The East Asian food production complex eventually came to dominate many of the northern and western islands of Southeast Asia as a result of migration out of southern coastal China into Taiwan around 3500 bc. This was followed by a crossing from Taiwan into the northern Philippines around 2200 bc (Figure 8.4).32 The complex underwent considerable transformation after 1500 bc in the eastern islands of Indonesia, with the loss of rice and millet cultivation33 and a switch to tree and tuber crops such as bananas, coconuts, breadfruit, taro and yams. Some of these were of Island Southeast Asian and western Pacific origin, and they lead us into the western Pacific food production complex, which evolved quite independently of that in East Asia. The western Pacific complex contributed to the major migrations by Austronesian-speaking populations into Oceania after 1300 bc, culminating after ad 800 in the colonization of the remotest islands of eastern Polynesia.
The most important economic players in the genesis of this western Pacific agricultural complex were the ancestors of the indigenous Papuan-speaking populations of New Guinea and adjacent islands. The New Guinea highlands, in particular, form a unique environment in world terms, being well watered, high in altitude, equatorial in climate (hence without marked rainfall seasons), and well protected by mountain walls from external population intrusion. They are also high enough in altitude to be malaria-free, an important consideration given the evidence for the rather disastrous effects of malaria in some Southeast Asian lowland Neolithic societies, as well as in lowland regions of Indonesia and Melanesia today.34 The New Guinea Highland archaeological record indicates that people there were cultivating bananas, taro and perhaps sugar cane by at least 4500 bc. But no cereals or domesticated animals were present, so levels of productivity compared to those in China were quite low.35
There is no genetic or linguistic evidence that suggests significant Holocene population expansion from New Guinea towards Asia, unlike the huge body of evidence that points to movement the other way. But, as far as the western Pacific as a whole is concerned, the genetic and cultural contributions from western Melanesia were quite dramatic. Within New Guinea itself and adjacent islands, the New Guinea Highland variant of the western Pacific food production complex probably gave rise to the radiation of the early Trans New Guinea languages, which ultimately came to dominate most of the island, especially the interior.36 Although many coastal New Guinea and Melanesian island populations today speak Austronesian languages as a result of language shifts in prehistory, the modern populations themselves, especially in
Figure 8.4 Holocene population movements through Island Southeast Asia and across Oceania, according to archaeological and comparative linguistic data. This map is also published in Bellwood 2013. Original by Multimedia Services, ANU.
Southeastern Indonesia, and in the Solomon, Vanuatu and New Caledonia island groups, are genetically of mainly indigenous Melanesian origin.37 We return to this issue later, since there was also a degree of Melanesian genetic input into the genomes of Polynesians, who are otherwise a population very clearly of Asian rather than Melanesian origin.
As for as the initial outflow of Neolithic populations off the Asian mainland into Island Southeast Asia, an agricultural way of life with pottery, stone adzes, rice, millets, pigs and dogs had spread southwards by 3500 bc, through the coastal provinces and offshore islands of southern China, to reach Taiwan. A long pause in migration then occurred in Taiwan until around 2200 bc, when the archaeological record reveals the commencement of a spread of related cultural complexes, with red-slipped and often stamped pottery, domestic animals, and similar kinds of stone and shell artifacts, through the Philippines into Indonesia and the islands of the western Pacific. This migration reached Guam by 1500 bc and western Polynesia (Tonga and Samoa) by about 900 bc.
Important discoveries related to the earliest Neolithic way of life in Island Southeast Asia have recently been excavated at Nanguanli in southwestern Taiwan, associated with the Dabenkeng phase of Taiwan prehistory. Rescue excavations here into waterlogged deposits 7 m below ground level, dating between 3000 and 2500 bc, have produced pottery with cord-marked, red-painted and red-slipped decoration, stone bark cloth beaters, perforated slate projectile points, shouldered stone adzes, baked clay spindle whorls, tanged shell reaping knives, and shell bracelets and earrings. Found with these artifacts were complete dog burials, pig bones, and carbonized grains of japonica rice and foxtail millet. The Dabenkeng complex was introduced decisively into Taiwan about 3500 bc, replacing a very sparse flaked stone assemblage known as the Changbinian, associated with earlier hunter-gatherer settlement.
In Neolithic Taiwan, there is very good evidence for considerable population growth between 3500 and 2000 bc, with a 20-fold increase in site numbers during this period in one region near Taipei. There was also a 20-30 times increase in total site area, with individual sites reaching a maximum extent of 60 ha. Similar data are available for eastern Taiwan, indicating that the period from 3500 to 2000 bc was one of considerable growth throughout the island.38 It is striking that the movement of Neolithic populations from Taiwan into the northern Philippines, at around 2200 bc, coincided with such a high population density in Taiwan. It also coincided with some potential evidence for landscape degradation, especially in the Penghu Islands that lie between Taiwan and Fujian.39 A desire for new agricultural land emerges as a possible motive, fuelling further migration.
Beyond Taiwan, the Neolithic settlement of the Philippines is well recorded in the archaeology of the Batanes Islands and the Cagayan Valley in the north of the country Finely cord-marked pottery of a type dated to before 1500 bc in southern Taiwan occurred first, in Batanes only, then red-slipped pottery of southern Taiwan ancestry replaced the cord-marked and spread into Luzon.40 Taiwan jade occurs in Neolithic sites in both Batanes and the Cagayan Valley, imported from the Fengtian source near Hualian in eastern Taiwan. The Cagayan sites have red-slipped punctate and dentate stamped pottery which is difficult to trace to any particular source in Taiwan, but
This type of decoration occurs in Dabenkeng contexts at Nanguanli and also has very distinctive parallels in Yangzi basin Neolithic complexes such as Daxi (circa 4000 bc).
South of the Philippines, archaeological traces of Neolithic expansion during the middle and late second millennium bc have been found in a number of rock shelters and open sites in Malaysian Borneo, Sulawesi, eastern Java, Timor, and the northern Moluccas.41 These sites contain occasional evidence for rice, with mostly undecorated red-slipped pottery and polished stone adzes. The archaeological record has many gaps, but it is possible that two cultural streams might have been involved in the movement southwards from the northern Philippines. An eastern one with mainly plain red-slipped pottery and a diminishing reliance on rice eventually entered the Moluccas and Nusa Tenggara (the Lesser Sundas). A western one, via Palawan and western Borneo, ultimately reached Sumatra and Java with paddle-impressed pottery and continuing rice cultivation (Figure 8.3 and Figure 8.4). But for the moment this is merely supposition, as additional field archaeology is required in these regions.
The Colonization of Oceania
In Micronesia, the first settlers to reach the Mariana Islands were evidently derived directly from the Philippines, according to parallels in the earliest red-slipped and stamped pottery in these two regions.42 They also perhaps took rice (but not pigs or dogs) with them, which would have been the only occasion this crop was transported into Oceania. The open-sea crossing to the Marianas from the Philippines, at least 2300 km, justifiably ranks as the first crossing of this distance known in Oceanic or even world cultural history, given that the archaeological record so far fails to indicate any other route except for the most direct one. The Palau Islands were settled about 1000 bc, but they reveal no direct archaeological links with the earliest Marianas assemblages. The other islands of southern Micronesia, especially the Carolines (mostly atolls), were settled after 500 bc since the coral reefs responsible for atoll formation did not provide habitable land above the Holocene sea level until this time (Figure 8.4).
Between 1350 and 900 bc, Neolithic colonists in Island Melanesia (excluding New Guinea) and western Polynesia (Tonga, Samoa, and adjacent islands) left an extremely clear-cut trail of sites belonging to the Lapita cultural complex (Figure 8.5).43 In New Caledonia, Vanuatu, Fiji, Tonga and Samoa, Lapita marked the first documented human settlement. Of course, the more westerly islands from New Guinea out to as far as the Solomons had already been long settled by pre-Neolithic indigenous populations, some practicing horticulture or arboriculture, but these groups did not use pottery or keep domesticated animals before Lapita times. Interestingly, early Lapita sites have never been found on New Guinea itself, although there are a few sherds from the northern coast and some new discoveries of rather late Lapita sites on the southeastern coast (McNiven et al. 2011). Lapita settlers appear to have had preferences for coastlines with fringing coral reefs and lagoons, and such environments were rare around most of New Guinea proper due to rapid rates of uplift along the narrow northern coastline and the rapid alluvial sedimentation in the vast swamplands in the south.
Figure 8.5 The distribution of Lapita sites in Oceania. This map has also been published in Spriggs 2013, and is reproduced with the permission of Matthew Spriggs.
This impressive Lapita migration involved Neolithic populations whose archaeological origins have long been assumed to be in eastern Indonesia. However, the closest parallels in terms of decorated pottery and shell fishhooks occur much further north, in the just-discussed Cagayan Valley sites in the Philippines and in the Mariana Islands. One current view44 is that an identifiable population movement occurred from the Philippines, via the Marianas, and directly south into the Bismarck Archipelago, the latter serving as the immediate homeland of Lapita within Melanesia. From the Bismarcks, subsequent movements prior to 1000 bc must have gone west into eastern Indonesia, generally avoiding New Guinea itself but carrying black volcanic obsidian from sources in the Bismarck Archipelago to as far west as Sabah in northern Borneo (this movement is indicated in Figures 8.3 and 8.4). At the same time, groups possibly moved the other way, from eastern Indonesia to the Bismarcks (again avoiding New Guinea), although the only evidence for this at the moment appears to be linguistic and not archaeological.
Lapita pottery is well known for its decoration, often red-slipped, with horizontal zones of intricately incised and dentate-stamped rectilinear, curvilinear and even anthropomorphic motifs, the latter perhaps indicating the concern with ancestors common to all Austronesian speaking populations. Late Lapita pottery tended to have simpler designs. Dentate stamping faded in popularity after 750 bc in favor of plain ware in western Polynesia, although other styles of incised, applique, and carved paddle-impressed pottery continued until late prehistory in some Melanesian archipelagos and parts of coastal New Guinea. Apart from pottery, other items of Lapita material culture similar to those found in contemporary sites in the Philippines and eastern Indonesia included stone and shell adzes, and a range of shell ornaments including beads and arm rings. Lapita shell fishhooks, however, find their closest ancestral parallels in the Marianas.
Lapita village settlements, in some cases of stilt houses over shallow lagoons, averaged about 1 ha in size (maximum 7-8 ha) in coastal and small offshore island
Locations. Economically, Lapita was based on a mix of horticultural and maritime subsistence, but without rice or other cereals. Plant remains from waterlogged sites in the Bismarck Archipelago include taro, coconut, candlenut, pandanus, canarium nuts and bananas, most exploited by pre-Lapita indigenous populations in western Melanesia. Pigs,45 fowl and dogs were all present in Lapita sites, although not all sites or island groups have yielded the same results. Lapita settlers were initially distracted away from their domesticated food supplies by prolific wild resources (littoral and avian) in the areas they colonized, until these became reduced through extinction and local extirpation.
Beyond Melanesia, Lapita colonists reached Tonga and Samoa in western Polynesia by about 900 to 850 bc (Burley and Dickinson 2010). As in Melanesia, the decorated forms of Lapita pottery lasted for only a few centuries, eventually turning into a plain ware of increasing thickness before the eventual demise of pottery in Samoa and southern Micronesia by ad 300. Without rice, Polynesians did not need pots in which to boil their food, and managed instead with earth ovens and open hearths. Together with the earlier losses, prior to reaching Oceania, of rice and millet, loom weaving, and the associated clay spindle whorls for spinning fibers, this eventual loss of potterymaking suggests bottleneck cultural simplification as small groups pushed ever further east, gradually losing contact with their more complex homeland cultures and leaving behind aspects of cultural knowledge.
Nevertheless, although Polynesians might well have lacked rice, pottery, and woven cloth, they (or some of their immediate ancestors) reversed cultural simplification in a very impressive way by inventing the double sailing canoe, a remarkable construction that allowed the discovery and planned colonization of islands located thousands of kilometers over the horizon. Indeed, double canoes could have aided the much earlier colonization of the Mariana Islands from the Philippines, although this form had been replaced by double outrigger canoes in Island Southeast Asia by historical times. Polynesians also developed complex forms of terraced-field and canal-fed taro irrigation, as well as dry stone platform monuments, massive carved stone statues (on Easter Island), and palisaded earthwork fortifications, the latter reaching an apogee in Maori New Zealand (Bellwood and Hiscock 2013).
The settlement of the islands of central and eastern Polynesia - the Marquesas, Societies, Cooks, Australs, Tuamotus, Hawaiian Islands, Easter Island, New Zealand, and many others - occurred after a long migratory pause in western Polynesia. None were settled before ad 700, and most not until after ad 1000 (Wilmshurst et al. 2011). Easter Island with its famous stone statues was long thought to have been first settled almost 2000 years ago, but recent research has brought this date forward to circa ad 1200. Why the rapid spread into eastern Polynesia occurred after such a long pause in Tonga and Samoa is not clear, but many Oceanic atolls were still below sea level until as recently as 1000 years ago, hence making some sea crossings much longer. However, eastern Polynesia was not the only region of Oceania to see quite intensive seaborne colonization around ad 1000, since many atolls and formerly settled high islands in Melanesia and southern Micronesia were also settled at this time by a process of reverse 'Polynesian Outlier' migration, assisted in its general east to west direction by the
Prevailing trade winds.46 These movements led to replacement or assimilation of earlier Malayo-Polynesian-speaking resident populations in many cases.
Apart from the settlement of the Mariana Islands and eastern Polynesia, another extraordinary feat of long-distance colonization carried Austronesians across the Indian Ocean westwards to Madagascar and the Comoro Islands (the latter now Bantuspeaking), probably in the mid-1st millennium ad. Madagascar itself was evidently reached by colonists from southern Borneo, perhaps with Malay - or Javanese-speaking leaders, presumably during the existence of a Sumatra-based early Indic kingdom known as Srivijaya. Interestingly, the most recent evidence in support of an Indonesian origin for the Malagasy comes from genetics and linguistics.47
Why did all this island migration come about? Simply looking for new islands for agricultural land or other resources does not explain everything, given the huge sizes of many of the islands of Southeast Asia, even now underpopulated in some remote equatorial areas. Periodic increases in the frequencies of westerly winds could have encouraged sailing to the east (Anderson et al. 2006), but this factor is unlikely to have started and maintained the whole migration process. My preference is for founder rank expansion, based on those younger sons and their families, able to found only lineages of junior rank at home, who sought to found new senior lines through the colonization of new territories. The genealogically ranked societies of Micronesia and Polynesia provided ideal circumstances for this process to operate.
But one question remains. Why did the spurts of migration occur when they did? The history of Neolithic migration into Oceania clearly occurred in two rather explosive phases, one dating to 1500-1000 bc, the other to around ad 1000-1200. It is almost as if the first migration became exhausted in the face of increasing interisland distances, and the second then depended on some lucky travel-related advantage. Increased incidence of winds blowing from the west, the invention of the double canoe, even the emergence of atolls above sea level can all be suggested as promoters of the second push into Polynesia, but I find none of these very convincing in themselves. The answer might be pure chance. Perhaps no one was lucky enough to find new land and return home to tell the tale after western Polynesia was settled about 900 Bc. Someone else might have been much luckier on a chance voyage eastwards around ad 1000, and the floodgates opened. Founder rank enhancement reigned anew, at least for a while. This would also explain why the later voyages took place not just over vast distances into eastern Polynesia, but also over much shorter Polynesian Outlier distances back westwards and northwards into Melanesia and southern Micronesia. It might have been the social desire to find new land that was revolutionary, if short-lived, and not just the means of doing it.
The History of the Austronesian Language Family
Of all human ethnolinguistic dispersals prior to ad 1500, that of the Austronesian-speaking peoples was the most extensive. It can be related directly to the archaeological spread of Neolithic populations through Island Southeast Asia and the Pacific, as described earlier. Austronesian-speaking founder populations eventually migrated
Figure 8.6 The Austronesian languages and their major subgroups, as classified by linguist Robert Blust (also published in Blust 2013). Map production by Multimedia Services, ANU.
Through more than 210° of longitude, to as far as Madagascar, Easter Island and New Zealand, through an astounding west to east 22,000 km spread of ocean and islands (Figure 8.6). The whole dispersal required more than 4000 years for completion.48
Today, there are more than 1000 Austro nesian languages, making it the second-largest language family in the world in number of languages, after the Niger-Congo family of Africa. The language family first crystallized in Taiwan, where several of its most deeply rooted subgroups still exist today. But the most remarkable observation made about Austronesian languages is that all languages outside Taiwan, from Madagascar right across Island Southeast Asia to Polynesia, belong to only one subgroup, termed Malayo-Polynesian by the linguist Robert Blust.49 This subgroup is sufficiently homogeneous to support a hypothesis of extremely rapid initial settlement of the Philippines, Indonesia, and Island Melanesia, down to and including the phase of Lapita colonization, a hypothesis supported very strongly by the archaeological record and associated radiocarbon dates. Islands such as Madagascar, those of eastern Polynesia, and Micronesia beyond the Mariana and Palau Islands took much longer to reach, according to both the linguistic and the archaeological perspectives.
Comparative linguistic reconstructions make it very clear that the earliest identifiable Austronesian speakers in Taiwan (Proto-Austronesians) were agriculturalists of immediate southern Chinese origin who grew foxtail millet, sugarcane and rice, together with some tubers and fruits. They made canoes and rafts, probably used sails, and lived in timber houses.50 They kept pigs and dogs (chickens are uncertain, and unattested archaeologically before Lapita times), used bows and arrows, had some form of loom (backstrap?) for weaving, and made pottery. They did not cast copper or smelt iron. In archaeological terms, these were Neolithic societies with a material culture similar to that of many Pacific Island Austronesian communities who survived without knowledge of metallurgy until European contact.
Because the Philippines are tropical, whereas most of Taiwan and southern China are temperate, the earliest Malayo-Polynesian-speaking settlers who arrived in the Philippines after 2000 bc would have found new crops such as breadfruit, coconut and bananas to add to their food supplies. Yams and aroids (Colocasia and Alocasia taro) were also certainly cultivated by this time, as presumably was sugarcane. In the prevailing equatorial, ever-wet and nonseasonal climates of the southern Philippines and Indonesia, these indigenous fruits and tubers came to dominate the economy that spread into the Pacific beyond the Solomons, and rice and millet faded from popularity Some of these tree and tuber crops might have been domesticated earlier in New Guinea and adjacent regions, as related earlier, and carried westwards into Indonesia by Papuan-speaking populations. It is also possible that some were domesticated in pre-Austronesian times in Island Southeast Asia (Donohue and Denham 2009), although so far no archaeological evidence has come forward in support of this suggestion.
Indeed, the pre-Neolithic archaeological record in Island Southeast Asia generally suggests a presence of only very small scattered populations of hunter-gatherers, whose records survive mostly in the form of flaked stone tool assemblages in caves. Shell fishhooks have been claimed from pre-Neolithic contexts in Timor, but they remain unique so far across the whole of Island Southeast Asia at this time level.51
Given the exiguity of pre-Neolithic occupation in these islands, it is not surprising that the progress of Austronesian colonization was rapid and linguistically dominant.
One of the factors that distinguishes Island from Mainland Southeast Asia in linguistic terms is that only Austronesian language are spoken in the islands, beyond the Papuan-speaking regions in and close to New Guinea. The Southeast Asian mainland has, on the contrary, seen many episodes of linguistic overlay, with Austroasiatic tribal languages now submerged beneath various layers of Tai, Tibeto-Burman, Khmer, and Vietnamese. Although it has been argued that other languages, especially Austroasiatic, were spoken widely in Island Southeast Asia before the arrival of Malayo-Polynesian speakers (e. g., Blench 2010), there is no convincing archaeological, linguistic, or place-name evidence to give this opinion much credibility If Austroasiatic languages were once widespread, one would have to ask why and how they were entirely replaced by so many small and tribal Austronesian languages, given that the process of Austronesian expansion was certainly not associated with state authority. Of course, non-Austronesian languages of some kind must have been spoken in Island Southeast Asia before the arrival of Austronesian languages, but they were more likely affiliated with Papuan languages than with mainland Austroasiatic.
Moving now into Oceania, most linguists accept that the protolanguage of the Oceanic subgroup of Malayo-Polynesian, that includes all the Pacific Island Austronesian languages apart from Chamorro (Marianas) and Palauan, originated during Lapita times and most probably in the Bismarck Archipelago. The emergence of Proto-Oceanic has long been considered a result of population movement from Halmahera and the western tip of New Guinea into the Bismarcks, where a small degree of borrowing from Papuan languages occurred (Pawley 2007). But the archaeological evidence for links between these regions is not strong at all, far weaker than the Lapita archaeological links with the northern Philippines and Marianas. The number of linguistic innovations that link Halmahera with Oceanic languages is also rather few (Blust 1993). These could reflect later contacts, albeit quite early in relative date, rather than relationships defined at the very point of initial settlement.
Indeed, the present-day Malayo-Polynesian subgroup patterning might not reflect directly the very first years of colonization at all. This is because people at that time of significant migration, around 1500 to 1000 Bc, were still speaking a relatively coherent Proto-Malayo-Polynesian language, essentially then still a single language, just as modern British, Australians and North Americans still speak inter-comprehensible versions of English.52 Linguistic subgroups take time to emerge, and need a fair number of uniquely shared innovations if they are to be identified at all. This situation implies that the first Malayo-Polynesian-speaking migrants to travel beyond the Philippines could have gone in several directions at the same time - to Borneo, eastern Indonesia, and the Mariana Islands, for instance - with their inter-comprehensible versions of Proto-Malayo-Polynesian. It might not be possible to identify evidence for such contemporary movements purely from comparisons of the linguistic subgroups that exist today
The spread of the early Malayo-Polynesian languages and their speakers through the Philippines, Indonesia, and into the western Pacific during the latter part of
The second millennium Bc can therefore be traced in both the linguistic and the archaeological records. The flow was slowed sometimes by an existing presence of other non-Austronesian agricultural populations, especially by Austroasiatic speakers on the Asian mainland who grew rice, and around coastal New Guinea. Otherwise, the Malayo-Polynesian dispersal was only impeded by scattered hunter-gatherers, or by no people at all beyond the Solomon Islands. Beyond Tonga and Samoa in western Polynesia, settled by 900-850 Bc, ever-increasing ocean expanses within Polynesia intervened to slow the migration process down, for almost two millennia. As indicated by the archaeological record, most of eastern Polynesia was not colonized before ad 1000.
The eventual result of all this migration, spread through 4000 years from Taiwan to eastern Polynesia, was that the Austronesian and Malayo-Polynesian-speaking populations formed the most widespread ethnolinguistic group in the world prior to ad 1500. Laurent Sagart (2005b) has also suggested a Taiwan or northern Philippine ultimate origin for the Tai language family, as a cousinly subgroup to Malayo-Polynesian (Figure 8.3). This is most interesting but difficult to assess archaeologically, owing to the relative lack of Neolithic information from the key Chinese provinces of Hainan, Guangdong and Guangxi, the provinces in which most other linguists locate Tai origins. Early Neolithic links in artifact assemblages between Taiwan and the Pearl Delta region of Guangdong have been suggested by Tsang Chenghwa (2005: 71), but the archaeological record at this stage is too thin to allow any real testing of Sagart's hypothesis for the Tai languages.
Biological Anthropology and the Austronesians
Possession of an Austronesian language need not mean shared biological identity. Austronesian speakers in the Philippines, Indonesia, and the western Pacific are extremely varied genetically, ranging from Asian to Melanesian in their DNA affinities. As Johann Reinhold Forster asked in 1774, “What occasions the inhabitants of O-Taheitee to be so much distinguished from the Mallicolese?”53 Philippine and western Indonesian populations are mainly of recent Asian descent, but Philippine Negritos (Agta, Aeta) are not, yet they also speak Austronesian languages as a result of language change in prehistory (Reid 1994). Likewise, many Austronesian languages are spoken in and around New Guinea by populations who certainly are not of recent Asian descent, but who have shifted from Papuan to Austronesian languages in the past. One of the attractions of early Malayo-Polynesian dialects as spoken around 1500 bc would have been that they were remarkably inter-comprehensible over vast areas. This could have made bilingualism or even language shift attractive options for some indigenous populations of Island Southeast Asia and western Oceania, doubtless linguistically diverse after at least 40,000 years of prior settlement.
In recent years, geneticists have offered much insight into this admixture phenomenon, which was apparent not only to Forster but also to the famous naturalist Alfred Russel Wallace in the mid-nineteenth century. In the last few years, a great deal of new information has come from an analysis of the autosomal genome across many
Populations in Southeast Asia and Oceania. The results of these new analyses confirm that Asian populations penetrated Island Southeast Asia in large numbers between 2000 and 1000 bc, admixing increasingly towards the island of New Guinea with native Melanesian peoples, who still occupy the latter island almost exclusively Beyond New Guinea and Melanesia, however, Asian populations were the main colonizers of the uninhabited islands of Micronesia and Polynesia.
Genetic research into Island Southeast Asian and Oceanic peoples came to many dead ends in the late twentieth century owing to a dependence on mtDNA and very error-prone molecular clocks. In one case it was claimed that Polynesians were unrelated genetically to other Austronesian speakers in Taiwan and most of Island Southeast Asia, and evolved in mtDNA terms in the remote Paleolithic of eastern Indonesia, about 17,000 years ago.54 Such historical reconstructions are untenable in the light of archaeological and linguistic data, and require impossible explanations based on l anguage and culture shift through trade and other unlikely processes. They also require a complete disjunction between population and language history.
The current picture, focusing on the most recent analyses of whole genome autosomal DNA, together with considerations of both NRY and mtDNA data, is summarized by Murray Cox (2013). He refers to the new methods that use multiple ancestry-informative markers, drawn from across the human genome, as the 'gold standard'. Cox's general summary of the situation amongst Austronesian speakers is that Formosans carry an entirely Asian autosomal ancestry, whereas amongst Polynesians it is around 80%. Conversely, amongst western Pacific populations in and around New Guinea, including both Papuan and Austronesian speakers, the percentage of overall Asian autosomal ancestry drops to about 20%. Within Indonesia itself, a major shift from Asian to Melanesian genotypes occurs across the region between Bali and Flores, with a rise in Melanesian ancestry from almost zero in Bali to 50% in Alor, near Timor.55
This picture of genetic admixture between populations of Asian and Melanesian origin through Indonesia is given very strong support by an important new genetic study that uses whole-genome data from modern groups located along the Sunda island chain and in Borneo.56 The results indicate significant levels of population migration from the north and west into eastern Indonesia, dated by an autosomal molecular clock to around the second millennium bc. This is in excellent agreement with inferences based on Austronesian languages and archaeology. The results underline the great impact of Austronesian migration from Taiwan and the Philippines into Borneo, Sulawesi, and the Indonesian islands to the west of Bali and Lombok. They also underline how significant has been the indigenous western Pacific population in the prehistory of the eastern regions of Indonesia and the Pacific Islands. Forster and Wallace sensed the same result from their observations so long ago.
This new genetic perspective naturally gives considerable importance to Austronesian migration through Island Southeast Asia. However, this need not mean that huge numbers of colonists were sailing constantly forth in their canoes, unlike the recent European shipborne colonizations of Australia and the Americas. It is likely that actual colonizing groups were quite small in numbers, as many geneticists suggest,
But then expanded very rapidly once they found new and fertile terrain to support their large families. The more females in the colonizing parties the better. The same situation is being revealed by genetic research into the European Neolithic (Chapter 7).
Finer detail is available for the genetic makeup of individual regions. Western Indonesian populations carry mtDNA and NRY haplogroups that were evidently brought in by Neolithic migrants from the Asian Mainland, in some cases via Taiwan (e. g., NRY haplogroup O). Eastern Indonesian populations carry more indigenous Melanesian haplogroups, as expected, but also carry a large number of Asian haplo-groups derived via Neolithic Taiwan. Eastern Indonesia might also have been the source of the specific mutation in mtDNA haplogroup B that was carried almost to 100% fixation by early Polynesian migrants as a result of extreme founder effects.57 This is the 'Polynesian motif', B4a1a1a, recently identified from ancient DNA as the main haplogroup amongst the first New Zealand Maoris (Knapp et al. 2013). Polynesians otherwise acquired little of their genetic ancestry from Melanesia, apart from some of their Y-chromosomes, and in the early years of colonization there may have been relatively little genetic admixture between the two populations, explaining why the archaeological signature from Lapita seems to suggest long-distance leap-frogging towards the east over regions already settled by Papuan populations (Sheppard 2011). A recent craniofacial analysis also support a Polynesian ancestry firmly in Southeast Asia rather than Melanesia, albeit with some regional variation that might be mainly a reflection of founder effects (Buck and Vidarsdottir 2012).
The haploid lineages, however, do suggest one interesting point about Oceanic colonization. Polynesians and eastern Indonesians carry mtDNA that is mainly of Asian female derivation, although the occurrence is not quite 100%.58 On the other hand, the NRY distribution amongst Polynesians and eastern Indonesians reveals a greater presence of indigenous male Melanesian haplogroups, albeit still with strong evidence for NRY and autosomal links between Polynesians and some Formosan groups such as the Amis.59 These results have suggested to some geneticists that Austronesian migration involved large numbers of Austronesian-speaking women of Asian genetic origin, who founded local lineages and thereafter stayed matrilo-cally at home, sometimes attracting husbands from outside communities. Some of these husbands in the western Pacific were perhaps drawn from indigenous Papuanspeaking communities.
The Austronesian situation is unusual if compared to the Eurasian and African patterns discussed in Chapters 7 and 9. The Neolithic migrations of farmers in Eurasia and in Bantu-speaking Africa appear to have been associated more with Y-chromosome (male) than mtDNA (female) transference by the initial farming groups. MtDNA hap-logroup distributions in these regions incorporate high levels of female admixture from populations of indigenous hunter-gatherer descent, and ancient DNA in the case of Europe tells us that much of this admixture took place long after the Neolithic migrations actually took place. This rings a small warning bell for me as far as the Pacific region is concerned, because the idea that many of the Austronesian migrant males were of indigenous local western Pacific descent, but most of the women were Asian, makes little sense if applied to the migration situation itself. After all, it is
Difficult to imagine why women would wish to migrate into new territories, and men not, in the context of small-scale and sea-borne Neolithic societies.
It seems far more likely to me that the differing male and female patterns apparent in both haploid and autosomal markers amongst modern Austronesians relate to population movements and intermarriage during the whole of prehistory over a period of 3500 years, and may not have much significance for understanding the actual foundation migrations themselves. It is very likely for instance, that Melanesian populations, especially males, spread genetically through Melanesia, to as far as Fiji, long after Lapita times, simply through continuous processes of small-scale but gradually compounding movement. Similar movement might well have occurred from New Guinea westwards into eastern Indonesia. A good example is given by Lansing et al. (2011) of how women in certain Austronesian-speaking communities in Timor still marry Papuan-speaking men at a rate greater than the converse.60
World History
