Asia and Africa
Virtually no ancient DNA evidence relevant to the Neolithic transition is available for any other part of the world. Several areas have, however, been subjected to intensive study from the perspective of modern human variation, and two in particular include certain checks and balances in the procedure that were lacking in the case of Europe.
Both Southeast Asia/Pacific and sub-Saharan Africa are dominated by a single wide-spread language family - Austronesian and Bantu respectively - and both regions underwent recent and well-documented colonization processes at one extreme of the range. The language distributions and archaeological/historical evidence can therefore exert greater discipline upon the genetic models, and provide an independent corroboration for the genetic dating. At the same time, the linguistic evidence in particular has often led to quite simplistic models of demographic expansion across the entire range of the language families,[59] whereas the genetic evidence has suggested a much more complex pattern preceding the final phase of expansion, in the Remote Pacific and southern Africa respectively.In Southeast Asia, the main focus has been on island Southeast Asia (ISEA) and the Pacific islands, where Austronesian languages have been spoken, but the model of agricultural expansion (fuelled by the development of rice agriculture) ultimately also takes in central and southern China and mainland Southeast Asia. However, by and large, genetic studies have worked from the periphery - beginning with the Remote Pacific and working backwards towards the mainland - because of the relative simplicity of the situation in Oceania. Early studies confirmed the supposition, based on linguistics and archaeology, that genetic diversity was extremely low in the Remote Pacific, as the result of major founder effects in the recent past - in particular generating the so-called mitochondrial ‘Polynesian motif, a single lineage spread throughout the Remote Pacific at very high frequencies.
The earliest studies saw evidence in this for the consensus farming/language dispersal hypothesis: an expansion from China/Taiwan through Southeast Asia beginning c. 6 ka and reaching the Remote Pacific by c. 3 ka, largely bypassing New Guinea, which had evolved separately from Southeast Asia for tens of thousands of years. Much genetic work continues to lay the main emphasis on the Pacific, and to interpret the genetic data in these terms.However, even from the earliest days it was clear that ISEA was much more genetically diverse than Micronesia and Polynesia. Further scrutiny of the mtDNA evidence, and the emergence of MSY evidence as well, indicated that the so-called ‘express-train' model of rapid dispersal all the way from China/Taiwan was not holding up. Both systems pointed to a major role for eastern Indonesia/Near Oceania (the New Guinea region) as a source for the Remote Pacific islanders. The major proximal source for the mtDNA ‘Polynesian motif appears to be the Bismarck archipelago in the early to mid-Holocene, with a deeper ancestry further back in island Southeast Asia,[60] but also with minor arrivals from ISEA in the late Holocene, preceding the expansions into Remote Oceania. This intertwining of ISEA and the New Guinea region throughout the Holocene renders simplistic admixture accounts that apportion different levels of ‘Asian' and ‘Melanesian' ancestry to Pacific islanders (often equating the former with ‘Austronesian') somewhat outmoded, and suggests that archaeological models that emphasize ancient voyaging may have a significant role to play.[61]
Similarly, it seemed clear early on that much of the MSY variation, too, traces to eastern Indonesia/Near Oceania. A minority may derive from Taiwan or Borneo, potentially tracing a male-mediated ‘farming/language dispersal' from the north and west, but the details are unclear and similar caveats to those we mentioned with regard to the European studies are appropriate here too.[62] Both mtDNA and some autosomal analyses have suggested, for example, that Taiwan, rather than being a major source for ISEA diversity as suggested by the language distributions, may rather have been largely a recipient for lineages from further south.[63] There are some candidates for minor mtDNA lineages, as well as MSY lineages, that may have transmitted Austronesian languages and/or Neolithic industries through ISEA, but the great majority - including those ancestral to the Polynesian motif - have a deep insular ancestry and appear to have been largely shaped by the sealevel rises accompanying the end of the last Glacial period.[64] The spread of the
Neolithic and of the Austronesian languages appears to have had a minor impact on population structure and may owe more to the establishment of social networks and spheres of interaction.[65] This pattern of predominantly late Glacial and post-Glacial but pre-Neolithic growth is also seen in Chinese mitogenomes,[66] although a full phylogeographic picture has yet to be worked out.
The pattern in mainland Southeast Asia - where the main language family is Austroasiatic - is rather different from ISEA.
There are clear links to southern China, but whether the expansions into Indo-China accompanied the spread of rice farming or the earlier spread of ‘coastal Neolithic' huntergatherer populations driven by climatic changes remains debatable. It has been suggested that the horticulturalist Senoi of the Malay peninsula are the product of both farmer-dispersals from Indo-China and assimilation of the indigenous people of the peninsula, and may have been the conduit that brought the Austroasiatic languages south. MSY and autosomal patterns suggest that the Austroasiatic languages were spread from mainland Southeast Asia west into India, rather than in the opposite direction as has been suggested. The Indian subcontinent in fact received lineages from both Southeast Asia (mainly on the male side) and Southwest Asia (from both lines of descent), which might possibly signal the diffusion of rice and wheat/barley into the east and west respectively - although work to date suggests that those from the west, at least, may predate the Neolithic.[67]In Africa, the spread of agriculture through much of the central and southern part of the continent went hand in hand with the dispersal of the Bantu languages from west Central Africa, within the last few thousand years. Rather like in Remote Oceania, the situation in southern Africa is relatively straightforward and reasonably well understood in outline. The indigenous Khoe and San populations speak ‘Khoisan' click languages and have traditionally had foraging or herding economies. They have a very characteristic suite of mtDNA and MSY lineages, and highly distinctive autosomal profiles. Bantu-speakers in southern Africa, by contrast, show signs of a maternal founder effect c. 2 ka, and they share many mtDNA and MSY lineages, and an autosomal profile, with Bantu-speakers from across Central Africa.[68]
Again, however, southern Africa lies on the far edge of the Bantu expansion, and the situation further north is much more complex.
Bantu-speakers spread both south and east from their core area around west Cameroon/ southeast Nigeria, and the eastern stream encountered settled populations with cereal crops in East Africa. Somewhere in this region, or in east Central Africa, they appear to have assimilated new lineages, at least on the maternal line of descent, so that southeast Bantu-speakers have a mixed ancestry including lineages from west Central Africa, East Africa, and southern Africa, with the levels of ‘Khoisan’ ancestry increasing substantially into southern Africa.[69]Southwest Bantu-speakers, on the other hand, have a preponderance of Central African lineages barely seen in southeast Bantu-speakers, presumably assimilated in the forest zone of Central Africa by the western stream of dispersal. Intriguingly, these lineages are far more diverse in Bantu-speakers than in Central African forest-dwelling foragers - a counter-intuitive pattern that might imply large-scale assimilation of lineages into the Bantu-speaking groups and heavy losses through genetic drift among the foragers. Similar complexities are evident on the male line of descent, and indeed in the autosomal patterns of, for example, East Africa.[70]
It seems clear that the Bantu dispersals hugely reshaped the genetic landscape of Africa, but as with much of what we have described in this chapter, working out the processes in detail still lies ahead. We have only outlined here some of what we consider to be the more persuasive arguments among what has become a huge and growing area of research, but even so it seems clear that archaeogenetics has barely scratched the surface of its potential.
A huge amount of genetic data is pouring out at an accelerating rate, but it is often still treated in quite a broad-brush way. This is in part because of the coarse resolution of genetic studies that have mostly relied on present-day data, where the signals from the Neolithic can be obscured by millennia of subsequent migrations, making nuanced regional studies difficult, but it is beginning to look as if ancient DNA analyses might have the potential to overcome this limitation. Perhaps it is time for renewed collaborations between geneticists and archaeologists in order to refocus the analyses away from the grand narratives of the past and bring the subject down to earth.