Monday, February 11, 2008

The many dimensions of history   posted by Razib @ 2/11/2008 06:02:00 PM

T'ang China - The Rise of the East in World History:
Interconnections, one may agree, are both old and wide. Initially there was the genetic connection. Although the new alliance between genetics and history is in its infancy, it has already established that, at least since sapiens replaced Neanderthal, habilis and erectus, there have been no true human races. The human genome is unitary and the genetic differences within it are not that large. Cavalli-Sforza states: 'According to that first estimate, the woman from whom all modern human mitochondria descend lived about 190,000 years ago', adding 'this first attempt was not so bad'. Subsequently, Bryan Sykes of the Oxford Institute of Molecular Medicine has concluded that 'almost all native Europeans belong to one of seven distinct clans each with a founding mother ... But there is virtually no consistent pattern to the way the clans are distributed in modern Europe: in the thousands of years of European history they have become thorough mixed. Since Europe was originally populated from Asia, and to a lesser degree from Africa, similar continuities are likely to be traceable there. As Teilhard de Chardin saw, the absence of human branching is striking. The continued convergence rather than divergence is not easily explained by the first three principles of evolution - mutation, natural selection and drift - but only by the fourth, flow, especially the mobility of women, which in turn implies contact between groups....

I've read a great deal of Samuel Adshead's work; though perhaps overly ambitious, like William McNeill he approaches big questions in a with a wide and diverse toolkit. It's thinly populated territory, but one which needs to be explored. Unlike some humanists Adshead has no problem with including the natural sciences as part of his toolkit; the reference to population genetic parameters such as selection, drift, mutation and gene flow imply a level of familiarity far beyond the ken of expectation. That being said, I would obviously object to many of Adshead's characterizations. Biology is it not physics, and like history sentences such as "The human genome is unitary and the genetic differences within it are not that large" need to be placed in their proper context. And within an area such as human genetics one needs to be cautious about citing works and opinion from even the 1990s; one needs to read the journals as opposed to paying attention what L. L. Cavall-Sforza or Bryan Sykes reported as to the state of consensus a decade ago.

A truly synthetic take on the history of our species needs to consider all population genetic parameters. The necessary subtly and depth of understanding is, I believe, properly modeled in the sort of dynamic system which is illustrated in William McNeill's Plagues and Peoples, though even more so in terms of its complexity. The lactase persistence story over the past two years illustrates the power of gene-culture coevolution; but too often it is framed in terms of the power culture has in shaping genetics, but what about the impact that genetic revolutions may have upon history? The Out-of-Africa movement itself may be a more viable template than we could have imagined, after all, a common hypothesis is that evolutionary/genetic changes gave our African predecessors a fitness advantage over other archaic hominid species. The LCT region is the largest portion of the European genome cleaned of genetic variation because of powerful selection; and it all occurred within the last 7,000 years! History as traditionally understood by cothe rise of Sumer is 5,000 years ago, so on the scale of populations the domain of evolutionary genetics does impinge upon that of other humane studies.

Much of Ashdead and McNeill's work hinges upon networks and flow of information. I would hold that one of the great stories soon to be told about human history over the last 10,000 years are the changes in the nature and extent of gene flow due to the rapid increase in population (triggered by the agricultural revolution). A change in density may have had a natural affect on the weighting of various alternative population genetic parameters. Here you have a case where a cultural innovation, agriculture, results in necessary shifts in evolutionary dynamics, which themselves may set in motion historical processes!

But to understand and infer historical processes, I do believe that the insights and knowledge of scholars of promiscuous inclinations such as Samuel Adshead will be essential in putting the pieces in place.

This is a map of the distribution of two alleles of SLC24A5. In case you don't know, this locus has an ancestral and derived variant on one SNP. A few years ago one group noted that Africans & Asians were nearly fixed for the ancestral variation, while Europeans were fixed for the derived on. Through an admixture study on an African American population they determined that the variation on SLC24A5 was responsible for 1/4 to 2/5 of the trait value different between Africans & Europeans. A recent study with South Asians shows that this locus is responsible for about the same proportion of the within population variance among this group. The frequency of the derived variant is still high in South Asia as you can see on this map...but since it is sampled from the northwest one can't get a fix on the cline within South Asia. Another paper tells us more, among a sample of Tamils in Sri Lanka the derived variant is extant at frequencies around 25%, while the Sinhalese the frequency was 50%. I've seen other samples which suggest that among central Indian populations the frequency could be as high as 80%. The point is that though the frequency of the derived variant drops as you go south it remains rather high. Population number 30 on the map are the Uighers. Historical and other genetic studies suggest that the Uighers are a recent hybrid population; a pre-Turkic, likely Indo-European speaking, substrate seems to have been absorbed over the last 1,000 years. I point this out because Uighers have about equal measures of the ancestral and derived variants of SLC24A5; one might posit that this is natural noting the location of Xinjiang, but the Uighers as they are today are a new population which did not exist during the distant past. In fact, the earliest "European" mummies which have been found in Xinjiang date to 4,000 years ago. The circumstantial evidence is that for several thousand years a population of West Eurasian provenance was numerically dominant along the oases around the edge of the Tarim Basin. A late date for the initiation of the sweep which resulted in the high frequencies of SLC24A5 across Western Eurasia is 6,000 years in the past; that doesn't leave much time. To make a long story short it seems like at some point within the last 10,000 years the derived variant of SLC24A5 was strongly selected from Norway to Kerala to Morocco, but it stopped at the Himalayas and Altai.

Why was SLC24A5 selected? I've asked many, many, people. Geneticists who are studying this locus. No one seems to really know! Yes, it causes a change in skin color...but the people of South India are very dark-skinned for a reason. I have a few ideas, as some of you may have gathered, but, to generate these ideas I'm having to do a lot of reading. Men like William McNeill and Samuel Adshead have done a lifetime's worth of reading and become collectors of odd facts and obscure trends. If a great deal of the evolution over the past 10,000 years was due to adaptation to cultural change, then human scientists whose bread & butter are these topics need to join the conversation. History and Geography of Human Genes needs to become more than just a footnote, it needs to turn into a window onto another domain of knowledge which they must become fluent in so that they can help generate a better picture of the past.