Imagining an orgasm

Colin McGinn

  • Mind and Cognition: A Reader edited by William Lycan
    Blackwell, 683 pp, £14.95, April 1990, ISBN 0 631 16763 3
  • Acts of Meaning by Jerome Bruner
    Harvard, 179 pp, £15.95, December 1990, ISBN 0 674 00360 8
  • Modelling the mind edited by K.A. Mohyeldin Said
    Oxford, 216 pp, £25.00, August 1990, ISBN 0 19 824973 X

The more philosophically interesting a science, the less secure or transparent are apt to be its theoretical foundations, given that philosophy thrives on perplexity. It is some time since chemistry produced much of a reaction in philosophers, but biology can still get their juices flowing, though not so freely as in the days of the Bergsonian élan vital. Quantum physics is a contemporary focus of philosophical attention – despite the suspicion of some that it is only a dispensable anti-realism that generates the putative puzzles. Mathematics induces periodic bouts of fascination, even of deep distrust – as with Brouwer and Wittgenstein – but its rigour and finality tend to keep the perplexities at bay. In the case of psychology, however, philosophical interest reaches its highest pitch, and never more so than at present: perhaps to the chagrin of practising psychologists, philosophers are now very interested in what they are doing – or at any rate in what they ought to be doing.

The reason for this intense scrutiny can be summed up in two words: ‘meaning’ and ‘consciousness’. These crop up with increasing frequency in psychological writings, after a long period during which they were anathematised. And the topics they refer to have never ceased to occupy philosophers; indeed, the theory of meaning might justly be regarded as the central concern of 20th-century philosophy. Together, the two concepts are definitive of what we ordinarily mean by ‘mind’. It is largely because psychology is turning again to these constitutive marks of mentality that philosophers have once more become intrigued by that science. They found little to grip them while psychology perversely defined itself as the ‘science of behaviour’, entirely eliminating the notions of meaning and consciousness from its purview. Put differently, now that scientific psychology is acknowledging its continuity with common-sense or folk psychology, in which philosophers have maintained a steady interest, psychological theories contain concepts that provoke difficult philosophical questions. There is no shame for the scientists in this: it was misguided to defenestrate the mind just because the concepts that characterise it are philosophically rich and demanding. On the contrary, it is good to see one of the more philistine legacies of anti-philosophical positivism finally melting from the scene.

I do not mean to imply that the notions of meaning and consciousness are in good odour with all philosophers of psychology: they are certainly not. But the philosophical issues that surround these notions are now part of what a reflective psychologist needs to be sensitive to: they can no longer be left to those reactionary old philosophers. For these issues determine the shape and content of empirical theories. A central question here is whether theories that make serious use of these notions can be properly ‘scientific’ – whether, that is, their employment calls for a distinctive methodology. Specifically, can the study of meaning and consciousness conform to the theoretical paradigm set by the natural physical sciences? The physical sciences deal with quite different sorts of phenomenon, at least on the face of it: does this mean that a psychology so conceived cannot take the form assumed by physical theories – with their laws, causes mechanisms? What happens to the structure of psychological theories, and the empirical procedures that lead to them, when you make psychology go consciously semantical? How does a psychology of belief and desire compare to a physics of gravity and electric charge?

There are basically three schools of thought on this issue, with much variation within them. One school, which we may call the nomothetic realists, holds that (ideal) psychology consists of an explanatory set of content-involving causal laws: psychology is just one more special science, but one in which intentional properties are the domain of interest, as factual and nomic as geology or biology. General statements like, ‘if an agent desires that p and believes that making it the case that q is a good way to bring it about that p, then that agent will, ceteris paribus, bring it about that q,’ are thus comparable in status to causal laws like ‘Free-falling bodies accelerate to earth at a rate of 32 feet per second squared.’

This is the school of which Fodor is the most forthright and formidable member: on his view, psychological attributions are made true by real internal sententially-structured states of the subject which stand in certain kinds of reference-creating nomic relations to environmental contingencies – in a word, by a meaning-endowed language of thought. According to nomothetic realism, there is no good reason why invoking meaning should disqualify psychology from taking its place among the other natural sciences. Fodor is less robustly sanguine about consciousness, however: he tends not to mention it at all.

In opposition to the Fodorean school, stand eliminativists and instrumentalists of varying degrees of boldness. The Churchlands, following Feyerabend and Quine, assert that folk psychology is a discredited proto-theory, and that the language of neuroscience states the only psychological facts worthy of the name: there simply are no beliefs and desires, no meanings, no states of consciousness – period. Somewhat less drastically, Stich has suggested that psychosemantics in the style of Fodor should give way to unadulterated psychosyntax: cognitive science should restrict its theoretical concepts to purely formal or structural features of the internal code, regarding semantic interpretation as so much pointless myth-making. One reason for this recommendation is the difficulty of seeing how referential properties of internal symbols could exert a causal hold over the subject’s behaviour, once it is granted that such properties are not super venient upon the total internal physical state of the subject: if meanings are not ‘in the head’, then they are not where the causes of bodily movements are located. An ostensibly weaker position undertakes to define a type of meaning that does not diverge from the causal taxonomy determined by the inner syntax, so that folk psychology comes out as approximately fifty percent true. In any case, there is to be no room for the ordinary notion of meaning in theories of cognition.

Dennett, for his part, rejects the psychosyntactic story about intentional states, with or without a semantic component, preferring to construe our ascriptions of belief and desire in an instrumentalist spirit: folk psychology is merely a ‘stance’ we adopt towards the behaviour of people, animals and computers – a useful scheme that enables us to predict what they will do. It isn’t an attempt to depict an inner landscape of the mind. His most recent thesis, indeed, is that human intentionality is just as derivative as the intentionality we attribute to our artifacts: the nearest thing to original intentionality in the world is exemplified by the blind processes of natural selection. The about-ness of our conscious beliefs is thus merely ‘as-if’: we can retain such talk as a convenient heuristic, but we should not credit it with more factual solidity than this.

The third school might be called ‘interpretationalism’: talk of meaning and content is literally true, and truly literal, but it is not talk that lends itself to canonical law-like formulation. The concepts of belief and desire are agreed to be causal concepts, and the denoted states taken to be as real as reality gets, but they are concepts that belong to a kind of understanding that differs from that typical of the natural sciences: their business is Verstehen, not the expression of nomic regularities. One mark of this specialness is their essential involvement with normative notions – consistency, consequence, good reason – from which it follows that there can be no reduction of folk psychology to anything purged of the normative. This type of view is championed by Davidson and those influenced by him, as well as (latterly) by Putnam. If psychology is to immerse itself in the intentional, then it must expect to sacrifice the kind of scientific rigour for which its exponents have hankered. It must join the intellectual B-stream of history, anthropology, literary criticism, corner gossip. Each of these schools, which I have so crudely summarised, has its representatives in the volumes under review. Mind and Cognition, in particular, offers a broad sample of philosophical opinion on these matters, though it seems biased towards the more ‘hardnosed’ end of the spectrum.

What should become clear to theorists of cognition once the issue of meaning is explicitly raised is that the computer model of mind is a good deal less straightforward than it might have seemed. I suspect that the enthusiasm of cognitive scientists for the computational conception of human cognition was nourished by a certain unclarity about what precisely a computer program is. For, as Bruner observes in Acts of Meaning, the ‘cognitive revolution’ in psychology was welcomed precisely became it was not properly understood: it was made to seem like less of a departure than it really was. With the idea of computer program as their inspiration and paradigm, psychologists felt that they could speak of human ‘information-processing’ without fear of departing from standards of scientific purity. But the picture becomes substantially less reassuring when the idea of a program is scrutinised more carefully. As Searle never tires of reminding us, a program is a list of purely syntactic or formal instructions; in itself it contains no semantic interpretation for the symbols manipulated. We cannot therefore expect that program rules will explain to us what meaning is, thus rendering semantics scientifically reputable. Indeed, it seems highly plausible that such meaning as computer codes have is conferred by their operatives, and is therefore presupposed rather than explained. Once we enquire what could determine the semantics of a machine code independently of human interpretation, all the usual perplexities about meaning begin to surface. The idea that a program is both well-understood in virtue of its purely formal character and at the same time accurately simulates human thinking is thus illusory: it will simulate thought only if it carries genuine intentionality (pace Dennett) – but then the program cannot be defined purely formally. So the computer model doesn’t answer the deep problems about meaning raised by the cognitive turn in psychology: it simply presupposes an answer to them, or else passes them by. In other words, you can’t expect AI programming competence to let you off doing the messy philosophical work – not if you are serious about a semantics-based psychology.

I have said little so tar about consciousness. That is because even philosophers find this one a bit of a hot potato. Whereas it is possible to say a lot of quite interesting things about meaning, it is hard to say more than a little about consciousness (beyond the merely rhetorical) – and most of this is a touch too interesting. On the one hand, there are those who insist, startlingly enough, that, appearances notwithstanding, states of consciousness reduce without residue to neural states or physical causal roles of such states. On the other hand, there are those who stoutly declare it as self-evident that no amount of physical information about the brain could ever imply the possession of a state of consciousness, so that conscious experience falls radically outside the domain of physical science. This dispute has recently centred on the question of what one’s own experiences teach one about consciousness that could not be taught otherwise: Mind and Cognition contains a useful section on this. Nagel and Jackson hold that there are real features of experience that only direct acquaintance with it can reveal: these features cannot then be comprised in physical information about the experiencer, which can be taught discursively. Nemirow and Lewis, on the other hand, contend that undergoing an experience confers only in ability to imagine experiences; it does not reveal special non-physical properties of experience that are accessible only by acquaintance. The question dividing these disputants is whether what is referred to by one side as an irreducible subjective state can be exhaustively explained by the other in terms of an ability to imagine. As a student of mine remarked, the latter thinkers hold, in effect, that the feeling of an orgasm is equivalent to imagining an orgasm – an equation that she felt (perceptively) not to be very plausible.

As to the place of consciousness in theoretical psychology, it is almost a reflex among psychologists to cry ‘Epiphenomenalism!’ and reach for Occam“s razor. However, it is far from clear that conscious events suffer causal inertness in any sense beyond that true of events described in any of the special sciences – biology, geology, even chemistry. From the explanatory universality of basic physics we cannot infer that other modes of explanation fail to capture causally significant patterns in nature: the hierarchical arrangement of the sciences should not be confused with epiphenomenalism about all but the bottom layer. What is wanted here is not a priori dismissal but a serious investigation into the properties and processes of consciousness; its developmental history, both phylo- and onto genetic; its contribution to our modes of cognitive processing; the nature of its pathologies (e.g. blindsight). Psychologists should shed their dated philosophical hang-ups about consciousness, as in fact they are now beginning to do, and apply themselves to carrying out some empirical work on what their forebears regarded as taboo. If nothing empirically worthwhile turns up, that will be the time to abjure interest in the topic and leave it to the philosophers to puzzle over.

A question seldom raised in these discussions is whether meaning and consciousness are susceptible of deep investigation by human knowers. (Admittedly, I discuss it myself at some length in The Problem of Consciousness.) A properly general naturalism should leave open the possibility that human cognition is not designed in such a way that we can gain any real scientific insight into the underlying workings of our own minds (or those of other animals). Certainly, it is painfully plain that we have not achieved in this area anything like the theoretical depth we have attained in understanding the physical world: there appears to be a systematic elusiveness about the ultimate science of mental phenomena. Despite our fairly advanced understanding of brain function, for example, we seem no nearer than Descartes was to explaining how conscious states result from neural excitations. As an ancillary investigation to the science of meaning and consciousness, then, we should also try to develop a higher-order science of our capacity to understand these phenomena – a science of our ability to arrive at psychological knowledge. It may be the case, not that meaning and consciousness are in themselves suspect or mythical since our science cannot plumb their depths, but rather that our science, as a natural product of human cognitive capacities, has the wrong kind of structure to take in all that the world objectively contains. Transcendent nomothetic realism may in the end be the truth of the matter. The philosophical interest of mental notions might thus be an artifact of the human inaccessibility of the ultimately correct mental science. Psychology might be philosophically boring after all, objectively speaking, if only we could come to know its deep principles. But given the limiting parameters of human cognition, it is possible that the science of mind is condemned to perpetual philosophical interest.