rd e r o f scapulae fo r p ro d u c tio n o f w h a t are term ed re
sh ap ed scapu la awls: percussion o r pecking th ro u g h the th in blade o f the
scap u la ju s t a n te rio r to the trian g u lar-in -cro ss-sectio n p o ste rio r b o rd e r, an d
(p ro b ab ly b u rin cut) g ro o v e -an d -sp lin ter th ro u g h the thin scap u la blade.
F u n c tio n a l analysis can exam ine, o f course, the use-w ear attrib u tes.
In a ra th e r in n ovative analysis, S hip m an (1989:322) notes th a t, based on
ex p erim en tal d a ta , utilized b o n e to o ls should display a clear distin ctio n
betw een utilized edges a n d unused edges, a n d th a t (m icroscopic) high po in ts
(convex surfaces) o f utilized edges will be m o re highly polished th a n low p o in ts
(concave areas). She exam ined 116 possible bone to o ls fro m O lduvai G orge,
a n d identified 41 o f th em as utilized to o ls based o n m icroscopic use-w ear
a ttrib u te s. She arg u es th a t fo u r o f the 41 to o ls w ere used as anvils a n d display
“ a series o f p ecu liar p u n ctu re s o r depressed fractu res o n a single, b ro a d , gently
curving, n a tu ra l su rface” (S h ip m an 1989:325). All m ark s she reco rd ed on the
anvils are unlike ca rn iv o re to o th m ark s a n d are fo u n d o n large specim ens. F o r
the rem ain ing 37 too ls, S hip m an (1989:326) fo u n d no statistically significant
ten den cy fo r p a rtic u la r skeletal elem ents to show a p a rtic u la r kind o f use-w ear.
Specim ens display ing use-w ear have three tim es m ore “ flaked fra c tu re s” th a n
bo n e specim ens fro m O lduvai th a t do n o t display use-w ear, a n d use-w orn
bones a p p e a r to have been b ro k e n an d flaked w hen fresh. T he use-w orn bone
to ols seem to have been regularly m ad e from h um eri, scapulae, a n d fem ora,
an d , m o st o f the b o n e to o ls are from large ra th e r th a n sm all m am m als o r n o n
m am m alian tax a (S h ip m an 1989:328).
Bone tools and skeleta l p a rt frequencies

O ne o f th e ta p h o n o m ic effects o f m ak in g a n d using bo n e to o ls is th e influence
such a process has o n frequencies o f skeletal p arts, a q u a n tita tiv e v ariab le th a t
is q u ite im p o rta n t in m o d ern v erte b rate ta p h o n o m y (C h a p te r 7). O n one han d ,
if bones are sufficiently m odified d u rin g the m a n u fa c tu rin g o f a tool, they will
n o t be id entifiable to skeletal elem ent, a n d th u s they will be analytically ab sen t
T able 8.10 M N E frequencies o f bison bones fr o m

the Phillips Ranch site (fro m W hite 1952c)
Skeletal p art M N E unm odified M N E tools T o tal M N E
scapula 12 56 68
P hum erus 10 5 15
D hum erus 20 0 20
P radius 15 0 15
D radius 11 0 11
P ulna 9 0 9
P fem ur 8 1 9
D fem ur 5 1 6
P tibia " 1 8
D tibia 23 0 23
even th o u g h they m ay be lying on the la b o ra to ry table. T h ere is little we can do

a b o u t this type o f bias in tallies o f skeletal p a rt frequencies. O n the o th e r han d ,
som e b ones m ade in to tools are still identifiable to skeletal elem ent. T he
difficulty in analysis here is th a t it m u st be determ in ed w h eth er o r n o t these
to o ls sh o u ld be inclu ded in co u n ts o f skeletal p arts. T h a t is, the an aly st m ust
decide w h eth er these m odified skeletal p a rts ow e their presence in an assem
blage to th eir use as to o ls (w hich m ay have little to do w ith such things as the
utility indices listed in C h a p te r 7), o r if they owe th eir presence to having been
p a rt o f an an im al exp lo ited fo r food. W ith o u t such d eterm in a tio n , analysis o f
skeletal p a rt frequencies can be seriously co m p ro m ised , as the follow ing
exam ple m ak es clear.
Several o f th e late p re h isto ric peoples w ho occupied the m idw estern U nited
S tates p racticed h o rticu ltu re. O ne o f th eir co m m o n ly m ade a n d used to o ls was
a hoe m ad e from the scap u la o f a bison (Bison bison), b u t they utilized o th er
b ones o f the bison as m ateria l for o th e r to o ls as well. T he M N E per skeletal p art
for 10 p a rts recovered from one site a n d re p o rted by W hite (1952c) is given in
T ab le 8.10. F o r this sam ple, conversion o f the M N E values to M A U values is
unnecessary because all included skeletal p a rts are paired bones, th u s all M N E
values w ould be divided by 2 a n d the results w ould th u s n o t change. N o te th a t
for this p a rtic u la r site (Phillips R an ch ), well over h a lf o f th e bison scapulae
recovered h ad been m ad e in to tools, a n d one th ird o f the p ro x im al h um eri h ad
also been m ad e into tools. F ig u re 8.14 presents a p a ir o f sca tte rp lo ts o f the
M N E frequencies p er skeletal p a rt, w ith a n d w ith o u t the to o ls included. T he
differences are re m a rk ab le, the m o st n o ta b le one being th a t w ith o u t the M N E
o f to o ls scapu lae ra n k fo u rth in a b u n d a n c e b u t w ith the M N E o f to o ls scapulae
ra n k first in ab u n d a n ce. If the to o ls are n o t included in the M N E tallies, the
skeletal p a rt frequencies are co rrelated w ith the bison food utility index
(rs= —0.59. P = 0.07) w hereas if the to o ls are included in the M N E tallies the
skeletal p a rt frequencies are n o t co rrelated w ith th a t index (rs = 0.13, P = 0.73).
■o
o
o
li -
MN E without tools
0 20 40
MNE with tools
Figure 8.14. Seatterplot o f M N E frequencies o f selected bison bones ag ain st the
bison food utility index (from T able 8.10). a, bones m odified into tools excluded;
b, bones m odified into tools included. N o te especially the relative position o f the
scapula in the two plots.
In clu d in g the b o n e tools w eakens the co rre la tio n coefficient betw een skeletal
p a rt frequencies a n d the s tru c tu ra l density o f the skeletal p a rts (from
rs = - 0 . 3 6 , P = 0.3 to rs = 0.02, P = 0.95).
Several zo o a rch a eo lo g ist colleagues have suggested th a t the u tilizatio n o f
bones as to o l m aterial m ay well be p ro d u c in g the L -shaped curves like th a t seen
in F ig ure 7.2 an d p erh ap s as well the positive statistical re la tio n sh ip betw een
bon e density an d skeletal p a rt frequencies such as th a t seen in F ig u re 7.12a. I
Age Class (age in African Elephant years)

Figure 8.15. D em o g rap h y o f m o rtality o f m asto d o n carcasses rep o rted by Fisher
(1987). Age is in A frican elephant years.
h ave n o t yet seen an y d a ta , eith er p reh isto ric o r actualistic, to in d icate th a t this
is an ac cu ra te assessm ent o f th e situ atio n , b u t th e p o in t is well tak en . O n this
basis, a n d on the basis o f the d a ta in F ig u re 8.14.1 suggest th a t if it can be show n
th a t p a rtic u la r skeletal p a rts w ere being used as to o ls m o re often th a n o th er
skeletal p arts, then all o f the fo rm er should be om itted fro m study o f skeletal
p a rt frequencies. T h a t w ay, the biasing effects o f p re h isto ric h o m in id s m o d ify
ing som e skeletal p a rts b eyond recognition will be at least p artia lly elim inated,
an d th e ad d itio n a l biasing effects o f p re h isto ric peoples selecting a n d c u ratin g
p a rtic u la r skeletal elem ents fo r use as to o ls will also be at least p artially
elim inated from analyses o f skeletal p a rt frequencies.
Butchering, breakage, and bone tools
O ne o f the finest exam ples o f ta p h o n o m ic analysis th a t considers evidence for

b u tch erin g , fractu rin g o f bones by ho m in id s, a n d the m a n u fa c tu re o f bone
to ols is fo u n d in F ish e r’s (1984a, 1984b, 1987; S hipm an et al. 1984a) study o f
late P leistocene m a sto d o n (M a m m u t am ericanum ) rem ains in N o rth A m erica.
T he 19 sites he exam in ed each h ad th e rem ains o f one individual m a sto d o n , bu t
o f th o se sk eleto ns F ish e r believes ten h ad been b u tch ered by h u m an s a n d the
o th e r nine w ere n o t b u tch ered . H e describes vario u s lines o f evidence to
su p p o rt his in te rp re ta tio n s. T he d em o g ra p h y o f the b u tch ered m a sto d o n s is
different from th a t o f the n o n -b u tch e red m a sto d o n s (F ig u re 8.15). T he seasons
o f d e a th fo r th e tw o sam ples are com pletely different (F ig u re 8.16), suggesting
to F ish er (1987:359) th a t the bu tch ered m a sto d o n s w ere h u n ted because the
□ butchered EH not butchered
Number of
2
I ndi v i dual s
0
Fall Winter S pr i ng Summer
Se a so n of Death
F igure 8.16. Season o f m o rtality o f m asto d o n carcasses rep o rted by F isher

(1987).
seasons o f d e a th o f the bu tch ered individuals do n o t ov erlap w ith the late

w in te r-e a rly sp rin g p a tte rn o f n a tu ra l m o rtality .
T he evidence fo r b u tch erin g consists o f cut m ark s on som e bones, a n d m ost
im pressively, th ere are several bones th a t w ere a rtic u la te d in life w hich display
m ark s on th eir c o n a rtic u la tin g surfaces m ad e by an in tru siv e im plem ent. T he
bones o f in d iv iduals th a t a p p e a r to have been b u tch ered are m o re intensively
a n d extensively b ro k e n th a n the bones o f indiv id u als th a t a p p e a r n o t to have
been b u tch ered . A s well, som e o f the b ones o f the fo rm e r g ro u p are bu rn ed
w hereas no n e o f the b o nes o f the la tte r g ro u p are b u rn ed . Som e o f the b ro k e n
bones display fra c tu re m o rp h o lo g ies suggestive o f b re ak ag e w hile the bones
w ere fresh, a n d several also have percussion m ark s ad jac en t to the fractu re
edge. T h ere are som e c a rn iv o re gnaw ing m ark s o n bones o f the bu tch ered
g ro u p b u t n o t o n bon es o f th e n o n -b u tc h e re d g ro u p . W hile there are no stone
tools associated w ith the skeletons, there are several b o n e specim ens associated
w ith the b u tch ery -m ark e d carcasses w hich display po lish suggestive o f th eir use
as expediency to o ls, a n d several specim ens have h a d m ultiple flakes rem oved
from th eir m argins (F ish er 1984a, 1984b, 1987; S hip m an et al. 1984a).
W h a t is p erh ap s m o st im pressive a b o u t F ish e r’s analysis is the detailed
d o c u m e n ta tio n o f m u ltiple attrib u te s, all suggested o n the basis o f actualistic
research to be exclusively o r nearly exclusively p ro d u c e d by hom inids. T he
presence o f only b u tch erin g m ark s, o r only green b o n e fractu re s, o r only
b u rn e d bo nes, o r only a couple o f specim ens w ith po lish restricted to one area,
w ould n o t p ro v id e as com pelling a n a rg u m e n t fo r h u m a n in terv e n tio n as the
m u ltip le lines o f evidence m arsh alled by F isher. In c o n ju n ctio n w ith the
d em o g rap h ic d a ta , the b o n e m o d ificatio n d a ta allow him to b uild a convincing
arg u m en t. T he a ttrib u te s re p o rte d by F ish e r are n o t u n am b ig u o u s sig n atu re
criteria diag n o stic o f only h o m in id activity, fo r exceptions to each individual
a ttrib u te ca n be fo u n d . B one ca n be b u rn e d , fo r exam ple, w ith o u t h u m a n
activity. W h a t is co n vincing a b o u t the b u tch ery m ark s is th eir p u rposefulness
(especially th o se on tigh tly b o u n d jo in ts a n d fo u n d on c o n a rtic u la r surfaces)

an d th eir lack o f prim itiveness. T he p ro b a b ility th a t the c o m b in a tio n o f
a ttrib u te s F ish er describes could have been p ro d u c ed by n a tu ra l processes
seems m in u te indeed.
A n o th e r reason F ish e r's (1984a, 1984b, 1987) arg u m e n ts are so com pelling is
p erh ap s less obvious. It concerns the fact th a t he is dealing w ith faunal
collections th a t have relatively sim ple ta p h o n o m ic histories, in p a rt because
they each represen t a single m o rta lity event a n d a single individual anim al.
S au n d ers’ (1977) collection o f the rem ains o f som e 30 m a sto d o n s fro m a single
site h ad u n d erg o n e v ario u s ta p h o n o m ic processes (such as tram p lin g ) th a t
F ish er's in div idu al anim als h ad not. F ish e r’s single-carcass assem blages m ade
it m uch easier fo r him to establish the interd ep en d en ce o f skeletal p a rts (they all
cam e from the sam e in d ividual anim al). In B in fo rd ’s (1980, 1981b) te rm in o
logy, F ish er is dealing w ith “ fin e-g rain ed " assem blages w hereas S aunders is
dealing w ith a “ p alim p sest” o r “ c o a rse-g rain ed ” assem blage (see the G lo ss
ary). T his sim ple fact u n d ersco res the n o tio n s th a t som e ta p h o n o m ic problem s
will be m uch easier to solve if the sam ple a t h an d has experienced a tap h o n o m ic
h isto ry th a t does n o t o bscure ta p h o n o m ic traces critical to p ro b lem solution,
an d th a t ta p h o n o m ic h istories are cum ulative. A lso, som etim es tap h o n o m ic
pro b lem s will only be solved w ith great difficulty, only be p artia lly solved, o r
n o t be solved at all.
Summary
H un ting m an alone ad ap ts som e p arts o f the carcase w hich he is him self unable to
eat to o th er ends, serviceable to his living, his com fort, his vanity o r his whim.
(I. W. C ornw all 1968:88)
In this c h a p te r I have focused largely on h o m in id -related b io stratin o m ic

processes; C h a p te r 9 focuses on n a tu ra l b io stratin o m ic processes. T he reason
for this k ind o f se p a ra tio n in topics is th a t I suspect m an y w ho read this book
will be arch aeo lo g ists a n d zo o arch aeo lo g ists, an d they will have a stro n g
in terest in h o m in id m odifications o f fau n al rem ains. H om inids variously
b u tch er an im als, a n d they m ay b reak bones d u rin g b u tch erin g o r d u rin g the
process o f m ak in g b o n e tools. It is a p p ro p ria te then, to conclude this c h a p te r
w ith a re ite ratio n o f som e o f the issues involved in establishing h o m in id s as
ta p h o n o m ic agents th a t affected a p a rtic u la r b o n e assem blage.
H aynes a n d S ta n fo rd (1984:217) suggest there are three levels o f “ evidence"
req u ired fo r inferences o f the h o m in id u tilizatio n o f an im al carcasses. F irst, the
contem poraneity o f h o m in id s w ith the tax a m u st be established. W ere the
an im als rep resen ted alive an d in the geo g rap h ic area w hen the h u m an s were
there? S tra tig ra p h ic m ixing m ay result in an archaeological asso ciatio n o f
an im al a n d h u m a n rem ains, a n d m u st be considered a t this level. A nd, o f
course, the an aly st sh o u ld keep o p en the possibility th a t fossils m ay have been
picked u p a n d used by p re h isto ric peoples. T he second level o f evidence

involves estab lishin g the association o f h u m a n s w ith the an im al rem ains. A re
th ere m od ification s to th e rem ains th a t only h o m in id s could have p roduced?
T hese include such th in gs as b u tch erin g m ark s. T he th ird level o f evidence
involves asc ertain in g w h eth er o r n o t the h o m in id s utilized th e anim al. D id they
b u tch er a n d eat th e m eat, o r did they sim ply m ake bo n e tools? O bviously,
different kinds o f evidence are req u ired a t each level, b u t as well the three levels
are n o t m u tu ally exclusive. F ish e r's (1984a, 1984b, 1987; S hip m an et al. 1984a)
studies o n th e late P leistocene u tilizatio n o f m a sto d o n (M a m m u t am ericanum )
carcasses by N o rth A m erican P aleo -In d ia n s w a rra n t careful scrutiny an d , in
my o p in io n , e m u latio n if the an a ly st is in terested in establishing th a t a hom inid
ta p h o n o m ic ag en t w as p a rt o f the ta p h o n o m ic h isto ry o f a bone assem blage.
But even then, a n d p erh ap s especially th en , there are n u m ero u s n a tu ra l
b io stratin o m ic processes th a t sh o u ld be considered in analysis. It is the n a tu ra l
aspects o f b io stra tin o m y th a t we tu rn to in the next ch a p te r.
9
OTHER BIOSTRATINOMIC FACTORS
Introduction
M an y ta p h o n o m ic processes m ay affect an im al carcasses a n d bones betw een

the tim e o f an im al d e a th and b u rial o f the carcass o r bones. Several o f the m ajo r
h u m an b io stratin o m ic facto rs are discussed in C h a p te r 8. In this c h a p te r o th er
b io stratin o m ic facto rs, several o f w hich are n a tu ra l processes, are reviewed.
D iscussion is lim ited to th ose facto rs th a t have been m ore o r less extensively
d ealt w ith in the literatu re. As well, several basic co m p arativ e analytic
tech niq ues are described at the end o f the ch ap ter.
Weathering
The degree o f brittleness o f the skeleton gives no in fo rm atio n as to age; the n atu re o f
the place w here it is fo u n d m ust be taken in to account. T he m ore the bones are
exposed to air, the m ore quickly they disintegrate. T he q u an tity o f p recipitation, the
nu m b er o f days below freezing, covering with clay, burial in sand o r loam - all these
factors play an im p o rta n t role in forensic medicine.
(J. W eigelt 1927/1989:18)
B ehrensm eyer (1978:153) defines the w eath erin g o f bo n e as “ the process by

w hich the original m icroscopic o rg an ic an d in o rg an ic co m p o n en ts o f b o n e are
sep a rated from each o th e r a n d destro y ed by physical a n d chem ical agents
o p eratin g on the bo n e in situ , eith er on the surface o r w ithin the soil z o n e .”
W e ath erin g involves th e d eco m p o sitio n an d d estru c tio n o f bones “ as p a rt o f
the n o rm al process o f n u trie n t recycling in a n d on soils” (B ehrensm eyer
1978:150). M iller (1975:217) in dicates w eath erin g refers to “ the effects on bone
o f s a tu ra tio n , d esiccation, a n d te m p e ra tu re ch an g es.” S aunders (1977:104)
reco rd s “ w eath erin g indices” th a t are “ subjective in te rp re ta tio n s o f the
co n d itio n s o f at least th ree sm all a n a to m ica l crests o r ru g o sities” p er specim en.
H e lists fo u r kinds o f w eathering ("u n w e a th e re d , slight, m o d era te, a n d severe” )
bu t describes none o f them (S au n d ers 1977:108). B ehrensm eyer (1978:161)
defines six stages o f w eath erin g o f b ones o f m am m als > 5 kg in b o d y w eight in
su b aerial/su rface co n tex ts, a n d uses readily observed, m acroscopic criteria to
distin g u ish those stages an d to en h an ce th eir reco g n itio n in the field (T able 9.1.
F ig u re 9.1). T o reco rd w eath erin g d a ta , the an aly st records the m axim um
w eatherin g d isplayed over p atch es larger th a n 1 c m 2 on each bone specim en,
an d “ w henever possible shafts o f lim b bones, flat surfaces o f jaw s, pelves.
354
9.1 W eathering stages in large ( after Behrensm eyer 1978) and sm a ll ( a fter Andrew s 1990) m am m als. A dditi
ge m a m m a l descriptions fr o m Johnson (1985) in parentheses
LARG E M AM M ALS SM A LL M A M M A LS
ering R ange in years R ange in

D escription since death D escription since dea
greasy, no cracking o r flaking, p erh ap s with 0-1 no m odification 0-2

skin o r ligam ent/soft tissue attach ed
(m arrow edible, bone still m oist)
cracking parallel to fiber stru ctu re (longitudinal); 0-3 slight splitting o f bone parallel to fiber 1-5
a rticu lar surfaces p erh ap s w ith m osaic cracking o f structure; chipping o f teeth and splitting
covering tissue an d bone (split lines begin to o f dentine
form , low m oisture, m arro w sours an d is inedible)
flaking o f o u ter surface (exfoliation), cracks are 2 -6 m ore extensive splitting but little 3-5 +
present, crack edge is an g u lar flaking; chipping and splitting o f
(m arrow decays, split lines well developed) teeth leading to loss o f parts o f crow n
rough hom ogeneously altered co m p act bone 4-15 deep splitting and som e loss o f deep 4 -5 +
resulting in fibrous texture; w eathering segm ents o r “ flakes” between splits;
penetrates 1-1.5 m m m axim um ; crack edges are extensive splitting o f teeth
rounded
coarsely fibrous an d rough surface; splinters 6-15
o f bone loose on surface, w ith w eathering
penetrating in n er cavities; open cracks
bone falling a p a rt in situ, large splinters present, 6-15
bone m aterial very fragile
Figure 9.1. Bone w eathering stages described by B ehrensm eyer (1978). a,

w eathering stage 1 (lower) on tw o deer (Odocoiteus sp.) fem ora show ing split line
cracks on b o th specimens; b, detail o f w eathering stage 1 (u p p er specim en in a)
show ing m osaic cracking; c, w eathering stage 2 on a horse (Equus caballus)
O ther biostratinom ic fa c to r s 357
C ap tio n for fig. 9.1. (cont.).

m etacarpal show ing flaking o f o u ter surface and initial stages o f exfoliation; d,
w eathering stages 3 an d 4 o n a horse fem ur, show ing extensive exfoliation (near
ends) and deep, m ultiple cracks (m id-shaft).
v erteb rae, o r ribs are used, n o t edges o f bones o r areas w here th ere is evidence
o f physical d a m a g e ” (B ehrensm eyer 1978:152).
A n drew s (1990:10-11) sum m arizes ex perim ental results o f w eathering o f
sm all m am m al bones, a n d co m p ares them to B ehrensm eyer's (1978) fo r large
m am m als (T able 9.1). H e n o tes th a t bones em bedded in a n owl pellet are
p ro tec ted fro m w eathering. T ee th show som e splitting a fte r tw o years o f
exposure, “ p ro b a b ly d u e to the differential c o n tra c tio n o f enam el a n d d e n tin e ”
(A ndrew s 1990:11). Splits an d crack s develop betw een collagen fibers, bones o f
the skull sep a rate alo n g su tures, a n d teeth fall o u t o f th eir alveoli as w eathering
progresses. C hem ical w eath erin g begins on the b o n e surface an d progresses
in to th e bo n e tissue m ass (B rom age 1984).
B ehrensm eyer’s (1 9 7 8:161)hope w as th a t once sufficient c o n tro l studies w ere
available, bo n e w eath erin g featu res m ig h t “ give specific in fo rm a tio n co n c ern
ing surface expo sure o f b o n e p rio r to b u rial a n d the tim e perio d s over w hich
bones a c c u m u la te d .” W hile she ca u tio n e d th a t h er d a ta w ere p relim in ary an d
th eir significance w as co n jectu ral, an aly sts have used them to infer the
fo rm a tio n a l h isto ry o f bone assem blages (e.g., B unn a n d K roll 1987; P o tts
1986, 1988). B one w eath erin g is an im p o rta n t b io stratin o m ic a n d tap h o n o m ic
variab le, a n d I d ev o te som e tim e to review ing it.
W eathering and tim e
T he w eath erin g o f bo n e is a h isto rical process. T he c a p ita liza tio n o f “ T im e”

signifies th e passage o f so lar years. I d istin g u ish this kind o f tem p o ral
m easu rem en t from “ tim e” o r “ ta p h o n o m ic tim e .” In the latter, tim e is
m easu red on an o rd in a l scale; th a t is, we are able to say p h e n o m e n o n A is older
th a n p h en o m en o n B b u t we are u n ab le to say how m any years o lder A is th a n B.
T h e goal o f analysis o f bo n e w eath erin g d a ta is m easu rem en t o f ta p h o n o m ic
tim e, o r Tim e.
T he stage o f w eath ering displayed by a b o n e m easures, in p a rt, the ra te and
the d u ra tio n o f w eathering. By rate I m ean how quickly a b o n e passes th ro u g h
the w eath erin g stages. A t least th ree facto rs co n tro l the w eath erin g rate. F irst,
“ sm all, co m p ac t b ones such as p o d ials a n d ph alan g es w eath er m o re slowly
th a n o th e r elem ents o f the sam e sk eleto n ” (B ehrensm eyer 1978:152). Second,
bones o f different tax a, especially th o se o f different b o d y size, w e ath er at
different rates (B ehrensm eyer 1978, 1982; G ifford 1981). T h ird , “ the less
eq u ab le the im m ed iate e n v iro n m e n t (in term s o f te m p e ra tu re a n d m o istu re
flu ctu atio n s) o f th e bo ne, the fa ste r it sh o u ld w e a th e r” (B ehrensm eyer
1978:156). T he duration o f w eath erin g concerns the span o f T im e over w hich a
b o n e is exposed to w eath erin g agents. E x p o su re to th o se agents begins once
soft tissues d etac h fro m the bones, a n d th u s the m an n er in w hich hide, m uscle
m asses, etc. are rem oved c o n tro ls w hen exposure begins. B ecause bones m ay,
by definition, w eath er in su b su rface co n tex ts, identifying the te m p o ra l end o f
exposu re is difficult w hen it d epends o n the degree to w hich a bone is w eathered.
O ne critical aspect o f analyzing a n d in terp re tin g bo n e w eath erin g d a ta

involves th e co n v ersio n o f th a t d a ta in to a t least ta p h o n o m ic tim e if n o t Tim e.
B ehrensm eyer (1978:157) accom plished th a t co n v ersio n because she h ad
know ledge o f th e years since d e a th o f the a n im al c o n trib u tin g the w eathered
bones. She fo u n d th a t the w eath erin g stage displayed by the m o st w eathered
b o n e o f a c arcass’ skeleto n w as related to th e years since d eath . T he c o rrelatio n
betw een each categ o ry o f w eath erin g stage a n d the years since d e a th re p re
sented in h er sam ple is stro n g a n d significant (rs = 0.802, P = 0.002). T h a t
c o rrelatio n p ro m p te d B ehrensm eyer’s suggestion th a t bo n e w eath erin g d a ta
m ay reveal insights to th e d u ra tio n o f p re h isto ric b o n e assem blage fo rm a tio n .
T he c o rre la tio n B ehrensm eyer (1978) fo u n d can be expressed w ith the
eq u atio n :
WS = f(YD ) [9.1]
w here W S is th e m ax im u m w eath erin g stage observed fo r an an im al carcass,

a n d Y D is th e years since th a t an im al died. L y m an a n d F o x (1989:312) argue
th a t w hen th e an a ly st is dealing w ith an assem blage o f b ones consisting o f
m u ltip le in d iv id u al anim als recovered fro m a single stra tu m , in o rd e r to
in te rp re t b o n e w eath erin g d a ta validly the e q u a tio n th a t m u st be solved is:
YD„ E D ,, AH,j = fTWS.j, SE,r T X ir ME,,) [9.2]
in w hich Y D a n d W S are defined as in e q u a tio n [9.1], E D is the exposure

d u ra tio n , A H is the ac cu m u latio n h isto ry , SE is th e skeletal elem ent, T X is the
ta x o n rep resen ted, M E is the d ep o sitio n al m icro e n v iro n m e n t, i is th e individual
carcass th a t c o n trib u te d the bone, a n d j is the p a rtic u la r b o n e o f carcass i th a t is
u n d er study.
E q u a tio n [9.2] c a n n o t in fact be solved fo r several reasons. It is, how ever, a
useful heuristic th a t highlights the v ariables w hich influence the ra te and
d u ra tio n o f w eath ering. T he variables on the rig h t-h an d side o f e q u a tio n [9.2]
a n d the su b scrip ts u n d ersco re the fact th a t p re h isto ric bone assem blages
typically c o n ta in specim ens o f vario u s skeletal elem ents from m ultiple c a r
casses rep resen tin g m u ltip le tax a, often surficially d ep o sited across different
m icro en v iro n m en ts a t different tim es. V ariables on the left-h an d side o f
eq u a tio n [9.2] are related to the passage o f tim e an d un d ersco re the fact th at
different carcasses will have died a t different tim es relative to one a n o th e r, th a t
different b o nes (b o th w ithin a n d betw een carcasses) will have different
expo sure d u ra tio n s, a n d th a t different b o n es (b o th w ithin a n d betw een
carcasses) m ay h av e been ac cu m u lated at different tim es. T he an a ly st m ay
re co rd th e w eath erin g stage (W S) observed for, say, only left fem o ra o f one
tax o n , th ereb y analytically co n tro llin g SE a n d T X (L ym an a n d F o x 1989). T he
o th e r fo u r v ariables, how ever, c a n n o t be analytically co n tro lled in an y strict
sense. F o r exam ple, if one desires to in fer the T im e over w hich a bone
assem blage a c cu m u lated , one m u st assum e th a t all the b ones w ere accu m u lated
w hen they w ere fresh o r in w eath erin g stage 0. W hile th a t m ay well have been
the case, th ere are no actu alistic d a ta to su b sta n tia te th a t a ssu m p tio n a n d in
fact lim ited d a ta in d icate it m ay n o t be w a rra n te d . T o e la b o ra te an d clarify the
in terp retiv e issues raised by e q u a tio n [9.2], 1 discuss each variable in th a t
e q u a tio n in som e d etail, beginning w ith th o se variables on the rig h t side o f the
equ als sign.
W eathering stage
B ehrensm eyer’s (1978) w eath erin g stages (T able 9.1) each represent a p o in t in
T im e alo n g th e co n tin u o u s process o f bo n e d e te rio ra tio n . B ehrensm eyer
(1978:152-153) states, fo r exam ple, th a t “ th e six w eath erin g stages im pose
a rb itra ry divisions u p o n w h a t w as observed to be a co n tin u o u s sp e c tru m .”
Jo h n so n (1985:187), in fact, defines six w eath erin g “ p h ase s” sp an n in g B ehrens
m eyer’s first th ree w eath erin g stages. G ifford (1977:291) re p o rts a sam ple o f
b ones w hich prog ressed th ro u g h w eath erin g stages 0 an d 1 to stage 2 w ithin one
to tw o years, b u t w hich then rem ain ed in stages 2 a n d 3 fo r several years, an d
suggests th a t we m ay be u nable to assign “ ab so lu te T im e values to different
w eath erin g stag es” (G iffo rd 1981:418). T h a t o b serv atio n , plus B ehrensm eyer’s
(1978:157) statem en t th a t “ w eath erin g stages are m ost useful in pro v id in g an
estim ate o f th e m in im u m n u m b e r o f years since d ea th (o r ex p o su re),” indicates
th a t th e w eath erin g stages are a t best a n o rd in al scale m easu rem en t o f time;
they do n o t clearly m easu re Tim e.
B ones w eath er, by d efinition, in b o th surface a n d su b su rface contexts
(B ehrensm eyer 1978; F riso n a n d T o d d 1986). G iven an interest in only the
w eath erin g th a t tak es place on the surface o f the g ro u n d , it is su rp risin g th a t so
little effort has been m ade to d istinguish su b aerial (surface) from subsurface
w eath erin g (e.g., F riso n an d T o d d 1986; M ehl 1966; T o d d 1987b: T o d d et al.
1987). W hile b u ried b o nes seem to w e ath er m uch slow er th a n exposed bones,
the fo rm er do som etim es w eather. T hus, until a reliable way to distinguish
su b aerial a n d su b su rface w eath erin g is developed, w hen w eath erin g d a ta are
in terp re ted it m u st be w ith the a ssu m p tio n th a t su b su rface w eath erin g is
insignificant.
G iffo rd -G o n z alez (1989a: 192) re p o rts d am ag e to bo n e specim ens th a t
superficially resem bles w eath erin g cracks b u t w hich she believes w as created by
h eatin g the b ones, such as w hen they are exposed to a fire. S hipm an (1981 b: 177)
states th a t h ea tin g b o n e results in the “ d e n a tu rin g ” o f collagen fibers w hich are
n a tu ra lly “ u n d e r ten sio n in the b o n e;” the im p licatio n is th a t the ten sio n in
these fibers th a t is released by h eatin g causes cracks to form , cracks w hich are
" p e rp e n d ic u la r to the directio n o f the collagen fibers a n d the long axis o f the
b o n e ” (see th e discussion o f “ B u rn in g ” below ). T his o b serv atio n n o t only
u n d ersco res the care th a t m u st acco m p an y th e identification o f a w eathering
stage displayed by a bo n e, b u t the fact th a t w eathering, as defined, is sim ply one
fo rm o f b o n e d eterio ratio n .
O ther biostratinom ic fa c to rs 361
S ke le ta l elem ent
B ehrensm eyer (1978) an d T o d d et al. (1987:68-70) suggest th a t different
skeletal elem ents will w eath er at different rates, p e rh a p s d u e to v aria tio n in
th eir s tru c tu ra l density (L y m an a n d F o x 1989:297; see C h a p te r 7 fo r discussion
o f s tru c tu ra l density). W e sim ply do n o t k now , how ever, w hich skeletal
elem ents w e ath er fa st a n d w hich ones w e ath er m ore slowly. T o c o n tro l this
v ariab le, th en , one co u ld re co rd the m ax im u m w eath erin g stage displayed by
one k in d o f skeletal elem ent, say, only th e h u m eri o r fem ora, to c o n tro l fo r
p o te n tia l v a ria tio n in th e ra te a t w hich different skeletal elem ents p ass th ro u g h
th e w eath erin g stages.
Taxon
B ehrensm eyer’s (1978:153) w eath erin g stages w ere defined on the basis o f
m am m als w ith > 5 kg b ody w eight. G iffo rd (1981:417) observes “ bones o f
ro u g h ly like-sized m am m als o f different ta x a m ay w e ath er at som ew hat
different rates due to c o n stru c tio n a l differences,” an d in h er co n tro l sam ples
“ m o re heavily co n stru c te d equid bones w eath ered a t a so m ew h at slow er ra te
th a n h o m o lo g o u s bovid b o n e s.” S chafer (1962/1972:24) re p o rts fences an d
to m b sto n e s co n stru c te d o f w hale b o n e last fo r centuries in H o llan d . A gain,
p erh ap s th e stru c tu ra l density a n d /o r the p o ro sity o f skeletal elem ents
influences the w eatherin g rate. T o co n tro l fo r p o te n tia l tax o n o m ic v a ria tio n in
b o n e w eath ering, one sh ould re co rd the w eath erin g stage displayed by the
rem ains o f each ta x o n sep arately (as well as each skeletal elem ent separately).
T his w ould pro v id e in d ep en d e n t sam ples o f d a ta o n the sam e p h en o m en o n .
D epositional m icroenvironm ent

T h e m icro e n v iro n m e n t in w hich bones are d ep o sited is discussed a t tw o spatial
scales by B ehrensm eyer (1978). A t a large scale one considers the general
v eg etatio n al h a b ita t o f the d ep o sitio n al area, a n d a t a fine scale one considers
th e specifics o f the m icro e n v iro n m e n t at th e lo catio n w here a b o n e is d eposited.
T he fo rm er co ncerns assem blages o f bones as these o cc u r over large spatial
u n its w h ereas single b on es each occupy a p a rtic u la r sp atial p o in t. It is relevant
to co n sid er each scale o f d ep o sitio n al m icro en v iro n m en t.
Vegetation habitat in an area B ehrensm eyer (1978) presen ts d a ta o n 1,534

carcasses in six v eg etatio n al h a b ita ts, a n d re p o rts th e m o st ad v an ced w e a th e r
ing stage per carcass. She fo u n d all six w eathering stages rep resen ted in all six
h a b ita ts, b u t the freq uency d istrib u tio n s o f re p re se n ta tio n o f w eath erin g stages
differ betw een som e o f the h a b ita ts. L ym an a n d F o x (1989:298) calculated
K o lm o g o ro v -S m irn o v tw o-sam ple D statistics betw een all possible p airs o f
hab itat-specific w eath erin g d a ta (T able 9.2). W hile th o se statistics indicate
som e in te r-h a b ita t v a ria tio n in th e frequency d istrib u tio n s, B ehrensm eyer
T able 9.2 K olm ogo rov-S m irnov D statistics between all possible pairs o f
carcass assem blages fr o m m ajor habitats, based on data in B ehrensm eyer
(1 9 7 8 ). N u m b er in parentheses n e x t to D statistic is the w eathering stage
where D occurs (fro m L ym a n and F ox 1989)
D ense w oodland £> = 0 .1 4 (1 )

P < 0 .0 1
O pen w oodland £> = 0.16 (1) £> = 0.04 (3)
£><0.01 £■>0.1
Bush £> = 0.35 (1) £> = 0.21 (1) £> = 0.1 9 (1 )
/><0.01 £'< 0.01 £><0.01
Plains £> = 0.23 (1) £> = 0 .0 9 (1 ) D = 0.07 (2) £> = 0.12 (1)
£><0.01 />>0.1 />>0.1 0.1 > £ > 0 .0 5
Lake bed £> = 0 .26(1) £> = 0 .1 2 (1 ) £> = 0.1 0 (1 ) £> = 0.09 (1) £> = 0.06 (3)
(N = 213) />< 0.01 £>= 0.05 />>0.1 £ > 0 .1 £ > 0 .1
Sw am p D ense w oodland O pen Bush Plains
(N = 322) (N = 312) w oodland (N = 184) (N = 248)
(N = 255)
(1978:159) ca u tio n s th a t som e o f th a t v aria tio n m ight also be a ttrib u ta b le to

“ changes in p a tte rn s o f h a b ita t utilizatio n a n d /o r m o rta lity ” o f the anim als in
th ose h a b itats. T his m ean s th a t the ac cu m u latio n h isto ry o f b ones m ay vary
betw een h ab itats.
M icroenvironm ent o f a spatial point: B ehrensm eyer (1978:158) argues th a t the

“ localized co n d itio n s (e.g., veg etatio n , shade, m o istu re) are m o re im p o rta n t to
bo ne w eath erin g th a n overall characteristics o f the [different vegetation]
h a b ita ts .” T he su b aerial d ep o sitio n al m icro e n v iro n m e n t o f a b o n e m ay in h ib it
o r ex acerb ate the ra te o f w eathering. B ehrensm eyer’s d a ta indicate th a t the size
and in tern a l density o f v egetation p atches in w hich bones are deposited are
im p o rta n t v ariables, a n d o th e r d a ta in d icate th a t the m ag n itu d e o f seasonal
changes in w eath er an d d u ra tio n s o f the seasons are also im p o rta n t variables,
as are th e m oisture c o n ten t, te m p eratu re, a n d tex tu re o f the sedim ent on w hich
a b o n e is d ep o sited (B rain 1967a; C o o k 1986; M iller 1975). T o d d (1983a,
1983b) presents d a ta ind icatin g th a t the presence o f m ultiple, closely spaced
carcasses tends to create a different m icro en v iro n m en t th a n w hen a carcass is
isolated from o th e r carcasses. F u rth e r, in a late Pleistocene cluster o f m a sto d o n
(M a m m u t am ericanum ) bones S aunders (1977:104) fo u n d th a t “ th e bones
occurring at the p erip h ery a n d at the to p o f the bone bed w ere generally poorly
preserved in d icatin g severe w eath erin g p rio r to final burial. [As well] the o u ter
lim it o f the bo n e bed w as defined by b ones a n d tusks in generally p o o r
c o n d itio n .” L ym an (1988a: 103) a n d L am (1992:401) d e m o n stra te th a t bones
dep o sited in a ro ck sh elter o r cave are less w eath ered th a n b ones deposited
ou tsid e o f the ro ck sh elter o r cave; little else w as different betw een the
d ep o sitio n al en v iro n m en ts o f these specim ens.
G ra n tin g th at m icro e n v iro n m e n ta l c o n d itio n s m ay influence the rate o f

w eath erin g , we still d o no t k now the sp atial scale at w hich th a t m icro e n v iro n
m en tal v a ria tio n will create significant v a ria tio n in w eath erin g rates; th a t is, we
do n o t know how far a p a rt tw o bones m u st be (assum ing they are o f the sam e
elem ent a n d o f th e sam e tax o n ) in h o riz o n ta l space to ensure th a t they occu r in
d ep o sitio n al m icro en v iro n m en ts th a t are sufficiently different to cause one o f
them to w eath er significantly fa ste r th a n the o th er. T his p ro b lem seem s to be
dealt w ith analytically by assum ing th a t all skeletal specim ens w ithin an
assem blage w ere d ep o sited in the sam e d ep o sitio n al m icro en v iro n m en t, o r at
least in one w ith v a ria tio n th a t insignificantly influenced the ra te o f w eathering.
T his m ay be re aso n ab le as m o st assem blages are (analytically) specified as
co m in g fro m som e h o rizo n ta lly a n d vertically d elim ited sp atial u n it, typically a
s tra tu m o r som e o th e r d ep o sitio n al unit. H ow ever, assum ing th a t the m icro en
v iro n m en t did n o t v ary across th a t sp atial unit p ro b a b ly becom es m ore
ten u o u s as its size increases.
Years since death

A b o n e begins to w e ath er afte r the an im al co n trib u tin g th a t bone dies an d the
bo n e is freed fro m the soft tissues in w hich it is em bedded. T hus, b o n e 1 should
be m o re w eath ered th a n b o n e 2 if b o n e 1 “ d ie d ” before b o n e 2 a n d b o th were
exposed w ithin, say, a few days o f d e a th a n d all o th e r variables in e q u a tio n [9.2]
are eq u al fo r the tw o bones. In tra -c a rc a ss v a ria tio n in th e w eath erin g stages
displayed by b o nes is a ttrib u ta b le to v a ria tio n in w hen th o se bones w ere
exposed (freed from soft tissues). T h u s the stro n g est c o rre la tio n betw een the
years since an im al d e a th a n d w eath erin g stage sh o u ld be fo u n d using th e m o st
ad v an ced w eath ering stage displayed by the bones o f a carcass, w hich is
precisely th e basis fo r the c o rrelatio n betw een w eath erin g stage an d years since
d e a th described by B ehrensm eyer (1978).
Y ears since a n im al d e a th can only v ary betw een ind iv id u al anim als; it c a n n o t
v ary b etw een the b o n es o f a n anim al. T h a t re n d ers years since an im al d e a th a
significantly different kind o f ta p h o n o m ic tim e th a n exposure d u ra tio n and
ac cu m u latio n h isto ry (discussed below ). T o infer the years since an im al d eath
the an a ly st m u st d istin g u ish w hich bones m ak e u p a p a rtic u la r carcass; th a t is,
the in terd ep en d en ce o f th e b ones m u st be d eterm in ed an d each set o f bones
fro m each carcass rep resen ted in a n assem blage m u st be so rted fro m the
com p lete collection. H esse a n d W a p n ish (1985:88) n o te th a t “ since different
b on es are likely to w e ath er a t different rates, w eath erin g stage d a ta c a n n o t be
used fo r g ro u p in g specim ens o f th e sam e sk eleto n .” As well, such a p ro ced u re
w ould certain ly be tau to lo g ical. A n a to m ica l refitting (C h a p te r 5) m ay be the
m o st p ro m isin g tech n iq u e fo r c o n tro llin g th e in terd ep en d en ce o f skeletal p arts,
b u t as yet we lack sufficient c o n tro l d a ta to allow its ap p lica tio n . T he analyst
m ig h t c o n tro l specim en in terd ep en d en ce a n d th u s m easu re v a ria tio n in the
years since d e a th o f th e an im als rep resen ted by re co rd in g th e m axim um
w eath erin g stage displayed by specim ens o f a single categ o ry o f skeletal

elem ent, such as left fem ora, realizing th a t the left fem u r fro m a carcass m ay
d isplay a w eatherin g stage different from th a t displayed by the rig h t fe m u r o f
th a t carcass (see below an d L ym an a n d F o x 1989:299-300 fo r a d d itio n al
discussion).
E xposure duration
B ecause expo su re begins w hen in su latin g soft tissues are rem oved fro m bones,
tw o b ones w hich die a t th e sam e tim e b u t w hich vary in w hen they are exposed
m ay o r m ay n o t experience sim ilar exposure d u ra tio n s even if b o th are fro m the
sam e carcass. Sim ilarly, tw o bones w hich die a t the sam e tim e a n d w hich are
exposed sim u ltan eo u sly m ay n o t experience sim ilar ex posure d u ra tio n s if they
are bu ried at different tim es. T he influence o f exposure d u ra tio n as influenced
by v a ria tio n in w hen b o n es are first exposed is clear in a sam ple o f 20 p airs o f
left a n d rig h t do m estic cow (Bos taurus) fem ora described by T o d d (1983b). All
p airs died a t th e sam e tim e, all w ere in th e sam e d ep o sitio n al m icro e n v iro n
m ent, a n d all w ere in su b aerial co n tex t w hen w eathering d a ta w ere reco rd ed . O f
the 20 p airs only eight d isplayed th e sam e w eath erin g stage; the m em bers o f
eight o th e r p airs differed by a w eathering stage o f one, the m em bers o f three
p airs differed by tw o stages, a n d th e m em bers o f one p a ir differed by three
w eath erin g stages. F u rth e r, all six w eath erin g stages w ere rep resen ted by the 40
to ta l fem o ra even th o u g h all ind iv id u al an im als died a t the sam e tim e.
E x p o su re d u ra tio n can v ary betw een the bones o f one carcass as well as
betw een the bones o f m u ltip le carcasses because ex posure d u ra tio n concerns
in d iv id u al bones. E x p o su re d u ra tio n m easu red as T im e (years) will alw ays be
less th a n o r equal to th e years since d ea th , a n d th u s it is a different kin d o f
ta p h o n o m ic tim e th a n years since d eath . T his is especially so w hen factors
influencing an o rg a n ism ’s m o rta lity (e.g., agent o f d ea th ) are in d ep en d en t o f
facto rs influencing ex po sure o f a n o rg a n ism ’s b ones (e.g., scavengers). I f an
an aly st w ishes to in fer years since d e a th from a n a n im a l’s w eathered bones,
they m u st assum e th a t a ca rca ss’ b ones w ere exposed im m ediately afte r d e a th in
o rd e r th a t th e m axim ally w eath ered bo n e o f a carcass be tightly co rrelated w ith
years since d eath ; in term s o f e q u a tio n [9.2], E D m u st equal Y D .
V a ria tio n in the expo sure o f bones o f a single carcass can be illu stra te d by
ex am in atio n o f th e w eathering profile displayed by a ca rca ss’ bones th ro u g h
tim e. A weathering profile is “ the percen tag e frequencies o f bo n e specim ens in
an assem blage disp lay in g each w eath erin g stag e” (L ym an a n d F ox 1989:300).
Such profiles often serve as the focus o f in terp re tin g w eathered bones (e.g.,
Boaz 1982; P o tts 1986, 1988). W hen they are c o n stru c te d for com plete
assem blages, they o ften consist o f variously in terd e p en d en t (bones from sam e
carcass) an d in d ep en d en t bones (those from different carcasses). T here are no
w eath erin g d a ta fo r individual carcasses th a t allow us to exam ine how the
bones o f a carcass pass th ro u g h th e w eath erin g stages in the fo rm o f a
CD
0 1 2 3 4
Weathering Stage
Figure 9.2. W eathering profiles for carcasses dead 0.5 to 1 yr, 2.5 to 3 yr. 4 to 10
yr, and 10 to 15 yr (after G ifford 1977, 1984).
w eath erin g profile. G iffo rd 's (1977, 1984) d a ta for sets o f m ultiple carcasses,
each set having a k n o w n years since d ea th , are p lo tte d as w eath erin g profiles in
F igure 9.2. If all b ones o f a new ly d ead carcass display w eath erin g stage 0, an d
all b o n es o f a long d ead carcass display w eath erin g stage 5 (this kind o f
w eath erin g profile will p ro b a b ly be ra re as bones in this stage tend to
d isin teg rate to d u st), th e w eath erin g profiles in F ig u re 9.2 suggest a “ w ave
m o d el” (L y m an a n d F o x 1989:300) for the w eath erin g profile o f the bones o f a
carcass th ro u g h tim e. T he bones o f the carcass w eath er progressively, b u t each
b o n e w eath ers a t a slightly different ra te , a n d each experiences a slightly
different ex p osure h istory. T he ideal m odel th u s takes the form o f a u n im o d al
w ave w hich m oves across the w eath erin g stages fro m left (w eath erin g stage 0)
to rig h t (w eathering stage 5). D ev iatio n s fro m th e ideal m odel are expected, o f
course, because b o nes o f a carcass will experience v aried ex posure histories,
will w eath er at different rates, a n d will p ro b a b ly be differentially dispersed and
d ep o sited in different m icro en v iro n m en ts.
A ccum ulation history

H ere we are p a rtic u la rly co n cern ed w ith active ac cu m u latio n o f b o n es (see
C h a p te r 6). T h e w ave m odel o f b o n e w eath erin g fo r a single carcass (F ig u re 9.2)
m akes it clear th a t b on es collected from a carcass d ead only a few h o u rs an d
th a t are d ep o sited in a site a n d b u ried sh o rtly th e re a fte r will have a w eathering
profile different from th a t displayed by a carcass d ead a n d exposed for a year-
an d -a-h alf. T h u s, in te rp re tin g the d u ra tio n over w hich a bo n e assem blage
ac cu m u lated d em an d s th a t th e an a ly st assum e all b ones w ere ac cu m u lated

w hen fresh (w eath ering stage 0). Is th a t a reaso n ab le assu m p tio n ?
1 am u n aw are o f any study focusing on the w eath erin g stage displayed by
b o nes collected by in d ividual b o n e-acc u m u la tin g agents. A n e cd o tal d a ta
in d icate b ones in w eath erin g stage 0 are typically collected by ca rn iv o res and
scavengers, bu t th o se d a ta also indicate bones in w eathering stage 1 are
o ccasion ally collected a n d b ones displaying w eathering stage 2 are som etim es,
alth o u g h rarely, collected by such bo n e ac cu m u lato rs (L ym an a n d Fox
1989:303). B rain (1980) indicates A frican p o rcu p in es p re fer bones th a t are in
w eath erin g stage 1 an d also ap p a re n tly collect bones in w eathering stage 2 (see
C h a p te r 6). F riso n (1982) suggests h u m an s m ay have collected b ones in
w eath erin g stage 1 ra th e r th a n use fresh bones (in w eathering stage 0) to
m a n u fa c tu re bone tools. F o r the present, it m ay be re aso n ab le to assum e th a t
the m ajo rity o f the bones used to co n stru c t a w eath erin g profile w ere
ac cu m u lated w hen fresh (displayed w eath erin g stage 0), b u t the an aly st should
be aw are th a t som e o f th e b ones m ay h av e been so m ew h at w eathered w hen
accu m u lated . N o te th a t B ehrensm eyer (1990:233) re p o rts th a t “ w eathered
bones th a t have lost th eir o rg an ic m aterial an d becom e cracked are less
b u o y a n t” in w ater, a n d th u s less susceptible to fluvial tra n s p o rt th a n fresh,
g reen bone. T his is p ro b a b ly because bo n e sh rin k s as it dries, th ere b y reducing
its p o ro sity a n d increasing its bu lk density. T h u s a w eathering profile c o n
stru cted fro m fluvially ac cu m u lated bones w ould have am b ig u o u s significance
for in ferrin g the d u ra tio n o f ac cu m u latio n .
B rain (1981:115-116) p erfo rm ed an experim ent in w hich he placed the b ones
o f a pig (S u s sp.) in a su b aerial co n tex t a n d let them w eath er for seven years.
Som e o f the bones w ere placed in the open a n d th u s w ere exposed to direct
su nlight, o th e r b ones w ere placed in the shade. B rain (1981:116) re p o rts th a t
even a fte r seven years the difference in w eathering stage is rem ark ab le; the
m an d ib le placed in the sh ade “ re tain ed en o u g h fa t afte r seven years fo r dust to
be ad h e rin g to it [was w eathering stage 0], T he d efattin g process o f the fully
exposed m an d ib le was com plete w ithin one year, a n d [perhaps] w ithin three
m o n th s. In a sh ad ed situ atio n , d efattin g m ay tak e d ec ad es.” T his o b serv atio n
p ro m p te d Yellen (1991 b) to w orry th a t such fat-rich bones o n a b a n d o n e d sites
m ight be exp loited by scavenging carnivores. T hese o b serv atio n s in d icate how
th e w eath erin g stage displayed by a bo n e m ay n o t be related to th e years since
an im al d e a th , b u t it also indicates th a t b ones th a t have been d ead fo r som e tim e
m ay be ac cu m u lated by carnivores.
L y m an a n d F o x (1989:304-307) explore som e o f the effects o f an ac cu m u
latio n h isto ry involving the collection o f b ones in varied w eathering stages and
suggest th a t as th e w eath erin g stage increases (fro m 0 to 5) a t the tim e o f bone
ac cu m u latio n , the resulting w eath erin g profile will suggest lo n g er a c cu m u
latio n histories, lo n g er ex posure d u ra tio n s, a n d /o r m o re years since anim al
d eath . A c cu m u la tio n h isto ry is th u s yet a n o th e r kind o f ta p h o n o m ic tim e,
d istin ct fro m ex po sure d u ra tio n a n d years since anim al d eath .
W eathering an d spatial data

B ehrensm eyer (1978) suggests th a t the an a ly st m ay be able to begin to co n tro l
fo r v a ria tio n in the a c cu m u latio n h isto ry a n d d ep o sitio n al m icro en v iro n m en t
o f in d iv id u al w eathered b ones by stu d y in g th eir sp atial d istrib u tio n s an d
association s. She suggests th a t (1) if bones in an assem blage display all
w eath erin g stages a n d are “ hom o g en eo u sly m ixed in a single d e p o sit” the
assem blage p ro b a b ly rep resents a lo n g -term ac cu m u latio n , a n d (2) if only one
w eath erin g stage is displayed by all o f the bones an d those bones are spatially
clu stered , then th e assem blage m ay re p resen t a sh o rt-te rm single ac cu m u latio n
event o r, if the d ep o sitio n al m icro e n v iro n m e n t varied betw een the individual
w eath ered bones, the clu ster m ay re p resen t a lo n g -term series o f ac cu m u latio n
events (B ehrensm eyer 1978:161).
B ones are d istrib u te d w ithin th ree-d im en sio n al space (h o riz o n ta l a n d verti
cal). T his in tu rn d em an d s th a t the an aly st d eterm in e if the bones being studied
com e fro m a single, stratig ra p h ic ally defined surface (there w as m inim al
v a ria tio n in vertical lo catio n ), o r if they com e fro m a th ree-d im en sio n al
s tra tu m (there w as v a ria tio n in b o th the h o riz o n ta l a n d the vertical lo catio n s o f
bones). A s well, the an aly st sh o u ld co n sid er the sp atial scale o f the d istrib u tio n .
W ere th e b o nes scattered over an a rea 10 x 10 m , o r 100 x 100 m horizontally?
B ut as n o te d above, we sim ply d o n o t yet kn o w how m u ch sp atial v a ria tio n in
the lo catio n s o f in d iv id u al b ones, eith er h o riz o n ta l o r vertical, m ig h t signify the
p o ten tial fo r different d ep o sitio n al m icro en v iro n m en ts. T h u s n o t only are the
tw o sp atial areas ju s t suggested so m ew h at a rb itra ry , they are, fo r in terp retiv e
p u rp o ses, p ro b a b ly am b ig u o u s (see L ym an a n d F o x 1989:309-311 fo r a d d i
tio n al discussion).
T h ere seem to be tw o w ays to rem ove som e o f the arb itra rin e ss o f the spatial
units. T h e first sh o u ld be o b vious to an arch aeo lo g ist. B ecause fau n al
assem blages tend to be spatially defined, the o bvious sp atial unit to sta rt w ith is
a d ep o sitio n al stra tu m . Stein (1987:340) no tes th a t geologists define a deposit as
rep resen tin g “ one d ep o sitio n al event [but] the d u ra tio n o f such a d ep o sitio n al
event is n o t often know n. A single d ep o sit m ay re p resen t eith er co n tin u o u s o r
a b ru p t d ep o sitio n o v er eith er long o r sh o rt perio d s o f tim e .” I suggest,
n on etheless, th a t because a d ep o sitio n al s tra tu m (o r p e rh a p s an archaeological
featu re) rep resen ts a single geological event o f d ep o sitio n , it w ould seem to be a
logical w ay to define a n assem blage o f fossils fro m w hich a w eath erin g profile
m ay be derived.
T h e m ean in g o f a w eath erin g profile derived from a single d ep o sitio n al unit
m ay be d eterm in ed by co n sid erin g a second w ay to a p p ro a c h spatial d a ta . T he
second a p p ro a c h is exem plified by B ow er et al. (1985). T hey c o n stru c t
w eath erin g profiles fo r tw o M iddle to L ate S tone A ge bo n e assem blages from
E ast A frica an d exam ine th e relatio n betw een b o n e size an d severity o f
w eath erin g . F o r th e w eath erin g profile labeled “ E ” in F ig u re 9.3 B ow er et al.
(1985) n o te th a t m o st b ones displaying the m o st ad v an ced w eath erin g stages
Weathering Stage
Figure 9.3. W eathering profiles for tw o assem blages o f bones described by Bower
et al. (1985).
are larg e r th a n less w eath ered bones, a n d th o se heavily w eath ered b ones are all
o f the sam e ta x o n a n d p ro b a b ly from the sam e individual (an instance o f
tax o n o m ic refitting). T hey also n o te th a t w eath erin g profile “ E ” could be
in terp re ted one o f tw o ways: the heavily w eathered large bones h ad been lying
o n th e surface lo n g er th a n th e less w eath ere d b ones o f sm aller anim als due
eith er to (a) the g re ater sed im en tatio n necessary to b u ry the large b ones, o r (b) a
te m p o ra l d isasso ciatio n betw een th e ac cu m u latio n o f the large b ones (accu m u
lated first an d th u s exposed longer) a n d the sm all bones (accu m u lated
som etim e afte r the large bones). T hey co n clu d e th a t “ it is im possible to select
the m o re likely o f these tw o ex p la n a tio n s” (B ow er et al. 1985:51), b u t th eir use
o f b o n e size d a ta lead th em to a m o re d etailed c o n sid e ra tio n o f the accum ula-
tio n al h isto ry o f th e b o n e assem blage th a n is possible using only the w eath erin g
stages displayed by th e bones.
B ow er et al. (1985) also exam ine th e re la tio n o f bo n e size to how w eath ered a
bo n e is in th eir d iscussion o f th e w eath erin g profile labeled “ D ” in F ig u re 9.3.
T hey n o te th a t all sizes o f b ones are equally d istrib u te d across all w eath erin g
stages in profile “ D ” an d in te rp re t this profile as ind icatin g m o re o r less
sim u ltan eo u s d ep o sitio n o f all b ones in th a t assem blage. F inally, they suggest
th a t th e b ones in w eath erin g profile “ D ” w ere exposed lo n g er th a n th o se in
w eath ering profile “ E ” even th o u g h the bones exhibiting w eath erin g stages 2
an d 3 in profile “ E ” m ak e u p only 22% o f the to ta l w hereas bones exhibiting
w eath ering stages 2 a n d 3 in profile “ D ” m ak e up 44% o f the to tal. T he basis for
this co n clusion resides in the fact th a t som e large b ones an d som e sm all bones
in w eath erin g profile “ D ” display all w eath erin g stages, so o n average a bo n e in
profile “ D ” m u st h av e been exposed to w eath erin g lo n g er th a n an average bo n e
in profile “ E .”
W h e th e r o r n o t B ow er et al. (1985) have offered co rrec t in te rp re ta tio n s is

irrelev an t (for exam ple, th eir inferences d em an d the a ssu m p tio n th a t all bones
w ere ac cu m u lated w hen they displayed w eathering stage 0). W h a t is o f interest
is th eir use o f bo n e size as a w ay to co n sid er the vertical dim en sio n o f d ep o sitio n
a n d ex p o su re h istory. As well, they a tte m p t to c o n tro l specim en in te rd e p e n
dence. It is p e rh a p s in dicative o f the absence o f sufficient in terp retiv e analogs
th a t B ow er et al. (1985) c a n n o t choose betw een altern ativ e in te rp re ta tio n s.
C o n sid e ra tio n o f the v ariables influencing the d u ra tio n an d ra te o f bone
w eath erin g (defined in e q u a tio n [9.2]) m akes it clear th a t there are also various
m eth o d o lo g ical w eaknesses in how th o se v ariables m ight be analytically
c o n tro lled w hen th e an aly st seeks to in te rp re t w eath erin g profiles. A m ore
d etailed exam ple m ak es these p o in ts clear.
A nalysis o f weathering data
P o tts (1986, 1988:48-56) provides the m o st d etailed analysis a n d m o st ex ten

sive in te rp re ta tio n o f bo n e w eath erin g d a ta yet available. D etailed review o f his
stu d y pro vid es im p o rta n t insights to the difficulties o f analyzing a n d in te rp re t
ing such d ata. M y review is co u ch ed w ithin the discussion p resen ted above, an d
while P o tts ’ in te rp re ta tio n m ay well be co rrect, there are reasons to suspect it
m ay n o t be co rrect.
P o tts (1986, 1988) presents d a ta fo r w eathered bo n e recovered fro m six Plio-
Pleistocene sites at O lduvai G orge. D a ta fo r the F L K Z injanthropus site are
su m m arized in F ig u re 9.4, alo n g w ith d a ta P o tts presents fo r bones recovered
from a hyen a d en a n d the sam ple o f carcasses scattered across the landscape
re p o rted by B ehrensm eyer (1978). P o tts (1988:51) n o tes th a t the la tte r tw o
sam ples re p resen t “ a ttritio n a l, possible g ra d u a l, accu m u latio n s o f bones ra th e r
th a n o ne m ajo r, sh o rt-term pulse o f b o n e a c c u m u la tio n .” S im ilarities betw een
th e tw o co n tro l sam ples a n d the F L K Zinjanthropus assem blage p ro m p t P o tts
(1988:51) to suggest th a t “ the tim e sp an rep resen ted by an a c cu m u latio n o f
b ones m ay be in ferred from the d istrib u tio n o f w eath erin g stag es” a n d th a t the
F L K Z injanthropus assem blage p ro b a b ly represents “ a t least a 4-year p erio d
[and perhaps] at least a 5- to 10-year p e rio d ” (P o tts 1988:54). T he frequency
d istrib u tio n s show n by the three assem blages are n o t statistically significantly
different (K o lm o g o ro v -S m irn o v tw o-sam ple D statistics are: Zinj to L a n d
scape, D = 0.33, P = 0.28; Zinj to H yena den, D = 0.17, P = 0.39; L an d scap e to
H yena d en, D = 0.33, P = 0.28). P o tts (1988:55) notes th a t the F L K Z injanthro
p us b o n e assem blage o ccu rred “ as a layer w ithin a th in , paleosol h o rizo n a b o u t
as thick as the b ones them selves,” a n d th ere w as n o p a tte rn e d v a ria tio n in
w eath erin g stages show n by bones in different h o riz o n ta l contexts. H e in te r
p rets this to in d icate th a t differential b u ria l (E D in e q u a tio n [9.2]) has n o t
influenced the bo n e w eathering p attern s.
B unn a n d K ro ll (1987:97) re p o rt th a t w ithin the F L K Zinjanthropus
assem blage they fo u n d “ co n jo in in g specim ens o f the sam e original bo n e th a t
0 1 2 3 4 5
Weathering Stage
Figure 9.4. Frequency d istrib u tio n o f percentages o f bones per w eathering stage
in three assem blages o f bones (after P o tts 1988:52-53, Figure 3.1; courtesy o f the
a u th o r and A ldine de G ruyter).
exhibit different w eath erin g stag es.” B unn a n d K ro ll (1987:97) th u s argue,

c o n tra ry to P o tts (1986,1988), th a t the w eath erin g d a ta m easu re b u rial tim e, o r
E D in e q u a tio n [9.2], ra th e r th a n the d u ra tio n o f tim e over w hich bones
ac cu m u lated . G iven this d eb ate, it is im p o rta n t to exam ine fu rth e r the d a ta
P o tts (1982, 1986, 1988) provides for the O lduvai bone assem blages. P otts
(1986, 1988) a tte m p ts to ascertain the tem p o ra l d u ra tio n over w hich six
O ld u v ai b o n e assem blages ac cu m u lated (T able 9.3). H e concludes th a t the
b ones m ak in g u p th e three assem blages fo u n d in th in stra tig ra p h ic layers
ac cu m u lated over a 5 to 10 y ear interval w hereas the bo nes m ak in g up the three
assem blages fo u n d in relatively thick stra tig ra p h ic layers ac cu m u lated over a
m inim um o f 5 to 10 years. P o tts (1986:30) suggests th a t all o f the O lduvai bone
assem blages re p resen t sites used over lo n g er periods o f tim e th a n sites used by
m o d ern h u n te r-g a th e re rs w ho “ ten d to reoccupy cam psites tran sien tly fo r no
lon g er th a n several m o n th s b u t only afte r several m o n th s o f n o n -o c c u p a tio n .”
G iven th a t all six o f the assem blages consist o f bones o f m ultiple carcasses an d
n u m ero u s tax a, it is logical to su p p o se th a t m ore th a n one b o n e ac cu m u latio n
event is rep resen ted if it is assum ed th a t early h o m in id s w ould typically
accu m u late little m o re th a n one carcass o r a p o rtio n o f one carcass (an d thus
one tax o n ) p er d ay. B ut this does n o t itself signify the te m p o ra l d u ra tio n o f the
fo rm a tio n o f the b o n e assem blages.
P o tts (1986:29) re p o rts th a t th ere is no co rre la tio n o f the sp atial d istrib u tio n
o f b ones a n d the w eath erin g stages they display, a n d th u s he suggests in tra
assem blage v ariab ility in w eath erin g “ reflects th e length o f tim e b ones were
exposed o n th e lan d su rfa ce.” T o infer the tem p o ra l d u ra tio n o f ac cu m u latio n
T ab le 9.3 Frequencies o f w eathered bones in s ix assem blages fr o m Olduvai

Gorge (fro m P o tts 1 982:99-100). Frequencies are listed as N I S P ( % o f total
N IS P )
W eathering stage
Site (deposit thickness, cm) 0 1 2 3 4 5
F L K N o rth -6 (50)
all bones 403 (76) 77(14) 20 (4) 3 2(6) 0 0
long bones only 13 (27) 18 (38) 2 (4 ) 15(31) 0 0
F L K N N L /2 (24)
all bones 105(46) 59 (26) 24(10) 39(17) 2 (1 ) 1 (0.5)
long bones only 9 (1 2 ) 27 (36) 12(16) 23 (31) 2 (3 ) 1 (1)
DK L/2 (68)
all bones 208 (52) 84 (21) 3 5 (9 ) 74(18) 0 0
long bones only 17(17) 20 (20) 16(16) 45 (46) 0 0
F L K "Z in j" (9)
all bones 771 (76) 147(14) 63 (6) 3 6(4) 0 1(0)
long bones only 66 (4 4 ) 43 (29) 22 (1 5 ) 18 (12) 0 1 (1)
F L K N N L/3 (9)
all bones 188(64) 62 (21) 26 (9) 19(6) 1 (0.5) 0
long bones only 11 (32) 9 (2 6 ) 6 (1 8 ) 7 (2 1 ) 1 (3) 0
D K L/3 (9)
all bones 281 (60) 86 (1 8 ) 71 (15) 29 (6) 0 5 (1 )
long bones only 29 (34) 24 (28) 17(20) 15(18) 0 0
o f th e bo nes, P o tts n o tes th e m o st ad v an ced w eath erin g stage in each

assem blage, an d th en suggests th a t the m in im u m n u m b e r o f years necessary to
a tta in th a t w eath erin g stage b ased on B ehrensm eyer’s (1978) d a ta (T able 9.1)
pro vid es a n estim ate o f the m in im al d u ra tio n o f b o n e ac cu m u latio n . T his o f
co u rse presu m es a m o re o r less co n tin u o u s a n d o n g o in g a c cu m u latio n o f
carcasses a n d bones d isplaying w eath erin g stage 0 over the tim e req u ired for
th e first ac cu m u lated b ones to reach w eath erin g stage 3 o r 4. A single o r several
tem p o ra lly c o n tig u o u s ac cu m u latio n events over a single year co u ld pro d u ce,
fo r exam ple, th e F L K Z injanthropus w eath erin g profile (F igure 9.4), as
B ehrensm eyer (1978) n o ted , given sufficient v a ria tio n in the exposure histories
o f the bones.
G iffo rd ’s (1977, 1984) d a ta on the w eath erin g stage exhibited by bones for
w hich the years since d e a th are k n o w n (F ig u re 9.2) are p lo tte d o n a three-pole
g ra p h , alo n g w ith th e six O lduvai assem blages, in F ig u re 9.5. G iffo rd ’s co n tro l
assem blages fo r carcasses d ead 4 to 10 years an d th o se d ead 10 to 15 years all
a p p e a r to be m u ch m o re w eath ered th a n the com plete assem blages o f O lduvai
bones. P erh a p s this is because G iffo rd ’s assem blages all derive fro m surface
co n tex ts w hereas all o f the O lduvai assem blages h a d been bu ried fo r n early tw o
m illion years. T his leads us to a c o n sid e ra tio n o f th e geological co n tex t o f the
O ldu v ai specim ens.
P o tts (1982:102-103. 1986:30) suggests the bone assem blages fo u n d in thin
Weathering Stage
3 - 5 ( 1 0 0 * )
Figure 9.5. T hree-pole g rap h o f bone w eathering d a ta for six assem blages from
O lduvai G orge (after P o tts 1982) an d G ifford's (1977, 1984) con tro l assem blages
o f carcasses dead for know n num bers o f years.
dep osits ac cu m u lated over a sh o rte r tim e p erio d th a n the bo n e assem blages
recovered from th ick d eposits. T he basis fo r th a t suggestion resides in p a rt in
P o tts ' (1986:30) belief th a t “ b ones lying higher in the [thick] deposits m ay have
sta rte d w eath erin g years later th a n [bones] b u ried [m ore deeply].” W hile this
presum es th e sam e ra te o f d ep o sitio n fo r th in a n d thick deposits (e.g., 1 cm per
decade), it really only suggests th a t th ick er depo sits to o k longer to accu m u late
th a n th in d ep o sits because the fo rm er are th ick er th a n the latter; it is thus
in d ep en d en t o f th e w eath erin g d a ta . It does not ac co u n t fo r v aria tio n betw een
the w eath erin g profiles displayed by the assem blages o f bones in thin a n d thick
d ep o sits. T h a t v a ria tio n is clear in F ig u re 9.6a w here it is show n th a t, if all
bones are included, th ere are p ro p o rtio n a te ly m ore heavily w eathered bones in
th e su m m ed thick d ep o sit assem blages th a n in the sum m ed thin deposit
assem blages (this p a tte rn holds if each assem blage is considered individually).
A K o lm o g o ro v -S m irn o v tw o-sam ple D statistic betw een the tw o is w eak b u t
significant (D = 0.078, P = 0.05), a n d in d icates th a t the tw o are statistically
different.
D ifferences betw een the sum m ed thin d ep o sit assem blages an d the sum m ed
thick d ep o sit assem blages are en h an ced if we follow P o tts (1986:23, 1988:53)
a n d o m it all b u t the lim b bone d iaphyses (F ig u re 9.6b; D = 0.217. P < 0.01). In
elim inatin g all b u t th e lim b b ones P o tts is follow ing B ehrensm eyer’s (1978:152)
suggestion th a t sm all, n o n-lim b bones w eath er m ore slow ly th a n lim bs. In so
Weathering Stage
Weathering Stage
Figure 9.6. C um ulative percent frequency d istrib u tio n s fo r w eathering stages o f
bones in sum m ed assem blages o f O lduvai G o rg e thin deposit sites an d sum m ed
assem blages o f O lduvai G orge thick deposit sites, (a) all bones included; (b) limb
bones only included.
d o in g P o tts (1986:29) ho pes to “ rem ove the bias to w a rd w eathering stage 0 by

exam inin g only the identified long bones w ith d iap h y ses.7' T h a t this om ission
p ro c ed u re in fact p ro d u c es w eath erin g profiles th a t a p p e a r to have been
w eath ered for lon ger tim es is clear in F ig u re 9.5. T here, th e co m p lete O lduvai
assem blages o f b o nes are to the low er right o f the g ra p h w hereas the lim b-bone
only O lduvai assem blages are to the u p p er right. T he q u estio n begged by
o m ittin g all b u t lim b b ones is w h eth er the om ission results in w eathering
profiles th a t a p p e a r to re p resen t longer a c cu m u latio n tim es sim ply because

slo w er-w eathering bon es o f the sam e carcass are o m itted , o r fo r som e o th er
reason.
R ecalling th a t G iffo rd (1981, cited above) no tes equid bones w e ath er slow er
th a n bovid bones, I used P o tts ’ (1982) d a ta to d eterm in e th a t in the th in dep o sit
assem blages, bovid specim ens o u tn u m b e r eq u id specim ens 16.7 to 1 w hereas in
thick d ep o sit assem blages bovid specim ens o u tn u m b e r equid specim ens 28.3 to
1. T h erefo re the th in d ep o sit assem blages sh o u ld a p p e a r less w eathered th a n
the thick d ep o sit assem blages sim ply because equid bones are relatively m ore
a b u n d a n t in the th in d ep o sits th a n in the thick d ep osits. F u rth e r, the n et change
in w eath erin g profiles fo r thick assem blages is g re ater th a n the change in
w eath erin g profiles fo r thick deposits w hen n o n -lim b bones are om itted
because, on average, m o re b ones w ere o m itted from the thin d eposits th a n from
th e thick d eposits, a n d this en h an ces the a p p a re n t difference betw een the tw o
sets o f assem blages (co m p are F igures 9.6a a n d 9.6b). T hese co n sid eratio n s
suggest th a t v a ria tio n betw een thick an d th in d ep o sit assem blages in term s o f
tax a rep resen ted in each (T X in e q u a tio n [9.2]) a n d in term s o f the bones
included in each (SE in e q u a tio n [9.2]) m ay ac co u n t fo r a t least som e o f the
perceived v a ria tio n in th e tw o sets o f w eath erin g profiles, ra th e r th a n solely
v a ria tio n in a c cu m u latio n (o r exposure) histo ries o f the tw o sets.
A final co m m en t co n cern s P o tts ’ (1986:30) suggestion th a t his inferred
“ lo n g -term a c cu m u latio n o f b ones a t each O lduvai site c o n tra sts w ith ”
h isto rically d o cu m en te d p a tte rn s o f site use by h u n ter-g ath ere rs. A s yet, we do
n o t have sufficient eth n o arch ae o lo g ic al d a ta o n (a) the d u ra tio n o f o c c u p atio n
of, (b) the h isto ry o f re o ccu p a tio n of, a n d (c) th e w eath erin g profiles o f bones
ac cu m u lated in cam p s a n d villages utilized by m o d ern h u n ter-g ath ere rs. W hile
m y discussion m akes it a b u n d a n tly clear th a t I am skeptical o f P o tts ’ (1982,
1986, 1988) conclusio ns re g ard in g the b o n e-w eath erin g d a ta he presents, his
analysis is extrem ely im p o rta n t because it helps un d ersco re vario u s w ays to
stren g th en inferences b u ilt u p o n such d a ta . T o co n clu d e o u r co n sid eratio n o f
bo ne w eathering, I now describe o th e r kinds o f d a ta th a t should be recorded in
c o n ju n ctio n w ith n o tin g the m ax im u m w eath erin g stage d isplayed by a skeletal
specim en.
Discussion
B ehren sm ey er’s (1978) sem inal w ork on bo n e w eath erin g reveals th a t w e ath er
ing is a process, a n d th u s it reflects the passage o f T im e. H e r actu alistic d ata
in d icate th ere is a stro n g co rre la tio n betw een the g re atest w eath erin g stage
displayed by th e b ones o f a carcass a n d the n u m b e r o f years since th e anim al
died. H ow ever, as e q u a tio n [9.2] m akes clear, the re la tio n o f a w eathering
profile derived fro m a fossil assem blage a n d th e passage o f T im e is obscure
because o f th e several kinds o f ta p h o n o m ic tim e th a t are included. A ctualistic
research is necessary to allow us to d isentangle the effects o f these variables on

o u r inferences.
B ehrensm eyer (1978:152-153) suggests th a t an aly sts sh o u ld reco rd the m o st
ad v an ced w eath erin g stage over areas g re ater th a n 1 c m 2 on each bone
specim en. L y m an a n d F o x (1989:314) suggest such a p ro c ed u re m ay m ask
precisely th o se d a ta m o st relev an t to assessing the v ario u s aspects o f the
fo rm a tio n a l h isto ry o f a fossil assem blage. F o r exam ple, S aunders (1977:104)
re p o rts th a t “ in all instances w here the effects o f ex posure w ere a p p a re n t, the
exposed surfaces (sky w ard o rie n ta tio n ) o f th e bo n e exhibited g re ater m odifica
tio n th a n the unexpo sed surfaces” in a collection o f late P leistocene m a m m a
lian rem ain s he studied. B ehrensm eyer (1978:153) notes th a t “ bones are usually
w eath ered m o re o n th e u p p e r (exposed) th a n the low er (g ro u n d co n tac t)
surfaces.” T h u s, F riso n an d T o d d (1986:39) suggest th a t bones subjected to
m in im al p o st-d ep o sitio n al a n d p o st-b u rial m o v em en t m ig h t n o t be unifo rm ly
w eath ered o n all surfaces (see also B ehrensm eyer 1981:611). I f w eath erin g d a ta
are to be an aly zed , then, w hen bones are collected the w eath erin g stage
displayed by the u p p e rm o st surface o f each b o n e sh o u ld be recorded alo n g w ith
th e w eath erin g stage d isplayed by the lo w erm o st surface. I f a b o n e displays a
b ro a d ran g e o f w eath erin g stages over v ario u s o f its surfaces, it m ay have
u n d erg o n e slow b u rial, o r possibly, m u ltip le exposures a n d burials. O ne m ay
ev entually reco rd a tran se ct o f w eath erin g stages parallel to the long axis o f a
bon e th a t deviates fro m the h o riz o n ta l to help estim ate sed im en tatio n and
b u rial rates. In c o n ju n ctio n w ith o th e r evidence we now tu rn to, bone
w eath erin g m ay ultim ately prove to be a highly significant source o f ta p h o n o
mic d a ta (see also C h a p te r 11).
R oot etching
T he ro o ts o f m an y p lan ts excrete hum ic acid, a n d o ften “ d en d ritic p a tte rn s o f
shallow g ro o v es” o n b o n e surfaces “ are in terp re ted as the resu lt o f d isso lu tio n
by acids asso ciated w ith the g ro w th a n d decay o f ro o ts o r fungus in direct
c o n ta c t w ith b o n e su rfaces” (B ehrensm eyer 1978:154). M o rlan (1980:56-57)
a n d G ra y so n (1988:30) indicate, how ever, th a t the etching m ay be caused by
acids secreted by fungi associated w ith deco m p o sin g p lants. “ F u n g i can no t
only d eco m p o se o rg an ic m a tte r u n d er relatively dry co n d itio n s, b u t can also
p ro d u ce a w ide variety o f o rg an ic acids d u rin g the p ro cess” (G ray so n 1988:30).
P erh a p s it is a trivial d istin ctio n w h eth er the ro o ts them selves o r the fungi
asso ciated w ith d eco m p o sin g ro o ts secrete th e acids th a t etch bones, b u t for
reaso n s n o ted below , this m ay n o t be a trivial distin ctio n . In the follow ing, I
refer to th e etch in g as “ ro o t etch in g ” sim ply fo r convenience. F u rth e r, I
co n sid er ro o t etching here u n d e r b io stra tin o m y because while such etching can
occu r afte r b ones have been b u ried , som e m osses a n d lichens grow on bones
p rio r to b u rial a n d can result in p re -b u rial ro o t etching.
Figure 9.7. R o o t etching o n a sheep m andible. R eproduced w ith perm ission

from : B inford, L. R. Bones: ancient men and modern m yths, Figure 3.07. New
Y ork: A cadem ic Press, Inc. C o p yright 1981 by A cadem ic Press, Inc.
T he w avy, “ d e n d ritic ” (M o rla n 1980:56), “ sin u o u s” (A ndrew s an d C o o k

1985:685), “ sp ag h etti-lik e” (H esse an d W apnish 1985:85) p a tte rn s o f the
in div id ual ro o ts in c o n ta c t w ith the b o n e are etched in to the b o n e surface
(F ig u re 9.7). E ach ro o tle t lies in a g roove “ w hich p resu m ab ly w as form ed by its
ex u d ates o r by th e m icro o rg an ism s associated w ith ro o tlet m etab o lism ”
(M o rla n 1980:57). S tain in g o f th e etched g ro o v e som etim es is n o different in
co lo r fro m the un etch ed surface (M o rla n 1980:57), som etim es the etched
g roove is lig hter th a n th e un etch ed surface (M o rla n 1980:57), a n d som etim es
the g ro o v e is d a rk e r th a n th e etched surface (B inford 1981 b:50). M icroscopic
insp ection o f ro o t etch ing m a rk s in dicates they are “ b ro a d , sm o o th -b o tto m e d ,
U -sh ap ed [in cross-section] g ro o v e s” (A ndrew s an d C o o k 1985:685) th a t are
in tern ally etched (C o o k 1986:157). O ccasionally ro o t etching has been in te r
p reted as h u m a n -g e n era te d (B inford 1981 b:49—51).
T he presence o f ro o t etching indicates the b o n e existed in a p la n t-su p p o rtin g
sed im en tary e n v iro n m e n t fo r a t least p a rt o f its ta p h o n o m ic history. C o o k
(1986:157) re p o rts th a t ro o t etching “ affects b ones as they are being b u ried an d ,
in th e case o f lichen, in d icates a p erio d o f at least p a rtia l exposure w ith o u t m uch
d istu rb a n c e ." A n drew s (1990:19) im plies th a t ro o t etching occurs su b sequent
to th e b u rial o f a bone. H ow ever, we d o n o t yet k now precisely w hich kinds o f
p la n t ro o ts create the etching, o r even if it is the ro o ts o r associated fungi th a t
create the etching, th u s we do n o t kn o w if bones m u st be buried, a n d if so how
deeply they m u st be b u ried , to be susceptible to ro o t etching. F u rth e r, we do n o t

know th e ra te o f ro o t (or fungi) etching; th a t is, we do n o t know how long it
tak es fo r a ro o t in c o n ta c t w ith a bo n e to etch a n obvious groove (an d th u s we
do n o t k n ow if different species o f p lan ts etch bones a t different rates). T h u s as I
n o ted ab o v e, the d istin c tio n betw een w h eth er the ro o ts d o the etching o r fungi
asso ciated w ith deco m p o sin g ro o ts do the etching m ay be a critical distin ctio n
if “ ro o t etch in g ” d a ta are to be used to infer exposure history. If we knew the
kinds o f things in d icated a b o u t the agent responsible fo r the etching, this d ata
m ay serve as a check on bone w eathering d a ta an d aid in the d e te rm in a tio n o f
the ex p osu re h isto ry o f an assem blage.
T w o o th e r w ays th a t ro o t etching d a ta m ay prove analytically useful concern
the d istrib u tio n o f such etching in an assem blage. G ra y so n (1988) n o tes th a t the
b o n e assem blages asso ciated w ith five s tra ta in a U ta h cave display different
p ro p o rtio n s o f etched bones. T he deepest s tra tu m co n tain s no etched bones
w hereas from 1 to 17% (avg. = 5 .4 ± 6.6) o f the specim ens m ak in g up the
assem blages from th e fo u r shallow er s tra ta are etched. G ra y so n (1988:30)
re p o rts th a t the deepest stra tu m is “ the only stra tu m in th e cave ch aracterized
by extrem ely low o rg an ic c o n te n t,” an d , given his u n d e rsta n d in g o f the
m echanism w hich creates this kin d o f m odification to bo n e surfaces he suggests
“ if this d istrib u tio n is related to differential fungal activity th ro u g h tim e, the
d istrib u tio n itself m ig h t have p a leo e n v iro n m en tal significance.”
W h ite (1992:119) suggests th a t “ the presence o f ro o tm a rk s on fractu re
surfaces o r on the in tern a l surface o f lim b -b o n e shafts can be essential clues
ab o u t the relative tim ing o f bone fra c tu re .” If ro o t etching is p resen t on a
fractu re surface, th en the bo n e w as b ro k e n p rio r to ro o t etching, a n d th u s
p erh ap s p rio r to d ep o sitio n . But, again, n o t kno w in g the (rate or) tim ing o f
ro o t etching relative to tim e o f d ep o sitio n a n d tim ing relative to bone burial,
precludes fine reso lu tio n reg ard in g the tim ing o f b o n e fractu re. T his is
p artic u la rly th e case w hen it is realized th a t ro o t g ro w th ca n fractu re bones if
th e ro o t(s) g row th ro u g h the b o n e (B ehrensm eyer 1978).
D istin g u ish in g ro o t etching fro m o th e r kinds o f acidic co rro sio n such as th a t
created by acidic sed im ent m atrices involves re co g n itio n o f th e individual
ch ann els o r gro oves fo rm ed by the ro o ts (A ndrew s 1990:19). D igestive an d
sed im en tary co rro sio n do n o t p ro d u c e such grooves, a lth o u g h som etim es,
a p p a ren tly , ro o t etch ing can be quite extensive an d m ay resem ble the fo rm er
tw o k in d s o f co rro sio n (A ndrew s 1990:19). A nd, as n o ted above, ro o t etching
m ark s are sin uou s a n d have sm o o th . U -sh ap ed cross sections, allow ing the easy
d istin c tio n o f th em fro m h u m a n -c re a te d b u tch ery m ark s (C h a p te r 8).
Trampling
A ctu alistic study o f the effects o f tram p lin g by anim als, including h u m an s, on
b ones has ten d ed to focus on three things: the c reatio n o f m ark s o n bones, the
fractu rin g o f b ones, a n d the sp atial d isp lacem en t o f bones. T ra m p lin g also
S u rfa c e to 0.9 9
1.0 to 2.9 15.8
§ 3.0 to 4.9 22.5
S' 5.0 to 6.9 27.6

O
7.0 to 8.9 20
9.0 to 10.9
— -------r------------- T------------ t--------- — r ----------- ............. r
0 10 20 30
Percent Frequency
Figure 9.8. V ertical frequency d istrib u tio n o f tram p led artifacts. N o te the
approxim ately norm al d istrib u tio n (d a ta from G ifford-G onzalez et al. 1985).
ap p a re n tly a b rad es b ones (e.g.. B rain 1967a), b u t I reserve discussion o f th at

kind o f m o d ification fo r the next section o f this c h a p te r because tap h o n o m ic
processes o th e r th a n tram p lin g ca n also a b ra d e bones.
M ovem ent
G iffo rd -G o n zalez et al. (1985) sum m arize m uch o f the lite ratu re on the effects
o f tram p lin g on the vertical d istrib u tio n o f artifa cts, a n d they perform ed
ex p erim ents on tho se effects using sto n e a rtifa cts an d bones. W hile th eir results
co ncern the fo rm er, th o se results are nonetheless intriguing. T hey re p o rt th at
m an y p reh isto ric assem blages th o u g h t to have been tram p led a n d several
ex perim en tally tram p led assem blages display an essentially n o rm al frequency
d istrib u tio n o f item s against d e p th from surface (F ig u re 9.8). Such a d is trib u
tion o ccurs w hen the item s w ere first tram p led while they lay o n the surface, and
w hen th e su b strate w as n o t to o fine, co m p ac t, o r cem ented. W hen the tram p led
item s are located ju s t b en e ath the g ro u n d surface p rio r to tram p lin g , the
frequency d istrib u tio n o f item s afte r tram p lin g tends to “ m o re closely resem ble
a P o isso n d istrib u tio n in vertical co n fig u ra tio n ” (G iffo rd -G o n zalez et al.
1985:816). T here is also som e h o riz o n ta l displacem ent o f item s, b u t G ifford-
G o n zalez et al. p resen t no d a ta on th a t displacem ent, p erh ap s because it seems
to be m inim al. O lsen a n d S hip m an (1988:536) re p o rt th a t “ h o rizo n ta l m ove
m ent seem s to be related to the c o m p ac tio n o f the soil. A h a rd su b strate enables
b ones to stay o n th e surface longer, w hich increases the p ro b a b ility o f
h o rizo n ta l m o v em en t.” B uried b ones are less susceptible to h o rizo n ta l m o v e
m ent resu ltin g from kicking (see also Y ellen 1991b).
V ertical m ov em en t o f bones seem s to be a typical, b u t v ariab le result o f

tram p lin g . It is n o t restricted to d o w n w a rd m ovem ent; it can also be u p w a rd as
“ w hen a fo o t placed im m ediately ad jac en t to the [bone] sinks deeply in to the
su b strate a n d displaces the [bone] a n d ad jac en t sedim ents b o th laterally a n d
u p w a rd ly ” (O lsen an d S hip m an 1988:537). V ertical m ovem ent seems to
d ep en d o n th e in ten sity o f tram p lin g , th e co m p actn ess o f the sedim ents, the
ex tent to w hich bones a re alre ad y b uried w hen stepped on, a n d the size (w eight)
an d sh ap e o f the bone. D o w n w ard m o v em en t a n d resu ltin g b u rial seem s to
o ccur rap id ly in soft, san d y sedim ent b u t is slow er a n d less extensive in silt
m ixed w ith gravel, th e gravel ap p a re n tly actin g m u ch like a “ p a v e m e n t” (O lsen
an d S hip m an 1988:537). T ra m p le d objects m ay so rt by size and surface area,
w ith sm all objects b ecom ing m o re deeply b u ried th a n large objects o r objects
w ith large surface areas (G ifford 1977:183). L arge b ones such as m an y u ngulate
skulls m ay be sim ply stepped a ro u n d . T he o rie n ta tio n a n d plunge o r dip o f long
bones m ay be altered by tram p lin g , as w hen one end o f a long b o n e is stepped
on a n d forced b en e ath the surface w hile the o th e r end raises u p off o f the
g ro u n d surface (see C h a p te r 6 fo r discussion o f bone o rie n ta tio n a n d plunge).
F ragm entation
H ay n es (1991:253) im plies th a t tra m p lin g m ay “ d estro y ” som e skeletal ele
m ents, especially th o se th a t are so m ew h at w eath ered a n d easily b ro k en . T his
seem s likely w hen it is realized th a t frag m en tatio n , regardless o f the ta p h o n o
m ic process o r ag en t d o in g the fractu rin g , serves to reduce skeletal elem ents
in to pieces only som e o f w hich m ay be identifiable to skeletal elem ent. Such
“ an a ly tical ab sen ce” o f skeletal p a rts (L y m an an d O 'B rien 1987) is equivalent
to the ac tu a l d e stru c tio n o f a skeletal elem ent. Y ellen (1991 b: 165) suggests th a t
“ fro m an arch aeo lo g ical perspective d e stru c tio n occurs eith er (1) w hen a
specim en becom es so fragile th a t it c a n n o t be exposed o r rem oved fro m the
g ro u n d fo r iden tificatio n o r (2) w hen it frag m en ts in to pieces to o sm all to be
recovered by n o rm a l arch aeo lo g ical m ean s.” T h u s th ere are several ways to
d estro y bones b u t th ere are m an y w ays to end up w ith skeletal p a rts being
ab sen t fro m o n e ’s analysis.
A nd rew s (1990:8-10) tram p led single ow l pellets in sealed plastic bags and
fo u n d th a t excessive tram p lin g resulted in the d isin te g ratio n o f the pellets,
frag m en tatio n o f the m andibles a n d m axillae in the pellets (and th u s tram p lin g
p ro d u c ed m an y iso lated teeth), a n d sm aller lim b b ones rem ained essentially
in tact w hile larg er ones w ere so m ew h at fragm ented. M an y m axillae were
d estro yed . H e no tes th a t w hile encased in the pellet, ro d e n t bones ten d to be
fairly well p ro tected from the tram p lin g forces. T hus, b o th large (elephant)
bones (H ay n es 1991) a n d sm all (ro d en t) bones (A ndrew s 1990) can be bro k en ,
a n d th u s be effectively destro y ed fo r analy tical p u rp o ses, as a result o f
tram p lin g . Y ellen (1 9 9 lb: 165) suggests b o n e shape m ay influence the suscepti
bility o f b o n es to fractu re by tram p lin g ; “ sp h erical” specim ens should be less

p ro n e to b re ak ag e w hereas “ p la te -sh a p e d ” a n d “ cylinder-like” specim ens
sh ou ld be m ore p ro n e to b reakage. W hile b o n e shape can be readily m easured
using techn iques described in c o n ju n ctio n w ith F igure 6.6, I kn o w o f n o d a ta
b earin g o n Y ellen’s suggestion.
Several a u th o rs suggest tra m p lin g can b re ak bones, b u t only if certain
co n d itio n s are m et. M yers et al. (1980:487), on one h an d , were able to break
“ slightly w e ath ere d ” long b ones o f an ad u lt dom estic cow ( Bos taurus), an d
n o te th a t fresh bo n e is a “ to u g h , d u ra b le m aterial a n d is difficult to b re ak by
tra m p lin g ” (see also O lsen an d S hip m an 1988:537). S aunders (1977:105), on
the o th e r h a n d , arg ues th a t the fractu re m o rp h o lo g y o f a t least one m a sto d o n
(M a m m u t am ericanum ) b o n e recovered fro m a late P leistocene site in M issouri
suggests th e fractu re o ccurred w hile the b o n e was fresh a n d w as the result o f
tram p lin g . All o f these re p o rts in d icate (see also C h a p te r 8) th a t bones are m ore
likely to be b ro k e n by tram p lin g su b seq u en t to som e w eath erin g (w eathering
stage 1 o r g reater) th a n w hen still fresh (w eathering stage 0), a lth o u g h fresh
hones can be b ro k e n by tram p lin g . T he critical aspect o f analysis here, then, is
to discern w h eth er a b o n e was b ro k e n while fresh o r afte r it w as w eathered (see
C h a p te r 8 fo r ad d itio n a l discussion). T o aid such analyses, one sh o u ld also
a tte m p t to refit b o n e frag m en ts m echanically (C h a p te r 5), an d m easure the
d istan ce betw een refit fragm ents. T he in terp retiv e assu m p tio n w ould be th a t
closely co n tig u o u s refitting frag m en ts w ere fractu re d afte r th eir d ep o sitio n an d
th u s by tram p lin g o r possibly sed im en tary o v erb u rd en pressures ra th e r th an
th eir hav in g been b ro k e n p rio r to final d ep o sitio n (see C h a p te rs 8 a n d 11 for
a d d itio n a l discussion).
O lsen a n d S h ip m an (1988:537) suggest th a t “ since b reak ag e from tram p lin g
ten ds to o ccu r in the w eakest p a rts o f the bone, as it w ould in m o st n a tu ra l
circum stan ces, th ere does no t a p p e a r to be an y th in g p artic u la rly diagnostic
a b o u t the type o r p a tte rn in g o f break s created by this p ro cess.” W hile they are
p erh ap s co rrect, I suggest th a t because bones becom e stru c tu ra lly w eaker as
they becom e progressively m o re w eathered, the tem p o ra l placem ent o f w hen a
bo ne is b ro k e n relative to its w eath erin g stage m ay be an im p o rta n t bit o f
ta p h o n o m ic in fo rm a tio n fo r u nravelling th e exposure d u ra tio n (bones o n the
g ro u n d surface are m ore likely to be b ro k e n by tram p lin g th a n b u ried bones)
an d th e a c cu m u latio n h isto ry o f a bone (fresh bones are m ore likely to be
ac cu m u lated by biological ag en ts such as p re d a to rs a n d scavengers th a n
w eath ered bones).
M a rk s on bones created by tram pling
B ehrensm eyer et al. (1986, 1989) a n d F iorillo (1989), while not w ith o u t som e
n o n -ex p erim en tal p reced en t (e.g., A ndrew s a n d C o o k 1985), w ere a m o n g the
first to d e m o n stra te experim entally th a t tram p lin g can p ro d u c e scratch m ark s
on bones. T h e p ro b lem being addressed w as th a t the scratch m ark s created by

tram p lin g a p p e a r to be m orp h o lo g ically (on a m icroscopic scale) sim ilar to
sto ne to o l-g en erated c u t m ark s (m o rp h o lo g ical a ttrib u te s o f these m ark s are
discussed in d etail in C h a p te r 8). It suffices here to n o te th a t tram p lin g m ark s
ten d to be m ore ra n d o m ly orien ted o r m u ltid ire ctio n a l relative to b u tch erin g
m ark s. T ra m p lin g m a rk s tend to be located on the shafts o f long b ones ra th e r
th a n on the ends, a n d are shallow relative to b u tch erin g m ark s (A ndrew s an d
C o o k 1985; O lsen a n d S hipm an 1988).
Sum m a ry
D o c u m e n ta tio n o f changes in b ones in eth n o arch ae o lo g ic al a n d experim ental
co n tex ts has in d icated th a t the fossil reco rd can be greatly m odified from its
d ep o sitio n al co n d itio n by tram p lin g . Bones m ay be m oved, b ro k e n , an d
scratch ed by tram p lin g . Y et reco g n itio n o f the effects o f tram p lin g in p re h is
to ric co n tex ts h as n o t often been rep o rted . F io rillo (1989) suggests the shallow ,
su b-parallel scratch m ark s he observed on M iocene-aged equid bones from a
paleo n to lo g ical site in N e b ra sk a rep resen t tram p lin g m ark s because they m ore
o r less m atch experim entally g enerated tra m p lin g m arks. T he age o f the bones
F io rillo stu d ied precludes a h o m in id agent. S tahl an d Z iedler (1990) infer th a t
tram p lin g by h o m in id s resulted in the fra g m e n ta tio n a n d d o w n w ard d isplace
m ent o f b ones on the floor o f a 4000 y ear old house in E cu ad o r. B ones on th a t
flo o r w hich w ere n o t tram p led w ere larg er a n d h ad n o t been p u shed d o w n w ard
th ro u g h th e floor surface. S tahl a n d Z ied ler’s inferences, like F io rillo ’s, are
fo u n d ed on n e o ta p h o n o m ic a n d c o n tex tu a l d a ta , a n d analogical reasoning.
T h eir studies illu stra te well the kinds o f analyses necessary to in ferrin g th a t
tram p lin g o f a b o n e assem blage has tak en place.
A brasion
C. K. B rain (1967a) w as one o f the first m o d ern tap h o n o m ists to describe the
n a tu ra l a b ra sio n o f bones. He re p o rts th a t a collection o f m am m al bones
scattered on san dy sedim ent a ro u n d a w aterh o le h ad been tram p led by goats
(C apra hircus) a n d people com ing fo r w ater. Such d istu rb an ce . B rain
(1967a:98) believes, “ serves to co n stan tly a b ra d e the w eathered surface o f the
b o n e as it develops, p ro d u c in g a sm o o th n ess a n d polish o f the so rt th a t one
w ould n o rm ally associate only w ith h u m an agency.” H e also fo u n d n atu ra lly
a b ra d e d b on es a ro u n d g o a t c o rrals a n d alo n g p ath s, w hereas ab ra sio n was n o t
evident o n b ones lo cated in areas in freq u en tly tro d by anim als o r people. B rain
(1967a) suggests th a t eo lian -related ab ra sio n does n o t seem to be the re sp o n
sible ta p h o n o m ic process because b ones collected fro m sand d unes show
“ severe etchin g ra th e r th a n sm o o th in g o r polish. T he [bone] surface ten d s to be
selectively [abraded] as a result o f c o n sta n t b o m b a rd m e n t by the [w ind-blow n]
sand g ra in s” (B rain 1967a:99). N a tu ra l a b ra sio n is “ ch aracteristically fairly

general a n d n o t specifically restricted to an y one p a rt o f the b o n e;” b o n e tools
d isplay ab ra sio n (from m a n u fa c tu re a n d /o r use) over restricted areas o f the
b o n e surface (B rain 1967a:99).
W hile a serious flaw can be fo u n d in B rain ’s (1967a) p a p e r (he did n o t know
w h eth er th e b ones he suspected to have been a b ra d e d by tram p lin g w ere not
a b ra d e d p rio r to d ep o sitio n , a n d he did n o t see the inspected bones be
tram p led ), la te r ex p erim ental research confirm s his o b serv atio n s. It has been
fo u n d , fo r exam ple, th a t fluvial tra n s p o rt o f b ones ten d s to a b ra d e the entire
surface o f a b o n e specim en w hereas ab ra sio n by eolian activity seems to a b ra d e
only th e exposed o r to p surface(s) o f specim ens (e.g., S hip m an an d R ose
1983a). B ehrensm eyer (1990:234) re p o rts th a t w eath ered bones are m ore
v ulnerab le to ab ra sio n a n d b re ak ag e in fluvial tra n sp o rt situ atio n s th a n fresh,
green bone. T ra m p lin g creates deep scratches in b o n e surfaces (O lsen and
S hip m an 1988), so m eth in g fluvial ab ra sio n does n o t p ro d u c e (see C h a p te r 6).
S hip m an a n d R ose (1983a:79) re p o rt th a t, based on th eir experim ents o f
tu m b lin g bones in a b arrel w ith w ater a n d sedim ent, “ (1) sed im en tary a b rasio n
o f an y significant d u ra tio n will o b lite rate [stone-tool g enerated] slicing m arks,
(2) sed im en tary ab ra sio n will n o t p ro d u c e m a rk s th a t m im ic slicing m ark s, and
(3) sed im en tary ab ra sio n will o ccasionally p ro d u c e m ark s th a t m im ic carn iv o re
to o th scratch es.” S h ip m an a n d R ose (1988:328) fo u n d th a t eolian ab rasio n o f
large m am m al b o n e w ith loess does n o t create either a fine polish o r scratches
like th o se resu lting fro m extensive u tilizatio n by h om inids; fo r such ab rasio n to
p ro d u c e significant ro u n d in g o f b ro k e n edges o f bones, they suggest a great
deal o f tim e w ould be necessary (see C h a p te r 6 fo r a d d itio n al details).
G iv en th a t m a n u fa c tu rin g - a n d u se-related d am ag e to a b o n e should
p ro b a b ly o ccu r n e a r the fractu re , S ad ek -K o o ro s (1975) lists the follow ing
possible categ ories fo r the d istrib u tio n o f d am ag e displayed by a specim en,
especially a b ro k e n long bone: (1) faint, irreg u lar, ra n d o m ly located over the
surface o f th e specim en; (2) over the en tire [exposed] surface o f the specim en;
(3) o v er th e entire b ro k e n end o f the specim en [presum ing the o th e r end is an
a rtic u la r end]; (4) o v er only the tip o f the b ro k e n end; (5) over the en tire fractu re
edge o r surface; a n d (6) over the b ro k e n end a n d the sh aft, b u t n o t the artic u la r
end. Few an aly sts h ave stu d ied the d istrib u tio n o f ab ra sio n m odification across
a sam ple o f bones, b u t alo n g w ith B rain a n d S ad ek -K o o ro s, o th ers (L ym an
1984b; M yers et al. 1980) suggest the d istrib u tio n o f ab ra sio n an d polish should
be m o re spatially restricted o n bo n e specim ens used as to o ls th a n on n atu ra lly
a b ra d e d bones.
W h ite (1992) b ro k e a n d then boiled several m ule deer ( Odocoileus hem ionus)
m etap o d ials in a replica o f a p re h isto ric ceram ic vessel. W hile the w a te r in the
vessel never vig orously boiled, the bones w ere co o k ed fo r three h o u rs and
o ccasionally stirred. W hite (1992:124) inspected the b o n e frag m en ts after
co o k in g a n d fo u n d p o t polish o n 29 o f 69 (42% ) fragm ents. L o n g er specim ens
( > 3 cm ) w ere m o re likely to display p o t polish. A n d , the polish w as restricted

to “ p ro jectin g e n d s” o f the frag m en ts (W hite 1992:122), w here beveling an d
ro u n d in g o f th e ends as well as m icroscopic stria tio n s w ere fo u n d . W hite
(1992:122) suggests th a t this “ a b ra s io n ” occurs due to “ the c o n ta c t betw een the
b ro k e n edge o f the b o n e a n d th e p o t side” w hich creates “ facets” th a t are visible
to the n ak ed eye; th e facets are “ shiny relative to the rem ain d er o f the b ro k e n
bo n e edge; they are placed at the m ost p ro jectin g p o in ts o f the bone; a n d they
are m o st easily recognized as m o d ificatio n to sh arp b ro k e n edges.” These
ab rasio n -created facets m o re readily reflect light (an d a p p e a r shiny) th an
u n a b ra d e d bo n e surfaces. Im p o rta n tly , W hite (1992:124) no tes th a t the
a m o u n t o f p o t polish in a collection “ will exist alo n g a c o n tin u u m d epending
u p o n th e length o f co o k in g tim e, th e a m o u n t o f grit in the co o k in g vessel bed
lo ad , th e ro u g h n ess o f the vessel’s in n er surface, a n d the a m o u n t o f stirrin g .”
D oes W h ite ’s (1992) p o t polish w eaken o r co m p ro m ise the validity o f the
d istrib u tio n o f ab rasio n d am ag e over the surface o f a specim en as an in d icato r
o f th e use o f b ro k e n b on es as tools? W hite (1992:324) suggests it m ight. W h a t
seem s to be called for, no n etheless, is m o re intensive a n d extensive stu d y o f the
d istrib u tio n o f ab ra sio n d am ag e across the surfaces an d edges o f bone
specim ens w ith k n o w n ta p h o n o m ic histories in o rd e r to establish the range o f
v a ria tio n one m ig h t find in such d istrib u tio n s. T his necessity is highlighted by
G ilb e rt’s (1979:185) o b serv atio n th a t “ h an d lin g o f b o n e to o ls w ould likely
raise a sheen in p ro p o rtio n to th e len g th o f tim e the to o l w as u sed .” I f G ilb e rt is
co rrect, th e n u se-w ear related ab ra sio n o n b o n e to o ls m ay have a d istrib u tio n
on a specim en distinct fro m the d istrib u tio n o f ab rasio n o r polish created by
how a n d w here the specim en w as held.
All reference to ab ra sio n to this p o in t concerns m acroscopic attrib u tes.
B rom age (1984) re p o rts a series o f experim ents he p erfo rm ed to determ ine the
m icroscopic effects o f v ario u s ab rasiv e forces on form ing bone. D efining
abrasion as “ the resu lt o f an y ag en t th a t erodes the bo n e surface th ro u g h the
a p p lica tio n o f physical fo rce,” B rom age (1984:173) notes th a t all o f the
abrasive forces he exam ined rem ove incom pletely m ineralized collagen fiber
bundles. Such m icroscop ic ab ra sio n ca n occu r in p re -d ep o sitio n al, d e p o sitio
nal, a n d p o st-d ep o sitio n al contexts. B rom age (1984:175) fo u n d th a t sliding
ab rasio n (w ith ab rasiv e pap er), b ru sh in g (w ith a to o th b ru sh ), ru b b in g (w ith
th e fingers; n o te G ilb e rt’s [1979] co m m en t in the preceding p a ra g ra p h ), an d
w eight (such as sedim ent o v e rb u rd en ) all p ro d u c e sm o o th a b ra d e d surfaces
w ith o b lite rated details o f v ascu lar canals. R o u g h a b ra d e d surfaces result from
p articles a n d w ater tra n sm itte d u n d er p ressu re to bone surfaces; these surfaces
a p p e a r shiny m acro sco p ically as a resu lt o f “ increased reflectance fro m m any
su rfaces” (B rom age 1984:164). B rom age (1984:175) ca u tio n s th a t the precise
ta p h o n o m ic significance o f his experim ents requires fu rth e r stu d y a n d the
estab lish m en t o f ties w ith studies o f m acroscopic ab rasio n . O ne can w onder,
fo r exam ple, h ow m u ch ab ra sio n force (stren g th , d u ra tio n ) o f each p a rtic u la r
k in d B rom age identifies is req u ired to rem ove traces o f o th e r ta p h o n o m ic

processes such as b u tch erin g m ark s o r gnaw ing m arks.
A final o b serv atio n by M artill (1990:282) is im p o rta n t; he suggests th a t
ab rasio n m ay be m ed iated by “ the slightly elastic surface o f [fresh] bo n e which
retain s its o rg an ic m atrix , th u s a b so rb in g som e o f the shock o f im p actin g sand
g ra in s.” T he significance o f this o b serv atio n is crucial because it im plies th at
bones d isplaying w eathering stage 2 o r 3 will suffer ab rasiv e d am ag e m uch
m ore quickly th a n b ones displaying w eath erin g stage 0 sim ply because o f
differences in th e degree to w hich the b ones are w eathered. If M a rtill’s
suggestion is co rrect, then analyses o f ab rasio n d am ag e should no t be
p erfo rm ed w ith o u t co n sid eratio n o f the degree to w hich bones are w eathered.
In co m b in atio n , p erh ap s these tw o variables will reveal details o f the tra n sp o rt
an d ex p o su re h isto ry o f the bo n e assem blage. I am n o t aw are, how ever, o f any
actu alistic research on the c o v a ria tio n o f w eathering a n d ab rasio n .
Burning
It is im p o rta n t in an y discussion o f bu rn ed bo n e to clarify w h at is m ean t by

“ b u rn e d .” M arsh all (1989:17) notes th a t cooking involves the p re p a ra tio n o f
food fo r eating by heatin g th a t food by boiling, ro astin g , bak in g , o r the like.
H eating an object involves m ak in g th a t object w arm o r hot. Burning results
from excessive heat an d m odifies o r d am ag es the heated object. Excessive heat
can involve high tem p eratu res, tem p o rally long exposure to h eat, o r b o th . To
suggest, th en , th a t c o o k in g will p ro d u c e b u rn e d b o n e is, p erh ap s, to o sim plistic
an d m ay be unrealistic.
In terest in w h eth er p a rtic u la r bones have been b u rn e d o r n o t p ro b a b ly has
ro o ts in th e typical use o f such m o d ificatio n to infer th a t the b ones were
d ep o sited by people a n d re p resen t the rem ains o f co o k ed m eals (see C h a p te r 6),
an d in atte m p ts to d eterm in e w hen early ho m in id s first began to c o n tro l fire.
R egard in g the latter, Jam es (1989) re p o rts th a t in 34 low er an d m iddle
Pleistocene sites in the O ld W orld, 11 kinds o f evidence o f the use o f fire are
cited. O ne o f th o se kind s o f evidence is b u rn e d bone, w hich is re p o rte d at 10 o f
the 34 sites, an d it is th e exclusive kind o f evidence at five sites. B oth o f these
uses o f b u rn e d b o n e are w h a t I co n sid er interpretive.
Before b u rn in g can be used in an in terp retiv e sense one m ust kn o w the
a ttrib u te s used to d eterm in e w h eth er o r n o t a bo n e specim en h as been b u rn ed . I
include discussion o f b u rn in g in this c h a p te r on b io stra tin o m y because m ost
b ones are p ro b a b ly b u rn e d som etim e betw een the d e a th o f an o rg an ism and
b urial o f th a t o rg a n ism 's bones. H ow ever, it is im p o rta n t to note th a t while
b ones m ay typically be ch a rre d p rio r to d ep o sitio n a n d burial o r after
d ep o sitio n an d p rio r to b u rial (e.g., G ra y so n 1988; L ym an 1988a), they m ay
also be b u rn ed afte r b u rial if the m atrix they are buried in is rich in organic
m aterial a n d dry, such as is fo u n d in m an y cave sites in the w estern U n ited
S tates (see discussion in Jam es 1989:9-10). T h u s, there seem to be m inim ally

tw o steps to in terp re tin g b u rn e d bone: identifying bo n e as b u rn e d , an d
id entifying w hen the b o n e w as b u rn e d relative to its d ep o sitio n al a n d burial
history. W e co n sid er each in tu rn .
A ttrib u tes o f burning

O n the basis o f ex perim ents “ w ith bones in the ashes o f cam pfires” B rain
(1981:54) suggests “ there are tw o d istin ct stages in the c h a rrin g o f b o n e .” As
collagen is ca rb o n ized , the bo n e tu rn s black. W ith c o n tin u ed heating, the black
c a rb o n is oxidized a n d the b o n e becom es w hite a n d h as a chalky consistency.
B rain (1981:55) labels th e b lack stage “ ca rb o n iz e d ” a n d the w hite stage
“ calcin ed .” Jo h n so n (1989:441) d istinguishes fo u r b u rn in g stages: u n b u rn e d ,
“ sco rched (superficial b u rn in g ), c h a rre d (blackened, to w ard s ch arco al), an d
calcined (blue-w hite, loss o f all org an ic m ateria l, plastically d efo rm ed ).” T he
differences betw een B rain ’s a n d Jo h n s o n ’s b u rn in g stages u n d ersco re th a t
h eatin g o f a b o n e is a process because the b o n e ’s te m p e ra tu re (o r p erh ap s the
d u ra tio n th a t it is exposed to h eat) m u st increase fo r it to p ro g ress from
“ c h a rre d ” to “ calcin ed .” It is m y im pression th a t typically it is the co lo r o f a
specim en, a n d less freq u ently a specim en’s lack o f h ard n ess w hen w hite a n d its
b rittlen ess w hen b lack, th a t an aly sts use to d istin g u ish b u rn e d b o n e from
u n b u rn e d bone.
K iszely (1973) suggests th ere are three basic stages o f change in bone as it is
h eated. W a te r escapes fro m the b o n e a t a p eak ra te w hen it atta in s a
te m p e ra tu re o f 137°C a n d slows to a m inim um ra te a t 220°C. A t the latter
te m p e ra tu re th e second stage o f change tak es over; it involves the liquification
an d d ec o m p o sitio n o f org an ic m atter. T he second stage peaks a t a b o u t 330°C
an d is essentially com p leted w hen the b o n e atta in s a te m p e ra tu re o f 380°C.
F in ally , virtually all org an ic m a tte r is b u rn e d aw ay a t 600°C. K iszely’s
ex p erim en tal results w ere o b ta in e d w ith p o w d ered co rtical bone. Y on E n d t and
O rtn e r (1984) re p o rt th a t the stru c tu ra l density o f b o n e influences the accessi
bility o f th e m o lecu lar c o n stitu en ts o f b o n e m aterial. T hus, the tem p eratu re
m in im a a n d m axim a p er stage m ay be low er fo r K iszely’s m aterials th a n for
u n p o w d e red b ones a n d b o n e fragm ents.
O n the basis o f experim ents in w hich bones w ere heated in a m uffle fu rn ace/
kiln S h ip m an et al. (19 8 4 b :3 14) conclude th a t co lo r is a p o o r in d ic a to r o f the
precise te m p e ra tu re to w hich a bo n e was heated due to difficulties in recording
co lo r accu rately a n d because b ones m ay ch ange co lo r diagenetically a n d th u s a
specim en’s co lo r m ay have n o th in g to d o w ith w h eth er o r n o t the bo n e was
h eated. T hey suggest, how ever, th a t a specim en’s co lo r can be used as an
in d icatio n o f th e range o f te m p e ra tu re s to w hich a b o n e w as h eated if diagenetic
processes h ave n o t altered the specim en’s co lo r a n d it is clear th a t the bone w as
h eated (S h ip m an et al. 19 8 4 b :3 14). L ightly heated bones (< 4 0 0 ° C ) tend
crem ation
g ra ss fi re (> 6 5 ° C , < 6 mi nutes) pyres
Heat
Source /
camp fire s oak fi re coals
open forest (<
hydroxyapatite la rg e r crysta l size

S t r u ct ur e
------------ >
crackin g, epiphyses honey-combed
blue-
Shipm an yellow ish re d -b ro w n black g re y -w h ite white
Color
1 1 ^
Davis dark brown b lu e -g re y white
to black to light
grey
Degrees | | | | | | | | | | |
Celsius 0 >0 0 200 300 400 500 600 700 800 900 1000
Figure 9.9. Sum m ary o f changes to bone subjected to heating. R edraw n after
Shipm an (1988a:279) w ith add itio n s and m odifications a fte r D avid (1990).
to w ard n eu tra l a n d yellow colors (F ig u re 9.9). B ones heated betw een ab o u t

300°C a n d 800°C ten d to be yellow -red, an d red to pu rp le. Intensively heated
b ones ( > 600°C) tend to be p u rp lish -b lu e a n d blue. B ones th a t are “ com pletely
in cin erated o r calcined [can be] described as bluish-w hite o r gray in c o lo r”
(S h ip m an et al. 1984b:308).
S h ip m an et al. (1984b) re p o rt th a t the m ic ro m o rp h o lo g y o f bone, enam el,
a n d d en tin changes w ith progressively g re ater tem p eratu res, a n d these changes
in d icate recry stallizatio n o f b o n e m ineral a n d possibly the m elting o f h y d ro x
y ap atite. U sing scan n in g electron m icroscopy they fo und th a t a m a jo r change
in h y d ro x y a p a tite cry stal size occurs a t a b o u t 645°C. B rain a n d Sillen
(1988:464) re p o rt th a t in bovid bones “ heated to 300-400°C the lam ellar
stru c tu re w as greatly a c c e n tu a te d ,” b u t d o n o t describe w h at this m eans in
term s o f ap p e ara n ce. S h ip m an et al. (19 8 4 b :3 2 1) suggest th a t the d ec o m p o
sition o f th e o rg an ic c o m p o n e n t o f b o n e “ p ro b a b ly occurs betw een 360°C and
525°C;” this ran g e is hig h er th a n th a t re p o rte d by K iszely (1973) a n d described
above, alth o u g h it is n o t in co m p atib le w ith his results because K iszely used
p o w d ered b o n e a n d th u s the a lte ra tio n he observed m ay h ave o ccurred a t low er
tem p eratu res th a n th a t observed by S hip m an et al. (1984b). F inally, S hipm an
et al. (1984b) re p o rt th a t b o n e sh rin k s m ore as it is heated to progressively
h igh er tem p eratu res. T hey also in d icate th a t the a m o u n t o f sh rin k ag e m ay be
d ep e n d en t on the ra tio o f spongy to co m p ac t bone in the m easured section, the
a m o u n t o f sh rin k ag e increasing as the a m o u n t o f spongy bone increases.
G ilch rist an d M ytu m (1986) d o cu m en t an average sh rin k ag e o f 5 to 30% for
cow ( Bos sp.) a n d sheep (O vis aries) bones, w ith the sm aller (and a p p a re n tly
on to g en ically y oun g er, an d th u s less m ineralized a n d m o re p o ro u s) bones o f

the la tte r tax o n sh rin k in g m ore th a n the b ones o f the form er. T he o b serv atio n s
re p o rted by S hip m an et al. (1984b) are sum m arized in F igure 9.9, a n d are
su p plem en ted by o b serv atio n s re p o rte d by B rain a n d Sillen (1988), D avid
(1990), an d G ilch rist a n d M y tu m (1986).
B u ik stra a n d Swegle (1989) review d a ta th o u g h t to in d icate the co n d itio n o f
the b o n e at th e tim e o f in cin eratio n . T hey re p o rt th a t “ m o st descriptive an d
experim ental w o rk has focused on d istinguishing betw een bones b u rn e d in one
o f th ree co n d itio n s: fleshed, green (defleshed sh o rtly before b u rn in g ), a n d d ry ”
(B uik stra a n d Swegle 1989:248). E x p erim en tal d a ta in d icate th a t bones b u rn e d
w hen d ry disp lay surficial checking a n d cracking, a lack o f lo n g itu d in al
splitting, a n d no w arping; bo n e th a t w as fleshed an d green w hen b u rn e d does
n o t display these attrib u tes. F leshed (crem ated) bo n e tends to have “ serrated,
tran sv erse fractu res to tally th ro u g h it, alo n g w ith d iag o n al crack in g ac c o m p a
nied by w a rp in g ” (T h u rm a n a n d W illm ore 1981:281). Bone th a t w as bu rn ed
while green (m oist, o r w eathering stage 0) displays “ serrated fractu re s n ea r
epiphyses b u t o therw ise parallel-sided fractu re s th ro u g h it (along checking
lines), a n d less p ro n o u n c ed w a rp in g ” (T h u rm a n a n d W illm ore 1981:281).
B uikstra an d Swegle (1989:249) em phasize th a t the a ttrib u te s used to d e te r
m ine if b o n e was fleshed, green, o r d ry w hen b u rn e d are ten tativ e because they
depen d in p a rt on the te m p e ra tu re o f the fire, the length o f tim e the b o n e is
heated , an d the a m o u n t o f flesh covering the b o n e (see also D avid 1990 and
below).
B u ik stra an d Swegle (1989) p erfo rm ed experim ents to ev alu ate the criteria
review ed in the preceding p a ra g ra p h . T o a p p ro x im a te “ o p e n -a ir” heating, they
h eated som e o f each o f fleshed, defleshed a n d green, a n d d ry b ones in a gas
in cin erato r, a n d o th e r v ariously fleshed, green, a n d d ry bones in a w o o d fire.
T h eir term in o lo g y fo r b u rn in g stages is: “ u n b u rn e d , sm oked, a n d calcin ed ”
(B uikstra an d Swegle 1989:250). I presum e th eir “ sm o k ed ” stage is equivalent
to B rain ’s (1981) “ ca rb o n iz e d ” stage because they in d icate th a t sm oked bo n e is
black an d displays “ in com plete co m b u stio n o f org an ic m a te ria ls” (B u ik stra
an d Swegle 1989:252). T he w ood fire was b uilt over b ones laid on the g ro u n d to
a p p ro x im a te h u m a n m o rtu a ry practices. T h eir conclusions a n d o b serv atio n s
can be sum m arized as follows:
1. only defleshed bone is uniform ly sm oked (blackened);
a. dry bone has insufficient organic substance to becom e uniform ly sm oked;
b. flesh on bone insulates the covered areas which retain an un b u rn ed colo r w hereas
exposed surfaces becom e blackened;
2. fleshed bone th a t has been calcined can n o t be distinguished on the basis o f color
from green bone th a t has been calcined;
a. b o th fleshed and green bone th a t has been calcined exhibits deep longitudinal
fissures (cracks), an d transverse splitting is com m on on both; b o th are white,
blue, a n d /o r gray;
b. dry bone th a t has been calcined exhibits shallow longitudinal fissures, and
transverse splitting is present b u t rare; calcined dry b one is light brow n or tan.
T hey co nclu d e th a t ta p h o n o m ists m ay distinguish b o n e th a t w as b u rn e d w hen

dry from bone th a t w as eith er fleshed o r green, bu t the la tte r tw o c a n n o t be
reliably distin g uish ed . T he ease w ith w hich the p re -b u rn co n d itio n m ay be
d iag n o sed is d ep e n d en t on the intensity o f the burning: sm oked bones are m ore
difficult to d iag n ose th a n calcined bones.
Tim ing and agent o f burning
D eterm in ing w hen a bo ne w as b u rn e d relative to its ac cu m u latio n a l and

d ep o sitio n al histo ry m ay be critical to analysis, a n d m ay help the an aly st
d eterm in e th e m echanism o f b u rn in g . It is im p o rta n t, then, first to o u tlin e how
a bon e m ight be bu rn ed . F ollow ing B uikstra a n d Swegle (1989). D avid (1990),
an d Jam es (1989), th ere are several ways th a t bones m ight be bu rn ed . T hese can
be su m m arized as follows:
a. hum anly burned bone (bu rn in g is intentional);

1. cooking;
2. disposal o f food w aste (perhaps to reduce attractiveness to scavengers; G ifford-
G onzalez 1989a: 187);
3. fuel fo r anth ro p o g en ic fires (for w arm th a n d /o r pro tectio n from predators);
4. crem ation (generally o f hum an rem ains);
b. natu rally bu rn ed bone (bu rn in g is accidental o r unintentional);
1. nearness to one o f the an th ro p o g en ic kinds o f fires above;
2. brush (grass, forest) fire;
3. in situ burn ing o f organic m atrix.
T he m ajo r d istin c tio n is th a t betw een n atu ra lly b u rn t bone an d h u m an ly b u rn t

bone. O nce this d istin c tio n is m ade, the an aly st m ay w a n t to d eterm in e if
h u m an ly o r in ten tio n a lly b u rn t b o n e was b u rn t d u rin g crem atio n , disposal o f
food w aste, o r c o o k in g o f food. I believe few w ould disagree w ith m y suspicion
th a t h u m an ly in ten d ed b u rn in g s tend to o cc u r m ore frequently w hen bone is
fleshed o r green th a n w hen bone is dry, a lth o u g h dry bo n e m ay occasionally
have been used as fuel fo r a fire (th u s the co n tex ts o f bo n e specim ens m ay be
im p o rtan t). Bones b u rn e d u n in ten tio n a lly o r by n a tu ra l fires m ay be fleshed,
green, o r dry. T h u s, d eterm in in g the co n d itio n o f a b u rn e d bone p rio r to
b u rn in g m ay p ro v id e a clue to the tim ing o f b urning.
D av id (1990) re p o rts three experim ents: the b u rn in g o f bones exposed to a
b ru sh fire, th e b u rn in g o f b ones in an a n th ro p o g e n ic h e a rth fo r 25 m inutes
(m axim um heat was 84°C fo r 15 m inutes), an d the b u rn in g o f bones in an
a n th ro p o g e n ic h e a rth in w hich the fire actively b u rn ed (flam es visible) for 65
m inutes a n d then sm old ered (no flam es visible) fo r an a d d itio n al five ho u rs
(sam e fire as the second experim ent). His results (D avid 1990:68, 71) indicate
th a t “ ca lc in atio n (as evidenced by grey, w hite, blue, o r bluish-green tin ts to a
b o n e)" involves the o x id atio n o f the c a rb o n created d u rin g B rain ’s (1981)
c a rb o n iz a tio n stage, a n d th a t ca lcin atio n tak es lo n g er h eatin g tim es, higher
tem p eratu res, o r b o th , relative to c a rb o n iz a tio n . C alcin atio n requires “ tem p er
atu re s o f o ver 450°C to 500°C, o r h ea tin g fo r over 3 to 4 m inutes, o r a
co m b in a tio n o f b o th ” (D avid 1990:69). In the o rd e r listed, the three experi
m en ts p ro d u c ed the follow ing p ro p o rtio n s o f the surface area o f bones
disp laying a ttrib u te s o f being u n b u rn t, ca rb o n ized , a n d calcined:
brush fire 25 min. in h earth 6 hr. in h earth
u n b u rn t 1.1 0.0 0.0
carbonized 98.9 75.5 5.0
calcined 0.0 24.5 95.0
D av id (1990:75) th u s con cludes th a t “ n a tu ra l c o n d itio n s will regularly c a rb o n

ise b o n es b u t will rarely calcine them . W hen large p ro p o rtio n s o f the surface
area o f a b o n e are calcined, one can safely infer (a n th ro p o g en ic) p ro lo n g ed fires
u n d e r high te m p e ra tu re s.” T h u s th e degree to w hich a bo n e has been bu rn ed
m ay p ro v id e a clue to the ta p h o n o m ic ag en t responsible fo r the b u rn in g .
G iffo rd -G o n zalez (1989a: 193) suggests th a t “ a bo n e b u rn e d all over was
clearly subjected to b u rn in g afte r flesh h a d been rem oved, eith er by p rio r food
co n su m p tio n o r by intensive in cin eratio n . B urning on a rtic u la r surfaces only
co uld have o ccu rred w hen the rest o f the b o n e w as p ro tec ted by soft tissues, as
w hen a jo in t o f m eat is ro a s te d .” T h a t is, the d istrib u tio n o f b u rn in g d am age
across skeletal elem ents m ay help one d eterm in e if the b o n e w as b u rn e d d u rin g
co oking. I suggest th a t such a n inference requires som e know ledge o f the
b u tch erin g p a tte rn , especially how an im al carcasses w ere d isarticu lated (see
C h a p te r 8). Jo h n so n (1989:441) suggests th a t the d istrib u tio n o f b u rn in g o ver a
b ro k e n skeletal elem ent m ay in d icate if the bo n e w as b ro k e n before, o r after,
b urning . E vidence o f b u rn in g o n fra c tu re surfaces, on the in te rio r (m edullary
cavity) o f specim ens, o r tw o co n jo in in g pieces only one o f w hich is b u rn t are
g o o d in d icatio n s th a t a b o n e w as b ro k e n (o r d isarticu lated ) before it was
b u rn ed . K n ig h t (1985:10) fo u n d , fo r exam ple, th a t co m p lete deer (Odocoileus
sp.) b on es b u rn e d in an open pit m ay be calcined on th eir ex terio r surfaces b u t
only ca rb o n ized o n th eir in te rio r surfaces.
O ther effects o f burning on bone

Several o f th e a u th o rs cited in preceding sections suggest th a t b u rn e d bones
ten d to be m o re frag m en ted th a n u n b u rn e d specim ens (e.g., Jo h n so n 1989).
T his is p ro b a b ly because b u rn e d bo n e is m o re b rittle th a n u n b u rn e d bo n e due
to th e rem oval o f o rg an ic m a tte r (collagen fibers) fro m the form er; the effects o f
recry stallizatio n (S h ip m an et al. 1984b) o n the brittleness o f a b o n e are
u n k n o w n , b u t experim ents re p o rte d by K n ig h t (1985) suggest several things.
H e fo u n d th a t com plete bones o f deer (Odocoileus sp.) becam e m o re fractu red
a n d w ere m o re o ften fractu re d by h ea tin g th a n com plete bones o f beaver
(■C astor canadensis) a n d m u s k ra t (O ndatra zibethicus) w hen all w ere heated
u n d e r sim ilar co n d itio n s. T his m ay, how ever, be a fu n c tio n o f the fact th a t the
-o- Fresh-pH 6
Fresh-pH 3
-a - Burned-pH 6
- a - Burned-pH 3
0 2 4 6 8
Day
Figure 9.10. C um ulative percentage o f w eight loss o f fresh an d bu rn ed bone
specimens placed in acid solutions (after K night 1985).
deer b ones w ere h eated fo r six h o u rs, the b eav er skeletons fo r three h o u rs, an d
the m u sk ra t sk eletons fo r one a n d a h a lf h o u rs, a lth o u g h K n ig h t (1985:8)
re p o rts th a t all w ere “ co m pletely in c in erated .”
K n ig h t (1985:22) m easu red the cru sh in g lo ad o r com pressive force (lb /in 2)
necessary to fractu re com pletely calcined skeletal elem ents. H e fo u n d th a t
generally, b u rn e d b ones fro m o ntogenically old individuals require m ore
com pressive force to fractu re th a n b u rn e d bones o f y o u n g individuals. T he
com pressive force req u ired to fractu re b u rn e d bones ten d s to be co rrelated w ith
the stru ctu ra l density o f a b o n e (see C h a p te r 7 for discussion o f stru ctu ra l
density). F ro m this he concludes th a t “ even afte r the destru ctiv e process o f
in cin eratio n each skeletal elem ent has the sam e p reserv atio n p o te n tia l relative
to th e o th e r elem ents in the skeleton th a t it h ad in a fresh co n d itio n . A dense,
fresh b o n e is still a dense b o n e a fte r in c in e ra tio n ” (K n ig h t 1985:73).
K n ig h t (1985) fo u n d th a t b u rn e d bone tends to dissolve m o re ra p id ly th an
fresh bo n e in acid so lu tio n . H e placed com pletely calcined long b o n e diaphysis
pieces o f beaver ( C astor canadensis) in acid so lu tio n s o f p H 6, pH 5, pH 4, an d
pH 3. H e placed sim ilar pieces o f fresh, u n b u rn e d b o n e in solutions o f the sam e
p H , a n d reco rd ed th e daily w eight loss o f each specim en over eight days.
R esults fo r the specim ens placed in the least acidic an d the m o st acidic solutions
are p lo tte d in F ig u re 9.10 as cu m u lativ e p ercen tag e o f w eight loss over the eight
days. N o te th a t n o t only does the m o re acid so lu tio n result in m ore ra p id w eight
loss fo r b o th fresh a n d b u rn e d bone, b u t th a t th e b u rn e d b o n e in th e pH 6
so lu tio n lost w eight m o re ra p id ly th a n the fresh bone in the pH 3 solution.
B ased on their experim ents. B rain a n d Sillen (1988:464) re p o rt th a t bu rn ed
bones c o n ta in m o re c a rb o n (w h at they term “ org an ic c h a r” ), w hich develops

betw een 300°C a n d 400°C, th a n u n b u rn e d bone. F u rth e r, the ca rb o n -to -
n itro g en ra tio in fresh collagen ranges from 2.9 to 3.6 w hereas in their
exp erim en tally b u rn e d b ones th a t ra tio ranges fro m 4.2 to 6.0. T hey used these
criteria as well as histological ones to d istin g u ish b u rn e d bones fro m those
blacken ed by diagenetic in c o rp o ra tio n o f m anganese dioxide. W hile they do
n o t p resen t the req u isite d a ta fo r statistical testing, the frequency d istrib u tio n
o f b u rn e d b on es across vertical space in an ex cav atio n they re p o rt seem s to be a
fun ctio n o f sam ple size o r the frequency d istrib u tio n o f all bone specim ens
across vertical space (B rain an d Sillen 1988:465). T h a t is, the m o re bone
specim ens they fo u n d p er vertical ex cav atio n unit, the m o re b u rn e d bones they
fo un d . I n o te th a t the significance o f such a frequency d istrib u tio n (or, the lack
o f a co rre la tio n betw een the frequency o f b ones a n d th e frequency o f b u rn ed
bones) m ig h t reveal d etails re g ard in g the b u rn in g h isto ry o f the specim ens.
G ilch rist a n d M y tu m (1986) re p o rt th a t 10 to 50% o f the bones they heated
were “ d estro y e d .” K n ig h t (1985) suggests bones are no t literally destro y ed by
b u rn in g , b u t ra th e r becom e extrem ely fragm ented. G iven th a t intensive
frag m en tatio n reduces the p ro b a b ility th a t a specim en can be identified
(L ym an a n d O 'B rien 1987), b ones are p ro b a b ly n o t literally d estro y ed by
b u rn in g b u t ra th e r are analytically destroyed; th a t is, th eir pieces m ay be
recovered b u t be u n identifiable d u e to th eir sm all size.
A nalysis o f burned bones
T here seem s to be a belief a m o n g at least som e arch aeo lo g ists th a t bu rn ed

bones will preserve b e tte r th a n u n b u rn e d bones. T his belief ap p a re n tly resides
in in d icatio n s th a t carb o n ized rem ains are chem ically in ert because they consist
m ostly o f c a rb o n (G ilch rist a n d M ytum 1986:30), carb o n ized rem ains are m ore
re sistan t to b io d e g ra d a tio n th a n u n ca rb o n ized organics (e.g., B ow er 1986:94),
o r b o th . K n ig h t's (1985) ex perim ental results indicate the fo rm er is false (e.g..
F igure 9.10). T he la tte r m ay well be tru e co n sid erin g th a t it is the organic
fractio n o f b o n e tissue th a t is ca rb o n ized a n d a t least som e biological
organism s, especially m icroscopic ones, feed on the org an ic fractio n . T he
q u estio n rem ain s, how ever, once an assem blage o f fau n al rem ains have been
so rted in to b u rn e d an d u n b u rn e d specim ens, w h at is the an aly st to d o next?
A n alysts o ften present co u n ts o f b u rn e d a n d u n b u rn e d specim ens, o r
p ro p o rtio n s o f the to ta l N IS P fo r each tax o n th a t are b u rn t. T hese d a ta are
little m o re th a n descriptive, how ever, especially if b ones b u rn e d in ten tio n a lly
h ave n o t been so rted fro m th o se b u rn e d accidentally. A s em phasized by several
a u th o rs (e.g., G ra y so n 1988; W hite 1992), d eterm in in g the p ro p o rtio n o f
p a rtic u la r skeletal p a rts (e.g.. distal hum eri, first p h alan g es) th a t is b u rn e d m ay
be m ore in fo rm ativ e o f the b u rn in g processes, a n d w h eth er bones w ere fleshed
o r n o t w hen they w ere heated. A re only sm all specim ens th a t c a n n o t be
identified as a p a rt o f a p a rtic u la r skeletal elem ent bu rn ed ? D o m o re b u rn e d

bones disp lay m o re b u tch ery m ark s (C h a p te r 8) o r evidence o f carnivore
gnaw ing th a n u n b u rn e d specim ens? A re skeletal elem ents th a t are only covered
by th in layers o f so ft tissue in life, such as the m etap o d ials o f un g u lates, m o re
heavily a n d frequently b u rn e d th a n skeletal elem ents covered by thick layers o f
soft tissue, such as th e h u m eri o f ungulates? If m ultiple assem blages o f bu rn ed
bones are recovered from a site, are the sam e skeletal elem ents alw ays b u rn e d in
all o f the assem blages?
If th e stra tu m in w hich bones are fo u n d is org an ic rich an d c h a rre d , an d the
b o n es in th a t s tra tu m are also c h a rre d , th e n it is likely the bones w ere bu rn ed
w hen th e m atrix in w hich they are em b edded bu rn ed . If m ost o f the bu rn ed
b o n e specim ens in an assem blage are spatially associated w ith fire h earth s, fire
pits, a n d /o r c o n c e n tra tio n s o f th erm ally fractu red rock w hereas u n b u rn ed
bo n e specim ens are n o t associated w ith o th e r evidence o f fire o r heat, then it is
likely th e b u rn e d b ones w ere b u rn e d by h u m an s, a n d p ro b a b ly in ten tio n ally . If
the d istrib u tio n o f b u rn e d a n d u n b u rn e d bones is identical, then one m ight
arg u e th a t b ones w ere b u rn e d p rio r to th eir final d ep o sitio n (e.g., S tahl and
Z eidler 1990).
T here are few goo d exam ples o f analyses o f b u rn e d bones th a t are am enable
to su m m ary here (b u t see G iffo rd -G o n zalez 1989a; W hite 1992 fo r extended
discussions). It sh o u ld nonetheless be clear th a t the significance often ascribed
to b u rn e d b ones d em an d s intensive an d extensive analysis o f fau n al rem ains
th a t have been b u rn ed . W hile actualistic research has gone fa r tow ards
solidifying the m od ification a ttrib u te s th a t signify bones w ere b u rn ed , th at
sam e research h as only been m inim ally d irected to w ard s developing analytical
tech n iq u es for u n ravelling the ta p h o n o m ic histories o f assem blages w ith varied
frequencies o f b u rn e d skeletal p arts. Sim ilarly, d etailed analyses o f b u rn ed
b ones w hich include the kind o f d a ta alluded to in preceding p a ra g ra p h s are
few in n u m b e r at this tim e. T hus, it seem s we need to explore the v ariab ility o f
b u rn in g m o d ificatio n alo n g a n u m b e r o f dim ensions (e.g., w hich skeletal p arts
are b u rn e d , w here are b u rn e d b o n es fo u n d a n d w h a t are they regularly
asso ciated w ith) to ascertain w hich o f th o se dim ensions are archaeologically
visible, a n d w hich ones m ay help us identify h u m a n behaviors.
Other biological agents of bone modification
A skull is fo und covered w ith m ud firmly stuck on, and w ith the traces o f the white
a n ts’ [termites] tunnels run n in g th ro u g h . If the m ud is rem oved, large areas o f the
cranial walls m ay be fo u n d to be d isappeared altogether. In less exaggerated cases,
holes will be seen w ith w hite, gnaw ed edges, o r p erh ap s only the surface o f the bone
has been attack ed . The cranial sutures are a favourite site for the com m encem ent o f
the term ites' operations.
(D. E. D erry 1911:245)
In earlier c h a p te rs we see th a t ca rn iv o res a n d ro d en ts, in p artic u la r, variously

con su m e a n d m o d ify v e rte b ra te fa u n al rem ains. T here are several o th er
biological agen ts o f b o n e m odification, m o st o p e ra tin g d u rin g the b io s tra tin o

m ic ph ase o f a ta p h o n o m ic h isto ry , th a t sh o u ld be m en tio n ed . T he m a jo r one
zo o a rch aeo lo g ical ta p h o n o m ists have been co n cern ed w ith is h erbivores
(m am m alian ca rn iv o res are covered elsew here in this volum e), b u t there are
o th ers as well. M artill (1990:279), for exam ple, refers to studies o n snails
k n o w n to eat w hale bones. In the follow ing I restrict discussion to three o f the
b etter k n o w n kinds o f bone consum ers.
Insects
D erry (1911) was one o f the first to re p o rt on the d am ag e to bones th a t could be

created by th e gn aw ing actio n o f insects, especially term ites. B ehrensm eyer
(1978:154) b ro u g h t the possibility o f such d am ag e to the a tte n tio n o f m o d ern
tap h o n o m ists w hen she illu stra te d grooves gnaw ed in h o rn cores by m o th
larv ae w hich feed o n th e o rg an ic co m p o n en ts o f h o rn s. T he grooves she
illu stra te d are p erp en d icu lar to the g rain o f the h o rn core (B ehrensm eyer
1978:156). In the only ex p erim en tal study o f w hich I am aw are, W a tso n an d
A b bey (1986) fo u n d th a t A u stra lia n term ites gnaw b o n e a n d create “ scratch es”
in b o n e surfaces. C an cellous b o n e is m o re d am ag e d th a n co m p ac t bo n e, an d
“ o n co m p ac t bone, d am ag e is co n c en trate d o n ro u g h e n ed surfaces o r along
edges” (W atso n a n d A b bey 1986:250). C ancellous b o n e is som etim es “ tu n
neled in to ” (W atso n a n d A bbey 1986:250). O verall, the chew ing o n bone by
term ites is “ superficial” a n d W a tso n a n d A bbey (1986:251) suggest bo n e is
sim ply “ explo red by chew ing in the sam e w ay th a t term ites explore a range o f
h a rd plastic m a te ria ls.” W a tso n a n d A b b ey (1986:253) fo u n d “ no c o rrelatio n
betw een th e presence o f visible o rg an ic m ateria l a n d the lo catio n o r severity o f
term ite d am age, b u t th e high a n d u n ifo rm c o n c e n tra tio n o f n itro g en in the
bones m ay have been a ttra c tiv e ,” a lth o u g h it ap p e a rs th a t term ites gnaw
“ relatively fresh b o n e ” m o re intensively th a n “ old bones in [hum an] o c c u p a tio
nal d ep o sits.”
O ver eighty years ago S m ith (1908:524) described w h at he in terp re ted to be
gnaw ing d am ag e to h u m a n b ones created by beetles. H e re p o rts th a t the
gnaw ed b ones typically show gnaw ing d am ag e on the “ u n d er surfaces o f the
bones as they h a p p e n e d to lie in the g ro u n d . . . especially o n th o se p a rts w hich
are pressed tightly ag a in st the soil,” a n d o n this basis he concludes th a t the
d am age (originally th o u g h t to rep resen t a n a n te m o rte m p a th o lo g ic a l c o n
dition ) h ad o ccu rred p o stm o rtem . (E rz in ^ io g lu [1983:58] states th a t in forensic
e x am in atio n s, “ the soil b en e ath the corpse sh o u ld be exam ined fo r larvae an d
p u p a ria .” ) S m ith (1908:524) also re p o rts th a t “ a w hite p ow der, consisting o f
pulverized bone, is o ften fo u n d sp rin k led over the d am ag ed p a rt a n d the
ad jo in in g soil; in m any cases this is obviously fresh . . . T he b u rro w s (usually
a b o u t 1 m m in d iam eter) o f sm all an im als can alw ays be seen leading to the
[dam age trace].” T hese b u rro w s som etim es co n tain e d the rem ains o f beetles.
Sm ith (1908:524) in d icates th a t “ the little grooves p ro d u c ed by the scraping o f
the beetles are d istinctly visible” o n the edge o f the d am ag e traces w ith the aid
o f a m agnifying lens, an d sedim ent is often caked a ro u n d the traces.
M an y zoologists w ho stu d y an im al skeletons clean co m p a ra tiv e skeletons by
using d erm estid (D erm estes sp.) beetles (H ild eb ran d 1968:21-23 w ith refer
ences). A c a u tio n o ften expressed w hen describing this tech n iq u e o f skeletal
p re p a ra tio n is th a t as the bones are progressively defleshed, the skeleton should
be closely checked to ensure th a t it is n o t dam ag ed by th e feeding beetles. H efti
et al. (1980:45) w rite th a t skeletons sh o u ld be rem oved fro m the beetle colony
w hen th e skeleton is free o f soft tissues “ since longer ex p o su re results in p a rtia l
d estru c tio n o f p a rts o f the skeleton, the beetles atta c k in g bone w hen they are
dep riv ed o f o th e r fo o d .” T hey d o n o t describe the d am ag e su stain ed by bones
so a tta c k e d , b u t re p o rt th a t the d am ag e “ is easily seen by eye, since the process
is such th a t pieces o f co n sp icu o u s size becom e d e ta c h e d ” (H efti et al. 1980:47).
K itch in g (1980) illu strates holes in fossil bovid b ones he believes w ere
gnaw ed by d erm estid beetles. F ollow ing his lead. Jo d ry a n d S ta n fo rd
(1992:111-113) illu stra te a n d describe holes in bison b ones recovered fro m a
late P leistocene-early H o lo cen e site in C o lo ra d o . T hese holes are 9 to 14 m m in
d iam eter, o ccu r singly on a specim en, typically orig in ate from a n a tu ra l
a p e rtu re such as a fo ram en , a n d a p p e a r to have been b o re d o r chem ically
dissolved ra th e r th a n p u n ctu re d . Jo d ry an d S ta n fo rd (1992:113) state th a t the
holes th ey observed are larg er th a n th o se gnaw ed by derm estids (w hich they
in d icate are ca. 6 m m d iam eter), a n d suggest, follow ing K itch in g (1980), th a t
the holes in the fossils w ere gnaw ed by c a rrio n beetle p u p a ria . H oles b o re d o r
eaten th ro u g h b o n e by insects “ ca n be distinguished fro m p u n ctu re s m ad e by
carn iv o re teeth by th eir larg er size a n d the absence o f crushed bo n e in the
b o tto m ” (H esse an d W ap n ish 1985:85).
Several kind s o f evidence m u st be reco rd ed in o rd e r to infer the ac tio n o f
g naw ing insects. R ogers (1992), fo r exam ple, describes the size o f b o re d holes
asso ciated w ith grooves in several d in o sa u r bones. H e n o tes th a t the holes are
filled w ith the sam e sedim ent as the " h o st m atrix o f the b o n e b e d ,” th a t there
are n o scratches o r grooves o n th e walls o f the holes, a n d th a t fossil beetle
p u p a ria l cases w ere stratig ra p h ic ally associated (R ogers 1992:528-529). T he
lo catio n o f the b o rin g s a n d /o r grooves in the d ep o sitio n al co n tex t (e.g., on the
u p w a rd o r d o w n w a rd surface o f the bone) a n d on the bo n e (e.g., in co m p ac t o r
tra b e c u la r bone) are also im p o rta n t co n sid eratio n s.
Herbivores
It w as p ro b a b ly Sutcliffe (1973) w ho first b ro u g h t to the a tte n tio n o f tap h o -

n o m ists th a t h erb ivo res, p a rtic u la rly un g u lates, gnaw bones a n d an tlers, an d
create w h a t m ight be m istak en fo r h u m an ly -created tools. Sutcliffe's (1973)
sem inal p a p e r has been follow ed by a n u m b e r o f o th e r re p o rts o f various
h e rb iv o ro u s tax a gnaw in g bones a n d an tlers (B rothw ell 1976; G o rd o n 1976;

K ra u sm an a n d B issonette 1977; Sutcliffe 1977; W ika 1982; B ow yer 1983;
Jo h n so n a n d H ay nes 1985; W arrick a n d K ra u sm a n 1986; G reenfield 1988).
Som e o f these are by zo ologists in terested in the n u tritio n a l aspect o f bone
gnaw ing by herbiv ores. T hey believe such b eh av io r, called osteophagia, is
d irected to w a rd alleviating n u trie n t deficiencies, especially p h o sp h o ru s and
calcium . R egardless o f the re aso n th a t h erbivores chew bones, the effects o f
such chew ing are tap h o n o m ically significant.
Sutcliffe (1973:428,430) suggests th a t a b o n e is g rasped in the cheek teeth “ in
a ‘cig ar-lik e’ m a n n e r” a n d “ chew ing is w ith a sidew ays m o v em en t o f the ja w s .”
T his chew ing ac tio n results in “ p lan in g off the to p a n d b o tto m [of a long bone
shaft] u n til the m arro w cavity o r a n tle r core is reached, leaving only the sides
in tact, [p ro d ucing a] fork -like re m n a n t” (Sutcliffe 1973:430). T he p ro n g s o f the
fo rk h av e a zigzag surface th a t “ m atches th e a lte rn a tin g u p p er a n d low er cheek
teeth o f th e ch ew er” (Sutcliffe 1973:430). G o rd o n (1976:121) describes the
p ro n g s o f th e fo rk as “ u n d u la tin g a n d [having progressively] th in n ed e n d s.” H e
no tes th a t w ithin the u n d u la tio n s a n d o n the crests betw een th em there are
“ striae o r w ear p a tte rn s p arallel to th e valleys a n d crests; i.e., p erp en d icu lar to
the lo n g itu d in al axis” o f the specim en, a n d this in tu rn suggests the specim ens
w ere “ chew ed by a tran sv erse g rin d in g m o tio n o f the cheek teeth - a side to side
m ovem ent n o rm al to u n g u lates” (G o rd o n 1986:122). B rothw ell (1976:182)
suggests th e chew ing is a “ g razing-saw ing” m o tio n th a t can p ro d u c e “ m ultiple
a n d p arallel g ro oved m a rk s” o n som e bones.
M icroscopic organism s
U sing histo lo gical techniques, researchers have fo u n d vario u s m icroscopic
m odificatio n s to skeletal tissues ap p a re n tly caused by th e ac tio n o f m icro
o rganism s. F o r exam ple, it ap p e a rs th a t som e fungi p e n e tra te b o n e w ithin 2 5 -
30 day s a fte r ex p osu re (M a rch iafav a et al. 1974). F u n g i create tu n n els in bone
a b o u t 1 to 8 jim in d iam eter (one o r m icro n = 1/1000 m m ) as a result o f
a tte m p ts to gain access to collagen fo r c o n su m p tio n a n d m etab o lism (H a ck ett
1981; M arch iafa v a e t al. 1974). H y d ro x y a p a tite is red ep o sited in th e tu n n els as
a new m in eral - b ru sh ite - in the fo rm o f a “ cuff, 3 to 6 ^.m thick, in the bone
su rro u n d in g m o st tun n els, fro m 5 ^ m u p w a rd s” (H a c k e tt 1981:247). Piepen-
b rin k (1986:418) re p o rts th a t “ fungi can [also] d ecom pose dead bo n e by m ore
extensive d e stru c tio n o f h a rd tissues” (see also G a rla n d 1987, 1988; G a rla n d et
al. 1987; H a n so n a n d B u ik stra 1987:554). In som e experim ents he fo u n d th a t
“ a lth o u g h in regions w ith extensive fungal g ro w th the co rtical b o n e surface is
p artia lly ero d ed , th e fungi never actu ally p e n e tra te d o r tunnelled th ro u g h the
b o n e tissue by m ean s o f d istin ct focal d e stru c tio n ” (P iep en b rin k 1986:421).
Like H a c k e tt (1981) before him , he fo u n d “ recry stallizatio n in the tunnelled
areas [of] previously dissolved calcium p h o s p h a te .” H a c k e tt (1981:247) notes

such “ m ineral red ep o sitio n is n o t seen in p ath o lo g ical processes” a n d Piepen-
b rin k (1986:424) suggests it “ is never fo u n d in osteolytic processes th a t occur
d u rin g life;” th e m in eral red ep o sitio n can th u s be used to distinguish these
p o stm o rtem changes from p ath o lo g ical ones.
G a rla n d (1987:113) illu strates fungi w ithin p reh isto ric bo n e tissue an d
im plies th eir presence results in the “ d isin te g ratio n , disag g reg atio n an d
d isso ciatio n o f o steon s. It is still possible to recognize th e m o rp h o lo g y o f the
lam ellar bon e ad jac en t to the H av ersian can al b u t the bone to w ard s the
p erip h ery o f the osteo n has an a m o rp h o u s a p p e a ra n c e .” B one m ineral is lost,
an d a “ g ra n u la r” a p p e ara n ce to bone tissue is visible in th in sections as a result
o f fungal activity. All o f these m odifications “ were fo u n d only in the o u ter
co rtical zo n e” o f th e specim ens G a rla n d (1987:113) exam ined. F u n g i, rhizo-
m o rp h s, a n d b ac te ria lying w ith in e m p ty H a v ersian canals, m ed u llary cavities,
a n d trab e cu lae w ere fo u n d in som e p re h isto ric b o n e specim ens (G a rla n d
1987:118). T hese as well as foreign m in eral m ateria l in such spaces are term ed
inclusions, th e “ ex tra n eo u s m a te ria l w ith in bo n e spaces” (G a rla n d 1987:122).
H a c k e tt (1981:264) suggests b ac te ria invading bo n e will create “ tubules
a b o u t 300 nm [nan o m eters, 1 n m = 1 b illio n th o f a m eter = 1/1000 o f a m icron]
in d ia m e te r” (see also G a rla n d 1988), a n d th u s b ac te ria-fo rm ed tu n n els can be
distin g u ish ed fro m th o se fo rm ed by fu n g al activity o n the basis o f size
differences. P ie p en b rin k (1986:424-426) suggests th a t fu n g al activity can also
result in staining o f skeletal m aterial. G a rla n d (1987:118-120) re p o rts terra
stain in g w hich h a d p e n e tra te d in to the co rtex [of p re h isto ric b o n e specim ens] to
v ario u s d e p th s.” B ecause the stain in g (and o th e r m odifications, see above) only
o ccu rred in th e o u te r co rtical area o f the b o n e tissue o r “ the zone o f b o n e lying
in closest c o n ta c t w ith the soil, there m ay be suggested a physico-chem ical
aetiology to the n a tu re o f the staining, w ith a necessary co n d itio n being the
presence o f g ro u n d w a te r” (G a rla n d 1987:121).
H a n so n a n d B u ik stra (1987:554—559) describe a n d illu stra te v ario u s stages
o f b o n e d estru c tio n caused by m icro-organism s. F ocal destruction (the c reatio n
o f tunnels) as described by H a c k e tt (1981; see also G a rla n d 1987:118) is the first
stage, a n d th e tu n n els “ g ra d u ally coalesce to fo rm large p atches o f resorbed
co rtical b o n e ” (H a n so n a n d B u ik stra 1987:555). T he coalescence re p o rte d by
H a n so n a n d B u ik stra m ay be the a m o rp h o u s osteo n b o rd e rs re p o rted by
G a rla n d (1987, see above). F o cal d e stru c tio n ap p e a rs to begin in trac o rtic ally in
H a v ersian bone, a n d g ra d u ally sp read s to w a rd the p erio steal a n d endosteal
surfaces. “ S econd ary H a v ersian b o n e is m echanically w eak er a n d less m in era l
ized th a n in terstitia l a n d circum ferential lam ellar bo n e . . . M icro-organism s
a p p e a r to fa v o r the less m ineralized tissue fo r initial b o rin g . . . T he ab u n d a n ce
o f v asc u la r spaces in the m id -co rtex facilitates th e sp re ad o f o rg an ism s”
(H a n so n a n d B u ik stra 1987:559).
T he im p o rta n c e o f these o b serv atio n s is th a t the sm all tu n n els created by
log (live w eight)
F igure 9.11. R egression o f log 10 o f live w eight ag ain st log 10 o f the ratio o f
num ber o f individuals expected (N exp) to n u m b er o f individuals observed (N
obs) (after B ehrensm eyer an d B oaz 1980).
fun g al a n d b acterial activity reduce th e s tru c tu ra l density a n d increase the

p o ro sity o f th e tissue (Bell 1990; H a n so n a n d B uikstra 1987). T his m ay
ex acerb ate the effects o f d iagenetic processes such as cru sh in g fro m sedim ent
o v erb u rd en w eight a n d exchange o f chem ical ions (G a rla n d 1987:122; see
C h a p te r 11). D o c u m e n ta tio n o f geological o r biological stain in g u n d ersco res
the fact th a t care sh o uld be ta k e n w hen a tte m p tin g to identify b u rn e d bone
sim ply on the basis o f co lo r (F ig u re 9.9).
Preservation and size biasing
C o m p a riso n o f surface-collected assem blages o f n a tu ra lly accu m u lated

m o d e rn m a m m alian rem ains w ith the living co m m u n ity from w hich the
assem blages w ere derived suggests th a t th e rem ains o f sm all anim als preserve
significantly less well th a n th e rem ains o f large anim als (B ehrensm eyer 1981;
B ehrensm eyer a n d B oaz 1980; B ehrensm eyer et al. 1979). T his is show n in
F ig u re 9.11, w here it is clear th a t as th e log 10 o f the average live w eight
increases fo r th e 10 m am m alian tax a p lo tte d , the log 10 o f the ra tio o f the
observed to expected frequency o f carcasses p er an im al size class also increases
(/• = 0.77, P = 0.009). T h is p ro b a b ly applies as well in tra -ta x o n o m ic a lly to sm all
individuals, p artic u la rly juveniles o f a tax o n (B ehrensm eyer 1981:600) due to
the low er stru c tu ra l density o f th eir bones. B ehrensm eyer et al. (1979:17)
suggest th a t this difference in p re serv atio n is largely due to g re ater d estru c tio n
o f b o n es o f sm all m am m alian tax a by ca rn iv o res a n d scavengers, a n d m ore
ra p id w eath erin g a n d fra g m e n ta tio n due to tram p lin g o f sm all bones. In o th er

w ords, they po sit b io stratin o m ic processes as the cause o f the p re serv atio n al
b ias ag a in st the rem ain s o f anim als o f sm all size, w hich is re aso n ab le given th a t
th eir stu d y is fo u n d e d on su rface-o ccu rrin g bones th a t have n o t yet u n d erg o n e
d iagenetic processes. R etallack (1988:338), how ever, suggests a sim ilar p re ser
v atio n al bias ag a in st rem ains o f anim als o f sm all b o d y size m ay result from
d iagenetic effects because “ b ones o f sm aller m am m als h ave a h ig h er surface-to-
volum e ra tio , a n d so are m o re p ro n e to acidic d isso lu tio n th a n are th o se o f large
m a m m a ls.”
R egardless o f w ho is co rrec t in the case above, a n d b o th m ay in fact be
co rrect, D a m u th (1982:441) argues th a t the slope o f the regression line derived
by B ehrensm eyer et al. (1979), w ho o m itted the tw o p o in ts fa rth est ab o v e the
line in F ig u re 9.11 to derive a slope o f 0.68, w as sufficiently close to the surface-
to -v o lum e ra tio o f 0.67 to suggest to him th e su rface-to -v o lu m e ra tio w as
c o n tro llin g th e relatively p o o r p re serv atio n o f sm aller bones. S tudy o f fossil
assem blages w ith this p o te n tia l p re serv atio n al bias in m ind m u st, o f course,
p resu m e bones o f b o th sm all a n d large m am m als w ere being ac cu m u lated an d
d ep o sited in ab u n d a n ces p ro p o rtio n a l to th eir frequencies in th e biotic
co m m u n ity if th e an a ly st w an ts to explain a relative p au c ity o f sm all m am m al
rem ain s as being th e resu lt o f the su rface-to -v o lu m e ra tio bias, regardless o f the
m ech anism o f differential p reserv atio n . A s n o ted above, sim ilar p re serv atio n al
p ro b lem s m ay accrue in tratax o n o m ica lly ; sm all skeletal elem ents o f a n in d iv id
ual an im al m ay preserve less well th a n th a t in d iv id u a l’s large skeletal elem ents,
b u t this has n o t yet been stu d ied in th e sam e w ay th a t B ehrensm eyer et al.
(1979) stu d ied in tertax o n o m ic v a ria tio n in p reserv atio n .
Comparative analytic techniques

B ecause we are still learn in g a b o u t w hich ta p h o n o m ic processes a n d agents
p ro d u c e p a rtic u la r kind s o f m odifications to assem blages o f fa u n al rem ains,
som e an aly sts have d eveloped w ays sim ply to c o m p are vario u s m o d ificatio n
a ttrib u te s betw een assem blages w ith o u t specific reference to the tap h o n o m ic
ag en ts w hich m ay h ave created the m odifications. I review several o f these here
to p ro v id e the re ad er w ith a feel fo r som e o f the a ttrib u te s o f bone assem blages
th a t m ig h t be studied.
Percentage difference in long bone ends

In his stu d y o f th e effects o f ca rn iv o re d e stru c tio n o f bovid lo n g bones,
R ic h a rd so n (1980) develops a n e q u a tio n fo r calcu latin g the percen tag e differ
ence betw een th e frequ ency o f p ro x im al a n d distal ends o f each in d iv id u al long
bone. T h a t e q u a tio n is:
(N complete! + N proxim al endSj) —(N complete; + N distal endsO
[ 9 . 3]
(N complete! + N proxim al endsi) + (N complete! + N distal endsi)
T ab le 9.4 Frequencies o f bone p a rts in selected sites (fro m T odd and Rapson
1988:310; Todd 1987a:235; B inford 1981b: 1 7 4 -1 7 5 ). Upper ratio o f
p roxim a l:d istal is raw values; lower ratio is standardized
% difference M N E p ro x im ak M N E distal
Site hum erus tibia hum erus tibia
Jones-M iller 55.33 21.55 44:153 91:141

28.8:100 59.5:92.2
H o rn er II 40.98 17.29 38.5:61.5 47.5:55.0
62.6:100 77.2:89.4
C asper 40.74 12.50 20:47.5 31.5:24.5
42.1:100 66.3:51.6
O lsen-C hubbuck 1.42 2.04 86.5:89.0 84.0:87.5
97.2:100 94.4:98.3
L am b Spring 54.55 17.24 2.0:7.5 5.5:8.5
23.5:88.2 64.7:100
W olf kills 60.00 37.50 2.5:9.5 5.0:11.0
22.7:86.4 45.5:100
w here i is th e skeletal elem ent o f co n cern (generally the h u m eru s, radius,

m etac arp a l, fem ur, tib ia, o r m etatarsal), N com plete is the n u m b er o f com plete
specim ens o f skeletal elem ent i, N p ro x im al ends is th e n u m b e r o f specim ens
th a t are only the p roxim al end o f skeletal elem ent i, a n d N distal ends is the
n u m b e r o f specim ens th a t are only the d istal end o f skeletal elem ent i.
T o d d an d R ap so n (1988) use e q u a tio n [9.3] to co m p are differences in
p ro p o rtio n a l frequencies o f lo n g b o n e ends in several bo n e assem blages, an d
suggest w hen th e differences are m inim al, especially fo r th e tib ia a n d h u m eru s,
lim ited d e stru c tio n o f b ones by carn iv o res is indicated. A s an exam ple, the d a ta
for p ercen t difference values fo r hum eri a n d tibiae from the six assem blages
given in T ab le 9.4 (fro m T o d d an d R a p so n 1988:310) are p lo tte d in F ig u re 9.12.
T h a t g ra p h in d icates th e “ w o lf kills” assem blage h as g re ater differences in the
a b u n d a n ces o f b o th p roxim al a n d distal h um eri a n d p ro x im al a n d distal tibiae,
unlike th e O lsen -C h u b b u c k assem blage w hich has essentially eq u al a b u n
dances o f all a n d w hich has a p p a re n tly (on the basis o f o th e r evidence such as
th e p au c ity o f gnaw in g m ark s) u n d erg o n e m in im al c a rn iv o re a ttritio n .
B inford (1981 b :2 19) uses a different tech n iq u e fo r g ra p h in g d a ta like th o se in
F ig u re 9.12. H e p lo ts the stan d ard iz ed frequencies o f p ro x im al versus distal
h u m eri, a n d p ro x im al versus distal tibiae, o n the g ra p h in F ig u re 9.13a. As
show n there, if the p lo tte d p o in ts fall w ithin the area labeled “ Z o n e o f N o
D e stru c tio n ” th e n th e conclusion is the assem blage has u n d erg o n e very little
d en sity -m ed iated d estru c tio n (see C h a p te r 7) w hereas if the p lo tte d p o in ts fall
in the a rea labeled “ Z o n e o f D e stru c tio n ” th en th e an aly st concludes the
assem blage h a d in fact u n d erg o n e som e d en sity -m ed iated a ttritio n (in m an y
artio d a cty ls the distal end o f b o th the h u m eru s an d the tib ia is denser th a n its
p ro x im al end). T he frequencies o f p ro x im al a n d distal ends are stan d ard iz ed to
40 -I
Wolf kills ■
.
CD 30 -
03
1q
i-
Jones-Miller
CD ■
o
c 20 -
o
k_
Horner M b ■
Q)
3= Lamb Spring
b
Casper ■
vO 10 -
■
uOlsen-Chubbuck
0 10 20 30 40 50 60 70
% D iffe re n c e H um eri
F igure 9.12. S eatterp lo t o f % differences in frequencies o f proxim al and distal

hum eri against % differences in frequencies o f proxim al and distal tibiae (from
T able 9.4).
allow p lo ttin g o f b o th skeletal elem ents o n the sam e g ra p h . T his is acco m

plished by d eterm in in g the M N E (o r M A U ) fo r the proxim al end a n d for the
distal end o f each bo ne, a n d th en dividing all fo u r values (p ro x im al hum erus,
distal h um erus, p ro x im al tibia, distal tibia) by the largest o f the fo u r values an d
m ultiplying the results by 100. B inford (1981 b :2 19) labels these " ra tio values,”
exam ples from T ab le 9.4 (d a ta from B inford 1981 b: 174—175 an d T o d d
1987a:235) o f w hich are p lo tte d in F igure 9.13b. T his figure in dicates th a t the
O lsen -C h u b b u ck assem blage h as u n d erg o n e v irtu ally n o d estru c tio n w hereas
the w o lf kills assem blage h as u n d erg o n e som e d estru ctio n .
L ong bone shafts

T o d d a n d R ap so n (1988:314-317) suggest using a sta n d a rd m easu rem en t o f
th e w idth o f the artic u la r end to calcu late a sh aft-to -en d ra tio as a m easure o f
th e a m o u n t o f long b o n e sh aft a tta c h e d to the end. T his p ro c ed u re ac co u n ts for
in tra ta x o n o m ic sexual d im o rp h ism a n d individ ual size v aria tio n . B lum ens
chine (1988:487) suggests re co rd in g the "lim b seg m en t” represented by a long
b o n e frag m en t using the follow ing categories:
1) epiphyseal fragm ent: has all o r a p o rtio n o f the proxim al o r distal a rticu lar surface;
2) near-epiphyseal fragm ent: lacks any articu lar surfaces, b u t has trab ecu lar bone on
the m edullary surface indicating p roxim ity to an epiphysis;
3) m idshaft fragm ent: lacking a rticu lar surfaces and trab ecu lar bone.
T o d d a n d R ap so n (1988:322) suggest differentiating betw een p ro x im al and

distal sh aft p o rtio n s (see also H offm an 1988).
Ratio Value of Proximal Humerus and Tibia
Ratio Value of Distal Humerus and Tibia

Ratio Value of Proximal Humerus and Tibia
Ratio Value of Distal Humerus and Tibia

Figure 9.13. Bone d estru ctio n graphs, a, m odel g rap h o f destru ctio n o f proxim al
and distal hum eri and tibiae (after B inford 198lb:219. F ig u re 5.07); b,
d estruction grap h for six assem blages in T able 9.4: JM , Jones-M iller; H N ,
H o rn er II; C P, Casper; O C, O lsen-C hubbuck; LS, L am b Spring; W K . W olf kills;
h, hum erus; t, tibia..
T h e frequency o f sh aft frag m en ts gen erated by hom in id agents o f bone

fractu re has been discussed by B inford (1978) an d B unn (1989) (see C h a p te r 7).
T o d d a n d R ap so n (1988) suggest the analyst m ay sim ply co m p are the N IS P
frequency o f sh aft frag m ents p er skeletal elem ent w ith the N IS P frequency o f
artic u la r ends fo r th a t skeletal elem ent. In so doing, they identify w h a t they
tak e to be tw o different kinds o f frag m en tatio n , one in w hich the N IS P o f bone
ends is inversely co rrelated w ith the N IS P o f shaft frag m en ts a n d a n o th e r in
w hich the N IS P o f bon e ends is positively co rrelated w ith the N IS P o f shaft
fragm ents. W hile the precise ta p h o n o m ic significance o f such m easures is
u n clear (a lth o u g h they are p ro b a b ly at least in p a rt a fu n ctio n o f the intensity o f
frag m en tatio n , see C h a p te r 8), such values do provide a basis for co m p arin g
assem blages o f b ro k e n bones.
Discussion
In this section I have m entioned som e o f the a n aly tic techniques an aly sts use to
reco rd m o d ifications to bone assem blages, a n d how the extent an d frequency o f
those m odificatio ns m ight be c o m p ared betw een assem blages. I have, in fact,
barely scratch ed the surface o f this topic. It should be obvious, if the re ad er has
read ev erything betw een C h a p te r 5 an d this sentence, th a t there are n u m ero u s
m o d ification a ttrib u te s th a t a ta p h o n o m ist m ight ch o o se to reco rd an d
analyze. T he n u m b e r o f such a ttrib u te s is lim ited by only tw o things: (a) the
actu al m odification s d isplayed by a n assem blage o f an im al rem ains, a n d (b) the
a n a ly st’s im ag in atio n . T h a t is, there is no set o r stan d ard iz ed p ro c ed u re o f
tap h o n o m ic analysis; because ta p h o n o m ic processes are historical a n d to som e
degree cum u lativ e, each assem blage o f bones is to a g re ater o r lesser degree
tap h o n o m ically uniq ue a n d th u s requires unique analyses. O ne w orks w ith
w h at one has, a n d w h at o ne has is a m odel o f a com plete living skeleton, an d a
set o f b ones an d teeth th a t are variously different from th a t living skeleton.
R eco rdin g th o se differences is the first step in ta p h o n o m ic analysis, a n d I have
to u ch ed on w h at m ight be recorded in this section. I have also described various
w ays th e recorded d a ta m ight be m an ip u lated d u rin g analysis. A ssigning
ta p h o n o m ic m eanin g to th e differences a n d an aly tical results is the second step.
W h a t created the m odifications? W hy w ere they created? H o w d o these
m o difications a d d to o r su b tra c t from d a ta th a t are relevant to the in terp retiv e
qu estio n s one w ishes to answ er? I have to u ch ed o n how to answ er these
qu estio n s in o th e r sections.
Summary
B io stratin o m y is th a t p o rtio n o f a ta p h o n o m ic h isto ry w hich occurs betw een
the d ea th o f an o rg an ism a n d the b u rial o f th a t o rg a n ism ’s rem ains. T he
ta p h o n o m ic aspects o f d e a th are considered in C h a p te r 5, a n d b u rial a n d p o s t
b u rial ta p h o n o m ic processes are described in C h a p te rs 10 a n d 11, respectively.

B io stratin o m y has occupied th e discussion fo r three c h a p te rs because it tends
to be the m ost stud ied aspect o f tap h o n o m ic histories. T his is especially so for
zo o arch aeo lo g ists w ith in terests in h u m a n b eh av io rs because th o se b ehaviors
are generally b io stratin o m ic . T h u s we d ev o ted som e tim e to tw o o f th e m ost
im p o rta n t h om in id b io stratin o m ic processes in preceding c h ap ters, differential
tra n s p o rt o f skeletal p a rts (C h a p te r 7) a n d b u tch erin g (C h a p te r 8). In this
c h a p te r o th e r significant b io stra tin o m ic processes have been review ed.
B one w eath erin g is, generally, the m echanical an d chem ical d e te rio ra tio n o f
bo n e, m ainly in su b aerial o r surface contexts. Bones ten d to becom e m ore
w eath ered th ro u g h tim e, b u t the re la tio n o f h ow w eath ered th e bones in an
assem blage are to the fo rm a tio n a l h isto ry o f th a t assem blage is n eith er sim ple
n o r stra ig h tfo rw a rd . R o o t etching m ay occu r as bones lie o n the surface o f the
g ro u n d , a n d study o f such m o d ificatio n m ay help us m o re fully u n d e rsta n d the
significance o f b o n e w eathering. T ra m p lin g ten d s to disperse b ones h o riz o n
tally a n d vertically, to scratch them , a n d to fractu re them . B ones m ay be
a b ra d e d by several different m echanism s, including tram p lin g . B ones m ay be
b u rn e d to v arious ex ten ts a n d by several different m echanism s, an d m icro
o rg an ism s such as fungi, b ac te ria, a n d insects can all consum e skeletal tissue.
W hile all o f these b io stratin o m ic processes variously m odify bones, few
analyses w hich in teg ra te th em all have been p erfo rm ed . A s well, w hile m an y o f
these processes h ave been experim entally d o cu m en te d , th eir precise ta p h o n o
m ic significance is u n clea r because little has been w ritten a b o u t how a n d why
these v ariab les sh o u ld be re co rd e d o r q uantified in p re h isto ric contexts. In fact,
the d ev elo p m en t o f an aly tic techniques fo r co m p arin g a ttrib u te s o f bone
m o d ificatio n seem to have o u tp ace d o u r u n d e rsta n d in g o f w h a t th o se a ttri
b u tes signify re g ard in g ta p h o n o m ic histories. I re tu rn to this issue in C h a p te r
13. It is n o w tim e, how ever, to tu rn to th e process o f b u ria l a n d its effects on
v erte b rate fa u n al rem ains.
10
B URI AL AS A T A P H O N O M I C P R O C E S S
A sym m etrical bedding on either side o f the carcass is o f special significance, and we
m ust pay atten tio n to it w hen we study fossil m aterial.
(J. W eigelt 1927/1989:95)
Introduction
M o st fa u n al rem ain s p aleo b io lo g ists a n d zo o a rch a eo lo g ists stu d y are recov
ered fro m su b su rface contexts. A n im p o rta n t ta p h o n o m ic p ro b lem then,
co n cerns gainin g an u n d e rsta n d in g o f b u rial processes. F o r exam ple, h u m an s
do not seem to be the only biological agent th a t buries an im al p arts. Sexton
beetles (M icrophorus sp.) b u ry ro d e n t carcasses (M ilne a n d M ilne 1976),
fo sso rial ro d e n ts o ften die in th eir b u rro w s a n d th u s are alread y bu ried , an d
tram p lin g (C h a p te r 9) by an y n u m b er o f biological ag en ts can result in the
b u rial o f an im al rem ains. A n d there are n u m ero u s geological processes w hich
result in th e b u rial o f an im al carcasses a n d rem ains (B ehrensm eyer a n d H o o k
1992). T h e b u rial process is im p o rta n t because n o t only are bones a n d teeth
p laced in the sed im en tary m atrix in w hich they u n d erg o diagenetic processes
(C h a p te r 11), b u t they m ay be v ario u sly m oved, reo rien ted , b ro k e n , a n d /o r
ab ra d e d d u rin g b u rial by the ta p h o n o m ic agents o f b u rial (C h a p te r 6).
B urial as a ta p h o n o m ic process has u n d erg o n e little intensive stu d y relative
to b io stra tin o m ic processes th a t m odify v e rte b ra te rem ains. S traus (1990:261)
suggests th a t d ep o sitio n al a n d fo rm a tio n a l processes w ere initially ig n o red by
arch aeo lo g ists in fa v o r o f the m o re im m ediate an d in terestin g task o f building
cu ltu ra l ch ro n o lo g ies once the deep an tiq u ity o f hom inids w as established in
the last h a lf o f the n in eteen th century. R etallack (1988:342) indicates th a t in
p aleo n to lo g y th e p ercep tio n has, u n til recently, been th a t fossil c o n c en tratio n s
are th e result o f ca ta stro p h ic b u rial events such as flash floods w hich sim u lta
neously kill, d epo sit, a n d b u ry large n u m b ers o f carcasses a n d bones. C hanges
in this p ercep tio n d u rin g the early p a rt o f the tw en tieth c e n tu ry h ave resulted in
th e necessity o f a lterin g o u r view point fro m one o f conceiving o f fossiliferous
d ep o sits as sed im en tary u n its to conceiving o f th em as pedogenic un its in o rd e r
to u n d ersta n d fossilization processes. F o r exam ple, R etallack (1984, 1988:342)
no tes th a t it is q u ite unlikely th a t fossils em bedded in paleosols (fossil soils)
have been tra n s p o rte d far “ since a paleosol w ould be destro y ed beyond
reco g n itio n if tra n s p o rte d elsew here,” a n d th e degree o f dev elo p m en t o f
paleosols can be used as an in d icatio n o f the tim e span o f fossil assem blage
404
Burial as a taphonom ic process 405
fo rm a tio n a n d d ep o sitio n al hiatuses. A n im al rem ains m u st be d ep o sited to

b ecom e p a rt o f th e fossil record; once bu ried , fossilization processes (chem ical
a lte ra tio n , see the G lo ssary ) affect them . B ones a n d teeth, th en , begin as
sed im en tary clasts o r particles. W e are co n cern ed here w ith th eir b u rial, having
discussed th eir ac cu m u latio n a n d d ep o sitio n in earlier ch ap ters. W e re tu rn to
d iagenetic (o r pedogenic) processes in C h a p te r 11.
O f m a jo r significance fo r ta p h o n o m ists is the co n c ep t o f a taphonom ically
active zone, o r T A Z , p ro p o se d by tap h o n o m ists w ith in terests in m arine
m olluscs. D avies et al. (1989:208-209) define the T A Z as the “ se d im e n t-w a te r
in terface a n d the b io tu rb a te layer ju s t b en e ath it,” a n d suggest this is the
d ep o sitio n al a n d stra tig ra p h ic zone w here “ m o st ta p h o n o m ic loss is co n cen
tra te d ” (see P a rso n s a n d B rett 1990 fo r a n overview o f the T A Z an d the
in v erte b rate fossil record). A sim ilar terrestrial T A Z can be designated for
v erte b rate rem ains. T a p h o n o m ic processes such as ab rasio n , w eathering, an d
tra m p lin g o ccu r a t th e se d im e n t-a ir interface. S cavengers from b ac te ria to
h yenas to h u m an s variously m ove an d m odify an im al carcasses th a t are
d ep o sited on th e g ro u n d surface o r a t the se d im e n t-a ir interface. A s should be
clear fro m p receding ch a p te rs, su b seq u en t to the initial d ep o sitio n o f a bone
(u p o n a n a n im a l’s d ea th ), the so o n er th a t a bone is buried b en e ath the
se d im e n t-a ir interface fo r g o o d , the b e tte r preserved (m ore like it w as in life) it
will o ften be.
M an y o f the principles developed in studies o f the b u rial o f in v erte b rates also
ap p ly to v erteb rates, w h eth er th o se v erte b rates live in a q u a tic o r terrestrial
en v iro n m en ts. In this c h a p te r som e o f the m a jo r aspects o f b u rial are review ed,
w ith a focus o n the geological a ttrib u te s a n d co n tex t o f the fossil record. A ny
serious stu d e n t o f ta p h o n o m y should have a basic u n d e rsta n d in g o f g eom or-
phic processes a n d ideally will have a geologist in the field to study the
d ep o sitio n al co n tex ts o f the fossils being collected. As W o o d a n d Jo h n so n
(1978:315) n ote, the co n tex t a n d sp atial asso ciatio n s o f arch aeo lo g ical (and
zo o a rch aeo lo g ical a n d paleo n to lo g ical) rem ains is the “ fo u n d a tio n o f o u r
discipline. If we fail to record the co n tex t, o r if we m isread o r m isin terp ret th a t
co n tex t, p ro p e r arch aeo lo g ical [and tap h o n o m ic] in te rp re ta tio n is im p o ssib le.”
C lark an d K ietzke (1967:1 17) suggest there are at least “ six m a jo r facto rs o f
b u rial th a t can , th eo retically at least, result in a difference betw een the
[deposited] assem blage an d the [buried] fossil assem blage.” T hese are: (1) the
tim e in terv al betw een episodes o f sed im en tatio n ; (2) th e thickness o f sedim en
ta ry increm ents; (3) the velocity o f d ep o sitio n al forces in c o n ta c t w ith bones o r
corpses; (4) the n a tu re o f the sedim ent, such as the a m o u n t o f co m p ac tio n an d
g ra in size; (5) th e p o st-d ep o sitio n al ac tio n o f ro o ts a n d b u rro w in g anim als; an d
(6) th e p erm eab ility o f the sedim ent a n d chem ical n a tu re o f the p erm eatin g
so lu tio n s. In this c h a p te r I co n sid er basically the first fo u r, reserving discussion
o f th e last tw o fo r C h a p te r 11 as they largely co n cern diagenetic o r p o st-b u rial
processes.
Deposition and burial

I f burial is to be considered a significant aspect o f ta p h o n o m ic histories, th e n it
is im p o rta n t to d istingu ish tw o periods in the ta p h o n o m ic h isto ry o f faunal
rem ains. T h e deposition o f fa u n al rem ains refers to th eir d ynam ic placem ent
either on a land surface o r in an existing sed im en tary unit. D e p o sitio n is
sim u ltan eo u s w ith b u rial only in the la tte r case; an exam ple w ould be the
ca ta stro p h ic burial o f an anim al by volcanic ash o r the d e a th o f a fossorial
ro d e n t in its b u rro w . Burial refers to the covering o f fa u n al rem ains w ith
sedim ents, w hich can be eith er m ineralogical o r biological, such as san d o r
v eg etation , respectively. T he d istin c tio n betw een d ep o sitio n a n d b u rial m ay
seem p ed an tic, b u t it is n o t. T o o often one read s a b o u t “ p o st-d e p o sitio n a l”
ta p h o n o m ic processes th a t in fact refer to “ p o st-b u ria l” processes (e.g., K lein
an d C ru z-U rib e 1984:70-75). W hile it is tru e th a t p o st-b u rial processes are also
p o st-d ep o sitio n al ones, the reverse is n o t necessarily tru e. A b o n e can be
d ep o sited yet rem ain o n the g ro u n d surface a n d u n b u rie d fo r som e tim e; it is
this asp ect o f the tim in g o f b u ria l th a t allow s us to ta lk a b o u t such things as
bo ne w eath erin g in sub aerial co n tex ts an d , in p a rt, tra m p lin g (see C h a p te r 9).
T he d istin c tio n o f d ep o sitio n a n d b u rial is w h a t allow s us to distinguish
diagenetic (p o st-b u ria l) processes fro m p re -b u rial processes, a n d w ithin p re
b u rial processes we can d istinguish b io stra tin o m ic processes th a t o ccu r p rio r to
d ep o sitio n (such as bu tch erin g ) a n d those th a t occur afte r d ep o sitio n (such as
tram p ling ). T h e ta p h o n o m ic process betw een d ep o sitio n a n d diagenesis, then,
is b urial.
T he b u rial process m ay influence how deeply fau n al rem ains are b u ried as
well as th e types o f sedim ents in w hich they are buried. Som e diagenetic (p o st
bu rial ta p h o n o m ic) processes are d ep en d en t on the d ep th o f b u rial (e.g.,
sedim ent o v erb u rd en w eight). T h u s n o t only are b u rial a n d rates o f d ep o sitio n
im p o rta n t to a ta p h o n o m ic history, so to o are events o f erosion o r sedim ent
rem oval (H e n d erso n 1987). S edim ent is n o t sim ply layered over bones. This
m ay h ap p e n , b u t so m ig h t m o v em en t o f sed im en tary particles occu r d u rin g
sed im en tatio n such th a t sedim ents lay over bones only briefly p rio r to their
rem oval a n d rep lacem en t by o th e r sedim ents. B ones m ay becom e sedim entary
p articles o r clasts in som e settings. T h u s w hile I focus here on sedim ent
d ep o sitio n , the p o te n tia l th a t ero sio n a n d (re)exposure o f fau n al rem ains
occurs as p a rt o f the b u rial process sh o u ld be k e p t in m ind.
Sedimentation
T he ra te a n d m o d e o f sed im en tatio n seem to be the m a jo r variables th a t
influence th e co n ten t (the p a rtic u la r fossils represented) a n d stru c tu re (spatial
arra n g e m e n t o f indiv idu al fossils) o f the fossil record. S ed im en tatio n can be
slow o r it can be rap id ; sedim ents ca n be fluvially d ep o sited o r they can be
B urial as a taphonom ic process 407
T able 10.1 S ta n dard sedim ent size classes
Size class nam e Particle diam eter (mm)
B oulders >256
C obbles 64 to 256
Pebbles 2 to 64
very coarse 32 to 64
coarse 16 to 32
m edium 8 to 16
fine 4 to 8
very fine 2 to 4
Sand 1 16 to 2
very coarse 1 to 2
coarse 1/2 to 1
m edium 1/4 to 1/2
tine 1/8 to 1/4
very fine 1/16 to 1/8
Silt 1/256 to 1/16
Clay 1/4096 to 1/256
d ep o sited by w ind. E ach ten d s to p ro d u c e p a rtic u la r ta p h o n o m ic signatures.

W e th ere fo re need to begin o u r discussion o f sed im en tatio n w ith several basic
geological term s a n d concepts.
“Sedim en ts are p a rtic u la te m a tte r th a t has been tra n sp o rte d by som e process
fro m one lo catio n to a n o th e r” (Stein 1987:339). A geological deposit o r stratum
is a th ree-d im en sio n al u n it d istin g u ish ab le fro m o th er such u nits o n the basis o f
u n iq u e physical p ro p e rtie s, a n d is a n aggregate o f sed im en tary particles (Stein
1987:339, 344). G ra in size (T able 10.1) (grain size d istrib u tio n , g ra in o rien
ta tio n o r fabric, g ra in shape, a n d g ra in surface m arkings), co m p o sitio n
(m ineralogy), a n d stru c tu re (sm all-scale v aria tio n s in g ra in size, g rain shape,
co m p o sitio n , o r p o re space) o f sedim ents are influenced by the d ep o sitio n al
en v iro n m e n t (Stein 1987:357). S edim ent g ra in size helps ascertain the m ag n i
tu d e o f th e energy o f the sedim ent d ep o sitio n m echanism , a n d is m en tio n ed
below . T h e d ep o sitio n o f sedim ents is “ n o t an in d iscrim in ate process th a t can
h a p p e n a n y w h ere” (B u tzer 1982:44). S edim ent supply, n a tu re o f g ro u n d cover,
to p o g ra p h y , a n d o p erativ e g eo m o rp h ic processes influence if a n d w here
d ep o sitio n occurs (B utzer 1982:44).
B u tzer (1982:56-57) p resen ts a n o u tlin e o f d ep o sitio n al settings th a t includes
basic details re g ard in g the sedim ents a n d sed im en tary un its (facies) th a t result
in these settings. Som e o f his o b serv atio n s are sum m arized in T ab le 10.2. I
h asten to n o te this tab le is no su b stitu te fo r hav in g a geologist on the site d u rin g
recovery o f fossils. W h a t I in ten d by p resen tin g it is to p ro v id e the re ad er w ith a
general im p ressio n o f som e o f th e v a ria tio n in sed im en tary en v iro n m en ts in
w hich fossils m ig h t be fo u n d , a n d som e idea o f the kinds o f sed im en tary d a ta
T ab le 10.2 D epositional settings and attributes o f sedim ents and sedim entary
units (fro m B utzer 1982:56-57)
D epositional energy B edding; texture Sorting
Spring low to high lenticular heterogeneous, m ay p o o r to good

be co n to rted ; organic m uck,
sand, precipitates
K arst low p rim arily m assive, often m ainly p o o r
heterogeneous; hum ic loam s or
gravelly w ash, precipitates
Cave d etritus from ro o f fall, varied generally p o o r
fluvial, eolian
S eacoast highly variable thin to m assive, com plex generally good at
facies; clays to cobbles and below w aterm ark,
variable above
L akeshore low to m oderate th in to massive; clays to p o o r to good
and m arsh sands, organic m atte r
D elta edge variable extensive an d massive; m o d erate to good
clay to silt, som e sand
F loo d plain variable com plex vertical an d lateral m oderate to good
sequences; clays to gravels
Eolian low to m o d erate well stratified, massive; silt usually excellent
to sand
Slope variable thin to m assive, poorly generally p o o r
stratified; silt to rubble
V olcanic variable massive; mixed m oderate to excellent
th a t sh ou ld be re co rd ed (texture, size, a n d shape o f sed im en tary unit) if the

ta p h o n o m is t h o pes to gain som e u n d e rsta n d in g o f b u rial processes th a t m ight
h av e m odified the recovered fossils. T hese attrib u te s, a n d o th ers, are im p o rta n t
w hen it com es to u n ravelling th e ra te a n d m o d e o f sedim ent d ep o sitio n , w hich
are, afte r all, th e facto rs th a t resu lt in the b u rial o f fa u n al rem ains.
R a te o f sedim entation
K idw ell (1985, 1986) discusses th e ta p h o n o m ic significance o f the ra te o f
sed im en tatio n . H e r discussion centers a ro u n d the density (frequency p er unit
area) o f fossils in a geological d ep o sitio n al unit. T he density o f fossils is a
fu n ctio n o f the ra te o f fossil in p u t relative to the ra te o f sed im en tatio n ; fossils
occu r in dense c o n c e n tra tio n s if the ra te o f fossil in p u t increases relative to the
ra te o f sed im en tatio n , a n d fossils display a scattered , n o n -dense d istrib u tio n if
the ra te o f fossil in p u t decreases relative to the ra te o f sed im en tatio n . T h u s n o t
only is th e ra te o f sed im en tatio n im p o rta n t, b u t so to o is the ra te a t w hich
in d iv id u al fossils are in p u t to the geological reco rd . T h e ra te o f fossil in p u t is a
fu n ctio n o f th e ra te a t w hich fossils are supplied to the d ep o sitio n al locus an d
the ra te a t w hich th o se fossils are rem oved by v ario u s ta p h o n o m ic processes.
T h e sed im en tatio n ra te in tu rn is a fu n c tio n o f th e volum e o f sedim ent

dep o sited p er u n it o f tim e a n d the ra te o f erosion o r volum e o f sedim ent
rem oved p er un it o f time.
A ssum ing a c o n sta n t rate o f fossil in p u t, K idw ell (1985, 1986) distinguishes
fo u r types o f fossil c o n c en tratio n s. F irst, fossils m ay increase in density w ith
d ecreasing d ep th w ith the u p p e r term in a tio n o f the fossil co n c e n tra tio n at a
stra tig ra p h ic b o u n d a ry created by d isco n tin u ed d ep o sitio n o f sedim ent. This
d en o tes a decreased ra te o f sed im en tatio n u ntil it stops. S econd, fossils m ay
increase in density w ith decreasing d e p th w ith the u p p er term in atio n o f the
fossil c o n c e n tra tio n m ark e d by a n ero sio n a l surface. In this case, ero sio n has
rem oved sedim ent an d the co n c en trate d fossils occu r as a lag dep o sit left
b eh in d by ero sio n a l processes th a t rem oved sedim ents b u t n o t bones. In v ertin g
the first tw o scenarios, the th ird case finds fossils increasing in co n c en tratio n
w ith increasing d ep th ; the low er b o u n d a ry o f th e co n c e n tra tio n is m a rk e d by an
a b ru p t in itia tio n o f sed im en tatio n a n d increasingly h ig h er rates o f sedim en
ta tio n w ith decreasing d ep th . T he fo u rth scenario h as the low er b o u n d a ry o f
th e fossil c o n c e n tra tio n m a rk e d by a n ero sio n a l surface, densely p ac k ed fossils,
a n d fossil den sity decreases w ith decreasing d e p th as sed im en tatio n rates
increase. K idw ell (1986:11) suggests the first tw o types o f c o n c en tratio n s
consist o f fossils th a t are m o re a b ra d e d , frag m en ted , ero d ed , a n d w eathered
th a n th e fossils in th e o th e r tw o types d u e to the g re ater exposure d u ra tio n o f
fossils in th e fo rm e r tw o c o n c en tratio n s.
T a p h o n o m ists w o rk in g w ith fossil collections fro m w ithin the tem p o ra l
ran g e o f ra d io c a rb o n d atin g ca n o b ta in m u ltip le d ates on a fossil-rich s tra tu m
in o rd e r to e x tra p o la te a ra te o f sed im en tatio n a n d a ra te o f fossil deposition.
R etallack (1984:60) suggests th a t co m p arin g the degree o f d evelopm ent o f
fossil soils (paleosols) w ith the te m p o ra l d u ra tio n o f th e fo rm a tio n o f sim ilar
m o d ern soils “ m ay give an in d icatio n o f the d u ra tio n o f break s in d ep o sitio n
w ith in a [stratigraphic] sequence o f fossil soils” a n d th u s pro v id e estim ates o f
sedim ent a c cu m u latio n rates in sequences o f m ultiple paleosols. In c o n ju n ctio n
w ith K idw ell’s (1985, 1986) m odels, the ta p h o n o m ist m ay be able to use such
d ep o sitio n al rates to help distinguish passive m ass a c cu m u latio n s th a t form ed
over lo n g tim e spans in a n area experiencing low sed im en tatio n rates fro m
active m ass accu m u latio n s th a t fo rm ed over sh o rt tim e p erio d s such as are
fo u n d in v arious m ass kill sites (e.g., O lsen 1988; T o d d 1987a; see C h a p te r 8 for
discussion o f b o n e a ccu m u latio n ). K n o w in g the ra te o f sed im en tatio n m ay also
help u n rav el the ta p h o n o m ic significance o f b o n e w eath erin g profiles based on
specim ens fro m th ick stra tig ra p h ic un its (see C h a p te r 9).
M o d e o f sedim entation
G eo lo g ical sedim ents can be dep o sited by a n u m b e r o f geological processes.
T he tw o m o st stu died in the co n tex t o f ta p h o n o m y are fluvial a n d eolian
d ep o sitio n (S au n d ers 1977:68). M o st studies o f d ep o sitio n al co n tex ts in the
service o f p aleo n to lo g ical ta p h o n o m y have involved floodplains, deltas, and

river ch an n els (e.g., B ehrensm eyer 1975a, 1975b; C lark et al. 1967; V oorhies
1969); th o se con cerned w ith arch aeo lo g ical ta p h o n o m y have been focused on
differences betw een b o n e assem blages d ep o sited in on-site a n d off-site
lo catio n s (e.g., B unn et al. 1991). T here are, o f course, o th e r processes and
settings o f sed im en tatio n . In the follow ing I co n sid er som e o f these, citing
exam ples o f m an y o f th em along w ith som e o f the key a ttrib u te s tap h o n o m ists
use to help u n d e rsta n d b u rial processes. As will becom e clear, o ften if one can
identify the ag ent o f b o n e ac cu m u latio n (see C h a p te r 6), clues to the n a tu re o f
the b u rial process will be suggested.
Fluvial deposition
H yd ro lo g ical processes are a m a jo r fa c to r in th e fo rm a tio n o f m an y a rc h a e o lo
gical sites (Schiffer 1987:243-256). F luvial d ep o sitio n in ch annels has been
stu d ied by B ehrensm eyer (1988b, 1990 a n d references therein). She n o tes th a t a
ra p id ra te o f fluvial sed im en tatio n ca n be m ark e d by sedim ents o f m ixed
tex tu re (fine to co arse sed im en tary particles) w hereas slow fluvial sedim en
ta tio n rates will be m ark e d by sedim ents o f fine tex tu re, a lth o u g h fine sedim ents
m ay also be d ep o sited relatively ra p id ly (B ehrensm eyer 1988b). F a u n a l
rem ains “ m ay be o v erta k en by m oving b ed fo rm s (ripples, sand w aves), an d
sco u r o n th eir d o w n stre am [or leew ard] side also p ro m o te s b u ria l” (B ehrens
m eyer 1990:235). T his is sim ilar to W eigelt’s (1927/1989:95) o b serv atio n over
60 years ago: a sizable carcass “ restricts the flow o f onco m in g w ater, causing an
increase in its velocity, w hich in tu rn m eans increased force; consequently, the
w ater o n th e sh o rew ard side o f the carcass can dig o u t a deep hollow , while
sedim ent accu m u lates o n th e o th e r side. T his often causes the carcass to lie
c ro o k ed o r to sink in .”
F lu v ial ac tio n ca n b u ry bones, a n d it ca n also m ove o r tra n s p o rt them
(C h a p te r 6). B ones n o t tra n s p o rte d by fluvial actio n ten d to be ro b u st, heavy
specim ens th a t m ay be v ariously a b ra d e d a n d b ro k e n (see also F ig u re 6.5),
p artic u la rly w hen sedim ents are coarse. H igh-energy en v iro n m en ts o f fluvial
d ep o sitio n such as ch annel fills a n d lag deposits tend to have high ra tio s o f
teeth -to -v erteb rae w hereas low -energy en v iro n m en ts such as d eltaic an d
lacu strin e settings ten d to have low ra tio s (B ehrensm eyer 1975b). A b ra sio n and
the fluvial tra n sp o rta b ility o f b ones are th e m a jo r a ttrib u te s tap h o n o m ists
exam ine in fluvial d ep o sitio n al environm ents.
B oaz (1982) p ro v ides an extensive overview a n d results o f experim ents w ith
fluvial processes. C alling on the w ork o f V oorhies (1969), B ehrensm eyer
(1975b). K o rth (1979), H a n so n (1980), an d o th ers, she recorded the follow ing
variables in h er stu d y o f fluvially tra n sp o rte d a n d b u ried bones: sedim entary
en v iro n m e n t, tax o n o m ic co m p o sitio n a n d relative ab u n d a n ces o f tax a in the
assem blage, d em o g ra p h y o f the fossil p o p u la tio n , degree o f a rtic u la tio n ,
degree o f sp atial asso ciatio n a n d dispersal o f skeletal p arts, w eathering,
skeletal p a rt frequencies an d th eir re la tio n to stru c tu ra l density a n d tra n s p o r ta

bility, h o riz o n ta l a n d vertical d istrib u tio n s o f skeletal p a rts, o rie n ta tio n o f long
axes o f specim ens, a n d ex ten t a n d types o f d am ag e such as ab ra sio n a n d
ca rn iv o re gnaw ing. She notes th a t fluvial tra n s p o rt a n d b u rial can be recog
nized by p a tte rn e d o rie n ta tio n s o f bones (F ig u re 6.7), the spatially restricted
occu rren ce o f lag dep osits (F ig u re 6.5), a n d a b ra sio n causing the exposure o f
trab e cu lae w ith san d a n d gravel em bedded in fossae a n d trabeculae. B oaz
(1982:219) no tes th e low er the stru c tu ra l density a n d the g re ater the tra n s p o rt
d istan ce o f a skeletal p a rt, the low er the p ro b a b ility th a t the p a rt will survive the
rigors o f fluvial tra n sp o rt.
Eolian deposition
E o lian sedim ents ten d to be fine, a n d m o st ta p h o n o m ic stu d y o f th em concerns
the ab rasiv e effects o f w in d -b o rn e sed im en tary particles o n bones (see C h a p te r
9). E o lian processes, like hy d ro lo g ical ones, are often im p o rta n t c o n trib u to rs
to th e fo rm a tio n a l h isto ry o f archaeological sites (S chiller 1987:238-243).
W ind n o t only deposits sedim ents, it rem oves them a n d creates lag deposits o f
fossils. It is in th e creatio n o f lag d ep o sits th a t w ind m oves bones, an d this
m o vem en t is largely d o w n w a rd . Som e d o w n slo p e m o v em en t m ight result from
w ind activity an d en tail h o riz o n ta l as well as vertical m ovem ent, b u t this is
surely grav ity aided. I suspect th a t typically only very sm all, light bones are
m oved significant distances by w ind actio n , a lth o u g h S chafer (1962/1972:37)
re p o rts seeing “ v erteb rae o f Phoca [a seal] being carried by the w ind, racing an d
ju m p in g over the slightly salt-en cru sted surface o f the [flat sand beach].”
H o m in id deposition
H o m in id s, inclu ding som e early a n a to m ica l form s o f m o d ern H o m o sapiens,
seem to be u n iq u e in th e an im al k in g d o m fo r ritu alistic disposal o f conspecifics.
O ften, such disp osal included burial. P u rp o sefu l b u rial is m ark e d geologically
by th e o ccu rren ce o f th e h u m a n rem ains in a stratig ra p h ic ally d istin ct unit,
u sually a p it feature. W h a t a p p e a r to have been pets o f p re h isto ric h u m an s,
such as dogs ( C anisfam iliaris), w ere also som etim es p u rposefully b u ried in pits.
In ten tio n a l b u rial, th en , seems to be m ark e d by intrusive pits co n tain in g anim al
(including h u m an ) rem ains. T h e d isp o sal o f fo o d rem ains by h o m in id s seems to
be a typical m o d e o f ad d in g v e rte b ra te skeletal p a rts to the fu tu re zo o a rch a eo -
logical reco rd . Such a d d itio n s are p erh ap s m o re o ften b u ried by n a tu ra l
sed im en tatio n processes such as eolian a n d fluvial d ep o sitio n th a n by hom in id
activities.
H o m in id s “ d eterm in e who is b u ried a n d when b u t also how a n d where”
(H e n d erso n 1987:49). As n o ted earlier in this c h a p te r an d in C h a p te r 11,
extrinsic facto rs such as the chem istry a n d p o ro sity o f sedim ents can influence
b o n e p reserv atio n . D id p re h isto ric peoples dispose o f th eir d ead in well-
d ra in e d sedim ents o r in sw am ps? T h a t sim ple difference can result in m ajo r
p re serv atio n al v aria tio n . A s well, crem atio n , m a n u a l defleshing, ex posure o f

bodies to the elem ents for a p erio d o f tim e p rio r to b u rial, in term en t in a coffin
o r c o n ta in e r o f som e so rt, a n d o th e r p re -b u rial facto rs all influence how well
h u m a n rem ains are preserved (H e n d erso n 1987). Such facto rs should, o f
course, be con sidered w hen dealing w ith the hom in id d ep o sitio n an d b u rial o f
any fa u n al rem ains.
Fossorial animals
B u rro w ing o r fossorial anim als, such as m an y ro d e n t tax a, som etim es die o f
n a tu ra l causes in th eir b u rro w s. T he o ccurrence o f an essentially com plete
sk eleton o f a fossorial ro d e n t, th a t is to som e degree artic u la te d , a n d in a
b u rro w (w hen filled w ith sedim ent, it is term ed a krotovina) is typically inferred
to re p resen t such a d e a th an d n a tu ra l burial. O ne m u st th erefo re be well aw are
o f th e geological co n tex t o f the rem ains a n d the sp atial relatio n sh ip s o f
in d iv id u al bones.
O ther burial processes

O ne o f th e m o re u niq u e, if n o t ra re, types o f b u rial concerns the e n tra p m e n t o f
an im als in bogs, m arsh es, springs, o r ta r pits a n d th eir su b sequent d e a th and
b u rial via th e carcass sinking in to the sedim ent. T h e L a B rea ta r pits are one
fam o us exam ple (e.g., Stock 1956). V arious spring sites in N o rth A m erica have
p ro d u c ed m an y w ell-preserved fossils o f anim als th a t becam e m ired in the
sedim ent a n d eventually buried. T h e 26,000 year old H o t S prings M a m m o th
site in S o u th D a k o ta represents a fossil assem blage form ed w ithin “ a springfed
p o n d w ithin a k a rst d ep ressio n ” (A g e n b ro ad 1984:119). M o st o f the rem ains
belong to the C o lu m b ian m a m m o th (M a m m u th u s columbi). T ra n s p o rta tio n o f
skeletal elem ents w as m inim al, a n d seem s to have involved m ostly g ravity an d
p erh ap s b lo ated carcasses floating in the w a te r w ithin the k a rst (A g en b ro ad
1984. 1989, 1990). T h e m ajo rity o f the rem ains are c o n c en trate d a ro u n d the
edges o f th e k arst, suggesting th e anim als fell in to the steep-sided depression,
co uld n o t clim b o u t, a n d eventually died a t the base o f the slope. N early
com p lete skeletons are co m m o n in th e deposits, b u t th e ex ten t o f a rtic u la tio n
varies. T he H o t S prings site ap p e a rs to rep resen t a ttritio n a l m o rta lity an d
g ra d u al a c cu m u latio n o f skeletons o f m ostly im m a tu re individuals, p erh ap s
b ac h elo r m ales (A g e n b ro ad 1990). C arn iv o re d am ag e to the bones is m inim al,
b u t som e o f th e b o nes have been fractu re d by tra m p lin g from o th e r en tra p p ed
m am m o th s, slope collapse a n d rock fall o n to bones, a n d p o st-b u ria l sedim ent
co m p ac tio n an d slippage (A g e n b ro ad 1989).
T h e ca. 16,000 y ear old B oney S prings site in M issouri co n tain s the rem ains
o f 31 m a s to d o n ts (M a m m u t am ericanum ) th a t ap p a re n tly died o f d ro u g h t an d
n u tritio n a l deficiency a ro u n d th e o u tlet o f an arte sia n spring (S au n d ers 1977).
T he d em o g ra p h y o f the d ead anim als seem s to re p resen t an in stan ce o f
c a ta stro p h ic m o rtality . T he long axis o f m o st b o n e specim ens w as parallel to
the h o riz o n ta l b ed d in g p lane (show ed n o plunge) a n d no p referred com pass

o rie n ta tio n was ap p a re n t. B ones rested directly on each o th e r a n d o ccurred in
“ n early circu lar p lan s o f co n c e n tra tio n , dim inishing in d iam eter w ith each
successively d eep er level” (S au n d ers 1977:73). O nly th ree instances o f a rtic u
lated b ones w ere n o te d in the sam ple o f 517 elem ents. Sm all elem ents o f low
s tru c tu ra l density w ere less a b u n d a n t th a n large elem ents o f g re at stru c tu ra l
density. In situ d ec o m p o sitio n o f soft tissues is suggested by the presence o f
rem ains o f carrio n -feed in g co leo p te ro u s insects. P a tte rn e d fractu rin g o f bones
is a ttrib u te d to “ lith o static p re ssu re” (S au n d ers 1977:104), w eathering, an d
tram p lin g . S ed im en tatio n w as som etim es g ra d u al, som etim es ra p id , a n d was
in te rm itte n t as in d icated by bo n e w eathering. C overing sedim ent o rig in ated
fro m flood events o f a n ad jac en t river (S au n d ers 1977:109).
Burial processes
T here are o th e r aspects o f b u rial th a t w a rra n t com m ent. W as the b u rial o f the
fossil assem blage u n d e r study d ynam ic a n d sh o rt-term , o r w as it g ra d u a l an d
long-term ? W h a t k in ds o f specific d ep o sitio n al processes lead to burial? A fter
all, a skeletal p a rt m u st m inim ally be dep o sited on the g ro u n d surface in o rd e r
fo r it to becom e buried. Som e o f these issues are discussed in this section.
T ra m p lin g as a process o f b u rial is discussed in C h a p te r 9.
K ra n z (1974a, 1974b) follow s B ro n g ersm a-S an d ers (1957) a n d defines an
anastrophe as a c a ta stro p h e o f lim ited scope a n d a rea th a t p ro d u ces m ass
m o rta lity in the affected area. A n a n a stro p h ic b u rial event is th u s a c a ta
stro p h ic, sh o rt-term , b u rial event p o ten tia lly p ro d u c in g w h a t is typically
term ed a c a ta stro p h ic m o rtality profile (C h a p te r 5), b u t this m o rtality p a tte rn
m ay n o t be realized. As K ra n z (1974b) em phasizes, several facto rs o th e r th a n
extrem ely ra p id d ep o sitio n play a role. In p a rtic u la r, the o rg an ism being b uried
m ay h ave a chan ce to escape the en to m b in g sedim ents. T he life h ab its a n d size
o f th e o rganism s, the type o f sedim ent cover, a n d the d ep th o f b u rial all
influence w h ether an ind ividual org an ism m ay escape (K ran z 1974b).
M o st arch aeo lo g ists are aw are o f th e site o f P om peii th a t w as an a stro p h i-
cally b u ried in A .D . 79 by the volcanic e ru p tio n o f M t. V esuvius. L ym an
(1989b) a n d V oorhies (1981) re p o rt on sep a rate instances o f a n a stro p h ic burial
o f large v erteb rates resu lting fro m volcanic eru p tio n s. Such an a stro p h ic ally
b uried assem blages o f an im al rem ains tend to display m inim al carnivore
gnaw ing w hen com p lete carcasses are b u ried , a n d the carcasses are relatively
co m plete an d artic u la te d . T hese o b serv atio n s on ra p id burial sh o rtly afte r the
orig in al d ep o sitio n o f a carcass o r bones tend to confirm m o st ta p h o n o m is ts’
belief th a t such b u rial effectively buffers fa u n al rem ains from the m yriad
b io stratin o m ic processes th a t m ight otherw ise m odify them . T his does not,
how ever, m ean th a t ta p h o n o m ic processes sto p m odifying b ones u p o n their
b u rial (e.g., see the discussion o f bone w eathering in C h a p te r 9). R ath er, the
diagenetic processes th a t m odify bones afte r b u rial are sim ply different from ,
an d ten d to have less m acro sco p ic affects on the bones th a n b io stratin o m ic
processes, as we see in C h a p te r 11.
A n a stro p h ic b u rial by ro o f fall events in caves can result in the fractu re o f
b ones (e.g., T h o m as a n d M ay e r 1983) a n d p e rh a p s th eir scarification (e.g.,
D ix o n 1984). C aves a n d ro ck sh elters are u n u su al d ep o sitio n al env iro n m en ts
b ecause th ey are spatially b o u n d ed , a n d as such S trau s (1990:273) refers to
them as “ b o n e b o x es.” In such settings, d ep o sitio n an d p reserv atio n tend to
o u tstrip ero sio n a n d lack o f p re serv atio n due to the b o u n d e d sp atial u n it
defined by a cave o r ro c k sh elte r a n d the d ep o sitio n al en v iro n m e n t being
p ro tec ted fro m clim atic facto rs in p a rtic u la r. T his does n o t, how ever, m ean
th a t caves are the best places to recover fa u n al rem ains th a t have u n d erg o n e
m in im al ta p h o n o m ic m o d ificatio n resulting fro m b io stratin o m ic , b u rial, o r
d iagenetic processes (see S trau s 1990 fo r a m o re com plete discussion).
Sites w here an im als fell o r ju m p e d to th eir d ea th s m ay have the first-
d ep o sited carcasses b u ried by carcasses d ep o sited m o m en ts later (e.g., W h e at
1972). T h e d istin c tio n o f a fall fro m a ju m p is an im p o rta n t one. “ A ju m p is an
in ten tio n a l leaping ac tio n w hich cu lm in ates in an a tte m p te d in n ate an d
p red ictab le lan d in g p o sitio n . A f a l l is a descent w ith a ra n d o m im p act p o s itio n ”
(H ughes 1986:55). In h er stu d y o f carcasses o f w apiti (Cervus elaphus) th a t
ju m p e d off a 20 m high cliff, H ughes (1986) argues th a t ju m p s from high ledges
will result in b ro k e n fro n t legs, lu m b ar v erteb rae, a n d ribs, an d bruised faces;
falls will p ro d u c e less p a tte rn in g in the lo catio n o f fractu re s a n d injuries.
B ehrensm eyer a n d H o o k (1992) describe som e very general a ttrib u te s o f
fossil d epo sits in 34 different kinds o f d ep o sitio n al env iro n m en ts. M an y o f
these p arallel th o se n o te d o r im plied in T ab le 10.2. F o r exam ple, B ehrensm eyer
an d H o o k (1992:21) describe fo u r kinds o f “ c o a sta l” s e ttin g s -o ffs h o re , beach,
lago o n , estu ary - a n d n o te th a t v erte b rate fossils seem to be rare in the first two,
co m m o n in the th ird , are “ p re se n t” in the fo u rth , a n d all consist o f rarely
artic u la te d rem ains o f m ainly a q u a tic taxa. A t a very general level, th eir schem e
is useful fo r co m p arin g fossil assem blages fro m sim ilar d ep o sitio n al contexts.
H ow ever, they list “ arch aeo lo g ical sites” as o ne o f the 34 kinds o f co n tex ts, and
note th a t v erte b rate fossils in such co n tex ts are “ very co m m on, [consist of]
w hole a n d b ro k e n p a rts, [and represent] allo c h th o n o u s [assem blages]” (B eh
rensm eyer an d H o o k 1992:61). T his ch a rac te rizatio n is far to o sim plistic to
describe th e rang e o f v a ria tio n in bo n e assem blages fro m different kinds o f
archaeolo gical sites o r from different d ep o sitio n al co n tex ts (e.g., stra ta ,
features) w ithin a site. B ehrensm eyer an d H o o k 's (1992) schem e is nonetheless
suggestive. It indicates th a t zo o a rch a eo lo g ists should be co g n iza n t o f w hen
they co m p are, say, an assem blage o f bones from a kill site w ith one from a
h a b ita tio n site, o r an assem blage from a cave w ith an assem blage from an open
site, given differences in the n a tu ra l a n d cu ltu ra l d ep o sitio n al a n d b u rial
histo ries o f the tw o. P erh a p s we should follow B ehrensm eyer a n d H o o k ’s lead
an d develop a list o f kinds o f arch aeo lo g ical d ep o sitio n al co n tex ts in an

a tte m p t to n o te w ith in -co n tex t ta p h o n o m ic sim ilarities a n d betw een-context
ta p h o n o m ic differences o f fa u n a l rem ains.
Spatial distribution o f faunal remains

S tu dy o f th e sp atial d istrib u tio n o f v erte b rate fa u n al rem ains trad itio n ally w as
a m a tte r o f n o tin g w hich tax a occur in p a rtic u la r stra ta , p erh ap s w ith their
ab u n d a n ce relative to o th e r tax a also being n oted. T his vertical perspective w as
su pp lem en ted in the 1960s a n d 1970s co in cid en t w ith the shift in archaeological
focus fro m ch ro n o lo g ical issues to sy nchronic v ariab ility w ithin archaeological
cultures. Z o o arch ae o lo g ists b egan to study v a ria tio n in the h o riz o n ta l d istrib u
tio n s o f tax a a n d skeletal p a rts, positing such things as k ill-red istrib u tio n or
sh arin g (L y m an 1980), c o v a ria tio n in the kin d o f bo n e refuse d ep o sited a n d the
tim e o f site a b a n d o n m e n t (P ozorski 1979), a n d co v a riatio n o f dep o sitio n al
co n tex ts a n d differential p re serv atio n (M ead o w 1978). A lm ost im m ediately
e th n o arch ae o lo g ic al studies w ere published w hich suggested c a u tio n w as due
w hen in terp re tin g the h o rizo n ta l d istrib u tio n s o f faunal rem ains (e.g., B inford
1978; K e n t 1981), a lth o u g h som e m o re recent n eo ta p h o n o m ic research in d i
cates these early in te rp re ta tio n s m ay well have been c o rrec t (e.g., B a rtra m et al.
1991; M arsh all 1993).
S patial d istrib u tio n s o f fa u n al rem ains have n o t been studied w ith the sam e
intensity th a t skeletal p a rt frequencies a n d b u tch erin g m a rk s have been.
A rch aeo lo g ists regu larly reco rd detailed d a ta co n cern in g the sp atial locations
an d asso ciatio n s o f th e fau n al rem ains they recover, a n d virtually every
in tro d u c to ry te x tb o o k on arch aeo lo g ical m eth o d s a n d tech n iq u es devotes a
c h a p te r o r tw o to basic field recovery an d m ap p in g techniques. D etailed spatial
d a ta are now reg u larly recorded in p aleo n to lo g y (e.g.. A b ler 1984). G iven th a t
b u rial processes d eterm in e the final sp atial p o sitio n o f v e rte b ra te rem ains it is
ra th e r surp risin g th a t these processes have seen so little study. P erh a p s the
d e a rth o f such research is, as Y ellen (1 9 9 lb : 154) notes, a fu n c tio n o f the fact
th a t the lo n g -term stu d y o f p o st-d ep o sitio n al a n d p o st-b u rial processes is
“ rarely p o ssible.” O r p erh ap s the lack o f research on the sp atial d istrib u tio n o f
bones a n d teeth is due to a generally sh ared p ercep tio n th a t “ it is likely th at
m uch o f th e [spatial] p a tte rn in g evident in bo n e refuse can only be u n d ersto o d
in term s o f featu res th a t are unlikely to be archaeologically d etec tab le”
(B artram et al. 1991:143). W h atev er the case, as I have no ted th ro u g h o u t this
volum e, sp atial a n d c o n tex tu a l d a ta are im p o rta n t to m an y analytical
p ro blem s, a n d m ay becom e im p o rta n t to new ones as the volum e o f n e o ta p h o
n om ic d a ta increases. In p a rtic u la r, stu d y o f the h o riz o n ta l a n d vertical
d istrib u tio n s, o rie n ta tio n s, a n d asso ciatio n s o f v e rte b ra te fa u n al rem ains m ay
tell us m u ch a b o u t the processes responsible fo r th eir burial. As n o ted in o th er
ch a p te rs, sp atial d a ta in fo rm us a b o u t processes o f dispersal a n d ac cu m u latio n
(e.g., C h a p te r 6), a n d o ften these processes (e.g., fluvial ac tio n ) are also burial
processes.
Summary
It is an inviolable rule th a t collections from different areas an d levels m ust be kept

separate.
(I. W. C ornw all 1956:241)
I have review ed som e o f the m a jo r geological processes a n d effects o f b u rial in

this c h a p te r. W h a t I h av e fo u n d tru ly am azin g is th a t d espite the fact th a t while
fau n al rem ain s typically com e to us in bu ried fo rm , th ere is very little w ritten
a b o u t th e b u rial process itself. T he brevity o f this c h a p te r is a g ood in d icatio n o f
th a t d e p a u p e ra te literatu re; th ere w as sim ply little on the ta p h o n o m ic effects o f
th e b u ria l o f v erte b rates fo r m e to sum m arize here. G iven the geological m ode
o f o ccurrence o f fossils, the m echanism s by w hich fau n al rem ains a tta in a
geological m ode o f occurrence w ould seem to w a rra n t m u ch m o re discussion
th a n they have received by zo o a rch a eo lo g ists in p artic u la r.
A nd rew s (1992:34) argues th a t ta p h o n o m ic p ro b lem s can be divided “ into
tw o issues, th e n a tu re o f th e physical en v iro n m e n t o r sedim ents in w hich anim al
rem ains are fo u n d , a n d the n a tu re o f the fossil assem blage itself.” T he la tte r has
received m uch atte n tio n . P erh a p s the lack o f lite ratu re o n th e fo rm e r results
fro m its less th a n readily visible n atu re ; th a t is, w hile we h ave a m odel o f a living
skeleton to w hich we can co m p are the fossils we find to d eterm in e how they
differ from living b ones, no such m odel is available fo r the sedim entological
reco rd . T he geological co n tex t o f fossils is stu d ied largely by sedim entological
analyses focusing o n processes o f sedim ent ac cu m u latio n , dispersal, an d
diagenesis (A ndrew s 1992:35-36). T h o ro u g h tap h o n o m ic analysis, then,
d em an d s c o n su lta tio n w ith a sedim entologist a n d /o r geologist.
T he b u rial o f an im al rem ains can be studied by ex a m in a tio n o f the geological
con tex t o f the fossils. R apid burial sh o rtly afte r anim al d ea th m ay result in
so rtin g (see th e G lo ssary ) o f the rem ains, b u t it also rem oves the rem ains from
b io stratin o m ic processes m an y o f w hich ten d to m odify b ones a t m acroscopic
scales. T h u s, relatively com plete, a rtic u la te d , a n d largely unm odified skeletons
will result. T he lo n g er the tim e span betw een d ep o sitio n o f the fa u n al rem ains
an d th eir bu rial, th e g re ater the chance th a t th o se rem ains will be m odified by
b io stratin o m ic processes. O nce an im al rem ains are bu ried , diagenetic p ro
cesses tak e over. S edim ent chem istry, p o ro sity , an d w eight are im p o rta n t
ta p h o n o m ic factors, a n d are co n tro lled in p a rt by th eir original d ep o sitio n rate
a n d m ode. It is to these topics th a t we tu rn in the next ch ap ter.
11
DIAGENESIS
T he n a tu re o f the bones, th a t o f the soil, its dryness o r hum idity, its perm eability by
air and w ater, the m ore o r less ancient d ate o f burial, the d ep th a t w hich they lie,
have a considerable effect o n the co n d itio n o f the bones.
(N . Joly 1887:88)
Introduction
S ed im en tary p etro lo g ists define diagenesis as the “ a lte ra tio n o f sedim ents after
d e p o sitio n ” (R etallack 1990:129). It is, how ever, som etim es tak en to m ean
only " a lte ra tio n afte r b u ria l” (R etallack 1990:129). T h ro u g h o u t this volum e I
tak e diagenesis to h ave the la tte r m eaning fo r v erte b rate fa u n al rem ains. T he
im p o rtan ce o f this d istin c tio n resides in the m any p o st-d ep o sitio n al and p re
b u rial ta p h o n o m ic processes th a t can m odify v erte b rate rem ains, alo n g w ith
m o dification s resultin g fro m the b u rial process itself.
O nce an im al rem ain s have been buried, a n u m b e r o f ta p h o n o m ic processes
can act o n them . Som e o f the m o re fam iliar ones are m in era liza tio n a n d
d efo rm atio n . In this c h a p te r, these an d o th e r diagenetic processes are d is
cussed. D iagenesis o f skeletal tissues is affected by intrinsic fa c to r s o f the tissue
specim en, such as its size, p o ro sity , chem ical a n d m o lecu lar stru ctu re , an d by
extrinsic fa c to r s such as sedim ent p H , w a te r a n d te m p e ra tu re regim es, an d
b acterial ac tio n (V on E ndt an d O rtn e r 1984). In trin sic facto rs such as
hydrolysis o f collagen by bo n e w a te r m ay ex acerb ate o r buffer extrinsic factors.
T he p o st-b u rial h isto ry o f an im al rem ains involves th eir p re serv atio n as fossils
(e.g., S ch o p f 1975), a n d th eir chem ical a n d m echanical a lte ra tio n o r d e stru c
tion. Z o o arch ae o lo g ists have explicitly recognized the p o te n tia l effects o f
d iagenetic processes fo r at least tw o decades (e.g., C h ap lin 1971), a n d m o st
zo o a rch aeo lo g ical collections have u n d erg o n e at least som e diagenetic
m o dification.
M artill (1990:285-287) suggests th a t diagenesis as a process can be co n v e
niently con sidered as tak in g place in th ree “ en v iro n m e n ts:” the bo n e tissue
itself, the p o re spaces an d cavities w ithin bones, an d the sedim ent su rro u n d in g
the bone. In the first, ionic su b stitu tio n o f a p a tite m inerals w ith o th e r m inerals
regu larly occurs. In th e second, p o re spaces a n d trab e cu lae are filled w ith
diagenetic m inerals; c o m p ariso n o f these w ith th o se o f the su rro u n d in g
sed im en tary m atrix m ay in d icate tra n s p o rt p rio r to final burial o f the bone if
the tw o sets o f m inerals differ. I f the tw o sets o f m inerals differ, the process is
417
som etim es referred to as “ p re-fo ssilizatio n ” (M a rtill 1990:288). F inally, the

sed im en tary m atrix in w hich v e rte b ra te rem ains are b u ried affects (a) the o u ter
surface o f th e bo n e, (b) th e n a tu re o f g ro u n d w ater (b o th its chem ical m ak eu p
a n d tra n sp o rta b ility ), a n d (c) w h eth er b u ried v erte b rate rem ains are crushed or
d efo rm ed based o n th e presence o r absence o f diagenetic cem ents in the
sedim ent.
R olfe a n d B rett (1969:233) a rra n g e diagenesis into three phases. T he first,
syndiagenesis, involves b ac te ria in shallow sedim ents m etab o lizin g organic
m a tte r in th e skeletal tissue. Anadiagenesis “ is the deep b u rial phase o f
co m p ac tio n a n d ce m e n tatio n . . . in o rg an ic chem ical reactio n s p re d o m in a te ”
(R olfe a n d B rett 1969:233). F inally, epidiagenesis can resu lt in the d isa p p e a r
ance o f th e fossil itself leaving only a m old; this phase seem s to be o f little
co n cern to m ost zo o a rch a eo lo g ists a n d is n o t d ealt w ith fu rth e r here. D istin c
tio n o f th e first tw o phases m ay also be o f little co n cern to zo o arch aeo lo g ists,
but I suspect w ould becom e m ore im p o rta n t as the age o f the fossil assem blage
increases.
O ne m ig h t th in k o f diagenesis as being m odeled by the e q u a tio n
D = f (M , C, D , S, T) [11.1]
in w hich D is the sum o f th e diagenetic processes influencing a p a rtic u la r fossil,

M is th e orig in al physical a n d chem ical co m p o sitio n o f the m aterial o f concern
(see C h a p te r 4), C is th e clim ate o f d ep o sitio n , D is the depositional m ode, S is
the n a tu re o f the sedim ent in w hich the m aterial is buried, a n d T is the tim e span
o ver w hich a specim en is buried. I briefly co n sid er each v ariab le in tu rn .
T he p o ro sity o f the skeletal tissue can greatly influence diagenetic chem ical
changes in th o se tissues. T he “ high density a n d low p o ro sity o f enam el m akes it
a n attra c tiv e c a n d id a te fo r m inim al [chem ical diagenetic] a lte ra tio n ” (Sillen
1989:212). T h e p o ro sity o f d en tin e m ak es it fairly susceptible to such alte ra tio n
(C arlso n 1990:545). T h e g re ater the p o ro sity o f bone tissue the g re ater the
susceptibility o f th a t tissue to diagenetic change a n d the m ore ra p id the
chem ical io n exchange betw een the tissue a n d the sedim entary m atrix in w hich
it is em b ed d ed (H a n so n a n d B u ik stra 1987:552).
C lim ate m ay have stro n g diagenetic influences, as in the c o n tra st betw een
arctic, tem p erate, tro p ica l, a n d d esert en v iro n m en ts. B eing frozen, altern ately
w et a n d dry, o r co n stan tly w et o r dry, all have m a jo r effects on the ra te an d
types o f chem ical reactio n s th a t can tak e place. F o r exam ple, chem ical
reaction s d o u b le in ra te fo r every 10°C increase in te m p e ra tu re (e.g., V on E n d t
an d O rtn e r 1984). Sillen (1989:220) suggests the ac tio n o f m icro-organism s
such as fungi, k n o w n to a tta c k collagen, is less p ro n o u n c e d in d ry e n v iro n
m ents th a n in m oist en v iro n m en ts. W here a bo n e is dep o sited influences how
quickly it will be b u ried afte r d ep o sitio n , a n d bones th a t are w eathered
su b aerially p ro b a b ly resp o n d differently to diagenetic processes th a n u n w e ath
ered bones. T he acidity, chem istry, m o istu re c o n ten t, perm eability, a n d o th er
Diagenesis 419
p ro p e rtie s o f the em b ed ding sed im en tary m atrix , as well as pedogenic (soil

form ing) processes if any, are also im p o rta n t (e.g., R etallack 1984; T u ro ss et al.
1989). F inally, as one m ight expect, a sh o rt tem p o ra l d u ra tio n o f burial m ay
result in a bo n e n o t being exposed to vario u s diagenetic processes th a t it w ould
be exposed to w ere it bu ried fo r a long tim e. F o r exam ple, Sillen (1989:212)
w rites, “ th e d u ra tio n o f in term en t is a critical v ariab le in diagenesis, [but]
d iagenetic changes in a given e n v iro n m e n t m ay n o t b e a r a linear re latio n sh ip to
tim e.” H e goes o n to m ak e a n o b serv atio n w hich m ay relate to the d istin ctio n
betw een syndiagenesis an d anadiagenesis: “ In the early p o st-d ep o sitio n al
period , the m in eral [of skeletal tissue] m ay be p artia lly shielded fro m diagenesis
by th e o rg an ic p h ase . . . since the a p a tite crystals becom e m o re exposed to
w a te r a n d soil io n s” w ith the d eco m p o sitio n o f the org an ic ph ase (Sillen
1989:219).
E q u a tio n [11.1] is a sim plistic re n d itio n o f w h a t is a com plex suite o f
p o te n tia l ta p h o n o m ic processes. C arlso n (1990:545) notes, fo r exam ple, th a t
the diagenesis o f d en tin e in teeth m ay be linearly related to tim e, b u t given the
v ariab ility in d iagenetic processes th a t m ig h t affect dentine, the re la tio n o f tim e
a n d diagenesis m ay n o t be statistically linear. A t the o u tset I am com pelled to
n o te th a t m y experience w ith diagenetically altered fossils is lim ited. B ut, m any
o f the v erte b rate rem ains zo o a rch a eo lo g ists deal w ith h ave u n d erg o n e m inim al
(and p ro b a b ly insignificant) diagenetic m o dification. (T he m o st extensive a n d
intensive analysis o f arc h a e o fa u n a l rem ains recovered from P lio-P leistocene
co n tex ts a t O ldu v ai G o rg e does n o t even list diagenesis o r m in era liza tio n in the
index [P otts 1988].) A n d , b io stra tin o m y is certain ly m u ch easier (an d m ore
fun?) to stud y eth n o arch ae o lo g ic ally due to the often long tim e spans involved
in a n d the sub surface (an d th u s largely invisible) n a tu re o f diagenesis. M y
tre a tm e n t o f diagenesis m ay th u s seem less d etailed th a n it m ig h t be. F o r
exam ple, I do n o t discuss the significant research on the effects o f diagenetic
processes on analyses o f bo n e trace elem ent c o n te n t (e.g., Price et al. 1992;
W h itm er et al. 1989), analyses w hich are becom ing increasingly im p o rta n t in
studies o f h u m a n diet. T he interested re ad er is en co u rag ed to p u rsu e the topics
in tro d u ce d here, a n d to read the references cited fo r a d d itio n a l in fo rm atio n . A
set o f ra th e r geochem ically-detailed references o n diagenesis an d m in eraliza
tion is fo u n d in A llison an d Briggs (1991a), Lucas a n d P revot (1991), and
T u ck e r (1990).
Mineralization, leaching, enrichment

In the process o f fossilization the b o n e’s com ponents are m odified as follows: (a)
gradual d isappearance o f all organic structures, i.e., the osteocytes and the
[collagen]; (b) their replacem ent by m aterial carried in the w ater o f the ground; (c)
substitution o f the chem ical elem ents co nstituting the crystalline lattice o f the
apatite.
(A. Ascenzi 1969:527)
Basics and background

A p erm ineralized o r p etrified fossil is form ed by in filtratio n o f m in eral-b earin g
so lu tio n s in to p o res in skeletal tissue a n d d ep o sitio n o f m inerals in those
spaces. T his process is also k n o w n as cellular p erm in e ralizatio n o r p etrifaction.
A m ineralized fossil is fo rm ed by the replacem ent o f the original h a rd p a rts o f
an o rg an ism w ith o th e r m inerals; the skeletal tissu e’s chem ical c o n stitu en ts are
dissolved a n d rem o ved w hile m inerals dissolved in g ro u n d w a te r are sim u lta
neously dep osited in th eir place. T his process is also kn o w n as authigenic
p re serv atio n o r rep lacem ent (M a tth ew s 1962; S ch o p f 1975). B oth m in eraliza
tio n a n d p etrificatio n result in a sto n y h a rd replica o f the skeletal specim en.
Leaching is the rem o val o f soluble m atter; enrichm ent is th e ad d itio n o f soluble
m atter. T h u s m in era liza tio n a n d p etrification, in a w ay, are en rich m en t
processes. H ydrolysis is o ften a leaching process involving th e c o m b in a tio n o f
w ater w ith o th e r m olecules.
B artsio k as an d M id d leto n (1992:68) suggest the term “ fossilization (o f bone)
m ay be defined as a n a tu ra lly o ccu rrin g physico-chem ical process th a t pre
serves th e gross m o rp h o lo g y by ionic exchange a n d an increase in the size o f the
a p a tite crystallites; this process also involves th e loss o f o rg an ic m ateria l o f the
bo n e a n d the filling o f voids w ith m in era ls” (em phasis in original). T hey
ind icate th a t fossilization refers to the “ sto n in ess” o f bone, a n d th a t the G reek
w ord apolelithom enon, m eaning tu rn e d to stone, was used to describe fossils
over 2,000 years ago. T h u s they arg u e th a t the term diagenesis is n o t a good
d esc rip to r o f fossilization processes, a n d suggest instead the term s apolithosis
to m ean fo ssilization a n d the term apolithotic effects to refer to the changes
b ro u g h t a b o u t by fossilization processes. W hile it rem ains to be seen if their
revisionist term in o lo g y is a d o p te d , w h a t is p erh ap s m ore im p o rta n t is th a t they
d o cu m en t, as o th ers have before (e.g., T u ro ss et al. 1989), th a t skeletal tissue
increases its “ crystallin ity (w hich p ro b a b ly reflects a n increase in [apatite]
cry stallite size)” d u rin g fossilization, a n d th a t m ay a c c o u n t fo r the “ high
co m p actn ess, h ard n ess, stiffness a n d generally 's to n y ’ a p p e ara n ce o f heavily
fossilized b o n es” (B artsio k as an d M id d leto n 1992:67). P reserv atio n o f the
histo log ical s tru c tu re o f skeletal tissue o f reptiles a n d fish d u e to fossilization
processes is review ed by A scenzi (1969).
A process we ca n term encrustation involves the p re c ip ita tio n o f soluble salts
on the surface o f a b o n e o r to o th . T he soluble salts are derived from sedim ents
an d tra n s p o rte d by g ro u n d w ater. C alcification - the p re c ip ita tio n o f calcium -
c a rb o n a te salts - is typ ical o f som e archaeological rem ains recovered from
so m ew h at arid areas w here m o istu re is insufficient to flush the salts fro m the
sed im en tary m atrix . T h e co a tin g o f bones w ith such salts can o bscure the
details necessary to iden tify th em a n d requires intensive cleaning (S tahl and
B rinker 1991). B rain a n d Sillen (1988:464), fo r exam ple, re p o rt b o th th a t fossils
Diagenesis 421
they exam in ed w ere stain ed a d a rk co lo r fro m “ diagenetic in c o rp o ra tio n o f

m an g an ese d io x id e” a n d th a t fossil surfaces w ere “ ob scu red by heavy en c ru s
ta tio n o f calcium c a rb o n a te a n d m anganese d io x id e.”
D iag en etic processes n o t only result in the creatio n o f replicas, b u t they can
significantly alter the chem ical co m p o sitio n o f buried skeletal tissues. W h itm er
et al. (1989:244—245) o u tline a useful m odel o f such chem ical m odifications by
using th e co n cep t o f diffusion-"a. process th a t o p erates w henever m obile
[chem ical] elem ents o ccur in differing co n c en tratio n s. In this event, a co n c en t
ra tio n g ra d ie n t is estab lish ed w ith the result th a t elem ents tend to m ig rate from
areas o f high to low c o n c e n tra tio n . T his process co n tin u es betw een the areas
until a c o n c e n tra tio n eq u ilib riu m is estab lish ed .” T h u s, theoretically, if the
skeletal tissue (M in e q u a tio n [11.1]) differs in chem ical co m p o sitio n from its
su rro u n d in g sed im en tary m atrix (S in e q u a tio n [11.1]) a t the tim e o f b u rial, the
b o n e will ten d to lose chem ical elem ents in w hich it is rich (the b o n e will be
leached o f these elem ents), a n d gain elem ents in w hich it is p o o r relative to the
sedim ent (be enriched). P ate a n d H u tto n (1988) describe one m eth o d o f
stu d y in g v a ria tio n in the ch em istry o f the fossil a n d the chem istry o f the
sed im en tary m atrix.
W h itm er et al. (1989) pro v id e an excellent overview o f the study o f trace
elem ents a n d skeletal tissue chem istry th a t goes fa r b eyond the scope o f
coverage necessary in this volum e; in terested read ers are en co u rag ed to
exam ine th eir acco u n t. It suffices here to sum m arize som e general o b serv atio n s
th a t are readily ap p reciate d by th o se w ith m inim al expertise in chem istry.
G eneral observations
C h ap lin (1971:16-18) suggests th a t at least fo u r p ro p e rtie s o f the d ep o sitio n al
m atrix affect skeletal tissue preserv atio n : p H . a e ra tio n , w a te r regim e, an d
b acterial action . H e suggests (a) b acterial ac tio n will d estro y the organic
fractio n o f skeletal tissue (see C h a p te r 5), (b) bacterial actio n is inhibited in
acidic sedim ent, (c) acidic sedim ent will dissolve th e m in eral fractio n o f skeletal
tissue an d the ra te o f d isso lu tio n depends on the degree o f acidity a n d am o u n t
o f p erco latin g w ater, (d) skeletal tissue will preserve fo r only a few m illennia in
acidic sedim ent, an d (e) p reserv atio n is b e tte r in basic sedim ents. C haplin
(1971) n o tes th a t the ra te o f decay a n d d isso lu tio n is d ep e n d en t o n the p o ro sity
o f th e skeletal tissue; m o re p o ro u s tissue decays m o re ra p id ly th a n less p o ro u s
tissue. T h u s the stru c tu ra l o r bu lk density o f b o n e tissue (C h a p te r 7) is an
im p o rta n t v ariable. T he chem ical b re a k d o w n a n d leaching processes progress
fro m exposed surfaces into the bo n e tissue (C ussler a n d F e a th e rsto n e 1981).
R olfe a n d B rett (1969:226, 229) suggest (a) highly basic sedim ents inhibit
b acterial activity, (b) the o rg an ic fractio n o f skeletal tissue decays m ore rapidly
in an aero b ic th a n a n a ero b ic e n v iro n m en t, a n d (c) p h o sp h atic, k eratin o u s, an d

org an ic skeletal tissues preserve in acidic sedim ents. H a re (1980:209) no tes th a t
“ if a b o n e is left in a so lu tio n o f a w eak acid the m in eral m ateria l will dissolve
an d th ere will be left a p seu d o m o rp h ic m odel o f th e b o n e form ed by its organic
m a trix .” B rom age (1984:168, 175-176) re p o rts th a t b ones so ak ed in acetic acid
h ave surflcial m in eral m ateria l dissolved a n d lacu n ae, canaliculi, a n d vascu lar
a n d cap illary spaces are enlarged. E ven tu ally th e b o n e tak es on a p itte d or
etched ap p e ara n ce as a result o f th e en larg em en t o f n a tu ra lly p re sen t spaces.
K n ig h t (1985) re p o rts th a t b u rn e d bo n e dissolves fa ste r th a n u n b u rn e d
m am m al b o n e w hen placed in an acidic so lu tio n (F ig u re 9.10).
H a re (1980:213-214) re p o rts th a t w a te r leaches soluble p ro tein c o m p o u n d s
from skeletal tissues by b re ak in g collagen d o w n in to p olypeptides (“ su b p ro tein
u n its” ) w hich are sub seq u en tly leached. G ro u n d w ater leaches the organic
fractio n o f skeletal tissues at a ra te th a t depends o n how sa tu ra te d the w a te r is
a n d the ra te o f g ro u n d w ater m o v em en t (flushing) th ro u g h the sedim ent m atrix
(H are 1980). G ro u n d w ater m ay also leach the m ineral co m p o n en t o f skeletal
tissues if th a t w ater is n o t sa tu ra te d w ith the m inerals m ak in g up b o n e tissue
(D o d d a n d S ta n to n 1981:128). Such leaching ten d s to require perm eable
sedim ent a n d a sed im en t m atrix th a t is m ineralogically different fro m th e bone
tissue (D o d d a n d S ta n to n 1981:128). “ I f eith er the o rg an ic m atrix o r the
m ineralized co m p o n en ts o f bo n e are p artia lly rem oved, the h ard n ess and
stren g th o f b o n e will d ecrease” (H a re 1980:217).
W h ite a n d H a n n u s (1983) suggest th a t h y d ro x y a p a tite h as low solubility in
alkaline (p H > 7 .5 ) to slightly acidic (pH = 6.0) aq u eo u s system s, w hile in
highly acidic en v iro n m en ts (p H < 6 .0 ) its solubility is very high. “ C hem ical
w eath erin g o f b o n e is p ro b a b ly in itiated by acids created as m icro o rg an ism s
d eco m p o se co llag en ” (W hite an d H a n n u s 1983:321). T hey suggest th a t heating
o f b o n e ap p a re n tly does n o t accelerate chem ical w eath erin g a n d a p p e ars to
increase th e p ro p o rtio n o f p h o sp h o ru s in bo n e (b u t see F ig u re 9.10 an d
asso ciated discussion).
In a series o f experim ents, V on E n d t an d O rtn e r (1984) used fresh cow (Bos
sp.) b o n e crush ed to different sizes to assess the effect o f te m p e ra tu re a n d bone
size o n th e ra te a n d n a tu re o f b o n e d isin te g ratio n (see also O rtn e r et al. 1972).
T hey fo u n d th a t sm all bo n e particles tend to preserve less well th a n large bone
particles, a n d th u s they suggest th a t the rem ains o f sm all anim als m ay be less
well preserved th a n th e rem ains o f large anim als (see also C h a p te r 9). T hey
fo u n d th a t th e p ro te in in b o n e is lost m o re quickly a t high tem p eratu res. V on
E n d t a n d O rtn e r (1984:252) also n o te th a t th e p o ro sity o f the skeletal tissue
exerts a stro n g influence on its p reserv atio n ; m o re p o ro u s tissue allow s qu ick er
diffusion o f g ro u n d w a te r into the tissue a n d th u s the q u ick er the p ro tein -to -
m in eral b o n d s will be altered (see also C ussler a n d F e a th e rsto n e 1981).
Diagenesis 423
Analysis o f chemically altered bone

C hange is the essence o f the process o f fossilization . . . the cardinal prerequisite [is]
th a t the object still be recognizable as the physical form o f a p lan t o r anim al, o r one
o f its parts.
(S. F. C ook e ta l. 1961:356)
G o rd o n a n d B u ik stra (1981) exam ine the c o rrelatio n betw een the preservatio-
nal c o n d itio n o f h u m a n b ones recovered fro m sites in Illinois a n d the p H o f the
sed im en t fro m w hich th e bones w ere recovered. T hey reco rd ed the p reserv atio -
nal c o n d itio n o f b o n e specim ens as belonging to one o f six categories. Strong
com plete bone displays n o evidence o f p o stm o rte m m o d ificatio n a n d specim ens
are w hole. Fragile bone has ex tern al surfaces th a t m ay show som e etching,
co uld be frag m en ted , h as som e superficial d estru c tio n , a n d epiphyseal ossifica
tio n cen ters o f on to gen ically y o u n g b o n e are ero d ed . F ragm ented bone consists
o f crack ed a n d b ro k e n skeletal elem ents, a n d surfaces are heavily etched an d
crack ed . E x tre m e ly fra g m e n te d bone consists o f specim ens th a t are extensively
a n d intensively b ro k e n (see C h a p te r 8) a n d q u ite eroded. T he last tw o
categ ories are sim ilar; the bone m eal a n d ghost categories include bone tissue
th a t co n sists o f a pow d ery su b stan ce th a t will n o t h o ld its a n a to m ica l shape
w ith o u t su rro u n d in g sedim ent.
G o rd o n a n d B u ik stra (1981:569) fo u n d th a t o ntogenically y o ung b ones were
less well preserved th a n o ntogenically old bones. T hey fo u n d a stro n g inverse
co rre la tio n betw een sedim ent p H a n d the frequency o f ontogenically m atu re
bones p reserved w ithin p a rtic u la r p reserv atio n al categories (r= - 0 .9 2 ,
P < 0.001). T h a t re la tio n w as also negative fo r the ontogenically y o u n g bones
(/• = —0.48, P < 0.005). T hey suggest the la tte r coefficient is less stro n g th a n the
fo rm er d u e to the g re ater range o f stru c tu ra l density (see C h a p te r 7) displayed
by the o nto g en ically y o u n g bones. T o g eth er, sedim ent p H a n d o n to g en etic age
w ere stro n g ly co rre la te d w ith p re serv atio n (r= —0.87, P < 0.001).
G o rd o n an d B u ik stra ’s (1981) stra ig h tfo rw a rd analysis is exem plary, w ith
the ex ceptio n th a t they did n o t p u b lish the d a ta from w hich th eir c o rrelatio n
coefficients w ere derived. T hey identified p re serv atio n al v a ria tio n , designed a
w ay to categorize a n d q u an tify th a t v a ria tio n , a n d so u g h t a n d fo u n d a
co rre la tio n betw een th a t p re serv atio n al v a ria tio n a n d a w ell-know n diagenetic
p ro p e rty . A ll such analyses o f the diagenetic effects o f ta p h o n o m ic processes
sh o u ld be so clear (a n d include all relevant d ata).
Sediment overburden weight

As fa u n al rem ains becom e m o re deeply bu ried , the w eight o f the sedim ent
overlying them becom es g re ater an d u nderlying sedim ents becom e m ore
co m p ac t a n d o f g re ater bulk density because there are sm aller a n d few er p o re
spaces betw een sed im en tary particles. K idw ell (1986:7) suggests differential
c o m p a c tio n fro m sedim ent o v e rb u rd e n w eight can c o n c e n tra te fossils, a n d this

p ro b a b ly results from the g re ater bulk density (low er p o ro sity ) o f the sedim en
ta ry m atrix in general; m ore fossils p er u n it volum e o f sedim ent w ould result.
T his suggests th a t one m ay w an t to estim ate the bulk density (porosity) o f the
sed im en tary m atrix alo n g at least the vertical dim ension o f an ex cav atio n to
ascertain if g re ater fossil density (frequency p er u n it volum e) at g re ater d ep th is
a fu n c tio n o f decreased p o ro sity w ith increasing d e p th (e.g., R etallack
1990:135).
R etallack (1990:132) suggests th a t the best evidence fo r co m p ac tio n o f
sedim ents fro m o v e rb u rd e n w eight is “ folding o f vertical crack s [in paleosols]
filled w ith d istinctiv e soil m aterial; co m p ariso n o f the len g th o f these [distorted
a n d filled cracks] w ith th e ir len g th in a stra ig h t line can yield a c o m p a c tio n ratio
th a t is useful fo r re co n stru ctin g fo rm e r soil thickness, density, a n d chem ical
c o m p o sitio n .” T he follow ing e q u a tio n p rovides a w ay to d eterm in e the degree
o f co m p actio n :
Cf= Vf/Vs = Tf/T s [11.2]
w here Cf is the diag en etic co m p ac tio n ra tio o r fractio n o f original thickness o f

the soil d u e to b u rial, V f is the volum e o f the fossil soil o r h o rizo n in cm 3
inclu din g foreign m a te ria l in the filled cracks, V s is the volum e o f th e fossil soil
o r h o riz o n in cm 3 fo rm ed fro m p a re n t m ateria l (excludes foreign m aterial), T fis
the th ickn ess o f a fossil soil o r h o rizo n in cm derived fro m the soil o r soil
h o rizo n , a n d T s is th e thickness o f th e soil o r soil h o riz o n in cm derived from
p aren t m ateria l (R etallack 1990:69-70). C o m p a c tio n a n d d e fo rm a tio n m ay
also be ind icated by d isto rte d ro o t casts a n d crushed b o tan ica l m aterials
(R etallack 1990:132-135).
S edim ent o v erb u rd en w eight m ay also d efo rm o r fra c tu re fossils. S hipm an
(1981 b: 172) suggests w hen th e d ep o sitio n o f o v e rb u rd e n sedim ent is ra p id , the
pressu re from th e w eight m ay crush o r b re ak the b u ried fau n al rem ains;
“ slow er d ep o sitio n m ay p ro d u c e plastic d e fo rm a tio n , in w hich the sh ap e or
d im ensions o f the o rig inal [fossil] are ch an g ed w ith o u t b re a k in g .” F in k s
(1979:330) in d icates th a t “ m echanical d isto rtio n is accom plished by plastic
flow o f th e [skeletal] m a te ria l (m ainly m ic ro fra c tu rin g a n d recrystallization).
Sim ple co m p ac tio n o f the sedim ent afte r b u rial m ay d isto rt fossils by s h o rte n
ing them in the d irectio n p erp en d icu lar to the bedding p la n e .” B ecause som e
degree o f d e fo rm a tio n precedes fra c tu re (see C h a p te r 4), I co n sid er d e fo rm a
tio n p rio r to cru sh in g a n d frag m en tatio n .
D eform ation
T h e deform ation o f v erte b rate rem ain s refers to th eir d isto rtio n ; th a t is, tw o o r
m o re a n a to m ica l p o in ts on a single skeletal elem ent change th eir spatial
lo catio n s relative to one a n o th e r (see the G lossary). T he degree a n d kind o f
Diagenesis 425
d e fo rm a tio n are d eterm in ed by several facto rs, including b o th intrinsic an d

extrensic facto rs. Intrinsic fa c to r s include the original m o rp h o lo g y , m o d u lu s o f
elasticity (see F ig u re 4.6), a n d o rie n ta tio n in b o th h o riz o n ta l a n d vertical
planes o f th e skeletal p a rt being defo rm ed . H e n d erso n (1987:44), fo r instance,
n o tes th a t the skull, w hen considered as a u n it, is ra th e r susceptible to
d e fo rm a tio n a n d cru sh in g due to it being a large, hollow sphere. E xtrinsic
fa c to rs include th e n a tu re a n d tim ing o f diagenetic fa cto rs such as m in era liza
tio n a n d leaching (w hich alte r th e m o d u lu s o f elasticity), a n d the g rain size an d
co m p o sitio n o f th e sedim ent. “ C o arse r-g ra in ed sedim ents are m ore re sistan t to
co m p ac tio n resu lting from the su p p o rtin g effect o f the grains a n d the low er
p o re -w ate r v o lu m e” (Briggs 1990:244).
D e fo rm a tio n can tak e several form s. H om ogeneous deform ation involves
ch ang in g th e m o rp h o m e try o f a b o n e b u t m a in tain in g th e relative o rie n ta tio n
o f all planes; “ all lines o r areas o f b o n e in one p a rtic u la r o rie n ta tio n are
uniform ly leng then ed o r sh o rte n e d ” (S h ip m an 198lb : 179). B asically, the
an aly st superim p o ses an u n d efo rm ed m odel o f the fossil over a grid, and,
d eterm in es th e “ strain ellipse” (S h ip m an 1981 b: 179) th a t w ould be p ro d u c ed if
the u n d efo rm ed fossil w ere a perfect circle a n d the d efo rm ed fossil w ere an
ellipse. D e te rm in a tio n o f the forces necessary to create the ellipse fro m the
circle in d icates th e forces necessary to d efo rm th e fossil in to its p resen t
m o rp h o m e try (see D o v e to n 1979; Sdzuy 1966; W ellm an 1962 fo r m ore
co m p lete d escrip tio n s o f th e m atrix alg eb ra p rocedures). C o m p u te r-g ra p h ic
p ro g ram s ca n be used to help d eterm in e the original, p re -d e fo rm a tio n shape o f
the fossil (Briggs 1990). O th e r form s o f d e fo rm a tio n include m o v em en t o f
a n a to m ica l p o in ts in different d irections relative to one a n o th e r, such as
b ending, w hich causes p o in ts on the concave side o f the b en d to becom e closer
a n d p o in ts on the convex side to becom e fa rth e r a p a rt.
F o r th ose less m ath em atica lly inclined. Briggs (1990:245) suggests one can
tak e p h o to g ra p h s o f u n d isto rte d h o m o lo g o u s skeletal p a rts from different
angles to derive a m odel o f th e original m o rp h o m e try o f the d isto rte d fossil;
“ these p h o to g ra p h s are directly an a lo g o u s to flatten in g the organism o n to a
bed ding p la n e .” C o m p a riso n o f the d isto rte d fossil w ith the p h o to g ra p h s can
help the an aly st d eterm in e the direction(s) in w hich d isto rtio n forces were
applied. M irro r im aging o f p aire d , bilaterally sy m m etrical skeletal p a rts is also
possible (S h ip m an 1981 b: 179).
C rushing and fra g m en ta tio n

In C h a p te r 8 I describe analyses o f frag m en ted bones p erfo rm ed by V illa an d
M ah ieu (1991). O ne o f the things they n o te in th eir analysis is the p o st-b u rial,
diagenetic fractu rin g o f som e specim ens. T hey arg u e th a t bones b ro k e n by
sedim ent o v erb u rd en w eight are in d icated by co n jo in in g frag m en ts lying
ad jac en t to, a n d o ften in c o n ta c t w ith, one a n o th e r, a n d b ro k e n b ones lying on
convex o r concave surfaces (th u s the b ones w ould be subjected to bending

forces). A s well, the b asic m o rp h o lo g y o f fractu re s created afte r b u rial ten d s to
be different th a n th e m o rp h o lo g y o f fractu res gen erated by b io stratin o m ic
processes. P o st-d ep o sitio n a l a n d p o st-b u rial m ovem ent o f bo n e frag m en ts due
to diagenetic facto rs a n d tu rb a tio n processes can be m easu red by refitting o r
con jo in in g pieces (see C h a p te r 5). F o r ad d itio n a l discussion o f analytically
detecting p o st-d ep o sitio n al fra g m e n ta tio n by stu d y o f fra c tu re kinds see
C h a p te r 8. T he re m a in d er o f this section is devoted to o th e r analytical
techniques fo r detecting p o st-d ep o sitio n al d estru c tio n o f bones.
L y m an a n d O ’B rien (1987) suggest th a t fra g m e n ta tio n o f b ones ca n resu lt in
th eir “ an aly tical ab sen ce.” A s a skeletal elem ent is b ro k e n in to sm aller pieces,
the p ro b a b ility it will be identifiable in th e set o f frag m en ts decreases as
frag m en t size decreases. T his is so because th e p ro b a b ility th a t any p a rtic u la r
frag m en t will have an a n a to m ic a l la n d m a rk th a t is diag n o stic o f the skeletal
elem ent decreases as th e frag m en t becom es sm aller. T h u s, a collection o f bone
frag m en ts m ay c o n ta in pieces of, say, fem ora, b u t if those pieces are sufficiently
sm all a n d th e an a to m ic a l la n d m a rk s are sufficiently incom plete, th en the
an a ly st will be u n ab le to identify th e fem ur. It is in this sense th a t fra c tu rin g o f
skeletal elem ents “ d estro y s” them . T he b o n e tissue is n o t destro y ed , b u t the
an a to m ica l in teg rity o f th e skeletal elem ent is sufficiently co m p ro m ised th a t the
effect is th e sam e as if the b o n e tissue w ere d estro y ed w hen th e an a ly st is
atte m p tin g to id entify skeletal p a rts. It is in this sense o f “ d e s tru c tio n ” o f bones
- th eir analy tical absence d u e to intensive fra g m e n ta tio n - th a t we tu rn to
a n o th e r m e th o d o f detectin g p o st-d e p o sitio n a l d estru c tio n o f skeletal p arts.
Based on a stud y o f several p reh isto ric assem blages w ith u n k n o w n ta p h o n o
m ic h istories, K lein a n d C ru z-U rib e (1984:70-75) suggest th a t “ p o st-d ep o sitio n -
a l” d e stru c tio n ca n be detected by stu d y o f a suite o f attrib u te s. I suspect th a t
w h a t they are a tte m p tin g to m easu re is actu ally b e tte r labelled “ p o st-b u ria l”
d estru c tio n , given th e ir co m m en t th a t the ta p h o n o m ic processes involved are
“ th e ex ten t to w hich a sam ple h as been leached a n d th e ex ten t to w hich it has
been c o m p ac ted in th e g ro u n d [or] p o st-d ep o sitio n al leaching co upled w ith
very slow sed im e n ta tio n ” (K lein a n d C ru z-U rib e 1984:70, 72). T he a ttrib u te s
they stu d y include the h ard n ess a n d /o r co m p actn ess o f the stra tu m on w hich
th e b o n es lay because if they lay o n h a rd su b strates, p re -b u rial tram p lin g will be
m o re likely to cru sh th e bones. A s well, they suggest an a b u n d a n c e o f isolated
teeth a n d p au c ity o f id entifiable m an d ib les a n d m axillae, a n d a p le th o ra o f
sm all, dense b ones such as carp als, tarsals, sesam oids, a n d p h alan g es indicate
th e b o n e assem blage “ h as p ro b a b ly suffered greatly fro m p o st-d ep o sitio n al
d e s tru c tio n ” (K lein a n d C ru z-U rib e 1984:71). H a rd e r su b strates m ay p ro d u c e
(a) relatively hig h a b u n d a n ces o f iso lated teeth a n d to o th -b e a rin g skeletal
elem ents, a n d (b) h igh a b u n d a n ces o f ca rp als a n d o th e r sm all, dense bones,
d u rin g early phases o f an assem blage’s p o st-d ep o sitio n al ta p h o n o m ic history,
assum in g all skeletal elem ents w ere originally p resen t to begin w ith (essentially
Diagenesis 427
T ab le 11.1 R a tio s o f N IS P : M N I p er skeleta l p a rt in two assem blages (from

Klein an d C ruz-U ribe 1984)
Skeletal p a rt Level 4 Level 6 Skeletal p art Level 4 Level 6
m axilla 5.92 3.37 m an d ib u la r condyle 8.00 2.24

m andible 7.41 3.81 hyoid 3.50 3.14
atlas 5.00 5.00 axis 1.00 1.50
cervical 0.00 10.00 thoracic 6.00 11.00
lum bar 2.00 6.00 sacrum 3.00 1.00
rib 19.67 18.00 innom inate 3.33 2.50
scapula 2.00 2.00 P hum erus 1.00 5.00
D hum erus 2.00 1.80 P radius 3.60 2.62
D radius 7.00 1.33 P ulna 2.14 1.43
D ulna 1.60 1.25 carpal 6.83 7.00
P m etacarpal 6.00 5.45 D m etacarpal 2.00 1.93
P fem ur 2.00 2.33 D fem ur 4.00 3.00
P tibia 1.50 1.00 D tibia 2.67 2.31
naviculo-cuboid 6.00 2.25 astragalus 1.60 2.10
calcaneum 2.78 3.00 small tarsals 2.92 3.14
P m etatarsal 7.00 6.75 D m etatarsal 1.33 2.00
1st phalanx 24.13 22.11 2nd phalanx 25.67 16.62
3rd phalanx 16.33 9.75 sesam oid 16.33 16.20
com p lete sk eleto ns w ere deposited). But I w o n d e r if d u rin g later phases o f th a t

h isto ry even sm all, dense bones a n d teeth will becom e destroyed. D a ta an d
arg u m en ts p resen ted elsew here (L y m an 1991c, 1993b) suggest th a t w ith
sufficient tim e, m an y bones will be destro y ed by d en sity -m ed iated processes.
F o r exam ple, th e d a ta in F ig u re 7.13 suggest p reh isto ric, bu ried assem blages
will be m o re p ro n e to display such d estru c tio n , because m o re p re h isto ric
assem blages (75 o f 143; 52.4% ) th a n eth n o arch ae o lo g ic al (9 o f 41; 22.0% )
assem blages are positively co rrelated w ith density.
W e re tu rn to th e issue o f th e frequency a n d p re serv atio n o f sm all, dense
skeletal elem ents below . F irst, how ever, it is im p o rta n t to review the o th er
a ttrib u te K lein a n d C ru z-U rib e (1984:71) suggest will be displayed by post-
d ep o sitio n ally d estro y ed assem blages: high N IS P :M N I ra tio s p er skeletal p art.
T hey in d icate th a t tw o assem blages o f b ones fro m stratig ra p h ic ally d istin ct
co n tex ts in the L o w er M ag d a le n ian El Ju y o C ave h av e ra th e r different
N IS P :M N I ra tio s p er skeletal p a rt. T he ra tio s fo r the tw o to ta l assem blages
are: Level 4 - 97.76 (1662 N IS P : 17 M N I), a n d Level 6 - 38.47 (1462 N IS P :38
M N I). R atio s fo r m an y o f th e in d iv id u al skeletal p a rts are sum m arized in T able
11.1. It is im p o rta n t first to n o te th a t these are not N IS P :M N E ra tio s as
described in C h a p te r 8. R ath er, they are N IS P :M N I ra tio s p er skeletal p art,
w ith M N I d eterm in ed fo r each skeletal p a r t (w hether a specim en is fro m the left
o r rig h t side is ta k e n in to acco u n t). T he im p o rtan ce o f this is th a t, as show n in
C h a p te r 8, the N IS P :M N I ra tio s m ay n o t m easure fra g m e n ta tio n very well if
Level 4
Figure 11.1. B ivariate seatterp lo t o f N IS P :M N I ratio s p er skeletal p a rt for tw o
bone assem blages from the sam e site (from T able 11.1).
th ere are m a jo r differences in frequencies o f left a n d rig h t specim ens o f a

skeletal p a r t a n d /o r if th ere is a high p ro p o rtio n o f com plete skeletal elem ents
per an a to m ica l p a rt in th e collections. N onetheless, the ra tio s in T able 11.1 are
in fo rm ativ e, a n d they are one o f the criteria used by K lein a n d C ru z-U rib e
(1984) to arg u e th a t th e Level 4 assem blage has u n d erg o n e m o re “ post-
d e p o sitio n a l” d estru c tio n th a n the Level 6 assem blage.
F ig u re 11.1 show s th a t there are m o re hig h N IS P :M N I ra tio s a m o n g the
Level 4 skeletal p a rts (p o in ts below the d iag o n al line) th a n am o n g the Level 6
skeletal p a rts (p o in ts ab o v e th e d iag o n al line). O n the basis o f such differences,
Klein a n d C ru z-U rib e (1984:75) co n clu d e th a t N IS P :M N I ra tio s are “ a
succinct a n d re aso n ab ly objective m easure o f fra g m e n ta tio n , p artic u la rly w hen
they are p resen ted fo r each skeletal p a r t.” N o te here th a t these ra tio s alone do
n o t in d icate w h eth er fra g m e n ta tio n w as b io stratin o m ic , p o st-d ep o sitio n al, or
p o st-b u rial. B ut in co m b in a tio n w ith relatively high ab u n d a n ces o f sm all,
dense b o n es a n d iso lated teeth, K lein a n d C ru z-U rib e (1984:71) believe these
ra tio s are in dicative o f “ p o st-d e p o sitio n a l” d estru c tio n . T h e bones w ith high
N IS P :M N I ra tio s w ere b ro k e n by p o st-d ep o sitio n al processes w hereas the
sm all, dense b ones a n d teeth “ are p a rtic u la rly likely to survive m echanical
cru n ch in g ag ain st a h a rd s u b s tra te ” (K lein a n d C ru z-U rib e 1984:71). T he
N IS P :M N I ra tio s fo r m andibles a n d m axillae are h ig h er in the Level 4
co llection (7.41 a n d 5.92, respectively) th a n in the Level 6 assem blage (3.81 and
3.37, respectively), a n d th e fo rm e r assem blage h as v irtu ally n o m an d ib les o r
m axillae b u t has m an y isolated teeth. F inally, the sm all, co m p ac t bones m ake
up m o re o f the Level 4 to ta l N IS P (919, o r 55.3% ) th a n o f the Level 6 to ta l
N IS P (544, o r 37.2% ; arcsine ts= 10.13, P < 0 .0 0 1 ).
Diagenesis 429
In a later analysis ad d ressin g the sam e issue, K lein (1989:374-375) suggests

skeletal elem ents o f sm aller tax a
are m ore likely to retain their integrity durin g butchering an d food p rep aratio n o r
during kicking an d tram p lin g across the surface o f repeatedly occupied sites [than
the skeletal elem ents o f larger taxa]. Bones th a t are less dam aged before burial are
also m ore likely to survive leaching an d profile com p actio n afterw ards, w hich
m eans th a t even a sm all p re-depositional difference in relative durab ility will be
m agnified post-depositionally.
T h a t is, K lein (1989) is a d d in g the variab le o f skeletal elem ent size to the
e q u a tio n . H e fu rth e r suggests th a t skeletal p a rts w ith high stru c tu ra l density
(C h a p te r 7) ten d to w ith stan d p o st-d ep o sitio n al d estructive forces b e tte r th a n
tho se w ith low s tru c tu ra l density (K lein 1989:378). It is im p o rta n t to note, then,
th a t th e skeletal p a rts listed in T ab le 11.1 are all fro m red d eer (C ervus elaphus),
a n d th u s skeletal elem ent size is n o t influencing the results (e.g., F ig u re 11.1).
A n d , w hile skeletal p a rt a b u n d a n ces as m easu red by M N I in b o th assem blages
are b o th co rrelated w ith b o n e stru c tu ra l density, the Level 4 assem blage is less
stron g ly co rrelated w ith density (rs = 0.58, P = 0.001) th a n the Level 6 assem
blage (rs = 0.67, P < 0.001) (frequencies o f skull p a rts a n d m an d ib les are n o t
included in the c alcu latio n o f these statistics). P erh a p s as m u ch as 10% m ore o f
the v ariab ility in skeletal p a rt frequencies in the Level 6 assem blage m ight be
a ttrib u te d to d en sity -m ed iated d estru c tio n th a n m ight be so a ttrib u te d in the
Level 4 assem blage, w hich is the o p p o site o f w h at m ight be expected on the
basis o f th e N IS P :M N I ratios.
M a re a n (1991) p u rsu es the them e s ta rte d by K lein a n d C ru z-U rib e (1984;
K lein 1989). H e focuses on the intensity o f b o n e fra g m e n ta tio n (as defined in
C h a p te r 8), b u t offers th ree im p o rta n t qu alificatio n s th a t m u st be m et if this
a ttrib u te is to be used to m easu re p o st-d ep o sitio n al a n d p o st-b u ria l d e stru c
tion. F irst, the b o n es used to m easu re d e stru c tio n in the fo rm o f fra g m e n ta tio n
sho u ld be th o se “ th a t are never o r very rarely frag m en ted by people o r anim als
atte m p tin g to e x tra ct n u tritio n ;” second, th e b ones “ sh o u ld be in d ep en d e n t o f
the b o n e tra n s p o rt b eh a v io r o f bone co llecto rs;” a n d th ird , th e m eth o d o f
calcu latin g the degree o f fra g m e n ta tio n “ should be in d ep en d e n t o f the
ca lc u latio n p ro c ed u re fo r arch aeo zo o lo g ical m easures o f elem ent a b u n d a n c e ”
(M a rean 1991:680). M a re a n ’s (1991:6 8 1) experim ents suggest th a t ca rp als an d
tarsa ls (excluding calcan ea) w ould best m eet the first req u irem en t. C alcan ea
were som etim es frag m en ted by carn iv o res a n d display gnaw ing m ark s, a n d
carp als a n d tarsals b ro k e n by h u m an s w ith h am m ersto n e s an d anvils display
p ercussion m ark s (see C h a p te r 8); th u s specim ens displaying gnaw ing m ark s o r
percu ssio n m ark s sh ould be excluded fro m analyses o f p o st-d ep o sitio n al
d estru ctio n .
A fte r feeding a set o f ta rsa l-m e ta ta rsa l a n d ca rp a l-m e ta c a rp a l lim b sections
o f do m estic sheep ( Ovis aries) to sp o tted hyenas (C rocuta crocuta), M a re a n
(1991:684-685) co n cludes th a t this carn iv o re sw allow s carp als a n d tarsals, an d
th o se b ones m ay show digestive co rro sio n b u t only rarely display gnaw ing
magnum ZZZZZZZZZZZL
scaphoid '/ / / / / / / / / ///////A
1
lunate V / / / / / / / / / 7 / / / / / / / A
cuboid YZm ZZZZZZZZZZZZZZZZA H GvJm22

0 GvJm46
external cun.
naviculo-cub. / 7 7 Z 1 }
astragalus ZZZZ2____________
0 20 40 60 80 100
Completeness Index
Figure 11.2. B ar g rap h show ing v ariatio n in com pleteness index values across
seven sm all, co m p act bones from tw o sites (m odified from M arean 1991:688,
Figure 2).
dam age. As well, a statistic M a re a n calls the “ com pleteness index” is typically
high ( > 9 2 % ) fo r ca rn iv o re-d ep o site d sets o f ca rp als a n d tarsals. T his statistic
is calcu lated by “ estim atin g fo r each specim en the fractio n o f the original
co m p ac t b o n e th a t is presen t, sum m ing the values, a n d dividing th a t by the
to ta l n u m b e r o f specim ens ascribed to th a t bone a n d tax o n . M u ltip lica tio n by
100 pro v id es a m ean p ercentage co m p leten ess” (M a rean 1991:685). T h u s, if the
an aly st has a co m p lete unciform , h a lf o f a n unciform , a n d one th ird o f an
uncifo rm , one calcu lates the com pleteness index fo r the unciform as:
100[( 1 + 0 .5 + 0 .3 3 )/3 ]= 100[ 1.83/3] = 100[0.61] = 61 % .
M a re a n (1991) co m p ares his experim entally derived com pleteness index
d a ta w ith co m pleteness index d a ta fo r tw o sites in K enya. Site G vJm 22 is a
ro ck shelter; site G vJm 46 is an o p en site. T he sm all, co m p act, dense bones from
b o th sites display n o evidence o f percu ssio n m ark s, ch o p p in g m ark s, or
gnaw ing d am age. M a re a n (1991:687) concludes th a t “ virtually all o f the
fra g m e n ta tio n is d u e to p o st-d ep o sitio n al d e s tru c tio n ” a n d th a t the bone
assem blage from "G v Jm 4 6 has u n d erg o n e su b stan tially m ore p o st-d ep o sitio
nal d e stru c tio n th a n [the b o n e assem blage from ] G v Jm 2 2 .” A s show n in F igure
11.2, th e com pleteness index values in d icate the G vJm 46 frag m en ts are, on
average, sm aller th a n th e G vJm 22 fragm ents.
M arean (1991: 687-690) follow s K lein (1989) a n d suggests th a t bones o f
sm aller tax a will be b e tte r preserved th a n bones o f larg er taxa. T his is readily
show n fo r M a re a n 's tw o K enyan sites in F ig u re 11.3. T his b iv ariate sca tte rp lo t
is betw een the com pleteness indices fo r six sm all, dense b ones o f tw o size classes
Diagenesis 431
B ovid S iz e 1
Figure 11.3. B ivariate scatterp lo t o f com pleteness index values for six sm all,
com pact bones for tw o size classes o f anim als (from M arean 1991:689).
o f bovids; size class 1 an im als are 7 to 27 kg live w eight an d size class 2 bovids
are 50 to 220 kg live w eight. T he p lo tte d bones are the astra g alu s, calcaneum ,
n av icu lo -cu b o id, ex tern al cuneiform , in tern al cu neiform , a n d distal fibula. If
fra g m e n ta tio n is sim ilar fo r the tw o size classes, then all p lo tte d p o in ts should
fall on o r qu ite n ear the d iag o n al line. In fact, fo u r o f th em fall below it an d only
one falls ab o v e it. W ith size 1 bovids p lo tte d on the x-axis and size 2 bovids
p lo tte d on the y-axis, th e p o sitio n s o f the p lo tte d p o in ts indicate th a t fo u r o f the
p lo tte d skeletal elem ents have high com pleteness indices fo r the size 1 bovids
relative to size 2 bovids, an d only one skeletal elem ent has a high com pleteness
index fo r th e size 2 bovids relative to the size 1 bovids. T h a t is, overall, the
co m pleteness indices are higher for the size 1 (sm aller) bovids th an fo r the size 2
(larger) b ovids, in d icatin g b ones o f the fo rm er are b e tte r preserved th a n bones
o f th e la tte r tax a, ju st as M a re a n a n d K lein suggest they should be.
P o st-b u rial cru sh in g an d fra g m e n ta tio n have been little studied by v erte
b ra te ta p h o n o m ists a n d zo o arch aeo lo g ists. P erhaps, as M arean (1991:678)
suggests, th e re aso n fo r this resides in an y o r all o f the follow ing: a lack o f
aw areness o f th e p ro b lem , a belief th a t p o st-b u rial d estru c tio n is insignificant,
a n d /o r the lack o f a readily applied analy tical technique fo r detectin g such
d estru ctio n . T he last is no lo n g er ap p licab le as M a re a n 's (1991) tech n iq u e and
K lein a n d C ru z -U rib e 's (1984; K lein 1989) criteria fo r recognizing such
d estru c tio n are now available. M o re freq u en t ap p lica tio n o f these analytical
tools m ay show precisely how w idespread p o st-b u rial d estru c tio n is in various
tem p o ra l, sp atial, a n d geological contexts, a n d th u s how significant a pro b lem
it is.
T ab le 11.2 Turbation processes influencing burial, exposure, and m ovem ent o f

fo ssils (fro m W ood an d Johnson 1978)
Process class Sedim ent mixing
F au n a ltu rb a tio n anim als, especially burrow ing

F lo raltu rb atio n plan ts, ro o t grow th, an d treefall
C ry o tu rb atio n freezing an d thaw ing
fro st heaving
m ass displacem ent
frost cracking, ice wedges, sand wedges
sorting
G rav itu rb atio n mass w asting
solifluction, creep, subsidence
m udflow s, earthflow s, avalanche, landslide
A rgilliturbatio n swelling an d shrinking o f clays
A ero tu rb atio n gas, air, wind
A q u atu rb atio n w ater
C ry staltu rb atio n grow th an d w asting o f salts
S eism iturbation earth q u ak es
Post-burial movement
T h u s far discussion has focused on chem ical a n d m echanical m odifications o f
skeletal p arts. A n o th e r categ o ry o f m odification does n o t influence the skeletal
tissues b u t ra th e r m odifies the lo catio n s o f skeletal p arts. T he d etectio n o f p o st
b u rial m o vem ent sh ould be a m a jo r p a rt o f any analysis o f the spatial
d istrib u tio n o f fau n al rem ains. T his is so fo r any categ o ry o f artifa ct o r ecofact,
as m ad e clear by W o o d a n d J o h n s o n ’s (1978) extensive list o f turbation
processes (T able 11.2). B urial does n o t rem ove fau n al rem ains o r artifa cts from
processes th a t m odify th eir locations. T he discussion o f b u rial processes in
C h a p te r 10 m akes it clear th a t b u rial can be a tra n sito ry p h en o m en o n .
G eological processes th a t b u ry bones m ay also re-expose them . But even w hen
bones are b u ried a n d n o t re-exposed, v ario u s forces o p eratin g b en e ath the
surface ca n m ove fa u n a l rem ains.
P erh a p s th e easiest w ay to d eterm in e if th ere h as been som e p o st-b u rial
m o v em en t o f bones is the reco g n itio n o f spatially disassociated refitting
skeletal p arts. B ut how does one d eterm in e if the sp atial disasso ciatio n o f
refitting skeletal p a rts is the result o f p re -b u rial o r p o st-b u rial processes? As
n o ted in C h a p te r 5, this can be accom plished w ith stone to o ls fo r w hich the
o rd e r in w hich flakes have been rem oved fro m a core d ictates the vertical
o rd erin g o f flakes w ith in a stra tig ra p h ic sequence: flakes rem oved first lying
stratig ra p h ic ally ab o v e flakes rem oved last m u st have been d istu rb ed after
b urial. As also n o ted in C h a p te r 5, Villa et al. (1986) suggest th a t post-
d ep o sitio n al d istu rb an ce s result in v aria tio n in the vertical lo catio n s o f refitting
pieces o f b o n e, b u t w h eth er th o se d istu rb an ce s w ere also p o st-b u ria l is n o t
Diagenesis 433
clear. B ones m ay also m ove h o rizo n ta lly a fte r b u rial, b u t th a t m ay be m ore

difficult to detect. It m ay be th a t evidence o f vertical m o v em en t ca n also be
tak en as evidence o f h o riz o n ta l m ovem ent.
O ne o f the m o st intensive a n d well d o cu m en te d studies o f refitting o f bone
frag m en ts in an a tte m p t to discern p o st-d ep o sitio n al (n o t necessarily p o st
bu rial) m ov em en t o f m am m alian rem ains is W h ite ’s (1992:67-83) analysis o f
h u m a n rem ains recovered from a pu eb lo in the so u th w estern U n ited S tates. H e
suggests ‘"p ost-d epo sition al d istu rb a n c e ” w as m inim al because m ost refits
w ere betw een specim ens w ithin a single ro o m a n d often w ithin a single
co n c e n tra tio n o f fragm ents; very few b etw een -ro o m refits w ere fo u n d (W hite
1992:83). W hite (1992:82) im plies the n u m ero u s refits he identified indicate
m inim al p o st-d ep o sitio n al d istu rb an ce.
A n o th e r w ay to detect p o st-b u ria l m o v em en t w ould be to n o te if the m ost
w eath ered surfaces o f b ones w ere alw ays the u p p e rm o st surfaces. In the
absence o f p o st-b u rial d istu rb an ces, this should be the case (B ehrensm eyer
1981). As well, fau n al rem ains th a t are differentially en cru sted o r co a te d w ith
calcium c a rb o n a te m ay in d icate p o st-b u rial m ovem ent. C alcium c a rb o n a te
co atin g s o ften fo rm o n th e low er o r d o w n w a rd -o rie n te d surface first. T h u s, for
b o th w eath erin g a n d c a rb o n a te co a tin g o b serv atio n s to in fo rm one a b o u t p o s t
burial m o v em ent, th e fossils m u st be collected w ith a p p ro p ria te o rie n ta tio n a l
an d p ro v en ien ce d a ta .
Summary
W hen a fresh bone becom es buried in the earth it undergoes chem ical changes,
differing in n atu re an d degree w ith the chem istry o f the su rro u n d in g m atrix.
(I. W. C ornw all 1956:205)
T he p o st-b u rial m o d ificatio n o f fa u n al rem ains can tak e several form s.

P etrification , m in era liza tio n , co rro sio n , d e fo rm a tio n , fractu rin g , a n d m o v e
m en t seem to be the m ajo r ones. T ypically these are analytically detected by
chem ical a n d m icroscopic analyses. T he influence o f these processes o n analysis
depends, o f course, o n th e analy tical result desired. M in era liz atio n can resu lt in
a perfect replica o f th e o riginal w hereas d e fo rm a tio n ca n p reclude the recording
o f ac cu ra te m o rp h o m etric d a ta . F ra c tu rin g can variously analytically rem ove
specim ens by m ak in g th em unidentifiable, a n d it m ay o bscure traces o f h u m an
b u tch erin g activities an d o th e r processes.
Because m o st p o st-b u rial ta p h o n o m ic processes are co n tro lled in p a rt by the
co n tex t o f th e fau n al rem ains, th a t co n tex t should be considered d u rin g
analysis. T his is tru e o f virtually an y o f the ta p h o n o m ic processes one wishes to
stu dy , b u t w a rra n ts em phasizing in th e co n tex t o f diagenesis because the
geological co n tex t o f fau n al rem ains is the u ltim ate extrinsic fa c to r o f d iag en e
sis, th e m o re p ro x im ate extrinsic facto rs being sedim ent chem istry, texture,
p o ro sity , a n d the like.
12
T A P H O N O M Y OF FI SH, BI RDS,
R EP TI LE S , A N D A M P H I B I A N S
Introduction
In preced in g c h a p te rs I review v ario u s an aly tic techniques fo r assessing the

ta p h o n o m ic h isto ry o f v e rte b ra te fa u n al rem ains. T h a t discussion focuses on
m am m alian rem ains because th a t tax o n o m ic g ro u p has received the m ost
a tte n tio n in the literatu re. But m am m als are n o t the only v erteb rates w ith
w hich zo o a rch a eo lo g ists deal. Birds, reptiles, am p h ib ian s, a n d som e fish are
also verteb rates. M an y o f the an aly tic techniques developed fo r m am m alian
rem ains are also ap plicable to these o th e r v erte b rate taxa. T o illu strate this, in
this c h a p te r I review, w ith less a tte n tio n to detail an d few er exam ples th a n in
previous ch a p te rs, m uch o f the lite ra tu re on n o n -m am m alian v erte b rate
ta p h o n o m y . T he re ad er w ho has ingested a n d digested (keeping this som ew hat
tap h o n o m ic) th e c o n te n t o f p revious c h a p te rs will see m an y parallels am o n g
v ariables studied by tap h o n o m ists w h atev er the tax o n o m ic subject. These
include skeletal com pleteness, n a tu ra l d isarticu latio n sequences, inherent
p ro p erties o f skeletal elem ents such as size, shape, an d stru c tu ra l density, an d
v ario u s k in d s o f m odifications to bones.
T his c h a p te r is n o t m ean t to im ply th a t n o n -m am m alian v erteb rates are less
im p o rta n t tap h o n o m ically th a n m am m alian rem ains. F o r exam ple, it m ay
p ro v e very in terestin g to c o m p are th e ta p h o n o m ic histories fo r each v erte b rate
categ o ry in assem blages rich in fish, birds, a n d m am m als. I am u n aw are o f any
such co m p arativ e study, b u t this form o f co m p arativ e analysis m ay prove
en ligh ten ing b ey ond the details of, say, sm all m am m al versus large m am m al
ta p h o n o m y . If this c h a p te r p ro m p ts th a t kind o f study, if it p ro m p ts o th ers to
p erfo rm m o re n e o ta p h o n o m ic research on n o n -m am m alian v erte b rate
rem ains, o r b o th , it will have fulfilled a g re ater p u rp o se th a n m y illu stra tio n o f
the in tertax o n o m ic parallels in v erte b rate tap h o n o m y .
Fish taphonomy
F ish carcasses are m ore vulnerable to decay th an o th er vertebrates.

(W. Schafer 1962/1972:49)
Fish skeletons and natural processes

T h ere h av e been p ro b a b ly few er th a n tw o dozen articles explicitly concerned
w ith th e ta p h o n o m y o f arch aeo lo g ical fish rem ains p u b lish ed in th e last tw o
434
T aphonom y o f fish, birds, reptiles, and am phibians 435
decades (C olley 1990b). T h a t is so, I suspect, because w hile fish rem ains are
co m m o n in som e arch aeo lo g ical co n tex ts, they ten d to be ra re relative to
m am m al rem ains in m ost sites. Som e arg u e th a t this is so because fish bones are
m o re v ulnerab le to the effects o f differential p re serv atio n th a n m am m al bones
due to the fo rm e r’s relative fragility (e.g., W heeler a n d Jo n es 1989:63), o r the
generally sm all size o f fish rem ains results in failure to recover th em unless
special tech n iques are em ployed. B ut it is also clear th a t som e skeletal elem ents
o f fish preserve extrem ely well, a n d som e fish elem ents are large (C olley 1990b),
ju s t as is fo u n d w ith m am m al a n d b ird skeletons. W h atev er the case, Colley
(19 9 0 b :2 15) is certain ly a t least generally co rrec t w hen she states, “ fish are
subject to th e sam e processes o f ta p h o n o m y a n d site fo rm a tio n as o th e r fau n al
re m a in s.” In p a rt fo r th a t reason, m y c o n sid e ra tio n o f explicitly piscean
ta p h o n o m y is relatively brief.
T o begin, it is im p o rta n t to n o te th a t som e (b u t n o t all) fish are n o t only
v erteb rates, a n d th u s possess h a rd en d o sk eleto n s, they also have scales an d
o to lith s as h a rd p a rts th a t m ight preserve a n d be recovered from archaeological
co n tex ts (C asteel 1976). Scales are form ed in the d erm a l layers, a n d in teleost
fishes are co m p o sed o f a th in sheet o f bonelike m ateria l th a t is so m ew h at soft
a n d flexible (H ild eb ran d 1974:99; R o m er a n d P arso n s 1977:158). O to lith s are
“ e a r sto n es” o f th e m in eral a ra g o n ite th a t h ave a n eq u ilib riu m -related fun ctio n
(C asteel 1976; C olley 1990b; H ild eb ra n d 1974). T he an aly tic an d interpretive
p o ten tial o f fish rem ains is detailed in several p u b licatio n s (e.g., C asteel 1976;
C olley 1990b; W heeler an d Jo n es 1989).
W heeler a n d Jones (1989) p re sen t a useful synopsis o f significant ta p h o n o
m ic facto rs th a t influence fish rem ains (see also C olley 1990b). T he facto rs they
co n sid er can be su m m arized in a fash io n sim ilar to th o se th a t influence o th er
v erte b rate rem ains: th o se th a t alte r th e m o rp h o lo g ical o r physical a n d chem ical
ch aracteristics, o r com pleteness o f bones; a n d th o se th a t affect the spatial
d istrib u tio n o f bones. T hese facto rs are m ed iated by such things as the “ n a tu re
o f the m aterial fo rm in g the h a rd tissue” (W heeler a n d Jo n es 1989:62). W heeler
an d Jo n es n ote m ineralized cartilag e does n o t preserve well, w ithin b o n y fishes
som e b o n es are m o re re sistan t to d e te rio ra tio n a n d m o d ificatio n th a n others,
a n d the bones o f som e species are m o re re sistan t to m o d ificatio n processes th a n
the b ones o f o th e r species. S chafer (1962/1972:56) re p o rts th a t “ carcasses o f
different species o f fish do n o t [naturally] decay in the sam e w ay despite alm ost
equal ex tern al co n d itio n s,” a n d suggests “ the ra tio o f the size o f the b o d y cavity
to the m ass o f th e b o d y determ ines the m o d e o f d ecay .”
W heeler a n d Jo n es (1989:63) suggest fish bo n e rarely preserves in acidic
sedim ent, w hereas sedim ents th a t are n e u tra l o r basic “ are conducive to fish
bo n e survival [but] o to lith s rarely persist except in base-rich d ep o sits.” C oarse
sedim ents can a b ra d e fish rem ains (see also S m ith et al. 1988). S chafer (1962/
1972:61) re p o rts o to lith s are unlikely to be tra n s p o rte d fa r by fluvial processes
due to “ th eir co m p ac t sh a p e ” a n d M artill (1990:271) n o tes th a t “ en h anced
c o n c e n tra tio n s o f o to lith s relative to fish bone are th o u g h t to be a ttrib u ta b le to

the ab ility o f o to lith s to resist acid digestion in fish guts, w here b o n e readily
dissolves.” P o st-d ep o sitio n a l a n d p o st-h u m a n b eh av io ral facto rs include
co n su m p tio n o f h u m a n w aste by scavengers, tram p lin g , a n d w eathering, an d
these v ario usly frag m en t, co rro d e, o r rem ove b ones fro m h u m a n c o n su m p tio n
sites. T hese sam e sites can have fish rem ains ad d ed to them by vario u s n o n
h u m an processes, such as birds a n d o tters dep o sitin g rem ains o f th eir ow n
m eals o f fish o n arch aeo lo g ical sites, o r in fish-consum ing sea m am m al
sto m ach s if the h u m a n o cc u p an ts exploited such m am m als.
M cG rew (1975) exam ines several a ttrib u te s o f a p aleo n to lo g ical co n cen
tra tio n o f fish rem ains. H e suggests the degree o f d isarticu latio n is an in d icatio n
o f h o w d ecom p o sed fish carcasses w ere p rio r to burial; the g re ater the
d isarticu latio n , th e lo n g er the carcasses w ere exposed to b ac te rial ac tio n while
u n d erw ater b u t on the surface o f the lake b o tto m . T he extrem ely high
c o n c e n tra tio n o f fish carcasses in several stra tig ra p h ic layers indicates c a ta s
tro p h ic m o rta lity to M cG rew (1975); he suggests th a t a p erio d o f clim atic
arid ity increased the salinity o f the lake b eyond the p o in t at w hich the fish could
survive, p ro m p tin g non-selective m ass m o rta lity (see A n tia 1979 fo r o th er
exam ples).
A n tia (1979:140) im plies th a t the w eath erin g stages described by B ehrens
m eyer (1978; see C h a p te r 9) can be detected on o to lith s. A n tia (1979:143)
argues th a t p a rtic u la r elem ents o f fish skeletons will, like th o se o f m am m al
skeletons, w eath er at different rates. F ish teeth a n d scales seem to w eath er m ore
slowly th a n fin spines, headshields, a n d en d o sk eletal bones, a n d A n tia
(1979:143) hypothesizes th a t “ the cyam ine o u te r co atin g s o f the teeth an d
scales m ay [help] to shield them from ra p id w eath erin g a n d d isin teg ratio n ,
while th e less well p ro tec ted a n d m o re fibrous fin spines a n d b ones [will]
w eath er m o re ra p id ly .” A n tia (1979:144) p ro p o ses th a t co n stru c tin g a bone
w eath erin g profile (e.g.. F ig u re 9.2) m ay p ro v id e the an aly st w ith “ an estim ate
o f th e m in im um residence tim e o f [fish bones] on the su b stra te su rface.”
E ld er a n d Sm ith (1988) suggest the in ferred ta p h o n o m ic h isto ry o f a
n atu ra lly depo sited fish fau n a can be tested w ith relevant paleoecological d a ta
w hen su b stan tiv e u n ifo rm itarian ism (F igure 3.1) is used to infer the ecology
a n d b eh a v io r o f lo n g -d ea d fishes. A s we have seen in p revious ch a p te rs, the
ecology a n d b eh a v io r o f v e rte b ra te tax a do, in fact, influence the ta p h o n o m ic
h isto ry o f th o se tax a. A fte r review ing how one uses su b stan tiv e u n ifo rm ita ria
nism to re co n stru ct the p aleoecology o f fossil fish. E lder a n d Sm ith (1988) tu rn
to “ th e use o f ta p h o n o m y ” w hich, given th eir discussion, ad h eres to the
principle o f m eth o d o lo g ical u n ifo rm itarian ism .
E lder a n d S m ith (1988) exam ine carcass (soft tissue) decay, tra n s p o rt, and
b u rial, a n d focus o n biological activity (e.g., b acterial decay) an d h y d ro d y n am ic
energy as en tro p ic forces. T hey p o in t o u t how fish fa u n as are influenced by
n ea r-sh o re versus off-shore ecological p redilections o f the tax a, how salinity,
Taphonom y o f fish, birds, reptiles, and am phibians 437
te m p eratu re, ox y g en atio n , a n d o v e rtu rn o f w ater layers influence m o rtality ,

how scavenging o f fish carcasses by b acteria (decay), crayfish, a n d insects can
d isarticu late carcasses, a n d how gas p ro d u c tio n w ithin a d ead fish results in
floating, d riftin g carcasses; w aves, cu rren ts, a n d w inds m ove floating fish
carcasses. T hey n o te th a t u n b u ried fish carcasses will n o t rem ain on the b o tto m
o f a lake unless w a te r te m p e ra tu re rem ains cold, a n d carcasses in n ear-sh o re
lo catio n s w here w ater ten ds to w arm m ay float an d disperse, d ro p p in g skeletal
elem ents as b ac te rial activity rem oves soft tissues (see also S chafer 1962/1972).
T hey suggest th a t m easu rin g the distances betw een p aired skeletal elem ents
allow s d e te rm in a tio n o f tra n s p o rt directio n . W hen a fish carcass has been
scavenged, b ones are ra n d o m ly d istrib u te d a n d have n o p referred o rien ta tio n s.
F lu vial c u rren ts ten d to disperse all elem ents o f p aire d bones in one direction,
a n d m an y elem ents display a p referred o rie n ta tio n . Such o b serv atio n s can (and
sh o uld ) be su p p lem en ted w ith sedim entological, co n tex tu a l, a n d asso ciatio n al
d a ta (C h a p te r 10).
In a series o f experim ents Jo n es (1984, 1986, 1990) fo u n d th a t fish bones fed
to can id s m ay o r m ay n o t be preserved afte r passing th ro u g h the digestive trac t
(see also C asteel 1971; L yon 1970). “ T he m o st co m m o n signs th a t bones have
been ea te n are the presence o f to o th m a rk s a n d crushed v erteb rae. A n o th e r
featu re is surface ero sio n cau sed by acid so lu tio n o f o to lith s, v erte b rae an d
o th e r e le m e n ts . . . O to lith s show ero sio n clearly. T hey have a glossy surface a n d
have lost m o st o f th eir surface relief fe a tu re s” (Jones 1990:143). F ish bone is
v ario u sly d estro y ed by tram p lin g , som e b ones surviving b e tte r th a n others.
F inally, “ fish b o n e loses m uch o f its m echanical stren g th w hen b o iled ” (Jones
1990:144).
C ultural taphonom ic fa c to rs
W heeler a n d Jo n es (1989) a n d C olley (1990b) list m ultiple h u m an tap h o n o m ic
facto rs th a t can influence the k in d s o f fish rem ains fo u n d a n d the species o f fish
rep resen ted in arch aeo lo g ical contexts. T hese include fishing techniques (w here
in a b o d y o f w ater, technology used), b u tch erin g techniques, c o o k in g an d
c o n su m p tio n p ractices, a n d discard practices. Jo n es (1984, 1986) show s th a t
c o n su m p tio n a n d d igestion o f fish b o n e by ra ts, a dog, a pig, a n d a h u m a n result
in the loss o f som e b o nes, fra g m e n ta tio n o f som e bones (see also C asteel 1971),
an d only a few (a b o u t 13% ; W heeler a n d Jones 1989:72) o f the originally
ingested b o nes w ere recognizable u p o n recovery fro m feces. M asticatio n
v ariou sly results in fra g m e n ta tio n a n d d isto rtio n , a n d digestion results in
c o rro sio n o f bones. R ich ter (1986) show s th a t h ea tin g fish b o n e to 60°C w ith
eith er so ft tissue scrap ed off o r soft tissue ad h e rin g to the b o n e results in m elted
ends o f collagen fibers. " A t 80°C the m elted areas alo n g the fibrils h ad increased
a n d sh o rt sticks o f collagen w ith sw ollen ends w ere seen. Very few u n d am ag ed
fibrils w ere p re se n t” (R ic h te r 1986:479). B oiling rem oved all evidence o f native
collagen fro m fish bone. R ich ter (1986:480) concludes by n o tin g th a t diagenetic
processes m ay also rem ove traces o f h eat-m odified collagen (e.g., M arch iafa v a
e ta l. 1974).
R etrieval o f fish rem ains is n o to rio u sly difficult because m an y fish bones are
relatively sm all. W hile som e fish b ones are large en o u g h n o t to require
screening, “ experim en ts have show n th a t sm aller fish rem ains can easily be
m issed unless [sedim ent] is passed th ro u g h m inim um m esh sizes ranging
betw een 0.5 a n d 2 m m ” (C olley 1990b:208). S m ith e ta l. (1988) suggest th a t one
reaso n fossil fish fa u n as are less tax o n o m ically rich th a n m o d ern fish fa u n as is
because o f the difficulty o f recovering the (sm all) rem ains o f sm all tax a a n d it is
precisely th e sm all fish tax a th a t o ften m ak e u p m u ch o f th e tax o n o m ic richness
o f a fish fa u n a. Several techniques have been d escribed fo r co n ten d in g w ith the
differential recovery o f fish rem ains (C olley 1990b; W heeler an d Jo n es 1989).
F ish seem to die attritio n a lly in typical situ atio n s, b u t can die in large
n u m b ers o r ca ta stro p h ically u n d e r som e co n d itio n s (S chafer 1962/1972:51;
W eigelt 1927/1989:163). V an N eer a n d M uniz (1992) re p o rt an e th n o a rc h a e o
logical case in w hich fisherm en cleaned th eir nets o f fish to o sm all to w a rra n t
processing, a n d in so d o in g a “ fish m id d e n ,” a dense co n c e n tra tio n o f
artic u la te d fish sk eletons piled u p o n one a n o th e r a n d a b o u t 3 x 3 m in
h o rizo n ta l extent, w as created. T hey n o te th a t such a n th ro p o g e n ic ac cu m u
lation s, if excavated from a p re h isto ric co n tex t, m ight be confused w ith m ass
m o rta lity o f fish tra p p e d in a seasonally desiccated p o o l as b o th kinds o f
ac cu m u latio n ten d to have few associated artifacts. V an N eer a n d M uniz
(1992) suggest th a t these tw o k in d s o f a c cu m u latio n m ay be d istin g u ish ed by
the fact th a t an a n th ro p o g e n ic ac cu m u latio n will have a g re ater n u m b e r o f
species rep resen ted in it th a n a n a tu ra l accu m u latio n .
W heeler a n d Jo n es (1989:78) re p o rt th a t “ n a tu ra l agencies d ep o sitin g fish
bones rarely p ro d u c e su b sta n tia l co n c e n tra tio n s o f the kinds o f fish preferred
as h u m an fo o d .” B utler (1987, 1990) has, how ever, d o cu m en ted ju s t such a
case. T h e season al sp aw ning h ab its o f a n a d ro m o u s Pacific salm o n (O neorhyn-
chus spp.) o ften result in the seasonal ac cu m u latio n o f n u m ero u s rem ains o f
these fish in q u ite sm all areas alo n g river b an k s a n d islands. B utler (1987, 1990)
w as specifically co n cern ed w ith developing an aly tical techniques a n d criteria
th a t w o u ld allow h er to d istin g u ish such n a tu ra lly created c o n c e n tra tio n s o f
salm o nid rem ain s fro m cu ltu ra lly created c o n c e n tra tio n s as salm o n w ere (and
are still) a m a jo r food source fo r m an y h u m a n g ro u p s alo n g th e Pacific rim o f
n o rth w e ste rn N o rth A m erica. T o w a rd th a t end she collected all salm onid
rem ain s fro m a sm all p o in t b a r in th e m iddle o f a river in w estern W a sh in g to n
state. She th e n stu d ied th o se rem ains, c o m p a rin g v ario u s a ttrib u te s o f this
n a tu ra l assem blage w ith a ttrib u te s o f a sta n d a rd salm onid skeleton and
a ttrib u te s o f several arch aeo lo g ical assem blages.
B utler (1987, 1990, 1993) d em o n strate s th a t b o th the degree o f carcass
co m pleteness a n d th e frequency o f p a rtic u la r skeletal elem ents allow d istin c
c
o
tr
Cranial
8.
O □ Post-cranial
CL
Figure 12.1. P ro p o rtio n a l frequencies o f salm onid cranial an d p o st-cranial

rem ains in a sta n d a rd salm onid skeleton, a n atu ral dep o sit o f salm onid rem ains,
and five archaeological sites (m odified from B utler 1990:189, Figure 8.13).
tio n betw een cu ltu ra l a n d n a tu ra l deposits o f salm o n id rem ains. N a tu ra l

d ep o sits c o n ta in relatively com plete carcasses a n d th u s a b o u t equal p ro p o r
tions o f c ran ial a n d p o st-c ra n ia l elem ents. C u ltu ra l (archaeological) deposits
co n tain relatively inco m plete carcasses a n d p o st-cran ial elem ents are m uch
m ore a b u n d a n t th a n c ran ial elem ents. She calls u p o n the stru c tu ra l density (see
below ) a n d size o f b ones as m ed iatin g n a tu ra l ta p h o n o m ic processes, a n d the
d istrib u tio n o f edibile tissue, a n d b u tch ery a n d disposal p ractices as influencing
cu ltu ra l ta p h o n o m ic processes. D espite th e fact th a t she exam ined th o u sa n d s
o f fish b o n es from b o th n a tu ra l a n d cu ltu ra l d eposits, she fo u n d no evidence o f
b u tch ery m ark s, ab ra sio n from fluvial processes, o r gnaw ing m ark s o f c a rn i
v o ro u s scavengers, b u t she d id find a n o n ra n d o m p a tte rn o f b u rn in g o f skeletal
elem ents suggesting co o k in g a t one arch aeo lo g ical site. B ones less deeply
b u ried in th e soft tissues o f the body ten d ed to be m ore o ften b u rn e d th a n bones
m o re deeply b u ried , suggesting co o k in g w as o f carcasses w ith the flesh a ttac h ed
to th e skeleton.
B utler (1990, 1993) co m p ared the frequencies o f cran ial an d p o st-cran ial
elem ents betw een a s ta n d a rd skeleton, the n a tu ra l dep o sit, an d five a rc h a e o lo
gical d ep o sits (F ig u re 12.1). T he n a tu ra l d ep o sit varies in relative frequencies o f
cran ial a n d p o st-cran ial skeletal elem ents fro m a sta n d a rd skeleton due, in
p a rt, to fluvial w innow ing, w eath erin g , a n d scavenging. T w o o f the a rc h a e o lo
gical d ep o sits (45K I23, 4 5 D 0 2 1 1 ) have relatively low frequencies o f cran ial
p a rts due, B utler (1990:130) suggests, to “ w hole carcass d ep o sitio n a n d in situ
T ab le 12.1 A verage ske le ta l com pleteness ratios fo r

various sized horizontal units and sites (fro m Butler
1990)
Spatial u n it size
A ssem blage 4x4 m 4x2m 2x2m 2 x 1m
N atu ral A ssem blage 0.65 0.68 0.71 0.63

45K I23 0.44 0.47 0.45 —
45D 0211 — — — 0.64
4 5D 0 2 8 5 — — 0.08 0.13
45LE222 0.85 0.66 0.72 —
35WS5 0.42 0.44 0.54 —
d e stru c tio n .” T h e arch aeo lo g ical d ep o sit w ith high frequencies o f cran ial p a rts
(4 5 D 0 2 8 5 ) resulted fro m “ alm o st exclusive d ep o sitio n o f salm o n h e a d s.” T he
assem blage recovered from a ro ck sh elter (45LE222), unlike the o th e r a rc h a e o
logical assem blages, is in a non -riv erin e setting a n d ap p e ars to be sim ilar to the
n a tu ra l d ep osit, leading B utler (1990:159) to suggest the ra tio o f cran ial to p o s t
cran ial skeletal p a rts m ay be d iag n o stic o f cu ltu ra l an d n a tu ra l accu m u latio n s
only in riverine settings, such as is the case w ith the o th e r fo u r archaeological
collections. T he final arch aeo lo g ical assem blage (35W S5) is the assem blage
th a t p ro m p te d B u tle r’s research, a n d she concludes th a t it ap p e ars to be a
cu ltu ra l d ep o sit, a lth o u g h vario u s p ro b lem s w ith the sam ple available fo r her
analysis preclude a firm conclusion.
B utler (1990:131, 1993) m easu red skeletal com pleteness by calcu latin g the
ra tio o f h ead to tru n k M A U s w ith the eq u atio n :
MAU based on most frequent cranial element:MAU based on most frequent post-cranial element [12.1]
w ith the largest o f tw o values used as the d e n o m in a to r (B utler 1987, 1990:30).

R atio s a p p ro a c h in g one indicate essentially com plete carcasses w hereas ra tio s
a p p ro a c h in g zero suggest dissim ilar p ro p o rtio n s o f fro n t a n d re a r ends. B utler
(1990:131) p red icted th a t fluvially (n a tu rally ) created deposits sh o u ld have
h ig h er co m p leteness ra tio s th a n c u ltu ra l deposits, w ith the im p o rta n t q ualifica
tio n th a t “ ra tio s w o u ld be larg er a t sm aller sp atial scales in n a tu ra l sites th a n in
cu ltu ra l d ep o sits.” T h a t is, n a tu ra l deposits a t v irtu ally an y sp atial scale (in so
far as th e sp atial u n it is n o t significantly less th a n the average size o f an
in d iv id u al fish) sh o u ld h ave co m p lete carcasses, while com pleteness ra tio s
ap p ro a c h in g one sh o u ld be fo u n d m o re freq u en tly a t large th a n at sm all spatial
scales in c u ltu ra l d ep o sits due to the b u tch ery a n d b o n e-dispersal activities o f
h u m an s. G iven the lim itatio n s o f the d a ta av ailab le to her, B utler co u ld n o t
calculate co m pleteness ra tio s a t the sam e sp atial scales fo r all assem blages.
Basically, how ever, b u t n o t w ith o u t exception, h er p re d ic tio n is b o rn ou t
(T able 12.1). T w o arch aeo lo g ical assem blages have low er average co m p lete
T aphonom y o ffis h , birds, reptiles, and am phibians 441
ness ra tio s th a n the n a tu ra l assem blage, b u t tw o o th e r arch aeo lo g ical assem

blages h av e com pleteness ra tio s equal to o r g re ater th a n the n a tu ra l assem
blage, p ro m p tin g h er to conclude th a t the com pleteness ra tio is “ n o t alw ays a
clear in d ic a to r o f d ep o sit o rig in " (B utler 1990:134).
B utler (1990:48-49, 1993) also assessed the p o te n tia l ta p h o n o m ic effects o f
skeletal elem ent size, m easu red as the m axim um length o f a specim en. She
distin gu ish ed five size classes by 1.5 cm intervals; size class 1 includes elem ents 0
to 1.5 cm in length, size class 2 includes elem ents 1.5 to 3 cm in length, an d so
fo rth . She fo u n d n o c o rre la tio n betw een the frequency o f each size class as
rep resen ted in h er n a tu ra l sam ple, an d size class, co n clu d in g “ sm all elem ents
have a b o u t the sam e survivorship as large o n es” (B utler 1990:69). Sim ilarly,
she fo u n d no statistically significant relatio n betw een these tw o variables in the
arch aeo lo g ical assem blages, suggesting elem ent size h ad n o t played a role in
th eir ta p h o n o m ic histories.
B utler (1990:46-47) m easu red the stru c tu ra l density fo r 51 elem ents o f one
skeleton, using the sam e p h o to n d en sito m etry techniques used by L ym an
(1984a; L y m an et al. 1992a) a n d K re u tzer (1992) for m am m als (T able 12.2;
S tew art [1991] p ro v id es m easures o f b o n e density fo r th ree elem ents o f each o f
several tax a, b u t h o w th o se m easu rem en ts w ere derived is n o t clear). B utler’s
density values are, how ever, n o t the sam e as th o se available for u n g u lates as she
did n o t divide the density value by bo n e thickness, th u s h er values are pro p erly
called “ lin ear d en sities” (L ym an 1984a). She fo u n d a w eak b u t statistically
significant co rre la tio n betw een skeletal elem ent frequencies in the n a tu ra l
d ep o sit a n d bo n e density (rs = 0.308; 0.05 > P > 0.02 [B utler 1990:68]), suggest
ing to m e th e p o te n tia l th a t som e w innow ing o f less dense elem ents by fluvial
processes h ad o ccurred. She also fo u n d statistically significant relations
b etw een b o n e density a n d skeletal p a rt frequencies in tw o o f fo u r o f the
arch aeo log ical assem blages (P < 0.001), suggesting d en sity -m ed iated su rv iv o r
ship influenced b o n e frequencies in th o se tw o assem blages.
S m ith et al. (1988), like B utler (1987, 1990, 1993), call u p o n the rep ro d u ctiv e
b eh a v io r o f fish as a n im p o rta n t ta p h o n o m ic facto r, p a rtic u la rly th e tem p o ra l
co rresp o n d en c e o f sp aw n ing m ig ra tio n a n d river level flu ctu a tio n . T hey fo u n d
the p ro b a b ility th a t fish b ones will be b u ried in flo o d p lain sedim ents goes up as
the a b u n d a n c e o f fish increases (p o ten tial m o rta lity is high, especially w ith taxa
like a n a d ro m o u s salm o n id s th a t spaw n in high frequencies a n d die afterw ard )
a n d the ra te o f sed im en tatio n increases. P re d a tio n o f y o u n g m ig h t reduce the
p ro b a b ility th a t th eir rem ains will survive to be deposited. P hysiological
v a ria tio n betw een fish tax a m ay result in a high p ro b a b ility for som e tax a to
en ter th e fossil reco rd a n d a low p ro b a b ility fo r o th er taxa. F o r exam ple, d u rin g
perio d s o f d ro u g h t, fish th a t need high levels o f oxygen in the w ater are m ore
a p t to die th a n catfish w hich can survive w a te r w ith little oxygen. S m ith et al.
(1988) suggest stru c tu ra lly den ser bones are m o re p ro n e to survive a b ra sio n
a n d acidic co rro sio n , a n d also ten d to be the m o st tax o n o m ically distinctive.
T ab le 12.2 S tru ctu ra l density (g /c m 2) o f coho salm on fO n c o rh y n ch u s

k istu c h j skeleta l elem ents (fro m B utler 1990)
E lem ent D ensity Elem ent D ensity
ang ular 0.063 basioccipital 0.023

basisphenoid 0.048 ceratohyal 0.040
dentary 0.042 ectopterygoid 0.027
epihyal 0.027 epiotic 0.067
exoccipital 0.142 frontal 0.044
hyom andibula 0.056 hypohyal, upper 0.049
hypohyal, low er 0.053 interopercle 0.024
lingual plate 0.056 m axilla 0.076
m esopterygoid 0.027 m etapterygoid 0.016
opercle 0.036 opisthotic 0.017
o tolith (sagitta) 0.072 p alatine 0.040
parasphenoid 0.064 prefro n tal 0.037
prem axilla 0.062 preopercle 0.038
prootic 0.077 p terosphenoid 0.057
pterotic 0.059 q u ad rate 0.084
sphenotic 0.054 subopercle 0.021
supraethm oid 0.023 supraoccipital 0.070
urohyal 0.062 vom er 0.053
clei thrum 0.037 coracoid 0.039
m esocoracoid 0.036 p ecto ral fin ray 0.036
postcleithrum a 0.020 p o sttem p o ral 0.048
scapula 0.060 supracleithrum 0.023
basipterygium 0.080 v erteb ra type 1 0.185
vertebra type 2 0.220 v erteb ra type 3 0.203
vertebra type 4b 0.128 cau d al bony plate 0.033
hy pural0 0.080
Notes:
11 m iddle elem ent o f series.
b average o f vertebrae 4a an d 4b.
c for H 2.(B utler [1990:40; 1993:8] distinguished fo u r basic types o f vertebrae. F ro m an terio r
to po sterio r these are: T ype 1 is the first [anterior-m ost] vertebra in the colum n an d has two
prom in ent d o rso -an terio r facets for articu latio n w ith the exoccipitals. Type 2 lack
parapophyses an d fused neural and haem al spines, and have tw o distinct orifices on their
dorsal an d ventral surfaces. T ype 3a has fused neural spines, type 3b has fused neural spines
an d abbreviated haem al spines, type 3c h as fused neural an d haem al spines. T ype 4a has
fused neural spines th a t are m ore ro b u st th a n those o f 3a, type 4b lacks neural an d haem al
spines and has a single orifice on b o th d orsal an d v entral surfaces.)
T h u s th o se tax a w ith dense b ones will be m o re readily identified a n d m o re likely

to be preserved in the fossil record.
S tew art (1989:79) w as in terested in developing criteria th a t w ould allow h er
to d istin g u ish cu ltu ra lly from n a tu ra lly d ep o sited lacu strin e fish-bone assem
blages. She collected three assem blages fro m g ro u n d surfaces alo n g the
sh oreline o f L ak e T u rk a n a in A frica. T w o w ere n a tu ra l deposits given the
d esig n atio n s PS1 (from an area 90 x 90 m ) a n d PS2 (90 x 110 m ), the th ird w as a
site w here h u m an s h a d ro a ste d a n d ea te n fish a n d w as given the d esig n atio n
AS1 (11.4 x 9.5 m ). T h e frequencies o f collected fish bones p er assem blage w ere
910, 360, a n d 651, respectively. S tew art exam ined fo u r a ttrib u te s o f each
assem blage: (1) frequency o f b ones p er u n it area; (2) skeletal p a rt frequency; (3)
tax o n o m ic ab u n d a n ce; a n d (4) tax o n o m ic diversity.
T h e frequency o f b o nes p er m 2 w as hig h er in the c u ltu ra l a c cu m u latio n (6.01
b ones p er m 2) th a n in eith er o f the tw o n a tu ra l assem blages (PS 1 = 0.11;
PS2 = 0.03). T he frequencies o f skeletal p a rts w ere c o m p ared to th eir freq u en
cies in a fish skeleton, an d tax o n o m ic ab u n d a n ces a n d diversity w ere co m p ared
to th o se values as reflected in the lake by living fishes. S tew art (1989:80-81)
suggests th a t a n average fish in L ake T u rk a n a w ould consist o f 65 (54.2% )
cran ial elem ents. 40 (33.3% ) verteb ral elem ents, a n d 15 (12.5% ) n o n -v e rteb ra l
p o st-cran ial elem ents. T he first value w as reduced from the average o f 102
cran ial elem ents by o m ittin g th o se elem ents S tew art deem ed unlikely to survive
because o f th eir th in , delicate stru ctu re . As well, rays, ribs, a n d b ra n ch ial
elem ents w ere n o t included. S te w art’s (1989:82) o p eratin g assu m p tio n s w ere
th a t in n a tu ra l assem blages, skeletal p a rts will occu r in p ro p o rtio n to their
n a tu ra l occu rren ce in an average fish, a n d processing a n d c o n su m p tio n
practices o f h u m an s will result in p ro p o rtio n s o f skeletal p a rts different fro m an
average fish.
S tew art (1989) fo u n d th a t the ra tio o f c ran ial to v erteb ral elem ents w as 1.6
for a n average fish sk eleton, 1.4 fo r the n a tu ra l assem blages (PS 1 a n d PS2), an d
9.5 fo r th e cu ltu ra l assem blage (AS 1). T his results because the b o n e -a c c u m u la t
ing h ab its o f peo p le lead to denser c o n c e n tra tio n s o f fish bones a n d d u e to the
m ultiple o cc u p atio n s o f h a b ita tio n sites (S tew art 1989:96). In the n a tu ra l
assem blages, n o n -v e rteb ra l p o st-cran ial elem ents tend to be well represented
relative to cran ial elem ents a n d verteb ral elem ents, an d the la tte r tend to be
rare. In th e cu ltu ra l assem blage c ran ial elem ents a n d n o n -v e rteb ra l p o s t
cran ial elem ents are relatively a b u n d a n t a n d v erte b rae are rare. O n the basis o f
these o b serv atio n s, S tew art (1989:92) concludes th a t high p ro p o rtio n s o f
cran ial elem ents ch aracterize cu ltu ra l assem blages a n d d istinguish th em from
n a tu ra l ones. T he ra tio o f cran ial to v erteb ral elem ents is sim ilar betw een the
n a tu ra l dep o sits a n d her average fish, a n d “ is useful for ch a racterizin g the
skeletal elem ent co m p o sitio n o f a n atu ra lly -d ep o site d assem b lag e” because
the ra tio is low er fo r n a tu ra l dep o sits th a n cu ltu ra l ones (S tew art 1989:96).
S tew art (1989:97-98) concludes th a t n a tu ra l deposits te n d to reflect the

n a tu ra l fish p o p u la tio n in term s o f relative tax o n o m ic ab u n d a n ces a n d
tax o n o m ic diversity, w ith the im p o rta n t cav eat th a t fish tax a w hich seldom
exceed 35 cm in length are ra re a n d m ay be lost " in the fossilization pro cess.”
M o d ern fisherm en ex p loiting L ake T u rk a n a catch fish “ a p p ro x im a te ly in
p ro p o rtio n to th eir presen t a b u n d a n c e .” U sing S te w art's (1989:94-95) d a ta for
the m o d ern co m p o sitio n o f L ake T u rk a n a ’s fish fa u n a in term s o f biom ass, and
h er values fo r th e biom ass o f fish rep resen ted by the c u ltu ra l d ep o sit (A S1) and
th e sum m ed n a tu ra l d eposits (PS1 + PS2), if only tax a fo u n d in the AS1
assem blage are included, S p e a rm a n 's rh o betw een p airs o f assem blages are:
m odern catch r, = 0.90
P = 0.07
cultural deposit r, = 0.60 0.30
P = 0.23 0.56
n atu ral deposit m odern catch
These coefficients suggest th e n a tu ra l assem blage (PS1 + PS2) is sim ilar to the
m o d ern fish fau n a o f the lake, a n d the cu ltu ra l assem blage (A S 1) is dissim ilar to
b o th the m o d ern an d n a tu ra l assem blages. If only tax a fo u n d in the n atu ra l
assem blage are used, coefficients are:
m odern catch rs = 0.86
7^ = 0.01
cultural deposit r, = 0.68 0.37
P = 0.01 0.33
n atu ral deposit m odern catch
I f all tax a fo r w hich S tew art calcu lated bio m ass are included, the c o rrelatio n
coefficients are:
m odern catch r, = 0.83
/> = 0.008
cultural deposit /-s = 0.78 0.50
/ >= 0.01 0.11
n a tu ral deposit m odern catch
These coefficients suggest th a t sam ple sizes are influencing the results in such a
m a n n e r as to m ake the significance o f these co m p ariso n s am b ig u o u s. Species
richness p e r assem blage, fo r exam ple, is 15 fo r th e m o d e rn sam ple o f live fish
fro m th e lake, th e su m m ed n a tu ra l assem blages have 9 tax a, a n d th e cu ltu ral
assem blage h as 6 tax a. T he coefficients increase due to ties in th e ra n k o f
m issing tax a, a n d the significance levels increase due to the increase in the
n u m b e r o f categ o ries being co rrelated . T h u s the cu ltu ra l assem blage is n o t
co rrelated w ith th e m o d e rn assem blage in an y o f the th ree co m p ariso n s, a n d is
only significantly co rrelated (P < 0 .0 5 ) w ith the n a tu ra l assem blage w hen all
tax a are included.
In a m o re extensive study th a t included a d d itio n al n a tu ra l deposits o f fish
rem ains, S tew art (1991) again focuses o n establishing criteria distinctive o f
T able 12.3 S u m m a ry o f criteria fo r distinguishing culturally fr o m naturally

deposited assem blages o f fish rem ains around large lakes (fro m S tew a rt 1991)
N atu rally deposited assem blages C ulturally deposited assem blages
1. contain no sm all individuals 1. co n tain “ m edium -sized” fish

a. sm all taxa an d sm all individuals are a. size distrib u tio n s o f individual fish
absent ( < 35 cm length o f individual fish) dependent on w here fishing done and
technology used to take fish
2. density-m ediated attritio n strongly reflected 2. density-m ediated attritio n weakly
a. m ore strongly w ith increasing age reflected, if at all
b. bones o f low stru ctu ral density
destroyed,
tax a w ith only low density elem ents no t
represented
3. tax a living in near-shore zone m o st frequent 3. m ay be near-sh o re taxa, o r others
4. m any epaxial elem ents, b u t ab u n d an ce 4. relatively few axial elem ents
relative to cranial and axial elem ents sim ilar a. relatively m o re cranial and epaxial
to relative abundances in an average fish elem ents th a n axial elem ents due to
differential butchery and tran sp o rt
5. high taxonom ic diversity 5. low taxonom ic diversity
a. slightly low er diversity than in living a. n o tab ly low er th an in living fish
fish fauna fauna
low density o f fish rem ains p er u n it area* 6. high density o f fish rem ains per u n it area
m ore com plete skulls 7. m o re fragm ented skulls
no b u rn ing or b u tchering m arks 8. som e elem ents b u rn ed an d som e have
b utchering m arks
N o te :
a B ut see discussion here o f B utler (1987, 1990), an d M cG rew (1975).
n atu ra lly a n d c u ltu rally d ep o sited assem blages. H er results fo r fish fa u n as from
large lake h a b ita ts are sum m arized in T ab le 1 2 .3 .1 em phasize th a t the criteria
deem ed d iag n o stic by S tew art (1991) are fo u n d ed on lim ited d a ta ; th a t is, the
n u m b er o f cases she (or an y o n e else, including B utler [1987, 1990]) has
ex am in ed is n o t large, n o r are these cases rep resen tativ e o f th e m y riad possible
v arian ts. T he d a ta in T ab le 12.3 (along w ith those described by B utler [1987,
1990]) do, how ever, rep resen t a m a jo r ad v an cem en t in o u r know ledge ab o u t
p iscean tap h o n o m y .
B oth B utler a n d S tew art follow ed the lead o f tap h o n o m ists focusing on
m am m als an d explored a ttrib u te s th a t w ould allow them to distinguish
c u ltu rally from n atu ra lly d ep o sited assem blages o f fish bones. T h eir results are
som ew h at d issim ilar as reg ard s ra tio s o f c ran ial to p o st-cran ial elem ents,
p erh ap s because B u tler stu d ied a riverine assem blage w hereas S tew art exa
m ined a lake-sh ore assem blage o r because the w ay the tw o o f them defined
“ p o st-c ra n ia l” differs. T hey b o th have m ad e significant co n trib u tio n s to o u r
u n d ersta n d in g o f fish ta p h o n o m y th a t are in need o f a d d itio n al actualistic
research fo r clarification an d c o rro b o ra tio n .
Avian taphonomy
The limb bones are usually b ro k en in to either tw o o r three pieces . . . M an y o f the

broken lim b bones are ch arred a t the b ro k en end . . . T he b ro k en state o f the bones
m ay indicate attem p ts a t m aking w histles o r o th er artifacts, or the m an n er o f
p rep aring the birds to eat.
(H. H ow ard 1929:311, 384)
In te rp re ta tio n o f b ird rem ains o ften m irro rs th a t fo u n d in analyses o f

m am m alian rem ains. T h ere has been p erh ap s less ta p h o n o m ic research on bird
rem ains th a n on fish rem ains, a lth o u g h som e im p o rta n t d a ta gen erated by
w ildlife biologists are available. P erh a p s th ere is a p au c ity o f ta p h o n o m ic
research on b ird rem ain s because they tend to be rare relative to m am m alian
rem ain s in m o st arch aeo lo g ical sites, b u t it is n o t clear w h a t the overall typical
relative ab u n d a n ces o f fish a n d b ird s are in such sites. B ird bones are often
fo u n d in sites, if in low ab u n d a n c e , a n d u n d e r exceptional p re serv atio n al
co n d itio n s such as d ry cave deposits even feath ers a n d quills have been fo u n d
(L ivin g ston 1988). A s w ith m o st v e rte b ra te gro u p s, screening sedim ent results
in the recovery o f rem ain s o f sm all bird tax a th a t will n o t otherw ise be found
(P ark er 1988).
S chafer (1962/1972:41) re p o rts th a t bird m o rtality can be c a ta stro p h ic “ at
tim es o f extrem e w e ath er c o n d itio n s especially in the spring a n d a u tu m n w hen
m an y bird s are m ig ra tin g .” W hen a b ird dies a n d its carcass is on land, the soft
tissues d ry a n d sh rin k , th u s the h ead a n d neck “ bend b ac k w ard a n d the tail
w ith its feath ers bend u p ” a n d the wings tu rn aw ay fro m th e b o d y (S chafer
1962/1972:46, 47), a p h en o m e n o n re p o rte d thirty-five years earlier by W eigelt
(1927/1989:105-106). A s bird carcasses d e te rio ra te u n d e r n a tu ra l co n d itio n s,
the hin d lim b s sep a rate from the tru n k , a n d then the pelvis from the lu m b a r
vertebrae; b ones o f the forelim b a n d sh o u ld er girdle alo n g w ith the sternum
“ co n tin u e to h old to g e th e r as a u n it fo r a long tim e” (S chafer 1962/1972:48).
H a rg rav e (1970:59) suggests th a t d u rin g ex cav atio n o f av ian rem ains “ care
sho u ld be ta k e n to collect som e trac h eal rings (w indpipe), since [he] considers
them to be p ro o f th a t the b ird w as bu ried in the flesh. In like m an n er, the
presence o f calcified ten d in al splints o f the leg m uscles o f tu rk ey s (M eleagris
gallopavo) also in dicate th a t the tu rk ey w as bu ried in the flesh.” H e is no d o u b t
correct if these a n a to m ica l p a rts are in p ro p e r an a to m ica l position. H arg rav e
(1970:59) also suggests th a t the presence o f a rtic u la te d w ing bones, b u t “ n o t the
h u m eru s . . . co uld in d icate th a t a w ing w as used as p a ra p h e rn a lia o r fo r an
o rn a m e n t” by h u m an s.
R ich (1980) suggests th a t v a ria tio n in skeletal p a rt frequencies she observed
in a T e rtia ry d ep o sit in S o u th A frica w as d u e to fluvial so rtin g an d w innow ing.
She suggests co raco id s, ta rso m e ta ta rs i a n d tib io tarsi are a b u n d a n t th ere due to
th eir relatively high s tru c tu ra l density, a n d sm all, elo n g ate a n d slender
elem ents, a n d those w ith high su rface-to -v o lu m e ra tio s are ra re d u e to their
g re ater p o te n tia l o f fluvial tra n s p o rt. R adii, ca rp o m e ta c a rp i, fibulae and
T aphonom y o f fis h , birds, reptiles, and am phibians 447
fu rculae m ay be ra re in R ich ’s sam ple d u e to th eir relatively g re ater fragility

a n d the difficulty o f identifying them if they are fra g m en tary (L ivingston
1989:539).
R ich (1980) h ad virtually no actu alistic d a ta to inform h er in te rp re ta tio n s.
R elev an t actu alistic research w as p erfo rm ed by B ickart (1984), w ho d o c u
m ented th e decay, d isarticu latio n , dam age, a n d fluvial tra n s p o rt o f tw enty-
eight b ird carcasses o f th ree species. H e fo u n d th a t unless p ro tec ted in cages o r
ra p id ly b u ried , b ird carcasses are quickly rem oved by scavengers. U n d istu rb e d
carcasses “ becam e firm ly stu ck to the su b stra te w ithin several days, possibly by
a c o m b in a tio n o f bo d y fluids a n d g ro u n d m o istu re ” (B ickart 1984:527).
“ S tu c k ” carcasses w ere n o t sub seq u en tly m oved by fluvial ac tio n except w hen
“ the m o st severe sto rm o f the sea so n ” stru ck , a n d even then the b ird b ones w ere
“ little m oved fro m th eir original p o sitio n s” (B ickart 1984:528). W ith in a year
a b o u t h a lf o f the b o nes o f four, relatively un m o v ed carcasses w ere buried.
D ecay a n d d isarticu latio n varied across individuals, ta k in g a b o u t tw o weeks
for som e ind ivid uals a n d six m o n th s for others. B ickart (1984:528) re p o rts th a t
d isarticu latio n to o k th e follow ing order: (1) ribs fro m stern u m ; (2) hindlim b
jo in ts; (3) v ertebrae; (4) p ecto ral girdle (sternum -coracoid?); (5a) proxim al
w ing jo in ts (co raco id a n d sca p u la -h u m eru s first; th en h u m eru s a n d radius-
u ln a, th en co raco id -scap u la); a n d (5b) distal w ing jo in ts (m anus). B ird bones
d am ag ed by scavenging ca rn iv o res display “ green fractu res typical o f fresh
b o n e ,” a n d 84% o f th e recovered long b ones h ad one o r b o th artic u la r ends
rem oved (B ick art 1984:531). A fte r one y ear o f exposure, w eath erin g d am age
w as v irtu ally n o n -e x istan t, alth o u g h the b ones w ere in relatively m oist, shaded
settings so we sh o u ld n o t expect them to be w eath ered (see C h a p te r 9).
R osene an d L ay (1963) re p o rt th a t o f sixty b o b w h ite quail (Colinus
virginianus) carcasses they placed in ag ric u ltu ral fields in A lab a m a a n d T exas,
2 5 -5 0 % d isap p ea red com pletely w ithin fo u r days d u e to the activities o f
scavengers. Som e indiv idual q u ail d isap p ea red com pletely w ithin 24 ho u rs,
a n d afte r 30 days there w ere v irtu ally no traces o f any o f the 60 quail. B alcom b
(1986) placed carcasses o f 78 so n g b ird s (m ainly passerines) in co rn fields a few
days a fte r p lan tin g . A fte r five days, 72 carcasses h ad been rem oved by
scavengers, an d from this a n d o th e r d a ta B alcom b (1986:819) concludes the
average tim e a carcass survived was a b o u t 1.2 days, the rate o f carcass
d isap p ea ran c e w as g re atest d u rin g the first 24 h o u rs, a n d over h a lf th e carcasses
were to tally rem oved w ith o u t a trace (i.e., n o t even a fe ath er rem ained). T obin
an d D o lb eer (1990) fo u n d th at bird carcasses d isa p p e a r from cherry a n d apple
o rc h a rd s w ithin 8 -10.5 days, on average, due to scavenging activity. G ro u n d
cover did n o t affect the d u ra tio n o f carcass survival. W o rk in g w ith bird
carcasses placed in catta il m arshes, Linz et al. (1991) found th a t m o re carcasses
were rem oved by scavengers as carcass density increased. As well, a g re ater
p ro p o rtio n o f carcasses w as rem oved fro m shallow w a te r ( < 1 5 cm ) d e p o sitio
nal loci th a n fro m deep w ater ( > 30 cm ) loci.
G ra h a m a n d O liver (1986) stu d ied a set o f over 200 A m erican co o ts (Fulica
am ericana) th a t froze in to a lake. T his p ro d u c e d a classic in stan ce o f c a ta

stro p h ic m o rtality . C o o t carcasses w ere first scavenged by gulls (L arus sp.), b u t
while c o o t b ones w ere exposed by this scavenging few were b ro k e n o r
d isarticu lated . H ead s did fall off, an d m an y carcasses a p p a re n tly d isarticu lated
in to a n te rio r (cervical v ertebrae, wings, co rac o id /ste rn u m com plex) a n d p o s
terio r (synsacrum , h ind lim bs) halves. L ate r scavenging by m am m als p ro d u ced
m o re b o n e b reak ag e a n d scattering, w ith “ p o in ts o f a tta c k usually a t the b reast
o r h in d lim b s.”
E ricson (1987) c o m p ared the ra tio s o f a n te rio r lim b elem ents to p o sterio r
lim b elem ents in 54 assem blages, including tw o m o d e rn beach collections, one
p aleo n to lo g ical co llection a n d 51 arch aeo lo g ical collections. H e calcu lated the
ra tio as th e n u m b e r o f w ing elem ents (hum erus, u ln a, ca rp o m etac arp u s)
divided by th e to ta l o f w ing a n d h in d lim b elem ents (fem ur, tib io tarsu s,
ta rso m e ta ta rsu s), tim es 100 to derive a ra tio called the p ercentage o f forelim b
elem ents statistic. C alcu latin g the ra tio for each o f several tax a o f birds,
E ricso n (1987) fo u n d th a t in n a tu ra lly d ep o sited assem blages forelim b ele
m ents te n d to be m o re a b u n d a n t th a n h in d lim b elem ents b u t in archaeological
assem blages hin d lim b s tend to be m o re a b u n d a n t th a n forelim b elem ents. H e
in terp re ts this to in d icate h u m an s w ere focusing th eir ex p lo itatio n o f avian
carcasses o n th e m o re m eaty hindlim bs. T his is different fro m a suggestion
m ade by a n o th e r zo o a rch a eo lo g ist (cited in B ram w ell et al. 1987), w ho suggests
th a t ra p to rs w ould d ep o sit m an y ca rp o m e ta c a rp a ls, ta rso m e ta ta rsa ls, an d
co raco id s (distal lim b elem ents) relative to p ro x im al lim b elem ents, w hereas
h o m in id s w ould d ep o sit m an y hum eri, fem ora, a n d co raco id s (p ro x im al lim b
elem ents) relative to d istal lim b elem ents o f th eir av ian prey. B ram w ell et al.
(1987) p resen t d a ta th a t suggest the la tte r p a tte rn o f b o n e frequencies m ay
result fro m b o n e-d ep o sitin g activities o f the golden eagle (A quila chrysaetos).
T hey also re p o rt this ra p to r d am ages the p o ste rio r p o rtio n o f the stern u m an d
break s th e keel o f th e stern u m o f avian prey.
L iv ing sto n (1988,1989) uses several lines o f evidence a n d analytic techniques
to assess w h e th e r large assem blages o f avian rem ains recovered fro m a rc h a e o
logical sites in N e v a d a rep resen ted c u ltu ra lly o r n atu ra lly d ep o sited m aterials.
O ne o f th o se sites is a cave, in w hich the h o riz o n ta l d istrib u tio n a n d density o f
av ian rem ains suggest n o n -h u m a n ac cu m u lato rs o f av ian rem ains m ay d ep o sit
m o re o f tho se rem ain s n e a r th e cave e n tra n ce a n d ju s t o u tsid e the en tra n ce an d
depo sit few er rem ain s inside the cave (L ivingston 1988). She also no tes th a t
m an y a v ian rem ains, in cluding fe ath er bundles, skins, a n d bones, have been
m ad e in to artifacts, suggesting people co u ld h av e ac cu m u lated a n d d ep o sited
som e o f the unm o dified avian rem ains. T h ird , L ivingston (1988) co m p ares the
relative ab u n d a n ce o f w a te rb ird s a n d m am m als from h er cave site w ith the
relative a b u n d a n ces o f th o se ta x a in o th e r cave a n d o p en sites in the region,
n o tin g th a t h er cave co n tain s m o re rem ains o f large w a te rb ird s th a n o th e r caves
b u t a b o u t th e sam e relative frequencies as observed a t several op en sites in
m arsh -sid e locatio n s. T he presence o f feath ers a n d av ian bones a n d skin in

h u m a n co p ro lites recovered fro m th e cave, a n d the fact th a t local coyotes
(Canis latrans, one p o te n tia l n o n -h u m a n a c c u m u la to r o f the avian rem ains in
the area) ten d to focus th eir subsistence p u rsu its on lag o m o rp h s, lead her to
co nclu d e a t least som e o f the b ird rem ains in the cave w ere ac cu m u lated an d
d ep o sited by h u m an s.
T h e av ian rem ains fro m th e cave site a n d a n ea rb y open, m arsh-side site
ex hibit m in im al w eath erin g a n d ro o t etching (L ivingston 1988). L ivingston
(1988, 1989) follow ed E ricso n ’s (1987) lead a n d exam ined the relative freq u en
cies o f sets o f skeletal elem ents. She fo u n d th a t fo r b o th h er cave site a n d open
m arsh -sid e site av ian fo relim b a n d hindlim b elem ents occurred, on average
across all tax a, in virtually identical frequencies. O ne individual bird tax o n th a t
displayed significantly low ab u n d a n ces o f forelim b elem ents ap p e are d , on the
basis o f o th e r evidence, to h av e been cu ltu ra lly dep o sited , as E ricso n ’s (1987)
m odel pred icts. B ut a n o th e r b ird ta x o n h a d relatively hig h b u t statistically
insignificant a b u n d a n ces o f forelim b elem ents; over one th ird o f th e rem ains o f
th a t ta x o n seem to h av e been d ep o sited by n a tu ra l m echanism s (ra p to rs) on the
basis o f o th er evidence. T h a t plus the fact th a t R ich ’s (1980) paleo n to lo g ical
assem blage h ad a low a b u n d a n c e o f forelim b elem ents lead L ivingston
(1989:542) to co n clu d e th a t E ricso n ’s (1987) percent o f forelim b elem ents ratio
“ m ay n o t be a reliable in d ic a to r o f h u m an activity in the d ep o sitio n al p ro cess.”
T ak in g a different an aly tic a p p ro a c h , L ivingston (1988, 1989) co m p ares the
av ian assem blages fro m h er cave site a n d open m arsh-side site a n d finds th a t
relative tax o n o m ic ab u n d a n ces are sim ilar, suggesting to h er th a t sim ilar
a c cu m u latio n a l a n d d ep o sitio n al processes m ay be a t w o rk a t b o th sites. She
argues th a t the percen t o f forelim b elem ents statistic ten d s to be high fo r avian
tax a in h er sam ples th a t are stro n g fliers w ith m o re ro b u st w ing elem ents th a n
av ian tax a th a t w ade, have tro u b le tak in g off fro m w ater, a n d have lightly built
w ing elem ents. T ax a w hich do n o t m eet these ex p ectatio n s have eith er sm all
an d th u s u n reliab le sam ples, o r h ad w ings th a t w ere used to m a n u fa c tu re
decoys by the p re h isto ric o cc u p an ts o f the sites. She concludes th a t “ if. in fact,
there is an u n d erly in g p ro p e rty co n tro llin g elem ent su rv iv o rsh ip in avian
rem ains, derived m easu res o f su rv iv o rsh ip based o n elem ent frequencies will
also reflect th a t p ro p e rty ” (L ivingston 1989:545), a n d suggests stru c tu ra l
density m ay be th a t u n d erlying p ro p e rty a n d th u s m ay v ary w ith the fun ctio n al
a n a to m y o f th e tax o n .
U n p u b lish ed research cited by P a rk e r (1988:201) indicates th a t “ m icro
scopic ex a m in a tio n ” o f b ird skeletal p a rts ca n reveal w h eth er o r n o t the
rem ain s “ w ere in fact eaten by h u m a n s,” b u t he does n o t in d icate the n a tu re o f
th e m icro sco p ic a ttrib u te s. M a n y specim ens in a large sam ple o f b ird bones
from a site on the coast o f O regon th a t I exam ined display cut o r b u tch erin g
m ark s, in d icatin g th a t birds in som e p re h isto ric co n tex ts w ere subjected to the
sam e intensive level o f b u tch ery as the m am m al rem ains w ith w hich they were
asso ciated (L y m an 1991a, 1992c, 1993a). S tallibrass (1990:158) re p o rts u n p u b

lished ex p erim ents in d icatin g d om estic cats (Felis cattus) d am ag e b ird bones;
tw o -m o n th -o ld k itten s can “ consum e w hole long bones o f juvenile chickens
(G allusgallus) a n d com p lete chicken carcasses ca n be co n su m ed by a d u lt c a ts.”
Reptilian and amphibian taphonomy

If little is k n o w n a b o u t piscean a n d avian tap h o n o m y , still less is kn o w n a b o u t
rep tilian (w ith the possible exception o f d in o sau rs; e.g., D o d so n 1971; S ander
1992) a n d a m p h ib ia n ta p h o n o m y . T h a t m ay be because rem ains o f these tax a
are seldom fo u n d in a b u n d a n c e in arch aeo lo g ical sites, a n d /o r because so few
o f these tax a have been studied actualistically. F o r exam ple, W eigelt’s (1927/
1989) volum e is still the classic one fo r learn in g a b o u t the m ore im m ediately
p o stm o rtem effects o f cro co d ilian ta p h o n o m ic histories.
M eyer (1991) bu ried tw o m arin e (haw ksbill) tu rtle (E retm ochelys im bricata)
carap aces in in tertid al sands o f a lagoon. D e co m p o sitio n o f ligam ents holding
in d iv idu al elem ents o f th e c arap aces to g eth er allow ed d isarticu latio n a n d wave
actio n dispersed in d iv id u al elem ents w ithin th ree w eeks. M eyer (1991:92)
fo u n d th a t o rie n ta tio n s o f the long axis o f ind iv id u al skeletal p a rts fro m o th er
tu rtles p rim arily align p e rp en d icu lar to the co a st an d secondarily p arallel to it
due to sh o re-p arallel cu rren ts. M eyer (1991:92) re p o rts th a t “ direct p re d a tio n
by large v erteb rates can be deduced from scratch m ark s o n the shell.” M eyer
con cludes th a t ra p id b u rial is req u ired if tu rtle carcasses are to be fo u n d in
a rtic u la te d co n d itio n .
S chneider an d E v erson (1989) sum m arize the biology of, a n d archaeological
an d e th n o g ra p h ic in fo rm a tio n con cern in g the desert to rto ise (X erobates
agassizii) o f so u th e rn C alifo rn ia a n d ad jac en t N ev ad a a n d A rizo n a. T hey list
b eh a v io ral tra its a n d ecological pred ilectio n s o f this tax o n , as well as the
m an n ers in w hich people utilized it, including fo r subsistence, cerem ony, ritual,
m edicine, a n d tech n olo g ical an d sym bolic p u rp o ses. S chneider a n d E verson
(1989:189) p o in t o u t th a t d istinguishing cu ltu ra lly fro m n atu ra lly deposited
desert to rto ise rem ain s is difficult; they co n clu d e b u rn in g is n o t a reliable
in d ic a to r b u t “ there is no q u estio n th a t g ro u n d , drilled, d ec o rated , o r otherw ise
m odified specim ens are an in d icatio n o f cu ltu ra l use.” T h eir study is valuable
n o t so m uch fo r the ta p h o n o m ic in fo rm a tio n they provide, w hich is actually
m inim al (e.g., they d o n o t describe an y o f the m an y w ays people m odify tu rtle
shells), b u t ra th e r fo r the fact th a t it illu strates th a t tu rtles at least, if n o t all
reptiles, are subject to the sam e kinds o f h u m a n ta p h o n o m ic facto rs as
m am m als, birds, a n d fish.
Summary
P erh a p s the m o st strik ing th in g in n o n -m am m alian v e rte b ra te ta p h o n o m y is
the fact th a t we k n o w so little a b o u t it, alth o u g h serious efforts are being m ade
to learn a b o u t th e ta p h o n o m y o f fishes. A n o th e r very strik in g th in g is th a t

v irtu ally all o f th e research being u n d e rta k e n o n the ta p h o n o m y o f fish a n d
birds by zo o arch aeo lo g ists is follow ing in the fo o tstep s o f tap h o n o m ists
w o rk in g w ith m am m als. T he a ttrib u te s a n d variables the la tte r have studied,
such as skeletal p a rt frequencies a n d extent o f d isarticu latio n , are precisely the
variab les chosen fo r stud y by tap h o n o m ists w o rk in g w ith n o n -m am m alian
verteb rates. W hile th a t is p e rh a p s to be expected, an d certain ly is w a rra n te d , I
w o n d e r if such m im ick ing is lim iting w h a t we can learn a b o u t the ta p h o n o m y o f
these o th e r kinds o f organism s. T he logical startin g p o in t for all tap h o n o m ic
analysis (regardless o f tax o n ) is the m odel o f a com plete, living organism . W h a t
I am th u s ask in g is, sh o uld the different life h ab its o f fish, w hich are aqu atic,
a n d b ird s, w hich are aerial a n d o ften a rb o re a l, a n d m am m als, w hich are
v ario usly terrestrial, arb o re a l, a n d aq u a tic , be ta k e n in to a c c o u n t in o u r
ta p h o n o m ic analyses? C ertain ly since the sem inal w o rk o f S chafer, W eigelt,
a n d o th ers, we sh ould be at least th in k in g in these term s. F ish are m ore likely to
be subjected to fluvial ta p h o n o m ic processes th a n te rre stria l anim als. Sim ilarly,
alpine anim als are likely to be subjected to fluvial tra n s p o rt a n d ab ra sio n from
sed im ent d u rin g fluvial tra n s p o rt w hereas desert-dw elling anim als are less
likely to be subjected to such actio n . K n o w in g w h e th e r the fa u n al rem ains one
is stu d y in g derive from arb o rea l, terrestrial, o r a q u a tic tax a can th u s be o f
m a jo r im p o rta n c e d u rin g ta p h o n o m ic analysis. A n d , o ften these distinctions
parallel tax o n o m ic lines.
13
D IS C U S S I O N AND C O NC L US I O N S
Bones are docum ents as are p o tsh erd s an d dem an d the sam e scrupulous a tten tio n
b o th on the site a n d in th e lab o rato ry .
(M . W heeler 1954:192)
Introduction
M o d e rn ta p h o n o m ic analysis is (som etim es excrutiatingly) detailed, it is

extensive, a n d it is intensive. T h e n u m b e r o f variables th e an a ly st sh o u ld an d
p erh ap s m u st co n sid er is large, a n d ten d s to increase as th e com plexity o f an
assem b lag e’s ta p h o n o m ic h isto ry increases. I n o ted earlier, fo r exam ple, th a t
assem blages rep resen tin g one o r a few in d iv id u al o rganism s signifying one
ac cu m u latio n a l a n d d ep o sitio n al event often te n d to be easier to in te rp re t th a n
lo n g -term accu m u latio n s co n sistin g o f m u ltip le ta x a a n d m u ltip le individuals.
It seem s th a t th e la tte r k in d o f assem blage is m o re co m m o n th a n the form er,
th erefo re the tap h o n o m ist m ay typically be faced w ith a collection o f v erte b rate
rem ain s th a t h a d a com plex ta p h o n o m ic h isto ry th a t m ay n o t be analytically
discernable.
“ T a p h o n o m ic ch ang e is seq u e n tia l” (A ndres 1992:39). B ecause ta p h o n o m ic
processes are h isto rical, they are cu m u lativ e (effects o f som e processes obscure
o r d estro y effects o f earlier processes) a n d in som e cases, one process is
d ep e n d en t o n a p recedin g one. F o r exam ple, c a rn iv o re gnaw ing m ay o b literate
b u tch erin g m ark s o n bones, o r a highly w eath ered bo n e is unlikely to be
tra n s p o rte d to the d en o f a scavenging carn iv o re. W h a t a ta p h o n o m is t studies
is how a collection o f fossils differs from the living skeletons o f anim als
rep resen ted by th o se fossils, a n d if a n d how the p o p u la tio n o f skeletons differs
from th e n a tu ra l b io tic p o p u la tio n o f skeletons. F o r exam ple, A ndrew s
(1992:34) suggests “ all ta p h o n o m ic p ro b lem s can be subsum ed u n d e r the
general q uestion: w h a t is the origin o f the fossil assem blage, w h a t are the
[accu m u lation al, d ep o sitio n al, preserv atio n al] processes involved, a n d how
m ay these processes h ave ch an g ed the original co m p o sitio n o f the [fossil]
assem blage?”
T o an sw er the q u estio n ju s t p osed, A ndrew s (1992:33) echoes earlier a u th o rs
(e.g., G iffo rd 1981: S h ip m an 1981b) a n d suggests fo u r criteria are necessary.
F irst, ta p h o n o m ic analysis sh o u ld be p ro b lem orien ted in o rd e r th a t, fo r
exam ple, we k now w hich ta p h o n o m ic ch aracteristics o f a fossil assem blage
signify b ias a n d w hich ones signify in fo rm a tio n relev an t to th e inferences we
452
Discussion and conclusions 453
w ish to m ake. Second, relev an t n e o ta p h o n o m ic o r actu alistic d a ta (d eter

m ined, in p a rt, by the p ro b lem ) m u st be av ailab le to help in te rp re t ta p h o n o m ic
a ttrib u te s o f the fossils. T h ird , ta p h o n o m ic a ttrib u te s m u st be well defined so
th a t they m ay be reco rd ed a n d tallied. A n d fo u rth , ta p h o n o m ic histories o f
fossil assem blages, w ith ta p h o n o m ic events in p ro p e r sequence, m u st be
w ritten . T he last - w riting ta p h o n o m ic histories - is, o f course, w h at ta p h o
nom y is all a b o u t. W h a t one does w ith th a t history, once it is w ritten, depends
on th e k in ds o f inferences one w ishes to m ake a b o u t the bone accu m u lato rs, the
b o n e m odifiers, o r th e p aleoecological c o n d itio n s in d icated by the fossils.
A re th e b ones d isarticu lated ? W h a t is the d em o g ra p h y o f the represented
organism s? A re co m p lete skeletons presen t, o r only certain p a rts o f the
skeleton? D o som e b ones have ca rn iv o re g naw ing m ark s o r b u tch erin g m arks?
H ave som e b ones been w eathered, b u rn ed , b ro k e n , o r ab rad ed ? W h a t are the
o rie n ta tio n s o f the lo n g bones? H av e som e bones been crushed by sedim ent
o v erb u rd en ? T hese qu estio n s un d ersco re A n d re w s' (1992) th ird crite rio n and
su m m arize th e gist o f m uch o f w h at has been said in preceding ch a p te rs. T he
q u estio n th a t rem ain s concerns h ow an extensive a n d intensive tap h o n o m ic
analysis m ig ht be p erfo rm ed . In p revious ch a p te rs I p resen t exam ples o f
analysis th a t h av e as th eir focus one, tw o, o r (less often) three variables or
m o d ificatio n attrib u te s. H o w d o we in teg ra te these single variables in to a
m u lti-v ariate analysis?
Multi-variate taphonomic analysis

Relatively u n am bitio u s tap h o n o m ic studies m ust assem ble extensive m ultivariable
descriptions, often in the form o f dishearteningly large tables . . . If archaeologists
are to learn from tap h o n o m y it is clear th a t they are going to have to handle
descriptions in the hyperspace o f m any variables. [We need to determ ine] w hether
som e observed variables consistently co-associate an d o thers consistently avoid
each other.
(R. W right 1990:260, 262)
T ap h o n o m y as a w hole becom es m ore useful the m o re we th in k a b o u t processes

beyond descriptive details. N a rrativ e anecdotes an d num erical o r m em orial d a ta
sheets are the necessary ingredients, b u t they m ake a m eal only w ith conceptual
recipes.
(A. Seilacher 1992:123)
G iffo rd -G o n zalez (1991:217) un d ersco res the significance o f m u lti-v ariate

ta p h o n o m ic analysis w ith her d istin ctio n betw een the 1970s an d 1980s research
focus on identifying ta p h o n o m ic agents a n d the recent em erging focus on
“ m o re com plex analyses a n d m o re am p liativ e inferences a b o u t th e life relatio n s
o f p re h isto ric h o m in id s.” T he fo rm er ph ase o f ta p h o n o m ic research has been
q uite successful in several respects, as evidenced by earlier c h a p te rs o f this
volum e. T h e second a n d new er p h ase will be m o re difficult, G iffo rd -G o n zalez
(1991) co n ten d s, fo r tw o reasons. O ne involves the search fo r a p p ro p ria te levels
an d scales o f analysis, a n d clarification o f d a ta categories. I suggest, as G ifford-

G o n zalez (1991) does, th a t this p ro b lem is easily rem edied by m ore in terp lay
betw een eth n o arch ae o lo g ic al o r n e o ta p h o n o m ic research a n d the analysis o f
p re h isto ric v erte b rate fau n as. T o o o ften I read eth n o arch ae o lo g ic al research
results th a t seem to have been w ritten by som eone w ho h as never d ea lt w ith a
co llection o f p re h isto ric fa u n al rem ains. M a n y o f the p u b lish ed e th n o a rc h a e o
logical d a ta c a n n o t be used as fram es o f reference fo r co m p ariso n w ith
pre h isto ric m aterials because they are p resen ted in a w ay th a t c a n n o t be
rep licated in an arch aeo lo g ical co n tex t. F o r instance, Y ellen (1991b: 172)
re p o rts th a t “ even a few days o f c o n tin u ed site o cc u p atio n afte r d iscard [of
b ones a n d bon e fragm ents] will serve to p ro te c t the frag m en ts fro m p re d a to rs
a n d increase the lik elih o od o f lo n g -term su rv iv al.” D oes this m ean th a t every
tim e g o o d bone p re serv atio n is fo u n d bo n e d iscard to o k place a w eek o r so
p rio r to site ab a n d o n m e n t? C ertain ly n o t, b u t to use th is eth n o arch ae o lo g ic al
o b serv atio n we w ou ld need to be able to d eterm in e the re la tio n betw een the
tim e o f b o n e d ep o sitio n a n d site a b a n d o n m e n t, a re la tio n th a t surely is beyond
the reach o f m o d e rn arch aeo lo g ical analysis.
A c co rd in g to G iffo rd -G o n zalez (1991:218), the second reaso n the new ly
em ergent p h ase o f m o re com plex analyses a n d a ttem p ts to derive inferences o f
h o m in id b eh av io rs will be m o re difficult th a n its a n te c e d a n t agent-identifica-
tio n p h ase is th a t the desired inferences involve “ cau sality in b iological system s
. . . because h om inids a n d the co n tex ts in w hich they exist are biological
e n tities.” T herefo re, she tu rn s to how p aleobiologists a n d ev o lu tio n ists deal
w ith cau sality , a n d co n cludes th a t w hile th ere are n o im m u tab le laws o f
ta p h o n o m y th a t are n o t o b vious o r trivial (b u t see G o u ld 1986), m o st
inferences these researchers derive are p h ra se d in p ro b a b ilistic term s; “ ex p la
n a tio n becom es a p ro b a b ilistic ac co u n t in w hich ce rta in tactics m ay be
em ployed to reduce u n c e rta in ty ” (G iffo rd -G o n zalez 1991:241). T he " c e rta in
tactics” involve “ in d ep en d e n t lines o f evidence, derived fro m d istin ct system s
o f ca u satio n , m obilized to challenge a n d /o r s u p p o rt one a n o th e r, [and lead] to
m o re stron gly w a rra n te d inferences re g ard in g the p a st life re la tio n s th a t
p ro d u c ed certain co n fig u ratio n s o f [zooarchaeological a n d tap h o n o m ic trace]
m aterial in o u r sites” (G iffo rd -G o n zalez 1991:243). But, G iffo rd -G o n zalez
(1991:243-245) c a u tio n s th a t we sh o u ld n o t be m isled in to th in k in g th a t these
tactics sim ply involve m o re ag en t-id en tificatio n research. R a th e r, th e “ fu n c
tio n al lin k ag es” in th e observed p a tte rn s o f bo n e m o d ificatio n a ttrib u te s need
to be assessed eth n o arch ae o lo g ic ally a n d experim entally, a n d then “ relatio n al
an alo g ies” (see C h a p te r 3) betw een th em a n d p re h isto ric cases are co n stru c te d
(G iffo rd -G o n zalez 1991:245). As n o ted in earlier ch a p te rs (see especially the
first tw o exam ples o f ta p h o n o m ic analysis in C h a p te r 3 a n d the “ B utchering,
b reak age, a n d b o n e to o ls” section o f C h a p te r 8), som e o f w h a t I th in k is the
best ta p h o n o m ic research p erfo rm ed to d a te precisely follow s G iffo rd -G o n z a-
lez's (1991) recom m ended research p ro g ram .
I liken m o d ern ta p h o n o m ic analysis to a huge analysis o f variance, o r

A N O V A , in statistics. D o no t be terrified by the m e ta p h o r, n o r be m isled into
th in k in g ta p h o n o m ic analysis necessarily involves statistics o r even interval-
scale m easu rem en ts. T he sim ilarity o f ta p h o n o m ic analysis w ith A N O V A I
n o te here is m ean t to u n d ersco re the fact th a t, like A N O V A , ta p h o n o m ic
analysis to d a y involves th e study o f m an y sam ples o f m an y variables, and
c o m p ariso n s betw een the m eans a n d v ariances o f (tap h o n o m ic traces or)
a ttrib u te s d isplayed by fa u n al rem ains. A N O V A is m e a n t to d eterm in e the
statistical significance o f v a ria tio n displayed w ithin a n d betw een gro u p s
(assem blages) o f one o r m o re v ariables m easu red in a set o f m u ltip le sam ples.
A n d , sim ply p u t, th a t is w h a t ta p h o n o m ic analysis involves to o , except the
tap h o n o m ist is co n cern ed w ith d eterm in in g the ta p h o n o m ic (ra th e r th an
statistical) significance o f the observed v a ria tio n fo r the p u rp o se o f w riting
ta p h o n o m ic h istories, strip p in g aw ay the ta p h o n o m ic o v erp rin t, o r a sc ertain
ing so m eth in g a b o u t paleoecology from the ta p h o n o m ic a ttrib u te s (e.g.,
W ilson 1988).
In o rd e r to m ak e th e p receding arg u m e n t clear, co n sid er the basic analytic
p ro c ed u re follow ed d u rin g ta p h o n o m ic analysis o f a collection o f v erte b rate
rem ains. F irst, we reco rd v ario u s a ttrib u te s displayed by the individual
specim ens. Is a b o n e gnaw ed, h ow w eath ered is it, is it b u rn ed , w h a t w as its
o rie n ta tio n in th e g ro u n d , is it an incom plete skeletal elem ent (is it b ro k e n , an d
if so, how )? G en erally , each v ariab le th a t m ig h t be reco rd ed has m ore th a n one
possible v alue th a t it can display on a specim en: g n aw ed /n o t gnaw ed, w e ath er
ing stage 0, 1,2, 3 ,4 , o r 5; etc. (F o r discussion, I ignore the fact th a t a specim en
m ay disp lay m o re th a n one possible value fo r a given variable, such as the
sk y w ard surface o f a b o n e displaying w eath erin g stage 3 w hereas the g ro u n d
w ard surface o f th a t b o n e displays w eath erin g stage 1.) A n d , each specim en
m ay disp lay one o r m o re a ttrib u te s o f m odification. A single specim en m ay be
b ro k en , b u tch ery m ark e d , an d w eathered, fo r exam ple. T he p o in t here is, to
p a ra p h ra se G ifford (1981:385-386), the co n d itio n o f an y individual fossil
specim en is the end p ro d u c t o f a series o f ta p h o n o m ic events w hereas a fossil
assem blage is an aggregate o f individual end p ro d u cts. R egularities in ta p h o
nom ic m odification s displayed in specim en afte r specim en m ay be reflected in
the assem blage, o r reg u larities in p ro p o rtio n s o f specim ens displaying a
p a rtic u la r kind o f m od ification m ay be fo u n d in assem blage afte r assem blage.
T hese reg u larities are o ften fo u n d by inductive p a tte rn reco g n itio n (B inford
1984b:9—15). A scrib ing ta p h o n o m ic m ean in g to the recognized p a tte rn s ty p i
cally involves m atch in g them w ith actualistically observed p attern s.
T h e preced ing p a ra g ra p h is ac cu ra te b u t quite general a n d sim plistic. If we
ad d d etails, th e m ag n itu d e o f the com plexity o f the issue - the A N O V A -like
c h a ra c te r o f m u lti-v a riate ta p h o n o m ic analysis - becom es clear. G iven th a t we
reco rd th e ta p h o n o m ic m o d ificatio n a ttrib u te s displayed by each individual
specim en, it shou ld be obvious th a t any set o f ta p h o n o m ic d a ta represents a
palim p sest, even if th e assem blage is one specim en displaying tw o o r m ore

a ttrib u te s o f m o d ificatio n (e.g., a b u tch erin g m ark an d an ad v an ced w eathering
stage). F o r exam ple, m y typing the letter “ F ” in th e first w o rd o f this sentence
m ay be co nsid ered a ra th e r fine-grained event a t one scale, b u t a t a finer scale I
h ad to m ove one finger to the “ sh ift” key, p u sh th a t key d ow n, ho ld it dow n,
m ove a second finger to the “ f ’ key on the k ey b o a rd , strike the “ f” key, then
release the “ sh ift” key before typ in g the second letter o f the sentence. T h e p oint
here is th a t ta p h o n o m ists m u st alw ays be co g n izan t o f the scale a t w hich they
conceive o f fine-grained a n d co arse-g rain ed o r p alim psest assem blages o f
fossils a n d a ttrib u te s o f ta p h o n o m ic m odification.
G en erally, arch aeo lo g ical ta p h o n o m ists w a n t (a) to identify the bone-
ac cu m u latin g agent, in o rd e r to assess differences betw een w h a t is p resen t in the
assem blage a n d w h at m ight have been p resen t (given kn o w n o r suspected
b eh av io rs o f the a c cu m u lato r), a n d /o r (b) to identify the ta p h o n o m ic agents
th a t m odified th e b o n e assem blage in o rd e r to assess differences betw een w h at
w as av ailab le fo r ac cu m u latio n , w h a t w as deposited, w h a t w as bu ried , and
w h a t w as collected, a n d /o r (c) to tease o u t the h u m a n b eh a v io ral o r paleoecolo-
gical significance o f m odified fa u n al rem ains. T he first tw o desires are
in terre lated because identified agents o f b o n e m o d ificatio n are o ften ta k e n also
to be ag ents o f b o n e ac cu m u latio n . In fulfilling an y o r all three o f these
an aly tical desires, ta p h o n o m ists deal w ith m u ltip le a ttrib u te s (e.g., T ab le 2.1).
B ecause in d iv idu al b o n e specim ens m ay be m odified by m o re th a n one
ta p h o n o m ic agen t a n d process a n d th u s m ay display m ore th a n one a ttrib u te o f
m od ificatio n, it is im p o rta n t to n o te w hich specim ens are, for exam ple, b o th
b u rn e d a n d b u tch ery m ark e d , o r b o th w eath ered an d b ro k en . A t its best, then,
the p o te n tia l m o d ificatio n a ttrib u te s sh o u ld be arran g e d p arad ig m atic ally in
o rd e r th a t th e an aly st can tally how m an y h u m eri o f a p a rtic u la r tax o n are, for
in stance, b u tch ery m ark e d , b u rn e d , a n d b ro k e n , a n d how m an y fem o ra o f th a t
tax o n are n o t b u tch ery m ark e d , n o t b u rn e d , a n d com plete; th a t is, to m o n ito r
c o v a ria tio n o f th e different ta p h o n o m ic a ttrib u te s.
A p a ra d ig m a tic classification consists o f dim ensions w hich are sets o f
m u tu ally exclusive v ariables. E ach dim ension consists o f several a ttrib u te
states o r th e values th a t a p a rtic u la r dim ension o r v ariab le m ay display. F o r
exam ple, the dim ension “ b u rn in g ” m ay h av e the a ttrib u te states “ u n b u rn e d ,
ch a rre d , c alcin ed ” a n d th e dim ension “ w e ath erin g ” co u ld h ave the six w e ath er
ing stages as its a ttrib u te states. M an y o f the dim ensions o f the p arad ig m atic
classification alo n g w ith som e o f the possible a ttrib u te states for each d im en
sion are given in T ab le 13.1, a n d are variously su p p lem en ted by figures an d
tables cited therein. T his classification can a n d sh o u ld be m odified for
p a rtic u la r assem blages (e.g., if n o b u rn e d specim ens are fo u n d , th e n th a t
dim en sion can be o m itted ). T his classification allow s the an aly st to identify
each specim en as belong ing to a p a rtic u la r class, w here each class is defined by
in tersectio n o f the d im ensions. T h e classification is in th a t reg ard , truly
T able 13.1 D im ensions and a ttribute states f o r taphonom ic analysis
D im ension
A ttrib u te states per dim ension (partial)
Specim en attrib u tes

a. taxon
1. fam ily, genus, species
2. anim al live w eight size
3. ethology and ecology o f taxon
b. skeletal p a rt (e.g., T able 4.1)
c. context/asso ciatio n
1) horizon tal an d vertical position (e.g., d ep th from surface)
2) depositional unit attrib u tes (e.g., sedim ent texture, chem istry)
d. an atom ical d istrib u tio n o f dam age o r a ttrib u te
1) u pw ard (skyw ard) o r dow nw ard (gro u n d w ard ) surface
2) anterior, po sterior
3) lateral, m edial
4) proxim al, distal
e. w eathering stage (T able 9.1)
f. gnaw ing dam age: presence/absence
1) carnivore
2) rodent
3) insect
4) som e co m binatio n o f 1, 2, a n d /o r 3
g. p o lish /ab rasio n /co rro sio n
h. fracture type (e.g.. Figure 8.4)
i. com pleteness
1) fractu red o r com plete
2) articulated w ith o th er bones in situ
3) refits an d conjoins
j. butchery m arks (type, o rien tatio n , location, frequency o r density p er area)
k. azim u th /o rie n ta tio n (e.g., Figure 6.7)
1. plu nge/dip (e.g., Figure 6.8)
m. b u rn e d /n o t b urned
A ssem blage attrib u tes
n. frequencies (absolute an d p ro p o rtio n al) o f b, d, e, f, etc., above
o. m ortality p attern s (e.g., Figure 5.1)
p. skeletal p a rt frequencies
q. density p er un it space
unw ieldy. T here are, fo r in stan ce, 13 dim ensions listed u n d e r “ specim en
a ttrib u te s ” in T ab le 13.1. T here co u ld be m o re dim ensions, b u t let us assum e
fo r sake o f discussion th a t this is a com plete list. L et us fu rth e r assum e th a t each
dim ension has fo u r a ttrib u te states; there co u ld be few er o r there co u ld be m ore
fo r an y given dim en sion. T he 13 dim ensions w ith fo u r a ttrib u te states each
m eans th ere are 4 13, o r 67,108,864 possible classes o f bones displaying various
co m b in atio n s o f ta p h o n o m ic traces! I kn o w o f no p racticin g ta p h o n o m ist who
reco rd s a ttrib u te s o f b o n e m o d ificatio n in this fashion, alth o u g h a n increasing

n u m b e r are b eginning to a p p ro a c h this k in d o f system o f d a ta re co rd in g (i.e.,
m ore th a n tw o o r th ree a ttrib u te s p er specim en are recorded). W h at T ab le 13.1
does, how ever, is u n d ersco re the m u lti-v a riate n a tu re o f m o d ern ta p h o n o m ic
research, an d the p o ten tial com plexity o f th a t research (see S tiner 1992; W hite
1992 fo r recent exam ples w hich a p p ro x im a te this m u lti-v ariate com plexity).
Several q uestion s are begged by T ab le 13.1. W h ich dim ensions o r attrib u te s
o f b o n e m od ificatio n are relevant o r applicable to w hich questions? T h a t is,
how d o we en sure c o n c o rd an c e betw een th e d a ta we reco rd , the analyses we
p erfo rm , a n d th e q u estio n s to w hich we seek answ ers? T he answ ers reside in
actu alistic o r n eo ta p h o n o m ic research: “ [T aphonom ic] agencies have been
stu died in m o d ern co n tex ts fo r the diag n o stic p a tte rn in g th a t each p ro d u c es in
the co m p o sitio n a n d ch a ra c te r o f bone accu m u latio n s. T he d iagnostic p a tte rn
ing is in effect a sig n atu re o f g re at in terp re tiv e value, because it is possible to
identify th e sam e p a tte rn in g in an cien t b o n e assem blages a n d to infer eq u iv a
lent cau sativ e agencies” (B unn a n d K roll 1986:432). B ut as should be obvious
from preceding ch a p te rs, there is a p le th o ra o f p a tte rn s th a t m ight be identified
fo r any given variab le th a t m ig h t be m o n ito red , a n d th ere is also a p le th o ra o f
variables. G iven the facts th a t (a) a ta p h o n o m ic h isto ry is cu m ulative in the
sense th a t ag ents a n d processes th a t w o rk late in th e h isto ry m ay be m ed iated o r
influenced by agents a n d processes th a t w ork early in the history, a n d vice versa
(e.g., o b lite ra tio n o f traces form ed early by traces th a t form late), a n d (b) som e
o f th e v ariables are am en ab le to interval-scale m easu rem en t, o th ers are
am en ab le to only o rdin al-scale m easu rem en t, an d still o th ers ca n only be
m easu red in no m in al-scale term s, w h a t is th e m o d e rn ta p h o n o m ist to do?
B ehrensm eyer (1991:317, 321) p ro p o ses a grap h ic so lu tio n to the qu estio n
ju st posed (a sim ilar so lu tio n for in v erte b rate ta p h o n o m y is illu stra te d by
P arso n s a n d B rett 1990). A n exam ple o f h er grap h ic technique is given in
F ig u re 13.1, an d the v ariables included there are defined a n d th e p lo tte d values
are given in T able 13.2 for tw o fictional assem blages o f v erte b rate rem ains. This
grap h ic tech niq u e “ p ro vides a visual overview [of ta p h o n o m ic a ttrib u te s
displayed by an assem blage] th a t can be used as a basis for stan d ard iz ed
co m p ariso n s am o n g different a c c u m u la tio n s,” an d “ helps to sum m arize the
com plex a rra y o f ta p h o n o m ic in fo rm a tio n av ailab le in each b o n e ac cu m u
la tio n ” (B ehrensm eyer 1991:315). T he g ra p h s accom plish tw o goals: “ (1) they
show th e b eh a v io r o f different ta p h o n o m ic variables in re la tio n to one a n o th e r
w ithin each assem blage, a n d (2) they en co u rag e stan d ard iz ed re p resen ta tio n o f
d a ta , allow ing co m p arativ e analysis o f different assem blages” (B ehrensm eyer
1991:315). N o te, fo r exam ple, th a t it is possible to d eterm in e by b rie f study o f
F ig u re 13.1 th a t A ssem blage 1 is sm aller, less tax o n o m ically rich, m ore
a rtic u la te d , an d has few er skulls, few er w hole bones, few er w eathered bones,
few er b u rn e d b o nes, a n d m o re carn iv o re-g n aw ed bones th a n A ssem blage 2.
G ra p h s like th a t in F ig u re 13.1 m ig h t be gen erated fo r an y assem blage o f
M u lti var iat e Examples: A ss e m b la g e Assemblage 2
N IS P
CD
MNI
*->
<o
o Ri c h n es s
a>
cn Diversity
<o
E Body Size
a>
t/>
u> Age Spectrum
<
ISD
Cranial:Postcran.
Si z e of Accum.
<D
(D Density
Q
ZD _ Orientation
L_ 03 --------------
L_
CD
<D Dip
CL
CO
P at c hi n e ss 0.1
% Complete 100
CD
CD
Q % Weathered 100
C
o % Abraded 100
(D
O % Carn. Gnawed 100
•o % Burned 100
O
n
% Butchered 100
Figure 13.1. E xam ple o f g raphic technique for sum m arizing an d com paring
tap h o n o m ic d ata for m ultiple assem blages (after B ehrensm eyer 1991:317,
Figure 2).
T ab le 13.2 Definition o f variables and listing o f values per p lo tte d variable in

Figure 13.1
P lotted values
p er assem blage
V ariable D efinition 1 2
A ssem blage d ata

N IS P n u m b er o f identified specimens 1000 2000
MNI m inim um nu m b er o f individuals 30 70
R ichness nu m b er o f taxa identified 15 27
D iversity taxonom ic diversity 1.5 1.
Body size body size (kg) o f organism s w ith m odal M N I 100 100
Age spectrum % o f M N I th a t are adults (skeletally m ature) 85 100
ISD index o f skeletal disjunction (see C h ap ter 5) 50 75
C ranial:P ostcran. ratio o f M N E o f cran ia to M N E o f p ost-cranial bones 0.1 25
Q u arry d ata
Size o f accum . area (m 2) over w hich rem ains are recovered 100 150
D ensity average N IS P p er m 2 10 13.3
O rientation % o f long bones w ith sim ilar o rien tatio n 0 0
D ip % o f long bones deviating > 5 ° from h o rizontal 2 5
Patchiness sta n d a rd deviation o f density 0.1 5
M odification d a ta
% com plete % of specim ens th a t are w hole skeletal elem ents 5 50
% w eathered % of specim ens th a t display > stage 2 5 20
% ab rad ed % of specim ens th a t display ab rasio n 0 20
% earn, gnaw ed % of specim ens th a t display carnivore gnaw ing 25 10
% burned % of specim ens th a t have been burned 5 15
% butchered % of specim ens th a t have been butchered 25 20
v erte b rate fa u n al rem ains. T he key to co m p arativ e analysis o f m ultiple

assem blages is, o f course, th a t th e scales used for p lo ttin g the observed
variab les be id en tical fo r the co m p ared assem blages. O ne o f the ad v an tag es to
such a g ra p h ic tech n iq u e is th a t v irtu ally a n y suite o f variables m ig h t be
p lo tted . F o r exam ple, I have n o t discussed tax o n o m ic richness o r diversity in
this volum e, b u t these m ight be im p o rta n t variables fo r helping the analyst
u n d e rsta n d th e ta p h o n o m ic h isto ry o f an assem blage. N o r h ave I discussed the
“ p atch in e ss” o f b o n e d istrib u tio n s (e.g., ra n g e an d average o f frequency o f
fau n al rem ains p er u n it o f space) b u t this v ariab le also m ay be im p o rta n t. As
well, the an a ly st m ight ch oose to p lo t the ra tio o f axial N IS P to ap p e n d ic u lar
N IS P ra th e r th a n the ra tio o f cran ial N IS P to p o st-cran ial N IS P , a n d to om it
the p ro p o rtio n o f b u rn e d specim ens a n d a d d the p ro p o rtio n o f ro d en t-g n aw ed
specim ens. T h e set o f A frican bovid size classes (e.g., B unn an d K roll 1986;
K lein 1989) co u ld be used to develop a different scale fo r p lo ttin g the live w eight
size o f tax a in the assem blage. O ne could include plots o f the frequencies o f
specim ens th a t are b o th b u rn e d a n d display b u tch erin g m ark s. T he grap h ic
tech n iq u e is th u s q u ite flexible in term s o f th e in fo rm a tio n th a t it m ig h t include
Discussion an d conclusions 461
o r exclude. C o m p a riso n o f a g ra p h o f th e a ttrib u te s o f a n e o ta p h o n o m ic

assem blage w ith a k n o w n ta p h o n o m ic h isto ry w ith a g ra p h o f a p reh isto ric
assem blage could help determ ine th e ta p h o n o m ic h isto ry o f the latter, a n d m ay
help th e an a ly st asc ertain w hich ta p h o n o m ic traces are th e result o f diagenetic
processes a n d w hich are the resu lt o f b io stra tin o m ic processes.
G ra p h s like th a t in F igure 13.1 necessarily sim plify reality. T hey p lo t central
tendencies (e.g., m odes an d averages) o r p ro p o rtio n s, a n d om it v aria tio n (e.g.,
ranges a n d sta n d a rd d ev iations). M y im p ressio n fro m the lite ra tu re is th a t as
m ore actu alistic research is co m p leted a n d re p o rte d , m o re a n d m o re v aria tio n
is being d o cu m en te d (e.g., L am 1992). T h a t v aria tio n m ight eventually becom e
eq u al in im p o rta n c e to the m easures o f cen tral tendency p lo tte d in F ig u re 13.1
an d used by v irtu ally all tap h o n o m ists w ho p erfo rm co m p arativ e analyses
(e.g., Stiner 1992). A n o th e r bit o f in fo rm a tio n th a t is n o t readily included in
such g ra p h s is th e re la tio n o f som e o f the assem blage d a ta a n d som e o f the
m o d ificatio n d a ta . It has been d e m o n stra te d tim e a n d again, fo r instance, th a t
the tax o n o m ic richness a n d diversity o f an assem blage ten d to co rrelate w ith
the size o f th a t assem blage m easu red as N IS P (e.g., G ra y so n 1984) and
m easu red as the size o f th e ex cav atio n (e.g., L y m an 1991a; W o lff 1975).
A ttem p ts to circum vent this p ro b lem have n o t yet been successful (see the
discussion in C ru z-U rib e [1988] a n d M eltzer et al. [1992]). C are is called for
then, w hen co m p arin g assem blages o f ra th e r different size. T here are several
statistical tech n iques av ailable fo r d eterm in in g if the differences in p ro p o rtio n s
o f m odified specim ens is a fu n c tio n o f sam ple size, o r if they validly m easure
differences in ta p h o n o m ic histories (for exam ple, see th e discussion o f the
b u tch ery m a rk d a ta in T able 8.4).
B ehrensm eyer’s (1991; F igure 13.1) g rap h ic so lu tio n is an in trig u in g one th a t
can use n o m in al, o rd in al, a n d in terv al scale d a ta . S tiner (1992:436) uses a
sim ilar g rap h ic tech n iq u e fo r co m p arin g eleven a ttrib u te s displayed by 28
assem blages o f bones. H er tech n iq u e can em ploy only nom inal a n d o rd in al
scale d a ta , b u t accom plishes the sam e fu n c tio n as B ehrensm eyer’s. S tin er’s
tech n iq u e, how ever, is n o te w o rth y because she arran g e s the a ttrib u te s she
re co rd ed in to th ree categories: th o se only a ttrib u ta b le to ho m in id s, th o se only
a ttrib u ta b le to carn iv ores, a n d those p o ten tially a ttrib u ta b le to b o th hom inids
a n d carn ivo res. I h ave little d o u b t th a t this ch aracterisitic o f S tin er’s graphic
tech n iq u e can be in c o rp o ra te d in to B ehrensm eyer’s tech n iq u e to pro v id e
su m m ary illu stra tio n s o f the ta p h o n o m ic d a ta re co rd ed fo r archaeological
assem blages.
T h e list o f ta p h o n o m ic v ariables in T ab le 13.1 is extensive, b u t it is
incom plete. A ll o f th o se variables could be stru c tu re d so as to be included in
g ra p h s like th a t in F ig u re 13.1. O nce these g rap h s are gen erated , they can be
co m p ared . T o re tu rn to the an alogy betw een m u lti-v a riate tap h o n o m ic
analysis a n d A N O V A , each g ra p h m ight be co n sid ered a n A N O V A tab le (w ith
the n o ta b le exception m en tio n ed above th a t the g ra p h s p lo t m easures o f cen tral
ten d en cy ra th e r th a n v a ria tio n a n d it is the la tte r th a t is m easu red in A N O V A ).

T w o q u estio n s m u st be answ ered a t this p o in t in the discussion. F irst, how do
we solve the g ra p h as certain ly it is n o t an A N O V A tab le w hich can be solved
a n d a su m m ary statistic generated? A n d second, once th e g ra p h is solved,
p resu m in g it can be solved, w h a t does the so lu tio n m ean? B ehrensm eyer
(1991:318-319) arg u es th a t a so lu tio n is derived by c o m p a riso n o f g ra p h s for
p reh isto ric assem blages w ith g ra p h s for n eo ta p h o n o m ic assem blages. Sim ilar
g ra p h s suggest sim ilar ta p h o n o m ic histories. F o r exam ple, m o st assem blages
w ith b o n es d isplaying p referen tial o rie n ta tio n s, ab ra d e d surfaces a n d ro u n d ed
fractu re edges, a n d a p au c ity o f bones w ith low s tru c tu ra l densities, have been
fluvially tra n s p o rte d (B ehrensm eyer 1991:318). T hus, we can sim ultaneously
solve the g ra p h s a n d d eterm in e w h a t the so lu tio n m eans in term s o f ta p h o n o
m ic h istories by c o m p ariso n w ith actu alistic assem blages a n d arg u m e n t by
an alo g y th a t sim ilar com binations o f traces o r a ttrib u te s o f b o n e m odification
are indicative o f sim ilar ta p h o n o m ic histories. T his is, o f course, the so lu tio n
offered by G iffo rd -G o n zalez (1991) no ted above, except th a t h er so lu tio n
explicitly considers the c o v a ria tio n o f variables. T he im p o rtan ce o f this
co v a riatio n ca n n o t be overem phasized. F o r exam ple, Y ellen (1 9 9 lb : 186)
suggests th a t “ long b ones b ro k e n d u rin g b u tch erin g lose th eir a ttra c tio n [to
carnivores] m o re ra p id ly th a n th e ir com plete c o u n te rp a rts a n d u n d e r certain
circu m stan ces are m o re likely to be preserved fo r p o te n tia l archaeological
discovery. B roken b ones subjected to extended boiling are m ore com pletely
ren d ered o f m a rro w [and grease] th a n th o se w hich are n o t a n d co n seq u en tly are
o f less in tere st to c a rn iv o res.” T h ere m ay be a sim ple rule here th a t b ro k e n
b ones will survive c a rn iv o re a ttritio n b e tte r th a n u n b ro k e n b ones (see also
B lum enschine 1988; M arean et al. 1992), b u t w h at is im p o rta n t is the
covariation o f carn iv o re scavenging o f b ones a n d the presence o f exploitable
n u trien ts in bones: th a t c o v a ria tio n explains why the bones are o r are not
exploited by the scavenging ca rn iv o res (see below ). T o ap p ly the rule in
p reh isto ric settings th e ta p h o n o m is t m u st d eterm in e if the degreasing o f the
bones o ccu rred p rio r to d ep o sitio n (due, p erh ap s, to co o k in g an d boiling) o r
afte r d ep o sitio n (due to su b aerial w eathering o r diagenetic processes such as
leaching). T h e fact th a t the bones are b ro k e n (o r n o t) a n d ca rn iv o re g naw ed (or
n o t) are only two p a rts o f solving a m u lti-v a riate problem .
I suspect m u lti-v a riate analyses o f ta p h o n o m ic d a ta will becom e m ore
freq u en t a n d th e n u m b e r o f v ariables included in these analyses will also
increase as we learn m o re fro m actu alistic research a n d find th a t tw o o r th ree o r
fo u r v ariab les are sim ply n o t en o u g h to u nravel a n d w rite the com plex
ta p h o n o m ic histories a n d infer the h u m an b eh a v io ral o r paleoecological
significance o f th e co arse-g rain ed zo o a rch aeo lo g ical collections th a t m an y o f
us study. S olution s will be fo u n d using the co m p arativ e g ra p h ic tech n iq u e only
as the n u m b e r an d com plexity o f actu alistic cases th a t are re p o rted increase. If
th e m ajo rity o f e th n o -z o o arc h aeo lo g ica l research co n tin u es to focus o n only
Discussion an d conclusions 463
tw o o r th ree arch aeo log ically visible variables, an d to ig n o re the reality o f the
fossil reco rd , we m ay find th a t we are re co rd in g m uch ta p h o n o m ic d a ta for
p re h isto ric assem blages b u t th a t we do n o t kn o w w h at th o se d a ta signify in
term s o f th e ta p h o n o m ic h isto ry , h u m a n b eh a v io ral, a n d paleoecological
significance o f a n assem blage. T his ad m itte d ly p erso n al o b serv atio n p ro m p ts
m e to raise a final issue. C a n we p erh ap s co n stru c t a general th e o ry o f
ta p h o n o m y th a t m ig h t help us in o u r w o rk w ith n e o ta p h o n o m ic a n d p re h is
to ric assem blages alike? Such a th eo ry w ould p e rh a p s aid in the selection o f
variables fo r study, the variables th a t should covary in significant fashions, an d
the an aly tic techn iques fo r m an ip u la tin g th e o b serv atio n s we m ake.
A general theory of taphonom y?

T aph ono m y is a science a b o u t the unlikely: the survival o f organic m aterials and
form s in spite o f their general drive to w ard s recycling by biological, physical and
chem ical degradation .
(A. Seilacher 1992:109)
In a n earlier p a p e r I (L ym an 1987e) offered w h a t I th o u g h t at the tim e w as a

reaso n ab le o u tlin e o f w h at m ig h t eventually, w ith som e m odification, co n sti
tu te a g eneral th eo ry o f tap h o n o m y . M u ch o f th a t o u tlin e is p resen ted in the
discussion o f ta p h o n o m ic histories (“ O n th e stru c tu re o f ta p h o n o m y : a
p erso n al view ” ) in C h a p te r 2. W ith the benefit o f h in d sig h t I now perceive th a t
discussion as c o n stitu tin g n o t a sta te m e n t o f th eo ry , ra th e r it is a d escrip tio n , in
very general term s, o f w h a t ta p h o n o m ic histories lo o k like. T h u s it is
re aso n ab le to w o n d e r if a general th eo ry o f tap h o n o m y , sim ilar to th e general
th eo ry o f biological evolution, can be w ritten.
T a p h o n o m ic research , as we have seen th ro u g h o u t this volum e, focuses on
the physical, chem ical, an d m echanical differences betw een fossil rem ains o f
organ ism s a n d w h a t th o se o rganism s a n d th eir c o n stitu en t p a rts lo o k ed like
w hen th e o rg an ism w as alive. W e can w rite law s (if this is to o stro n g an
ep istem ological term , su b stitu te “ ru les” ) co n cern in g cause-effect re la tio n s o f
v ario u s p ro p e rtie s o f an o rg a n ism 's tissues a n d how th o se tissues will respond
to a p a rtic u la r ta p h o n o m ic process. F o r exam ple, I d o u b t th a t an y v erte b rate
ta p h o n o m is t w o u ld disagree th a t, given a set o f skeletal elem ents rep resen tin g a
com plete bo vid skeleton, if th o se skeletal elem ents are subjected to gnaw ing by
carn iv o res th en th e least stru c tu ra lly dense bones will be destro y ed (consum ed)
first a n d the b ones o f g re atest s tru c tu ra l density will h ave th e g reatest chance to
survive. T his sta te m e n t m ig h t be th o u g h t o f as a ta p h o n o m ic law o r rule. T here
are no d o u b t o th e r sim ilar law s o r rules con cern in g cause-effect relatio n s; if a
p a rtic u la r bo n e is subjected to a p a rtic u la r ta p h o n o m ic process th en a
p a rtic u la r e n d -p ro d u c t will result. T hese are all law s co n cerning m echanical,
chem ical, a n d p hysical properties o f v e rte b ra te h a rd p arts; they give m o d ern
ta p h o n o m y its decidedly “ m id d le -ra n g e” flavor (e.g., B inford 1981b; D u n n ell
1992; Sim m s 1992). T h e law s ta p h o n o m ists te n d to use (a) are ahistorical and
p ro d u c e little m ore th a n w h a t m ight be term ed a “ physics o f ta p h o n o m y ”
(after D u n n ell 1992:212) a n d (b) do little to explain the fossil reco rd (after
Sim m s 1992:190-191) in the sense o f helping us u n d e rsta n d why tap h o n o m ic
processes o ccur in the first place o r why ta p h o n o m ic processes o p erate the way
an d in th e tem p o ra l o rd e r th a t they do. T he fo rm er m ay be self-evident a n d the
la tte r will p ro b a b ly be perceived as heresy in som e q u a rte rs. L et m e elab o rate,
th en , o n w hy I ho ld th e view th a t tap h o n o m y , as yet, does n o t seem to be
e x p la n a to ry in som e u ltim ate sense.
W e h ave com e fa r in the p a st tw o decades; we can stu d y a set o f b ones an d
p osit a reaso n ab le set o f ta p h o n o m ic agents an d processes th a t acted on them
based on th e a ttrib u te s they display (G iffo rd -G o n zalez 1991). W e can say m uch
a b o u t w hat h ap p e n ed to an assem blage o f v erte b rate rem ains an d how it
h ap p en ed ; I view these as p ro xim a te causal explanations o f th e fossil reco rd . W e
can even, in som e cases, w rite fairly d etailed ta p h o n o m ic histories a lth o u g h we
m ay o ften be th w arte d in o u r efforts to p u t the precise ch ro n o lo g ical sequence
o f ta p h o n o m ic processes to g eth er, such as is fo u n d in the p resen t d eb a te over
th e su p erp o sitio n o f ca rn iv o re gnaw ing m ark s a n d h o m in id -g en erated b u t
chering m ark s on som e P lio-Pleistocene aged bovid bones from A frica (e.g..
Bunn a n d K ro ll 1986, 1988; P o tts 1988; S hip m an 1986a, 1986b). B uilding the
ch ro n o lo g ical sequence o f ta p h o n o m ic events a b o n e collection experienced is
ra th e r like the co n stru c tio n o f a h o m in id phylogenetic lineage fo r the last 3-5
m illion years. C o n stru c tio n o f the p hylogenetic lineage is fo u n d ed on principles
o f n e o -D a rw in ia n ev o lu tio n ary th eo ry , w hich unifies co ncepts such as n a tu ra l
selection, a d a p ta tio n , in h eritan ce, prim itive ch a rac te rs, derived characters,
an d th e like. A phylo g en etic lineage a rran g e s a series o f fossils in to a genetically
related , tim e-transgressiv e series on the basis o f ev o lu tio n ary th eo ry , the
ultim ate causal explanation fo r th a t arran g e m e n t. T he ta p h o n o m ist to d a y has
no such unifying th eo ry an d th u s to o often w rites their p artic u la ristic ta p h o n o
m ic h isto ry w ith, it seem s, little co n cern fo r the reaso n s w hy th a t history
occu rred in the first place a n d w hy th a t p a rtic u la r h isto ry has the p a rtic u la r
fo rm o r ch ro n o lo g y o f events it does. I suspect this is so because we lack a
c o h e ren t th eo ry o f ta p h o n o m y a n d th u s we ten d to focus o n the w h a t a n d how
ra th e r th a n the w hy o f ta p h o n o m y .
W hile this is n o t th e place to delve deeply in to this issue, the re ad er certainly
deserves a n an sw er to th e question: w here d o we find a sta rtin g p o in t o r fram e
o f reference w ith w hich to build a general th eo ry o f tap h o n o m y ? I suggest basic
ecological th eo ry is a g o o d place to start. A s we have seen th ro u g h o u t this
volum e, m an y ta p h o n o m ic processes co n cern the m echanical, chem ical, and
physical b re a k d o w n o f an im al carcasses a n d skeletal tissues. A nim al carcasses
are sim ply one stage in the biogeochem ical cycle o f m a tte r a n d energy. A
biogeochem ical cycle can be defined as the “ m o re o r less c ircu lar p a th s chem ical
elem ents follow fro m th e en v iro n m e n t to organism s a n d b ack to the e n v iro n
m e n t” (O d um 1971:86). A fter the d ea th o f a v erte b rate, its c o n stitu en t p a rts are

“ recycled” b ack in to th e en v iro n m en t, a n d it is th a t recycling th a t ta p h o n o m ic
research m o n ito rs. As B ehrensm eyer a n d H o o k (1992:90) note, “ d ead o rg a
nism s are d y n am ic co m p o n en ts o f the ecological w eb th a t su p p o rts living
p lan ts a n d anim als. Fossils are possible only w hen som e o f the org an ic rem ains
are n o t recycled.” . E arlier in this volum e, fo r instance, I suggested th a t b u rn e d
bones m ay n o t be gnaw ed by carn iv o res because the o rg an ic chem ical n u trien ts
in b ones th a t ca rn iv o res n o rm ally exploit - the m arro w a n d grease - have
alread y been cycled b ack in to the e n v iro n m e n t by th eir co m b u stio n . It is, then,
the p a rtic u la r biogeochem ical p a th follow ed by th o se w ith in -b o n e n u trien ts
th a t results in th eir n o t follow ing the different p a th o f being co n su m ed an d
m etab o lized in to energy w ithin a n o th e r organism . Sim ilarly, w hen bones are
tra n s p o rte d by geological processes, they are v ariously w eathered a n d ab rad ed
as sed im en tary p articles the co n stitu e n t p a rts o f w hich becom e p a rt o f the
p o te n tia l geological reservoir o f n u trie n ts fo r feeding p lan ts an d anim als. T he
recycling o f chem ical elem ents from the e n v iro n m e n t to o rganism s can be
fo u n d in th e feeding o f organism s on an im al carcasses, a n d the m echanical an d
chem ical b re ak d o w n o f th ose carcasses in to m o re basic n u trie n ts p rio r to their
co n su m p tio n .
T he re ad er m ight p ro te st th a t I have n o t tru ly suggested a w ay to shift o u r
perspective from th e m iddle ran g e o f chem ical, physical, a n d m echanical laws
u n d erp in n in g a physics o f ta p h o n o m y . T he re ad er m ight well be co rrect in such
a p ro test. I w ould no netheless arg u e th a t w h a t tap h o n o m ists are in fact
stu d y in g are p o rtio n s o f such ecologically fo u n d e d (an d ecologically ex p lain
able) biogeochem ical cycles. A n d th a t is, in fact, w h a t u n d erp in s eith er studies
th a t focus on the biasing aspects o f ta p h o n o m ic histories w ith reg ard s to
p a rtic u la r research q uestions, o r studies th a t utilize ta p h o n o m ic d a ta to gain
insights to the paleoecology o f the p ast, w h eth er o r not th a t paleoecology
con cern s hom inids. A s we m ove in to the tw enty-first century, p e rh a p s m y
suggestion reg ard in g a general th eo ry o f ta p h o n o m y will be deb ated . W h eth er
it is o r n o t, given the present state o f affairs in p aleobiology a n d arch aeo lo g y , I
d o u b t th a t the necrolog y (and ta p h o n o m y ) o f ta p h o n o m y will ever be a subject
p u rsu e d by h isto rian s o f p aleo n to lo g y o r zooarch aeo lo g y .
BIBLIOGRAPHY
A bler, W. L. 1984. A three-dim ensional m ap o f a paleontological q u arry . University o f W yom ing

Contributions to Geology 23:9-14. Laram ie.
1985. Skulls in fossil m aterial: one m echanism co n trib u tin g to th eir rarity. Journal o f
Paleontology 59:249-250.
A genbroad, L. D . 1984. H o t Springs, South D a k o ta: en trap m en t an d tap h o n o m y o f C olum bian
M am m oth. In (P. S. M a rtin and R. G . K lein, eds.) Q uaternary extinctions: a prehistoric
revolution, pp. 113-127. T ucson: U niversity o f A rizona Press.
1989. Spiral fractured m am m o th bone from n o n h u m an tap h o n o m ic processes a t H o t Springs
M am m o th Site. In (R . B onnichsen an d M . H . Sorg, eds.) Bone m odification, pp. 139-147.
O rono: U niversity o f M aine C en ter fo r the Study o f the F irst A m ericans.
1990. T he m am m o th p o p u latio n o f the H o t Springs Site an d associated m egafauna. In (L. D.
A genb ro ad, J. I. M ead, an d L. W. N elson, eds.) M egafauna and man: discovery o f A m erica's
heartland, pp. 32-39. H o t Springs: T he M am m o th Site o f H o t Springs, S outh D ak o ta , Inc.,
Scientific Papers Vol. 1.
A lexander, R. M cN . 1985. Body su p p o rt, scaling an d allom etry. In (M . H ildebrand, D . M.
B ram ble, K . F. Liem , an d D . B. W ake, eds.) Functional vertebrate anatom y, pp. 26-37.
C am bridge: B elknap Press.
A llison, P. A . 1990. D ecay processes. In (D. E. G . Briggs an d P. R. C row ther, eds.)
Palaeobiology: a synthesis, pp. 213-216. O xford: Blackwell Scientific P ublications.
1991. T ap h o n o m y has com e o f age! Palaios 6:345-346.
A llison, P. A. an d Briggs, D . E. G . 1991a. T ap h o n o m y o f nonm ineralized tissues. In (P. A.
A llison an d D . E. G . Briggs, eds.) Taphonomy: releasing the data locked in the fo s sil record,
pp. 25-70. T opics in G eobiology Vol. 9. N ew Y ork: Plenum Press.
1991b. (eds.) Taphonomy: releasing the data locked in the fo s s il record. T opics in G eobiology
Vol. 9. N ew Y ork: P lenum Press.
A nderson, K . M . 1969. E th n o g rap h ic analogy and archeological in terp re tatio n . Science 163:
133-138.
A ndrew s, P. 1990. Owls, caves and fossils. C hicago: U niversity o f C hicago Press.
1992. T he basis for tap h o n o m ic research on vertebrate fossils. In (S. F ern an d ez Lopez, ed.)
Conferencias de la reunion de tafonom ia y fosilizacion, pp. 33—43. M adrid: E ditorial
C om plutense.
A ndrew s, P. and C ook, J. 1985. N a tu ra l m odifications to bones in a tem p erate setting. M an 20:
675-691.
A ndrew s, P. an d Evans, E. N . 1983. Small m am m al b one accum ulations pro d u ced by
m am m alian carnivores. Paleobiology 9: 289-307.
. A ntia, D . D . J. 1979. B one-beds: a review o f th eir classification, occurrence, genesis, diagenesis,
geochem istry, palaeoecology, w eathering, and m icrobiotas. M ercian Geologist 7:93-174.
Ascenzi, A. 1969. M icroscopy an d prehistoric bone. In (D . B rothw ell and E. Higgs, eds.) Science
in archaeology, pp. 526-538. L ondon: T ham es an d H udson.
A scher, R. 1961a. A nalogy in archaeological in terp retatio n . Southwestern Journal o f A nthropo
logy 17: 317-325.
466
Bibliography 467
1961b. E xperim ental archeology. Am erican A nthropologist 63: 793-816.

1968. T im e’s arrow an d the archaeology o f a co n tem p o rary com m unity. In (K . C. C hang, ed.)
Settlem ent archaeology, pp. 43-52. Palo A lto: N a tio n a l Press Books.
A scher, R. and A scher, M . 1965. R ecognizing the em ergence o f m an. Science 147:243-250.
A vebury, L ord (Jo h n L ubbock). 1913. Prehistoric times as illustrated by ancient remains and the
manners and customs o f modern savages, 7th edition. N ew Y ork: H enry H o lt an d C om pany.
Avery, G . 1984. Sacred cow s o r jack a l kitchens, hyena m iddens an d bird nests: som e im plications
o f m ulti-agent co n trib u tio n s to archaeological accum ulations. In (M . H all, G . Avery, D. M.
A very, M . L. W ilson, and A. J. B. H um phreys, eds.) Frontiers: Southern A frican archaeo
logy today, pp. 344-348. British A rchaeological R ep o rts In tern atio n al Series 207.
Badgley, C. 1986a. C o u n tin g individuals in m am m alian fossil assem blages from fluvial
environm ents. Palaios 1: 328-338.
1986b. T ap h o n o m y o f m am m alian fossil rem ains from Siwalik rocks o f Pak istan . Paleobio
logy 12: 119-142.
Bailey, G . 1983. H u n ter-g ath erer b eh av io u r in prehistory: problem s an d perspectives. In (G.
Bailey, ed.) Hunter-gatherer econom y in prehistory, pp. 1-6. C am bridge: C am bridge
U niversity Press.
B aker, J. and Brothw ell, D . 1980. A nim al diseases in archaeology. L ondon: A cadem ic Press.
Balcom b, R. 1986. Songbird carcasses d isap p ear rapidly from agricultural fields. A u k 103:
817-820.
Balme, J. 1980. A n analysis o f charred bone from D evil’s L air, w estern A ustralia. Archaeology
and Physical A nthropology in Oceania 15: 81-85.
Barlow , J. 1984. M o rtality estim ation: biased results from u n biased ages. Canadian Journal o f
Fisheries and Aquatic Science 41: 1843-1847.
B arnosky, A. D . 1985. T ap h o n o m y an d herd stru ctu re o f the extinct Irish Elk (M egaloceros
giganteus). Science 228: 340-344.
1986. ‘Big gam e’ extinction caused by late Pleistocene clim atic change: Irish elk (M egaloceros
giganteus) in Ireland. Quaternary Research 25: 128-135.
B artram , L. E., K roll, E. M ., and B unn, H . T. 1991. V ariability in cam p stru ctu re an d b o n e food
refuse p attern in g a t K u a San h u n ter-g ath erer cam ps. In (E. M . K ro ll and T. D . Price, eds.)
The interpretation o f archaeological spatial patterning, pp. 77-148. N ew Y ork: Plenum
Press.
>-Bartsiokas, A . and M iddleton, A. P. 1992. C h aracterizatio n an d d atin g o f recent an d fossil bone
by X -ray diffraction. Journal o f Archaeological Science 19:63-72.
Becking, J. H . 1975. T he u ltrastru ctu re o f avian eggshell. Ibis 117:143-151.
Behrensm eyer, A. K. 1975a. T ap h o n o m y an d paleoecology in the H om inid fossil record.
Yearbook o f Physical Anthropology 19: 36-50.
1975b. T he tap h o n o m y and paleoecology o f Plio-Pleistocene vertebrate assem blages east o f
L ake R udolf, K enya. Bulletin o f the M useum o f Comparative Zoology 146: 473-578.
1978. T ap ho nom ic and ecologic in fo rm atio n from bone w eathering. Paleobiology 4: 150-162.
1979. T he h a b ita t o f Plio-Pleistocene hom inids in E ast A frica: tap h o n o m ic an d m icro-
stratig raphic evidence. In (C. Jolly, ed.) Early hominids o f A frica, pp. 165-189. L ondon:
D uckw orth.
1981. V ertebrate paleoecology in a recent E ast A frican ecosystem . In (J. G ray, A. J. B oucot,
and W. B. N. Berry, eds.) C om m unities o f the p a st, pp. 591-615. S troudsburg: H utchinson
Ross Publishing Co.
1982. Tim e resolution in fluvial v erteb rate assem blages. Paleobiology 8: 211-228.
1983. P attern s o f n a tu ra l bone d istrib u tio n o n recent land surfaces: im plications for
archaeological site fo rm atio n . In (J. C lu tto n -B ro ck an d C. G rigson, eds.) A nim als and
archaeology: 1. hunters and their prey, pp. 93-106. British A rchaeological R ep o rts In te r
n atio nal Series 163.
468 Bibliography
1984. T ap h o n o m y and the fossil record. Am erican Scientist 72: 558-566.

1987. T ap h o n o m y an d hunting. In (M . H . N itecki an d D. V. N itecki, eds.) The evolution o f
human hunting, pp. 423—450. N ew Y ork: Plenum Press.
1988a. T he pull o f the recent analogue. Palaios 3: 373.
1988b. V ertebrate preservation in fluvial channels. Palaeogeography, Palaeoclimatology,
Palaeoecology 63: 183-199.
1990. T ran sp o rt-h y d ro d y n am ics: bones. In (D. E. G . Briggs and P. R. C row ther, eds.)
1991. T errestrial v ertebrate accum ulations. In (P. A. A llison an d D . E. G . Briggs, eds.)
Taphonom y: releasing the data locked in the fo s sil record, pp. 291-335. T opics in G eobiology
Vol. 9. N ew Y ork: Plenum Press.
B ehrensm eyer, A. K. an d Badgley, C. 1989. F orew ord. In W eigelt 1989, pp. vii-x.
B ehrensm eyer, A. K . an d Boaz, D. E. D. 1980. The recent bones o f A m boseli N atio n al P ark,
K enya, in relation to E ast A frican paleoecology. In (A. K . B ehrensm eyer an d A. P. Hill,
eds.) Fossils in the m aking, pp. 72-92. C hicago: U niversity o f C hicago Press.
B ehrensm eyer, A. K ., G o rd o n , K. D ., an d Y anagi, G . T. 1986. T ram p lin g as a cause o f bone
surface d am age and pseudo-cutm arks. N ature 319: 768-771.
1989. N o n h u m an bone m odification in M iocene fossils from P akistan. In (R . B onnichsen and
M. H. Sorg, eds.) Bone m odification, pp. 99-120. O rono: U niversity o f M aine C en ter fo r the
Study o f the F irst A m ericans.
Behrensm eyer, A . K . an d Hill, A. P. 1980. (eds.) Fossils in the making: vertebrate taphonom y and
paleoecology. C hicago: U niversity o f Chicago Press.
Behrensm eyer, A. K. an d H oo k , R. W. 1992. P aleoenvironm ental contexts an d tap h o n o m ic
m odes. In (A. K . Behrensm eyer, J. D . D am u th , W. A. D iM ichele. R. P o tts, H .-D . Sues, and
S. L. W ing, eds.) Terrestrial ecosystem s through time: evolutionary paleoecology o f terrestrial
plants and animals, pp. 15-136. Chicago: U niversity o f C hicago Press.
Behrensm eyer, A. K. and K idw ell, S. M . 1985. T a p h o n o m y ’s co n trib u tio n to paleobiology.
Paleobiology 11: 105-119.
Behrensm eyer, A. K ., W estern, D ., an d D ech an t Boaz, D . E. 1979. N ew perspectives in
vertebrate paleoecology from a recent bone assem blage. Paleobiology 5: 12-21.
Bell, L. S. 1990. P aleopathology an d diagenesis: an SEM evalu ation o f stru ctu ral changes using
b ackscattered electron im aging. Journal o f Archaeological Science 17: 85-108.
Berger, J. 1983. Ecology and c atastro p h ic m o rtality in wild horses: im plications fo r interpreting
fossil assem blages. Science 220: 1403-1404.
Berger, W. H . 1991. O n the extinction o f the m am m oth: science an d m yth. In (D. W . M uller, J. A.
M cK enzie, and H . W eissert, eds.) Controversies in m odem geology, pp. 115-132. L ondon:
A cadem ic Press.
B ettinger, R. L. 1991. Hunter-gatherers: archaeological and evolutionary theory. N ew Y ork:
Plenum Press.
B ickart, K . J. 1984. A field experim ent in avian tap h o n o m y . Journal o f Vertebrate Paleontology
4: 525-535.
Biddick, K . A. an d T om enchuck, J. 1975. Q uantifying co n tin u o u s lesions an d fractures on long
bones. Journal o f Field Archaeology 2: 239-249.
B inford, L. R. 1964. A con sideratio n o f archaeological research design. American A ntiquity 29:
425-441.
1967. Sm udge pits and hide sm oking: the use o f analogy in archaeological reasoning. American
A ntiquity 32: 1-12.
1968. A rchaeological perspectives. In (S. R. B inford an d L. R. B inford, eds.) N ew perspectives
in archaeology, pp. 5-32. C hicago: A ldine.
1972. A n archaeological perspective. N ew Y ork: Sem inar Press.
B ibliography 469
1977. G eneral in tro d u ctio n . In (L. R. B inford, ed.) For theory building in archaeology, pp. 1-
10. N ew Y ork: A cadem ic Press.
1978. N unam iut ethnoarchaeology. N ew Y ork: A cadem ic Press.
1980. W illow sm oke an d d ogs’ tails: h u n ter-g ath erer settlem ent system s and archaeological
site form atio n . Am erican A ntiquity 45: 4-20.
1981a. B ehavioral archaeology an d the “ Pom peii Prem ise.” Journal o f Anthropological
Research 37: 195-208.
1981b. Bones: ancient men and m odern m yths. N ew Y ork: A cadem ic Press.
1984a. B utchering, sharing, an d the archaeological record. Journal o f Anthropological
Archaeology 3: 235-257.
1984b. Faunal remains fr o m Klasies River M outh. O rlando: A cadem ic Press.
1986. C om m ent on B unn an d K roll 1986. Current Anthropology 27: 444-446.
1987. R esearching am biguity: fram es o f reference and site structure. In (S. K ent, ed.) M ethod
and theory fo r activity area research: an ethnoarchaeological approach, pp. 449-512. New
Y ork: C olum bia U niversity Press.
1988. F act an d fiction ab o u t the Zinjanthropus floor: d ata, argum ents, and interp retatio n s.
Current Anthropology 29: 123-135.
B inford, L. R. and B ertram , J. B. 1977. Bone frequencies - and a ttritio n a l processes. In (L. R.
B inford, ed.) For theory building in archaeology, pp. 77-153. N ew Y ork: A cadem ic Press.
B lum enschine, R. J. 1986a. C arcass con su m p tio n sequences an d the archaeological distinction o f
scavenging an d hunting. Journal o f H uman Evolution 15: 639-659.
1986b. Early hom inid scavenging opportunities: implications o f carcass availability in the
Serengeti and Ngorongoro ecosystem s. British A rchaeological R ep o rts In tern atio n al Series
283.
1987. C h aracteristics o f an early hom inid scavenging niche. Current A nthropology 28:
383-407.
1988. A n experim ental m odel o f th e tim ing o f hom inid an d carnivore influence on archaeologi
cal bone assem blages. Journal o f Archaeological Science 15: 483-502.
Blum enschine, R. J. and C aro, T. M . 1986. U n it flesh w eights o f som e E ast A frican bovids.
African Journal o f Ecology 24: 273-286.
B lum enschine, R. J. an d C avallo, J. A. 1992. Scavenging an d h u m an evolution. Scientific
American 267(4): 90-96.
Blum enschine, R. J. an d Selvaggio, M . 1988. Percussion m arks o n bone surfaces as a new
diagnostic o f hom inid behavior. N ature 333: 763-765.
1991. O n the m ark s o f m arro w bone processing by h am m erstones and hyaenas: their
an atom ical p attern in g an d archaeological im plications. In (J. D. C lark, ed.) Cultural
beginnings: approaches to understanding early hom inid life-ways in the African savanna, pp.
17-32. U n io n In tern atio n ale des Sciences Prehistoriques et P ro to h isto riq u es M o nogra-
phien B and 19.
Boaz, D. 1980. Fossils in the m aking: tap h o n o m y an d verteb rate paleoecology. Current
Anthropology 21: 404-406.
1982. M odern riverine taphonom y: its relevance to the in terp retatio n o f Plio-Pleistocene
h om inid paleoecology in the O m o Basin, E th io p ia. Ph. D. dissertation, U niversity o f
C alifornia, Berkeley. A n n A rbor: U niversity M icrofilm s In tern atio n al.
Boaz, N . T. and Behrensm eyer, A. K . 1976. H om inid taphonom y: tra n sp o rt o f hu m an skeletal
p arts in an artifical fluviatile environm ent. Am erican Journal o f P hysical Anthropology 45:
53-60.
B ohannan, P. and G lazer, M . 1988. (eds.) In tro d u ctio n . In High Points in Anthropology, 2nd
edition. N ew Y ork: M cG raw -H ill, Inc.
B onnichsen, R. 1973. Som e op eratio n al aspects o f h u m an an d anim al bone alteratio n . In (B. M.
470 Bibliography
G ilbert, ed.) M am m alian osteo-archaeology: N orth America, pp. 9-24. C olum bia: M issouri
A rchaeological Society, U niversity o f M issouri.
1975. O n faunal analysis an d the au stralopithecines. Current A nthropology 16: 635-636.
1979. Pleistocene bone technology in the Beringian Refugium . A rchaeological Survey o f
C an ad a P ap er N o . 89, M ercury Series. O ttaw a: N a tio n al M useum o f M an.
1989. A n in tro d u c tio n to tap h o n o m y w ith an archaeological focus. In (R . B onnichsen an d M.
H. Sorg, eds.) Bone m odification, pp. 1-5. O rono: U niversity o f M aine C en ter fo r the Study
o f the F irst A m ericans.
B onnichsen, R. an d Sanger, D . 1977. In teg ratin g faunal analysis. Canadian Journal o f Archaeo
logy 1: 109-133.
B onnichsen, R. and Sorg, M . H . 1989. (eds.) Bone modification. O rono: U niversity o f M aine
C enter for the Study o f the F irst A m ericans.
B onnichsen, R. and W ill, R. T. 1980. C u ltu ral m odification o f bone: the experim ental ap p ro ach
in faun al analysis. In (B. M . G ilbert, ed.) M am m alian osteology, pp. 7-30. M issouri
A rchaeological Society, U niversity o f M issouri-C olum bia.
B orrero, L. 1990. F ueg o -P atag o n ian b one assem blages a n d the p roblem o f com m u n al guanaco
hunting. In (L. B. D avis and B. O. K . Reeves, eds.) H unters o f the recent p a st, pp. 373-399.
L ondon: U nw in H ym an.
Bower, J. 1986. In search o f the past. C hicago: D orsey Press.
Bower, J. R. F., G ifford, D . P., and L ivingston, D . 1985. E xcavations a t the L oiyangalani Site,
Serengeti N atio n al P ark , T an zan ia. N ational Geographic Society Research Reports 20:
41-56.
Bowyer, R. T. 1983. O steoph ag ia an d an tler breakage am ong R oosevelt elk. California Fish and
Game 69: 84-88.
Boyd, D . W . an d Newell, N . D . 1972. T ap h o n o m y an d diagenesis o f a P erm ian fossil assem blage
from W yom ing. Journal o f Paleontology 46: 1-14.
B rain, C. K . 1967a. Bone w eathering an d the p roblem o f bone pseudo-tools. South African
Journal o f Science 63: 97-99.
1967b. H o tte n to t food rem ains an d th eir bearing on the in terp re ta tio n o f fossil bone
assem blages. Scientific Papers o f the N am ib Desert Research S tation 32: 1-7.
1969. T he co n trib u tio n o f N am ib D esert H o tte n to ts to an u n d erstand in g o f australopithecine
bone accum ulations. Scientific Papers o f the N am ib Desert Research Station 39: 13-22.
1974. Som e suggested procedures in th e analysis o f b one accum ulations from southern
A frican Q u atern ary caves. Annals o f the Transvaal M useum 29: 1-8.
1976. Some principles in the in terp re tatio n o f bone accum ulations associated w ith m an. In (G.
L. Isaac and E. R. M cC ow n, eds.) H uman origins: Louis L ea key and the East African
evidence, pp. 97-116. M enlo Park: W. A. B enjam in. Inc.
1980. Som e criteria for the recognition o f bone-collecting agencies in A frican caves. In (A. K.
B ehrensm eyer and A. P. Hill, eds.) Fossils in the m aking, pp. 107-130. C hicago: U niversity
o f C hicago Press.
1981. The hunters or the hunted? A n introduction to African cave taphonom y. Chicago:
U niversity o f C hicago Press.
1989. T he evidence for bone m odification by early hom inids in so u th ern A frica. In (R.
B onnichsen and M . H . Sorg, eds.) Bone m odification, pp. 291-297. O rono: U niversity o f
M aine C en ter for the Study o f the F irst A m ericans.
B rain, C. K . and Sillen, A. 1988. Evidence from the S w artk ran s cave fo r the earliest use o f fire.
N ature 336: 464-466.
Bramwell, D ., Y alden, D . W ., and Y alden, P. E. 1987. B lack grouse as the prey o f the golden
eagle at an archaeological site. Journal o f Archaeological Science 14: 195-200.
B rett, C. E. 1990a. D estructive tap h o n o m ic processes an d skeletal du rability. In (D . E. G . Briggs
Bibliography 47 1
and P. R. C row ther, eds.) Palaeobiology: a synthesis, pp. 223-226. O xford: Blackwell
Scientific Publications.
1990b. O b ru tio n deposits. In (D. E. G . Briggs and P. R. C row ther, eds.) Palaeobiology: a
synthesis, pp. 239-243. O xford: Blackwell Scientific Publications.
Brett. C. E. and Speyer, S. E. 1990. T aphofacies. In (D . E. G . Briggs and P. R. C row ther, eds.)
Breuil, H. A. 1932. Le Feu et l’industrie de pierre et d ’os d an s le gisem ent du “ S in an th ro p u s” a
C hou K ou Tien. Anthropologie 42: 1-17.
1938. T he use o f bone im plem ents in the O ld Paleolithic period. A ntiquity 12: 56-67.
1939. Bone and antler industry o f the Choukoutien S in an th ro p u s site. P alaeontologica Sinica,
Series D , New Series N o. 6(117). Peiking: G eological Survey o f C hina.
Brewer, D. J. 1992. Z ooarchaeology: m ethod, theory, an d goals. In (M . B. Schiffer, ed.)
Archaeological m ethod and theory vol. 4, pp. 195-244. Tucson: U niversity o f A rizona Press.
Briggs, D . E. G . 1990. F lattening. In (D. E. G . Briggs an d P. R. C row ther, eds.) Palaeobiology: a
synthesis, pp. 244-247. O xford: Blackwell Scientific P ublications.
Briggs, D . E. G . and C row ther, P. R. 1990. (eds.) Palaeobiology: a synthesis. O xford: Blackwell
Scientific Publications.
Brink, J. and Dawe, B. 1989. Final report o f the 1985 and 1986fie ld season at H ead-Sm ashed-In
buffalo ju m p , Alberta. A rchaeological Survey o f A lb erta M an u scrip t Series N o. 16.
E dm onton.
B riuer, F. L. 1977. P lan t and anim al rem ains from caves an d rockshelters o f C hevelon C anyon,
A rizona: m ethods for isolating depositional processes. P h.D . dissertation, U niversity o f
C alifornia-L os Angeles. A nn A rb o r: U niversity M icrofilm s In tern atio n al.
B rom age, T. G . 1984. In terp retatio n o f scanning electron m icroscopic im ages o f ab rad ed
form ing bone surfaces. American Journal o f Physical Anthropology 64: 161-178.
B rom age, T. G . and Boyde, A. 1984. M icroscopic criteria for the determ in atio n o f directionality
o f cutm ark s on bone. American Journal o f Physical A nthropology 65: 359-366.
B rongersm a-Sanders, M. 1957. M ass m o rtality in the sea. Geological Society o f A m erica M em oir
67(1): 941-1010.
B rothw ell, D. 1976. F u rth e r evidence o f bone chew ing by ungulates: the sheep o f N o rth
R onaldsay, O rkney. Journal o f Archaeological Science 3: 179-182.
B row n, R. D . 1983. (ed.) A ntler development in Cervidae. Kingsville, TX: C aesar K leberg W ildlife
R esearch Institute.
B uckland, W. 1823. Reliquiae diluvianae, or, observations on the organic remains contained in
caves, fissures, and diluvial gravel, and on other geological phenom ena, attesting to the action
o f an universal deluge. L ondon: M urray.
B uikstra, J. E. and Swegle, M. 1989. Bone m odification due to burning: experim ental evidence.
In (R. B onnichsen and M . H. Sorg, eds.) Bone m odification, pp. 247-258. O rono: U niversity
o f M aine C enter for the Study o f the F irst A m ericans.
B unn, H . T. 1981. A rchaeological evidence for m eat-eating by Plio-Pleistocene hom inids from
K oobi F o ra and O lduvai G orge. N ature 291: 574—577.
1982. M eat-eating and h u m an evolution: studies on the diet an d subsistence p attern s o f Plio-
Pleistocene hom inids in E ast A frica. Ph. D. dissertatio n , U niversity o f C alifornia, Berkeley.
A nn A rb or: U niversity M icrofilm s In tern atio n al.
1983. Evidence on the diet an d subsistence p attern s o f Plio-Pleistocene hom inids a t K oobi
F o ra, K enya, an d O lduvai G orge, T an zan ia. In (J. C lu tto n -B ro ck an d C. G rigson, eds.)
A nim als and archaeology: 1. hunters and their prey, pp. 21-30. British A rchaeological
R ep o rts In tern atio n al Series 163.
1986. P attern s o f skeletal rep resen tatio n an d h om inid subsistence activities at O lduvai G orge,
T anzania, and K oobi F o ra , K enya. Journal o f H um an Evolution 15: 673-690.
472 Bibliography
1989. D iagnosing Plio-Pleistocene h om inid activity w ith b one fractu re evidence. In (R.
Bonnichsen and M . H . Sorg, eds.) Bone m odification, pp. 299-315. O rono: U niversity of
M aine C en ter for the Study o f the F irst A m ericans.
1991. A tap h o n o m ic perspective on the archaeology o f h u m an origins. Annual Review o f
Anthropology 20: 4 3 3 ^ 6 7 .
B unn, H . T., B artram , L. E., an d K roll, E. M . 1988. V ariability in bone assem blage form atio n
from H ad za hunting, scavenging, an d carcass processing. Journal o f Anthropological
B unn, H . T. and B lum enschine, R. J. 1987. O n “ th eoretical fram ew ork an d tests” o f early
hom inid m eat an d m arrow acquisition: a reply to Shipm an. American Anthropologist
B unn, H . T. and K roll, E. M . 1986. System atic butchery by Plio/Pleistocene hom inids at O lduvai
G orge, T an zania. Current Anthropology 27: 431-452.
1987. R eply to P otts. Current A nthropology 28: 96-98.
1988. R eply to B inford. Current Anthropology 29: 135-149.
B unn, H . T ., K roll, E. M ., a n d B artram , L. E. 1991. Bone d istrib u tio n on a m odern E ast A frican
landscape an d its archaeological im plications. In (J. D. C lark, ed.) C ultural beginnings:
approaches to understanding early hom inidlife-w ays in the African savanna, pp. 33-54. U nion
In tern atio n ale des Sciences P rehistoriques et P ro to h isto riq u es M o n o g rap h ien B and 19.
B urgett, G . R. 1990. T he bones o f the beast: resolving questions o f faunal assem blage form ation
processes th ro u g h actualistic research. P h.D . dissertation, U niversity o f N ew M exico,
A lbuquerque.
B urr, D. B. 1980. The relationships am o n g physical, geom etrical an d m echanical p roperties o f
bone, w ith a note on the properties o f n o n h u m an prim ate bone. Yearbook o f Physical
B utler, V. L. 1987. D istinguishing n a tu ra l from cu ltu ral salm onid deposits in the Pacific
N orthw est o f N o rth A m erica. In (D . T. N ash an d M . D . Petraglia, eds.) N atural form a tio n
processes and the archaeological record, pp. 131-149. British A rchaeological R eports
In tern atio n al Series 352.
1990. D istinguishing n atu ra l from cu ltu ral salm onid deposits in Pacific N o rth w est N o rth
A m erica. P h.D . d issertatio n , U niversity o f W ashington. A n n A rb o r: U niversity
M icrofilm s.
1993. N a tu ra l versus cu ltu ral salm onid rem ains: origin o f T he D alles R o ad cu t bones,
C olum bia River, O regon, U .S.A . Journal o f Archaeological Science 20: 1-24.
Butzer, K. W . 1982. Archaeology as human ecology. C am bridge: C am bridge U niversity Press.
C adee, G . C. 1990. T he history o f tap h o n o m y . In (S. K. D o no v an , ed.) The processes o f
fossilization, pp. 3-21. N ew Y ork: C olu m b ia U niversity Press.
C ahen, D. an d M oeyersons, J. 1977. S ubsurface m ovem ents o f stone artefacts an d their
im plications for the p rehistory o f C en tral A frica. N ature 266: 812-815.
C arlson, S. J. 1990. V erteb rate dental structures. In (J. G . C arter, ed.) S k ele ta l biomineralization:
patterns, processes and evolutionary trends, pp. 531-556. N ew Y ork: V an N o stran d
R einhold.
C asteel, R. W . 1971. D ifferential bone d estruction: som e com m ents. A merican A ntiquity 36:
466-469.
1976. Fish remains in archaeology and environmental studies. N ew Y ork: A cadem ic Press.
C aughley, G . 1966. M o rtality p attern s in m am m als. Ecology 47: 906-918.
1977. A nalysis o f vertebrate populations. L ondon: Jo h n W iley & Sons.
C avallo, J. A. and B lum enschine, R. J. 1989. T ree-stored leopard kills: expanding the hom inid
scavenging niche. Journal o f H um an Evolution 18: 393-399.
C ham bers, A. L. 1992. Seal bone m ineral density: its effect on specim en survival in archaeologi
Bibliography 473
cal sites. U np u blished H o n o rs thesis, D ep artm e n t o f A nth ro p o lo g y , U niversity o f M is

souri, C olum bia.
C haplin, R. E. 1965. A nim als in archaeology. A ntiquity 39: 204—211.
1971. The study o f anim al bones fr o m archaeological sites._L ondon: Sem inar Press.
C h arlto n , T. H . 1981. A rchaeology, eth n o h isto ry , and ethnology: interpretive interfaces. In (M .
B. Schiffer, ed.) Advances in archaeological m ethod and theory vol. 4, pp. 129-176. New
Y ork: A cadem ic Press.
C hase, P. G . 1985. O n the use o f B in fo rd ’s utility indices in the analysis o f archaeological sites.
P .A .C .T . 11: 287-302.
C hom ko, S. A. and G ilbert, B. M . 1991. Bone refuse an d insect rem ains: th eir p o ten tial for
tem p o ral resolution o f the archaeological record. American A ntiq u ity 56: 680-686.
C lark, J., Beerbow er, J. R ., and K ietzke, K . K . 1967. Oligocene sedimentation, stratigraphy and
paleoclim atology in the Big Badlands o f South D akota. F ieldiana G eology M em oir 5.
C lark, J. an d G uensberg, T. E. 1970. P o p u latio n dynam ics o f L eptom eryx. Fieldiana Geology 16:
411-511.
C lark, J. an d K ietzke, K. K . 1967. Paleoecology o f th e L ow er N o d u la r Z one, Brule F o rm atio n ,
in the Big B adlands o f S outh D ak o ta . In (J. C lark, J. R. Beerbow er, and K . K . K ietzke)
Oligocene sedimentation, stratigraphy and paleoclim atology in the Big Badlands o f South
D ako ta, pp. 111-137. F ieldiana G eology M em oir 5.
C lark, J. D . 1972. Paleolithic butchery practices. In (P. J. U cko, R. T ringham , an d G . W.
D im bleby, eds.) M an, settlem ent and urbanism , pp. 149-156. C am bridge: D u ck w o rth .
C lason, A. T. 1975. (ed.) A rchaeozoologicalstudies. A m sterdam : N o rth H ollan d P ublishing Co.
Coe, M. J. 1978. T he decom position o f elephant carcasses in the T savo (E ast) N atio n al Park,
K enya. Journal o f A rid Environm ents 1: 71-86.
1980. T he role o f m o dern ecological studies in the reco n stru ctio n o f p aleoenvironm ents in sub-
S ah aran A frica. In (A. K . B ehrensm eyer and A. P. Hill, eds.) Fossils in the m aking, pp. 5 5 -
67. C hicago: U niversity o f C hicago Press.
Coe, R. J., D ow ning, R. L., and M cG innes, B. S. 1980. Sex and age bias in hunter-killed w hite
tailed deer. Journal o f Wildlife M anagem ent 44: 245-249.
Colley, S. M . 1990a. H u m an s as tap h o n o m ic agents. In (S. Solom on, I. D avidson, and D.
W atson, eds.) Problem solving in taphonom y, pp. 50-64. T em pus Vol. 2.
1990b. T he analysis and in terp retatio n o f archaeological fish rem ains. In (M . B. Schiffer, ed.)
Archaeological m ethod and theory vol. 2, pp. 207-253. T ucson: U niversity o f A rizona Press.
Collier, S. an d W hite, J. P. 1976. G et them young? Age an d sex inferences o n anim al
dom estication in archaeology. Am erican A ntiquity 41: 96-102.
C onybeare, A. and H aynes, G . 1984. O bservations o n elep h an t m o rtality and bones in w ater
holes. Quaternary Research 22: 189-200.
C ook, J. 1986. The application o f scanning electron m icroscopy to tap h o n om ic and arch aeo lo g i
cal problem s. In (D . A. R oe, ed.) Studies in the Upper Paleolithic o f Britian and northwest
Europe, pp. 143-163. B ritish A rchaeological R ep o rts In tern atio n a l Series 296.
C ook, S. F ., B rooks, S. T., an d E zra-C ohn, H . C. 1961. T he process o f fossilization. Southwestern
Journal o f Anthropology 17: 355-364.
C ornw all, I. W. 1956. Bones fo r the archaeologist. L on d o n : Phoenix H ouse.
1968. Prehistoric animals and their hunters. L ondon: F a b e r an d F aber.
C rader, D . C. 1983. R ecent single-carcass bone scatters an d the p roblem o f “ b u tch ery ” sites in
the archaeological record. In (J. C lu tto n -B ro ck an d C. G rigson, eds.) A nim als and
archaeology: 1. hunters and their p rey, pp. 107-141. British A rchaeological R eports
C raig, G . Y. an d O ertel, G . 1966. D eterm inistic m odels o f living an d fossil p o p u latio n s o f
anim als. Q uarterly Journal o f the Geological Society o f London 122: 315-355.
474 Bibliography
C raw ford, H . 1982. A nalogies, anom alies an d research strategy. Paleorient 8: 5-9.
C ribb, R. 1985. The analysis o f ancient h erding systems: an ap p licatio n o f co m p u ter sim ulation
in faunal studies. In (G . B arker an d C. G am ble, eds.) B eyond dom estication in prehistoric
Europe, pp. 75-106. L on don : A cadem ic Press.
1987. T he logic o f the herd: a co m p u ter sim ulation o f archaeological herd structure. Journal o f
Anthropological Archaeology 6: 376—415.
C ruz-U ribe, K . 1988. T he use an d m eaning o f species diversity an d richness in archaeological
faunas. Journal o f Archaeological Science 15: 179-196.
1991. D istinguishing hyena from hom inid bone accum ulations. Journal o f Field Archaeology
18: 467—486.
C urrey, J. 1984. The m echanical adaptations o f bones. Princeton: P rin ceto n U niversity Press.
1990. B iom echanics o f m ineralized skeletons. In (J. G . C arter, ed.) S keleta l biomineralization:
patterns, processes and evolutionary trends, pp. 11-25. N ew Y ork: V an N o stra n d R einhold.
C ussler, E. L. and F eathersto n e, J. D . B. 1981. D em ineralization o f p o ro u s solids. Science 213:
1018-1019.
C zaplew ski, R. L., C row e, D . M ., an d M cD o n ald , L. L. 1983. Sam ple sizes an d confidence
intervals fo r wildlife p o p u la tio n ratios. W ildlife Society Bulletin 11: 121-127.
D aly, P. 1969. A p proaches to faunal analysis in archaeology. American A ntiquity 34: 146-153.
D am u th , J. 1982. A nalysis o f the preserv atio n o f com m unity stru ctu re in assem blages o f fossil
m am m als. Paleobiology 8: 434-446.
D ’A ndrea, A. C. an d G o tth a rd t, R. M . 1984. P red ato r an d scavenger m odification o f recent
equid skeletal assem blages. A rctic 37: 276-283.
D aniel, G . 1981. A short history o f archaeology. L ondon: T ham es an d H udson.
D art, R. A. 1949. The p re d a to ry im plem ental technique o f the au stralopithecines. American
Journal o f Physical Anthropology 7: 1-16.
1956a. C u ltu ral status o f the S o u th A frican m an-apes. Sm ithsonian Institution A nnual Report
fo r 1955, pp. 317-338.
1956b. M yth o f the bo n e-accum ulating hyena. Am erican A nthropologist 58: 40-62.
1957. The osteodontokeratic culture o f A u stralo p ith ecu s prom etheus. T ran sv aal M useum
M em oir N o. 10. P reto ria.
1958. B one tools and p orcupine gnaw ing. Am erican Anthropologist 60: 715-724.
1960. T he bone tool m an u factu rin g ability o f Australopithecus prom etheus. American A nthro
pologist 62: 134—138.
D avid, B. 1990. H ow w as this bone b u rn t? In (S. Solom on, I. D avidson, an d D . W atson, eds.)
Problem solving in taphonom y, pp. 65-79. T em pus vol. 2.
D avies, D . J., Powell, E. N ., and S tan to n , R. J., Jr. 1989. Relative rates o f shell d issolution and
net accum ulation - a com m entary: can shell beds form by the g rad u al accu m u latio n o f
biogenic debris on the sea floor? Lethaia 22: 207-212.
D avis, K . L. 1985. A tap h o n o m ic ap p ro ac h to experim ental bone fractu rin g an d applications to
several S ou th A frican Pleistocene sites. P h .D . dissertation, State U niversity o f N ew Y o rk at
B ingham ton.
D avis, S. J. M . 1987. The archaeology o f animals. N ew H aven: Y ale U niversity Press.
D aw kins, W . B. 1869. O n the d istrib u tio n o f British postglacial m am m als. Q uarterly Journal o f
the Geological Society o f London 25: 192-217.
1874. Cave hunting, researches on the evidence o f caves respecting the early inhabitants o f
Europe. L ondon: M acm illan & C om pany.
Deevey, E. S., Jr. 1947. Life tables fo r n atu ral p o p u latio n s o f anim als. Q uarterly Review o f
Biology 22: 283-314.
de R ousseau, C. J. 1988. Bone biology. In (I. T attersall, E. D elson, an d J. van C ouvering, eds.)
Encyclopedia o f hum an evolution and prehistory, pp. 95-96. N ew Y ork: G arla n d Publishing.
Bibliography 475
d ’E rrico, F ., G iacobini, B., an d Puech, P .-F . 1984. A n experim ental study o f the technology o f
bone-im plem ent m anufactu re. M A S C A Journal 3: 71-74.
D erry, D. E. 1911. D am age done to skulls an d bones by term ites. N ature 86: 245-246.
D ibble. D. S. and L orrain, D. 1968. Bonfire Shelter: a stratified bison kill site, Val Verde County,
Texas. Texas M em orial M useum M iscellaneous P apers N o. 1.
D ixon, E. J. 1984. C o n tex t an d enviro n m en t in tap h o n o m ic analysis: exam ples from A la sk a ’s
P orcupine R iver caves. Quaternary Research 22: 201-215.
sJD ixon, E. J. an d T h o rson, R. M . 1984. T ap h o n o m ic analysis an d in terp re ta tio n in N o rth
A m erican Pleistocene archaeology. Q uaternary Research 22: 155-159.
D od d, J. R. and S tan to n , R. J., Jr. 1981. Paleoecology: concepts and applications. N ew Y ork:
W iley-Interscience.
D odso n, P. 1971. Sedim entology and tap h o n o m y o f the O ld m an F o rm a tio n (C am panian),
D in o sau r Provincial P ark , A lb erta (C an ad a). Palaeogeography, Palaeoclim atology,
Palaeoecology 10: 21-74.
1973. The significance o f sm all bones in paleoecological in terp retatio n . University o f W yom ing
Contributions to Geology 12: 15-19. Laram ie.
1980. V ertebrate burials. Paleobiology 6: 6-8.
D od son, P and W exlar, D. 1979. T ap ho n o m ic investigations o f owl pellets. Paleobiology 5:
279-284.
D o novan, S. K . 1990. (ed.) The processes offossilization. N ew Y ork: C olum bia U niversity Press.
D oveton, J. H . 1979. N um erical m eth o d s fo r the reco n stru ctio n o f fossil m aterial in three
dim ensions. Geological M agazine 116: 215-226.
D uckw orth, W. L. H . 1904. N o te o n the dispersive pow er o f ru nn in g w ater on skeletons: with
p articu lar reference to the skeletal rem ains o f Pithecanthropus erectus. In (W. L. H.
D uck w orth , ed.) Studies fr o m the anthropological laboratory, pp. 274-277. C am bridge:
C am bridge U niversity Press.
D unnell, R. C. 1992. A rchaeology an d ev o lutionary science. In (L. W andsnider, ed.) Quandaries
and quests: visions o f archaeology’s fu tu re , pp. 209-224. S ou th ern Illinois U niversity C enter
for A rchaeological Investigations O ccasional P ap er N o. 20. C arbondale.
D uring, E. M . an d N ilsson, L. 1991. M echanical surface analysis o f bone: a case study o f cut
m arks an d enam el hypoplasia on a N eolithic cran iu m from Sweden. Am erican Journal o f
Physical Anthropology 84: 113-125.
E frem ov, I. A . 1940. T aphonom y: a new b ran ch o f paleontology. Pan-Am erican Geologist 74:
81-93.
1958. Som e consid eration s on biological bases o f paleozoology. Vertebrata Palasiatica 2(2/3):
83-98.
Eickhoff, S. an d H errm an n , B. 1985. Surface m ark s on bones from a N eolithic collective grave
(O dagsen, Low er Saxony): a study on differential diagnosis. Journal o f H um an Evolution 14:
263-274.
Elder, R. L. and Sm ith, G . R. 1988. Fish tap h o n o m y an d en v ironm ental inference in
paleolim nology. Palaeogeography, Palaeoclim atology, Palaeoecology 62: 577-592.
Elkin, D. C. and Z an ch etta, J. R. 1991. D ensitom etria osea de cam elidos - aplicaciones
arqueologicas. A d a s del X Congreso Nacional de Arqueologia Argentina 3: 195-204.
C atam arca.
Em erson, A. M . 1990. A rchaeological im plications o f variability in the econom ic an ato m y o f
Bison bison. P h .D . dissertation, W ash in g to n S tate U niversity. A n n A rbor: U niversity
M icrofilm s.
Enlow , R. D . 1969. T he bone o f reptiles. In (C. G an s, ed.) Biology o f the R eptilia vol. 1,
m orphology A , pp. 45-80. L on d o n : A cadem ic Press.
Ericson, P. G . P. 1987. Interp retatio n s o f archaeological bird rem ains: a tap h o n o m ic ap p ro ach .
476 Bibliography
Journal o f Archaeological Science 14: 65-75.

E rzin?lioglu, Y. Z. 1983. T he applicatio n o f entom ology to forensic m edicine. M edicine, Science
and the Law 23:57-63.
Evans, F. G . 1961. R elation o f the physical p ro p erties o f b one to fractures. American Academ y o f
Orthopedic Surgeons 18: 110-121.
E veritt, B. S. 1977. The analysis o f contingency tables. L ondon: C h ap m an an d Hall.
F ernandez-Jalvo, Y. and A ndrew s, P. 1992. Small m am m al tap h o n o m y o f G ra n D olina,
A tap u erca (Burgos), Spain. Journal o f Archaeological Science 19: 407-428.
F inks, R. M . 1979. Fossils an d fossilization. In (R . W. F airb rid g e an d D. Jablonski, eds.)
Encyclopedia o f paleontology, pp. 327-332. S troudsburg: D ow den, H u tchin son & R oss, Inc.
Fiorillo, A. R. 1988. A p ro posal for g raphic p resen tatio n o f o rien tatio n d a ta from fossils.
University o f W yom ing C ontributions to Geology 26: 1-4. Laram ie.
1989. A n experim ental study o f tram pling: im plications fo r the fossil record. In (R.
B onnichsen and M . H . Sorg, eds.) Bone m odification, pp. 61-71. O rono: U niversity o f
M aine C enter fo r the Study o f the F irst A m ericans.
Fisher, D . C. 1981. C rocodilian scatology, m icrovertebrate concen tration s, an d enamel-less
teeth. Paleobiology 7: 262-275.
1984a. M asto d o n butchery by N o rth A m erican P aleo-Indians. N ature 308: 271-272.
1984b. T apho nom ic analysis o f late Pleistocene m asto d o n occurrences: evidence o f butchery
by N o rth A m erican Paleo-Indians. Paleobiology 10: 338-357.
1987. M a sto d o n t procurem en t by P aleoindians o f the G reat Lakes region: h u n tin g or
scavenging? In (M . H . N itecki an d D. V. N itecki, eds.) The evolution o f human hunting, pp.
309-421. New Y ork: Plenum Press.
Fisher, J. W ., Jr. 1992. O bservations on the late Pleistocene bone assem blage from the Lam b
Spring Site, C o lorad o. In (D . J. S tan fo rd an d J. S. D ay, eds.) Ice Age hunters o f the Rockies,
pp. 51-81. N iw ot: D enver M useum o f N a tu ra l H isto ry an d U niversity Press o f C olorado.
Flinn, L., T u rn er, C. G ., II, an d Brew, A. 1976. A d d itio n al evidence fo r cannibalism in the
Southw est: the case o f LA4528. American A ntiquity 41: 308-318.
Francillon-V ieillot, H ., de Buffrenil, V., C astan et, J., G erau d i, J., M eunier, F. J., Sire, J. Y.,
Zylberberg, L., an d de Ricqles, A. 1990. M icro stru ctu re an d m ineralization o f vertebrate
skeletal tissues. In (J. G . C arter, ed.) S kele ta l biomineralization: patterns, processes and
evolutionary trends, pp. 471-530. N ew Y ork: V an N o stra n d R einhold.
F reem an, L. G ., Jr. 1968. A theoretical fram ew ork for interpreting archeological m aterials, in
(R. B. Lee and I. D eV ore, eds.) M an the hunter, pp. 262-267. C hicago: Aldine.
Fritz, J. M . 1972. A rchaeological systems fo r indirect observ atio n o f the past. In (M . P. Leone,
ed.) C ontem porary archaeology, pp. 135-157. C arbondale: S ou th ern Illinois U niversity
Press.
F rison, G . C. 1974. A rchaeology o f the C asper Site. In (G . C. F riso n , ed.) T he Casper Site. a Hell
Gap bison kill on the High Plains, pp. 1-111. N ew Y ork: A cadem ic Press.
1982. Bone butchering tools in archaeological site. Canadian Journal o f Anthropology 2:
159-167.
F rison, G . C. and T o d d , L. C. 1986. The Colby M am m oth Site: taphonom y and archaeology o f a
Clovis kill in northern Wyoming. A lbuquerque: U niversity o f N ew M exico Press.
F rostick, L. an d Reid, I. 1983. T ap h o n o m ic significance o f sub-aerial tra n sp o rt o f vertebrate
fossils on steep sem i-arid slopes. Lethaia 16: 157-164.
Fiirsich, F. T. 1990. Fossil co n cen tratio n s an d life an d d eath assem blages. In (D . E. G . Briggs
an d P. R. C row ther, eds.) Palaeobiology: a synthesis, pp. 235-239. O xford: Blackwell
Scientific P ublications.
G allow ay, A.. Birkby, W. H., Jones, A. M .. H enry, T. E., and P arks, B. O. 1989. D ecay rates of
h um an rem ains in an arid environm ent. Journal o f Forensic Sciences 34: 607-616.
Bibliography All
G am ble, C. 1978. O ptim ising in fo rm atio n from studies o f faunal rem ains. In (J. F. C herry, C.
G am ble, an d S. S hennan, eds.) Sam pling in contem porary British archaeology, pp. 321-353.
British A rchaeological R ep o rts B ritish Series 50.
G arlan d , A. N . 1987. A histological study o f archaeological bone decom position. In (A.
B oddington, A. N . G arlan d , and R. C. Janaw ay, eds.) Death, decay and reconstruction:
approaches to archaeology and foren sic science, pp. 109-126. M anchester: M anchester
U niversity Press.
1988. C o n trib u tio n s to palaeohistology. In (E. A. Slater and J. O. T ate, eds.) Science and
archaeology: Glasgow 1987, pp. 321-338. British A rchaeological R ep o rts British Series
196(ii).
G arlan d , A. N ., Janaw ay, R. C., an d R oberts, C. A. 1988. A study o f th e decay processes o f
h u m an skeletal rem ains from the P arish C h u rch o f the H oly T rinity, R othw ell, N o rth a m p
tonshire. O xfo rd Journal o f Archaeology 7: 235-252.
G arvin, R. D . 1987. R esearch in Plains taphonom y: the m a n ip u latio n o f faunal assem blages by
scavengers. U npublished M .A . thesis, D ep artm en t o f A rchaeology, U niversity o f C algary,
C algary, A lberta.
Geniesse, S. C. 1982. A study o f m an u factu re an d w ear on bone artifacts. U n published M .A .
p aper, D ep artm en t o f A n thro p o lo g y , U niversity o f W ashington, Seattle.
G ibbo n, G . 1984. A nthropological archaeology. N ew Y ork: C olum bia U niversity Press.
G ibert, J. an d Jim enez, C. 1991. Investigations in to cu t-m ark s o n fossil bones o f L ow er
Pleistocene age from V enta M icena (O rce, G ra n a d a province, Spain). H um an Evolution 6:
117-128.
G ifford, D . P. 1977. O bservations o f m o d ern h u m an settlem ents as an aid to archaeological
in terp retatio n . P h .D . dissertation, U niversity o f C alifornia-B erkeley. A n n A rb o r: U niver
sity M icrofilm s Internatio n al.
1981. T ap h o n o m y and paleoecology: a critical review o f arch aeo lo g y ’s sister disciplines. In
(M . B. Schiffer, ed.) Advances in archaeological m ethod and theory vol. 4, pp. 365—438. N ew
Y ork: A cadem ic Press.
1984. T aph on om ic specim ens, L ake T u rk an a . N ational Geographic Research R eports 17:
419^128.
G ilford, D . P. and Behrensm eyer, A. K . 1977. O bserved fo rm atio n and burial o f a recent hum an
occupation site in K enya. Quaternary Research 8: 245-266.
G ifford, D . P., Isaac, G . L., and N elson, C. M . 1980. Evidence for p re d a tio n an d pastoralism at
P rolonged D rift: a p asto ral N eolithic site in K enya. Azania 15: 57-108.
G ifford-G onzalez, D . P. 1989a. E th n o g rap h ic analogues for in terp re tin g m odified bones: some
cases from E ast A frica. In (R . B onnichsen an d M. H . Sorg, eds.) Bone m odification, pp. 179-
246. O rono: U niversity o f M aine C enter for the Study o f the F irst A m ericans.
1989b. M od ern analogues: developing an interpretive fram ew ork. In (R . B onnichsen and M.
H . Sorg, eds.) Bone modification, pp. 43-52. O rono: U niversity o f M aine C en ter for the
1989c. S hipm an’s shaky fou n d atio n s. American Anthropologist 91: 180-186.
1991. Bones are n o t enough: analogues, know ledge, an d interpretive strategies in zooarchaeo-
logy. Journal o f Anthropological Archaeology 10: 215-254.
G ifford-G onzalez, D. P., D am rosch, D. B.. D am rosch, D. R., P ryor, J., and T hunen. R. L. 1985.
The th ird dim ension in site structure: an experim ent in tram p lin g an d vertical dispersal.
American A ntiquity 50: 803-818.
G ifford-G onzalez, D . and G arg ett, R. H . n.d. A n exercise in utility: com m ents on bone density,
in situ destruction, and utility indices. U n p ublished m anuscript.
G ilbert, A. S. 1979. U rb a n tap h o n o m y o f m am m alian rem ains from the B ronze Age o f G odin
Tepe, w estern Iran . P h.D . dissertatio n , C olu m b ia U niversity. A n n A rb o r: U niversity
478 Bibliography
M icrofilm s In ternation al.

G ilbert, A. S. and Singer, B. H. 1982. Reassessing zooarchaeological q u antification. W orld
G ilbert, B. M . 1969. Som e aspects o f d iet an d butchering techniques am ong prehistoric Indians
o f So uth D ak o ta. Plains Anthropologist 14: 277-294.
G ilbert, B. M . an d Bass, W . M . 1967. S easonal d atin g o f burials from the presence o f fly pupae.
Am erican A ntiquity 32: 534—535.
G ilbow , D. W. 1981. Inference o f h u m an activity from faunal rem ains. U n published M . A. thesis,
D ep artm en t o f A n thropo lo g y , W ash in g to n State U niversity, Pullm an.
G ilchrist, R. and M ytum , H . C. 1986. E xperim ental archaeology an d b u rn t anim al b one from
archaeological sites. Circaea 4: 29-38.
G ilm ore, R. M. 1949. T he identification an d value o f m am m al bones from archeologic
excavations. Journal o f M am m alogy 30: 163-169.
G o o d m an , N . 1967. U niform ity an d sim plicity. In (C. C. A lb ritto n , Jr., ed.) U niform ity and
sim plicity, pp. 93-99. G eological Society o f A m erica Special P ap er N o . 89.
G o rd o n . B. C. 1976. A ntler pseudo-tools m ade by caribou. In (J. S. R ay m o n d , B. Loveseth, C.
A rnold, an d G . R eardo n , eds.) Prim itive art and technology, pp. 121-128. C algary:
U niversity o f C algary A rchaeological A ssociation.
G o rd o n , C. C. and B uikstra, J. E. 1981. Soil pH , b one preservation, a n d sam pling bias at
m o rtu ary sites. American A ntiquity 46: 566-571.
G oss, R. J. 1983. Deer antlers: regeneration, fu n ctio n , and evolution. N ew Y ork: A cadem ic Press.
G ould, R. A. 1978. (ed.) E xplorations in ethnoarchaeology. A lbuquerque: U niversity o f New
M exico Press. 1980. Living archaeology. C am bridge: C am bridge U niversity Press. 1990.
Recovering the past. A lbuquerque: U niversity o f N ew M exico Press.
G ould, R. A. and W atson, P. J. 1982. A dialogue on the m eaning an d use o f analogy in
ethnoarchaeological reasoning. Journal o f Anthropological Archaeology 1: 355-381.
G ou ld, S. J. 1965. Is uniform itarian ism necessary? American Journal o f Science 263: 223-228.
1967. Is un ifo rm itarian ism useful? Journal o f Geofegical Education 15: 149-150.
1979. A gassiz’s m arginalia in Lyell’s Principles, o r the perils o f un iform ity and the am biguity
o f heroes. Studies in the H istory o f Biology 3: 119-138.
1982. H u tto n ’s purposeful view. N atural H istory 91(5): 6-12.
1984. T ow ard the v indication o f p u n ctu a tio n a l change. In (W. A. Berggren an d J. A. van
C ouvering, eds.) Catastrophes and earth history: the new uniform itarianism , pp. 9-34.
P rinceton: P rinceton U niversity Press.
1986. E volution an d the triu m p h o f hom ology, o r why history m atters. American Scientist 74:
60-69.
G rah am , R. W ., H olm an, J. A ., an d P arm alee, P. W. 1983. Taphonom y and paleoecology o f the
Christensen Bog m astodon bone bed, H ancock C ounty, Indiana. Illinois State M useum
R eports o f Investigations 38. Springfield.
G rah am , R. W. and K ay, M . 1988. T ap h o n o m ic com p ariso n s o f cu ltural an d n o n cu ltu ral faunal
deposits a t the K im m sw ick and B arn h art sites, Jefferson C ounty, M issouri. In (R . S. L aub,
N. G . M iller, and D. W. S teadm an, eds.) Late Pleistocene and early Holocene paleoecology
and archeology o f the Eastern Great L akes Region, pp. 227-240. Bulletin o f the Buffalo
Society o f N a tu ra l Sciences 33.
G rah am , R. W . an d O liver, J. S. 1986. T ap h o n o m y o f ice-trapped co o t p o p u latio n on Spring
Lake, Tazewell C o unty , Illinois. Am erican Quaternary Association 9th Biennial M eeting
Program and Abstracts, p. 84.
G rayso n, D. K. 1978a. M inim um num bers an d sam ple size in v erteb rate faunal analysis.
1978b. R econstructing m am m alian com m unities: a discussion o f S hotw ell’s m eth o d o f
Bibliography 479
p a le o e c o lo g ic a l a n a ly s is . Paleobiology 4: 7 7 - 8 1 .
1979. O n the qu an tification o f v erteb rate arch aeo fau n as. In (M . B. Schiffer, ed.) Advances in
archaeological m ethod and theory vol. 2, pp. 199-237. N ew Y ork: A cadem ic Press.
1981. A critical view o f the use o f archaeological vertebrates in p aleoenvironm ental recon
struction. Journal o f Ethnobiology 1: 28-38.
1983. T he paleontology o f G atecliff Shelter: sm all m am m als. In (D . H. T h o m as, ed.) The
archaeology o f M onitor Valley 2: G atecliff Shelter, pp. 99-135. A m erican M useum o f
N a tu ra l H istory A nth ro p olo g ical P apers 59(1).
1984. Q uantitative zooarchaeology: topics in the analysis o f archaeological faunas. O rlando:
A cadem ic Press.
1986. E oliths, archaeological am biguity, an d the gen eratio n o f “ m iddle-range” research. In
(D. J. M eltzer, D . D . Fow ler, an d J. A. Sabloff, eds.) Am erican archaeology:past andfuture,
pp. 77-133. W ashington: Sm ithsonian In stitu tio n Press.
1987. T he biogeographic h istory o f sm all m am m als in the G re at Basin: o bservations o n the
last 20,000 years. Journal o f M am m alogy 68: 359-375.
1988. D anger Cave, L ast Supper Cave, and H anging R ock Shelter: the fa unas. A m erican
M useum o f N a tu ra l H istory A n th ro p o lo g ical P ap ers 66(1): 1-130.
1989. B one tran sp o rt, bone destru ctio n , an d reverse utility curves. Journal o f Archaeological
Science 16: 643-652.
G reenfield, H . J. 1988. Bone co n su m p tio n by pigs in a co n tem p o rary Serbian village: im pli
cations fo r the in terp re tatio n o f prehistoric faunal assem blages. Journal o f Field Archaeo
logy 15: 4 7 3 ^ 7 9 .
G uilday, J. E ., P arm alee, P. W ., and T an n er, D . P. 1962. A boriginal bu tchering techniques at the
E schelm an site (36LA 12), L an caster C ounty, Pennsylvania. Pennsylvania Archaeologist 32:
59-83.
G uilday, J. E., H am ilton, H. W ., A nderson, E., an d P arm alee, P. W . 1978. The Baker B lu ff Cave
deposit, Tennessee, and the late P leistocenefaunal gradient. Bulletin o f the C arnegie M useum
o f N a tu ra l H istory 11. P ittsburgh.
G ustafson , C. E., G ilbow , D ., and D au g h erty , R. D . 1979. The M anis M asto d o n Site: early m an
on the O lym pic Peninsula. Canadian Journal o f Anthropology 3: 157-164.
^-G uthrie, R. D . 1967. D ifferential preserv atio n an d recovery o f Pleistocene large m am m al
rem ains in A laska. Journal o f Paleontology 41: 243-246.
1990. Frozen fa u n a o f the M am m oth Steppe: the story o f Blue Babe. C hicago: U niversity o f
C hicago Press.
H ackett, C. J. 1981. M icroscopical focal destru ctio n (tunnels) in exhum ed h u m an bones.
M edicine, Science and the Law 21: 243-265.
H aglund, W . D. 1991. A pplications o f ta p h o n o m ic m odels to forensic investigations. P h.D .
d issertation . U niversity o f W ashington, Seattle.
H aneberg, W . C. 1983. A parad ig m atic analysis o f D a rw in ’s use o f u n ifo rm itarian ism in The
origin o f species. The Compass o f Sigm a Gamm a Epsilon 60: 89-94.
H anson , C. B. 1980. F luvial tap h o n o m ic processes: m odels an d experim ents. In (A. K.
B ehrensm eyer and A. P. Hill, eds.) Fossils in the m aking, pp. 156-181. C hicago: U niversity
o f C hicago Press.
H anson, D . B. and B uikstra, J. E. 1987. H istom orphological a lteratio n in buried h u m an bone
from the Low er Illinois Valley: im plications for palaeodietary research. Journal o f Archaeo
logical Science 14: 549-563.
H are, P. E. 1980. O rganic geochem istry o f bone an d its relation to the survival o f bone in n atu ral
environm ents. In (A. K . B ehrensm eyer an d A. P. Hill, eds.) Fossils in the m aking, pp. 108—
219. C hicago: U niversity o f C hicago Press.
H argrave, L. L. 1970. M exican macaws: comparative osteology and survey o f remains fr o m the
480 Bibliography
Southw est. A nthro pological P apers o f the U niversity o f A rizona N o. 20. Tucson.
H aynes, G . 1980a. Evidence o f carnivore gnaw ing o n Pleistocene an d R ecent m am m alian bones.
1980b. Prey bones an d pred ato rs: p o ten tial ecologic in fo rm atio n from analysis o f b one sites.
Ossa 7: 75-97.
1982. U tilization and skeletal d isturbances o f N o rth A m erican prey carcasses. A rctic 35:
266-281.
1983a. A guide for differentiating m am m alian carnivore tax a responsible fo r gnaw d am age to
herbivore lim b bones. Paleobiology 9: 164-172.
1983b. Frequencies o f spiral and green-bone fractures on ungu late lim b bones in m odern
surface assem blages. American A ntiquity 48: 102-114.
1984. Age profiles in elep h an t and m am m o th b one assem blages. Quaternary Research 24:
333-345.
1987. P roboscidean die-offs an d die-outs: age profiles in fossil collections. Journal o f
Archaeological Science 14: 659-668.
1988a. L ongitudinal studies o f A frican elep h an t d eath and bone deposits. Journal o f
1988b. M ass d eaths an d serial p redation: co m parative tap h o n o m ic studies o f m o d ern large
m am m al death sites. Journal o f Archaeological Science 15: 219-235.
1991. M am m oths, mastodons, and elephants: biology, behavior, and the fo s sil record. C am
bridge: C am bridge U niversity Press.
H aynes, G . and S tanford, D . 1984. O n the possible utilization o f Camelops by E arly M an in
N o rth A m erica. Q uaternary Research 22: 216-230.
H efti, E., Trechsel, U ., R tifenacht, H ., an d Fleisch, H . 1980. U se o f derm estid beetles fo r cleaning
bones. Calcified Tissue International 31: 45-47.
H enderson, J. 1987. F a cto rs determ ining the state o f p reservation o f h u m an rem ains. In (A.
B oddington, A. N . G arla n d , an d R. C. Jan aw ay , eds.) D eath, decay and reconstruction:
approaches to archaeology and foren sic science, pp. 43-54. M anchester: M anchester
U niversity Press.
Hesse, B. and W apnish, P. 1985. A nim al bone archeology: fr o m objectives to analysis. M an u als on
A rcheology N o. 5. W ashington: T araxacum .
H ew son, R. and K olb, H . H . 1976. Scavenging o n sheep carcases by foxes ( Vulpes vulpes) and
badgers (M eles meles). Journal o f Zoology 180: 496-498.
H ildebrand, M . 1968. A natom ical preparations. Berkeley: U niversity o f C alifornia Press.
1974. Analysis o f vertebrate structure. N ew Y ork: Wiley.
Hill, A. 1976. O n carnivore an d w eathering dam age to bone. Current Anthropology 17: 335-336.
1978. T ap ho no m ical b ack g ro u n d to fossil m an - problem s in palaeoecology. In (W. W.
Bishop, ed.) Geological background to fo s sil man, pp. 87-101. E dinburgh: Scottish A cadem ic
Press, L td.
1979a. B utchery and n atu ra l disarticulation: an investigatory technique. American A ntiquity
44: 739-744.
1979b. D isarticu latio n and scattering o f m am m al skeletons. Paleobiology 5: 261-274.
1980. E arly po stm ortem dam age to the rem ains o f som e co n tem p o rary east A frican m am m als.
In (A. K . B ehrensm eyer an d A. P. Hill, eds.) Fossils in the M aking, pp. 131-152. Chicago:
U niversity o f Chicago Press.
1988. T ap ho no m y. In (I. T attersall, E. D elson, an d J. van C ouvering, eds.) Encyclopedia o f
human evolution and prehistory, pp. 562-566. N ew Y o rk : G arlan d Publishing.
Hill, A. an d Behrensm eyer, A. K . 1984. D isarticu latio n p attern s o f som e m o d ern E ast A frican
m am m als. Paleobiology 10: 366-376.
1985. N a tu ra l d isarticu lation and bison butchery. American A ntiquity 50: 141-145.
Hill, A. an d W alker, A. 1972. Procedures in v erteb rate tap h o n o m y : notes on a U g an d a M iocene
Bibliography 481
fossil locality. Journal o f the Geological Society o f London 128: 3 9 9 ^ 0 6 .

H iscock, P. 1990. A study in scarlet: tap h o n o m y o f inorganic artifacts. In (S. Solom on, I.
D avidson, and D . W atson, eds.) Problem solving in taphonom y, pp. 34-49. T em pus vol. 2.
H ockett, B. S. 1989a. A rchaeological significance o f ra b b it-ra p to r interactions in southern
C alifornia. North Am erican Archaeologist 10: 123-139.
1989b. The concept o f “ carrying range:” a m ethod fo r determ ining the role played by
w o od rats in co n trib u tin g bones to archaeological sites. N evada Archaeologist 7: 28-35.
1991. T o w ard distinguishing h um an an d ra p to r p attern in g on leporid bones. American
A ntiquity 56: 667-679.
H odder, I. 1982. The present past: an introduction to anthropology fo r archaeologists. L ondon: B.
T. B atsford. Ltd.
H offm an, R. 1988. T he co n trib u tio n o f ra p to rial birds to p attern in g in sm all m am m al
assem blages. Paleobiology 14: 81-90.
H offm an, R. an d H ays, C. 1987. The eastern w ood ra t (N eotom a floridana) as a tap h o n o m ic
facto r in archaeological sites. Journal o f Archaeological Science 14: 325-337.
H ofm an, J. L. 1981. T he refitting o f chipped-stone artifacts as an analytical and interpretive tool.
Current A nthropology 22: 691-693.
1986. V ertical m ovem ent o f artifacts in alluvial an d stratified deposits. Current Anthropology
27: 163-171.
H oltzm an, R. C. 1979. M axim um likelihood estim ation o f fossil assem blage com position.
H ooykaas, R. 1970. Catastrophism in geology, its scientific character in relation to actualism and
uniform itarianism . A m sterdam : N o rth H o llan d Publishing C om pany.
H orw itz, L. K . and Sm ith, P. 1988. T he effects o f striped hyaena activity on h u m an rem ains.
Journal o f Archaeological Science 15: 471—481.
1990. A rad iographic study o f the exten t o f v ariatio n in cortical bone thickness in Soay sheep.
H ow ard. H . 1929. T he avifau na o f Em eryville Shellm ound. University o f California Publications
in Zoology 32: 301-394. Berkeley.
H u b b ert, M. K . 1967. C ritique o f the principle o f uniform ity. In (C . C. A lb ritto n , ed.) U niformity
and sim plicity, pp. 3-33. G eological Society o f A m erica Special P ap er N o. 89.
H udson, J. L. 1990. A dvancing m eth o d s in zooarchaeology: an ethnoarchaeological study
am ong the A ka. P h.D . dissertation, U niversity o f C alifornia, S an ta B arbara.
1991. N onselective sm all gam e hu n tin g strategies: an ethnoarchaeological study o f A ka
Pygmy sites. In (M . C. Stiner, ed.) H uman predators and p rey m ortality, pp. 105-120.
B oulder: W estview Press.
1993. (ed.) From bones to behavior: ethnoarchaeological and experim ental contributions to the
interpretation o f faunal remains. C enter for A rchaeological Investigations O ccasional P aper
N o. 21. C arb o ndale: S outh ern Illinois U niversity.
H uelsbeck, D . R. 1989. Z oo archaeological m easures revisited. H istorical Archaeology 23:
113-117.
1991. F au n al rem ains an d consum er behavior: w hat is being m easured? H istorical Archaeo
logy 25: 62-76.
H ughes, A. R. 1954. H yaenas versus au stralopithecines as agents o f bone accum ulation.
Am erican Journal o f Physical Anthropology 12: 467-486.
H ughes, S. S. 1986. A m o d ern analo g to a bison ju m p . The W yom ing Archaeologist 29: 45-67.
H ulbert, R. C ., Jr. 1982. P o p u latio n dynam ics o f the three-toed horse Neohipparion from the
L ate M iocene o f Flo rid a. Paleobiology 8: 159-167.
H u tto n , J. 1795. Theory o f the earth (2 vols.). L ondon: C adell, Ju n io r an d D avies; E dinburgh:
W illiam Creech.
Irving, W. N . and H a ringto n , C. R. 1973. U p p er Pleistocene rad io carb o n -d ated artifa cts from
482 B ibliography
the n o rth ern Y ukon. Science 179: 335-340.

Irving, W. N ., Jopling, A. V., an d K ritsch -A rm stro n g , J. 1989. Studies o f b one technology and
taphonom y, O ld C row Basin, Y u k o n T erritory. In (R. B onnichsen an d M . H . Sorg, eds.)
Bone m odification, pp. 347-379. O rono: U niversity o f M aine C en ter fo r the Study o f the
F irst A m ericans.
Isaac, G . L. 1967. T o w ard the in terp re tatio n o f o ccupational debris: som e experim ents and
observations. Kroeber Anthropological Society Papers 37: 31-57.
Jam es, S. R. 1989. H om inid use o f fire in the L ow er an d M iddle Pleistocene. Current
A nthropology 30: 1-26.
Janaw ay, R. C. 1990. E xperim ental investigations o f th e burial en vironm ent o f inh u m atio n
graves. In (D. E. R obinson, ed.) Experim entation and reconstruction in environmental
archaeology, pp. 147-149. O xford: O xbow Books.
Jodry, M. A. and S tanford, D. J. 1992. S tew art's C attle G u a rd site: an analysis o f bison rem ains
in a F olsom kill-butchery cam psite. In (D . J. S tan fo rd and J. S. D ay, eds.) Ice Age hunters o f
the R ockies, pp. 101-168. N iw ot: D enver M useum o f N atu ra l H istory an d U niversity Press
o f C olorado .
Jo hn son, D . 1975. Seasonal an d m icroserai v ariatio n in the insect p o p u la tio n o n carrion.
American M idland Naturalist 93: 79-90.
Jo hn son, D . L. and H aynes, C. V. 1985. C am els as tap h o n o m ic agents. Quarternary Research 24:
365-366.
Jo hnson, E. 1982. P aleoindian bone expediency tools: L u b b o ck L ake an d Bonfire Shelter.
Canadian Journal o f Anthropology 2: 145-157.
1983. A fram ew ork fo r in terp re tatio n in bone technology. In (G . M . L eM oine and A. S.
M acE achern, eds.) Carnivores, human scavengers & predators: a question o f bone technology,
pp. 55-93. Calgary: U niversity o f C algary A rchaeological A ssociation.
1985. C u rren t developm ents in b one technology. In (M . B. Schiffer, ed.) Advances in
archaeological m ethod and theory vol. 8, pp. 157-235. N ew Y ork: A cadem ic Press.
1987. C u ltu ral activities and interactions. In Lubbock L ake: L a te Quaternary studies on the
Southern High Plains (E. Jo h n so n , ed.), pp. 120-158. College S tation: Texas A&M
U niversity Press.
1989. H u m an m odified bones from early so u th ern P lains Sites. In (R . B onnichsen an d M . H.
Sorg, eds.) Bone m odification, pp. 431-471. O rono: U niversity o f M aine C en ter for the
Joly, N . 1887. M an before m etals. L ondon: K egan Paul, Trench.
Jones, A . K . G . 1984. Som e effects o f the m am m alian digestive system on fish bones. In (N.
D esse-Berset, ed.) 2nd fish osteoarchaeology m eeting, C N R S , pp. 61-65. C en ter de
recherches archeologiques. N otes et M o nographies T echniques 16.
1986. Fish bone survival in the digestive system s o f the pig, dog and m an: som e experim ents. In
(D. C. B rinkhuizen an d A. T. C lason, eds.) Fish and archaeology, pp. 53-61. British
A rchaeological R ep o rts In tern atio n al Series 294.
1990. E xperim ents w ith fish bones an d otoliths: im plications for the reco n stru ctio n o f p ast diet
and econom y. In (D . E. R obin so n , ed.) Experim entation and reconstruction in environmental
archaeology, pp. 143-146. O xford: O xbow Books.
Jones, K . T. an d M etcalfe, D. 1988. B are bones archaeology: b one m arro w indices and
efficiency. Journal o f Archaeological Science 15: 415—423.
K ay, R. F. 1988. Teeth. In (I. T attersall, E. D elson, an d J. van C ouvering, eds.) Encyclopedia o f
human evolution and prehistory, pp. 571-578. N ew Y ork: G arla n d Publishing.
K eepax, C. A. 1981. A vian egg-shell from archaeological sites. Journal o f Archaeological Science
8: 315-335.
K ehoe, T. F . a n d K ehoe. A. B. 1960. O bservations o n the b utchering technique a t a prehistoric
bison-kill in M o n tan a . American A ntiquity 25: 420-423.
Bibliography 483
Kelley, J. H . an d H anen, M . P. 1988. Archaeology and the m ethodology o f science. A lbuquerque:

U niversity o f N ew M exico Press.
K ent, S. 1981. The dog: an archaeo lo g ist’s best friend o r w o rst enem y - the spatial d istrib u tio n o f
faunal rem ains. Journal o f Field Archaeology 8: 367-372.
K erbis-Peterhans, J. C. an d H orw itz, L. K . 1992. A b o n e assem blage from a striped hyaena
(H yaena hyaena) den the in the Negev D esert, Israel. Israel Journal o f Zoology 37: 225-245.
K idw ell, S. M . 1985. P aleobiological an d sedim entological im plications o f fossil concentrations.
N ature 318: 457-460.
1986. M odels for fossil concentrations: paleobiologic im plications. Paleobiology 12: 6-24.
K idwell, S. M . an d Bosence, D . W . J. 1991. T ap h o n o m y an d tim e-averaging o f m arine shelly
faunas. In (P. A. A llison and D. E. G . Briggs, eds.) Taphonomy: releasing the data locked in
the fo s s il record, pp. 115-209. Topics in G eobiology vol. 9. N ew Y ork: Plenum Press.
K ing, F. B. an d G rah a m , R. W. 1981. Effects o f ecological an d paleoecological p attern s on
subsistence and paleoenvironm ental reconstructions. American A ntiquity 46: 128-142.
Kiszely, I. 1973. D erivatographic exam in atio n o f subfossil an d fossil bones. Current Anthropo
logy 14: 280-286.
K itching, J. M . 1980. O n som e fossil A rth ro p o d a from the L im ew orks, M ak ap an sg at,
Potgietersrus. Palaeontologia A fricana 23: 63-68.
K itts, D. B. 1977. The structure o f geology. D allas: S ou th ern M eth o d ist U niversity Press.
Klein, R. G . 1975. P aleoanthrop o lo g ical im plications o f th e nonarcheological bone assem blage
from Sw artklip I, south-w estern C ape Province, S outh A frica. Quaternary Research 5:
275-288.
1976. T he m am m alian fau n a o f the K lasies R iver M o u th sites, sou th ern C ape Province, South
A frica. South African Archaeological Bulletin 31: 75-98.
1982a. A ge (m ortality) profiles as a m eans o f distinguishing h u n ted species from scavenged
ones in Stone A ge archaeological sites. Paleobiology 8: 151-158.
1982b. P attern s o f ungu late m o rtality and ungu late m o rtality profiles from L angebaanw eg
(E arly Pliocene) an d E landsfontein (M iddle Pleistocene), South-w estern C ape Province,
South A frica. Annals o f the South A frican M useum 90: 49-94.
1989. W hy does skeletal p a rt representation differ betw een sm aller an d larger bovids at K lasies
R iver M o u th and o th er archeological sites? Journal o f Archaeological Science 16: 363-381.
K lein, R. G . and C ruz-U ribe, K . 1984. The analysis o f anim al bones fr o m archeological sites.
C hicago: U niversity C hicago Press.
K lippel, W . E., Snyder, L. M ., an d Parm alee, P. W. 1987. T ap h o n o m y an d archaeologically
recovered m am m al bone from so u th east M issouri. Journal o f Ethnobiology 7: 155-169.
K night, J. A. 1985. D ifferential preservation o f calcined bone at the H iru n d o Site, A lto n , M aine.
U npublished M .S. thesis, Q u atern ary Science, U niversity o f M aine at O rono.
K och, C. P. 1989. Taphonom y: a bibliographic guide to the literature. O rono: U niversity o f M aine
C enter for the Study o f the F irst A m ericans.
K oike, H . an d O htaishi, N . 1985. P rehistoric hu n tin g p ressure estim ated by the age com position
o f excavated sika deer (Cervus nippon) using the an n u al layer o f to o th cem ent. Journal o f
Archaeological Science 12: 443^156.
K ooym an, B. 1984. M o a utilisation a t Ow ens F erry, O tago, N ew Z ealand. N ew Z ealand Journal
o f Archaeology 6: 47-57.
1990. M o a procurem ent: com m unal o r individual hunting? In (L. B. D avis an d B. O. K.
Reeves, eds.) H unters o f the recent past, pp. 327-351. L ondon: U nw in H ym an.
K o rth , W . W. 1979. T ap h o n o m y o f m icro v erteb rate fossil assem blages. Annals o f the Carnegie
M useum 48: 235-285.
K o rth , W . W. and E vander, R. L. 1986. T he use o f age-frequency distrib u tio n s o f m icro m am
m als in the d eterm inatio n o f a ttritio n al an d c atastro p h ic m o rtality o f fossil assem blages.
Palaeogeography, Palaeoclim atology, Palaeoecology 52: 227-236.
484 Bibliography
K ram er, C. 1979. (ed.) Ethnoarchaeology: implications o f ethnography fo r archaeology. N ew

Y ork: C olum bia U niversity Press.
K ranz, P. M . 1974a. C o m p u ter sim ulation o f fossil assem blage fo rm atio n u n d er cond itio n s o f
anastrop hic burial. Journal o f Paleontology 48: 800-808.
1974b. T he an astro p h ic bu rial o f bivalves an d its paleoecological significance. Journal o f
Geology 82: 237-266.
K rausm an , P. R. and B issonette, J. A. 1977. Bone-chew ing b ehavior o f desert m ule deer.
Southwestern N aturalist 22: 149-150.
K reutzer, L. A. 1988. M egafaunal b u tchering a t L u b b o ck L ake, Texas: a tap h o n o m ic reanalysis.
Quaternary Research 30: 221-231.
1992. Bison an d deer bone m ineral densities: com parisons an d im plications for the in terp re
tatio n o f archaeological faunas. Journal o f Archaeological Science 19: 271-294.
K rum bein, W. C. 1965. S am pling in paleontology. In (B. K um m el an d D . R au p , eds.) H andbook
o f paleontological techniques, pp. 137-150. San F rancisco: W . H . F reem an & Co.
K uhn , T. S. 1970. The structure o f scientific revolutions, 2nd edition. C hicago: U niversity o f
C hicago Press.
K um m el, B. and R au p , D . 1965. (eds.) H andbook o f paleontological techniques. San Francisco:
W. H. Freem an & Co.
K urten, B. 1953. O n the v ariatio n an d p o p u la tio n dynam ics o f fossil an d recent m am m al
populatio ns. A cta Zoologica Fennica 76: 1-122.
1958. Life an d d eath o f the Pleistocene cave bear. A cta Zoologica Fennica 95: 1-59.
1983. V ariatio n and dynam ics o f a fossil antelo p e p o p u latio n . Paleobiology 9: 62-69.
K usm er, K . D . 1986. M icroverteb rate tap h o n o m y in archaeological sites: an exam ination o f owl
deposition and the tap h o n o m y o f sm all m am m als fro m Sentinel C ave, O regon. U n p u b
lished M . A. thesis, D ep a rtm en t o f A rchaeology, Sim on F raser U niversity, B urnaby, British
C olum bia.
1990. T ap h o n o m y o f owl pellet deposition. Journal o f Paleontology 64: 629-637.
L aB arbera, M . 1989. A nalyzing body size as a facto r in ecology an d evolution. A nnual Review o f
Ecology and System atics 20: 97-117.
L am , Y. M . 1992. V ariability in the beh av io u r o f spo tted hyaenas as tap h o n o m ic agents. Journal
o f Archaeological Science 19: 389^406.
L andals, A. 1990. The M aple L e a f site: im plications o f the analysis o f sm all-scale bison kills, in
(L. B. D avis an d B. O. K . Reeves, eds.) H unters o f the recent past, pp. 122-151. L ondon:
U nw in H ym an.
Lange, F. W. 1980. P rehistory an d h u n ter/g ath erers: the role o f analogs. M id-C ontinental
Journal o f Archaeology 5: 133-147.
L artet, E. 1860. O n the coexistence o f m an w ith certain extinct qu ad ru p ed s, proved by fossil
bones, from various Pleistocene deposits, bearing incisions m ade by sh arp instrum ents. M .
G. S. Q uarterly Journal o f the Geological S ociety o f London 16:471^179. (English tran slatio n
published in 1969 [R. F. H eizer, ed.] M a n ’s discovery o f his past, pp. 122-131, P alo Alto:
Peek Publications)
Law rence, D . R. 1968. T ap h o n o m y an d in fo rm atio n losses in fossil com m unities. Geological
Society o f Am erica Bulletin 79: 1315-1330.
1971. T he n atu re an d stru ctu re o f paleoecology. Journal o f Paleontology 45: 593-607.
1979a. B iostratinom y. In (R . W . F airb rid g e an d D . Jablonski, eds.) Encyclopedia o f
paleontology , pp. 99-102. S troudsburg: D ow den, H u tch in so n & R oss, Inc.
1979b. D iagenesis o f fossils - fossildiagenese. In (R . W. F airb rid g e an d D . Jablonski, eds.)
Encyclopedia o f paleontology, pp. 245-247. S troudsburg: D ow den, H u tch in so n & R oss, Inc.
1979c. T aphon om y . In (R . W . F airb rid g e an d D . Jablonski, eds.) Encyclopedia o f paleonto-
logy, pp. 793-799. S troudsburg: D ow den, H u tch in so n & R oss, Inc.
L eM oine, G . M . and M aeE achern , A . S. 1983. (eds.). Carnivores, human scavengers <£predators:
Bibliography 485
a question o f bone technology. C algary: U niversity o f C algary A rchaeological A ssociation.

Levine, M . A. 1983. M o rtality m odels an d the in terp re tatio n o f horse p o p u la tio n structure. In
(G. Bailey, ed.) H unter-gatherer econom y in prehistory: a European perspective, pp. 23—46.
L ondon: C am bridge U niversity Press.
Linz, G . M ., Davis, J. E., Jr., E ngem an, R. M .. O tis, D. L., and Avery, M. L. 1991. E stim ating
survival o f bird carcasses in cattail m arshes. W ildlife Society Bulletin 19: 195-199.
L ivingston, S. D. 1988. T he avian an d m am m alian fau n as from L ovelock C ave and the
H u m b o ld t L akebed Site. P h .D . dissertation, U niversity o f W ashington. A n n A rbor:
U niversity M icrofilm s.
1989. T he tap h o n o m ic in terp retatio n o f avian skeletal p a rt frequencies. Journal o f Archaeolo
gical Science 16: 537-547.
Lowe, V. P. W . 1980. V ariatio n in the digestion o f prey by the taw ny owl (S trix aluco). Journal o f
Zoology 192: 283-293.
Lucas, J. and Prevot, L. E. 1991. P h o sp h ates an d fossil preservation. In (P. A. A llison an d D. E.
G . Briggs, eds.) Taphonomy: releasing the data locked in the fo s sil record, pp. 389-409.
T opics in G eobiology vol. 9. N ew Y ork: P lenum Press.
L ym an, R. L. 1978. P rehistoric b u tchering techniques in the L ow er G ran ite Reservoir,
southeastern W ashington. Tebiwa, Miscellaneous Papers o f the Idaho S ta te University
M useum o f N atural H istory 13. Pocatello.
1979a. A rchaeologicalfaunal analysis: a bibliography. O ccasional P apers o f the Id ah o M useum
o f N atu ral H istory N o. 31. Pocatello.
1979b. A vailable m eat from fau n al rem ains: a co n sid eratio n o f techniques. American
A ntiquity 44: 536-546.
1979c. F au n al analysis: an outline o f m ethod and theory w ith som e suggestions. N orthw est
Anthropological Research N otes 13: 22-35.
1980. Inferences from bone distrib u tio n s in prehistoric sites in the L ow er G ra n ite R eservoir
area, so uth eastern W ashington. N orthw est Anthropological Research N otes 14: 107-123.
1982a. A rchaeofau n as and subsistence studies. In (M . B. Schilfer, ed.) Advances in archaeolo
gical m ethod and theory vol. 5, pp. 331-393. N ew Y ork: A cadem ic Press.
1982b. The tap h o n o m y o f vertebrate arch aeo fau n as: b one density and differential survivor
ship o f fossil classes. P h.D . dissertation, U niversity o f W ashington, Seattle. A n n A rbor:
U niversity M icrofilm s In tern atio n al.
1984a. Bone density an d differential survivorship o f fossil classes. Journal o f Anthropological
1984b. B roken bones, bone expediency tools, an d bone pseudotools: lessons from the blast
zone aro u n d M o u n t St. H elens, W ashington. Am erican A ntiquity 49: 315-333.
1985a. Bone frequencies: differential tra n sp o rt, in situ destru ction , an d the M G U I. Journal o f
1985b. The paleozoology o f the A vey’s O rch ard Site. In (J. R. G alm an d R. A. M asten, eds.)
A v e y ’s Orchard: Archaeological Investigations o f a L a te Prehistoric Columbia River
C om m unity, pp. 243-319. E astern W ash in g to n U niversity R ep o rts in A rchaeology and
H istory 100-42. C heney.
1987a. A rchaeofau nas and butchery studies: a tap h o n o m ic perspective. In (M . B. Schiffer, ed.)
Advances in archaeological m ethod and theory vol. 10, pp. 249-337. San Diego: A cadem ic
Press.
1987b. H u n tin g fo r evidence o f Plio-Pleistocene h om inid scavengers. A m erican Anthropologist
89: 710-715^
1987c. O n the analysis o f vertebrate m o rtality profiles: sam ple size, m o rtality type, and
hunting pressure. American A ntiquity 52: 125-142.
1987d. O n zooarchaeological m easures o f socioeconom ic position an d cost-efficient m eat
purchases. H istorical Archaeology 21: 58-66.
486 Bibliography
1987e. Z ooarchaeo logy an d taphonom y: a general consideration. Journal o f Ethnobiology 7:

93-117.
1988a. W as there a last sup p er a t L ast Supper Cave? In (D . K . G ray so n , ed.) D anger Cave, L ast
Supper Cave, and H anging Rock Shelter: the fa unas, pp. 81-104. A m erican M useum o f
N atu ral H istory A nth ro polo g ical Papers 66(1).
1988b Z ooarchaeology o f 4 5D 0 1 8 9 . In (J. R. G alm a n d R. L. L ym an, eds.) Archaeological
investigations at River M ile 590: the excavations at 45D O 189, pp. 97-141. E astern
W ashin gton U niversity R ep o rts in A rchaeology an d H istory 100-61. C heney.
1989a. Seal an d sea lion hunting: a zooarchaeological study from the so u th ern N orth w est
C oast o f N o rth A m erica. Journal o f Anthropological Archaeology 8: 68-99.
1989b. T ap h o n o m y o f cervids killed by the 18 M ay 1980 volcanic eru p tio n o f M o u n t St.
H elens, W ashington. In (R . B onnichsen an d M . Sorg, eds.) Bone m odification, pp. 149-167.
O rono: C enter fo r the Study o f the F irst A m ericans, U niversity o f M aine.
1991 a. Prehistory o f the Oregon coast: the effects o f excavation strategies and assemblage size on
archaeological inquiry. San Diego: A cadem ic Press.
1991b. Subsistence change and pinniped hunting. In (M . C. Stiner, ed.) H uman predators and
prey m ortality, pp. 187-199. B oulder: W estview Press.
1991c. T aph on om ic problem s w ith archaeological analysis o f anim al carcass utilization and
tran sp o rt. In (J. R. P urdu e, W. E. K lippel, an d B. W . Styles, eds.) Beamers, bobwhites, and
blue-points: tributes to the career o f Paul W. Parmalee, pp. 125-138. Illinois S tate M useum
Scientific P apers vol. 23. Springfield.
1992a. A natom ical consideratio n s o f utility curves in zooarchaeology. Journal o f Archaeologi
cal Science 19: 7-22.
1992b. Influences o f m id-H olocene A ltitherm al clim ates o n m am m alian faunas an d hum an
subsistence in eastern W ashington. Journal o f Ethnobiology 12: 37-62.
1992c. P rehistoric seal a n d sea-lion b u tchering on the so u th ern N o rth w est C oast. American
A ntiquity 57: 246-261.
1993a. A study o f v ariation in the p rehistoric b utchery o f large artiodactyls. In (E. Jo h n so n ,
ed.) Ancient peoples and landscapes (in press). L ubbock: Texas Tech U niversity Press.
1993b. D en sity -m ed iated attritio n o f b one assem blages: new insights. In (J. H u d so n , ed.)
From bones to behavior, pp. 324-341. C en ter for A rchaeological Investigations O ccasional
P ap er N o. 21. C arbo n dale: Sou th ern Illinois U niversity Press.
n.d.a. Z ooarchaeology o f 45C H 302. In (J. R. G alm , ed.) Archaeological investigations at
45C H 302, E astern W ashington U niversity R ep o rts in A rchaeology and H istory. C heney (in
press).
n.d.b. Z ooarchaeology o f 45G R 445. In (J. R. G alm , ed.) The archeology o f Salishan M esa,
E astern W ashington U niversity R ep o rts in A rchaeology an d H istory. C heney (in press).
L ym an, R. L. an d Fox, G . L. 1989. A critical ev aluation o f bone w eathering as an indication o f
bone assem blage fo rm ation. Journal o f Archaeological Science 16: 293-317.
L ym an, R. L. and O ’Brien, M . J. 1987. Plow -zone zooarchaeology: frag m en tatio n and
identifiability. Journal o f Field Archaeology 14: 493^498.
Lym an, R. L., H o u g h to n , L. E., an d C ham bers, A. L. 1992a. T he effect o f stru ctu ral density on
m arm o t skeletal p a rt rep resen tation in archaeological sites. Journal o f Archaeological
Science 19: 557-573.
L ym an, R. L., Savelle, J. M ., an d W hitridge, P. 1992b. D erivation an d applicatio n o f a food
utility index for P hocid seals. Journal o f Archaeological Science 19: 531-555.
L yon, P. J. 1970. D ifferential bone destruction: an eth n o g rap h ic exam ple. American A ntiq u ity 35:
213-215.
M acG rego r, A. 1985. Bone, antler, ivory and horn: the technology o f skeletal materials since the
Rom an Period. L ondon: C ro o m Helm.
Bibliography 487
M ackin, J. H . 1963. R atio n al and em pirical m eth o d s o f investigation in geology. In (C. C.

A lb ritto n , Jr., ed.) The fa b ric o f geology, pp. 135-163. R eading, M A .: A ddison-W esley
Publishing Co.
M aguire, J. M ., P em berton, D ., an d C ollett, M. H . 1980. T he M ak a p a n sg a t L im ew orks G rey
Breccia: hom inids, hyaenas, hystricids o r hillw ash? Paleontologia Africana 23: 75-98.
M altby, J. M . 1985a. Assessing variatio n s in Iro n A ge an d R o m an butchery practices: the need
for q uantification. In (N . R. J. Fieller, D. D . G ilb ertso n , an d N . G . A. R alp h , eds.)
Paleobiological investigations: research design, m ethods and data analysis, pp. 19-30. British
A rchaeological R ep o rts In tern atio n al Series 266.
1985b. P attern s in faunal assem blage variability. In (G . B arker an d C. G am ble, eds.) Beyond
dom estication in prehistoric Europe, pp. 33-74. L ondon: A cadem ic Press.
M arch iafava, V., Bonucci, E., and A scenzi, A. 1974. F u n g al osteoclasia: a m odel o f dead bone
resorption. Calcified Tissue Research 14: 195-210.
M arean, C. W . 1991. M easuring the po st-d ep o sitio n al destru ctio n o f b one in archaeological
assem blages. Journal o f Archaeological Science 18: 677-694.
M arean, C. W . and Spencer, L. M . 1991. Im p act o f carnivore ravaging on zooarchaeological
m easures o f elem ent abundance. Am erican A ntiquity 56: 645-658.
M arean , C. W ., Spencer, L. M ., B lum enschine, R. J., an d C ap ald o , S. D . 1992. C aptive hyaena
bone choice an d destructio n , the schlepp effect an d O lduvai arch aeo fau n as. Journal o f
M arshall, F. 1986. Im plications o f b one m odification in a N eolithic faunal assem blage for the
study o f early hom inid b utchery an d subsistence practices. Journal o f H um an Evolution 15:
661-672.
1993. F o o d sharing and body p a rt rep resen tatio n in O kiek faunal assem blages. Journal o f
Archaeological Science 20 (in press).
M arshall, F. and Pilgram , T. 1991. M eat versus w ithin-bone nutrients: a n o th e r lo o k at the
m eaning o f body p a rt represen tatio n in archaeological sites. Journal o f Archaeological
Science 18: 149-163.
M arshall, L. G . 1989. Bone m odification an d “ the laws o f b u rial.” In (R. B onnichsen an d M . H.
Sorg) Bone modification, pp. 7-24. O rono: U niversity o f M aine C enter for the Study o f the
F irst A m ericans.
M artill, D . M . 1990. Bones as stones: the co n trib u tio n o f vertebrate rem ains to the lithologic
record. In (S. K . D o nov an , ed.) The processes o f fossilization, pp. 270-292. N ew Y ork:
C olum b ia U niversity Press.
M artin , H . 1910. L a percussion osseuse et les esquilles qui en derivent ex perim entation. Societe
Prehistorique Franqaise Bulletin 7: 299-309.
M artin, P. S. an d K lein, R. G . 1984. (eds.) Q uarternary extinctions. T ucson: U niversity o f
A rizona Press.
M atthew s, W. H.. III. 1962. Fossils. N ew Y ork: Barnes & N oble.
M cG rew , P. 0 . 1975. T ap h o n o m y o f E ocene fish from Fossil Basin, W yom ing. Fieldiana Geology
33: 257-270.
M cK enna, M . C. 1962. C ollecting sm all fossils by w ashing an d screening. Curator 5: 221-235.
M cK inney, M . L. 1990. C om pleteness o f the fossil record: an overview. In (S. K . D o n ov an , ed.)
The processes o f fossilization, pp. 66-83. N ew Y ork: C olum bia U niversity Press.
M eadow , R. H . 1978. Effects o f context o n the in terp re tatio n o f faunal rem ains: a case study. In
(R. H . M eadow an d M . A. Zeder, eds.) Approaches to fa u n a l analysis in the M iddle East, pp.
15-21. Peabody M useum Bulletin 2. C am bridge.
1981. A nim al bones - problem s for the archaeologist to g eth er w ith som e possible solutions.
Paleorient 6: 65-77.
M edlock. R. C. 1975. F au n al analysis. In (M . B. S chifferand J. H. H ouse, eds.) The Cache River
488 Bibliography
archeological project: an experim ent in contract archeology, pp. 223-242. Fayetteville:

A rkansas A rcheological Survey R esearch Series N o. 8.
M ehl, M . G . 1966. T he D om ebo m am m oth: vertebrate p aleom ortology. In (F. C. L eonhardy,
ed.) D omebo: a Paleo-Indian m am m oth kill in the Prairie-Plains, pp. 27-30. C o n trib u tio n s o f
the M useum o f the G re a t Plains 1. L aw to n , O klahom a.
M einke, D . K . 1979. Bones and teeth. In (R . W. F airb rid g e an d D . Jab lo n sk i, eds.) Encyclopedia
o f paleontology, pp. 122-131. S troudsburg: D ow den, H utch inso n & R oss, Inc.
M ellet, J. S. 1974. Scatological origins o f m icro v erteb rate fossils. Science 185: 349-350.
M eltzer, D . J., L eonard, R. D ., an d S tratto n , S. K . 1992. T he relationship betw een sam ple size
and diversity in archaeological assem blages. Journal o f Archaeological Science 19: 375-387.
M en goni-G onalons, G . L. 1991. L a llam a y sus p ro d u cto s prim arios. Arqueologia: R evista de la
Seccion Prehistoria 1: 179-196.
M etcalfe, D . an d Jones, K . T. 1988. A reconsideration o f anim al body p a rt utility indices.
M eyer, C. A. 1991. B urial experim ents w ith m arine tu rtle carcasses an d th eir paleoecological
significance. Palaios 6: 89-96.
M icozzi, M . S. 1986. E xperim ental study o f p o stm o rtem change u n d erfield conditions: effects o f
freezing, thaw ing, an d m echanical injury. Journal o f Forensic Science 31: 953-961.
1991. Postm ortem change in human and anim al remains: a system atic approach. Springfield:
C harles C. T hom as.
M iller, G . J. 1969. A study o f cuts, grooves an d o th er m ark s o n recent an d fossil bone: I. anim al
to o th m arks. Tebiwa, Journal o f the Idaho S ta te University M useum 12: 20-26. Pocatello.
1975. A study o f cuts, grooves, an d o th er m arks on recent an d fossil bones: II, w eathering
cracks, fractures, splinters, an d o th er sim ilar n atu ra l phenom ena. In (E. Sw anson, ed.)
Lithic technology, pp. 212-226. T he H ague: M o u to n .
M iller, S. J. 1989. C haracteristics o f m am m o th bone red u ctio n a t Owl C ave, the W asden Site,
Idaho. In (R . B onnichsen an d M . H . Sorg, eds.) Bone modification, pp. 381-393. O rono:
U niversity o f M aine C enter for the Study o f the F irst A m ericans.
M ilne, L. J. and M ilne, M . 1976. T he social beh av io r o f burying beetles. Scientific American
230(2): 84-89.
M odell, W. 1969. H orns an d antlers. Scientific American 220(4): 114-122.
M onks, G . G . 1981. Seasonality studies. In (M . B. Schiffer, ed.) Advances in archaeological
m ethod and theory vol. 4, pp. 177-240. N ew Y ork: A cadem ic Press.
M ontelius, O. 1888. The civilization o f Sweden in heathen times. L ondon: M acm illan.
M oore, K . L. 1985. Clinically oriented anatom y, 2nd edition. Baltim ore: W illiam s & W ilkins.
M o rlan, R. E. 1980. Taphonom y and archaeology in the upper Pleistocene o f the northern Yukon
Territory: a glimpse o f the peopling o f the N ew World. A rchaeological Survey o f C an ad a
P ap er N o. 94, M ercury Series. O ttaw a: N atio n al M useum o f M an.
1983. Spiral fractures on lim b bones: w hich ones are artificial? In (G. M . L eM oine an d A. S.
M acE achern , eds.) Carnivores, human scavengers and predators: a question o f bone techno
logy, pp. 241-269. C algary: U niversity o f C algary A rchaeological A ssociation.
1984. T ow ard the definition o f criteria fo r the recognition o f artificial bone alterations.
Q uarternary Research 22: 160-171.
1988. Pre-C lovis people: early discoveries o f A m erica? In (R . C. C arlisle, ed.) Am ericans before
Columbus: Ice-Age origins, pp. 31^43. E thn o lo g y M o n g rap h s 12. P ittsburgh: U niversity o f
P ittsb u rg h D ep a rtm e n t o f A n thropology.
M o rlot, A. 1861. G eneral views on archaeology. Sm ithsonian Institution A nnual Report fo r I860,
pp. 284-343.
M orse, D . 1983. The skeletal path o lo g y o f trau m a. In (D . M orse, J. D u n can , and J. Stoutm ire,
eds.) H andbook o f forensic archaeology and anthropology, pp. 145-185. Tallahassee:
Privately published.
Bibliography 489
M uller, A. H . 1951. Grundlagen der biostratonomie. D eutsche A kadem ie d er W issenschaften zu

Berlin, A bh an dlu n gen , 1950, N o. 3.
1963. Lehrbuch depalaozoologie, band 1. allgemeinegrundlagen. Jena: G u stav Fischer Verlag.
M undy, P. J. and Ledger, J. A. 1976. G riffon vultures, carnivores an d bones. South African
Journal o f Science 72: 106-110.
M unson, P. J. 1969. C om m ents on B inford’s “ Sm udge pits an d hide sm oking: the use o f analogy
in archaeological reason ing” . Am erican A ntiquity 34: 83-85.
M unthe, L. and M cL eod, S. A . 1975. Collection o f taphonomic inform ation fr o m fo s sil and recent
vertebrate specimens with a selected bibliography. PaleoB ios 19. C o n trib u tio n s from the
U niversity o f C alifornia-B erkeley M useum o f Paleontology.
M iinzel, S. 1986. C oding system for b one fragm ents. In (L. H . van W ijngaarden-B akker, ed.)
D atabase m anagem ent and zooarchaeology, pp. 193-195. P A C T 14.
M urray, T. an d W alker, M . J. 1988. Like W H A T ? a practical question o f analogical inference
and archaeological m eaningfulness. Journal o f Anthropological Archaeology 7: 248-287.
M yers, T ., V oorhies, M . R ., an d C orner, R. G . 1980. Spiral fractures an d bone p seudotools at
paleontological sites. Am erican A ntiquity 45: 483^489.
N airn , A. E. M . 1965. U nifo rm itarian ism an d en vironm ent. Palaeogeography, Palaeoclimato-
logy, Palaeoecology 1: 5—11.
N elson, D. E., M orlan, R. E., Vogel, J. S., S ou th o n, J. R ., and H a rin g to n , C. R. 1986. New dates
on n o rth ern Y u k o n artifacts: H olocene n o t U p p er Pleistocene. Science 232: 749-751.
N im m o, B. W . 1971. P o p u la tio n dynam ics o f a W yom ing p ro n g h o rn co h o rt from the Eden-
F arso n Site. 48SW 304. Plains Anthropologist 16: 285-288.
N itecki, M . H . an d N itecki, D . V. 1987. (eds.) The evolution o f human hunting. N ew Y ork: Plenum
Press.
N oe-N yg aard. N . 1974. M esolithic h u n tin g in D en m ark illustrated by bone injuries caused by
h u m an w eapons. Journal o f Archaeological Science 1: 217-218.
1975a. Bone injuries caused by h u m an w eapons in M esolithic D en m ark . In (A. T. C lason, ed.)
A rchaeozoological studies, pp. 151-159. A m sterdam : N o rth H o llan d Publishing Co.
1975b. T w o shoulder blades w ith healed lesions from S tar C arr. Proceedings o f the Prehistoric
Society 41: 10-16.
1977. B utchering an d m arro w fractu rin g as a tap h o n o m ic facto r in archaeological deposits.
1987. T ap h o n o m y in archaeology w ith special em phasis o n m an as a biasing factor. Journal o f
Danish Archaeology 6: 7-52.
1989. M an-m ade trace fossils on bones. H um an Evolution 4: 461—491.
N oe-N ygaard, N . an d R ichter, J. 1990. Seventeen wild b o a r m andibles from S ludegards Som ose
- offal o r sacrifice? In (D . E. R o b in so n , ed.) E xperim entation and reconstruction in
environmental archaeology, pp. 175-189. O xford: O xbow B ooks.
O ’C onnell, J. F., H aw kes, K ., an d Jones, N . B. 1988. H ad za hunting, butchering, an d bone
tra n sp o rt and their archaeological im plications. Journal o f Anthropological Research 44:
113-161.
1990. R eanalysis o f large m am m al body p a rt tra n sp o rt am ong the H adza. Journal o f
1992. P attern s in the distrib u tio n , site stru ctu re an d assem blage com position o f H ad za kill-
butchering sites. Journal o f Archaeological Science 19: 319-345.
O ’C onnell, J. F. and M arshall, B. 1989. A nalysis o f k an g aro o body p a rt tra n sp o rt am ong the
A lyaw ara o f C entral A ustralia. Journal o f Archaeological Science 16: 393-405.
O dum , E. P. 1971. Fundamentals o f ecology, 3rd edition. Philadelphia: W . B. S aunders C om pany.
O ’G ara, B. W. and M atso n , G . 1975. G ro w th an d casting o f h o rn s by p ro n g h o rn s and
exfoliation o f ho rn s by bovids. Journal o f M am m alogy 56: 829-846.
Oliver, J. S. 1989. A nalogues and site context: bone dam ages from Shield T rap C ave (24CB91),
490 Bibliography
C arb o n C ounty, M o n tan a , U .S.A . In (R . B onnichsen an d M . H . Sorg, eds.) Bone

m odification, pp. 73-98. O rono: U niversity o f M aine C enter fo r the Study o f the F irst
A m ericans.
Olsen, S. J. 1968. Fish, am ph ibian an d reptile rem ains from archaeological sites, p a rt I:
southeastern and southw estern U nited States. Papers o f the Peabody M useum o f Archaeo
logy and Ethnology 56(2). C am bridge: H arv ard U niversity.
Olsen, S. L. 1988. The identification o f stone an d m etal to o l m ark s o n bone artifacts. In (S. L.
Olsen, ed.) Scanning electron microscopy in archaeology, pp. 337-360. British A rchaeologi
cal R ep o rts In tern atio n al Series 452.
1989a. O n distinguishing n a tu ra l from cu ltu ral d am age on archaeological antler. Journal o f
1989b. Solutre: a theoretical ap p ro ach to the reco n stru ctio n o f U p p e r Paleolithic hunting
strategies. Journal o f H um an Evolution 18: 295-327.
Olsen, S. L. an d Shipm an, P. 1988. Surface m odification on bone: tram p ling versus butchery.
O lson, E. C. 1952. The evolution o f a P erm ian v erteb rate ch ro n o fau n a. Evolution 6: 181-196.
1958. F au n a o f the Vale and C hoza: 14. sum m ary, review, an d integration o f the geology and
the faunas. Fieldiana Geology 10: 397^148.
1980. T ap ho no m y: its history an d role in com m unity evolution. In (A. K . B ehrensm eyer and
A. P. Hill, eds.) Fossils in the m aking, pp. 6-19. C hicago: U niversity o f C hicago Press.
O rtn er, D. J. and P utsch ar, W . G . J. 1981. Identification o f pathological conditions in human
skeletal remains. S m ithsonian C o n trib u tio n s to A n th ro p o lo g y N o . 28.
O rtn er, D. J., V on E ndt, D . W ., and R o b in so n , M . S. 1972. T he effect o f tem p eratu re on protein
decay in bone: its significance in n itrogen d atin g o f archaeological specimens. American
A ntiquity 37: 514^520.
Park er, A. J. 1988. T he birds o f R o m an B ritain. O xfo rd Journal o f Archaeology 1: 197-226.
Parsons, M . K . and B rett, C. E. 1990. T ap ho n o m ic processes and biases in m odern m arine
environm ents: an actualistic perspective on fossil assem blage preservation. In (S. K.
D onovan , ed.) The processes o f fossilization, pp. 22-65. N ew Y ork: C olu m b ia U niversity
Press.
Pate, F. D. an d H u tto n , J. T. 1988. The use o f soil chem istry d a ta to address p ost-m ortem
diagenesis in bone m ineral. Journal o f Archaeological Science 15: 729-739.
Payne, J. A. 1965. Sum m er carrio n study o f the b aby pig Sus scrofa. Ecology 46: 592-602.
Payne, J. A ., K ing, R. E., and B einhart, G . 1968. A rth ro p o d succession an d decom position o f
buried pigs. N ature 219: 1180-1181.
Payne, S. 1972a. O n the in terp re ta tio n o f bone sam ples from archaeological sites. In (E. S. Higgs,
ed.) Papers in economic prehistory, pp. 65-81. C am bridge: C am bridge U niversity Press.
1972b. P artial recovery an d sam ple bias: the results o f som e sieving experim ents. In (E. S.
Higgs, ed.) Papers in economic prehistory, pp. 49-64. C am bridge: C am bridge U niversity
Press.
Peale, T. R. 1871. O n the uses o f the b rain an d m arro w o f anim als am ong the Indians o f N o rth
A m erica. Sm ithsonian Institution A nnual Report fo r 1870, pp. 390-391.
Pei, W .-C. 1932. Prelim inary n ote o n som e incised, cu t an d bro k en bones found in association
w ith Sinanthropus rem ains and lithic artifacts from C h o u k o u tien . Geological Society o f
China Bulletin 12: 105-108.
1938. L e role des anim aux et des causes naturelles dans la cassure des os. P alaeontologica Sinica,
Series D , N ew Series 7 (118). N anking: G eological Survey o f C hina.
Perkins, D . and D aly, P. 1968. A h u n te r’s village in N eolithic T urkey. Scientific Am erican 219(5):
96-106.
Perzigian, A . J. 1973. O steo porotic bone loss in tw o p rehistoric In d ia n p o p u latio n s. American
Journal o f P hysical Anthropology 39: 87-96.
Bibliography 491
P iepenbrink, H. 1986. T w o exam ples o f biogenous dead bone d ecom position and their
consequences for tap h o n o m ic in terp retatio n . Journal o f Archaeological Science 13:
417-430.
Plotnick, R. E. an d Speyer, S. E. 1989. (eds.) D eath, decay, disintegration: the newsletter fo r
research on taphonom y 1: 1-33. C hicago: U niversity o f Illinois.
1990. (eds.) Death, decay, disintegration: the new sletter fo r research on taphonom y 2: 1-22.
Plotnick, R. E. an d W alker, S. E. 1991. (eds.) D eath, decay, disintegration: the newsletter fo r
research on taphonom y 3: 1-18. C hicago: U niversity o f Illinois.
Plug, I. 1978. C ollecting p a tte rn s o f six species o f vultures (Aves: A ccipitridae). Annals o f the
Transvaal M useum 31(6): 51-63.
Polacheck, T. 1985. T h e sam pling d istrib u tio n o f age-specific survival estim ates from an age
distribu tion. Journal o f Wildlife M anagem ent 49: 180-184.
P otts, R. 1982. L ow er Pleistocene site fo rm atio n an d hom inid activities at O lduvai G orge,
T anzania. P h.D . dissertation, H arv a rd U niversity. A n n A rb o r: U niversity M icrofilm s.
1984. H om inid hunters? Problem s o f identifying the earliest hu n ter/g ath erers. In (R . Foley,
ed.) H om inid evolution and com m unity ecology, pp. 129-166. L ondon: A cadem ic Press.
1986. T em poral sp an o f bone accum ulations a t O lduvai G orge an d im plications for early
hom inid foraging behavior. Paleobiology 12: 25-31.
1988. E arly hom inid activities a t Olduvai. N ew Y ork: A ldine de G ruyter.
P otts, R. an d Shipm an, P. 1981. C u tm ark s m ade by stone tools o n bones from O lduvai G orge,
T an zania. N ature 291: 577-580.
P ozorski, S. G . 1979. L ate prehistoric llam a rem ains from the M oche Valley, Peru. Annals o f the
Carnegie M useum 48: 139-170.
Price, T. D ., Blitz, J., B urton, J., and Ezzo, J. A. 1992. D iagenesis in prehistoric bone: problem s
and solutions. Journal o f Archaeological Science 19: 513-529.
R ackham , J. 1983. F a u n a l sam ple to subsistence econom y. In (M . Jones, ed.) Integrating the
subsistence econom y, pp. 251-277. B ritish A rchaeological R ep o rts In te rn atio n al Series 181.
R agir, S. 1967. A review o f techniques for archaeological sam pling. In (R. F. H eizer an d J. A.
G rah am , eds.) A guide to fie ld m ethods in archaeology: approaches to the anthropology o f the
dead, pp. 181-197. P alo A lto: N atio n a l Press Books.
R apson, D. J. 1990. P a tte rn and process in intra-site spatial analysis: site stru ctu ral an d faunal
research at the B ugas-H olding Site. P h.D . dissertation, U niversity o f N ew M exico,
A lbuquerque. A nn A rb o r: U niversity M icrofilm s In tern atio nal.
R ead, C. E. 1971. A nim al bones and h u m an behavior: ap p ro ach es to faunal analysis in
archeology. P h.D . dissertation, U niversity o f C alifornia-L os A ngeles. A nn A rb o r: U niver
sity M icrofilm s In tern atio nal.
R ead-M artin , C. E. an d R ead, D . W . 1975. A ustralo p ith ecin e scavenging and h u m an evolution:
an ap p ro ach from faunal analysis. Current Anthropology 16: 359-368.
R eed, C. A. 1963. O steo-archaeology. In (D . Brothw ell an d E. Higgs, eds.) Science in
A rchaeology, pp. 204-216. N ew Y ork: Basic Books.
R eher, C. A. 1974. P o p u latio n study o f the C asper Site bison. In (G . C. F rison, ed.) The Casper
Site: A H ell Gap bison kill on the High Plains, pp. 113-124. N ew Y ork: A cadem ic Press.
R eher, C. A. and Frison, G . C. 1980. The Vore Site, 48CK302, a stratified buffalo ju m p in the
W yom ing B lack Hills. Plains A n th ro p o lo g ist M em o ir 16.
Reitz, E. J. 1986. V ertebrate fau n a from L ocus 39, P u erto R eal, H aiti. Journal o f Field
R ensberger, J. M . and K rentz, H. B. 1988. M icroscopic effects o f p re d a to r digestion on the
surfaces o f bones an d teeth. Scanning M icroscopy 2: 1541-1551.
R ensch, B. 1960. Evolution above the species level. N ew Y ork: C olu m b ia U niversity Press.
R etallack, G . 1984. C om pleteness o f the rock an d fossil record: som e estim ates using fossil soils.
492 Bibliography
1988. D ow n -to -earth approach es to v erteb rate paleontology. Palaios 3: 335-344.

1990. Soils o f the past: an introduction to paleopedology. Boston: U nw in H ym an.
R ich, P. V. 1980. Prelim inary rep o rt on the fossil avian rem ains from late T ertiary sedim ents at
L angebaanw eg (C ape Province), So u th A frica. South A frican Journal o f Science 76:
166-170.
R ichardson, P. R. K . 1980. C arn iv o re dam age to antelo p e bones an d its archaeological
im plications. Paleontologia Africana 23:109-125.
R ichardson, P. R. K ., M undy, P. J., an d Plug, I. 1986. Bone crushing carnivores and their
significance to osteo dystro p h y in griffon v ulture chicks. Journal o f Z oology 210: 23—43.
R ichter, J. 1986. E xperim ental study o f h eat induced m orph o lo g ical changes in fish bone
collagen. Journal o f Archaeological Science 13: 477-481.
R ichter, R. 1928. A ktuopalaontologie und Palaobiologie, eine Abgrenzung. S enkenbergiana 19.
R obison, N . D. 1978. Z ooarchaeology: its history an d developm ent. Tennessee Anthropological
Association Miscellaneous Paper 2: 1-22. K noxville.
R odriguez, W. C. an d Bass, W . M . 1983. Insect activity an d its relatio n sh ip to decay rates o f
hu m an cadavers in east Tennessee. Journal o f Forensic Sciences 28: 423—430.
Rogers, R. R. 1992. N o n -m arin e borings in d in o sau r bones from th e U p p e r C retaceous Tw o
M edicine F o rm atio n , no rth w estern M o n ta n a. Journal o f Vertebrate Paleontology 12:
528-531.
Rolfe, W. D . I. an d B rett, D . W . 1969. F ossilization processes. In (G . E glinton an d M . T. J.
M urph y, eds.) Organic geochem istry: m ethods and results, pp. 213-244. Berlin:
Springer-V erlag.
R om er, A. S. and P arsons, T. S. 1977. The vertebrate body, 5th edition. P hiladelphia: W . B.
S aunders C om pany.
R osene, W ., Jr. and Lay, D . W . 1963. D isappearance an d visibility o f quail rem ains. Journal o f
Wildlife M anagem ent 27: 139-142.
R udw ick, M . J. S. 1971. U n iform ity an d progression: reflections on the stru ctu re o f geological
theory in the age o f Lyell. In (D . R oller, ed.) Perspectives in the history o f science and
technology, pp. 209-227. N o rm an : U niversity o f O k lah o m a Press.
1976. The meaning o f fossils, 2nd edition. N ew Y ork: N eale W atso n A cadem ic Publications.
Russell, M . D . 1987. M o rtu a ry practices a t the K rap in a N ean d ertal site. American Journal o f
Physical Anthropology 12: 381-397.
Russell, M . P. 1992. Review o f “ T he processes o f fossilization,” edited by S. K. D ono v an .
Palaios 1: 331-332.
S adek-K ooros, H . 1972. Prim itive bone fracturing: a m eth o d o f research. American A ntiq u ity 37:
369-382.
1975. In ten tio n al fractu ring o f bone: d escription o f criteria. In (A. T. C lason, ed.) Archaeozoo-
logical studies, pp. 139-150. A m sterdam : N o rth H o llan d P ublishing Co.
Salm on, M . H . 1981. A scribing functions to archaeological objects. Philosophy o f the Social
Sciences 11: 19-26.
S alm on, W . C. 1953. The uniform ity o f n atu re. Philosophy and Phenomenological Research 14:
39-48.
1963. Logic. E nglew ood Cliffs: Prentice-H all, Inc.
1984. Scientific explanation and the causal structure o f the world. Princeton: P rinceton
U niversity Press.
Sander, P. M . 1992. T he N o rian Plateosaurus bonebeds o f central E u ro p e an d th eir tap h o n om y .
P alaeogeography, Palaeoclim atology, Palaeoecology 93: 255-299.
Sanders, H . L. 1968. M arine benthic diversity: a com p arativ e study. Am erican N aturalist 102:
243-282.
Saunders, J. J. 1977. L ate Pleistocene v ertebrates o f the w estern O zark H ighland, M issouri.
Illinois S ta te M useum Reports o f Investigations N o. 33. Springfield.
Bibliography 493
Schafer, W. 1962. Aktuo-palaontologie nach Studien in der Nordsee. F ra n k fu rt am M ain: Verlag

W aldem ar K ram er.
1972. Ecology and palaeoecology o f marine environments. C hicago: U niversity o f C hicago
Press. (English tran slatio n o f Schafer 1962 by I. O ertel, edited by C. Y. Craig).
Scharnberger, C. K ., B ushm an, J. R., an d Shea, J. H. 1983. C om m ents an d reply on ‘Twelve
fallacies o f un iform itarian ism .’ Geology 11: 312-313.
Schick, K. D ., T oth , N ., and D aeschler, E. 1989. A n early paleontological assem blage as an
archaeological test case. In (R. B onnichsen an d M . H. Sorg, eds.) Bone m odification, pp.
121-137. O rono: U niversity o f M aine C en ter for the Study o f the F irst A m ericans.
Schiffer, M. B. 1972. A rchaeological context an d systemic context. American A ntiquity 37:
155-165.
1976. Behavioral archaeology. N ew Y ork: A cadem ic Press.
1978. M ethodological issues in ethnoarch aeo lo g y . In (R. A. G o u ld , ed.) Explorations in
ethnoarchaeology, pp. 229-247. A lb uquerque: U niversity o f N ew M exico Press.
1983. T o w ard the identification o f fo rm atio n processes. American A ntiquity 48: 675-706.
1985. Is there a “ Pom peii Prem ise” in archaeology? Journal o f Anthropological Research 41:
18^11.
1987. Formation processes o f the archaeological record. A lbuquerque: U niversity o f New
M exico Press.
Schild, R. 1976. The final Paleolithic settlem ents o f the E u ro p ean Plain. Scientific American
234(2): 88-99.
Schneider, J. S. and E verson, G . D . 1989. T he desert tortoise (Xerobates agassizii) in the
prehistory o f the southw estern G re a t Basin an d ad jacen t areas. Journal o f California and
G reat Basin A nthropology 11: 175-202.
Schopf, J. M . 1975. M odes o f fossil preservation. Review o f Paleobotany and Palynology 20:
27-53.
Schulz, P. D . an d G u st, S. M . 1983. F au n al rem ains an d social statu s in 19th century
Sacram ento. H istorical Archaeology 17: 44—53.
Schum m , S. A. 1985. E xplanatio n an d ex trap o latio n in geom orphology: seven reasons for
geologic un certainty. Transactions o f the Japanese Geomorphological Union 6(1): 1-18.
Scott, G . H. 1963. U n iform itarian ism , the uniform ity o f n atu re, an d paleoecology. N ew Zealand
Journal o f Geology and Geophysics 6: 510-527.
Sdzuy, K . 1966. A n im proved m eth o d o f analysing d isto rtio n in fossils. Palaeontology 9:
125-134.
Seilacher, A. 1990. T ap h o n o m y o f fossil-L agerstatten: overview. In (D . E. G . Briggs an d P. R.
C row ther, eds.) Palaeobiology: a synthesis, pp. 266-270. O xford: Blackwell Scientific
P ublications.
1992. D ynam ic taphonom y: the process-related view o f fossil-lagerstatten. In (S. F ernandez
Lopez, ed.) Conferencias de la reunion de tafonom ia y fosilizacion, pp. 109-125. M adrid:
E ditorial C om plutense.
Sem enov, S. 1964. Prehistoric technology. Bath: A dam s & D art.
Shea, B. T. 1988a. A llom etry. In (I. T attersall, E. D elson, an d J. van C ouvering, eds.)
Encyclopedia o f human evolution and prehistory, pp. 20-22. N ew Y ork: G arlan d Publishing.
1988b. O ntogeny. In (I. T attersall, E. D elson, an d J. van C ouvering, eds.) Encyclopedia o f
human evolution and prehistory, pp. 401. N ew Y ork: G arlan d Publishing.
1992. D evelopm ental perspective o n size change an d allom etry in evolution. Evolutionary
Shea, J. H . 1982. Twelve fallacies o f uniform itarianism . Geology 10: 455^160.
S hennan, S. 1988. Q uantifying archaeology. San Diego: A cadem ic Press.
Shipm an, P. 1979. W h at are all these bones doing here? C onfessions o f a tap h o n o m ist. H arvard
M agazine N o v .-D e e ., pp. 42-46.
494 Bibliography
1981a. A pplications o f scanning electron m icroscopy to tap h o n o m ic problem s. In (A. M.

C antw ell, J. B. Griffin, an d N . A. R othschild, eds.) The research potential o f anthropological
museum collections, pp. 357-385. A nnals o f the N ew Y o rk A cadem y o f Science 376.
1981b. L ife history o f a fo s sil: an introduction to taphonom y and paleoecology. C am bridge:
H arv ard U niversity Press.
1983. E arly hom inid lifestyles: hu n tin g an d gathering o r foraging an d scavenging? In (J.
C lu tto n-B ro ck an d C. G rigson, eds.) Anim als and archaeology: 1. hunters and their prey, pp.
31^49. British A rchaological R ep o rts In tern atio n al Series 163.
1986a. Scavenging o r hun tin g in early hom inids: theoretical fram ew ork and test. American
Anthropologist 88: 27-43.
1986b. Studies o f h o m in id -fau n al in teractio n s at O lduvai G orge. Journal o f H uman Evolution
15: 691-706.
1987. R esponse to R. Lee Lym an. Am erican Anthropologist 89: 715-717.
1988a. A ctualistic studies o f anim al resources an d hom inid activities. In (S. L. Olsen, ed.)
Scanning electron microscopy in archaeology, pp. 261-285. British A rchaeological R eports
1988b. D iet and subsistence strategies at O lduvai G orge. In (B. V. K ennedy an d G . M.
L eM oine, eds.) D iet and subsistence: current archaeological perspectives, pp. 3-12. C algary:
U niversity o f C algary A rchaeological A ssociation.
1989. A ltered bones from O lduvai G orge, T anzania: techniques, problem s, and im plications
o f their recognition. In (R . B onnichsen an d M . H . Sorg, eds.) Bone modification, pp. 317—
334. O rono: U niversity o f M aine C en ter fo r the Study o f the F irst A m ericans.
Shipm an, P.. Bosler, W ., and D avis, K . L. 1981. B utchering o f g iant geladas at an A cheulian site.
Current Anthropology 22: 257-268.
Shipm an, P., Fisher, D . C., and R ose, J. 1984a. M asto d o n butchery: m icroscopic evidence o f
carcass processing and bone tool use. Paleobiology 10: 358-365.
S hipm an, P., F o ster, G ., an d Schoeninger, M . 1984b. B u rn t bones an d teeth: an experim ental
study o f color, m orphology, crystal stru ctu re an d shrinkage. Journal o f Archaeological
Science 11: 307-325.
S hipm an, P. an d Phillips, J. E. 1976. O n scavenging by hom inids and o th er carnivores. Current
S hipm an, P. and Phillips-C onroy, J. E. 1977. H om inid tool-m aking versus carnivore scavenging.
American Journal o f Physical Anthropology 46: 77-86.
Shipm an, P. an d R ose, J. 1983a. E arly hom inid hunting, butchering, an d carcass-processing
behaviors: approaches to the fossil record. Journal o f Anthropological Archaeology 2: 57-98.
1983b. Evidence o f butchery activities a t T o rralb a an d A m brona: an evalu atio n using
m icroscopic techniques. Journal o f Archaeological Science 10: 465-474.
1984. C u tm ark m imics on m odern an d fossil bovid bones. Current Anthropology 25: 116-117.
1988. Bone tools: an experim ental ap p ro ach . In (S. L. Olsen, ed.) Scanning electron microscopy
in archaeology, pp. 303-335. British A rchaeological R ep o rts In tern atio n al Series 452.
Shipm an, P. an d W alker, A. 1980. Bone-collecting by harvesting ants. Paleobiology 6: 496-502.
Shotw ell, J. A. 1955. A n ap p ro ach to the paleoecology o f m am m als. Ecology 36: 327-337.
Sillen, A. 1989. D iagenesis o f the in organic phase o f cortical bone. In (T. D. Price, ed.) The
chem istry o f prehistoric human bone, pp. 211-229. C am bridge: C am bridge U niversity Press.
Simms, S. R. 1992. E thnoarchaeology: o bnoxious spectator, trivial p u rsu it, o r the keys to a time
m achine? In (L. W andsnider, ed.) Quandaries and quests: visions o f archaeology's fu tu re, pp.
186-198. S outhern Illinois U niversity C en ter for A rchaeological Investigations O ccasional
P aper N o. 20. C arbon d ale.
Sim pson, G . G . 1961. Som e problem s o f vertebrate paleontology. Science 133: 1679-1689.
1963. H istorical science. In (C. C. A lb ritto n , Jr., ed.) The fa b ric o f geology, pp. 24-48.
S tanford: Freem an, C ooper, & Co.
Bibliography 495
1970. U niform itarianism : an inquiry in to principle, theory, an d m eth o d in geohistory and

biohistory. In (M . K. H echt and W . C. Steere, eds.) Essays in evolution and genetics in honor
o f Theodosius D ohzhansky, pp. 43-96. N ew Y ork: A ppleton.
Sim pson, T. 1984. P o p u latio n dynam ics o f m ule deer. In (G . C. F riso n an d D. N . W alker, eds.)
The D ead Indian Creek Site: an Archaic occupation in the A bsaroka M ountains o f
northeastern W yom ing, pp. 83-96. T he W yom ing A rchaeologist 27.
Singer, D . A. 1985. The use o f fish rem ains as a socio-econom ic m easure: an exam ple from 19th-
century N ew E ngland. H istorical Archaeology 19: 132-136.
Skinner, J. D . and van A arde, R. J. 1991. Bone collecting by brow n hyaenas H yaena brunnea in
the C en tral N am ib D esert, N am ibia. Journal o f Archaeological Science 18: 513-523.
Sm ith, B. D. 1974. P re d a to r-p re y relationships in the south eastern O z a r k s - A .D . 1300. Human
Ecology 2: 31—43.
1976. “ T w itching:” a m in o r ailm ent affecting h u m an paleoecological research. In (C. E.
C leland, ed.) Cultural change and continuity, pp. 275-292. N ew Y ork: A cadem ic Press.
1977. A rchaeological inference and inductive confirm ation. American A nthropologist 79:
598-617.
1979. M easuring the selective utilization o f anim al species by prehistoric hum an p opulations.
Am erican A ntiquity 44: 155-160.
Sm ith, G . E. 1908. The alleged discovery o f syphilis in p rehistoric E gyptians. Lancet 2: 521-524.
Sm ith, G . R.. Stearley. R. F., and Badgley, C. E. 1988. T ap h o n o m ic bias in fish diversity from
C enozoic fioodplain environm ents. Palaeogeography, Palaeoclim atology, Palaeoecology
63: 263-273.
Sm ith, R. J. 1984. A llom etric scaling in com p arativ e biology: problem s o f concept an d m ethod.
American Journal o f Physiology 246: R 152-160.
Sokal, R. R. and R ohlf, F. J. 1969. Biom etry. San F rancisco: W. H . F reem an and C om pany.
Solom on, S. 1990. W h a t is this thing called tap h o n o m y ? In (S. Solom on, I. D avidson, and D.
W atson, eds.) Problem solving in taphonom y, pp. 25-33. T em pus vol. 2.
Solom on, S., D avidson, I., an d W atson, D . 1990. (eds.) Problem solving in taphonomy:
archaeology and palaeontological studies fr o m Europe, A frica and Oceania. Tem pus:
A rchaeology and M aterial C u ltu re Studies in A n th ro p o lo g y vol. 2. St. Lucia: U niversity o f
Q ueensland A n th ro p o lo g y M useum .
Solom on, S., M innegal, M ., an d D wyer, P. 1986. Bower birds, bones an d archaeology. Journal o f
Solounias, N . 1988. Evidence from h o rn m o rphology on the phylogenetic relationships o f the
p ro n g h o rn (Antilocapra americana). Journal o f M am m alogy 69: 140-143.
Speth, J. D. 1983. Bison kills and bone counts. C hicago: U niversity o f C hicago Press.
1991. T aphono m y an d early hom inid behavior: problem s in distinguishing cu ltural an d non-
cultural agents. In (M . C. Stiner, ed.) H um an predators and prey m ortality, pp. 31—40.
Boulder: W estview Press.
Speth, J. D. an d S p ielm an n , K. A. 1983. Energy source, p rotein m etabolism , and h u n ter-g ath erer
subsistence strategies. Journal o f Anthropological Archaeology 2: 1-31.
Spiess, A. E. 1979. Reindeer and caribou hunters: an archaeological study. N ew Y ork: A cadem ic
Press.
Stahl, P. W. and B rinker, U . H . 1991. R em oval o f calcareous in cru statio n from bone a n d teeth
with acid an d u ltraso u n d . Journal o f Field Archaeology 18: 138-140.
Stahl, P. W . an d Zeidler, J. A. 1990. D ifferential bone-refuse accu m u latio n in fo o d -p rep aratio n
and traffic areas on an early E cu ad o rian house floor. L atin American A ntiq u ity 1: 150-169.
Stallibrass, S. 1984. The distinctio n betw een the effects o f sm all carnivores an d hum ans on p o st
glacial faunal assem blages. In (C. G rigson an d J. C lutto n -B ro ck , eds.) Anim als and
archaeology: 4. husbandry in Europe, pp. 259-269. British A rchaeological R ep o rts In te r
nation al Series 227.
496 Bibliography
1990. C anid dam age to anim al bones: tw o cu rren t lines o f evidence. In (D . E. R obinson, ed.)
Experim entation and reconstruction in environmental archaeology, pp. 151-165. O xford:
O xbow Books.
Stein, B. R. 1989. Bone density an d ad a p ta tio n in sem iaquatic m am m als. Journal o f M am m alogy
70: 467^176.
Stein, J. K. 1987. D eposits for archaeologists. In (M . B. Schiffer, ed.) Advances in archaeological
m ethod and theory vol. 11, pp. 337-395. San Diego: A cadem ic Press.
Stevens, S. S. 1946. O n th e theory o f scales o f m easurem ent. Science 103: 677-680.
S tew art, K . M . 1989. Fishing sites o f north and east A frica in the late Pleistocene and Holocene:
environmental change and human adaptation. British A rchaeological R eports In tern atio n al
Series 521.
1991. M o d ern fishbone assem blages a t L ake T u rk an a , K enya: a m ethodology to aid in
recognition o f hom inid fish utilization. Journal o f Archaeological Science 18: 579-603.
Stiles, D . 1977. E thnoarchaeology: a discussion o f m eth o d s and applications. M an 12: 87-103.
Stiner, M . C. 1990a. T he ecology o f choice: p ro cu rem en t an d tra n sp o rt o f anim al resources by
U p p er Pleistocene hom inids in w est-central Italy. P h.D . dissertation, U niversity o f New
M exico, A lbuquerque. A nn A rb o r: U niversity M icrofilm s In tern atio n al.
1990b. T he use o f m ortality p attern s in archaeological studies o f hom inid p red ato ry
ad ap ta tio n s. Journal o f Anthropological Archaeology 9: 305-351.
1991 a. A n interspecific perspective o n the em ergence o f the m od ern h u m an pre d a to ry niche. In
(M . C. Stiner, ed.) H uman predators and prey m ortality, pp. 149-185. Boulder: W estview
Press.
1991b. F o o d p rocu rem en t an d tra n sp o rt by h u m an an d n o n -h u m an p red ato rs. Journal o f
Archaeological Science 18: 455—482.
1991c. (ed.) Human predators and prey m ortality. B oulder, CO: W estview Press.
1991 d. Introd u ction: actualistic an d archaeological studies o f prey m o rtality. In (M . C. Stiner,
ed.) H um an predators and p rey m ortality, pp. 1-13. Boulder: W estview Press.
1991e. T he faunal rem ains from G ro tta G u a ttari: a tap h o n o m ic perspective. Current
1992. O verlapping species “ choice” by Italian U p p er Pleistocene p red ato rs. Current A nthro
pology 33: 433-451.
S tirton, R. A . 1959. Time, life and man: the fo s sil record. N ew Y ork: Jo h n W iley an d Sons, Inc.
Stock, C. 1956. R an ch o L a Brea: a record o f Pleistocene life in C alifornia. L os Angeles C ounty
M useum o f N atural H istory Science Series N o . 20, 6th edition.
Straus, L. G . 1990. U n d erg ro u n d archaeology: perspectives o n caves an d rockshelters. In (M . B.
Schiffer, ed.) Archaeological m ethod and theory vol. 2, pp. 255-304. T ucson: U niversity o f
A rizona Press.
Styles, B. W . 1981. Faunal exploitation and resource selection: early L a te W oodland subsistence in
the Lower Illinois Valley. N o rth w estern U niversity A rchaeological P ro g ram Scientific
P aper N o . 3, E van ston , Illinois.
Sutcliffe, A. J. 1970. S potted-hyena: crusher, gnaw er, digester and collector o f bones. N ature 227:
1110-1113.
1973. Sim ilarity o f bones and antlers gnaw ed by deer to h u m an artifacts. N ature 246: 428^130.
1977. F u rth e r notes on bones an d an tlers chew ed by deer an d o th e r ungulates. Deer 4: 73-82.
T appen, N . C. 1969. T he relatio n sh ip o f w eathering cracks to split-line o rien tatio n in bone.
American Journal o f Physical A nthropology 31: 191-198.
T appen, N . C. an d Peske, G . R. 1970. M icroscopic exam in atio n o f h u m an fem urs buried at
W ashington Island. W isconsin. Am erican A ntiquity 35: 463^465.
T edford, R. H . 1970. Principles an d practices o f m am m alian geochronology in N o rth A m erica.
In (E. L. Y ochelson, ed.) Proceedings o f the N orth Am erican paleontological convention, pp.
666-703. Lawrence: A llen Press.
B ibliography 497
T hom as, D . H . 1971. O n distinguishing n a tu ral from cu ltu ral bone in archaeological sites.
T hom as, D . H . an d M ayer, D . 1983. B ehavioral faunal analysis o f selected horizons. In (D. H.
T hom as) The archaeology o f M onitor Valley 2: G atecliff Shelter, pp. 353-391. A m erican
M useum o f N a tu ra l H isto ry A n th ro p o lo g ical P ap ers 59(1).
T hom as, R. D . K . 1986. T apho n o m y : ecology’s loss is sedim entology’s gain. Palaios 1: 206.
T h u rm an , M . D . an d W illm ore, L. J. 1981. A replicative crem ation experim ent. N orth American
Archaeologist 2: 275-283.
T ipper, J. C. 1979. R arefaction an d rarefiction - the use a n d abuse o f a m eth o d in paleoecology.
Paleobiology 5: 423^134.
T obin, M . E. and D olbeer, R. A. 1990. D isap p earan ce an d recoverability o f songbird carcasses
in fruit orchards. Journal o f Field O rnithology 61: 237-242.
T odd, L. C. 1983a. T aph o nom y: fleshing o u t the d ry bones o f P lains prehistory. The W yom ing
Archaeologist 26(3-4): 36-46.
1983b. T he H o rn e r Site: tap h o n o m y o f an early H olocene bison bonebed. P h .D . dissertation,
U niversity o f New M exico, A lbuquerque.
1987a. A nalysis o f kill-butchery bonebeds an d in terp re ta tio n o f P aleo in d ian hunting. In (M .
H. N itecki and D. V. N itecki, eds.) The evolution o f human hunting, pp. 225-266. N ew Y ork:
Plenum Press.
1987b. T ap ho n om y o f the H o rn er II b one bed. In (G. C. F rison an d L. C. T o d d , eds.) The
Horner Site: the type site o f the C ody Cultural C om plex, pp. 107-198. O rlando: A cadem ic
Press.
T odd, L. C. and R apson, D . J. 1988. L ong b one frag m en tatio n an d in terp retatio n o f faunal
assem blages: approaches to com p arativ e analysis. Journal o f Archaeological Science 15:
307-325.
T odd, L. C ., W itter, R. V., an d F rison, G . C. 1987. E xcavation an d d o cu m en tatio n o f the
P rinceton an d Sm ithsonian H o rn e r Site assem blages. In (G . C. F riso n an d L. C. T o d d , eds.)
The H orner Site: the type site o f the C ody C ultural C om plex, pp. 39-91. O rlando: A cadem ic
Press.
T oots, H . 1965a. O rien tation an d d istrib u tio n o f fossils as enviro n m en tal indicators. N ineteenth
Field Conference o f the W yom ing Geological Association, pp. 219-229.
1965b. R an d o m o rien tatio n o f fossils an d its significance. University o f W yom ing Contribu
tions to Geology 4: 59-62. Laram ie.
1965c. Sequence o f d isarticu latio n in m am m alian skeletons. University o f W yom ing Contribu
tions to Geology 4: 37-39. L aram ie.
T orbenson, M ., A ufderheide, A ., and Jo h n so n , E. 1992. P u n ctu red h u m an bones o f the Laurel
C u ltu re from Sm ith M o u n d F o u r, M in n eso ta. Am erican A ntiquity 57: 506-514.
T ou rn al, M . 1833. G eneral consid eratio n s on the p h en o m en o n o f b one caverns. A nnales de
Chimie et de Physique 25: 161-181. (English tra n slatio n by A. B. Elsasser p rin ted 1959 in
Kroeber Anthropological Society Papers 21: 6-16; rep rin ted in 1969 in [R. F. H eizer, ed.]
M a n ’s discovery o f his past, pp. 84-94, P alo A lto: Peek Publications).
Trigger, B. G . 1989. A history o f archaeological thought. C am bridge: C am bridge U niversity
Press.
T ringham , R. 1978. E xperim entation, ethnoarch aeo lo g y , an d the leapfrogs in archaeological
m ethodology. In (R . A. G o u ld , ed.) Explorations in ethnoarchaeology, pp. 169-199.
A lbuquerque: U niversity o f New M exico Press.
T ucker, M . E. 1990. The diagenesis o f fossils. In (S. K . D o n o v an , ed.) The processes o f
fossilization. pp. 84—104. N ew Y ork: C olum bia U niversity Press.
T urn er, A. 1983. The quantification o f relative ab u n d an ces in fossil and subfossil bone
assem blages. Annals o f the Transvaal M useum 33: 311-321.
1984. Identifying bo n e-accum ulating agents. In (M . H all, G . A very, D . M . A very, M . L.
498 Bibliography
W ilson, an d A. J. B. H um phreys, eds.) Frontiers: Southern A frican archaeology today, pp.

334-339. B ritish A rchaeological R ep o rts In tern atio n al Series 207.
1989. Sam ple selection, schlepp effects an d scavenging: the im plications o f p artial recovery for
in terp retatio n s o f the terrestrial m am m al assem blage from K lasies R iver M o u th . Journal o f
T urner, C. G ., II. 1983. T ap h o n o m ic reco n stru ctio n s o f h u m an violence an d cannibalism based
on m ass burials in the A m erican Southw est. In (G. M . L eM oine an d A. S. M acE achern,
eds.) Carnivores, human scavengers & predators: a question o f bone technology, pp. 219-240.
C algary: U niversity o f C algary A rchaeological A ssociation.
T uross, N ., Behrensm eyer, A. K ., an d E anes, E. D . 1989. S tro n tiu m increases an d crystallinity
changes in tap h o n o m ic an d archaeological bone. Journal o f Archaeological Science 16:
661-672.
Tyler, C. 1969. A vian egg shells: th eir stru ctu re an d characteristics. International Review o f
Genetics and Experim ental Zoology 4: 81-130.
U erpm ann , H. P. 1973. A nim al bone finds an d econom ic archaeology: a critical study o f “ osteo-
archaeological” m ethod. W orld Archaeology 4: 307-332.
V an C ouvering, J. A. H . 1980. C om m unity evolution in east A frica d u rin g the late C enozoic. In
(A. K . B ehrensm eyer an d A. P. Hill, eds.) Fossils in the m aking, pp. 272-298. C hicago:
U niversity o f C hicago Press.
V an N eer, W. an d M uniz, A. M . 1992. “ Fish m iddens:” an th ro p o g en ic accum ulations o f fish
rem ains an d their bearing o n archaeoichthyological analysis. Journal o f Archaeological
Science 19: 683-695.
V ehik, S. C. 1977. Bone fragm ents an d bone grease m an u factu re: a review o f their archaeological
use and potential. Plains Anthropologist 22: 169-182.
Villa, P. 1991. M iddle Pleistocene preh isto ry in southw estern E urope: the state o f o u r know ledge
and ignorance. Journal o f Anthropological Research 47: 193-217.
Villa, P. an d M ahieu, E. 1991. B reakage p attern s o f h u m an long bones. Journal o f Human
Evolution 21: 27-48.
Villa, P., Bouville, C ., C o u rtin , J., H elm er, D ., M ahieu, E., Shipm an, P., B elluom ini, G ., and
B ranca, M . 1986. C annibalism in the N eolithic. Science 233: 431—437.
V on E ndt, D . W . and O rtn er, D . J. 1984. E xperim ental effects o f b one size an d tem p eratu re on
bone diagenesis. Journal o f Archaeological Science 11: 247-253.
V oorhies, M . 1969. Taphonom y and population dynam ics o f an early Pliocene vertebrate fauna,
K nox C ounty, N ebraska. U niversity o f W yom ing C o n trib u tio n s to G eology Special P aper
N o. 1. Laram ie.
1970. Sam pling difficulties in reconstructing late T ertiary m am m alian com m unities. In (E. L.
Y ochelson, ed.) Proceedings o f the N orth American paleontological convention, pp. 454-468.
Lawrence: Allen Press.
1981. A ncient ashfall creates a Pom peii o f p rehistoric anim als. N ational Geographic 159:
66-75.
W ainw right, S. A ., Biggs, W . D ., C urrey, J. D ., an d G osline, J. M . 1976. M echanical design in
organisms. N ew Y ork: Jo h n W iley & Sons.
W alker, P. L. an d L ong, J. C. 1977. A n experim ental study o f the m o rphological characteristics
o f tool m arks. Am erican A ntiquity 42: 605-616.
W all, W. P. 1983. T he co rrelatio n betw een high lim b-bone density an d a q u atic h ab its in R ecent
m am m als. Journal o f Paleontology 57: 197-207.
W alters, I. 1984. G o ne to th e dogs: a study o f bone a ttritio n at a cen tral A u stralian cam psite.
M ankin d 14: 389^400.
1985. B one loss: one explicit q u an titativ e guess. Current Anthropology 26: 642-643.
W arm e, J. E. and H antzschel, W . 1979. A ctualistic paleontology. In (R . W . F airb rid g e an d D.
Bibliography 499
Jablon ski, eds.) Encyclopedia o f paleontology, pp. 4-1 0 . S troudsburg: D ow den, H u tch in so n
& R oss, Inc.
W arrick, G . a n d K rau sm an , P. R. 1986. B one-chew ing by desert b ig h o rn sheep. Southw estern
N aturalist 31: 414.
W atson, J. A. L. an d A bbey, H . M . 1986. T he effects o f term ites (Isoptera) o n bone: some
archeological im plications. Sociobiology 11: 245-254.
W atson, J. P. N . 1972. F rag m en tatio n analysis o f anim al bone sam ples from archaeological sites.
A rchaeom etry 14: 221-228.
W atson, P. J. 1979. T he idea o f ethnoarchaeology: notes and com m ents. In (C. K ram er, ed.)
Ethnoarchaeology: implications o f ethnography fo r archaeology, pp. 277-287. N ew Y ork:
C o lum bia U niversity Press.
W atson , R. A . 1966. D iscussion: is geology different: a critical discussion o f “ T he fabric o f
geology." Philosophy o f Science 33: 172-185.
1969. E xplanation and prediction in geology. Journal o f Geology 77: 488-494.
1976. Inference in archaeology. A merican A ntiquity 41: 58-66.
W eigelt, J. 1927. R ezente wirbeltierleichen und ihre palaobiologische bedeutung. Leipzig: M ax
W eg Verlag.
1989. R ecent vertebrate carcasses and theirpaleobiological implications. C hicago: U niversity o f
C hicago Press. (English tran slatio n o f W eigelt 1927, by J. Schaefer).
W ellm an, H . W . 1962. A graphical m eth o d fo r analysing fossil d isto rtio n caused by tectonic
deform ation. Geological M agazine 99: 348-352.
W estern, D . 1980. L inking the ecology o f p ast an d presen t m am m al com m unities. In (A. K.
B ehrensm eyer an d A. P. Hill, eds.) Fossils in the m aking, pp. 41-54. Chicago: U niversity o f
C hicago Press.
W heat, J. B. 1972. The O lsen-Chubbuck site: a Paleo-Indian bison kill. Society fo r A m erican
A rchaeology M em oir 26.
1979. The Jurgens Site. P lains A n th ro p o lo g ist M em oir 15.
W heeler, A. and Jones, A. K . G . 1989. Fishes. C am bridge: C am bridge U niversity Press.
W heeler, M . 1954. Archaeology fr o m the earth. M iddlesex: Penguin Books.
W hewell, W. 1832. Review o f vol. 2 o f C harles Lyell’s Principles o f Geology. Q uarterly Review 47:
103-132.
W hite, E. M . and H an nus, L. A. 1983. C hem ical w eathering o f bone in archaeological soils.
Am erican A ntiquity 48: 316-322.
W hite, J. A., M cD o n ald, H . G ., A n d erso n , E., an d Soiset, J. M . 1984. L ava blisters as carnivore
traps. In (H. H. G enow ays an d M . R. D aw son, eds.) C ontributions in Quaternary vertebrate
paleontology: a volume in m em orial to John E. Guilday, pp. 241-256. C arnegie M useum o f
N a tu ra l H isto ry Special P ublication N o. 8. P ittsburgh.
W hite, T. D . 1992. Prehistoric cannibalism at M ancos 5 M T U M R -2 3 4 6 . P rinceton: Princeton
U niversity Press.
W hite, T. E. 1952a. A nim al bone an d Plains archeology. Plains Archaeological Conference
N ew sletter 4: 46-48.
1952b. O bservations o n the butchering technique o f som e aboriginal peoples: no. 1. American
A ntiquity 17: 337-338.
1952c. Suggestions on the butchering technique o f the in h ab itan ts o f the D o d d an d Phillips
R anch sites in O ahe R eservoir A rea. Plains Archaeological Conference N ew sletter 5: 20-25.
1953a. A m eth od o f calculating the d ietary percentage o f various food anim als utilized by
aboriginal peoples. American A ntiquity 19: 396-398.
1953b. O bservations on the butchering technique o f som e aboriginal peoples: no. 2. American
A ntiquity 19: 160-164.
1953c. Studying osteological m aterial. Plains Archaeological Conference N ew sletter 6: 58-67.
500 Bibliography
1954. O bservations on the butchering technique o f som e ab original peoples: nos. 3 ,4 , 5, a n d 6.

American A ntiquity 19: 254—264.
1955. O bservations on the butchering technique o f som e ab original peoples: nos. 7, 8, an d 9.
1956. T he study o f osteological m aterials in the Plains. Am erican A ntiq u ity 21: 401-404.
W hitm er, A. M ., R am enofsky, A. F ., T h o m as, J., T hibo d eau x , L. J., Field, S. D., and M iller, B.
J. 1989. Stability o r instability: the role o f diffusion in trace elem ent studies. In (M . B.
Schiffer, ed.) Archaeological m ethod and theory vol. 1, pp. 205-273. T ucson: U niversity o f
A rizon a Press.
W ika, M . 1982. A ntlers - a m ineral source in Rangifer. A cta Zoologica 63: 7-10.
W ilkinson, P. F. 1976. “ R a n d o m ” h u n tin g and the com position o f faunal sam ples from
archaeological excavations: a m o d ern exam ple from N ew Z ealand. Journal o f Archaeologi
cal Science 3: 321-328.
Will, R. T. 1985. N ineteenth C entury C o p p er In u it subsistence practices o n B anks Island, N. W.
T. P h.D . dissertation, U niversity o f A lberta, E d m o n to n .
W illey, G . R. and Sabloff, J. A. 1980. A history o f American archaeology, 2nd edition. San
F rancisco: W. H . F reem an an d C om pany.
W ilier, D . and W ilier, J. 1973. S ystem atic empiricism: critique o f a pseudoscience. Englew ood
Cliffs: Prentice-H all. Inc.
W ilson, L. G . 1967. T he origins o f C harles Lyell’s u niform itarianism . In (C. C. A lb ritto n , Jr., ed.)
U niform ity and sim plicity, pp. 35-62. G eological Society o f A m erica Special P ap er N o. 89.
W ilson, M . V. H. 1988. T ap h o n o m ic processes: in fo rm atio n loss and in fo rm atio n gain.
Geoscience Canada 15: 131-148.
W ilson, T. 1901. A rrow w ounds. A m erican Anthropologist 3: 513-531.
W ing, E. S. an d B row n, A. B. 1979. Paleonutrition: m ethod and theory in prehistoric foodw ays.
N ew Y ork: A cadem ic Press.
W intem berg, W. J. 1919. A rchaeology as an aid to zoology. The Canadian Field-Naturalist 33:
63-72.
W olberg, D . L. 1970. T he hypothesized o steo d o n to k eratic culture o f the A ustralopithecinae: a
look at the evidence a n d the opinions. Current Anthropology 11: 23-37.
Wolff, R. G . 1973. H ydrod y nam ic sorting an d ecology o f a Pleistocene m am m alian assem blage
from C alifornia (U .S.A .). Paleogeography, Palaeoclim atology, Palaeoecology 13: 91-101.
1975. Sam pling and sam ple size in ecological analysis o f fossil m am m als. Paleobiology 1:
195-204.
W ood, W. R. 1962. N otes on the bison b one from the Paul Brave, Huff, an d D em ery Sites (O ahe
R eservoir). Plains A nthropologist 7: 201-204.
1968. M ississippian hu n ting and butchering pattern s: bone from the V ista Shelter, 23SR-20,
M issouri. American A ntiquity 33: 170-179.
W ood, W . R. an d John son , D. L. 1978. A survey o f d istu rb an ce processes in archaeological site
form ation. In (M . B. Schiffer, ed.) A dvances in archaeological m ethod and theory vol. 1, pp.
315-381. N ew Y ork: A cadem ic Press.
W right, R. 1990. T aphonom y: its applications an d im plications for archaeology. In (S. Solom on,
1. D avidson, and D. W atson, eds.) Problem solving in taphonom y, pp. 260-264. T em pus vol.
2.
Wylie, A. 1982a. A n analogy by any o th er nam e is ju s t as analogical: a co m m entary on the
G o u ld -W a tso n dialogue. Journal o f Anthropological Archaeology 1: 382—401.
1982b. Epistem ological issues raised by a stru ctu ralist archaeology. In (I. H o d d er, ed.)
Sym bolic and structural archaeology, pp. 39—46. C am bridge: C am bridge U niversity Press.
1985. T he reaction against analogy. In (M . B. Schiffer, ed.) Advances in archaeological m ethod
and theory vol. 8, pp. 63-111. N ew Y ork: A cadem ic Press.
Bibliography 501
1988. ‘Sim ple’ analogy and the role o f relevance assum ptions: im plications o f archaeological
practice. International Studies in the Philosophy o f Science 2: 134—150.
1989a. A rchaeological cables an d tacking: the im plications o f practice for B ernstein’s ‘o ptions
beyond objectivism an d relativism .’ Philosophy o f the Social Sciences 19: 1-18.
1989b. T he interpretive dilem m a. In (V. Pinsky an d A. W ylie, eds.) C ritical traditions in
contem porary archaeology, pp. 18-27. C am bridge: C am bridge U niversity Press.
W ym an, J. 1868. A n acco u nt o f som e kjoekkenm oeddings, or shell-heaps, in M aine and
M assachusetts. The American N aturalist 1: 561-584.
Y ellen, J. E. 1977. C u ltu ral p attern in g in faunal rem ains: evidence from the !K ung B ushm en, in
(D . Ingersoll, J. E. Y ellen, and W. M acD o n ald , eds.) E xperim ental archaeology, pp. 2 7 1 -
331. N ew Y ork: C olum bia U niversity Press.
1991a. Sm all m am m als: !K ung San utilization a n d the p ro d u ctio n o f faunal assem blages.
Journal o f Anthropological Archaeology 10: 1-26.
1991b. Small m am m als: post-d iscard p attern in g o f !K ung San faunal rem ains. Journal o f
Anthropological Archaeology 10: 152-192.
Z angerl, R. 1969. T he tu rtle shell. In (C. G ans, ed.) Biology o f the Reptilia vol. I , morphology A,
pp. 311-339. L ondon: A cadem ic Press.
Ziegler, A. C. 1965. T he role o f faunal rem ains in archeological investigations. Sacrem ento
Anthropological Society Papers 3: 47-75.
1973. Inference from prehistoric fau n al rem ains. Addison- Wesley M odule in Anthropology N o.
43.
Z ierhut, N . W. 1967. B one breaking activities o f the C alling L ake Cree. Alberta Anthropologist
1(3): 33-36.
GLOSSARY
P alaeontologists ap p aren tly have never been very strict in the use o f definitions and
m any w ords have evolved to have m eanings quite different from the original.
(G. C. C adee 1990:4)
These term s should be clearly defined by researchers w ho use them , tak in g into
acco u nt b o th the d a ta a t h a n d an d the questions being addressed.
(A. K. B ehrensm eyer 1991:324)
O ne o f the earliest lists o f ta p h o n o m ic term s was published by C lark et al.

(1967:155). Since th a t tim e, the list h as g ro w n considerably. T he follow ing is no
d o u b t inco m p lete, b u t I believe it co n tain s m o st o f the n ow regularly used
term s. It is in ten d ed to serve as an in tro d u c tio n to the term s one will e n c o u n te r
w hen re ad in g th e lite ra tu re o n ta p h o n o m y , w h e th e r th a t lite ra tu re concerns
v erteb rates o r in v erteb rates, a n d also includes term s th a t are specifically
arch aeo lo g ical b u t relev ant to discussion o f ta p h o n o m ic p ro b lem s in zo o ar-
chaeology. D ifferent a u th o rs som etim es offer slightly different definitions, an d
th u s w here possible th e source o f a p a rtic u la r definition is given. U n a ttrib u te d
definitions are m y ow n, gleaned from the literatu re. T he o rd e r o f m ultiple
definitions fo r a term is n o t m e a n t to d en o te p rim a ry a n d secondary, or
p referred a n d n o n p re ferred definitions, ra th e r th e o rd e r o f definitions is sim ply
ch ro n o lo g ical. It is th u s im p o rta n t to n o te th a t I have m ad e no intensive effort
to trac e the etym ology o f each term a n d find its original definition. A n analysis
o f th e ch ro n o lo g y o f term in tro d u c tio n m ig h t illu m in ate vario u s in terestin g
aspects o f th e h isto ry o f ta p h o n o m ic research, b u t is b eyond the scope o f this
volum e.
ab rasio n : (1) “ th e resu lt o f an y ag e n t th a t ero d es th e b o n e surface th ro u g h the

ap p lica tio n o f physical fo rce” (B rom age 1984:173); (2) “ th e rem o v al o f
b o n e m ateria l cau sed by the im p act o f sedim entary p articles” (S hipm an
a n d R o se 1988:323); (3) rem o v al o f edges a n d /o r surfaces o f an im al
rem ain s by physical ero sio n (as o p p o sed to chem ical d isso lu tio n ) (B eh
rensm eyer et al. 1989); (4) “ physical grin d in g an d polishing, resu ltin g in
ro u n d in g o f skeletal elem ents an d loss o f surficial d etails” (B rett
1990a:224); (5) “ the w earing-dow n o f skeletons d u e to th eir differential
m o v em en t w ith respect to sed im en ts” (P arso n s a n d B rett 1990:38); (6)
“ th e loss o f ex tern al surfaces a n d p ro jectio n s th ro u g h physical o r chem ical
e ro sio n ” (B ehrensm eyer 1991:309); (7) see also “ b io e ro sio n ,” “ co rra-
sio n ,” a n d “ c o rro sio n /d isso lu tio n ”
502
G lossary o f taphonom y term inology 503
actualism : “ the m e th o d o lo g y o f in ferrin g the n a tu re o f p a st events by analogy

w ith processes o b serv ab le a n d in a c tio n a t the p re se n t” (R udw ick
1976:110)
ac tu o p a leo n to lo g y : (1) the study o f m o d e rn o rganism s a n d en v iro n m en ts fo r

ap p lica tio n to p aleo n to lo g ical p ro b lem s (K ra n z 1974b; see also Schafer
1962/1972; R ich ter 1928); (2) em phasizes the idea th a t u n d ersta n d in g the
p o st-m o rte m h isto ry o f one fossil g ro u p o ften requires know ledge o f the
life histo ries o f asso ciated a n d in tera ctin g o rganism s (L aw rence 1968); (3)
“ th e a p p lic a tio n o f the u n ifo rm ita ria n principle to p aleo n to lo g ical
p ro b le m s” (W arm e a n d H an tzsch el 1979); (4) see also “ n e o ta p h o n o m y ”
a n d “ n e o n to lo g y ”
ag en t o r agency: (1) “ the general g ro u p to w hich th e p h e n o m e n o n causing

ch an g e in the o rig in al state o f th e b o n e o r a b o n e assem blage can be classed
(e.g., biological, geological, h o m in id )” (Jo h n so n 1985:158); (2) the
“ im m ed iate physical ca u se” o f m o d ificatio n to bones (G iffo rd -G o n zalez
1991:228); (3) see also “ effect" a n d “ p ro cess”
allo ch th o n o u s: (1) a fossil assem blage w hich h as been tra n sp o rte d fro m the
area w here th e rep resen ted an im als died a n d p re su m a b ly lived; (2)
"fo reig n , th a t p a rt o f the assem blage w hich did n o t in h ab it the e n v iro n
m en t o f d ep o sitio n an d arriv ed th ere a fte r active o r passive tra n s p o r t”
(S au n d ers 1977:68); (3) “ rem ains th a t have been m oved from the site o f
d e a th a n d o u t o f the original [life] h a b ita t” (B ehrensm eyer a n d H o o k
1992:19); (4) see also “ a u to c h th o n o u s ,” “ c h th o n ic ,” and
“ p a ra u to c h th o n o u s ”
anadiagenesis: (1) th e second ph ase o f diagenesis, involves “ the deep b u rial

ph ase o f co m p ac tio n a n d c e m e n tatio n . . . in o rg an ic chem ical reactio n s
p re d o m in a te ” (R olfe a n d B rett 1969:233); (2) see also “ diagenesis” an d
“ sy ndiagenesis”
an a sta sio n o m y : “ processes a n d events involved in ex posure o f fossils by

n a tu ra l o r h u m a n m eans a n d su b seq u en t m ateria l w eath erin g o r h u m a n
pro cessin g ;” from the G reek anastasis m eaning “ re su rre c tio n ” (G . G.
S im pson, cited in S aunders 1977:68)
an a stro p h e: a c a ta s tro p h e o f lim ited scope a n d area, generally p ro d u c in g m ass

m o rta lity in the affected a rea (K ra n z 1974b)
an ataxic: (1) “ the w eath erin g a n d ero sio n a l fa cto rs w hich act to d estro y a fossil
d u rin g the d e g ra d a tio n o f the rocks in w hich it occu rs” (C lark et al.
1967:155); (2) “ th e sum o f th e p h e n o m e n a w hich o p erate to expose an d
d estro y a fossil” (C lark a n d K ietzke 1967:118)
apo lith o sis: (1) “ fossilization; a n a tu ra lly o cc u rrin g physico-chem ical process
th a t preserves the gross m o rp h o lo g y [of bones] by ionic exchange an d an
504 G lossary o f taphonom y term inology
increase in the size o f the a p a tite crystals; also involves the loss o f organic
m aterial o f th e bo n e an d the filling o f voids w ith m in era ls” (em phasis in
original), from the G reek apolelithom enon m eaning tu rn e d to stone,
“ ap o lith o tic effects refer to the changes b ro u g h t a b o u t by fossilization
p rocesses” (B artsio k as a n d M id d leto n 1992:68); (2) see also “ fossiliza
tio n ,” “ m in e ra liz a tio n ,” “ p erm in e ra liz a tio n ,” a n d “ p e trifa c tio n ”
a rch ae o fau n a : (1) fa u n as recovered fro m arch aeo lo g ical sites (G ray so n 1979);
(2) see also “ p aleo n to lo g ical fa u n a ” a n d “ local fa u n a ”
artic u la tio n : (1) “ th e occurrence o f fossilized elem ents in the sam e an a to m ica l
sequence a n d p ro x im ity as in the living o rg a n ism ” (B adgley 1986a:332);
(2) “ artic u la te d b ones are th o se th a t re ta in th eir exact an a to m ica l
p o sitio n s relative to one a n o th e r” (B ehrensm eyer 1991:302); (3) see also
“ d isa rtic u la tio n ”
a ttritio n : loss o f fossil in fo rm a tio n by n o n -p reserv a tio n (after L aw rence 1968)
a ttritio n a l m o rtality : (1) d iach ro n ic d e a th assem blage, th e d ea th s o f different-

aged anim als o v er a p ro lo n g e d p erio d , th a t indirectly reflects th e age-
specific surviv o rship o f a p o p u la tio n (V oorhies 1969; G ifford 1981); (2)
“ d eath s (due to p re d a tio n , disease, o r o th e r causes) th a t occu r as a p a rt o f
th e n o rm al, lo n g -term tu rn o v e r o f n a tu ra l an im al p o p u la tio n s over the
areas in h ab ited by such p o p u la tio n s” (B ehrensm eyer 1983:94)
au to c h th o n o u s: (1) a fossil assem blage w hich is fo u n d w here the rep resen ted
an im als died an d p re su m a b ly lived; fossils th a t experienced life, d e a th , an d
b u rial w ithin the sam e place o r locale; (2) “ indigenous, th a t p a rt o f the
assem blage w hich is fossilized in situ in th e sed im en t” (S au n d ers 1977:68);
(3) “ skeletal rem ains are preserved close to the place o f d ea th , w ith o u t
significant tra n s p o rt by physical o r biological processes” (B ehrensm eyer
1983:94); (4) “ rem ains are p reserved a t the d ea th site o f a n org an ism o r the
site w here p a rts are discarded; the d e a th o r d iscard site is usually w ith in the
o rig in al [life] h a b ita t” (B ehrensm eyer a n d H o o k 1992:19); (5) see also
“ a llo c h th o n o u s,” “ c h th o n ic ,” a n d " p a ra u to c h th o n o u s ”
autolysis: en zym atic b re a k d o w n o f soft tissue by th e enzym es in the (once-

living) o rg anism , enzym es th a t assisted in m etab o lic fu n ctio n s (H ag lu n d
1991)
b ack g ro u n d : (1) fau n al m aterials c o n trib u te d to geological dep o sits at c a rn i

vore kill sites a n d carn iv o re lairs an d by “ n a tu ra l d e a th s” (B inford
1981 b:20); (2) “ the low est density o f b ones [per unit o f space, such as m 2
rep resented by a ttritio n a l d ea th s a n d kills dispersed th ro u g h o u t an
ecosystem ” (B ehrensm eyer 1987:429)
biocoenose: (1) the life assem blage o f organism s; (2) consists o f a living
p o p u la tio n ; (3) “ a biocoenose encom passes a b io to p e a n d a c o m m u n ity o f
all o rg anism s living in i t ” (S chafer 1962/1972:453); (4) an ecological unit

(or living co m m u n ity ) consisting o f an in teg ra te d living co n g reg atio n o f
diverse organ ism s w ith b o th biotic a n d ab io tic ch aracteristics
b io ero sio n: (1) biological processes o f ero sio n o f skeletal m ateria l (B rett
1990a); (2) “ co rrasiv e processes by o rganism s [such as] b o rin g an d
g razing; results from the search fo r food a n d /o r sh elter” (P arso n s an d
B rett 1990:38, 45); (3) see also “ a b ra s io n ,” “ c o rra sio n ,” an d
“ c o rro sio n /d isso lu tio n ”
b io stratin o m y : (originally b io stra to n o m y ) (1) the stu d y o f pre- a n d syn- burial
in te rre la tio n s betw een d ead organism s a n d th eir ex tern al en v iro n m en t
(L aw rence 1968); (2) the stu d y o f p re -b u rial ta p h o n o m ic fa cto rs, e.g.,
th o se processes affecting an o rg an ism betw een d e a th an d final burial
(L aw rence 1979a); (3) “ the sedim entary h isto ry o f the fossil until b u ria l”
(C adee 1990:3); (4) see also “ n ecro lo g y ” a n d “ p e rth o ta x ic ”
b io tu rb a tio n : d istu rb a n c e by o rganism s, e.g., ch u rn in g , m ixing, tram p lin g ,
b u rro w in g , ro o t p e n e tra tio n (B ehrensm eyer et al. 1989)
bo n e technology: b ro a d ly , “ the h o m in id u tiliza tio n o f bo n e, the m odification
o f w hich can be p u rp o se fu l o r accidental. T h e focus is on the techniques
selected fo r m odifying th e in h ere n t m o rp h o lo g y o f a b o n e ” (Jo h n so n
1985:162)
burial: (1) an im al rem ains are in terre d below the g ro u n d surface an d encased in
sedim ent; rem ains m ay be subjected to a lte ra tio n by sub-surface ta p h o n o
m ic processes, o r exh u m ed /ex p o sed a n d u n d erg o fu rth e r m o d ificatio n
p rio r to re b u rial (B ehrensm eyer et al. 1989); (2) see also “ final b u ria l”
c a ta stro p h ic m o rtality : (1) sy nchronic d e a th assem blage; (2) a rep resen tativ e
sam ple o f all living age classes killed m o re o r less in stan tan eo u sly , form ing
a “ s n a p sh o t” o f a living p o p u la tio n stru ctu re (V oorhies 1969; G ifford
1981)
chrisocoenosis: th e assem blage o f bones created by the p o st-m o rte m use o f
b on es by h u m an s (A. S. G ilb e rt 1979)
ch th o n ic: p erta in in g to things in the g ro u n d o r in the e a rth (from the G reek

“ c h th o n ” m ean in g e a rth )
coarse-g rain ed: (1) “ an assem blage th a t is th e ac cu m u lated p ro d u c t o f events

sp an n in g a n entire y e a r . . . re so lu tio n betw een arch aeo lo g ical rem ains a n d
specific events is p o o r” (B inford 1980:17); (2) an assem blage th a t
“ accu m u lates over a co n sid erab le p erio d o f tim e a n d /o r d u rin g perio d o f
ra p id 'tu rn o v e r' o f events, resulting in the asso ciatio n o f item s, debris,
features, la n d surfaces, a n d the like th a t w ere differential p a rtic ip a n ts in
different events d u rin g the course o f the o c c u p a tio n ” (B inford 1981 b:20);
(3) see also " fin e-g ra in ed ,” “ g ra in ,” a n d “ p alim p sest”
cop ro co en o sis: a fossil assem blage derived fro m scats a n d owl pellets (M ellet
1974)
co rrasio n : (1) an y co m b in a tio n o f m echanical a b ra sio n , b io ero sio n , and

c o rro sio n /d isso lu tio n (B rett 1990a); (2) see also “ a b ra sio n ,” “ b io e ro
sio n ,” a n d “ c o rro sio n /d isso lu tio n ”
co rro sio n : (1) results fro m chem ical in stab ility o f skeletal m inerals [im plied
n o n -b io lo g ical m echanism ] (B rett 1990a); (2) see also “ c o rra s io n ” an d
“ d isso lu tio n ”
cu ltu ra l filter: h u m a n beh av io rs th a t c o n trib u te to o r resu lt in the fo rm a tio n o f

an a rc h a e o fa u n a l reco rd (a fter D a ly 1969)
d e a th assem blage: (1) “ th e sum to ta l o f corpses w hich arrive u p o n a surface

betw een th e d ep o sitio n o f tw o successive increm ents o f sedim ent, o r
betw een the in cep tio n o f one p erio d o f in cre m en tatio n a n d its te rm in a
tio n ” (C lark et al. 1967:155); (2) see “ life assem blage”
decay: (1) u tiliza tio n o f d ead organism s as a fo o d source by m icrobes (e.g.,

fungi a n d b acteria) (A llison 1990); (2) d eco m p o sitio n o f p ro te in u n d e r
aero b ic co n d itio n s (H a g lu n d 1991); (3) ae ro b ic d e g ra d a tio n o f soft tissue
(M icozzi 1991); (4) see also “ scavenging”
d efo rm atio n , plastic: p o st-d e p o sitio n a l change in the shape o f a b o n e w ith o u t

b reak ag e, usu ally caused by c o m p a c tio n fro m th e w eight o f overlying
sedim ents (S h ipm an 1981 b:200)
diagenesis: (1) “ fossildiagenese” (M uller 1951); (2) stu d y o f the p o st-e n to m b

m en t h istories o f org an ic rem ains (L aw rence 1968); (3) the study o f p o s t
b u ria l ta p h o n o m ic fa cto rs, i.e., betw een b u ria l a n d recovery (L aw rence
1979b); (4) chem ical a n d physical changes th a t o ccur in an im al rem ains as
they alter from th eir original state a n d becom e fossilized (B ehrensm eyer et
al. 1989); (5) “ the chem ical a n d m echanical alte ra tio n s w ithin the sedi
m e n t” (C adee 1990:3); (6) “ a lte ra tio n a fte r dep o sitio n ; som etim es tak en to
m ean a lte ra tio n afte r b u ria l” (R etallack 1990:129); (7) “ all the processes
affecting a sedim ent a n d its c o n ta in e d fossils fro m d ep o sitio n u n til
m e ta m o rp h ism tak es over a t elevated te m p e ra tu re s a n d p re ssu res”
(T u ck er 1990:84); (8) see also “ an a d ia g en esis,” “ a n a ta x ic ,” “ syn d iag en e
sis,” a n d “ ta p h ic ”
dip: “ th e angle o f d eclin atio n o f an object o r p la n e ” (S h ip m an 1981 b:200); syn.

plunge
d isarticu latio n : (1) the generic process a n d result o f loss o f a n a to m ica l integrity
(H ill 1979b); (2) “ th e d isin te g ratio n o f m ultiple-elem ent skeletons along
pre-existing jo in ts o r a rtic u la tio n s ” (B rett 1990a:223); (3) “ d isarticu lated
b u t asso ciated b o n es are sep a rated fro m each o th e r b u t in close p roxim ity,
Glossary o f taphonom y term inology 507
an d they re ta in th eir id en tity as p a rts o f in d iv id u al a n im a ls” (B ehrens

m eyer 1991:302); (4) see also “ d isp ersa l” a n d “ sca tte rin g ”
dispersal: (1) th e generic process a n d result o f sp atial m ovem ent o f individual

skeletal elem ents fro m a single o rg an ism (n o te th a t tw o elem ents m ay
becom e m o re, o r less, spatially co n tig u o u s) (H ill 1979b); (2) “ dispersed
bones m ay be (a) associated, scattered over areas several tim es the size o f
the artic u la te d skeleton, b u t can be related to the sam e individual using
a n a to m ica l ch aracteristics [e.g., a n a to m ica l refitting], or, (b) isolated,
w idely se p a ra te d from o th e r bones belonging to the sam e sk eleto n ”
(B ehrensm eyer 1991:302); (3) see also “ d is a rtic u la tio n ” a n d “ sc a tte rin g ”
dissolu tion : (1) “ ch an ges in chem ical co m p o sitio n [of skeletal tissues] usually
ap p e arin g as p ittin g a n d c o rro sio n o f the skeletal su rfa ce” (P arso n s a n d
B rett 1990:38, 43); (2) see also “ c o rro sio n ” an d “ leach in g ”
distal co m m un ity : (1) the species in a n assem blage w hich lived far fro m the site
o f d ep o sitio n , these species have low er th a n average skeletal com pleteness
in the assem blage (a fter S hotw ell 1955; b u t see G ra y so n 1978b); (2) see
also “ a u to c h th o n o u s ,” “ a llo c h th o n o u s ,” “ p a ra u to c h th o n o u s ,” an d
“ p ro x im al c o m m u n ity ”
d urab ility : “ the relativ e resistance o f skeletons to b re a k d o w n a n d d estru c tio n

by physical, chem ical, a n d biotic ag e n ts” (B rett 1990a:223)
effect: (1) “ the m o d ificatio n o r a lte ra tio n one sees o n a bo n e o r in the bone
assem blage as a w hole as the result o f changes th e bo n e o r assem blage has
u n d erg o n e” (Jo h n so n 1985:159); (2) th e static result o r trace o f a ta p h o n o
m ic process h av in g acted on bones, in cluding the p hysical a n d /o r chem ical
m o d ificatio n o f a b o n e o r bo n e assem blage (see C h a p te r 1); (3) see also
“ a g e n t” a n d “ p ro cess”
elem ent: an an a to m ica lly com plete bo n e o r to o th (a discrete a n a to m ica l o rg an )

in th e skeleto n o f an an im al
enrichm en t: (1) [“ ac c u m u la tio n ” is] “ the d ep o sitio n o f an y sub stan ce [either

o rg an ic o r inorganic] a t a level o r c o n c e n tra tio n h ig h er th a n th a t found
while the b o nes w as in the living s ta te ” (C o o k et al. 1961:360, 358); (2) the
ad d itio n of soluble m atter; (3) see “ m in e ra liz a tio n ” and
“ p e rm in e ra liz a tio n ”
equifm ality: th e p ro p e rty o f allow ing o r hav in g the sam e effect o r resu lt from
different events {W e b ste r’s T hird International Unabridged D ictionary)
erosion: g eneral term fo r the physical o r chem ical rem oval o f original m aterial
from edges a n d /o r surfaces o f an im al rem ains (B ehrensm eyer et al. 1989)
eth n o arch aeo lo g y : stu d y o f living peoples w ith the aim o f elu cid atin g a rc h a e o
logical prob lem s; i.e., a discipline w ith the goal o f establishing a n d
clarifying the relatio n sh ip s betw een m ateria l vestiges o f h u m an beh av io r

a n d th e living system s w hich generate them (G ilfo rd 1977)
fauna: som e specified set o f an im al tax a in close sp atial a n d tem p o ra l

asso ciation; usually qualified by som e geographic, tem p o ra l a n d /o r
tax o n o m ic crite rio n (a fter O d u m 1971:366-367; see also T ed fo rd 1970)
fidelity: th e degree to w hich a fossil fa u n a accu rately reflects a p a leo fau n a, in

term s o f tax a rep resen ted , tax o n o m ic richness, a n d /o r tax o n o m ic d iv er
sity (K idw ell a n d Bosence 1991)
final burial: (1) “ o p e ra tio n a lly defined as b u rial to sufficient d ep th w ithin a

sed im en tary su b stra tu m fo r a h a rd p a rt to b o th a tta in a refuge from
fu rth e r sm all-scale episodes o f e x h u m atio n a n d exposure at the d e p o sitio
n al in terface a n d to escape destru ctiv e early diagenetic processes” (K id-
well 1986:7); (2) an im al rem ains are in terre d below the g ro u n d surface,
encased in sedim ent, a n d n o t exh u m ed /ex p o sed to ab o v e -g ro u n d ta p h o
n om ic processes (B ehrensm eyer et al. 1989); (3) see also “ b u ria l”
fine-grained: (1) “ an assem blage ac cu m u lated over a sh o rt perio d o f tim e . . .

reso lu tio n betw een debris o r b y -p ro d u cts a n d events [is g o o d ]” (B inford
1980:17); (2) an assem blage in w hich “ all the included item s, features, and
la n d surfaces relate to a very few events; th a t is, all asso ciated a rch ae o lo g i
cal ch a rac te ristic s o f the d ep o sit are the consequences o f basically the sam e
ev en ts" (B inford 1981 b:20); (3) see also “ c o a rse -g ra in e d ,” “ g ra in ,” and
“ p alim p sest”
fossil: (1) “ fossils are th e rem ains o r traces o f any recognizable organic
stru ctu re s p reserved since p re h isto ric tim es” (S tirto n 1959:18); (2) any
c o n te m p o ra ry trac e o r rem a in o f a n o rg an ism th a t died a t som e tim e in the
p a st (M a tth ew s 1962); (3) an y specim en d e m o n stra tin g physical evidence
o f th e occurren ce o f an cien t life; generally d istinguished from R ecent or
n o n /su b -fo ssil rem ains on the basis o f its (the fossil’s) geologic m ode o f
o ccu rren ce (S ch o p f 1975); (4) the identifiable rem ains o f (once) living
o rg anism s o r o f th eir activities preserved in the sedim ents by n a tu ra l
processes (F in k s 1979)
fossil assem blage: (1) an aggregate o f in d iv id u al elem ents (th a t in tera ct w ith
v ario u s m o d ificatio n agents in statistical fash io n , w ith co n siderable
p o te n tia l for v a ria tio n in traces they ultim ately m ay bear) (G ilfo rd 1981);
(2) spatially asso ciated fossils th a t becam e em bedded to g eth er after
b io stratin o m ic processes affected them a n d th a t w ere influenced by
d iagenetic processes (F iirsich 1990:237)
fossilization: (1) “ ch an ge is the essence o f the process o f fo ssilizatio n ” (C o o k et

al. 1961:356); (2) the m ain ten an c e o r a lte ra tio n o f chem ical p ro p e rtie s o f
o rg an ic m aterials by n a tu ra l processes (F in k s 1979); (3) see also “ ap o lith -
osis,” “ m in e ra liz a tio n ,” “ p erm in e ra liz a tio n ,” “ p o st-fo ssilizatio n ,” an d

“ p re -fo ssiliz atio n ”
fo ssil-L ag erstatte (sing.; L a g e rsta tte n - pi.): (1) “ any rock [stratum ] co n tain in g
fossils w hich are sufficiently well preserved a n d /o r a b u n d a n t to w a rra n t
ex p lo itatio n , if only fo r scientific p u rp o se s” (S eilacher 1990:266), (a)
co n c e n tra tio n L a g e rsta tte n c o n ta in a b u n d a n t fossils, (b) co n se rv atio n
L ag e rstatten co n tain w ell-preserved fossils (Seilacher 1990); (2) see also
“ o b ru tio n d e p o sits”
fossil record : (1) som e set o f rem ains o f organism s, p la n ts a n d /o r anim als,
h av in g a geological m o d e o f o ccu rren ce in som e defined g eo g rap h ic space
an d th e geological co n tex t a n d sp atial d istrib u tio n s o f th o se fossils
(m odified fro m L y m an 1982a); (2) see also “ local fa u n a ”
frag m en tatio n : m echanical d isasso ciatio n o f skeletal elem ents alo n g n o n

a rtic u la tio n planes o r n o n -jo in t-re la te d planes
grain: (1) “ the relative c o n tex tu a l com plexity o f an assem blage fro m the
perspective o f events o ccu rrin g d u rin g the course o f c o n tin u o u s o cc u p a
tio n a n d d eriv ativ e p ro d u c tio n o f a n arch aeo lo g ical assem blage” (B inford
1981 b:20); (2) increasing coarseness o f grain reduces interassem blage
v ariab ility (B in ford 1980:18); (3) see also “ co a rse-g rain ed ,” “ fine
g ra in e d ,” a n d " p a lim p se st”
hydrolysis: (1) o ften a leaching process involving the co m b in a tio n o f w a te r w ith
o th e r m olecules; (2) see "le a c h in g ”
iso tap h o n o m ic: (1) " h a v in g the sam e ta p h o n o m ic features an d biases” (B eh
rensm eyer 1991:326); (2) a fo rm o f analysis involving the co m p ariso n o f
fossil assem blages from the sam e “ ta p h o n o m ic m o d e ” (B ehrensm eyer and
H o o k 1992:18)
lag deposit: (1) “ a n a c cu m u latio n o f b ones th a t are to o dense to be m oved by

w ind o r w a te r c u rre n t” (S h ip m an 1981 b :2 0 1); (2) “ the heavier, less
[fiuvially] tra n s p o rta b le b o n e s” (B ehrensm eyer 1990:233); (3) see also
“ so rtin g ,” “ V oorh ies G ro u p ,” an d “ w in n o w in g ”
leaching: (1) the rem oval o f soluble m atter; (2) see also “ d isso lu tio n ”
life assem blage: (1) “ th e to ta l n u m b e r o f indiv id u als w ithin a specified area

d u rin g a specified tim e, the term m ay be lim ited to an y tax o n o m ic g ro u p ”
(C lark et al. 1967:155); (2) see “ d e a th assem b lag e”
liptocoenosis: re m n a n t (fossil) assem blage (R olfe a n d B rett 1969)
local faun a: (1) the fa u n a rep resen ted by one o r several geographically,
geologically a n d tax o n o m ically sim ilar fossil sam ples; i.e., m ay be re p re
sented by fossil sam ples fro m a single site o r several sites in close
g eo g rap hic a n d stra tig ra p h ic asso c ia tio n (n o t necessarily rep resen tativ e o f
a bio co eno se, a n d n o t necessarily im plying an y p aleoecological reality)
(T ed fo rd 1970); (2) see also “ fossil re c o rd ”
M A U : th e m in im u m n u m b e r o f an im al u n its necessary to a c co u n t fo r the

specim ens observed, calcu lated as
M N E C-r n u m b er o f tim es e occurs in one skeleton = M A U C
w here M N E is as defined in this glossary a n d e is a p a rtic u la r ca te g o ry o f

skeletal p a rt such as the p ro x im al h a lf o f a fem u r (p o rtio n o f a skeletal
elem ent), a m an d ib le (com plete skeletal elem ent), o r the rib-cage (several
an a to m ica lly related com plete skeletal elem ents)
M G U I: th e m odified general utility index (one o f a series o f indices m e a n t to

m easu re th e fo o d value o r utility o f th e soft tissues associated w ith o r
a tta c h e d to th e different skeletal p a rts in a carcass)
m in eralizatio n : (1) “ the ac cu m u latio n o f in o rg an ic so lu tes” (C o o k et al.

1961:361); (2) th e replacem ent o f the original h a rd p a rts o f an organism
w ith o th e r m inerals; the skeletal tissu e’s chem ical co n stitu en ts are dis
solved a n d rem oved w hile m inerals dissolved in g ro u n d w a te r are sim u lta
n eously d ep o sited in th eir place (M a tth ew s 1962); (3) also k n o w n as
au th ig en ic p re serv atio n o r rep lacem en t (S ch o p f 1975); (4) see also “ apo-
lith o sis,” “ fo ssilizatio n ,” “ p e rm in e ra liz a tio n ,” an d “ p e trifa c tio n ”
M N E : (1) th e m in im u m n u m b e r o f skeletal elem ents necessary to a c c o u n t fo r

(to h ave co n trib u te d ) the specim ens observed; the categ o ry o f skeletal
elem ent som etim es is a com plete skeletal elem ent o r discrete a n a to m ic a l
o rg a n a n d o th e r tim es is a p o rtio n o f a co m p lete skeletal elem ent such as
th e p ro x im al h a lf o f the fem ur; (2) see also “ elem en t”
M N I: th e m in im u m n u m b e r o f (com plete) in d iv id u al an im als necessary to

a c c o u n t fo r (to h av e c o n trib u te d ) the specim ens observed
m u m m ificatio n : “ all n a tu ra l a n d artificial processes th a t b rin g a b o u t p re serv a

tio n o f the b ody o r its p a r ts ” (M icozzi 1991:17)
necrology: (1) alte ra tio n s o f an im al carcasses p rio r to diagenesis (W arm e and

H an tzsch el 1979:5-6); (2) stu d y o f “ the d e a th a n d d eco m p o sitio n o f an
o rg a n ism ” [necrosis, necrobiosis] (C adee 1990:3); (3) see also “ b io strati-
n o m y ” a n d “ p e rth o ta x ic ”
neon to lo g y : (1) th e “ p a le o n to lo g y ” o f living an im als including the “ p aleo e co

logy” o f m o d ern en v iro n m en ts (W arm e a n d H an tzsch el 1979); (2) see also
“ n e o ta p h o n o m y ” a n d “ a c tu o p a le o n to lo g y ”
n eo tap h o n o m y : (1) involves relev an t e x p e rim en tatio n o r o b serv atio n s o n the

c o n d itio n o f m o d e rn v e rte b ra te rem ains in closely defined env iro n m en ts,
designed to test ta p h o n o m ic co njectures a n d to suggest consequences for

p aleoecolo gical in te rp re ta tio n n o t visible in the fossil record such as the
absence o f a ta x o n o r the stru c tu re a n d co m p o sitio n o f a p aleo co m m u n ity
from certain kinds o f fossil rem ains (H ill 1978); (2) the study o f c o n te m
p o ra ry processes o f d e a th , decay, a n d d ep o sitio n o f o rganism s (S hipm an
1981b); (3) see also “ a c tu o p a le o n to lo g y ,” “ n e o n to lo g y ,” an d
“ p a le o ta p h o n o m y ”
N IS P : (1) th e n u m b e r o f identified specim ens in a collection, w here “ identified”

u su ally m ean s identified to ta x o n b u t m ay m ean identified to skeletal
elem ent represen ted ; (2) see also “ specim ens”
o b ru tio n deposits: (1) d ep o sits o f w ell-preserved a n d a rtic u la te d fossils ow ing

th eir p re serv atio n a n d a rtic u la tio n to ra p id b u rial (B rett 1990b:239); (2)
from the L atin obruere, m eaning to sm o th er (S eilacher 1992:116); (3) see
also “ fossil L a g e rsta tte ”
o rien ta tio n : “ the az im u th , o r co m p ass d irectio n , o f the long axis o f a bone, as

m easured on a h o riz o n ta l p la n e ” (S hipm an 1981 b:202)
oryctocoenose: rem ain s th a t w ere fo u n d to g e th e r in an o u tc ro p (L aw rence

1979c)
paleoecology: (1) th e stu d y o f e n v iro n m e n ta l re la tio n s o f fossil organism s

betw een th eir b irth an d d e a th (L aw rence 1968); (2) a discipline focusing on
in terre latio n s w hich o ccu rred in the geologic p a st betw een living o rg a
nism s a n d th eir su rro u n d in g s (L aw rence 1971)
p aleo m o rto lo g y : (1) th e ac cu m u latio n a n d in te rp re ta tio n o f all d a ta bearin g on

the d e a th a n d p o stm o rte m events o f skeletal finds th a t offer som e chance
o f d e m o n stra tin g “ m a n -a n im a l” asso ciatio n (M ehl 1966:27); (2) see also
“ a rc h a e o fa u n a ”
paleo n to lo g ical fauna: (1) “ the m axim um g eo graphic a n d tem p o ral lim its o f a
g ro u p o f o rg an ism s sh arin g a suite o f co m m o n species (as evidenced by the
fossil re c o rd )” (T ed fo rd 1970); (2) fau n as recovered from p aleontological
sites; (3) see also “ a rc h a e o fa u n a ” a n d “ local fa u n a ”
p aleo ta p h o n o m y : (1) o b serv atio n s on fossil assem blages (H ill 1978); (2) study
o f th e fossil reco rd (S hipm an 1981); (3) see also “ n e o ta p h o n o m y ” and
“ ta p h o n o m y ”
palim psest: (1) an ac cu m u latio n o f arch aeo lo g ical m ateria ls “ ch aracterized by

an accretio n al fo rm a tio n a l h isto ry ” (B inford 1980:9); (2) a dep o sit o f
arch aeo lo g ical m ateria ls “ deriving fro m a variety o f events o r actions o f
b o th m an a n d a n im a ls” (B inford 1981 b:9); (3) see also “ co a rse-g rain ed ,”
“ fin e-g rain ed ,” a n d “ g ra in ”
p a ra u to c h th o n o u s : (1) “ rem ains th a t are tra n s p o rte d fro m the d e a th o r discard

site b u t stay w ith in the original [life] h a b ita t” (B ehrensm eyer a n d H o o k
1992:19); (2) “ tra n s p o rt o f rem ains w ithin the b o u n d arie s o f an o rg a n ism ’s
h a b ita t” (B ehrensm eyer an d H o o k 1992:78); (3) see also “ a llo c h th o n o u s,”
“ c h th o n ic ,” a n d “ a u to c h th o n o u s ”
p e d o tu rb a tio n : v ario u s processes o f h o m o g en iza tio n (o r h ap lo id izatio n ),

w hich im pede soil h o rizo n fo rm a tio n ; soil m ixing (W o o d an d Jo h n so n
1978); m ay be m echanical o r chem ical (fa u n a ltu rb a tio n ; flo ra ltu rb atio n ;
c ry o tu rb a tio n ; g ra v itu rb a tio n ; arg illitu rb atio n ; a e ro tu rb a tio n ; a q u a tu r-
b atio n ; c ry sta ltu rb a tio n ; seism itu rb atio n )
p erm in eralizatio n : (1) th e in filtratio n o f m in eral-b earin g so lu tio n s in to pores in
skeletal tissue a n d th e d ep o sitio n o f m inerals in those spaces (C o o k et al.
1961:361; M atth ew s 1962); (2) also kn o w n as cellular p erm in e ralizatio n o r
p etrifa ctio n , results in petrified fossils (S ch o p f 1975); (3) see also “ ap o lith -
osis,” “ fo ssilizatio n ,” “ m in e ra liz a tio n ,” a n d “ p e trifa c tio n ”
p erth o tax ic: (1) “ the facto rs w hich act to d estro y an im al corpses w ithin any
p a rtic u la r en v iro n m e n t” (C lark et al. 1967:155); (2) “ clim ate a n d ex posure
are th e ch ief v ariab les c o n tro llin g a p erth o ta x ic system ” (C lark and
K ietzk e 1967:117); (3) ta p h o n o m ic facto rs w hich o p erate betw een the tim e
o f an o rg a n ism ’s d e a th a n d the tim e o f its b u rial, including b u t n o t lim ited
to scavenging a n d w eathering; (4) see also “ b io s tra tin o m y ” an d
“ n ecro lo g y ”
p erth o tax is: " a d ea th assem blage w ith the an im al corpses in v ario u s stages o f
d e stru c tio n by th e set o f processes n o rm ally o p erativ e u n d e r the e n v iro n
m en t c o n c ern e d ” (C lark et al. 1967:155)
p e rth o ta x y : “ th e m o re o r less o rd erly d e stru c tio n o f a n im a ls’ corpses by the

co m b in ed actio n o f n a tu ra l processes” (C lark et al. 1967:155)
p etrification: (1) extrem e p erm in e ralizatio n (C o o k e ? al. 1961:361); (2)“ cellular

p e rm in e ra liz a tio n ” o r p erm e atio n o f cells a n d in terstices (n o t re p la ce
m en t) by m ineral m atrix a t o r very soon afte r d ep o sitio n (S ch o p f 1975); (3)
see also “ a p o lith o sis,” “ fo ssilizatio n ,” a n d “ p e rm in e ra liz a tio n ”
plunge: see “ d ip ”
post-fo ssilizatio n: (1) refers to the p erio d in the ta p h o n o m ic h isto ry a fte r the
orig in al physical a n d chem ical ch a rac te ristic s o f an im al rem ains h ave been
altered by fo ssilization (B ehrensm eyer et al. 1989); (2) see also “ fossiliza
tio n ” a n d “ p re-fo ssilizatio n ”
p re-fossilization: (1) refers to th e p erio d in the ta p h o n o m ic h isto ry p rio r to

a lte ra tio n o f th e original ch a rac te ristic s by the process o f fossilization
(B ehrensm eyer et al. 1989); (2) see also “ fo ssilizatio n " and
“ p o st-fo ssilizatio n ”
preserv atio n : (1) “ d u rip a tric (h a rd p a rt) p re s e rv a tio n '4; original h a rd p a rts are
p reserved d u e to resistance to o x id a tio n a n d physical d am ag e (S ch o p f
1975); (2) “ au th ig en ic p re se rv a tio n 44; fossil is encased by cem enting
m in erals w hich preserve surface co n fig u ratio n o f o rg an ic p a rts while
in te rn a l o rg a n iz a tio n is lo st o r d eg rad ed (S ch o p f 1975)
p rim a ry deposit: (1) d eposits “ p ro d u c ed by th e p rim a ry c o n c e n tra tio n o f

d isarticu lated v e rte b ra te m aterial; (a) co n d e n sa tio n deposits are form ed
ov er a long p erio d o f tim e by the w ith h o ld in g o f sedim ent fro m a n area
[m o rtality is lo n g term ], (b) m ass m o rta lity d ep o sits are fo rm ed over sh o rt
p erio d s o f tim e by the m ass m o rta lity o f v e rte b ra te o rg a n ism s” (A n tia
1979: 108-109); (2) see " se c o n d a ry d e p o sit”
process: (1) “ th e ‘h o w ’ a b o n e becam e m odified (e.g., scored by a to o th

scratch ing the co rtical surface) a n d the ‘w h a t' th a t caused the m o d ificatio n
(e.g., large ca rn iv o re)” (Jo h n so n 1985:158-159); (2) the d ynam ic ac tio n o f
a ta p h o n o m ic ag en t o n b ones (e.g., d o w n slo p e m ovem ent, gnaw ing, o r
fractu rin g ) (see C h a p te r 1); (3) see also “ a g e n t” a n d “ effect”
p ro x im al co m m unity : (1) the species o f a co m m u n ity w hich lived in close

sp atial p ro x im ity to the site o f the d ep o sitio n o f th eir rem ains, species o f
this co m m u n ity have g re ater th a n average skeletal com pleteness in the
assem blage (S hotw ell 1955; b u t see G ra y so n 1978b); (2) see also “ a u to c h
th o n o u s ,” “ distal c o m m u n ity ,” a n d “ p a ra u to c h th o n o u s ”
p u trifac tio n : (1) b ac te rial b re a k d o w n o f p ro te in u n d e r a n a ero b ic c o n d itio n s

(H a g lu n d 1991); (2) a n a ero b ic d e g ra d a tio n o f soft tissues (M icozzi 1991)
q u a rry site: localized c o n c e n tra tio n s o f fossil bones; v ary greatly in density o f
m ateria ls a n d to ta l volum e; vary in degree o f re p re se n ta tio n o f biocoenose
(Shotw ell 1955)
scattering: the increase in disp ersio n o f skeletal p a rts (H ill 1979b); see also
“ d isp ersa l” a n d “ d isa rtic u la tio n ”
scavenging: (1) the use o f d ead o rganism s as a source o f food by " m a c ro

o rg a n ism s” (A llison 1990); (2) see also “ d ecay ”
scratch m ark : general term fo r lin ear, sh arp ly defined grooves in b o n e surfaces;
usually less th a n 1 m m deep a n d 0.5 m m wide (B ehrensm eyer et al. 1989)
seco n d ary deposit: (1) d ep o sit “ p ro d u c ed by rew o rk in g a n d co n c e n tra tio n

fro m o ld er sedim ents; (a) refers to a p rim a ry v e rte b ra te dep o sit w hich has
been b u ried , diagenetically altered , excavated, frag m en ted , a n d th en
fu rth e r c o n c e n tra te d to fo rm a b o n e-b ed , (b) fo rm ed by the rew orking o f
o ld er v e rte b ra te d ep o sits” (A n tia 1979:108, 111); (2) see also “ p rim ary
d e p o sit”
sig n atu re c rite rio n /p a tte rn : “ a crite rio n th a t is c o n s ta n t a n d u n iq u e a n d th a t

discrim inates one m odifying ag e n t o r set o f agents fro m a n o th e r” (B inford
1981b; G o u ld 1980)
skeleto nizatio n : (1) rem oval o f soft tissue from a carcass to p ro d u c e a skeleton;
(2) see also “ au to ly sis,” “ d ecay ,” “ p u trifa c tio n ,” a n d “ scavenging”
specim en: an arch ae o lo g ic al/p a leo n to lo g ic al p a rt o f a skeleton th a t can consist

o f a co m p lete b o n e o r frag m en t thereof, a com plete to o th o r frag m en t
th ereo f, o r a b o n e (such as the m andible) w ith teeth in it
sorting: (1) th e h y d ro d y n a m ic process o f sep a ratin g b o n es o f skeletons into

th o se read ily m o v ed by w a te r a n d th o se n o t readily m oved by w a te r (after
W olff 1973); (2) “ se p a ra tio n o f b o d y p a rts acco rd in g to [fluvial] tra n sp o rt
ra te s” (B ehrensm eyer 1990:233); (3) see also “ V oorhies G ro u p s ”
sullegic: (1) “ th o se fa cto rs influencing the collecting o f fossils w hich d eterm in e

w h eth er o r n o t a n y p a rtic u la r fossil at th e surface will find its w ay in to a
co llectio n ” (C lark et al. 1967:155; C lark an d G u e n sb e rg 1970:440); (2)
facto rs influencing collections; i.e., w h eth er o r n o t a p a rtic u la r fossil is
collected; includes a re a o f site chosen a n d /o r site chosen, sam pling design
(w here you collect), collection p ro c ed u re s (e.g., h an d -p ic k in g versus
screening [and m esh size] versus flo tatio n ) (M ead o w 1981)
syndiagenesis: (1) “ th e first phase o f diagenesis, the early b u rial phase o f

b ac te rial activity in w hich o rg an ic m a tte r p rovides the n u trie n t for
b ac te rial m etab o lism . . . occurs w ithin th e to p few m eters o f a new ly
fo rm ed sed im en t” (R olfe an d B rett 1969:233); (2) n o t explicitly defined,
b u t used by A n drew s a n d C o o k (1985; see also A ndrew s 1990; C o o k 1986)
a p p a re n tly to signify ta p h o n o m ic processes th a t occu r sim u ltan eo u s w ith
b u rial (see F ig u re 2.8); (3) see also “ d iagenesis,” “ an ad iag en esis,” and
“ b u ria l”
sy n tap h o n o m y : “ the sum o f ta p h o n o m ic h isto ries” o f each specim en in a fossil

assem blage (S eilacher 1992:110)
tap hic: “ fa cto rs d eterm in in g w h e th e r o r n o t a b o n e will be b u rie d ” (C lark et al.

1967:155); (2) th e w hen, w here, how a n d w hy o f b u rial
tap h o co en o sis: (tap h o co en o se) (1) assem blage o f organic m aterials w hich are
b u ried to g eth er (L aw rence 1979c); (2) “ d e a th assem blage, g ro u p in g o f
o rg an ism s fo u n d to g eth er in fossil assem blages” (B rett a n d Speyer
1990:258); (3) “ h a rd p a rts o f organism s w hich becam e em bedded to g eth er
afte r hav ing been subject to b io stratin o m ic processes” (F iirsich 1990:237);
(4) syn. = a llo c h th o n o u s th a n a to c o e n o sis
tapho facies: (1) “ suites o f rock ch aracterized by p a rtic u la r co m b in atio n s o f

p re serv atio n al featu res o f the c o n ta in e d fossils; defined on the basis o f
co n sisten t p re serv atio n al fe a tu re s” (B rett a n d Speyer 1990: 258); (2) see
also “ ta p h o n o m ic m o d e ”
ta p h o n o m ic m ode: (1) “ a re cu rrin g p a tte rn o f p re serv atio n o f org an ic rem ains

in a p a rtic u la r sedim entary co n tex t, acco m p an ied by ch a rac te ristic ta p h o
nom ic fe atu res” (B ehrensm eyer 1988:183); (2) “ a set o f fossil occurrences
th a t result fro m sim ilar physical, chem ical, a n d biological processes;
processes m ay be only b ro a d ly sim ilar, o r m o re specifically defined
acco rd in g to detailed sedim entological a n d ta p h o n o m ic evidence” (B eh
rensm eyer a n d H o o k 1992:15); (3) see also “ tap h o fa cies”
ta p h o n o m ically active zo n e (T A Z ): th e se d im e n t-w a te r in terface in a q u a tic

settings, th e se d im e n t-a ir interface in terrestrial settings, w here the
m ajo rity o f ta p h o n o m ic processes are active (a fter D avies et al. 1989)
ta p h o n o m y : (1) th e science o f the law s o f em b ed d in g o r burial; the stu d y o f the

tra n sitio n , in all details, o f organics fro m th e bio sp h ere in to th e lith o s
p here (E frem o v 1940); (2) the stu d y o f differences betw een a fossil
assem blage a n d th e com m unity(ies) fro m w hich it derived; the n ebulous
region o f co n jectu re c o n stitu tin g h y p o th etical assertio n s a b o u t the causes
o f the observed bias in fossil assem blages (H ill 1978)
th an atic: (1) “ the facto rs su rro u n d in g an a n im a l’s d e a th w hich determ ine

w h eth er o r n o t its b o d y will arrive u p o n th e surface as a m em b er o f a d ea th
assem b lage” (C lark et al. 1967:155); (2) facto rs o r variables p erta in in g to
th e d e a th o f a n organism ; (3) circum stances in ducing d e a th am o n g
in d iv idu als o f a biocoenose
th an ato c o en o se: (1) th e d ea th assem blage derived fro m a biocoenose (b io co e

nose th a n a to c o e n o se ^ fossil assem blage); (2) m ay n o t be fro m o n e bu t
several com m un ities (Shotw ell 1955); (3) o rg an ism s th a t died to g eth er
(L aw rence 1979c); (4) “ in terp re ted as variously biased derivatives o f once
living com m un ities, o r b io co en o ses” (B rett a n d Speyer 1990:258); (5)
“ d e a th assem blage, th e preserved elem ents o f a life assem blage a fte r its
d e a th a n d d ecay ” (F iirsich 1990:237)
tim e-averaged: (1) fossiliferous stra tig ra p h ic units th a t rep resen t extended
perio d s o f tim e a n d m ixing o f o rganism s fro m different h a b ita ts (B ehrens
m eyer 1982); (2) “ a ttritio n a l m o rta lity co m b in ed w ith c o n tin u o u s in p u t o f
som e skeletal rem ains in to the sedim ent over long periods o f tim e”
(B ehrensm eyer 1983:94); (3) tim e av eraging “ refers to the m ixing o f
skeletal elem ents o f n o n -c o n te m p o ra n e o u s p o p u la tio n s o r co m m u n itie s”
(F iirsich 1990:237); (4) “ ta p h o n o m ic tim e a v e rag in g ” co n cern s the fact
th a t all species in a stra tu m , fo r exam ple, m u st be “ reg ard ed as c o n te m
p o ra n e o u s only w ithin the in ferred tim e in terv a l” rep resen ted by the
d ep o sitio n al tim e sp an o f the stra tu m , w hereas “ analy tical tim e av e rag
in g ” results fro m lu m p in g m o re o r less co n te m p o ra n e o u s collections or
assem blages in to one an a ly tical unit (B ehrensm eyer a n d H o o k
1992:75-76)
tra n sp o rt: (1) loss o f fossil in fo rm a tio n by physical m ovem ent o f fossils aw ay
from the site o f th e o riginal biocoenose (a d a p te d from L aw rence 1968); (2)
m o v em en t o f an im al rem ains by the physical ac tio n o f w ater, w ind,
people, o r ca rn iv o res (m odified fro m B ehrensm eyer et al. 1989)
trephic: (1) “ facto rs incident to c u ra tin g a n d identifying a specim en w hich

d eterm in e w h eth er o r n o t a fossil in a collection becom es available for
[analytical] use” (i.e., becom es o r p rovides d a ta ) (C lark et al. 1967:155;
C lark an d G u e n sb u rg 1970:440); (2) includes d eterm in in g w hich bones
th a t w ere recovered have been identified an d recorded (skill o f an aly st),
an aly tic p ro ced u res (sam pling), a n d the published fo rm o f d a ta (M ead o w
1981)
V oorhies G ro u p : (1) a set o f skeletal elem ents gro u p ed on the basis o f th eir
sim ilar p o ten tial for h y draulic tra n s p o rt (a fter S hip m an 1981 b:205); (2)
“ three distin ct g ro u p s for m edium to large m am m als” o f bones based on
th eir p o ten tial fo r fluvial tra n s p o rt (B ehrensm eyer 1990:233); (3) see also
“ lag d ep o sit,” “ w in n o w in g ,” a n d “ s o rtin g ”
w eathering: (1) refers to “ the effects on bo n e o f s a tu ra tio n , d esiccation, an d

te m p e ra tu re ch a n g es” (M iller 1975:217); (2) “ the process by w hich the
orig in al m icrosco pic org an ic a n d in o rg an ic co m p o n en ts o f bo n e are
sep a rated from each o th e r a n d destro y ed by physical a n d chem ical agents
o p eratin g on the b o n e in situ , eith er on the surface o r w ith the soil z o n e ”
(B ehrensm eyer 1978:153), “ p a rt o f the n o rm al process o f n u trie n t recy
cling in a n d on soils” (B ehrensm eyer 1978:150)
w innow ing: (1) selective rem oval o f an im al rem ains fro m a n assem blage due to
c u rre n t (w ater) actio n ; can occu r w hen all bones are in m o tio n o r w hen
som e are s ta tio n a ry (B ehrensm eyer et al. 1989); (2) “ rem oval [by fluvial
processes] o f lighter skeletal elem ents” (B ehrensm eyer 1990:233); (3) see
also “ lag d ep o sit,” “ so rtin g ,” a n d “ V oorhies G ro u p ”
INDEX
Abbey, H. M „ 393 background scatters, 189-191, 504

Abler, W. L„ 144 bacteria, 137-144, 150, 396,405,417-418,421,437
abrasion, 37, 39, 42^13, 86, 174, 185-187, 193, 212, Bailey, G., 57
345-346, 378, 381-384, 405, 410-411, 441, 502 Balcomb, R., 447
accumulation, 13, 18, 22, 32, 61, 162-168, Ballybetagh Bog, 42
189-220,358, 365-366, 370-374 Balme, J., 216
active, 114, 162-166, 190-193,409 Barnosky, A.D., 42^43, 131, 135
fluvial, 163, 166, 193 Bartsiokas. A., 420
mass, 191-192,409 Bass, W. M „ 135
palimpsest, 129, 352.455^156, 511 Behrensmeyer, A. K... 18, 23, 25, 27, 30, 31, 33,
passive, 114, 162-166, 190-192, 409 144, 162-166, 172-174, 176, 180, 185,
acidity, 138-139, 390, 418, 421^123, 435 190-191, 237, 354, 358-375, 380, 382, 393,
see also pH 397-398, 410, 414, 436, 458, 461 462, 465
actualism, 41, 46, 49-64, 503 Berger, W. H „ 115
actualistic paleontology, actuopaleontology, 16, Bertram, J. B„ 284-285, 288
1 8 ,4 6,49,52,503 Bezouce site, 328-331
actualistic research, 2, 17-18, 20-21, 24, 43, 45, bias, 1, 4, 6, 12, 18, 23, 26-27, 31-33, 60, 373
60-63, 69, 112, 142, 146, 153, 205, 315, Bickart, K. J..447
374-375, 445, 447, 453 Biddick, K. A., 322-324, 338
Agenbroad, L. D „ 345 Binford, L. R„ 10, 45, 54-57, 60. 104-110, 150,
Allison. P. A.. 9, 139, 419 188, 219-220, 225-229, 231, 234, 256-257,
allometry, 70-71 266, 274-275, 279, 281, 284-285, 288, 295,
am phibian bone, 78, 450 298-301, 303, 309, 313, 340, 343, 352,
anadiagenesis, 418, 503 399^400, 402
analogs, 15,43,45, 53-56 biogeochemical cycles, 464-465
analogy, 13-14, 24, 41, 43, 49, 53-54, 57, 64-68 biomechanics, 83-86
formal, 58, 66, 69 biostratinom y, 16-17, 22, 33, 419, 505
relational, 58-59, 64, 66-69 bird bones, 14, 77, 87, 89, 446-450
analytical absence, 277, 281-282, 379, 426 birds, as taphonom ic agents, 127, 195-205,
anastrophe, 413, 503 333-338, 358, 379
Anavik site, 227-229, 256-257 bloating. 139-141. 412
Andrews, P.. 10, 193, 199-205, 358, 376, 379, 416, Blumenschine, R. J„ 147, 187-189, 213, 276. 284,
452-453 400
antem ortem , 95-97 Boaz, D., 174, 186,410-411
Antia, D. D. J., 436 bone grain, 86
antler, 72, 80-81, 83, 112, 346 bone shape, 39, 78, 87, 172, 176-177, 379-380, 435
apatite, 72-74, 79, 83, 417,419-420 bone tissue, 71-78, 95
see also hydroxyapatite Boney Springs site, 412
archaeofauna, 4-6, 16, 34, 37 38, 116, 221, 504 Bonnichsen, R., 22, 30, 31, 33
archaeological record. I, 5, 26, 54 56, 59 Bower, J. R. F„ 367-369
argum ent by elimination, 62, 64 Brain, C. K.. 10, 36, 112, 194-195, 198, 235-237,
arguments o f relevance, 55 239, 255, 288, 366, 381-382, 385-387, 390, 420
articulation, 7, 20, 137, 143, 150, 153-154, 170, Bram well, D., 448
410, 504 breakage, 14, 15, 20, 86, 150, 158, 195, 202-204,
Ascenzi, A.. 420 274, 276, 279, 389.410
Ascher. R„ 54 see also fragm entation
assemblage, 8, 35 Brett, D. W .,418, 421
attritional processes, 14 Breuil, H. A., 20
autolysis, 140 Brewer, D. J., I l l
Avebury, Lord (Sir John Lubbock), 15 Briggs, D. E. G „ 9, 139, 419, 425
517
518 Index
Bromage, T. G., 383, 422 conjoining, 154-160, 328, 369, 425

Brongersma-Sanders, M., 413 see also refitting
Brothwell, D. R., 395 context, 218—219, 405
Buckland, W., 13, 15 behavioral, 68, 170, 300
Buikstra, J. E., 387-388, 396, 423 distributional, 7, 22, 174, 362-363
Bunn, H. T„ 103-104, 219-220, 274-275, 298, 303, ecological, 68
307-309, 311, 313, 319, 369-370, 402 geological (stratigraphic, sedimentological), 4,
Burgett, G. R„ 152-154 37, 42, 128, 135, 137, 139, 173, 178, 185, 187,
burial, 1, 3, 16, 20, 36, 61, 96, 137, 144, 162, 190, 216-217, 328, 343, 345, 371-372, 393-394,
370, 375-379, 404-416, 432-433, 505 411,416-418,426-427
burning, 11. 44, 216-217, 351, 384-392, 397, 439, prehistoric, 10, 13, 32, 170
465 spatio-tem poral, 9, 343, 367
see also calcination; charring; cooking; heating see also distribution, spatial
Burr, D. B.. 85 Cook, J., 376
butchering, 6, 23. 33-34, 67, 105, 147, 150, 156, cooking, 216, 295, 301, 384, 388-389, 437
170, 223-225, 294-315, 351 see also burning; calcination; charring; heating
pattern, 296, 301-302, 307, 389 coprolites, scats, 14, 43, 199, 205-206, 280, 437
process, 296 Cornwall, I. W., 23
butchering m arks, 7, 15, 44-45, 132, 169, 217-218, corrasion, 506
296-299, 351, 377, 384, 449, 452 corrected joint frequency, 151-152
location of, 45, 276, 297, 307, 309 corrosion, 195, 198-199, 204-206, 210-211, 280,
orientation of, 45, 297 377, 429, 437,441,506
quantification of, 301-306 see also digestion
types of, 297-299 coyotes, 154, 200, 205-206. 449
see also cut marks; flake scars; percussion marks see also canids; hyenas; wolves
Butler, V. L„ 438-441, 445 crocodilians, 43—14, 59
Butzer, K. W„ 407 crushing, 39, 249, 328, 390, 413, 424-431, 437
Cruz-Uribe, K., 39, 60-62, 103, 214-216, 426-429
Cadee, G. C„ 17, 52 cultural filter, 33-34, 506
Cahen, D., 154 culturally deposited bone, 5, 7, 18, 23, 35, 189,
calcination, 385-389 216-218
see also burning; charring; cooking: heating Currey, J.. 78, 82-83,316
canids, 14-15, 205-206, 210, 235, 277, 411 cut m arks, 296-299, 307-314, 449
see also coyotes; hyenas; wolves
carcass consum ption sequence, 147-150, 187-189 Daly, P., 224-225. 292
Carlson, S. J., 419 D am uth, J., 398
cartilage, 78-79 D art, R., 20, 224,315, 324
Casper site, 151-153 David, B„ 387-388
catastrophism , 47, 49-50 Davies, D. J„ 405
causal links, 45^16, 58-59, 63, 454 da Vinci, L., 17
cause (proximate, ultimate), 464 Davis, K. L„ 317-321, 324
cause-effect, 44-46, 50, 53-58, 60-63, 66, 463 Davis, S. J. M „ 78
see also necessary and sufficient causal relations Dawkins, W. B., 13, 14
Chaplin, R. E„ 23,421 death. 16. 114-135, 363-364
charring, 216-217, 384-385 location of, 162-163, 171-172, 187
see also burning; calcination; cooking; heating see also m ortality
Chom ko, S. A., 135 decay, 16, 23, 140-142, 435, 506
Clark. J.. 17,21,27, 405, 502 decom position, 137, 140-144, 385
climate, 8, 39, 140, 146,418 deform ation, 417, 424^125, 506
see also depositional environment; environment density, frequency per unit of area, 166. 169,
cohort, 116, 132 189-192. 237, 408-409, 424. 436, 443
Colby site, 175-176 density-mediated attrition, 235-258, 260, 268-269,
collagen, 72, 74, 85-86, 217, 325, 358, 360, 383, 273, 278-279, 285, 288-289, 427
417^118,422,437 deposition, 405—413
Colley, S. M ., 33, 435,437 depositional environment, 16, 37, 138-139,
compaction, compression, 72, 79, 85-86, 324, 390, 141-142, 144, 146, 162, 178, 184, 190-191,
424,426 358-359, 361-364, 367, 407, 414, 418, 447
comparative approach, 6, 36, 60-61, 104, 398^402, Derry. D. E.. 393
434, 458 diagenesis, 16, 22, 73, 97, 217, 385, 397-398, 406,
completeness index, 430 414. 506
configurational properties, 47, 50-52, 54-56, 61 diffusion, 421
Index 519
digestion, 8, 43, 195, 198-199, 204-205, 210-212, flake scars, 212, 298-299, 304, 314, 326, 335, 345
215, 280, 377,436-437 see also percussion m arks
see also corrosion FLK Zinjanthropus site, 103-104, 271-272,
dip, 39, 178, 181-182, 506 307-314, 369-374
see also plunge fluvial processes, 20, 31, 37-38, 166, 177-183,
disarticulation, 16, 18, 33, 36-38, 137, 141-147, 185-187, 193, 234-235, 239, 409-111, 435, 437
150-154, 156, 161, 298, 307, 436, 506 see also accumulation, fluvial; transport, fluvial
sequence, 38, 144-150 fluvial transport index, 174-175
dismember(ment), 104, 158, 295, 300-301 focal destruction, 396
see also disarticulation Fontbregoua site, 328-331
dispersal, 32, 36-38, 154-160, 162, 168-189, fossil, 4, 16, 32, 508
219-220, 507 fossil assemblage, 5-6, 8, 10, 26, 32-33, 35-36, 40.
fluvial, 171-176,410 60-62, 508
see also scattering; transport; fluvial fossil fauna, 4, 38
distribution, spatial, 4, 7, 14, 22, 33, 35, 38-39, fossil record, 4-6, 12, 18, 27, 34-36, 38, 60, 63,
367-371,415-416 191,219
see also context fossil-Lagerstatten, 20, 509
Dixon, E. J.. 195 fossildiagenese, 16
Dodson, P., 63, 139, 173, 177, 203 see also diagenesis
Dolbeer, R. A., 447 fossilization, 11, 16, 36-37, 39, 405, 418, 420, 508
Donovan, S. K., 9 see also mineralization; permineralization;
dry bone, 86, 194-195, 198, 316-317, 320, 328, petrifaction
330, 338, 387 Fox, G. L„ 359. 361,366, 375
Duckworth, W. L. H.. 20 fracture, of bones, 315-338
agents of, 20, 37, 186, 193, 197-199, 324-331,
Efremov. I. A.. 1, 12, 16, 30, 31, 32. 33, 52 334,413
Elandsfontein site, 129-131 callus, 74, 96
Elder, R. L„ 436 features of, 318-333
El Juyo Cave site, 427—128 healed, 74
Emerson, A. M., 158 mechanics of, 315-318
empirical generalization, 59, 63, 69, 99 patterns of, 15, 22, 318-324
encrustation, 420, 433 traum atic, 96
enrichment, 420 types of, 318-325, 328-333, 426
environment, 12, 16, 30, 118, 138-139, 142, 146 fragm ent size, 14, 22, 39, 194-195, 197, 206, 217,
see also climate; depositional environment 279, 329, 334-335, 429
enzymes, 140-141 fragm entation, 15, 23, 37, 39, 111,203,217,249,
eolian processes, 187, 381-382, 409, 411 254, 280-281, 379-380, 391, 398, 427-430,509
epidiagenesis, 418 extent of, 333-334
equifinality, 38, 63, 258, 299, 507 intensity of, 333-338
Ericson, P. G. P., 448^149 see also breakage; fragment size; N ISP:M N E
erosion, 185, 408^109, 507 ratio
etching, 210, 381 Frison, G. C., 174-176, 178, 366, 375
ethnoarchaeology, 2, 41, 49, 54, 112, 150, 220, 239, Fritz, J. M., 55
261, 270, 293, 295, 301-302, 307, 314, 419, Frostick, L., 176
454, 507 fungi, 137, 140, 375, 395-396, 418
ethnographic analogy, 15, 53-54, 56, 218
Everson, G. D., 450 G arland, A. N., 96-97, 396
expediency tools, 343-344 Garvin, R. D., 277-278
see also tools Gatecliff Shelter site, 300
experimental research, 15, 20, 23-24, 60-61, 179, Geniesse, S. C., 86
186, 239, 382, 387, 389-391 Gifford, D. P., 26, 46, 61-62, 64, 360-361, 365,
371,374, 455
Gifford-Gonzalez, D „ 64, 67, 300-301, 318, 360,
fauna, 4, 508 378, 453—154, 462
Fernandez-Jalvo, Y., 205 Gilbert, A. S., 383
felids, 14, 206,214,218,450 Gilbert, B. M „ 135
fidelity, 12, 508 Gilchrist, R., 386-387, 391
Finks, R. M „ 424 gnawing, o f bones, 306, 384
Fiorillo, A. R„ 181, 380-381 by carnivores, 11, 14, 21, 22, 32, 129, 206-216,
fish bone, 78, 420, 434-445 < — 235, 272, 276-279, 282, 325-326, 328, 351,
Fisher, D. C., 43^15, 59, 62, 350-353 429, 452
520 Index
gnawing, of bones, (cont.) insects, 135, 137, 139-144, 150, 193, 393-394, 404,
by rodents, 32, 194-195, 252 437
see also tooth m arks invertebrate taphonom y, 3, 17, 405
G oodm an, N., 48 Irish elk, 42^13. 131, 135
G ordon, B. C.. 395
G ordon, C. C., 423 James, S. R.. 384, 388
Gould, R. A., 56 Jodry, M. A., 394
Gould, S. J., 48, 52 Johnson. D.. 141-142
gradualism, 47 Johnson. D. L., 405. 432
G raham , R. W., 5, 187, 447 Johnson, E„ 86, 155-156, 316, 318-320, 324, 346,
Grayson, D. K., 9, 13, 216-217, 260-262, 375, 377 385, 389
green, fresh bone, 297, 316, 320, 324, 326, 328, joints, 36, 38. 45, 142-146. 152-154, 160, 307,
330, 338, 351, 366, 382, 387 309-313
Guilday, J. E„ 21,288, 298 Jones, A. K. G „ 435, 437^138
Guthrie, R. D„ 135, 235 Jones, K. T., 265, 271-272
Hackett, C. J., 395-396 Kay, M ., 5, 187

Haglund, W. D.. 140 Kelley, J. H „ 67
Hanen, M. P.. 67 Kent, S., 212, 277
H annus, L. A., 422 Kidwell, S. M „ 18, 25, 27, 30, 31, 33, 4 0 8 ^ 0 9 , 423
Hanson, C. B.. 410 Kietzke. K. K.. 405
Hanson, D. B.. 396 Kiszely, I.. 385-386
Hare, P. E.. 422 Kitching. J. M., 394
Hargrave, L. L.. 446 Kitts, D. B„ 48-49
Haynes, G „ 10, 146-147, 191-192, 212-213, 297, Klasies River M outh Cave, 45, 129-131, 261
345, 352, 379 Klein, R. G., 39, 60-62, 103, 121, 124, 128-131,
Hays, C., 193-194 270, 426-431
heating, 360, 384, 422, 437 Knight, J. A., 389-391,422
see also burning; calcination; charring; cooking Koch, C. P., 16, 25, 30, 33
Hefti, E„ 394 Koike, H.. 117
Henderson, J., 425 Kooym an, B., 303
Hesse, B.. 27, 363 Korth. W. W„ 63. 173. 185-186, 410
Hill, A. P., 26, 38, 39, 60, 112, 144-147, 151-152, Kranz. P. M ..413
154, 162, 320 Kreutzer. L. A., 179-180, 288, 441
Hiscock. P., 33 Kroll, E. M.. 103-104, 303, 307-309, 311, 313,
Hockett, B. S., 127, 194, 198 369-370
Hodder, I., 58 krotovina. 7, 218, 412
Hoffman, R., 63, 193-194, 198, 200, 202-203 Kummel, B . 21
Hook, R. W„ 414, 465 Kurten. B„ 122. 135, 324
horn, 80-81 Kusmer, K. D „ 63, 203
H orner II site, 151-153, 157-159
Horwitz, L. K., 288-289 La Brea tar pit site, 191, 412
H ot Springs M am m oth Site, 345, 412 lag deposit, 172, 411, 509
Hudson, J., 127 Lam, Y. M., 362
Hughes, S. S., 414 Lartet, E., 13, 15
Hulbert. R. C„ Jr., 122 Lawrence, D. R.. 17
hunting, 128-129 Lay, D. W.. 447
H utton, J., 47 leaching, 420-422, 426, 509
H utton, J. T., 421 Levine, M. A.. 169
hydrolysis, 141. 417, 420, 509 life table, 116-117
hydroxyapatite, 72, 79, 386, 395 Linz, G. M „ 447
see also apatite Livingston, S. D., 448-449
hyenas, 3, 13-14, 50, 190, 206-216, 267, 270, 273, Lubbock Lake site, 179-180
275, 405, 429 Lucas. J.. 419
see also coyotes; wolves Lyell, C„ 13,47
Lyman, R. L„ 117, 127, 137, 144, 171, 183-184,
immanent properties, 47, 50-53, 55-56, 60-61 237-238, 252-254, 265-266, 280-281.
inclusions, 396 288-289, 294. 309, 314, 325, 346, 359,
index of fragment disjunction, 156-160, 168 361-362, 366, 375. 413, 426, 441
index of skeletal disjunction, 156-160, 168
Ingsted site, 257 M acgregor, A., 340
Index 521
Mahieu. E., 328-331. 334. 425 natural traps, 135, 191-192

M aguire, J. M., 195,212, 325 naturally deposited bone, 5, 7, 23, 35, 44, 189,
M akapansgat, 315 216-218. 443
M arean, C. W.. 103-104, 267, 270-273, 275, necessary and sufficient causal relations, 50, 62,
288-292, 429-431 318
Marshall, F„ 281 necrology, 33, 510
Marshall, L. G „ 318, 384 neotaphonom y, 2, 36, 49, 270, 415, 453, 510
Martill, D .M .. 384, 417,435 Nimmo, B. W „ 124
M artin. H.. 20 nonprogression, 47
M A U, 104-110, 219, 227, 254, 258-260, 266, 270, NISP, 100-102, 111,201,255, 274, 281,304,
285, 291,440,510 310-311, 331, 391,402, 460-461, 511
Mayer, D., 300 N ISP:M N E ratio, 101-102, 214, 274, 281-283,
McGrew, P. O., 436 336-338, 427
Meadow, R. H., 27
mechanical alteration, 36-39 O'Brien, M .J., 281,426
mechanical properties, 82-87, 315, 4 6 3 ^ 6 4 obrution deposits, 20, 511
Medlock, R. C„ 23 Ohtaishi, N.. 117
Mellet, J. S., 63 Old Crow locality, 21
Metcalfe. D„ 265. 271-272 Olduvai Gorge, 306-314, 347, 369-374
Meyer, C. A., 450 Oliver, J. S., 192, 345,447
Micozzi, M. S., 78, 138, 141, 144, 146 Olsen, S. L„ 168, 297, 346, 378, 380
middle-range research, 2, 41, 55 Olsen-Chubbock site, 6, 184, 399
middle-range theory, 64 Olson, E. C., 17
M iddleton, A. P., 420 ontogeny, 22, 50, 70-73, 79, 112, 116, 132,
Miller. G. J.. 354 154-155, 169, 238, 288, 390, 423
m ineralization, 85, 217, 338, 417, 420, 510 orientation, 511
see also fossilization; permineralization; of bones, 18, 27, 39, 177-183,411-413
petrifaction of carcass(es), 139, 183-185
M N E, 102-104, 111, 201, 203, 227, 239, 251, 255, see also position, carcass
260, 266, 270-272, 274-275, 281, 291, 400, 510 polarity, 180-181, 184
M N I, 100-102, 104-110, 153, 199-200, 236, 239, preferred. 178, 180-181,413,437
249, 255, 281.427-429,510 Ortner, D. J., 96, 385, 422
m onitoring perspective, 162-163, 168, 300-301 osteodontokeratic, 20-21
Monks, G. G., 132 osteophagia, 394-395
Moore. K. L„ 143, 146 otoliths, 82, 4 3 5 ^ 3 7
M orlan, R., 212, 298, 320, 375 overburden weight, 39, 249, 324, 328, 423^131
M orlot, A., 14 Owens Ferry site, 303
Morry Dog Yard, 285-287 owl pellets, 198-199, 205-206, 358, 379
mortality. 45, 114, 137, 350-351, 364, 437
attritional, 43, 118-127, 160, 163, 215, 412, 438, paleobiology, 5, 12, 20, 27, 31
504 paleoecology, 2-8, 12, 18, 26, 33, 38, 46, 121, 436,
catastrophic, 118-127, 132-135, 160, 163, 511
168-169, 171, 215, 324, 412, 436, 438, 446, paleontological fauna, 4, 6, 34
505 Parker, A. J., 449
demographics of, 116-118, 126-132 Pate, F. D.. 421
mass, 118, 120, 164, 191,436, 438 Payne, J. A., 135, 140, 142
see also death Peale, T. R.. 15,21
mortality pattern, 45, 116, 127 Pei, W. C„ 20
mortality profile, 45, 118-132, 135,215 percentage of potential articulations, 153
m ortality, season of, 132-135, 160 percentage of surviving articulations, 153-154
see also seasonality percussion m arks, 213, 275-276, 279, 282, 299,
M ount St. Helens, 125-126 326, 335, 429
Muller, A. H „ 16, 17 see also flake scars
mummification, 138, 141, 510 perim ortem , 95-97
mummy, 138 Perkins, D „ 224-225, 292
Muniz, A. M., 438 perm ineralization, 420, 512
M unson, P. J., 54-55 see also fossilization; mineralization
Miinzel, S., 323 petrifaction, 420, 512
M urray, T„ 57, 60, 67 see also fossilization; mineralization
Myers, T., 380 pH , 390,417, 421-423,435
Mytum, H. C , 386-387, 391 see also acidity
522 In d ex
Piepenbrink, H., 395-396 Sadek-Kooros, H., 382

Pilgram, T., 281 Salishan Mesa site, 254
plausibility considerations, 55 Salmon, W. C„ 49
plunge, 178, 181-182, 412-413 sample population, 54, 111-112
see also dip sampling, 4-5, 11, 18, 38, 54, 124. 161, 223-224
porosity, 417-418, 4 2 1 4 2 2 , 424 see also representative sample
o f bone, 39, 72, 76, 85, 237, 239, 397, 417 Sanger, D., 22
o f enamel, 79 saponification. 141
position, carcass, 162, 446 Sarrians site, 328-331
passive, 139 saturated weight index. 175-176
contorted, 139 Saunders, J. J., 352, 354, 362, 375, 380
postm ortem , 3, 4, 95-97, 137, 142, 184 scales
pot polish, 382-383 am phibian, 81
Potts, R., 298, 320, 369-374 fish, 81-82. 435
preservation, 1, 16, 23, 27, 30, 513 reptilian, 81
Prevot, L. E., 419 scattering, 144, 150, 154-162, 168-189, 513
proboscideans, 80, 146, 174-176, 178, 299, 345 see also dispersal
m am m oth, 14, 175-176, 345, 412 scavengers, 8, 23, 144, 147, 162, 187, 397, 405, 447
m astodon, 34-35, 350-353, 362, 380, 412 scavenging, 8, 42, 128, 140, 224, 307, 366, 447-448.
Prolonged D rift site, 307 452,513
proportional completeness index, 254 Schafer, W., 17, 18, 137, 139, 361, 411, 435, 446
provenience, 154-158, 177 Schiffer, M . B . 54
proximal comm unity, 18 Schild, R., 154
puncture, 197, 206, 212-214, 325, 394 schlepp effect, 224
putrefaction, 138, 140-141, 513 Schneider, J. S., 450
Putschar, W. G. J., 96 seasonality, 121, 124, 132-135, 146, 160, 288-289,
346, 351,438
quantification o f skeletal parts, 97-112 seasonality profile, 132
quantitative measures, 6, 97-100 sedimentation, 368, 405
quantitative units, 9, 70, 98-112, 263-275 rate, 137,405-409,411
see also M AU; M NE; M NI; NISP type, 405-407, 409-41 1
sediments, 173, 186-187, 193, 378-379, 382,
Rapson, D. J., 270, 303, 399-402 405^107, 417, 422, 424-425, 435
Raup, D., 21 Selvaggio, M., 213. 276
Read, C. E., 266 Semenov, S., 340
refitting, 154-160, 168, 171,426, 432-433 settling velocity, 173-174
anatom ical, 155, 171, 363 Shea, J. H „ 48
mechanical, 155, 380 shell
see also conjoining egg, 77, 82
refutation strategy, 57, 67 turtle, 78
Reid, I . 176 Shield T rap Cave, 192
Rensberger, J. M., 205 Shipman, P., 60-61, 63, 178, 180-181, 186, 193,
representative sample, 5-6, 23, 54, 125, 201, 445 206, 210, 237, 297, 307, 318, 344-347, 360,
see also sampling 378, 380, 382, 385-387
reptilian bone, 77-78, 87, 89, 420, 450 Shotwell, J. A., 18, 37
Retallack, G., 398, 404, 409, 424 shrinkage, 86, 139, 316, 366, 386-387
Rich, P. V., 446 447, 449 signature criterion, pattern, 55-56, 60, 69, 514
Richardson, P. R. K.. 214, 398 Sillen, A., 386-387, 390, 418-120
Richter, J., 437^t38 simplicity, 48
Richter, R., 16, 46, 52 Simpson, G. G., 4 7 4 8 , 51
rigor mortis, 139 Simpson, T., 124, 135
rodents (as taphonom ic agents), 37, 193-195, 366 size biasing. 191, 289, 397-398
see also gnawing, rodent skeletal completeness, 7, 140, 218, 438 4 4 1
Rogers, R. R., 394 skeletal element, 87, 95, 100-101, 265
Rolfe, W. D. I., 418,421 appendicular, 87, 147
root etching, 306, 375-377 axial, 87, 147
Rose, J., 186, 344-346, 382 skeletal part frequencies, 6, 20, 102, 111, 188, 190,
rose diagram, 177-181 199-201,219, 347-350,411
Rosene, W., Jr., 447 skeletal part profile, 223
Rulland Dog Y ard, 285-287 skeletal portion, 266-267
Russell, M. D., 295 skeletal tissue, see bone tissue
Russell, M. P., 26 skeletonization, 114, 135-150, 160-162
Index 523
Smith. B. D., 55, 127 338-352, 366, 382

Smith, G. E„ 393 see also expediency tools
Smith, G. R., 436, 438,441 tooth marks, 13-14, 42, 206-216, 275-277, 282,
Smith, P., 288-289 297-299, 307, 326, 335, 437
Solomon, S., 52, 195 see also gnawing
Solutre, 168-169 T oots, H., 18, 177, 181
sorting, 27, 172, 177, 514 torsion, 316-317, 320
specimen, skeletal, 100-101, 265, 514 Tournal, M., 13
Spencer, L. M „ 103-104, 267, 270-273 tram pling, 42, 162, 187, 297, 328, 377-382, 398,
sphericity, 176-177 404-406, 426, 436
Stahl, P .W ., 381 transport, 8, 16, 162, 219, 516
staining, 217, 396-397, 421 fluvial, 20, 37, 87, 162, 171-177, 185, 187, 252,
Stallibrass, S., 205, 450 366, 411
Stanford, D., 352, 394 see also accumulation; fluvial; fluvial
Stein, J. K., 367 processes
Stenson, N „ 17 hum an, 37, 104-105, 158, 162, 224-234, 252,
stereographic projection, 177, 181-183 256-257, 269, 295,299
Stewart, K. M., 443-445 see also utility indices
Stiner, M . C ., 45-46, 64, 112, 127 131,216, 266, Tucker, M. E., 419
273-274, 280-281,461 turbation, 39, 432-433
strain, 83, 315 Turner, A., I l l , 284
Straus, L. G., 404, 414
stress, 83, 315 U erpm ann, H. P., 283-287
structural density (g/cm3), 14-15, 63, 72, 79, uniformitarianism , 10, 16, 41, 46-49, 52, 56, 60
85-86, 110, 169, 172, 215, 234-258, 293, 349, methodological, 47^19, 52, 56-57, 61, 436
385, 397,411,421,423,441,449 substantive, 47, 50, 52, 54-57, 61, 436
bulk density, 72, 166, 237, 239 utility (economic) indices, 110, 158, 225-234, 252,
true density, 237 256, 258-259, 293
sublimation, 138 see also transport indices
survivorship, o f skeletal parts, 104, 112, 235-239,
249-258 Van Neer, W., 438
Sutcliffe, A. J., 270, 394-395 Vehik, S. C., 217
syndiagenesis, 418, 514 Villa, P., 155-156, 328-331, 334, 345, 425
Von Endt, D. W„ 385, 422
taphofacies, 20, 515 Voorhies, M „ 17, 21, 63, 137, 172, 180, 235, 410,
taphonom ic 413
agent, 3, 20, 33-34, 44, 67-68, 82, 503 Voorhies groups, 172, 193, 516
effect, 3, 33, 35-36, 38, 40, 64, 507
history, 3-7, 10, 23, 27, 35^tt), 46, 53, 453 Walker, A „ 63, 193
pathway, 3, 35-39 Walker, M. J., 57, 60, 67
process, 3, 27, 33, 35, 38^10, 44, 63-64, 80, 82, W apnish, P., 27, 363
513 W atson, J. A. L„ 393
strength, 82 W atson, R. N .,48, 53
trace, 3, 27, 35, 68 weathering, 32, 37, 43, 80, 86, 144, 217, 252, 306,
taphonom ically active zone, 405, 515 317, 320, 328, 354-375, 380, 382, 384, 405,
taphonom y, multivariate, 4 5 3 ^ 6 3 413,433,436, 439,452, 516
Tappen, N. C., 86 profile, 364-365, 374, 409, 436
target population, 111-112 rate, 358, 360-361, 363, 398
taxonomic abundances, 6, 111, 444 stages. 354, 360
teeth, tooth, 14, 79-80, 83, 112, 203-205, 358, 419, Weigelt, J„ 16, 17, 52, 115, 137, 139, 410, 446, 450
426, 436 Wexlar, D.. 63
tem perature, 39, 138-141, 146, 354, 385-386, W heat, J. B , 6, 170, 184
388-389, 391, 417-418, 437 Wheeler, A., 435, 437^138
tensile (tension), 72, 85-86, 316-317, 319, 360 White, E. M., 422
Thom as, D. H„ 150, 158, 216, 218, 300 White, J. A., 192
Thomas, R. D. K., 9 White M ountains site, 253-254
time averaged, 164, 166, 515 White, T. D., 10, 377, 382-383, 433
Tobin, M. E.. 447 White, T. E„ 104-110, 223-224, 279, 292, 348
Todd, L. C„ 146, 152, 155, 174-176, 178, 361-362, W hitmer, A. M „ 421
364, 375, 399-402 Wilson, M. V. H „ 33
Tomenchuk. J„ 322-324, 338 winnowing, 168, 172, 174, 516
tools, 86-87, 217-218, 224, 252, 279-280, 315, winterkill, 42
524 In d ex
Wolff. R. G., 172

Yellen, J. E„ 300, 366, 379-380, 415, 454, 462
wolves. 127, 147. 206, 399
see also coyotes; hyenas
Wood. W. R„ 107, 405, 432 Ziedler, J. A., 381
Wylie, A., 60, 66-67 zooarchaeological record, 1, 2, 23, 26, 27
W yman, J., 14

(Cambridge Manuals in Archaeology) R Lee Lyman-Vertebrate Taphonomy-Cambridge University Press (1994)

Uploaded by

Copyright:

Available Formats

(Cambridge Manuals in Archaeology) R Lee Lyman-Vertebrate Taphonomy-Cambridge University Press (1994)

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

(Cambridge Manuals in Archaeology) R Lee Lyman-Vertebrate Taphonomy-Cambridge University Press (1994)

Uploaded by

Copyright:

Available Formats

CAMBRIDGE MANUALS IN ARCHAEOLOGY

D on B rothw ell, University o f London

J.D . R ichards an d N .S. R yan, D ata processing in archaeology

Cambridge M anuals in Archaeology are reference h an d b o o k s designed for an in tern atio n al

© C am bridge U niversity Press 1994

First published 1994

P rinted in G reat B ritain at the U niversity Press, C am bridge

A catalogue record fo r this book is available fr o m the British Library

Library o f Congress cataloguing in publication data

ISBN 0 521 452155 h ard b ack

4 STR U C TU R E AND Q U A N TIFIC A TIO N OF

P ro p erties o f skeletal tissues a n d ta p h o n o m y 82

7 FREQ UENCIES OF SKELETAL PARTS 223

8 BUTCHERING. BONE FRA CTU RING, AND

B utchering, break ag e, a n d b o n e to o ls 350

9 OTHER BIOSTRATINOMIC FACTORS 354

12 T A P H O N O M Y OF FISH, BIRDS, REPTILES,

13 DISCUSSION AND CONCLUSIONS 452

M o deled re latio n o f stress a n d strain, Y o u n g ’s m o d u lu s o f

5.15. O rd e r o f jo in t d isarticu latio n at C asp er a n d H o rn e r II as

6.15. R o d e n t gnaw ed bones. 196-7

fro m the S alishan M esa site ag a in st bo n e m in eral density

diaphysis frag m en ts fo r tw o assem blages o f deer bones. 335

12.1. P ro p o rtio n a l frequencies o f salm o n id c ran ial an d post-

K inds o f ta p h o n o m ic d a ta th a t should be recorded for page

6.7. F luvial tra n s p o rt index values a n d sa tu ra te d w eight index

C arcass resources ex p lo itab le by a fa u n al p ro cesso r o r

W hen I sta rte d m y studies o f v e rte b ra te fa u n al rem ains recovered from

this b o o k has is a trib u te to all o f those people w ho know ingly a n d u n k n o w ­

a n d k ep t m e u p -to -d a te w ith w h at w as com ing off the presses. M ichael B.

F o r perm ission to re p ro d u ce figures, I th a n k the follow ing:

C h a p te r 2 review s th e h isto ry o f ta p h o n o m ic research an d assesses its cu rren t

On the analysis o f archaeological faunal remains

analysis a n d in te rp re ta tio n (R a c k h a m 1983; e.g., L ym an 1992b). A nalyses o f

the physical a n d /o r chem ical m o d ificatio n o f a bone. As we will see, ta p h o n o ­

h u m an s h ad a role in the ta p h o n o m ic h isto ry o f a p a rtic u la r fossil assem blage.

Goals o f taphonomic analysis in zooarchaeology

studies, b u t [it is] a m eans to fa c to r o u t biases a n d fu rth e r o u r fu n d a m e n tal

The challenge o f taphonomy

U n d e rsta n d in g how ta p h o n o m ic processes affect q u a n tita tiv e m easures o f

elem ents are all affected by ta p h o n o m ic processes (B adgley 1986a; G ra y so n

Taphonomv’s contribution to zooarchaeology

sam p le o f an im al rem ain s, a n d n o t d ietary b re a d th . Inferences th a t h u m an s

Terminology used in this book

What this book is and what it is not

This b o o k is m ean t to review m an y o f the p o ten tially useful a n d in fo rm ativ e

carn iv o re chew ed a n d som e as b u rn e d , a n d it is such analytical m a n ip u la tio n

T ap h o n o m y is now in its salad days.

A brief history o f taphonomic research

h isto rical la n d m a rk o f a discipline’s d ev elopm ent) w ith zo o arch aeo lo g ical

N ineteenth-century zooarchaeological taphonom y

G ra y so n (1986) d o cu m en te d th a t research n o t unlike ta p h o n o m y has a deep

T he significance o f this o b serv atio n w as n o t lost o n m an y p re h isto ria n s

traces o f gn aw in g on b ones signifying hyenas in p a rtic u la r to signifying

N o te th a t T o u rn a l ex ten ded the an alo g y by lo o k in g n o t ju st at the gnaw ing

B u ck lan d ’s a p p ro a c h , involving ex a m in a tio n o f the a ttrib u te s o f m o d ern

cal reco rd to th e tw en tieth -c en tu ry focus o n establishing the te m p o ra l relations

T aphonom y in the early and m iddle tw entieth century

Figure 2.1. G eneral relations o f the subdisciplines o f tap h o n o m y relative to an

T h eir coalescence in th e 1970s w as a m a jo r historical event in the developm ent

this b o o k has is a trib u te to all o f those people w ho know ingly a n d u n k n o w

the physical a n d /o r chem ical m o d ificatio n o f a bone. As we will see, ta p h o n o

elem ents to display p a tte rn e d o rien ta tio n s). T he positive aspects o f ta p h o n o

resulted in d etailed co m p arativ e analyses o f m a sto d o n rem ains clearly asso

conclu sion s sim ply “ m easures the increase in o u r u n d e rsta n d in g o f ta p h o n o

T h e exam ples review ed sh are several featu res o f the fo u n d a tio n s o f ta p h o n o

G iffo rd (1981) o utlin es the co m m o n p ercep tio n o f the th eo retical a n d m e th o d

used to stu d y p a st p h e n o m en a , a n d co n fu sio n o f su b stan tiv e a n d m e th o d o lo g i