There’s a fascinating new article in PNAS, ‘Stone tools and foraging in northern Madagascar challenge Holocene extinction models‘, outlining a long chronology for the settlement and early habitation of Madagascar. The traditional wisdom is that Madagascar was uninhabited until around 500 CE when Austronesian speakers from southern Borneo migrated several thousand kilometres westward, and Bantu-speaking East Africans crossed the Mozambique Channel, producing a civilization of iron-using swidden farmers and creating an ecological catastrophe in which many native species went extinct. The discovery that the Malagasy language is most closely related to the Southeast Barito languages of Borneo, proposed systematically for the first time by Otto Dahl in the 1950s, is one of the most significant and surprising findings in historical linguistics of the past century, given the enormous geographic distance between the two regions. Later, Dahl helped to establish that Malagasy also has an important Bantu linguistic substratum, and more recent genetic evidence confirms that both African and Southeast Asian migration was involved.
This new study, whose first author, Robert Dewar, unfortunately passed away before its publication, shows the situation is significantly more complex, and that there is a history of hunter-forager habitation in at least some parts of Madagascar going back up to 4,000 years (i.e. 2,500 years more than previously acknowledged by the traditional hypothesis). I’ve always wondered how it was that Madagascar, which is not that far from the East African coast, could remain entirely uninhabited by humans for so long. The new study, based on fieldwork conducted a few years ago at two rock shelters in the northern part of the country, shows a vibrant hunting-foraging adaptation with microlithic tool technology to have existed far earlier than previously suspected. This tool tradition has similarities with both East African and Middle Eastern traditions of the same period, but not with Southeast Asian ones (unsurprisingly). What this tells us is that there was a previously-unidentified pre-Bantu, pre-Austronesian population on the island, probably of East African ancestry for millennia before the extinctions of Madagascar’s megafauna began in earnest, It requires that we rethink the model that sees the arrival of humans on Madagascar as the simple direct cause of the extinctions, and forces us to instead ask what sorts of human-environment interactions cause effects, and how.
Glossographia 
The fact that previous Occidental scholarship preferred to imagine that people had to travel all the way from South Borneo to populate Madagascar, and that people from Africa could not have done so previously, fits very nicely with typical racist scholarship. That this anomaly prompted you to wonder is a credit to you, of course.
Simply per Occam’s razor, it makes much more sense that Africans had already settled on Madagascar and the South Borneo travelers found them there upon their arrival–if (a more radical or tentative hypothesis) the Africans did not have some hand in contacting South Borneo in the first place and establishing a reason for the migration at all. I don’t know, in that context, if we really need to speak of the original settlers from Africa necessarily as pre-Bantu. If an original African settlement on Madagascar led to cultural separation, then we may now simply have two “varieties” of Bantu: the Madagascar variety, the mainland variety, and the (not actually really) proto-variety (so to speak) from which both now derive. I apologise if this amounts speaking in gross generalities, since I don’t have any specialized knowledge about this topic specifically. I’m interested to know though of course.
Also, we know (as, for example, from the history of the Roman empire) that destructive expansion often accompanies empire, colonization, and the like. While it is difficult to imagine such a massive contingent of South Borneo emigres arriving to physically overthrow whatever civilization prevailed on Madagasccar at the time, we might also look at the European invasion of Mesoamerica, where a relatively tiny portion of new arrivals caused continental change (thanks to gobs of factors). So we needn’t rule out of hand as simply absurd that the arrival of South Borneo interlopers into an already established culture did not, in fact, contribute to the sort of extinction described above.
These points are simply to say that history gives evidence of wide-scale social change that happens (in the right social conditions) through the arrival of small groups of people, and that these sorts of wide-scale changes could, after all, have a hand in the extinctions that occurred. The arrival of humans might exactly provide a direct cause, but we would more exactly identify which humans or–more precisely still–humans arriving with what purpose? If interlopers arrive somewhere with intentions like the Europeans in Mesoamerica, things histroically tend not to turn out well.
I’m reasonably confident that the early African populations in Madagascar were non-Bantu – really the historical linguistic evidence along with the archaeology suggest that Bantu-speaking peoples displaced, intermarried with, conquered, etc., the pre-Bantu populations of much of eastern Africa around the same time that they were settling Madagascar. I will also say, in fairness, that Madagascar is an *extremely* challenging place to find evidence for hunter-foragers in the archaeological record, given the speed at which material decays. Archaeologists would, however, have found evidence of large fortifications, towns, etc., that would have characterized the large-scale Bantu societies of eastern and central Africa, if there were such evidence to be found. So in part what’s going on here is certainly a colonialist and potentially racist mentality, but not simply that ‘Africans couldn’t have settled Madagascar’, but rather ‘small-scale foragers couldn’t have settled Madagascar’. It will be very interesting to see where this goes in the next ten years – often after an initial find like this, there is a wave of research – that may show continuous habitation from 4000 years ago to the present, or it may show more sporadic and strategic use of Madagascar’s resources by African-based foragers, or it may show something else entirely unexpected.