When these lectures were first published eight years ago (in a collection), they stood analytic philosophy on its ear. Everybody was either furious, or exhilarated, or thoroughly perplexed. No one was indifferent. This welcome republication in a separate volume (with a helpful new preface, but no substantive changes) provides a chance to look back at a modern classic, and to say something about why it was found so shocking and liberating.
Since Kant, philosophers have prided themselves on transcending the ‘naive realism’ of Aristotle and of common sense. On this naive view, there is a right way of describing things, corresponding to how they are in themselves, to their real essences. Scientists, philosophers like to say, are especially prone to adopt this unreflective view. They think they are discovering the secrets of nature, but philosophers know that they are really constituting objects by synthesising the manifold of intuition, or predicting the occurrence of sensations, or wielding instruments to cope with the flux of experience, or something equally pragmatic and anthropocentric. This condescending attitude towards common sense, Aristotle and science has been shared by people as far apart as Russell and Bergson, Whitehead and Husserl, James and Nietzsche, Carnap and Cassirer.
Until Kripke came along, almost the only exceptions to this consensus were the Catholics and the Marxists. Between the two Vatican Councils, neo-Thomists tried to explain that the ‘naive’ Aristotelian view was the sound intuitive belief of the common man, and that Cartesian subjectivism, Kantian transcendental idealism and positivistic empiricism were successively more virulent forms of a mad modern heresy. But nobody listened, and after the aggiornamento the neo-Thomists pretty well gave up. Old-time Marxists, who had cut their teeth on Lenin’s Materialism and Empirio-Criticism, used to argue that Russell was just the latest English version of the ‘bourgeois formalism’ which Hegel had diagnosed in Kant. But nobody listened to them either, and after the discovery of the young, humanist, pragmatist Marx they, too, gave up. Just when it seemed that the dialectic which Kant began had culminated in universal acceptance of the relaxed pragmatism of Wittgenstein and Quine, Kripke exploded his bomb.
In a hundred pages of sinewy colloquial prose, Kripke offered a realistic, anti-Kantian, anti-pragmatist way of treating the concepts of ‘meaning’, ‘reference’ and ‘truth’ which Frege and Russell had treated in a Kantian way. The Kantian picture is that we decide what counts as an ‘object’ by putting ideas together. We build a world inside our minds by tying concepts together so as to package sensations more conveniently. Frege and Russell updated this picture by handing the mind over, with a shrug, to empirical psychology and putting the point in terms of language – talking about words instead of ideas. In Russell’s view, names are shorthand for descriptions – lists of the qualities which we have decided to use to identify occasions on which we shall use a name. Russell took his cue from Frege’s distinction between a term’s meaning and its reference, a distinction between such a list of qualities and the portions of the world to which the term is correctly applied. Meaning, Frege claimed, determines reference. This claim is uncontroversial if you think of the world as an undifferentiated manifold waiting for us to structure it. But, as Marxists and neo-Thomists insist, such a view smacks of idealism. It leads fairly quickly to the pragmatist view that science, and human inquiry generally, makes truth rather than finds it – that we did not discover sub-atomic particles, but rather discovered that it was helpful to package the flux under such labels as ‘positron’.
Philosophers raised on Frege and Russell are so habituated to this Kantian way of thinking that the very idea that some properties of a thing are necessary ones – properties a thing could not lack while remaining the same thing – has seemed merely a Gothic curiosity, the last enchantment of the Middle Ages. It has seemed clear that if you describe Aristotle as ‘the author of the Metaphysics’, then it is essential to him that he was literate and accidental that he was male, whereas if you describe him as the father of Nicomachus the converse will be true. Essentiality and necessity had seemed obviously relative to one’s choice of description, rather than a matter of what some particular hunk of space-time had to be like. There is something giddy and intoxicating about this Kantian freedom to redescribe, and thus to create new essences, new necessities, new structures. Philosophers of language in the Frege-Russell-Quine tradition share with Fichte the thought that we make worlds as we make poems and pictures – a thought which finds full expression in Nelson Goodman’s recent Ways of Worldmaking.
Kripke tries to sober us up by denying that meaning determines reference. Rather, we name things by confronting them and baptising them, not by creating them out of a list of qualities. Names are not, pace Russell, shorthand for such lists. They are not abbreviations for descriptions, but (in Kripke’s coinage) ‘rigid designators’ – that is, they would name the same things in any possible world, including worlds in which their bearers did not have the properties we, in this world, use to identify them. The non-philosopher can hardly imagine how shocking these Kripkean claims sounded. Analytic philosophers had been brought up to believe that semantics was constituted as ‘first philosophy’ (and had thus become the glory of our century) precisely by accepting Frege’s and Russell’s way of looking at things – just as metaphysics had once been constituted as ‘first philosophy’ by accepting Aristotle’s (and had thus become the putative glory of the 13th century). To have the brightest young philosopher of language denounce not merely the details of Russell’s Theory of Descriptions but the Kantian world-view which that theory articulated was like having a newly-elected Pope denounce the Incarnation. The whole idea of what it was to be an analytic philosopher, what it was to be sophisticated about the relation of thought to the world, began to totter. For a moment, nobody could quite believe that a leading modal logician should seriously commend the Aristotelian way of looking at things. Perhaps it was merely affected Gothicising?
Kripke was, however, entirely serious. Naming and Necessity lays out a way of thinking about the relation between language and the world which permits just as formal and rigorous a treatment of notions like ‘meaning’, ‘truth’ and ‘reference’ as had Russell’s and Frege’s. Nobody would have believed that the neatness – what Kripke calls ‘the marvellous internal coherence’ – of Frege-Russell semantics could be duplicated after everything was turned upside down. But Kripke showed how to do it, and now philosophers are busily rewriting all of semantics (and a good deal of epistemology) in Kripkean terms. The basic idea which Kripke develops is that we pick out the bearers for proper names like ‘Aristotle’ and ‘Brezhnev’, or common names like ‘gold’ or ‘positron’, by saying ‘We’ll call that “X”,’ rather than by saying: ‘We’ll call something “X” if it meets the following criteria ... ’ The coarseness and brutality of this sort of semantics can only be appreciated by those who have lived in lotus-land, exulting in their Kantian-Goodmanian freedom to create worlds. Kripke showed that one could be coarse about how language works while being delicate and polished in the theory that spells out the details.
It would be impossible, in this space, to exhibit just how Naming and Necessity did this. Instead, it might be well to ask: now that we know that Aristotle as well as Kant can be successfully updated, how are we supposed to decide between them? How do we choose between the attractions of the Gothic and the modern, between naive realism and quasi-idealism? Sometimes Kripke suggests that there are simple arguments which show he is right – as when he urges that no Russell-like account can pick out a unique bearer for a proper name. But his opponents could easily give up Russell’s claim that uniqueness is required, and urge that a Russellian account in terms of descriptions need merely pick out a (possibly non-unique) referent which can be agreed on for present conversational purposes. This pragmatical cop-out would leave the argument up in the air. Kripke’s usual strategy, however, is not to look for knockdown arguments. Instead, he appeals to the greater ‘intuitive’ appeal of his view. He asks his reader to cleave to untutored intuition and resist the false sophistication of the schools.
It has even been suggested in the literature, that though a notion of necessity may have some sort of intuition behind it ... this notion [of a distinction between necessary and contingent properties] is just a doctrine made up by some bad philosopher, who (I guess) didn’t realise that there are several ways of referring to the same thing ... It is very far from being true that this idea [that a property can meaningfully be held to be essential or accidental to an object independently of its description] is a notion which has no intuitive content, which means nothing to the ordinary man. Suppose that someone said, pointing to Nixon, ‘That’s the guy who might have lost.’ Someone else says ‘Oh no, if you describe him as the winner, then it’s not true that he might have lost.’ Now which one is being the philosopher here, the unintuitive man? It seems to me obviously to be the second. The second man has a philosophical theory. The first man would say, and with great conviction, ‘Well, of course, the winner of the election might have been someone else ... So, such terms as “the winner” or “the loser” don’t designate the same objects in all possible worlds. On the other hand, the term “Nixon” is just the name of this man.’
But do ordinary men really have views about what they would call by what name in counterfactual situations – situations in which the various qualities are shifted about the various space-time points in all sorts of funny ways? Do they have ‘a direct intuition of the rigidity of names, exhibited in our understanding of the truth conditions of particular sentences’ (page 14)? Suppose ‘Aristotle’ was the name of an unreflective scribe who copied out the various treatises making up what we call ‘the Aristotelian corpus’, each treatise having been composed by a different member of a committee, only one of whom was also named ‘Aristotle’. What, intuitive reader, are the truth-conditions of ‘Aristotle was less religious than Plato’? Philosophers of language have to supply truth-conditions in such puzzle-cases; it is their job. But it is not clear that the man in the street is going to be of much help to either side in the controversy.
Still, even if we have no intuitions about truth-conditions in weird counterfactual situations, maybe we do have intuitions about realism v. idealism, Aristotle v. Kant? Or is this a matter for the cultivated taste of those who have savoured both? Is the success of Kant among practically everybody from Schiller to Goodman a matter of his seductive appeal to our more irresponsible impulses? Or does Kant have some sturdy intuitions on his side too? If semantics really is (as Michael Dummett has claimed) ‘first philosophy’, should we try desperately to make the right decision between Russell and Kripke, so as to know whether it will be necessary to reconstruct the rest of philosophy and culture? Modern thought on everything from politics to literature to religion is, after all, shot through with Kantian assumptions. Maybe a thorough house-cleaning is in order? Can we just shrug the quarrel between Kripke and the Frege-Russell-Quine tradition off as a ‘technical’ issue, or is there more to it?
On the narrow ‘technical’ ground of how to explicate concepts like ‘meaning’ and.‘reference’, the Russell-Kripke issue is probably a stand-off. One can play it either way, and develop a system from either starting-point with equal completeness and elegance. In either case, the budget of paradoxes will be about equally long, though much will depend upon what one has been brought up to find paradoxical. On the slightly larger ground of reflection on the nature of science, and of inquiry generally, it is hard to see that the new possibilities within semantics change the traditional stand-off between the pragmatist who says, ‘Talking about positrons gets us what we want,’ and the realist who says: ‘There really are positrons.’ The same goes for the more sophisticated stand-off between the pragmatist claim that ‘are really there’ just means ‘talk about it gets us what we want’ and those who think that there is more to reality than that. On still larger questions, it is very doubtful indeed that the Kantian ideas which are taken for granted in our culture are going to be refuted by anything that philosophy professors do. The senses in which philosophy is ‘queen of the sciences’ and in which semantics is ‘first philosophy’ are not strong enough to permit any straight forward deductions from new semantical theories to consequences for culture as a whole.
Nor, of course, need Kripke claim that such deductions are possible. It is enough for his purposes in this book to have pointed semantics in a new direction. Still, the connection between traditional Russell-Frege semantics and the larger world-picture it adumbrates leads one to reflect on the relation of developments in ‘technical’ philosophy to philosophy’s larger role. Such reflection suggests that Kripke may have demoted philosophy of language in the course of revolutionising it. As long as we took for granted Kant’s notion that we structure the world by representing it, the study of the nature of representation (of Mind in the 19th century, of Language in this century) took pride of place. For in studying the activity of representation philosophy takes itself to discover ‘formal’, ‘conceptual’, ‘structural’ truths – truths higher and purer than those produced by science. If we lose our grip on the Kantian picture, this structure-content distinction begins to evaporate. So does the notion of philosophy as the armchair study of the nature of representation. So, a fortiori, does semantics as the study of how language relates to the world. By rediscovering (or, as disgruntled Kantians would say, by reinventing) ‘metaphysical necessity’, as opposed to Kant-Russell description-relative ‘epistemic necessity’, Kripke has revoked the charter of modern philosophy.
Further, by showing that semantics need not be committed to Kant – that one can have a philosophy of language which plays it the other way – he has made it easier to dismiss the question of which way to play it as ‘merely technical’. This attitude is encouraged if one reads the attempts in the philosophical journals to resolve conundrums such as the one about Aristotle set out above – attempts which make irresistible the epithet ‘decadent scholasticism’. A similar insouciance is encouraged by neo-Quinean philosophers of language like Donald Davidson, who object to ‘building block’ approaches of either the Russellian or the Kripkean sort. Davidsonian ‘holistic’ semantics, which may emerge as tertius gaudens, presents itself as an empirical theory of linguistic behaviour, not as a successor subject to Kantian epistemology. It eschews theories of how language either structures or copies the world. Given all these ways in which the old consensus is breaking up, ‘philosophy of language’ is losing the ideological overtones it once shared with ‘linguistic philosophy’. It may become the sort of academic speciality which neither wants nor needs links with, or readers in, the larger world.
Perhaps, however, such speculations are premature. At the moment, everybody is waiting for Kripke to drop the other shoe, to till us in on his new way of seeing things, beyond the few tantalising details revealed in Naming and Necessity. That work resembles Wittgenstein’s Tractatus in hinting at a larger context and higher motives. Since Kripke changed the tone of analytic philosophy with ideas formulated before his 25th year (see page 3), he may well have surprises in store.