David Sloan Wilson’s book Darwin’s Cathedral is a kind of follow-on to the book he wrote with Elliot Sober, defending group seleciton and the possibility of real altruism. DCis about religion considered as an agent of group selection. So he has to start by establishing the real possibility of group selection and so rehabilitate holism, or at least point out the limits of reductionism. This leads him to show that all arguments from design, including the standard “Blind Watchmaker” functional arguments of modern biology, from Maynard Smith, through Dawkins and Dennett, are much less reductionistic than they seem. The concept of adaptation is itself holistic relative to molecular genetics, because it’s agnostic about the particular mechanisms involved in an adaptation.
The most relevant form of holism for our purposes is based on the concept of adaptation. Consider all the artificial selection experiments that have been conducted on fruit flies (Drosophila): they have been given short wings, long wings, no wings, many bristles, few bristles — the list goes on and on. The question ‘Why do these particular fruit flies have this particular phenotype’ has two answers. First, every phenotype is caused mecha- nistically by genes that interact with each other and their environment during development. Second, the phenotype exists because of a history of selection for that phenotype, coupled with heritable variation. These two explanations are usually labeled ‘proximate’ and ‘ultimate’ respectively, and there is a sense in which the latter is more fundamental than the former. As one classic example, Cohan (1984) divided a single population of fruit flies into a number of isolated smaller populations. He then selected the same trait (wing vein length) in each isolated population, examining the response to selection and the underlying genetic mechanisms in each case. It turned out that the same phenotypic trait of long wing veins evolved by different genetic mechanisms. A single ultimate-level explanation sufficed for all the populations, but different proximate-level explanations were required. More generally, the natural world is full of species that have evolved similar solutions to life?s problems (e.g., hard exteriors as protection from predators), even though they are composed of different genes and physical materials (e.g., chitin for beetles, cellulose for plant seeds, and calcium carbonate for snails). In short, heritable variation and selection provide a solid foundation for the holistic claim that the parts permit the properties of the whole but do not cause the properties of the whole. A lump of clay permits but does not cause the form provided by the sculptor. To the extent that the physical make-up of organisms provides heritable variation, it becomes a malleable clay that can be sculpted by selection. Evolutionary biologists rely upon this kind of holism all the time. They confidently predict that desert animals should be sandy colored, a bird’s wing should be aerodynamically efficient, and small fish should be timid in the presence of predators with- out any reference whatsoever to the genes or physical material that make up these organisms. Proximate explanation (‘the sandy color in this species is caused by chromatin granules, which are coded by genes on the fourth chromosome’) complements ultimate explanation (‘the sandy color is caused by a history of selection favoring cryptic coloration’) but never substitutes for it.The reason this caught my eye, apart from its cleverness, what that his distinction between proximate and ultimate explanaitons was exactly the distinction I was trying to make in the Darwin Wars beween “functional” and “analytical” genes: The functional genes supply proximal explanations; the analytical genes are those postulated to provide “ultimate” explanations. It’s clear that this was a distinction that gave a lot of readers trouble. It gave me trouble when I tried to explain it, though I understood it quite clearly when I was writing. I think that the Wilson argument might make it clearer.