Conceptual Revision and the Identity Theory

are thoughts and sensations nothing over and above brain processes?


An excerpt from the dissertation submitted in 1968 to the graduate school at Yale University by Gilbert Scott Markle, in partial satisfaction of the degree requirements for a PhD, in Philosophy. Markle was already teaching at that time at Clark University, in Worcester, Massachusetts.

Editor's Note:


    By an 'identity theory" of mind and body we will understand any of those theories which hold that, in a suitably specified sense, the items of our mental life such as thoughts, sensations, urges, anxieties, and so forth, are identical with certain spatio-temporally describable states obtaining in our brains, or, more generally, in our central nervous systems.  A typical claim of psychophysical identification would be of the form

    "Searing pain IS an x-y-z stimulation of thalamic C-fibers."

    Although identity theories began, with Hobbes and Spinoza, as explicitly metaphysical accounts of the relation holding between the mental and the physical, current formulations of the identity theory would present themselves as purely scientific hypotheses. Contemporary identity theorists maintain that correlated psychic and physical events are empirically identical — that their identity is a simple matter of fact.  In this way, it is suggested that empirical science — not metaphysics — is able to provide us with an explanation, or ground, for the puzzlingly reliable correlations obtaining between certain psychic and physical events. The identity theory would explain these correlations in pointing out that entities of the latter sort have been simply correlated with themselves. A more simple explanation could hardly be desired.

    To say, however, that a claim of psychophysical identity is a purely scientific or empirical claim, is to make things much too simple. For, the proposed identifications are not just factual identifications of the sort expressed by the sentence "Scott is the author of Waverly," but are also interpretative in character. They give expression to the belief — an assuredly metaphysical belief — that the world in general, and the world of psychophysical phenomena in particular, is just and no more than the way our physical theories say it is. This view of things is called scientific realism, and is to be distinguished from other accounts of the cognitive status of theories, according to which it does not necessarily follow from the success of a theory in predicting future events and occurrences that the theoretical entities implicitly defined by the theory really exist — that the world is in its intrinsic nature the way the theory says.

    If there are, as we say, alternatives to scientific realism, it is nevertheless the case that science is infected with this one view of itself. To be sure, it is generally presupposed by scientists that theories 'work' only becausethey correspond, in some sense, to the objective features of a real world. The astronomer, for example, will scoff at the suggestion that his sightings are made as though of celestial bodies. He will attach little importance to the fact that no expression appears in his theories for 'celestial body', but only for hypothetical relational properties that celestial bodies are thought, in the light of his theories, to possess. For him, the predictions made possible by astronomical theory are confirmed by experience precisely because there really are such theoretical objects as planets and asteroids, having such and such properties. For the scientific realist, it is the world that makes theoretical science possible; not, certainly, the other way around. Wilfrid Sellars gives concise expression to this feeling in claiming that "to have good reason for holding a theory is ipso facto to have good reason for holding that the entities postulated by the theory exist."

    The same attitude — often unexamined by its possessors — is rife within the empirical sciences. Microbiologists, for example, reason from the success of a virus theory in explaining the contraction of disease, or the souring of milk, to the real existence of viruses. It becomes a scientific 'fact' that there are such things as viruses. A micro-physicist, for his part, suggests the existence of a new fundamental particle in order that he might retain a simple and elegant conservation principle, and accords to this theoretical explainer the same sort of existence enjoyed by perceptible material objects such as tables and chairs. The claim that there is such a thing as the new fundamental particle is taken to be as empirically meaningful, and indeed as true, as the scientific principle whose continued utility is at stake. The logical status of the existential claim is in this way conflated with that of the conservation principle. The two hypotheses are imagined to sink or swim together on the weight of the same empirical evidence.

    In all these cases, the pattern is the same. Just so long as a hypothesis 'H' stands to be confirmed or disconfirmed at the hands of experience; i.e., just so long as 'H' is judged to be empirically meaningful, it is assumed that the existence of the theoretical entity or entities of which the hypothesis 'H' treats is an empirical matter as well. It is this assumption that we shall be anxious to challenge. We will claim that questions of existence are metaphysical in nature — that they cannot be decided on the basis of the empirical 'facts'. For, it is not required of the empirical scientist, quaempirical scientist, that he accord 'real existence' to the basic particulars of his successful theories, and that he deny such status to explainers figuring in his less successful ones. It is not asked that he add to his calculations the claim that the world really is the way his theories say, and that it is not the way disproven theories, or superstitions, say. It is asked of the theoretician only that he generate successful theories — theories that are useful in the prediction and retrodiction of physical events and occurrences. Whether or not scientific theories that succeed at this task mirror or correspond to a real, permanent structure of the universe is not, we shall insist, a question that is decidable by science itself.

    An identity view of mind and body comes into being whenever this question is unwittingly decided in the affirmative by scientists, or by philosophers of science, with respect to the contemporary and ever more successful theories of neurophysiology. This decision is made in two steps. The first step — which the identity theorist makes sans le savoir — is to reason from the success of neurophysiological theory to the real existence of its objects, which are brain processes, or ultimately, the basic particulars of microphysical theory. This inference, as we have said, is metaphysical in character. The second step, motivated by the realist's conviction that theoretical talk is not only definitive of, but exhaustive with respect to, the intrinsic nature of that part of the world concerned by it, is to hypothesize that brain processes and psychic states are empirically identical; i.e., to suggest that entities of the latter sort are 'nothing over and above' entities of the former sort. It is this second step which occupies the attentions of the identity theorist, and which creates the general impression that identity theory is an empirical matter alone. It should be clear, however, that this is not so. It is scientific realism that provides the identity theory with something to identify thoughts and sensations with, and scientific realism is as metaphysical a doctrine as one can find.

    All this would be not so bad were the identity theorist to realize that, at bottom, it was metaphysics he was doing, and not just empirical science. Unfortunately, this is rarely the case. It is more common, as we have indicated above, to see the identity theory presented by its authors as an expression of simple empirical fact — as a scientific solution to the problems and perplexities of psychophysical parallelism. And it is equally commonplace, in the face of objections made to an identity theory, to see this theory defended as though it were an empirical claim like any other.

    In what follows here, we will be concerned to evaluate the various stratagems used by the identity theorist in defending his purportedly empirical claim of psychophysical identity. One such stratagem, which constitutes more of a thematic for the defense of identity theory than a defense per se, will be called the "argument from conceptual revision." This argument suggests, on behalf of the identity theory, that the logical status of an empirical claim (its relative truth, falsity, or factual meaningfulness) is a function of the theoretical context wherein such logical status is assayed. Different theoretical contexts, it is said, determine different classes of propositions as empirically meaningful, and hence as possibly true, propositions. Furthermore, the identity theorist may argue that since theoretical contexts nave been known to change in the past, his own presently objectionable (i.e., factually meaningless, if not false) claims of psychophysical identity might, another day, take on a reasonability all their own. Ptolemy's hypothesis was perhaps as peculiar within the well-established context of Greek mythology as was Copernicus' hypothesis within the Ptolemaic conceptual frame. Yet, hypotheses are sometimes rendered plausible when frameworks shift beneath the facts, and in so doing 'change' the facts. (Thomas Kuhn and Norwood Russell Hanson have been among the most recent to argue that the history of empirical science cannot be adequately understood save from such a contextualistic point of view.)

    J.J.C. Smart, for example, a scientific materialist, must deny that phenomenal reporting terms refer, as they appear to refer, to irreducibly psychic entities. As we will see, he does this, somewhat arbitrarily, by relocating the objective meaning of phenomenal reporting terms in the occurrence of certain physical processes, where such a relocation of factual reference is suggested in terms of the Fregian distinction between Sinn and Bedeutung. The conceptual revisionist, however, would be anxious to understand Smart's theory as recommending more than a mere relocation of the meaning of psychic terms. Also recommended, he would say, is a shift from the conceptual frame of psycho-physical dualism to another conceptual 'set', wherein the possible arbitrariness of the disputed relocations of meaning is masked, or neutralized. Smart, indeed, counts himself guilty of no verbal chicanery in maintaining that the sole reference of phenomenal reporting terms lies in the occurrence of certain brain states. For, within the materialist's world view, or logical space, there are only physical events and processes. Nothing is left out, for there is nothing to be left out.

    When presented in such a light, even the most iconoclastic of identity views take on shades of nuance and subtlety. They ask only to be treated as empirical claims, in order that, in consequence of a 'shift' in frame that such treatment will bring about, they might take on empirical content. The identity theory asks only to be given a chance.

    It will be to give the identity theory such a 'chance' that we will address ourselves, in the light of the thesis of conceptual revision, to the materialistic monism of J.J.C. Smart, and the 'pan-quality' view of Herbert Feigl — the two main formulations given to the identity theory in current discussions. The first of these theories maintains, in sum, that the objective reference of phenomenal reporting terms lies in the occurrence of certain brain states. The second suggests that the objective reference of certain physical describing terms lies in qualitative features of thoughts and sensations. In each case, our job as conceptual revisionists will be to ask, what modifications of conceptual framework would have to be admitted in order that the identity claim in question might come to express a 'matter of fact'?

    The results of our experiment will be significant ones, for in each case it will turn out that only a modification of a metaphysical context will do. In Smart's case, for example, the requisite shift in frame will be seen to involve a conflation of the logics of descriptive and self-ascriptive discourses — a conflation which could never come about, so we will reason, so long as there are subjects in the world whose subjecthood consists in their consciousness of external objects. In Feigl's case as well, we will see that the identification in question — that of the theory-like denotata of certain physical terms and the non-theory-like raw feels of sentient experience — is one which only a metaphysical theory, and a defective metaphysical theory at that, could make. And from all this, we will conclude that identity theories, given their chance to perform as empirical claims, only reveal themselves as having been metaphysicaltheories, and only metaphysical theories, all along. What's more, we will decide that these are metaphysical theories which, for one reason or another, must he rejected.

    It will be at this point that our essay will turn from exposition to speculation. We will remind ourselves, first of all, that in rejecting the several formulations of the identity theory, we have been in fact rejecting scientific realism — a metaphysical account of the cognitive status of theories which maintains that the world in general, and the world of psychophysical phenomena in particular, is just and no more than what our physical theories say it is. It is this view of things that is rendered untenable by our conclusion that there is at least one part of the world — that part directly known by ourselves in the having of thoughts and sensations — that is not, and could never be, the way our physical theories say.

    There are alternatives, of course, to scientific realism, and we will adopt one of them in our attempt to show that science, once properly understood, is able to treat of the problems of psychophysics in an interesting and useful fashion. The alternative adopted will resemble in many ways the instrumental conception of science, and of the cognitive status of theories. To begin, it will not require, as did the realist view, that we reason from the successes enjoyed by a theory in predicting or retrodicting phenomena to the real or factual existence of the theory's basic particulars. For, it will be argued, the truth of existential statements made on behalf of explainers is not a necessary condition for the explanatory successes of a theory in which these explainers might appear. It is logically possible, for example, that the Copernican hypothesis should be every bit as well confirmed as it is today, yet that there be no such things as the planets Venus, Mars, and so forth. Logically speaking, scientific explanations and scientific ontologies are separable.

    For us, the proper concern of science will be seen to lie only with the generation and logical analysis of relations. Scientific knowledge begins, we will say, with relational properties abstracted from the world, or imputed to the world, which properties are logically inter-related by our scientific theories in such a fashion as to permit the formal derivation of certain of these properties from others. On this understanding of things, the cognitive significance of our scientific theories extends 'no deeper' into the world than to such relational properties. Beneath these relational properties — to the intrinsic nature of underlying structures and realities — it is simply not the business of our scientific theories to go. Science, in this sense, does not know its objects. It deals only with foci of relations. A person, thinking metaphysically, is required in order to provide the interpretation, or qualitative construction, by virtue of which a focus of relations becomes an object. And so the logical point made above; namely, that scientific explanations and scientific ontologies are separable, will have for its ground our more fundamental distinction between scientific explanation and ontology in general.

    It is of course a metaphysical speculation to say that science does not know its objects, and that thus science has no natural right to the tempting metaphysic of scientific realism. We do not for a moment suggest that such a speculation is scientific in character; i.e., empirically meaningful, whereas those of the realist are not. Our own view is as metaphysical as his. The difference is, however, that scientific realism gives a less adequate account of science, conceived en large, than does our own view. Our first chapter, for example, will treat of the difficulty of espousing the doctrine of scientific realism, and at the same time giving a reasonable account of the history of science — a history marked by conceptual 'shifts' in consequence of which certain well-confirmed theories are replaced by new ones. Our second and third chapters, as we have already indicated, will point out the unreasonability of the realist position with particular respect to those scientific explananda which appear to be qualitative in character. Still further difficulties with the realist view will be discussed in our fourth and fifth chapters. Our own account of the cognitive status of theories, on the other hand, which in effect denies to the theoretical entities of science any ontological status save that of 'explainer', succeeds nicely with those issues that are embarrassing to the realist view, and thus will be embraced in preference to that view. Let us make clear, however, that the considerations out of which the realist view will be rejected, and our own retained, are not scientific in character, but are metaphysical considerations. Our interest is not to make metaphysics scientific, or to replace metaphysics with empirical reasonings, but to adopt a metaphysical position in terms of which science, and its history, might be easily and gracefully understood.

    There will be consequences, of course, of our adoption of one metaphysical account of science, and of our rejection of another. Chief among these will be a change in our attitude concerning what counts as a scientific explanation. Whereas, for the scientific realist, a scientific explanation for X had ultimately to place in evidence those real structures and objects out of which X might be thought to be composed; i.e., to show what X really was, it now becomes a matter only of putting X, in terms of its relational properties, into useful relation with the rest of the world. The aim of science is to render the world intelligible, we will say; not necessarily to show how and out of what the world is 'made'.

    As concerns the science of psychophysics, there will be specific consequences of our reinterpretation of the cognitive status of scientific explanations. We will maintain that this science, one relieved of the constraints imposed upon it by a realist metaphysic, will no longer be bent on uncovering what, at bottom, psychic events and occurrences really are;e.g., how these events and occurrences might be 'reduced' to brain processes, but will content itself to relate the properties abstracted from or imputed to psychic and physical events in any useful fashion at all, whether or not the relations involved bespeak a reduction of one sort of entity to the other. We will envision, in particular, one such 'abstractivist' approach to the problems of psychophysics — an approach with which science has already made a good bit of progress. This approach involves the quantization of psychic and physical entities (the latter being quantized already; the former being candidates for such quantization) and the eventual logical inter-relation of these symbolically expressed quantities within a theory. The basic particulars of this theory need not be objects, we will say, nor even capable of being interpreted as objects, but might play the role of 'construct' alone. We will suggest that one such construct might emerge as a sort of dummy variable intervening between the expressions for psychic and physical abstracta, thus providing us with a sense in which the relationship holding between these abstracta might be thought through identity. We will greet this 'identity view' of the mental and the physical as a scientific explanation, although now according to our new, and more adequate understanding of what a scientific explanation is.


 All original material copyright © Gilbert Scott Markle. All rights reserved.