Razib Khan comments on the current round of Henry Louis Gates ancestry programming: "Finding fake roots", and "Reification is alright by me! Razib notes that the criteria that tell many subjects that their ancestry is a mixture of different populations are conditioned on assumptions that don't work at all for South Asians. From the latter:
In my post below some commenters argued that obviously implausible inferences from a thin set of reference populations are acceptable considering Henry Louis Gates Jr’s target audience. But that really wasn’t my main point. Rather, it was that he was eliding the distinction between uniparental markers, and the clusters generated by modeled based ancestry assignment algorithms, and ascribing the phylogenies of the former to the latter. It is important to note that categories like “Europeans” are only approximations. But they’re damn good approximations today! Nevertheless, note the qualification of time: they may have basically no meaning at some point in the recent past. They’re powerful when it comes to precisely partitioning modern variation, but they don’t tell us the history of that variation.
The uniparental marker "interpretations" given to people doing genealogical work has become increasingly comical in its distance from what we now know about ancient variation. For example, I carry mtDNA haplogroup H, and here's what the Genographic Project tells me about that history in their "Atlas of the Human Journey":
Around 15,000 to 20,000 years ago, colder temperatures and a drier global climate locked much of the world's fresh water at the polar ice caps, making living conditions near impossible for much of the northern hemisphere. Early Europeans retreated to the warmer climates of the Iberian Peninsula, Italy, and the Balkans, where they waited out the cold spell. Their population sizes were drastically reduced, and much of the genetic diversity that had previously existed in Europe was lost. Beginning about 15,000 years ago -- after the ice sheets had begun their retreat -- humans moved north again and recolonized western Europe. By far the most frequent mitochondrial lineage carried by these expanding groups was haplogroup H. Because of the population growth that quickly followed this expansion, this haplogroup now dominates the European female landscape.
Here, a very common mtDNA haplogroup today is given its own origin myth, complete with a glacial refugium and massive expansion and dispersal. The text goes on to explain how this European haplogroup spread right out of southern Europe into central Asia, where today -- surprisingly -- it is even more variable and shows less sign of expansion. Notice how precise the story sounds, a fleshed-out history for people looking to connect their roots to European prehistoric events.
Why do I say comical? We have ancient mtDNA from all over Europe now, from Neolithic and pre-Neolithic people, showing that haplogroup H was barely there before farming.
I don't mean to single out Genographic for this issue, in fact the whole edifice of genealogical interpretation is built on assumptions about history that are currently known to be false. We can do much better than this, I think. But many of the same characters who failed five or six years ago keep plugging at it, persisting in describing a distorted version of human history.
UPDATE (2012-05-08): The thing that really bugs me, is that the amount of money spent producing a season of one of these programs would be more than enough to get some of us to straighten some of these problems out. Population genetics is a lot cheaper than media. Or, to put it in a more inspiring way: any media organization that is willing to spring for a couple of postdocs along with their program can show some real science instead of making stuff up. Just saying...