Some caveats about cultural transmission models

Abstract

The symposium papers at hand apply a population-biological modeling approach to the problems of cultural transmission and cultural evolution. The papers by Boyd and Richerson, by Chen et al., and by Pulliam are conceptually very similar, each advancing a variant on the additive transmission model of Cavalli-Sforza and Feldman (1981), in which the probability that pseudoparticulate "traits" [essentially similar to Dawkins' (1976) "memes" and Lumsden and Wilson's (1981) "culturgens"] will be replicated in a focal individual is an additive function of their separate presences in that individual's "culture parents." Durham characterized this approach well by noting that the models view "cultural change as the outcome of an individual-level process of differential transmission." (p. 304). Perhaps many customs, techniques, neologisms, and so forth are propagated by processes resembling those modeled, but for the sake of injecting a little controversy into the discussion, I will argue that this approach is based on a weak analogy. Such analogic models may be heuristic at early stages of analysis, but they can become at best irrelevant and at worst impediments to progress as the analysis of process continues. The second set of critical comments applies only to Durham's paper. Despite his laudable and sometimes persuasive efforts to analyze real examples of complex cultural change, I believe that Durham's analysis of "coevolution" is fundamentally wrong-headed. Finally, I will briefly suggest that evolutionary biology indeed has something to offer students of social influence and cultural change, something other than an analogy with evolutionary population genetics. Invited to comment on the four papers in this symposium, I feel compelled to a dialectical stance: such commentary, if it is to be at all constructive, must be critical.