Sunday, February 17, 2008

A revival of functionalism?   posted by Razib @ 2/17/2008 12:11:00 AM

Human Culture Subject To Natural Selection, Study Shows:
The Stanford team studied reports of canoe designs from 11 Oceanic island cultures. They evaluated 96 functional features (such as how the hull was constructed or the way outriggers were attached) that could contribute to the seaworthiness of the canoes and thus have a bearing on fishing success or survival during migration or warfare.

They also evaluated 38 decorative or symbolic features (such as the types of carved or painted designs). They analyzed mathematically the rates of change for the two groups of canoe design traits from island group to island group. Statistical test results showed clearly that the functional canoe design elements changed more slowly over time, indicating that natural selection could be weeding out inferior new designs. This cultural analysis is similar to analyses of the human genome that have been successful in finding which genes are under selection.

The study is coming out on the 19th in PNAS (so that means it will show up on the website at some time after that date). As most of you know in the 1960s the neutral theory of molecular evolution emerged in response to the finding that there was a great deal of extant genetic variation on allozyme loci (OK, to be fair neutralist ideas predate the empirical results; but I think it is clear that those results made the model intellectually far more compelling). Prior to this there were two broad schools of evolutionary genetic thought; one group accepted that there would be low levels of polymorphism due to balancing selection, and another assumed that there would be little to no polymorphism because of selective constraint. No matter the rearguard attempts by the likes of Richard Dawkins to argue that molecular variation "doesn't count," I think the neutralist (or nearly neutralist) insights are important in giving us a better understanding of the nature of evolutionary dynamics on the genomic scale. In The Origins of Genome Architecture Mike Lynch argues that low effective population sizes have had a strong role in shaping the character of genomic variation in more complex organisms. In other words, we are all non-adaptationists now!

What does any of that have to do with the paper above? Peter Richerson & Robert Boyd, L. L. Cavalli-Sforza and Marcus Feldman and E. O. Wilson & Charles Lumsden have all attempted to show how evolutionary processes are relevant to our understanding of human soceties. Unfortunately, as L. L. Cavalli-Sforza observes, cultural anthropologists are less interested in understanding humans as opposed to interpreting them. Formal frameworks to accompany the mass of empirical observations are simply neglected or seen as unnecessary. This is an unfortunate overreaction to the hubris of earlier generations of anthropologists who attempted to shoehorn all human variety into a set of functional adaptations. Instead of a happy medium where skepticism is balanced with empiricism and rationalism, anthropology has swung from a total lack of critical analysis toward one where positive assertions are eschewed on principle (unless, of course, those assertions are directed toward Western culture).

In Darwin's Cathedral David Sloan Wilson tries to make an argument for resurrecting a functional understanding of cultural traits as adaptations. I think that this sort of work is hard-going, at least beyond the level of triviality (e.g., the rationales for why the Inuit dress the way they do is rather straightforward). That is because "culture" is a very broad and ill-defined term and the selective pressures are myriad; the environment, the social matrix and the correlations with other traits are all critical. Wilson's methodology in Darwin's Cathedral was to use case studies; I don't think that that will cut it. Rather, massive surveys of collected data tested via statistical methods are probably more useful in extracting out the adaptive trends as a function of time and space. I do not, for example, think it is a coincidence that over the last 2,500 years all the complex cultural traditions on the World Island became associated with what we would call "Higher Religions," roughly, the fusion of supernaturalism with philosophy and institutional structures. But were these parallel developments a function of the specific adaptive needs of these complex societies? Or where they perhaps inevitable byproducts of the sufficient intersections of modal human psychology with the rise of the novelties of mass post-tribal society?

These are big complex questions. I think that are certainly functionally significant cultural adaptations. That being said, I am not sure sure that they are responsible for the preponderance of between cultural variation. To go back to the example of Higher Religions, I think one can plausibly argue that some sort of synthesis between intuitively appealing extant supernaturalism with the intellectual & institutional abstracting tendencies of complex societies made them inevitable, necessary perhaps. Societies which were united by a common religious ethos may very well have been more fit than societies still characterized by a welter of tribal gods uncomfortably corralled under one political dispensation (though the dynamic might usually have been played out within an intrasocietal context; e.g., the introduction of Buddhism to Tibet and Japan by a particular faction at court and the subsequent nativist reaction with failed). But the specific nature of the Higher Religions may very well be arbitrary, neutral so to speak, because like a synonymous substitution they have no functional significance.

Obviously the paper above targets the law hanging fruit. Engineering is not contingent upon the caprice of human social dynamics; it works, or it doesn't, by the grace of Mother Nature. But it's a start, as it is a reality check upon those who would argue that the full sample space of cultural possibilities are theoretically at play, and equally likely. The next step is to start examining traits not so strongly constrained by physical conditions.

Labels: ,