دوموز 2

Emerging Complexity on the Kahramanmaraş Plain, Turkey: The Domuztepe Project,
1995-1997
Author(s): Stuart Campbell, Elizabeth Carter, Elizabeth Healey, Seona Anderson, Amanda
Kennedy and Sarah Whitcher
Source: American Journal of Archaeology , Jul., 1999, Vol. 103, No. 3 (Jul., 1999), pp.
395-418
Published by: Archaeological Institute of America
Stable URL: http://www.jstor.com/stable/506968
REFERENCES
Linked references are available on JSTOR for this article:
http://www.jstor.com/stable/506968?seq=1&cid=pdf-
reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
Archaeological Institute of America is collaborating with JSTOR to digitize, preserve and

extend access to American Journal of Archaeology
This content downloaded from

94.184.169.35 on Sun, 06 Dec 2020 06:02:23 UTC
All use subject to https://about.jstor.org/terms
Emerging Complexity on the Kahramanmara
Plain, Turkey: The Domuztepe Project, 1995-1997
STUART CAMPBELL, ELIZABETH CARTER, ELIZABETH HEALEY,
SEONA ANDERSON, AMANDA KENNEDY, AND SARAH WHITCHER
Abstract have their roots in the Late Neolithic and Chalcolithic

The fifth millennium is a key period in the develop-of the fifth millennium. The fifth-millennium develop-
ment of complex societies in the Near East. Domuztepe, ments themselves have antecedents going back to at
situated in southeastern Turkey on the northwestern
least the seventh millennium, suggesting not only a
edge of the traditional heartlands of the Halaf, is one of
continuous spectrum of social development but also
the largest sites known from this period. The investiga-
that a division between the traditional Neolithic and
tion of this large (20 ha), central site is providing new
details of the organization of society at the site andUrban
its Revolutions may be misleading. Indeed, the
relationship with the surrounding environment. The
early appearance of large settlements very soon after
settlement seems to have been a focus of long-distance
exchange, with evidence for the manufacture and ma- the emergence of agriculture is difficult to fit into the
classic schemes of social evolution2 that still underlie
nipulation of status items. Stamp seals occur remarkably
frequently and ceramics seem to have been used in a of our explanatory framework for this period.
much
complex way, indicating shifting external relations overIt has been clear for some time that there was an
time. There is also evidence for economic intensifica-
increased degree of regional integration in northern
tion, notably the possible use of secondary products.*
Mesopotamia just before 5000 B.C., with the appear-
INTRODUCTION
ance of what is conventionally called the Halaf.3 A
In the Near East, the sixth and fifthconsistent
millennia B.C.
style of painted pottery is found over an
may have been understudied in the area past from
yetnorthern
they en- Iraq almost to the Mediterra-
compass a series of key social transitions.' The period
nean, accompanied by distinctive circular buildings,
opens with the developed Neolithic and the
figurine types,appear-
and stamp seals. The Halaf is gener-
ally recognized as
ance of sites of surprising size and complexity, lasting
such as in this area until the mid-
fifth millennium
'Ain Ghazal, (atalh6yiik, and Abu Hureyra. It ends whenon it is replaced by the Ubaid,
linking however,
the verge of urban society. We know little, northern and
ofsouthern Mesopotamia into a
the intervening processes. It seems single cultural
possible, unit.in-
and The social models we should in-
deed increasingly probable, that many
ferof
arethe perceived
less clear, but we might look to the develop-
innovations of urbanism in the fourth millennium B.C. ment of larger, regionally based, social groupings.4
* We wish to express our appreciation to the Turkish 'All dates are uncalibrated B.C.
Ministry of Culture's Directorate-General of Monuments ' E.g., E. Service, Primitive Social Organization: An Evolu-
and Museums for permission to conduct this research. Wetionary Perspective (New York 1962); M.H. Fried, The Evolution
are equally indebted to the directors and staff of the Kah-
of Political Society (New York 1967).
ramanmara? Museum and the museum representatives, 3 E.g., S. Campbell, "Problems of Definition: The Ori-
who have greatly contributed to the success of this gins of the Halaf in North Iraq," in M. Lebeau ed., About
project. We are also grateful for help provided by the di-Subartu: Studies Devoted to Upper Mesopotamia (Brussels
rector and staff of the Gaziantep Museum. During the1998) 39-52; P.M.M.G. Akkermans, Villages in the Steppe:
1995-1997 seasons, funding was generously provided byLater Neolithic Settlement and Subsistence in the Balikh Valley,
the British Academy, the University of Manchester, the So- Northern Syria (Ann Arbor 1993).
ciety of Antiquaries, the British Institute of Archaeology at 4 Cf. S. LeBlanc and P.J. Watson, "A Comparative Statis-
Ankara; the National Geographic Society, the Wenner-tical Analysis of Painted Pottery from Seven Halafian
Gren Foundation for Anthropological Research, the Uni-Sites," Paleorient 1 (1973) 119-36; T.E. Davidson, Regional
versity of California, Los Angeles, the American Philo-Variation within the Halaf Ceramic Tradition (Diss. Univ. of
sophical Society, the American Research Institute in Tur-Edinburgh 1977); L. Copeland and F. Hours, "L'expan-
key, and various private donors. The people who have sion de la civilisation halafienne: Une interpretation de la
contributed most, however, are the teams of archaeolo- repartition des sites," inJ.-L. Huot ed., Prehistoire de la Meso-
gists that have worked on the project during these years.potamie (Paris 1987) 209-20; S. Campbell, Culture, Chronol-
Without their dedicated hard work and skill we would
ogy and Change in the Later Neolithic of North Mesopotamia
have nothing to report on here. (Diss. Univ. of Edinburgh 1992).
395
American Journal of Archaeology 103 (1999) 395-418

94.184.169.35 on Sun, 06 Dec 2020 06:02:23 UTC
396 STUART CAMPBELL ET AL. [AJA 103
Domuztepe Kazane Takyan

Mersin Sakqe G6zu o
e Tell Turu Tell Sabi Abyab o Tell Aqab0

The Amuq A h
Yarim Tepe
Shams ed-Din
* More than 1Oha

o Less than lOha
Fig. 1. Distribution of selected sites with Halaf affinities
than 2-3 ha, and settlement hierarchies have been

There are signs that different redistributive systems
were developing, at least for some items of long- difficult to detect. It may be possible to accommo-
distance exchange such as obsidian.5 We have also date the range of evidence for complexity and inte-
known since the 1930s that seals and sealings, with gration in a model where social and economic
their implications concerning resource access and changes are not reflected in site size. Alternatively,
we may infer the existence of larger central sites.
control over property, were already in use.6 The date
of this development has been pushed back, and theJ. Oates, for example, postulated the existence of
complexity of sealing practices has been demon- "bridgehead" sites, largely buried under later occu-
strated by the spectacular discoveries in the burnt pation, at key locations such as Nineveh, Erbil, and
level 6 of Tell Sabi Abyad.7 These indicators of socialTell Brak.9 Until very recently, however, it has been
differentiation are not seen in burial practices,difficult to demonstrate the existence of such large
which, although varied, provide only sparse evidence sites, much less study the level of complexity occur-
for hierarchy.8 ring within them. Evidence for large sites has now
To date, it has proven difficult to present both astarted to accumulate. Takyan is a 10-ha site in the
coherent and convincing picture of fifth-millennium Cizre-Silopi plain in southeastern Turkey that seems
society in the northern arc of the Fertile Crescent to have been entirely occupied in the Halaf.10 Ka-
(fig. 1). In general, sites are quite small, rarely more
zane, on the southern outskirts of Urfa, is larger still,
5 Campbell (supra n. 4) 151-56; E. Healey, The Role of Ob- Survey," in O.M.C. Haex, H.H. Curvers, and P.M.M.G. Ak-
sidian in the Halaf (Diss. Univ. of Manchester, in preparation). kermans eds., To the Euphrates and Beyond: Archaeological
6 M.E.L. Mallowan and J.C. Rose, "Excavations at Tall Studies in Honour of Maurits N. van Loon (Rotterdam 1989)
Arpachiyah, 1933," Iraq 2 (1935) 1-178; A. von Wickede, 75-88; Campbell (supra n. 4) 182.
Prdhistoriche Stempelglyptik in Vorderasien (Munich 1990); 9J. Oates, "Prehistoric Settlement Patterns in Mesopota-
Campbell (supra n. 4). mia," in P.J. Ucko, R. Tringham, and G.W. Dimbleby eds.,
7 P.M.M.G. Akkermans and M. Verhoeven, "An Image of Man, Settlement and Urbanism (London 1972) 299-310.
Complexity: The Burnt Village at Late Neolithic Sabi Ab- 10 G. Algaze et al., "The Tigris-Euphrates Archaeological
yad, Syria," AJA 99 (1995) 5-32; Akkermans, Tell Sabi Ab- Reconnaissance Project: A Preliminary Report of the
yad, the Late Neolithic Settlement (Istanbul 1996). 1989-90 Seasons," Anatolica 17 (1991) 175-240.
8 P.M.M.G. Akkermans, "Halaf Mortuary Practices: A

94.184.169.35 on Sun, 06 Dec 2020 06:02:23 UTC
1999] THE DOMUZTEPE PROJECT, 1995-1997 397
and its Halaf settlement is now under excavation.11 A-B (ca. 4500-4300 B.C.); chronological details are
discussed below.
More certainly exist. The distribution of these sites is
to the north rather than in the traditional Halaf
ENVIRONMENTAL AND CULTURAL SETTING
heartlands of northern Iraq and northeastern Syria.
This distribution may be an artifact of survey and siteThe Kahramanmara? region is well wa
burial under later occupations, but it may, in fact,fertile, located on the northwestern edge
conceal significant regional variations in settlementtile Crescent as it runs up to the Taurus M
structure and social organization. fringing the Anatolian plateau. It cont
Similarly, although it has been possible to identify
through passes to the plateau and is also s
regional variations within the Halaf,12 they have an important north-south route running
rarely been the focus of analysis.'3 There has beenrift
a valley from the northern Levant to
tendency to see the Halaf as a single entity ratherregion. Domuztepe is set on the edge of a
than as an aggregate whose constituents are not nec-
vial plain with a range of low limestone
essarily integrally related to each other. The answers
hills immediately adjacent to it on the
to questions about the development of society in this
plain is very flat and, although fertile, w
period probably lie in detailed local and subregional
subject to flooding. Indeed, the alluvial de
studies and they may well also lie outside the tradi-
have built up around the site since its preh
tional areas of archaeological activity such as north-
cupation have been waterlogged at some d
ern Iraq and northeastern Syria. In these areas, ourto relatively recent drainage around the s
established typologies will not necessarily fit and was
we extensive marshy ground immedia
need to confront their contradictions.14 west of the site.
Domuztepe was located during the Kahraman- Pollen cores suggest that from 8000 to 4
mara? regional survey conducted by E. Carter in the oak-pistachio steppe-forests covered th
summer of 1993.15 A preliminary surface investiga- which Domuztepe lies.18 From 4000, th
tion was conducted in the summer of 1994 that suc- forests gave way to expanding oak forests
cessfully demonstrated that the site was primarily oc- natural vegetation of the region today
cupied from early in the sixth to the late fifth human interference). The Domuztepe
millennium and that its size was ca. 20 ha.16 A de- mains recovered to date fit with such an environ-
tailed surface collection was carried out in 1995, to- mental setting. On a local scale, wetland taxa include
gether with limited soundings.'7 Full excavation sea- Scirpus sp., Carex sp., Cladium sp., and Schoenus sp.
sons took place in 1996-1998 as a joint project of Other taxa that may be found in wet conditions in-
the University of California, Los Angeles, and the clude Astragalus sp. and Malva sp. (along with a few
University of Manchester under the direction of possible, cf. Trigonella sp. and Juncus sp. seeds).
Carter and S. Campbell. A further season is planned These all suggest that Domuztepe may have been sit-
for the summer of 1999 with a study season in 2000. uated near a marshy area, although there are too few
The present article summarizes work carried out in seed remains at present on which to base any reli-
1995-1997 and is an interim statement on the site- able conclusions. The frequency of pig bones in the
an overview of the main results and an indication of faunal assemblage also suggests that the environ-
the ways in which the project is contributing to an- ment was wet enough to keep pigs at or near the site.
swering some of the questions posed above. The The majority of Halaf sites investigated to date are
major prehistoric phases investigated at the site are located in northern Iraq and the Khabur area of
Late Halaf (ca. 4750-4500 B.C.) and Post-Halaf Syria. The northern and western extension of the
11 P. Wattenmaker, personal communication 1996. 16 E. Carter, "The Kahramanmara? Archaeological Sur-

12 E.g., A.-L. Perkins, The Comparative Archaeology of Early vey Project: A Preliminary Report on the 1994 Season,"
Mesopotamia (SAOC 25, Chicago 1949) 43-45; Davidson XIV Arastzrma Sonuflan Toplantzsi (Ankara 1997) 289-305.
(supra n. 4). 17 E. Carter et al., "Report on the Archaeological Work
13 Notably LeBlanc and Watson (supra n. 4). in Domuztepe and Its Environs in 1995," XVIII. Kazz
14 S. Campbell, "Archaeological Constructs and Past Re- Sonulanr Toplantzsi (Ankara 1997) 173-87.
ality on the Upper Euphrates," in Archaeology of the Upper 18 W. van Zeist, "Reflections on Prehistoric Environ-
Syrian Euphrates (Tishreen Dam Area) (Barcelona, in press). ments in the Near East," in P.J. Ucko and G.W. Dimbleby
15 E. Carter, "The Kahramanmara? Archaeological Sur- eds., The Domestication and Exploitation of Plants and Animals
vey Project from 24/9/93-11/11/93," XIII. Arastzrma (London 1969) 35-46.
Sonuflanr Toplantzsi (Ankara 1996) 331-41.

94.184.169.35 on Sun, 06 Dec 2020 06:02:23 UTC
900- ???????? ?""?"""~" "" \ I I \
800-T ..... ..
Operation III
700-
Operation IV
600 w 1
Operationn I
500-
200 300 400 500 600 700
Fig. 2. Domuztepe: contour plan
such ephemeral
Halaf is less well known, although s n
creasing attention foothills
in the Euphrates on the ed
suggests
lar. The only excavated site that near more Dom
G6zi, some 19 km to the south; relat
SURFACE SURVEY
is known of its relevant levels.19 Br
rary sites have beenToday, thelocated
site of Domuztepe in th
covers an
survey. Only threeproximately
have 20 haproduced
and rises to a maximu la
pottery: sites 69 and
ca. 12 m 125above thehave Halaf
alluvial plain (fig. 2).
67 has Post-Halaftentmaterial thatsomewhat
of the site was probably is br
seems very
rary with the final probable that alluviation
prehistoric occu h
tepe. Other contemporary
around the site to some sites prob
depth since the f
their deposits arenium; buried
a Roman sherd was beneath
found at a dep
and only a scattering
below the plain in of sherds
a test pit southeast ofh
Nonetheless, it tivation
seems has also encroached on the site.
clear that D
the center of a system that
Realistically, it is impossible has
to excavate morenot
than
dense pattern of a tiny
settlement remain
portion of a site the size of Domuztepe. Exca-
nities may have vated
left minimal
areas can be no more than minute andarch
largely
perhaps due to subsequent
random keyholes into the overall alluviat
settlement. The
the ephemeral nature of
only way to try to tie transhum
them together into an overall
19J. Garstang, "Excavations at Sakje-Geuzi in North 1911," Liverpool Annals of Archaeology and Anthropology 24
Syria: Preliminary Report for 1908," Liverpool Annals of Ar-
(1937) 119-40; J. du Plat Taylor et al., "The Excavations at
chaeology and Anthropology 1 (1908) 97-117; Garstang,
Sakce G6zil," Iraq 12 (1950) 53-138.
"Third Report on the Excavations at Sakje-Geuzi, 1908-20 A. Garrard, personal communication 1996.

94.184.169.35 on Sun, 06 Dec 2020 06:02:23 UTC
900 N
800
800 100m
700
75
600 00
35
60
.. ............... 2 015
4004
2003
200 300 400 500 600 700
Fig. 3. Density of Halaf pottery on the surface of

age of total ceramics collected
understanding of the development of

both by pottery and by fragments the
of roof tiles andsite is to
maximize our use of the artifacts scattered over the bricks. Three identified coins date to the first half of
surface. This process was begun in 1995 with an in- the fourth century A.D.21 This late occupation is
tensive and systematic collection of material from largely concentrated to the southwest and extends in
the surface of the site but is an ongoing study, with a band running northeast across the central area of
repeated collection in subsequent seasons and ex- the site. It probably covers no more a third of the
amination of the distribution of artifacts in the tepe. Given the substantial remains of plowed up ma-
plow zone. sonry (including a column section), the settlement
The density of artifacts in the limited systematic may have been of some significance during the first
collection of 10 m2 areas in 1994 suggested that, to millennium A.D., possibly associated with the bridge
gain significantly large samples of the full range of that crossed the stream to the west of the site until
ceramics, in particular, we would need to collect ma- it was largely destroyed by the construction of a ma-
terial in quite large units. Circles 4.37 m in radius jor drainage canal in the 1970s. The excavations
were used, giving areas of 60 m2. They were laid out have also revealed a cemetery on the south mound
in a stratified random pattern with one collection that had not been indicated by surface remains.
area in each 25 x 25 m block. Omitting areas that This later activity is not, however, the focus of the
still had crops on them gave us a total of 115 collec- current article.
tion areas, approximately 5% of the total area avail- Prehistoric pottery appears across the entire sur-
able for collection. Additional off-site surveys have face of the site in large quantities (fig. 3). Points of
led to a better understanding of site boundaries. lower density are probably attributable to overlying
First-millennium A.D. occupation is indicated late occupation rather than to any absence of prehis-
21 Thanks are due to Robert Gurval, University of California, Los Angeles, for this identification.

94.184.169.35 on Sun, 06 Dec 2020 06:02:23 UTC
check the highest, and

toric occupation. Our understanding ofpresumably
the detai most
historic
the prehistoric material is remains on the southern mound in final
preliminary; an area exa
nation of the material remains from the site will take where there appeared to be a concentration of
place only after further excavation. The latest prehis-Ubaid pottery on the surface. Its excavation has
toric material on the site immediately postdates the been continued in subsequent seasons and it is now
Halaf and occurs over the entire area. If we take the largest exposure because of the presence of well-
the occupied area as 20 ha, there may have been a
preserved wall footings at a reasonable depth below
peak population in the order of 1,500-3,000 people.
the surface. Operation II was also begun as a sound-
This sort of estimate is very approximate and there
ing inis1995, taking advantage of agricultural leveling
already some indication that there are morethat cut into
com- the base at the southeast corner of the
mound, to investigate what were expected to be ear-

plex spatial patterns. One example is an area of higher
density of thick-burnished sherds in the southeast- lier deposits. We continued work there with a lim-
ern part of the site, particularly to the north ofited exposure because, although the levels are not
oper-
ation II. If this type is characteristic of the LateasHalaf
early as expected, it remains the only area where
and Post-Halaf A ceramic phases, this portion of wethe
are able to excavate Late Halaf deposits. Opera-
mound was not occupied intensively during thetion Post-
III was first excavated in 1996 to investigate an
Halaf B phase (see below for discussion of the ce-
apparent concentration of obsidian and other lithic
ramic phases). In fact, given the presence of debris
Lateon the surface and has been continued be-
Halaf ceramics close to the level of the surrounding
cause of more fortuitous discoveries. Operation IV
plain in operation II, it is possible that this portion
was opened in 1997 to test the deposits near the top
of the mound was a relatively short-lived expansionof the mound at a point where we suspected that the
of the site.
edge of the first-millennium deposits overlaid the pre-
The point at which the site grew to its current size historic. The main evidence is discussed below in
is unclear. The deep deposits across most of the site, chronological order.
however, suggest that its final extent (ca. 20 ha) was
maintained for a considerable time. Pottery from the Late Halaf
earlier phases of the site is obviously masked by ma- Operation II has provided a very compressed
terial from close to the surface. Individual sherds stratigraphy of at least eight distinct levels in an
that could be broadly categorized as Early Halaf (late
area of 85 m2 and a maximum depth of around 2 m.
sixth millennium) occur in both the northern and The architecture, a single level of which was never
southern parts of the site, and it may be that the full
present in the entire trench, varies between rectilin-
size had been reached early in the fifth millennium.ear and circular, with at least three examples of clas-
The very small quantities of Ceramic Neolithic pot- sic circular Halaf tholoi found. The building tech-
tery are more common in the southern part of the niques vary too, with some buildings having stone
site. A broken Amuq point also came from this por- foundations and others not, generally with a pise su-
tion of the site. These finds may suggest an earlierperstructure, although in the case of the earliest tho-
but possibly substantial occupation in this area, los
al- the superstructure was probably of mudbrick.
though a variety of other explanations are possible.
The pottery from all of these levels seems to be Late
EXCAVATIONS
Halaf, but there are probably developments not yet
detected because of the relatively small sample col-
The site of Domuztepe does notlected
appear
from anyto have
one phase.
suffered significant erosion in recent centuries
Probably the most complex al- building level exca-
though it has been plowed and very extensive
vated is that illustrated insoil
figure 4. It was founded on
formation has affected the upper parts of
a terrace the
that archae-
had truncated a child burial. A hybrid
ological deposits. The top 0.7-1 m ofbuilding
deposit has al-
was constructed with a circular outer wall
most entirely lost its original soil structure,
(approximatelyincluding
6 m in diameter) incorporating a
all traces of pise walls and earth surfaces, although
square central room, the corners of which touched
stone foundations, plastered floors, the and larger
interior of the arti-
circular wall. It remains unclear
facts are in their original positions. whether this double wall was functional22 or whether
Four areas have been opened for some
excavation (fig.extended to the height of
of the walls never
2). Operation I was begun as a sounding in
the building. 1995
One to especially likely given
possibility,
22 Possible parallels may be from Yarim Tepe: N.I. Mer- Yarim Tepe III: Settlement in North-western Iraq, 1978-
pert and R.M. Munchaev, "Soviet Expedition's Research at 1981," Sumer43 (1984) 54-68, figs. II-III.

94.184.169.35 on Sun, 06 Dec 2020 06:02:23 UTC
...............:i:::
X:~'i_-:~%:_-i:i~:m : ::;il:i~~: ~::~l:~~r~~:~~li_ i(. - ::i--Ii 8~i::?::
.............::--:
Fig. 4. Operation II, Late Halaf tholos in its initial phase

Note the group of thick-burnished jars built into the wall (
the very thick walls of the square room, is that the mal bones, mainly lying on or adjacent to a curved
double wall was intended to provide insulation for area of tightly packed pebbles (fig. 5). Over 90% of
some particular function. Later, the circular wall was the bone fragments (52 of 57 bones) recovered from
built over and the structure became a more regular this concentration in 1996 were those of cattle. More
building of small rectangular rooms. Finally, the than half of the cattle bones belonged to meat-bearing
building was abandoned and the central room was parts of the carcass, making this an unlikely butchery
used as a rubbish dump. Among the material found deposit. These bones represent at least four individ-
dating to the final phase was a complete painted ves- uals. As the deposit had little other refuse material
sel, a seal possibly made of malachite, a concentra- such as sherds, and the bones were not highly frag-
tion of animal bones, and a human jaw. Among mented and came mainly from a single species, one
many important structural details was a cache of six explanation may be that this is primary refuse from
thick-burnished jars built into the base of the circu- consumption. These data, possibly suggesting feast-
lar wall (fig. 4). These appear to have been used as if ing activities, would have exciting implications in a
they were stones, rather than for any ritual purpose. society that, on the classic social evolutionary scale, is
Just outside the south wall of this building is a cor- usually considered a chiefdom. Additional dense de-
ner of an exterior surface projecting into the trench. posits of bone in a pit in the northeast part of the op-
On it was knapping debris in situ, mainly of chert eration suggest that this area was used for refuse
but including a few pieces of obsidian. Not only may disposal over a long period of time. Although exca-
this material provide new information about knap- vation and analysis are not complete, it is clear that
ping technology in the Late Halaf, but the obsidian there is not a simple pattern of disposal; some indi-
debitage suggests routine working of obsidian at vidual clusters of bones resemble butchery refuse
the site. while others suggest feasting.
Operation IV was opened in 1997. The prime ob-
Post-Halaf A jectives were to establish the depth at which the pre-
Although in both operations III and IV there is a historic deposits occurred and to understand how
paucity of well-stratified ceramics because of soil for- the late pottery on the surface in this area (ca. 40%
mation processes, the general assemblages can be of all diagnostics) relate to the subsurface material.
broadly dated to the Post-Halaf A phase, sometime Approximately 70 cm of later deposit, roughly first
between the phases represented in operations II and millennium A.D. in date, from probable building de-
I. Operation III was opened as a 5 X 4 m trench in bris rather than in situ occupation overlay intact pre-
1996 and was extended to a 9 X 5 m trench in 1997. historic deposits. The soil was only partially devel-
Alongside later prehistoric pits and traces of archi- oped, which gives us some idea of the rate of soil
tectural remains was a dense concentration of ani- development on the site. Below this was prehistoric

94.184.169.35 on Sun, 06 Dec 2020 06:02:23 UTC
-- -- - - - - - - I"
- - - - - -- .q
- - - - - - - i
- - - - V"
- - ------
X~
i
63- i
..o
its
S\\
A
I ci::? , 41
-T--
A
VA
S:96.2
I ccb ) .._.
01)
ID o
o r G,96.1!
" --4---
96.5 0 T1L 4
7--KeyA
w 9696.
I 96.5(2 0 96.4l .Am

.
145
A
[--
Thick[
r~1~996.2
plaster -_ [ Sherd
- - - - - -
+ ____
- - - - - - - - - - . . . . . . . . .
,Darkburnth
-
S/ ( 1Posthote
- - - - - - . . . .
cm
. . . . . . . .
Fig. 6
s s
burial
cately
sel (se
pleted
Post-H
A sma
to a d
est fe
S
(fig.
148 li
Fig. 5. Plan of Post-Halaf A architecture and b
in operation III sonab
structed as follows. A wide, shallow cut, much
material, the uppersteeper on the north than
portion of the south,
whichwas first made
had
veloped buried soilin horizon.
the slope of the mound,Architectur
and a steep-sided, but
were discovered inrelatively
the trench,
shallow, pit was cut into it.includin
The edges of
of tholoi and a hearth containing
the wide pit have not yet been excavated buta con
since
of seeds from the they extend out of the
Vicieae excavated area(they
tribe on all sides, a
served well enough the to determine
pit must have covered a large area. Thewheth
sloping
lentil, vetch),
edge ofpresumably pea,
the pit, together with the positionthe of laterresu or
dental of structures,
food into
suggests the
that the edges spilling
may not fire.
lie too A
other finds were 13 far clay
outside our excavated
sling area butballs
that would still
from
of deposits and 18 stone bowl
give a north-south dimension fragments,
of at least 10 m. The
23 Further excavation in the summer of 1998 supports a much more complex process of deposition of human
the following interpretation but makes it clear that the bone. Further excavation is required in 1999 to elucidate
events reconstructed here took place as the final stages of
this pattern.

94.184.169.35 on Sun, 06 Dec 2020 06:02:23 UTC
pit was approximately 4 mpitwide at

are generally theastop,
unburnt, 1.6
are the sides m
of the pit,
wide at the bottom, and 0.8
so the m deep;
burning certainlyit
tookappears,
place elsewhere, per-
from the portion excavated, to have
haps associated been structure
with a mudbrick roughly since burnt
circular. mudbrick also occurs in the fill. Two carbon samples
A small oval area of burnt earth (45 X 60 cm) in from the ashy fill have been identified as a mixture
the base of the pit suggests that in situ burning took of parenchyma (probably from a fleshy type of fruit)
place in the pit soon after its excavation but that it and chaff. Fragments of charcoal in the upper por-
was very restricted. It did, however, take place before tion of the ash come from juniper and oak. One
any deposits were made in the pit and may represent skull placed near the top of the ashy fill still pos-
a key stage in the funerary ritual. A tightly packed sessed a mandible; in contrast to most of the human
layer of stones was then placed at the base of the pit, remains, it may represent the burial of a head rather
providing a level upper surface. Several skulls were than a skull. The shallow cut finally filled up with
found on or adjacent to the stones at the bottom of the silty mudbrick or pise debris falling or washing in
pit and some long bones on the north edge of the pit, from the north. The find of an extremely worn hu-
although not enough to match the number of skulls. man jawbone in this later fill may not be accidental
At least one of the skulls was carefully placed on a and may suggest a continuing knowledge of the fu-
stone but others seem to have been placed with less nerary nature of this area. Another find, although
care. Until a full anthropological study has been fragmentary and redeposited in the later infill of the
completed it is impossible to be sure of the original pit, is worthy of attention: a fragment of wall plaster,
number, but it is provisionally estimated that be- still with a piece of pise or mudbrick on the reverse,
tween six and nine skulls are represented, appar- with painted lines on it. It appears to be a fragment
ently from both children and adults. The skulls were of wall painting with a geometric design.
damaged in most cases, sometimes with substantial It would appear that we are dealing with a com-
portions missing. The clear absence of later distur- plex burial rite. In general, the skulls certainly repre-
bance, the generally good state of bone preserva- sent secondary burials of individuals who may have
tion, and the missing fragments suggest that this died at varying times, to judge from the widely differ-
damage occurred outside the pit. The degree of ing damage to the skulls. The burials can also be as-
wear of some of the skulls suggests that they may sociated with wider ceremonies outside the pit. The
have been damaged a considerable time before their ash came from a large fire, which would certainly
burial. It is provisionally suggested that the fragmen- have been widely visible within the settlement and
tary human remains represent not the bones of the perhaps beyond, and may have formed one element
recently dead or simply the results of excarnation of a ceremony involving a significant portion of soci-
but were assembled over a longer period, conceiv- ety. If the burials represent a single family, possibly
ably dug up from older burials. more than one generation, this ceremony may have
Nine animal bones were found associated with the been of primary concern to a family group. If the
human bones in the pit. Six of the bones were thosebones did not come from related skeletons, a much
of cattle, all representing meat-bearing parts from at
wider segment of society may have had an interest
and involvement in the ceremonies.
least two individuals (two right humeri, two halves of
one left tibia, one left radius, one right rib). The ra- Mass burials in the later Halaf are also known at
dius and tibia were nearly complete, and half of theTepe Gawra24 but they are of articulated skeletons,
rib is preserved. Whole cattle bones such as these are presumably buried very soon after broadly contem-
rare in deposits relating to consumption. It is there-porary deaths and, therefore, not directly compara-
fore likely that these cattle bones served another ble to burial 148. Other elements of the Domuztepe
purpose, perhaps as food offerings placed in the pit burial pit are mirrored in northern Iraq, although
with the human bones. there must be some doubt as to how relevant the
Fine ash and burnt brick debris were dumped into similarities are. At Arpachiyah, one burial contained
the pit together with further skeletal material as partan isolated skull in a vessel while another con-
of a single phase of backfill; there are no recuts visi-tained four skulls in a unique bowl.25 A single skull
ble and multiple depositional episodes in the ashy
was also found, probably in a Halaf context, in Mal-
matrix would have left clear traces. The bones in the lowan's excavations.26 Skull burials also occur at
24 A.L. Tobler, Excavations at Tepe Gawra 2 (PhiladelphiaArch 10 (1978) 125-28.

1950). 26 Campbell (supra n. 4) 176.
25 I. Hijara, "Three New Graves at Arpachiyah," World-

94.184.169.35 on Sun, 06 Dec 2020 06:02:23 UTC
Fig. 7. Operation I, Post-Halaf B. Tholos w
Yarim Tepe II. In twoThere

caseswas probably
singlea period skulls
during which had
no
buried and, in another buildings
three, skulls
were constructed overwere
the area of buried
the burial
gether.27 In none ofpit.these
Certainly itscases were
upper portions grave
appear to have been
filled gradually were
found, except where vessels by erosion and
useddumping asfrom the
conta
at Arpachiyah. Seven cremations were
north. Finally, it was sealed found
and the area leveled by a a
rim Tepe II, although dense
most depositwere
of sherds in
and bones,
the deliberately
earlier
of that site.28 Burning
dumped generally
and probably intendedtook
to consolidate
place
an
where, in one case in aarea
specially
that was prone to flooding.
built On oven.29
top of this there In
cases, the grave goodswere
werea series ofrich. A single
poorly preserved, crem
packed earth floors,
also occurs at Chagar Bazar.30
occasionally with aThese
scattering of examples
plaster. Nothing sub-of
mations fit into a wider spectrum
stantial appeared until a serieswhere the
of well-constructed cir-de
ate burial of burnt objects was
cular buildings important;
were built. exam
The only completely exca-
vated tholosTepe
can be seen both at Yarim is about 2.5IIm in
anddiameter and probably
Arpachiy
had a door to the
The burial pit at Domuztepe maysouth (fig. 7). The floor is mark-dra
therefore
widespread ritual elements,
edly concave and was andmade bypossibly
first packing a clayeybe
from across northern Mesopotamia.
subsoil, It perh
then putting down a layer of tightly packed
should be seen in a broad ritual context rather than
pebbles, and finally by adding a thick and very hard
a primarily funerary one. The burial may have more layer of plaster on top. This process would have
provided an extremely strong floor. The plaster was
to do with establishing a relationship with ancestors
not preserved as far as the wall, leaving a channel
and negotiating group identities within society than
simply being a mechanism whereby mortal remains about 20 cm in width at the base of the walls, filled
were disposed of in an appropriate way. with fine silt. It seems likely that the floor originally
27 N.I. Merpert, R.M. Munchaev, and N.D. Bader, "So- 29 Merpert et al. (supra n. 27) 51-52.
viet Investigations in the Sinjar Plain," Sumer 34 (1978) 27- :0 M.E.L. Mallowan, "The Excavations at Chagar Bazar,
70. and an Archaeological Survey of the Habur Region, 1934-
28 N.I. Merpert, R.M. Munchaev, and N.D. Bader, "Inves- 35," Iraq 3 (1936) 1-59, fig. 6.6.
tigations of the Soviet Expedition to Iraq, 1973," Sumer 32 11 Campbell (supra n. 4) 182-83.
(1976) 25-62; Merpert et al. (supra n. 27).

94.184.169.35 on Sun, 06 Dec 2020 06:02:23 UTC
I?
... i... .. :: I I ..........
.
: .
I
.. .. t.
. ..
.
.. . .
. .
. . .. i:.. I
, . +iI!! :; .... '
: : ,li-: . . .
: :..-I""& . ... ... ..., r .. :; ' .. ..
I~ : : : ; ... -?
... ,. - ?. _.,.,. -/, ?.-
Fig. 8. Operation I, Post-Halaf B. This plan represents a slightly arbitrar

phases of building. Most of the buildings were probably contemporary a
sloped upward over an empty channel to meet

The rather the
complex series of founda
wall. There are no precise parallels known
up this for this
phase is ar-
summarized, in a clar
rangement elsewhere; but it seems figure
likely that
8. It the ex- be a mistake to see
would
planation is functional, possibly having
staticto do withplan.
building air Instead, it repres
circulation. Only small portions of other plastered
sequence of modification and rebui
tholoi were excavated but they seem seven phasesover-
to slightly of construction were
lie each other, rather than being contemporary.
1996 and, although the final stratigra
The main exposure area of almoststill
400 being
m2 incompleted
oper- for 1997, we ant
ation I lies directly over this tholos and has
fying been
several ex- subphases. Thus, t
more
cavated to reveal as much as possible
lows of a coherent
us not only to examine a section
architectural plan, dating fromplan
the but
Post-Halaf B how space was uti
also to see
phase. Because soil formation has destroyed
ified overmuch
time. of
As the exposure wide
sequence
the overlying soil structure, we have will
been able to be
ex-increasingly impor
cavate this area relatively rapidly standing
and economically.
how space was structured in
It is only on the level of this architecture,
settlementbetween
and in obtaining a better u
0.6 and 1 m below the surface, thatofmore
howsubtle soil
residential and functional
differentiation has been reasonably preserved,
through time. allow-
The central
ing us to start to isolate the surfaces architectural
associated feature of the area exca-
with
the building. With a few exceptions vated appearspise
where to be su-
a U-shaped range of rooms, al-
perstructure has survived, we are mainly
though the dealing with are not perfect and the
wall alignments
stone wall foundations. It wouldtechniques
appear, ofhowever,
foundation construction vary. Small
that the walls were generally of pis6 built on
rectangular roomstop
are of a
present in the arms of the U,
in the order
single aboveground foundation course. In of the
2 x 4 m, and in its earliest form the
only
case where we have a well-preserved pise
top of the superstruc-
U, facing south, seems to have been open;
later,
ture, there is a thin external white a line coating.
plaster of stones, either a boundary or a true

94.184.169.35 on Sun, 06 Dec 2020 06:02:23 UTC
WIW t5 cm
..
/ i/ i
?~i , _ ... .
Fig. 9. Stone bowls from ope
wallfoundation, the main building

was laid as part of a large and it.
across complex I
the enclosed
central
integrated unit.space was roo
Evidence for
tempting to describe the final as
this phase of
a prehistoric
variant occupa-
plan that becomes commonplace
tion has been a
largely lost through erosion on the
Mesopotamia during south mound. theSeveral pits
laterwere found, fifth
especially in
Clearly this scenario the northwest would
area of operation be I, cut down
comp from
evidence for Ubaid influence in the ceramics, al- surfaces that are either eroded or destroyed in the
though it is too early to make definitive compari- plow zone. Although they are certainly prehistoric in
sons. Given the width of some of the foundation date, the ceramics in the pit fills seem to be clearly
walls, up to 1.1 m wide, a second story seems possi-
redeposited, probably from Post-Halaf A contexts. A
ble. Other narrow rectangular rooms to the series northof wall segments probably also relate to this
and east of this building may be discovered phase. to be- Although they do not make up clear plans as
long, after further excavation, to similar structures,
yet, they appear to be, in part, rebuildings of the op-
or they could have been freestanding. Although the I architecture; we may be dealing with a con-
eration
initial construction phase contained large open tinuous building process rather than an entirely dis-
tinct later phase of occupation. It seems unlikely that
spaces, these were gradually filled in, and by the latter
stages the architecture seems to be tightly packed.we will be able to excavate architecture of the final
phase of occupation at the site, which will limit

There appears to be rather different architecture
in the southeast part of operation I. The walls are
speculation on the reasons for the apparently rapid
thinner-typically a single or double row of founda-decline in settlement leading to abandonment of
the site.
tion stones- suggesting different wall constructions
or, perhaps, single rather than double stories. There A number of miniature spouted stone bowls were
are also two oven bases in this area consisting of found
thinin the upper levels of operation I (fig. 9.2).
They may have been associated with layers now de-
floors of burnt clay built on a foundation of pebbles
stroyed
and edged by a single line of small stones, probably a by soil formation or they may have come
foundation for the superstructures. Furthermore, from destroyed burials; certainly the discovery of two
human mandibles in similar deposits strengthens
one of the rooms has a thin, albeit poorly preserved,
plaster floor. All of these features make this
thearea
latter possibility. These bowls are very finely
distinct and suggest a functional segmentation,
made
al- items, produced mainly from locally available
though it remains to be seen whether theseserpentine
fea- and manufactured with great skill. Stone
tures are in any sense communal or belong bowls
with occur on Halaf sites regularly but in small
32J.-D. Forest, "La grande architecture obeidienne: Sa Roaf, "Ubaid Houses and Temples," Sumer43 (1984) 80-90.
forme et sa fonction," in Huot (supra n. 4) 385-423; M.

94.184.169.35 on Sun, 06 Dec 2020 06:02:23 UTC
pottery that
numbers; there is an unusually highis to be expected alongside it. This
concentration
at Domuztepe (over 140 fragments
would not be ofnoteworthy
rims in itself,
andyetbases
a substantial
have been found so far). It is tempting
quantity to think
of less usual pottery that
in a surprising variety
of wares is also found.manufactured
they belong in a group of high-status It seems clear that the Halaf
style below).
items, including obsidian (see has been adopted and developed to carry a
range of meanings in a local context and is used as
CERAMIC CHRONOLOGY only one part of a wider functional and symbolic as-
semblage. It seems useful, therefore, to summarize
The general ceramic categorization and prehis-
the main categories.
toric chronology at Domuztepe are outlined here
partly to substantiate the chronology of the
Ceramic Typessite but
Painted and
also because the ways in which the assemblage unpainted Halaf (fig. 10.1-5). These
differs
from classic Halaf assemblages encourage a are
two categories moreclearly related to Halaf pottery to
subtle approach to our social interpretation
the south and of
east.the
The fabrics typically range be-
ceramics. The scheme presented here, both
tween forand
orange cate-
buff, sometimes with a gray core.
gories and phases, is strictly provisional. Ithighly
They are is based
and evenly fired and usually have a
on 23,705 sherds recorded by "medium-level pro-
clear orange to buff surface color. They often have
cessing" (relatively rapid analysis more
of all gritdiagnostic
temper than is traditional in Halaf ceram-
sherds). It is expected to change, possibly signifi-
ics, but an unbroken pot would look very similar to
cantly, in light of future excavation one
and from the classic Halaf heartlands. The sherds
study.
The surface collection suggests that the site
are painted was of colors, from red through
in a range
occupied in two broad phases: during orangethe to sixth
dark brown andand black, and most vessels
fifth millennia B.C. and in the first millennium A.D. were probably decorated to some extent. Most of the
The latter occupation is not a focus of this article decoration fits comfortably into the Halaf tradition
and much of the former is only hinted at. Surface but there are recognizable stylistic traits that are
and residual pottery probably dating from the Ce- more regional.
ramic Neolithic includes vegetable-tempered sherds Classic Ubaid. This category is very rare. Sherds do
with a fugitive red wash and a few coarse sherds with occur with high-fired fabrics, often with a greenish
applied decoration, both types known from survey tinge, fine grits, and decorated in matte dark paint
sites of this area and with good external parallels. Awith distinctive Ubaid motifs, but they are infre-
single Samarran sherd is not particularly surprisingquent at best.
given the presence of Samarran pottery at Sak;e Bichrome (fig. 11.2, 4). This group is composed of
G6zfi to the south; it fits with the presence of a fewall sherds with painted decoration in two or more
very Early Halaf sherds with horizontal cross- colors. Although there are a few genuine Halaf
hatched decoration. Domuztepe thus appears to bichrome sherds (including examples with white
have been part of the very wide zone that adopted paint) from operation II, they are very rare. Much
well-fired pottery with painted decoration late in more common are bichrome sherds that carry some
the sixth millennium, leading to the traditional Ubaid traits. The fabrics are usually very well fired,
Halaf style.33 If so, this extends the known area inorange, and rarely with a gray core. There is often al-
which the Halaf developed well to the west of the most no visible temper; in fact, the fabrics are gener-
Euphrates basin for the first time. There are further ally closer to those of classic Halaf than those in our
hints that the subsequent Halaf sequence is also rep- general Halaf category. The surface most often has a
resented at Domuztepe. Most clearly present, at or thin orange slip (at times almost a thin wash) and
close to the surface, are Late Halaf and Post-Halaf the paint colors are dark brown and dark orange or
ceramics, probably dating to the third quarter of red. Shapes include bowls with a sinuous profile and
the fifth millennium. Although there is a small clear Ubaid parallels; motifs are also rather different
Ubaid component postdating the Halaf material, it from those on standard Halaf pottery and may have
is not large and no later prehistoric pottery has been Ubaid links.
recognized so far. Painted-orange. This type is probably closely linked
Much of the pottery from the Late and Post-Halaf to bichrome pottery. Only a single color of paint is
contexts at Domuztepe is classic Halaf or the coarse used but the fabrics, shapes, and motifs are similar. A
3 Campbell (supra n. 3).

94.184.169.35 on Sun, 06 Dec 2020 06:02:23 UTC
~(i(( ((?(? 1 of pffoopoppi tot I I??

(((.I I??.?? II its 1111111~:
,-- 0. 1111 - t -to -1 7,7
19 it 9-v i # I I I
~(?(( rr??? rrr
~5 cm
6 7
Fig. 10. Classic Halaf pottery (1-5) and th
dark brown or black paint is usually

a glossy surface used on a s
of strikin
face that has had a thin oforange
shapesor is red
present,
slip appli
shal
Red-, brown-, and black-burnished
turned rims are (fig. 11.3, 5
particula
These three groups areIncised
closely linked.
(fig. 12.4,Indeed
6). Alt
probably form a spectrum cised sherds
rather among
than threethe H
clea
visions. Technologically, this
been group is distinct
restricted to a diffe f
the Halaf. Fabrics are ally
more heavily
this type has grit-temp
a semic
and firing is less high. with
Gray prominent
cores are the grit
ruletem
an
the case of black-burnished pottery,
were used often the
for cooking bu
is heavily reduced throughout.
A variety of The surface
simple geomof
pots is often only lightly burnished.
erally made by In some
deep V-prca
however, the burnishingThick-burnished
is exceptionally (fig.fine, 10gi

94.184.169.35 on Sun, 06 Dec 2020 06:02:23 UTC
1
2
3 4
5 cm
Fig. 1
assemblage can be divided into three main phases

tery
gener
(fig. 13), each probably at most 150 years long. One
comp
of the aims of the project is to provide finer chrono-
logical precision. To date, these phases have not
proce
tion,
been identified through stratigraphic relationships.
ric is
Pottery assigned to each phase occurs in different op-
orang
erations and can be arranged in a logical sequence; as
thick
different developments all seem to support the same
usuall
arrangement, we can probably approach it with con-
siderable confidence. It does mean, however, that
The v the
propo
phases are artificially distinct. If we had the missing
which
strata that would link them together, we would al-
that
most certainly discover a very gradual transition be- a
mens
tween them and we might well find that our divisions
diame
are not the optimal ones. Therefore, the terms used
mum
here for the three phases are strictly provisional and
and
may be superseded in the final report. 1
large
southeast surface of the mound.
Late Halaf Pottery
Other ceramic groups. A range of other distinctive In the Late Halaf phase, documented in operation
types of decoration and fabrics can be seen (fig. 12).
II, the ceramics are dominated by classic Halaf types
Pattern-burnished pottery is present, usually associ-
but with significant quantities of less typical varieties.
ated with a gray reduced fabric. Sherds with incisedHalaf pottery is the major group in this phase, at
or molded textile imitations are very rare but also ex-
65% of the total (painted sherds being more than
tremely striking. Other decoration includes exam-
four times as common as unpainted). Sherds with
ples where the painting is done with a very distinc- possible Ubaid links, bichrome, or painted-orange
tive fine brush.
decoration are effectively absent. The rare bichrome
On the basis of these types, the excavated ceramic painted sherds have a closer relation to classic Late

94.184.169.35 on Sun, 06 Dec 2020 06:02:23 UTC
/, _.
/.. I/
--- --. 3-
S . .... .....
IL
i 5er5
_5 -
Fig. 12. Unu
Halaf small bichro

handlesjoining the lower neck to the shoulder,
with white
and p
bowls with sinuous profiles) are paralleled in the
rare Ubaid in northern Syria and Iraq. The ways some
examples
the motifs are used is also clearlystag
latest influenced by Ubaid
brown-,
styles. Classic Ubaid sherds,and
however, are extremely
of therare and total,
the fabric of the vessels showing Ubaid in-
sherds and
fluence is generally p
the same as that of Halaf vessels.
small Although
number there does seem to be genuine Ubaid oc-
burnished
cupation at some of the survey sites in sh the region,
the term Halaf-Ubaid Transitional, used elsewhere, (c
quantities
12% of the
seems inappropriate here since it is very unclear what a
vided betwee
transition may be taking place at Domuztepe. Accord-
ing to our current understanding, this phase would
actually seem to be a continuation
Post-Halaf A of the Halaf ce-Po
The term Post-Halaf is used with deliberation. ramic tradition into a time period contemporary with
There are rare but clear indications of Ubaid influ- the appearance of the northern Ubaid assemblage
ence in the painted ceramics of this phase. Somefurther
of south.
The assemblage has a slightly lower quantity of

the vessel shapes (e.g., bow-rim jars, especially with
34 Campbell (supra n. 4).

94.184.169.35 on Sun, 06 Dec 2020 06:02:23 UTC
0 10 20 30 40 50%
Late Halaf I I I I
Post-HalafA I I I
Post-HalafB I I I I
S- c "w w -1 0 0
- D - 0 0
a )
CD =3 7i~
-~CD- w CD CD
CD0-
=3- C
CDm - "7 "7 " 0
CD-
CD C
Fig. 13. Relative p
Halaf-type sherds
this phase, is (b
a large, straight-sided basin, with a flat-
tened outturned
assemblage, rim (fig. 12.1). Thick-burnished
with pai
mon as sherds are almost absent and may well be present
unpainted).
orange only as residual material. Pattern
sherds appearburnish also almost
most). disappears and may only be residual in this phase
Burnished she but
bers (ca. 20%)
incised decoration remains with
as a rare but constant type b
cised and (1.5%). There are notable changes too in the coarse
pattern-bu
ties pottery. It makes
similar to up about those
20% of the assemblage but i
with now grit-tempered sherds
pattern are about three times more
burnish
cised. common than vegetable-tempered ones.
Thick-burnishe
6-9%) and coarse vege
Summary
contribute equally to
The ceramic development outlined above is plau-
pottery. It is clear t
sible and there are interesting and consistent trends
lated to the Late Hal
in its detail. In the Late Halaf phase, although the
tinct changes in som
ceramic assemblage has substantial local elements,
in Halaf types and
most of its obvious external parallels lie to the
ties of Ubaid, painte
southeast, in northern Syria, linking it to the north-
Post-Halaf ern
B Mesopotamian plain. This does not seem to be
Pottery
The Post-Halaf
the case in the Post-Halaf. The developments that B as
clearly linked to
take place in northeastern Syria, for instance at Tell th
same general types
Aqab,35 do not occur at Domuztepe. There is no ob- p
however, viousin phase that parallels
some the Halaf-Ubaid transi- re
ther (slightly less
tion. Furthermore, the most distinctive new types of th
sherds twice as common as unpainted sherds). pottery (as opposed to changes in the relative pro-
Ubaid-related, bichrome, and painted-orange sherds portions of existing types) are ones that do not
now make up 4-5% of the assemblage and compen- seem to be present either at the end of the Halaf or
sate for the apparent decline in painted Halaf. Bur- in the full Ubaid in northern Syria. Bichrome and
nished sherds make up about 27% of the assemblage painted-orange wares do not occur at sites in north-
and there is a marked shift in color toward reds and eastern Syria such as Tell Aqab,36 or at Tell al-'Abr on
browns. A notable shape, which seems confined tothe Euphrates.37 They are, however, regularly present
35 Davidson (supra n. 4); T.E. Davidson and T. Watkins, 37 H. Hammade and Y. Koike, "Syrian Archaeological
"Two Seasons of Excavations at Tell Aqab in the Jezirah, Expedition in the Tishreen Dam Basin: Excavations at Tell
Northeast Syria," Iraq 43 (1981) 1-18. al-'Abr 1990 and 1991," DM9 (1996) 109-75.
36 Davidson (supra n. 4).

94.184.169.35 on Sun, 06 Dec 2020 06:02:23 UTC
Table 1. Relative Proportions

in assemblages in the northern Levant of Faunal
in Remains
the Amuq
at Domuztepe
and at Ras Shamra;39 less precise parallels also exist
Mersin.40 More surprisingly, there seemPost-
Late Post- to be par
lels much further south in the Levant. A series of Halaf Halaf A Halaf B
bichrome sherds at Tell Tsaf in the Jordan Valley41
have never fitted well in the local Wadi Rabba tradi- Sheep/goat 49% 47% 49%
Cattle 23% 24% 22%
tion, but neither do they have close parallels in the
Pig 23% 28% 25%
Halaf in general. They are, however, similar to the Other 5% 1% 4%
bichrome pottery at Domuztepe and in the northern Sample 359 2081 4364
Levant. Tentatively, one might suggest that an axis of
Data are based on the 1997 field so
ceramic influence and probably cultural interaction,
which had extended from Domuztepe into northern pise structures. Free-threshing
Mesopotamia, began to be supplanted, or supple- also present. A range of other
mented, by one reaching far down the Levantine coast. dicated: bitter vetch, lentil, and
pea among the pulses; fruits s
ECONOMY
with the fleshy and stone rema
An extensive sampling policy is in use
and at linseeds,
Domuz- which may have b
tepe, not only for botanical and faunal
oilremains but
crop (the evidence is inconc
ultimately to investigate environmental and eco-
A substantial sample of the an
nomic questions as well as the characteristics of dif-
Domuztepe has now been ana
ferent depositional circumstances as they relate to
consistent pattern has emerg
site formation processes. An effort was made to sam-
semblage is overwhelmingly do
ple most securely defined excavation units (lots), in very small numb
represented
and a whole-earth sample of between fox,
40 and 60 liters possible wild cattle
gazelle,
was collected from each. The samples were dog,
mestic pro- and birds. Sheep/
cessed using an electric motor and a flotation tank.42
half the animals represented
The flotation residue was collected in 0.25-mm mesh both contributing about 25%
bags. Seed densities are generally low, in part due to though the relative proportio
soil formation processes in the less deeply buried lev- would have been rather differen
els in which our largest exposures have concen- ticular much more significant.
trated; the preservation of seeds seems to be improv- pig were probably herded close
ing as we excavate deeper layers.43 goat may well have had a wider
The excavated levels of Domuztepe appear to parts present suggest that the s
have possessed a full range of domesticates. Samples types probably occurred on the
are currently too small to substantiate inferences of biases were sometimes visible in the distribution
chronological change, and no attempt at differentia- within the site (see operation III, above). Pig kill off
tion is made here. Emmer is the most commonly patterns indicate that meat was the main concern, as
found grain, followed by lesser quantities of einkorn. would be expected from such an efficient meat-
In general, the remains seem to be from crop pro- producer. Bone fusion data from cattle indicate a
cessing rather than primary storage and may have kill off of about a third of the sample by the age of
been deliberately reused as temper in mudbrick or four years, the point of optimal growth for cattle.
38 R.J. Braidwood and L.S. Braidwood, Excavations on the Prince Takahito Mikasa ed., Essays on Ancient Anatolia and
Plain ofAntioch I (Chicago 1960) figs. 131.11-19 and 158. Its Surrounding Civilizations (Wiesbaden 1995) 75-97.
9 H. de Contenson, Ras Shamra-Ougarit VIII: Prihistoire 43 Responsibility for retrieval was taken in 1996 and
de Ras Shamra (Paris 1992) figs. 211-12; see also figs. 1997 by Seona Anderson, who also was primarily responsi-
191.8-11 and 192.4 for incised and painted decoration. ble for analysis of the botanical remains in 1996. Amanda
40J. Garstang, Prehistoric Mersin: Yumuk Tepe in Southern Kennedy analyzed the botanical remains from the 1997
Turkey (Oxford 1953) fig. 73.3-10, 12-14, 18, and 34-40. season. In both seasons, Mark Nesbitt contributed to pol-
41 R. Gophna and S. Sadeh, "Excavations at Tell Tsaf: An icy decisions and gave expert advice on identification.
Early Chalcolithic Site in the Jordan Valley," Tel Aviv 15- 44 Sarah Whitcher has supervised collection of the
16.1 (1988) 3-36.
bones in the field and their later analysis at the University
42 Based on plans in M. Nesbitt, "Recovery of Archaeo- of Edinburgh. Valuable assistance has been provided by
logical Plant Remains at Kaman-Kaleh6yfik," in H.I.H. Nicola Murray.

94.184.169.35 on Sun, 06 Dec 2020 06:02:23 UTC
S2
75 6 7
5 cm
Fig. 14. Seals from D
Whether thisSEALS
indicates the in
cant kill off after four years
Stamp seals are well known as a regular but
of cattle into adulthood
artifact fo
from fifth-millennium sites.46 Signif
slaughtering them for their
semblages have only been excavated where in
mined due to tions
a lack of
were on a particularlyteeth
large scale-as at
tle beyond four years. Fortu
Tepe-or where particular locations were ex
bers of sheep/goat bones
notably the Burnt House p
at Arpachiyah with
which to base10 age estimates.
seals in a single building. One might supp
sheep and goats,
seals were valued itemsbased
and that their use was re-on
epiphyseal fusion and
stricted to certain people in specific man
contexts. We
data, shows that animals
know from Arpachiyah, Tepe Gawra, and, more w re-
older ages, typically
cently, Sabi Abyad47 that seals were used to th
more mark
to be exploited for
property. One might, milk, wo
therefore, argue further that
important to access
remember that
to seals, and perhaps the control of property,
tion may leadwas to
restrictedunderestim
to an elite group who may have used
such an early concentration
this as a mechanism of exercising control in society.
would be extremely
One of our initial hypothesesinteres
at Domuztepe was that
though the kill-off pattern
this behavior might be d
more marked at a larger set-
sive concentration on one pro
tlement because of the possibility of a more devel-
maintenance of a
oped hierarchy. large numb
sence of juveniles
Twenty-nine stampmight in
seals have come from Domuz-
important secondary product
tepe (fig. 14), collected from the surface in two sea-
ported by the extensive
sons of rang
excavation. Operation II has yielded 10 seals,
cluding spindle whorls, bone
most from stratified deposits; eight seals from opera-
ulas, found at the
tion I and twosite.
seals from operation III are from less
47 P.M.M.G. Akkermans and K. Duistermaat, "Of Stor-

45 Cf. S. Payne, "Kill-off Patterns in Sheep and Goats: The
Mandibles from Asvan Kale," AnatSt 23 (1973) 281-303. age and Nomads: The Sealings from Late Neolithic Sabi
46 von Wickede (supra n. 6). Abyad, Syria," Palorient 22.2 (1997) 17-44.

94.184.169.35 on Sun, 06 Dec 2020 06:02:23 UTC
secure contexts; and the Table 2. Concentration of Obsidian

remainder arein from
Different the su
face. The majority of Phases Domuztepe
the at Domuztepe stamp seals
made in geometric forms (e.g., rectangle,
Total % squ
circle, cone, trapezoid, pyramid,
Phase lozenge,
Lithics Obsidian Obsidian trian
and quatrefoil) but there is also a seal in the shap
Late Halaf 3906 434 11%
a four-fingered "hand." These seals are small, sel
Post-Halaf A 6037 449 7%
exceeding 2 cm in length, width,
Post-Halaf B 5453 or
1039diameter,
19%
carry incised geometric patterns on one face. On
other side are drilled In the Post-Halaf B building phase, th
loops that occasionally s
sidian is ca. 16%, but in the levels be
traces of polish, suggesting they
burial pit 148, the were strung
percentage is 22%.
worn as a personal ornament. There are numer
parallels for these seals unique,
in northern Mesopotamia
deserve further description
number of the seals also pleshow signs
(fig. 14.5) of top
is the long use a
fragmen
even of recutting or recycling, which
seal whose original may
shape maybe co
have
strued as an indirect record of their value. One cir- slightly convex sides. The seal face
cular stamp seal without a hole drilled in the backwith two rows of seven (preserv
appears unfinished and suggests that at least some offrame a central panel of four (prese
these objects were made at Domuztepe. To date, crania. The second (fig. 14.6) is a
only a single fragment of a sealing has been recov- block seal whose original shape w
ered at Domuztepe, but this is likely to be a result ofenge. The bottom and three of t
poor preservation and it is a reasonable assumption well smoothed. The top is deeply ca
that sealings were used in a manner similar to that that may represent some kind of
documented at other sites.
fringes preserved on the long edge
Some indication of the numbers of stamp seals in allels are two fragmentary carved b
use at Domuztepe can be gained from the six exam- daidah, phase B in the Amuq,48 wh
ples found during the systematic surface collection. diamond patterns.
Given that the collection represents a ca. 5% sample
of the site, there should be over 100 seals on the sur- CHIPPED STONE
face at any one time. If we assume that 20% of the ar- A variety of raw materials were exp
tifacts in the plow zone were on the surface at any chipped stone tools at Domuztepe. Che
given time, this would suggest 500 seals in the plow common. Several varieties of chert were u
soil of the site alone. The implication is clear that, atbly from a number of sources. The prefer
the end of the Halaf at Domuztepe, seals were notseems to have been an opaque chert wit
rare items. As a property-marking device, sealing is worn cortex, in various shades of brow
often thought to have been encouraged by societies from the local Aksu gravels. Chert of rath
that have become too large for face-to-face commu- quality is also found in the limestone outcr
nication. This may be the case at Domuztepe and itadjacent hills to the west. A gray chert, sim
may account for why seals are so pervasive. To afound in the limestone near Sakge G6zil
greater or lesser degree, almost everyone may have sionally employed as is a much rarer,
needed access to a seal to mark property; everyonegrained, dark chocolate brown to black fli
may have had their own seal. This does not mean may not be local. Quartz, which occurs
that the elite would not have manipulated as well as chert in the local limestone, was also us
used this incipient bureaucratic technology, by con- sional flakes of igneous rock occur, althoug
trolling rights to its use or by elaborating its usage. Itlittle evidence that igneous rock was used f
does, however, reinforce the suggestion that previ- ditional range of tool types. Obsidian is the
ous excavations on smaller Halaf sites have not en-
ily documented nonlocal resource the s
couraged us to understand the degree and nature of which depended on long-distance exchange
regional social organization. up between 7% and 19% of the chipped sto
Two carved pieces of stones (fig. 14.5-6) are prob- depending on context (table 2). The range
ably also from seals and, since they are potentially seen in transmitted light suggests tha
48 Braidwood and Braidwood (supra n. 38) fig. 66.6-7.

Saklag6zil Region, North Levantine Rift Valley
49A. Garrard et al., "The Early Prehistory of the
1995 Survey Season," AnatSt 66 (1996) 58, 67.

94.184.169.35 on Sun, 06 Dec 2020 06:02:23 UTC
sources were exploited: on present

are, for evidence
the most part, segments the
of blades retouched
green, peralkaline obsidian along the back andeastern
is from ends and with a Anatolia
denticulated
(probably Nemrut Dag or functional
Bing6l), the
edge with a band ofsemitranslu-
luster. Two pieces
cent gray material is from have traces of bitumen,
either used as an adhesive,
eastern to fix
or central
Anatolia, and the mahoganythem
obsidian isis possibly
to the haft. There from
also a variety of piercers
the Kars area. and drills, burins, and scrapers. Projectile points are
The chert industry seems predominantly focused rare but bifacially flaked points (two examples made
on flake production (ca. 60% of the total), withfrom the chocolate brown flint) and trapezes are
blades accounting forjust over 20%. A much higher present. There are also some larger bifacially flaked
percentage of the blades, however, have been re-leaf-shaped objects. All the forms are consistent
touched. All stages of flake production are present with the range of forms found in contemporary in-
as indicated by flakes with cortex consistent with the
dustries elsewhere. The number and types of re-
reduction of unprepared nodules and core trim- touched pieces made of obsidian are more restricted;
ming elements, as well as cores. Blade production only 5% of the obsidian is retouched, with edge-
also took place, though perhaps from preprepared retouched blades most common. Formal types are
cores, as to date there is little recognizible classic
unusual, but include notched arrowheads (one with
blade-core shaping debitage. The absence of a pre- differential weathering suggestive of binding), tra-
determined reduction system for flakes is empha- pezes, and points, together with some large pieces
sized by several small nodules that seem to have been
with semi-invasive retouch.
flaked on the Late Halaf working floor excavated in The distribution of obsidian at contemporary sites
operation II. Unlike other evidence for core reduc- in northern Mesopotamia is distinctive. Although
tion, this deposit seems to represent a fairly ad hocsites have high percentages of obsidian (up to 80%
affair. With the chert tools are a few pieces of obsid-
of the lithics at Tell Aqab but more usually around
ian debitage, suggesting that obsidian manufacture 20-40%), primary manufacturing debris is absent;
took place at the same location, possibly by thethe vast majority of the obsidian artifacts are blades,
same person. sometimes retouched. The usual interpretation is
Obsidian, on the other hand, was obtained and that these sites obtained their raw material from a
worked rather differently. There is some evidence center that carried out initial manufacture of blades,
that green obsidian came to the site in a semipre- or possibly blade cores. These central sites would
pared form, since flakes with patches of cortex are have had access to a different exchange network,
present. Gray obsidian, however, may have arrived in bringing raw material, directly or indirectly, from the
the form of preprepared cores ready for blade pro- obsidian sources themselves. Only Arpachiyah shows
duction. The grinding of butts, the thinness and reg- clear evidence of being such a redistributive center.
ularity of the blades, and the shape of their termina- There is rare but consistent evidence in the Halaf for
tions suggest that some blades were produced by manufacture of decorative objects by grinding and
pressure flaking-a technique largely confined to polishing obsidian. Again, Arpachiyah is the excep-
obsidian.50 We may note too a considerable differ- tion with rather large quantities.51 This association
ence in the economy of obsidian usage even at the of a redistributive center and more elaborate pro-
general level of blade to core ratios. Every obsidian cessing of obsidian is surely not coincidental. The
core could be expected to produce between 70 and supply of large pieces of raw material was probably
150 blades, while chert cores generally produced less necessary, expertise may have been greater, and it is
than 30 removals each. While part of the reason for exactly the type of context in which one might ex-
working obsidian differently can be accounted for by pect the appearance of status items.
the different properties of the raw material, we may Although Domuztepe is on the edge of the distri-
also suspect the presence of skilled or semiskilled bution of northern Mesopotamian sites and the par-
specialists. ticular circumstances are different, significant paral-
A wide range of retouched forms is present, mak- lels exist. Obsidian is relatively common but, more
ing up about 10% of the chert artifacts. Most com- importantly, all phases of manufacturing are present.
mon among the chert artifacts are edge-retouched There is also a growing number of instances of ob-
blades. Next in frequency are (sickle?) blades that sidian being ground and polished. Several pierced
50 M.-L. Inizan, "Technologie et prehistoire recente en l'&conomie de l'obsidienne," in Huot (supra n. 4) 205-315.
Mesopotamie: L'exemple du debitage par pression et de 51 Mallowan and Rose (supra n. 6).

94.184.169.35 on Sun, 06 Dec 2020 06:02:23 UTC
.......... iiiiiiiiiiil iiiiiiiiii::-:ii i::: ::i:i:::-::liii:: i

2 / 4
!iii-iii!} iiiiiiiililiiiiii
5cm
ii;!iiiii~i~ii~ii~iiii!iiiii~iii!iiii
i 'iii~iiiiii~iiiiiiiill
i il ii~~iiiiiii% ~i iiiiiii~iiiii~iiiiiiiii iiiiii iiiiii-iiiils:-. ~ ~ iiiiiiiii-iii~iiii~:i?:,'?i~i~iiiiiiiiiiii'iiii
2::::::: 3-::::~i:::i
5
Fig. 15. Obsidian objects from Domuztepe
plaques or beads wereFig.found. The

16. Remains of obsidian bowls from plaques
Domuztepe (1- are lo
enge-shaped, made from 4) and computer-generated
blades,reconstruction and (5)finely groun
and polished (fig. 15.2); one partly completed exam
a localized origin
ple demonstrates that they werein northern Iraq or northern
being made at th
site. One unique bead is
Syria, pierced at one
as has generally been supposed, it seems toend and
a deep incised spiral running down
have arisen initially through to ofits tip (f
the local adoption
15.1). There are also two miniature
Samarran-style axes
painted pottery, followed by themade
evo- of ob
sidian (fig. 15.3). Even more
lution of closely remarkable
related types of Early Halaf paintedare severa
fragments from obsidian vessels
pottery over a very wide area, (fig.
within which we16).now There i
one example of an obsidian vessel
need to include Domuztepe. from
In all probability, after the Bur
House at Arpachiyah52 butthat point it
in the is thick-walled
late sixth millennium, Domuztepe and wi
a rough, pecked exterior. The
was able to maintain a four examples
set of links reaching far to the from
Domuztepe are completely different.
southeast, which enabled it both to receive All and are th
walled (3-4 mm) and highly
transmit informationpolished.
on innovations in materialFrom
cul- th
small fragments, they appear
ture. By the Late to have
Halaf, seen in been
operation II,symmet
these
cal cups of an unparalleled
links are evidentquality (seeand
in pottery, architecture, comput
glyptic.
generated reconstruction,
The changes infig. 16.5)
material culture and
were clearly are th
adopted
most surprising find at inthe site.
a way that Regional
suited and was modified byvariation
the local
in itself, insufficient to explain
context. the
This pattern begins presence of th
to change in the latter
highly crafted items. They
half of themay be more
fifth millennium. specifical
In a recent publication,
the product of status competition. If,
Catherine Breniquet argues that the as
Halaf seems
cultures to
the case in the later Halaf,
of northernobsidian wasbecause
Mesopotamia disappeared in theyregular
supply and relatively common despite
were absorbed by the the
technologically superiorlong
Ubaid di
tance over which it was obtained,
culture, we
whose origins are to be found may
in southernsugg
that grinding and polishing was
Mesopotamia.53 She argues athatway of of
the knowledge addi
value to it, and enriching its
mudbrick potential
architecture for
and irrigation status
agriculture was d
play, by using a technology that
superior to Halaf couldandbe
pise architecture restrict
dry-farming
to a specific group. technology. The Ubaid cultures, because they could
support larger populations, simply absorbed the Halaf
CONCLUSIONS
populations, who left little, if any, trace in the ar-
Domuztepe sits on the edge of the chaeological
traditional dis-
record. Although it is unclear whether
tribution of Halafian sites. Far from the
these Halaf
cultures having
ever existed in the past in a form di-
52 Mallowan and Rose (supra n. 6) pl. Vc and fig. 44.15. (Paris 1996).
53 C. Breniquet, La disparition de la culture de Halaf

94.184.169.35 on Sun, 06 Dec 2020 06:02:23 UTC
rectly equivalent to our archaeological constructs,

determine the position of this central settlement. If
Breniquet's thesis is not clearly the survey evidence is at
supported an accurate
Domuz- indication, the
tepe. Instead, the Late Halaf surrounding is succeeded settlements
by may have been sparse. It
a Post-
Halaf phase of some substance seems, likely, minimal
with however, that the survey evidence only
(albeit
detectable) traces of Ubaid influence. Instead
gives a partial picture. of Do-groups, par-
More ephemeral
muztepe remaining in step with ticularlynorthern
nomadic and seminomadic
Mesopota- ones, were prob-
mia, we can see, to some extent, ably partaofchange
the constituency
in the that Domuztepe
axes served
of culture interaction toward the northern Levant. and exploited. The presence of these groups may be
This shift may suggest the critical influence of suggested
local by the evidence for the use of secondary
factors, not so much environmental as social. The products at the site, consistent with an important ele-
Halaf did not disappear in the Kahramanmara? re- ment of extensive herding in the economy.
gion in the same way as in northern Syria because it Domuztepe played a key role in the redistribution
remained advantageous in its particular local con- of goods. This role is visible in the case of obsidian: if
text and, possibly, because different options were we extrapolate from the evidence known from
northern Mesopotamia, the site distributed the raw
available for its manipulation. Although it is difficult
to document the end of the settlement of Domuz- material, to which it had special access, in processed
tepe precisely, it seems likely that it was largely aban-
or partly processed form. We may expect obsidian to
doned before the end of this Post-Halaf phase. Itbe is only one example of a wider range of materials.
These may have been raw materials obtainable only
tempting to suggest that this timing may not be coin-
cidental. There may be a link between the changing through long-distance exchange networks. Prestige
nature of transregional interaction and the long- goods were being produced at Domuztepe itself, in-
term stability of a settlement structure centered on
cluding items made from obsidian obtained through
Domuztepe. The subtle changes in ceramic stylelong-distance exchange. Such functions would pro-
may-and it remains no more than a hypothesis--
vide the settlement, and particularly elite groups
underlie a period when the social landscape manip-
within the settlement, with an important mechanism
ulated by elites to maintain their positions may by
havewhich to maintain power through economic and
shifted, initially in a manner that could be accommo-
status manipulation, and encourage the integration
dated but ultimately to an extent that required of aa regional entity. Equally, the need to negotiate ac-
general realignment of settlement in the region.cess
It to distant and widely separated sources of raw
was only at this point that the settlement at Domuz-
materials would provide multiple axes of increased
tepe was abandoned. interaction and social comparison through which
Our understanding of the social organization at
transregional relations could develop.
Domuztepe and its role in the region can only be
Internally, there are only hints at the social sys-
tentative. By far the largest site in the area, Domuz-
tems that permitted a site of this size to exist. We may
tepe would certainly have been the major social and,
suspect increased hierarchy but there is little detect-
probably, political and ritual center in the region.
able sign of it-probably the result of the small area
One puzzling aspect concerns the location of excavated.
the Realistically, we may never excavate an
site, on the edge of a plain rather than at its center
elite area of the site and need to find types of expla-
and with a substantial portion of its immediate catch-
nation that do not require this. The ways in which so-
ment covered by relatively barren hills. Initial calcu-
ciety was segmented may be a key. Kinship and com-
lations of the carrying capacity of the catchment
mon ritual performances such as burial rites and
group feasting may have united individual social
around the site suggest that it would have been very
difficult to support the population (estimated ongroups.
the These subgroups would have had their own
basis of site size) from local agriculture.54 Our esti-
interrelationships through which integration of the
mates may be flawed, there may have been a greater
entire community would have been achieved. In-
dependency on herding, or foodstuffs may have creasingly formal mechanisms to control relation-
been brought into the site on a large scale. This
ships between individuals and groups and over
does, however, emphasize that the location ofproperty
the ownership would have been needed as
site was not chosen purely on economic grounds.the number of individuals in the settlement would
Other social, economic, or ritual reasons helped tobeen far too great for each individual to have
have
54 C. Eissenstat, chapter in Domuztepe and Its Neighbors: Southeast Anatolia (Diss. Univ. of California, Los Angeles, in
Community Integration and Social Organization in Prehistoric preparation).

94.184.169.35 on Sun, 06 Dec 2020 06:02:23 UTC
418 S. CAMPBELL ET AL., THE DOMUZTEPE PROJECT, 1995-1997
personally negotiated relationships. A whole range ELIZABETH HEALEY

of techniques may have been adopted but they DEPARTMENT OF ART HISTORY
are represented particularly in extensive posses-

AND ARCHAEOLOGY
sion of seals. UNIVERSITY OF MANCHESTER
Domuztepe is certainly one of the largest pre-

MANCHESTER M 13 9PL
fourth millennium sites in the Near East and it al- ENGLAND
most certainly sat at the center of a rather extensive

ELIZ .HEALEY@BTINTERNET. COM
regional system of settlement. More important, how-

ever, is the evidence that it is providing for a com- SEONA ANDERSON
plex economic and social system for integrating INSTITUTE OF ARCHAEOLOGY
and maintaining this system. This evidence addsUNIVERSITY

a COLLEGE LONDON
crucial new element to our understanding of post-

31-34 GORDON SQUARE
Neolithic adaptations and should contribute to our
LONDON WC 1H OPY
understanding of critical social developments inENGLAND
the Near East. [email protected]
STUART CAMPBELL AMANDA KENNEDY

DEPARTMENT OF ART HISTORY AND
45 RIVERSIDE
ARCHAEOLOGY GUIDEPOST, SHOPPINGTON
UNIVERSITY OF MANCHESTER
NORTHUMBERLAND NE62 5PP
MANCHESTER M13 9PL ENGLAND
ENGLAND [email protected]
STUART. [email protected]
SARAH WHITCHER
ELIZABETH CARTER DEPARTMENT OF ARCHAEOLOGY
DEPARTMENT OF NEAR EASTERN LANGUAGES UNIVERSITY OF EDINBURGH
AND CULTURES INFIRMARY STREET
UNIVERSITY OF CALIFORNIA, LOS ANGELES EDINBURGH EH1 1LT
LOS ANGELES, CALIFORNIA 90095 SCOTLAND
[email protected]. EDU SWHITCH@HSY 1 .SSC.ED.AC.UK

94.184.169.35 on Sun, 06 Dec 2020 06:02:23 UTC

دوموز 2

Uploaded by

Copyright:

Available Formats

دوموز 2

Uploaded by

Document Information

Original Description:

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

دوموز 2

Uploaded by

Copyright:

Available Formats

Emerging Complexity on the Kahramanmaraş Plain, Turkey: The Domuztepe Project,

Stable URL: http://www.jstor.com/stable/506968

Archaeological Institute of America is collaborating with JSTOR to digitize, preserve and

This content downloaded from

SEONA ANDERSON, AMANDA KENNEDY, AND SARAH WHITCHER

Abstract have their roots in the Late Neolithic and Chalcolithic

This content downloaded from

Domuztepe Kazane Takyan

e Tell Turu Tell Sabi Abyab o Tell Aqab0

* More than 1Oha

Fig. 1. Distribution of selected sites with Halaf affinities

than 2-3 ha, and settlement hierarchies have been

This content downloaded from

11 P. Wattenmaker, personal communication 1996. 16 E. Carter, "The Kahramanmara? Archaeological Sur-

This content downloaded from

900- ???????? ?""?"""~" "" \ I I \

200 300 400 500 600 700

Fig. 2. Domuztepe: contour plan

This content downloaded from

200 300 400 500 600 700

Fig. 3. Density of Halaf pottery on the surface of

understanding of the development of

This content downloaded from

check the highest, and

mound, to investigate what were expected to be ear-

This content downloaded from

X:~'i_-:~%:_-i:i~:m : ::;il:i~~: ~::~l:~~r~~:~~li_ i(. - ::i--Ii 8~i::?::

Fig. 4. Operation II, Late Halaf tholos in its initial phase

This content downloaded from

I 96.5(2 0 96.4l .Am

This content downloaded from

pit was approximately 4 mpitwide at

24 A.L. Tobler, Excavations at Tepe Gawra 2 (PhiladelphiaArch 10 (1978) 125-28.

This content downloaded from

Fig. 7. Operation I, Post-Halaf B. Tholos w

Yarim Tepe II. In twoThere

This content downloaded from

... i... .. :: I I ..........

Fig. 8. Operation I, Post-Halaf B. This plan represents a slightly arbitrar

sloped upward over an empty channel to meet

This content downloaded from

wallfoundation, the main building

phase of occupation at the site, which will limit

This content downloaded from

3 Campbell (supra n. 3).

This content downloaded from

~(i(( ((?(? 1 of pffoopoppi tot I I??

,-- 0. 1111 - t -to -1 7,7

~(?(( rr??? rrr

Fig. 10. Classic Halaf pottery (1-5) and th

dark brown or black paint is usually

This content downloaded from

assemblage can be divided into three main phases

This content downloaded from

Fig. 12. Unu

Halaf small bichro

The assemblage has a slightly lower quantity of

34 Campbell (supra n. 4).

This content downloaded from

X:~'i_-:~%:_-i:i~:m : ::;il:i~~: ~::~l:r:~~li_ i(. - ::i--Ii 8~i::?::