What Darwin Got Wrong by Jerry Fodor and Massimo Piatelli-Palmarini (UK: Profile, US: Farrar, Straus and Giroux) £20/$26 (hb)

Neo-Darwinism is, very roughly, the claim that natural selection is by far the most important explanation of biological form, the particular characteristics of particular kinds of organism. It usually includes a commitment to gradualism (the idea that evolution occurs in small steps), and often involves attributing central importance to genes as the units that natural selection selects, or at any rate as the objective measure of evolutionary change. Versions have been prominently defended in recent years by such authors as Richard Dawkins, Daniel Dennett and Jerry Coyne.

Neo-Darwinism is, however, a perspective under ever-growing pressure, not (or not only) from the antiscientific assaults of the religious, but from the advancement of science. The decline of this intellectual monolith is generally to be welcomed, not least because it may be expected to bring down with it some of its less appetising academic fellow travellers, most notably Evolutionary Psychology. At the same time those contributing to the demise of neo-Darwinism must be aware of the risk, especially in the United States, that they will provide succour for fundamentalist Creationists and aficionados of so-called Intelligent Design.

Fodor and Piatelli-Palmarini’s (henceforth FPP) book is intended as a contribution to the critical task just mentioned, and they are well aware of the potential hazards. Sadly, however, the book is an almost tragic failure: it is unlikely to be taken seriously as a contribution to the dismantling of neo-Darwinism and it has been, and will continue to be, picked up by the fundamentalist enemies of science.

The first half of the book does a decent job of summarising the recent scientific insights responsible for the growing difficulties facing neo-Darwinism. Neo-Darwinism, by virtue of its emphasis on natural selection, sees evolution as driven from outside, by the environment. Central among the difficulties that FPP emphasise are crucial respects in which evolution is constrained, or even driven, by internal features of the organism. This realisation has been promoted by evolutionary developmental biology (“evo-devo”), which has also highlighted the unacceptable black-boxing of development in mainstream evolutionary theory, a concomitant of the exclusive focus on external determinants of change. Also crucial has been a gradual move away from excessively atomistic views of organisms and an appreciation of the necessity of treating them as integrated wholes, illustrated by the impossibility of analysing the genome into a unique set of discrete elements, “genes”. And equally important has been the disclosure of the complexity of the relations between genomes and phenotypes.

While much material is presented that does indeed reveal the dire straits in which neo-Darwinism finds itself, the overall argument is generally elusive. I speculate that this is because there are two quite different conclusions in the offing. One, to which I shall return, is in my view eminently sensible: evolution is a historical process subject to many causal influences; natural selection is just one of these. The other, which is the source of the tragic failure mentioned above, is the thesis that natural selection is, more or less, a wholly incoherent idea. Unsurprisingly, this radical argument has attracted the majority of the considerable attention the book has received. And the argument is, to be blunt, almost embarrassingly broken-backed.

FPP’s assault on natural selection has been thoroughly dissected by other competent authorities, such as Ned Block and Philip Kitcher in the Boston Review, and I won’t go into great detail. In essence, they claim that natural selection cannot distinguish a trait’s being selected from a trait that is perfectly correlated with the first. So, for instance, if polar bears that match the colour of their environment are selected, so also are bears that are white, since these two traits are perfectly correlated. FPP think the only way of distinguishing these traits is in terms of counterfactuals (e.g., if the environment turned green, bears that matched their environment would be selected, white bears wouldn’t); and natural selection can’t see counterfactuals.

It’s hard to know what to make of this. Matching the environment is what causes bears to do well—their prey doesn’t see them coming. Being white causes them to match their environment, if that is white. These facts could in principle be tested by dying polar bears green, by spraying the ice green, and so on. It is true that testing these propositions can be seen as a way of testing what, without the tests, would have been counterfactuals. This can be said of any experimental test whatsoever. Investigating causality may indeed necessarily involve counterfactuals; but that doesn’t mean that causation in the world does. The metaphor of selection may suggest an agent that decides what causes survival, but to take this seriously is simply to be misled by a metaphor. There is no such agent. Camouflage causes successful hunting; whiteness enables successful camouflage in white environments.

Of course there is more to it, and this is gone through in detail by Block and Kitcher (see reference above). More interesting, perhaps, is why FPP are tempted by this feeble argument. One reason is that they are much impressed by an analogy between behaviourist psychology and natural selection, and take both to commit the error of failing to take account of the intentionality of the phenomena they address. (The difference is that the phenomena of interest to psychology really are intentional; natural selection, they think, unwittingly applies an intentional model to a quite inappropriate domain of phenomena.)

A second explanation is that FPP assume a central role for laws of nature in both science and causality. So, for example, they devote several pages to arguing that there are no laws of selection, laws that connect certain properties (being large, being fast, etc.) to selective success. Of course there are no such laws: sometimes it’s good to be big, sometimes small. But since I doubt whether any serious biologist has ever held that there were such laws, this is all rather puzzling. If FPP had engaged with the literature in philosophy of biology they would be aware that most contributors to this field have given up the idea that there are laws in biology. Evolutionary biology, at least, is better seen as the construction of families of models, rather than as the futile search for implausible laws.

After this disastrous diversion, FPP revert to a sensible and even important set of conclusions. Evolution is a historical narrative a proper telling of which requires reference to many causal processes. (Nothing here that Darwin would disagree with, least of all the pluralism – one is tempted to agree with those who have suggested that the inappropriate title was an attempt to appeal to the creationist market.)

I recently reviewed a quite different book criticising neo-Darwinism for American Scientist, John O. Reiss’s Not by Design: Retiring Darwin’s Watchmaker. This subtle and informed work offers a far more interesting and plausible account of what Darwin got wrong than do FPP. Reiss traces a tradition of teleological thinking that goes back to Aristotle and a genuinely materialist alternative originating with the Epicureans. Darwin’s mistake is his alignment with the former. The latter tradition, according to Reiss, reaches a seminal modern statement in Georges Cuvier’s idea of the conditions for existence: the fundamental explanation of the traits of an organism is that they couldn’t and wouldn’t exist without them. Survival of an organism, and even more of a lineage, is a very difficult trick. Darwin’s mistake was to assume a gap between existence and adaptation that is constantly pulling the organism to some superior state. But there is no such gap. This account provides a context for a convincing account of the scope and important limits of Darwinian natural selection

FPP come very close to such a position in the last few pages of their book, but sadly the misguided assault on natural selection is likely to prevent its being much noticed. The opportunity for a real advance in public understanding has been squandered. Reiss’s book is long and challenging, whereas FFP’s is short and written in an accessible, even racy style. The latter has received vastly more attention. I fear that modern publishing does not encourage the reader to pursue the long and winding road.

John Dupré is the Director of the ESRC Centre for Genomics in Society (Egenis) at the University of Exeter

Read all tpm’s reviews as soon as they are published, and much more, by subscribing to the magazine