Southern Africa is kind of a big deal. Not because it is the seat of human origins; I am beginning to think that question is “not even wrong.” Nor because it contains the “oldest human population” in the world; we are all the oldest human population in the world. Rather, the genetic variation one can find across a small region in southern Africa is incredible, and, it is one of the few regions of the world where hunter-gatherers have persisted in a culturally pristine fashion. By this, I mean that there is no evidence that the hunter-gatherers of southern Africa practiced another way of life (i.e., that they are marginalized agriculturalists or nomads), or, that their language was adopted from agriculturalists (as is the case with the Pygmies of central Africa). In other words, the continuity of the peoples of southern Africa is more notable not for their genetics, but their culture. That being said, the cultural conditions under which the KhoeSan peoples existed are of genetic interest, because their high degree of variation may reflect aspects of population dynamics common to hunter-gatherers as a whole.
In light of all this a new preprint on bioarxiv is quite interesting, Fine-scale human population structure in southern Africa reflects ecological boundaries. The title rather says it all, but I’ll admit that it’s hard to keep track of all the populations. It certainly strikes me as plausible, as the genetics suggests that there’s a lot of structure that built up over the years. Of note: “To contrast this with Europeans, the ≠Khomani and the Ju/’hoansi may have diverged over 30,000 ya but live only 1,000 km apart, roughly the equivalent distance between Switzerland and Denmark whose populations have little genetic divergence.” They report trans-Kalahari F st values on the order of 0.05, which is quite high, on an order of magnitude or so greater than what can be found in Northern Europe. But the latter is more comprehensible when you consider the genetic character of the North European plain only arrived at its current state ~4,000 years ago.
Since the above is a preprint, some critiques are in order. The authors use ancestry tract lengths to assess admixture of Bantu, Asian, and European, elements into the KhoeSan. The implication is that these were separate admixture events. Some of them certainly were. But I’m a little skeptical of the power of these methods to distinguish admixture between the last two non-African components (also, I think it is probably advised for a population genetics paper to dispense with the cultural construct of Asian and make the clear distinction between South and East Asians, since the former are often genetically closer to Europeans, and this previously inflated the European proportion among Cape Coloureds). It’s been many years since I read A History of South Africa, but one of the more interesting aspects that I recall from this book was the cultural distinctiveness of what has now come to be called the Cape Coloured people, and their role as mediators along a fluid cultural frontier with the KhoeSan people. The history of the Griqua in particular shed light on how one might imagine European and Asian (South and East) ancestry arrived into the KhoeSan. Though racially mixed, the Griqua resembled the Dutch in formal and institutional aspects of their culture by the 19th century. But, as a semi-nomadic and pastoralist group they also had affinities with their African neighbors (and often, they played the role of predators with the Africans and their herds), with whom they clearly shared ancestry. It was not an unknown phenomenon to have Griqua scouts “go home to their mother’s people.” Even if their literal mother was also a Griqua, they seem to have had a sense that their maternal ancestors were invariably non-European, and often derived from the KhoeSan (many Griqua still spoke the now extinct Cape Khoi language, though they were shifting toward a Dutch Creole). The probability of assimilation of unadmixed Europeans, South Asians, and East Asians, into KhoeSan groups is not zero, but it strikes me as quite low. On the other hand, Griqua, who were mixed between all the populations of southern Africa, and culturally quite at home in the semi-desert wilderness, seem ideal candidates for the population which could serve as the vectors for transmission of these ancestral elements into the KhoeSan.
A second issue with this preprint is that I’d like to see more methods. E.g., three and four-population tests and TreeMix in particular, since with 320,000 SNPs these should be totally feasible with genotype data. These are not groups that most people have much familiar with, so PCA and ADMIXTURE tend to overwhelm. To develop a decent intuition about what’s going on trees and tables are often helpful (I don’t like tables usually, but often when you are focusing on a finite number of populations per row they can allow for greater focus). Also, there are now interesting ways to analyze spatial genetics beyond Mantel tests.
Ultimately these results confirm what I already held as a prior. It strikes me that the relationship we see between language and genes today is largely a function of the reality that much of the population genetic structure we see around us is a recent phenomenon. That is, massive migrations due to cultural changes (e.g., agriculture) were accompanied by both language and genes, and only a few thousand years are simply not enough time to allow for linguistic differentiation to obscure common origins. In contrast, if, as seems plausible, many of the KhoeSan people were resident in southern Africa for tens of thousands of years, then correlations between language and genes should slowly decouple (in part because deep linguistic affinities may not be discernible). That being said, I would not be surprised if ancient DNA from southern Africa at some point overturns conventional wisdom that these peoples are truly primal….