An Introduction To Indus Writing
An Introduction To Indus Writing
An Introduction To Indus Writing
by
Bryan Wells
A THESIS
SUBMIïTED TO THE FACULTY OF GRADUATE STUDIES
IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE
DEGREE OF MASTER OF ARTS
DEPARTMENT OF ARCHAEOLOGY
CALGARY, ALBERTA
J M A R Y , 1998
The author has granted a non- L'auteur a accordé une licence non
exclusive licence dowing the exclusive permettant à la
National Liirary of Canada to Bibliothèque nationale du Canada de
reproduce, 10- distriiute or sell reproduire, prêter, distribuer ou
copies of this thesis m microform, vendre des copies de cette thèse sous
papa or electronic formats. la forme de microfiche/nlm, de
reproduction sur papier ou sur format
électronique.
This thesis uses the conjunctive approach to anaiyze inscriptions of the Indus
Valley Civilization (1600-3000 BC), presented in five chapten: Chapter 1 The
Archaeology of the Indus Civilization; Chapter 2 Inscribed Indus Artifacts; Chapter 3
Indus Sign List; Chapter 4 Analyzing Indus inscriptions; and Chapter 5 Reading Indus
Signs.
The purpose of interrelating these lines of evidence is to resolve basic issues
concerning the cultural and linguistic identity of the Indus (Harappan) people. The
intemal workings of Indus inscriptions are examined.
1conclude that the Language of the Indus inscriptions has many characteristics of
Proto-Dravidian as reconstructed. The subject matter of the majority of texts is
economic, and the script is logo-syilabic. Major ciifferences in sign distribitions can be
used to identiQ syntactic elements in Indus texts. The * k W a n reading in Chapter 5 is
used to relate case endings in PDr with specific signs.
Acknowledgments
The process of creating the Indus sign list used in this thesis took many hours of
collecting, editing, entering, checking, and double checking a i l data base entries. The
helpful suggestions and long suffering patience of Margaret Thornason of Calgary made
the Indus sign list data base possible. Thank you Dr. Radhakrishnan, University of
Calgary, Deparmient of Archaeology for your advice on Chapter 5. Cost of thesis
production and data input were offset by fianciai contributions by Neale E. Smith of
Douglas, Arizona and by Early Sites Research Society, Inc. (James P Whittall Jr,
Director). Thank you Susan Bonten for the loan of your cornputer which was used in al1
aspects of thesis research and production. To my mother Hazel who always clipped the
archaeology stories from the newspapers and magazines for me, and who helped me
through the tough times, thank you. Special thanks to Vladimir and Margaret Markotic
for their generous gift of Indus books and articles.
It was a pleasure and an honor to snidy with Dr. David H. Kelley, Professor
emeritus, University of Calgary, Sak Itzat. Thank you for your help and guidance.
Dedication
for
Vladimir Markotic who got me started
and
David Kelley who wouldn't let me quit
TABLE OF CONTENTS
Approvai Page.................................................................................................................... ... ii
Absmct.................... .....o......................................C...t............................... ........................ UI
Acknowledgments
. ......
.... ...... .
... .......................iv
Dedication........................................................................................................................... v
Table of Contents......................C....C,...........o..........o.............................. ................... ...
vi
List of Tables.............,L.......,,..,,...................C.................o.C..~................ .........
......... ..................
VU
List of Figures............ ......... ................................................................................... ix
CHAPTER 1: INDUS (3vLIZAnON............................................................................ 1
Introduction......... .........................*..........
.................................................... 1
The Archaeology of the Indus Vaiiey ....................... . . ....................... 2
CHAPTER 2: INSCRIBED INDUS ARTIFACTS...........
........................................... 20
Introduction............................ ................................................................ 20
Inscribed Artifact Types........................................................................... 22
Seais.......................................................................................................... 25
Type A...................... .......................................................................... 26
Field Objects................... . . ................................................. 27
Cult Objects.......................... ................................................ 28
Type B .................. .
... .................................................................. 29
Type C.......................................*.....*.....*.............................................29
Type D...................... . ................*...*...*.............................................. 30
Type E.................................................................................................31
Type F.................... . ...................*........*..**..................................... 31
Tablets.................... ....*...*...*....*..............*...*........*....*.*............................. 32
Type G a-b.......................... ........................................................... 32
Type G c-e ...................... . . . ........................................................... 33
Type J............................ ....~............. ....................................... 33
MisceUaneous Obiec . ts.............................................................................. 33
Type K ..................................................... .......................... 33
. . ,Vessels
Ceramic . . .......................
...... 33
Type L............................................................................................ 35
Type M............................................................................................... 35
Type N/O............................................................................................ 35
CHAPTER 3: INDUS.SIGN . .............................
. .LIST. . . . . . . 42
Approaches To The Indus Script............... .
..........................................43
Stroke Signs.............................................................................................. 47
The Sign List Data Base.................
...
................................................ 48
Expanding Parpola's Critena............. . ...*............................................ 49
Mahadevan's Sign 15.................................................... .................... 51
AppIying the Expanded Sign Critena...................................................... 52
Collateral R d t s.................... 54
Changes In OnentationTo Save Space: Sign414.................................. 54
The Range Of AUographic Variation In Sign 4....................................... 55
Compound Signs...................................................................................... 56
Internai Hatching...................................................................................... 57
Concl~on 59
CHAPTER 4: ANALYZING INDUS INSCRIPITONS.... ............................................. '78
The History Of 'IZie ShuchiralAnalysis of The lndus Script...-.....-..,- 85
. . .Inscriptions................................................
The Structure of The Indus
ExamhiqParaldI p s c n p h ~ ~ l ~ ~ , , , ~ . , . - . . . . . ~ . ~ - ~ ..--.....
91
91
. ..............................................................................
Column Analysis
Field Shiftuig ........ , , .-..-. ., ....
91
94
Field Contents........................................................................................%
EieldL-U-U..--I------------U1.I- ..-. -.... ..-.......-.. 97
Field II.......... ...................................................
YI
. .t+.eo. 99
..rir-.--.-..tUI..-.l-.W.-U.LI.....U.CIIYU.- . 1ûû
Conclusions............................................................................................ 101
CHAITER 5: READING INDUS SIGNS...................~..+.............................................. 114
Methods For Analyzing Inscriptions........ . ...
.... .......................... 114
Specific Cases...................................................................................... 116
Conclusionc..c
... 120
List of Tables
from the DKG Area and that seals are not distriiuted evenly across the site.
6
Chapter 2 makes the point that seds and tablets may have different functions, in
terms of social behavior, and that these difierences are reflected in the inscriptions.
Patterns in the distribution of seals through time (Figure 1.4) c m be interpreted using a
deiailed knowledge of the descriptions given in the text of Volume 1. For example,
MacKay ( 1938:43) d e s c n i s his excavations of the earliest deposits in Block 7 House 1.
Archaeological data from these early levels is sparse outside Block 7 (Figure 1.5).
Inscnbed artifacts are given special attention by MacKay (1938),and this is the largest
collection of usable raw data concerning the inscriptions of Mohenjo-daro.
. The Early levels of Block 1 ars extremely weli documented by MacKay (1938:44)
and his observations are often important:
1 have already stated that water level was reached at 38.5 ft. below
datum, but h m 35 fi. downwards a Iayer of stiff clay with occasional
pockets of grey sand is clear evidence of a flood. It is, however,
impossible to estimate the depth of the deposit Iefi by this inundation,
for it certainly extends well below the lowest water-level of the present
&y. The subsidence of the LittIe buiIding describeci above is in itself
proof of the occurrence of a flood of contmporary date, which cleariy
was the reason for the construction of the overlying pladorm of
sundried bricks before further building in this quartet was atternpted.
The deep excavations of DK.G Area were limited to the area immediately
surroundhg Block 7, and this included part of Block 1. MacKay's data can be used in
conjunction with other data to examine the disaibutional characteristics of artifacts. The
maps in -
,iF 1.4 show two major trends: one is a mesure of clustering (îegend bar
graph), the other is the locations of seais by house. Figure 1.4 a has an obvious clustering
7
of seais in Houses 1 and 2 of Block 1. This is the area at the center of MacKay's "wide
pit". Seai ftequencies here are elevated by virtue of the depth of excavation at this point
and the fact that these houses are foci of seal concentrations in several stratum of the
DK-G area.
MacKay gives the depth of al1 building platforms and foundations in his
discussion of the house by house description of his excavations. He relates these to the
Penods and Phases proposed by Marshail (1931), but he argues (1938:1-6) for a different
placement of the Period boundaries based on the relationships between architecturai
features, and evidence of inundations (Figure 1.3).
The decentralized pattern given in Figure 1.4 b is probably more representative of
the large scaie pattern of seal distribution (even for the earlier penod). The locations of
thick mud brick walls, lanes, and other feahires suggests that there were two types of
building clusters. The fmt of these consists of buildings of various sizes associated with
a courtyard and enclosed by a thick waU (Block 1 and 7, Figure 1.5). Four of the five
concentrations of seais in Figure 1.4 b are associated with this fust class of building
cluster (compound). 1 would suggest that these compounds are controlled by social
groups involved in speciaiized craft production. The basis of this social unit canot be
determined at present. but kin based systems seem the most probable given what is
known about the organization of other ancient cultures aud of later cultures in this region.
)the DKG Area are found in association with the
in
Kilns in use during this period (III
courtyards of enclosed compounds.
The second class of building group consists of rows or clusters of buildings with
near identical shapes, sizes, and plans. These row houses are not associated with
courtyards and enclosing walls. This simple dichotomy accounts for most of the
buildings in the DKG Area of Mohenjo-daro. None of the seal concentrations in Figure
1.4 b are associated with this second class of building cluster, although seais are found in
Rao's 1973 report of the excavation of Lothal supplies no raw data, although Rao's
observations and conclusions are informative about Late Harappan frontier sites.
Additionally, his excavations supply us with the Iargest collection of tags (Type N and 0)
from a single Indus site, and detailed information regarding the organization of
specialized production. The Indus inscriptions may annotate either destinations of
shipments of goods andlor descriptions of the goods to be transponed, or both of these.
Evidence for this cornes from the seal impressions from Lothal. Many of the seal
impressions at Lothal have impressions of a coarse cloth on their reverse sides. The sheer
numbers of tags from Lothal indicate these objects were in common use. Lothal had a
wharf, workers barracks, an elite residence and a bead factory. This is essentiaiiy the
pattern descnid for Mohenjdaro, but at a much smaller scale.
Published archaeological data has ken supplemented rnost recently by the 1986-
1990 excavations of Harappa (Dales and Kenoyer 1991, and othea). This publication
(Meadow 1991 (ed.)) gives the first processuaiist led (Dales et al 1977, 1986, 1990, and
1991; and Kenoyer 199la, 1991b) multi-disciplinary excavation of a major Indus site.
These reports make avaiiable the fmt comprehensive forensic and faunai data as weii as
snidies of craft speciaiization and aide. Kenoyer (199lb:364) tells us that inscribed
artifacts were found "...dong major access routes and main streets" at Harappa Yet
Belcher (1991: 109, fig. 8.2) found 10 inscnbed artifacu in his 5 m by 10 m unit at
Mound E, Area C, of Harappa. This unit exposed a work area amund a kiln.
Damage to Indus sites is si-cant and widespread. Mohenjo-daro and
especialiy Harappa have semed as sources for baked mud bricks both for local people and
for the construction of British Indian railways. At Harappa brick robbing has removed
10
most of the late deposits. At Mohenjo-daro the high water table prevents excavation of
the earliest deposits. These circumstances leave us with incomplete sequences at both
sites. Further, there are large paps in the sequence for which there are no archaeological
data. For this reason the chronology of the Indus Tradition is somewhat disjointed and
chronologicai models are necessarily general (Figure 1.3).
Research into the nature of the Indus script over the last 25 years has examined
many aspects of the script including its stmctun. and the possible linguistic identity of
the people who inhabited the Indus Valley before the Vedic age. Most researchers, for
various reasons. have assumed that the language of the script was some early form of
Dravidian.2 This is by no means the only solution to the question of root language, but
given present evidence it does seem the most viable solution. Zvelebil(1972b), McAipin
(198 1) and othen (Parpola 1994; Fairservis 1992; and Knorosov 1968) present linguistic,
historical, and archaeological evidence supporting the Draviciian solution, while Hemphi11
et al ( 1991) present metric and non-rnetric anaiysis of the hurnan remains from Harappa
that agee in general terms with the linguistic, historical, and archaeological data.
Earlier models of invasions of Vedic Aryans (Wheeler 1959; Piggott 1952) have
not s u ~ v e drecent research. These models relied heavily on accepting specifc sets of
data (the massacre in Deadman's Ailey Mohenjo-daro) as support for the Aryan invasion
model. As mon details of Indus prehistory were published it became evident that these
models were conceptuaily too simple. Additionaily, increasing control of the chronology
and archaeology has made it clear that the corning of the Aryans was 500 or more yean
after the end of the Integration Era (Table 1.1). During this period (Localkation Era) the
population density of the Indus valiey was much reduced due to deurbanization of Major
Indus centers. Cultural development was diverse with localized styles and technology
developing dong distinct trajectories. The Indus civilization had long disappeared when
the nrSt Vedic honemen arrived to stay in the Indus d e y .
11
A recent summary of Indus valley archaeology is Kenoyer's (1991a:331-385)
article 'The Indus Valley Tradition of Pakistan and Western India' which suggests a
general chronological frame work for the development of the Indus civilization. The
chronology of the Indus Valley Tradition as given by Kenoyer (1991a:333) is
summrized in Table 1.1.
Table 1. I Chronology of the Indus Valley Tradition.
Indus Tradition
Localization Era 1900 to 1300 BC
htegration Era 2600 to 1900 BC
Regionalization Era ca. 5000 to 2600 BC
Early Food Roducing Era ca, 6500 to 5000 BC
A chronological model can be created using several lines of evidence. This model
would have Proto-Dravidian speakers moving to the Indus Valley from the Iranian
Plateau between 6500 BC and 4500 BC. Archaeological evidence for this early
occupation of the indus Valley cornes from the site of Mehrgarh (ca 6000 BC). Matenal
excavated from aceramic deposits (pre 4500 BC) at Mehrgarh include six-row barley,
einkom, emmer, and durum-bread wheat (Costanthi: 1984). Later levels show an
increase in wheat utilization, and the introduction of ceramics and other technological
innovations which lead from the Early Food Producing Era (pre 5000 BC) to the
Regionalization Era (5000 - 2600 BC) of the Indus Tradition (6500 - 1300 B. C.)
(Kenoyer, 1991x341).
The process of regionalization resuits in the in situ development of regional
variations of Indus culture. This process includes increasing urbanism, social
stratification, craft speciabzation, inter and intra regional trade, and a reliance on
agriculture. There is a comsponding decrease in regional variations in cultures over
t h e . These trends culminate in the Integrution Era (2600 - 1900 BC) with the
establishment of recognizably Indus culture (Harappan Phase) including state ievel
12
organization and writing. The excavated deposits at Mohenjo-daro al1 date to the
Integration Era or later.
The lndus tradition (Harappan Civilization) is recognizably distinct fiom
contemporaneous West and Southeast Asian cultures, and represents an in situ
development over 4000 years. That is not to Say that the Indus tradition developed in
isolation. There is ample evidence for contacts between Mesopotamia and the Indus
valley fiom the time of Sargon (c. 2300 BC), if you accept the equivalence of the Indus
valley and Meluhha (Parpola, 1994: pp. 12-15). The fact that al1 nferences to Meluhhans
in Akkadian texts give hem Sumerian names may point to even earlier contacts between
these two areas.
According to Hemphill et al (1991) following the Chaicholithic settlernent of the
Indus Valley (6000 - 4500 BC), there is a long period of biological continuity which las&
(at Harappa) until between 800 and 200 BC. The biological model suggests bat.
following the end of the Indus Tradition (c. 1300 BC), there was a period of continuity in
human populations at Harappa which was not interrupted until after 800 BC. The
population discontinuity is likely related to the arrivai of the geneticaily different Indo-
European speaking Vedic people. The fact that the biological model presented by
Hemphill et al (1991) and the linguistic mode1 proposed by Zvelebil (1972) and othen
agree so closely lends credence to both lines of evidence.
3 Type G anifacts (Fi- 2 1: Bas Relief Tablets) are most common at Harappa, while Type J artifacir
(Copper Tables) are found only at Mohenjodm.
13
environmental and social circumstances, and these difierences may include regional
didects of Proto-Dravidian.
Figure 1.1 Indus Region and Indus Texts By Site
Figure 1.2 Some Applications For Data Frorn MacKay 1938
Years
B.C. Mesopotamia indus Vrilley Mohenjo-daro
ParpoIa( 1994) PacpoIa(l994) (Wells 1997)
Iron Age
Middle
Babylonian
Post
Harappan
.. . .. . . --
Old Late Harappan
Babvlonian
Period IiI
Period 11
Period 1
Unexcavated Unexcavated
Uruk IV
Figure 1.4 The Distribution of Seals For DK.G Area Mohenjo-daro
c. Period IV d. AU Seals
1 -. i 7
Fi,pre 1.6 Seals and Weights From DK.G, Block 7, House II
Spaulding (1953), Rouse (1960) and others have in part inspired the new
archaeology and the quantitative generation of multivariate cornputer analysis and
electronic data bases, which dominated the Iiteraaire of the 1970s and 80s. Among these
22
Christenson and Read (1977) stand out as the most austere. However, the new
archaeology found its most explicit expression in Essays On Archaeologicol Typology,
edited by WhaUon and Brown (1982). Here the application of statistical techniques
dominates. Papen by Spaulding and by Hodson extol the use of amibute-association vs.
object clustering techniques respectively, while Cowgill's paper compares the pros and
cons of such methods.
Yet in spite of this comprehensive set of papers, the sum of their agreement
amounts to the following definition of type as: "...a group or class of items that are
intemaily cohesive and separated from other groups by one or more discontinuities "
(Whallon and Brown 1982:.xvii). The most important contribution of this work (in the
editon' opinion) was that it made: "...some fundamental lines of disagreement explicit ...
and highlighted: the role of such differences in shaping approaches and methodologies in
this subject." (ibid). Adams considers this debate a false dichotomy in that:
A11 types are essential in the sense that they are objectivcly definable,
but insaumcntal in that WC would not retain them if they did not serve
some purpose; most of them are based to some extent on initiai gestalts
that are subsequently objectified by rational analysis...and they have
usually evoIved through continual dialectic, or feedback, betwetn
induction and dcduction, object clustering and attribute-clustering,
lumping and splitting. (Adams,1988:45)
So it seems that whüe we aU use types and typologies in our everyday lives, the
explicit description and dennition of the typological process is more elusive. Adams and
Adams (19915) argue: " It is therefore impossible to talk about types and typologies
except in subjective t e m . We cannot speak of the concept but only our concepts."
I Marshall (1937: 576-7) describes faience as a coime paste with a greenish glare. Chemically it is 58%
silica and 28%magnesia. A relateci material vitreous pare is distinguishabfe by its fine grain silica and the
absence of Magnesia visible in faience as btack specks.
Boih BRTs and Tags are baked seai impressions but the Tags are not mold-made. They are commonly
roughiy shaped chunks of bumt cfay with seai imp&ons on the obverse and textile or rced impressions
on the reverse. Unlike BRTs, Tags are usudly nred accidentally.
24
Lothal stands out from the usual pattern of Type A seals being the sole dominant
form. Type C seals are unusuaiiy abundant fÎom this site, being twice as cornmon
(36.29% of all seals) as compared to either M o h e n j W o (14.35%)or Harappa (15.22%).
This shift rnay be related to the relatively late date of Lothal or may reflect a regionai
25
variation in seal usage. Other possibilities are that the Cult and Field objecd mark
ownenhip or place names. These possibilities &es sense in that there are signincant
variations in the design of both Cult and Field objects (Figure 2.2 and 2.4). Tags are also
very abundant at Lothal(32.40%). Harappa has a large number of Type G, Type H, and
Type J artifacts (Tablets). These locally cornmon artifacts are present at other sites, but
represent a relatively small percentage of the total artifacts outside these areas of
concentration,
This typology defines sets of artifacts that have sidar physical fonn, assuming a
linkage of axtifact form and artifact uses in antiquity. The hope is that their inscriptions
are concemed with sirnilar topics and will share parallel structures. These parailel
structures are the subject of Chapter 4. Thus, this typology is the fust step toward the
structural analysis of the Indus inscriptions.
Seals
For the purposes of this discussion a seal is defined as any artifact with an
intaglio4 inscription or design. The implication is that all these objects were used to rnake
impressions, and that these impressions were subsequently read. Seals cari be made fiom
a variety of materials. The most cornmon materials are: steatite, faience, vitreous paste,
and siiver.
The vast majority of inscribed artifacts corne to us from the early excavations
carried out by the Archaeologicai S w e y of India. MacKay (1938: 325-326) follows
Marshall's (1937) typology for the seals excavated nom the DICG area at Mohenjo-daro.
Marshall's typology represents the fint effort dirrcted at identifying and classifying seals.
Table 2.2 gives the combined tallies of seals within Marshali's system for the 1927 to
1931 excavations at Mohenjo-daro:
As defined in the following sections labled 'Cult Objects' and 'Field ObjectS.
carved in negative image.
Table 2.2 Seals excavated from Mohenjo-daro 1927 to 1931.
MacKav arshdl Total
(a) Cylinder seais 2 5 7
(b) Square seals with perforated boss 558 328t 886
(c) Square seals with no boss 12 12 24
(d) Rectangular seais with no boss 17 6 23
(e) Button seals with Linear designs 9 22 31
(f) Rectangular seais with perf. convex back 81 64 145
(g) Cube Seals 3 5 8
(h) Round seai with perforated boss 1 3 4
(i) Rectangular seals with perforated boss 2 1 3
QI Round seal with no boss 1 1 2
Totals 686 447 1133
iMarshall's "perfect specimens" (Marshall, 1937: 372).
The main drawback of Marshall's typology is that it ignores the seal's design
elements, and focuses on the seai's handle (boss) instead. This approach places Type B
seals and Type A seals in the same category. This grouping is unacceptable given the
difference between these types. Other attempts (Shah and Parpola 1987; Joshi and
Parpola 1991) at classiQing inscribed artifacts have likewise focused on the shape or
handles of the seals, while ignoring the faces of the anifacts.
Seals were used to make impressions in wet clay as a means of seaiing shipments
of goods (Kelley and Wells 1995). We know this from the many tags found at
Kalibangan, Lothal, and Mohenjo-daro. Regardless of boss style the majority of seals
were designed to be stning together for use as needed. That several seais were needed to
seal a single leoad of goods raises the possibility that individual seals contain ody part of
the necessary information. The implication is that each seal contains only part of the
message and that whole messages are consmicted by combininp several seal impressions.
This matter is p m e d M e r in Chapter 4.
Field objects
Field objects are nomiaily animais or groups of animals, often in profile, depicted
either standing or crouched to feed, and bearing recognizable sets of
2.2). For example, the Type 1Buii has markings similar to an upper case W.
markings are repeated at all major Indus sites. Neck markings Vary from simple
necklaces to necklaces combined with cornpiicated sets of curved parailel and parallel
wavy Lines. The two horned buil (Type 3 Field object) is also a Type A seal.
Corpus of Indus Seals and Insaiptions. Joahi and Parpola 1987 and Shah and Parpola 1991
28
There are significant differences in the distribution of Field objects
pographicaiiy. Expressed as the percentage of al1 seals h m each site, Type 1 Bulls are
proportionaily most common at Harappa, while Type 2 Buils (which are rare at Harappa)
are most common at Mohenjo-daro. Kalibangan and Chanhujo-daro have approximately
equal proportions of Type 1 and 2 bulls.
Beyond these two types of Field objects there are 23 other reco,@zable motifs
(Figure 2.2). These Field objects depict animals (Types 1-19) and humans involved in
various activities (Types 20-25).
Cult Objects
The identity or meaning of cult objects is more enigmatic, while their
classification is easier than the classification of Field objects. The goal of this
classification of Cult objects is two-fold: 1) to defke the design elements that are used on
Cuit objects; and 2) to identiq combinations of elements that CO-occur(Figure 2.4).
Cult objects consist of two components: Tops and Bottoms. Tops can be either
rounded or square and cm bear various patterns. Rounded Tops (15%) and square Tops
(85%) have similar percentages from both Mohenjo-dam and Harappa. Cult object
Bottoms can have one of three shapes and may or may not have nills on the lower edge.
These variations create six categories into which Bottoms can be placed and two
categories for Tops. Additionally, Bottoms can have one of several designs. There are
375 seals fiom Mohenjo-daro and 128 seals fiom Harappa that have recognizable and
complete Cult objects. There are 408 possible combinations of shapes and patterns for
Tops and Bottoms of cult objects, but only 75 combinations are used at Mohenjo-daro
and 46 combinatioas at Harappa. This indicates the preferential selection of certain
combinations. Figure 2.4 gives the number of occurrences for Top and Bottom
combinations expressed as a percent of the total Cult objects for Mohenjo-daro and
Harappa
29
The most common Top and Bottom combinations for cult objects are:
The significance of these combinations is not clear. They may be simple stylistic
variations or they may carry some meaning.
The relationship between Cult and Field object is not certain. They do
coliocate preferentially and then are certain combinations that are far more common.
This suggesu that they individually convey information which can only be combined in
certain ways. This information must relate to the huiction of seals within the Indus made
network, as Cult and Field objects rarely occur except on Type A and Type B seals.
materiais: steatite 59; faience 12; vitreous paste 1; silver 1. The shapes of both Type A
and Type B seals are distinctive as the foliowing graphs show:
O ols i 2 :
HorizonmiS i In M e s
The distinctive shapes of these seals are a function of the carvings on their faces.
The Type C seals contain only inscriptions without the iconographie elements typical of
Type A B seals. Consequently, Type C seals are wider than they are high. This type of
seal occun at most major sites including Mohenjo-duo (168), Harappa (86),and Lothal
(45). Chapter 4 wili demonstrate that the sign sequences of Type C Seals are different
from Type A sign sequences. This is sufficient evidence to postdate a difference in
subject matter.
arappa are the most varied in design having 14 different motifs. A tag
fiom Lothal (L-174)has an impression identicai to B119 (pictund here).
Only Types A, B,C and D seal impressions are found on tags.
Type E: Cylinder seals.
There are four cylinder seals from Indus sites: three from Mohenjo-ciaro and one
from Chanhujo-daro. Only M-1370has an inscription. Parpola (1994: 1 8 4188) tells us
that Indus style cylinder seals are found widely in the Near East (Tell as-Sulema, Teil
Asrnar, Ur, and Susa). There is a stronger and older cylinder seal tradition in
Mesopotamia than there is in the Indus Valley.
Tablets
Tablets can be defhed as inscrïbed objects with positive, as opposed to intaglio,
images a d o r inscriptions. These objects have no recognllable utilitarian or decorative
function (as do 2.3 Misc. Objects). Tablets are often mold made,7 incised (Type Hii),
,
Bibby T.G. (1972) for descriptions of Dilmun seais.
Bas Relief
32
etched in copper (Type J), or impressed (Type G.a and b). The following table (2.3)
gives the distribution of tablets by type.
Table 2.3 Tabulation of Tablets
Mohenio-daro Haran~a Tot4
Type G (Bas Relief) (121) (309) (430)
a. Round Impression 14 O 14
b. Square Impressions 14 3 17
c. Rect. Mold (no icon.) 56 159 315
d. Rect. Mold (twisted) 4 8 12
e. Rect. Mold (iconography) 5 6 11
f. Cylinder Mold 1 52 53
g. Ovate 5 18 23
h. Prismatic 13 17 30
i. Cube With Grid 7 O 7
j. Round Imp. (not Type A) 2 48
Type H (Incised Rect.) il 139
46 150
Type 1 (Shaped) 1 40 41
Type J (Copper Wafer) 237 O 237
Total 370 488 858
made, and fied. They have signs andor iconography on al1 t h e faces. Type G.i are
cube shaped fïred tablets with a simple grid design. Type G.j round seal impressions(not
Type A sed) cm also have the v I
and one to four signs on their reverse.
Type H tablets are incised and without iconography. This type of tablet is very
cornmon (n=139) at Harappa. They are sirnilar to Type C seals but the inscriptions are
not intaglio. Type 1 tablets are also hcised but the artifacts themselves have distinctive
and specific shapes. Some of these are zoomorphic, mostly shaped like a hare or fish.
Ceramic vessels fkom Indus sites bear a variety of inscriptions ran,@g from a
single sign (presumed potter's marks), to lengthy inscriptions, and seal impressions. The
longest inscriptions on ceramic vessels (6 signs) are of the incised variety and corne nom
(Figure 2.3).
The implication of the longer incised inscriptions is that the potters who scratched
them into the vessels were literate enough to compose sign strings of comparable lenath
to the seal inscriptions. This implies wide-spread literacy among at least this segment of
the Indus population. niese vessels are not elite ware but utilitarian ceramics.
Stamped vessels are found at both Mohenjo-daro and Harappa, although they are
more common at the former site. These inscriptions are surprisingly uniform in form but
not in content. Al1 of the stamped inscriptions lack the iconography common to the Type
A and B seals. hstead these impressions nsemble Type C seals, but with much shorter
inscriptions (usuaily thee signs). There are 13 of these inscriptions Mohenjo-duo
and two fiom Harappa With the exception of M-1382 and M-424 these inscriptions do
not repeat The most probable explmation for the meaning of these inscriptions is that
they are the names of the ownen. Because the inscriptions are mostly unique, it is
unlikely that they refer to the contents of the vessels. Pot sherds with painted Indus signs
exist but are very rare.
35
Type L: Adornments
Bangles and beads also bear Indus inscriptions. While most of these are only one
or two signs in length, some (5) are longer. The longest inscription on a bangle is five
signs long. These sign sequences are very different in content from either seais, tableu,
or ceramics.
Inset
Type A Inscription
r Key
2.1 S& 2.1.1 TjpcARccmoguiar(wFBtC) 23Misc 23.1 Type K Utiiïtarïan
31.1-1 field objets 2.3.1.1 Type Ica Concs
2 1. I 3 Cuit objccts 23.12 Type K b CcrPmics
313 ~ y p B e RccrimgPlar (w neld ody) 1- S m d
2 . 1 7 1 TyptBnAnimalwII1EC. 2. G d t i
2.133 TypcB.b ûtkr 23-13 Type Kc Coppcr Objccts
313 Typc C RccmgurPr(no F&C) 232 T- -m L A d o m n s
31.4 Type D Square Design '32.1 Type La Bmgits
2-13 Type E Cyiinder 2.322 Type L b Bcnds
= TabIcts 21.6 Type F Roud
33 1 Type G Bas Relief
3 2 i . I TypeGa RomdwTypeAhp,
233 TypcM Olha
733.1 Type M Ivory Rods
2
.4 T;igs",Q,l Type N Multiple Irnprrssions
7 2 1 3 TmG.b SqtuutwTypcAlmp. 24.1.1 Typt Na Type A impression
3 3 1 3 Type G.c Rcct w-out Iconography 24-13 Type N.b Typc D impression
7221.4 T m G.d Twisaed 2.43 Typc O Siagle impression
2'>13T+G.C -W rtooography 2.42 1 Typc 0.a Type A impnssion
2.2.1.6 Type G.f Cylindcr 2433 Typc 0.b Typc D impression
2 2 - 1 3 TypcG.gOYPtC
331.8 T w G.h Rismshapd
3213 TmG.i Sqpp~wgrid
' 2 1.10 Type G-j Round moldcd
2?2 Type H InciKd b
22.3 Type 1 Irtciscd specific ShapC
3 2 4 Type J Coppcrwafér
Figure 2.2 Field Objects - Part 1 :Type A, Type B and Type F Seals
Type Z
Minor Sites
Chapter 3
Indus Sign List
This sign List (Table 3.1 and 3.2) is based primarily on the Corpus of Indus Seuls
and Inscriptions (Joshi and Parpola: 1987; Shah and Parpola: 1991). The corpus
represents the largest collection of Indus inscriptions currentiy available. With the
exception of the copper tablets fiorn Mohenjo-darot signs were generated mostly from
the photographic corpus. For copper tablet inscriptions, the published drawings
(Marshall: 1931; MacKay: 19%; and Parpola, 1994: 111-2) were used because
deterioration to the original artifacts (as seen in the photographic corpus) has removed
much of their surface detail. The corpus photographs were used to confum the
reliabüity of the drawings of the copper tablets where possible.
Unfortunately some anifacts are missing from the corpus, for exarnple SD
2172: .(Manhall 1937:PI.CXII, 385). Manhall tells us that whiie excavating the
foundation of Chamber 36 in the Great Bath Area of Mohenjo-duo: "There was
unearthed a steatite seal (SD 2172) bearing an unusual device" (Marshall:Volume 1,
p.136). Other minor omissions exist and the original site reports were used to
supplement the corpus where these omissions were recognized. Regardless of its
shortcomings the Photogrophic Corpus offers superior photography of both the original
seals and their impressions, and so remains the best source of Indus inscriptions. The
planned third volume of the Photographic Corpus is not yet available and is to include
inscriptions fiom pnvate collections and excavated objects omitted from the f î t two
volumes. The corpus is linked to the original site reports using the artifact field number.
This system is cumbenome and effectively isolates the artifacts from their
archaeological provenience. This makes andysis of temporal and sitebased studies
nifncult, if not impossible, in most cases.
43
In the following pages inscriptions are referenced using the photographic corpus
system. This consists of a site code and a sequence number. For example, M-1 is the
f i t seai in Volume 1 of the corpus (Joshi and Parpola, 1987). M is its site code, in this
case Mohenjo-daro, and 1 is the artifact sequence number. The following table lis& site
codes for major sites fiom both volumes of the photographic corpus.
Table 3.3 Site Codes and Artifact Numbers
Code Site Name Volume 1 Volume 2 Total
M Mohenjdaro M-1to 620 M-595.621 to 1659 1659
H Hara~~a H-1to 264,278 to 382 H-266 to 275,356,383io 1019 1019
L Lothal L-1 to 290 290
K Kaiibangan K-1 to 122 122
C Chanhujo-dato C-1 to 50 50
B Banawaii B-t to 37 37
Rdh Rahman-dheri Rhd-1 to 270 270
Pk Pirak Pk-1 to 49 49
Minor sites: PP 350 to 363 52
stands for them in the graphemic sense when analyulg context. This factor reduces the
detaii of the structurai analysis based on these generalized sign Est.. In the foliowing
discussion of sign contexts I give details of why Mahadevan's (1977) sign list of 417
signs does not adequately represent some signincant and demonstlable clustering of sign
44
sub-varieties in the Indus inscriptions. The second problem is that he relies too heavily
on visual sirnilarities of sign graphs in denning his sign list.
More recently Asko Parpola (1994) has tried to fonnalize techniques for
analyzing the Indus script. Parpola (1994:68-79) outhes three criteria 'Tor idenuQing
two or more graphs as variants of a single grapheme":
2) Two signs meeting the above condition occur only in almosr identical
contexîs.
See Qiapter~.
* ~hapter4
Mirror signs may indicate a different ieaduig order (lefi to right) or may signai a reversai of the normal
graphemic value of that sign.
46
signs with multiple markings. The sign rypology (Figure 3.1) is multi-tiered, consisting
of sign Class and Type. Sign classes are defined as foilows:
Simple Signs (SIM): A simple sign consists of a defiable Indus grapheme
without any elaboration. They are the simplest form of a sign. There are 127 simple
Indus signs. Simple signs are further divided into Stroke and Other signs.
Complex Signs (CMX): An elaborated f o m of a simple sign using elements
which are not themseives signs. Intemal hatching is a common type of elaboration.
There are 175 complex Indus signs.
Compound Signs (0Two
) known
: signs combine to form a single sign.
There are 135 Indus signs that fa11 into this class. These signs can be Attached,
Conflated, Doubled, Wied, Minor, Multiple or Other.
Marked Sigas: (MKD): The addition of a fmed set of elements to a simple,
complex, or compound sign. There are I l marks found in Indus inscriptions, and 146
marked Indus signs. Marked signs behave differently fiom their unmarked counterparts.
In 4 cases more than one sign is enclosed by markings: 126 Rh), 3fi:?
145 226
Indus signs cm be given one of the 18 different sets of markings used in the
inscriptions. Some signs (148: 3)do not exist in their unmarked state. In s o w cases
the boundaries betweeo marked and compound signs are blvrred Sign 405 @is sign
392 O (compound). The same logic can be applied to ovates such as sign 357
(marked) and sign 32 A@(compound). 'A
S
Seveo signs have more than one typt of rnarking. For example sign 119 t&l is
marked with both a superfix and a cage. Chapter 4 examines in detail the effects of the
addition of various markings to Indus signs. In tems of the sign list the typology as
aven in Figure 3.1 applies.
47
Marker Signs (MKR): Marker signs were recognized by Parpola (1994:88-92)
as being elements that separate gmups of signs within inscriptions. There are t h e well
known markers: 193, 196 and 23 1 (Fi,- 3.1). Signs 193 (n= 180) and 196 (n=û2)are
the most problematic of this class of signs as they may be markers in one context and
numencal in another. Sign 23 1 (n=120) has two varieties and can Vary greatly in size.
In addition to the descriptions above, some signs have examples in rnirmr image,
and these occur only in the fmt four classes given above. Minor signs rnay be the result
of a revend of reading order, but this accounts for only 5 of the 42 examples of mimr
image signs. For 10 examples reading order cannot be determined. The majority of
mirror signs (27) are found in inscriptions with normal reading order (right to Mt). If
these are syllabic signs, then reversed graphs may have the CV4 order reversed to create
VC syllables.
In 162 inscriptions 26 signs are doubled. These signs are defined within the sign
typology as a separate Type of compound sign. The function of doubling is unknown
but it may be a way of creating CVCV or CVC syllables. Both the mirrored and
doubled signs need M e r study.
Stroke Signs
There are 37 types simple stroke signs in the Indus inscriptions. Stroke signs can
be categorized into three groups: long strokes (signs 198, 194,195?205,209,220, 221,
222, -24, and 2 12); short stmkes (signs 193, 196,197,199,200,20 1,202,203,206,207,
208, 210, 21 1, 217, 218, 219, and 227); and staggered strokes (229 and 230). Long
strokes seem to represent numerals fiom one to six ody. Sign 194 can take attachments
(221,222, and 228) and superfixes (220). Short stroke signs representing numerals cm
be either linear (196, 193, 197,200,202,217,210,214, and 215) representing 1 through
7, or stacked (206,218, 21 1, 203, 199, 208,219) representing 1 through 9. Some of
48
these signs can also receive markings (204, 214, 215. 216, 223, and 225). Staggered
kz
signs represent the numerals 8 and 9 only. 1 wouid argue that these various types of
simple stroke signs represent values in the Indus system of weights and measures, and
not numerals per se.
These sorts of searches can target specific signs. For example, sign 113
the foliowing distribution: n=179; Mohenjo-daro 85 (47.49%); Harappa 78 (4358%);
Lothd 6 (3.35%); Minor sites 10 (559%). These nurnbers show that sign 113 is much
is found primarily on bas relief tablets, which are more cornmon at that site. Signs 289
and 241 show higher than expected fkquencies at Harappa, but this cannot be Iinked to
a speciai case as with sign 289. Signs 289 and 241 may be more frequent than expected
at Harappa for a variety of reasons. Regional differences in trade practices or dialect are
the most likely reasons. These maps disprove the long held opinion that the Indus scnpt
is homogenous with standardized uscge.
examples of sign 11 96 are found on bas relief tablets fiom Harappa (Figure 2.1: Type
G.f) and four of these tablets arr fkom the same mold. There are eleven inscriptions but
51
only 7 separate examples of this sign for that reason. This weights consideration of sub-
variety unifonnity when defining standard sign graphs (Table 3.1). Examinations of the
context of specifc inscriptions m u t include their archaeological contexts as far as these
are known -particularly the artifact type (Figure 2.1). Structural contexts should be
evaiuated after the grouping of inscriptions by artifact type.
Cnterion 3. The existence of ligatures which behave in the same way as their
component signs are rare in Indus inscriptions. Many ligatures in the Indus script are
accompanied by shifts in the location of the resultant sign to a different field? For
example: sign 582
A
t shifts right when it is infiied in sign 213 O to form the marked
W? When sign 194 is rnarked with a single 'caget (sign 204) .0.it moves left
b A b
sign 194
in the inscription. Analysis of marked signs and their contexts demonstrates that
specific additions to basic sign graphs result in predictable changes in sign positioning
within inscriptions. This critenon should be reworded to read: Graphemic elements will
d e r n o m t e the same range of allographic variation regardless of their Class and Type.
CriteBon 4. This criterion extends considerations of context to include the geographic
and temporal provenience of an inscription as far as they are known. One possibility is
that signs with limited distributions are regional aliographs. Systems of wnting undergo
changes over t h e and space. Some sigps have unique variations which are tempordy
andlor geographicaily resmcted.
Mahadevan's Sign 15
Mahadevan's sign #15lO is often described as the ligature of three graphic
elements: U+ k + O0 . ~ h fci t two are common-=cius signr. m e 67)never
OCCLUS atone, the graphically similar signs @, OP,do,0,and Sa do occw as
independent signs.
-- into four groups (Figure3.2). Three of these groups require further analysis:
be divided
Sign 5 @- those with amis and a canying pole. The arms can take 6 forms, and there
tF
is no noticeable Link between arm shape and context; Sign 8 - those with no arms,
w
but which have a carrying pole; and sign 11 - those with neither arms nor carrying
pole.
2) The contexts of these inscriptions have strong patterns of association:
Sign 5 Shares contexts with sign 8 ., but not with sign 1 1:
3e
H-152
unique contexts:
M-379
3) Analysis of ligatures is not applicable for signs 5, 8, and 11 but is for sign 343
discussed under Coliaterai Resuits.
4) There are definite locational preferences in terms of sign use and artifact type (Figure
3.3). This correlation of sign to artifact type is not perfect but is noticeable in Figure
3.3. Sign 5 occurs most often on Type A seals nom Mohenjo-daro (n=17). There is
only one example of this sign that is not from a Type A seal and that is found on a Type
G tablet from Mohenjo-duo. At Harappa sign 5 occurs 7 times. There are ten
occumnces of sign 8 from Harappa, 7 from Mohenjo-daro, and 2 from Lothal. Most
Harappan examples of sign 8 are korn Type B seais. Sign 11 is also most common at
Harappa (n=9). Whether these patterns are related to regional variations in style or to
differences in the subject matter of the inscriptions is not clear.
We know fiom Chapter 2 that bas relief tablets are more abundant at Harappa
(309) than at Mohenjo-daro (121). Whatever the use of tablets in antiquity, this practice
was far more important at Harappa than at Mohenjo-daro. If Parpola is right (Parpola
1994:10%109) then this disparity may indicate differences in religious practices within
the Indus valIey. If bas relief tablets are exchange tokens, then the difference might
indicate that Harappa was more highly organized economically, and politically more
The sign sequence: 00b" occurs five thes at Mohenjo-dam and 17 times
at Harappa, with si= 8 and 11 CO-varying,and signs 436 to 446 occurring in most
varieties. Sign 5 never occurs in this context The most constant element in this sign
-
cluster is the middle sign (343) . This is never repiaced by sign 344
confirmation that signs 343 and 344 are different graphemes and not allographic
variations as Parpola (1994:17) suggests. Likewise signs 262
different signs, using the expanded criterion number three.
The variations in signs 436 to 446 suggest that (sign 235) or /' (sign M O ? ) is
the basic sign and that infixes change the quantity, but not the quality of the basic sign.
If so, then the basic signs are Iogographs and are various quantities of /' .
This basic sign receives double cage markhgs one example.
sign represents, its' variants teii us something very important about Indus inscriptions.
Sign 414 is wider than it is high and, thnefore. uses a lot of horizontal space on the
inscribed surface (Figure 3.4). Because seals are din size, the horizontal space used
by 414 creates a shortage of space when long inscriptions are to be carved. Therefore,
orientation of Indus signs may not be meaningfûi in the graphemic sense.
The Range of Aiiographic Variation In Sign 4
Sign 4 has sixteen varieties, the most varieties of any Indus sign. As can be seen
fiom the sign graphs above, most variations are in the design and placement of the arm
and leg elements, and the objects suspended from the carrying pole. This is typical of
the aliographic variation found in Indus inscriptions, and these sorts of variations are not
Iinked to specific context. For example. the foilowing inscriptions demonstrate different
varieties of sign 4 in similar contexts:
Set 1
M-783
M-20 1
M-664
Set 2
M- 1O82
M-834
M-43
Set 3
M- 160
M-969
H-416
56
Compound signs
Compound signs consist of two or more simple or complex signs combined in
various ways (Figure 3.1) to form a single sign cluster. Several questions need to be
answered in the analysis of these signs. Fint, what are the component signs? The
answer to this question is not always simple, and has traditionaily k e n expressed using
a formula. For example: sign 275 f consists of signs 282 . The question of
Should this formula read 275 = 282 + 262, or 275 =
reading order is stU not nsolved.
262 + 282 ? This question is further complicated by signs such as 328
therefore: u y 8. (iead h m right to lefi). This sign requence does not oceur
independently, making verification of the reading order impossible. The same is m e for
sign 285
l i t
UIIMld. ' P i 3 sign may be a cornbination of signs 282 + 348 0+ 4 14
This sequence is the likely reading order given the bown distributions of signs
282 and 414.
Some compound signs (25 X"' for example) present less of a problem because
they are arranged in a way that requires they be read in order (nght to left). In cases
where the reading order can be worked out, the chances of working out readings for
these signs is improved.
The question remaîns: why compound signs? Several answers are possible.
First, inscriid artifacts are typically small in size and compounding might nsult fiom
the need to conserve space, as rotation does. If this is the source of compounding, then
b
of associating signs clusters. Some cornpound signs (436 @')may be conflations of a
noun and an adjective. These relationships are not necessarily rnutually exclusive and
several factors may infiuence the construction of compound signs.
The main purpose of the sign list (Table 3.2) is to bring together for aoalysis aU
contexts of allographic variants of a single sign. The extraction and analysis of parailel
inscriptions from the locations given in the s i s list aifows the details of sign behavior to
Other relationships cm also be seen. For example, the sign cluster DEoccw
.X
only on tablets at Harappa The onl time these signs occur as a set is at Mohenjo-daro
contexts. The lack of syntagmatic relationships between graphically similar signs, and
the existence of parallel inscriptions with distinct preferences for specific sets of sign
variations, ailows the identification of graphemes which might have been grouped as a
presence or absence of intemal hatching seems to be the important factor, while the
specinc f o m of hatching has no noticeable effect on sign behavior.
59
Conclusion
The discussion given in this chapter has defined forrnal criteria for differentiating
ailographic fiom graphemic variations found in hdus signs. The examples given in this
chapter were selected to demonstrate how these criteria might be applied to sets of
possible aliographs. The examples have s h o w that minor graphic variations of signs
may mark different graphemes. Figure 3.6 lists 24 examples where previous sign lists
have not adequately defined demonstrable graphemes. These previous attempts lacked a
comprehensive corpus, and a detailed sign list.
The decipherment of the Indus inscriptions is not an epigraphic problem alone.
It needs to encompass the archaeological, linguistic, geographic, and epigraphic
dimensions of the inscriptions. The Indus civilization has often been described as
hornogenous. As Chapten 1,2 and 3 of this thesis have shown, this is not the case upon
detailed exarninations of the data. There are some characteristics which seem
widespread and uniform in very pneral tems. The use of seals and tablets, site
settlement patterns, and a relatively uniform system of weights and measures are among
the more standard attributes of Indus Culture. Regardless of similarities, there is no
evidence that the entire Indus region was unified politically or theologically. As Chapter
4 shows there are some interesting differences in the stmcnin of the inscriptions which
Toial Hnrappa
Toîsl h t h a l Toial Number If % t * r
of Occurrences 7
Total Olher 3 1
Total Sign
Vmricrler Sign Frcqucncy
a
802 Dy Siic
Toirl
62 7
"
Toial
21
var.
I
Toral
14 2
Total
Il 8 1
,
020
Toral Total var.
4 4 2
[y t,i Lm?
----- - -
Toial var.
3 1
Total
2 I
Total
2 2
03'
Toial
2
4 var.
2
Total
I I
Total
I
"
Toial
I
var,
1
044
Toiel
I
var.
I
g z 1- -
a Cœ Ir- 1s s- Ir-
'FI 3 -
-IJ ==lx-
-- II- =-A
Total
3 I I
4 var.
1
O
'
Total
I
A 1
262
Toial
116
y var,
2
Total Total
48 3
Total
2
267
" Y I
dn var, 269
Tolal
3
y var,
1
14 56 1
275
Toial
3 I
Toirl
I I
Toial y Var. Total
1 I
1 y y
-
277
Toid v~r. Total 279
Total va. Toid va,. Total Toril
I I I I I I I I I 1 1O 6
285
Total Total Var.
2 I 1 W I
----.
3W
Total Y , Tolal Total
27 7 6 I 5
438
Total var. 439
Toial var. Total
2 I 2 I I
Total
3
Total
2 vr Total
2 Mcv;" Total
1 I
Total
I
var*
I
Total
1
467
Total
I
pg Var.
I
Total
99 1
"' &
Total ;v 474
Total
I
& var.
I
Total
27
48'
Toial var.
I I
484
Total
I
var,
1
Total
4 2
'*'
Total
1
var.
I
Toir l
I
494
Toial
27
$ var.
6
504
498
Total var. Toral Total Total Var.
17 I 14 3 I I I I
505
Total
I
var.
I
Total
5 mm vu.
I
Var.
I
5 14
Toi81
I
Var.
I
Total
l
Var.
I
516
Toial
I
Vu.
I
l7
Taial
18
8] var,
I
523
Toial var. 527
Total var,
I I 1 1
Toiol
2
Vw
I
528
Total
20
Q var,
I
529
Toial
16
Q var.
I
531
Total
4
1" var.
I
537
Toial
I
var.
1
Toial
I
Q var.
I
539
Toiol
I
Q var.
I
@b 4i Jf i& aA aB + aS: g@. :&i g@u
1tw ion
VVWUVUEUVYM
tji r:i 1 t:t ts r:b 1) ::a t3 ta tuc
uuOP$YTYYYS
III ai:
Figure 3.1 vpology of Indus Signs.
Unmarked
443
u R m
Kvrio
Unique
1
I
Sign 8 Sign 11
With A r m s
n=29
I
Criterion 2
L
I
I Criterion 3 I
I No Ligatures
I
L
Sign 11 occurs twice
Sign 8 is found on several
Other than M-1405 arcifact types h m Mohenjo-dan,. on Type G tabIets
al1 examples of Sign 5 hHmppa (excIudingH-27) fiom Mohenjwhm.
are h m Type A se& a11 examples are fiom At Harappa it occurs
Type B seals. on 1 Type A seal, and
8 Type G tablets
Figure 3.3 Variations in Signs 5 . 8 and 11
N-4
F-1
None
None
L-7
None
1-4.5
G-Il ro 15
Nonc
UNJ
t b c J-7
J-9
14
Chapter 4
Analyzing Indus Inscriptions
The purpose of this chapter is to examine in detail the interna1 structures of the
Indus inscriptions. This study is based on four major fields of data: archaeological
reports, epigraphic analysis, linguistic reconstructions, and the corpus of inscriptions.
Reports of excavations of Indus sites consists of efforts by Marshall (1931) and MacKay
(1938) at Mohenjo-daro and Vats (1940) at Harappa These initial excavations of major
Indus sites stiil remain the main sources of archaeological data relating to inscribed
artifacts. Major contributions were later made by Dales (1962, 1965, 1971 1979%Dales
et al 1977, 1986, 1990, and 1991) and Kenoyer (1989, 1991a and 1991b) at both
Mohenjo-daro and Harappa. These reports have made available the first professional
excavations at these sites. The summary of the 19864990 excavations at Harappa reports
the fmt processualist led multi-disciplinary excavation of a major Indus site. Other major
excavations carried out between 1940 and 1990 were either not widely reported,
completely unreported, or iimited in the scope of their reporting. In addition to the
archaeological data, various attempts at decipherment of Indus inscriptions have been
published fiom Waddell in 1925 to Parpola in 1994. These include the work of
Mahadevan (1977 and 1986), Parpola (1970,1975, 1976,1986a and Parpola et al 1969%
1969b, and 1970), and Knorozov (1968; Knorozov et al 1981, and 1984). The third
source of data is Linguistic reconstructions. These corne in the form of a senes of journal
articles (Zide and ZvelebiI, 1976; Zvelebil 1965, 1970, 1972% 1972b, 1977, 1990;
McAlpin, 1974; Andronov, 1970 and 1976), and related publications (McAlpin, 1981;
Burrow and Emenau, 1961 and 1968). The final major source of information is the
inscriptions themselves. The photographs published by Marshall, MacKay, and Vats are
of mixed quaiity and give only the impressions of the seals. Other insaibed artifacts are
occasionally drawn and only photographs of exceptionai artifacts were published. No
79
major improvernent in the quality of data occurred until the publication of the Corpus of
Indus Seds and Inscriptions (Joshi and Parpola, 1987 and Shah and Parpola, 1991).
W of the publications listed above are used to various degrees in the following
pages to argue for my own analysis of the Indus script. In Chapter 1 and 2 the
archaeological data were presented in an attempt to show the necessity of considenng
inscriptions as having a function, in a social economic sense, as weU as bearing some
message in the unknown language of the Indus people. Knowing the use an artifact was
put to gives us a clue to the subject matter of its inscriptions. Readings of inscriptions
should in some way coincide with the expectations raised by their archaeologicai
contexts. In Chapter 3, 1 created a sign list and computer data base which brought
together groups of inscriptions with the same signs. The sign typology and sign list were
used to set up the parameten of the andysis presented in this chapter, and in Chapter 5.
There are still several questions about the Indus script which need to be answered
before the analysis of its formal structure can be undertaken. Fint, in which direction
(left to right; or right to lefi) are the inscriptions intended to be read in? As with most of
the fbndamental questions about the Indus script there is no simple answer. Parpola
(1991: 64-67) points out thah although the normal direction of reading was right to Ieft,
both left to right and boustmphedon are recognizable. Figure 4.1 (a, b, and c) presents
three seal impressions which demonstrate that the Indus script was normaUy read from
right to left. The nIst example (M-735) shows crowding at the left of the inscription
where the carver ran out of space. The second example in Figure 4.1 (MacKay 1938:
Pl.LXXXTV:83) has a large unused space at the left of the second Iine of script. This
space would have k e n lefi over at the end of the inscription in a right to left system. The
third example (MacKay 1938: PlXCVL521) contains a two line inscription, with the first
iine situated to the extreme right of the space available for carvhg. This inscription can
be reconstructed by anaiogy with the top line to the right. For examples of
80
boustrophedon inscriptions see M-66 and M-892. For these reasons Indus inscriptions
wiil nonnally be read right to lefi in the anaiysis pnsented in this p a p a
The question of what wrïting system is used in the Indus inscriptions affects
structural analysis, and this question needs to be resolved before the results of structural
analysis can be interpreted Coe (1992:26) tells us:
Scripts are usually divided into four classes: alphabctic, syllabic, logo-syilabic,
and logographic. There are examples of intermediate systems. The number of signs in a
given script gives a clue to the class of script we are dealing with. For example, known
alphabetic scripts have 20 to 36 signs and syliabic scripts have 40 to 87 signs. More than
100 signs indicates that the script is logographic or logo-syllabic (Coe, 1992:43). The
Indus script has about 600 signs (very close to Sumerian) and is therefore most probably
a logographic or logo-syllabic script.
These general descriptions of ancient scripts do not address issues of the
mechanical details of sign use. Many ancient scripts employ special sets of signs, known
as determinatives, which are not pronounced when inscriptions are read and consequentiy
have no linguistic counterparts. The purpose of these special case logographs is to
classify words in order to avoid confusion over homophones, or to idenw the gender of
objects, or to classify objects. For exampie, in Egyptian hieroglyphic writing: "open",
"hurry", "mistake", "become bald", "light", and the city of Hermopolis are aii written
identicdy except for their final sign (a determinative in each case) which clarifies their
meaning (Zauzich 1992:28-29). Figure 4.2 shows the way in which signs of several
ancient scripts c m be categorized. These scripts, while alI logo-sylIabic, have very
different smtegies for combining elements into word Egyptian hierogiyphic has four
81
distinct types of syllabic signs and relies heavily on the use of logograms and
deterrninatives. Maya hieroglyphic does not have hue detemiinatives, but instead uses
phonetic complements to cl* the meaning of logograms. Maya too has unditeral and
bilited signs, and some CV syilabic signs can occasionally be used for their vowels with
the consonants left unpronounced. Sumenan cuneiform is a logosyllabic script, but
syllabic spelIings represent a relatively smdi part of these inscriptions. Wheo this script
was adopted by the Akkadian scribes syliabic spellings became much more cornmon and
the use of logographs diminished. Ugaritic cuneiform is alphabetic, but derived nom its
Akkadian predecessor.
There is no reason that the Indus script could not be another variation in the
application of the basic elements as aven in Figure 4.2. Which elements are present in
the Indus script, and how they are combineci, cannot be laiown with confidence. We can
be fairly certain, however, that the Indus script will use some or dl of the elements used
in other ancient script. The strategies employed in cornbining these elements Vary, but
the basic inventory seems limited to those defined in Figure 4.2.
What kind of system does the Indus script use? It is probably a logosyllabic script
because there are about 600 Indus signs. Yet the details of its form cannot be known
simply on the basis of this identification. Figure 4.2 compares the structure of four
ancient scripts; al1 are logo-syilabic, but no two have identical structures. There does
seem to be a fixed nurnber of components (phonetic, logographic, and determinative
signs) used in ancient scripts. Taken individuaily, signs c m Vary in fuoction. Signs
sometimes are used as word building signs or phonetic complements, or sometimes as
logograms or determinatives. In undeciphered scripts this can lead to a great deai of
confusion about sign function and meaning. We c m expect that the Indus script WU
have some or all of the sign classes identified in Figure 4.2. We c m also expect that
some signs will have more than one function (syllabic and logographic). In order to
identQ these signs and assess what their fiinctions might be, detailed contextual analysis
82
Separation could not have been earlier than 10,000 BC and more Iikely
wouId be much closer to 6000 or 7000 BC. Givcn the numerous
cognates for animal hwbanàry and the close association of animal
husbandry with agriculture in West Asia, the unificd culture must also
have taken part in the agricultural portion of the rcvolution.
PED and PDr have an inventory of cognates that suggest they were part of a
unifïed culture at a time when these lexical items were coming into use. This sets an
earliest limit on the date of separation at 6 0 to 7000 BC. The latest limit is suggested
by the lack of a s h e d PED and PD word for writing. The words for writing in PED
(*ta1' to push in') and in PDr (various words meaning 'to paint' or 'to draw')3 suggest that
the split between these languages was pre-literate. McAlpin ( 1981:134) places the split
no later than 3000 BC . Further, McAlpin (198 k134)suggests that the fifth millennium
BC is the most likely period of separation, noting that this fits well with the
archaeological data but gives no deuils.
McAlpin's (198 1) argument is detailed and points to a very ancient and definable
relationship between Dravidian and Elamite that is not demonstrable for any other
laaguage family in the List of proposed Indus langages.
Indo-European has quite a Merent relationship to Dravidian having several clear
cases of borrowing. The most convincing argument against Indo-European as being the
laquage of the Indus people is made by Parpola (1994: 167). Parpola presents the work?
of Tikkanen which examines the presence of the "type A retroflex system" (ParpoIa 1994:
167). Tikkanen maps the presence of the fmt peson plural pronoun (inclusive and
îhere is one word in Tamil w u 'to write, painl draw; become indenad by pressuret thar suggests this
question is not hlly resolved
This work is not pubüshed outside of Parpola, so we are essentially dealing with Parpoia's summary of
Tikkanen-
85
exclusive)5 which is a Dravidian sub-stratum present in the Indo-European languages of
the Indian sub-continent. Parpola (1994: 167) describes the relevance of Tikkanen's work
to the identification of the Indus language a s follows:
Tikkanen's anaIysis suggests that Dravidian had once been spoken also
in dl those parts of the type A retroflex system area which are Indo-
Aryan-speaking now. This distribution rnakes Dravidian the most
likely language to have been spoken by the Harappans
Figure 4.3 combines the boundaries of Dravidian language groups and the type A
retroflex system. It is clear from this map that the Indus vailey fails within the area
defmed by Tikkanen. Parpola's conclusion that aU Indo-European speakers within the
type A retroflex system area were infiuenced by Dravidian speakers in a way that implies
population infusion is reasonable, but not conclusive. Much of the argument for
Dravidian as the Indus language is based on this type of suggestive but inconclusive
evidence. Yet no other proposed language within this limited geographic area shows
borrowings of the type and with the same time depth as those borrowed into Indo-
European from Dravidian. W e this does not demonstrate that Dravidian is the
language of the Indus inscriptions, it does demonstrate that Indo-European came to the
Indian sub-continent at a time when Dravidian was in widespread use. Given the time
depth of the Indus civilization, and the relationship of Dravidian to both proto-Elamo-
Dravidian and Indo-European, there is linle doubt that the Indus people were Dravidian
speakers. The exact form of Dravidian spoken by the Indus People is not known nor is it
reconstructable fÎom the inscriptions as they stand. The best starhg place for matching
the structures evident in the inscriptions to a language is proto-Dravidian (PDr),and that
assumption is the basis of Chapter 5.
r
we=*you and and we='all of us'
86
Knorozov (1968) was the b t researcher to use computer technology in an
attempt to define the intemal stnicnires of the Indus inscriptions. His work defmes the
linguistic structures and patterns of sign replacement, using a computer program and,
"About 350 signs [that] compose the Roto-Indian alphabet" (Knorozov: 1968: 13). As
Chapter 3 demonstnttes, this is too few signs to account for ail the significant sign
variations recognizable in Table 3.2. Regardless of the shortcomings of his sign list,
fiorozov's analysis manages to define an impressive set of strucnual elements that were
previously unknown - blocks in his terminology. Knorozov (1968: 10) offers the
following interpretation of his results:
Theoretical considerations led to a supposition that according to the
technique used the blocks will represent not only word forms plus the
corresponding word- and form- building affixes, but those plus
prepositions, particles, conjunctions etc. and even unchangeable
attributs (if they are present) devoid of their morphological formats.
Such attributes are practicaily indistinguishable fiom subordinate
elements in unknown texts.
In short he was able to define syntactic elements through the analysis of sign
order. He further breaks blocks into constant, variable, and semi-variable signs. He gives
the following description of these elements (Knorozov 1968: 1 1):
Constant signs are retained in al1 cases when a givcn block occurs, they
render root morphemes in al1 probability. Thrce groups are
distinguished among variable signs. The fmt group is composed of
properly variable signs (the most flexible). They occur at the end of the
blocks, i.e. in the position u s d l y oceupied by fonn-building suff'i'ies
and othcr morphological fermants. Other variable signs should
properiy be called serni-variable, because they occupy an intermediate
place between the variable signs proper and the constants (they are
closer to variable signs). The scmi-variables signs are subdivided into
two marktdly differcnt groups. Ttie semi-variables of the first p u p
always precede the variable signs proper (if these are present), and
imrnediately follow the constant signs, i.e. they occupy a position
usualIy taken up by derivative suffixes. The semi-variables making up
the second group aIways precedt the constant signs; togethcr they
compose a combination which remains stable within one
microparadigm-
See Figure 4.4 for bis examples of several microparadigms. Knorozov goes on to
descrik the place of numerals in his system identifying them both as group II semi-
variables and as counts of various uniu of meanue. To this point Knorozov's method is
87
robust. He reads the texts as proto-Dravidian or Proto-Indian in his terminology. His
The object identifed as H 305 has no relation to the H-305in the photogaphîc coipus.
88
readings, Knorozov's structural analysis offen a very good starting point for M e r
attempts at structural analysis.
In 1970 Mahadevan (1970, 1977, and 1986) began his examination of the Indus
script. He c o ~ e c t e dhis structural analysis to Dravidian morphology in 1970, and in
1977 Mahadevan published The Indus Script: Tms,Concordnnce and Tables. This was
the first attempt at a computerized sign list and font, and brought together for the first
tirne large nurnbus of replacement sets. As demonstrated in Chapter 3, Mahadevan's sign
list (417 signs) was too generalwd to allow detailed structural analysis. Yet Mahadevan
was able to demonstrate both large scale structures in the script and a link to patterns in
Dravidian morphology.
The next attempt at formal structurd analysis was carried out from 1970 through
1982 by Koskenniemi and Parpola (1979, 1980, and 1983), Koskenniemi (198 l), and
Koskenniemi, Parpola and Parpola (1970, 1973). Their publication of Corpus of Texts in
the Indrrs Script in 1979 mimicked Mahadevan's corpus in many respects. First, they
used a computer based font for their sign list (401 signs). Second, they provided a
concordance of related inscriptions using their sign font. There were some notable
differences in their sign list compared to Mahadevan's. Koskenniemi et al discuss these
differences, arguing that several signs be collapsed because of their occurrences in near
Y v
identical contexts. Notable among these sign were the graphs 1 and 1 . The
collocations of these signs is discussed in Chapter 3 where it is shown that there are
si,pificant differences In their collocations, when al1 contexts are considered.
Convenely, they argue for the separations of several sets of allogniphs which Mahadevan
had grouped under one These minor differences aside, both Mahadeva. and
Koskenniemi et al fouad large scaie structures which they related to Dravidian
morphology. Both combined all inscriptions, regardless of artifact type, into one
analytical set, and both used very generalized sign lists and sign replacement fonts. The
The rows and columns of the grid correlate the syntagmatic and
paradigrnatic relationships of single signs and sign sequences. The
word syntagrn(a) mcans 'a string of elements fonning a syntactic unit' ...
A Paradigmatic relationship is the 'the relationship of substitution
between one Iinguistic unit and other comparable units at a particular
place in a structure'.... in other words, the investigation of the
syntagmatic refationships determined fiom these grids leads to the
estabIishmcnt of particular 'functional' of 'grammatical' units, whereas
the investigation of paradigmatic relationships leads to the
establishment of the range and structure of these particular 'functional'
or 'grammatical' units.
Using paraiiel inscriptions Parpola defines three dots in his grid (1 through m),
which he believes hold distinct elements of syntax (Figure 4.4). His slots are not the
equivalent of the fields in Figure 4.6, although there is some agreement in the placement
of the boundary between Fields 1and II. Using these slots as the basic uni& he compares
them to the few very long Indus inscriptions noting that these long inscriptions are
comprised of strings of sign sequences found in shorter inscriptions. These long
inscriptions consist of t h e to five dots which Parpola groups into cycles. Parpola
( 1994:90) believes that "a major syntactic boundary is likely to occur between the
cycles". In brkf, the short inscriptions are single syntactic units, while long inscrîptions
are combinations of several of these units.
90
Parpola goes on to relate his slots and cycles to words and sentences, using the
Dravidian (S)OV syntactic order. He works out a paradigm of inflectional suffixes
(including case endings) (Parpola 1994:94),8 and gives the Soviet (Knorozov 1968%1)
'micro-paradigm of variable signs'. In summarizing what is known about the Indus
inscriptions Parpola makes the assumption that most sed inscriptions are noun phrases.
His analysis of what is possible is based on this assumption. For example, he States "a
good candidate for a verb would be something found at the end of longer text that does
not occur in shorter ones" (Parpola 1994:96). As shown in Chapter 2, seais were meant
to be combined to form long inscriptions, but sometimes only one seal impression is
needed. The short inscriptions may thenfore contain any of several elements of syntax,
not just nouns, and long inscriptions contaîn ail necessary elements of syntax.
Inflectional paradigms based on the flawed assumption that these are only nouns are
bound to have serious functional problems.
Fairservis (1992:117) notes bat the shortness of Harappan texu "makes the
identification of syntax difficult if not impossible since sentences per se do not exist in
the texts discovered to date." This seems to be a controlling premise rather than a
satisfactorily based conclusion. FairseMs does not consider any of the very long seal
texts nor the Type N a tags from Lothal, Kalibangan and other sites. These texts clearly
contain enough signs to dIow the analysis of syntagmatic relationships as defhed by
Parpola (1994). Fairservis presents a List of identifications of signs and readings of
inscriptions which reflect his belief that the seals were sewn to clothing as a means of
identifkation. To Fairservis (1992) the Indus inscriptions contain the name, title, status,
lineage, or occupation of the individual within the social and politicai, and economic
secton of the culture. He suggests that "the tablets were used as badges or even maniage
tali." (Faitservis, 19925). Like Knorozov, the readings proposed by Fairservis are often
cryptic or even nonsensicai (Keiley and Wells 1995:22).
8 He abandoned this paradigrn in 1969 because of some inconsistemes but mentions it again as the fkst
resuits of his structurai analysis.
The Structure of the Indus inscriptions
Building on the work of Parpola, Mahadevan, and Knorozov, my goal is to use a
mo-ed form of structurai analysis to examine the details of sign relationships. 1 have
concluded (on the basis of multiple seal impressions on tags and the length of Type G
inscriptions) that Type A/B seals with short inscriptions (5 signs or less) are fragments of
messages that need to be recombined. Longer inscriptions (6 signs or more) are more
likely to contain whoIe inscriptions or at least larger fragments. Figure 4.5 shows both a
long and short Type A seal inscription. The inscription on M-37 consists of three signs
that are repeated on M-650,but with four additional signs (also see Figure 4. Id).
Column Analysis
CoIumn analysis is a method whereby the relative position (lefi or right ) of a sign
(in relation to al1 other signs with which it occurs) is assessed. For example, does sign
414 consistendy locate left or right of sign 189 /! when they occur in the rame
inscriptions? These relationships can be used to arrange signs in columns. Iniîial
attempts to use al1 Indus inscriptions to generate a matrix of a l l Indus signs sorted by
relative positioning within texts resulted in several noticeable groups for which relative
positions could not be resolved. These unresolved relationships pointed to a need for the
selection of a more cohesive group that could be used to establish a base line which the
rest of the inscriptions could be compand to. For this purpose 1 selected the Type A/B
seal inscriptions fkom Mohenjo-daro which contain sign 288 tf. There are 327
inscriptions which match this description. The result of comparing the dative positions
of signs within this limited set of inscriptions was the eIimination of most questionable
relationships (Figure 4.6). The signs in Figure 4.6 are arranged more or less by Set
verticaily through the matnx. Horizontal (relative) locations of signs reflects their
sequence of occurrence within inscriptions containing sign 288 Uas specified above.
Columns are grouped into 3 fields (1-III)and eight sub-fields (A-H). These divisions
reflect the syntagmatic relationship between groups of signs within inscriptions. It is my
contention that the fields mark the subject, object, and verb boundaries within
inscriptions, and that the sub-fields mark the boundaries between sub-elements of syntax.
Colurnn 8 contains signs which cannot be placed in Field II or III, and Iücely contains
signs that belong to both fields. Some signs appear in more than one field and this may
be an indication that either they have multiple values or are syllabic signs. Al1
inscriptions contain one or more fields dependent on the number of signs in the
inscription.
Defining the field and sub-field units for a specific inscription can be
accomplished by Looking for sequential breaks in their column numben. Table 4.1
compares seven inscriptions.
Breaks in the sequence indicate boundaries between syntactic units. In some
cases these breaks are less well defined (M-595) and in these situations parallel
inscriptions must be examined to clariQ the boundaries of syntactic units. In the case of
M-595 it is not clear if sign 78 3should be part of Field II or Field III based on its
column number alone. Through the examination of the 12 other inscriptions from
Mohenjo-duo which contain the 288/1/264
that sign 78
u ?s i s sequencc
3 is in Field II. For M-289sign 78 is in Field III
WC can determine
as part of the sign 288/78
U2 pairing. Here sign 78 replaces for rigns 1,264. It seems likely that sign 78 is a
loaograph while signs IR64 are syuabic signs. This interpretation assumes that sign 288
is the case ending. It is also probably a ryllabic rign because case endings in Roto-
Dravidian tend to be monosyllabic. 1propose that any sign that occurs alone in a field, or
colIocates with siga 288 as the only signs in a field, is a logograph.
93
Many signs occur in more than one field, most often Field I and either Field II or
I
Field m. For example sign 469 b occurs in Field 1 and Field III,
W
but not in Field 11.
f
Other signs are restricted to specific fields. For example, sign 119 occun ody in Field
II.
The process of identifying syntactic units for specific inscription is expedited in
many cases by analyzing the signs in the inscription for their column numbers. Funher,
many uncertain boundaries can be clarified by examining paraiiel inscriptions, which can
provide clearer examples of interna1 divisions. The analysis of inscriptions using these
techniques yields a list of Indus words which can be identified at the Subject-Object-Verb
level.
Before examining the content of the various fields and the arrangement of signs
within fields it is necessary to examine some general charactenstics of the syntax and
morphology of Proto-Dravidian. McAlpin (198155) tells us:
It is diffkult to speak mcaningfully of a syntax which can be labeled
Proto-Dravidian. Most Dravidian languagcs have almost identical
syntactic structures, but the problem is that so do Indo-Aryan Marathi
...
and Singhala In othcr words, there is a syntactic pattern typicai of
Dravidian languages, but it is very difficult to determine how much of
it is inherited and how much is areaily influenced.
Field Shifting
Fish, as represented in the Indus script, consist of 31 separate signs which
combine into 46 different combinations. In its simplest form, sign 131 I\ , it is rare
Including Tienin's type A retrofiex system. the fkt person 'we' Indo-Arp substratum proposed by
Parpola 1994: 167 (Figure 43).
occming only twicc on pots h m Kalibangan. The common variant, sign 112
W
8, and
113 # have wider distribution (Figure 4.7). This basic sign undergoes elaboration
creathg 29 variantdo Variants have different positional charactenstics dependent on the
form elaboration takes. The fmt set of additions to the basic signs (112 and i 13) consist
of both horizontal, vertical, and diagonal infixing of the Bsh's body, resulting in three
separate signs.ll These five basic signs can then receive additional markings. In some
cases this results in a movement Ieft in column positions through the inscription.l2 For
'n 4l
example, if sign 114 has cage markings added to form sign 119 IX8 the column number
changes from 4 to 7. This pattern is repeated for signs 116 'a'
and 122, with column
numbers of 5 and 7 respectively. This is a phenornenon 1 c d field shifting and it is not
limited to fish signs, nor to the cage markings alone. Brackets, enclosures, ovates,
diamonds, and occasionally doubled cages enclose13 other signs changing their position
in both directions in Figure 4.6.
The implication of field shifüng is that these markings and additions to the basic
s i p s change their syntactic function. In the example above the cage rnight change a
syllabic sign to a logograph (stem), analogous to transforming a noun stem to a verb
stem. An example can be seen in Tamil: ka! (vb.) @ED1090) 'to leam, snidy, pnictice
(as arts)', kaIIun (n.) 'unleamed, illiterate person', kalvi (vb.)'studying, leaming'. In this
example the addition of -an and -vi changes both the meanhg and the part of speech or
tense of the verb respectively. In a paraiiel example, the fkshwater shark (Wallago am)
is called valai in Tamil. Volai dso means 'net' whiie valaiyan glosses as 'fisherman'.
Morphological elements in the language are manifested in the script with a series of
markers. The sign typology (Figure 3.1) shows the complexity of the Indus marking
system. The 146 rnarked signs can be divided into 11 types (Figure 3.1). This defines
Tamil and other Dravidian languages. Several suggestions have been made by Knorozov
(1968), Mahadevan (1973:50), and Parpola (1994:96-97) of how various noun cases may
be represented in the inscriptions. In most approaches to the Indus script marked signs
are unexplained and suspected case markers are usudy identified as independent signs.
While the identification of syntactic eiements and in some cases sign classes~4
can be made fkirly securely, the next step, assigning values to signs, is more difficult. In
Chapter 5, readings of some signs are suggested but the analysis of inscriptions is not far
enough advanced to derive readings directly fiom the inscnptions alone. Therefore, the
focus of the following discussion of fields is limited to more general identifications of
functional charactenstics as related to Roto-Dravidian syntax and word construction.
Field Contents
In the following discussion I will examine each of the fields in detail and suggest
some possible identifications of their function in the inscriptions. For the moment we are
still discussing the inscriptions analyzed to create Figure 4.6, although al1 inscriptions
follow similar patterns. 1 assume hereafter that fields 1-IlI approximate the subject,
object, and verb components of Dravidian syntax. Numenc (simple saoke) signs occur in
the inscriptions (Figure 4.6) in various positions in texts, but are most commonly found in
Field II. Numben are amonp the few forms of adjectives that are reconstructable for
PDr. McAlpin (1994:40 cf. Table 1.7) tells us that "An invariant compounding base is
readily reconstructable for one through eight." He gives the following reconstructions:
Compounding Nouns
Base Variants Human Neuter
or oni, or omvanre(m)
0ItItti (f)
on onre
okka okkanre
ol
ir iru, ir iruvar irante
mu(N) mu, muC muvar munre
nal nan naike
cayN CaYn caynte
care care Cam
ez ezu, ezu eze
en entte
col ton
Field 1
8 II
Field 1is defined as the sign clusten which end with the markea , ,and As
with much of the morphologicai patternhg in the inscriptions, hard and fast d e s of
I
usage are scarce. Sign 196 ,seems to function both as a Field 1 terminator and as a more
pneral word separator. There is no clear context where sign 196 is being used as a
II
numeral. Sign 193 is found used as a field 1 terminator and a numeral. Evidence for
the numeric context cornes from the comparative structural analysis of the following
inscriptions given in Table 4.2
These distributions prompted Parpola (1979: 19-20) to create separate dographic
entries in his sign List for sign 199 in the top, middle, and bottom registen. In the case of
sign 199 context is more important than vertical placement. The third common Field 1
temiinator is sign 231 and Y which occur 120 times in the inscriptions. Aiiographic
variations are great, but structural analysis shows a i l variants in identical contexts. The
foilowing signs ofien collocate with Markets (signs 199,196, and 23 1): 341 8
342,@,
543 x, and 571 P. Some enamples have longer sign sequences but ofien terminate with
one of these four signs and any of the three Markers. Roblematic contexts aside, markers
are fiequently the most obvious feature in an inscription. There are large numben of
Mahadevan (197350)pointed out that 7546 of Old Tamil male names end with
-an . Most of the rest of these male names end with -i or ai.16 Female names commonly
end with - a l . These endings may be the equivalents of any of 290-292or the marken.
The structure of Field 1 is suggestive of the same mechanisms of construction as
observable in Field II.
Locatives are more cornplex. Location nouns are inherent locatives. While there
are no locative cases in PDr there are specific agglutinative elements which mark
locatives. McAlpin (1984:37) gives cornmon Roto-Dravidian locative postpositions as
follows:
*-in Comparative
*+)tu fixed in space
*-(k)ku location ih tirne, or motion towards
*-ul inside, within
*-il place, house
Another possible component of Field 1 is penonal pronouns. McAlpin (1984:37)
gives the folIowing list of personal pronouns for PDr
van17 1s T
%i 2s 'you'
%im 2p 'you'
*tan 3dp 'sewselves'
I I I I l or I [ ' ',la and he reads seal H-9 as scven star and equates 'seven stars' with
the Pleiades. 1 agree that sign 112 is most WceIy a representation of a fish. Given the
archaeological context (economic) of seab and tags it seems more likely to me that in this
case a fish is just a fish. Supporthg evidence cornes nom the detailed analysis of fish
remains fkom Harappa (Belcher 1991). Belcher (1991:1 18) reports: "Fish appear to have
been an important protein source for some populations of the Harappan civilization."
Most fish remallis were hshwater catfish and carp. However, there were rernains of
Etyrnological Dicîionary (Burrow and Emeneau 1960). Whiie min (DED 4060, Ta.)
means 'fish' in a general sense, eri-valai (DED 764, Ta.) refen to the Wallcrgo attu
specificdly.2o Kayal and cel (DED 1050, Ta.) are specific terms for 'carp' in am il.*^
Any of these are just as valid a reading for sign 112 as min. Additionaliy, any reading of
sign 112 must also account for the other fish signs and the effects of markings (field
shifting) and elaboration.
Fied Iïï
This field shouid contain verbs and verb clauses according to the mode1 of PDr
syntax aven previously in this chapter. What smicniral analysis gives us is a series of
sign clusters which can be identified as either verbs or nouns. Methods of constructing
verbs are agglutinative and typically verbs consist of root+augment+case+(pplcl). This
19 Sign I 12 occurs in conru*i where it is the oniy sign (K-9) othn than moke signs in sn inscription.
20 Watlago rrrm accouna for 42% of the e h remaino found in the 1986.1990 excavations of Harappa
(Belchet 1991:I 13)-
* Carp represent 255% of Belcher's (199 1) sample.
101
mode1 fits well with the resuits of structural aoaiysis. For example, the Field DI sign
cluster U cm be interpreted as a stem (root+augment) X"I" plus the case marker u.
Conclusions
Given the anaiysis presented in Chapters 3 and 4 the inscription M-221can now
be analyzed as follows:
Field III Field II Field 1
case+s tem(root+augment)
While many uncextainties remain regarding readings of specific signs, the ability
to identiw the part of speech of a given sign cluster limits the phonemic values it cm
have. For example, sign 288 u has been identified as a verbal case marker and therefore
1s *-en
IP *-em
2s *-&+Jy
2~ *-ir
3s (masc) *-anre
3s (non-masc) *-(V)te
These sorts of iimits can act as critena for the verifkation of future readings. Sign
288 Ucan be expected to have one of the values listed above, and readings which do not
match this List can be discounted?
Column andysis cm be extended to other types of inscriptions. The tablet M-495
is inscriid on three sides. The foUowing sign sequence is found on side B:
22 ,
Chapter 5 for sign 288 as *-q.
IO2
Field Iïï Field II Field 1?
case + verb stem noun ~hrase
- .
Tags c m also be analyzed in this way. The tag M-425 has three seal impressions
sign # 282
col.# 14
sign #
COI.#
U
sign # 288
col-# 13
Breaks in column sequences mark the boundary between parts of speech, but the
application of these methodological techniques is not advanced enough for this to be a
simple mechanical process. The extension of column analysis beyond its original
boundaries requires some detailed knowledge of the inscriptions.
As Figure 4.8 demonstrates, the utilization of the methodology descnbed in this
chapter allows the definition of sign function on several levels. At the grossest level
inscriptions can be broken fields (1-m). These elements c m be identified as subjects,
objects, or verbs because of their order in the inscriptions. The definition of these
elements is a necessary first step in the anaïysis of an inscription. Verifkation that these
elements are hctional uni& used in Indus writing can be seen in Figure 4.8% where sign
103
clusters can be found as interchangeable sets which maintain their field positions in ail
inscriptions in which they are found.
Column analysis M e r differentiates sub-structures within fields. These sub-
structures can be identified in some cases as logographic or syllabic in function. The
same eIements have morphological functions as well (Root (stem+medial), case marker.
postposition) and these can be ascertained through the positional analysis of sign
sequences (Figure 4.8b).
The continuous interaction of different lines of analysis is necessary during the
process of deciphennent. It requins the researcher to shift among data sets while
allowing the results of ongoing analysis to interact. Epigraphic analysis is therefore a
synergetic and dialectic process in which multiple Iines of evidence are used to
undentand the meaning of inscriptions. The process of decipherment is complicated by
Our impoverished data and the lack of certainty with respect to key issues. When ancient
scripts are deciphered, major advances are usudy the result of the discovery of bihgual
texu or very long inscriptions- neither of these are available for the Indus script. The
decipherment of Maya hieroglyphic texts was stalled for years because of
misunderstanding of the subject matter and mechanics of the script.
Table 4.1 Column Numbers and Sign Frequencies Compared
M-900
sign #
column #
frequencu
M-1353
sign #
column #
frequency
M-671
sign #
colwnn #
fiewency
M-28 1
sign #
column #
frewency
M-289
sign #
colwnn #
fiequencv
M-221
sign #
colwnn #
Fequency
M-803
sign #
M-658
sign #
H-141
sign #
H-472
sign #
Figure 4.1 Space Utilization and Direction of Reading.
Crowding
\n
Runs out of space before
reaching the end of the seai
and shifts orientation and
alignment of last sign.
d.
Figure 4.2The Function of Signs in Four Ancient Scripts
Egyptian
Hieroglyphic
Phonograms Ideograms
II
Word Building phonetic logo gram^ detexminatives
Signs complements
Maya
Hieroglyphic
I
Phonograms
Ideopms
Word
m
Building phonetic
Signs compiements
uniliterais
(vowels)
-
I
biliterals -1 Sumerian
Hieroglyphic
Phonograms
I
Ideograms
Word Building
mphonetic
1
deteminatives
~ o P o ~ F ~ S
Signs complements
uniliterals
(vowels)
-
biliterals J
Phonograms
tIdeograms
I
Word Building I
Signs loIPw=
biliterals
triliterals
Figure 4.3 Map of Dravidian laquage boundaries and the extent of the
type A retroflex system (after Parpola 1994: Figures 8.8 and 9.3)
Figure 4.4 Previous Methods of Structural Analysis.
K-$9 X 14
* pp/d is an abreviatioa which stands for postpositiodclitic. Clitics are gramatical morphemes which
follow words or phrases and descn'be qualities of inclusion, exclusion. uniquenm etc (McAIpin. 1981q3Z)).
Postpositions in this paper are denned as morphmes which foIlow the mt,bat which are neither a case
nor a clitic.
These few general d e s are testable to sorne degree by comparison with positional
patterns of signs within the Indus inscriptions.
This analysis is complicated by the nature of the artifacts. The vast majority of
Indus inscriptions are found on intaglio seals. As argued elsewhere (Kelley and Weiis,
1995) the multiple seal impressions on load tags (Joshi and Parpola: 1987, pp. 316-318)
demonstrate that some individual seals contain only part of the message and m u t be
combined to fonn a complete message. Conversely, artifacts such as sealings (tablets) and
copper tablets contain whole messages, albeit short ones. Seals cm be expected to contain
information relating to the msportation and identification of shipments of trade items.
Tablets give us some idea of the structure of whole messages, and they can be used for
comparison to seals to help in their recombination.
Traditional methods of analyzing the Indus inscriptions have combined al1
inscriptions into a single set. This has Iead to some irreguiarities in results. There is a
snong possibility that different classes of artifacts address different topics a n d therefore,
use different signs and sign combinations. Further, there are noticeable regional
differences in the constniction of inscriptions between those excavated fiom Mohenjo-daro
and those excavated h m Harappa. These minor differences may be related to variations in
Indus didects, or to regional differences in the methods of using seals, or CO regional
differences in w n ~ practices.
g These possibiüties recommend a comparative rather than
comprehensive approach to the inscriptions. My own research (column analysis) indicates
that, among intaglio seai inscriptions, there is a good deal of regularity in relative
containing sign replacement sets. The h t approach is difficuit with Indus inscriptions as
depictions are rare. The second approach requires a knowledge of the value of the sign
k i n g compared. Analysis is also complicated by the possibility that replacement sets have
different values, for example: '2 days' as compared to '7 days'. If the signs for two and
seven are not recognized as numerical, then they may be wrongly identified as allographic
variations of a single sign. This already m ~ ~ lsituation
t is M e r complicated by the fact
that signs can be logographs, syllables, detexminatives, or have some other as yet
uadefined function. Some signs may have multiple functions depending on their context.
In sorne cases, for example the Maya tw, sign, signs in a script can have several different
values. In these cases meaning often depends on context. Not a l l scripts use all these
components, nor is every component used in the same way fiom script to script. The lack
of a widely accepted Est of Indus signs prevents the complete differentiation between
allographic and graphemic variations of signs, and issues of sign functions have
traditionally been ignored. Consequently, decipherments of Indus inscriptions have
presented readings which are untestable.
specific Cases
There is one set of Indus inscriptions w k e signs and images replace each other -
namely the copper tablets fiom Mohenjo-dm. These copper wafers are engraved on both
sides. The reverse bean an inscription, the obverse bean either an image or a sign set
(usuaUy several conflated signs).
This raises the possibility that this sign has the meaning of 'haret. in Tamil the= are t ~ o
words for hare. Of these possibiiities ceviyan (DED16493 is one and muyai (DED4071)
is the other.
Most researchers, for reasons obvious fkom Table 4.1. reconsrnct *muyal as the
Proro-Dravidian word for %are'. Yet the strong association of cevi to words relating to
hearing and ear. makes ceviyan (PDr *km-) worth a closer look. If it is assumed for
H-47& H-58),but always with the pair UV. That ir to say that
marker.
1am convinced by the summary of arguments presented by ParpoIa (1994: pp. 64 -
67) chat the Indus script reads nght to le& and follow that practice in idenufying the fields
of the inscriptions that follow. The following five inscriptions are very similar in terms of
their content :
Verb Numeric I Obiect/Subiect Fields l
Values I I
1111
111
I
1111
\\\\
lifr
lllltl
The two examples fiom Harappa occur in siightly different conrext than those from
Mohenjo-daro. These variations in usage may point to differences in dialect or regional
differences in sed usage.
119
pp/d? Verb Nc- I Obiect/Subject Field I
Vdues I 1
H-47
H-58
Given the suggested * k W a n reading above, we cm expect that the verb in these
seven inscriptions wiU begin with kevi and end with a comrnon verbal case marker. The
only match for these expectations fiom the DED is #2265: cewai = 'correctness, fitness,
accuracy, straightness, evenness, smoothness, sound condition as of rnind, body'. The
inscriptions above folIow the pattern predicted by Dravidian syntax in that the obëct/subject
7 Fê d
is separated from the verb by a numeral. In inscriptions ending with , sign 288
terminates the inscriptions, and verSies that sign 306 can be foiiowed by a case rnarker.
V (306)and @ (342) rignr ocçur inciependcntly and may have the foiiowing
values: = *kev(i)(DED #1645). 'ear' (Ta & Ma); = yan (DED #4234), fmt peson
oblique ending, and personal pronoun 'r.
7F
The *kewai reading would give V a value of *-ay [or *-ai] (PDr) accusative case
(McAlpin, 1981:122; DED(s) #2830), fmt person singular appellative, a word for 'cow'
(DED(s) 283) , and R.D 'mother' (McAlpin, 1981:122). Sign 288 usuaIly locates
terminaily or semi-tefminaliy, however, it does locate mediaiîy in five inscriptions:
M-495A
With the exception of L-122 these inscriptions occur on bas-relief tablets. In these
cases sign 288 lJ may be functioning as a syllabic sign [ay or aiJor as the basic Roto-
II
sign 195 .). The Type G (tablet ) inscription found on H-182 (shown here) combines
both these elements, perhaps reading *-iraymeaning 'great peaon'. The value of sign 345
@ is unknown, but both the accompanying pictvre of the h e e r and graphic design
of sign 345 suggest that a reading meaning drum or dnunmer is possible. The only word
for dnun in the Dravidian Etyrnological Dictionary which meets the expectations outhed
above is DED 3319parai (Ta) ' d m $ a measure of capacities'; para (Ma) 'dnim, a rice
measure, disk, circle'; pare (Ka. and Kod) 'dnun(a large double headed d a m beaten by
Mede)'; par (Ko. and To.) ' d m . The root par 'dnim'and the affixes *-iray may read
pariray 'great dnunmef.5
ConcIusion
In the above discussion I have appiied some simple epigraphic techniques to
DNIIL in this case is a musical instrument as di of the 22 iisthgs in the DED as various specific words
for dnrm and drummhg.
Pmoi @ED 33 18) is a relatcd word meaning 'to speak, say'.
121
reading Indus signs. These techniques have been used with good results by Mayanists in
the decipherment of ancient Maya inscriptions. The readings resulting from the application
of these techniques to Indus inscriptions support the hypothesis that the Indus people spoke
and wrote a language related to Roto-Dravidian (and Roto-Elamo-Dravidian). The resuits
of this study also confirm that the Indus script is logographic and syllabic, and that signs
may have slightly different values dependent on context (i.e. m= y= *kev4; *k&-
yan). Further, the general niles of syntax and word construction outlined in Chapter 4 for
Dravidian laquages fit with The *kevWan reading agrees with the image
a Signs
a b c d
Bibliography
Adams, E.W.
1988 Archaeological Classifcation: theory vs. practice. Antiqtciiy, 63:40-59
Adams, W.Y.and E-W. Adams
1991 Archaeological Typology a d Pructicul Real* :a didectic approach to
artifact classincation and sorting. Cambridge University Press,
Cambridge.
Agrawal, D.P.
1964 Harappa Culture: New Evidence for a Shorter Chronology. Science. 143:
950-2.
1969 Dilmun and the Gulf of Cambay. Antiquiiy. 43:3 15-7.
Ailchin, B. and Ailchin, R.
1968 The Birth of Indian Civilization. Hamiondsworth.
Andronov, M.S
1970 Dravidian Languuges. Moscow.
1976 Case Suffixes in Dravidian: A Comparative Study. Anthropos. Vol.
71~716-37.
Belcher, W.R.
1991 Fish resourtes in an Early Urban Context at Harappa. In R.H.Meadow
(ed.) Monographs in World Archaeology, No.3: 107-20.
Bibby, T.G.
1958 The "Ancient Indian Style" seals from Bahrain. Antiquity. 32: 243-46.
With comments by Gordon and Wheeler.
1972 Looking for Dilmwt .Harmondsworth. London.
Brunswig, RJI.
1975. Radiocarbon Dating and the Indus Civilization: Calibration and
Chronology. East and West. 25: 11 1-45. Rome.
Bninswig, RH., A. Parpola and D.Pot&
1983 New Indus and related seals fiom the Near East. In D.T. Potts
(ed.)Dilmun :101-115.
Burrow, T.and M.B. Emeneau
1961 A Dravidian Erymological Dictiunury. (2nd edition) Oxford.
1968 A Dravidian Etymoiogical Dicno~ry.: Supplement. Oxford.
Casal, J .MI
1966 Nindowari: a Chalcoiithic Site in South Baluchistan. Pakistan
Archaeology, Vol. 3: 10-21.
126
Chakrabarti, D.K
1977 India and West Asia: An alternative approach. Man and Environment 1:
25-38
Christenson, A.L. and D.W. Read
1977 Numericai Taxonomy, R-Mode Factor Analysis, and Archaeological
Classification. Arnen'can Antiquity, VOL42, No. 2: 163-79.
Coe, M. D.
1992 Breuking the Maya Code. Thames and Hudson. New York, N.Y.
Costantini, L.
1984 The beginning of agriculture in the Kachi Plain: The eveidence from
Mehrganh. In Allchin, B. (ed.), South Asim Archneology, 1981.
Cambridge University Press, Cambridge, 29-33.
Crawford, V E.
Swnerian Economic Texsfromthe First DyMsty of lsin. New Haven.
m.
A Search for Ancient Seaports. Erpedition 4% 10; 44.
Lambrick, H.T.
1967 The Indus Flood-plain and the 'Indus' Civilization. Geographical Journul.
133:483-95.
130
Lambrick, H.T.
1970 Stratigraphy at Mohenjo-daro. Journal of the Oriental Inninte. Baroda.
20:363-9.
MacKay, D.
1945 Ancient River Beds and Dead Cities. Antiquity. 19: 135-44.
MacKay, E.J.H.
1925 Report on the Excavation of the "A" Cemetery at Kish, Mesopotamia
Part 1. Chicago.
1931 Further Links between Ancient Sind, Sumer and Elsewhere. Antiquity.
5459-73.
1932 An Important Link between Ancient India and Elam. Antiquity. 6:356-7.
Macffiy E. 1938. Further Excavations at Mohenjo Daro. New Delhi. (2
vols.).
McAlpin, D.W.
1981 Proto-Elamo-Dravidiaa: The Evidence and its Implications. Transactions
of the American Philosophical Society. Vol. 71 Part 3. Philadelphia.
Mahadevan, 1.
1970 Dravidian Parallels in Proto-Indian Script. Journal of Tamil Studies: Vol.,
Ii, No. 1.
1973 Method of paralielisms in the interpretation of the Proto-Indian script.
International Association of Tamil Research, Proceedings of the Third
International Conference Seminar, Paris 1970. Pondicherry Institute of
Linguistics and Culture,Publication No. 50:44-55. Pondicheny.
1977 Ihe Indus Script: T e s , Concordance and Tables. New Delhi.
Memoirs of the Archaeological S w e y of India No. 77.
1986 Dravidian Models of Deciphennent of the Indus Script: a case study.
T m i l CM'Iizations4 (3-4): 133-43
Marshall, J.
1937/1973 Mohenjo-daro and the Indus Civilization. Reprinted by Indological Book
House. Delhi
Meadow, R.H.
1973 A Chronology for the Indo-Iranian BorderIands and Southem Baluchistan:
4000-2000 B.C.in D.P. Agrawal and A. Ghosh (eds.), Radiocarbon mrd
Indian Archaeology. Bombay. 190-204.
Meadow, RH. (ed.)
1991 Harappa Excavations 1986-1990: A Multidisciplinary Approach to Third
Mîllennium Urbanism. Monograph in Wotld Archaeology 3. Prehistory
Press
Mughal, MX.
1972 AS v of Excavations and Explorations in Pakistan (1971 and 1972).
Paki&m Archaeology. 8: 113-58.
Mughal, M.R.
1974 New evidence of the Early HarappanCuiture form Jalilpur, Pakistan.
Archaeology 27: 106-113
Pande, B.M.
1973 Inscriid Copper tablets fkom Mohenjo-dam. in D.P.Agrawal and A.
Ghosh (eds.), Radiocarbon and Indiun Archaeology . Bombay. 305-22.
Tata Institute of Fundamnetai Research.
Parker, B.
1955 Excavations at N i d , 1949-1953: Seals and Seal Impressions. Iraq.
17:93-125.
Parpola, A.
1970 The indus Script Deciphement: the situation at the end of 1969. Journal.
of Tamil Studies. 2:W- 109.
1975 Tasks, methods amd results in the study of the Indus script. Journal of the
Royal Asiatic Sociery, Vol. 2: 179-209. London.
1976 Interpreting the Indus script, II. S M a O~entalia,
Vol. 45:125-60.
Helsinki
1986 The Indus Script: a chailenging puzzle. World Archaeology. Vol. 17, No.
3:399-419
1994 Deciphering the Indur Script. Cambridge Univenisr Press. Cambridge.
Parpola, S., A. Parpola and R H . Brunswig
1977 The Meluhha Village: evidence of acculturation of Harappan traders in
late third miliennium Mesopotamia? Jouml of the Econornic and Social
History of the Orient. Vol. 20. No. 2: 129-65.
Parpola, A., S. Koskenniemi, S. Parpola, and P. Aalto
1969a Deciphennent of the Proto-Dravidian imcriptions of the Indus
Civilization. The Scandinavian Institute of Asian Studies. Copenhagen.
Special Publication 1.
1969b Progress in the deciphennent of the Proto-Dravidiun script The
Scandhavian Institute of Asian Studies. Copenhagen. Special Publication
2.
1970 Further Progress in the deciphennent of the Proto-Dravidian script The
Scandinavian Inshtute of Asian Saidies. Copenhagen. Speciai Publication
3.
Piggott, S.
1952 Prehisroric India to LOOO B.C. Penguin Books Ltd., Harmondswonh. UK.
Possehl. GL.
1967 The Mohenjo-Daro Floods: a Reply. Americm Anthropologist, 69: 32-40.
132
1977 The End of a State and Continuity of a Tradition: a Discussion of the Late
Harappan. in R. Fox (ed.), R e a h and Region in Tradironal India. New
Delhi. 234-54.
1990 Revolution in the Urban Revolution: The emergence of Indus urbanism.
AnnuuZ Review of Anthropology, Vol. l9:26 1-282.
Raikes, R.
1964. The End of the Ancient Cities of the Indus. Arnerican Anthropologist,
66:284-99.
1965 The Mohenjo-daro Flwds. Antiquity. 39: 196-203.
1968 Kalibangan: Death h m Nahual Causes. Antiquity. 2: 2869 1.
Raikes, R. and G. Dales,
1968 The Moheajo-daro Floods: a Rejoinder. American Anthropologist, 70:957-
61.
Raikes, R. and RH. Dyson
1961 The Prehistonc Climate of Baluchistan and the Indus Valley. American
AnthropologLn, 63:265-8 1.
Ramaswamy, C.
1968 Monsoon over the Indus Vdey during the Harappan period. N a m .
2 17:628-9.
Rao, S.R.
1963 A "Penian Gulf Seal fkom Lothal. Antiquiîy. Vol. 37:96-9.
1965 Shipping and Maritime Trade of the Indus People. Expeditiun. 7:30-7.
Zvelebil, K.V.
Harappa and the Dravidians: an old mystery in a new light. New Orient,
Vol.4, No. 3:65-9.
Comparative Dravùiiun phonology. Janua Linguanim, Series Practica,
No. 80. The Hague.
Dravidian case-suffies: attempt at a reconsmction. Journal of the
American Oriental Society. Vol. 92, No. 2:272-6.
The decent of the Dravidîans. International Journal of Dravidian
LUlguistics.Vo1. 1., No. 257-63.
A sketch of comparative Dravidian morphology., I. Janua Linguarum,
Series Practica, No. 180. The Hague.
Dravidian linguistics: an introduction. Pondicheny Instiîute of
Linguistics and Culture,Publication No. 3. Pondicherry.
Appendix 1
Table 3.2
Sign # Sign # sel
001 T d 6 2 Vakh 7 1 002 T"a149 vtritria 4 1
Ll*
C-i
M o R c a j ~a ) M-79. M-91. M-199. M-211. M-231. M-648.M-709, Mabuijodam a) M-7. M-8. M-53. M-IlS. M-708,M-796, M - l m . b)
M-727. M-772. M-869. M-980. M-1034. M-103s. M-1045. M - I W . M-74,M-243, M-634, M-726, M-1055. M-1098. M-1108. M-1630 (Bal);e)
M-1108. M-1122. M-132. M-1221; b) M-116: C) M-la). M-144. M427 M-14?. M-647. M-762. M-1016. M-1114. M-l IV. d i M-2A9. M-1067.
(Tw. Ma%. M4W. M-739. M-8% M - l m . M-1009, M-1418 (&SR d) M-116% M-1173. M-1369: Type C-> a) M-3%. M412. M-1%: b)
M.248. hi-647.M-839: Typc C.> b) M-37% M-376, M.402. M-1264. M-12'71. M-1332~)M - l m . Q M-3SJ
M-1298: d)M-357. M-363:Coppafobko-> M l : OCYm 3 H.nppi: a) H-29:b i H-153: Typc C-> a) H-659: b) H-5, H-62. H-420.
Hmppc a ) H-ISû (Bu).H476. H-799(Bas): b) H-23.H-31. if-62. H a . H-50s. H m
H-757 ( h l . H-758 (Bor): C) H-76. H-4% H M (BOSE.C) H-80;T y p C-> L o W r ) L-237(Fut). L-t 80 0
'- Type Cs> dl Ln)
b) H-441. H-684. Quihopdrro: b) C-1: c) C-5. C-30:
t#hrk a ) t 3 8 , L-21 1 Cfw. c) L-180Cfe e) L4. IWibuifP# C) K-18: d) K 4
Chsrhojodam e) C-1: 0 C-17.
$FI
MobcajIduo: a) M-102; c) M-1372 (Potk e) M m . M-89e. g) M-37%
M-1324 ( B u k Type C-> a) M-L2IQ.
nPnppi: b) H-22 T w C-> r ) H-1514, C) H-132 H-643; d) K- in.f) H-tSf
h) H-143. H-196;i ) H-160
torhjl: b) L-Io:Typ C-> d)L-102
a b
Sign # set
012 ' faal 8 V& 2 1.75
cks
Sign #
Mohenjo-âmx M-71. M42S 0:Type C->M-403.M-10% Coppcr Mohsajo-dam M-ln. M-478 (Bu),M-479 (Bu).M-480 (Bu).M-8%.
T & b > .%kJby:XCm 1; M-915
Hpnppr.Typc C-> Hd60 Hanppo: H-103
Loth& L.1.L-12
a b c a b c
Sign # set Sign # s ~ t
01s 1- VacrYI 3 1 016 Toril 7 valk~k~3 1
M d r i i j d a n x a)M-67. M-95.M-721
Hurpp:a) KST3
LabJ:b) L-94;cl L-26. L-218(Bas)
Sign #
017 T d 6 vlntr~~
I
0.05
Mohatjdam .U-87. M-1110:Type C-> M-361. M-366 Moktnj&mc M-7. ç1-184 M-1372(Pa)
Hpnppp: Type c->H-680 tkrPpprH.45
Loilummb l Q u n h o m w c-l
Sign #
019 4 v~iitiitr 1
a b
Sign # kt
024 T a 3 vpricria 2. 1
Sign # set Sign #
025 3 vatacs I I 026 3 VWoir 1
--
M o h c n j 0 - d ~M-262.
~ Type C-> W-383
Hprrppo: H-197(Ba)
Sign #
f Set Sign #
X Set 143
029 7 0 ~ 13 vtric~ia I t 030 T d 2 v* 1 1
di3 &
* Hiinppr H-76
&
K a f k q a W K-2s
Sign #
031 rd 2 v d 1
I
Sign #
033 T a 2 valjc~ks I
Set
1
a b
Sign # set Sign # set 145
037 T d 2 vyitacr 2 1
r"i
,~~ a) M-669
BYiPimü:b) 517
Sign # a Sign # kt 146
041 rd 1 1
V~U~CS I 042 Tacai 1 v~cdaI 1
t*i &
& r"i
A
Sign # Set
044 1 vuyas 1 t
Sign # S, Sign # set 147
045 1 vsictia 1
Sign # Set
047 Toal 1 vjritrics 1 1
Sign #
050 T W f vuiedrr 1
Sign # Sct
051 fd 1 viricoa 1 1
Sign #
055 ' VIrictYI 1
Sign # kt Sign# 150
057 1 V& 1 1-75 05% T d l Vtiitlrx 1 1.75
& &
63 6%
M o b t n m M-35
Sign # k
t Sm# set
075 T a 3 vali~~ksl 1.7 076 T m 1 1.7
& [+El
M a h m m M-1397 (BW H4SS
Hptnppe-H-17s(-1
CbrnbIErn- C-35
r~J2q~x
a b r k/g&r%
c d c p h i
Sign #
077 1 viricria I
Sign #
A set Sign# set 156
081 T d 1 viricrar 1 3 082 r d 1 v m ~ ~ t tr 3
set 157
OS5 1 VatMUI 4 O86 TocJ 9 VX~&S J s
& a
r"i Mnhai- JI M-11s. M-639. M-53:b) M4I. C)
B
M-lS3, M-W. M - l m .
4M-735
Lastul:W
a b a b
Sign # Set Sign #
087 vtnetr 2 s 088 T a 3 vuicritr 1
Sign #
OS9 2 vtiitrier 1
Sign # sct
092 1 v;rriccicr I 6
r"
ihi
Sign # set Sign # set
095 f Vu#iÉr 1 6 096 T d 9 vmcria I 7
a
M o & a w M-31. M-153.
.&'-lh M-9'2'2. M-12DQ.M-1127
HinppP:H-580
labrl:L-147 K 4 )
Kabbgam K 4
Sign # kt Sign# a 161
101 -rd ' VarierYI i 7 102 = d l v&fks 1 1
*[1 &
Mdwnpdnnx M-843 hnppcH-457
Sign #
106 Taal 1 vuieder I
& Gai
a
M d K a j d a r a M-S. M-7. M-24. M-29. M-30. M-38. M-39. M41. Ma42
M46. M49. M-W. M-79. M-81. M-84. M-93. M-129, M-133, M-141.
Mobea- M-5. M I O . M-23.M-24. M-21.M-29. M-30. M-38. M-39,
M a l . M46. M-47. M-51. M-52. M-56. M a . M-89. M-104. M-127.
M-144. M-166. M-186. M-199. M-208. M-236. M-240. M-280. M-285. M-129. M-234. M-t39, M-246. M-266. M-277. M-ML.M-314. M-323.
M-189. M-297, M-319. M-629. M a l . M-638. M-684, hi-1. MM. M-326, M a . M-86. M-638. M-647. M653. M-67t. M m - M-675.
M-671. M-705. M-706. M-721, M-758. M-759. M-793. M.801. M a . M-6116. M-703. M-72% M-735, M-757. M-7û9. M-777. M-788. M-793,
M-814. M-819. M-827.M-833, M-858. M-860.M4#. M-888. M-941. M a . M-814, M-818,M-865. M-897. M-934. M-962. M-96S. M-1001.
M-951. M-958. M-960. M-962. M-970.M-979.M-IOOS. M-1096. M-1099. M-IM1, M-1064. M-1081. M-IUB2. M-1096. M-1113. M-1146. M-1152.
M-1136. ,M-1150. M-1166: Bor Tabla->M-464 COM-467. M-492 M-1415. M-1190. M-445W). M 4 6 (Bu). M-L386 (Te . am), M-1576
M-1415
M-1416. M-1418: Typt C-> M-357. M-358. M-377. M-393. M-405. (Pol): Type C-> M-355. M-359. M-360. M-377. M-3%. M-399. M a .
.CI-1265. M-1269. M - l m C o p p T . b h i o -y: XClü 1Q. W k M-407. hi-1266. M-1269. M-1310. M-1343. M-1351: Coppa Tablct~->
ml9: M d a y X Q n 7:ManlIaaCYm 2
H-12. H-14. H-17. H-26. H-31. H-35, H-39. H-51. H a . H-76. H-12. H-21. H-S. H-39. H a , H-42. H a . H-M. H-58. H-76.
H-89. H-102. H-272 H-273. H-408, H a l 1. H42t. H-423, Ha6 H 4 1 .
H a . H.449. H4SO. H-569. H-598. H-610. H.649. H-#)6 (BU). H-231
F1-389. H-391. H-126. HG
H-150 t~ H-754. H-768. H-7 69, H-7% H-807:T y ~ C->
.
H478: B.t Trbk&> H-694. H-733 H-734.
e H- 129. H- I4 1.
(BUS). H-237 (Bu). H-?r18 (Bu).H-252 (Biir)U, H-276 (Bir), H-743 (Bu), H-148.K149. H-639. H-657,H-667
H-745 (BU). H-762 (Bu).H-808 (Brr). H421 (Btr).H-83 (-1. II-9 t~ hduL- L.9. I.23,L51. L.89. tt 11
a10(BpI1):TypC->H-131.H-148. H-659. HboO Q i n h o w C4. C. 14
h b i & 1-10, L-26. L-38. L-83. L-1 Il. L-211 (Ty).
Kdibuyin: K-5. K-16, Lob110)~tal
a
Mohcnjodam M-12. M-31. M-33. M-35. M41. M e . WU. M-111.
&- Rnmt 0.88
W o b c D j d ~ ~M4-
:
0.65 0.83
sel 166
121 T a 7 Vtrirrb 2 8 122 T d 7 v* 2 8
1 - 3 1 1 &El
A
Sign # set Sign # set
123 6 vuicria t 8.5 124 T w 5 vtraoa t a
M o b e n j w h ~ ~M-82.
: M-MO; Type C-> M-1291:Capper Ta&> MwK.y: Mobeapdw: M-IQ1
xcm 7 H-Pl=-
Sign #
129 2 v ~ h i t 1~
Sign # sct Sign # set
135 TotJ1 1 vsric~cr 1 8 136 ta 1 vtrictia 1 8
&
Sign # set
144 T a i v- I 3
Sign # Scr
148 T d 49 varietm 4 9
Sign #
M set
152 1 vtrroa I 9
fi
Sign # Sct
155 viriroa 2 10
Sign # set 176
r"i Clorr
ICMP_I
iii 6%
\\\
Set Sign # set
11 166 TaQl vtrrria 1 11
m
25
Sign # kt
168 Td vmicria 1 11
Sign #
Sign # set
172 v~ricricr I II
da
i-i
a b c d
Sign # kt Si@# SU 179
173 T d 5 v~cricr 4 12 174 toQt5 vai&~ 3 12.5
& [SEl
M ~ b e r i M-Zï4,
m M-1 164, M-1201 MabeaWM44. M-308
H-t6l Hinpprm
Qirnhirmpdira c-3. C-9
Sign #
177 T d 4 v~ I
M~UI~O&IUM-ML M-976
Sign #
181 Tual ' vuiccitr 1
Sign # k
t Sign # set
183 Topri vpricrier 1 123 184 1 V
- 1 12.75
&a
Sign # Sign# 182
185 1 vlricticr I 12 186 1 v&&a 1 12s
Sign #
187 v~ricticr t
Sign #
189 T d l s vmttia 1
Sign #
191 TorJl 1 vllicria~
Sign #
Set
VIricritr I 13.2
MdKnjo-dam M-32. M-72. M-80. M-93. M-W. M-95. M-tû2. M-132, MohmpdProi M-9.M-65. M-90. M-135. M-1 M. M-1%. M-160. M-188.
M-161. M-162. M-ln. M-179. M-212. M-215. M-251. M-292. M-318. M-201. ,W-229. M-263. M-268. M-670. M-678. M-726. M-739. M-837.
M-326. M-429 (Bu). M-458. M-712. M-838. M-922. M-9%. M - l m . M-875. M-928. M-937. M-941. M-991. M-992M-993, M - l a M-1160.
M-11OB M-1 t97. M-1232.M-1388 (BU). M-1429 (Bu), M-1433 (Bu) t~ M-1161: Type C-> M-389.M-412
M-1439 (Bu):Type C-> M-358. M-366. M-374. M-380. M-1287. M-t322. H 4 . H67, H-389.H.391. H a . H-53û.H 4 M : Typ~C-> H-129
M-1355. M-1364: Coppet T&~cEI-> MPcK;ry: XQll 2, 5. 6: CUI 4: Lo<hit:L I 8
MdmlkW14 Kaliùmgm~K-1: Type C s > K-59
H-2% H-48. H-49,H-54. Ha. H d Z H-68.H-70. H-81. H a . C-f s
H-450. H457. H-468,H499. H-536.H-585. H-921 (Inc) io H-923 (k): D m Dlv- t
TypeC-> H-134, H-141, H a 7 Kbarx Kn-1
wlra Bk-1
Nourhoro: N e 7
AIIPhdiaa Ad-2
1111
Sign # set
200 Taal 34 VPriclLr f 13
[$El
MW- M-75.M-96.M-103. M-2%. M-278. M-315. M-331. M 4 7 8
(6~M ) . O (Bor). M 4 (BU), M692 M-710-M-749. M-909.M-984.
M-tW3. M-14ZS (BU); T y p C-wM-3s. M-386. M413. .M-1262. M-IMS.
M-mo
H-55. H a . H-518. H-m.H-719 (-1. H-778(Bu);T y p C->
H-665
b h n L L i 9 3 (Ta@
l c d h q m c K-50
BmImrü:&9
sdcamtc Strtd-1
1111
Ill1
lrll
Sign # ser Sign #
201 30 V&&$ 1 13
63
j*[
M o h c r i w M-18. M-29.M a . M-1 If. M-252.M-638. Ma36. M-7. M d m p d m n M-20.M a . M-138. M-187, M-193, M-194. MG.
M-2S4,
M988. M-1053. M-1081. M-1369 (But); Type C-> M-368. M-381. M-397. M m . M-109%. M-125. M-1375 (Pa): Type C-> AM-378.M W . M-1368
M-399. M-1265. M-1%. M-1273. M-LW H-512. H-589. H-993 @OC)
H-75.H- 131. H-209(Ba). H-210, H-386;Typ Cs> H a 1 Lahak T y p C - > L a L m
Loctul:LM. L I 9 0 (Teg). 1-191 (l'y) Kdiknp~ K-19
Kjliipn:K-30
Sign # sa Sign # 187
205 toQtI5 vazick~ t 13.2 206 T d 11 vafw~I 13
& $El
$FI
Mobenjo-dam M-16. M-265. M-614 (Pot). M.699. M-835. M-1112. MobeDpduocM-39.M-00. M-116. M-= Type C-> M 4 û
M.I m.M-i rso. M-1202 FZmppw H e . H-924(Irwk Typc C-> FI-154
Harappo: H-Z19(IIJt). KM. H-801(b) t#hrl:L-22
LdUk L-12 IWibtirfrn:type C->K-59
Kalibongm KA. K-94(Pot) M p a :Dtpl
1 - -'
038
MdlLDjo&u M-331
Loitiok Type C-> L-97
Type C->K-59
QM
I b- c-30
Sign #
215 'faal 1 v* I
lltlll II
Sign # sct Sign # set 190
217 r d 1 v-1 13 218 rd 1 Vujuks 1 13
Sign # k
t
220 1 V& I 13.7
Sign #
IY Sian#
IY 191
221 1 Virrotr 1 13.7 222 T d 1 VJrriYI 1 133
Sign #
223 T a 1
KT
1111
Vorittitr 1
%
13.7
Sign#
224 T d f
: I:'
VPrictitr 1
Sa
13.7
Sign # set SignR kr 192
225 1 V&&s t 13.7 226 Totd t VuYtia I 13.7
111 I f ! II
I l 1 (Il 111
II II III
a b c
Sign #
229 *& 2 v a i ~ i2~
.
,
I
1-
- 1
Sign #
235 4 v~ricw I
1
Y i
m
&
Mohcnjo-dam M-114. M-140.M-149.M-627.+Ci-709. M-746.M-888.
M-967
Hhpft& H-2. H-7. H-90.Hall. H-757(Bu).H-758(BU).H-761
w:L45
ItrlibonQn:K-28.K40
C-22
A&h&mAb8
Wpa:Dlp2
Sign #
239 T w 3 v~ricritr 1
WIIILDI
Sign #
264 V~ 3 19
Clut
[SIMI
[BI
35
Mohjo-dim: M-ZS. M-91. M-114,M-182. M-938. M-971. M-1139.
M-1222' t y p ~C-> M-1x7. M-1284
Hiripp: H47S. H -mH-T7û
Larhil:L5
RbomitR: Wm-l
a b c d t f g
Sign # Sign #
269 T d 3 v- 1 19 270 Tm 3 vuiccnr t
Sign #
271 TorJl 3 V~~~CUC$I
mn
Sign # Set sign # Set
275 Ta vuktj~s I 19 276 T m 1 vtntcia 1 19
c"i [-mj
Sign #
277 Tw 1 V~ I
a &q
tahl:L-28 Labrt: L-36
Sign #
281 T d 1 vuimtr 1
sa
19
Sign#
282
-
Total143
-
--- Variaka 6
-
w<
20.5
206
1 & &
Mobcojodm: M-123. M-73. M-145. M-214. M-2û2. M-218. M-231.
M-263. M-290. M-7W. M-710. M-732. M-748. M-770,M-830. M-836.
a
M-874. M-879.M-916. M-W. M-%1. M-1û7S. M-1133. M-1135. M-12%
BU T & b > M468, M m , M-470. M-4722 M.978 B M a . M42.
M488. M 4 e . Typ~C-> M-3%. M-1278. M-1%. M-1311. M-1332
tirrrppr:H-75. H- 451, H-471. H-530, H-584; Bu T & b > H-172 H-174.
H-183.H-184. H-187 b H-190. H-2û3. H-Zû4.H-t07. H-211. H-213. H-214.
H-215.H-226 iû H-229. H-232. H-233. H-293, B-245 H-US. H--51.
H-286,H-296 ai H-303.H a , H-705. H-722. H-740. H-742. H-747 io
H-755. H-761. H-763. H-767. H-775.H-789. H-790. H-792. H-793. H-794.
Ha8012 W l 3 . H-816. H-817. H-841. H-842 H-WS. H-849. H-874; Incired
W> H-313 io H-318. H-819 î~ H-88t. H-893 iû 895. H-912. H-916.
H-917. H-943tû H-W. H m . H.961,H-Q16. H-978;Type C-> Hm160
lçlübmtur: K-U.K .% K a
QtRRojcdamc c-1s
K i Nd.2
Sign # Sian #
*-
sct set
283 T d 4 vuyoa 1 20.5 284 fa? vyicoa I 20.5
Mohenpdpro:M - l m
&
Hinppr H-271
Moheojoduo: M-7. M-12 M-14. M-15. M-18. M-21. M-2?. M-23, M-24.
M-26.M-2%. M-29.M-30. M-31. M-33. M-34. M-35. M-36.M-38.MAI.
M43, M-44.M-46.M-Çl. M-50.M-51. W-52. M-53. M-56, M-58.M a .
M a . M-69.M-72. M-78. M-79. M-82.M-84. M-89. M-90. M-91. M 9 2
M-W. M-98. M-99. M-100. M-101. M-lm. M-106. M-107. M-109. M-Il2
M-113. M-114 M-t 16. M-118. M ' M-121, M-124. M-127. M-128.
M-130. M-131. M-136, M-1" e f-143. M-144. M-145. M-146.
M-14. M-149. M-1%
M-161. M-170. V - pbr?474.
. IbISî. M-159. M-162 M-164.
,180. M-1, M-184. iK-189.
M-194. M-198. L A. M-203. M-m. M-211. M-212 M-213.
M-111. M-221. h <228. Km.M-nl. M-23U. M-m.M-238.
M-239. M-241. M-24t. M-295.M-W. M-249. M-255. M-2S7. M-160.
M-263. M-264.M-26% M-266. M-267. M-269,M-274. M-275. M-277.
M-280. M-281. M - a . M-286. M-288. M-289, M-291. M-294, M-297.
M-MO. M-301. M-m. M-M6. M-UI-MT. M-309. M-310. M-314. M-319.
W-320. M-322 M-324. M-326..M*â27. M-5%. M-&?S. M-626. iH428.
M-û29. M-434. M d 3 t . M-8. .W.M-647. M a . M-648, M-655.
M-665. M-671. M472. M m . M-678. M-682. M a . M m . M-699.
Sign #
vvua b c
set Sign #
289 Tom1183 v ~ 3 21s
w
Sign # set Sign # set
291 v~rictia 1 21.53 292 T d i S Vyiciitr 1 2153
&
hfohcn-
si
M-59. M-68.M-105. M-274. M-304. M-709. M-758. M.=
h M 9 Z M-115 1. Si-1631 (Bn& Type C->M-369.M-387. M a . M-1272.
hi-1%. hl-13%
Hmjp~a:H-36. H-270
Lathal: L-79
IW~brapn:K43. K-91 (Pot)
a b
Sign # sec Sign # ,210
297 T d 2 vifiebcr 1 21.3 298 r d 2 II- 1 213
GiI
Mo&njdmx M-180(BW.Coppcr Table&* Maeky: Cm 2
El
a b c d
Sign # sa Sign # sa 211
301 Td 3 215 302 T d 46 v~ 4 t t5
$El r"i
Moheojodam M-4. M-44. M-50. M-54. M-12.M-86, M-IOO. M-117.
P-i
Mo&apdmr: i) M-8, M-IO. M-U. M-314. M-621. M a t , M-674. M a S .
M-118. M-122 M-128.M-135. W248. M-276, M-286. M-303. M a 2 6 M-677. M-747. M-753. M-781, M-7%. M-= M-1014. M-1079. M-1091:
(Tau. M-670. M-677. M-714. M-7% M-757. M-78% M-797. M-814. b) M-1633 ( B aC) M a . Ma. M-234,M-1 IOl. M-1203; d) M-781: T
m
M-840. M-860. M-863. M-864, M-928. M-932. M-9S6. M-957. M-959. C-> I ) ) wmo. M-1290. M-na. M-1359 CI M-1'49:CO- ~rb~ctt-,
M-985. M-997. M-1057. M-lW7. M-1168, M-1155. M-1191. M-1200. M a m t d k m Il
M-12%. M-1629 (BqkType C-> M-355. M-359. M.365, M-1299. M-1328. Hmppa a) H-5, H-8. H-30, Ha. H-53. H-268. H-388: b) H417. H 4 S .
M 1355. kt-1364: Coppa TrblaI> Murha:ÇXW S. 6: CXVm 2 TypcC-ai)fi-131. H-149. H-153. H a . H a H-682
HPtcippo: H-24. H-25. H-383. H-455.H a . H-550. H-175 (Bir). H-76l -alb1
(BU), H-'fi (Spi): T y ~Cce > H-136. H-137. H-641.H-649
&thL L-I0.L-II.L14J C'l'y).L-l46CTit)
Kdi- K-89
C k o n h u j d C-7.
~ ~ C-9. C-38
y 0-2
Rakhigarhr w-2
P b c
Sign # set Sign # set
303 rd v~icricr 3 21.7 304 T d 5 vali~k 7 21s
chu
icHpl da
6%
M o b e n j w MJ7. M-55. M-149. M-239. M-27s. .-284. M-314. M-700.
3%
M-733. W-735. M-&3. M d 6 2 M-1070. M-1071. M-1079. M-1100.
M-1103.M-1155. M-1163.M-il77
H-13. H-101. H-161. H-593. FI-9731-(
Loth& L-S. L-211 (Tm)
Kdibuign: K-24.K4û
x i Nd- 1
a b c
Sign # ser Sign # Sct 212
305 -rd2 vnicrier 4 21.7 306 Tou l3 vuit~ia 3 21.7
[SI
Sign #
311 T d 3 v- 1
a b
Sign # set 2 14
313 Tw 2 viritcitr 2
a b
Sign # set
315 2 vtricricr 1 11.8
Sign # SU sign a set 215
317 T d 2 vpima 2 11.7 318 TdQl 1 Viricpa 1 11.7
&a
a
Sign #
319 facrf 1 -v- I
Sign # sa
321 TotJ1 vmerrr I 21.7
&
Kdiboiyirn:K-IS
il
Sign # Sip# a 217
325 mai 1 V&i&s 1 21.7 326 ratai 1 Vmcria 1 21.7
r"i m
Sign # sa Sign # k
t
331 *d 1 vacries 1 21.8 332 T d 1 vuitoa 1 215
&
v
Sign # 219
334 'rcQl1 v;inttitr 1 215
&
El
Sign # set Sign # sci 220
337 h d 1 Vuiak 1 21.7 338 T d 2 vliietia I 215
1
I
m
scc PIIpd.(1992112$iï7.14)CktrCZd B7forFcplvrmcntse~
Sign # set
340 Td 1 vpieirr 1 21.7
Sign #
341 Tocrl. 127 Vuittia 1 -
SU
n
Sign #
VtrrDer 2 22f
r"i
Mohcnjoauo: M-14. M-32. M42. M-58. M-75. M-80. M-90.M-101. M~h#i- Ma. M-16. M-31. M-41. M a . M-54. M-86. M-103.
C-i
M-110. M-116. M-141. M-144. M-148, M-164, M-165. M-177. M-198, M-108.M-117. M-157. M-160, M-163.M-174. M-205. M-200.M-202.
M.204. M-221. M-25. M-36. M-259. M-278.M-282, M-289, M-314, M-214, M-235. M-26û. M-268. M-288. M-300.M-3û3. M-314. M-322.
M U . Md46, M-650. M-658. M-692. W-703. M-722. M-722,M-7= M-326. M-437(Bu). M-595, M-621. M d 3 4 M-649. M-W 1. M-662.
M-776. M-792 M-803. M.816. M-819. M-8?3. M - m . M-835. M-839. M-6û4. M-700.M-727. M-726. M-728, M-7%. M - a . M-865. M-875,
M-851.M-853. M-855, M-868. M-89S. M-897. M-929. M-940. M-944. M-889. M-900. M-931.M-998. M- 1045. M-1066. M-lû88. M-1103.
M-963. M-976. M-984. M-1085. #-1110. M-1126. M-1166. M-1177. M-1 LW. M-1109. M-1119. M-1134. M-1136. M-1137. M-1138, M-1139.
M-1226. M a 7 1 (b) . (Tw);Tm C-> M-362. M-364. M a .
M-1391 M-1118. M-1369 (But). M-1640 (En@. Typ C-a M-356. M-351. bM-361.
M-1268. M-1299. M-1311 M-37Z, M-375. M-381. M-384. M-385. M-390. M a . M-1264. M-1186.
H arom H-13. H-15. H 4 X H-54.H a . H-266. H-388. H-407. H-456. M-1190. M-1306 M-130% M- 132% M-1346. M-1350
H 4 9 . H.978. H W . H-501.H-609. H-694(Bor). He695 (-1. .
H-697 (Bu), HPnppi: H-1 H-IO. H-18, H-19. H-26. H-45. H-6A H-80. H-88. H - M .
H-778 (Bir). H-825(Bu).H-901 (Bu) iû H-9C# (Bu). H-300 (trc),H-987 H-390,H-472. H-Sû6. H-Xn. H-514. H-558. H-S74. H-598. H-688. H-176
(ta);Trpe C-> H-137. H-140. H-154. H a 5 8 (BW.H-231(tkr). H-833 @W.H-829 (Bu).H-900(1n~). H m (Potk Typ
tachol: L-1 1. t 4 1 . L-84.L-92. L l 6 l (Twio L-171 CTw c-* H-669. Hbn
IWfbanpn: K-27. K-78 (Bu).Cbuihuja-dam C-4. C-IO. C-18. C-23. Latrtl:L I & L-21, LW CTW.L M 8 (Toi). L-208CiW.Type C-> L-122
WwJli: &S. 5 7 . E l S . Dcnlpor: DIpi. Miol.via DVI-2 Jbrilar: nt-2 Kolibrnpn: K-6. K-16. K-23. Km78 (Bu). W i g a r ù i : Rgr-3 (751).
Nmdowpiiawrb:Nbt S u t k o m k Sktd-1. Sktû-3 (Pot). Sktd-4 (Pot). Nawhoro: Ns-5. N e f .
UnhKRRC?-5
&
MohtnjIduo: M-IO. M-29. M-30. bM-173. M a 7 2 (Bir). M-656. M478.
M-686. M-759. M-916. M-1- T w C-> M-357. M-358. M 9 . M-370.
M-371. M-391. hl-392 M-393. M a . M-009. M-1265. M-1287, KI%.
M-1319. M-1323. M-1324. M-1339. C o p ~ e r T & b >MrKnp XCnt 3.
12.14: ,Mmùdl: O(Vm 1.2
H-20. H-21, H-22 H-25.H-27. H-58. H-78. H - l a . H-388. H-3%.
H-l90@W. H-a)i) -1. H-724 (8rr). H-753 -1. H-TI5 T m C->
H-130. H-1 3 2 H-135. H-143. H-145, H-146.H-152. FI-154. H-160. -9.
H a . H-661. H-667.H m . H-679. H-681
Loth& L-19. L a . L-83
Sign # set 222
346 rw 17 vpiciri 2 22
w
Sign # set
347 =orJi17 vimicr I n
a
MOhUIf*
Tablas-> Mx-
a
l
Type C->M-354.M-391.M-lm.M-1295.M-13% Coppr
XCüI 3.4. IO. 1s:Mmùdk aMf S. 6.9: CXVIII 5
Typc C-> H-132. H-639.H-659
Loihil: L-86
Sign #
351 fa 9
Wlih -
bchcii
V~ 3
Sign # Set Sign # set 224
353 fml 7 vuictitr 1 22.7
P b
Sign # set Sign # set
359 4 Vatkh 1 1.7 360 T a 3 VUMCS I 223
&
Sign #
361 T a 3 v m c t ~ ~1
Sign #
363 T d 3 vmec~r 1
Sign # set Sign #
365 fa 2 vti~oa I m
r"i
&
M a b a i p d m a M-13
H-184
Sign #
369 T d 1 v ~ i ~ i 1a
Sign # Set
371 1 valida 1 1.7
Sign # kt
375 =aJi ' V W t 22
Sign # Set
379 nd 1 VIritricr 1 ??
Sign #
381 T d 1 v* L
Sign #
383 T d 1 vtiima I
I
/@O\
Sign #
385 vliicria I
Sign # a Sign #
a
Mdwnjdara M-104. M-184. M-1%. M-85. W - a l . M-8SZ M.123 M d w n j o d a x Md. M- 145. M-22%M 4 J : Type C-> M-1318
Sign # SC< 235
398 T a 2 VX~C&S I 24
a
a
Sign #
399 ' v* f
Lœlul I
I/
Sign # Signif set
403 T d 1 vpr~lrr 1 B 404 T d 1 vmicria 1 24
d.u
a
il
et iiej
Sign # Sign #
/IO11 ,237
405 T d 1 vpirdcs 1 21) 406 1 vrriecitr 1 24
a &
a rEei
Sign CI Sct
24
408 Td I v m 1
Sign # set 238
Sign # sa
i
Patarr
Moiwn-
0.64 0.74 0.83 0 3
1a
M-14. M-W. M-M. M-58. M-IOI. M-245.M-246. M478.
M-809.M-859. M-861. M-937.M-961, M-109% T m C-> M-362. M.374.
M-1307. M-1318. M-1353.M-1363: C- TrMeo-> MYK.r: XClll 9:
ClII 1: MinhPII: QÉM 7.8.12; c%wrx 4
Hprpppi: H-13. H-139. H-157. H-If4 (BU),H-LS7(BU). M. Hm.
H-521.H-599. H-7W (BU) tû 7 10 (hl. H-712 (BU): Type C-> H a l .
H-666
Loth& L- 10. L-20.L-78
IWiôunp~Kt89
Ahmimtw.A68
Sign # set Sip# set 239
414 Tom133 vuitt*f 4 2S.7 415 ?dm v~titr 1 255
Mahtnjoau# M-20. M-48, M-51. M-5% M-58, M-92. M-US. M-281. Mo&njodam M-23.M-280. M-638. M-788. M-866. M-1001. M-1168.
M-379. M471 (BU). M 4 S . M-683.M-723.M-809.M-84% M-870. .H-lsil(BW
M-8% M-936. M.937. M-961. W-1085. M-1133. $4- 1 1 9 2 Type C-> Hmpp& H-47.H-199 (Bas) ta H-2û2 (Bas), H-216 (Bas). H-270. H-tlt
H414.H-743(Buh H-745 (BU)
.
M-1307. M-1318. M-13S3
H-8n H416. H-51% W72S (hl. (EUS)
H-729 tahrl:Lt3
Lahnl.Lm10
Knlikuigaa K-18
Sign # Set
421 vtric&r 1 255
Sign # set
422 3 VSIWC$I 2s.7
a
M-219; Type C-> M-1260. M-1271
Sign #
424 TOQI 3 v w 1
r"i
Sign #
430 f voryria 1
Sign #
432 rd 1 vtrict~~I
! I
Sign #
434 1 voirpia 1
Sign # set 246
442 1 v ~ ~ oI a 26
Sign # sa
445 1 vpitricr 1 25
Sign # set
448 ma 14 vatc# 1 27
GiI
&
Mohajoasta: M a l . M-636.M a l . M-934.M-940. M-tOn. M-1320. MW- M-127. M-237.M-246.M.832 M-1107
M-1 Ul: Type C->M-368 Hmppc Type c->H-679
HPnppP:H-21.H-386
Lahnt. L-9.1-51. L-89
Sign # ,248
451 T d S variaia I a
&
MW- M-1% M-220.M-243.M-1173
FtrtppP:
M-119
a
Sign # set Sign # ,252
466 1 vmctia 1 n 467 1 v* 1 a
a a
a 23
63
Mohtn-
23
M-98. M-120. M-722. M-776.M-798. M-1112. M-1138;
Typ~C-> M-13s; Co~pcrT & b % XCIII 1-8. 11: MM1LPII:
OCYlU 2 3 : M V I I 3
H-3. H-268.K-383.H-446.H 4 2 T m C-> H-136
Kdibonfrnt K-2 K-13
Sign #
a
Sign #
486 4 v a 2
Sign #
488 1 v* I
1
I
Sign # se< Sign #
490 T d 3 vrritm 1 3t
1=1
Set
31
I
II u
Mobcajo-duo: a) M-la. M-736 cl M-34. M-755. Type C-w c)
M-372. M-1271
HInpps: a) H-17a (Bas). H-218(Bu). H-37(Ir),H-817 (Bu): b) H-818
( h l . H-892(Inch C ) H-216 (Bn),H441.H-893
Wibaagax e) K44
a
Mohcnj-
a
M d l . M-849.M-937;Type C-> M-1-ur; CoppaTablar-w
,%kKayzXCm 9;MnnbillrMYII 8.IZ;CXYm 4
1
Il u
Mobeajwlam M-1101. M-1110; Type C-> M-374. M-1363: Coppcr
TrMcrIwhhîmkCXM 7
" q p - 3 HPJppl' H-13. H-2B (Bu):Type C-> H-t S7
Cbinhumpdwrc-XI
Sign # Set Sign #
502 T d 2 vurticr 1 345
Sign # k
t Sign f set
504 ToQL 1 V~ 1 345 505 1 Vliictia 1 345
+k$
L d & L-86
Sign #
506 T d 6 variaiu I
Sign # k
t
528 1- 3 v a I 35.75
63
Happa:
B
LodukL-45
Biih-lrot Bk-2 B W
Sign #
532 3 vurcia I
set
35.75
sign #
533 2
El
Vmcricr 1
Sa
35.7 5
$!El
Sign #
534 rd 1
r"i E
$
iJ
E h p pH-51
Sign # S, Sign #
538 TasJ vmctia I 3.7 539 TOPl 1 Vaaiaka 1
a b c d c
Sign # Sign # set
540 TacJ1 f va~iaia t 541 v~ricritr 5 M
&a
iroj
M a k a j w h œ Q M-218;Type C->a) M4Qt; Q M-1322 CopprTibka->
d ) M C X M 14
b) H-48
Loth& 1) L-36:t) L-90
a b
Sign # SU Sign #
542 T d 40 V* a n 543 36 Varhiu 6 37
mi, 63
M-38. M- 176. M- 425 ffw.
M- 4% (BuLM-626. M- 628
rQi 635
Mober,pduo: M-21. M-23. M-26. M-69.M-70. M-134. M-17 1. M-234.
M- 671. M-682. M-771. M-816, M- 8S7. M- %O, M- l m . M- 1156. M- M-626. M-666. M-751. M-994. M-llll. M-1127, M-1181. M-1189.
1340. M- 1444 (k) .1445 (IIIc); Tm C-> M-W. M- 37% C-
M- M-1206
TIbkfc-wMmbakOCVn 13 H-1. H-7. H-M. H-91. H.237 (-1. H-267. H-300 (Inc). H-360
Hmippr H-8. K I 2 Km. H-61. H-S09 ( h l . H-236(k) H-278
. (EU) i~ (inch a363 Cine). H423. Hal (Bas) IO H-8% (Bas). Hg18 (BU): Type
284 (Bst). H-321-1. H-580. H-783 (EW,H-783 -1. H-917 (Iir). H-466 C m > H-142 Ha65
(Irr) (ta
Ladrnl:L-S. L-199
Kdi- K-IO Bamwak B.21
Sign #
544 Total36 v~ 1
set
n
Sign #
545 T d 9
X vuitlia I
Set
n
&
Mohai- M-38, M-QI. M42S
a
M 4 8 (PNTsb). M a . M a .
M-671. M W - M-816. M-950. M-9W. M-993. M-11%. M-119. M - l m .
Mohaijdam M-267.M-393. M-629. M-634.M-667. M-t 149: Type C->
M-IQT-
Typ C-> Ici-369. M-377. M-1- C o p p c r T W > kbhdk CXVIl 13 H-35. H 4 l
Hamppl: H-12. H-m. H-61. H-278 @as) m H-2& (Bas). H-m.H-m@a).
H-598 (EW. H-7U3 (Bu). H-761 (B@
t#h;rl: L-11
biibsnpui: K-10
Sign #
X ser Sign #
'X
II
kt 272
546 T W 3 vtntcicr I n 547 T d 2 v~pmo 1 37
Sign f sct
549 faai i vwiaia 1 37
Sign # Sign # set 273
550 fd 1 v- I 37 551 Tadl VirWricr 1 37
1*
Mobaijodam M-261
Sign #
552 1 VIncotr I
1
Sign # k
t Sign # set
556 T d 1 vtritm t n 557 1 VPieria 1 n
& &
25 da
Sign #
558 1 v ~ i ~1 k
Sign # Set
560 T d +r vptictitr 1 38
Sign # set
564 T d varwc~ 1 38
tabal
Sign # Sign# kt
568 1 vpiericr 1 38 569 TOQII 4 v m t 39
Sign # a Sign#
P sct 278
570 Td 1 vllieiier 1 39 571 T a a Vtricdtr 1 40
aao
8
mi
3 fi&
M o b j d a m M-23. M-44. M a , M-53. M-59. M-68. M-M. M-105.
M-Ill. M-113. M-136, M-146. M-173, M - 2 9 l M-326. M-427 (Tau.
M-7. M-709. M-754. M-755. M-7611. M-820. M-841. M W . M-870.
M - m . M-966. M-974. M-997, M-IWO. M-1078. M-1114. M-1129.
M-1165. M-1190. M-1601 (PMk Type C-> M-356.M-363. M-369. M-380.
M-387. M-398, M - l n l . M-1280. M-1289. M-13%. M-1MO
Happ:H-7. H-39. H4.H-89. H-268.H-380(CO). H-391. H411 (Bu)
Lacbrl: L-28,L I 9 0 (Ta& Trpc Cs> L-114
Korikagax K-4
Qirihtijodmx C-3. C-Il. C-12. C-15. C-16
Dtrolpor:D l p l
El
ibtohenjo-duo: M-67. M-102. M-119. M-121. M-138. M-1s. M-1%. M o & n m M-7. M-33. M-1%. M-128. M-170- AM-171. Y-174. M-677.
M-195. M-W. M m . M-893. M-937. M-958. M.9fi. Typ C-> M-374. M-72û. M-846. M-917. M-945.M-968. M-1153. M-11% T m C-> M 4 .
M-W. M-1225 M-1358. M-1368. C- tabla^> XClH 3. S. M a . M-1314.M-1364. M-l4Uk C o p p r T i b b > m CXM 11
1.1 H-32H-931(Int)
H-25. H- MI. H-4û8. H 4 2 2 H-512. H-589. H-927[ t i r ) H-930 CarimBhimGrbl
tlrr,
I W i h p m K-9. K- Il
Sign # set Sign #
574 =al' Vtrieoa 1 40
r"i
1 EI
M d K n m : M-24.M-M. M-133: T m C-> M-3%
KllibOngDR:K-30
Sign #
582 1
'fd v;writr 1
MdwnjPdaio:W-210
Sign # set 282
586 T d 1 v- 1 40
Mobcopd.iior M-843
8. b.
Sign # Set
588 T d l O variaies 2 21
Total Set
I Mohenio-Dam Harama Lothal
58
et 855 21
Sign # 1Rquency 472 29 1 34
varicties
288 Percent by Site f 1.53 13.51 9.44 1 1.09
i
Mohenjo-dam: M-7, M-12, M-14, M-15, M-18. M-21, M-22 M-23,M-24. M-26. M-28. M-29, M-30, M-31, M-33. M-34, M-35. M-36, M-38, M41, M-43.
M-44. M-46. M47. M-50. M-SI. M-52, M-53, M-56. M-58, M-67. M-68.M-69, M-7% M-7% M-79,M-82 M-84. M-89, M-90. M-91, M-92 M-W, M-98, M-99.
M-IOO. M-101, M-104. M-106, M-107, M-109, M-112, M-113, M-114, M-116, M-118. M-120. M-121. M-124, M-127. M-128. M-130. M-131, M-136. M-138,
M-140. M-143. M-IM. M-145, M-146, M-148, M-149, M-150. M-152. M-155, M-157, M-159, M-162, M-164, M-165, M-170. M-171. M-173, M-174. M-180,
M-1822 M-184. M-189. M-194, M-198, M-199, M-200. M-203, M-209, M-211. M-212 M-213. M-217, M-221. M-223, M-228,M-230, M-231, M-234.M-237,
M-38. M-239. M-241, M-242 M-245, M-248, M-249. M-255. M-257.M-260, M-263, M-264, M-265,M-266. M-267, M-269, M-274, M-275, M-277. M-280.
M-281, M - W . M-286. M-288. M-289, M-291, M-294.M-297, M-300, M-301, M-304, M-3M. M-307, M-309, M-310, M-314, M-319. M-320, M-322 M-324,
M-326. M-327. M-595, Md=, M-626, M-628, M 4 W . M-634, M-637. M-638, M-644, M-647, Md46. M W . M-655, M W M671. M-672, M-677. M478.
M d 8 2 M-683, M-693, M-699. M-700, M-706. M-708, M-709. M-714, M-717, M-720, M-722, M-723.M-727. M-128, M-733, M-335, M-736,M-737 M-738.
M-746, M-750. M-754. M-755. M-756, M-758. M-761, M-771, M-780, M-782. M-785. M-786,M-792. M-801. M-808. M-8 10. M-812 M-8 14, M-816, M-818.
M a , M-832, M-833, M-835. M-837, M-839, M W . M-847. M-850, M-851, M-853, M-856.M-857. M-859, M-861, M-862. M-863, M-865, M-866. M-870,
M-878. M-879. M-880. M-882, M-889, M-890, M-892, M-893, M-895, M-900.M-903, M417. M-921. M-926. M-932, M-935. M-933. M-937, M-942 M-943,
M-944, M-945. M-W6, M-950. M-954, M-953, M-957, M-958. M-963.M-966, M-965, M-968, M-971, M-972, M-979,M-980, M-982, M-986. M-994, M-1001.
M-1005. M-1003, M-1006, M-1009, M-1010, M-1015. M-1017. M-1018, M-1019. M-1020. M-1021. M-1027. M-1028, M-1034. M-1045. M-1052. M-105%
M-1062. M-1069, M-1070. M-1071, M-1079, M-1077, M-1085. MM-1088.M-1096, M-1100, M-1104, M-1107. M-1108, M-11 IO, M-1111, M-1112 M-1114.
M-I IlS. M-1119, M-1127, M-1133. M-1135. M-1136. M-1138. M-1139, M-1146, M-1152 M-1153. M-1155. M-1156, M-1163, M-1165, M-1166. M-1168.
M-1 177, M-1 179, M- 1 181, M-1194. M-1195. M-1202. M-1200. M-1206. M-1221, M-1225, M-1224. M-1233; Bas Rdief Tableu-> M-445,M-450, M-453.
M-456. M459. M-457. M-460 to M469, M471, M-472, M-473,M 4 Q M 4 8 6 M-495, M-540, M-1402, M-1406, M-1409. M-1413, M-1412 M-1415,
M- 1416. M-1418. M-1424. M- 142% Pou-> M-1379, M-1382 M-1577. M-1578, M-1592, M-1591, M-1601; Bangles-> M-1647; Ivory R&> M-1650, M-1651;
Tap-> M-1383. M-1385, M-1386; Inciscd Tablets-> M-1438, M-1439, M-1441. M-1442 M-1443, M-1444. M-1445. M-501: Type C-> M-354, M-355. M-357.
M-359. M-360, M-362. M-363, M-364, M-369, M-370. M-374. M-375. M-376, M-377, M-380. M-381. M-384. M-388, M-389. M-391. M-392, M-393. M-394,
M-397, M-398. M-399, M405, M-407. M-409. M-1263, M-1264, M-1265, M-1266, M-1267.M-1268, M-1273. M-1276, M-1288. M-1291. M-1294. M-1304.
M- 1301. M-1306. M- 1307, M- 1308. M-13 10. M-1314. M-13 18. M-1321. M-1323, M-1326. M-1328. M-1337. M- 1340. M-1342 M- 1343. M- 1344, M-1353.
M-1361, M-1364, M-1366. M-1368; CopperTablcu-> MacKay Cm 1.3.4.7; XCIIl 2 3 . 4 5. 6.7.8.9,lZ 13. 14; Marshail: CXVII 1,3,3,7,8.11.12 14.
15; C X W 1.4. Hnrappa: H-2 H-3, H-8. H-12 H-13, H44. H-20. H-22, H-23, H-24, H-25, H-26, H-27. H-30. H-3 1. H-32. H-36, H-37. H-39, H 4 , H 4 2
H-45, H-46, H-47, H-50. H 3 1 , H-56. H-58. H-59, HdO. H d l . Hd2, H-63, H a . H a . H-69. H-75, H-76, H-79, H-87, H-88. H-101. H-102, H-103. H-268,
H-270. H-283.H-286. H-289. H-291. H401. H48, H 4 l l . H417. H-420. H-421. H432, H431, H M , H441. H a . H449. H-451, H 4 5 2 H-453, H-454,
H-457. H-461, H-462. H-465. H e . H-468. H-469, H-471, H472. H475, H-478. H479, H a , H49..205H- H-5 10. H-5 1 2 H-515. H-5 19. H-521. H-523.
H-524, H-533. H-537. H-546. H-565, H-563. H-569, H-577. H-580. H-586, H-589, H-592 H-593. H-597, H-598. H-599, H-612; Bas Relief Tableu-> H-171.
H-174, H-176. H-178. H-179, H-la H-187. H-188, H-199, H-200.H-201. H-203. H-204, H-206. H-211. H-216. H-217. H-218. H-229, H-233, Y-236.H-245.
H-247. H-248, H-250, H-278 to H-284. H-694, H-695. H-696. H-697. H-699, H-705. H-707, H-708, H-723, H-722. H-733, H-734. H-740. H-742. H-743.
H-750. H-75 1. H-752 H-753. H-754, H-755, H-757. H-758, H-763. H-764, H-765, H-767, H-768. H-773.H-78% H-788, H-790, H-800. H-806, H-807,H-808.
H-811. H-817. H-818, H-823. H-841. H-844H-846, H-847, H-848,H-849, H-852. H-854, H-8SI. H-874. H-877.H-875: Incùod Tablets-> H-285, H-287,
H-290. H-189. H-29 1. H-294. H-296, H-298, H-300, H-301. H-302 H-303. H-304.H-305, H-306. H-308. H-309, H-310, H-3 12. H-313, H-3 14, H-3 15, H-316.
H-317, H-318. H-321, H-323 H-323. H-325, H m , H-880, H-890. H-892, H-894. H-893, H-903, H-905. H-918, H-91% H-921. H-924, H-925, H-935. H-936.
H-954 H-959. H-960. H-961, H-962 H-964H-966, H-967. H-%9, H-970.H-973, H-975, H-978. H-977. H-984, H-987. H-988, H-341,H-343,H-345.H-350,
H-35 1. H-352, H-353, H-3%. H-355. H-356. H-357 H-360, H-361, H-362; PO&> H - W . Type C-> H-129, H-132 H- 133, H-134. H-135, H-137. H-138, H-140,
H-147 H-143, H-147. H-150. H-156, H-158, H-161. H-639. H-641. H-645. H a . H654, H-656, H-652 H-659. H a l . H-666,H-669, H-8, H-684. hW:
L-5. L-14. L-22. L-23. L-S.L26. L-28. L-31. M. L-55. L-58.L-56, L-59. L-63. t6rl; Type C-> L-95. L-1 I l , L-112 L-1TZ; Tags-> L-140. L-141. L-143.
L-149. L-161 to L-170, L-180, L-190, L-189, ï-191. L-193.1-20s. L-206. î-211: BasRclidTiblets->L-217. KPlibmgan: K-2 K-9. K-10, K-Il, K-13, K-15,
K-16. K-17. K-23. K-24.K43. K 4 K-61, KdZ K-63: Bas RcIief Toblcrs-> K-77. K-78; Tags-> K-81, K-87,K-88, K-89; Pou-> K-90.K-LOO. K-108. K-106.
K-105, K-104: Coppet Objcct-> K-121. Chanhujodaro: C-1. C-2, C-3, C-5, C-8. C-15. C-17. C-29.C-33. B~iawaü:B-1, Chandigark Ch-1 (Pot), Ch-2 (Pot).
DPimPbad: Dmd-1 (Bui). Desaipur: Type C-> Dlp-1: Hufas: HIs-1 (Tag). hpkar: Jk-2. KErsim Km2 L o h u m j h Lh-1- Raldiiphi: Rgr-1. Sltrllotak
Sktd-2. AUlrMino: Ad-1, Ad-4, Ad-6. A&& Gharo Bhrro:Grb-1. NauJharo: Ns-8, NiS. Wtndowari-damb: Nd-1. Pi& Pk-1 URirnown: 1 6
- -
Total Set
l
1
MohenibDaro Harauoa Lothal 0th-
I 187 8
Sip # ErÉqucnc~ 118 46 8 15
1 Varicties
112 Percent by Site 2.8 8 2. 14 2-22 2.87
L .
Mohcnjwiam: M-23, M-37, M-36, M-38, M46, M-49. M-52 M-53. M d l . M-65, M-91. M-94, M- 108. M-117. M-124, M-130, M-133. M-136, M-140. M-147,
M- 151. M-154. M-163. M- 1732 M- 174. M-211. M-218, M-221. M-234,M-237. M-240. M-260.M-279, M-304. M-308, M-309. M-319, M-323,M-623. M-627.
M d 3 1. M-632. M-634, M-636, M-648, M-650. M-651, M-655, M-661. M-707, M-715, M-720, M-722. M-723, M-726, M-733. M-768, M-856, M-888,M-900,
M-9 14. M-92 1. M-943, M-967, M-999, M-1020. M-1031. M- 1044, M-1052, M- tOS'?. M-1112, M-1121. M-1148. M- 1159. M-1169, M- 1206, Type C-> M-359.
M-375. M-38 1. M-392. M-395, M 4 1 Z M414. M-1265, M- 1275. M-1294, M-1295, M-13 12, M-1329. M-1330. M- 1350. M-1362; Bas Tabler M a , M-453,
M464 to Er1469. M-1426. M-1427. M-1429. M-1439 to M-1442; C Tabltu->MacKay: XQn 1,2,4,5,6,7, Il; Miirsball:CXVm 5
Harappc H-4. H-9. H-IZ H-18. H-20. H-68, H-82, H-85. H-92 ~-"1%5f~-320 (Incl, H-364 W ) ,H-268. H-270. H-385. H-396. H 4 1 . H-412. H-423.H-514.
H-53. H-569. H-58 1. H-592. H 4 l . H-609;Bas Tableu-> H-203. H-230. H-278 ra H-284.H-747, H-748, H-761, H-767. H-789, H-807. H-815; Type C->
H-128. H-133, H-134, H-640, H-669
L o W : L-66.L-87.L-98. L-114. L I 4 3 (Tag). i-208 flag), L-211 (Ta@, L-219 (Tag). Kaliban- K-7, K-8. K-9, K-15, K-25. K-32. Chanhujo-dam: C-IO,
C- 13. C-33, C-38 (CO)+Banawali: 5 1, B-3. Pabumatb: Pbm-1 N i n d o w i u i ~ b Nd-1.
: Nawharo: Ns-S