Update: The preprint is out. End update
… and also after that, when the sons of God came in unto the daughters of men, and they bare children to them, the same became mighty men which were of old, men of renown.
An emanation from the one most high…uh, I mean, David Reich, has given his talk at Oxford. Thanks to Jean Manco we have a pretty good report of what he said. The core element seems to be that a paper will soon be published using ancient DNA results to conclude that Indo-European languages came to Europe from the Yamna culture of the Pontic Steppe ~4,000 years ago. Roughly, the argument laid out by David Anthony in The Horse, the Wheel, and Language: How Bronze-Age Riders from the Eurasian Steppes Shaped the Modern World. Where Reich’s conclusions differ from those that Anthony presented in the book is that this eruption from the Eurasian heartland resulted in a genetic-demographic transformation of Europe ~4,000 years ago. Manco confirms that the genetic data from the ancient sites that Reich’s team has access too indicate that the two dominant Y chromosomal haplogroups in Europe, R1b, and R1a, arrived with the Yamna people. This is not surprising, as previous ancient DNA exhibited a surprising death of these two lineages from both hunter-gatherer and early to late Neolithic populations in Europe. And, recent whole genome sequencing of Y chromosomes indicates that both R1a and R1b lineages across Eurasia have undergone recent explosive demographic expansion on a Holocene timescale (closer to 5 than 10 thousand years). In terms of total genome ancestry it looks as if the transition to Yamna successor cultures in Central Europe (e.g., Corded Ware) was accompanied by substantial genetic turnover. In the initial Corded Ware burial grounds 60-80 percent of the ancestry seems to have derived from the Yamna. The modern Yamna-derived proportion seems to be closer to ~50 percent in a region like Germany.
The Yamna themselves are a compound population, a mix of ancient hunter-gatherer groups (analogous to the modern Karelians), and an intrusive population with Near Eastern affinities, likely from the Caucasus. I am not clear whether the Near Eastern group had “Ancestral North Eurasian” ancestry, but the hunter-gatherers almost certainly did. The 1 to 18 percent “Ancestral North Eurasian” ancestry across continental Western and Central Europe today dates to the arrival of these Indo-European speakers. The very low fractions in Southern Europe, and its near absence in Sardinia, may suggest that the Indo-Europeans were demographically more significant in Northern Europe, even though they were clearly culturally effectual along the northern rim of the Mediterranean, and into Anatolia. In line with the Mal’ta results Manco relays that Reich believes that the R lineages, which are the dominant ones across Indo-European speaking populations from the North Sea down to South Asia, came into Europe via the Yamna, but ultimately derive from an “Ancestral North Eurasian” group.
There are some phenotypic tidbits in the talk apparently. The Yamna were tall in terms of their genomic potential. Additionally, the very high frequency of lactase persistence may date to their arrival in Europe (there is some lack of clarity here). I doubt the high lactase persistence frequency and genotypes which result in greater final height are together by coincidence. Large people need a larger nutrient pipe, and adult digestion of lactose sugar would enable that.
There are two aspects which are not widely address in this talk. First, what was the exact dynamic of how the Indo-Europeans replaced the original populations? The idea of “demic diffusion” by waves of “demographic advance” promoted by Colin Renfrew seem to gradual and continuous to be responsible for this. This is basically an argument predicated on individual fitness, summed over groups. In this case I suspect that a better analogy may be the future that Genghis Khan had in mind for Northern China before his adviser Yelü Chucai dissuaded him: the North European plain was cleared out of people and turned over to pastureland. Genghis Khan and his Mongols were convinced of the value of Chinese as tax paying peasants, who could support the Mongol elite with their surplus. I suspect in a pre-state society such considerations were less relevant, as the institutional frameworks which would allow for the smooth absorption of subordinate groups were less elaborated, or even non-existent.
In both the Late Neolithic and after the Bronze Age the Reich group alludes to a return of the primal populations which were marginalized by the farmers, and later the Indo-European agro-pastoralists. One way to look at this is that there were larger migrations which were overlain upon the palimpsest. But, I believe one might also consider a model whereby there is ascertainment bias in the sorts of burial sites being explored and sampled, and one might be witnessing a patchy occupation of the landscape by intrusive cultures. For example, the newcomers might monopolize the rich bottom-lands for thousands of years, but huge swaths of the hinterland might be occupied by marginalized and less developed people, who over time drift into the core and become culturally absorbed. Instead of imagining the expansion of these people as purely ones of a vast uniform wave front, it might be better to conceptualize them as penetrating into virgin territory along the optimal avenues of settlement, and producing a patchy archipelago of habitation.
Second, there is the issue that though Reich and company focus on Indo-Europeans and the Yamna culture, the genetics leaves may loose threads that are difficult to tie back up. At ASHG Mait Metspalu express to me some misgivings about the term “Ancestral North Eurasian.” How do we truly know the locus and distribution of this ancestral component across Eurasia ~10,000 years ago? The Kalash of Pakistan exhibit signals of admixture with this group as high as Northern Europeans, so it is not limited to West Eurasia proper. The highest fractions today seem to be found in the North Caucasus, among many groups which are not Indo-European. If R1a was brought by Indo-Europeans to Europe, it is harder to conclude that this was the case in South Asia. Though the frequency of these lineages is higher in the Indo-Aryan North, there are relatively high fractions of R1a even among some South Indian tribal populations. R1b is found in appreciable fractions in Sardinia and among the Basques (one argument for the old idea that R1b was the legacy of European hunter-gatherers!). Obviously some of this could be due to admixture between Indo-Europeans and non-Indo-Europeans. But I think a major issue here is that Indo-European groups were a synthetic population which arose in a world where there were many synthetic populations, with ancient and recent affinities to them. I doubt the “Ancestral North Eurasian” ancestral component was limited purely to Indo-Europeans. So it seems unlikely that the R1 lineages would be purely Indo-European, even if recent expansion of some of their sub-lineages is a function of the Indo-European cultural explosion.
Of course there are only so many ancient DNA samples one can retrieve from a finite number of sites. The age of new genomic discoveries will start to close over the next few years as the paleo-demography inferred will start to exhibit some predictable solidity. That means that a deep knowledge of the archaeology, and what history there is, is essential.