Ever since Aristotle declared that philosophy is “the first and last of the sciences,” philosophers have tended to take a very exalted view of the importance of their subject. They will find it much harder to do so after reading Professor Richard Rorty’s disturbing and brilliantly argued book. He opens his attack on the traditional pretensions of philosophers by considering the reasons for their longstanding confidence. The explanation for this confidence is said to lie in their continuing acceptance of the seventeenth-century idea that a philosopher is someone who knows “something about knowing which nobody else knows so well.” This image is in turn said to owe its plausibility to the work of Descartes, who introduced the key concept of the mind as a species of inner space, a consciousness in search of indubitable knowledge about the external world. He thereby suggested the central task of modern philosophy: the attempt to determine, by analyzing the concept of mind itself, what forms of knowledge are susceptible to being acquired with certainty.

As a result, his successors readily came to see themselves as exponents of the pivotal cultural discipline. Since culture is the assemblage of claims to knowledge, and since philosophy is pictured as the arbiter of such claims, the philosopher appears as a kind of “cultural overseer” with the job of “keeping the other disciplines honest, limiting their claims to what can be properly ‘grounded.”‘

This image of philosophy, Rorty roundly asserts, is not only absurd but has already been decisively overthrown by “the three most important philosophers of our century—Wittgenstein, Heidegger, and Dewey.” As Rorty puts it,

These writers have kept alive the suggestion that, even when we have justified true belief about everything we want to know, we may have no more than conformity to the norms of the day. They have kept alive the historicist sense that this century’s “superstition” was the last century’s triumph of reason, as well as the relativist sense that the latest vocabulary, borrowed from the latest scientific achievement, may not express privileged representations of essences, but be just another of the potential infinity of vocabularies in which the world can be described.

These are Rorty’s heroes, and in his own work he is largely content to present himself—though with excessive modesty—as a historian of the movement toward the “deconstruction” of philosophy which, he maintains, they successfully initiated.

The singling out and yoking together of this improbable troika is of course intended as a shock tactic, but it produces the one structural weakness of Rorty’s book. No interpretations are put forward in justification of the claim that these are the leading philosophers of the age, and the fact that Bertrand Russell’s name is not mentioned leaves Rorty open to the suspicion that he has covertly defined philosophical genius as an ability to undermine central traditions of philosophy. Moreover, even if we concede that the project of deconstruction has given rise to the finest philosophical work of recent times, there is surely one other name that ought to appear on Rorty’s list—that of R.G. Collingwood.

As early as 1924, in Speculum Mentis, Collingwood mocked the idea of the philosopher as an “international boundary commissioner,” officiously mapping out the permitted limits of the empirical sciences. And in his Essay on Metaphysics, published in 1940, he went on to repudiate the entire Cartesian program of uncovering the indubitable foundations of thought. Every science, he contended, proceeds by asking and answering questions, and every set of questions eventually leads us back to the “ultimate presuppositions” of the science concerned. Such presuppositions are not themselves questioned, and cannot be assessed as either true or false; they simply furnish the frame within which the given science happens to be conducted at a given historical period. There are thus no ultimate categories of thought for the metaphysician to lay bare. There are only shifting paradigms, changing questions, new sets of answers, all inevitably limited by the scope of the prevailing disciplines. These are precisely the arguments that Rorty’s heroes mount, so it is surprising that he never mentions Collingwood’s work.

As well as introducing us to his heroes, Rorty’s opening chapters contain a much more sophisticated form of intellectual history. History, he proposes, may be capable of serving as therapy: by returning to the historical moments at which our current epistemological delusions arose, we may be able to liberate ourselves from their grasp. The main point at which Rorty employs this strategy is in examining what he takes to be the principal delusion of post-Cartesian philosophy—the belief that a certain range of ideas about the mind and its powers of understanding is conceptually indispensable. Liberation is said to come when we recover the questions to which these ideas were originally propounded as answers, and in consequence recognize that the distinctions involved, far from mirroring the essential nature of things, are merely reflections of various parochial and thoroughly unfortunate linguistic developments.


The first victim of this approach is the seemingly inescapable distinction between states of consciousness and events in the external world. As Rorty argues in his opening chapter, the drawing of such a distinction was so foreign to ancient Greek philosophy that there was simply no vocabulary for expressing it, and hence no temptation to divide the world up in a fashion which, in modern epistemology, has often been represented as essential to any system of thought.

A similar strategy is employed in chapter three, which surveys the epistemological tradition that has dominated philosophy since Descartes. Rorty turns first to Locke’s contention that when we speak of knowing something, we must be pointing to some relationship between ourselves and the object we claim to know. If there is to be any certainty about our knowledge, it follows that this must be a product of the way in which certain objects come to be apprehended. The suggestion Locke went on to make was that, if an object is presented to our senses, we cannot doubt its existence and we can thus claim to know it with absolute assurance.

Rorty’s first comment on this argument is that it confuses explanation with justification. Locke needs to show why we are justified in holding certain beliefs with particular tenacity; but all he succeeds in showing is how certain of our beliefs arise. Rorty’s main point, however, is again about the liberating power of history. By recovering the question Locke was trying to answer, we can see how his confusion arose; and by distancing ourselves from his assumptions in this way, we can avoid any temptation to accept his conclusions.

Rorty next turns to Kant, who shared Locke’s dilemma and proposed a new solution to it. Kant conceded that we cannot hope to acquire indubitable knowledge from mere sensory acquaintance with objects, and argued that this leaves us with only one possibility. If we are capable of knowing anything with complete assurance, this must be due to the process of interpretation that goes on in our own minds when we examine the raw data we receive from the outside world. It is because our minds in effect constitute the world that we can claim to know it with certainty.

Rorty again insists that an understanding of history serves to free us from the grip of such arguments. Once we see that Kant is responding to a confused question inherited from Locke, we can also see that he is merely providing a more elaborate answer to a puzzle that ought never to have been raised. Concluding his survey, Rorty spells out what he takes to be its implications for the whole project of trying to ground our knowledge on certainty. “The moral to be drawn,” he briskly asserts, “is that if this way of thinking about knowledge is optional, then so is epistemology, and so is philosophy as it has understood itself since the middle of the last century.”

Rorty completes his case, in chapter four, by adding some contemporary history. He observes that modern analytical philosophy, ignorant of its own past, has largely continued to uphold the Kantian belief in a realm of privileged and even indubitable truths, and has done so in two distinct ways. A special status still tends to be assigned to the claims that agents make about their own internal states—claims such as “I am in pain.” And philosophers still tend to assume that so-called analytical propositions—propositions of the form “no bachelor is married”—must be altogether indubitable, since their truth is not dependent on experience, but is guaranteed by the meanings of the terms involved.

These twin pillars of the analytical movement, Rorty contends, have both been dynamited in recent years by two influential American philosophers, Wilfrid Sellars and W.V.O. Quine. Sellars attacks the supposition that assertions of the form “I am in pain” offer automatically accurate descriptions of experience. To endorse this “myth of the given,” he maintains, is to put the argument back-to-front. The reason why we generally accept such first-person reports is not that they possess any privileged attachment to reality; it is rather, Sellars argues, that their authority stems from the fact that we cannot usually see how to place them in doubt.

Finally, Quine is said to have driven the last nail into the coffin of analytical philosophy by denying that any categorical distinction can be drawn between matters of definition and matters of fact. He challenges us to explain how we could possibly tell, in the case of an alien culture, whether its native speakers assent to propositions of the form “no bachelor is married” because they feel compelled by their language, or merely because they can think of no counter examples in their experience. Quine thus arrives at the view that allegedly “analytical” truths may simply be those for which no one has yet given us any interesting alternatives that might lead us to question them. And this destroys the Kantian dream of making a certain class of propositions indubitable in virtue of their purely linguistic character.


This central chapter—which Rorty regards as the most important in his book—is a true tour de force, executed with unflagging dialectical skill and a dazzling virtuosity of style. It not only provides a clear and vivid account of Quine’s and Sellars’s complex arguments but also contains a splendidly assured commentary in which their theories are partially recast in order to enlist them more securely in Rorty’s revolutionary cause. As a result, Rorty is able to restate his fundamental thesis in uncompromising terms: the search for the indubitable foundations of our thought is no more likely to succeed than the search for the unicorn, and ought immediately (and for the same reason) to be called off.

This is a book of exceptional originality and importance, and its central argument is presented with immense persuasive force. Does this mean, then, that philosophy is actually at an end? Has Rorty really succeeded in burying (and not praising) an entire academic discipline?

Rorty himself draws back from this apparent implication of his argument. At the end of the book he reassures us that he is not attacking the possibility of philosophy, but only its current practice. We generally conceive of philosophy as a matter of raising puzzles, taking up positions, trying to find out the truth about some particular theme. We ought instead to think of it on the model of a conversation; seeking to compare and widen our experiences rather than to win arguments. Once we approach the subject in what Rorty calls this “edifying” way, we shall see that—in Michael Oakeshott’s phrase—there will always be a place for philosophy in the conversation of mankind.

It is a strange paradox, however, that although Rorty ends with these precepts, he scarcely practices them at all in the body of his book. Asked to comment on the claim that some of our beliefs are indubitable, one might expect a merely “edifying” philosopher to pass by on the other side. But as we have seen, Rorty in fact responds in the best traditions of analytical philosophy. He treats the contention as an argument to be judged; he takes up a bold position about its status, declaring that it misconceives the nature of our concepts; and he ends by proclaiming the truth (at last) about “foundational” epistemology—that it rests on a series of essentially linguistic mistakes.

Nor is it surprising that Rorty fails to present himself wholeheartedly as an edifying philosopher. For it is clear, in spite of his reassurances, that he sees little future for the role. The only reason he offers for attending to the voice of philosophy is that it remains capable of warning us against the pitfalls of epistemology. As he concedes, however, such a function is obviously parasitic on the continuing existence of philosophy centered on epistemology. But this means that, once we are thoroughly persuaded that it is indeed misguided to look for the foundations of our beliefs, the voice counseling us against this rarified temptation will be left with nothing to say. So Rorty does seem committed to the view that, once the proper business of philosophy is successfully discharged, philosophy itself will come to an end.

However, the suggestion that we may be “in a fair way to dispensing with philosophy” is surely a considerable overstatement. There are many branches of the subject which remain content to investigate a variety of technical puzzles arising out of our prevailing structures of thought. They consider such disparate questions as the assumptions underlying our theories in the social and physical sciences, the semantic issues connected with the description of natural languages, the dilemmas arising out of what we take to be the process of rational choice, and so on. These topics have all been at the center of much discussion in recent philosophy, but they involve no necessary pretense of mirroring the ultimate reality of things, and therefore stand in no danger from Rorty’s assault on “foundational” theories of knowledge.

It is arguable, moreover, that some of the philosophical structures which undoubtedly suffer extensive damage from Rorty’s attack could perhaps be rebuilt in a less vulnerable style. The most obvious example is the current idiom of moral and political philosophy. Consider for a moment a typical instance of the genre, Robert Nozick’s book Anarchy, State and Utopia.1 Nozick confronts us at the outset with the assertion that “individuals have rights, and there are things no person or group may do to them (without violating their rights).” Nozick’s entire discussion then proceeds on the assumption that we possess our rights in roughly the same sense that we possess our arms and legs. They are properties of our persons, facts about ourselves that some forms of social organization are much better at recognizing than others. This is instantly vulnerable. As Rorty points out in some of his most telling asides, such a style of thought is woefully unprotected against two lethal possibilities: that our intuitions about the rights we possess may be nothing more than residua of certain linguistic practices; and that the actual list of our rights may be nothing more elevated or imposingly grounded than the list of the liberties our rulers may for the moment have chosen to bestow on us.

This line of criticism certainly calls in question a good deal of what Nozick and other like-minded moralists have supposed themselves to be doing. However, there is no reason why they should not respond by conceding that, like the rest of us, they doubtless derive their key moral intuitions to some extent from their social experience. Admittedly this would undercut their prestige, and would effectively place them on a par with the journalists, novelists, politicians, and other day-to-day commentators whose aim is to persuade us to their own point of view. But this would at least offer such philosophers (pace Rorty) a continuing function, and one that would not betray them into delusions of grandeur about the possibility of intuiting The Right Thing to Do or the Nature of the Just State.

If it does not follow, then, from endorsing Rorty’s epistemological skepticism that philosophy is dead, what does follow? The first implication to be drawn, according to Rorty himself, is that the connections between philosophy and the other disciplines have commonly been misunderstood. Philosophers have been apt to think of themselves as practitioners of a subject “which stands apart from empirical inquiry and explains the relevance of the results of such inquiry to the rest of culture.” But once we see that this position is founded on a misconception about the grounding of our beliefs, we can see that the position itself amounts to little more than a piece of academic imperialism.

A number of philosophers who have already commented on Rorty’s book have unhesitatingly dismissed this argument as a manifest absurdity. As Professor Alasdair MacIntyre has retorted, “If I am doomed to spending the rest of my life talking with literary critics and sociologists and historians and physicists, I am going to have to listen to a great deal of philosophy, much of it inept.” To ward off such an unthinkable prospect, he insists that we are bound to acknowledge “the need for a synoptic and systematic discipline concerned with the overall problems of justification and representation,” and are bound in consequence to uphold the traditional claims of philosophy.2

This resolutely misses one of Rorty’s most convincing points. It is true that Rorty lays himself open to such misunderstanding by failing to support his case with adequate examples. But it is fairly clear that what he has in mind is the sort of argument we encounter in a philosophical work such as MacIntyre’s own book, The Unconscious.3 This presents us with a standard empiricist account of the part played by explanatory concepts in helping us to deduce instances from regularities and predict their occurrence. It is then pointed out that Freud’s analysis of the unconscious fits this pattern unsatisfactorily. The right conclusion to be drawn, we are then assured, is that since Freud’s central concept is so dubious, he cannot really be regarded as providing explanations at all, but ought instead to be classified among the imaginative describers of human behavior.

One of the most notable achievements of Rorty’s book is to shatter whatever confidence we may have had in this type of conceptual analysis. The limitation inherent in MacIntyre’s approach is that it prevents us from even considering a possibility which obviously needs to be seriously raised: that a new theory such as Freud’s might force us to revise some of our most general beliefs—in this case, our beliefs about what constitutes a sound explanation. It is of course clear why MacIntyre is unable to countenance such a possibility. To do so he would have to give up one of the governing assumptions of his book (and indeed of his profession), the assumption that philosophers already know the answers to the basic “conceptual” questions (such as what counts as an explanation) and are thus in a position to criticize the exponents of the empirical sciences (such as Freud) for doing such a poor job. But as Rorty points out, to insist that all the empirical disciplines must be made to fit a pre-existing conceptual grid is to incur a most ironic risk. While posing as the vigilant guardians of rationality, such philosophers expose themselves to the perpetual danger of elevating our local and fallible canons of argument into a set of imperishable truths. A more serious affront to rationality would be hard to imagine.

The other major implication Rorty derives from his central thesis is that none of the empirical sciences has a privileged foundation in the sense of being able to claim a unique, essential grasp of reality. The empirical sciences all describe different realities, each with its own internal structure, as indeed do the humanities in Rorty’s view. It follows that all of them must be “on a par,” each capable of offering us a “repertoire of self-descriptions,” but none capable of offering anything more.

This commitment leads Rorty to defend a radical sequence of arguments, some of which are reminiscent of those associated with Professor Paul Feyerabend. Since we cannot stand outside our current language and structure of thought, the only test for any of our theories must be coherence with the rest of our beliefs. It follows that the much-prized “observations” on which our physical theories are based cannot amount to anything more than “what we can agree on these days.” And it follows from this that it must be an illusion to suppose “that physics is ‘objective’ in some way in which poetry or politics may not be.” All these disciplines are concerned with truth, but none of the truths they discover is more privileged in its foundations than any of the rest.

There is undoubtedly something highly salutary about this chain of reasoning. It reminds us that there is indeed a crucial sense in which the work, say, of a political theorist may be strongly analogous to that of an experimental scientist. It is not only that both may be judged on grounds of internal consistency and coherence with our other beliefs. It is also that, as Rorty likes to put it, both are “ways of coping,” and both are subject to stringent pragmatic tests. A good theory of mechanics enables us to construct bridges which are not likely to collapse; a good theory of politics enables us to construct states of which the same may be said with similar confidence.

But does this really mean that the two activities are altogether on a par? One objection to such a comparison is that it appears to underestimate the precision as well as the scope of the predictions that a well-developed physical theory licenses us to make. But the main objection is that the very idea of issuing such predictions presupposes that there is something in the external world for such theories to capture and encapsulate in a more or less satisfactory way. By insisting that all our claims to knowledge are nothing more than claims about existing canons of argument, Rorty seems to risk eliding one very obvious but vital point: that scientific theories are tested not simply by their coherence with our other beliefs, but also by their capacity to explain and control the phenomena of nature. It may well be that empiricist philosophies of science underestimated the extent to which the practice of science can appropriately be portrayed as a conversation rather than an inquiry; but Rorty seems in danger of drawing the illegitimate inference that such conversations cannot be inquiries at the same time.

Rorty has not I think demonstrated the falsity of supposing that scientific theories have a representational character. To this extent, the most startling implications of his argument must to some degree remain in doubt. However, his general view of what philosophers cannot any longer say with any confidence is developed with so much power and persuasiveness that I am well prepared to believe that my residual expressions of doubt may amount to little more than whistling to keep my philosophical spirits up. And as F.P. Ramsey immortally replied to Wittgenstein, “What we can’t say we can’t say, and we can’t whistle it either.”

This Issue

March 19, 1981