Read Ned Block and Philip Kitchers review of What Darwin Got Wrong, by Jerry Fodor and Massimo Piattelli-Palmarini.
See Block and Kitchers rejoinder to Fodor and Piattelli-Palmarini, below.
Jerry Fodor and Massimo Piattelli-Palmarini:
When we were writing our book, it occurred to us that there was a kind of misinterpretation of which a very incautious reader might be guilty, and which we ought to do our very best to block. We did do our very best; but to no avail. Ned Block and Philip Kitcher make precisely the mistake that wed dreaded. Worse, they then proceed to commit several other misreadings, the possibility of which, we admit, had not occurred to us. Well now do our very best to correct their mistakes, but time and space are pressing, and the opportunities for misinterpretation are, it appears, boundless.
First misreading: Block and Kitcher think we argue, erroneously, that with respect to correlated traits in organismstraits that come packaged togetherthere is no fact of the matter about which of the correlated traits causes increased reproductive success. They then speculate that we are making the very ambitious claim that whenever there are correlated traits there is no fact of the matter about which of the traits causes any effect.
But, of course, we dont believe, still less endorse, either of these theses. In fact, we think both are preposterous. We therefore spent our whole seventh chapter discussing a number of ways in which the causal roles of confounded variables can be, and routinely are, successfully distinguished. There are many such, the most obvious of which is perhaps John Stuart Mills method of differences. In effect, you run an experiment in which one but not the other of the correlated variables is suppressed. If you still get the effect, then it must be the variable you didnt suppress thats doing the causing. (If you think its maybe the ice rather than the alcohol that makes you tipsy, try taking one or other out, drink whats left, and see what happens). People, scientists very definitely included, do this sort of experiment all the time. And often it works fine; we report lots of cases in our book. All this is familiar from Philosophy 101; do Block and Kitcher really believe that, old and case-hardened as we are, we could have failed to notice this?
However, the questions of whether there is a fact of the matter about which variable causes what, and whether this fact of the matter is epistemically accessible, really must not be confused with the question of whether natural selection, as neo-Darwinians understand it, is able to distinguish the causes of fitness from their local confounds. You and I can work out what caused what in all sorts of ways: we can use Mills method, or we can take the system of causes and effects apart and find out what mechanisms operate inside it, or we can ask the guy who built the thing (if somebody did) how it works, or we can look it up in the scientific journals, or we can ask Google, and so on and on. But natural selection cant do any of these things. It cant look inside, and it cant run experiments, and it cant contrive theories about internal structure, and it cant consult the intentions of the builder. Doing any of those requires having a mind, and, by general consensus, natural selection doesnt have one. All natural selection can do is respond to correlations between phenotypic traits and fitness. And that doesnt help because, by assumption, if either of the confounded traits is correlated with fitness, so too is the other.
(We think, by the way, that its no accident that the notion of selection Darwinians favor is so thin. What they want is a theory of evolution that applies to any creature in any ecology. We doubt that such a theory is possible. We think that how a creatures trait evolved in its ecology is likely to be very sensitive to which creature it is, and which trait it is, and which ecology it is. Such facts belong to natural history which, like historical explanations in general, reveals the causal antecedents of phenomena, not the empirical necessities that subsume it. That, however, is another story; one that we try to tell a little in our book).
Darwin needs an account of selection-for that distinguishes phenotypic traits that cause fitness from traits that are merely correlated with traits that cause fitness.
Second misreading: The last part of Block and Kitchers review offers an account of the opacity (or otherwise) of cause. Its not an account that we favor, since we think its most implausible that causation is a relation between properties: causes and effects have dates and times, but properties do not. (Roughly, we think causes and effects are events which consist of the instantiation of properties. Properties dont cause anything, though their instantiation often does. The reader, however, may feel that only a philosopher could care, and we suspect is right to feel that.)
Anyhow, it simply doesnt matter to the present issues. What matters here is that, by stipulation, selection-for is sensitive to distinctions among locally coextensive states; i.e., it can distinguish between As causing Cs and Bs causing Cs even when every actual instance of an A is accompanied by an actual instance of a B and vice versa. Darwins theory of natural selection just is that phenotypic traits are selected for their effects on fitness. So Darwin needs an account of selection-for that distinguishes phenotypic traits that cause fitness from traits that are merely correlated with traits that cause fitness. Just as Block and Kitcher say, natural selection make[s] a causal claim: having the trait causes greater reproductive success. That being the claim, Darwin is in need of a mechanism that responds differently to A-caused Cs on the one hand and to B-caused Cs on the other, and that does so with no access to the endogenous structures of organisms; indeed a device that responds selectively to A-caused Cs and B-caused Cs even though it has access to nothing at all except correlations between phenotypic traits and fitness.
We dont think that there is or could be such a mechanism. We have heard of many ways of distinguishing confounded variables, and none of them will do the job. In any case, this line of argument has precisely nothing to do with our (and Darwins) occasional indulgence in heuristic personification; and precisely nothing to do with the
thesis . . . that natural selection can make the same finely graded discriminations available to a human (or divine?) observer. Where on earth did that come from?
We think whats emerging is a quite deep disagreement over what the theory of natural selection is supposed to be and do. Our view on this agrees with what practically every Darwinist weve come across has claimed: natural selection is supposed to be a theory of adaptation; i.e., its supposed to explain how adaptive phenotypic changes become diffused in a population; i.e. its supposed to explain what instances of adaptation have in common as such. This is the traditional, mainline, view of what theories are for (Plato: a theory of beds would explain what beds have in common as such; Saul Kripke: science discovers essences; etc.) A mere chronicle of instances of adaptation would not therefore amount to a theory of adaptation. It would just be natural history. We havent the slightest doubt that Darwin thought that he had discovered a theory of adaptation. It was, to be sure, a pretty thin theory, as it would have to be in order to apply to evolving creatures as such, whatever their phenotypes and whatever their ecologies. But its quite thick enough to be substantive; indeed, its quite thick enough to be false.
The outlines are familiar:
There is random variation of phenotypic traits.
There is some ecological variable that is sensitive to the strength of the correlation of such traits with fitness.
There is some mechanism that alters the relative frequency of the trait in the population so that, all else equal, it varies with the strength of the correlation between the trait and fitness.
For example, suppose random variation produces a trait that tends to make its bearers invisible to their predators. Then, all else equal, the predators gobble up the creatures that dont have it, and the relative frequency of the trait in the population increases from generation to generation. To repeat, we havent the slightest doubt that this is the process that Darwin called natural selection and that Darwinists have always believed in some or other version. In fact, it sounds pretty good. It sounds like it ought to work.
But it doesnt. A way to see that it doesnt (not by any means the only way) is to consider confounded (linked) phenotypic traits, one but not the other of which is fitness-enhancing. Both traits are then correlated with fitness, so both should count as adaptations according to the formulation of natural selection given above. But only one of them is a cause of selection, so only one of them is an adaptation, and, though both are selected, only one is selected-for. Thus the free-rider problem. Prima facie, free riding is a counterexample to natural selection.
As far as we can see this objection to natural selection is decisive. If so, there are two ways in which the dialectic can now proceed. First, one might try for some alternative to natural selection, one that doesnt have a free-rider problem. We think there is unlikely to be such an alternative simply because there is no inference from correlation to causation, and, by definition, its the latter, not the former, that natural selection requires of the traits it selects for. The other possibility is to give up on the project of finding a mechanism for evolution and study the fixation of adaptive traits case by case. Since all the evidence suggests that they are extremely heterogeneous, this should keep evolutionary biologists busy well into the indefinite future. It is what we think evolutionary biologists actually do for a living (despite their inaccurate claim to believe in the theory of natural selection.) We hope not to sound petulant, but the last couple of chapters of our book go on about all this at considerable length. We do hope you will have a look at them.
One final point. Block and Kitcher keep suggesting that we dont know enough about biology to criticize a theory that so many biologists hold dear. We are unmoved. For one thing, there is, as far as we can tell, much less consensus among experts than Block and Kitcher suggest. Certified evolutionary biologists have aired their disagreements in many publications over the years. Some are unwavering in their allegiance to the modern synthesis (the fusion of Darwinism with genetics); some advocate revisions and extensions, even radical revisions and radical extensions; and some have declared that the modern synthesis is dead. Quite likely, the neo-Darwinists constitute a majority. But so what? Surely, its common ground that the truth isnt to be found by counting noses, not even expert noses; experts have been known to make mistakes.
Everybody makes mistakes. Even biologists do; even lots of biologists assembled together do; even great biologists like Darwin do. You dont need an advanced degree to make a mistake or to spot one. You dont need an ornithologist to tell a hawk from a handsaw. If, therefore, you think a mistake has been made, its a good thing to say so, loudly and in public, so that it (or you) can be corrected. We would have thought this a consensus view among scientists, philosophers, and everybody else who cares at all about distinguishing whats true from what isnt. The parochial is the enemy of the true and should not, therefore, be encouraged by its friends.
Ned Block and Philip Kitcher respond:
We are pleased to see that there is now so much agreement between Jerry Fodor and Massimo Piattelli-Palmarini and us. They now concede (at least in the second paragraph of their reply) that one but not the other of two correlated variables can have a differential effect on reproductive success. To use our example, dark color in the famous moths can be causally responsible for their reproductive success in polluted environments even though correlated properties have no effect on reproductive success.
Since selection-for just is identical to having a differential effect on reproductive success, one might suppose that Fodor and Piattelli-Palmarini would have to accept that dark color was selected for whereas properties correlated with the dark color were not. But they hedge their acceptance of this standard identification. For they go on to say: Darwin needs an account of selection-for that distinguishes phenotypic traits that cause fitness [i.e. reproductive success] from traits that are merely correlated with traits that cause fitness. We find their odd demand frustrating because as the pertinent notions are used in evolutionary theory and in philosophy of biology, traits that cause fitness just are traits that are selected for. Selection for a particular trait requires nothing further.
What is the additional requirement on selection-for over and above differential effect on fitness? Fodor and Piattelli-Palmarinis answer: Darwin is in need of a mechanism that responds differently to A-caused Cs on the one hand and to B-caused Cs on the other. This is either to contradict the common understanding of selection-for as differential causation of fitness, or to take back their concession that just one of two (or more) correlated properties can differentially affect fitness.
Fodor and Piattelli-Palmarini are imposing a requirement on natural selection that no biologist or philosopher whom we know of has ever suggested.
Consider the moths. That dark color has a differential effect on fitness is sufficient to establish that dark color is selected for. There is no need for a detector mechanism to register that dark color is the cause of the increased reproductive success. Fodor and Piattelli-Palmarini write as if Darwinian evolutionary theory requires some analog of litmus paper that turns red when there is selection-for and blue when a trait is merely a free rider. But that supposed requirement is ludicrous. We mentioned Fodor and Piattelli-Palmarinis repeated personifications of natural selection, derived from Darwins unfortunate attempts to explain the idea. Thus they say that natural selection (or evolution) is insensitive to the distinction between traits selected-for and their free-riders. Our critique charitably construed these personifications as part of a philosophical argument designed to show a supposed intensional fallacy. Their reply reveals a confusion much simpler than the ones we criticized in our review.
Fodor and Piattelli-Palmarini are imposing a requirement that no one elseno biologist or philosopher whom we know ofhas ever suggested in the history of evolutionary thinking. The idea that, in order for dark color rather than some correlated trait to be selected for, there is a need for a device that responds selectively to one rather than the other is simply an eccentric invention. It is as if, having abandoned Mother Nature, Fodor and Piattelli-Palmarini think there has to be a mechanism that takes her place. As they say, We dont think that there is or could be such a mechanism. Who ever thought otherwise? They continue: We have heard of many ways of distinguishing confounded variables, and none of them will do the job. But the job is done by causation itself. Dark color promotes progeny; no other mechanism is needed. To repeat: there is selection for a (heritable) property just in case having that property causes increased reproductive success. If Fodor and Piattelli-Palmarini have some other job in mind, they have to explain why Darwinian evolutionary theory needs that job to be done. They claim to be criticizing natural selection, as neo-Darwinians understand it, but instead they are criticizing a strange version of natural selection, invented solely so that it can be demolished.
So far, we would not characterize this as a deep disagreement. However, there is a disagreement described in Fodor and Piattelli-Palmarinis response that is worth discussing. It concerns the idea that a mere chronicle of instances of adaptation would not . . . amount to a theory of adaptation. It would just be natural history.
Fodor and Piattelli-Palmarinis charge against Darwinian evolutionary theory reduces to the observation that there is no general implementation of natural selection, no all-purpose mechanism by which one trait rather than another differentially affects fitness. In consequence, according to them, there is no theory of natural selection. Many biologists and philosophers would want to use the term theory rather differently from the usage Fodor and Piattelli-Palmarini favor. Some would wince at the idea that Darwin and his successors were in the business of finding some analog of a Platonic theory of beds. One of the greatest twentieth-century evolutionary theorists, Ernst Mayr, pointed out, repeatedly, that Darwins ideas are profoundly anti-essentialist, based on population thinking. But let us simply grant Fodor and Piattelli-Palmarini the word. There is no theory of natural selection. So what?
Evolutionary biology, from Darwin on, is concerned to understand the history of life. Evolutionary biologists want to know what transitions have occurred, how species are related, what general types of factors have played a role, and what particular causes have operated in particular instances. That is what the subject is about. Does it require any theory of the kind Fodor and Piattelli-Palmarini tendentiously demand?
In a word, no. For 150 years everybody in the business has known, very clearly, that cases of natural selection are diverse. The point is elementary: the natural selection of moths in polluted woods results from the causal process in which dark coloration makes for effective camouflage; the natural selection of finches in the Galapagos stems from causal processes in which beaks of different sizes and shapes enable birds to crack open seeds and thus obtain food; the natural selection of (male) guppies in streams results from causal processes in which spots on the fish are large enough to attract mates but not so large as to interest predators; and on and on and on. Any search for a universal piece of machinery implementing natural selection would be futile. Yet that makes not the slightest difference to the Darwinian project. The explanation of the history of life can proceed in just the ways it has done and continues to do, exploring the famously problematic cases (the origins of the cell, the evolution of sex), finding some general types of processes that occur with considerable frequency (kin selection, balanced polymorphisms), tracing the causes of particular transitions about which scientists (and others) are curious (increased brain size in the hominid line, the evolution of flight).
We entirely agree that these investigations can benefit from considerations of the type that Fodor and Piattelli-Palmarini review in Part I of their book (constraints, epigenetics, attention to properties of the genome, and so forth). One of us (Kitcher) has argued for decades against the naïve adaptationism that assumes the untrammeled operation of natural selection. Recognizing the ways in which the possibilities of Darwinian explanation have been enriched in recent years is, however, a far cry from declaring that natural selection is unnecessary or that the idea is incoherent. Retelling familiar things does not debunk natural selection.
The argument of the book thus depends on the charge that natural selection is incoherent. In the book itself, this is presented in terms of an intensional fallacy, aimed, apparently, at supposed difficulties in separating correlated traits from genuine causes of fitness. In our critique we responded to these difficulties in the most prominent forms in which Fodor and Piattelli-Palmarini present them. After our critique, Fodor and Piattelli-Palmarini have apparently decided that the crucial point is the lack of a theory of natural selection. But, as we have noted here, nobody needs a theory of the type they demand. Instead of the flawed reasoning that seemed to be leading them to a challenging conclusion, they have now shifted to an absurd demand for some uniform process that will be common to all cases in which traits cause increased reproductive success.
As old and case-hardened veterans, Fodor and Piattelli-Palmarini ask their readers to understand that they are aware of points made in introductory philosophy courses. We do not accuse them of not knowing (for example) about Mills methods. The trouble is that they are commenting on an area about which they know so little that they fail to see how basic philosophical points apply. Even the slightest awareness of fallibility might incline them to wonder if their supposedly bold attempt to spread enlightenment is really ignorant blundering that can be exploited by people whose causes they oppose.
Jerry Fodor is State of New Jersey Professor of Philosophy at Rutgers University and coauthor with Massimo Piattelli-Palmarini of What Darwin Got Wrong.
Massimo Piattelli-Palmarini is Professor of Cognitive Science, Linguistics, and Psychology at the University of Arizona and coauthor with Jerry Fodor of What Darwin Got Wrong.
Ned Block, Silver Professor of Philosophy, Psychology, and Neural Science at New York University, is author of Consciousness, Function, and Representation.
Philip Kitcher, John Dewey Professor of Philosophy at Columbia University, received a Lannan Foundation Distinguished Book Award for Living With Darwin.
Ned Block and Philip Kitcher, Misunderstanding Darwin