Jerry Fodor

  • Consilience: The Unity of Knowledge by Edward O. Wilson
    Little, Brown, 374 pp, £18.99, September 1998, ISBN 0 316 64569 9

Suppose God took it into his head to make another world just like ours; if one is good, why wouldn’t two be better? There’s a lot he’d have to see to; dividing the light from the dark and the seas from the dry land would hardly make a start. He’d need to conjure up another Milky Way, for example, that’s exactly counterpart to ours, and arrange the very same number of stars in the very same relative locations. There would have to be the same number of planets circling these stars as circle ours; and the same number of moons circling the planets … and so on down to the least significant particles of asteroidal debris. All of which he’d have to set moving, at just the right velocity, away from duplicates of all the other galaxies.

Similarly, on a local scale: he’d need a New Pacific Ocean, with just the right amount of salt in just the right amount of water, to wash the shores of a New California with just the right amount of smog in the air. Also, item, another tree just like the one that I can see from here, and another me to look at it. And this new JF would have to have, along with my slightly guilty passion for Wagner, a worry, just like mine, about whether he’d remembered to feed the cat and lock the apartment when he left for work this morning. Likewise, mutatis mutandis, for you; and for the tree that you can see from wherever it is that you are; and likewise for all your friends and relations and all their trees and cats. So many states of affairs, and they all seem to be so unconnected. First blush, anyhow, the world is a great democracy of facts, each set up in business on its own.

But maybe not. Maybe there are basic facts and, if God settles them, the rest fall into place; he takes care of the pennies, and they take care of the pounds. That something of the sort is so, and that the basic facts concern the physics of things that are very small, seems central to the view of the world that has been emerging from scientific inquiry for the last five hundred years. Put the quarks and the protons in their places, along with whatever else it is that basic physics talks about, and you don’t also have to worry about placing the planets, or the Pacific, or me, or my anxieties. Physics determines chemistry; chemistry determines biology; biology determines brain science; and brain science determines my neuroses, my fondness for Wagner and the rest of my mental life. Yours too. All the facts that there are, including all the facts that there are about minds, ‘supervene on’ the facts of basic physics. So the story is supposed to go.

As things stand, it’s far from clear that any version of this story will prove to be true. The determination of the facts of psychology by facts about the brain, in particular, is less a confirmed hypothesis than an article of faith some scientists share. Still, it’s not an irrational faith and there are cases where at least the outlines of a comprehensive physicalistic reductionism seem to be emerging. For example, it seems increasingly certain that the laws of heredity supervene on the biochemistry of genes; if that is so, it’s a remarkable vindication of the general physicalist worldview. Extrapolating from such successes, one might well suppose that some kind of mind/brain supervenience is likewise on the cards. The case for a bottom-up ontology is good enough, even as it stands, to warrant thinking hard about what it means to us if that turns out to be the sort of world we live in.

The distinguished biologist E.O. Wilson has been thinking hard about this; which is a fine thing. But not to very great effect; which is too bad. The key issue is this: if physics fixes all the facts there are, does it follow that all the explanations that there are are physical explanations? Would the physicalistic unity of being entail the physicalistic unity of knowledge? Wilson clearly thinks so. ‘I have argued that there is intrinsically only one class of explanation. It traverses the scales of space, time and complexity to unite the disparate facts of the disciplines by consilience, the perception of a seamless web of cause and effect.’

Consilience is an epistemological thesis: roughly, it says that all knowledge reduces to basic science. This would appear to be very different from the metaphysical thesis that all the facts supervene on the facts of basic science. In particular, it is by no means obvious that the epistemological kind of physicalism follows from the metaphysical kind. And if it doesn’t, then an enthusiast for the second might consistently – even plausibly – reject the first. Wilson’s failure even to notice this possibility makes a shambles of his book.

If the case for metaphysical physicalism is induction over its past successes, the case against consilience is induction over its lack of them. In fact, there are very few examples so far in which it has turned out that the explanatory apparatus of a higher-level science can be paraphrased in the vocabulary of some science further down. To be sure, there used to be a lot of interest in a research programme that the Logical Positivists called ‘the unity of science’: from astronomy to zoology, all scientific vocabulary was (sooner or later) to be defined in that of basic physics. ‘Science is physics plus abbreviations,’ so such Positivists said. But hardly anyone believes this any more.

Reviving the unity of science programme is most of what consilience amounts to. Though Wilson is aware that it’s now pretty much a dead issue among philosophers of science, he thinks ‘its failure, or put more generously, its shortcoming, was caused by ignorance of how the brain works. That in my opinion is the whole story.’ If philosophy is your line of work, this will likely raise your eyebrows, since you probably think (as the Positivists also did) that the burden of the unity of science is maybe methodological or ontological, but certainly not psychological or neurological. Offhand, I can’t imagine what kinds of fact about the brain would have saved the Positivist philosophy of science, and Wilson doesn’t say.

Wilson thinks consilience is in disrepute because philosophers don’t take science seriously. On the contrary, it’s in disrepute because they do. It’s attending to how the scientific edifice is actually organised that makes the eventual reduction of the rest of science to physics seem so unlikely. Here, for once, ‘don’t think, look’ sounds like good advice; one could wish that Wilson had taken it. For what one sees when one looks doesn’t at all suggest a structure that is collapsing into its basement. If the unity of the sciences is true, then there ought to be fewer sciences every day, as basic physics absorbs them one by one. But what’s going on seems to be quite the reverse: an accelerating proliferation of new disciplines; the damned things multiply faster than college deans can keep up with them.

I think one should be moved by this apparent failure of consilience. I think it poses a deep, deep question about the way our physicalistic ontology comports with the pluralism of scientific discourse to which any college prospectus bears testimony. I think we ought to pay attention to what the structure of the scientific institution seems to be trying to tell us. Not so Professor Wilson. Wilson is in a pique with the structure of the scientific institution. ‘The ongoing fragmentation of knowledge and resulting chaos in philosophy are not reflections of the real world but artifacts of scholarship.’ That is a claim to conjure with, and one would like to see some arguments. Wilson says – and this is really all he says that’s pertinent – that its ‘best support … is no more than an extrapolation of the consistent success of the natural sciences’. But what the natural sciences have been successful at is not to Wilson’s purpose. They have been remarkably, impressively, gloriously, good at explaining things. But they generally do so in their proprietary dialects, not in the language of basic physics. The success of the sciences is one thing: the unity of science is quite another.

It is, by the way, characteristic of Wilson’s book that he fails to notice the difference between what one might call vertical and horizontal consilience. Cases of the former (the molecular theory of heat; the physical theory of the chemical bond) provide the paradigms for the unification programme. Far more frequent, however, is the joining forces of scientific disciplines at more or less the same explanatory level; and in these cases, no reduction need be achieved or intended. Rather, conjoining the experimental and theoretical armamentarium of several sciences allows explanations and systematisations of phenomena that none of them is able to handle on its own. This really is a robust tactic of scientific investigation: it’s what spawns the host of ‘hyphenated’ disciplines that have become increasingly familiar, especially in the biological and social sciences – physical anthropology; developmental psycholinguistics; acoustic phonetics; palaeobiology, evolutionary psychology and so on. The point to notice is that when this sort of thing happens, you end up with more sciences than you started with, not fewer: developmental psychology and linguistics and developmental psycholinguistics, as the case might be. The web of causal explanation is extended; but sideways, not up and down.

Ironically, the best and most striking example of the failure of sciences to unify vertically, in spite of heroic efforts and massive expenditure, is exactly the one that Wilson has bet the most on: the reduction of psychology to neurology. To hear Wilson describe it, the real breakthrough in understanding the mind is ‘cognitive neuroscience’, the attempt to model the neurological mechanisms of intelligence. Now, it is possible for people who are sensible and well informed to disagree about how much has been found out about the way the brain implements the cognitive mind. My view of the matter is, I admit, extreme: namely, that the progress has been negligible to nil. Suffice it to say, on the one hand, that the examples Wilson has on offer are pretty thin (‘disturbance of particular circuits of the human brain often produces bizarre results’); and, on the other, that the most important advances in 20th-century psychology are arguably Turing’s proposal for a computational theory of thought and Chomsky’s discovery of the mathematical structure of language. Neither of these emerged from, or was so much as influenced by, neurological investigations. And neither is mentioned by Wilson.

In fact, Wilson appears to have swallowed whole the depressing combination of recidivist Associationism with engineering jargon that is characteristic of much of the recent brain science literature: memory

re-creates not just moving images and sound but meaning [sic] in the form of linked concepts simultaneously experienced. Fire is connected to hot, red, dangerous, cooked, the passion of sex, and the creative act, and on out through multitudinous hypertext pathways selected by context, sometimes building new associations in memory for future recall. The concepts are the nodes or reference points in long-term memory … Recall with images from the long-term banks with little or no linkage is just memory. Recall with linkages, and especially when tinged by the resonance of emotional circuits, is remembrance.

The psychology endorsed here is no advance on Hume or Mill, and the exposition is markedly less sophisticated. The talk about nodes, linkages, long-term banks and resonating circuits is entirely meretricious and adds nothing but the show of connections with the computational sciences (which, by the way, Wilson appears to have misunderstood; the kinds of model that make play with nodes and linkages and such generally deny that there are also ‘long-term banks’ and such). If this is what consilience is like, I recommend the assorted antipasti.

It is actually worse still with Wilson’s book than I have been suggesting. He has ambitions that the Positivists didn’t dream of, and consilience’s writ runs not just to the relation between the special sciences and physics, but also to ‘art, ethics and religion’. I hasten past this part because I don’t understand how it’s supposed to work. It is, after all, entirely possible to doubt that ‘art, ethics and religion’ are primarily in the business of explaining things: not, anyhow, in anything like the way that geology and biology and physics seek to do. In which case, it’s hard to see how the putative unity of scientific explanations could be a model for consilience between science and ‘the humanities’. Wilson himself appears vaguely aware that there may be trouble here, and he’s not very clear about what the programme comes to. ‘Neither science nor the arts can be complete without combining their separate strengths. Science needs the intuition and metaphorical power of the arts, and the arts need the fresh blood of science.’ So characterised, the idea seems a bit generic, and one does wonder what on earth ‘completing’ the arts would be like.

Sometimes what Wilson wants to unify is science and art criticism. Sometimes consilience consists in science illuminating the origin of art (or of ethics). Sometimes the idea is that science will tell us what human nature is, so that ethics, psychology and economics can be rendered ‘compatible’. Sometimes, rather touchingly, Wilson wants consilience to provide an ‘ultimate purpose to intellect … Order, not chaos … beyond the horizon.’ I suspect what Wilson wants from unifying science with the humanities is the prospect of some sort of happy ending. When he’s in this mood, his sense of things is much closer to the Baptism he tells us he was raised in than to the ‘Ionian’ worldview that he tells us he admires. That the news about the human condition might turn out to be simply awful is not a possibility that he appears prepared to contemplate.

Actually, he seems to have a tin ear for this sort of stuff; what he says about art, religion and ethics is often embarrassing. Wilson on Kant: Kant’s ethics

has a comforting feel to it, but it makes no sense at all in terms of either material or imaginable entities, which is why Kant, even apart from his tortured prose, is so hard to understand. Sometimes a concept is baffling not because it is profound but because it’s wrong. It does not accord, we know now, with the evidence of how the brain works.

Wilson on art and edification: ‘Works of art that prove enduring are intensely humanistic.’ Wilson on Goethe: ‘In the philosophers’ empyrean, I imagine Bacon has long since lectured Goethe on the idols of the mind. Newton will have lost patience immediately.’ Still, ‘Goethe can be easily forgiven. After all, he had a noble purpose.’ I’m so glad, because I do rather like some of his poems.

Science is the characteristic product of our culture. Similarly, understanding where science fits in – metaphysically, epistemologically, morally, aesthetically and otherwise – is our characteristic philosophical problem; we’ve been working on it since Descartes. As of now, the hardest part is to reconcile a physicalistic ontology with the apparently ineliminable multiplicity of discourses that we require when we try to say how things are. Wilson thinks this appearance of tension is unreal. He suspects that if we resist consilience, that’s because we’re suffering from pluralism, nihilism, solipsism, relativism, idealism, deconstructionism and other symptoms of the French disease. Well, maybe, but I for one plead not guilty. It seems to me that scientific Realism is quite compatible with the view that events fall into revealing and reliable patterns not just at the level of microstructure but at many different orders of aggregation of matter. The heterogeneity of our discourse would then correspond to the heterogeneity of levels at which the world is organised, and both might well prove irreducible.

Everything is physical perhaps, but surely there are many different kinds of physical things. Some are protons; some are constellations; some are trees or cats; and some are butchers, bakers or candlesticks. For each kind of thing, there are the proprietary generalisations by which it is subsumed, and in terms of which its behaviour is to be explained. For each such generalisation, there is the proprietary vocabulary that is required in order for our discourse to express it. Nothing can happen except what the laws of physics permit, of course; but much goes on that the laws of physics do not talk about. It would not be entirely surprising if the explanatory apparatus that our higher-level theories require in order to say the sorts of thing that physics doesn’t, cross-classifies the taxonomy that physical explanation employs. Maybe this kind of picture is a viable alternative to consilience. Or maybe it’s not. Or maybe both are wrong. Or maybe it’s still too soon to tell.