The triffids would have eaten everybody if it wasn’t for those neanderthal alleles that made us a bit smarter. Oh wait a minute, that can’t be, we are all equal.

Or with a lack of it.

You’re wrong. This is exactly what you’d expect to see for neutral Neanderthal alleles. The overall amount of admixture does not change – law of large numbers – but the frequency of individual alleles drifts. Since the base frequency is only about 2.5%, most Neanderthal alleles drift to zero frequency, while a few drift to high frequency. If two populations have been separated for a long time, the high-frequency Neanderthal alleles in the two populations will mostly be different ones.

This requires that the effective population sizes be small enough that drift matters, but that seems to have been the case for much of the period we’re talking about, considering the low population density of hunter-gatherers, nasty glacial climates, intertribal warfare, Triffids, etc. This drift effectively stops when populations become large, as they are today. Assuming that they stay large.

The story would be quite different for Neanderthal alleles with a noticeable selective advantage.

While you could have a common set of Neaderthal genes from a single admixed population that then splits, the more time elapsed between the admixture events and the ultimate population split, and the larger the effective population sizes of the diverging populations at the time of divergence, the less likely the kind of distribution Hawks describes seems. If a single admixed population were a single population genetic population for several thousands years before dividing, the Neanderthal components would have reached very close to uniform distributions in the entire population – you’d need admixture to be something less than several hundred years old and quite small diverging effective population (closer to several hundred than ten thousand), or strongly endogamous tribal structures prior to divergence, to get the kinds of genetic drift effects Hawks is suggesting.

There are lots of mechanisms that could keep the percentage of admixture similar without the source population being a common one. After all, the social structure of the modern human population and the Neanderthal populations would be very similar to each other in a two admixing population scenario, so the level of and nature of their interactions would have been similar, thus producing similar admixture percentages from independent events.

In the same vein, assimilation patterns of immigrant Hispanics in the United States today are quite similar to the assimilation patterns of Irish and Italian immigrants a century and a half earlier.

Also striking is the near absence of Neanderthal genes that are common in both Europe and China. This argues against adaptative value being important. Adaptive genes for modern humans in general would likely be high frequency in both populations. Whether there was one admixing population, or more than one, I would agree that Hawks is on target in suggesting genetic drift from the X chomosome data set as opposed to something else.

Panadaptationist. ;