The main points of the linguistic and archaeological records as they relate to early Austronesian dispersal, and possible reasons for it, can now be summarized. Austronesian-speaking agricultural colonists underwent a fairly continuous expansion (albeit divided into periods of relative stasis punctuated by rapid movement), over a period of about 4000 years, from the agricultural heartland region of southern China through many thousands of kilometres of coastline and across increasingly wide sea gaps eastwards into the Pacific. This expansion, which seems to have ignored island interiors in its early stages, met with stiff cultural resistance only in regions with prior histories of agriculture, these being restricted to mainland Southeast Asia and western Melanesia, the latter area being one where archaeology has indicated the existence of a prior and independent development of agriculture (Golson 1985; Golson and Gardner 1990).
The rate at which the early Austronesian colonization occurred must surely be one of the most rapid on record from the prehistoric agricultural world, although admittedly much of it was across sea rather than into large and absorbent land masses. It was probably not caused simply by an over-reliance on land-hungry shifting cultivation, an explanation which I have favoured in the past (Bellwood 1980b), but by a number of different stimuli. These include, not necessarily in order of significance:
- continuous population growth based on an agricultural food supply, allowing a continuous generation-by-generation “budding-off” of new families into new terrain (cf. Ammerman and Cavalli-Sforza 1984 for a European model);
- the inherent transportability and reproducibility of the agricultural economy to support colonizing propagules, especially on resource-poor small islands;
- the presence of a deep and absorbent “frontier zone” available for colonization adjacent to the area of early Austronesian agricultural development, occupied purely by foraging populations (i.e. Taiwan and the Philippines in the early days of expansion), most of whom would presumably have shown little interest in adopting a systematic agricultural economy for themselves (Bellwood 1990b, 1991);
- a developing tradition of sailing-canoe construction and navigation (see Adrian Horridge in this volume);
- a predilection for rapid coastal movement and exploration, probably to find the most favourable environments for cultivation and sheltered inshore fishing, and thus promoting a colonization pattern of wide-ranging settlement followed, often only centuries later, by territorial infilling;
- a culturally-sanctioned desire to found new settlements in order to become a revered or even deified founder ancestor in the genealogies of future generations (presumably this evolved hand-in-hand with the colonization process itself — see Bellwood, in press a);
- a desire to find new sources of raw materials for “prestige goods” exchange networks (Friedman 1981; Hayden 1983; Kirch 1988).
Not all of these stimuli were present before the process of Austronesian expansion began; those listed at (4) and (6) in particular surely evolved in part as a result of the process itself, as might (7) if it was of major significance as an agency of colonization (which I doubt; Bellwood in press a). However, it is my suspicion that the tap-root of the expansion process, a sine qua non, was the possession of a systematic agricultural economy capable of supporting continuous population growth.
The sceptic here may ask why, if agriculturally-induced population growth was so important, all early agricultural peoples did not simply expand in this way. I would reply that perhaps the majority of them did, and this becomes highly significant if one takes the view that early agriculture was an uncommon development in a primary form (i.e. as a result purely of local evolution from an indigenous foraging cultural base), restricted to only a very few specific environmental and floral/faunal regions of central Africa, southwestern Asia, China, New Guinea, Mesoamerica, the northern Andes and the Mississippi basin. As I have already noted, it is perhaps these early expansions, flourishing in a lightly populated, healthy and resource-rich world,[3] which laid the bases for the distributions of many of the major language families of the Old World today.