Reply #18: Another article on this that I read made a crucial point. [View All]

That they can only know what current gene distributions tell us. We can't know what distributions were 100k years BCE, esp. if the population was fairly thin.

They added something important, however. The San have a fairly complex genetic distribution--it's been assumed for a few years that they're likely to be more similar, in many ways, to early humans, and that other "tribes" branch not from the Africans living in E Africa but from their stock or kindred stock.

It's also fairly certain that the San were eliminated from a fair amount of Africa by the Bantu expansion, which happened fairly late. The assumption was that the people before the Bantu were San, and the linguistic evidence (and maybe some archeological evidence) points this way. The assumption is that the San were as close to indigenous to those areas as any group could be--first settlers, followed by 10s of millennia of isolation in a fair degree of isolation--which would in itself produce a fair amount of genetic diversity.

This study points out that there's a fair genetic/linguistic correlation (most members speaking language X will be identifiably people of ethnicity X). They also point out that most of the language/ethnic groups show considerable mixing ... except one. The Bantu (which is just a language group, and like the linguists' use of "Aryan" not a single ethnicity). However, it's precisely the Bantu who overlay areas that were formerly occupied by the San. So do Bantus actually show significant mixing, but we can't spot it, or did they swamp the local population so that the substratum genetics are hard to spot, or did they simply dispose of the San?

For an example of how the location could be a spurious conclusion, what would happen if more diffuse groups, showing less concentrated genetic diversity, were driven into a much smaller area by another group's expansion? You'd get greater genetic diversity in a small area, and satisfy the criterion for that being the "homeland". As another example, consider what would happen if the Bantu were to (as is happening, for the most part) completely wipe out the San. Suddenly there'd be scant trace of this diversity and suddenly the method used would conclusively point to another place. Since much of the San area *has* been overwritten, we have to simply acknowledge this as a flaw and assume the uncertainty can't be resolved.