Charles Taylor (1995)

Charles Taylor

Overcoming Epistemology


Source: Philosophical Arguments, Harvard University Press, 1995. Just the first essay reproduced here.


Epistemology, once the pride of modern philosophy, seems in a bad way these days. Fifty years ago, during the heyday of logical empiricism, which was not only a powerful movement in philosophy but also immensely influential in social science, it seemed as though the very center of philosophy was its theory of knowledge. That was clearly philosophy’s main contribution to a scientific culture. Science went ahead and gathered knowledge; philosophical reflection concerned the validity of claims to knowledge. The pre-eminence of epistemology explains a phenomenon like Karl Popper. On the strength of his reputation as a theorist of scientific knowledge, he could obtain a hearing for his intemperate views about famous philosophers of the tradition, which bore a rather distant relation to the truth. It is reminiscent of a parallel phenomenon in the arts, whereby the political opinions of a great performer or writer are often listened to with an attention and respect that their intrinsic worth hardly commands.

Of course, all this was only true of the Anglo-Saxon world. On the Continent the challenge to the epistemological tradition was already in full swing. Heidegger and Merleau-Ponty had a wide influence. It would be too simple to say that this skeptical stance has now spread to the English-speaking world. Rather it seems true to say that epistemology has come under more intensive critical scrutiny in both cultures. In France, the generation of structuralists and poststructuralists was if anything even more alienated from this whole manner of thinking than Merleau-Ponty had been. In England and America, the arguments of both generations of continental thinkers have begun to have an impact. The publication of Richard Rorty’s influential Philosophy and the Mirror of Nature (1979) helped both to crystallize and to accelerate a trend toward the repudiation of the whole epistemological enterprise.

In some circles it is becoming a new orthodoxy that the whole enterprise from Descartes, through Locke and Kant, and pursued by various nineteenth- and twentieth-century succession movements, was a mistake. What is becoming less and less clear, however, is what exactly it means to overcome the epistemological standpoint or to repudiate the enterprise. just what is one trying to deny?

Rorty’s book seems to offer a clear and plausible answer. The heart of the old epistemology was the belief in a foundational enterprise. What the positive sciences needed to complete them, on this view, was a rigorous discipline that could check the credentials of all truth claims. An alleged science could be valid only if its findings met this test; otherwise it rested on sand. Epistemology would ultimately make clear just what made knowledge claims valid, and what ultimate degree of validity they could lay claim to. (One could, of course, come up with a rather pessimistic, skeptical answer to the latter question. Epistemology was not necessarily a rationalist enterprise. Indeed, its last great defenders were and are empiricists.)

In practice, epistemologists took their cue from what they identified as the successful sciences of their day, all the way from Descartes’s infatuation with mathematics to the contemporary vogue for reduction to physics. But the actual foundational science was not itself supposed to be dependent on any of the empirical sciences, and this obviously on pain of a circularity that would sacrifice its foundational character. Arguments about the source of valid knowledge claims were not supposed to be empirical.

If we follow this description, then it is clear what overcoming epistemology has to mean. It will mean abandoning foundationalism. On this view, Quine would figure among the prominent leaders of this new philosophical turn, since he proposes to “naturalize” epistemology, that is, deprive it of its a priori status and consider it as one science among others, one of many mutually interacting departments of our picture of the world. And so Rorty does seem to consider him, albeit with some reservations.

But there is a wider conception of the epistemological tradition, from whose viewpoint this last would be a rather grotesque judgment. This is the interpretation that focuses not so much on foundationalism as on the understanding of knowledge that made it possible. If I had to sum up this understanding in a single formula, it would be that knowledge is to be seen as correct representation of an independent reality. In its original form, it saw knowledge as the inner depiction of an outer reality.

The reason why some thinkers prefer to focus on this interpretation, rather than merely on the foundationalist ambitions that are ultimately (as Quine has shown) detachable from it, is that it is bound up with very influential and often not fully articulated notions about science and about the nature of human agency. Through these it connects with certain central moral and spiritual ideas of the modern age. If one’s aim is, in challenging the primacy of epistemology, to challenge these ideas as well, then one has to take it up in this wider-or deeper-focus, and not simply show the vanity of the foundational enterprise.

I would like now to trace some of these connections. One of them is evident: the link between the representational conception and the new, mechanistic science of the seventeenth century. This is, in fact, twofold. On one side, the mechanization of the world picture undermined the previously dominant understanding of knowledge and thus paved the way for the modern view. The most important traditional view was Aristotle’s, according to which when we come to know something, the mind (nous) becomes one with the object of thought. Of course this is not to say that they become materially the same thing; rather, mind and object are informed by the same eidos. Here was a conception quite different from the representational model, even though some of the things Aristotle said could be construed as supporting the latter. The basic bent of Aristotle’s model could much better be described as participational: being informed by the same eidos, the mind participates in the being of the known object, rather than simply depicting it.

But this theory totally depends on the philosophy of Forms. Once we no longer explain the way things are in terms of the species that inform them, this conception of knowledge is untenable and rapidly becomes almost unintelligible. We have great difficulty in understanding it today. The representational view can then appear as the only available alternative.

This is the negative connection between mechanism and modern epistemology. The positive one obtrudes as soon as we attempt to explain our knowing activity itself in mechanistic terms. The key to this is obviously perception, and if we see it as another process in a mechanistic universe, we have to construe it as involving as a crucial component the passive reception of impressions from the external world. Knowledge then hangs on a certain relation holding between what is “out there” and certain inner states that this external reality causes in us. This construal, valid for Locke, applies just as much to the latest artificial-intelligence models of thinking. It is one of the mainsprings of the epistemological tradition.

The epistemological construal is, then, an understanding of knowledge that fits well with modern mechanistic science. This is one of its great strengths, and certainly it contributes to the present vogue of computer-based models of the mind. But that’s not all this construal has going for it. It is in fact heavily over-determined. For the representational view was also powered by the new ideals of science, and new conceptions of the excellence of thought, that arose at the same time.

This connection was central to Descartes’s philosophy. it was one of his leading ideas that science, or real knowledge, does not simply consist of a congruence between ideas in the mind and the reality outside. If the object of my musings happens to coincide with real events in the world, this doesn’t give me knowledge of them. The congruence has to come about through a reliable method, generating well-founded confidence. Science requires certainty, and this can only be based on that undeniable clarity Descartes called évidence. “Toute science est une connaissance certaine et évidente,” runs the opening sentence of the second rule in Rules for the Direction of the Mind.

Now certainty is something the mind has to generate for itself It requires a reflexive turn, where instead of simply trusting the opinions you have acquired through your upbringing, you examine their foundation, which is ultimately to be found in your own mind. Of course, the theme that the sage has to turn away from merely current opinion, and make a more rigorous examination that leads him to science, is a very old one, going back at least to Socrates and Plato. But what is different with Descartes is the reflexive nature of this turn. The seeker after science is not directed away from shifting and uncertain opinion toward the order of the unchanging, as with Plato, but rather within, to the contents of his own mind. These have to be carefully distinguished both from external reality and from their illusory localizations in the body, so that then the correct issue of science, that is, of certainty, can be posed — the issue of the correspondence of idea to reality, which Descartes raises and then disposes of through the supposition of the malin génie and the proof of his negation, the veracious God.

The confidence that underlies this whole operation is that certainty is something we can generate for ourselves, by ordering our thoughts correctly — according to clear and distinct connections. This confidence is in a sense independent of the positive outcome of Descartes’s argument to the existence of a veracious God, the guarantor of our science. The very fact of reflexive clarity is bound to improve our epistemic position, as long as knowledge is understood representationally. Even if we couldn’t prove that the malin génie doesn’t exist, Descartes would still be in a better position than the rest of us unreflecting minds, because he would have measured the full degree of uncertainty that hangs over all our beliefs about the world, and clearly separated out our undeniable belief in ourselves.

Descartes is thus the originator of the modern notion that certainty is the child of reflexive clarity, or the examination of our own ideas in abstraction from what they “represent,” which has exercised such a powerful influence on western culture, way beyond those who share his confidence in the power of argument to prove strong theses about external reality. Locke and Hume follow in the same path, although Hume goes about as far in the direction of skepticism as any modern has. Still, it remains true for Hume that we purge ourselves of our false confidence in our too-hasty extrapolations by focusing attention on their origin in our ideas. It is there that we see, for instance, that our beliefs in causation are based on nothing more than constant conjunction, that the self is nothing but a bundle of impressions, and so on.

This reflexive turn, which first took form in the seventeenth- and eighteenth-century “way of ideas,” is indissolubly linked to modern representational epistemology. One might say it presupposes this construal of knowledge. If Plato or Aristotle were right, the road to certainty couldn’t be inward — indeed, the very notion of certainty would be different: defined more in terms of the kinds of being that admit of it, rather than by the ordering of our thoughts. But I believe there is also a motivational connection in the other direction: the ideal of self-given certainty is a strong incentive to construe knowledge in such a way that our thought about the real can be distinguished from its objects and examined on its own. And this incentive has long outlived the original way of ideas. Even in an age when we no longer want to talk of Lockean “ideas” or of “sense data,” where the representational view is reconstrued in terms of linguistic representations or bodily states (and these are perhaps not genuine alternatives), there is still a strong draw toward distinguishing and mapping the formal operations of our thinking. In certain circles it would seem that an almost boundless confidence is placed in the defining of formal relations as a way of achieving clarity and certainty about our thinking, be it in the (mis)application of rational choice theory to ethical problems or in the great popularity of computer models of the mind.

The latter is an excellent example of what I called the “over-determination” of the epistemological construal. The plausibility of the computer as a model of thinking comes partly from the fact that it is a machine, hence living “proof” that materialism can accommodate explanations in terms of intelligent performance; but partly too it comes from the widespread faith that our intelligent performances are ultimately to be understood in terms of formal operations. The computer, it has been said, is a “syntactic engine”. A controversy rages over precisely this point. The most perspicuous critics of the runaway enthusiasm with the computer model, such as Hubert Dreyfus, tirelessly point out how implausible it is to understand certain of our intelligent performances in terms of a formal calculus, including our most common everyday actions, such as making our way around rooms, streets, and gardens or picking up and manipulating the objects we use. But the great difficulties that computer simulations have encountered in this area don’t seem to have dimmed the enthusiasm of real believers in the model. It is as though they had been vouchsafed some revelation a priori that it must all be done by formal calculi. Now this revelation, I submit, comes from the depths of our modern culture and the epistemological model anchored in it, whose strength is based not just on its affinity to mechanistic science but also on its congruence to the powerful ideal of reflexive, self-given certainty.

For this has to be understood as something like a moral ideal. The power of this ideal can be sensed in the following passage from Husserl’s Cartesian Meditations (1929), all the more significant in that Husserl had already broken with some of the main theses of the epistemological tradition. He asks in the first meditation whether the “hopelessness” of the current philosophical predicament doesn’t spring from our having abandoned Descartes’s original emphasis on self-responsibility:

Sollte die vermeintlich überspannte Forderung einer auf letzte erdenkliche Vorurteilslosigkeit abgestellten Philosophie, einer in wirklicher Autonomie aus letzten selbst erzeugten Evidenzen sich gestaltenden und sich von daher absolut selbstverantwortenden Philosophie nicht vielmehr zum Grundsinn echter Philosophie gehören?

The ideal of self-responsibility is foundational to modern culture. It emerges not only in our picture of the growth of modern science through the heroism of the great scientist, standing against the opinion of his age on the basis of his own self-responsible certainty-Copernicus, Galileo (he wobbled a bit before the Holy Office, but who can blame him?), Darwin, Freud. It is also closely linked to the modern ideal of freedom as self-autonomy, as the passage from Husserl implies. To be free in the modern sense is to be self-responsible, to rely on your own judgment, to find your purpose in yourself And so the epistemological tradition is also intricated in a certain notion of freedom, and the dignity attaching to us in virtue of this. The theory of knowledge partly draws its strength from this connection. But, reciprocally, the ideal of freedom has also drawn strength from its sensed connection with the construal of knowledge seemingly favored by modern science. From this point of view it is fateful that this notion of freedom has been interpreted as involving certain key theses about the nature of the human agent; we might call them anthropological beliefs. Whether these are in fact inseparable from the modern aspiration to autonomy is an open question, and a very important one, to which I will return briefly later. But the three connected notions I want to mention here are closely connected historically with the epistemological construal.

The first is the picture of the subject as ideally disengaged, that is, as free and rational to the extent that he has fully distinguished himself from the natural and social worlds, so that his identity is no longer to be defined in terms of what lies outside him in these worlds. The second, which flows’ from this, is a punctual view of the self, ideally ready as free and rational to treat these worlds — and even some of the features of his own character — instrumentally, as subject to change and reorganizing in order the better to secure the welfare of himself and others. The third is the social consequence of the first two: an atomistic construal of society as constituted by, or ultimately to be explained in terms of, individual purposes.

The first notion emerges originally in classical dualism, where the subject withdraws even from his own body, which he is able to look on as an object; but it continues beyond the demise of dualism in the contemporary demand for a neutral, objectifying science of human life and action. The second originates in the ideals of the government and reform of the self that have such an important place in the seventeenth century and of which Locke develops an influential version;” it continues today in the tremendous force that instrumental reason and engineering models have in our social policy, medicine, psychiatry, politics, and so on. The third notion takes shape in social-contract theories of the seventeenth century, but continues not only in their contemporary successors but also in many of the assumptions of contemporary liberalism and mainstream social science.

We don’t need to unpack these ideas any further to see that the epistemological tradition is connected with some of the most important moral and spiritual ideas of our civilization-and also with some of the most controversial and questionable. To challenge them is sooner or later to run up against the force of this tradition, which stands with them in a complex relation of mutual support. Overcoming or criticizing these ideas involves coming to grips with epistemology. But this means taking it in what I identified as its broad focus, the whole representational construal of knowledge, not just as the faith in foundationalism.

When we turn to the classic critiques of epistemology, we find that they have, in fact, mostly been attuned to this interpenetration of the scientific and the moral. Hegel, in his celebrated attack on this tradition in the introduction to the Phenomenology of Spirit, speaks of a “fear of error” that “reveals itself rather as fear of the truth."He goes on to show how this stance is bound up with a certain aspiration to individuality and separatedness, refusing what he sees as the “truth” of subject-object identity. Heidegger notoriously treats the rise of the modern epistemological standpoint as a stage in the development of a stance of domination to the world, which culminates in contemporary technological society. Merleau-Ponty draws more explicitly political connections and clarifies the alternative notion of freedom that arises from the critique of empiricism and intellectualism.” The moral consequences of the devastating critique of epistemology in the later Wittgenstein are less evident, since he was strongly averse to making this kind of thing explicit. But those who followed him have shown a certain affinity for the critique of disengagement, instrumental reason, and atomism.

It is safe to say that all these critics were largely motivated by a dislike of the moral and spiritual consequences of epistemology and by a strong affinity for some alternative. Indeed, the connection between the scientific and the moral is generally made more evident in their work than in that of mainstream supporters of the epistemological standpoint. But an important feature of all these critiques is that they establish a new moral outlook through overturning the modern conception of knowledge. They don’t simply register their dissidence from the anthropological beliefs associated with this conception, but show the foundations of these beliefs to be unsound, based as they are in an untenable construal of knowledge.

AH four of the men I have mentioned — whom I take to be the most important critics of epistemology, the authors of the most influential forms of critique-offer new construals of knowledge. Moreover, in spite of the great differences, all four share a basic form of argument, which finds its origins in Kant and which one might call “the argument from transcendental conditions.”

By this I mean something like the following. We argue the inadequacy of the epistemological construal, and the necessity of a new conception, from what we show to be the indispensable conditions of there being anything like experience or awareness of the world in the first place. just how to characterize this reality, whose conditions we are defining, can itself be a problem, of course. Kant speaks of it simply as “experience”; but Heidegger, with his concern to get beyond subjectivistic formulations, ends up talking about the “clearing” (Lichtung). Where the Kantian expression focuses on the mind of the subject and the conditions of having what we can call experience, the Heideggerian formulation points us toward another facet of the same phenomenon, the fact that anything can appear or come to light at all. This requires that there be a being to whom it appears, or whom it is an object; it requires a knower, in some sense. But the Lichtung. formulation focuses us on the fact (which we are meant to conic to perceive as astonishing) that the knower-known complex is at all, rather than taking the knower for granted as “subject” and examining what makes it possible to have any knowledge or experience of a world.

For all this extremely important shift in the center of gravity of what we take as the starting point, there is a continuity between Kant and Heidegger, Wittgenstein, or Merleau-Ponty. They all start from the intuition that this central phenomenon of experience, or the clearing, is not made intelligible on the epistemological construal, in either its empiricist or rationalist variants. That construal offers an account of stages of the knower consisting of an ultimately incoherent amalgam of two features: (a) these states (the ideas) are self-enclosed, in the sense that they can be accurately identified and described in abstraction from the “outside” world (this is, of course, essential to the whole rationalist thrust of reflexive testing of the grounds of knowledge); and (b) they nevertheless point toward and represent things in that outside world.

The incoherence of this combination may be hidden from us by the existence of things that seem to have feature (a), such as certain sensations, and even of states that seem to combine (a) and (b), such as stable illusions. But what clearly emerges from the whole argument of the last two centuries is that the condition of states of ourselves having (b) is that they cannot satisfy (a). This already began to be evident with classical empiricism in its uncertain shuffling between two definitions of the “idea” or “impression": on one reading, it was simply a content of the mind, an inner quasi-object, and it called for an object-description; on another, it had to be a claim about how things stood, and it could only be captured in a that-clause.

Feature (b) is what later came to be called in the Brentano-Husserl tradition “intentionality": our ideas are essentially of or about something. Here is another way of characterizing the central condition of experience or the clearing. What Kant calls transcendental conditions are conditions of intentionality, and the lines of argument that descend from Kant can be seen as exploring what these have to be.

Kant already showed that the atomistic understanding of knowledge that Hume espoused was untenable in the light of these conditions. If our states were to count as experience of an objective reality, they had to be bound together to form a coherent whole, or bound together by rules, as Kant conceived it. However much this formulation may be challenged, the incoherence of the Humean picture, which made the basis of all knowledge the reception of raw, atomic, uninterpreted data, was brilliantly demonstrated. How did Kant show this? He established in fact an argument form that has been used by his successors ever since. It can be seen as a kind of appeal to intuition. In the case of this particular refutation of Hume (which is, I believe, the main theme of the transcendental deduction in the first edition of the Critique of Pure Reason), he makes us aware, first, that we wouldn’t have what we recognize as experience at all unless it were construable as of an object (I take this as a kind of proto-thesis of intentionality), and second, that their being of an object entails a certain relatedness among our “representations.” Without this, Kant says, “it would be possible for appearances to crowd in upon the soul and yet to be such as would never allow of experience.” Our perceptions “would not then belong to any experience, consequently would be without an object, merely a blind play of representations, less even than a dream.”

I think this kind of appeal to intuition is better understood as an appeal to what I want to call our “agent’s knowledge.” As subjects effectively engaged in the activities of getting to perceive and know the world, we are capable of identifying certain conditions without which our activity would faH apart into incoherence. The philosophical achievement is to define the issues properly. Once this is done, as Kant does so brilliantly in relation to Humean empiricism, we find there is only one rational answer. Plainly we couldn’t have experience of the world at all if we had to start with a swirl of uninterpreted data. Indeed, there would be no “data,” because even this minimal description depends on our distinguishing what is given by some objective source from what we merely supply ourselves.

Now the four authors I mention push this argument form further, and explore conditions of intentionality that require a more fundamental break with the epistemological tradition. In particular, they push it far enough to undermine the anthropological beliefs I described earlier: beliefs in the disengaged subject, the punctual self, and atomism.

The arguments of Heidegger and Merleau-Ponty put paid to the first view. Heidegger, for instance, shows — especially in his celebrated analysis of being-in-the-world — that the condition of our forming disengaged representations of reality is that we must be already engaged in coping with our world, dealing with the things in it, at grips with them.” Disengaged description is a special possibility, realizable only intermittently, of a being (Dasein) who is always “in” the world in another way, as an agent engaged in realizing a certain form of life. That is what we are about “first and mostly” (zunichst und zumeist).

The tremendous contribution of Heidegger, like that of Kant, consists in having focused the issue properly. Once this is done, we can’t deny the picture that emerges. Even in our theoretical stance to the world, we are agents. Even to find out about the world and formulate disinterested pictures, we have to come to grips with it, experiment, set ourselves to observe, control conditions. But in all this, which forms the indispensable basis of theory, we are engaged as agents coping with things. It is clear that we couldn’t form disinterested representations any other way.

But once we take this point, then the entire epistemological position is undermined. Obviously foundationalism goes, since our representations of things — the kinds of objects we pick out as whole, enduring entities-are grounded in the way we deal with those things. These dealings are largely inarticulate, and the project of articulating them fully is an essentially incoherent one, just because any articulative project would itself rely on a background or horizon of non-explicit engagement with the world.

But the argument here cuts deeper. Foundationalism is undermined because you can’t go on digging under our ordinary representations to uncover further, more basic representations. What you get underlying our representations of the world-the kinds of things we formulate, for instance, in declarative sentences-is not further representation but rather a certain grasp of the world that we have as agents in it. This shows the whole epistemological construal of knowledge to be mistaken. It doesn’t just consist of inner pictures of outer reality, but grounds in something quite other. And in this “foundation,” the crucial move of the epistemological construal — distinguishing states of the subject (our “ideas”) from features of the external world — can’t be effected. We can draw a neat line between my picture of an object and that object, but not between my dealing with the object and that object. It may make sense to ask us to focus on what we believe about something, say a football, even in the absence of that thing; but when it comes to playing football, the corresponding suggestion would be absurd. The actions involved in the game can’t be done without the object; they include the object. Take it away and we have something quite different — people miming a game on the stage, perhaps. The notion that our understanding of the world is grounded in our dealings with it is equivalent to the thesis that this understanding is not ultimately based on representations at all, in the sense of depictions that are separately identifiable from what they are of.

Heidegger’s reflections take us entirely outside the epistemological construal. Our reflections on the conditions of intentionality show that these include our being “first and mostly” agents in the world. But this also ruins the conception of the agent as one whose ideal could be total disengagement. This turns out to be an impossibility, one that it would be destructive to attempt. We can’t turn the background against which we think into an object for us. The task of reason has to be conceived quite differently: as that of articulating the background, “disclosing” what it involves. This may open the way to detaching ourselves from or altering part of what has constituted it — may, indeed, make such alteration irresistible; but only through our unquestioning reliance on the rest.

And just as the notion of the agent underpinning the ideal of disengagement is rendered impossible, so is the punctual notion of the self. Heidegger and Merleau-Ponty both show how the inescapability of the background involves an understanding of the depth of the agent, but they do so by exploring the conditions of intentionality in complementary directions. Heidegger shows how Dasein’s world is defined by the related purposes of a certain way of life shared with others. Merleau-Ponty shows how our agency is essentially embodied and how this lived body is the locus of directions of action and desire that we never fully grasp or control by personal decision.

This critique also puts in question the third anthropological belief I singled out above, atomism. I have just mentioned how Heidegger’s notion of Dasein’s way of life is essentially that of a collectivity. A general feature of paradigm-setting critiques is that they strongly reject this third view and show instead the priority of society as the locus of the individual’s identity. But crucially this point is made through an exploration of the role of language. The new theory of language that arises at the end of the eighteenth century, most notably in the work of Herder and Humboldt, not only gives a new account of how language is essential to human thought, but also places the capacity to speak not simply in the individual but primarily in the speech community. This totally upsets the outlook of the mainstream epistemological tradition. Now arguments to this effect have formed part of the refutation of the atomism that has proceeded through an overturning of standard modern epistemology.

Important examples of arguments of this kind are Hegel’s in the first chapter of the Phenomenology of Spirit, against the position he defines as “sensible certainty,” where he shows both the indispensability of language and its holistic character; and Wittgenstein’s famous demonstrations of the uselessness of “ostensive definitions,” where he makes Plain the crucial role played by language in identifying the object and the impossibility of a purely private language.” Both are, I believe, excellent examples of arguments that explore the conditions of intentionality and show their conclusions to be inescapable.

It is evident that these arguments give us a quite different notion of what it is to overcome epistemology from those that merely eschew foundationalism. We can measure the full gulf by comparing any of the four — Heidegger, perhaps, or Merleau-Ponty — with the Quine of “Epistemology Naturalized.” It is plain that the essential elements of the epistemological construal have remained standing in Quine, and not surprisingly therefore the central anthropological beliefs of the tradition. Disengagement emerges in his “taste for desert landscapes”; the punctual self in his behaviorism; and atomism in his particular brand of political conservatism.” In face of differences of this magnitude, a question arises concerning what it means to overcome epistemology.”

A picture has been emerging here of what this ought to be-a tendentious one, I freely admit. It accepts the wider or deeper definition of the task: overcoming the distorted anthropological beliefs through a critique and correction of the construal of knowledge that is interwoven with them and has done so much to give them undeserved credit. Otherwise put: through a clarification of the conditions of intentionality, we come to a better understanding of what we are as knowing agents — and hence also as language beings — and thereby gain insight into some of the crucial anthropological questions that underpin our moral and spiritual beliefs.

For all its radical break with the tradition, this kind of philosophy would in one respect be in continuity with it. It would be carrying further the demand for self-clarity about our nature as knowing agents, by adopting a better and more critically defensible notion of what this entails. Instead of searching for an impossible foundational justification of knowledge or hoping to achieve total reflexive clarity about the bases of our beliefs, we would now conceive this self-understanding as awareness about the limits and conditions of our knowing, an awareness that would help us to overcome the illusions of disengagement and atomic individuality that are constantly being generated by a civilization founded on mobility and instrumental reason.

We could understand this as carrying the project of modern reason, even of “self-responsible” reason, further by giving it a new meaning. This is how Husserl conceived the critical project in his last great lectures on the “crisis of European science,” given in Vienna in 1935. Husserl thinks of us as struggling to realize a fundamental task, that of the “europäischen Geist,” whose goal is to achieve the fullness of reflexive clarity. We should see ourselves as philosopher-functionaries (“Funktionäre der neuzeitlichen philosophischen Menschheit”). The first foundation (Urstiftung) of the European tradition points to a final foundation (Endstiftung), and only in the latter is the former fully revealed:

nur von ihr [Endstiftung] aus kann sich die cinheitliche Ausgerichtetheit aller Philosophen und Philosophien erijffnen, und von ihr aus kann eine Erhellung gewonnen werden, in welcher man die vergangene Denker versteht, wie sie selbst sich me hätten verstchen können.

Husserl’s hope here sounds ridiculously overstated, which may have something to do with his having failed to push through his critique of foundationalism to the end. Overstatement has played an important role, as we will see, in casting discredit on the task as I have outlined it. But if we purge Husserl’s formulation of the prospect of a “final foundation” where absolute apodicticity would at last be won, if we concentrate merely on the gain for reason in coming to understand what is illusory in the modern epistemological project and in articulating the insights about us that flow from this, then the claim to have taken the modern project of reason a little farther, and to have understood our forbears a little better than they understood themselves, is not so unbelievable.

What reflection in this direction would entail is already fairly wen known. It involves, first, conceiving reason differently, as including — alongside the familiar forms of the Enlightenment — a new department, whose excellence consists in our being able to articulate the background of our lives perspicuously. We can use the word “disclosure” for this, following Heidegger. And along with this goes a conception of critical reasoning, of especial relevance for moral thinking, that focuses on the nature of transitions in our thought, of which “immanent critique” is only the best-known example.

In moral thought, what emerges from this critique is a rejection of moralities based purely on instrumental reason, such as utilitarianism; and also critical distance from those based on a punctual notion of the self, such as the various derivations of Kant. The critique of John Rawls’s theory by Michael Sandel, in the name of a less “thin” theory of the agent, is an excellent example of this. In social theory, the result is a rejection of atomist theories, of reductive causal theories (such as “vulgar” Marxism or sociobiology), and of theories that cannot accommodate intersubjective meaning. Social science is seen as being closer to historiography of a certain kind. In politics, the anti-atomist thrust of the critique makes it hostile to certain forms of contemporary conservatism, but also to radical doctrines of non-situated freedom. I believe there is a natural affinity between this critique, with its stress on situated freedom and the roots of our identity in community, on the one hand, and the civic humanist tradition on the other, as the works of a number of writers, from Humboldt to Arendt, testify.

It might seem now as though everything should run on smoothly, toward a set of anthropological conclusions with a certain moral-political hue. But in fact all this is hotly contested, not just by those who wish to defend the epistemological tradition, which would be understandable, but by those who consider themselves its critics. Foremost among these are a range of thinkers who have defined themselves in relation to a certain reading of Nietzsche. The most interesting and considerable of them, in my opinion, is Foucault. In keeping with the themes of this chapter, we can perhaps get most directly to the basis of their dissent if we go to the moral or spiritual outlook they wish to defend. In the case of Foucault this became relatively clear at the end of his life. He rejected the concept of the punctual self, which could take an instrumental stance toward its life and character — this is indeed what arises out of the practices and “truths” of the disciplinary society he painted in such repellent colors (whatever protestations of neutrality accompanied the depiction). But he couldn’t accept the rival notion of a deep or authentic self that arises out of the critical traditions of Hegel and, in another way, Heidegger or Merleau-Ponty. This seemed to him another prison. He rejected both in favor of a Nietzschean notion of the self as potentially self-making, the self as a work of art, a central conception of an “aesthetics of existence.”

Something analogous, but on a much more frivolous level, seems to animate some of the poststructuralist thinkers — Derrida, for instance. Paradoxically, for all the talk of the “end of subjectivity,” one of the strong attractions of this kind of position is precisely the license it offers to subjectivity, unfettered by anything in the nature of a correct interpretation or an irrecusable meaning of either life or text, to effect its own transformations, to invent meaning. Self-making is again primary.

Nietzsche’s insights into the way in which language imposes order on our world, into theory as a kind of violence, were crucial to all views of this kind. It offers an alternative to the kind of possible critique of epistemology in which we discover something deeper and more valid about ourselves in carrying it through — the kind I have been describing. Instead it attacks the very aspiration to truth, as this is usually understood. All epistemic orders are imposed, and the epistemological construal is just another one of those orders. It has no claim to ultimate correctness, not because it has been shown inadequate by an exploration of the conditions of intentionality, but because all such claims are bogus. They mistake an act of power for a revelation of truth. Husserl’s Urstiftung takes on a quite different and more sinister air.

Clearly this is the critique of epistemology that is most compatible with the spiritual stance of self-making. It makes the will primary in a radical way, whereas the critique through conditions of intentionality purports to show us more of what we really are like-to show us, as it were, something of our deep or authentic nature as selves. So those who take the Nietzschean road are naturally very reluctant to understand the critique as a gain in reason. They would rather deny that reason can have anything to do with our choices of what to be.

This is not to say that they propose the end of epistemology as a radical break. just as the critique through conditions of intentionality represents a kind of continuity — through-transformation in the tradition of self-critical reason, so the Nietzschean refusal represents a continuity-through-transformation of another facet of the modern identity — the primacy of the will. This played an important role in the rise of modern science and its associated epistemological standpoint; in a sense, a voluntaristic anthropology, with its roots in a voluntaristic theology, prepared the ground over centuries for the seventeenth-century revolution, most notably in the form of nominalism. It is a crucial point of division among moderns, what we think of primacy of the will. This is one of the issues at stake between these two conceptions of what it means to overcome the epistemological tradition.

Although this represents perhaps the most dramatic opposition among critics of epistemology, it is far from exhausting the field. Habermas, for instance, has staked out a position equivalent to neither. Against the neo-Nietzscheans, he would strongly defend the tradition of critical reason, but he has his own grounds for distrusting Heideggerian disclosure and wants instead to hold on to a formal understanding of reason and, in consequence, a procedural ethic, although purged of the monological errors of earlier variants. He has drawn heavily on the critique of epistemology in the four authors mentioned above, but fears for the fate of a truly universal and critical ethic if one were to go all the way with this critique.

How do we adjudicate this kind of dispute? How do we decide what it really means to overcome epistemology? I can’t hope to decide the issue here, only to make a claim as to how it should be settled. In order to define this better, I want to return to the most dramatic dispute, that between the neo-Nietzscheans and the defenders of critical reason.

It seems to me that, whoever is ultimately right, the dispute has to be fought on the terrain of the latter. The Nietzschean position also stands and falls with a certain construal of knowledge: that it is relative to various’ ultimately imposed “regimes of truth,” to use Foucault’s expression. This has to show itself to be a superior construal to that which emerges from the exploration of the conditions of intentionality. Does it?

Certainly the Nietzschean conception has brought important insights: no construal is quite innocent, something is always Suppressed; and what is more, some interlocutors are always advantaged relative to others, for any language .3 ‘ But the issue is whether this settles the matter of truth between construals. Does it mean that there can be no talk of epistemic gain in passing from one construal to another? That there is such a gain is the claim of those exploring the conditions of intentionality. This claim doesn’t stand and fall with a naive, angelic conception of philosophical construals as utterly uninvolved with power. Where is the argument that will show the more radical Nietzschean claim to be true and the thesis of critical reason untenable?

I regret to say that one hears very little serious argument in this domain. Neo-Nietzscheans seem to think that they are dispensed from it since it is already evident or, alternatively, that they are debarred from engaging in it on pain of compromising their position. Derrida and his followers seem to belong to the first category. The main weight of argument is carried here by an utterly caricatural view of the alternative as involving a belief in a kind of total self-transparent clarity, which would make even Hegel blush. The rhetoric deployed around this has the effect of obscuring the possibility that there might be a third alternative to the two rather dotty ones on offer; and as long as you go along with this, the Derridian view seems to win as the least mad, albeit by a hair.

Others try to argue on behalf of Foucault that he couldn’t enter the argument concerning construals of knowledge without abandoning his Nietzschean position, that there is nothing to argue between them. True enough, but then the issue whether there is something to argue itself demands some kind of support. Something can surely be said about that. Indeed, much has been said, by Nietzsche for one, and some also by Foucault-in talking for instance of “regimes of truth”; the question is whether it is really persuasive or involves a lot of slippery slides and evasion.

In short, the arguments for not arguing seriously are uniformly bad. And in fact Foucault did on one occasion make a serious attempt to engage with the exploration of the conditions of intentionality, and that was in the latter part (chapter 9) of The Order of Things, where he talks about the invention of Man and the “transcendental-empirical double.” This was admittedly prior to his last, much more centrally Nietzschean phase, but it can be seen as preparing the ground for this, as indeed Dreyfus and Rabinow see it.

The arguments here seem to me much more to build on the Heideggerian and Merleau-Pontyan critique against Kant rather than to challenge this critique. And the arguments, if valid, would have the consequence that nothing coherent could be said at all about the conditions of intentionality. I can’t see how this could fail to undercut the Nietzschean view as well. In The Order of Things Foucault takes refuge in a species of structuralism, which is meant to avoid this question altogether. But he abandons it soon afterwards, and we are left uncertain where the argument is meant to take us. In general among neo-Nietzscheans, however, an atmosphere reigns in which this issue is felt as already settled. We are exhorted by Lyotard not to take metanarratives seriously any more, but the argument for this seems to rely on caricature.

If I am right, the issue is far from settled. And yet at stake in this struggle over the corpse of epistemology are some of the most important spiritual issues of our time. The question, what it is to overcome epistemology, turns out to be of more than just historical interest.