kyoto for Cindy

13th Harvard University Round Table
ETHNOGENESIS OF SOUTH AND CENTRAL ASIA (ESCA)
Kyoto Session
Research Institute for Humanity and Nature (RIHN)
Kyoto, Japan
30 – 31 May 2009
南アジアと中央アジアにおける民族集団の形成
13th Harvard University Round Table
ETHNOGENESIS OF SOUTH AND CENTRAL ASIA (ESCA)
Kyoto Session, Research Institute for Humanity and Nature (RIHN)
30 – 31 May 2009
PROGRAMME
May 30
09:00-09:10 Introduction Narifumi Tachimoto
Discussion on the dispersal of agriculture and domestic animals in Asia, especially South Asia
09:10-09:15 Section summary Yo-Ichiro Sato
09:15-10:15 Late Harappan “collapse”, the opening of central Asia and long-distance
crop movements Dorian Q Fuller
10:15-11:15 Cropping Strategies and the Indus Civilization: New Crops, Regional
Variation, and Climatic Adjustments Steven A. Weber
11:15-12:15 Two very different millets: Setaria itralica and Spodiopogon formosanus, in
Asia Emiko Takei
12:15-13:15 Lunch
13:15-14:15 The Spread of Domestic Animals in South and East Asia Richard Meadow
Current Trends in Harappan Archaeology
14:15-14:20 Section summary Toshiki Osada
14:20-15:20 Cemetery Assemblages, Stratigraphy, and Chronology: A view from
Harappa Jonathan M. Kenoyer
15:20-15:35 Break
15:35-16:35 Harappa: The Role of an Urbanized Bronze Age Populace in the Population
History of South Asia Brian Hemphill
16:35-17:35 Human Burial Customs during 3rd and 2nd Millennia BC in Haryana
and Kuch: An Analytical Approach Vasant Shinde
The Harappan Burials in Gujarat P. Ajithprasad
May 31
Discussion on wide connection between South Asia and Gulf including issue of the Indus script
09:00-09:05 Section summary Toshiki Osada
09:05-10:05 The collapse of the Indus-script thesis, five years later: Massive non-literate
urban civilizations of ancient Eurasia Steve Farmer
10:05-11:05 Four World Quarters in the late 3rd millennium BC: Ur <> Shimashki <>
Meluhha <> Magan (and the bits in between) Daniel Potts
11:05-12:05 The Asiatic wild ass (Equus hemionus) in Harappan, Dravidian and Indo-
Iranian record Asko Parpola
12:05-13:00 Lunch
On the diversity of wheat varieties types and linguistic diversity in India: A hot spot of DNA
and languages
13:00-13:05 Section summary Michael Witzel, Yo-Ichiro Sato and Toshiki Osada
13:05-14:05 Genetic diversity of Afghan wheat landraces and their potential for future
breeding Tsuneo Sasanuma
14:05-15:05 Traditional management of agrobiodiversity of Rukai aboriginal peoples in
Taiwan Hsin-Fu Yen
15:05-15:20 Break
15:20-16:20 A hot spot of linguistic diversity in the Greater Hindukush/Pamir area: The
names of agricultural plants Michael Witzel
CONTENTS
Preface (M. Witzel)…………………………………………………………………… 1
Discussion on the dispersal of agriculture and domestic animals in Asia, especially South Asia
Late Harappan “collapse”, the opening of central Asia and long-distance crop movements
(D.Q Fu l l e r ) ………………………………………………………………………… 3
Cropping Strategies and the Indus Civilization: New Crops, Regional Variation,
and Climatic Adjustments (S.A. Weber)……………………………………………12
Two very different millets: Setaria itralica and Spodiopogon formosanus, in Asia
(E. Takei)…………………………………………………………………………………13
The Spread of Domestic Animals in South and East Asia (R. Meadow)………………16
Current Trends in Harappan Archaeology
Cemetery Assemblages, Stratigraphy, and Chronology: A view from Harappa
(J.M. Kenoyer)…………………………………………………………………………17
Harappa: The Role of an Urbanized Bronze Age Populace in the Population
History of South Asia (B. Hemphill)……………………………………………………19
Human Burial Customs during 3rd and 2nd Millennia BC in Haryana and Kuch:
An Analytical Approach (V. Shinde)………………………………………………………22
The Harappan Burials in Gujarat (P. Ajithprasad)…………………………………………24
Discussion on wide connection between South Asia and Gulf including issue of the Indus script
Four World Quarters in the late 3rd millennium BC: Ur <> Shimashki <> Meluhha <>
Magan(and the bits in between) (D. Potts)………………………………………………29
The collapse of the Indus-script thesis, five years later: Massive non-literate urban
civilizations of ancient Eurasia (S. Farmer)………………………………………………31
The Asiatic wild ass (Equus hemionus) in Harappan, Dravidian and Indo-Iranian record
(A. Parpola)………………………………………………………………………………33
On the diversity of wheat varieties types and linguistic diversity in India: A hot spot of DNA and languages
Genetic diversity of Afghan wheat landraces and their potential for future breeding
(T. Sasanuma)……………………………………………………………………………36
Traditional management of agrobiodiversity of Rukai aboriginal peoples in Taiwan
(H.-F. Yen)………………………………………………………………………………38
A hot spot of linguistic diversity in the Greater Hindukush/Pamir area:
The names of agricultural plants (M. Witzel)……………………………………………40
1
PREFACE
Our series of Round Tables began ten years ago with the cooperation between Harvard
archaeologists, linguists and textual scholars. Our aims stated in 1999 remain the same now:
we want to ascertain and describe the present state of the art in our respective disciplines
and we try to overcome the existing mutual incomprehension between our respective fields.
We aim to avoid using the partial results of disciplines other than our own without first
gaining a closer understanding of their background, procedures and limitations, as this has
lead, all too frequently, to circular feedback and argumentation. What is needed instead, we
feel, is more direct interaction, especially between archeologists, geneticists, anthropologists,
textual and linguistic scholars as each of our fields represents only one sector of the evidence
available for the historical problems at hand.
From the beginning, we have stressed that, preferably, an overlap of large sections
of the data available in our fields should be established; close interdisciplinary cooperation
would enable comparison of such data with as little initial interpretation as possible. Our
stress thus was and is on more, consistent cooperation.
During our meetings, we have always aimed at relative short presentations,
followed by extensive interdisciplinary discussion. At our Harvard meetings they famously
were nearly ‘endless’ ones – so that all participants were satisfied with the answers given to
their questions. While we will not have the same amount of time at this year’s meeting,
a significant section (20 minutes) as been set apart for detailed discussion. This time limit
must be respected for the sake of scholarly exchange, and presentations must be restricted to
the allotted time frame of 40 minutes.
Over the years, we have consistently expanded the areas covered by our Round
Tables: since 2001 by human and plant genetics and since 2004 by comparative mythology.
In 2005, due to the kind invitation of and skillful organization by RHIN
sanctioned by its then Director-General T. Hidaka, we could hold an international meeting
here at Kyoto. Its results have promptly been published by T. Osada -- something we had
not done before. We are very grateful for the invitation, organization, and ready assistance in
all practical matters of this memorable Round Table that covered most areas of Asia.
In some other years, on the contrary, we have dealt at length with just one
particular topic, such as in October 2006 in a meeting on Southeast Asian and Sahul Land
2
linguistic problems, on the respective remnant languages and substrates, or in May 2006 in a
meeting at Peking University devoted to comparative mythology; this topic has subsequently
been split off, after our section during the 2005 RHIN meeting, and is now dealt with at
length by a separate organization, the International Association for Comparative Mythology,
IACM. We will have a separate meeting about this topic at the Kokugakuin University,
Tokyo, on May 23/24 of this year.
Since 2006, we have also returned to our earlier format of Round Tables, and
have held a small but well attended Round Table at Harvard last year that was attended by
Professors T. Osada and Y.I. Sato -- an invitation that they kindly have returned to us now
by organizing the present Round Table and by bringing a large number scholars of Indus
archeology and related fields to Kyoto, for which we are very grateful. We wish this meeting
the same great success that RHIN has achieved in organizing the 2005 Round Table.
By constantly keeping up our dialogue between the various disciplines dealing
with early South and Central Asia (and often beyond), we have, to a large degree, succeeded
in a better mutual understanding of our individual aims, methods and the limits of our
respective fields, resulting in a better evaluation of our individual results.
It must be admitted, though, that in spite of constant interaction, our terminology
is not yet always the same -- as we will also notice in the present symposium-- and that some
engrained conceptions remain, due to the individual pathway dependencies of our respective
fields. Continuing with our dialogue thus is of great importance.
In addition, some serious issues remain unresolved among participants; some of
them will be picked up during this year’s Round Table. We hope that the present meeting
will lead to still a better mutual understanding between our respective disciplines.
We thank the authorities of RHIN and all those who helped in preparing this
Round Table for their extensive efforts, expressing our hope that the results of this Round
Table will be as significant, or even more so than those of the memorable 2005 one
organized by RHIN.
Michael Witzel
Preface
3
Framing a Middle Asian corridor of crop exchange and agricultural innovation
Dorian Q Fuller
Institute of Archaeology, University College London, UK
(currently visiting research fellow at Research Institute for Humanity & Nature, project 9)
This presentation explores the archaeobotanical patterns from Asia through Eastern
Africa at a large comparative scale, reviewing the evidence for exchanged of crops (cereals,
pulses fruits), and cropping systems, over long distances and between very different
cultural traditions. The “Middle Asian” region, stretching from the Arabian Peninsula
through the Iranian Plateau and Central Asia, represents both the frontier between
monsoonal and winter-rain climates, but also a corridor of smaller-scale more mobile
societies that seems to have played a key role to moving crops and innovations, starting
from the Third Millennium BC. It will assess the selective spread and adoption of crops
as they entered new agricultural and culinary worlds: wheat, barley and pastoralism
Figure 1. Map of general zones of early agricultural origins in the eastern Old World.
4
moved East into China before 2000 BC; select Chinese cultivars, including seed crops and
fruits, spread to Central Asia, the Indus region and beyond to Yemen and Africa; African
staple crops moved East to South Asia and beyond. At a later period, at the End of Bronze
Age or Iron Age, innovations in irrigation and labour intensive crops like cotton may have
followed similar pathways. The processes of selected crop spreads must been understood
within a local context of suitability to existing systems of food production and food
consumption, and the ability of systems to adapt existing systems of agricultural labour.
Archaeobotanical research has a key role to play in better documenting these processes,
and more effort is called for to fill in major geographical and chronological gaps in current
evidence. The background for this paper lies in our improved understanding of the separate
origins of agriculture across several parts of the Old World. A selection of domestication
centres is summarized on the map in Figure 1, indicating paired millet and rice foci in both
India and China, as well as two unconfirmed plausibly central Asian regions for additional
millet domestications. (Note that this map lumps some of the distinct sub-centres of early
agriculture in South Asia, which the author has outlined in detail elsewhere). In broad terms
the Old World grain cultures can be divided into the summer, monsoon-driven systems of
East Asia, South Asia and sub-Saharan African and the uniquely winter-rain-based system of
southwest Asia. There were also probably multiple eutropical vegecultural centres of origins
(like that indicated for New Guinea/Indonesia). Exchanges of crop species between these
centres of origin took place both across the seasonality frontier, i.e. as winter crops moved
east into China and India, but also between the different summer crop zones. This will
highlight our current understanding of the evidence for these crop transfers, which seem to
have been remarkably simultaneous, with exchanges starting sometime around ca. 2500 BC
and finished by perhaps ca. 1800-1600 BC.
The selective uptake of Western Domesticates in Chalcolithic China
Chinese agriculture today and historically includes an important component of
dry-crops, including both native millets and introduced wheat and barley. These are crops
native to Southwest Asia that spread eastwards with the initial Neolithic as far as Pakistan
and Turkmenistan, where they were established before 6000 BC. By contrast the first
appearance of wheat and barley in China is dated to ca. 2600 BC, and for much of central
China, where Chinese Shang civilization emerged, they appear only in the Bronze Age, from
13th Harvard University Round Table on the ESCA, Kyoto Session, RIHN, 30 – 31 May 2009
5
ca. 1800 BC. In addition, central China received domesticated taurine cattle and sheep from
the West, ultimately from southwest Asia. Both of these animal domesticates appear between
2500-2000 BC, i..e in the Longshan period. For all of these domesticates it is clear that their
adoption was a selective process, which must have been governed by local agricultural and
culinary preferences in China. Of the Near Eastern crops it is only bread wheat which is
widespread, tetraploids wheats (emmer and durum), so frequent in Western Asia are absent,
barley is extremely rare as an accompaniment to wheat, and the other Near Eastern founders
crops (pulses, flax) are entirely absent. Amongst animals, goats seem to have made little
headway in central China. This same process of diffusion is suggested to have introduced
copper metallurgy.
Chinese millets outside China
The earliest agriculture in China was based on broomcorn millet (Panicum
miliaceum) and common foxtail millet (Setaria italica), both probably domesticated by
ca. 6000 BC in northern China. It remains a point of some debate as to whether either of
these species had additional regions of domestication outside of China, such as in western
Central Asia, around the Caucasus perhaps (for Panicum) or northern Afghanistan (for
Setaria) (see Figure 1). One hypothesis suggests that Panicum miliaceum alone spread early
in the middle Holocene across the northern steppes of Asia to reach eastern Europe by the
Neolithic at around 5000 BC. While that hypothesis remains unsubstantiated, unambiguous
evidence for Panicum miliaceum in central Asia comes from the mid-late Third Millennium
BC. This same period also sees reports of Panicum miliaceum on the Arabia peninsula,
suggesting that it may have spread south from Iran. By ca. 1900 BC this millet is found in
Pakistan and possibly Gujarat. It also spread beyond Yemen to Africa, as indicated by finds
from the Kerma period in Nubia (by. 1700-1600 BC) and later Iron Age cultivation. The
period that this species first occurs in South Asia, after 2000 BC might also be when the
other Chinese millet, Setaria italica arrived there. This raises another area of continuing
uncertainly over the antiquity of Setaria italica in South Asia. While it has been suggested to
be in Gujarat as early as 2600 BC, difficulties with species level identification in the genus
Setaria, and between Setaria and Brachiaria ramosa raises reasons to doubt earlier reports,
which is equally true of early Setaria reports from Arabia, Iran and central Asia. I propose
the hypothesis that both Chinese millets saw their main period of spread from China to
Framing a Middle Asian corridor of crop exchange and agricultural innovation (D. Q Fuller)
6
central Asia and southwards towards India and Africa in the late Third and Early Second
Millennia BC, together with less ambiguous evidence for several other crop plants and some
technologies of Chinese origin.
Figure 2. A schematic map of crop exchanges between Africa-India,
China-India with approximate dates.
A Chinese horizon in Northwestern South Asia: Other crops & technology
Archaeobotanical evidence indicates that a few other crops, originating in China,
have their first appearances outside China in northwestern South Asia, during the Late
Harappan time Horizon (Figure 2). These include two fruit tree crops, apricot and peach,
both of which have a rich archaeobotanical record in Central China, from the Yellow River
to the Yangzte, from 6000 BC onwards. While it is unclear when they came to be cultivated
as opposed to collected, it seems likely that they were cultivated by later Neolithic/Longshan
times, and they must have been cultivars when these species began to planted beyond their
wild distribution (which extends westwards to the Tian Shan Mountains). Thus finds from
the Late Neolithic of Kashmir (at Burzahom and Semthan) which date to after 2000 BC
provide the earliest unequivocal evidence of their cultivation, and their transport outside
China. Interestingly these sites also provide evidence for a technology of Chinese origin, the
13th Harvard University Round Table on the ESCA, Kyoto Session, RIHN, 30 – 31 May 2009
7
hand held harvest knife, which traces its roots to the Cishan-Peiligang Neolithic in China,
but occurs outside China in the Late Harappan era both in Kashmir, where such tools had
earlier been absent, and in Swat. Although in China these are initially associated with millet
cultivation, and later with Yangtze rice, the spread to Kashmir represents a selective adoption
of this technology into a culture where agriculture was focused on wheat and barley.
Rice may also have spread from China at this time, although the history of rice
is complex and likely involves more than one source. In all likelihood early finds of rice
in Harappan contexts (e.g. Kunal) and in Swat from the Third Millennium BC are likely
to represent a spread of primitive indica cultivars, indigenous to the Ganges. However,
modern indica cultivars contain genes borrowed from East Asian japonica rice, indicating
that hybridization between these two lineages of rice was essential to improved productivity
and success of monsoon-adapted indica rices. The Late Harappan ‘Chinese’ horizon in the
northwest, which included finger knives, may have included japonica rice as well. Short
rice grains, a trait typical of temperate japonica and short-stalked bulliform phytoliths (also
typical of japonica) both occur at Late Harappan Pirak, Baluchistan, strongly suggesting that
some japonica rice strains had been introduced by ca. 1900 BC.
A final crop which makes its first appearance in India at this time, but which has
earlier origins in East Asia is Cannabis sativa. Genetic proxies suggest that East and South
Asian Cannabis varieties, especially for drug use, are related. There is plausibly a second
separate derivation in western Central Asia/ Eastern Europe. The earliest archaeobotanical
finds comes from near Obama Bay Japan, at the Torihama Shell Midden, before 3500 BC,
with Late Yangshao finds reported from northern China. In South Asia this species occurs at
Senuwar Period 2 (after 2000 BC) in the Middle Ganges, Phytoliths from Harappan Period
4/5 (after 2000 BC) and Harappan Kunal (Period 1C, but inadequate publication and
overlying later deposits suggest some caution).
For all of the above reasons, I am prone to reject more dubious claims for earlier
dispersals of Chinese millets and to suggest that these also came into northwest South Asia in
this same general “Chinese” horizon at the start of the Second Millennium BC, or perhaps
the late Third Millennium (Figure 2). The earliest evidence for possible contacts between
China and India via the Bengal-Assam-Yunnan route are later (perhaps from ca. 1200 BC?,
as indicated by the first Gangetic find of the fibre crop ramie), although archaeological
evidence from this region is extremely sparse.
Framing a Middle Asian corridor of crop exchange and agricultural innovation (D. Q Fuller)
8
The African horizon in savannah India
The importance of the early appearance of African domesticates in South Asia
has long been noted. These included Sorghum and hyacinth beans from the Eastern
savannahs, pearl millet and cowpea from the western Savannahs, and finger millet from
the East African hilly zones. Although there remain some debates over the antiquity and
identification of some finds, especially of finger millet, there is no doubt that several Africa
domesticates were established in parts of South Asia by Late Harappan times. A focus of
early African crop diversity outside Africa was Late Harappan Gujarat (2000-1700 BC)
and some of these species had spread to the South Indian savannahs by 1600-1500 BC,
confirmed by direct AMS dates on hyacinth bean. Early finds in the Ganges are perhaps a
couple of centuries later, although some reports from the Eastern Harappan zone may be
early but remains poorly documented. Some of these crops spread further east, including
finger millet to Yunnan and northern Southeast Asia and sorghum to China, but these
dispersal remain undocumented. As with the spread of Chinese crops, of Western crops
Figure 3. A generalized map of important cultural zones of the Eastern
Old World at ca. 3000 BC.
13th Harvard University Round Table on the ESCA, Kyoto Session, RIHN, 30 – 31 May 2009
9
to China, the adoption of African crops in India appears to have been a piecemeal process, as
they were selectively added as fairly minor supplements to established indigenous agricultural
systems. In the longer term they played an important role in increasing agricultural diversity
and adaptability.
A role for small scale societies: central Asian horsemen and Arabian boatmen
The evidence for the movement of domesticates, presumably through contacts of
long distance trade, raises the question as to who were the players in such trade and who
were the principle transporters. Similar questions surround the identity of those who moved
crops from Africa to India, and vice-versa, by ca. 2000 BC or shortly thereafter. While the
obvious candidates might seem to be the major urban civilizations, which we know to have
been part of trade networks, the hard evidence suggests that this was not the case. However,
I will argue that it was not the big players and better known urban states that were involved,
but rather the smaller-scale, less centralized and more mobile societies that operated on
the interstices. A comparison of the Asian works at around 3000 BC, the era of the Early
Figure 4. A generalized map of important cultural zones of the Eastern
Old World at ca. 3000 BC.
Framing a Middle Asian corridor of crop exchange and agricultural innovation (D. Q Fuller)
10
Harappan period, Sumer, the Late Yangshao and Liangzhu Neolithics of China, is one in
which there are fairly clearly defined foci on increasing population density, sedentism, and
trade of raw materials, mainly mineral, from peripheries. Trade links were relatively limited
(e.g. the upper Persian Gulf, the Iranian plateau, between Egypt, Nubia and the Red Sea;
between the Shandong Peninsula and the Lower Yangtze). By contrast the world at closer to
2000 BC is very different.
For one thing, the civilizations of the Indus and the Yellow river (the Longshan
horizon) had expanded, and trade in and around these spheres increased, as did that between
Egypt and Kerma. Also, however, there is important evidence for the spatial expansion of
cultural connections across the Asian steppe (the Andronov horizon) and in Arabia (with the
Wadi Suq/Dilmun horizon). The expanding connections of these latter, less-sedentary and
more mobile groups, aided no doubt by increased use of camel and horses on the on hand
and improved boats on the other, may have been the key component in transmitting crops
over long distances as part of expanding trade contacts. These mobile groups helped to stitch
together the previously separate worlds, of the jade-focused trading sphere of China (Late
Yangshao-Qiujialing-Dawenkou-Liangzhu) and the metal-trading sphere of western Asia
(in which tin and copper figured importantly). On the Arabia end of the world the incense
trade, which is at the roots of the later extensive spice trade, became linked to Middle Asian
exchanges. The fact that this period also witnessed the decline of the many urban/ proto-
Urban societies, such as the Harappan cities, is probably also important. It suggests that
innovations in trade and agriculture may have contributed to undermining established urban
orders.
13th Harvard University Round Table on the ESCA, Kyoto Session, RIHN, 30 – 31 May 2009
11
Figure 5. A schematic of the two major Asian trading spheres of the Third Millennium BC
which become increasingly inter-linked through mobile small scale societies in
between 2500 and 1900 BC. A similar role is implicated for Arabian coastal boating
communities for linking the Red Sea and Arabian Sea trade networks.
Framing a Middle Asian corridor of crop exchange and agricultural innovation (D. Q Fuller)
12
Cropping Strategies and the Indus Civilization: New Crops,
Regional Variation, and Climatic Adjustments
Steven A. Weber
Department of Anthropology, Washington State University, USA
Agricultural practices within the Indus Civilization were continually evolving like all aspects
of culture. The introduction of new crops alongside shifts in climate have either allowed
or required new cropping strategies to develop. When cultural, ecological and botanical
variables are taken into consideration the complexity of the Indus strategy becomes clear.
This paper explores this complexity and offers alternative avenues for archaeobotanical
research. A thorough examination of the archaeobotanical record will give us insights into
the link between ecology and culture, which in turn will help us better understand the
evolution of the Indus civilization.
Patterns in Harappan agriculture. Pie
charts represent the summed presence
of crops across the sites in three areas
of the Harappan civilization during the
Mature Period. Blue shades indicate
winter crops; red, summer crops.
Presence of major crops/crop groups is
indicated with whiskers. Locations of
sites with only Early or Late Harappan
evidence also indicated, but not
included in the sum. Western sites (1-
10), Northern/Eastern sites (17-25),
Southeasterrn sites (31-38). (Map by
D.Q Fuller)
13
Two very different millets: Setaria italica and Spodiopogon formosanus in Asia
Emiko Takei
Distribution and Communication Sciences, Osaka Gakuin University, Japan
In Asia, there are many small-grained cereal species known as millets. In this paper I compare
just two examples, the wide-ranging Setaria italica and the local-endemic Spodiopogon
formosanus.
Setaria italica also known as foxtail millet is an annual grass cultivated almost
throughout Eurasia, in cool temperate to fully tropical areas. It is historically important
as a crop that can be grown in very diverse environments. The wild type, Setaria viridis, is
itself widespread and easily dispersed as a weed species, and is now cosmopolitan. Within
foxtail millet, we can find many local landraces and varieties that display variation in plant
height, tillering, panicle number and size, daylight sensitivity, glume and seed color, and
starch qualities. As starch qualities there are distinct glutinous and non-glutinous forms. The
glutinous form is found in East and Southeast Asia, and diversified use is developed such as
sticky steamed cakes and fermented drinks. In the Ryukyu archipelago and Taiwan, sowing
and harvesting of foxtail millet were objects of major seasonal rituals.
Spodiopgon formosanus is a little-known endemic crop of Taiwan. This perennial
grass was first described by Rendle in 1904, on the basis of one specimen collected in
Southern Taiwan. The genus Spodiopgon has about ten known species, and ranges from
West to Northeast Asia. S. formosanus is the only species that has been reported in
cultivation, and is only cultivated by the Taiwan native people. This plant was collected
by Japanese botanists during the Japanese occupation period (1895-1945), but was not
well documented as cultivated plant, and was misidentified by ethnographers. For want
of a proper name, ethnographers referred to it as "hie", the Japanese name for Ehinochloa
utilis, Japanese barnyard millet, a crop that is common in Japan but never grown in Taiwan.
Through this and other confusions in naming, Spodiopgon formosanus has been almost
invisible, in literature, for over a century.
Two closely related wild species are known in Taiwan, S. cotulifer and S.
tainanensis. S. cotulifer is widespread in warm temperate areas from Northeast India, to
China, Taiwan, Korea and Japan. In Japan it is a common grass on sunny hillside with
14
moderate and infrequent human disturbance. On the other hand, S. tainanensis is endemic
to Taiwan, an alpine grass growing on mountain slopes. Both relatives have the same
chromosome number as S. formosanus, 2n=40, but only S. cotulifer is common in settled
areas, in ruderal habitats. From its proximity in habitat, and similarity in morphology, S.
cotulifer is an obvious candidate progenitor for S. formosanus, but this possibility has not
yet been investigated.
Historical and oral reports indicate that S. formosanus was grown in the shifting
cultivations of mountain villages from North to South in Taiwan, as well as foxtail millet.
Now it is cultivated in just a few villages in the central and southern mountains, by Bunun,
Rukai, Paiwan people. The plant is perennial, tillering, with non-shattering panicles, and has
oily stems and seeds. After dehusking, the grains are made into porridge. Although different
ethnic groups in different areas have been growing this plant, almost no variation is evident
among cultivars, except for the presence or absence of red pigmentation on the plant. The
starch is non-glutinous, and informants clearly state that the crop is not suitable for making
sticky steamed cakes. Agricultural rituals were previously recorded for this crop, but the
rituals have disappeared.
From archaeology, and from the differentiation of many landraces throughout
its range, we can see that foxtail millet must be an ancient crop even though we cannot be
sure exactly when and where it was domesticated. Although there is no comparable body of
information, Spodiopgon formosanus could be an ancient domesticate in Taiwan, but one
that never spread beyond the island. Most crops in the country are introductions of greater
or lesser antiquity. Foxtail millet is certainly one of the older introduced crops.
Despite their very different historical records, foxtail millet and Spodiopogon
formosanus may have been domesticated in similar ways. The likely progenitors of both
grasses can easily grow in disturbed, open habitats in the vicinity of settlements, and in both
grasses -- as in other cereal crops -- selection for the non-shattering seed habit may have been
of primary importance for domestication. However, because Spodiopogon formosanus has
remained endemic within Taiwan, it has not been exposed to selection in diverse physical
and social environments, and has had no opportunity to diversify into numerous local
landraces and varieties.
The continued survival of this crop is not certain. It is clear that its abundance and
range in cultivation declined throughout the 20th century. Over the same period, the most
13th Harvard University Round Table on the ESCA, Kyoto Session, RIHN, 30 – 31 May 2009
15
obvious changes in local economy have been the expansion of cash-cropping and a decline in
local production of staple foods. Among the many indigenous crops in Taiwan, Spodiopgon
formosanus may face the greatest risk of extinction. I can see some hope for this crop within
Taiwan indigenous movements to maintain, restore and develop cultural traditions.
Two very different millets: Setaria italica and Spodiopogon formosanus in Asia (E. Takei)
Setaria italica field, Kurnool District, India (Photo by D.Q Fuller)
16
The Spread of Domestic Animals in South and East Asia
Richard Meadow
Peabody Museum and Department of Anthropology, Harvard University, USA
Archaeological, zoorchaeological, and genetic research over the last decade shows that South
Asia was a significant player in bovid domestication and the development of pastoralism.
Recent analysis particularly of mtDNA from modern cattle and water buffalo indicate that
wild forms of these animals native to northwestern South Asia contributed in differing
degrees to the genomes of their domestic descendants. The spread of these domestic forms
into East Asia, as suggested by recent zooarchaeological research in China, underlines the
importance of South Asia in the development of pastoral practices in surrounding regions as
well.
Bos & Bubalus herd, Kurnool District, India (Photo by D.Q Fuller)
17
Cemetery Assemblages, Stratigraphy, and Chronology: A view from Harappa
Jonathan Mark Kenoyer
Department of Archaeology, University of Wisconsin, USA
Abstract
The main focus of this paper will be on the analysis of artifact assemblages from the
excavations of the Harappan cemetery (R-37) at Harappa undertaken from 1986-1988
and also in 1994. These excavations by the Harappa Archaeological Research project have
provided new insight into the types of artifacts included in Harappan burials, as well as
complex issues relating to the site formation processes of the cemetery itself. Due to the
intensive use of the Harappan period cemetery over approximately 700 years, many burials
were cut and disturbed by later interments. The problems of interpreting the stratigraphy
and also the chronology of the burials will be examined. Excavation of a Late Harappan
burial pot that was discovered in 2007 will also be discussed.
Summary
The site of Harappa is unique among all of the Indus settlements, because of the
fact that is has cemeteries dating from the Harappan period (2600-1900 BC) as well as
from the Late Harappan period (1700-1300 BC). Excavations carried out by Daya Ram
Sahni and M. S. Vats in the 1920s-30s focused on the Late Harappan burials in Cemetery
H, and Harappan burials in Area G. In Cemetery H they uncovered two layers of burials
representing two different types of burial traditions and also two different periods of burial.
K. N. Sastri who was the Curator of Harappa Museum discovered and excavated Cemetery
R-37 in 1937, but his report was never published. This cemetery however was clearly earlier
than the ones discovered by Vats and dated to the Harappan period. Some ten years later
in 1947, Sir Mortimer Wheeler excavated a trench between Cemetery H and R-37 in order
to resolve the chronology of the two cemetery areas, but many aspects of the pottery and
burial practices were still unclear. Dr. Muhammad Rafique Mughal also undertook small
scale excavations in 1966, focusing on the area of cemetery R-37 to the south and east of the
earlier excavations. His preliminary report provided some additional insights, but no final
report was published. Renewed excavations of a much more extensive area of cemetery R-37
18
were carried out by the Harappa Archaeological Research Project in collaboration with the
Department of Archaeology and Museums, Government of Pakistan between 1986-1988.
The initial excavations were under the direction of the late Dr. George F. Dales and Dr. J.
Mark Kenoyer. Additional excavations were carried out in 1994 under the direction of Dr.
Richard H. Meadow and Dr. J. Mark Kenoyer. Most recently, in 2007, Dr. J. M. Kenoyer
was able to excavate an urn burial of the Late Harappan period.
The major objectives of the renewed excavations of the Harappan cemetery by the
Harappa Archaeological Research Project were 1) to better define the assemblages associated
with the burials, including pottery and other artifacts; 2) to determine the chronology of
the burials using relative stratigraphic approaches as well as radiometric techniques; 3) to
obtain a wide range of skeletal materials representing the ancient populations of Harappa
in order to better understand their physical nature, health and mortality. In the following
paper I will summarize the findings from the first two categories of evidence. Four physical
anthropologists associated with the excavations have already published summaries of the
ancient populations and the full cemetery excavation report is soon to be published.
The detailed examination of stratigraphic relationships and post-depositional
processes have made it possible to understand that the cemetery area was continuously being
disturbed and reorganized by the excavation of new burial pits, the exhumation of earlier
burials and the redeposition of skeletal materials in collective burials or dumps. The pottery
and other artifacts associated with the burials can be related to specific occupations on the
various mounds of Harappa, indicating that the individuals in this cemetery lived in all the
major walled areas of the ancient city. Based on radiocarbon dates and ceramic comparisons,
the chronology of the cemetery as a whole can now be more precisely related to the rest of
the site and to other sites of the Indus Valley.
13th Harvard University Round Table on the ESCA, Kyoto Session, RIHN, 30 – 31 May 2009
19
Harappa: The Role of an Urbanized Bronze Age Populace
in the Population History of South Asia
Brian E. Hemphill
Centre for South Asian Dental Research and Department of Sociology and Anthropology,
California State University, USA
The Indus Civilization represents one of the signal developments in the prehistory of South
Asia. Linking together urban centers of the Indus Valley and beyond, it is logical to suspect
that this early urban development wielded significant impacts upon the biological structure
of human populations; not only of the Indus Valley, but beyond into peninsular India.
The current study examines the impact of Indus Civilization populations on the biological
history of human populations of South Asia through multivariate statistical analyses of three
genetically controlled systems of human variation: craniometry, odontometry, and dental
morphology.
Craniometric variation is compared between human remains recovered
from Cemeteries R37 and H at Harappa with those obtained from 29 skeletal samples
encompassing 1,505 individuals (845 males, 660 females). These samples range in antiquity
from the Early Bronze Age (c. 3500 B.C.) to the modern era and derive from the Indus
Valley, Iran, the Russo-Kazahk steppe, southern Central Asia, western China, Nepal and
Tibet. Odontometric variation of all permanent teeth except third molars is compared
between human remains recovered from Cemetery R37 at Harappa with data obtained
from 21 samples. Together, odontometric data encompasses a total of 2,166 individuals who
range in antiquity from the aceramic Neolithic (c. 6000 B.C.) to the modern era and derive
from the Indus Valley, the Hindu Kush highlands, southern Central Asia, and peninsular
India. Sample differences in frequencies of 17 dental morphology tooth-trait combinations
are compared between individuals recovered from Cemetery R 37 at Harappa with data
obtained from 20 additional samples. These samples account for a total of 2,105 individuals
ranging in antiquity from the aceramic Neolithic to the modern era and derive from the
Indus Valley, the Hindu Kush highlands, southern Central Asia, and peninsular India.
Results from neighbor-joining cluster analysis and principal coordinates analysis of
20
inter-sample variation in cranial dimensions are concordant and indicate affinities between
Mature Phase Harappans and Gandharan Grave Culture inhabitants of Timargarha, but
more distant affinities to those interred in Cemetery H at Harappa. Intriguingly, Cemetery
H individuals exhibit no affinities to Bronze Age inhabitants of either the Russo-Kazakh
steppe or southern Central Asia—the alleged homeland of purported “Indo-Aryan invaders”
into South Asia. Instead, Cemetery H individuals, and by extension other Indus Valley
inhabitants from Cemetery R37 at Harappa and Timargarha, show consistent affinities to
Late Bronze and Early Iron Age inhabitants of the Tarim Basin of Xinjiang, western China.
Results from neighbor-joining cluster analysis and principal coordinates analysis
of inter-sample variation in dental dimensions are also concordant and suggest a pattern of
long-standing biological continuity within the Indus Valley from aceramic Neolithic times
until the dawn of the Christian Era. However, Indus valley samples appear to have left little,
if any, genetic legacy among the living peoples of peninsular India, for the only affinities
between the peoples of these two regions is a prehistoric connection between mature phase
Harappans and the Late Jorwe occupants of Inamgaon in western Maharashtra. Living
inhabitants of the Hindu Kush highlands possess no affinities to prehistoric inhabitants of
the Indus Valley and in one instance, the Khowars of Chitral District, express close affinities
to the Bronze Age occupants of Oxus Civilization urban centers of southern Central Asia.
Results from neighbor-joining cluster analysis and principal coordinates analysis
of inter-sample variation in dental morphology trait frequencies yield somewhat discordant
results. The neighbor-joining tree identifies prehistoric Indus Valley samples as extremely
diverse, with Mature Phase Harappans exhibiting distant affinities to the Late Bronze/
Early Iron Age inhabitants of Sarai Khola and to living Dravidian-speaking inhabitants of
southeastern peninsular India. By contrast, the Neolithic inhabitants of Mehrgarh possess
affinities to living inhabitants of west-central peninsular India, while the Chalcolithic
inhabitants of this same site are identified as possessing affinities to living populations of
the Hindu Kush highlands. Principal coordinates analysis, on the other hand, identifies
greater biological homogeneity among prehistoric inhabitants of the Indus Valley, with
Mature Phase Harappans possessing far closer affinities to prehistoric samples that postdate
them (Timargarha, Sarai Khola), than to samples that antedate them (Neolithic, and
13th Harvard University Round Table on the ESCA, Kyoto Session, RIHN, 30 – 31 May 2009
21
Harappa: The Role of an Urbanized Bronze Age Populace in the Population History of South Asia (B.E. Hemphill)
especially, Chalcolithic Mehrgarh). In this latter case, it is the Chalcolithic inhabitants of
Mehrgarh, rather than Mature Phase Harappans, who are identified as possessing affinities to
Dravidian-speaking populations of peninsular India. Nevertheless, the Neolithic inhabitants
of Mehrgarh are once again identified as possessing affinities to living populations of westcentral
peninsular India.
Overall, the results of analyses of these three systems of biological variation
provide rather strong evidence that Indus Valley populations, including the Mature Phase
inhabitants of Harappa, appear to have played little role in the establishment of the living
populations of either peninsular India or even the Hindu Kush highlands. The ultimate
source of peninsular Indian populations remains unknown, but likely dates to the initial
dispersal of anatomically modern humans out of Africa during the Mid- Pleistocene. The
results of this analysis offer no support for any “Indo-Aryan” invasion during the midsecond
millennium B.C. Instead, what Central Asian biological impact may be found
among South Asian populations appears limited to the extreme northwestern periphery of
the subcontinent and likely reflect historic population movements that post-date the dawn
of the Christian Era.
22
Human Burial Customs during 3rd and 2nd Millennia BC in Haryana and Kuch:
An Analytical Approach
Vasant Shinde
Department of Archaeology, Deccan College,
Post-Graduate and Research Institute, Deemed University, India
The human burial tradition is very old in the Indian subcontinent dating back to the 7th
millennium BC. Though the burial custom started in the Mesolithic period in this part of
the world, it has remained steady from the Neolithic period and continued through the
Chalcolithic and Megalithic periods. The Neolithic and Chalcolithic people buried dead
bodies within the habitation, either beneath the living floor or in courtyards. However,
it was the Harappan culture which introduced the tradition of burying dead in separate
graveyards, slightly away from the habitation sites. It was one of the most important customs
of the Harappan people, who accorded lot of importance and significance to it as is evident
from many sites. The dead bodies were carefully buried in separate pits or clay coffins and
pots, ornaments and weapons were offered as burial goods. This no doubt indicates that
the Harappans believed in life after death. Though, archaeologists have not discovered
a cemetery in the proximity of every Harappan site, it should be safely presumed that it
existed but the remains have either not been located or they have been destroyed.
The Harappan burial custom was not uniform and one can see a lot of regional
variations. In the recently excavated Harappan cemetery at Farmana, there are three different
levels and three different customs. The site of Kalibangan, though located in the Ghaggar
Basin presents variations in the burial customs. The cosmopolitan city like Dholavira in
Kuch has numerous burial types indicating presence of a number of different groups within
the settlement who practiced different customs. Some of the burial types like Cist found at
Dholavira continued until the Megalithic times in many parts of the country. It is possible
to identify the social and economic status of families or communities on the basis of burial
customs and the quantity and quality of burial goods. There is neither uniformity in the
location of the cemeteries or in their customs which was thought to be uniform. Burial
is an important primary source of history as it contains not only human skeletal remains
23
Human Burial Customs during 3rd and 2nd Millennia BC in Haryana and Kuch: An Analytical Approach (V. Shinde)
but also numerous artifacts as burial goods, which remain in a good state of preservation.
The present paper shows similarities and differences between Mature Harappan and Late
Harappan burials customs in Haryana and Kutch and presents regional variations in the
burial customs.
Burials at Farmana (Photo by Akinori Uesugi, Indus Project, RIHN)
24
The Harappan Burials in Gujarat
P. Ajithprasad
Department of Archaeology and Ancient History, Faculty of Arts,
The Maharaja Sayajirao University of Baroda, India
Burials certainly were one of the most favoured modes of disposal of the dead among the
Harappans. It is also one of the easily discernable features in the archaeological record;
especially that dealing with rites, rituals and beliefs. One is not sure of, apart from different
types of burials, any other modes of disposal of the dead practiced by the Harappans.
Cremation of the body as a popular custom among the Harappan has been suggested by
several scholars, although no direct conclusive evidence for this has been reported from any
of the sites. Nevertheless, it is quite possible that cremation would have been practiced by
the Harappans.
Generally, the Harappan burials stand apart form contemporary Mesopotamian
and Egyptian burials in their simplicity and the matter of fact symbolism that one may
attach to them. The opulence and the architectural splendour of the pyramids or the royal
tombs of Mesopotamian rulers are almost completely absent in the Harappan burials. The
frugality of Harappan burials certainly has something to do with the customs and belief
systems the Harappans attached to death and the role of the dead in the society of the living
population, which would have been different from the contemporary Mesopotamian and
the Egyptian civilization. Whatever it may be, the apparent simplicity does not necessarily
suggest homogeneity or absence of any sort of variation. The fact that the Harappans
practiced two or three different types of burials in addition to the possible practice of
cremation bespeaks the inherent heterogeneity in their funerary practices and beliefs. The
difference and variation one comes across in the Harappan funerary practices are rather
subtle suggestions than loud statements about the society and the beliefs associated with the
disposal itself.
In view of these, burials are an important body of evidence that bear significant
25
information for reconstructing the Harappan society and its approach and attitude towards
fellow beings. Ideally speaking, most of the sites should have some burials associated with
them. Unfortunately, burials have actually been identified only from a few sites. This is
primarily due to the difficulty in locating burials, which normally does not leave behind
any clue for their identification. Besides, burial grounds are often segregated from the main
habitation area making the investigation quite speculative and based on a trial and error
method. No wonder many of the burials are actually discovered by sheer accident.
Burials are an important form of the funerary practices of the Harappans. They
are reported not only from sites in the Indus Valley proper in Pakistan but also from sites in
the Gagghar – Saraswati system in the east and sites in Kachchh and Gujarat in the south.
Generally the burial grounds or the cemeteries were segregated from the main habitation or
settlement, often choosing grounds some distance away from the habitation. Fore instance,
the Harappan burial R-37 is about 250m southwest of the main mound at Harappa. Recent
excavations at Fermana in Hariyana revealed a series of burials, probably the largest in
number in India almost a kilometre away from the site.
Gujarat has more than five hundred Harappan affiliated Chalcolithic sites,
primarily spread in the region of Kachchh, Saurashtra and north Gujarat. However, burials
are reported only from five or six sites: Dholavira and Surkotada in Kachchh, Nagwada,
Santhali and Loteashwar in north Gujarat and Lothal in Saurashtra (Fig.1). The cemetery
at Surkotada in Kachchh is about 300m northwest of the acropolis; so is the location of the
cemetery at Dholavira, the largest Harappan settlement in Gujarat. However, the burials
at Lothal are found closer to the settlement. On the other hand, in smaller and rural sites
such as Nagwada and Santhli the burials are found within the habitation area. It is therefore
apparent that the location of burials to some extend reflects the concerns of the city planners
in addressing the issues of space management and social concerns of the Harappans.
A fluxed burial reported from the Chalacolithic levels at Loteshwar probably
may be the earliest Chalcolithic burial in Gujarat. Although no Chalcolithic burial goods
have been reported with the skeleton, stratigraphically the burial appears to be part of the
Chalcolithic habitation that has been dated from 3600BC to 2900BC. The Harappan burials
The Harappan Burials in Gujarat (P. Ajithprasad)
26
Fig.1. The Harappan and other Chalcolithic burial sites in Gujarat.
from Lothal, Dholavira, Surkotada and the north Gujarat burials are certainly later in date.
They include both inhumation burials and symbolic pot burials. The latter ones do not have
any skeleton but a number of different pottery vessels deliberately buried in the burial pit.
A number of burials showing this feature have been reported from Nagwada, Surkotada and
Dholavira.
Dholavira in fact has yet another large tumuli-like hemispherical structure in
the cemetery area. This had a central square pit and spokes-like radiating walls. It also had
a series of rectangular chambers marked by stone slabs arranged in a radial fashion at the
periphery of the circular tumuli. They contained neither pottery nor skeletons. Some of the
pot-burials in Surkotada have either a stone slab or a small heap of rubbles demarcating the
burial pit.
The burials form Nagwada and Santhali in north Gujarat are particularly
interesting for the burial pottery. Of the five burials reported from Nagwada two are
inhumation burials and the remaining three are symbolic pot-burials (Fig. 2). These burials
13th Harvard University Round Table on the ESCA, Kyoto Session, RIHN, 30 – 31 May 2009
27
The Harappan Burials in Gujarat (P. Ajithprasad)
Fig.2. Pot-burial from Nagwada, North Gujarat.
have no burial goods except different pottery vessels whose number varied from just two to
twelve. Large storage jars of different size and shape, tulip shaped jars with convergent
rims, beakers, dish-on-stands with upturned straight rims, shallow bowls and medium size
pots with a constricted rim and squat profile etc. are the main vessels found in association
with the burials. These vessels are comparable in typo-technological features to the Early
Harappan pottery reported from Sindh and Baluchistan sites, especially Kot-Diji, Balakot,
Amri and Damb Sadaat etc. Moreover, the regular habitation levels at Nagwada do not
incorporate the pottery found in the burials. This certainly suggested that the pottery found
in the burials was earmarked for that purpose at the site.
Two more burials having similar pottery were excavated from Santhali, about
30km north of Nagwada (fig.3). The flimsy Chalcolithic habitation at this site had the same
kind of early Harappan pottery in the habitation level too. Yet another site that showed
substantial habitation deposit incorporating the Early Harappan pottery in the region is
Moti-Pipli. Nevertheless, there are no burials reported from this site. Besides, there are
about ten sites in north Gujarat that showed a similar early Harappan ceramic affiliation.
28
This is an indication of the fact that the food producing Chalcolithic communities from
Sindh and Baluchistan have already started moving towards south and had settled in
north Gujarat in the Early Harappan times. Burials from Surkotada present even on more
convincing picture of this movement as many of the burials in fact incorporated Early
Harappan pottery, although no Early Harappan pottery was reported from the regular
habitation layers. Probably, as at Nagwada, this pottery was an intentional choice as a burial
good and was produced for that purpose.
A close examination of most of the published burial pottery from Lothal in
Saurashtra shows that a large majority of them are of the Micaceous Red ware belonging
to the local tradition which is believed to original in the region. Some of the burials do
show the Harappan pottery and the Micaceous red ware together; probably suggesting the
level of integration that existed between the two traditions. Evidence from the burials is
therefore instructive in understanding the social configuration to a large extent. In so far as
the burials from Lothal, Nagwada and Surkotda are concerned, they not only illustrate the
heterogeneity of the Harappan society but also point towards its divergent roots.
Fig.3. The double burial from Santhali, North Gujarat.
13th Harvard University Round Table on the ESCA, Kyoto Session, RIHN, 30 – 31 May 2009
29
Four World Quarters in the late 3rd millennium BC:
Ur <> Shimashki <> Meluhha <> Magan (and the bits in between)
Daniel Potts
Department of Archaeology, The University of Sydney, Australia
The royal Akkadian rhetoric, according to which kings proclaimed themselves ‘king of
the four world quarters’, was first used by Naram-Sin, most probably after he had literally
campaigned to the north, south, east and west of Agade. As we learn more and more about
the world of the late 3rd millennium B.C., however, the notion of ‘four world quarters’
takes on new meaning. Excavations in Iran, Central Asia, the Indus Valley and the Persian
Gulf during the past few decades have begun to bring into focus what can be thought of as
a diamond-shaped region of interacting countries to the east of Mesopotamia. The rough
boundaries of this diamond may be drawn from Ur in the west, to Gonur Depe in the
north, to the Harappan sites of Gujarat in the east, and the Oman peninsula in the south,
and back up to Ur along a line running up the Persian Gulf. Naturally, this is a gross oversimplification,
and the world of the late 3rd millennium obviously extended westwards and
northwards as well, but there are abundant data now available attesting to the interactions
between the constituent parts of this ‘Eastern Diamond’. Ceramic, soft-stone, alabaster,
precious metal, base metal, glyptic and other indicators reflect exchanges between each
of these major poles, and between smaller regions within the broadly outlined diamond.
Clearly, traffic was not all one way, as we can see by the presence of both exports and imports
from each of these four areas in each of the other ones. Cuneiform texts, particularly of
the Ur III period, allow us to suggest names for some of the participating regions, moving
beyond the anonymity of the archaeological record. The identifications that can be proposed
allow us to consider many aspects of human behaviour - not merely political or commercial
- but social, ritual, and reciprocal as well, underlying the patterns of artifact distribution
that we can document archaeologically. Historical geography, sterile when pursued on its
own, becomes more compelling as we recognise just how many different types of materials
were moving between these different areas. It would be wrong, however, to assume a centreperiphery
model in seeking to understand the interactions between Sumer, Elam, Anshan,
Shimashki, Zabshali, Marhashi, Meluhha, Magan and Dilmun. The Eastern Diamond is
30
better conceived of as a mosaic than as a series of centres with concentric rings of influence.
This talk will illustrate the artifactual signatures of the interactions occurring within this
broad region, and make an attempt to place this material in the context of the historical
geography of the Eastern Diamond as imperfectly attested in Ur III cuneiform sources.
13th Harvard University Round Table on the ESCA, Kyoto Session, RIHN, 30 – 31 May 2009
Map by D.T. Potts 2003 "Anshan, Liyan, and Magan circa 2000 BCE." N.F. Miller, K. Abdi and
W.M. Summer (eds.) Yeki Bud, Yeki Nabud: Essays on the Archaeology of Iran in Honor of William M.
Sumner (Monographs Series (Cotsen Institute of Archaeology at Ucla), 48,) , pp.156-160.
31
The collapse of the Indus-script thesis, five years later:
Massive non-literate urban civilizations of ancient Eurasia
Steve Farmer*, Richard Sproat** and Michael Witzel***
The Cultural Modeling Research Group, Palo Alto, California, USA*
Center for Spoken Language Understanding,
Division of Biomedical Computer Science, Oregon Health and Science University, USA**
Department of Sanskrit and Indian Studies, Harvard University, USA***
Five years ago the three of us published “Collapse of the Indus-Script Thesis: The Myth of
a Literate Harappan Civilization” (reprint at http://www.safarmer.com/fsw2.pdf ). Our talk
today discusses developments in studies of the Indus symbol system in the half decade since
that paper was published and takes a quick look at the future. The talk is divided into four
parts.
1. The talk begins by discussing the often heated political and scholarly reactions
to our article, which has spawned a number of special colloquia and extensive if
distorted discussions in the press, over Internet, and in archaeological
conferences and academic studies. This part of the talk quickly reviews the best-
known attempts to defend the old script thesis, including claimed statistical data
introduced for that end in a recent paper in Science by Rao et al. It then
discusses new evidence that Harappan society was non-literate that has emerged
from analyses of the symbols over the past five years.
2. The paper continues by noting unexpectedly wide variation in symbol
frequencies that show up on ritual objects in different Indus regions and periods;
these data contradict older assumptions tied to the script model that picturm4 ed
the symbols as being largely uniform in use everywhere and “frozen” in time.
Discussion is raised of the light this evidence throws on apparent political
structures in Indus society and regional differences in agricultural rituals in
the various microecologies associated with different Indus regions (Weber, this
conference). Counterbalancing recent tendencies in the press to overemphasize
32
the Indus civilization as a third-millennium trading power, evidence is
underlined in the symbol system as a whole of the overwhelmingly agricultural
and predominantly local nature of the Indus economy.
3. The paper then expands discussion of a non-literate Indus society in light of a
wide range of continguous urban civilizations in Central Asia, SE Iran, and in
the Gulf (cf. D. Potts, this conference) — all regions that, despite occasional
claims otherwise, apparently remained non-literate from the third millennium
BCE well into the first millennium BCE. All these findings take on greater
significance in light of recent finds discussed in this conference by the Research
Institute for Humanity and Nature (RHIN) and their colleagues in Indus sites
distant from what has traditionally been viewed as Indus territories — making
the Indus the largest non-literate urban civilization of which we have evidence in
the new or old worlds.
4. We conclude by quickly listing popular myths about the Indus civilization
besides those involving the so-called script thesis that continue to distort Indus
studies; and take a quick look at the future by making a proposal, backed by
major private funding, of a collaborative project aimed at exploiting the massive
store of untapped data in Indus symbols to study the evolution of this unique
civilization in novel ways.
13th Harvard University Round Table on the ESCA, Kyoto Session, RIHN, 30 – 31 May 2009
33
The Asiatic wild ass in Harappan, Dravidian and Indo-Iranian record
Asko Parpola
University of Helsinki, Finland
This abstract summarizes my part of a longer paper written in collaboration with Juha
Janhunen (who deals with the Turkic, Mongolic and Tibetan terms), entitled "The Asiatic
wild asses (Equus hemionus & Equus kiang) and their vernacular names", to be published in
full in the Proceedings of this roundtable.
After an introduction on the taxonomy and geographical distribution of the
different ass species and subspecies, I discuss one grapheme of the Indus script (no. 46 in
the sign list of Parpola 1994: fig.5.1), proposing that it depicts the wild ass. The sign has
realistic (cf. fig.1 a & b) and schematic variants (fig.1 c). The wild ass is present in the
Harappan osteological record at least in Baluchistan, Sindh and Gujarat, but probably also
in the Punjab and Rajasthan. Moreover, there are terracotta figurines of the wild ass, but it is
not among the "heraldic" animals of the Indus seals, probably because the ass was already an
animal of ill omen: later on it was associated with Nirrti
'Destruction'.
The principal Harappan language, and apparently the only one in which the Indus
texts from South Asia were written, was Proto-Dravidian (cf. Parpola 1994). Attested in
13 Dravidian languages, representing all the subgroups except North Dravidian, is a word
for 'ass' (DEDR no. 1364). Bhadriraju Krishnamurti (2003: 12 and 525) reconstructs this
etymon for Proto-Dravidian as *kaz-
-ut-ay. Franklin Southworth (2005: 269-270) accepts
this recontruction, proposing that instead of the domestic ass, the word originally denoted
the wild ass, and that this animal was once present even in South India. This does not
seem impossible in view of the continuous belt of semi-arid thorn-desert and dry tropical
savannah from Kutch to Tamil Nadu, although there is little osteological support for this
hypothesis. The wild ass assumption is endorsed by a new etymology that I propose for the
word, as a Proto-Dravidian compound of *kaz-
- 'salt desert' (DEDR no. 1359 + Turner 1966
no. 2954) and *utay 'kick' (DEDR no. 616). Desert, especially salt desert, is the habitat of
the wild ass, and figures in the names of the onager in Sumerian (anše-eden-na) and Persian
(χar-e daštī). On the other hand, the ass is famous for its kicking, and represented as kicking
in the myth of the (wild) ass demon Dhenuka (cf. Harivamśa
57).
34
Sanskrit gardabha- 'ass' is very probably derived, with the animal name suffix
-bha- (of PIE origin but still productive in Indo-Aryan), from the Dravidian word for 'ass',
as proposed by Thomas Burrow and Murray Emeneau. One of the native terms for the
wild ass is khara-gardabha, not to be found in any Hindi, Urdu or Sanskrit dictionary, but
recorded as such in 1832 by James Tod and in 1834 by Alexander Burnes, and attested in
Matsya-Purāna
118,61 (turamgān
kharagardabhān).
Rāsabha- 'ass' is an archaic word, and the Rigveda uses it of the main animal of
the Aśvins. It pulls the chariot of the Aśvins and wins for them the prize-contest of Yama (RV
1,116,2). In another race, connected with the marriage of the solar maiden Sūryā, the asschariot
of the Aśvins surpasses the mule-chariot of Agni, the cow-chariot of Us4 as and even
the horse-chariot of Indra. The wild ass is the speediest of all equids, which is undoubtedly
reflected in this myth, even though the Aitareya-Brāhmana
(4,7-9) concludes by stating that
because the ass won, its speed was spent and it became the slowest of the beasts of burden.
Partly its connection with the Aśvins is due to their funeral function (cf. Parpola 2005: 29-
36; and in addition Rāmāyana
2,63,14-16). According to Herodotus (7,86), the chariots
of the Indians serving in the Persian army of Xerxes in 480 BCE were drawn by horses and
by wild asses (ónoi ágrioi). In the apadāna stairway of Persepolis two men from the province
of Hinduš, i.e. Sindh, bring a wild ass to the great king. In the Kharaputta-Jātaka, mules
(kharaputta) from Sindh pull the chariot of the king (rathe yuttasindhavā).
Khara- 'ass, donkey' is not in the Rigveda, but in the Paippalāda-Samhitā (20,39,2);
in Hindi it is khar. Ultimately probably a loanword from Semitic, cognates are found in
most Iranian languages, including Avestan χara-, Ossetic χæræg, Middle & New Persian and
Baluchi χar; from Baluchi χar has come to Dravidian Brahui as well. The zoological term
khur (1827) is, with anglicized spelling, Hindi khar or Persian χar.
In Middle and New Persian, the wild ass is usually called gōr or gōr-χar. James
Tod in 1832 mentioned gorkhur as a local name of the wild ass of Kutch, and gora-khara is
known from the Jaina text Pannavanāsutta
(1) written in Ardhamāgadhī. In the 9th century
Buddhist Hybrid Sanskrit glossary Mahāvyutpatti (no. 4797) gaura-khara is equated with
Tibetan rgyan4 'Tibetan wild ass'; this native Tibetan word is behind the zoological term kiang
for Equus kiang. Sanskrit gaurakhara in Matsya-Purāna
118,58 seems to refer to the kiang,
since it belongs to a list of wild animals near the source of River Ravi. The oldest Sanskrit
reference is in Vasist 4 4ha-Dharmasūtra 21,1-3, where krsn 4 4 4 a-khara, gaura-khara and śveta-khara
13th Harvard University Round Table on the ESCA, Kyoto Session, RIHN, 30 – 31 May 2009
35
The Asiatic wild ass in Harappan, Dravidian and Indo-Iranian record (A. Parpola)
are alternative mounts of a Brahmin woman involved in adultery; for purification she has to
ride the ass naked. Many authors since 1835 speak of ghor-khar; the aspiration seems to be
due to a folk-etymological transformation into 'horse-ass', although ghor-
-khar is not found
in Hindi dictionaries. The Yaśastilakacampū of the Jaina author Somadevasūri, written in
Ujjain in 951, records for 'wild ass' gaura-khura and khara-khura, ending in Sanskrit khura
'hoof' (possibly a loanword from Dravidian).
Gaura- is primarily an adjective, 'white, yellowish, pale red, fair-skinned',
etymologically probably 'cow-coloured', from gav- 'cow', with unusual long grade ablaut.
In Iranian, the only cognate in this meaning is Baluchi gōraγ (cf. *gaura-ka- > Hindi gōrā-).
Persian gōr 'wild ass' corresponds to Sanskrit gaura- m., which in Vedic texts does not denote
'the gaur bison (Bos gaurus)', as all dictionaries and translations have it, but 'wild ass'. In
Middle Vedic texts, gaura- or gaura-mrga-
is the wild counterpart of the horse (cf. e.g. VS
13,41-51 and ŚB 7,5,2,14-36), and in the Rigveda, gaura- is an animal that rushes down
to a water depression of saline soil to slake its great thirst (RV 8,4,3 yáthā gauró apā krtám
trs
4yann éty áverínam).
Fig.1a. Sign 46 in M-1097 Fig.1b. Sign 46 in M-290 A Fig.1c. Sign 46 in M-516 A bis
36
Genetic diversity of Afghan wheat landraces
and their potential for future breeding
Tsuneo Sasanuma
Faculty of Agriculture, Yamagata University, Japan
Landraces, the crop varieties cultivated for a long time in a local area, are considered to be
an important genetic resource in crop breeding because they possess the high level of genetic
diversity and the adaptability with the local environments. Numerous numbers of studies
have been conducted to reveal the genetic diversity in landraces of various crops. In some
cases, the landraces have practically contributed to modern breeding as a genetic resource.
One of the most famous contributions of the landraces to wheat breeding is the “Green
Revolution”, in which a Japanese landrace “Daruma” was used as a donor of the semi-dwarf
trait.
In our present research project, we focus the Afghan wheat landraces. Afghanistan
is a country located in the geographical and historical crossroad of several civilizations,
Arabic, Persian, Indian, and Chinese, so that it is expected that genetic variation should
accumulate in the country via human interactions among the civilizations. Furthermore,
Afghanistan is close to the place of the origin of wheat, thus it might be a center of diversity
of cultivated wheat. In spite of such importance, the wheat landraces in Afghanistan stand
on the edge of a precipite of genetic erosion caused by a social confusion and serious natural
disasters.
The Japanese genebank, the Plant Germ-plasm Institute in Kyoto University,
have maintained about 450 accessions of Afghan wheat landraces. These materials were
collected by three different expeditions in different periods, that is, the Kyoto University
Scientific Expedition in 1955, the British Scientist Thomas’ expedition in 1965, and the
Kyoto University Scientific Expedition to Southwestern Eurasia in 1978. Using these
Afghan wheat collections, we investigated the diversity of the morphological traits and the
component of high molecular weight (HMW) glutenin subunit that is a storage protein
of wheat grain essentially affecting the bread making quality. The morphological analysis
showed that they have primitive forms, that is, a tall plant height and a long and rough
spike. This result indicates that these collections refrect typical landraces, in other words,
37
they were not polluted by modern breeding, although the most recent materials were
collected about a decade later than the beginning of the wheat green revolution. The
analysis on the component of HMW glutenin subunit demonstrated that the majority of
the Afghan wheat landraces have subunits with a low bread making quality, which is the
typical component of the eastern Asian wheat landraces. As for the level of diversity, it
was unexpectedly revealed that the Afghan wheat contains a lower level of diversity than
the wheat landraces of the neighboring countries, Iran and Pakistan.
These genetic characteristics of the Afghan wheat collection seems unattractive
from the viewpoint of breeding. However, we have found a new type of the HMW
glutenin subunit in an accession of Afghan wheat. The micro-sedimentation test suggested
that this novel subunit has a unique and good bread making quality. This finding can
be strong evidence that the Afghan wheat collection has a potential as a genetic resource
for wheat breeding. Above all, we would like to emphasize that these landraces have
been cultivated in the local areas of Afghanistan for a long time and are adapted to
their individual environment. Since Afghanistan has a complex topography and various
climates within each of its counties, each of the wheat varieties should have adapted to
the regional climate. The fact that the Afghan wheat collection contains original endemic
genetic features suggests that they should have environmental adaptability in Afghanistan.
Throughout this research project, we have set as our final goal that we return the Afghan
wheat landraces maintained in a Japanese genebank to their home country. We try to
make them contribute towards reconstructing the sustainable agriculture destroyed by the
continuing wars and social upheaval in Afghanistan.
Genetic diversity of Afghan wheat landraces and their potential for future breeding (T. Sasanuma)
38
Traditional Management of Agrobiodiversity of Rukai aboriginal peoples in Taiwan
Hsin-Fu Yen
National Museum of Natural Science, Taiwan
There are 14 tribes of aboriginal peoples in Taiwan. The Rukai tribe lives in southern Taiwan
in the region of about 1000 meters altitude. They have many kinds of food crops which are
cultivated in the fields, for example lettuce (Lactuca sativa), sugar cane (Saccharum sinensis),
musky winter squash (Cucurbita moschata), chayote (Sechium edule), peanut (Arachis
hypogea), pigeon pea (Cajanus cajan), hyacinth bean (Lablab purpureus), cowpea (Vigna
unguiculata), Taiwan goosefoot (Chenopodium formosanum), Job's tears (Coix lacrymajobi),
foxtail millet (Setaria italica), Eccoilopus formosanus, sorghum (Sorghum bicolor),
corn (Zea mays), taro (Colocasia esculenta), tannia (Xanthosoma sagittifolium), sweet potato
(Ipomoea batatas), cassava (Manihot esculenta), yam (Dioscorea alata) and ginger (Zingiber
officinale), etc. Some food crops have a few cultivars, especially foxtail millet and taro.
Foxtail millet and taro are more important food crops for the Rukai tribe.
There are little cultivars of foxtail millet in other aboriginal tribes of Taiwan except for the
Rukai tribe. People of Rukai in Wutai have over 15 cultivars of foxtail millet, including
‘ababake’, ‘cipaerane’, ‘cipaerane ka ladwadwane’, ‘cipaerane ka sinamecyane’, ‘darangidangi’,
‘darangidangi ka ciparepare’, ‘darangidangi ka paedeane’, ‘eape ki talralreba’, ‘lrulrubungu’,
‘tirukale’, ‘makasasagarane’, ‘paedeane’, ‘palralraamu’, ‘saidhipane’ and ‘salalaai’. Incidentally,
they have over 11 cultivars of taro, including ‘gadaiyano’, ‘gulailaily’, ‘galalologo’, ‘gololalu’,
‘kalailaily’, ‘lilobo’, ‘logo’, ‘magapagalogalo’, ‘olisiliyuo’, ‘zago’ and ‘zasiliyo’.
There are two cultivars ‘davaliquan’ and ‘gacumusano’ in sugar cane, ‘zabyabyak’
and ‘dilololo’ in musky winter squash, ‘lalabon’ and ‘icuculo’ in hyacinth bean, ‘lamulam’
and ‘dalankili’ in ginger. Although there were many kinds of cultivars in sweet potato,
people forget them. The tannia is not a traditional food crop. The Japanese introduced it
from Southeast Asian a long time ago.
People sow the seeds of different foxtail millet cultivars in one field at the same
39
time, but harvest them by the character of cultivars at different times. The taro crops are
operated in the field following the same model. On the other hand, you can see many kinds
of crops cultivated in the same field at different seasons. For examples, taro, sweet potato
and foxtail millet are alternate cultivates at the same field.
People plant sweet potato in monoculture during June. They harvest the sweet
potato one by one six months after planting. They do not harvest sweet potato together at
the same time.
People have many experiments in taro cultivation. Typhoon damage has a relation
with yield and quality. If typhoons affect many times, the production and quality of taro
will go down. Although the production is increased by fertilizer, it will make the quality go
down, and pest, rot ratio will rise. People cultivated taro in upland fields, but some cultivars
will be suitable in paddy fields.
The Rukai aboriginal people have lived in the mountain area for a long time.
They still maintain many crops cultivars and biodiversity in agriculture. Their traditional
knowledge of agriculture will benefit modern agriculture.
Traditional Management of Agrobiodiversity on Rukai aboriginal peoples in Taiwan (H.-F. Yen)
40
A hot spot of linguistic diversity in the Greater Hindukush/Pamir area:
The names of agricultural plants
Michael Witzel
Department of Sanskrit & Indian Studies, Harvard University, USA / RHIN, Japan
This investigation is based both on our earliest texts (the Vedic Sanskrit (Rgveda), the Old
Iranian Avesta), as well as on the descendant languages of Old Iranian, Old Indo-Aryan
and Nuristani, such as modern Persian, Pashto, Hindi, etc. A study of their names for
domesticated plants indicates that the Rg-Veda contains just a few words that can be traced
back to Indo-European, but most of the others are of local origin. This is not surprising for a
mainly pastoral people such as the Indo-Iranians and Vedic Indo-Aryans.
Local (substrate) words can be isolated by linguistic means, including unusual
sounds and word structure, as well as the lack of a convincing Indo-European etymology.
This evidence can be counterchecked and expanded by their attestation in medieval (if any)
and in modern languages.
It then appears that the greater Hindukush/Pamir area was and is a hotspot of
linguistic diversity, which is also reflected by the names of domesticated plants of the area.
Beginning with the oldest texts, there is but a small number of Indo-European
terms, that are rapidly diminishing as we move further east from the home of the pre-Indo-
Aryans and Iranians, first in northern and then in southern Central Asia.
Some residue Indo-European and Indo-Iranian terms are still to be found in the
Hindukush-Pamir area in the local Iranian, Nuristani or the Indo-Aryan Dardic languages,
but they increasingly diminish in number, and finally disappear in the other Indo-Aryan
languages of the subcontinent.
Instead, the use of local plant names in Indo-Aryan languages is steadily increasing
when moving further into the subcontinent. They stem from the unknown prefixing
41
A hot spot of linguistic diversity in the Greater Hindukush/Pamir area: The names of agricultural plants (M. Witzel)
Indus language(s), from an equally unknown, generally North Indian substrate language
(“Language X,” as reflected in Hindi, etc.), as well as, later on, also from Dravidian (not
present in the Panjab until well after the post-Harappan period), and from Munda.
This result can now be correlated with the archaeobotanical study of plants as
carried out by Dorian Fuller (see detailed abstract, this conference).
Tharu
Indus
Language X
Brahui
Meluhhan
Proto-
Bhili Korku
Nahali
Malto
Mundari
Santali
Kharia
etc.
Sora
etc.
Proto-
Nilgiri
Vedda
Khasi
Kusunda
Burushaski
Nicobarese
Shompen
Andamanese
FORMER AUSTRO-ASIATIC AREAS?
FORMER DRAVIDIAN AREAS
T I B E T O - B U R M E S E
DRAVIDIAN
Isolated languages
Inferred languages
Other Austro-Asiatic
Munda languages
Dravidian
Tibeto-Burmese
After F. Southworth, 2005
Kurukh
42
Map of Indus civilization sites (Map by Hirofumi Teramura, Indus Project, RIHN)
Front and back cover: Faience tablet or standard from Harappa, Pakistan (photo by Harappa.com)
Courtesy of J.M. Kenoyer and R. Meadow
Edited by Hitoshi Endo (Indus Project, RIHN)
Published by Research Institute for Humanity and Nature (RIHN)
Small pot painted with a pair bulls, from Farmana, India
Drawn by Akinori Uesugi, Indus Project, RIHN
13th Harvard University Round Table
ETHNOGENESIS OF SOUTH AND CENTRAL ASIA (ESCA)
Kyoto Session
30 – 31 May 2009
at
Research Institute for Humanity and Nature (RIHN)
Kyoto, Japan