Austroasiatic Languages and Rice Cultivation

8/3/2019 Austroasiatic Languages and Rice Cultivation

http://slidepdf.com/reader/full/austroasiatic-languages-and-rice-cultivation 1/10

Chapter 18

Languages and Farming Dispersals:

Austroasiatic Languages and Rice Cultivation

Charles Higham

There were two major transitions to agriculture in

the Old World. One took place in the Levant and

involved wheat, barley, cattle and sheep. The otherwas centred in the Yangtze and Yellow River basins

of China, where rice and millet were brought under

cultivation in association with cattle and pig domes

tication. Both took place at about the same time and

under parallel climatic changes. In the western cen

tre, much research has been devoted to exploring

possible links between the expansion of agricultural

communities from the Near East and the present

distribution of Indo-European languages. Archaeo

genetic research has been deployed as a testing

mechanism for the broad models generated. East

and Southeast Asia lag well behind this move, but

the region is important not only on its own terms,

but also as a means of seeking possible similarities

with the spread of Indo-European languages.

This paper identifies first a series of cognates

for rice cultivation which link the Austroasiatic lan

guages of Southeast Asia and eastern India. It then

seeks archaeological evidence for the expansion of

rice farmers south and west from the centre of do

mestication in the Yangtze Valley, and finds an en

couraging conformity between the distribution of

Austroasiatic (AA) languages and the spread ofNeolithic settlement based on rice, and the raising of

domestic cattle, pigs and the dog. I t then considers

the possible adoption of Austroasiatic languages by

indigenous hunter-gatherer s. The concluding model

is proposed and means of testing it are explored.

AA languages fall into two major divisions,

Munda and Mon-Khmer, and are found from east

ern India to Vietnam, south to peninsular Malaysia

and the Nicobar Islands. The Kurku are the western

most group of AA speakers, living south of the

Narmada River in Maharashtra. Norman & Mei

(1976) have identified a possible AA substrate in

223

southern China which suggests that this language

family once had an even wider distribution. The

most northerly known AA language is P'u-man, recognized in 1899 in the village of Xiao Qin in Yunnan.

This is a particularly vital location, for it lies on the

strategic Mekong about 100 km south of lake Dali.

Apart from Vietnamese and Khmer, the national lan

guages of Vietnam and Cambodia, the distribution

of AA speakers consistently takes the form of iso

lated enclaves. This is, at least in part, due to more

recent, historically-documented intrusions. The Thai,

for example, have taken up much of the Chao Phraya

Valley, thus isolating the speakers of Mon (an AA

language) to remote, usually upland enclaves. The

Kuay people of the lower Thai provinces of the

Khorat Plateau are islands surrounded by speakers

of Lao. The Burmese have marginalized the Mon,

while Munda languages persist as enclaves sur

rounded by Indo-European languages. No AA speak

ers survive in Lingnan (southern China) in the face

of the expansion of Sino-Tibetan.

AA languages have, for almost a century, been

linked in various ways with other language families.

Schmidt (1906) was foremost in suggesting that AA

and Austronesian (AN) languages belong to a phy

lum he named Austric. This linkage was not widelysupported until Reid (1994) found evidence in the

Nancowry language of the Nicobar Islands for a link

based not so much on cognates, but on morphemes

in which conservative AN structures survived in AA

languages due probably to the remote island loca

tion. The notion that the Munda languages were

intrusive to India was suggested by Heine-Geldern

(1932), who further linked their arrival from South

east Asia with the distribution of the polished shoul

dered adze and the spread of agriculture. Wheeler

(1959) joined him in identifying an eastern source for

the Neolithic of eastern India.



o

• Palaungic

• Aslian NorthAslian

The word for dog, for

example, is likely to be im

portant. There is, in Southeast

Asia, no native wolf from

which to derive the domesti

cated dog. Yet the cranial

characteristics of the prehis

toric dog reveal, beyond

doubt, a lupine ancestry. The

nearest possible sources for

the wolf are Canis lUpus chanco

in China, and c.l. pallipes in

India. Figure 18.1 shows the

word for dog in a variety of

AA languages. I t is clear that

cognates are present over the

entire area of AA language

distribution, even into Central

India. The word for child (Fig.

18.2) is virtually identical be

tween Kurku in Central In

dia, and Bahnar on the eastern

seaboard of Vietnam, a dis

tance of almost 3000 km,

equivalent to that from the

Konya Plain in Turkey to

Skara Brae in Scotland. Fish

• Central Bahnaric m west Bahnaric is another key word for anyexpansionary group of farm

ers in Southeast and South

Asia. As can be seen in Figure

18.3, this too is clearly cog

nate across the area of AA lan

guages, linking small islands

of speakers.

But perhaps th e key

words in the vocabulary are those for rice in its

various forms. Luce (1985) has considered the word

for husked rice. In Old Mon it is sno', Old Khmerranko, Danaw ko, in the P'u-man language of Yunnan

it is 'n-k'u and in Khasi, it is khaw. The word for rice

plant (Fig. 18.4) in Sakai is ba'ba' or ba', in Stieng,

Biat, Gar and Bahnar it is ba, in Khasi is becomes kba

an d in Mundari, it is baba. Luce concluded with

these words: 'What can be the cause of this startling

diffusion? I can only think of one adequate explana

tion: wet rice cultivation' (Luce 1985, 3). At a time

when archaeological research had hardly begun, he

suggested that rice cultivation began in the Red River

Valley, whence agriculturalists moved up stream toYunnan, across to the headwaters of the Brahmaputra

and so into India. As will be seen, his first idea has

been largely sustained by further linguistic research,

ffiilij" .v......... Khmulc............'

'",

""

f i ~ ~ ~ ~ ~ : : ~ ~ ~ ' ......... , .

.......... . .;,.' " ...

...

.'

Munda• Monic

• Nlcobarese

• North Bannaric

"

o

KSEW

600

-

Pearic

[ill Khasian

D Soutn Bahnaric

!W!WiH Katuic

kilometres

200 400

- -

KSIA ~ ~

Austroasiatic words for dog

x Pu'man

KINSOR '?:>~

South Ashan

\:::::}::1 Khmaric

Chapter 18

Figure 18.1. Austroasiatic words for dog.

'-----'-----

224

Any consideration of this possible link between

AA languages and the spread of agriculture should

most logically commence by considering cognatewords for rice cultivation across the broad spectrum

of AA languages. As with all aspects of the South

east Asian past, such studies lag behind comparable

research in other parts of the world. However, in a

series of lectures delivered to the Ecole des Langues

Orientales Vivantes in Paris in 1966, Gordon Luce

provided a pioneering analysis of the implications

of the distribution of AA languages for the spread of

rice cultivation (Luce 1985). He began by considering a

number of key cognates linking the widely-scattered

speakers of AA. These form such a key platform forany further consideration of this issue that they need

to be briefly summarized. In doing so, I have chosen

certain words relevant to the spread of agriculturalists.

I I Vletic



- -


bu t his archaeological corre

lates need drastic revision.

Thus, Zide & Zide (1976)

have considered the Proto

Munda vocabulary, and com

pared the reconstructed wordswith those found in other AA

languages in Southeast Asia.

Their results reveal that, on the

basis of the reconstructed

Proto-Munda word list, th e

Munda were more advanced

agriculturally than archaeolo

gists had previously thought.

Whereas it was widely

assumed that the more ad

vanced Munda, speaking Sora,Mundari or Santali, received

their knowledge of agriculture

from intrusive Indo-Aryan

speakers, the linguistic evi

dence revealed that they

would have been rice farmers

at the time of their arrival in

eastern India. Indeed, the re

construction of plant names

provides a dimension to

Munda prehistory not available so far from archaeology.

Bamboo and bamboo shoot

have cognates between Sora

and Gorum in Munda, and in

Old Mon.

There are Proto-Munda

names for rice and uncooked

husked rice which have cog

;;....;:;

KHUN

. .J;'i{~ . ; : : " . ~ W N HON%. '

- -K)N "" •

" : . : : : ~ • \.. ..KON

· , ; , ~ , , · , ~ ,........................

ON '$:-.gO

... ..; .' .. : : - ~ ~ . .>, ....

........ '.' .

' ~ ~ ~ : 2 ~ ' ~ ~ ... .•...

KON (Kurku) I,

. D o

'".KUSN,

oAustroasiatic words for child

200 400 600 "

kilometres [ ~ ~ ~ ~ ~ m ~ ~ ~ 1 Katuic • North Bahnaric • Central Bahnaric .1m west Bahnaric0 Vietic

1::::::\1 Khmaric ,," . Pearic ............... Khmulc • Palaungic...............illMonic

Pu'man f:--::·:-::::·:-:..::·1 Khasian North Aslian • Aslian §§§§§ Munda

South Aslian t:::::I South Bahnanc• Nicobarese

Figure 18.2. Austroasiatic words for child.

nates in Mon-Khmer, Lawa, Rumai and Khmu. Lawa

is spoken in the Ping River valley of northern Thai

land, while Khmu speakers are found in upland Laos.The north Munda form has cognates in Kharia, Mon

Khmer, Khasi and Semang. The word for pestle might

be cognate in Kurku and Mon, Khmer and Proto

MK, while alcohol and inebriation have widespread

AA cognates. There is also a reconstructable word

for do g with cognate forms in Mon-Khmer. The

Munda word for bull seems to have been borrowed

from Indo-Aryan, whereas there is a possible cog

nate for cow with Proto-Munda and Mon-Khmer.

Zide & Zide have concluded that at least 3500 years

ago, at a conservative estimate, the Proto-Munda

speakers practised subsistence agriculture, cultivat

ing rice, millet and at least three legumes. They also

used husking pestles and mortars which go back to

Proto-AA. But they developed some cultigens or

plant resources in India, for there are no AA cog

nates for mango or turmeric.This situation is supported by Mahdi (1998),

who has found that the Proto-AA word for rice can

be reconstructed in Munda, Mon-Khmer, Palaung

Wa, Viet-Muong, Old Mon and Lamet. Pejros &

Shnirelman (1998) have also deployed linguistic evi

dence in suggesting that neither the Austroasiatic

nor Austronesian proto-languages reveal evidence

for a tropical origin, but rather point to inland begin

nings north of the tropical zone of eastern Eurasia.

They identify the middle Yangtze Valley as a likely

homeland, and feel that Proto-Austric began to di

vide in the ninth to eighth millennia Be. Within AA,

Munda and Mon-Khmer split from each other by the

end of the fifth millennium Be. By the end of the

225



o

m Aslian

• Palaungic

NorthAslian

tage of riverine routes of ex

pansion. From a source in the

upper Yangtze valley, he sug

gested that Proto-Munda

speakers followed the course

of the Brahmaputra River into

India, while speakers of Proto

Mon-Khmer followed the

Irrawaddy into Burma, the

Chao Phraya an d Mekong into

Thailand and Cambodia, an d

the Red River into Vietnam.

This proposal called upon pre

historians to review the ar

chaeological data available, to

see i f the evidence supported

such a model. This involves

consideration of the climatean d archaeological sequence

in the Yangtze Valley, based

on evidence only assembled

over the last few years.

The early Holocene cli

mate of the Yangtze Valley un

derwent a series of profound

changes incorporating the end

of the Pleistocene Ice Age an d

the oscillations in temperature

•

Central Bahnaric west Bahnaric

and rainfall. Recent evaluations of pollen spectra and

faunal assemblages there re

veal a climate 4-10°C cooler

and much drier than now be

tween 20,000 and 15,000 Be

(Higham & Lu 1998). Rainfall

was probably 1000 mm perannum below its present level of 1600 mm, account

ing for the predominance of drought-resistant plants

in the pollen spectra. From 15,000 to about 13,700

years BP the climate moderated, encouraging thespread of oak and pine, elm and willow. But thereaf

ter, and until 10,000 BP, there was a reversal to cold

conditions described across Eurasia as the Younger

Dryas period. Thereafter, it again became warmer

and moister. Broad-leaved trees colonized the Yang

tze Valley and the fauna became subtropical. In

creased rainfall fed rivers and lakes, and wild rice

spread out from refugia.

It is against this environmental kaleidoscope

that we can measure the significance of recent finds

from deep excavations in the caves which fringe the

lacustrine lowlands. The sequence at Diaotonghuan,

for example, spans the later Pleistocene into the early

• Monic Munda

• North Bahnaric

• Nicobarese

o

KHA .:;

600-t·:-::::·:-::::·:->1 Khasian

Pearic

I::::: :1 South Bahnaric

! ~ ~ m H ~ ~ ~ ~ 1 Kaluic

kilometres

200 400- -ustroasiatic words for fish

:It Pu'man

KAKU (Kurku)

South Aslian

Figure 18.3. Austroasiatic words for fish.

226

Chapter 18

DVielic

fourth millennium, Mon-Khmer began to divide into

Khmer, Bahnaric and Viet-Muong.

The linguistic evidence summarized above is

compatible with an original Austric homeland in themiddle Yangtze Valley, from which at least the an

cestors of the AA and AN languages originated and

spread, the former largely by land, an d the latter by

sea. Linguists seem to agree that a considerable time

depth is necessary to account for the differences be

tween the Munda and Mon-Khmer languages, and

rather less for the divergence between the individual

languages of the latter division.

Robert Blust (1996) followed Reid's conclusion

on the validity of Austric by proposing, purely on

linguistic evidence, that the distribution of AA lan

guages in South and Southeast Asia results from a

series of intrusive movements which took advan

1:::::::;:;::1 Khmaric



227

o

B west Bahnaric

Aslian

• Palaungic

NorthAslian

• Central Bahnaric

['''''1 Kh .............. mUle...............

I,

~ m ~ ~ : : L ! ' . '."" " .

.' ; ~ u ~ r F '•...

.'..

tI • \ . •.

•

Munda

• Monic

• North Bahnaric

• Nicobarese

,.,

KBA

o

marshes an d lakes, were also abundantly represented

in this settlement, together with hunted and prob

ably domestic animals, pottery vessels, wooden

spades and pestles, the foundations of pile dwellings and over 100 human burials. Bashidang is simi

la r in many respects to the settlement of Pengtoushan,

found only 20 km to the southwest. Here, we en

counter a cemetery in which the dead were interred

with complete pottery vessels and exotic stone orna

ments. The clay used for making pots was tempered

with rice chaff. Again, there are the remains of houses

and every sign of a successful adaptation to the rich

resources offered in the middle Yangtze Lakeland.

Two radiocarbon determinations obtained from the

rice used as a tempering agent are 6420-6990 BC and5780-6380 BC (calibrated: Crawford & Chen Shen

1998).

Katuic

60 0-

~

Pea ric

..»::··>::-I Khasian

):::::j South Bahnaric

kilometres

20 0 400

- -

' $:

.' ~ B A B A ~ ' "/

"'.

Austroasiatic words for rice plant

x Pu'man

!::::::;:;:j Khmaric

Figure 18.4. Austroasiatic words for rice plant.

period. This cavern

a small, swampy

in Jiangxi Province

The excavators

16 sequential

of occupation and resamples of rice phyto

the hard silica bodies

the rice plant. There

example, a surge in

e numbers of rice glume

in zone G, which

tentatively dated to the ter

Pleistocene. These are

evidence for the col

of wild rice during the

phase which character

that period. Rice

were extremely

during zone F, which cor

to the Younger

phase. However,

was again abundantly

during zone E,

is thought to date be

ween 10,000-8000 BP. About

th e sample conforms

a domestic variety of D Vietic

This context also pro

first evidence for

the form of very

sometimes cord

vessels which could SouthAslian

have been made in or

r to cook rice.

A lack of reliable radio

dates makes this a tentative framework,

ut it gains support from similar sequences in other

caves. Xianrendong is located only 800 m from

iaotonghuan, an d again has a Palaeolithic occupation under a Neolithic horizon containing rice

phytoliths. Yuchanyan also overlooks low-lying

etlands, an d has provided a sample of fish, turtle

an d mammalian bone as well as rice husks said to be

transitional to the domestic form. Potsherds from

this site are dated in the vicinity of 12,500 BP (Yuan &

Zhang 1999). Bashidang is a village site which cov

ers about three hectares. Its lower layers date to

about 8000 BP, and excavations in 1993-97 uncovered

waterlogged deposits which had preserved over

15,000 rice grains. These have been ascribed to acultivated variety (Pei 1998). Water caltrop and lo

tus, both of which can easily be propagated in




Chapter 18

This accumulating body of evidence indicates ing distance of all three. It has a deep stratigraphic

that the Yangtze Valley was one of the very few sequence, involving over four metres of accumu

areas in Eurasia that witnessed a Neolithic Revolu lated cultural material. The initial settlement has been

tion, the transition from hunting and gathering to dated to between 2400-2100 BC, and excavations over

agriculture. Population growth is a recurrent charac an area of 225 square metres have revealed the re

teristic of sedentary agricultural communities.As

mains ofeleven

houses and acemetery.

Many of thesettlements grow, there is a strong incentive for a human remains were found with no cranium, an d

segment to move and found a new community. This grave goods were also absent, but the pottery from

appears to have followed the establishment of such this phase was decorated with a distinctive series of

sites as Pengtoushan and Bashidang. Fenshanbao, patterns, incorporating parallel incised lines infilled

which was occupied within the period 8000-7500 BP, with impressions (YPM 1981). The nearby site of

lies east of Lake Dongting, and excavations have Dadunzi is rather later, the single radiocarbon date

revealed 50 burials and pottery tempered with rice. suggesting a mid second-millennium BC occupation.

To the west, we find agriculture spreading upstream Again, house plans were noted, often superimposed

to Chengbeixi in the Three Gorges. In an easterly over earlier structures, and 27 burials were encoun

direction, the famous site of Hemudu in Zhejiang tered. Adults were buried in extended positions with

Province was a base for lakeside rice cultivation by no preferred orientation, and infants were interred

7000 BP. in mortuary jars. The style of pottery decoration

This sequence ha s a strong bearing on the matched that found earlier at Baiyangcun.

Neolithic settlement of Southeast Asia, because it is Archaeological research in the major river val

now possible to trace the expansion of agricultural leys of Southeast Asia has revealed a compelling

communities progressively further to the south. Sev pattern in which ne w agricultural villages were es

eral rivers provide access from the Yangtze Valley to tablished between 2500-2000 Be. In the Red River

the rich hot lowlands of Lingnan. The Gan and Xiang valley, this phase is seen in many sites of the Phung

flow north to Lakes Poyang and Dongting, while the Nguyen culture. In the Mekong catchment, we find

Bei flows south. The first evidence we have for the Neolithic phases of occupation at Ban Chiang, No n

establishment of rice farmers is, no t unexpectedly, in Kao Noi, Ban Non Wat and Ban Lum Khao. In the

the headwaters of this last river, where the sites valley of the Chao Phraya River, Ban Kao, Non Pa

Shixia, Xincun, Chuangbanling and Niling date from Wai and Ban Tha Kae indicate settlement towards

the early third millennium Be. Shixia in its earliest the en d of the third millennium Be. A common inhu

phase included a cemetery in which grave goods in mation burial ritual, the bones of domestic pigs, cat

cluded jade cang (tubes) of a type known to have been tle and dogs, and a similar technique of decorating

of deep ritual significance in the Liangzhu culture to pottery vessels link these sites. In eastern India, rice

the north, as well as bracelets, pendants and split rings. remains and rice-tempered po ttery have been found

The subsequent Nianyuzhuan culture sites reflect a at Chirand, dated probably to the third millennium BC,

further spread of agricultural settlement, but began while Allchin & Allchin (1982) have described sites

to encounter and interact with rich hunter-gatherer further east, such as Sarutaru and Daojali Hading,

groups commanding the delta of the Zhu River. which contain cord-marked pottery recalling wares

The Bei is just one of the rivers which ultimately from Southeast Asia and southern China. There is,

connects the Yangtze Valley with Southeast Asia. In therefore, a consistent horizon of third-millennium

general, these rivers flow south and radiate out from BC settlement sites incorporating evidence for rice

a hub in the eastern Himalayan foothills. From east cultivation, from southern China to Eastern India. It

to west, they include the Red, Mekong and Chao is difficult not to see this pattern as being similar to

Phraya systems. Further to the west, this configura the expansion of the Linearbandkeramik sites of the

tion is repeated in the form of th e Irrawaddy, European loess lands.

Chindwin and Brahmaputra Rivers. Given the dense There is, however, as in Europe, a need to con

canopied forests that would then have dominated sider the presence of established hunter-gatherer

the lowlands of Southeast Asia, the rivers were the communities long since settled in the area which

principal arteries for communication and movement. saw such proposed intrusive Neolithic peoples. There

Yunnan is a key area for documenting any ex are at least two aspects to the hunter-gatherer settle

pansionary movement of this nature, because it has ment of mainland Southeast Asia. The first involved

links with the Yangtze, the Mekong and the Red settlement in the interior, where the remains are

Rivers. Baiyangcun is a site which lies within strik- largely confined to rockshelters, such as Lang

228




40°

35°

30°

25°

20°

15°

N

10°

o 95° 100° 105° 110°

Ii) Land above 180 m • Land above 2470 m

Figure 18.5. The distribution of sites mentioned in the text: 1) Diaotonghuan;2)

Xianrendong;3)

Yuchanyan;4) Bashidang; 5) Pengtoushan; 6) Fenshanbao; 7) Chengbeixi; 8) Hemudu; 9) area of the Tangjiagan culture; 10) Shixia; 11)

Xincun; 12) Chuangbanling; 13) Niling; 14) area of the Liangzhu culture; 15) Nianyuzhuan; 16) Balyancun;

17) Dadunzi; 18) Phung Nguyen; 19) Trang Kenh; 20) Lung Hoa, Xom Ren; 21) Ban Chiang; 22) Non Nok Tha;

23) Cu Lao Rua; 24) Cau Sat; 25) Ben Do; 26) XOIll Con; 27) Sanxingdui; 28) Erlitou.

229

115° 120°



Chapter 18

Rongrien in peninsular Thailand where the earliest working stone have been isolated. No evidence for

layers go back to about 38,000 BP. Recent investiga rice cultivation or animal domestication has been

tions, particularly in Vietnam, have identified nu found in this site, dated to about 2300 Be.

merous regional groups of hunter-gatherers, the earlier The form and decoration on the pottery vessels,

ones having considerable time depth. The Nguom in as well as the adze and bone industry at Nong Nor,

dustry is older than 23,000 BP, the Dieu sites date from are virtually identical with those from the base of a

30,000 BP and the Son Vi from 23,000-13,000 BP. Very much larger estuarine settlement known as Khok

few sites are found in interior river valleys, but this Phanom Di, 14 km to the north. This enigmatic site

could be the result of subsequent environmental modi was occupied from about 2000-1500 BC, and its pre

fication. The number of occupied inland rockshelters cise relationship to the intrusion of Neolithic groups

diminished markedly from the third millennium BC, into Central Thailand is no t yet finally resolved. The

but some sites continued in occupation, and forest material culture of the basal layers in all respects

hunter-gatherers continue to occupy small tracts of follows the local fisher-hunter-gatherer tradition.

peninsular Thailand and Malaysia. These hunter Over the ensuing five centuries, however, there were

gatherers present an interesting biological question, many developments. Burials followed the same pat

because in contrast to the agricultural population of tern as that seen in inland agricultural communities,

Southeast Asia they are short, dark and have a dis with extended inhumation replacing the former

tinctly Australo-Melanesian phenotype. seated, crouching position. Rice remains were found

The second hunter-gatherer adaptation was from fairly early in the sequence, but at a time when

coastal, and it has failed to survive into the present. local conditions would have either ruled ou t cultiva

However, the raised beaches which mark th e tion or made it highly marginal. A handful of sherds

Holocene high sea levels from southern China to the were tempered with rice chaff, bu t all were of exotic

Gulf of Siam harbour hundreds of former hunter origin. Initially, there were no dogs at the site, but

gatherer sites. The rich bio-productivity of the shore, these appeared after a century or so of occupation.

particularly where it forms an estuary, encourages Domestic dogs must have been derived from an ulti

permanent settlement, and some of these coastal sites mately exotic source that included native wolves.

are large an d deeply stratified. However, none ante The closest such source of wolves to Thailand is in

dates about 4000 BC, because prior to that period the China.

sea level was lower than today, but rising fast. The During the third an d fourth of the seven morarchaeological record is therefore confronted with tuary phases, local conditions saw a reduction in sea

coastal hunter gatherers who made pottery vessels level an d the formation of freshwater swamps. At

and polished adzes from the initial period of ar this juncture, the presence of hoes and reaping knives,

chaeological visibility. as well as changes in dental health, are compatible

Unfortunately, the situation has been confused with local rice cultivation. But a later rise in sea level

by the Vietnamese naming these groups 'coastal saw a return to marine conditions, and to the end of

Neolithic' on the basis of pottery making an d ground the reaping knives and hoes. While the potters of

stone tools rather than any biological evidence for Khok Phanom Di fashioned outstanding burnished

food production. What emerges from a considera mortuary vessels, and decorated them with incised

tion of the relevant sites is a series of regional hunter banded designs not totally dissimilar from the in

gatherer-fishers, some of whom lived long enough land repertoire, the forms of pot are quite differentat their base for a considerable depth of cultural from those of the inland farmers.

material to accumulate, who buried their dead by Again, parallels can be drawn with the situa

inhumation in a seated, crouched position, in asso tion in Northwest Europe, where expanding agricul

ciation with mortuary offerings. Very little is known tural groups met local hunter-gatherers. At present,

of the spatial organization within these sites, except Khok Phanom Di could be interpreted as a site where

for the site of Nong Nor, which has been almost there was a vigorous exchange in valued goods be

completely excavated (Higham & Thosarat 1998). tween coastal hunter-gatherers an d inland farmers,

This site was located on the shore of an extensive an exchange which certainly involved shell jewel

marine embayment of the Gulf of Siam. The faunal lery, stone adzes and ceramic vessels, bu t which

remains indicate deep-water fishing for large sharks could equally have incorporated people. The anvil,

and eagle rays, hunting marine mammals, as well as for example, associated with the richest female potfishing for smaller species and the collection of shell ter interred there, was made of an exotic clay and

fish. Specific areas for making pottery vessels and was inscribed with an owner's mark. Her presumed

230




daughter buried in an adjacent grave, aged 18 months

at death, was accompanied by a miniature anvil made

of the local clay (Vincent pers. comm.). I t would be

unusual if there were not such interactions at the

contact between two such different groups of peo

ple. The intriguing question posed concerns thecourse of language change under such circumstances.

Geoffrey Benjamin (1976) has reported on a de

tailed study of the languages spoken by the Aslian

(AA-speaking) hunter-gatherers of Malaysia. The

Semang are a group of Negrito hunter-gatherers

adapted to the inland forested habitat. They speak

AA languages (Aslian subgroup), and in particular,

their vocabularies for domesticated plants and ani

mals are derived from AA. Benjamin has suggested

that their ancestors originally would have spoken a

language related to Andamanese, an d adopted theirAA languages from intrusive agriculturalists, with

whom they would have been in exchange contact.

He turned to archaeology for the dating evidence

that suggests a beginning in the third millennium Be.

Reid (1994) adopted a similar interpretation for the

Nicobarese AA languages when he identified Nan

cowry as a conservative relic language, into which

the original Negrito inhabitants contributed much of

the non AA lexical component before being com

pletelyassimilated.

Conclusions

Bellwood (1993) has proposed a characteristically

succinct interpretation of a complex issue by sug

gesting that the original hunter-gatherers of South

east Asia now survive as Negrito groups in the

Andaman Islands, the Philippines and peninsula

Thailand and Malaysia. They may even be descended

from Hoabinhian occupants of the very caves where

to this day, hunter-gatherers still gather seasonally.

Their ancestral language is no t known but possibly

related ones could be investigated on the Andamans.

The intrusive agriculturalists were of southern Mon

goloid biological stock and introduced AA languages.

Acculturation in much of Southeast Asia then saw

the Widespread adoption of AA. A broad swathe of

interacting groups of AA agriculturalists, whose set

tlements stretched from Lingnan to Orissa, and from

Yunnan to southern Thailand, were later themselves

overtaken by other intrusive groups, including the Thais

(Austro-Tai languages), the Chams (Austronesian), the

Burmese (Sino-Tibetan) and the speakers of Indo

Aryan languages in India. Thus developed the kaleidoscope of languages spoken in Southeast Asia today,

a mix first noted by Simon de la Loubere in 1693.

This model stands for testing. I t has brought

some opprobrium on the author of this paper, but it

results from a genuine attempt to seek a consistent

and logical pattern. Critics are invited to provide an

alternative. However, testing must proceed, an d the

most promising avenue is seen in the new subject ofarchaeogenetics. Already, the study of dog DNA

hints at links between the prehistoric Southeast Asian

and Chinese canids. A research initiative to study

ancient human DN A is being planned.

Acknowledgements

I wish to thank Colin Renfrew an d Peter Bellwood

for inviting me to attend this meeting.

References

Allchin, R & B. Allchin, 1982. The Rise of Civilization in

India and Pakistan. Cambridge: Cambridge Univer

sity Press.

Bellwood, P., 1993. Cultural and biological differentiation

in peninsular Malaysia: the last 10,000 years. Asian

Perspectives 32, 37-60.

Benjamin, G., 1976. Austroasiatic sUbgroupings and pre

history in the Malay Peninsula, in Jenner et al. (eds.),

37-128.

Blench, R & M. Spriggs (eds.), 1998. Archaeology and Lan

guage, vol. II: Correlating Archaeological and Lingu istic

Hypotheses. London: Routledge.

Blust, R, 1996. Beyond the Austronesian homeland: the

Austric hypothesis an d its implications for archae

ology, in Prehistoric Settlement of the Pacific, ed. W.H.

Goodenough. (Transactions of the American Philo

sophical Society 86.) Philadelphia (PA): American

Philosophical Society, 117-40.

Crawford, G.W. & Chen 5.,1998. The origins of rice agri

culture: recent progress in East Asia. Antiquity 72,

858-66.

De la Loubere, 5., 1693. A New Historical Relation of the

Kingdom of Siam. London: Tho. Horne.

Heine-Geldern, R von, 1932. Urheimat und fruhestewanderungen der Austronesier. Anthropos 27, 543

619.

Higham, CF.W. & T.L.-D. Lu, 1998. The origins an d dis

persal of rice cultivation. Antiquity 72, 867-77.

Higham, CF.W. & R Thosarat (eds.), 1998. The Excavation

of Nong Nor: a Prehistoric Site in Central Thailand.

Oxford: Otago University Anthropology Department

and Oxbow Books.

Jenner, P.N., L.C Thompson & S. Starosta (eds.), 1976.

Austroasiatic Studies, vol. II. (Oceanic Linguistics Spe

cial Publication 13.) Honolulu (HI): University Press

of Hawaii.Luce, G.H., 1985. Phases of Pre-Pagan Burma. Oxford: Ox

ford University Press.

Mahdi, W., 1998. Linguistic data on transmission of South

231



Chapter 18

east Asian cultigens to India and Sri Lanka, in Blench

& Spriggs (eds.), 390-415.

Norman, J. & T. Mei, 1976. The Austroasiatics in ancient

South China; some lexical evidence. Monumenta

Serica 32, 274-301.

Pei, A., 1998. Notes on ne w advancements and revelations

in the agricultural archaeology of early rice domes

tication in the Dongting Lake region. Antiquity 72,

878-85.

Pejros,1. & V. Schirelman, 1998. Rice in Southeast Asia: a

regional interdisciplinary approach, in Blench &

Spriggs (eds.), 379-89.

Reid, L.A., 1994. Morphological evidence for Austric. Oce

anic Linguistics 33,323-44.

Schmidt, W., 1906. Die Mon-Khmer Volker: ein Bindeglied

Zwischen V61kern Zentralasiens un d Austronesiens.

Archiv der Anthropologie (Braunschweig) n.s. 5, 59-109.

Wheeler, R.E.M., 1959. Early India and Pakistan to Asboka.

London: Thames & Hudson.

YPM (Yunnan Provincial Museum), 1981. The Baiyangcun

site at Binchuan County, Yunnan Province. Kaogu

Xuebuo 1981, 349-68. [In Chinese.]

Yuan, J. & C. Zhang, 1999: The origins of pottery and rice

cultivation in China. Newsletter of the Grant-in-Aid

Program for COE Research Foundation of the Ministry

of Education. Science, Sports and Culture in Japan 2(1),

3-4.

Zhao, Z., 1998. The middle Yangtze region in China is one

place where rice was domesticated: phytolith evi

dence from Diaotonghuan cave, northern Jiangxi.

Antiquity 72, 885-97.

Zide, A.R.K. & N.H. Zide, 1976. Proto-Munda cultural

vocabulary: evidence for early agriculture, in Jenner

et al. (eds.), 1295-334.

232

Austroasiatic Languages and Rice Cultivation

Documents