Mind and World

With a New Introduction by the Author

John McDowell

• Publisher: Harvard University Press
• Year: 1996
• ISBN:
  9780674417908
• Topic: Philosophy
• Subtopic: Philosophy History & Theory

Modern philosophy finds it difficult to give a satisfactory picture of the place of minds in the world. In Mind and World, based on the 1991 John Locke Lectures, one of the most distinguished philosophers writing today offers his diagnosis of this difficulty and points to a cure. In doing so, he delivers the most complete and ambitious statement to date of his own views, a statement that no one concerned with the future of philosophy can afford to ignore.John McDowell amply illustrates a major problem of modern philosophy—the insidious persistence of dualism—in his discussion of empirical thought. Much as we would like to conceive empirical thought as rationally grounded in experience, pitfalls await anyone who tries to articulate this position, and McDowell exposes these traps by exploiting the work of contemporary philosophers from Wilfrid Sellars to Donald Davidson. These difficulties, he contends, reflect an understandable—but surmountable—failure to see how we might integrate what Sellars calls the "logical space of reasons" into the natural world. What underlies this impasse is a conception of nature that has certain attractions for the modern age, a conception that McDowell proposes to put aside, thus circumventing these philosophical difficulties. By returning to a pre-modern conception of nature but retaining the intellectual advance of modernity that has mistakenly been viewed as dislodging it, he makes room for a fully satisfying conception of experience as a rational openness to independent reality. This approach also overcomes other obstacles that impede a generally satisfying understanding of how we are placed in the world.

PART I

Davidson in Context

1. In the lectures, I use Davidson’s coherentism exclusively as a foil to the view of experience that I recommend. Here I want to place Davidson’s thinking about the epistemology of empirical thought in a historical context, constituted by a strand in the recent development of the American pragmatist tradition. I hope this will bring out something the lectures obscure: the extent to which I can count Davidson as an ally rather than an opponent.

2. The two dogmas of empiricism that W. V. Quine attacked in his celebrated essay of that title were, first, that there is a “fundamental cleavage” (p. 20) between the analytic, in the sense of statements true by virtue of meaning alone, and the synthetic, in the sense of statements whose truth is dependent not only on meaning but also on the world; and, second, that “empirical significance” can be parcelled out statement by statement among the body of statements that express our view of the empirical world.

What we ought to say instead of the second dogma, according to Quine, is this: “The unit of empirical significance is the whole of science” (p. 42: this is the source for the way I have put the second dogma). An alternative formulation is this: “our statements about the external world face the tribunal of experience not individually but as a corporate body” (p. 41). If these are two formulations of the same thought, Quine is implicitly glossing empirical significance in terms of being subject to the tribunal of experience. That makes it look as if Quine’s conception of “empirical significance” corresponds to the Kantian conception of empirical content, or bearing on the empirical world, that I defend in the lectures. According to that conception, the fact that something, say a belief, or—more congenially to Quine—a whole world-view, bears on the world, in the sense of constituting a stand taken as to how things are, depends on its being vulnerable to the world for a verdict on its acceptability; and the verdict can be delivered only through experience.

The rejected first dogma claims that the truth of a synthetic statement depends on two factors, meaning and the world; an analytic statement is one for which the “world” factor is null. Now Quine’s positive picture retains this duality of factors on which truth depends. He says (p. 36): “It is obvious that truth in general depends on both language and extralinguistic fact.” His claim is not that there are not those two factors, but just that we cannot separate them out statement by statement. In the context of an apparently uncontentious empiricism, the “world” factor is just the answerability to experience that is summed up as “empirical significance”. So Quine can reformulate the “obvious” dependence of truth on both language and extralinguistic fact, in an expression of the thesis that rejects the second dogma, by saying (p. 41): “Taken collectively, science has its double dependence upon language and experience; but this duality is not significantly traceable into the statements of science taken one by one.”

This validates an impression that is anyway given by the structure of Quine’s essay: that rejecting the second dogma is the fundamental point. Quine’s positive thinking in the essay is encapsulated in the thesis that the unit of empirical significance is the whole of science. With the duality of factors retained, the first dogma figures as something that can be correct only if the second dogma is, so that rejecting the second can suffice to reject both. The first dogma is the thesis that there are true statements that are analytic in the sense that for them the “world” factor—the factor of dependence on experience, “empirical significance”—is null. If “empirical significance” cannot be shared out among individual statements anyway, that undermines the very idea of a statement with no “empirical significance”. “No empirical significance of its own” could only be a special case of “some empirical significance of its own”. If it makes no sense to suppose a particular statement has its own positive quantity of “empirical significance”, a determinate share of the “empirical significance” of the total world-view in whose expression the statement figures, then it equally makes no sense to suppose there might be statements for which the quantity is zero.¹

3. As I said, Quine’s positive picture, his “empiricism without the dogmas” (p. 42), retains a counterpart to the duality that figures in spelling out the idea of analyticity. Truth—which we must now think of as primarily possessed by a whole world-view—depends partly on “language” and partly on “experience”. “Language” here labels an endogenous factor in the shaping of systems of empirical belief, distinguishable—though only for whole systems—from the exogenous factor indicated by “experience”. When we acknowledge this exogenous factor, we register that belief is vulnerable, by way of “the tribunal of experience”, to the world it aims to be true of. So “language” figures, in Quine’s holistic context, as a counterpart to “meaning” as “meaning” figured in the now debunked contrast between “true by virtue of meaning alone” (true in a way that involves no vulnerability to the world) and “true by virtue of both meaning and the way the world is”.

“Empirical significance” registers the exogenous factor in this contrast, the answerability to something outside the system. And the nearest thing in Quine’s positive thinking to the old notion of meaning is “language”, which stands on the other side of the retained duality, as the endogenous factor. “Empirical significance” is not meaning as it figured in the idea that there might be statements true by virtue of meaning alone; and it is not a functional descendant of meaning, so conceived, in the novel environment of Quine’s holism, but a functional descendant of what stood precisely in contrast to meaning in the old version of the duality.

The fact that Quine’s “empirical significance” is one side of a holistic counterpart to the old duality means that, in spite of the convergence between Quine’s talk of facing the tribunal of experience and my talk of a rational vulnerability to intuitions, we cannot gloss Quine’s “empirical significance” in terms of bearing on the empirical world in the sense in which I use that phrase: what stand one takes on how things are in the world when one adopts a belief or a world-view. For Quine, the two factors are distinguishable, even though only for whole systems, and that means that the “empirical significance” of a world-view cannot amount to its empirical content in the sense of how, in adopting the world-view, one takes things to be in the empirical world. That requires the other factor, the endogenous one, also.

So far, this might be merely a terminological oddity about Quine’s use of the phrase “empirical significance”. It is Quine’s own point that “empirical significance” does not amount to content, in the sense of what stand one takes on how things are in the empirical world. In the thesis that translation is indeterminate, which is meant to elaborate the moral of “Two Dogmas”, his aim is to stress “the extent of man’s conceptual sovereignty” in the formation of world-views:² that is—to put it in a way that brings Quine into explicit contact with Kant—the extent to which the content of world-views is a product of spontaneity operating freely, uncontrolled by the deliverances of receptivity. And from Quine’s point of view, it is a merit of the notion of “empirical significance” that it stands on the wrong side of the descendant duality to be a descendant of the old notion of meaning. Quine is no friend to the old notion of meaning, and the descendant notion, the notion of “language” as the endogenous factor, bound up as it is with “man’s conceptual sovereignty”, retains in Quine’s thinking some of the intellectual dubiousness of its ancestor. In contrast, “empirical significance” is an intellectually respectable notion, because it is explicable entirely in terms of the law-governed operations of receptivity, untainted by the freedom of spontaneity. To put it in a more Quinean way, “empirical significance” can be investigated scientifically. “The extent of man’s conceptual sovereignty”, the extent to which the content of a world-view goes beyond its “empirical significance”, is just the extent to which such a notion of content lies outside the reach of science, and therefore outside the reach of first-rate intellectual endeavour.

It is not just a verbal point that “empirical significance” is on the wrong side of the duality to be a descendant of the notion of meaning. We have to discount the rhetoric that makes it look, at first sight, as if Quine’s notion corresponds to the Kantian notion of empirical content. Quine speaks of facing the tribunal of experience, which seems to imply a vulnerability to rational criticism grounded in experience. But he conceives experience as “the stimulation of . . . sensory receptors”.³ And such a conception of experience makes no room for experience to stand in rational relations to beliefs or world-views. The cash value of the talk of facing the tribunal of experience can only be that different irritations of sensory nerve endings are disposed to have different impacts on the system of statements a subject accepts, not that different courses of experience have different rational implications about what system of statements a subject ought to accept. In spite of the juridical rhetoric, Quine conceives experience so that it could not figure in the order of justification, as opposed to the order of law-governed happenings. This is all of a piece with the idea that “empirical significance” is a topic for natural science.

At one point in “Two Dogmas” (p. 43), Quine writes: “Certain statements . . . seem peculiarly germane to sense experience—and in a selective way: some statements to some experiences, others to others. . . . But in this relation of ‘germaneness’ I envisage nothing more than a loose association reflecting the relative likelihood, in practice, of our choosing one statement rather than another for revision in the event of recalcitrant experience.” The only connection he countenances between experience and the acceptance of statements is a brutely causal linkage that subjects are conditioned into when they learn a language. It is not that it is right to revise one’s belief system thus and so in the light of such-and-such an experience, but just that that revision is what would probably happen if one’s experience took that course.⁴ Quine conceives experiences so that they can only be outside the space of reasons, the order of justification.

It may still seem that Quine is subject at most to criticism of his rhetoric. But his talk of facing the tribunal of experience is not just a slip, which we could easily excise; it has roots that go deep in Quine’s thinking. We cannot simply register that “empirical significance” is not genuinely a kind of significance if experience is not in the order of justification, and leave the substance of Quine’s thinking intact.

If experience is not in the order of justification, it cannot be something that world-views transcend or go beyond. But Quine needs that for his talk of “the extent of man’s conceptual sovereignty”. What a world-view might transcend, so that adoption of it might be an exercise of spontaneity or “conceptual sovereignty”, is evidence that tells less than conclusively in favour of it. But if experience plays only a causal role in the formation of a world-view, not a justificatory role, then it does not serve as evidence at all.

And if experience does not stand to world-views as evidence to theory, that puts in question the capacity of Quine’s picture to accommodate world-views at all. It is true that Quine wants the idea of a world-view to stand exposed as intellectually second-rate. But he does not want to abandon it altogether. That would be to abandon the point he wants to make by talking of “the extent of man’s conceptual sovereignty”. There would be nothing for the thesis of the indeterminacy of translation to be about if instead of talking about arriving at world-views we had to talk exclusively about acquiring propensities to feel comfortable with certain vocalizations. Now the idea of an interaction between spontaneity—“conceptual sovereignty”—and receptivity, which is Kantian so far as it goes, can so much as seem to make room for the idea of adopting a world-view only if the deliverances of receptivity are understood to belong with the adopted world-view in the order of justification. If we try to suppose that exercises of “conceptual sovereignty” are only causally affected by the course of experience, and not rationally answerable to it, there is nothing left of the idea that what “conceptual sovereignty” produces is something that is about the empirical world, a stance correctly or incorrectly adopted according to how things are in the empirical world. And if we lose that, there is nothing left of the idea that what is operative is “conceptual sovereignty”. The notion of a world-view, formed in an exercise of “conceptual sovereignty”, is not just the notion of a perturbation produced jointly by impacts from the world and by some force operating from within the subject, in a way that is partly (but only partly) determined by those impacts.⁵

If we clean up Quine’s formulations by eliminating the juridical rhetoric, we deprive him of the very idea of “conceptual sovereignty”, and the effect is to threaten the idea that we are in touch with the empirical world at all. It is not that this reading makes Quine suggest we may be wildly wrong about the world, like an old-fashioned philosophical sceptic. But without the “tribunal” rhetoric and its companion idea of “conceptual sovereignty”, which have emerged as strictly illicit by Quine’s own lights, he puts in question the very idea that we have the world in view at all, that anything that we do constitutes taking a stand, right or wrong, even wildly wrong, on how things are in it.⁶

4. The awkward position of experience in Quine’s thinking has implications for an otherwise attractive reading of what he is doing when he rejects the first of the two dogmas. In the lectures, I derive from Sellars, and trace to Kant, a rejection of the idea that something is Given in experience, from outside the activity of shaping world-views. The attractive reading has Quine making a counterpart point, rejecting the idea that something is Given from within the very structure of the understanding.⁷

Sellars says: “empirical knowledge, like its sophisticated extension, science, is rational, not because it has a foundation but because it is a self-correcting enterprise which can put any claim in jeopardy, though not all at once.”⁸ We must think of empirical rationality in a dynamic way, in terms of a continuing adjustment to the impact of experience.

To reject the idea of an exogenous Given is to follow this prescription in part. It is to refuse to conceive experience’s demands on a system of beliefs as imposed from outside the activity of adjusting the system, by something constituted independently of the current state of the evolving system, or a state into which the system might evolve. The required adjustments to the system depend on what we take experience to reveal to us, and we can capture that only in terms of the concepts and conceptions that figure in the evolving system. What we take experience to tell us is already part of the system, not an external constraint on it.

That is to say that nothing is Given from outside the evolving system of belief. The counterpart claim—the claim that nothing is Given from inside the understanding, the intellectual capacity that is operative in the continuing activity of shaping the system—is in Sellars too; it is implicit in the remark about the rationality of science that I have just quoted. It is true that when Sellars debunks the Myth of the Given in detail, he focuses on the supposed external constraint, but he begins “Empiricism and the Philosophy of Mind” by saying (pp. 253–4) that the idea of something Given in experience is a specific application of a conception that is much more general. It can easily seem that rejecting an endogenous Given requires us to say what Sellars says in the remark I have quoted: that any of our beliefs, including beliefs about structures that must be instantiated in intellectually respectable belief systems—the beliefs that implicitly or explicitly govern adjust...

Table of contents

• Cover
• Title
• Copyright
• Contents
• Preface
• Introduction
• Lecture I. Concepts and Intuitions
• Lecture II. The Unboundedness of the Conceptual
• Lecture III. Non-conceptual Content
• Lecture IV. Reason and Nature
• Lecture V. Action, Meaning, and the Self
• Lecture VI. Rational and Other Animals
• Afterword
• Part I. Davidson in Context
• Part II. Postscript to Lecture III
• Part III. Postscript to Lecture V
• Part IV. Postscript to Lecture VI
• Index