D. C. Dennett’s Content and Consciousness is an extraordinarily interesting and original book, and one which will raise the level of current discussion in the philosophy of mind. In this note, however, I should like to criticize one central thesis of the book – that puzzles in the philosophy of mind, and notably those about incorrigible knowledge, can only be cleared up by “an analysis of phenomena at the sub-personal level.” This thesis seems to me dead wrong, and I hope to show, by an analysis of what Dennett says about direct awareness and about incorrigibility, that the revival of Putnam’s notion of “functional state” (on which Dennett depends heavily in his defense of the thesis) is not a profitable strategy.

Dennett’s thesis is, roughly, that only by opening up the person (who has been treated as a sealed “black box” in traditional philosophy of mind) and saying something about his internal “functional organization” can we make the distinctions we need. I want to begin to cast doubt on this by taking up what Dennett says about the need to distinguish two senses of awareness. He regards such a distinction as the beginning of wisdom in this area of philosophy, and I think he is right. But I also think he makes the distinction in the wrong way.

One criticism which philosophers often make of their opponents is that they are “reductionists.” One gathers that a typical philosophic error is to infer from “X has the property Y” to “X is nothing but Y,” or to infer from “X is analogous to Y in respect to Z” that “X and Y are indistinguishable.” Another form of the same criticism is that philosophers take some feature of experience or reality or language to be paradigmatic, and are not then able to account for features which depart from this paradigm. Yet no such criticism can, without absurdity, object to the general procedure which these inferences illustrate. All rational inquiry is reductionist; all abstract thought takes selected aspects of a subject-matter as paradigmatic and ignores other aspects. Thought is reductionist or nothing, and the criticism only makes sense if it is narrowed down. When it is narrowed down, it usually turns out to be the claim that a reduction of X to Y is illegitimate because the very process of reducing presupposes some X that is not reduced. This claim, which we shall call “the appeal to self-referential consistency,” is the topic of this chapter. Our aim is to see what can be done to specify a point of diminishing returns in the reductive process, and thus to locate the limits of reductionism.

Introductory

Current controversies about the Mind-Body Identity Theory form a case-study for the investigation of the methods practiced by linguistic philosophers. Recent criticisms of these methods question that philosophers can discern lines of demarcation between “categories” of entities, and thereby diagnose “conceptual confusions” in “reductionist” philosophical theories. Such doubts arise once we see that it is very difficult, and perhaps impossible, to draw a firm line between the “conceptual” and the “empirical,” and thus to differentiate between a statement embodying a conceptual confusion and one that expresses a surprising empirical result. The proponent of the Identity Theory (by which I mean one who thinks it sensible to assert that empirical inquiry will discover that sensations (not thoughts) are identical with certain brain-processes) holds that his opponents’ arguments to the effect that empirical inquiry could not identify brain-processes and sensations are admirable illustrations of this difficulty. For, he argues, the classifications of linguistic expressions that are the ground of his opponents’ criticism are classifications of a language which is as it is because it is the language spoken at a given stage of empirical inquiry. But the sort of empirical results that would show brain-processes and sensations to be identical would also bring about changes in our ways of speaking. These changes would make these classifications out of date. To argue against the Identity Theory on the basis of the way we talk now is like arguing against an assertion that supernatural phenomena are identical with certain natural phenomena on the basis of the way in which superstitious people talk. There is simply no such thing as a method of classifying linguistic expressions that has results guaranteed to remain intact despite the results of future empirical inquiry. Thus, in this area (and perhaps in all areas) there is no method which will have the sort of magisterial neutrality of which linguistic philosophers fondly dream.

My title suggests that there are two well-defined schools of philosophers – the “analytic” and the “metaphysical.” I am aware of the sort of over-simplification which such labeling produces, but I think that it is sometimes justified. Despite the obvious differences between, say, Goodman, Wisdom, and Austin, and the equally obvious differences between Whitehead, Heidegger, and Maritain, some important metaphilosophical disagreements clearly exist between a position which is the least common denominator of the views of philosophers who have taken “the linguistic turn,” and a position which is the least common denominator of the views of those who have not. The labels “analyst” and “metaphysician” are at least brief, and using them is no more unfair to one side than to the other.

As my title also suggests, I find it very difficult to formulate these metaphilosophical disagreements. They are usually conceived as centering around a disagreement about what method to pursue. My principal thesis in this chapter will be that none of the familiar ways of characterizing such alternative methods are successful.

Introduction

In this chapter, I wish to argue for the following theses:

(A) None of the arguments about the possibility of a private language or about the privacy of sensations and thoughts which Wittgenstein advances in the Philosophical Investigations provide good reason for doubting

1. (a) that words like “toothache” and “pain” are the names (in a nontrivial sense) of sensations which people sometimes experience, or

2. (b) that when I assert truly “I have a toothache” or “I am in pain,” I am describing the state of my consciousness, or

3. (c) that when I assert of another person “He has a toothache” or “He is in pain” I claim that he is experiencing the same sort of sensation that I do when I have a toothache or am in pain.

(B) None of these arguments give good reasons for rejecting as senseless the claim that “sensations are private.”

(C) None of these arguments give good reasons for rejecting as senseless the claim that “I know that I am in pain because I feel it.”

I have drafted these theses with an eye to recent discussions of Wittgenstein’s views about the privacy of sensations, and in the belief that certain confusions committed by Wittgenstein or his interpreters – notably between “privacy” in the sense of “susceptibility to privileged access” and in the sense of “incommunicability” – have led sympathetic commentators to attribute unnecessarily paradoxical views to him, and hostile critics to attack him by attacking these paradoxes.

Richard Rorty published six volumes of his own selections from the papers he wrote between 1972 and 2006. Together with his two monographs, these now provide the main source for his views, and scholars may debate at leisure whether some of the papers Rorty did not select have been unjustly neglected as a consequence. However, it seems clear to us that there is a large body of Rorty’s work which has most certainly been unjustly neglected: the early work. For before he wrote “The World Well Lost,” the 1972 paper that opens Consequences of Pragmatism, Rorty was already an influential philosopher. Many still fondly recall this first phase of his career, and we have often heard say that Rorty was an excellent analytic philosopher back in the 1960s and early 1970s before he read some continental philosophy, became a postmodernist, and consequently went off the rails. Almost everything about this latter view is wrong, but there is nevertheless a kernel of truth to it. This is that Rorty wrote some classic papers in the 1960s and early 1970s which, for better or worse, would probably be of more interest to the average contemporary philosopher working on mind or language than his mature manifesto of replacing objectivity with solidarity and metaphysics with literature. But given that Rorty’s thinking never did radically change direction, there is also plenty to be found here for those interested in the mature Rorty. To neglect the papers in this volume is, in short, to neglect innovative ideas which retain their interest and relevance, as well as a significant part of the career of a highly significant thinker – the part you might well find you like best. The early Rorty has been languishing in old journals, obscure collections and faded memories for too long!

There are three theses connected with Quine’s views about the indeterminacy of translation which should be clearly distinguished from one another. These are:

1. (1) A person’s dispositions to accept sentences do not determine a unique interpretation of those sentences.

2. (2) The notions of meaning, propositional attitudes, etc., do not possess the explanatory power often attributed to them by philosophers. (In particular, one cannot explain the truth of a sentence by saying it is “true by virtue of meaning,” nor can one “explain why a person accepts a sentence by saying he accepts a proposition which the sentence expresses.”)

3. (3) “Though linguistics is of course a part of the theory of nature, the indeterminacy of translation is not just inherited as a special case of the underdetermination of our theory of nature. It is parallel but additional.”

I doubt that (1) is still a subject of debate. Quine’s opponents have by now been argued around to admitting that alternative systems of analytical hypotheses, producing non-equivalent translations, will predict dispositions to accept sentences equally well. (2) is a much more important, interesting, and controversial matter, and one which requires more argument than merely citing (1). The arguments for (2) consist mainly in exhibiting the poverty of the “explanations” provided by an appeal to the notions in question. (3) is often run together with (2). Thus, when replying to Chomsky’s doubts about (3) Quine himself seems to merge the two theses by saying that grasping the difference between the general “under-determination of our theory of nature” and the special indeterminacy of translation will lead to “a change in prevalent attitudes toward meaning, idea, proposition,” and that the difference escapes recognition “because of the uncritical persistence of old notions of meaning, idea, proposition.” The suggestion seems to be that accepting (2) should lead to accepting (3), or vice versa, or both.

There was a time when the slogan “Let us construct an extensional language which is adequate to say everything!” stirred the hearts of empiricists. It seemed evident that all that was required to accomplish the project was energy and discrimination, for surely to deny that the job could be done would be somehow tantamount to doubting the cardinal tenets of empiricism. In the course of some twenty or thirty years, however, extensionalism has fallen into disrepute. On the one hand, there is a history of failure in the attempts actually to get the job done. On the other, there is the suggestion that empiricism was led astray by its delight in a new toy, and that the use of an extensional calculus has nothing in particular to do with the tradition of Locke and Hume. American philosophers have been told by their British colleagues that they have picked a quite arbitrary straight-jacket in which to confine language, and have thereby created for themselves a host of pseudo-problems about areas where the straight-jacket pinches. The artificiality of suggested extensionalistic reconstructions of these areas of discourse has made many philosophers receptive to suggestions that they can be bona fide empiricists without being “reductionists.”

The purpose of this chapter is to show that there are indeed internal relationships between empiricism and extensionalism, to explain what these relationships are, and to show that neither empiricism nor extensionalism is internally related to reductionism, in the sense in which this term is usually understood. In other words, I want to disentangle three theses which empiricists have tried to establish and discuss what the history of recent empiricism teaches us about the hopes for establishing them.

In this chapter I argue, first, that various “topic-neutral” translations of mentalistic statements propounded by materialists are unsatisfactory in that they do not catch the specifically “mentalistic” element in these statements. I then go on to argue that to isolate this element one needs to insist on the incorrigibility of first-person reports of mental states. Finally, I consider whether this insistence is an obstacle to materialism.

We may begin by recalling that the origin of the attempt at “topic-neutral” translations of mentalistic statements was an attempt to avoid what we may call the “irreducible-properties objection” to the thesis that mental states are identical with brain states. This objection says that, even if the identity thesis frees us from nomologically dangling entities, it cannot free us from nomologically dangling properties – viz., those properties by which we originally identified the mental entities as such. Thus, for example, a sensation of yellow has the property “of yellow,” and the thought that p has the property “that p”; but it seems to make no sense for any brain-process to have either sort of property. So these properties seem irreducible. J. J. C. Smart originally tried to get around this objection for the case of sensations by saying that “I am having a sensation of yellow” was equivalent to (or could roughly be paraphrased as) “Something is going on in me like what is going on when I see something yellow.” More recently, D. M. Armstrong has employed the same technique in a program of translating (or paraphrasing) all statements ascribing mental states as statements containing the subject term “a state apt for the production of the following sorts of behaviour.”

Realist criticism of linguistic philosophy

Among contemporary realistic philosophers there is a tendency to see the history of philosophy from Descartes to Wittgenstein as one continuous process of garnering the wages of sin. One can identify the original sin as the representative theory of perception, as the confusion of noetic being with being as such, or in various other ways. In any case, the dialectical progression which in the course of three centuries has led philosophers away from things to ideas, and away from ideas to words, presents itself as the enactment of an agonizing drama of self-destruction.

Given the initial Cartesian sundering of cognition from the object of cognition, it was only to be expected that epistemology should claim priority over metaphysics. Once this happened, it was only a matter of time before metaphysics shriveled away and died. In due course, epistemology, finding itself unable to function as an autonomous discipline when deprived of metaphysical support, was pushed aside by its ungrateful child, linguistic analysis. The ultimate, foredoomed, result of this development is, according to such neo-Thomist descriptions of recent philosophy as that of Oliver Martin, the total confusion of the theoretic and instrumental orders, and the consequent replacement of philosophic controversy by ideological conflict. Linguistic analysis, Martin says, though not itself an ideology, is at the service of ideology through being “one of the many non-philosophical forms of anti-philosophy.” Probably, Martin continues, “It would not exist were it not for the fact that professors of philosophy who no longer believe in philosophy must do something if they are to continue to receive a salary for doing that which they profess to be impossible.”

The acclaim and controversy that has surrounded Richard Rorty’s later work, starting with Philosophy and the Mirror of Nature in 1979, has tended to eclipse the influence of his major contributions to analytic philosophy of mind in the 1960s and 1970s. Like Hilary Putnam, whose papers from that era remain classics in spite of their author’s abandonment of them, Rorty shaped the field for subsequent decades with a series of insightful, constructive, imaginative papers harnessing the insights of Sellars and Quine, and developing his own distinctive varieties of pragmatism and naturalism. Rorty’s version of eliminative materialism was both the first and the subtlest, and anybody interested in defending any variation today would be well advised to mine his work for insights. The same can be said for his delicate and undoctrinaire treatment of functionalism.

I would venture to say that today’s combatants in the skirmishes over property dualism, supervenience, mental causation, and their subsidiary issues would find that most of their moves and countermoves were anticipated and preempted by Rorty’s discussions more than forty years ago. “Mind-Body Identity, Privacy, and Categories” (1965), “Incorrigibility as the Mark of the Mental” (1970), “In Defense of Eliminative Materialism” (1972), and “Functionalism, Machines, and Incorrigibility” (1972) ought to be required reading for all aspiring philosophers of mind today, as they were for us when they first appeared, and they are only the best known of his works from that period. Fortunately, Rorty wrote admirably clear and vivid philosophy which holds up well at this remove in time, so this volume will be a joy to read, both for those who remember the impact these essays had when they were published and for those who know Rorty only as the lightning rod of the later culture wars.