The Contingency of Language

Richard Rorty

About two hundred years ago, the idea that A truth was made rather than found began to take hold of the imagination of Europe. The French Revolution had shown that the whole vocabulary of social relations, and the whole spectrum of social institutions, could be replaced almost overnight. This inspired a new sort of politics – revolutionary, utopian politics, the sort of political thought which sets aside questions about both the will of God and the nature of man and dreams of creating a new kind of human being. Simultaneously, the Romantic poets were showing what can happen when art is no longer thought of as imitation, but rather as self-creation. These poets made it plausible for art to claim the place in culture traditionally held by religion and philosophy, the place which the Enlightenment had claimed for science.

In the course of the last two centuries, these two tendencies have joined forces and have achieved cultural hegemony. For the contemporary intellectual, questions of ends as opposed to means – questions about how to give a sense to one’s own life or that of one’s community – are questions for art, politics, or both, rather than for religion, philosophy, or science. This development has led to a split within philosophy. Some philosophers have remained faithful to the Enlightenment, and have continued to identify themselves with the cause of science. They see the old struggle between religion and science as continuing, having now taken the form of a struggle between reason and all those forces within culture which take truth to be made rather than found. These philosophers take science as the paradigmatic human activity, and think of science as discovering truth rather than making it. They regard ‘making truth’ as a merely metaphorical, and thoroughly misleading, phrase. They think of politics and art as spheres in which the notion of ‘truth’ is out of place. Other philosophers, realising that the world as it is described by the physical sciences teaches no moral lesson, offers no spiritual comfort, have concluded that science is no more than the handmaiden of technology. These philosophers have ranged themselves alongside the political utopian and the innovative artist. Whereas the first kind of philosopher (a kind common in Britain and America, and exemplified by even such relatively liberated analytic philosophers as Thomas Nagel and Bernard Williams) contrasts ‘hard scientific fact’ with ‘the subjective’ or with ‘metaphor’, the second kind – common elsewhere in the world – sees science as one more human activity, not as the place at which human beings encounter a ‘hard’, non-human reality. On this view, great scientists invent descriptions of the world which are useful for purposes of predicting and controlling what happens, just as poets and political thinkers invent other descriptions of it for other purposes. But there is no sense in which any of these descriptions is an accurate representation of the way the world is in itself. These philosophers regard the very idea of such a representation as pointless.

Had the first sort of philosopher, the sort whose hero is the natural scientist, always been the only sort, we should probably never have had an autonomous discipline called ‘philosophy’ – a discipline as distinct from the sciences as it is from theology or from the arts. As such a discipline, philosophy is no more than two hundred years old. It owes its existence to attempts by the German idealists to put the sciences in their place, and to give a clear sense to the vague idea that human beings make truth rather than finding it. Kant wanted to consign science to the realm of second-rate truth – truth about a phenomenal world. Hegel wanted to think of natural science as a description of spirit not yet fully conscious of its own spiritual nature, and thus to elevate the sort of truth offered by the poet and the political revolutionary to first-rate status.

German idealism, however, was a short-lived and unsatisfactory compromise. For Kant and Hegel went only half-way in their repudiation of the idea that truth is ‘out there’. They were willing to view the world of empirical science as a made world – to see matter as constructed by mind, or as consisting in mind insufficiently conscious of its own mental character. But they persisted in seeing mind, spirit, the depths of the human self, as having an intrinsic nature – one which could be known by a kind of non-empirical super-science called philosophy. This meant that only half of truth – the bottom, scientific half – was made. Higher truth, the truth about mind, the province of philosophy, was still a matter of discovery rather than creation. What was needed, and what the idealists were unable to envisage, was a repudiation of the very idea of anything – mind or matter, self or world – having an intrinsic nature to be expressed or represented. For the idealists confused the idea that nothing has such a nature with the idea that the idea that space and time are unreal – with the idea that human beings cause the spatio-temporal world to exist.

If we are ever again to live in that dawn in which the young Wordsworth exulted, to recapture Schiller’s conviction that art and politics together may jointly shape a new humanity, we need to make a distinction between the claim that the world is out there and the claim that the truth is out there. To say that the world is out there, that it is not our creation, is to say, with common sense, that most things in space and time are the effects of causes which do not include human mental states. To say that truth is not out there is simply to say that where there are no sentences there is no truth, that sentences are elements of human languages, and that human languages are human creations. Truth cannot be out there – cannot exist independently of the human mind – because sentences cannot so exist, or be out there. The world is out there, but descriptions of the world are not. Only descriptions of the world can be true or false. The world on its own – unaided by the describing activities of human beings – cannot.

The suggestion that truth, as well as the world, is out there is the legacy of an age in which the world was seen as the creation of a being who had a language of His own. This suggestion runs together the truth that the world sometimes causes sentences to be true or false with the falsehood that what causes a sentence to be true is, somehow, itself true – that the world splits itself up, on its own initiative, into sentence-shaped chunks called ‘facts’. If one runs these together, it is easy to start capitalising the word ‘truth’ and treating it as something identical either with God or with the world as God’s project. Then one will say, for example, that Truth is great, and will prevail.

This conflation is facilitated by confining attention to single sentences as opposed to vocabularies. For we often let the world decide the competition between alternative sentences (for example, between ‘red wins’ and ‘black wins’ or between ‘the butler did it’ and ‘the doctor did it’). In such cases, it is easy to run together the fact that the world contains the cause of the sentence being true with the idea that that cause, that state of the world, is itself an example of truth. But it is not so easy once we turn from individual sentences to vocabularies as wholes. When we turn to examples of alternative language-games – the vocabulary of ancient Athenian politics as against that of Jefferson, the moral vocabulary of St Paul as against that of Freud, the jargon of Newton versus that of Aristotle, the idiom of Blake rather than that of Dryden – it is difficult to think of the world as making one of these better than another, of the world as deciding between them. When the notion of ‘description of the world’ is moved from the level of criterion-governed sentences within language-games to language-games as wholes, games which we do not choose between by reference to criteria, the idea that the world decides which descriptions are true can no longer be given a clear sense. It becomes hard to think that a vocabulary is somehow already out there in the world, waiting for us to discover it. Attention (of the sort fostered by intellectual historians like Thomas Kuhn and Quentin Skinner) to the vocabularies in which sentences are formulated, rather than to individual sentences, makes us realise, for example, that the fact that Newton’s vocabulary lets us predict the world more easily than Aristotle’s does not mean that the world speaks Newtonian. The world does not speak. Only we do. The world can, once we have programmed ourselves with a language, cause us to hold beliefs. But it cannot propose a language for us to speak. Only other human beings do that.

The realisation that the world does not tell us what language-games to play should not, however, lead us to say that a decision about which to play is arbitrary, nor to say that it is the expression of something deep within us. The moral is not that objective criteria for choice of vocabulary are to be replaced with subjective criteria, reason with will or feeling. It is rather that there are no criteria – that the notions of criteria and choice (including that of ‘arbitrary’ choice) are no longer in point when it comes to changes from one language-game to another. The mind of Europe did not decide to accept the idiom of Romantic poetry, or of socialist politics, or of Galilean mechanics. That sort of shift was no more an act of will than a result of argument. Rather, Europe gradually lost the habit of using certain words and gradually acquired the habit of using others. As Kuhn has argued in his book The Copernican Revolution, we did not decide on the basis of some telescopic observations, or on the basis of anything else, that the Earth was not the centre of the universe, that macroscopic behaviour could be explained on the basis of micro-structural motion, or that prediction and control should be the principal aim of scientific theorising. Rather, after a hundred years of inconclusive muddle, the Europeans found themselves speaking in a way which took these interlocked theses for granted. Cultural change of this magnitude does not result from applying criteria (nor from ‘arbitrary decision’) any more than individuals become theists or atheists, or shift from one spouse or circle of friends to another, as a result either of applying criteria or of actes gratuits. We should not look within ourselves for criteria of decision in such matters any more than we should look to the world.

The temptation to look for criteria is a species of the more general temptation to think of the world, or the human self, as possessing an intrinsic nature, an essence. That is, it is the result of the temptation to privilege some one among the many actual and possible languages in which we habitually describe the world or ourselves. As long as we think that there is some relation called ‘fitting the world’ or ‘expressing the real nature of the human self’ which can be possessed or lacked by vocabularies-as-wholes, we shall continue the traditional philosophical search for a criterion which will tell us which vocabularies have this desirable feature. But if we could ever become reconciled to the idea that reality is indifferent to our descriptions of it, and that the human self is created by the use of a vocabulary rather than being adequately or inadequately expressed in a vocabulary, then we should at last have assimilated what was true in the romantic idea that truth is made rather than found. What is true about this claim is just that languages are made rather than found, and that truth is a property of linguistic entities, of sentences.

I can sum up what I have been saying so far by redescribing what, on my view, the revolutionaries and poets of two centuries ago were getting at. What was glimpsed at the end of the 18th century was that anything could be made to look good or bad, important or unimportant, useful or useless, by being redescribed. What Hegel misdescribed as the process of spirit gradually becoming self-conscious of its intrinsic nature was the fact that European linguistic practices were changing at a faster and faster rate, that more people were offering more radical redescriptions of more things than ever before. What the Romantics expressed as the claim that imagination, rather than reason, was the central human faculty was the realisation that a talent for speaking differently, rather than for arguing well, was the chief instrument of cultural change. What political utopians since the French Revolution have sensed is not that an enduring, substratal, human nature has been suppressed or repressed by ‘unnatural’ or ‘irrational’ social institutions but that changing languages and other social practices can produce human beings with quite new natures, people of a sort who had never before existed. The German idealists, the French Revolutionaries and the Romantic poets had in common a dim sense that human beings whose language changed so that they no longer spoke of themselves as responsible to non-human powers would thereby have invented a new kind of human being.

The difficulty faced by a philosopher who, like myself, is sympathetic to this suggestion, one who thinks of himself as auxiliary to the poet rather than to the physicist, is to avoid hinting that this suggestion gets something right, that my sort of philosophy corresponds to the way things really are. For this talk of correspondence brings back just the idea which my sort of philosopher wants to get rid of, the idea that the world or the self has an intrinsic nature. From our point of view, explaining the success of science, or the desirability of political liberalism, by talk of ‘fitting the world’ or ‘expressing human nature’ is like explaining why opium makes you sleepy by talking about its dormitive power. To say that Freud’s vocabulary gets at the truth about human nature, or Newton’s at the truth about the heavens, is not an explanation of anything. It is just an empty metaphysical compliment which we pay to writers whose novel jargon we have found useful. To say that there is no such thing as intrinsic nature is not to say that the intrinsic nature of reality has turned out, surprisingly enough, to be extrinsic. It is to say that the term ‘intrinsic nature’ is one which it would pay us not to use, an expression which has caused more trouble than it has been worth. To say that we should drop the idea of truth as out there waiting to be discovered is not to say that we have discovered that there is no truth out there. It is to say that our purposes would be best served by ceasing to see truth as a deep matter, as a topic of philosophical interest, or ‘true’ as a term which repays ‘analysis’. It is to say that ‘the nature of truth’ is an unprofitable topic, resembling in this respect ‘the nature of man’ and ‘the nature of God’, and differing from ‘the nature of the positron’ and ‘the nature of Oedipal fixation’. But this claim about relative profitability, in turn, is just the recommendation that we in fact say little about these topics, and see how we get on.

On the view of philosophy which I am offering, philosophers should not be asked for arguments against, for example, the correspondence theory of truth or the idea of intrinsic nature. The trouble with arguments against the use of a familiar and time-honoured vocabulary is that they are expected to be phrased in that very vocabulary. They are expected to show that central elements in that vocabulary are ‘inconsistent in their own terms’ or that they ‘deconstruct themselves’. But that can never be shown. Any argument to the effect that our familiar use of a familiar term is incoherent, or empty, or confused, or vague, or ‘merely metaphorical’, is bound to be inconclusive and question-begging. For such use is, after all, the paradigm of coherent, meaningful, literal speech. Such arguments are always parasitic upon, and abbreviations for, claims that a better vocabulary is available. Interesting philosophy is never an examination of the pros and cons of a thesis, but, implicitly or explicitly, a contest between an entrenched vocabulary which has become a nuisance and a half-formed new vocabulary which vaguely promises great things.

This is to say that the method of philosophy is the same as the method of utopian politics or revolutionary science (as opposed to parliamentary politics or normal science). The method is to redescribe lots and lots of things in new ways, until you have created a pattern of linguistic behaviour which will tempt the rising generation to adopt it, thereby causing them to look for appropriate new forms of non-linguistic behaviour – e.g. the adoption of new scientific equipment or new social institutions. Philosophy, on this model, does not work piece by piece, analysing concept after concept, or testing thesis after thesis. Rather, it works holistically and pragmatically. It says things like ‘try thinking of it this way’ – or, more specifically, ‘try to ignore the apparently futile traditional questions; substitute the following new and possibly more interesting questions.’ It does not pretend to have a better candidate for doing the same old things which we did when we spoke in the old way. Rather, it suggests that we might want to stop doing those things and do something else. But it does not argue for this suggestion on the basis of antecedent criteria common to the old and the new language-games. For just insofar as the new language really is new, there will be no such criteria.

Conforming to my own precepts, I am not going to offer arguments against the vocabulary I want to replace. Instead, I am going to try to make the vocabulary I favour look attractive by showing how it may be used to describe a variety of topics. More specifically, I shall be describing the work of Donald Davidson in philosophy of language, and of Nietzsche, Freud and Harold Bloom in moral psychology, as so many manifestations of a willingness to drop the idea of ‘intrinsic nature’, a willingness to face up to the contingency of the language we use. I shall try to show how a recognition of that contingency leads to a recognition of the contingency of conscience, and how such recognition leads to a picture of the history of science, culture and politics as a history of metaphor rather than of discovery. This in turn leads to a sense of the poet rather than the priest, the philosopher or the scientist as the paradigmatic human being.

I begin with the philosophy of language because I want to spell out the consequences of my claim that only sentences can be true, and that human beings make truths by making languages in which to phrase sentences. I shall concentrate on the work of Davidson because he is the philosopher who has done most to explore these consequences. Davidson’s treatment of truth ties in with his treatment of language-learning and of metaphor to form the first systematic treatment of language which breaks completely with the notion of language as something which can be adequate or inadequate to the world or to the self. For Davidson breaks with the notion that language is a medium – a medium either of representation or of expression.

I can explain what I mean by a medium by noting that the traditional picture of the human situation has been one in which human beings are not simply networks of beliefs and desires but rather beings which have those beliefs and desires. The traditional view is that there is a core self which can look at, decide among, use, express itself by, such beliefs and desires. Further, these beliefs and desires are criticisable, not simply by reference to their ability to cohere with one another, but also by reference to something exterior to the network within which they are strands. Beliefs are, on this picture, criticisable because they fail to correspond to reality. Desires are criticisable because they fail to correspond to the essential nature of the human self – because they are ‘irrational’ or ‘unnatural’. So we have a picture of the essential core of the self located on one side of this network of beliefs and desires and reality located on the other side. On this picture, the network is the product of an interaction between the two, alternately expressing the one and representing the other. This is the traditional subject-object picture which idealism tried and failed to replace, and which Nietzsche, Heidegger, Derrida, James, Dewey, Goodman, Putnam, Davidson and others have, in their various ways, tried to replace without falling back into idealism.

Part of this effort of replacement has consisted in an attempt to substitute ‘language’ for ‘mind’ or ‘consciousness’ as the medium within which beliefs and desires occur, the interface between the self and the world. This turn toward language has been thought of as a progressive, naturalising move. It seemed that way because it seemed easier to give a causal account of the emergence of language-using organisms than of the emergence of consciousness out of non-consciousness. But in itself this substitution is ineffective. For if we stick to the picture of language as a medium, a third thing in between the self and the non-human reality with which the self seeks to be in touch, we have made no progress. We are still using a subject-object picture, and we are still stuck with issues about scepticism, idealism and realism. For we are still able to ask questions about language of the same sort we asked about consciousness.

These are such questions as: ‘Does the medium between the self and reality get them together or keep them apart?’ ‘Should we see the medium primarily as a medium of expression – of articulating what lies deep within the self? Or should we see it as primarily a medium of representation – showing the self what lies outside it?’ Idealist theories of knowledge, and Romantic notions of the imagination, can easily be transposed from the jargon of ‘consciousness’ into that of ‘language’. Realistic and moralistic reactions to such theories can be transposed equally easily. So the see-saw battles between romanticism and moralism, and between idealism and realism, will continue as long as one thinks that there is a hope of making sense of the question of whether a given language is ‘adequate’ to a task – either the task of properly expressing the nature of the human species, or that of properly representing the structure of non-human reality.

What we need to get off this see-saw, and what I think Davidson helps us get, is a conception of language which does not view it as a medium either for expression or representation, and which is therefore able to set aside the idea that both the self and reality have intrinsic natures, natures which are out there waiting to be known. Davidson’s view of language is neither reductionist nor expansionist. It does not purport, as analytical philosophers have sometimes purported, to give reductive definitions of semantical notions like ‘truth’ or ‘intentionality’ or ‘reference’. Nor does it attempt to make language into a kind of divinity, something of which human beings are mere emanations. As Derrida has warned us, such an apotheosis of language is merely a transposed version of the idealists’ apotheosis of consciousness.

In avoiding both reductionism and expansionism, Davidson takes his cue from Wittgenstein. Both philosophers treat alternative vocabularies as like alternative tools rather than like bits of a jigsaw puzzle. To treat them as pieces of a puzzle is to assume that all vocabularies are either dispensable, or reducible to other vocabularies, or capable of being united with all other vocabularies in one grand unified super-vocabulary. If we avoid this assumption, we shall not be inclined to ask questions like ‘What is the place of consciousness in a world of molecules?’ or ‘Are colours more mind-dependent than weights?’ or ‘What is the place of value in a world of fact?’ or ‘What is the place of intentionality in a world of causation?’ or ‘What is the relation between the solid table of common sense and the un-solid table of micro-physics?’ or ‘What is the relation of language to thought?’ We should not try to answer such questions, for doing so leads either to the evident failures of reductionism or the short-lived successes of expansionism. We should restrict ourselves to questions like ‘Does our use of these words get in the way of our use of those other words?’ This is a question about whether our use of tools is inefficient, not a question about whether our beliefs are contradictory.

‘Merely philosophical’ questions like Eddington’s question about the two tables are attempts to stir up a factitious theoretical quarrel between vocabularies which have proved capable of peaceful co-existence. The questions I have recited above are all cases in which philosophers see difficulties that nobody else sees. But this is not to say that vocabularies never do get in the way of each other. On the contrary, revolutionary achievements in the arts, in the sciences, and in moral and political thought, typically occur when somebody realises that two or more of our vocabularies are interfering with each other, and proceeds to invent a new vocabulary to replace both. The traditional Aristotelian cosmological vocabulary got in the way of the mathematised vocabulary being developed in the 16th century by students of mechanics. Again, young German theology students of the late 18th century, like Hegel and Hoelderlin, found that the vocabulary in which they worshipped Jesus was getting in the way of the vocabulary in which they worshipped the Greeks. And again, the use of Rossetti-like tropes got in the way of the early Yeats’s use of Blakean tropes.

The gradual trial-and-error creation of a new, third vocabulary – the sort of vocabulary developed by people like Galileo, Hegel or the later Yeats – is not a discovery about how old vocabularies fit together. That is why it cannot be reached by an inferential process – by starting with premises formulated in the old vocabularies. Such creations are not the result of successfully fitting together pieces of a puzzle. They are not discoveries of a reality behind the appearances, of an undistorted view of the whole picture with which to replace myopic views of its parts. The proper analogy is with the invention of new tools to take the place of old tools. To come up with such a vocabulary is more like discarding the lever and the chock because one has envisaged the pulley, or like discarding gesso and tempera because one has now figured out how to size canvas properly.

This Wittgensteinian analogy between vocabularies and tools has one obvious drawback. The craftsman typically knows what job he needs to do before picking or inventing tools with which to do it. By contrast, someone like Galileo, Yeats or Hegel typically is unable to make clear exactly what it is that he wants to do before developing the language in which he succeeds in doing it. His new vocabulary makes possible, for the first time, a formulation of its own purpose. It is a tool for doing something which could not have been envisaged prior to the development of a particular set of descriptions, those which it itself helps to provide. But I shall, for the moment, ignore this disanalogy. I want simply to remark that the contrast between the jigsaw-puzzle and the ‘tool’ models of alternative vocabularies reflects the contrast between, in Nietzsche’s terms, the will to truth and the will to self-overcoming. Both are expressions of the contrast between the attempt to represent or express something that was already there and the attempt to make something that never had been dreamt of before.

Davidson spells out the implications of Wittgenstein’s notion of vocabularies as tools by raising explicit doubts about the assumptions underlying traditional pre-Wittgensteinian accounts of language. These accounts have taken for granted that questions like ‘Is the language we are presently using the “right” language – is it adequate to its task as medium of expression or representation?’ or ‘Is our language a transparent or an opaque medium?’ – that questions like these make sense. Such questions assume that there are relations such as ‘fitting the world’ or ‘being faithful to the true nature of the self’ in which language can stand to non-language. This assumption goes along with the assumption that ‘our language’ – the language we speak now, the vocabulary at the disposal of educated inhabitants of the 20th century – is somehow a unity, a third thing which stands in some determinate relation with two other unities, the self and reality. Both assumptions are natural enough, once we accept the idea that there are non-linguistic things called ‘meanings’ which it is the task of language to express, as well as the idea that there are non-linguistic things called ‘facts’ which it is the task of language to represent. Both ideas enshrine the notion of language as medium.

Davidson’s polemics against the traditional philosophical uses of the terms ‘fact’ and ‘meaning’, and against what he calls ‘the scheme-content model’ of thought and inquiry, are parts of a larger polemic against the idea that there is a fixed task for language to perform, and an entity called ‘language’ or ‘the language’ or ‘our language’ which may or may not be performing this task. Davidson’s doubt that there is such an entity parallels Ryle’s and Dennett’s doubts about whether there is anything called ‘the mind’ or ‘consciousness’. Both sets of doubts are doubts about the utility of the notion of a medium between the self and reality – the sort of medium which realists see as transparent and sceptics as opaque.

In a recent paper, nicely entitled ‘A Nice Derangement of Epitaphs’, Davidson tries to undermine the notion of languages as entities by developing the notion of what he calls ‘a passing theory’ about the noises and inscriptions presently being produced by a fellow human. Think of such a theory as part of a larger ‘passing theory’ about this person’s total behaviour – a set of guesses about what she will do next. Such a theory is ‘passing’ because it must constantly be corrected to allow for mumbles, stumbles, malapropisms, metaphors, tics, seizures, psychotic symptoms, egregious stupidity, strokes of genius, and the like. To make things easier, imagine that I am forming such a theory about the current behaviour of a native of an exotic culture into which I have unexpectedly parachuted. This strange person, who presumably finds me equally strange, will simultaneously be busy forming a theory about my behaviour. If we ever succeed in communicating easily and happily, it will be because her guesses about what I am going to do next, including what noises I am going to make next, and my own expectations about what I will do or say under certain circumstances, come more or less to coincide, and because the converse is also true. She and I are coping with each other as we might cope with mangoes or boa constrictors – we are trying not to be taken by surprise. To say that we come to speak the same language is to say, as Davidson puts it, that ‘we tend to converge on passing theories.’ Davidson’s point is that all ‘two people need, if they are to understand one another through speech, is the ability to converge on passing theories from utterance to utterance’.

What is novel about this account of linguistic communication is that it dispenses with the picture of language as a third thing intervening between self and reality, and of different languages as barriers between persons or cultures. To say that one’s previous language was inappropriate for dealing with some segment of the world (for example, the starry heavens above, or the raging passions within) is just to say that one is now, having learned a new language, able to handle that segment more easily. To say that two communities have trouble getting along because the words they use are so hard to translate into each other is just to say that the linguistic behaviour of inhabitants of the one community may, like the rest of their behaviour, be hard for inhabitants of the other to predict. As Davidson puts it,

we should realise that we have abandoned not only the ordinary notion of a language, but we have erased the boundary between knowing a language and knowing our way around the world generally. For there are no rules for arriving at passing theories that work ... There is no more chance of regularising, or teaching, this process than there is of regularising or teaching the process of creating new theories to cope with new data – for that is what this process involves ... There is no such thing as a language, not if a language is anything like what philosophers, at least, have supposed. There is therefore no such thing to be learned or mastered. We must give up the idea of a clearly defined shared structure which language-users master and then apply to cases ... we should give up the attempt to illuminate how we communicate by appeal to conventions. [italics added]

This line of thought about language is analogous to the Ryle-Wittgenstein-Dennett view that when we use a mentalistic terminology we are simply using an efficient vocabulary – the vocabulary characteristic of what Dennett calls ‘the intentional stance’ – to predict what an organism is likely to do or say under various sets of circumstances. Davidson is a non-reductive behaviourist about language in the same way that Ryle was a non-reductive behaviourist about mind. Neither has any desire to give equivalents in Behaviourese for talk about beliefs or about reference. But both are saying: think of the term ‘mind’ or ‘language’, not as the name of a medium between self and reality, but simply as a flag which signals the desirability of using a certain vocabulary when trying to cope with certain kinds of organism. To say that a given organism – or, for that matter, a given machine – is a language-user is just to say that pairing off the marks and noises it makes with those we make will prove a useful tactic in predicting and controlling its future behaviour.

This Wittgensteinian attitude, developed by Ryle and Dennett for minds and by Davidson for languages, naturalises mind and language by making all questions about the relation of either to the rest of the universe causal questions, as opposed to questions about adequacy of representation or expression. It makes perfectly good sense to ask how we got from the relative mindlessness of the monkey to the fully-fledged mindedness of the human, or from speaking Neanderthal to speaking Post-Modern, if these are construed as straightforward causal questions. In the former case, the answer takes us off into neurology, and thence into evolutionary biology. In the latter case, it takes us into intellectual history, viewed as the history of metaphor. For my purposes, it is the latter which is important, and I shall spend the rest of this article developing an account of intellectual and ethical progress which squares with Davidson’s account of language.

To see the history of language, and thus of the arts, the sciences and the moral sense, as the history of metaphor is to drop the picture of the human mind, or human languages, becoming better and better suited to the purposes for which God or Nature designed them – e.g. ability to express more and more meanings, or to represent more and more facts. The idea that language has a purpose goes once the idea of language as medium goes. A culture which saw intellectual progress in this way would represent the triumph of those tendencies in modern thought which began two hundred years ago, the tendencies common to German idealism, romantic poetry and Utopian politics.

A non-teleological view of intellectual history, including the history of science, does for the theory of culture what the Mendelian, mechanistic account of natural selection did for evolutionary theory. Mendel let us see mind as something which just happened, rather than as something which was the point of the whole process. Davidson lets us think of the history of language, and thus of culture, as Darwin taught us to think of the history of a coral reef. Old metaphors are constantly dying off into literalness, and then serving as a platform and foil for new metaphors. This analogy lets us think of ‘our language’ – and thus of the science and culture of 20th-century Europe – as something that took shape as a result of a great number of sheer contingencies. Our language and our culture are as much a contingency, as much a result of a thousand small mutations finding niches (and a million others finding no niche), as are the orchids and the anthropoids.

To accept this analogy, we must follow Mary Hesse in thinking of scientific revolutions as ‘metaphoric redescriptions’ of nature rather than insights into the intrinsic nature of nature. Further, we must resist the temptation to think that the redescriptions of reality offered by contemporary physical or biological science are somehow closer to ‘the things themselves’, less ‘mind-dependent’, than the redescriptions of history offered by contemporary culture criticism. We need to see the constellations of causal forces which produced talk of DNA or of the Big Bang as of a piece with the causal forces which produced talk of ‘secularisation’ or of ‘late capitalism’. These various constellations are the random factors which have made some things subjects of conversation for us and others not, have made some projects and not others possible and important.

I can develop the contrast between the idea that the history of culture has a telos – the discovery of truth, or the emancipation of humanity – and the Nietzschean and Davidsonian picture by noting that the latter picture is compatible with a bleakly mechanical description of the relation between human beings and the rest of the universe. For genuine novelty can, after all, occur in a world of blind, contingent, mechanical forces. Think of novelty as the sort of thing which happens when a cosmic ray scrambles the atoms in a DNA molecule, thus sending things off in the direction of the orchids or the anthropoids. The orchids, when their time came, were no less novel or marvellous for the sheer contingency of this necessary condition of their existence. Analogously, for all we know, or should care, Aristotle’s metaphorical use of ousia, St Paul’s metaphorical use of agape, and Newton’s metaphorical use of gravitas, were the results of cosmic rays scrambling the fine structure of some crucial neurons in their respective brains. Or, more plausibly, they were the result of some odd episodes in infancy – some obsessional kinks left in these brains by idiosyncratic traumata. It hardly matters how the trick was done. The results were marvellous. There had never been such things before.

This account of intellectual history chimes with Nietzsche’s definition of ‘truth’ as ‘a mobile army of metaphors’. It also chimes with the description I offered earlier of people like Galileo and Hegel and Yeats, people in whose minds new vocabularies developed, thereby equipping them with tools for doing things which could not even have been envisaged before these tools were available. But in order to accept this picture, we need to see the distinction between the literal and the metaphorical in the way Davidson sees it: not as a distinction between two sorts of meaning, nor as a distinction between two sorts of interpretation, but as a distinction between familiar and unfamiliar uses of noises and marks. The literal uses of noises and marks are the uses which we can handle by our old theories about what people will say under various conditions. Their metaphorical use is the sort which makes us get busy developing a new theory.

Davidson puts this point by saying that one should not think of metaphors as having meanings. To have a meaning is to have a place in a language-game. Metaphors, by definition, do not. Davidson denies ‘the thesis that associated with a metaphor is a cognitive content that its author wishes to convey and that the interpreter must grasp if he is to get the message.’ On his view, to toss a metaphor into a conversation is like suddenly breaking off the conversation long enough to make a face, or pull a photograph out of your pocket and display it, or point at a feature of the surroundings, or slap your interlocutor’s face, or kiss him. Tossing a metaphor into a text is like using italics, or illustrations, or odd punctuation or formats. All these are ways of producing effects on your interlocutor or your reader, but not ways of conveying a message.

To none of these is it appropriate to respond with ‘what exactly are you trying to say?’ If one had wanted to say something – if one had wanted to utter a sentence with a meaning – one would presumably have done so. But instead one thought that one’s aim could be better carried out by other means. That one uses familiar words in unfamiliar ways – rather than slaps, kisses, pictures, gestures or grimaces – does not show that what one said must have a meaning. An attempt to state that meaning would be an attempt to find some familiar (that is, literal) use of words – some sentence which already had a place in the language-game – and to claim that one might just as well have that. But the unpara-phrasability of metaphor is just the unsuitability of any such familiar sentence for one’s purpose.

Uttering a sentence without a meaning is, as the positivists rightly said, to utter something which is neither true nor false – something which is not, in Ian Hacking’s terms, a ‘truth-value candidate’. This is because it is a sentence which one cannot confirm or disconfirm, argue for or against. One can only savour it or spit it out. But this is not to say that it may not, in time, become a truth-value candidate. If it is savoured rather than spat out, the sentence may be repeated, caught up, bandied about. Then it will gradually require a habitual use, a familiar place in the language-game. It will thereby have ceased to be a metaphor – or, if you like, it will have become what most sentences of our language are, a dead metaphor. It will be just one more, literally true or literally false, sentence of the language. That is to say that our theories about the linguistic behaviour of our fellows will suffice to let us cope with its utterance in the same unthinking way in which we cope with most of their other utterances.

The Davidsonian claim that metaphors do not have meanings may seem like a typical philosopher’s quibble, but it is not. It is part of an attempt to get us to stop thinking of language as a medium, and this, as I have said, is part of a larger attempt to get rid of the traditional philosophical picture of what it is to be human. The importance of Davidson’s point can perhaps best be seen by contrasting his treatment of metaphor with those of the Platonist and the positivist, on the one hand, and the romantic, on the other. The Platonist and the positivist share a reductionist view of metaphor: they think metaphors are either paraphrasable or useless for the one serious purpose which language has: namely, representing reality. By contrast, the romantic has an expansionist view: he thinks metaphor is strange, mystic, wonderful. Romantics attribute metaphor to a mysterious faculty called ‘the imagination’, a faculty which they suppose to be at the very centre of the self, the deep heart’s core. Whereas the metaphorical looks irrelevant to Platonists and positivists, the literal looks irrelevant to the romantic. For the former think that the point of language is to represent a hidden reality which lies outside it, and the latter thinks its purpose is to express a hidden reality which lies within us.

Positivist history of culture thus sees language as gradually shaping itself around the contours of the physical world. Romantic history of culture sees language as gradually bringing Spirit to self-consciousness. Nietzschean history of culture, and Davidsonian philosophy of language, see language as we now see evolution, as new forms of life constantly killing off old forms – not to accomplish a higher purpose, but blindly. Whereas the positivist sees Galileo as making a discovery, finally coming up with the words which were needed to fit the world properly, words Aristotle missed, the Davidsonian sees him as having hit upon a tool which happened to work better than previous tools. Once we found out what could be done with a Galilean vocabulary, nobody was interesting in doing the things which could be done with an Aristotelian vocabulary.

Similarly, whereas the romantic sees Yeats as having gotten at something which nobody had previously gotten at, expressed something which had long been yearning for expression, the Davidsonian sees him as having hit upon some tools which enabled him to write poems which were not just variations on the poems of his precursors. Once we had Yeats’s later poems in hand, we were less interested in reading Rossetti’s. What goes for revolutionary, strong scientists and poets goes also for strong philosophers – people like Hegel and Davidson, the sort of philosopher who is interested in dissolving inherited problems rather than in solving them. On this view, substituting dialectic for demonstration as the method of philosophy, or substituting a coherence for a correspondence theory of truth, is not a discovery about the nature of a pre-existent entity called ‘philosophy’ or ‘truth’. It is changing the way we talk, and thereby changing what we want to do and what we think we are.

But on a Nietzschean view, one which drops the reality-appearance distinction, to change how we talk is to change what, for our own purposes, we are. To say, with Nietzsche, that God is dead is to say that we serve no higher purposes. The Nietzschean substitution of self-creation for discovery substitutes a picture of the hungry generations treading each other down for a picture of humanity approaching closer and closer to the light. A culture in which Nietzschean metaphors were literalised would be one which took for granted that philosophical problems are as temporary as poetic problems, that there are no problems which bind the generations together into a single natural kind called ‘humanity’. A sense of human history as the history of successive metaphors would let us see the poet, in the general sense of the maker of new words, the shaper of new languages, as the vanguard of the species.

I should like to end by going back to the claim, which has been central to what I have been saying, that the world does not provide us with any criterion of choice between alternative metaphors, that we can only compare languages or metaphors with one another, not with something beyond language called ‘fact’. The only way to argue for this claim is to do what philosophers like Goodman, Putnam and Davidson have done: exhibit the sterility of attempts to give a sense to phrases like ‘the way the world is’ or ‘fitting the facts’. Such efforts can be supplemented by the work of philosophers of science such as Kuhn and Hesse. These philosophers explain why we cannot explain the fact that a Galilean vocabulary enables us to make better predictions than an Aristotelian vocabulary by saying that the book of the world is written in the language of mathematics. Rather than go over this ground again, I should like to go over some ground which has been traversed by intellectual historians like Hans Blumenberg, historians who have tried to trace the similarities and dissimilarities between the Age of Faith and the Age of Reason.

These historians have made the point which I mentioned earlier, the point that the very idea that the world or the self has an intrinsic nature – one which the physicist or the poet may have glimpsed – is a remnant of the idea that the world is a divine creation, the work of someone who had something in mind, who Himself spoke some language in which He described his own project. I think they are right that only if we have some such picture in mind, some picture of the universe as either itself a person or as created by a person, can we make sense of the idea that the world has an ‘intrinsic nature’. For the cash value of that phrase is just that some vocabularies are better representations of the world than others, as opposed to being better tools for dealing with the world for one or another purpose.

To drop the idea of languages as representations, and to be thoroughly Wittgensteinian in our approach to language, would be to de-divinise the world. Only if we do that can we fully accept the argument I offered earlier – the argument that since truth requires sentences, since sentences are products of vocabularies, and since vocabularies are made by human beings, so are truths. For as long as we think that ‘the world’ names something which we ought to respect as well as to cope with, something which is person-like in that it has a preferred description of itself, we shall insist that any philosophical account of truth save the ‘intuition’ that truth is ‘out there’. This intuition amounts to the vague sense that it would be hubris on our part to abandon the traditional language of ‘respect for fact’ and ‘objectivity’ – that it would be risky and blasphemous not to see the scientist (or the philosopher, or the poet, or somebody) as having a priestly function, as putting us in touch with a realm which transcends the human.

On the view I am suggesting, the claim that an ‘adequate’ philosophy must make room for our intuitions is a reactionary slogan, one which begs the question at hand. For it is essential to my view that we have no pre-linguistic consciousness to which language needs to be adequate, no deep sense of how things are which it is the duty of philosophers to spell out in language. Rather, all we have is a disposition to use the language of our ancestors, to worship the corpses of their metaphors. Unless we suffer from what Derrida calls ‘Heideggerian nostalgia’, we shall not think of our ‘intuitions’ as more than platitudes, as more than the habitual use of a certain repertoire of terms, more than old tools which as yet have no modern replacements.

I can crudely sum up the story which historians like Blumenberg tell by saying that once upon a time we felt a need to worship something which lay beyond the visible world. Beginning in the 17th century, we tried to substitute a love of truth for a love of God, treating the world described by science as a quasi-divinity. Beginning at the end of the 18th century, we tried to substitute a love of ourselves for a love of scientific truth, a worship of our own deep spiritual or poetic nature, treated as one more quasi-divinity. The line of thought common to Blumenberg, Nietzsche, Freud and Davidson suggests that we try to get to the point where we no longer worship anything, where we treat nothing as a quasi-divinity, where we treat everything – our language, our conscience, our community – as products of time and chance. To reach this point would be, in Freud’s words, to ‘treat chance as worthy of determining our fate’. In the next issue I shall try to describe how I think Freud, Nietzsche and Bloom do for our conscience what Wittgenstein and Davidson do for our language – namely, exhibit its sheer contingency.