Warren Dew wrote:It certainly contradicts the claim that all of the nonafrican diversity was a subset of African diversity, which was the original claim. Something that's only mostly a subset is not a subset.
But it's still mostly a subset and so even with interbreeding we still have a kind of population remplacement.
Warren Dew wrote:II think it's likely a little higher than that. Hammer's analysis of the modern genome, done before we had any ancient DNA test results, suggested slightly less than 5% ancient European ancestry and about 20% ancient Asian ancestry. He was right on the European ancestry, so I think he's probably right on the Asian ancestry as well - for a total of 25% non-recent-african ancestry for nonafricans, plus or minus 10 or 15 percent. Still a minority, of course.
Well Hammer's theory is certainly interesting but I think we should be cautious about the numbers considering (as you say here) the unequal distribution of "archaic diversity" across Eurasia.
Furthermore we can't discard the possible impact of ancien african population structure as mentioned previously by Moridin.
As Dienekes noted both ancien african population structure and archaic admixture probably played a role in what we are currently observing. So perhaps we really have 20% of "archaic ancestry" in some asian populations, but I think only in some relatively small "portions" of the various asian populations, I think for most of them we still have about 10% of archaïc ancestry.
Naturally I don't claim to be right that's just my bet!