|Home Page|Publications|
Forthcoming in Philosophical Books.

|Mind and World

By John McDowell

Harvard University Press, 1994. x + 192 pp. $27.50

How do rational minds make contact with the world? The empiricist tradition sees a gap between mind and world, and takes sensory experience, fallible as it is, to provide our only bridge across that gap. In its crudest form, for example, the traditional idea is that our minds consult an inner realm of sensory experience, which provides us with evidence about the nature of external reality. Notoriously, however, it turns out to be far from clear that there is any viable conception of experience which allows it to do the job. The original problem is to show that thought is rationally constrained by external reality. If sensory experience is to provide the solution--in particular, if it is to provide the answer to sceptical challenges--it must therefore meet two criteria. First, it must itself be `receptive'--i.e., appropriately constrained by external reality. Second, it must be the kind of thing that can enter into a logical or rational relationship with belief--it must already be `conceptual,' in other words. In arguing against the idea that anything could serve both roles, Wilfred Sellars termed this conception of experience "the Myth of the Given."

As McDowell points out in Mind and World, however, the Given is but one horn of a dilemma which goes back to Kant. Without the Given--without, that is, a viable conception of experience such that it could serve as the bridge between mind and world--we seem to be faced with a kind of blind internalism, in which our beliefs lack not merely justification, but content itself.

Some contemporary philosophers have been prepared to accept this second horn--Sellars himself, for example, and Davidson, whose "third dogma of empiricism" is again the idea that experience serves this foundational role. But McDowell regards this alternative as profoundly unsatisfactory. In this book--a product of his 1991 John Locke Lectures--McDowell's project is to try to make it plausible that, properly conceived, experience can indeed provide external anchors for belief. Properly construed, McDowell argues, experience can indeed combine the receptivity required for it to claim contact with the world with full participation in the conceptual realm. The trick is to see experience not as a kind of shell surrounding the space of reasons, but simply as the outer region of the space itself--the place where it makes immediate contact with the world. On McDowell's view, empiricism's mistake was to think that there is a gap to be bridged in the first place--to think of experience as an intermediary between the world and the rational mind, rather than as a region of direct contact.

As McDowell says, his view has antecedents, especially among post-Kantian thinkers such as Hegel, who reject Kant's notion of the unconceptualised `world in itself.' In this tradition, however, it has been taken to lead to idealism, and McDowell thinks this is a mistake. He sees the view as compatible with--indeed, eventually, the only acceptable form of--realism.

The position McDowell is trying to defend is thus an unusual and difficult one, threatened from several directions by a range of philosophical opponents. Indeed, it is not an easy position to bring clearly into focus, let alone to defend. McDowell's defence is not, in my view, especially convincing. But few contemporary philosophers tackle these issues, and McDowell is among the most able. His account of the dilemma posed by the notion of the Given is insightful and illuminating, not least for his careful attention to the distinction between appeals to experience which justify our beliefs, and appeals which merely explain--which merely exculpate, as McDowell puts it, by removing our beliefs from the realm of justification altogether. Although the book as a whole is often difficult, sometimes obscure, it rewards close study. It is often subtle and challenging, and deserves to be widely read--even if, as for me, its effect may be to convince the reader of the untenability of its main thesis.

In the remainder of this review I shall focus on what seem to me the two main difficulties for McDowell's argument. One is a broadly Kantian objection to McDowell's thesis whose significance McDowell seems to overlook. The other is a difficulty for what McDowell sees as the main objection to the response of Davidson, Quine and Sellars himself to the rejection of the Given. By way of introduction, however, I want to say a little more about the view that McDowell advocates.


McDowell speaks of experience as consisting in "openness to the world," but what does this mean? It is supposed to serve the traditional goal of allowing experience to provide a justificatory foundation for belief, but McDowell--wisely, of course--is not claiming that experience is infallible. He wants to allow the possibility of error, of cases in which appearances turn out to be mistaken--cases in which it our experience is as if P, even though it is not in fact the case that P. But how is it possible to allow this, without sliding back into the traditional sceptical worries?

As I interpret McDowell, something of the flavour of his reply is conveyed by the following analogy (which is mine, however, not McDowell's). If we consider paintings or photographs as pictorial representations of reality, the issue of error arises in a familiar way. Concerning any particular picture, there is always the possibility that it is not veridical: that the world is not as the picture represents it as being. In this context, general issues about the nature and accuracy of representation easily surface. Under what conditions are we justified in taking images to be veridical? And what does this mean? What is it for an image to `represent' a state of affairs? The possibility of error thus brings with it general issues about the whole notion of pictorial representation--of images being `about' the world.

Contrast this to the case in which we are dealing with paintings and photographs as works of art. Here the possibility of error also arises, but in a completely different way. With respect to any particular picture, there is a possibility that it may be a fake, and hence not the `genuine article.' However, this possibility sits quite happily with the fact that when an picture isn't a fake, we really do make contact with the real thing, with a genuine work of art. (There is no intermediary involved.) Unlike in the previous case, the relevant distinction is not between bad copies and good copies of reality, but between mere copies and originals.

As I read McDowell, he is suggesting that the right way to think of the way in which experience put us in touch with the world is, in effect, the second way. When it works, it gives us the real thing, directly: "When we are not misled by experience, we are directly confronted by a worldly state of affairs itself, not waited on by an intermediary that happens to tell the truth" (p. 143).

Explained in these terms, however, McDowell's view sounds like a rather Berkeleian form of idealism. The impression given is that we have `solved' the problem of the relationship between experience and the world by simply ignoring the (real) world, and reconstruing talk of the world in terms of what is immediately available to us in experience. McDowell denies that his view is a form of idealism, insisting that he takes it that experience does make contact with the real world, but this insistence has a familiar ring to it. Berkeley too insists that he agrees with the common man that there really are desks and chairs--provided, of course, that we interpret the common man in Berkeleian terms. We reach a familiar impasse, then. Both sides claim to do justice to common sense, but they seem to interpret common sense in very different ways.


My main objection to McDowell's project is that he seems to overlook, or at least to misread, a broadly Kantian strategy for breaking this impasse. One familiar way to take Kant's challenge to empiricism is something like this. Kant gives us a decisive reason to reject the naive Lockean picture of experience as copying--a process based on simple relations of similarity between things in the world and our mental impressions. The Kantian objection I have in mind is not based simply on Berkeley's point, viz. that the idea of similarity between mind and world is unintelligible. Rather, it depends on the realisation that the output of our perceptual apparatus is always processed: it depends on our own conceptual categories, as well as on the world.

So far, this may seem congenial to McDowell's project. Isn't the upshot of Kant's point that experience is already conceptual, which precisely is what McDowell wants to affirm, in arguing that the space of reasons extends all the way out to experience itself? Yes, but the Kantian point can also be read in the other direction: because experience is irremediably conceptual, it is too far in to provide the direct contact with the world that pre-Kantian (and non-Berkeleian) empiricists took for granted.

This point is hard to see if we simply identify the conceptual with the space of reasons, the realm of spontaneity, for there need be nothing reasoned or spontaneous about the way in which the mind imposes conceptual categories in perception. But the point is not merely terminological, of course. We could restrict the term `conceptual' to the space of reasons, but then we will simply have to find some other term to characterise what Kant taught empiricism about perception: namely, that it is never a matter of bare `copying.' What experience delivers is always a product not only of the world, but also of (contingent) features of our own constitution.

This application of Kant undermines the idea of direct contact between experience and reality in a familiar way. How can we accept both that colour vision puts us in direct touch with reality, and that the phenomenology of colour vision depends to a considerable extent on contingent features of our physical make-up, without falling into idealism: for example, into the view that colour-blind people inhabit a different reality? (It is no use appealing to the weight of numbers to say that colour blind people are mistaken, of course. Does the world change if the numerical balance shifts?)

The wider significance of this sort of Kantian consideration depends on how far it extends in our mental life. When Putnam suggests (e.g. in Reason, Truth and History, ch. 3) that Kant be interpreted as suggesting that all properties are secondary properties, he is suggesting, in effect, a completely global restriction on the tenability of the view that our thoughts make contact with pre-existing natural kinds in reality. Concerning McDowell's project, however, that wider issue can be left to one side: there are problems enough if the Kantian point is restricted to perception.

McDowell seems to overlook this difficulty. In contrasting his view to that of Sellars and Davidson, he says: "In my picture impressions are, so to speak, transparent. In the picture common to Sellars and Davidson they are opaque: if one knows enough about one's causal connections with the world, one can argue from them to conclusions about the world, but they do not in themselves disclose the world to one" (p. 145). As I have said, however, the rejection of the transparency of perceptual experience seems itself to be a Kantian lesson.

The explanation for McDowell's inattention to this point seems to lie in the belief--mistaken, in my view--that it rests essentially on Kant's transcendental perspective. In a section in which he compares his view to Kant's, McDowell says that if we reject Kant's transcendental story, and "restrict ourselves to the standpoint of experience itself, what we find in Kant is precisely the picture I have been recommending: a picture in which reality is not located outside a boundary that encloses the conceptual sphere" (p. 41). Once the transcendental frame is allowed, however,

the liberating thought cannot take proper shape. Once [Kant's supersensible reality] is in the picture, its radical independence of our thinking tends to present itself as no more than the independence that any genuine reality must have. The empirical world's claim to independence comes to seem fraudulent by comparison. We are asked to suppose that the fundamental structure of the world is somehow a product of subjectivity, in interaction with supersensible reality, which, as soon as it is in the picture, strikes us as the seat of true objectivity. But how can the empirical world be independent of us, if we are partly responsible for its fundamental structure? It does not help to be told that it is only transcendentally speaking that the fundamental structure of the empirical world is of our making. (p. 42)

It is true that McDowell recognises that the threat comes not so much from the transcendentalism as such, as from the fact that transcendental perspective is one example of what he calls a "sideways-on" perspective on mind and world. Throughout the book, in distancing himself from the position of Sellars and Davidson, he is concerned to show that a causal or naturalistic version of such a perspective cannot provide an adequate substitute for the normative justification we seek from the ordinary internal or participant's perspective. But when he touches on the possibility of replacing Kant's transcendental perspective with a naturalistic perspective, he seems to get the dialectic the wrong way round:

This picture tries to cast the realm of law in a naturalized version of the role Kant gives to the supersensible. But this is not how to correct what is unsatisfactory in Kant's thinking about the supersensible: keeping its basic shape, and merely naturalizing what lies beyond the conceptual. This way we lose the insight that Kant spoils by putting it in the framework of his talk of the supersensible; ... a way to make sense of empirical thinking as rationally answerable to the reality that it aims to be about. ... This kind of naturalism tends to represent itself as educated common sense, but it is really only primitive metaphysics. (p. 82)

This passage suggests that the naturalistic sideways-on perspective is a kind of life-support machine for the ailing remnants of Kant's transcendental perspective--a machine we should be more than happy to turn off, when we notice the attractions of McDowell's picture, in which experience puts us directly in contact with reality. If this were so then the naturalistic perspective would threaten McDowell's project only so long as it kept alive the project's sickly Kantian rival. As I have explained, however, the threat to McDowell's picture is much more direct. It is threatened by the observation--naturalistic, note, though Kantian in spirit--that in virtue of the processing that goes on in perception, experience is necessarily removed from reality.

While the transcendental perspective may be dismissed as ultimately incoherent--the product of an untenable metaphysical externalism--this naturalistic analogue is much more tenacious. In the end, indeed, it isn't "primitive metaphysics" but simple physiology which teaches us that Kant was right: what we get from our sensory apparatus depends on quite contingent features of our physical construction, as well as on the nature of the external world. Arguably, the same is true of our entire conceptual apparatus, but certainly it is true of experience. This product of a sideways-on scientific perspective is not a kind of comatose version of transcendentalism, but a plausible first-order theory about the way in which our brains are linked to their environment. Nor is it a kind of philosophical opening bid, which we can abandon on the grounds that it causes problems elsewhere in philosophy. To all intents and purposes it is a fact of modern life, within the constraints of which philosophy must operate.

This is not to say that this scientific perspective can, or should, replace the internal, hermeneutical, perspective we have on ourselves and our fellows. On this point, I agree completely with McDowell, who warns us against "aspiring to a sideways-on understanding of our own thinking, which we take to be the condition of someone else who understands us," (p. 35) and later denies that "it is in general possible to give sideways-on accounts on concepts" (p. 168). There is a kind of understanding of a concept used by another speaker which we only gain when, from the inside, we see that concept as equivalent to one of our own. However, this is not to say that the naturalistic sideways-on perspective is illegitimate, or mistaken so far as it goes--it is simply to say that when the project is interpretation, it doesn't go all the way. Interpretation doesn't invalidate sideways-on description, but simply supplements it. (Better still, it is simply in a different line of work.)

McDowell himself refers to the sideways-on perspective as the starting point for radical interpretation, but says that the radical interpreter "finishes with a theory whose point is exactly that it is not from sideways on: a theory that enables her to capture some of her subjects' relations to the world from their own point of view, though in her terms rather than theirs" (pp. 152-53). Indeed, but the project only goes through to the extent that the interpreter and her subjects are sufficiently similarly constructed so that they share points of view. If they have radically different visual systems, for example, the terms in which the subjects describe visual experience may be simply untranslatable. In this way, the Kantian dilemma McDowell overlooks emerges from within the project of radical interpretation: either we allow that the worlds we and our subjects inhabit are different, which leads to an implausible idealism; or we allow that in virtue of the dependence of perception (or cognition generally) on such contingent features of our physical make up, experience doesn't put us in direct touch with the world. True, a third possibility would be to deny that the subjects in question have any concepts in the untranslatable region of their (apparently) linguistic practice. This seems a trifle uncharitable, however, particularly as (i) the situation seems symmetric, so that they would be equally entitled to say the same about us, and (ii) their experts assure us that they do have concepts, in what--leaving the problem area to one side--we take to be well-translated subject-ese.

The fact that the difficulty arises in this natural way from within the project of radical interpretation should serve to counter a possible objection, namely that the supposed problem rests on an illegitimate externalist standpoint--a naturalistic `view from nowhere,' which abstracts away from the very contingencies to which the argument claims to appeal. On the contrary: the problem rests on something that we learn about ourselves from within, as we reflect--from an ordinary first-order standpoint--on the genealogy of our own concepts and practices. It is true that if the point is to be globalised, in the way that Putnam suggests--if all concepts are secondary in this sense--then the issue as to whether the view can be formulated without recourse to the idea of an unconceptualised `world in itself' becomes a question of some delicacy. For present purposes, however, when it is simply the status of perceptual experience that is up for grabs, this difficulty does not arise. Speaking in the language of natural science, we can describe the ways in which the view that experience gives us of the world described by natural science is conditioned by our own physical contingencies. This should suffice to convince us that experience doesn't provide direct access to that world, even though the question as to whether science itself is subject the similar contingencies--and, indeed, the question whether that view is even coherent--remain open.

To sum up: McDowell wants to argue that "reality is not located outside a boundary that encloses the conceptual" (p. 44). My point is that so long as we acknowledge the broadly Kantian discovery that the conceptual depends on contingent features of our physical make up, McDowell's view is bound to lead to idealism--to the view that reality, too, depends on such contingencies. Unless we are prepared to embrace such idealism, the upshot is not merely, as McDowell notes, that "there is no guarantee that the world is completely within reach." It is that there is a guarantee that at least in experience, the world is not within reach.


I want to finish with a briefer point, in defence of Quine, Sellars and especially Davidson. McDowell argues that having rejected the Given, and without his substitute conception of experience as openness to reality, these philosophers fall victim to the other horn of the Kantian dilemma. Without a basis in experience, "concepts are blind"--in other words, beliefs lose not only their claim to justification, but their claim to empirical content--their claim to be `about' the world at all. McDowell says this about Davidson, for example:

Davidson manages to be comfortable with his coherentism, which dispenses with rational constraint on thinking from outside it, only because he does not see that emptiness is the threat. He thinks that the only point of wanting a rational connection between intuitions and thoughts is reassurance that we are justified in endorsing the thoughts, as if we could take for granted in any case that they are thoughts, that they possess content. But if we do not let intuitions stand in rational relations to them, it is exactly the possession of content that is put in question. When Davidson argues that a body of beliefs is sure to be mostly true, he helps himself to the idea of a body of beliefs, a body of states that have content. And that means that, however successfully the argument might work on its own terms, it comes too late to neutralize the real problem ... . (p. 68)

However, I fail to see how this objection can get any grip within Davidson's picture, or, for that matter, within the subset that McDowell shares with Davidson. It would be easy enough to raise the objection from the kind of transcendental side-on perspective that McDowell and Davidson both reject: from that perspective, it would be much like a brain-in-a-vat point--just the suggestion that our (apparent) beliefs might not genuinely connect with the world. From that perspective, different thinkers might be observed to engaged in mutual interpretation, on Davidsonian lines, while all being equally disconnected from reality itself.

If we reject this transcendental perspective, however, what sense can we give to the idea of being disconnected? What more is there to the notion of belief about the world than is already built in to the notion of belief itself--i.e., built in to a notion whose main tie is to the notion of interpretation. What more is there to believing of the world that P, than being correctly interpreted as believing that P? In arguing that it is incoherent that all of our beliefs could be false, Davidson says, in effect, that it is a consequence of what we mean by `belief' that most of our beliefs are true--i.e., roughly, that it follows from what we mean by `believes that P' that :

(I have expressed the conclusion in these terms in order to emphasise how little role truth really plays here.)

Similarly, it seems appropriate for Davidson to say that it is a consequence of what we mean by belief that there is nothing more to really believing of the world that P than being correctly interpreted as believing that P, with the confines of the interpretative stance. In the end, the reason the transcendental stance is incoherent for Davidson seems to be that it requires that there be something else that could be meant by the question as to whether a belief is about the world. To put it another way, far from lacking a notion of content such that our mental states can be held to have content, to be `about' the world, Davidson's view claims to offer the only coherent notion of such a content. This claim might be challenged, of course, but only by someone who prepared to reject the view that content can only be ascribed from the interpretative stance--someone prepared to defend the ascription of content from some other, sideways-on stance. Many people in contemporary philosophy do defend such views, of course: those who seek naturalistic reductions of semantic content, for example. But McDowell is dismissive of these views, and of the sideways-on stance in general, at least for these purposes. As a result, I cannot see how it is open to him to claim that Davidson's position is deficient in this way.

McDowell begins Mind and World with a lucid account of the fundamental dilemma seemingly posed by the notion of the Given: on the one hand, no conception of experience seems able to do what the Given is required to do; on the other hand, if we reject the Given then we seemed to be faced with the conclusion that our beliefs are not only unjustified but empty of empirical content. But I think he fails to see that to follow Davidson (or, I think, Sellars or Quine) is not simply to opt for the second horn, while leaving the original dilemma in place. It is to argue that there really is no dilemma, in the sense that there is nothing that we really lose if we reject the Given. What we seem to lose turns out to be something that we never really had any need for or entitlement to in the first place, namely a notion of content distinct from that available to us from the interpretative stance.[1]

Huw Price
University of Sydney

1. I am grateful to Paul Redding for some illuminating discussions of Mind and World, and helpful comments on an earlier draft of this review.