More on Greek Hoplites: Darwinism and Social Evolution. (Comment)
Let me pick up the gauntlet that Hallpike threw down to Runciman in a recent issue of this Journal (Runciman & Hallpike 2001). Hallpike asserts that Runciman has failed to answer the critique of Darwinian analyses of human behaviour and socio-cultural systems that he offered in chapter 2 of his book, The principles of social evolution (Hallpike 1986). I have read chapter 2 -- as well as the rest of the book. I should say, incidentally, that I very much enjoyed the book and found much of it both interesting and congenial. However, I have to say that Hallpike's frank (if at times, ill-tempered) dismissal of the (neo-) Darwinian model seriously misrepresents what the Darwinian approach is all about. I will not elaborate in detail every incorrect or naive misreading of Darwinian theory, but simply draw attention to a selection of the more important ones.
(1) Hallpike seems to assume that genetics is the core of the whole Darwinian endeavour: it is not and never has been. As the work in animal behaviour, game theory, and cultural evolution modelling (and both Darwin's and Mendell's own original work) has emphasized, Darwinism (neo- or otherwise) is based on heritability (or fidelity of copying) between 'parent' and 'offspring', but it makes no claim whatsoever about the particular mechanisms that are involved (even though in one particular class of phenomena this happens to be DNA). Hallpike has fallen for the geneticists' public relations line that the new molecular biology is all there is of interest to study in the world.
(2) He assumes that natural (that is, environmental) selection is the sole process involved in Darwinian evolution (ignoring sexual selection and its ramifications, not to mention drift under neutral selection); this unfortunately leads him to make the egregious error of reifying the selection process into something that is 'Out there'. The environment happens to be one feature of the universe that imposes selection, but it is not the only one.
(3) He falls foul of the rather odd (and, among creationists, rather common) view that Darwinian evolutionary theory requires Boeing 747s to be assembled by random processes. There is nothing at all random about evolution. Hallpike has made the common mistake of confusing mutation (the random source of variation in genetic systems) with selection (as a factor directing the course of evolution). More importantly, mutations seldom arise de novo (as Goldschmidtian 'hopeful monsters') even in genetic systems, but rather as minor variants on an existing theme (which is one reason why evolution is always a bit of a Heath Robinson process). Halpike's description of how cultures evolve (by elaboration out of existing systems) is perfectly in keeping with the view of mutation deployed by Darwinian biologists, even when that elaboration is guided by some prescient human cogitation.
(4) He assumes that structural constraints are an alternative to natural selection: but in fact evolution (by natural selection or anything else) is always (implicitly or explicitly) subject to constraints. Evolution is always constrained by what is available (in terms of variation) and what is possible (in terms of the laws of physics, physiology, and growth). Were this not the case, butterflies would carry machine guns (to use an often quoted example from the biological literature). In any case, the existence of structural (for example, allometric or growth) laws in biology or sociology does not tell us why a particular evolutionary development took place -- merely that if an evolutionary process is going to happen, it must happen along a particular trajectory. Appeals to structural processes do not allow us to escape from the fact that we still require a mechanism to drive the system up its allometric trajectory, and that mechanism must either be drift (which can only happen when there are no costs involve d, which is rare when there are gradients involved) or selection. …