The Pattern of Austronesian Expansion

It is necessary to return again to the linguistic evidence in order to plot the geographical axes of expansion of early An-speaking peoples. Beginning with the Pre-Austronesian level, a number of claims have been made for ancient genetic relationships between An and other Asian mainland language families (see footnote 1). Perhaps the best-known of these claims is the Austro-Tai hypothesis of Paul Benedict (1975; Reid 1984-5), which postulates that the Tai-Kadai and Austronesian language families once shared a common ancestral language or chain of languages spoken on the southern Chinese mainland. Benedict has suggested a number of important vocabulary reconstructions for this ancestral language, including terms for field, wet field (for rice or taro), garden, rice, sugarcane, cattle, water buffalo, axe, and canoe. The overlap between this list and that presented above for Hemudu and other coastal southern Chinese Neolithic sites needs little emphasis. One has to consider very seriously the possibility that the initial expansions of Austronesian and Tai-Kadai languages (and probably also Austroasiatic) began among Neolithic rice-cultivating communities in China south of the Yangzi. The archaeological record agrees very well and provides a date range for initial developments between 5000 and 4000 BC.

Moving beyond Austro- Tai into Austronesian proper, the reconstruction of linguistic prehistory which is most widely used today is that postulated by Robert Blust (1984-5). This is based on a “family tree” of subgroups and a hierarchy of proto-languages extending from Proto-Austronesian (PAn) forwards in time. Reduced to its essentials, Blust’s reconstruction favours a geographical expansion beginning in Taiwan (the location of the oldest Austronesian languages, including PAn), then encompassing the Philippines, Borneo and Sulawesi, and finally bifurcating, one branch moving west to Java, Sumatra and Malaya, the other moving east into Oceania (see Darrell Tryon’s more detailed summary in this volume).

A wealth of linguistic detail can be added to this framework, but here I will restrict myself to some implications of broad historical and cultural significance. During the linguistic stage before the break-up of PAn it would appear that some colonists with an agricultural economy moved across the Formosa Strait from the Chinese mainland to Taiwan (Bellwood 1984-5, 1992). Here developed the Initial Austronesian language(s), and after a few centuries some speakers of one of these languages made the first moves into Luzon and the Philippines. This movement led to the division of Austronesian into its two major subgroups, Formosan and Malayo-Polynesian (or Extra-Formosan). The reconstructed PAn vocabulary, which relates generally to this early Taiwan-Luzon phase, indicates an economy well suited to marginal tropical latitudes with cultivation of rice, millet, sugarcane, domestication of dogs and pigs, and the use of some kind of watercraft.

As a result of further colonizing movements through the Philippines into Borneo, Sulawesi and the Moluccas, the Malayo-Polynesian (MP) subgroup eventually separated into its several lower-order Western and Central-Eastern branches. The break-up of Central-Eastern MP probably occurred initially in the Moluccas, and Eastern MP contains all the Austronesian languages of the Pacific Islands apart from some in western Micronesia. The vocabulary of Proto-Malayo-Polynesian (PMP), a linguistic entity which might have been located somewhere in the Philippines, is of great interest because it contains a number of tropical economic indicators which were absent in the earlier and more northerly PAn stage. These include, according to Blust (1984-5), Colocasia taro, breadfruit, banana, yam, sago and coconut. Their presences may reflect a shift away from rice, a plant initially adapted to sub-tropical latitudes, towards a greater dependence on tubers and fruits in equatorial latitudes (Bellwood 1980a, 1985). The PMP vocabulary also has terms for pottery, sailing canoes and several components of substantial timber houses (Zorc 1994).

It may now be asked how the archaeological record relates to this reconstruction of the directions and cultural components of Austronesian expansion. Specific archaeological cultures cannot logically be equated with specific ancestral languages in prehistoric time. However, the appearance of certain technological and economic components of PAn and PMP can be searched for profitably in the archaeological record of the area now occupied by Austronesian speakers. As already indicated, it is a reasonable inference that both PAn and PMP represent agricultural societies who, amongst other things, grew rice, made pots, lived in well-made timber houses and kept domesticated pigs. As it happens, direct material evidence for all these items survives in the archaeological records of the islands of Southeast Asia, and all of them make an initial appearance in widespread excavated sites between about 4000 and 1500 BC. Furthermore, their appearances (especially pottery) show a time trend — earliest in the northerly regions of China, Taiwan and Luzon, and progressively later as one moves southwards into equatorial Indonesia and western Oceania (Spriggs 1989). Given this seeming correlation between the linguistic and archaeological records (Bellwood 1985), we may hypothesize a direct association with the dispersal of the Austronesian language speakers, rather than dispersal of these cultural items by diffusion alone.

The Neolithic archaeological records in Taiwan began around 3000-4000 BC with archaeological assemblages of southern Chinese type, presumably carried initially by small groups of agricultural settlers across the Formosa Strait from Fujian (Tsang 1992). Characteristic artefacts of the oldest sites include cord-marked pottery, polished stone adzes and slate spear points. Other items such as slate-reaping knives and baked clay spindle whorls (for spinning thread for weaving) perhaps arrived a little later. By 3000 BC in Taiwan there is evidence for rice and, from pollen records, for inland forest clearance for agriculture.

Between 2500 and 1500 BC related archaeological assemblages with plain or red-slipped pottery, rather than the older Taiwan cord-marked type, appeared in coastal and favourable inland regions of the Philippines, Sulawesi, northern Borneo, Halmahera, and (with domestic pigs) to as far southeast as Timor. No sites of this period have yet been reported from the large islands of western Indonesia, but research on pollen history in the highlands of western Java and Sumatra suggests that some fairly intensive forest clearance for agriculture was underway there by at least 2000 BC, and probably earlier (Flenley 1988). In the equatorial latitudes of Indonesia there may also have been a shift away from rice cultivation towards a much greater dependence on the tropical fruit and tuber crops listed above for the PMP vocabulary. No cereals were ever introduced into the Pacific Islands, with the exception of rice to the Marianas.

By 1500 BC, therefore, agricultural colonists had spread from Taiwan to the western borders of Melanesia. The continuing expansion through Melanesia into western Polynesia, represented by the Lapita culture, seems to have been even more rapid than preceding movements, perhaps because food producing (as opposed to purely foraging) Papuan-speaking populations were already occupying some coastal regions of the large islands of New Guinea, the Bismarcks and the Solomons. Finely decorated Lapita pottery has been found in coastal or offshore island sites from the Admiralties in the west to Samoa in the east, a distance of about 5000 kilometres (see following chapter). This Lapita expansion occurred between 1600 and 1000 BC and to north and east of the Solomons it involved, for the first sustained period in Austronesian prehistory, the settlement of previously uninhabited islands. Between 1000 BC and AD 1000 the settlement of these uninhabited regions continued onwards (Irwin 1992), ultimately to incorporate all the islands of Polynesia and Micronesia and, on the other side of the world, Madagascar (Map 1).