The Indo-European Völkerwanderung

Update: Nick Patterson writes: Much ancient DNA genotype data is “pseudo-diploid” with just one allele given for a sample at each SNP. If you want to compute F_st values for such data, the easiest way is to run smartpca (see http://genetics.med.harvard.edu/reich/Reich_Lab/Software.html) with inbreed: YES set in the parameter file.

The downside is that this option requires at least 2 samples for each population while the default option (inbreed: NO) works on a single sample.

The above table is a selection of F_ST values I culled from the preprint Massive migration from the steppe is a source for Indo-European languages in Europe. To get an intuition, the F_ST value comparing Northern Europeans and Nigerians is ~0.15 using SNP data. Before I get to the meat of the argument above, let’s take that in for a moment. You see above that the pairwise values between WHG, “Western Hunter-Gatherers,” and the LBK, the first farmer culture of Central Europe, is on the order of ~0.10. That’s about the value between Europeans and East Asians today. We also know that it is possible there was a difference in color between the hunter-gatherers and the first farmers. The meeting of farmer and hunter-gatherer in early Neolithic Europe, and down to the Bronze Age, may best be thought of as analogous to a long term racial conflict and coexistence. Rather than a gradual wave of advance I envisage that the farmers hopped from point to point along fertile stretches of maritime littoral, and pushed their way up into the heart of Europe’s ancient forests by felling the wilderness around the great rivers which issue forth from the uplands. In a world of “isolation by distance” and “clines” this sort of recourse to a term like “race” would be anachronistic, but the model of genetic disruption being reported in these results using ancient DNA strong suggests punctuated demographic transitions across a wide range of localities which would result in biologically and culturally distinct groups persisting for many generations cheek by jowl.  Over time admixture resulted in amalgamation, but it was almost certainly a millennia long process.

More specifically, the authors report unequivocally that the arrival of cultures like the Corded Ware in Northern Europe 4,500 to 5,000 years ago was accompanied by massive demographic replacement. Not only were these bands of warriors traversing the landscape, but it was a whole people on the move, men, women, and children. These were akin to the Goths fording the Danube and bursting into a new landscape of conquest. But the lands of the first farmers were not like those of Rome, heavily settled, further human cattle for the steppe agro-pastoralists to extract rents from. The preprint is not clear as to the timescale of the arrival of the eastern genetic influence across Southern Europe, but in the North the conclusion is without nuance or qualification: during the early years of the Egyptian Old Kingdom the lands of the north were being roiled by migration.

But in this post I want to turn the focus away from Europe for a bit. In the text they note that “An interesting pattern occurs at K=8, with all the late LN/BA groups from central Europe and the Yamnaya having some of the “light green” component that is lacking in earlier European farmers and hunter-gatherers; this component is found at high frequencies in South Asian populations….” I’ve edited and uploaded a version of the admixture plot. One must be cautious when interpreting these plots, but with all the other information in the paper it is quite informative. As far back as Noah Rosenberg’s 2005 paper, and later on in the blogger Dienekes Pontikos’ analyses, there were suggestions of affinities between a subset of Europeans and South Asians, as well as a connection to the Caucasus. In the F_ST results above I note that the Sindhi population is closer to the Lezgins (Northeast Caucasians) and Yamna samples than they are to Armenians. The details are difficult to parse though. Otherwise they would have done so in this monumental paper.

I do want to add one final thing though. It’s been assumed in the past, including by me, that once farmers were established in a locale that future demographic perturbations were unlikely. By this, I mean that farmers are more useful alive to generate economic surplus for incoming elites than they are eliminated. But that is predicated on the idea of a complex specialized society where the elites view all non-elites in an almost Marxian sense of being objects of exploitation. It could very well be that this sort of cosmopolitan globalism only became common in complex societies later, and that in pre-state tribal groups even dense populations did not prevent extermination, because expropriation of vital resources necessary for survival was far more viable an option for these societies than exploitation.