In a recent polemic against Derrida, John Searle said that the present was a sort of ‘golden age of the philosophy of language’. This is certainly true. It is an era of system-building, in which dozens of immensely complex structures are being constructed. The older rhetoric of analytic philosophy, which decried system-building, big fat books (as opposed to thin, stiletto-like journal articles), and the development of philosophical ‘schools’, has been put aside. Nobody now talks about ‘teamwork’ or ‘bite-size problems’. Rather, every few years the problematic of philosophy of language is altered by the intrusion of yet another brilliantly original account of meaning and reference, one which starts off by denying a premise which had previously been assumed to be part of the rules of the game. Philosophy of language nowadays is an area in which a lot of extremely bright people, inspired by the challenge of friendly competition with equals, are busy creating schools – bodies of students prepared to defend the ‘central insight’ of their teacher by marvellously detailed accounts of modal contexts, conditionals, indexicals, and so on. No area of analytic philosophy demands, or gets, more concentrated intelligence. None generates more intellectual excitement.
This golden age is the product of the same sort of challenge which produced previous golden ages in philosophy – for example, the great ‘scholastic’ systems of the 13th and 14th centuries, or the great metaphysical systems of German Idealism. The challenge consists in the suggestion that we may somehow be locked in by our own limitations, unable to crash through to something with which it is very important to get in touch. In the Middle Ages, it was God who seemed out of reach. Human finitude might, or might not, manage to bridge the gap between itself and the infinite. Fabulous ingenuity was devoted to describing this gap, and to presenting or repelling suggestions about how it might be crossed. After Kant, it was the thing-in-itself which might, or might not, be beyond our grasp. Great feats of the metaphysical imagination produced systems explaining, or denying, the concealed identity of subject and object. After Frege, it has sometimes seemed as if the world of space and time were beyond the reach of human language. Kant’s vision of us as locked in with our own, merely phenomenal creations gave place to the suggestion that we may be locked in with our own intentions, unable to make our language-games latch onto the world. In recent years, the question of the relation between sense (what is in the heads of language-users) and reference (the relation which ties words to the world) has dominated analytic philosophy. It has been debated in the same tone of urgency in which the problem of ‘transcending our subjectivity’ used to be discussed.
Searle’s new book is an exceptionally elegant and incisive account of how we manage to, so to speak, be both inside and outside our heads at the same time. Here is his statement of purpose:
discussions like this can tend to degenerate into a fussy scholasticism which conceals the basic ‘metaphysical’ assumptions at issue ... My basic assumption is simply this: causal and other sorts of natural relations to language are only relevant to language and other sorts of Intentionality insofar as they impact on the brain ... Some form of internalism must be right because there isn’t anything else to do the job. The brain is all we have for the purpose of representing the world to ourselves and everything we can use must be inside the brain. Each of our beliefs must be possible for a being who is a brain in a vat because each of us is precisely a brain in a vat; the vat is a skull and the ‘messages’ coming in are coming in by way of impacts on the nervous system.
In Searle’s sense of the term, internalism contrasts with the Putnam-Kripke view that in order to figure out what somebody is referring to, talking about, you have to look outside her and see her in the context of causal relationships to the rest of the universe – relationships which she may know nothing about. Kripke had made Searle a whipping-boy for Frege, and had claimed that Searle’s account of proper names (in his earlier book, Speech Acts) discredited Frege’s idea that ‘sense determines reference’ – roughly, the internalist idea that what you are in touch with outside you depends on what is inside you. In Intentionality, internalism strikes back.
The dialectical situation is complicated, however, by the fact that Searle wants to kill off two quite different sets of opponents with one theory. He wants a general theory of Intentionality (always capitalised in his book), an account of ‘aboutness’ which will work for both mental entities (e.g. beliefs, desires, and intentions to act) and linguistic entities (e.g. sentences and words). His book is not only a defence of internalism in the philosophy of language (and especially of what he calls ‘descriptivism’ about proper names) but a rejoinder to functionalism (and to the very idea of ‘artificial intelligence’) in the philosophy of mind. He wants a single theory to show both why Ryle and Dennett are wrong and why Kripke and his allies are wrong. The unified theory he offers is a very beautiful one, and it is presented with Searle’s familiar, and unmatchable, lucidity and verve. But it is shaped by an idiosyncratic mixture of motives. Intentionality would have been an easier book to grasp and appreciate if Searle had included, as chapters or appendices, his previous attacks on Dennett and others who claim that mentality stretches over both organisms and computers.
There is not much point in asking, ‘Does Searle get Intentionality right?’, for most of us have no particular intuitions about aboutness-as-such, and would not know what to measure a theory of Intentionality against. So it is best to start out by asking, ‘Does Searle have effective criticisms of the people against whom he wants to use this theory?’ and then to ask: ‘What does his new account of intentionality contribute to these criticisms?’ I shall begin by summarising the gist of Searle’s attack on the-mind-as-computer, and asking how his new account of aboutness functions as a weapon.
Since Ryle, the received view of the mind among Anglo-Saxon philosophers has been that Cartesian talk of the-mind-as-contrasted-with-the-body was altogether misleading. To speak of the mind, on this view, is not to speak of an unobservable quasi-person inside ordinary persons, but rather a way of talking about the interaction of ordinary persons with their environment, and especially of what Ryle called ‘qualities of intellect and character’. This Rylean scepticism about mental entities was reinforced by the suggestion (offered first by Putnam and developed in detail by Dennett and others) that we should think of the relation between the mind and the brain as like the relation between a computer described as running through the steps of a program and the same computer described as redistributing electrical charges. The former sort of description is ‘intentional’ in the sense that it ascribes states to the computer which are ‘about’ something. Such descriptions of the computer – those which describe its functions rather than its internal springs and wheels – let us think of the computer as, at a given moment, having beliefs and desires in regard to things beyond itself. We say that the computer is presently calculating the payroll, or forming hypotheses about the occurrence of the enclitic de, or whatever, even though we might cheerfully admit that, under another description, it is simply sending electrical charges around complicated loops.
Dennett’s ‘functionalism’ in philosophy of mind amounts to saying that there cannot be any more of a philosophical problem about the relation between mind and body than there is about the relation between the computer’s program and its constituent hardware. To take an ‘intentional stance’ toward a machine or toward an organism – to describe it as having beliefs, or intentions to act, or as meaning something by its behaviour – is a useful, unproblematic thing to do. So there is no ‘ontological’ difference between our computers and us except that they are made out of hardware and we (as they say around MIT) of ‘wetware’. We can do things they cannot do, but merely because we are more complexly wired and programmed than we have as yet managed to wire and program them.
Searle has argued that computers simply do not have beliefs or intentions: to say they do is to use terms metaphorically, terms which apply literally only to organisms with the proper sort of brain, those which have what he has called ‘intrinsic intentionality’. Whereas functionalist philosophers of mind argue that, if there are any problems about aboutness, they are problems for the philosopher of language who wants to analyse ascriptions of belief and meaningfulness, Searle thinks that ‘the philosophy of language is a branch of the philosophy of mind. The capacity of speech acts to represent objects and states of affairs is an extension of the biologically more fundamental capacities of the mind (or brain) to relate the organism to the world ... ’ He thinks that any view which ascribes beliefs to thermostats is obviously false, whereas his functionalist opponents think that any attempt to make the difference between us and thermostats philosophically interesting must be wrong-headed.
The difficulty in knowing what side to be on in this argument is that nothing much seems to turn on whether ‘the thermostat believes it has gotten too cold in this room’ is metaphorical or literal. Both sides in the argument are anti-Cartesian materialists. Both can say, as Searle does, that ‘mental states are both caused by the operations of the brain and realised in the structure of the brain’ – although the functionalist will add that the same goes for the operation and structure of the thermostat. What is supposed to be at stake in all this?
Searle thinks that what is at stake is whether ‘there really are pains, tickles, itches, beliefs, fears, hopes’ etc. His functionalist opponents deny that this is at issue: for them, an account of beliefs in terms of behavioural dispositions no more shows that mental states do not exist than an account of electrons in terms of their dispositions to cause micro-structural changes would show that unobservable particles do not exist. Sustained debate between Searle and Dennett in the past has done little to break this impasse.
Nor does this book help. For the only element in Searle’s general theory which seems relevant to his quarrel with the functionalists is his claim that ‘the functioning of Intentional states’ has among its preconditions ‘a bedrock of mental capacities’ called ‘the Background’. The Background is ‘a set of non-representational mental capacities which enable all representing to take place’. It consists of unconscious know-how, of ‘skills, stances, pre-intentional assumptions and presuppositions, practices and habits’. Searle thinks that computers do not have such capacities, and that they therefore can’t have Intentionality either. Without the right Background, they can exhibit only pseudo-intentionality (just as base mechanicals can exhibit only pseudo-gentility).
Between the lines of the chapter on ‘The Background’ one can glimpse the argument offered by Searle’s colleague Hubert Dreyfus (in his What computers can’t do). This argument says that the stiff, awkward and blundering performance of computers, when contrasted with the suavity and skill of even the youngest mammals, indicates something more than a difference of complexity. But the general theory of Intentionality which Searle offers is irrelevant to this argument. A functionalist could take over Searle’s account of Intentionality intact, capitalising on his incautious admission that ‘there might be more biologically primitive Intentional states’ which do not require a Background. For the heart of Searle’s theory of Intentionality is not an account of how we can tell genuine intentionality from pseudo-intentionality, nor of why we should bother to do so, but rather an account of how intentional states can pick out particular things in the world rather than just kinds of things. That account bears against Kripke and his followers, but not against Dennett and his. The latter may be quite prepared to buy into Searle’s account of aboutness, while seeing no force in his claim that this account is incompatible with their own philosophical projects.
Why should it be a problem that mental states, or sentences, can pick out particular things? Why should there be what Searle calls a ‘problem of particularity’ for theorists of intentionality? Because, Searle tells us, his predecessors have assumed that what is in the head, the representational, intentional states, must be describable only in terms of universals rather than of particulars – adjectives and common nouns, roughly speaking, rather than pronouns and proper names. They have taken over the Platonic-Aristotelian idea that the mind contemplates essences rather than individuals – that the mind has commerce only with the non-spatio-temporal, leaving it to the body to relate us to spatio-temporal particulars. Searle thinks that this is why Fregean philosophy of language has seemed unable to account for the fact that two people might have ‘qualitatively indistinguishable’ states of mind and yet be thinking about different individuals.
Searle follows Putnam in making this problem vivid by imagining a far-away ‘Twin Earth’. On that planet, somebody just like you is petting a rabbit, just when you, back here, are petting an indistinguishable rabbit against an indistinguishable background. Indistinguishable as you all may be, you are seeing, feeling, having beliefs and desires about, and murmuring sentences about, one rabbit, and your counterpart is doing the same in respect to a different rabbit. What makes the difference in your intentional object, in what you’re believing or talking about, given that you both are in the same ‘state of mind’? That question poses ‘the problem of particularity’.
Searle’s answer is that you are not in the same state of mind, because your visual experience is of the form ‘there is a rabbit and it is causing me to have this visual experience.’ Your double, meanwhile, is having a visual experience of the same form. But her ‘it’ refers to a different rabbit. So when she believes or talks about this rabbit, her state of mind picks out – intends, as it were – a different rabbit than your state of mind. What Searle is suggesting is that we build the notion of ‘the particular thing which is causing me to be in the intentional state I am in’ into the intentional state itself, thus making the state self-referential. If we do so, we shall be able to be internalists. We can continue to assert, pace Kripke and Putnam, that ‘meanings are in the head.’ We can continue to say, with Frege, that sense determines reference, because our sense of what a sense is has now expanded to include a reference to a causal relationship and to a particular spatio-temporal region (viz. the one which contains the terminus of that relationship).
This solution to ‘the problem of particularity’ is ingenious, and Searle follows it up with a series of subtle and imaginative applications of the idea of self-referential mental states to all sorts of problems in the philosophy of language. He gives us an alternative way of handling all the puzzle cases which Kripke and his followers have raised for Fregean theories, and a very elegant and plausible way it is too. His strategy is to take all the features of the world outside the speaker’s head which Kripke insists must be regarded as determining reference and to put facsimiles of them inside the speaker’s head.
But, given that his ‘externalist’ rivals have their own gimmicks for accounting for those puzzle cases, it is as hard to figure out whose side to be on in this area as it was in the area of philosophy of mind. It is no good introspecting and asking: ‘Is Searle’s account of my visual experience (or his account of my use of proper names) faithful to my own experience (or to my own intentions when I use words)?’ It takes years of training in the subject to develop any intuitions about these matters, and your intuitions tend to depend on who taught you philosophy. The trouble is that ‘sense’ or ‘intention’ or ‘visual experience’ or ‘intrinsic mental character’ are like ‘gene’ or ‘quark’: they mean whatever the experts say they are supposed to mean. They are terms of art.
So the real question which Searle’s strategy raises is: is he playing fair? Can he change the terms of the problem, as it was previously thought to exist, by shoving causal relations inside experiences, after two centuries of philosophy (since Hume) have given us a sense of ‘experience’ in which causality is exactly the sort of thing which cannot fit inside experiences? Can Searle, all on his own, redefine ‘mental state’ in such a way as to give him what he needs to reply to Kripke? Granted that Searle is ready with an intriguing new (anti-Humean) view about causality to fit the occasion, and more generally that he is prepared to construct a complete philosophical system in order to have defensive breastworks against Kripkean rebuttal, how is one supposed to award the palm in this kind of free-style debate? Is Searle trivialising the issue, or is he (like Einstein changing the definition of ‘simultaneity’) pulling off a masterstroke?
I shall evade these large questions by raising a still larger one. I said at the outset that the current golden age of the philosophy of language was inspired by the sense that we may be trapped inside our language-games, cut off by some hard-to-define metaphysical hiatus from the world we should like to be talking about. This sense (like the definition of ‘in the mind’, or ‘given to consciousness’, as excluding causality which is common to Hume, Kant and pre-Searle Fregeans) is the latest manifestation of a dread of being on the wrong side of a gap, a dread that runs through Western philosophy. At the highest level of generality, or in the longest historical perspective, the question of whether sense determines reference, or whether meanings are in the head, is a question about whether human beings are somehow locked off from what they think they need, and about whether they really do need this farther thing. Should we, in thinking about Searle and his various rivals, start from this abstract and fuzzy question, or should we get down to nitty-gritty questions about whether Searle can get away with making intentional states intrinsically indexical?
Presumably we should go back and forth between these levels. If we stay too long at the bottom level we run the danger of our scholasticism becoming merely, in Searle’s phrase, ‘fussy’ (instead of its being, as it often is in the present golden age, bold and exciting). If we stick to the topmost level, we run the risk of falling in with French talk about ‘overcoming the hegemony of binary oppositions’: at this level, there is little to do except bat philosophical slogans back and forth across the Channel. But somewhere in between – the level at which, for example, Derrida offers us detailed analogies between current issues in philosophy of language and older issues in ‘onto-theology’ – might be a good place to begin. Given that philosophy is an area in which the experts begin by spotting gaps invisible to the laity, how seriously must we take these experts’ claim that the gaps were there beforehand?
Soner or later we are going to have to ask about contemporary systems of philosophy of language the sort of questions which eventually got asked about the great systems of German Idealism: granted that some of the best minds of the epoch took gap-filling, gap-crossing and gap-widening as their theme, was there a pay-off? Did any gap get crossed? And was the gap real or apparent in the first place? Do we have to carve things up into the subject and the object (or into language and world, intentional states and their objects, what is in the head and what is not) in the first place? Are Kripke-style causal theories of reference and Searle-style theories of intentional causation alternative theories about how nature works, or are they alternative patchwork repairs of a badly-functioning and obsolescent contrivance?