The cultural canoe

A new paper, unfortunately not yet available to nonsubscribers on PNAS's Early Edition, has done some remarkable work on the evolution of canoe designs, putting some meat onto cultural evolutionary models.

The paper is nicely reviewed by K. Kris Hirst here, however. And when we mere mortals can get it, the paper is listed at the bottom of that and this post.

What Rogers and Ehrlich (yes, that Paul Ehrlich) did was analyse 95 variables in the design of the canoes of the "Lapita Complex", a group of Polynesians regarded as having colonised their islands around 1400-900 BCE. They found that the physical features of the canoes that affected their reliability and efficiency were slower to evolve than the cultural features such as decorations.

In cultural evolution, or at any rate modern cultural evolution*, cultural "traits" that are adaptive to the nonsocial environment (such as travelling in open ocean) are expected to be more stable because the physical facts don't change, while cultural techniques have to become better rather than vary for variance's sake. On the other hand, cultural symbolisms can change arbitrarily, and so they ought to evolve more rapidly.

Or must they? Cultural markings are highly adaptive - they mark you out as a member of the social group, and therefore open to reciprocal altruism from members of the group. If there is any differentiation between regional settlements, we might expect that symbolic evolution is a faster adapter than technological evolution like canoe building techniques.

If we were to cast these as analogues for biology, we might think that technological evolution is like ecological adaptation, where the rate of change, once a satisfactory "solution" has been reached, will be very slow. Cultural/symbolic evolution, however, is rather more like sexual selection, where adaptation is about finding conspecifics with whom to propagate. Sexual selection can drive evolution to absurd lengths, as Darwin pointed out with the peacock's tail (and Wallace with the birds of paradise); so to can symbolic selection. This is because symbolism is a kind of marker of cultural conspecifics. If you can pronounce the word "Shibboleth" correctly, as the Bible examples (Judges 12:6), then you are Hebrew, and so on. Using the "right" designs attracts social benefits and benefactors.

We have to be careful about cultural evolutionary models for several reasons. One is that evolution is not always adaptive, and in fact the bulk of it may (I am unconvinced we can say anything quantitative yet) be contingent and "accidental", so we should perhaps take care assuming selection is occurring. A more serious issue, for me, is that we often assume only one level of selection, and treat other types as, by exclusion, non-selective. But it seems to me that if there are fitness variations between symbolic varieties, then they will be selected for as much as technological varieties - only on a faster time scale and perhaps more rapid rate of reproduction. The fact that bacteria evolve more rapidly than humans doesn't mean they are not subjected to selection pressures (nor that we aren't either).

I'm reading a book on evolution and religion right now, which I will blog about soon, I promise, in which it is argued whether religion is adaptive. One of the common arguments is that if it wasn't adaptive in some earlier point in evolution, it can't be adaptive now. This either-or theme is common in disputes over evolution, and it seems to me misplaced. Each "step" in the evolution of some feature has to have been at worst no less adaptive than its alternatives, and once it is in place, it will be subjected to one or the other kind of selection thereafter. There was no "canonical" period of evolution and adaptation; we are in our "natural" environment in urbanised societies as much as we were in the early Paleolithic.

I hope to say more on this paper when it's available to me. But one irony I note is that two recent works discussing cultural evolution have both used canoes as their exemplars. One is the paper "Memes Revisited" by the philosopher Kim Sterelny. The other is the recent book Not by Genes Alone: How Culture Transformed Human Evolution by Pete Richerson and Robert Boyd, who use on their cover Inuit canoes (kayaks).

Rogers, Deborah S. and Paul R. Ehrlich 2008 Natural selection and cultural rates of change. Proceedings of the National Academy of Sciences Early Edition 18 February 2008:1-5.

* Old style cultural evolution, from the nineteenth century anthropological tradition, was very unlike Darwinian evolution - it was linear, went through necessary stages, and had little to do with adaptation. For this reason a great many anthropologists actually hate the idea of cultural evolution, for the right reasons but the wrong targets. See this post for more information.