Elliott Sober. From a Biological Point of View: Essays in Evolutionary Philosophy. New York: Cambridge University Press (1994), x + 255pp., $59.95 (cloth), $18.95 (paper).
British Journal for the Philosophy of Science, 45, 1994, 1084-1087.
Biological knowledge has increased exponentially in the last century or so, and it would be surprising if some of this knowledge did not have implications for philosophy. In contrast with a good deal of Elliott Sober's best known work, which aims to bring philosophical methods to bear on issues within biology, the theme of this collection of essays is to explore some ways in which biological ideas, or more specifically evolutionary ideas, may be brought to bear on philosophical issues. Sober notes in his Introduction that there is no systematic theme beyond the attempt to explore the philosophical implications of taking evolutionary biology seriously. And certainly this is a diverse collection, especially with regard to the extent to which the various arguments of these papers depend on any particular content to evolutionary claims.
The first five essays in the collection do provide a reasonably cohesive set of arguments. All attempt to throw light, by appeal to evolutionary styles of argument, on some fairly traditional philosophical issues. This group of papers (and the book) begins with a nice piece on egoism. Starting with an unusually clear explanation of the distinction between evolutionary and psychological altruism, Sober argues convincingly that even if human traits are never altruistic in an evolutionary sense (i.e. all selection is individual selection) there is no reason to infer that evolved motivations must be psychologically egoistic. This is an important argument to make against a common tendency to suppose that evolutionary theory somehow underwrites a crudely egoistic rational choice approach to social and psychological questions. Also excellent is an illuminating, and generally skeptical, discussion of evolutionary ethics. I had more reservations about the remaining three papers in this group, though all raise interesting ideas. An essay on why evolution should not have made us into solipsists didn't manage to convince me that the possibility that we might have evolved as solipsists was sufficiently pressing or even coherent to motivate the discussion. Perhaps this is because I don't share Sober's "hunch" that there isn't much to Wittgenstein's private language argument (p.46, n.2). An ambitious paper attempts to analyze from an evolutionary point of view the circumstances under which learning or hard-wired belief ("prejudice") will be evolutionarily advantageous. My reservation here is a familiar one with such applications of evolutionary models. Do the abstractions involved in the modelling leave enough of the complexity of the issues in their familiar applications to human cognition to have much bearing on the traditional questions? I had similar doubts about an essay on the advantages of truth-telling and deception, though here I did find suggestive the difficulties Sober raises for the common philosophical argument that language necessarily presupposes the predominance of honesty. Despite some general doubts about the implications of such exercises (doubts to which Sober is not insensitive) anyone who is interested in this kind of evolutionary modelling will find valuable insights in these papers.
All of these first five papers attempt to show how biological thinking has concrete applications to philosophical questions. The remaining papers (with one exception to be noted below) appeal to biology rather as a source of examples in support of positions on a range of issues in the philosophy of science. I do not mean this characterization to be in any way disparaging: expanding the range of examples applied to a philosophical question may certainly lead to new insights; and it is still too common to base general theses in the philosophy of science solely on considerations from physics. Nevertheless, this use of the biological point of view may seem less exciting than that suggested in the introduction and illustrated in the earlier papers.
The first of these papers--and incidentally the one with the least biological input--proposes a compromise between realism and empiricism that Sober calls "contrastive empiricism." Anyone who is convinced that realism and empiricism are genuinely incompatible will probably feel that Sober concedes more to the latter than the former. The realism in his account is mainly the insistence that truth is a goal of science, and that it is generally possible to discriminate rationally between theories and a properly defined class of rival theories. He insists, as a good empiricist surely should, that there is no discriminating between empirically equivalent theories, for example between theories that postulate an external world and those that rely on evil demons. He doesn't say much about how likely it is that there are empirically equivalent theories which, unlike in the example just mentioned, are equally plausible from any broader perspective. In pursuing the more moderate discriminatory task Sober relies, here as in several of the following essays, on a broadly Bayesian approach, though he insists that he is not committed to the general availability of prior probabilities (p.139) The idea that priors are often available, on the other hand, reflects a central theme of these papers, the importance of background assumptions to the practice of science. Though this seems to me a laudable emphasis, at one point, at least, it seems to me that Sober is carried too far by his insistence on it. In the essay "Explanatory Presupposition" Sober is concerned to argue that any question of the form "Why P?" should be understood as implicitly involving a specific set of alternatives the occurrence of which is denied. Thus he considers the familiar example, "Why did Willi Sutton rob banks?" and notes that this is ambiguous between various readings including "Why did Willi Sutton rob banks?", Why did Willi Sutton rob banks"?, and "Why did Willi Sutton rob banks?", all with their distinct sets of contrasting alternatives. He goes on to answer that an adequate answer to any such question must causally explain both the truth of the explanandum, and the falsity of members of the contrast class; and in fact must do so by tracing back to a common cause of these two contingencies. As Sober is aware, the thesis that 'why-questions presuppose the existence of common causes" (p.181) is not of great interest if the cause of P is simply taken to be a common cause of not (not P). But it seems to me that this may be much more typical of Why questions than Sober suggests, a possibility obscured by Sober's failure to mention the most natural reading of "Why did Willi Sutton rob banks?' with no special emphasis. If the question had been, say, "Why did Willi Sutton take up badminton?" there would be no temptation to distinguish these various contrast cases since, as far as I can see, taking up badminton is not incompatible with anything much except not taking up badminton.
I mention, finally, the remaining paper which applies the specific content of biological ideas to philosophical issues, "Evolution, Population Thinking, and Essentialism." This is the oldest paper in the collection (originally published in 1980), but remains as clear and insightful an exposition as any I know of the sense in which evolutionary theory undermines essentialism in biology. (A tiny quibble: the suggestion that Galton discovered the standard deviation and the correlation coefficient (p. 217) sounds to me distressingly Platonistic.) Space prevents me form commenting on the remaining papers on Ockham's razor, common causes, temporally directed laws, and the apportioning of causal responsibility, though I will say that the last mentioned is essential reading for anyone who wants to talk sensibly about the nature-nurture debate.
As Sober admits, this is not a collection of essays with a strong unifying theme, and although the reader should be convinced that the biological point of view has considerable philosophical payoff, no simple characterization of that payoff emerges. And I did find some unevenness in the collection: most readers will find some essays more satisfying than others. On the other hand Sober is always lucid, and expounds both scientific and philosophical ideas with exemplary clarity and concision. All in all, any philosopher of science will find ideas and insights here worth the price of admission.
John Dupré
Stanford University