I

SOUTHEAST ASiA
Foundations for an archaeological history

The modern natioi,r t-: S,lj::.-::! -{sia contain a total popularion oiover -150 rrrillion people, spread thrc',r.: "r -:.:.:: n--::--:'ber of ditTerent ethnic and Linzuisnc gaoups (Figures 1.1 and 1.2). Some o: rl:..:=;:--s. Earricularly in coastal and rrverine uet rice-growing areas, are e-xrrel--::-"':.:::-:.,u: icdav and have had a history marked in recent centuries by the attain:r:::: .-: :::: -:';:.. of poLitical and socio-cultural complerrn: Such people include the Tr:.:,. L:-:ler:. \-irinamese, Burmans, Malays, Javanese and Balinese, to name so:::3 c: r,--. i.:=::s: rI. !-iirls of population. Others are extremely small, such as the bandr of i:::.r::;s '".,':rl .:, .unir-e in a rnarginal way in the ever-diminishing rainforests. Era;:.Lrl.. :: :h-.= ::. rFs i;r:iude the Agta of Luzon, the Semang of the Malay Peninsula, the P'',:::: ..: B:r::c;n.l rhe Mani of Southern Thailand. Bet'ween the numerically-

do::i:=:r: .---,",'-=r::rs aild the hunters, we have many other groups of farming po:'*-r::-r--. r:.r-L3nlr- located in rugged upland or remote areas, whose remarkable trajicon.. -::=':"',:s and patterns of diversity contribute to one of the richest er:;rc-:rr:j:r; ::--or& in the world. This last group is too numerous for a short and succri:r i-:.:. ,:ri: lirerallv hundreds of ethnolinguistic groups exrsr in the hinterlands of \I:in":i S:-:i:;.sr Asia and in the multitudinous islands rvhich stretch from Sumatra to

:::: \:'... Guinea. All c: :i-: rl:r\-3 peoples of Southeast Asia have ancestries, in terms of culture, language a:j S:ror.re. firmll' grounded in the Southeast Asian region. This becomes ven- clea:'.'. 1:::-....-: er:nine languages and physical appearance. At no time in the recorded Li.:.r:-,' :: So ':h:ast Asia, extending back for at last 2,000 years, can we see any
Luzo:,

major rep,a:-L-r-:--.-i c: i::rive languages or populations by ir-nmigrants from outside the region. lJndci:b:-t ..i:r:.r\ n'ith India and China through the past 2,000 years have "'. transfornred nr:--h .-: :::: --:.riral historl'of Southeast Asia and brought man\.settlers. So, too, have conta.rs-*"r:;: I...::tr po\\'ers after an 1200 and rvith European porvers after an 1500. But rn no .r:3 'r.'-:3 pcFulatrons from any of these regions able to settle in significant numbers. Tn: Sc'.'rh::.:r Asians themselves do not speak descendant forms of Sanskrit, Chinese or Ar;bii a. ri::r r-ernaculars and have never done so; neither are they biologically derived fronr Sou:h. E;': or \\'est Asia. As Alfred Crosbyl has so evocatively described, regions such as Sourhea:: -\sia u'ere far too densely settled and too tropical in environment to attract successlul Eu:opean settlement. Likewise, the Indian, Chinese and Arab trading states and enrpires. Ce'pire :h:ir undoubted contributions to Southeast Asian cultr-rre history, r.rever dispatche d large numbers of settlers into the region.

SOUTHEAST ASIA

Figure 1.1 The major language families and subgroups witl a selection of the major ethnolinguistic populations of modern Southeast Asia menrioned in the text-

Southeast Asia did not witness any truly independent of early agriculture, urban civilization or literacy, but it did witness the oldest recorded sea crossings by humans, the genesis of the greatest ethnolinguistic
developments
dispersal in human history (that of the Austronesians), and the evolution of some major historical civilizations. Southeast Asian populations during the Neolithic and Early Metal periods also contributed much to human achievements in agriculture, art,

As far as we know,

metallurgy, boat construction and ocean navigation. Southeast Asia, indeed, must have served as the proximal source during the past 60,000 years for all the populations of Australasia and the Pacific Islands, from the Aboriginal Tasmanians to the Hawaiians and
Easter Islanders.

So this book is about the history of the native peoples of Southeast Asia, as visibie i:-:::r-f il the data of archaeology, but with excursions when necessary into the fields of o..'i::ren'-ar-\'history art and architecture, language and human biology. We examine the flor','erl:ri l: rndigenous traditions across the millennia, often stimulated by external concaclj. ;.::: cien il turn stimulating developments in other regions. In order to put the follot-ins ch=:::r. il oenpective, the remainder of this chapter will be devoted ro three themes: rhe rr-'.'r::::::-n:s. rhe people, and the deeper foundation levels of prehistory before the bcgr:.:.-:.-. : ....:cmatic cereal agriculture and animal domestication.

].

:-.-:-:.: ::::-: --: '--.::-l: Kelantan, ?:::r=ari :-; ll Bel-,r-ood.)

Malaysia'

(b) Burme-.e l:lci-ls a:

F::::.. ?t--::aai'

I. C.

Gic.'-r:.

Bunak bards, dressed in ceremonial attire to

recite the Bei Gua,

Ancestors' before

the altar of war, East

or

'Itinerary

of

the

Timor. @hotograph by C. Friedberg-)

(d) People on the Xieng KhouanE Pla:ea'; northern Laos. (From Frangois Garnier 18--i.
Figure 1.2 Traditional and present da,v Southeast Asiaru.

SOUTHEAST ASIA

Southeast Asian environments

Southeast Asia comprises a remarkabie array of peoples and cultures located across two geographical divisions. These are Mainland Southeast ,4sla (Burma, Laos, Thailand, Cambodia, Vietnam, Peninsular Maiaysia) and Island Southeast Asia (Taiwan, Philippines,

East Malaysia, Indonesia, Brunei). These divisions have not always had the same geographical shapes or environmental characteristics as today. During the Pleistocene Epoch of geologrcal time (r. 1.8 million to 10,000 years ago), the eartht climate was drastically cooled bv a series of ice ages. At the peaks (or maxima) of such cold phases, land temperarures dropped on average by 5 or 6 degrees Celsius, sea levels fell to over 100 metres belou-rheir present levels, and rainfall and vegetation appear to have been greatly reduced. even in the tropics. The last glacial maximum occurred about 20,000 years ago, the previous one about 130,000 years ago, and between these glacial maxima the climates of the rvorld oscillated very frequently, generally being less warm and wet than those of today. 'We are now in the peak of an interglacial which has lasted for the past 10,000 ]*ears, a veritable paradise of climate and environment during which time ali human complex cultures, with farming and cities, have evolved. Without the existence of this current interglacial, life as we know it would not exist, hence, in part, our current fears over which way the climate will turn in the future. When sea levels fell during the long periods of cold climate during the Pleistocene, Main-land Southeast Asia was extended by land bridges to the Sunda Shelf islands of Sumatra, Java, Borneo and Bali, forming a huge southeasterly dry land extension to the Asian continent. This massive ancient land mass is termed "Sundaland" by biogeographers (Figure 1.3). Taiwan was also joined to the Chinese Mainland. However, the islands of the Philippines and eastern Indonesia (Sulawesi, Moluccas, Lesser Sundas) were never land-bridged either to Asia or to Greater Australia (Australia plus New Guinea and Thsmania), although the distances between them were sometimes much reduced. These eastern Indonesian islands belong to the biogeographical province of "Wal)acea", through which human migration beyond the Sunda Shelf has always involved sea crossings, generally between about 10 and 90 kilometres and thus too wide to allow migration by swimming alone. Boats or rafts of some kind must have been used, even for the first crossings to Australia berween 40,000 and 60,000 years ago. 'Wallace, a As biogeographers have long observed since the days of Alfred Russel contemporary of Charles Darwin, Sundaland continentality equates with richness in terms of native faunas and floras, whereas Wallacean insulariry equates with endemiciol and pauperization. The Asian Mainland, together with the Sunda Shelf islands to as far east as Bali and Borneo, has varied placental mammal faunas, including some very large species .uch as elephants, rhinos, cattle, deer, pigs and tigers. Greater Australia has a marsupial ---:ra. rvith animals such as wallabies, bandicoots and cuscuses (possums) occurring in \:'.'.' Guinea and adjacent islands. Colonists in'Wallacea found fewer animal resources, ''::-i':ii: a number of endemic mammal species existed in the PhiJippines and Sulawesi, r:: ::.= -=::er isiand even including wvo endemic genera of cuscuses whose ancestors rafted on ir-.-r:rird crust aeons ago from Gleater Australia. The Lesser Sundas and Moluccas, hos'e',-c:. :--:: precious few mammalian faunal resources in the Late Pleistocene until people lrar::: -l: r-aiues of animal translocation although stegodonts (elephant-like creatures. no'.,. ir::;r.-r\ had reached Sulawesi and the Lesser Sundas before modern humans enterej r:: reiion about 50,000 years ago. The first evidence for purposeful

'ft
_)l

Qt

z
F.-

o
(J G

q)

o-

= z
!

u)

t-u

it

-'.

l. lt,
:a.ccl

= a ) = = a
ri.t

l,;'
rI:E

o Oz.

lr)

3 =

-a--\- ;l-:-<---. '.-1.- -4=-... ____aj"r-_A\.

-

= ':?={tt\.
.-a_-

co

I
f

>.t

77

,rl
o
@

ds)
bs

U) 6

ti r\

\t

o
6 d

.o
o ! a

!:

=-,

\1?
a
:.i-

o
d
N

z <

.(!

tr

o
E
O
0)

c.

t.

SOUTHEAST ASIA

animal translocation falls benveen the late Pleistocene and middle Holocene, when
marsupials were taken &om Nerv Guinea to the northern Moluccas Islands,

Timor and

the Bismarck Archipelago.

The modern enr'-ironments of Southeast Asia are essentiaily tropical (Figure 1.4). There is increxing dry- sexon length as one moves polewards beyond 5 degrees north and south of the equator, along which rain falls during much of the year. Equatorial populacioru Lir-ing below these latitudes today tend to be small, and in eastern Indonesia often dependent for subsistence on tubers and tree crops such as yarris, taro, various palm
starches and bananas. The major cereals

of the Southeast Asian and Chinese Neolithic

culrures. especially rice and foxtail millet, are not well suited to hot and ever-wet equatorial climates and evolved beyond the equatorial zone in tropical to temperate monsoon regions with long dry seasons. Today, and in the historica-l past, the greatest population densities have occurred in these monsoon regions, especially in alluvial lorvlands in southern China and Mainland Southeast Asia north of the Malay Peninsula, and in alluvial/volcanic landscapes in Luzon, Java and Bali. As human prehistory in Southeast Asia progressed, so human relationships with landscapes and environments changed. Hunters, farmers, and urban dwellers tend to use different parts of the landscape, with increasing population density becoming focused through time in ever-smaller areas of high productive potential. Ten thousand years ago, human populations were mobile and dispersed, with periodic concentrations in rich coastal regions. By 1,000-1,500 years ago, wet-field rice landscapes and urban,/temple complexes occupied the richest and most accessible lowlands, village farmers occupied the less central valleys and slopes, and hunter-gatherers occupied the remorest interiors and coasts. By 1,000 years ago also, the ethnic patterns wtr-ich still join and divide the peoples of Southeast Asia were already well and truly in place.

The peoples of Southeast Asia

Since several chapters

in this book

extend well

into the hisrorical period, with

decipherable inscriptions and written records, it is necessary briefly to describe the main ethnolinguistic divisions within the recent and modern populations of Southeast Asia. This may best be done with reference to language, since there can be no doubt that language, on the large scale of the language family, is intimately correlated with the ct'ltural and geographical origins of different peoples. The language families of Southeast Asia will be discussed in more detail in the following rwo chapters, when we come to examine their suggested origins and early dispersal patterns. At this point, however, the main language families, together with brief statements on

'-tir
_+a

possible origins and histories, may be introduced. Approximate modern ::ic:':ucions are shown in Figure 1.1. Excluding minority languages which have .:r:;: ir recent centuries, particularly from Europe and china, we have five major

PETEIl. BELLWOOI) AND IAN (;LOVEI\ belonging to the agricuitural period, an advantage denied in the more cquatorial regions of Malaysia, Indonesia and the Philippines. Moving now to the broadest scale of prehictonc reconstruction, archaeology and allied disciplines tell us that the main roots of the pre-Colonrbian patterns of worldwide cultural and biological diversiry can be traced. subsequent ro the biological evolution of humaniry itself, to a number of human nrigraton'achieverrrents. One rvas the colonization of the Old World beyond the African homel.rnc. This prob,abh' spanned at least two distinct stages, the first involving an out-of -\iica mor-enrenr t'\'earh' Honto populations, with a

later one involving further nlovenr.nis. perhaps our of -{rrica again, by anatomically modern populations of Hottto -i,ryi61;-r. our direct ancestors. The eariy Flozo populations settled in Southeast Asia and rrt:nie d north\\'rrd\ an,l ea.nvards to China and Java by
perhaps 1.5 nrillion years aso. The nrost recent and extensir-e bursts

of colonization,

involving the anatomicalh' modern humans, extended across the sea gaps of eastern Indonesia to reach Ausrrali:. \es- Guinea and possiblr'the Philippines around 35,000 to

50,000 years ago (Figure i.l . The Anrericas s'ere reached nrore recently, perhaps about 14,000 years ago. The second ma-ior sriies of hurrran expansions resulted tiorn the establishment and growth of the earliesr agriculrural econornies. These arose independenth-. at various times between 12,0tttt and abour 5.000 years ago, in regions such as the Nliddle East, central China, the Nes' Guinea Highlands, and some tropical regions of Atrica and the Americas. Here rve s'itness. in the archaeologrcal record, the oldest systematic rvar-s of producing rather than collecring, or hunring food - in other words, agriculrure and animal husbandry. These developnrenrs a-ilorl'ed a spurt of human population growth in the agricultural latitudes of th.' earth unprecedented in the previous rwo million years of human evolution. The ethnolinguistic map of Southeast Asia today, with its major language
fanri,Lies

(Figure 1.1), is almost entirely a creation of this recent period of remarkable

social and economic deveiopment.

The first hurnans in Southeast Asia

Because thrs book focuses on the antecedents of living peoples and societies

in Southeast

or.r Honto trt.tts. But thev do deserve recognirior-r as the human pioneer species. Discovered in greatest tbssil numbers in the eroded anticline of Sangiran in centralJava, this earl1-human species has become known popularlv to the n'orld by the rather patriarchal term "far-a \Ian" (r'es, some were women too, and even children!), or Pitlrccanthropus, the latter ternr bestou-ed br.the original discoverer, Eugene Dubois, following his epochal discoveries at Trinii in 1891. Today, research on Honrtt erectus in East Asia is proceedine apace. Large popularions are now knorvn in tbssil fornr from sites in rnany parts of China and _Java. This species evolved in Java ior up*'ards of 1.5 million years, probabh' in senri-isolation and probably reaching extinction as recently as 50,000 years ago. Did nrodern humans apply the coupde-grice, as many suggest for the Neanderthals? 'W'e mav ner-er knorv, and it must be acknowledged here that opinions on the extinction issue are varied and often heated. There are many issues of chronology, species definition and genetic affinity which will need to be addressed before we can come close to an), fina1 conclusion. Extinct or not, Horno erectrts rvas clearly hunran. Bipedal like us, but somewhat smaller-brained, they rnade stone tools in China and possibly in Java, and also may have

Asia,

it is not necessary to focus in detail

t2

SOUTHEAST ASIA
made narrow sea crossings (see next section). They were perhaps hunters - at least, their fossils often occur rr'ith the bones of the large Early and Middle Pleistocene manunals of Java, species such as elephants, stegodonts, rhinos, hippos, u'ater buffalo, deer, pig and tiger. Most of these anirrrals favoured fairly open forest or grassland, and the rainforests of Java r.vith their orang-utans (no longer extant in Java) and gibbons emerged later, during the tlnal interglacial about 130,000 years ago. A hunter's life in 800000 nc, if indeed thev hunred, rvas perhaps a good one. The problem is that nrost fossils are found

deposits, rather than in primary circumstances, therefore "proving" an actir.rn' such as hunting, as opposed to scavenging, is not an easy matter. Apart from some ven' sinrple stone tools, we know nothing of the cultural capacities of these early hunrans. Did they create art, imagine the future, bury their dead, decorate their bodies? We do not know So let us move on a million years or so, to the first activities of the direct ancestors of r.nodern populations. The strong evidence for an anatomically modern human presence in Australia by 50,000 years ago makes it fairly certain that such populations u-ere also present at that time in Southeast Asia. A skull from Niah Cave in Sarawak dated perhaps to 40,000 years ago, and younger skull fragments from Tabon Cave in the Philippines. attest the Late Pleistocene presence in Island Southeast Asia of anatomically modern humans of a phenorype resembling ancient Australians and Melanesians. These populations travelled far across both land and sea.

in water-laid

The first crossings of the

sea

The abilin' to build a ratt or canoe. horvever primitive, in order to cross a body of u'ater is one n-hich has characrerized most coastal human societies in history. Gro*ing el-idence for the existence of such abiliry in the prehistoric record of humaniry makes us rvonder: horv long ago did people begin to launch themselves at sea, not just to cross a narrow channel to an offshore island but to reach places close to and even over the horizon? 'Were such crossings done deliberately? Data on the first ocean crossings of humaniry may be ambiguous about dates and degrees of intentionaLiry but current archaeological investigation is opening some remarkable new perspectives. In most areas of the world, the abiliry to cross oceans seenx to have developed within the time span of anatomicaliy modern humans like ourselves. The only exception to this generalization, publicized very recently and still under debate, is a claim that Homo erectus able to cross one or more sea gaps up to 19 km wide about 800,000 years ago in "vas order to travel from Bali across the Wallace Line to the eastern Indonesian island of Flores. This is a remarkable claim, so far unprecedented and obviously rather isolated, rten that no other ocean crossing can be positively attested until about 40,000 to 60,000 '. -i:i f,so. when humans probably reached Australia. But, if true, the Flores discovery .:- j-,-.::. _iust how much we have yet to learn about the capabilities of hominins very :::---:.-::: time, even though the sea crossings concerned in this instance were not out of

-..

,t].r.

r1o\1-

on the remarkable nerv information about the capaciry of early

berween 50,000 and 30,000 years ago, information recovered
Southeast Asia and

:-r:< io

cross sea

--:--,:

siies
1

, i:::::
l:.:.:

of Late Pleistocene age in the islands of

.3). Although most major islands in the Philippines and eastern =::-.'-:Sh to be visible to their neighbours, the crossing irom Roti or
13

I'ETEII I]ELLWOOI) ANI) IAN GLOVEI{

Timor to the Sahul shelf of northwestern AustraLia. perhaps via now-drowned intervening islands, would have involved one crossing oi at leasr 9{) km even at times of lorvest sea
level. This could have required sailing con.rpleteh'out of sight of land for a short period, and we know this crossing was done at least -+5.l'ir-rrl sears ago. Such relatively narrow sea gaps, mostly under 90 km wide, u-hich u'rie n-gotiable for Pleistocene seafarers, corltinued east of New Guinea to as far :s th: 'ouii-rtrn Solonrons, beyond which lay much rvider expanses of sea, crossed on ::-.rr eliience onlv bv Austronesians with sailing and outrigger canoe technolog,' :r::: -i.-i . '.'ear: ago.

years, archaeologistr:rj n=;-t:"-.,-t:niis:s u'orking in Wallacea and -.::::l: :::'''' huinan colonization of this rvhole region, from Sulau'est. ri. -.-:":j I...:'t,. rh: -\1c'u--c.:s. \es- Guinea, the Bismarck and Solonron Islanir. .:,i. ::::.::::1::r.:'-'u'. !r:illFi. *orlr Australia itself' In terms of the date of th: F-:s::::l-::1 hlnar: crossinq oithe \\''allace Line. ieaving So tar the Oldest eridence. Optical aside issues over -Flrtt;.-:rr--:i1-. -1'i:::x::ll) luminescence dates l:. :::::=:.---=:::::l-'i:s:i ttchnique) for quartz particles in sandy sediments lron nlo r-:.r ,:- :---= \i::hern Territon' suggest a human presence, with associated stone lia-{-:--,:1.. .:::lfs:cli:s and ground pieces of red ochre, by as much as 'W'ales are of a 50,000 years aso. \:..'.' j,:;. :in, Lake \lungo in western New South the islands of Sulawesi, similar antiquir,-. S: :;:. .:."-r'1.::,e radiocarbon dates suggest that Talaud, rhe -\1o-:-----. B:-.::t-,-*s and Solomons were first settled befvveen 28,000 and 35,001) sears ;E.-. T1::r: j;:e li.rributions raise the possibility that Australia rvas reached from rhl Le..=: S'-i::i:. perhaps 20,000 years before many of the islands in eastern Indonesia. f ::t:i::;.:nt riri. can be no more than a possibiliry subject to future testing br- drf:r:n: ;:!l:: ne:hod:. Dates for initial human occupation in the Philippines, a pro.e)s [:t:.1 ]-r:c:t certainll' occurred from Borneo rather than from Tbiwan, fall gener--r' -;"::i--;,-- tir- Fast 30,000 years and a former presence herc of Homo erectus seetr$ r-en' un *:il-. Le: ::s no-r.' .ook at s-hat is known of the lives of these first humans to cross the seas of \\ali;ce; arc \lelanesia. The precise types of $'atercraft used bv these first colonists will renain :lc;e'.'er unknorvn, but the large rafts made of lashed-toeether bamboo poles

In recent

Australasia have produced astound::rS:'.-,lJ:::--:

g-h1ch snl-

some Southeast Asian rivers for transporr purposes m,rght give some ciues. Lirhic t:chno1o5- certainly included simple flake and pebble tools. together with the use of red anc velo*'ochre for pigment and heated stones for cooking. Shellfish and inshore fish prouced protein. as did the rather limited mamrnal species of these island regions. A consunrh'-e\panding archaeological record tells us that the descendants of these early 'Wailacean niE anrs adapted rapidly to the very different inland enr-ironnlents of Australia

rh'

and Neu' Guinee. u'here mammal faunas were far more plenriful. The prehistory of Greater Ausralia. including New Guinea, is of course a topic tbr another venue, but the original island and coastal maritime lifestyle of the first settlers did continue in the small equarorial islan& of the N4oluccas and Bismarck Archipelagoes. located to west and east of New Guinea respecrivelr'. The record of this lifesryle over the past 35,000 years shows some very interesting developmentsIn the archaeological records of these two archipelagos there appear to be lvvo successive periods of actir-in'. The first, frombefore 35,000 down to about 20,000 years ago, reveals a fairly simple lifesryle of strandline exploitation with inshore fishing, shellfishing and crab collection, combined with only very limited evidence that communities were in frequent inter-island contact. For instance, in this early period rve
1A 1a

SOUTHEAST ASIA

of any trade in tool stones such as obsidian or chert, or any deliberate translocation of animals or plants berrveen islands. This ma1' reflect deficiencies of the record, but taken at face value the picture indicates that, once a nerv island was reached, people setded in, retaining mostly a localized form of mobiliry.
see few signs

Gradually, more regular inter-island contacts for trade purposes seem to have developed. A black volcanic glass called obsidian, collected from at least Nvo sources in central New Britain. rl-as uken by about 20,000 years ago, in small anrounts, to sites up to 350 km au'av in Neu' Ireland. Even more surprisingly, people began to carry marsupials, evidendv from the New Guinea Mainland, to those islands both west and east of Nerv Guinea u'here they did not occur in the native fauna. Cuscuses (Phalanger

orientalis) were taken to New Ireland by 15,000 years ago, bandicoots (Echynipera sp.) to the Admiralries br- 12,000 years ago, and wallabies (Thylogale brunil ro Nerv Ireland by possibly 7,500 y'ean ago. In the northern Moluccas, a species of wallaby (Dorcopsk nuellerl was translocared to Gebe Island, immediately west of New Guinea, at about 9,000 years ago, apparentlr'from a source on Misool Island. A related species of wallaby occurred on Halmahera. perhaps also translocated from New Guinea, although this remains uncertain. Sadl;', these u'allabies were extirpated for reasons unknown in the northern Moluccas b1'

800 lean ago. Although these animal movements were probably uncoordinated - currently they appear to have been drawn out over a period ofperhaps 10,000 years - they do indicate that conscious attempts to increase food supplies on relatively impoverished islands were

taking place. This record

of animal translocation is at present unique in

world

prehistory.a New tool forms also appeared after about 10,000 years ago in several islands. The most significant are adzes made from large reef shells such as Tiiilacna and Hippopus, ground with hollowed cutting edges, found in excavations in Gebe and the Admiralty Islands. These shell adzes could have been used in dug-out canoe construction, a development which might have led to the seeming florescence in human contacts around this time. Edge-ground stone axes also made an appearance as early as 25,000 to 30,000 years ago in northern Australia, Japan and the New Guinea Highlands, where they might have been used for forest clearance activities by hunter-gatherers. Other examples, less certainly dated, are known from Vietnam, Peninsular Malaysia and

Borneo. Did the earliest Palaeolithic colonists of Australia, New Guinea and the islands as far east as the Solomons, in some cases no doubt the contemporaries of late Neanderthals in Europe and western Asia, cross the sea with intent? This, naturally, is a question of i,r:ne contention which will perhaps never be answerable with great conviction. But as --:: ;olonization of these islands progressed, so the cultural practices required to make : - -::::zation successful were developed. These practices included translocation of food -:*::-i-i ard plants and an increasing sophistication of watercraft technology, perhaps : :: -:-.:-',-:ng use of the sail before Austronesian times. However, it is interesting that , : - - : : :.::- :onract the inhabitants of Australia, except for those in northern Australia :::.: of outrigger canoe distribution in Indonesia and New Guinea, did -: ::-,:rufacture large canoes. Thus, it is likely that advanced boat-building -- ;--, ri: uses of outriggers, sails and dug-out log hulls only evolved in -.-- '.i ::::s rou'ards the end of the Pleistocene, long after Australia and ::: .:,,::=,..-settled. The first settlers probably used paddles rather than

1:)

I'ETEII. I]ELLWOOI) ANI) IAN GLOVEI\

The Sunda Islands and the Southeast Asian Mainland

The above evidence for maritir-ne activities and sea crossings is of course best known to us fronr the island world of Waliacea and Melanesia rarher than from the continental regions, including the massive islands of Surnarra. Borneo and Java, which lie

inrmediately to its west. On the Southe.rsr Asian \lainland, cave occupation by populations assumed to be anator.rricalh' nroJtri: i:re. back to at least 40,000 years ago, particularly in Thailand and Vietnarl. althoush skel:ral nraterial older than 20,000 years has not yet been found. By the end of the P,:i.:..::::e. ai about 10,000 years ago, a series -re;:n: r'en' s-idespread in southern of lithic assemblages termed "Hoabinhr.rn" China and through the whole of Mairrland Scuil:i:.: -\:i.r:c;s trr as northern Sun.ratra. These are characterizedby unifacial or t'it;--i;- *:--rke :o.ris :r:dl on river pebbles and seem to

reflect a hunting and gathering -.onom\-. perhars u-ith some degree of plant nlanagenlent. Most interestir:S1..'. rh: Hoabinhians seen ro have been adept inhabitants of the interior rainforests of th: \1..:r- Peninsula. an enlironnent q'hich sureiy involved a substantial in-u'estment in pl.rr: er:-,rr=rion. Clear signs of arumal or plant domestication
are, however, entireh' at's:;ir. In the islands of r..-es:ern I:ion:.i; the contenrporan'popuiarions are represented by the "Deep Skull":c \iah ani br-a series of flake-based. rarher rhan pebble-based, lithic industries. Such in,lusrie. har-r been tound rvidely in caves in Sumarra, Java and northern Borneo, and the tools ;c::;er::eJ are essentially very similar to those manufacrured biz the contenlporan- sea-;cssir:E ropularions located to the east, in 'Wallacea. An unusual

industn(Figure

oi cher: bi:::e. or possible Late Pleistocene age from Tingkayu in Sabah i.5 doe. ri::.: to modifr'the general picture of lithic simplicity and uniformiry

an impressio:r u-hich ha-i suqqested to many that bamboos rather than stone was the real prorider of t-oriins eiges in this part of the rvorid. Of art rve have little, especially when compared ro rhe rich car-e upestries of northern Australia, but fairly schematic examples oi hunran ;nd arunal tigurative art, created trom ochre or charcoal br' late hunting populations. dcr oicur in caves and rockshelten scartered across the region. None can be secureh- ci:r:C. \eganr-e "hand stencils". an art lbrnr so s'ell knorr-n in Ausrralia and in the Palaeolirhic c:r-e art of Europe. are reported trom Sulas'esi. the southeastern Moluccas ani \\est Papua. and the -\1:nskalihar Peninsula of eastern Kalimantan (Borneo). susge-\nng perhaps a greater annquin'rha;r hirherto expected for some of the cave art ot the rcgion.' Durinq the nrilole Holocene. comner:cing about 7,000 years ago, the southwestern arm of the isiand of Sularr'esi s'itnessed a nicsr unusual effiorescence of lithic technoiogy, rypified bv backed bi;des and "nucrolirhs" a::i. a series of remarkable serrated-edged and hollow-based pro,ie ccle loints terrnei "\1:ros Points". This industry termed the Toalian (Figure 1.6), appears to have erisred i:: a rarher splendid isolation, rvith contemporary parallels found onh' to the south in Aurraiir. Perhaps there were links, although the absence of such technolosr- ir.l the inren'ening Lesser Sunda Isiands, including Timor, makes one wonder.

By 4,500 years ago. manr- ot the hunrer-satherer societies of Southeast Asia, particularly those located in northerh' re sions. must have begun to suspect that sorlething new was in the n-ind. Farninq societies had already been in existence for -1,000 years by this tinre in the rniddle Yrngzi resion of China. As farmers and their subsistence econolry slorvly spread south. reachinq the Pearl Delta and Taiwan by 5,000
1.6

I'ETEI{ I]ELLWOOI) ANI) IAN GLOVEI{

The Sunda Islands and the Southeast Asian Mainland

Tl-re above evidence for maritinre activrties and sea crossings is of course bestknown to us

fronr the island world of Wallacea and Melanesia rarher than from the continental regions, inciuding the massive islands oi Sumatra. Borneo and Java, which lie inrmediately to its west. On the Southeasr Asian Nlainland. cave occupation by populations assumed to be anaton.ricalh' modern dates back ro at least 40,000 years ago, particularly in Thailand and Vietnarr.r. although skeletal material older than 20,000 years
not yet been found. By the end of the Pleistocene, at about 1ll.{,)00 years ago, a series assemblages termed "Hoabinhi;rn'' became ven' s'idespread in southern China and through the whole of Mainland Southeast Asia to as far as nonhern Sumatra. These are characterizedby unifacial or bitaciai axe-like tools made on river pebbles and seem to reflect a hunting and gathering economv, perhaps rrirh some degree of plant
has

of lithic

nlanagement. Most ir.rterestinslr'. the Hoabinhians seem to have been adept inhabitants of the interior rainforests of the .\1aiav Peninsula, an environment s-hich sureh'involved a substantial investment in pianr erploitarion. Clear sigrs of animal or piant domestication are, however, entireh' absenr. In the islands oirr-esrern In,jonesia the contemporary populations are represented by the "Deep Skuil" at \iah and b'r'a series of flake-based, rather than pebble-based, lithic industries. Such indusrir. have been found widely in caves in Sumatra, Jar-a and northern Borneo, and the tooL concerned are essentially very similar to those manufacrured b,v the contemporan' sea-crossins populations located to the east, in 'W'allacea. An unusuai industn' oi cherr b,itlces of possible Late Pleistocene age from Tingkal'u in Sabah (Figure 1.5) doe. hrue ro modifi-the general picture of lithic simplicity and uniformiry an impression *'hich h,x suggested to many that bamboos rather than stone was the real pror-ider of s'orking edqes in this part of the rvorld. Of art we have little, especially when compared to rhe ri;h cale tapestries of northern Ausrralia, but fairly schematic examples of human anci anima-l t-igurative art, created from ochre or charcoal br' late hunting populations. do occur in caves and rocksheiters scattered across the region. None can be securelr- dateC. \eeacir-e "hand stencils", an art form so n-ell knot-n in Ausrralia and in the Palaeolithic car-e art of Europe, are reponed frorn Sulau'e.r. the southeastern Moluccas and \\ esr Papua, and the Mangkalihat Penrnsula of eartern Kalimantan (Borneo), susg3snns perhaps a greater anriquin than hirherto expecred for some of the cave art of the region.During rhe nriddle Hoiocene. commencing 3lr6u1 l.r,r ri t \-ears ago. the southrvestern arm of the island of Suias'esi u-itnessed a most unusual etiorescence of lithic technology, typified by backed blades and "microliths" and a series ot remarkabie serrated-edged and hollow-based pro-iecnie poinrs termed "lr4aros Points". Thrs indusrn'. rermed the Toalian (Figure 1.6), appears to have existed in a rather splendid isolarion. nith contemporary parallels found onlr- to rhe south in Australia. Perhaps there s-ere links, aithough the absence of such technoiogv in the rnten'ening Lesser Sunda Islands, inciuding Timor, nakes one wonder. By 4,500 years ago. manv of the hunter-gatherer societies of Southeast Asia, particularly those located in northerl_v regrons. musr have begun to suspect that something new was in the n-ind. Farming societies had already been in existence for -{,000 years by this time in the niddle Yangzi region of China. As farmers and their subsistence economy slorvly spread sourh, reaching rhe Pearl Delta and Taiwan by 5,000

1,6

PETER BELLWOOD AND IAN GLOVER

Notes
1 Crosby 1986.

2 For rice terminology

of this family,
see

Luce, G.

3 The authors have different opinions on this, with Bellwood favouring an agricultural/
Neolithic dispersal of early Austroasiaric speaken, together with Higham
whereas Glover prefers a pre-Neolithic dispenal. Island Melanesians, Oxford, Blackwell; and

amongst early speaken of Austroasiatic languages, and the general history H. 1985; Mahdi, W in Blench and Spriggs 1998; Peiros 1998.

in

Chapter 3,

4 On translocation, see Sprigg, M. (1997) The

Flannery T. et al. (1998) Mammals from Hoiocene archaeological sites on Gebe and Morotai Islands, northern Moluccas, Indonesia, Australian Mammalogy 20/3: 3914Q0. 5 There is an argument that the large bamboos which could have provided cutting tools on the mainland were taken by people from the Asian Mainland into Sahul and Wallacea in mid-Holocene times. This couid offer a partial explanation for the dichotomy berween pebble/core tools + bamboo on the mainland of Southext Asia, and sharp-edged flake tools in the islands to the east. Ho*'ever, Dr Soeyatmi Dransfield of the Royal Botanic Gardens, Kew notes that, "There are ser-eral large bamboos in Indonesia, Dendroulatnus giganteus, was introduced from Burma probably in the eariy nineteenth cenrury rvhi1e other large bamboos, such as Dendrocalanns a:per, Cigantochloa robusta, znd Gigantochloa pseudoarundinacea,havebeen in Java or Sumatra for a long time, either planted or natureiized, and their origins are not known. So far they have ner-er been recorded to occur irr Mainland Asia". 6 For Kalimanten hand stenciJs, seeJ. M. Chazine et aL at htrp://wwrv.kalimanthrope.com.

Select Bibliography
Peoplu of Southeast Asia: anthropology anil history
A-lbrecht, G. and \Ioser, J. (1999) "Recent Mani setdements in Sarun Province, South Thailand". Jounul oJ the Siam Sooery 86 (l&-2): 161-99. Bellrvood, P., Fox, J. J. and Tryon, D. (eds) (1995) The Austronesians: Comparatiue and Historical Perspeoira, Canberra: Dept. Anthropology, Research School of Pacific and Asian Studies,

Australian Nacional Univenity.

Chazine, J. NI. er al. arhttp://www.kalimanthrope.com. Crosbr', A. (1986) Ecologiul Inryerialism, Cambridge: Cambridge (Jniversiry Press. Garnier, E (1873) lbyage d'Exploration en Indochine, Paris: Hachette.

Geertz, H. (1963) "Indonesian culrures and communities", in R. McVey (ed.) Indonesia, New Haven: Yale Univeniry Press, pp. 24-96. Hoban, A., Ramsever, U. and Leemann, A. (1996) The Peopla oJ Bali, Ofrord: Blackwell. King, Victor T. (i993) The Peoples of Borneo, Oxford: Blackwell. Lebar, F M. (ed.) (1972-5) Ethnic Groups oJ Insular Southeat Asia, vols 7 and 2, New Haven:

HRAF

Press.

Lebar, F. Lee,

M., Hicker', G. C.

New Har-en: HRAF

and Musgrave, J.

K.

(1964) Ethnic Croups

oJ

Mainland Southeast Asia,

Press.

and Dal;', R. (eds) (1999) Tlrc Canfuidge Enryclopaedia oJHunters and Catherers, Cambridge: Cambridge lJnir-enin' Pres. Luce, G. H. (1985) Pha:es of Pre-Pagan Burna,2 vols, Ox{ord: Oxford UniversiryPress. Mabbett, I. and Chandler. D. (1995) The Khnrers, Oxford: Blackwell. Mahdi, 'f{ (1998) Linguistic daca on transmission of Southeast Asian culrigens to India and Sri Lanka, in R. Blench and.N1. Spriggs (eds), Archaeology and Language II, pp. 39G-415. London: Routledge. Peiras, C. (1996) The Bugrs, Oxford: Blackrvel. Spriggs, M. (1997) The Island lyfelanesians, Ot'ord: Blackwel].

18

SOUTHEAST ASIA

Enuironments anil pleistocene archaeology

Bellwood, P. (1997) Prehistorl, oJ tlrc Indo-Malaysian Archipelago,2nd edition, Honolulu: Universiry of Harvai'i Pres. Beilrvood, P. et al. (1998) "35,000 yean of prehistory in the northern Moluccas", in G-J. Bartstra et al.. Bid\ Head Approaches, Modern Quaternary Research in Southeast Asia 15; 23315.

Roncrd:m: Balkema. palagogeographic

Bergh. G. r'an den (1999) The l-zte Bergh. G.

Neogene Elephantoid-bearing.Fauna: o-f Indonesia and their Inplications, Leiden: Nationaal Natuurhistorisch Museum'

tr.r d.tr, de vos, J., Sondaar, P. a;nd Aziz, F. (1996). "Pleistocene zoogeog-raphic evolution ofJava (Indonesia) and glacio-eustatic sea level fluctuations: a background for the presence of Homo", Bulletin oJ the Indo-Pacifc Prehistory Association 14: 7-21 Bergir, G. van den, Mubroto, B., Aziz, F, Sondaar, P. and de Vos,J. (1996) "Did Hono ercctus reach the island of Flores?", Bulletin of the Indo-Pacifc Prehistory Association 14 27-36. Flannery T., Bellwood, P.,'lVhite,J. P., Ennis, T., Irwin, G., Schubert, K. and Balasubramanian, S. (1998) "Mammals from Holocene archaeological sites on Gebe and Morotai Islands, northern Moluccas, Indonesia", Australian Mammalogy 20 / 3: 397400. Keates, S. and Bartstra, G-J. (2001) "Observations on Cabengian and Pacitanian artelacts from
Island Southeast Asia", Quartiir 51 / 52: 9-32. Morwood, M., O'Sullivan, P.8., ,$ziz,F. andRaza, A. (1998) "Fission-track ages of stone tools and fossils on the east Indonesian island of Flores", Natwe 392: 17734. Simanjuntak, T., Praseryo, B. and Handini, R. (ed$ (2001) Sangkan; Mayt, Cuhure and Enuironnrertt ii Pleistocene Times,Jakarta: National Research Centre of Archaeology. 'w'allace, A. R. (1962) The Malay Archipelago (originally published 1869), New York: Dover. Watanabe, N. and Kadar, D. (ed$ (1985) Quaternary Geology oJ the Hominid Fossil Beaing Formatiotts inJava, Bandung: Geological Research and Development Centre.

Linguistic history in Southeast Asia

Blench,
Archaeology and l-anguage, vol. .2 ,(this volume contains on Southeast Asian language history), London, Roudedge. Blust, R. a.it^SSS) "The prehistory of the Austronesian-speaking peoples: a view from language",

R. and Spriggs, M. (ed$ (998)

several papers

Journal of Woild Prehistory 9: 453-570. pa*iey, A. k. (2003) "Th; Ausrronesian dispersal: languages, technologies and people", in p. Bell*ood and C. Renfrew (eds) Examining the Farming/Language Dispersal Hypothesis,

pp. 251-74, Cambridge: McDonald Insdrute for Archaeologicai -Researchpeiros, I. (1,998) Conparatiui Linguistics in Soutlrcast Asla, Canberra: Pacific Linguistics Series C.-142. \.urm, S. A. and Fiattori, S. Gdt (1983) I-anguage Atlas of the Paciic Area, Part II. Canberra: Australian Academy of the Humanities.

Holocene orchaeolog),, pre-neolithic

--: i:::.'-n. D. (1990) Lnng Rongrien Rockshelter, Philadelphia: University of Pennsylvania Museum.., j ,, ,": :. P. n992) "soitheast Asia before history", in N. Tarling (ed.) The Canbridge History qf

-'.,:::':-':

: , , ,: :. ? i -r93) "Cukutll and biological differentiation in Peninsuiar Malaysia: the last iO.[)t,t[t ',-:' .:-! 32 3740. ,: :. : . r'-- F)ehistory of the Indo-Malaysian Archipelago, 2nd edn, Honolulu: Univenin' of
': :.": i -:-

--{-,ia,

vol. I, pp. 55-136, Cambridge: Cambridge lJniversiry

Press.

''

'H-:man dispersals and colonizations in prehistory - the Southeast Asian data -irerns''. in K. Omoto and P. Tobias (eds) The Origins and Pa.r qi Jfodetn : :-- : (:-mDor. world Scientific. -.:.:: Burung 2: an Upper Palaeolithic rock shelter in South Sular'vesi'

t9

THE ORIGINS AND DISPERSALS

OF AGRICULTURAL COMMUNITIES IN SOUTHEAST ASIA
Peter Bellwood

After 5,000 years ago, the world of the hunter-gatherer economy across Southeast Asia and on the western edge of the Pacific was confronted rather suddenly by another economic system: that of the first agricultural populations. New populations and languages, with close links with contemporary populations in central and southern China, began to spread across the region to form the foundation of the pattern of ethnolinguistic groups which exists today, beneath the veneer of rwenty-first-century global civilization
(see Figure 1.1). This is not to state that there were ever mass nrigrations of early farmers who marched

their way across the landscape, eradicating all former hunter-gatherer presences. The reality was far more complex, and must be modelled demographically in terms of what we know of differing farmer and hunter population profrles under differing environmental circumstances. More importantly, we need to understand the historical processes concerned in terms of the distinct and independent data sets from archaeology, linguistics, genetics and palaeoanthropology.l It is not sufficient simply to accumulate all the data into a vagre concept of "migration", and then to visualize something akin to the observations ofJulius Caesar on the migrations of the Helvetii into Gaul in 58 sc, complete with their families, pack animals, wagons and seed corn. Of course, some ancient populations perhaps did migrate in this way. But we must also think hard about how archaeological and economic items (e.g. stamped pottery rice, domestic pigs), proto-languages, and specific "marker" genes might have originated and moved through space, sometimes in fairly independent ways. Archaeologists, linguists and geneticists are currently very interested in the question of :.c*- agriculture spread after its initial appearances in various parts of the world. Was it :-,::::gh the demographic increase and range expansion of farmers, via a process termed :=::-: diffusion" by geneticists? Or was it through hunter-gatherer adoption of

=:-:-----,.. pottery ornaments of stone and shell, ground stone tools, spindle ,:.:.:--.-.-: :::re\ and village-sized settlements - all spread far and u'ide within various .:.:*:.. -- :--,:-::':cs during the early Neolithic of Southeast Asia, essentially

lr'hori, :,:

don'...-:--,-.:

.--:*:;re? The problem is, rather often, that the relevant archaeoiogical, linguistic and :::-::-: "- does not always point with perfect precision to one or other of these polar .:.-::-: l.-, doubtless means that the realiry was somewhere in-benveen. 'l--= .:-----.:,:.cglcal assemblages we associate with early agricuiture - the crops,

21

PETER BELLWOOD
spanning the whole area between 3000 and 1500 ec (and being generally earlier in the north than in the south). The foundation nrovenrents of the major agriculturalist language families of Southeast Asia (Figure 2.1), especially Austroasiatic and Austronesian, also spread far and wide, carried in the main by populations of native speakers of the protolanguages concerned, although localized instances oflanguage shift doubtless occurred as

well. However, genetic data in Southeast Asia does not point clearly to total replacement of pre-agricultural peoples, and this is true in terms of nuclear DNA markers as well as mitochondrial DNA and Y chronrosome haplorypes (the last nvo being inherited without recombination throush temales and males respectivei-v).2 In fact, genes sometimes give away processes of population admixture far more clearly than do languages. Indeed, if rve think a'oout modern situations carefulh-, it is apparent that even recent colonial episodes oi rrulv nass migration into the Americas and Australasia have

Figure 2.

Putative honrelands oirhe nra_jor lanEuaEe tamilies of China and Southeast Asia. 22

AGRI CULTURAL COMMUNITIES

not completely removed all traces of previous peoples. Native peoples of course live on, very vigorouslv indeed in many regions of the recently-colonized world. So, with this inrroduction, we can perhaps begin to consider the Neolithic prehistory

of

Southeast Asia interacrions.

in

terms of real ancestral populations, their movements, and their

The first farmers

lxrh)- did transitions to agriculture occur, all within the past 12,000 \'ears or iess, sufficient to shake the human world irreversibly from its rwo-million-\'ear Pa-laeolithic state of relative hunter-gatherer demographic equilibrium? The "bonom-Iine" answer, if we u-ish to search for ultimate causes, probably iies in the posr-glacral climaric amelioration after 15,000 years ago. This led, as a result of warmer and \\-ener climates, to a massive increase in the productiviry of the earth's tropical and subrropical surface. at least from a human viewpoint. The three most significant of the indigenous rransirions to agriculture in the n'orld, in terms of population growth and impact on later histon', occurred in the Middle East,

China, and the tropical Americas. These region-s are all in fairly low latitudes rnith strongly seasonal rainfall distributions. rr'here conrbinations of domesticable animals and plants, especially annual large-grained cereais and other seed crops, proliferated in the warmer post-glacial conditiorx. In che Nfiddle East and China, we know that some populations turned to han-esrins these resources, particularly wheat and barley in the former region, and rice. fortail mrllet and broomcorn millet in the latter. In good times, such plentiful cereai foods u'ould have encouraged a degree of sedentism and population gro\\th. But the s-armth \\-as not continuous and ever-increasing; unstable climatic fluccuations are knon-n to have occurred, leading to periodic quite severe cold reversals. Perhaps, as a result of stresses in food supply because of these climatic fluctuations, some groups in the Levant and China began to select for domesticated features in their animals and plants. Such selection, whether conscious or unconscious, increased productivity via a number of changes promoting better yields and easier processing. The results would have benefitted pure subsistence as well as those competitive aspects of affiuence, such as feasting and wealth accumulation, which have always tended to characterize increasingly large and sedentary populations. Thus, both social and environmental factors could have reinforced each other within this very important developmental
trajectory.

The first such development of agriculture with morphologically domesticated cereals occurred in the Middle East about 10,500 years ago, followed closely by China by about tr.(:)00 years ago. The Americas followed much later, with their first major domesticated :,:o,i crops, especially matze, being dated to about 5,000 years ago. The absolute reasons : -: :::e many indigenous transitions to agriculture are still not totally understood by .:::-..:-ogists and each region has to be examined on its own merits, but one fact

:-::--:::-: ransparently obvious. These transitions v/ere fundamental for the present r:::'--J:::::-- siruation of humanity - they unleashed a cycle of popuiation growth \,,r:j-:'-: '.r,:-::r none of the ancient or modern civfizafions, nor the problems of o',-i:f -': *-.: -:- ::: j ro\-erw in the modern world, would exist.

23

PETER BELLWOOD

Early agriculturalists in China, Southeast Asia and New Guinea

The islands of Southeast Asia and the Pacitrc u-itnessed a nvo-part development and dispersal of agricultural populations. in rhe Highlands of New Guinea, agricultural systems based on the drainage of su-amps anC rhe shifring culcivation of tubers on surrounding slopes had developed ind:S:::o::sl-,- ':',' a: ,east 6.000 years ago. But the absences of cereals and domesticateC ariJne-< rn rl:l-. rslani meant that populations were generally small. Pigs were not introduc:i :o \eu- Guinea unii about 3,000 years ago, and the American sweet potato, a ma_ior h-ghland tood source rod-ar'. perhaps not until after AD 1550 (although it might har-e beer a hnle earlier. parucularir-if ir could have been transmitted from prehistorrc Pol',n*s:a r..'here it s-as present tre-:r: about 1.000 years ago). The relatively low demoerapiuc proli: of prehistoric \eu- Gurnea socienes meant that
no agricultural colonizanon ot Ausrraija. the last completelr- hunter-satherer continent to survive until European co:ii:--r. e.,-er occurred. This remarkabl3 ta.t cic not reflect any inherent deficiencr- in the -\::s::.-.:i enr-ironment. Rather. rr rei:cte; rhe absence of an,v sizeable agricultural po::^,::or rn the malarial southern los'lan& oi Neu' Guinea, combined rvith the geo::;r,h:;a- circun$tance whereb-v the mrgranons of Austronesian agriculturaLists into Oce;rri; o:curred aiong the northern side of Nes- Guinea, not the less hospitable sourherr s:Je. After 5,000 \-ears:lc. to:h rhe mainland and the islands of Southeasc Asia eventually became incorpo:a::J l::o a pha:e of increasing agricultural activi[', characterized by the spread ot a basr:aj','\eohthic set of artefacts of types generally absent in Australia and the Ner'.- Guine; Hrshlands (apart from ground stone axes). To understand these trends s-e neei io eramine the beginnings of agriculture in China, the second but by

far the most siEr:i.can; of the fivo regions of early agriculture in the Indo-Pacific
regron.

ago. the archaeological record reveals the fint village-agricultural in the \-erlos' and Yangzi river basins of China. These people grew foxtail and broomcorn nlr-i.ler in che north and rice in the south, and kept domesricated pigs, dogs and poultn. Ther- made pottery used polished stone tools, lived in subscanrid viliages of tinrber houses. and knerv such important technological innovarions as kiln-firing for potten' (in the \eLjorl. River valley), weaving, jade rvorkrng and skilled carpentry. By 5,500 )-ears aso the agricultural way of life had spread through the coastai provinces of China and sourhs'ards through interior valleys to the Pearl Delca and tiwan. It u'ould be a cardrnal mistake to regard all these people as "Chinese", in the sense of speaking ancestral Sinicic languages and being geneticalh' ancesrral onlv to modern Chinese ethnic erouF\s. "China" at 6,000 years ago was an ethruc mosaic, presumably with early Sino-Tibetan-speaking societies in the north, and socieries in the south ancestral to the present Tai. Ausrroasiatic and Austronesian peoples (Figure 2.1). At 5,500 years ago it is most un-likelv that the major modern language farnrlies u'ould have been

Bv

E.5(,'{-r \-ears

socienes

as they are no\\-. instead, a time-travelling linguist u'ould have noted a nerwork of languages and dialects of tribal size spread over the whole region, from only some of which are descended modern language subgroups and families such as Sinitic, Austroasiatic and Tai. These modern subgroups and families have arisen because, in certain places and under certain demographic conditions, some early societies underwent processes of expansion. The oldest of them spread with agricultural colonists into a world

delineated

peopled by fairly sparse groups of hunters and gatherers.

24

AGR] CULTURAL COMMUNITIES

This leads us to an important observation about the totaliry of language distribution in Southeasr Asra. Fisure 1.1 indicates that Austronesian languages have spread like a blanket over Isiand Sourheast Asia, u-ith no surviving remnants of eariier linguistic landscapes u'est ot rhe \toluccas. This siruation only changes as one approaches West Papua (lrian r..i:: Paiuan languages having a relatively restricted presence in parts of Halmahera -Ja)'at. anj th. i"s:-rn Lesser Sundas. As will be demonstrated belos'. the Austronesian dispersal

a Neolithic phenomenon, and since then only s'ithin-family linguistic :;:-t:.Ee nrenr have occurred, for instance the expansions of Malar- and Javanese. Hou-er-er, the Southeast Asian Mainland is a veritable mosaic of many language =n:,riies. often rather chaotically interdigitated, particularly in the mountain regions 're nr-een the major rivers in the northern zone. The Austroasiatic familv evidently spread tlnt, to judge from its fragmented distribution, and the more recent spreads of languages of early civilizations, such as Mon, Khmer and Vietnamese, still render the Austroasiatic fanrily one of major significance. The spreads of the Tai (especially Thai and Lao) and Hmong-Mien languages, together with Burmese, have taken place into Southeast Asia very much more recently, indeed mainly within the past millennium or so according to historical accounts. The mainland has thus had a much more complex ethnolinguistic history than the islands, possibly because of the greater number of large-scale sociecies here, from Iron Age times (c.500 nc) onwards, that have periodically reconfigured the political and ethnic landscape through colonization, assimilation, and on occasion downright conquest. In this sense, while the foundation dispersals of some of the major language families undoubtedly occurred in the Neolithic, much linguistic rearrangement on the mainland has occurred since, especially in the regions of Indic kingdom formation. In the remainder of this chapter, I wish to focus on the Austronesians and their colonization of Island Southeast Asia and Oceania, since the question of the contemporary Austroasiatic dispersal through the mainland of Southeast Asia is dealt with by Charles Higham in the following chapter. These rwo language families appear to have been the first to spread with agricultural populations through Southeast Asia, aithough it is also possible that the Tibeto-Burman languages could have commenced their spread into the Himalayan and Burma regions with early farmers - we simply do not have much information on this family and its possible archaeological record at
\r,';.) ::::i:rrv
present. Austronesian prehistory gives us one of our best-studied examples of a successful, and hugely long-distance, spread of Neolithic agriculturalists well over half way around

=e circumference of the world. The Austronesian dispersal

l"=
.

': . -- :

=:,-:.nsion of the Austronesian-speaking peoples undoubtedly involved a real --:::11 process by ethnolinguistically-coherent groups who spoke early Austrone..- -.:-=-:iij. As the most widespread ethnolinguistic population on earth prior to the : - - : : -- : : -::::zations of the last few centuries, the Austronesians settled more than half ' ": :.-)-,-. , :::;umference in tropical latitudes, only escaping into cooler climates to
,: :'- ;:.: and Easter Island. Today they number well over 350 million people t:-.: : ::: -*-;.: 1.000 difterent, although related, languages. Their dispersal, a

:--:--e in various stages befween 5,000 and 1,000 t'ears ago. took them

-r',;,::'r and Easter Island (Figures 2.2

25

.:..::: in the region of southern

China and Taiu'an to places as far

and 2.3'1.

PETER I]ELLWOOD

Figure

2.2 Overail disrribution of the Austronesian

language famili:

Research on the prehistorv of the Austronesians has now become a multidisciplinary industry involving the s'ork of many scholars; not just linguists and archaeologists, but also social anthropologists, geneticists, palaeoanthropologrsts, palynologists and environmental scientists of manv kinds.r From the outset, it is essencial to remember that the Austronesians are defined as such because they all speak Austronesian languages. This circumstance alone is a guarantee of a high degree of shared ancestry and history but it is easy to observe that Austronesians today are not a unified population in terms of biological a{finiry - the speakers of Austronesian languages share a wide range of Asian and Melanesian phenorypes. Neither are the Austronesians a single "ryp." in terms of
26

AGRICULTURAL COMMUNITIES

(*
\j --\-

!"<fi::ry-' 7>-s i\

f -)

/"Li

\/\

| !r

-,r ---.

Figure

2.j

Approximate dates from archaeological and linguistic data colonizations across Southeast Asia and the Pacific.

for initial Austronesian

sociefy or culture. Essentially, we must be aware that Austronesian languages, cultures and

biological gene pools have not evolved locked together as a single entity through
prehistoric times, but have blended together or drifted apart from time to time according to historical circumstances. The languages of the Austronesian family, and the cultures to which they belong, share to varying degrees an ultimate common ancestry located on the Asian Mainland at least 6,000 years ago. This can be seen not only through the languages, but also (admittedly to a hazier extent) through comparative archaeological, anthropological and art-historical observations. Since 5,500 years ago the Austronesians have undertaken an imrnense amount of colonization, leading to cultural divergence and adaptation to new, strange and often highly challenging environments. Their colonizations also brought them into varying degrees of contact with the pre-Austronesian populations of Island Southeast Asia and Melanesia - people whose ancestors had settled in these islands more than 35.000 years ago and who, for example, in parts of New Guinea, had already developed :-.'-.:erns of agriculture independently of the Austronesians. In the case of Indonesia and :,.= ?:::lippines it appears that the pre-Austronesian populations were hunter-gatherers, ::'.: i:rise areas the dominance of Austronesian languages and cultures at the present .'.',:". ::--::: srronger than is the case in New Guinea. In this chapter there is nor space :'.., -:-.: :::i3 issues of interaction in depth, but assimilation of, and interaction tvith, :,- ,-__:_,^_ f---1,-:..-- :.:.i :: --l-have _l__,-l been _f _:- --:c_-,---:^es L-__^ clearly l___,- of significance in regions as far east as the '.,.':,:ch no pre-Austronesian populations ever nranaged to penetrate. ::' , '.: : : -:- r- --:r--:. pre-Austronesian, colonization in Island Southeast and the

27

PETER BELLWOOD

The foundation of our knowledge of Austronesian prehistory must clearly come in the first instance from linguistic rather than biological or archaeological evidence. The Austronesian languages form a coherent phvlogenetic grouping, with only very rare examples which resist classification, characterized by a great deal of shared and inherited core vocabulary and many other aspects of grammar and phonology. In order to study the prehistories of unwritten languages. linguists favour a process of comparison involving the identification and geographrcai piornng of cognate items in living or recendy recorded languages (cognates are features shared bv languages as a result of shared corrunon origin rather than borrowing). Ther-are then able to identift genetic subgroups of languages by plotting the distributions of uniouelr' shared and innovated cognate features. These language subgroups can be oroered hierarchically, as in the case of the most widely accepted classification of the hish-order Austronesian subgroups, that by' Robert Blust, shown in Figure 2.4. The historical implic;.rio:r oi Figure 2.4 favours a southern coastal Chinese origin for the pre-Austroneslan rools ot the famii;', at which stage the pre-Austronesian languages concerned were perhaps reiated in some way &y borrowing or genetic affiniry) to the contemporalJ- roots oi che Tai. Austroasiatic and possibly also Sino-Tibetan language families. Thus. che deep roocs of Austronesian expansion were sown several millennia before the Auscronesians rhemselves began to differentiate as an ethnolinguistically-distinct people via their nisrarions to and beyond Taiwan. Ausffonesians, technically speaking, have on-lr' eristed in an ethnolinguistic sense since the period of Proto-Austronesian, iocated in Taiu-an about 5.500 years ago. Earlier populations are, by definition, PreAustronesian.

According ro the Linzuistic evidence, the pre-Austronesian phase was followed by a colonization of Tais-an (the linguistic homeland of Proto-Austronesian), and then a dispersal southu'ards through the Philippines into northern Indo-nesia. From here, in the
general region of northern Borneo, Sulawesi and the Moluccas, later population dispersals

tOn\ \
FORMOSAN
(nine primary subgroups)
? Western MP

Central and Eastern MP

Central MP

Eastern MP

South Halmahera/ West New Guinea

Oceanic

Figtre 2.4 The higher-level structure of the Austronesian language family (after Robert Blust). 28

AGRICULTURAL COMMUNITIES
proceeded wesrwards to the Malay Peninsula, southern Vietnam and Madagascar, and eastwards into Oceania. These colonizations followed the development, presumably in the Philippines and Indonesia, of sufficient maritime technoiogy to cross oceans easily and regularly. Horvever, despite the glamour of ethnographicaily-recorded Oceanic voyaging traditions, maritime skills alone were not the sole cause of Austronesian expansion. at least not in the earlier stages (although the status given to navigators was an increasinglv important reason for later voyages of discovery in Remote Oceania). Systematic agriculture and animal husbandry were probably of far greater overall significance in the earliest days of expansion, since they would have been indispensable for the rapid growth of a healthy population and indispensable later on for colonizing

Oceanic islands with limited terrestrial resources. One of the earliest mainstays of the Austronesian economy in Southeast Asia was rice, although this crop was never taken as far as Oceania and many tropical crops such as coconuts, taro, yams and sago were added to the economic plant roster as Austronesians expanded southwards into equatorial regions. Concerning the eariy Austronesian economy in more detail, it is crucial to note that the reconstructed Proto-Austronesian vocabularya indicates a people growing rice and millet, with domestic pigs and dogs (and perhaps chickens too), canoes, knowledge of tattooing, and use of the bow and arrow By very soon after the Proto-Austronesian phase, in the Proto-Malayo-Polynesian phase of Figure 2.4 (perhaps located c.4,5004,000 years ago in the northern Philippines), they were building substantial timber houses and using pottery sails and outriggers on canoes, looms (probably backstrap rype) for weaving, and chewing betel. They had also added the tropical tuber and fruit crops listed above to their diet. Indeed, most of these ProtoMalayo-Polynesian additions (except for the tropical crops) probably existed in the climatically more temperate Proto-Austronesian phase in Taiwan as well, but clear linguistic proof of this in the form of pan-Austronesian cognates has not always survived since many Formosan Aboriginal languages have now been replaced by Sinitic languages (Taiwanese and Mandarin). In terms of archaeology, the signatures of agricultural societies with sedentary villages, Comesticated pigs and dogs, pottery polished stone tools and the remains of rice are ciearlv marked in the coastal southern Chinese record, extending southwards from Z'ne\iang through Fujian to Guangxi, Guizhou and Guangdong Provinces, and Thiwan, :.::j. perhaps into northern Vietnam and Thailand too, by about 5,000 years ago. Some of ::.::e earlv Mainland Southeast Asian Neolithic assemblages cannot yet be associated -"::-*r.' s-ith evidence for agriculture, especially in Vietnam and Thailand and some ::::--i:esions of southern China, leading Charles Higham in the following chapter to .,, .-.: ":: ihem rvith foraging. However, we have firm dates for rice in Taiwan by at least I * '".:"::-> ago (see below), and the Taiwan situation indicates clearly that poor
i; '

-" - '-.'-: be imprudent to state that all Neolithic

---

f'

cultures

in the region

\\'ere

: j:
:

-:

" "-""

",, r
'- -:

.ti*:: ;:*,:;l;*#;x',alilT#:j#?
.-,.: ::. :ottel]' or in phytolith form,
-.-"_-:.- :--onomies in the future.
29

-" _',,.1.0:::I #i',l,::l

:. -

lliiliJilT::'fr1ii:';:[ ::':trffi :i
should increase greatly our

PETER BELLWOOD
the precise economies present in southern China before 5,000 years ago, it is again important to stress that such earlv East Asian Neolithic cultures cannot be equated with Austronesians per se. Horr-er-er, it is interesting to note that many artefact
rypes rypical of later lsland Southeast Asian and Oceanic archaeological cultures are very widespread on the Asian Mainland as n-ell. These include red-slipped and/or painted, stamped and cord-marked sn'ies of porten'. knee-shaped wooden adze handles, untanged or stepped and polished stone adzes. arr sn-les emphasizing the uses of spirals and circles, and bones of potentially-domesricared pigs, dogs and chickens (the wild ancestors of domesticated dogs are not natrve ro Isiand Southeast Asia, and neither are wild pigs native to islands east of Sulawesi). An e::cel-3nr assemblage of this earlv and potentially-ancestral rype conles from the r.7.frr ,, i ',';,:-cli sire of Hemudu, in the los-land coastal region of

'Whatever

Zhejiang Province, r.,'here it:s assocrated rvith prolific evidence for rice cultivation. In Taiwan, archaeologrcal :u-r::es identified by such features made an appearance at some uncertain time ben.'.'e :: j - - - and 3000 sc. The carriers of a Neolithrc economy to Taiwan and the P'ens-h:: P:, --. j -,.:cs) Islands made cord-marked and incised pottery and used polished stone ajz<,. ,-.:= >:-ar points and stone barkcloth beaters. Archaeoiogical evidence for rice cui::'.':::':: i) ::-.nt in Taiwan by 3000 ec in the form of carbonized grains in the Nan-ku::.-.' :::= ::ar Tainan,s and similar er.idence for rice is already attested before 2',','- : c r.: :., ihe south in the cave of Gua Sireh, located in the equatorial latiruje r: S-::.',',':!;. Brrneo. At Nan-kuan-li the rice u-as han'ested using
reaping knir-es ot s:r=*.

Bv about i, i,t :5 :,-. rhe Austronesian population which had colonized Taiu'an. presu::r;:-'"' r.;^.--l:: c,r lncorporating earlier populations of hunters and gatheren. r',-a-. rsaj'.' :c =:-.-: i'::ther. One of the next steps according to the dated archaecio*lcal :=co::. '.-.-,il :::: -:::{uistic record in strong support (see above), seems to har-e been a sou:i:r'.'.r* a:li i'.:,-.-;:* move which incorporated, around 2500-1500 sc, the Batanes Isla::c-s -:j ::= :=s: ci the Philippines, eastern Indonesia and northern Borneo. Br'15', E: ;:-;-=:"i::ached the Mariana Islands of western Micronesia across an absoluri^',- as:l:;::::-: l.i'ri-tl<rn of open sea to the east of Luzon. Soon aftenvards. parts oi ls^:::j l'1--.:-=,:3 ir.'ere reached (especially the Bismarcks, Solomons, Nerv Caledonia. \lnu;l;. F:-': . "::j rhen, by about 1000 to 800 nc, the islands of western Polr-nesia Tonga. S.;-;,.. \I--llis and Futuna). This southwards and eastwards move'uvas undoubreci.'''one o: rh: :::cs: asronishing bouts of colonization ever to occur in early human histon. and as 'u',.e .e:.rn more about the archaeologically less-known regions of *'estern Aus::ole.ra ir ::r:-." :*rn out to have been even more extensive than we now realize (dares for i:.r:ra. -L':s::,:,ne sian colonization in western regions such as Borneo, Sunratra an,J fa','a are sr;,' ;::.::.i:r. but are also perhaps likely to fali berween 2000 and 1000 ec). In a perioJ of u;-j=r '.' , rrl\-ears, Austronesian ancestors spread from Taiwan, through isiand Sourheast As:" ::r:-. \lelanesia and central Oceania, over about 7,000 km of land and sea. In order to ,jc r:rs rher- had to develop systematic strategies for the exploration and colonization oi tirav":.-.- rslands, strategies well described by Irwin
and Anderson.6

In the islands of

Sourheast Asia.

rhe earlr- Austronesians moved rapidly around

coastlines already settled, as in Tais'an. bv sparse qroups of hunters and gatherers. In New Guinea, however, the indigenous peoples (as nored above) seem to have already developed techniques of taro, sugar cane and tiuir lbanana. pandanus) agriculture, at least in the highlands, and possibly intenstr-e han-esrinq of vanrs. sago and other tree products in 30

{GRICULTURAL COMMUNITIES

rv io,,'.'.:.ni rigici:-i. For various reasons little understood, but perhaps reflecting both l:- F:::': ::r:::.:---- :i -:ree populations and environmental factors (nralaria?), New Gul;,:; ::i::jr::'-',::i r::::ch of the Soiomons seem to have been relatrvely avoided by the ;:ri;;::.\;,:::;r:.i:r coionizers, although our knowledge of prehistory in many of these :;;::-"; -; i:* s;:'ect ro constant revision. Later on, Austronesian descendants filled in :l::-: -:-::= :'"'-';bie gaps, except in the New Guinea Highlands, r'u'here Austronesian
ntf
:

: l-- : --- :

-.: :-:'.':: :Sneffated.

is that this amazingphase of coastal dispersal from Taiwan to :- :j-: '. S.:-::--'a took place at least 1,500 years before the better-knorvn phase of Eastern ,- ,. -- =",..:. --oionization and long-distance voyaging during the first millennium no (see = ---:= -,-' . in Island Southeast Asia, the archaeological traces of this expansion of the : : - - :.: ald early first millennia sc are visible in a number of sites, mainly caves and : - --.'-s::-iters, with red-slipped pottery shell tools and polished stone adzes, found through
".=

::::.: i,: remember

:--.: Baianes Islands and the northern Philippines, Sabah (northern Borneo), the Talaud I'--:::is. northern Sulawesi, easternJava and the northern Moluccas (Figures 2.5 and, 2.6). -a, sire called Andarayan in northern Luzon has yielded rice husks embedded in potsherds jared to 1500 Bc, and, as noted above, rice had already reached Sarawak before 2000 sc, a-chough here in association with a type ofpaddle-impressed pottery quite difterent from the eastern red-slipped ware. At the site of Bukit Tengkorak in Sabah, the Neolithic redshpped pottery here dated from about 1300 BC onwards, was found with two rather surprising occurrences of a lithic nature; an industry of ag te microblade drills used for Cnlhng shell artefacts, and obsidian imported from sources in New Britain in Melanesia, lccated over 3,500 km to the east and probably representing one of the longest-distance ::ansfers of a Neolithic commodiT in world prehistory. New research in the Batanes l'-ands and the Cagayan Valley in the northern Phfippines has produced many sites with ii;mped red-slipped pottery and stone artefacts related to contemporary sites in Taiwan, :,::d in the case of the Pamittan site in the Cagayan Valley (northeastern Luzon) the dates ;.:r red-slipped pottery range back to about 2000 nc.7 Palynological results from lakes in the highlands of tiwan, Java and Sumatra also :rCicate clearance for agriculture from 2000 BC onwards, although the exact dates for t:rese activities are not very secure. As noted, dates for Austronesian colonization in the
-a:ge islands ofBorneo, Sumatra andJava, and also the Malay Peninsula, remain uncertain :-.r-u1g to the sparseness of the archaeological record, but dates in the second or first

::----ennia BC seem very likely. One of the problems in these large islands with their : :'.'.';rNil erosion regimes is that the oldest coastal Neolithic sites are likely to be buried -- j=: huge depths of alluvium, and so extremely hard to find and excavate. The Malay

: :'-----i '-.. u'hich still today has many interior regions populated by Austroasiatic::.-,,-:.; -\Lian) populations, was probably only first settled, in coastal areas, by '"":: -r,:---s less than 2,500 years ago. These Austronesians would have found -.--::'i::. ancestors of some of the modern Aslians, already in occupation since at
i..''r : -

::: -,;"-:,: probably only reached in the mid-first millennium ao, evidentlv by -- , :-*- r,:-:ihern Borneo, perhaps with Malay- or Javanese-speaking leaders ;::-:- a:_ :::.:: -nzuistic analyses.s In the other direction, the furthest reaches of :r:i:rr. -'"--.': --: .. rl::olines, Marshalls, Kiribati) and eastern Polynesia (Societies, .'u1,::-:,:r ... ' :. , E,,::: Island) were reached at Various times betrveen 500 nC and au ::-l :::.-,:: _:_'. i:-.'ro 1200 in the case of New Zealand. . -. -"37

Figure

2.5 Sites.(in italics) with Neolithic red-slipped pottery in

Island Southeast Asia.

06m -t_tFd

A selection of dentate- and circle-stamped pottery from southern China to Melanesia.

-ill

sherds except a) are red-slipped and have

lime or white clay decorative infil.

Xiantouling, Guangdong coast, China (c. 3000 Bc). Courtesy Yang Yaolin.
]'1:eapit, Cagayan Valley, Luzon (1500-1000 nc). Courtesy Yoji Aoyagi. c) Yuanshan,

Tairvan (1000 nc). Courtesy National Taiwan Univeniry

d)

Nagsabaran,

l:: \allev (1500-1000 nc). Courtesy Tsang Cheng-hwa. e) and h) Barungan Cave, --erral Philippines (800 nc). Courtesy Social Science Research Insritute, c: Harr'ai'i. f) Kamgot, Anir lslands, Bismarck Archipelago (Lapiu - 1300 . G.:nn Summerhayes. g) Lapita (Site 13), New Caledonia (1000 nc).
=:.

::--:::: Sand. i) and j) Achugao, Saipan, Mariana Islands (1500 ec). B::.::. k) Bukit Tengkorak, Sabah, 1300 uc.

tr+ffi.grtt

-rl-

r:rfi{3 W 'trj i<r -,r^q

I)ETEI\ iJELLWOOI)

did the initial Austronesian dispersal occur? At this stage it is necessary before cor.rtinuing the record alter 1000 tc, to l]love back to the linguistic and archaeological er-idence in Southeast Asia to consider just what
pronroted all this movement and actir-in' into lightly settled. or, in the case of Oceania bcvond the Solorrlons, totallv r-rninh.rbired regions.Whl'did the Austronesians, like many peoples in history not sirnph' sir). .ri honre? In fact. the Austronesians were not particularly unusual in beine co.oiri.t. since the developnrent of agriculture allowed population expansions to occur rn njnv regions of the s'orld. The nrain claim to fame of the Austronesians is that thcir r: r:::!1ons ultinratelv crossed such vast ocean spaces. They were able to develop their r:r-r:::::::: skills for the sirnple reason rhat their l,onreland zone was located in a strategic lc:,i:r;:. close to the Asian end oithe greatest series of island chains in the u'orld. Ch"n:= j:;rr a eood geographical hand to the Austronesians, a chance which thev dr; i:.i rji ::-. Of course, therc rri:: .:-'.':r-r:- s:eciflc reasons s-hr-indirri'.:;, -\usironesran colonies were founded. Onc:::-:r::.:-:,-,,-::l reJ.son is that the process of coloruzation itself can increase the statu: c: ::- ::.i.-,.:i-:,' and his/her descendants. -\ficronesian and Poiynesian legends (e.g. the ir::-^,r'*! -'.1,,-,:: "Fleet" traditions) are tuil of the exploits of founderfigures, after r','rl::--,:::-- -:--:!'.'.'.re named and u-hose direcc descendants still form
aristocratic line .==s
.,-

.Why

j.-, S *,-- , "rcunder-focused

ideologr-"e u-ould ahr-avs be sufircient

to fonrenr a i:,::=.t

:::--:-,-:::-=-,-,-:erritory. In addition, some smali islands in Oceania nriqht alsc :r,-.-= :::::r-.: ,-..::::,.'-:,-,:d quite quickly, especiallv if rve think of birth-rates sinular to Lr,:r:i i..;;:-" ,:: l.:::,: E*::'pean coionists into fertile, relatively disease-free and onh'lich:l','rcru-.::c -=:rj ::-.'':s such as the more heavily coionized parts of Australia and Nor:h -\n.:r:-:, l-..-:::r:-:::i'.--n children per family in European Australia in the 1S-l',rs ani 155.'s. Tn: B::::::. -\luineen and their Austronesian (Tahitian) wives in Pitcairn recor,jeJ si::ii.ar..,- s:e ;:.--;1.r grou'rh rates. Yet overpopulation was clearly not the real reason ior Ausi:cr:c!:r.:. :i:la::sion. certainlv not in the huge islands of Indonesia and the Phihppines. and :ri:::r :::-s ri:opulation growth might have been the results rarher than the caLrses ol s-:.-i .,:=::-,:-t.. population dispersals. The nrain tut,ltrli'trtg rea:on tLrr ::r- -\:srronesian diaspora, the sine qua non (albeit not necessarih' tl're nrain inrnredlaii r.asor. :cr all individual population movements), was surelr- the possession of an efrlcieni aEricui:;ra1 econonx combined with reliable access to lneat sources. especiallv in the tori: ot lo::r:sticated animals and fish. A successful agriculturai econonn-is a s-av oiiite rranslor:ab-: ro anv neu. and suitable environment. Hunters and gatherers cannot colonrze r:.J p=:::r;:::nrlr'sertle snrall isolated islands on a continuing basrs u-ith gr,raranteed success: erile r :1:re rs srnrph- insutficient food to suDport a viable popularion. The Neolithic inhabitaiirs ot sor-i:h.rn co;lsral China, betrveen 5000 and 3000 BC, bv der-eloping both s;'stenr;rric tood producrion and also successful methods of maritime trar-el. s'ere able to put their trernen.ious porenrial for denrographic growth to rvork in nes'and u-elconring environnrents. Once colonization began, the only process to stop it, apart frorl geographical or technoloqical barriers, rvould have been other populations equallr/ empo\\'ered rvith demographic and cultural muscle. The early Austronesians of course met such populacions: the Austroasiatic speakers oi Peninsular Malaysia, Vietnanr, perhaps parts of Borneo, and the Papuan speakers of Ner.. Guinea. Such prior populations, rvhen strongh'entrenched like the Papuans, delaved rh= process of colonization and in the case of interior Neu' Guinea made it non-cxi:::r-.:

E4'w y
/f(1! J
.

.:

',' r

*i..[d[tr.i

A(;I\ICULTUI{AL COMMUN ITIES

prior to the period oirrrodern Ir.idonesian transinigration. Austronesians have only settled rrrirror pockets of coastline around Nerv Guinea, seenringll' only s'ithir-r the past 2,000 years according to borh archaeological and linguistic evidence. Indeed, the inrpressivc donrinance of the N4elanesian phenorype in the western Pacific indicates strongly that
sonle of the Austronesian languages of Melanesia spread via processes of language shift rather than colonrzarion. In other words, sonle coastal and island populations, descended geneticalh.tiorrr the initial colonists of Melanesia, also nrigrants frorn Southeast Asia, but over 35,000 r'ears aqo. abandoned their native languages to adopt the incornine languages of the Ausrronesians, in many cases at relatively early dates in the trajectory of Austronesi:in dispersal. Yet it nrr.rst bc renrenrbered that this kind of language shift, while conlnror.i in sonre regions. s'as probably rare in the totaliry of Austronesian history. It is not possrble to explain the r-ast extent and relative homogeneiry of the Austronesian famil1. of languages by a process of language shift alone. lJnmoving indigenous populations did not sinrprlladopt sonre fbrrn of early Austronesian language and abandon their own vernaculars all the s-av iron-r Sumatra to the Solomons; there are no known social mechanisnls at a tribal social ler-ei g'hich could account for such a phenonlenon. For instance, norle of the historical empires - Roman, Inca, Mongol nor any others - were capable of such a feat over such a large area. English and Spanish only spread through such vast regions because of deliberate and sustained policies of colonization by European powers. Also, there are no widespread tndrcations in Austronesian language history (e.g. strong traces of substrata or interference through shift) that such a massive process of language shift ever occurred. The dispersal of the Austronesian languages rnust be, at a guess, for at least 80 per cent of the geographical distribution of the family, the result of coionization by Austronesianspeaking founder communities. This is one of the greatest, perhaps tlrc greatesf, human dispersal on record since our cofiunon ancestors left Africa.

There remains one unanswered question. Why did the Austronesian

dispersal

ionrnlenc e when it did? As noted, we cannot trace back the early Austronesians directly :o rhe earliest farmers in central China, so some kind of "sealed" origin within an early :r.-e-cultivating Hemudu-type Neolithic at 5000 BC would not be a very viable -..':othesis. The main Austronesian migrations comrlenced quite late in terms of the --:::..:iolosv of Neolithic China, at about 3500 BC or later, with the main phase following lc. Agriculture had already been in existence in central China for 3,500 years by :. -) ::::-. The factors which allowed the Austronesian dispersai to occur - the agricultural - -: ::r::::irne skills, the demographic profile required to encourage expansion, the social ,:- -:i :-- i:ririate colonization - did not develop overnight and probably increased in :,.: ,: :.-.3 expansion process unfolded. It aiso seems likely that "domino effects" ilt ::--::ht have contributed, rvith rising agriculturai population densities in China er-en pushing, populations rn more peripheral areas into movement. Robert suqgested that the Polynesian nrigrations could have been stimulated bl' the - i...:rJ po\\'er in Bronze Age China after 1800 8C.10 We know now that Sr.rggs' :: : -- -- recent to explain Austronesian dispersal as a whole, but the general idea

-i ,-:..:. Humans then, as now were inexorably bound up in nets'orks of :-. -.:.-r: - r nrajor rearrangement ofpopulation in one region could have had .'. :.,i ::.: :::rnediate social horizon. .-.-,: :. :-:.:: -\'-r.rronesians since the "big bang" of colonizartotr fl-orrr the : ' ..'. '--,.., .:-. :i.: second millenniunl uc has obviousll'been contplex and
35

PETER BELL'WOOD
detailed, the more so because, as with the afterrnaths of all great population dispersals, the archaeological and cultural records simplv become nlore regionalized everywhere with time and so more difficult to read in ternrs of long-distance correlations. People clearly stopped moving very far once the initial phase of colonization in each area was over, as one might expect. Settling in became the norm. aibeit s'ith some continuing degree of trade and intermarciage, often over long distances (s-irness the extensive trade networks of the New Guinea region last cenrun. a-lberr in sicuations of remarkable cultural and linguistic heterogeneiry). For insrance. Island N{elanesia at 1000 BC was extensively occupied by culturally-homogeneous bearen o[ the Lapita culture, who probably spoke only one of a few closell' related Austronesian (Proto-Oceanic) dialects and made very similar rypes of decorated potten' iFrgure 2.6) and shell ornanrents across an island region almost 5,000km fronr \\-est to east. But by'ao 1000 the \{eianesian islands were dotted with literally hundreds oi "seccled-in" local cultures and languaees. despite the existence of trade and interaction. \\e see the same pattern in the later prehiston' of eastern Indonesia, Polvnesra. or u'herever \\-e look in the Austronesian u-orld be1'ond the influences of the relarir-elv honrogenizing Indianized and Islamic civi-lizarions. Initial population dispersals are rherefore much easier to describe than their ultimate regionalized producr-.. and the intensely diversified ethnographic record of Austronesia, today u'ith nrore thrn 1.r,t[rtr languages and cultures, gives a graphic picture ofjust how much dir-ersificanon ha-. occurred.

Sorne highlights of later Austronesian archaeology

The Ausrronesian colonization process must have passed through all significant coastal region-. oi Island Sourheast Asia during the second millennium BC, perhaps earlier in the Philippines and. oi course, Taiwan. After this time, despite the tendency to develop culrural dir-ersin, the region has seen a number of periodic but widespread stylistic horizons. The rnost significant of these is generally dated to berween 300 nc and AD 700, s'hen nrerallurqr'. s'idespread sryles of decorated pottery and the beEnrir$ of trade contact s-ith India and China spread through huge parts of the Indo-Malaysian Archipelago and adjacent regions of the main-land. The anceston oi the Malays and Chams (southern Vietnam) probably owed their arrivals on the Asian Mainland to the actir-ines of this phase, and by its end the first Ausrronesian migrants had reached Madagascar. Late in the first millennium AD we rn'itness the begrnnings of another very rvidespread sn-listic horizon, this time based on the import oi glazed ceramics from Chinese krlns ro all parts of Southeast Asia. The circunrtances surrounding the arrival of Indian contacti. iron metallurgy, glass beads and other e\otic trade items in Southeast Asia are touched upon in several of the following chapcen. In Island Southeast Asia, bronze and iron mecallurgy appear to have arrived together, perhaps soon after 300 nc. By the first century AD, the presence of plentiful sherds oi Indian Rouletted 'Ware and moulded ponery at Sembiran in Bali, together with a sherd graftito in Brahmi or Kharoshthi script, inform us that contact with India was well undenvay (Figure 2.7). Just who travelled where, of course, is hard to know, but it is possible at this stage that most movement was by Austronesians, who after ail :nvere soon to colonize Madagascar and perhaps to carry Southeast Asian crops such as bananas and taro to eastern Africa, from where they were incorporated into'West African
agricultural systems. Yet the graffito also hints at an Indian presence,
36
as

also do the oldest

{GR]

ULTURAL COMIUUNITIES

ffral:1wry arJ

Figure

2.7 Indian pottery from the

Sembiran

site, north Bali: Above left: Rouletted Ware; Above right: mold-made pottery of Wheelert Arikamedu type 10, stamped with a bird design; Left: a black slipped sherd with a grafiito in Brahmi or Kharoshthi script. an 100-200.

inscriptions in Southeast Asia, all in Sanskrit and some dating

as

early as AD 400. As noted

in the introduction to this volume,

Indians never colonized Southeast Asia, but the impressive Indic temples constructed after the seventh century could hardly have come into existence without considerable Indian knowledge and bodily presence. From the indigenous perspective, the period of initial contact with India was a period of considerable efflorescence of sryle and material culture. Not only did bronze and iron come into common use, but massive bronze drums ofDong Son sryle, mainly originating in northern Vietnam, spread rapidly down the length of Indonesia, particularly along the islands of the Sunda Chain and ultimately to as far east as the Bird's Head of New Guinea. Sulawesi and the Philippines have so far produced almost none of these drums, and they are rather rare in southern Vietnam. But all of this makes sense when we realize that

southern Vietnam during the Iron Ag. was the home of the Chamic-speaking -{';srronesian populations, close linguistic relatives of the Malays and, like the Malays, :.-:=..-ed to be of immediate origin in Borneo. The Dong Son drums themselves, :*..:.-:;.rllred most vigorously in northern Vietnam and Yunnan, were probably made -r:r ::l: aegls of lar- or Viet-speakins kingdoms wnOse tra0e an0 a[lance [nKS \\'ltn *- l:l :i: aesis oI Tai- Or Viet-speaking kinedoms whose trade and alliance links s'ith is Vlet-SpeaKrng KlngOOmS r:": :: l"larnland SoutheastAsia appear to have followedinterior riverine rather than "-:
-

- : * :. -:::

-rL.,-

**.-:1.^ ^^-!:-.-^--^ ^L^:^fl^-J -, ---::-- ;'-::;;.-]',,-a continuous chain of land. - :::-" '.'":::l the spread of these drums went the spread of beads of coloured : --:::---=:.-'"' of Indian inspiration (where carnelian and agate beads were

;:-:,--'

^- in Thailand, T ^^^ and'West Malar-sia. TLrus, Dong C^.^ drums are cofirrnon :-'rL^:l^--J Laos ^-l \Y/^^* ^i{^1^--^:^ Son -J-.-* ,::ead easily into Sumatra,Java and the Lesser Sunda Islands since these

-:

',:,:, .Jr -

u'itness their presence

37

in the Roval

Graves

of Ur in

9ETER BELLVOOD Mesopotanfi), but sometimes of

lool

manuf.cture. Local tradirions of bronzeworking

sprang up on the Southeast Asian MiiDland andJa!!/Bali, chancterized by large bels aDd "flasks" with spiral surf2cc dccoration, a larien of bronzc weapons (spcarheads and daggers). and kettledrumi in the spcial Pejeng sqte manufactured in Eestern Jaw and Dali, thcsc bcing rather difrerent from the more sundardized "Heger 1" rypes made in Vietnam and Yunnan. Many ofthese obj.crs in $es(e.n Indonesia come uncertain archeeologicel contexts. but sonlc ofrhe drunrs \\.re used as bone containc$ for burial in sites in Ja!.. In recent years, knowledge ofBronze-lron Age dc\?lopmenrs in Indonesia has grown

fro

Drost quickly in thc Borneo-Sula* es i-Ph,tipprn es r.gion He.e. ar noted many yean ago by Solheim, nruch ofdte poften oirhi pe.iod b.h|een 5 x ' B. rnd AD 11100 is intricately dconted by incision and stamprng. Ii h3, been termed the Sa Huynh-K,lanay Style" by Solheim, aftcr nvo sires in lou&ern Viemam and the centr2l Phdrppines respectively. Huynh-Kalami regionrl !:mrh.iRes subsumc the Philppjnes. soulhern VietnaDr aDd Sulawesi, $ith extenljon: ia5rern Borneo and th. \lolucc,!. The) reltecr sryle diffusion and trade rather 'iro migranon.,nd thus cannor be considercd a sirnilar phenomenon ro rh< srre:.i oire.i+t.rpped and incned/stamped po en Nith Austronesian

rh;

colonization 1,51{

\er^

b.rore. Bur rhey do indicate the sigmicance of inrcr-nhnd

relationships 2.arr, rean ago. puisaring in time with much vasre! Iinlages as Rome, India and Han Chinr tirt snnrLrlrted the global economy on the largesr scale. up ro rb.t tin1e, in hurnan hilron.

Acknowledgements

i \!rsh io rl--art JeDn! Shechan of thc Cartography Unir .t ANU for dnwing Fisures 2.1-2.5. Lln Schnudr and Dominique O'Dca of Archacology and N3rurel Hitory at
ANU dre$ Figure 2.6. Notes
Se. B.llnooJ:,{ laraelhvood ald Renfres 2003 on th6e isu6. S.e Huier lnr,i tbr. good dlcussion ofthe senetic hnrory ofAustrcnesian dispeFl3 Sune!5 of rhe !-u1( of somc of these disciplines are prescnted in Bellwd.r dt. 1995. ,{ Vocabuhd reconfuctions for Prcto-Au$ronesian are disc$scd bv Blust 1995.

'

5 Tsans Ch.n.s-hsi. p.don l comunicario!. 6 See Isin 1992r,AndeBon 2000.

7 Se.

Tanaka and

o.oso

2000 for Panirtan.

8 This boot do3s nor cortr thr societis and hnton of Nl1drsr5.r s?n rhoush rhe major popuiation! ih.r. aR'Au$rcresian. Fo. soDe recent ling!6t!c and ,r.hreolosicai obs.rano.s see Adelaar 1995: D.s"r and Wrighr 1993. 9 Fo. a discusron oathn in the A$trcnesian conrexr. 5ee BU\ood 199arb. l0 Sug$ 1960:::5-{,.

Select Bibliography

Eaq agi.t It

rc

- genudl

Bcllw@d, P (1996a) "The origin, rnd spread of agri.ulrure h the Indo-Pacific resion", iD D H ris (ed.) Tlk Orirnr dd S!rcad oI ASnd n@ d d P.sr dlk t in Eirdk, pp. 465-98. London: UCL Pr.s.
38

ACRICUlTURAL COMMUNITIES
Ddl$ood.

UellNood. rl " 'J CJ,,rbr'Ji(.

.lunrt P

l:,.i' lr

- Errl\ r$n.uIurl,{ popuhrio,, dulpohs? Fa'tu,ng. hnguages rnd teDc" 1N ,,t.1 tln,\,,t,.qr tn. t8l-207

Denham, T. P. Harberle, s. G., Lerrfer, C., Fulliser, R., Fi.ld, J., Thftin, M., potch, N_ rd winrbo'oush. B. (:003) Origins of.griculru.c at Kuk SNanlp in rhe higbttrlds ofNew Gunrr". Srnii l0!: 189-93. Gokon. -l !n.i (;r(incr I) (1990) "Asricukure rnd sociopolitical orslnisroon in Ne\a cuire.

I .r..-.1. \ .(dr t:,'u}) I:'oa'nrng tk rdn t R LtLrtdr( Dr\\al IlrDo Jd lr!rute for Ar.h*ologl(rl Resed<h.

Hwottt.,\.

H,cnirnJ, frehno4", ,,lflNa| Rcvitu' oI An 19.395117. "opolag L!. T.c\ Lk--d$ (9 ) 71( Td ta Fatuzi s to Fa nu a l rtt oqh oJ ArndttsE i,t ...... .]^'b'd: DAR lntelurioni Srri6 774 'itiot itlalo,,. C (cd.) (1998) "Sp.ci.l Sedion: Rice Do,r.sti.etion ., .:ln ri{rnt t::857-907. lrr\-(. r D. n,d Gel'aucr, A. A (edt {legs) itrr Hrrt.t Fn! Fan".t:. sinb Fe: School of

.{n)eri.
(

Snxth, B.

99s)

Rcse.rch. nr EucEa@ oI

Arndnis, New York: Scilriarc Aherj.an.

Axstuonesia
Adlaar,

otl'u

arcas oJ

liry .tk

histont

(1995) "Bornco s ! cro$-roids for.ompanri\r Austrcneshn linsunrici. in I B.lhlood. FoxJ J rt'd T4on D (edv (rv"it tl . .1,,.no , t r,.: Hi:rattut dti Cd,ra-\r n,'/l.h.r. pp. 75-qq. Crnb.'n. Depnrn)rn, of Ai,hropolo$r Re(c,rch Srhoot ot prcrtc rdd Anrn studrs, Aurh"Lirn NJnonJ Unnrr!nl 8-]'wood. P. Fox. J j 3nd lryon, D. {ed.) rlqoj, Th. qu)ftone'16. Compahtrve ,nd Historical Perspectivcs". Canbe'n: Delartlrenr olAnthropology, Res..rch Sch;ol ofpecific rnd Aqrn Siudie\, Au\rral.n Ndonrl UntrcDin. BlLrr. R. A (1484 5r. Ihc A,Nronrlrn l.un(h',1 J ling!6u< peLpecn\e'. Atia p.A'a\r.. 26: 45-$7 . Blust, R. A. (1995) "Thepehi,tory ofdle Austroneian-speaking peoples:, view from Lnsu.sc',, lo"nat .I Wa d Pt.hktory 9: 453 5\o. Blust, l\. A. (1996). A *ronesian culture hirtory: the indows of lanslags", in !r: H. Coodenough (ed.) PtelLkbtu Serte'"cnt oJ thc pdtij., pp. 28-35. phit dcl hia: Andc.n Philosophical Sodery collins, j. (1998) ll.t4l,X Wold taas'Msc, Ku 12 Ltmpu.r DMn Bab.sa dln pusraka. Drietu, _G. vrtr (1999) A new theory on rhe origid of Chirese", Bt\ctitl 4tte tido,k'tiIt PEhkfury Asqidtion 1A: $-58. Dltton,l (199s) "LlDsluse conta.r and ch.nsc in Melrneiir", in P Bellwood, J. J. rox and D Tryon (cdt 0995) Th Arxo sia s: Hitbn.al drd Can'pdnti,c Pespecti!4, ip. t92-2t3, c,nb<(J: Dcprnment ol" l,nhropolo$, Re$r.n School of Prci6c .nd As,,n Studres,

A.

AusFrli:n Nnronrl Unive6iry. M. (2003) 'Ce the hypothesis of Lnsuase,/asriculture .o-djrpe6al bc rdr./ wirh Archreoscncti.s? ', nr P Belrrcod and C. Ren&crv \e&) E ahhtinr t Faf,"iis/Ina a!. Diq ?6dtHtpo tsh, pp.299-31, Ca'nbridgr: McDonid lirtuute for Archaeologicrl Resa.c6. Prvlcy, A K. (2003) 'Thc A!kon$;n dnpe6al: lansu,ses, tecbnolosies .d; pcople", nr ll lrlh\ood and C. Renfrelv (cdt Fld trlr$ it. Fdh k!/Iingntc Di\pqsal Htpotn'n, pp.251-74, Crnbridger M.Donald Idsti tc fo. Archaeological Rekrrch. :, , l$r A. m,d Ilos, M. (1993) "Ausronesian hntoricil linsuistics and cuhure history'., A dl
Hurles,

R.tiir

oI

'topolos

22: 425-59.

Neolitltit archaeolog

- Mainlad Southeast Asid

1::!: P.r:rar,.t 32: 3740. . :- : ; 2 Er t Cutu rct t'J Mdnrldkd South.a.t Atia,Bangkok: Rncr Books. ., : . . . :.:. IL t 199q Pthi'htu Ttdik\l: Itu r .d/i, r.I/la'r.fl ,r, srllj.lli ri. Londor:

: ,!1, 'Cukunl and biologic.l difcrcnri.tion in peninsular Mal.ysia: the hsr I0.000

39

I'ETEII IJE!LWOOD
Neolithk al.hdeolog,

- kld

d Sorthcast Asia and Oeeania

C. Sathcr (edt Oifin', A c,r) d,d .tl/ifli.. ff 118-lr). C,nbcft; Deparrnrcnt of Anthopology, Reseirch SchooloaPa.rlr rri {ir Srud,.\. Alsnlian Nadonat Univc6iry Bellrvod, P (1997) oflk 1,,i.- rljr.:, J:.rj .l lnd edn. Honolulu: Unive6iry of
Bellwood, P (2001b) Southerr AnJ \.ohih,( (.ds) Ei(rtlo\dc.lia oJ Prehddrr, :.1 I Ej!

And.non, A. (2000) "SloN boats front chinr: h(ues I d1c prehnroN oftndo-t,acific seafari'rs", in S O Connor and P Verh (ed.r fd.,.illri,.. L{.. pp ll-50, Ronerdrm: BrtkcnD. Eellwood. P (1o9ob) HicD,.h\, found ideolor Jnd Au.ko, e.r exprnsion . in l. For rrd

B.lhrcod, .4rnror6id,

r d Erdr B:onz. . 1tr Il Peresrine and M. Ember .lr rJ o..r{,.:. rr 187 106. P (in p.6s) Formon. r?:,!:oa trnd A!$ri.*::r i:.eFrl . n, D Blund.ll {cd.)
hrdr.
Tiipe': 5\lC

"rrrnro,l

'r,,:ri-tr.

Chang, K. C. and Goodenouch. \'\: H ilr9.r) Archacoiogl oiSourh.rr Chin! and ift bearlng on the Auir.oDdnn honebnC . r n \\l H. Goodenough teC. P,.i:!: r.t Sdrld or .I thc Peija, pp. 36-56, Philadehhr: .l,ier);J. PhjloJophi.al Soo.N

Ipoi Dabn af,d Behrood.

th. In'la-Pdtiht Ir.lti:!.'", l:r..n.,rri 10: 336+05. INin, G.J. (199:) n. P'i;::i:r.: E\r,l.r."rr' n/ C.i.r'i:irt.ri .,arli n:.,i:. Crnlbridge: Curbridgc

I) 19!1

llc.entrsrarch

ar

cu

Sneh

r,.i Lu-br).-.tDgri . Brll

oJ

V (199: 'll:. Ll.:ij P..t/.i, O\{ordr Bhcksell. Kich, P V (2.,r O', r1,. R.!J.irlr. II'il,,iJ, Berketeyr Univesitr ofcatifornia Pre$. Solhei!\ $l G II 19'r- TM por(e^ rditions ofhte pr.historic nhes in Sourheat Atir", in F S. DFle .i S.nr.!r rn Hitnn.al, Arhd.olori@l dd Li rui.]i; Sltli.r or Sidltu Crt,/, Sr,r/.i,.1 : :.j i. F 't A.rt Ppg,"n, pp. tr 22. Hong Ko g: hong ko g Un^eF.h
Kircb, P
Sprisss. \1 Ie r lr:. IJ.,,J lIdd,.iid,r, orford: Blacklvcl Sussi. R. rrt6 1-t. Liad Cnili.atidf of Pol),n.ri,, New York N.w Amcric.n Libnry Trnala- K. JJ oroeo. ,{. B. (2000) 'The archacolosical excaurion ar rhe Parlitran 'ne",J'r,'dl .'/ Edtrr,.trrl Srrdi.r 8: 113+1,J.pan: Keiei Uni!6ity. Tsng. c-h il99i) Nes archkologic,l d.u 6oD botb ,ides ofrhe T.iwn Str.its", in p.J-k. Li, C'h. Tqng. \'-k. Huatrg. D-.- Ho ard C-t Tseng (eds),{rrroErid, sr.dB R.ldrtrf b 7:: ;',. tt IE5-216, Taipei: I$ri!!re ofHnrory and ph,iologl, Academii Slrica.

Uoive6iN

Prcs!

Post-Neolithk drchacololl

I'fu Kc l.dna s aJ So,rl,.d! .l!r. Rofi.rdJn: Brlkenu. D s. R E ind \tisht, H. T. (1993) "The culrue hEto6 ot \lrdrsscar", Jd!',"/ d t/o'h rrdti'ia r' llj 56. Clor(r. lriSrn,. B rlooJ) Thc Brcnze Agc rn southr"..i.D ,(. drt'nC.nd '..oF,i,ion. rec.nr ,eqrch-. \k, n l t nfr\ \n! lA/2 11 7t. Highi r. C (l99arr 71i Btr z. A!( h Sa ltul].4Jid, Cr l\nJ:.: Crorb.idge Unilersir\ Pre$.
Bcrne( Kenrpc6. A. J. (11)88)

40

MAINLAND SOUTHEAST ASIA FROT\,1 THE NEOLITHIC TO THE IRON AGE

Cltatlcs Highanr

The period covered by th's chiptcr witnessed a series otmijor cuttur.l ch.rrge; on rhc nr:irrlind ofSouthe:rst Asir which were ukiuutelt to ser the stagc for d)e derelopnrcnt oi drc fi$t civilizxtions. There are fo$r .rjor phrses. The firsr involved l)uDre6 .rn(l gathcrcn, some of s,honr lived relatively urobile lives in thc forestcd irxerior. $,hil! others occupied the rich coasts. A m.Jor chrngc occurred duLnrg the dird nillen,r[un tuc rvhcn rice farrres, rvho raised d.Dcsric cattle iDd pigs, iD6lr.ted rnd occupicd thr nrhnd rivcr vnlJeys. Their o.igin is not knorvr rvitb certrilq,, but present elidcDce dr:wn froDr both archaeology anl li.guistics points to aD nhimrte homeland in rhe nnddle reachcs of the Yengzi iaur:'. Behveen 1500 a d 1000 llc, these agricultural conun'rnities became fa,nili.r \;irh .he properties ofcopper and tilr. and began ninnrg, s elting and casting copper rnd oronze arteicrs. Agiin, rhe origirr oftlrs Bronze Age are not firniy esublished, :)ut nre accumulariig evidence points to stirnulus from Chira. Thc fourth phase invoi'"ed the adoption of iron rvorkiDg, the earliest forged iron implements being drted to about 500 Bc. Until recently, the vital rrillenniun fronr 500 Lc until Ar) 501!-l little knoNn, but a series ofexcavadons .re no$'begrnDng to ilhrninate this as a !'-riod ofdeep-seated cultural chrnge. This period ofprehistory in Tbailand and Vietnam lay thus be dividcd into :tn eirly period of hunting and gathernrg, followed by succesi'e Neofthic, Bro ze and IrcD Ages. thc latter being succeedcd in the errly centutr:s by the rnidrl enrergcncc ofthe

li.(oricil

statcx and civiliz:uions.

::irelce -.:rnlin

^D The priic;l,al features ofe:rch lvill be consid$ed b) to uujor represent.tivir sitcs. discrssing ho$ rhese harnlorize witlr lhe of :rccumulating e rioence.

T,opical hunters and gatherers of Central Thailand :::::r and grdreren ir tropical Irtitudcs ,re usudly accorded :r poor p.cs rr :: r::rr more afiuent tempcratc cousn$. Thcrc is i Nidespread $arDfrion ', : ::i:::rh on rhe move seeknrg food, s,ere not cndoscd snh rron'thrn -. :),'<r'-l cuhu'c rnd l(d s'nrplc. ntherwkrcl,.d L<. llrr' sa. :::.1 iertiitty in Southe t Asia, Nh.rc etid.ncc tbr hunring 'rot r :::. r::ter;or uplxnds h:rc iccuqrulated in rhe l'orcllon!. ol..xv.s.i ' :: : .:::. r.rovide evidence for inteflnitrc t o(.uf,tri,r), () rrkc
-+1

CHARLES

I6 HAM

.dvanlage ofdre local resour.es ofplrnrs and.ninrals, fron at lcast 35,000 yea6 ago. This docs not nrear thar there si re groups \\,ere nor ako trking idvanr.ge ofriver and streanr val-le-vs, but the many changes to tlrcse nrorc opcn and less protected habitats mean! that the archreological evidence within then has not 5unNed so well. Th coasts ofSoutheast Asia, and in parucular rrer estuaries. offer onc ofthe worldt richesr h,bitars in terms ofnatural boLrnm On a ..orld scene, huDter-gathercs living in such [rarine habirlrs caD secure rheir terflron lhroLigh peln nenr occupation and become lery .icb h social, technologrca! rnd economic !c.nrs. A very early Pottery tradirion known, for exanrple, conre' 6om the -lomon cuhlres of postglacial Japan, subnncn(e. alben po$ibl! \'r$ some incipient forrru of essenrially hunter'grtherer 'n cultiqtion and plant nianagcnr.'rr Tbe ethnograph;c Calu,J oi Florida, and the Kwakiud, Nootkr and Haidr oi Bnosh Columbia and \\'ashrnE'ron State. all essentielly hunters and githerers, susrarned ro\al elires, conrnlonels and sla!e!. The culture ofe:rlv hunrns and g:thering co3st:,l grouF.,n Ihrihnd and vietn:tm (Figure 3.1) has re.eDdr heert ,ilumnrtcd bl rwo excafaiiohs jn ttre GLrliofSirnr and by an increasnrg number oir*csrrg.:Ions at sites located on rhe raijed beaches ofVietnam. Nong Nor in Ce'rlralThrLjr.J \ai e\c:!ard over three seaio si,) I990-2.: h lics about 25 km behnrd the presenr shore oithe GuifofSi'.n, but durins ,!s 6rst occuPation phase it rvas iocated or tle eJ3. oian e\krni\ nrrine en$aymenr, no more than hafan hour by canoe 6on rhe outler !o rh. ofin ser (Fieure 3.2). The emb:ynEnt suppolted many other, sjDr .. sir::. \oJis \o. comprises wh:t looks lik a thick shli middeD, in which dre small cocHe .\l.r.n:r /l{,nr predominates. Carctirl e\.Iir:on ;nd analvsir ofthe finds fron Nong Nor have revealed how the occupans sFenr lherr ome. Thev Nent out to the open';a ;o hunt the eagle ray, bull aad tiger sha.Is. The\ hunred dolphins and a wide range of ilsh liom the inler and the oGhore cor:.I reerl onls one grale was found, th.t ofa rei,?.ively old wonun, buded in a crouched FoiiEon under a group ofpottery vessels. A roud u'hitc pebble sas found {rrh h.r boi!: Cloje inspection revealed that it had been u:-d tc rub and polish the suna(er o,poren lessels before they were fired. Many poa rve.e male and fired at the site. and lhe surtrces otrhe finer wares were decorared by potishing anci nded designs. l'en a* l;nJ mamnlih \ere hunted. lndeed, nearly aU rhe small sa iple of na!ffn l

bones rccor:red had beer converted into usetul tools. There sas no evidence for dorrestic anrmals. no rernains oithe domestic dog, ,nd a contlere absence ofrice. It is likelv rhat this sne Nas oDlv occr.rpied fur n season. lr orJ! hr! a rhin lrler ofocupation. Perh.ps lhe inhabi!3nrs rta'ed for dre dry seas<'n and mored to a moe permanent base for rhe resr ofrhe lear. According to rhe ndiocarbon d,,res, Nong Nor Nrs lisired in this way in abour :,3,ir Bc
Just such a pcnr]inenr base sire has becn found 11 kilomeoe! n, 'he north ofNong Nor rt Khok Phino r Di. a large site occupied beNeen 2rrrrll md 1iJ0 rc, during lvhich timc it5 cul(ural deposi* accumulated to a dcpth ofo!r 9 m.r The pottry ieltck and bone tools fronr rhe lo(es! layes at Khok Phanon Di rruror those from Nong Nor (Fisures 3.3-4). The econon,r szs aho similar. We 6nd dlick depos,ts ofshelish that prefer to live nr nuddr: cstuarine conditions. Tbere rras also ntucb nshing. and cnbs \ere collected in considerrble numben frorn the nearby mangroge foress. Tbe huntergadreren of Khok PheDont Di conmanded a river esnrary The thick layen ofash, the c:rches of the clry rnvik for shaping clay into pots, and tbe burnishing stones used to

decomte drc 6nnhed vesrcls rerdrl

! ligorous .eramic
42

industry. The polished adzes were

GULF
stAt\,1

;i{.

nk.d,tutu

- j ll- \oD Klo Noi, 12. Ufper SonSkh.am Valley, 13. Ban Lum Knro. 11. Srnrcng '. 15. \,: Pa \lti. 16. llin Kro, 17. Phu Lon, 18. Nil Kh.tu Haeng- 19. B$ Nr Di, :i.:,::i:.:l.DiBrt,22.Don8Son,23.HoaLoc,23.VictTri,2,l.BanDonTal)her, :5. .: : 16. \..n I1a Kok, 27. BinWing Hi, 28. Sx Hutnb. 29. And{or ljore,. 30. Ban l).r:i:- :31-\.!)irc.32.BinTikhong,33.Br Knbnr)sNok.3.l.I,hi r,.3s.Nocn L L i:: :: :.::i {::...36. Phronl Wrn.

1_:i:.n ofdie Southcst AsirD sit.s ncniioocd i', thc rext. 1- NonS Nor, 2. Khok Ph$oD, :: i. ?i!.g Go Bo.g.4. Co Lor, DoDg Ddr,5. Llng Hoa, Thanh I)cn. co ilIun. : l:, h)e^\gu\en, Xom l\en, 7. Ha Gi g, 8. Non Nok Th4 9. Brn Phrk Tof- 10. Ilan v!c- 6.

CHARLES HIGHAM

F&!E J.2 The r.cotutrucr.d cosdine in rhc vicinity of Nong Nor 4,500 y..r, .go. lr rcE
the sit.

rB

locat.d on th. .dc. of

lmin.

ts

rh.r

iDlcr. Such

infomtion

i5

lhe 6cld$oik of gEo'norlhotogi!8.

mdc posibL hy

irnponed poncry sherdr. Thc low-lying narinc or bnckish water h*iut of Khok Phlnom Di \\ould h.vc rulcd out local dce farming. Perhaps rhe inhabitano wcr cxchanging good! sith inlud ricc farmen. Well over 100 difiercnt speci$ ofshell6lh wcr ideniified ar Khok Phanom Di, as wclt .s microscopic fonrns 2nd ostrrcods, md difcrcnr lariecies ofreft or fr6hwltcr tuh .nd crabs. so ve can folow the history of rhi conununir] aj ir adaptcd nor onty ro envircnmcntal. but also social changcs. tn thc case of the hner the inlBbiranB had ro cope rvith $e chalengcs posed by incusive egricultunl goups in the hioterland. Paradoxicalt cemercries provide ihe best cvidence for assersing tbe Dilurc of this cxtinct sociery The dead at Khok Phanom Di wc.e intcrred in clustcrs, eech including the skeletons of men, rvomen, children and infants. The mortuary rites remaincd basicrly unaltercd through 20 or ro gcn.ntions ofoccupation. The corpse wrs wnppcd with a nnge ofgrave goods in an asbcstos or be.ten bark shroud and laid on e wooden bier:n a gr^c orienrared \o rhat rhe hcad poin(ed ro rhe rising sun. Red ochre wes powdercd ovr thc bddy, sometims in e considenble qurnrily. cnvc goo& includcd completc, very bcautitul ponery vesels, shell and fth bone berdr md brngles, es well as tools probably used by the jndividual in life, such as srone adzes, fish hook, clay potrcrs'
anvils and bumishing stones-

made ofstone gulrricd in th interior and wcrc gircn healy usc. Wc can imagine iheir impon2nce in clearing rhe mengrovc forcsr, and in meking houses end boa6. There is also som. tantaliing cvidencc for ricc cultivrtion in the frrm ofrice tempcr in e few

44

t,t{oilr I llI NIoLt'rHtc To TrlE \oN A(;E

t Khot Ph on, I)i. ft nrclud.s rroriG of no Di. Thc! \?r. nDd. sirl, ..fnn.r.,ll. hutr Dlitc ngurs f.mllelednr such nrhndsitcs cNpc i\e .l lirished trt burilhlii Thn r Khol (:brrccn Thc !c$cldia etcr n 20 cDr. eNr plcn!lcrhdilrr...

I:i{,m.,.J

Pottcry ve$cl

fbm nlorruary ph*e

l:d!tu J.l Mortlac r.$el fron, Xhok Ph!-

TnDe, holeever, aho wrought ch.nges. \isiblc i!) rhe rccord oi sclen iucccsnfe nrortuiry phases- Thc fil5t sxN scattercd burjrl! \i(h lcn i:rs gr;\c goodj. Thrs contmsrs wnh drc sccond and rhird pbases. in \\hich ,he derd $ere b !!cd ur dscretc clutcr r.conrpaDicd by potten vesels. shell bc:d: :nd t-,$ boJ,! jc\ellen: There sas a very high proport'oD of inf rts. tletr e:rl dratl: 1robrL,lr reflcctirrc an rnherited blood dnorder kro$n as thiltrsiirnr. Ti!. .o:rJrcon co'ri;rrc.1 sone resistance to malaria. bur could rlso induce anr:n: r'ber: inlcr:t:nce rr:s homongous for drc recesive alleles conc.rnel l)siFr. ih. nrir:l]i elxtrotrlnenr. rhe nren duruE drcse two phases had s:rcngh .1e!.;oFed urper boi! flnrscLrlitu.e and rheir tooth lvear direred from thar seen i,r \\'oDren ' Ir \\ould i.., lrkel! rhar thLse nrcn spent tinle rlvay from Khok PhaDom Di, perhaps ut cortrl or nrhnd tnsei. Their {bsisrencc w:rs strongly based upon huntuig,
gathering and fishnls. Sevei: changes rvere noted midrvay through dre third mortuary phase. The sea level fell. Shellfish, 6sh rnd crabs Dov show a $arp bi.s toserds freshwatcr species. At the slole rinre, grrnite boes werc found for the 6rsr tnDe, aloDg wilh shell knives that, iccordnrg to dre s'ear pattern on the blades, were used for curtmg cered stem!. The 6rst bones of the domesti. dog .lso appeared. Taken iD conJunction, tbese variibles are

compxt;ble Nith rhc local cultivation ofrice. This 6 dhg n firrther rupported by d1e recoi'eq of partirlly digested stonrach coDtents and hunrrn faeces ftom burials ofrhrs :bJie. The forDrer nrchrded rice chailand freshrva(er fish renuins, rhe htrer the rcD)ai : a:)ouse h.ir rd a beclle knosrr to nrlnbit ricc storcs. The ri.c corlrlnncd nrxv hrlc :::;: !,or!'d iD conditions s4rich harborrred sch vermnr and beetlcs. ' :::i ;nd soorcn ofnrortuary phises 3l] aDd .1 sere disdrguishcd on the blsrs ofgrr\c i .: :i:. rirrDlcrben interrcd rvith iargc nrrde shell ornaDrents. Wonren, on dre orher 1,.,:..: i:: ::rllrrnied in dea* by the clay anvils used aor shapjns porrerv ressek. Ar th. '.:: . :r::... -f: ilnd rhat nren we.e less powcrfi'lly brilt and did no! ltre !o long. \t'hric t:. :::::::::::. nor conpletely clcar, thesc chlnges nighr refleci a chang. \'hich sa\\ ric. .r::r.::: : :.Jr)inq nlore nnporr.nt at the expense ofishins .ri!t .ollc.riDg.

Th. nr.r.,::.: . .: .:. .j.!n

I ::.-.:::::::r

ro

hrlc nrirc.l a s;n'lrr d;er, for rhdr p:rtrn: oftooth

15

clr trtri\ t(,

\r!1

.rtclicrs. Spurdlc Nhorh aftcst to a \'(rrinq !)duirr! rDd bone hirpoons Nggcst sonre tbnn oi hunting. Thc strnr rd,.s t.rk( r !rn(r\ .r'r;r rs (Fistrc .1.(t. ()nl,v lbur ofthc shouldcrctt varicty wcre fou!)d, co,rt.neJ $!rh -:- e\.n4n.s ol. qurdmDguhr forrrl. sotrrc otwlti.h $crc sh:trpened so .or\iterrl\ rilt rhe\ ir('tlrorder than tlrey ae long. -or \|r.h (he 5hrfe rs indeterminare, i ror.l of whcD onc idds stone idze frrgln.Dr l.ll8.d,.{ or.d,e frrgDrenr\ \.re r..n:ir..i T1:.r( trrr rl{';9 nnxll stone chneh, sorttc *idr curting cdgcs only I0 nrlhnrer!(. '.r:.:: \l::r..I : ' sh.ril(r).s \!crc tbutd. Thcy bore groovc's thrt rcsrlt frorr dr:rrr::r:r::l !:or. rJz.. rnd chreh. Stole Projecdlc' poiDrs $crc nrer. Threc rerch rl: .ir::::::sr,::. r: : .r.:. ),r!r. drc reniitrdcr rR more likcl_v to L)c unged ]lro\herdt Tl)c inhibir.nts oal'hu'r! \:::.::; .,..r rir.Ir::c.1 .to]:. ::::--,. ::)n.rh 1r) nefhritc. Thc rottllsaDrple of5.tirsp..irn.c. lrtr. f:.n jubJNrie,j rnto ergi:r nrc.'b:.eJ on !hc shape of dr L' c.oss-scchor). Molr ire reir.:;rajlrr. bur tonre rrc'nruch morc rr.nr;lc'r. hrirq a rarge ofribs rnd 6rrrgc.. \ h:1. ..:::r ::. :rrse enoush k) ,,\ tr.:jll: :r::'r:.ii Lrlhcr\ hile lnrll dilll))ctcrs.nr.l \rr. :,:lr::.:..rlr)ed r-or childreo or t.ihrf' N. Ji.niJ! i\ errrnrg\. Stone bcilds \\.r. .,1*. :::::: :::.r.: berrg nrhrlrr ar rle:iunn: ur ro 1.-l ...tinrerr.i in lcDsth. Thcr. s.F :--. .i \,:..:,'u: tr.rJ,rion oiNorkrrg cl:r: Thcrr rrc prllct bo* pellets aDd clx! ncr \.!!]]rJ. :-:r ::r-r rrenrioD \Lrs;r(co.ded ro dre rrui:.ture ofcli! vessels. h i\ .n rh. t-ri:. ra .::rt]jlDg .tecorarile sqles on Foftc^ rh:rt the Vietnrmcsc archreologi.rs h.re::,,;::rr.t ph:ses ro this culture (Figu('3.7). The e.rrjie{.l;illed after rlrc sitc oiGr B-.:::. :. .hr:J.ierized by nlcised pira el bandr ini-,lcd snh rorvs of nntrejlort .rrr. r::i: r ronrc.l nnplement. The favoured nlotifrs in drc rhape cfan "S"

nkrnJ.r Tl. !::.::: !r: :.!r{' nrcorpomtes a range ofdesigns based. eccordinsto H' VaD Trn.'o:r S..::r.:.ii:!rnD.!ni. hrchedbi ds indlled rvith dentate impressions alternate \rrh b.!t]r.. :::: i'i::I. ro ibmr r scries of most atractive desigD fields. This form of dc.o rr:::r::.litr: frr,rlelt \\ith the Yunnrn Neolithic sites ofBaiyangcun and Dadunzi. The [rr PnLr']s \guren stylc of portcrv n found at thc site of Tr:rng Kcnh, nerr Hr::r!:: \-.::1:r KuD Dungl excalarions (herc hxle relealed r \otlrisricited

\i.i:h.f n.

.h).... .i::il :n)nr!. (i$s

r:). mrnut-r.rlrr. ofnephrite bmcele(r rnd b.rds, inrohin!a $orking lvith and crindirrs stonc{.'TIc rrdro. l.o)r drft. \ugg.st ihat the

ffil
Ii,r k .r.6
Sto,tc

*'

ize\

rc.o\err(l ionr l,hung Ngtry(

4li

FIlOAl THE NEOIITHIC TO THE JRON ACE

workshop d.res ro abou! l65li-1500 Bc, but the inircduction ofcraft skillt in working ncphotc :rnd Jrderc ornaDrents rcprcsc![s a key lnrk rvith the htcr Chinese Neotithic. Thc bangles Gortr phuDg Nguyen are the product ofhietJ){kiled lapidlries. The value attached to (one je\yeliery and weaponry is rlso in evidnce at Lung Hoa, a second late Phung Ngulen site characterized by deep gnves, somc cut over 5 me.res jnto the ground and equrppcd, as in the Lingnan Fubin culture otsouthern Chinr. ledges.trrTwo ".rrh grrles rnclude sto e bracelets, beads and earrings, as well as ceramrc resse)s, as morturry oiierin$. One f,c h:lberd is panicul,rly interesting because rt nutches rhose found nr southcrn Chnra and northwards into d1 Bronze Age early urbau states in Sichuan (Sanxinsdui).nd thc Huanghe basin (Erhou). The l,a.z,'d,( jade blade is also a central Iiature ofthe latter trvo Chnrese asse blagcs, and specinreDs have been recolered fro'n Phung Nguyen and XoDr ReD. Even at dlis late stage in the expension and esr.bhrhmenr

- -, i,::,:.: NgureD, Victmnl deco -:: . :::. :n \attrhic sircs iD

.+9

ted iD r \vidc

vtrr(!

oln,or1r-i

CHARLES HI GHAM

of Neolithic

communities

in

Southeast Asia. conracr \vas maiDtrined

with

thc

nrcreasingly cornplex societics ofthe Yangzi and Huanghc rivers.rl Archaeologicaj fieidwork in the uplands nonh of ihe Rcd River has resuked in an appreciarion that Ncolithic sites were nor conrincd !o the relarively open ternin abov Viet Tri. The l, River for example, errracr.d irnusir senlements characterized by a r]nge of stone adze forms, srone brngles an.i a c.ranxc These sites, known 'ndustry coilectively as the He Giang cukure. nng|r \eli hale been sources for rhe high qualiry stone found in such protusion at the mrnr Phune Ne!\en se(lements.r, Further south, recent excavations et the co3sral l*e Neolthr srre oi Con \ven hrve yielded tubulrr stone beads. stone b.nsles and adre h.a.l'. deconred porsher,I and rubular ccnmic feet for pottcry vcssels. The serlleflrenr oiCenlral \icrnr during !he iile r.veolirhic period is

nujor area of future research. The singular sryle of decofared Foften in rhese Phung tigu\rn srres is marched iurther up the Red tur.r !r!:! m rh. sourhern Chinese prolinc. ofYunnrn. A dcep s--quelrce xt Bai,1':lngcun r..orior:led lhe foundarions ofat l.an 11 holI!e!. iD addidon ro an inhumation cenreien Cr:eJ ro (he lecond halfofthe third nillennium rc. A second but rather later rite ar D;Junzi also rere.rled house foundarjons and a .enetery in which infants rvere interrcd jn L:rS. cere!. ja6.1tr These rwo sites are lqlhin srriking distance ofthe Mekong tur.r. :ri a5 $e fouo\ rhis grelr rboloughfare soulh, so \!e encounter more Frehilror,. j.illemen6 linked by a common econom).otrice ftrmjng and 'neny stock raising. jn .onJun.non sirh incised and bulnished ceramicsIn Thailrnd. rhe e:rlje'r lerrlenlenr phase ar Non Nok Tha in the oortheasr belongs ro this horiroD: h.re rsrin $e find inhumation graves associated with incised ceramics as morfuan oFerings:'Ban Phak Top has furnished spectacular examples of the black burnished and in.iied s1.es, but srdly this sire has been completely destroyed by looting. A handtul otgrales ar rhe nesrby sit ofBan Chiang belong ro rhe same tndirion and have been d:ted beoreen 2500-2000 Bc, while at Non Kao Noi, in the same area of the upper Songkhren V3llell tu.ther undated gnves harc bee. uncovered. The sne ofBan LuItl Khao in the upper Mun Valley, excavated in 1995-6, has revcaled a scries ofpics and ,$ocir.ed occut:rion rrurerial dated to about 1500 Bc. Again. ihe black jncised po$ery bclons ro rhe lame inruriye tradition, while rhe falrnal remiin! frcm rhe prls illuminate dre en!1ronn.n! !\'hich thse serders soughr. The majorin. of rhe remaiB come fronl Iish, sonc oiremarkable size Nhen comprred wirh their modern counrerparrs. The Dan Lunt Kiao viIlee *as locatd in a low-lying ri!rine habira! $hich altracred deer and \"terbuffalo. both of\\'hjcb wete hunted. Indeed, rhe large, (rld s"ter buffalo .ppear ro have been ripidl! eliminxed, for they o y appea. in the earlv phase ofoccupadon. Ban Non Uar lies abour 15 krlometres to rhe west of Ban Lum Khao. and rh.ee
a
se:sons ofexcavations there beginnjng in 2002 have uncolered NeoLithic occuparion and

cemetery daring beNen 2000--1300 Bc. Adult burj.ls $ere found Nith 6ne inched poteq lesreh, and bivalve shells thought to represent ferriliry One man rvas rouDd in a seated posidon \\tthin a largc, lidded cramic l'essel. A Noman had bcen interred with threc pig lkeletons, decorated pots, and co\\q shels pleced as ear ornaments. A ring offi\r lidded infant jar burials vcre found dnposeC round hcr head. One intanrjar burial coDrrincd a rice g.rin, while many shelfish in the occuparion lzyen nrdicate the exploit tion ofa \?rery environment. As src procecd rtill fur(her sourh, rhe sime portery srylc ind lonns wlrh incised aDd burnishcd dccomtion rre found at the site of Samrong Sen in Cambodia. where
and prjnted

50

ttl()r\l tHt lLoLnHtc T() THE tRo\

A(iE

prchisbri. rrcha.olo$ !) Sorrh.ast Asir h]d rts beginrin5r ore. r..nrurv igo. Thc\ occur rlso rt NoD l'jr \\'r -rust nordr ol' Lopburi. agJIr 1ll r lrnuliir trrhutrutior ceDrcier\:rt Su(h \crrlclrre!!\ nmst havc c.iercd the c\chrns. (,rbrr oi liHuenr .oasril setrlemcDt\ like lihok lJhrnotrl Di, bringnrg lvith theD knosledgc ot ri.e cuhivarion. doDresri. sro.k rnd dre dos. To dr..ouths.\r. in dr. \'allcys ofcentral ar:d \estern Thailmd. \.ollrhrc rerrlcmenr has lolg t rcn kloNn follovng SorenseDi pioncering Bl) Kao in Kir(hr )L'ufl 1'rolnrcc.r'Thc e.rlier ofthe t\\o major occupation r.h.rser inclLrded a \.rie. oi rrhun)ition gnves, whcreiD rhe dcad *'ere iccorrrpinied b! a \rrde r.n)gc of forr.^ \!srl aorms. Theic is litdc elidcDce here for rhe iDcised rDd burDr.h..l jr!le. dre hrl| rl, ol drc ccr.Dic iDdustry benrg r borvl raised oD ponrkd rflf(id rc.! Sunrlrr t;rlrs hrrc bce'r found to dre south into dre coast.l planr of Mahvsir. .rt such rrrcs rs J.nderam Hilir in Selangor.rT Radiocarbo dites rrom Ban Kao nrdicrtc tr,iriil sdrlenre r in ihc hre rh'rd millcnnium tc, lnd nvo dctenninations froDr Jcnderrn) Hilir sradJlc
2000 Dc. The snrihr dates ofthe eirly Neol;thic sitcs fron Guangdong to l,cnnxuhr Mrhlnr probably indicatc rchted exp:rn'ionary Dro\?nrents soudrwards, follo\vnis rhc corres of the n jor rivers. Linguists hlve identified nrh.rired snnihrirics in mrny ofthe langurges spokn o\q nruch ofMinrland Southeasr Asir, trorD easreln India to southcrn Chioa rnd south{"rds to Mrla}sii and the Nicobar klands,!r $mcient to classi4' then wrrhin rhc Ausfoasiatic fanljy (Lhich ercluder rhe Trj languiser. There rre three rnalor

subdilisions Nithin Ausrroasirrj.: \1u,rdii. in lndia. Mon-Khmr (which nrcludes \rietDrrrese !nd rhe Alliir Ln$rS.i oi\1:1,!\Dt. 3nd Nicobarese. Cogrates for rice lnd r\pecit oius cultiir!ion rre \norrn orer rnuch oftlis area. suggesting dut this crop \us sronr b\ rhe carhjt -{urro:,uor ;pcakers. The deep diferences btween the Mu'rdl brrn.h on tlie one hn:J;oC \lon-h:lnler on thc orher indic:rc a considcrable tnre deprh tbr Proto-Austro:siaric. nlexsur.d iD nxllenDi, rarher than ceniuries. Thh has
cncouraged both linguists and archaeologrts ro seek
ianguages hxd a

a pltiedr Nhereb)' rhe Austronsinric comnon origin in thc Yrngzi V.lle!: rheir presen! drrributioD reffccting the dhspora of agricultriral societies which begiD over 8,000 \'ealj ago. Thn expxrsionary model gathers force with dre appreciatioD th:u Austroasiatic and AustroDesian languiges probrbly have a coDnnon ancesrry iD dre Austric phylum. Whereas the former exprDded Gom the Yangzi VaUy by continental riverine router, the htter followcd an hland Soudleast Asian and ukinately Oceanic route. This model his drc advantage ofbeing testable through thc study ofancient DNA in bon, not o ly from rh. hunuD benBs deenred to h:r!e participrred bur also fronr their dogx. for dre latter

l:r

:::r.

s no ancestor iD Southeast Asi! or drc Pacific and n:ust have odginated uhnDatelv in \uch as Chin., \'here the *'olf is native.

The Bronze Age of Mainland Southeast Asia Tjr. :r-:i:i ,:r;c l4arnhnd Sourhc:rsr Asian D.onze Age have a hhton ofcontrortrst irr!ir-: :. :.:::r! drr;ns from dre 1960s and 1970s fo. remark.bly earl! dares I'or No sirc\ rn T;:::::j. \on \ok Tha and Ban Chiang.l'J These claims can noN be rir rskl.'. Ne\ a...:::.:-: ::::r-:rton drting technictucs allow f..g reDts oiric. ch i ujed io tcr+r.r rhr .:.. :...: : :::nrui:crue mortu.rv vesseh, to bc rl]red. Thr: pcrrrits thc prccse (lini! , ::.:: : ::: ;-L:rc.l conrr\ts in a prehisroric sirc Thc r.n,ll' rin NoD
51

CHARLES

IGHAM

o.igjn, or the knol'ledge oi snklftr:- and casring \!, inrroJuced trom an cstablished exotic source. Now that se can j.r :!,ie clajnx ibr a Sourler:r .1sDn bronze rndustry as early as tl1e fourth millennium rc. $hch rvould ha\e ruled olr an! iortu oiinrroduced knowledge, it is possible to cons,j:r fojlible oulsid. sour.ej The close{ and mos! likelv such source Lies nr rhe bron,e-c:nng rr.hrioD ofChin: \\i hrre rlread! rhar lhe 'rored Neolithic conmudties oisourh:rn China and Viernam obteured erori. norrhern Jades by exchange. There is alo ii3.n.. rhl these grours \ere ianxh,. srrh bron,es which originated sithjn lhe Sh:nS a:rd Zhou tradidon of casnng. The sire of Xing'an, for example, has turnished I l:re Shang period bronze yo! lessel.r: A b.onze halberd of sinilar ase has been lbund a! Xinjie in cuansE province.r rhe mainrenance of exchlnge rel3noni be(r.ed Lingnan Gourhern China) and dle Yang,i V ev dLlring rhe second millennium sc cou1J sr accounr for not only the fansni$ion ofjades and fiaished bron,es bur lhe \rLrl kno\\'ledge olrhe properties ofcoppe! and tin ore. One cd. ho\e\er. proceed a stage further. The origins ofthe ChiDese Bronze Age itselfhase long been engmahc, not least because ofthe recovery ofrhe earliest bronze artefacts in rhe .emore norlh\\'estern province ofGansu. Now, however, it is possible to face the e\Ta6ion eciois the Eurasian Sreppes ofbion,e industdes ftoDl west ro easrl rhus meking rhe Chinese redition itsef derivative 6om a western source.l \Vithn this broad patiern of rhe ditrusion of knowledge in irs many regronal bronze+vorking erTresions. rhe Mainhnd Southeast Asian province n dsdncr on jeler:l counrs. rhe nourh otthe Peari River and the general area o{Hong Kong has been knorvn as a centre ofbronze casting for th6 past 60 yqlrs. Schofield. tbr exampLe. recovered six bi!3]ve $ndsrone moulds tor casring socketed bronre ares ar Sh.k Pik, on Lannu Island, in 1937 Excavarions at Sham Wan in 1971 recovered b.on,. nshhooks and arrowhead made of a 10 pe. c.nr rin a1loy. Meacham has added furrher !o our understand;ng ofthe Hong Kong bronze industry through the recovery ofbirahe sandsroDe casring nloutds used as grare goods in a cemerery at Kwo Lo Wan.trr These s[es belong rvrrhlr ihe per]od 1300-1000 uc. Borb rh. lLnished bronzes and rhe rechniques of nunufacrure are precisely nuiched ro rhe Nesi and sourh in Thailand. Viern,n and Canlbodia. Our understanding ofthr arcnre Age rndirion has been grertlt enhanced by recent research in Thanand. borh at rhe copper mines rhenrehis and in dre viltages which acquired and cast the ingors. T\1o nining complexes have beeD examned. The first jies on the southern bar* ofrhe Mekong River at Phu Lon in norrliern Thailand, and the othcr n found in dle Khao Wong PrachaD Va ey in cenrral Thailand.,5 The rich veins of copper recognized on the surface by characterisric green staining, were folowed inro rhe
hillsides and the muied ore was crushed and concerrrated belore berns placcd nrto clay reactioD vesels. These rvere then probably covcred wirli ceranric chimDeys picrced to 52

reveal that the Bronze Age ceDreteN \"s in use Nnhin rhe period 1500-1000 Bc. Bronze Age burials at Ban Chiang belo.g to the same pe.iod. Freeing the issue fron chronological interesting a$esment of the Mainland Sourhearr Asian Bronze Age. The emerging pattern indicates dut, ftom Hong Kong and colsl Gurngdong ro rhe Chindwin Valey in Burma, a distinctive tmdition ofbronze casurg dereloped durng the second half of the second n1ilendum Bc':o TlN \1a! roored m lo.al Neolirhrc conuunities and poses the issue not only oforigins, bur oa rhe lnfhcanon5 ofme llurgcal skil. Unde6tanding origins invohes an a;preciatior oi bronze casting rndirions beyond Southeast Asia, because eithe. rhe lo.-l bronze \\orhns had a local and independent

Nok Tha

FI{OM THE NEOLITHIC TC) THE IIION

AC;E

adrlit tuydres, itr order to raise the furnace heat sufficientlr- for successful sn'relting. At Non Pa Wai in tlte Khao Wons l)racl-]ar-r Valley, snrelting operations covered an area of
5 hectares to a deptl.r of 3 i.iietres. Most of the resulting copper s'as poured into moulds to produce circular ingots, surely destined for exchange. Sonre. hon'ever. *'as also localIy cast in bivalve r.r.roulds. One of the graves, of a ysp11g nran. conrained the nrould for casting an axe. None of the copper artefacts found at Non Pa \\Iai conrained an,v tin. This intensive rlining activity has been dated to the second half of the second nrillenniurn tc, but it continued into the first rlillenniunr. as seen ar the very deep deposits of casting debris recovered at the neighbouring site oiNil Kham Haeng. Here, the thin horizontal lenses are thought to result fronr nretal-u'orkins debris being redeposited by nronsoon rains. Indeed, Andreu' Weiss has proposed that the smelting acdviry rvas itself a dry season activity undertaker.r bv the occupants of the rnany settlements which concentrate in the vicinin. of the rich copper deposits. Most Bronze Age settlements are found in the smal1 srreanr valleys which feed the major rivers of Mainland Sor-rtheast Asia. Thev n'ere linked by far-reaching exchange

networks which saw ntarine shell reaching settlements over 1,000 krn fiom the coast, and copper and tin entering villages lar renroved from the ore sources. One of the advantages of relatively extensive excavations in Bronze Age sites in Thailand has been the recovery of both cemeteries and the industrial areas in u'hich bronzes were cast. For example, a clay-lined furnace has been recovered at Ban Na Di in Udorn Province of the Khorat P1ateau.26 It was surrounded by crucible fragments, copper staining and a sandstone mould for casting a socketed axe. The ceramic crucibles still contained the remains of bronze resulting from a 10 per cent adnrkture of tin with the copper. This produced a tougher alloy than copper alone, and finished bronzes include axes as well as a range of arrowheads, socketed spears, fishhooks, bangles and anklets. A similar casting area was found at nearby Ban chiang, while at the Bronze Age cemetery of Non Nok Tha, several male graves contained either socketed axes or axe mouids in sandstone. Judicious interpretation of the mortuary remains from these Bronze Age cemeteries has provided some insight into social organization. This begins with the premise that a person's status may be indicated by the quanriry and quality of accompanying mortuary ofierings. At Ban Na Di, we find that men, women, infants and children were interred in groups, although the restricted area of the excavation ruled out the complete exposure of anv single cluster or row. Grave goods included ceramic vessels, some of which contained tood in the form of fish or pig bones, the complete limbs of cattle and pigs, exotic shell :rC stone beads and bangles, clay figurines of cattle, deer, elephants and people, and ::--cze jervellery in the form of bangles, anklets and possibly earrings, It was found that ,:,. oi the nvo groups of interments included most of the exotic offerings, all the -t:-:--::is and most of the bronzes. Since both groups were interred over the same period ':,:--:. i: ts possible that one commanded rather higher wealth and status than the other, : -. --:i :: : dominant degree. -: i-:-i ::::i hrpothesis would necessitate a much more extensive excavation. At Nong 1".. ,* : -: -"': -\-of the Bang Pakong River in central Thailand, a much larger area of a ::- --. *-:: --:::r-rerv has indeed been exposed.2T This overlay the hunter-gatherer :---:--:-- - ---,-:i :bove and over 160 Bronze Age graves were uncovered (Figure 3.8). --' -:"-::- ..:.' -r *-l:r Proportion had suffered disturbance. Nevertheless. ir is possible to -:- :r--:: . :-.-.' --- ::.:tn'tradition similar to that in northeastern Thailand. The dead ' ::: 1:--::::: .:. : , : .:::::ior1pS,,r'hiCl'r inCOrporated nten, wolt-Ien, inlantS and children,
53

FI{OM TIJE NEOLITHIC TO T}IE II{ON A(;E

in ordcr to raise the furrrace heat sufficientl), lor successftrl snrelting. At Non I)a Wai in thc Khao Wortg l)rachatt Valley, srrrelting operations covered an area of 5 hectares to :r dePrh oi3 iiietres. Most of the resulting."pp.. \\';rs pollrecl into nroulds to produce circular ittgots, surely destined for exchange. Sorne, lro...e.'.r, was also locally cast ill bivalve rrroulds. One of the qraves, of a young nlan, contained the for castillq ill'l i]\c. Notle of the copper artefacts found at Non I)a W.ri cont:rined'rould any tin. This irttensive rrrining activiry has been dated to the seconcl 6.rli of the second tnillctltttttlll B(., but it contirlued into the first rrrillenniunr, as secn lt the very deep deposirs oi casting debris recovered at the neighbouring site of Nil Kh.rrn Haerrg. Here, the thin horizotttal lertses are thou{rht to result fronr nretal-n.orkipq debril being redeposited bV trtotlsoolt rains. Indecd, Anclrcrv Weiss has proposcrl rSlr t6e s'relting actir-in' $'as itself a dry season activity undertaken by the o...,p,,r,r,, oi the ,1any settlelllents rvhich concentrate in the viciniry of the rich copper deposirs. Most Ilronze Age settlenrents are found in the snrall ,tr.r,r, r-.li.r-, u-hich teed rhe nrajor rivers of Mainland Southeast Asia. Ther' \\'ere linked br- f;rr-reachins exchange netn'orks which sarv urarine shell reaching setrlenrenrs o\-er 1.(rOirklr trorrr rh! coasr. and copperand tin enteringviliages far renroved frorn the ore soLlrces. Opc of r6e.l.ir-a'rages of relatively extensive excavations in Bronze Ase sires in Thailanci has b,een rhe.-.o.,J.1. of both cemeteries and the industrial areas in n-hich broirzes \\.ere cisr. For eranrple, a clay-lined furnace has been recovereC ai B-in \: Di in Uiorr: pr..r-urce of che Khorat Plateau.26 It rvas surrounded br- crucible f:-:=ienis. co:p.: >i;lnrnq ;rnd a sandstone rnould for castinq a sockete.l ax:. Tll ....r,i. .-r'.:.rt'le, sri^i conr;rined the remains of bronze resulring titr::l ; 1',::: --t::::i::t:_\ture ot nn s-rth the copper. This produced a tougher al'ov rh-:: .urlf3r '-r-:tj. :ni inished bronzes include a\eyas.r,.ll as a range of 'L=- j :-i! \1.-::*:- :-':r:i;j ::3::s. ::.:hooks. banqles and ank]ets. A sinrilar jr:::j -:: n:ar':r' B-::: chrans. s'hile ar rhe Bronze Age cemetery casting "ra"=*", of Non Nok Tha, s:r-eral nraie graves conraitred either socketed axes or axe moulds in sandstone. -fudicious interpretarion of the mortuary remains from these Bronze Age cemeteries has pror-ided sonre insight into social organization. This begins with the plemise that a personi statlrs ttrav be indicated by the quantiry and qualiry of acconrpanying mortuary offerings. At Ban Na Di, rve find that men, wonlen, infants and children were interred in groups' although the restricted area of the excavation ruled out the complete exposure of anl'single cluster or row Grave goods included ceramic vessels, some ofwhich contained food in the form of fish or pig bones, the complete limbs of cattle and pigs, exotic shell and stone beads and bangles, clay figurines of cattle, deer, elephants ,rrd p.ople, and ::onze jervellery in the form of bangles, anklets and possibly earrings. It was found that ':= of rhe nvo groups of internrents included most of the exotic offerilgs, all the i---:::-3s and nlost of the bronzes. Since both groups were interred over the rri'. period t:-::.:. :: ls possible that
adrrrit tttt'dres,

-: :._::t i dorninant degree. l: :--. j:-.:i hvpothesis would necessitate a much nlore extensive excavation. At Nong ': .- l -:-"= :1.'of the Bang Pakong River in central Thailand, a much larger area of a " :: :--- '--= :.::-::3n' has indeed been exposed.2T This overlay the hunier-gatherer :::--:::--:r.: :-::-,.:j abor-e and over 160 Bronze Age graves were uncol,ered (Figure 3.g). - ' :'::'.-':-" = --l-. ::oportion had suffered disturbance. Nevertheless. ir is possible to :- -- -: ':: . :::--- -'-- ::-.::''- tr:rdition sinrilar to that in northeastern Th.ril.rnd. The dead - ::. -:-:::::: -:- : ' .:-.,: -:oLtFi ri'hich incorporated rtren, \\'onren. int.rnts and
_

one conlmanded rather higher wealth and status than the orher,

children,

53

CHAI\LES HIGHAM
Excavation at Nong Nor, Central Thailand, 1 991 -3

l! aD O 6)

Male Female

S"r

unknown

child
lnlant jar burial

Figtre

J.8 The Bronze Age

cemetery of Nong Nor reveals nvo major groups. and a preference to

inter che dead in rows.

the infants often being interred in large lidded jars. Grave goo& included portery vessels, ornaments of exotic talc, carnelian, serpentine and marble, bronze and rin bangles, and animal bones. \X/hereas the people of Ban Na Di selected the lelr tbrelinrbs of cattle or pigs, the occupants of Nong Nor preferred a dogt skull for intermenr u'irh the dead. One man',,vas found associated with the horns of a large bull. This cemeren'formed over a period of several centuries, so identifying contemporaneous social variables is by no ltleans straightfonr-ard. Hos'ever, one group of five burials included a male interred in a particularly large grave. He s'as accompanied by a unique large bronze bangle and some fine pottery vessels. Moreover, the women buried in his immediate viciniry were unusually wealthy in terms of their multiple bangles from distant sources. As at Ban Na Di, there is a hint of social inequaliry but again not to a marked degree. The excavations of Non Nok Tha were also extensive and, rvhile some individuals were relatively well endon'ed rvith offerings, there is no evidence for a particular group defined by sex, age or preferred location being singled out for special attention. Indeed,
54

I:l{()M TIIh NEOLITHI(: T() THE III()N A(;E

lJrorrze Ag. approaches burial 15 lronr the hunter-qatherer conrnrurtin' of Khok l)hanonr l)i in ternrs of nrortuan' s'ealth. At Non Nok Tl-ra, the basic lJrottzc Age pattern recurs: the dead were set oLlt in ros-s. u'ith ceranric vessels, bronzes, and aninral bones, which in this site included the coniplete skeletons of dogs and the linrb bones of pies and cattle. At lJan Lunr Khao in Nakhon Ratchasinra Provrnce s'e asain trnd a sirnilar pattern. The dead u'ere laid out in rows. The pottery vessels at thls srte rncluded nranv particularly

no grave frortr the

attractive

re

d fornrs rvith trurllpet-shaped rinrs. anc anirlal bones included

the

extrenrities of pigs and, on occasion, conrplete brva\': sh:ll:i.h and the skeletons of flsh. Sorrre were buried with stone adze heads. t'il: ::o:r: ol rhe i 11 qraves trncovered irtcltrdc'd a sinsle bronze artefact. This rarin' of b:,-,:z: a: B.i:: Lunr Klieo. despite clear evidence for local casting, follorvs a s'ide patt.:l I3roi:ze: r..'cre als'avs rare in Thai Brotrze Age cenreteries. The recoven' oi nn or:.:::.irIS ar \ong Nor also reveals that copper-based alloys were but one aspcct of :h.' :::.:-.,,'.'n:ent of nretals in displav and ritual during this period. The Neolithic layers at llan Non Siat. irt rhe Llpper Mun Valley, lay beneath a Bronze Age ceuretery. Previous exposurcs of l3:onze Aqe setrlenlents in Thailand have been
relatively snrall. At Ban Non Wat. hor.,'ever. the opening of nearly 400 square nlerres over three seasons has uncovered sorne burral. of surprising and unparalleled wealth. While most of the dead rvere found s'ith a handrul oipottery vessels and some shell and marble ornaltlents, a fes' \\'ere interred ln ven' large graves with a rich array of offerings. One such grave nreasured 5 br'-1.5 r:lecres. Nearh'80 pottery vessels, some of considerable size. s'ere drsposeJ ln rur',r..! rouird the nvo burials. Both skeletons had been partially renroveC tio::r th: ,r:r'.'i soon atter death, and then carefully repiaced. Grave goods :r---ui:i socketej bronze axes. fine marble bangles, and thousands of shell beads. A s:.old such outstandrnglv rich burial matched the number and size of pottery vessels. One skeleton \\'as tound intact, the other had been removed and then replaced. The lacrer \\'as associated u-ith a fine socketed bronze alxe, the arms of the former were covered in large trochus shell and marble bangles. Two bronze axes and a socketed bronze knife \\'ere found in a third rich grave, along with 30 finely decorated pottery vessels. The bronzes at this site were restricted to these special burials. Although excavations are colltinuing, it is evident that this site requires a complete reappraisal of the social organization of the Bronze Age. Specifically, there is strong evidence for a socially elite group at this particular site by at least 1000 nc. There is no doubt that nlore extensive excavations in Vietnanr and Cambodia would treath' expand our kno'nvledge of this distinctive Bronze Age tradition. At present, the i-:-uence is best knorvn in Bac Bo, the area centred on the Red River Valley rvhere - , :-.: Dau is the nrost inrportant site, not least because of the unusually deep stratigraphic '.. :::' Ir is located to the east of nlost Phung Nguyen sites,'within sight of the Red :i,t .^ ,. ,.luare rlletres to a depth of between 5 and 6m. A fourth season in 1981 ----'- .'-::..:::o our understanding of the material culture of this phase.2e The nround .'. - . ::. .: ::: .j ha. Its basal cultural material has been ascribed to rhe final Phung -'-:-.:: :-:::..,.. and u'hile the pottery of the Dong Dau site reveals Phur-rg Nguyen :.-. .-. .=.: -:-. ,:.'.: and rtrode of decoration, we also find conrpelling evidence for a . -. .. : . : -. :: :.2'j industrv Ngo Si Honghas published half :r bivalve nrould for -'.-:.:.j - ,...-i--r rr :.1'j.rxc. u'hile Trinh Sinh3" has stressed that nl;rnv oithe Phung ---.
55

CHARLES HIGHAM Nguyen phase stone artefacts were copied in bronze. There are also intriguing parallels rvith rlaterial from the Northeastern Thai sites. nor onlv in the srrrall furnaces for casting bronze items, but in the ceramic cattle figurines. The bivalve moulds recovered bv Ha \ran Tan from a small (50 square metres) excavation in a Dong Dau context at Thanh Den are virtuallv identical with those from the Khorat plateau. The 30 or so frasnrenrs of srone and clav rnoulds were designed for
castirlg axes and fish hooks. The site has a culrural strarieraphv of only one metre, and the three radiocarbon dates match ven' cioselr' rhe dares obtained for the Bronze Age in Northeast Thailand. Bronze s'as also enrploved to make socketed spearheads, arrowheads and chisels. The Dong Dau phase a\es beean ro take on the initial pediform shape so diagnostic of later decorated eran:r'ies tiorn rhe Dons Son Iron Age. Thanh Den has also yielded nephrite bangles and rins\ u-hich recall Phung \guven Protorypes, some of which appear to have been nranuticrured at the sire.-': Towards the end of the s:co::i rnillenniunl BC. the Dong Dau phase developed into the Go Mun phase. Go \1'.r:: i, l,r..rted onlr'3 knr nonheasr of Phung Nguyen, and indeed the 30 or so knos-n Go \1u: sit3s are located u'ithin the sanre general area as those of the Phung Nguven anC Do:g Dau nhases. There have been four canrpaigns of excavations at Go Mun, conmencir:s in 1951 and finishing a decade larer. In all. 1.500 square metres were excavated. tl:e cui:ural scrariqraphv being only 1 mecre deep. As at Phung Nguyen,

stone adzes \\'er- 1:r ::.3 rnain quadrangular in section, and onh'one shouldered. Similar

stone chisels to tl:ose liorn Phung Nguyen were also encountered. The inventory of bronze artetacts rer':als that manl' forms in stone were copied in metal. There are also bir-alve nrou^c.. to: ;asring axes and arrowheads. Thirteen axes have been found, and seven chisels. The spearheads, arrowheads and bracelets were likewise rendered in bronze. Fishhooks s'ere the nost abundant bronze aftefact, followed by narrow projectile points. One sick-le \\-as recovered, and the figure of a seated individual which was presunrablr' cas: bv means of the lost wax technique. The Dong Dau pottery bears cun'ilinear anj rectanqular patterns which look like developed Phung Nguyen motifi, and ir is in rhese that Ha Van Tan32 sees models for the decoration later found on the s'eil-knos'n Dong Son drums. It is app.arent from the excavation carried out at Go Mun and related srres rhat bronze s-orkirrg was increasing in intensiry and range of artefacts. There \\'a-i also a coastal aspect to settlement in the Bac Bo area. Vietnamese archaeolosists see the earlv Neolithic site of Cai Beo and the Da But grouping of sites as ancestral to various so-called nriddle and late Neolithic cultures which occupied the coastal tract north aird souch of the Red River delta. Thus, the Ha Long culture of the islands in the northern secror of the Gulf of Bac Bo is seen as derivative from a Cai Beo context. Ther- base rhis opinion on the similarities in the rectangular and shouldered adzes recovered. Sires of this sroup also yield polished stone bracelets, beads and pendants not dissirnilar to Phune Nquven examples. Moving south oithe delu. \\'e encounter sites of another coastal group, called after the site of Hoa Loc. This eroup is knos'n for a most unusual form of decorated pottery with a quadrangular shaped rinr. Thev also used clay seals with deeply excised patterns resembling those found nruch later in Iron Age contexts in Northeast Thailand. Stone adzes and grinding stones are corunon and large stone hoes occur. The faunal associations from Hoa Loc are dominated bv marine species, but Vu The Long has also noted the remains of donresricared doe. pig and probably cattle. Le Van Thieu33 has indicated that sonre of the pigs' teeth displav a birnodal size distribution, probably
56

FI\OM THE NEOLITHIC TO THE IIT.ON A(;E

reflectins the presence oi both u'ild and domesticated aninrals. Hoa Loc itself is locate d on a raised sand L',ar nou' about -l kilonretres behind the shorelirre. It is e'vidertt that as the sea ler,el tell back tiorrr its post-Pleistocene high, so coastal settlelnent lollowed it. A ferv pieces of bronze \\'ere tbund at Hoa Loc when it was excavated in \976. The site seenls to equare. in rernrs of its nretallurgy, with the latest Phung Nguven subphase in the inland
area.

The geographrc exterlt of the Southeast Asian Bronze Age tradition is not yet defined. A.lrhouqh nlan)' sites are knorvn from the nrouth of the Mekonq to Yunnatr, and fronr Lingnan (southern China) to central Thailand, its rvestern linrits have onlv recently been exrended rvith the discovery of the Nyaunggan cenleten' in Burrnl.'" The Chindwin \:alley' is rich in copper ore, and Nyaunggan lies on the flank oi .rti old volcano overlooking its broad flood plain. Excavations there have rer-ealed an e\tensl\-c cenletery in which the dead were buried with many pottery r'essels. stone ornanlents and bronze spearheads which recall Mainland Southeast Asian forms. Once aq;rin. aninral bones acconlpanied the dead. This nrust be one of nranv such sites in the Chrndu'in and presunrably the Irrawaddy Valley, and a neu' chapter in defi-rine the Bronze Age oi Southeast Asia is now unfolding. The Bronze Age thus developed rvithin those late Neolithic communities exposed through exchange to knowledge and ideas originating in the early states of China. The economy continued to be based on rice cultivation, domestic livestock and hunting. fishing and collecting, and there is some linrited evidence for social distinctions based on inherited wealth and status. Copper and tin ores are relatively abundant, but have a restricted distribution far away from the nrajoriry of agricultural settlements. Hence, metal ingots were distributed by long-established exchange routes firmly related to passage by river. Marble, marine shell, serpentine, talc, carnelian, fine ceramics and doubtless many perishable items exchanged hands, all employed to exhibit status and to be immolated with the dead.

The Iron Ag" of rnainland Southeast Asia

The Bronze Age began within the period 1500-1000 nc and came to a close when iron entered the technological repertoire by approximately 500 BC. Iron ore is rlore rvidespread than copper or tin in Southeast Asia, and it is also available in the river valleys rhar attracted Neolithic and Bronze Age agricultural settlement. Relative abundance, coupled r,vith the superior qualities of iron once effectively worked, are qualities rvith :rofound cultural implications. The Iron Age of Southeast Asia can be divided into two distinct provinces. The first ...:-'.'ed societies that lay in the path of the predatory inrperial ambitions of the Han - ::-,:.' oi China. The second comprised those groups further south which u'ere onlr' ' ' ::----'. :fiected by Han expansion. By the first century BC, the Han had annexed

::-:.:- :reir irlperial schernes. The result was a marked surge in militarisnt. the ' . '--, -::-- -ocal chieftloms and the investment of much energy into s'ar. S/hen not : '-r: --- - ::-:,:.::::i the Chinese, these chiefcloms engaged in strife rvith their neighbours

i j-:- ' -- . ..::: -

".-:..,:-. :. :,os'here better iliustrated than in the I)ian chietdorrr oiYutrtratr. - - - -'j-:--::.:lis dating to the first century BC in \\'hicl-r tnen and wonlen

)7

CHAIILES HI(;IIAM

of the elite lineage were interred rvith

extraordin:lrv rllortuary wealth. Their large lacquered rvooden coffins were filled ri'irh outstanding bronzes, including drurns containing thousands of exotic co\\'n' shells. These drunrs conrnronly bore bronze figpres set in compositions on their n'rrr-

in scenes of s'ar and ritual. Thus we can adnrire the chiet^.

pana participating
resplendent in golden arnlorlr on his horse. surrounded by his follon'ers enraged in

vicious hand-to-hand contlict. One

loading a crossborv,
a

is

\\'eapon caken direcd,v

fronr the Chinese arnroun: Rirual scenes fulfilling the agriculrr-rral round can also be
recognized, often directed bv high-status wornen. Indeed. rhe roval cerneteries include the grar-es oi s'onren interred
'uvith

Figtrc

3.9 Bronze

superb bronze u-ear-ing inrplements. Again showing a nlan u'earing fine ornantents and a represented as eilded fizures, these female headband. leaders are seen receiving gifts, or ringed by a group of s'eaven. Pride of place among these scenes, however, nlust be accorded a nreetinq benveen the paramount and his subsidiary chiefs. Seated in a roofed pavilion surrounded bv cerenronial drums, the leader was the focus for the preparation of a feast involving che slaughter of cattle and pigs and the execution of war captives. Unfortunately, this luninous chietdonr, rvhich on its own accord stands out as one of the best documented from the prehistoric s,orld, was extinguished by incorporation as a pro\/ince of Han Dynasq' China.-'5 A similar florescence, followed by military defeat, afflicted the Dong Son culture of the Red, Ma and Ca valleys of northern Vietnanl. First recogiized on rhe basis of excavations at the cenletery and settlement of Dong Son. thrs group \s re\to\\r1ed for its giant ceremonia\ bronze drums. The drums \vere lavishlv decorated s'ith ritual and battle 'We are treated to vivid pictures of s'ar scenes. canoes brisciine s'ith plumed warriors brandishing spears an,1 firing bos's and arrorvs irom irghting p\atforms. There are ptocessrons. musrcal ensembles, rice threshing, eyen a scene of iour Dong Son drums set on a p\at{orm btir,,_:. beaten from above. The casting of such drunrs, together with the variery of large bronze vessels, nlust have required specia[st u-orkers under the control of

cemetery

of Lane Vac incorporating a hilt

dageer tronr rhe Dong Son

elite leaders. The huge settlenlent of Co Loa, which dominates the Red River plain north of Hanoi, likes'ise reflecrs control over a substantial labour force. Excavations within this site have uncovered a large ceremonial bronze drunr, many socketed bronze implements which might have been used as ploughshares or hoes, and a cache of bronze arrorvheads.3(' Several Dong Son cemeteries have also been exanrined. In the Red River valley, the.dead were often placed in cofhns fashioned from hollowed tree trunks. Their contents include weaponn' rvith the wooden hafts of bronze spears and axes, imported Chinese coins, rvooden bos'ls and the rernains of \r'oven clothine. The richest boat cofhn fronr Viet Khe includes nranv bronzes, sonle of rvhich were clearly of Chinese origin.'37
58

FRO\1 THE NEOLITHIC TO THE II\ON AGE

-.T.

:--:-:

:i.:::
:.1:

',r.'::

i :roiiteration

of bronzes in the Dong Son cemeteries, iron objects were

::-il-.':-'.

- --: \--_--=:: : .1i-' --, --r. "i---:ess of the Sky", and their energy for colonial expansion ebbed beyond ,'- :. -:-----::-- ::. j :.stern foothills. Beyond lay the rich lowlands of the Mekong and Chao , -,. . :.-.:: :'.'::3:ls. and a long coastline offering opportunities for regional and

.:'--.;:,-.--'' ::,i:.

Such trade could have been the spark which ignited the major social

--,--:.::. '.-..--:--:. rook place during the Iron Age of the Southeast Asian Mainland, : :: --: '. '.,.:::- the trading opportunities offered by the establishment of the Roman and ' -: ::::-.:::is. Horvever, the ready and local availabiliry of iron, and the potential for - :-:-:-'.': ernulation already apparent in the Bronze Age, should not be overlooked as - , .-::-: irvrrs to increasing social complexiry. -i- ;:ese variables are manifested in the wealth of the mortuan' offerings recovered :*:::S ihe excavation of the site of Ban Don Th Phet in cenrral Thaiiand.3E This site ,-,::r::ands the eastern approaches to the Three Pagodas Pass, a trade route linking the C:.o Phrava plains with the Gulf of Martaban and ultimatelr', India. It represents a .ieparture from earlier mortuary traditions in that it appears to have been a designated cemetery with no associated occupation or industriai remains, confined within a ditched enclosure. The available radiocarbon dates suggest an early fourth-century nc date for conunencement, by which time exchange contact with India brought in glass, carnelian
and agate ornaments, including a carnelian figurine of a rampant lion. Unfortunately the acidic soils have militated against the survival of human bone, so we have no means of identifying the age and sex of the burial assemblages. Yet the wealth of the grave goods far exceeds any encountered during the Bronze Age. There are iron spears, harpoons, axes, billhooks, bronze ornaments, and a rype of bronze bowl of probable Indian origin, fashioned from an alloy containing over 20 per cent of tin. This provides a golden finish, but also makes the metal so brittle that it is easily prone to fracture. These bowls, turned on a lathe to produce a very thin wall, were decorated with languid images of elegant women, animals, plants and structures quite at variance with the austere ceremonies and battles seen on the contemporary Dong Son drums. This bronze tradition, which must have called upon specialist craftspeople, also fed exchange nerworks. High tin bronzes have been found in Khorat Plateau sites such as Ban Na Di. A double-headed animal ornament of nephrite was also found at Ban Don Ta Phet, the parallels for which lie mainly to the east along the coast of Vietnam and into

:-e Philippines.3e ?::: Don Ta Phet is the best documented site in Central Thailand which exhibits Iron -.a- :::ial conrplexiry. The great limestone cavern of Ongbah was also used as a :--:i:1-.. the dead there being interred in hardrvood boat coffins.a0 'We knorv fes' - : .- . :'-: :he site 'uvas looted before the Danish archaeologist Per Sorensen could ---.-::-::: excavations, but the rvealth of the offerings gleaned from local villager
-.1

'.'.-.---.-- '.

:t- z-,-::::lanied by iron weaponry and exotic ornanrents.al :. r.. =r,J social comple*iry have long been recognizecl in large '.:.: . r:rl coast of Vietnanr. These sites have been nanred ,-::--:.=-: --::r-.:t:r.:
>::--...,r

.: -:::

for it eases castabiliry At Noen Ma Kok, - ::: --:i a range of artefacts which matches that from Ban Don Ta Phet.lr -.: ;.-. :' .:.'dine Hi, Pautreau has recovered a sample of Iron Aee graves in
--,: -:.::r: of Iron Age bronzes,

rr:r - '

59

C]HAI{LES HI(;tIAM

collectivell'after the villaee of Sa Huvnh. near u'hich Vinet discovered jar burials as long ago as 19O9. The rrrortuary renrains inclucle srrr.rll pottcn' r'cssels, irort tools, glass and carneiian beads, and sites extend fronr the H;ri \.rn l)ass to the viciniry of Saigon, a distance of over 700 kilometres.a'3 The crenration rite. in s'hich the incinerated bones were placed in lidded vessels, again rneAns thlr s'e knos' nothing of the age or sex of individual internrents. This nrortuarv ritr.r.rl .r)ntri\ti s'ith the s-idespread and enduring tradition of inhunration. It is highlv lrkeh rh:: ihe Sa Hr.rvnh people were directly ancestral to the Chanr civilization rr'hr;l ritr::r:r.riii ihis long coastlirle for a thousand years fronr the first ferv centuries.it'. Ti:e Ci:,::::;.::i*u;ges are Austronesian, with
closest relationships u,ith N4alav and c:i::.:n:,:.r::.::: I.l:::i Sc-rutl'reast Asia. At present, it seenrs likely that the S.r Hur-nl: --'.:.:'.::: rc:riii:r:i: l::.r::rillenniunr BC nrove into southern Vietnanr bv grou;''s s;:;k:::S i:: -\ustroneslJ;: i.:nr.r;ge.

Until recently our knou'lec:e .-rl rhe Iron Age in the \lekernc \':rllev, a vital area that later to witless t[e rrse e'.'jr].e .l:ie oiApgkor. s-as ii::ie r;r-r\,.-n. Tlie period between ap 1. and 500 in p.rrticul;: r'.'.: b:::,i'docunrcnted tirrough cx..r'.'a:ron. This situation is
rvas

now being addressed. Er--,','-::tr:S it the qreat urban centre of -\nSkor Borei, just above the delta, har-e rer-e.il.'J ::: Ir..i: ^\ge cenreter)' at the base of the se qr-le rtce (see Miriam Stark's chapter in c]:rs vt ^ui::e . A separate field progranxrre in the upper valley of the Mun River in ef,:ii:n Thailen,-l has involved the opening oi nr-o iarge Iron Ag. settlenrents u'ith s:artlr:S results. The broad iooopi;in of the Mun River suffers one of the longesr dn' seasons in N4ainland Sourheast Asia. r'et aerial photographs taken half-a-century ago reveal a dense distribution ol setrlenlent nrounds rvhich seenl to have been ringed by broad moats and raised banks. Sr:ral-l test excavations in several sites before 7996 had revealed evidence for Iron Age occupation. s'ith deep sequences and intriguing evidence for clay structures. At Ban Don Phlone. Nirtaaa uncovered a series of 1,7 iron-smelting furnaces overlying an Iron Age cenleten' in s-hich the dead were interred u'ithin clay-lined coflins which date to the last centuries nc. One man had been interred rvearing a necklace of agate and glass
beads. s-hile bronze bangles and rings were relativel)'abundant, especially rvhen compared u'ith the rturrtl.'ers in Bronze Age assemblages. At Non Yang, Nittat excavations revealed a series oitinrber and clay structures, the r.r,all of one roonl surviving to a height of 30 cm.

associated s-ith a curving row of post holes also ran across the excavated area. Excavations at Ban Takhong and Ban Krabuang Nok cornprised deep occupation layers, evidence tbr iron snrelting and burials rvithin lidded jars.a5 All these excavations eltcouraged a proeranrlle of fieldwork designed to uncover an extensive area, linked witli the exatnination of the moats and evidence for environnrental chanqe Such a proqrelntnc'began in 1995, in a defined studv area \\'est oithe later Angkorian centre of Phinrai. There is a dense concentration of Iron Aqe settlenrents in this region, ou occasiolt so close that they are alnrost 'uvithin hailing distance. Noen U-Loke was chosen for the closest attention, largely because an earlier exc:rvation b1'Wichakana had revealed a deep, scratified sequence containing inhunration burials (Figure 3.10). An area 15 by 14 metres has been opened to a depth of 5 nretres. u'hile long cuts up to 6 metres
.

A ditch

deep have been placed through the surrounding nloats

function.

to define their form

and

Noen U-Loke \\'as first occupied during the late Bronze Ag. and was finally
abandoned about au -l(X|-500. Most fortunately, the soil conditions lavour the survival of botte and 12(r btrrials have been rccovered, divided into five successivc phases (Fieure 3.1 1). 60

\E()LITHIC I () THE IRON

AGE

:
'

5O0m
rhe old channels.

tr EXCAVATION
lJ-Loke, showing the area excavated and

. -.:.: -, :lso a wealth of evidence for the technological and economic activities at the ': -:.:s. for the first time, it has proved possible to trace the fortunes of an Iron Age -'--:-.-:::in' over a period of a thousand vears. .:.: ;nost compelling evidence comes from the remains of the people and their - '::'*,:-; rituals. The deepest layers record the transition from the Bronze into the lron '..- '.'.:3n the site was still relatively small and surrounded by a braided nefwork of -:;--.--j a:rci sn'amps. One of the earliest graves contained a woman, interred with iron

-----. -:. tor a slightly later grave of a man included bronze bangles and neck rings, but r.- .:-:. .ocketed spearhead and a socketed hoe. He was accompanied by a tiger canine ,.- : :' - bronze spearheads, pottery vessels containing fish skeletons, shell discs for

- - -)::rs of graves, each containing the remains of nlen, women, children and -- ::=cple wore bronze ornaments including rings, bangles and anklets u'ith - =. :he earliest glass and agate jewellery in the sequence and the entire -:.. pigs. Carnelian was confined to the third phase graves, rvhich also
1

' ,
--

- ' '. . - : --:rnelian and agate ornaments (Figure 3.12) have been subjected to :- :. ind are thought to have conle from Thai quarry sites. These phases -.
i:-.:,::ics
LJC.

61

(]HAI\LES HIGHAM

FlFgrrrc
-1.

t-]
UP2

FItF]ahb^'b@
MP3 MP'
MPs

Pian oi the Noen U-Loke cenletery shou'ing the fir'e nrortuar)' phases.

Therc \\'ere nrajor changes rvith the fourth phase. Again. u'e find rice-filled graves, but the ranqe and quantiry of grave goods in each of the grave clusrers greath' increased, and

clar'-lined and lidded coflins were now favoured. At least one outstandingly rich individu;il s'as present in each cluster. One contained the renrains of a nrale rvith argtrabh' nlorc bronze than the accunru-

lated rotal oi all IJronze Ag"

graves

previousiv krtou'n in Thailand. He \\rore 15f) bronzc bangles. equall,,' distril'ruted on each arnr, bronz.' toe and finger rings, and three bronze br'lts. His ear coils were nrade of silver covered in gold foil and an iron knife lay beside his left hand. Further grave goods included several splendid pottery
vessels and slass beads bv the neck and ankles. I-]urial (r9 in a second cluster, also a ttt:tle, wore fottr bronze belts, heavl' bronze
Fi.ytrc
{rraves

t,

w,'
Higharrr.)

,t,tr:.l t

3.12 Agate jervellery lourrd in the later of Noen U-Loke. (Scale 1 cnr.) (Charles

62

Fir.()\1 I ttE NE()LITHI(l T() THE ll{()\

.{(;E

and nrllrv bangle s ancl rirrgs . Aqain, he u,as accolnp:lnied by

knite. The rvornan in a third cluster rron knife and vvore a necklace of .-:.: "nd agate beads, silver baneles and :: : zi' rir.tgs, anklets and bancles (Figure -: .j . Large, thin-rvalled pottery vessels .,.::j placed beyond her head. These ricli j:rlcs have been dated to between Al) 1(X) ,::J 3f)0, a period rvhen exchanqe' trp arrd iou'n the Mekonq arrd its tribut:rries qrc'u' ur response to the rapid developnrents in rlaritinre trade. In rrricrocosllr, we can see :rt Noen U-Loke the srvift rise of leaders. s'hose rvealth turned not only on nrorc

intensive agriculture, but also on the

production of salt on an industrial
scale

fl ll ,l
I

;l

and participation in burgeoning trade. The fifth and final nrortuar)/ phase at

Noen U-Loke saw the end of the ricefilled graves and tight clustering of burials. and a reduction in the u'ealth oi grar'.' goods. However, \\/e can also recogrrize other important changes. One nran \\'as

found interred with an arror.vhead through u'earing gold, agate, silver and bronze jewellery. his spine. This was also a period rvhen iron (Buriai 113.) (Charles Higham.) weaponry proliferated in the occupatiorl layers. Iron sickles were now placed as grave goods. There appears to have been a growth in militarism and agricultural intensiry ',vith the grou'ing population and foundation of new settlements in the valley. This final phase betore che site u'as abandoned brings us to the verge of early states in the region, rnanifested in the construction of brick temples dedicated to the ancestors, local gods and gods frorn the Hindu pantheon. For years, the nature of the moats around nlan)' of these Iron Age settlements in Mainland Southeast Asia has been enigmatic. Bv cutting deep sections through them at Noen U-Loke and other sires in the study area. Willianr I3oyd and his colleagues have identified their structure and date.a(' He has for.rnd that the banks round the site confined \\rater brought into the nloats by natural strearn channels. AMS radiocarbon dates fronr the banks and the channels thenrselves date these huge water-control structures to the late Iron Age, benveen abour nu 100 and 5()i). We still lack evidence for the purpose of the nroats, which nright r,vell have been ::rr-rltiple. Defence of the settlenrents is one possibiliry for this was the period s'hen rhe :-iunberof iron arro'nvheads increased nrarkedly. As populations gre'rv, so there n'ould also lri'e been a need for a predictable dornestic water supply in the dry season, n'hile extensir-e nroats u'ould also have provided a source of food. However, changine river parterns, :'-:ickened by deforestation and increased siltation, nlay well have caused rhe river ch:rnnels

Figrrr:

3.13 Noen U-Loke burial of a woman

-'-- ,:. nou' lies. utioccupied, sonrc distance frorn

63

the village of lJ.in Nonq Na and its

(]HAI{LES HI(;HAM
river. In this regard, the local snreltirrc of iron. castinq of bronze arrd production of salt had profound environtnental consequences. tor all led to increasinq deforestation. The upper Mun Valley contains tu'o nrajor rernple sires dacing to the period ofAngkor (fronr at> 800). Excavations at both have revealed earlier Iron Age occupation. At Phanonr 'Wan, an Iron Age cerneten' flrnks rhe central sarlctuary and contains a grave sinrilar to those from Noen U-Loke h-ins on rhe sf,nre orientation as the sanctuary above. A layer with Iron Age potten' r-rnderlies thc cenrral ternple at Phinrai.aT These sequences strongly support local origins tor the regional expressions in Thailand of the civilization of Angkor. [n central Thailand. sites like Ban Don Ta Pher anticipate the genesis of the state of Dvaravati, s'hile the Sa Hr.rvnh culture \\-A( ro develop into the civilization of Chanrpa. Thus can \\'e .rfaprccl.ltc hos- the trrst civilizarions oi So.'rtheast Asia had their origins deep in the prehiscoric pasr.

Notes
2 Highanr and Thosarat 1998: Bovd 1998; Mason 1998; O'Reillr-199,!. 3 Highanr and Bannanuraq 1990, Higham and Thosarat 1991. Pisnupone i993: Mason1.997. 4 Tavles 1999. 5 Highanr and Lu 1998. Pei 1998. 6 Lai Van Tor 1999. 7 Hoane Xuan Chinh and Nguyen Ngoc Bich 1978. 8 Ha \an Tan 1980. 9 \guven Kini Dung 1990, 1998.
1[t Hoane Xuan Chinh 1968. 11 Ha Van Tan 1993. 12 Bui Vinh 1995. 13 YP\,I t977 . t981. 1-l Bavard 1971. 1972a, 1972b. 15 Rispoli 1997. 16 Sorensen and Hatting 1967 . 17 Leong Sau Heng 1991. 18 DifHorh 199-t. 19 Solheinr 1968, Gorman and Charoenwongsa I976. 20 Meachanr 1993, Moore and Pauk 2001.

1 Anderson 199t,. -\ibrechr cr ,ri. n.d.

21 LiangJingin 1978. 22 GXBWG 198,+. 23 An Zhrmin 1998.

24 Meachanr 1975, 1994. 25 Natapintu 1988, Pigott ct al. 1997, Pigott and Weisgerber 1998.

26 Higham and Krjnganr 1984. 27 Highanr and Thosarat 1998. 28 Ha Van Phune 1979. 29 Ngo Si Hong 1987. 30 Trinh Sinh (1977). 31 Nguyen Kim Dung 1998. 32 Ha Van Tan (1980).
33 Le Van Thieu (1979). 34 Moore and U Pauk 2001.

35 Pnazzoli-t'Serstevens 197 4. 36 Nguyen Giang Hai and Neuyen Van Hung

37 VMH t965.

1983.

38 Glover 1980, 1983; Rajpitak and Seeley 1979.

64