I want to clarify a few points that I alluded to below. The fact that lactase persistence is nearly fixed in northern Europe can be used to make obvious inferences about the culture of this region. Cultural traditions clearly shaped via direct selection the gene frequencies and therefore the expressed phenotype. But most gene-trait relations that we are interested in are more complex, and the selection regime is less cut & dried. When it comes to "culture," broadly speaking (as opposed to something very precise such as the utilization of dairy in the daily diet), it is hard to make generalizations which yield us more information than we started with. One of the problems that crops up is that written texts, and avowed norms, can give a skewed perspective of the reality in the past and present.


If aliens looked at the preserved letters of Roman nobility such as Symmachus as the models for the nature of ancient Latin, they would likely be very mistaken. The likely reality is that modal Latin spoken by the masses was very different from elite Latin preserved in the extant literature, and can be gauged by the fact that Christian church made a proactive effort to recruit more of the elite into the priesthood because of the awkwardness that occurred when common preachers addressed upper class parishioners in "rough" dialect. There is also the issue that avowed norms are not always practiced widely. In Mother Nature Sarah Blaffer Hrdy observes that the lower caste vassals of the Masai espouse the same preference for sons that their overlords do, but mysteriously their mortality rates are inverted from that the Masai so that female children seem to do better than male children. The functional explanation is pretty obvious, female children can "marry up" into the Masai, while male children have far less chance of social mobility. This Trivers-Willard effect has been discerned in other situations, but, the fact that extant documentary evidence tends to skew toward the upper classes often masks the extent of daughter-preference among the lower orders (e.g., a survey of cemeteries of peasants in medieval Europe showed a bias toward youg male offspring). Additionally, modernization tends to result in the inevitable percolation of elite norms down to the masses, in India this can be seen in the spread of dowry during the early 20th century downward from the upper castes and the extinction of the practice of brideprice (where the family of the bride receives payment). But sometimes even in these societies modernization does only so much, ethnographic studies of the Khasi (a matrilineal society) and neighboring Bengalis (a patrilineal society) showed that both societies exhibited the maternal "grandmother effect" insofar as grandmothers favored the offspring of their daughters more than their sons. This should not surprise in regards to the Khasi, who are matrilineal, and where men and women are relatively equal in their relationship, but Bengalis notionally base their society around male descent groups. Like most north Indian societies women are strangers in their husband's houses (patrifocality is the norm, in contrast again to the Khasi), and their children are members of their husband's family. But the reality is often far different, in my own family my mother would wryly observe that it is somewhat indecent that her children have a closer relationship to her brothers (our maternal uncles) than to my father's brothers and sisters. Though this is not "decent," it is an ancient pattern attested within Indo-European societies. Anthropologically the explanation is often that maternal uncles can be assured of relation to their nieces and nephews by their sister to a greater extent than paternal uncles can be to their nieces and nephews by their brother. This is not a line of thinking which would be greeted positively by traditional Bengalis of course.

In any case, a few months ago I thought about the possible shallowness (or recency) of such avowed social norms when an acquaintance of mine told me the following story: basically, he had left a non-profit in Afghanistan. The reason was that his boss was having an affair with a woman who was also employed by this NGO. My acquaintance's boss was married to a close friend of his, and the "other woman" was engaged to a fellow Pashtun in Kabul. The short of it is that everyone knew of the affair in the organization, and some of them who were posted in Kabul were not happy with the possibility that a family tragedy might arise out of this, especially with the woman's impending marriage. Now, I bring this up because Pashtuns are pretty freaky about how much they control their womenfolk. When my father was a student in Pakistan the locals explained that whenever you saw one man chasing after another with a scythe or any other such dangerous implement it was likely to be a Pashtun. The locals thought it was great fun and seemed to act as if this was normal (this was Islamabad in the late 1960s). It was simply the "Pashtun way." And yet was this always the Pashtun way? The woman who I allude to above seemed to have sexual feelings which allowed her to indulge in very dangerous activities (in the context of Pashtun culture). For gene-culture coevolution to work the selection must either be persistent or powerful. Because of the variation of cultures over time in regards to particular traits persistence is often difficult. As to the strength or power of selection, I suspect that though it can be powerful in a small segment of the population, for most of history formalized norms codified in the texts which we analyze to make sense of the past were relevant for the elites. This means that the majority of the population lived "pre-civilized" lives as producers of surplus for the elites, who were the engines and consumers of civilization.

Addendum: I've re-edited the part about the NGO & Afghanistan to be more general so that individuals can't figure out who I was talking about (in case you saw version 1). I myself don't know.

Labels: culture, Evolution