It is a well-known fact that W.V. Quine sees in empiricism two dogmas: the analytic-synthetic distinction, on the one hand, and reductionism, on the other. To these two dogmas Donald Davidson adds a third which he calls scheme-content dualism (TI, 190).¹ One important use of this dualism is to give meaning to conceptual relativism where two incommensurable conceptual schemes are counted as conceptual schemes because they are schemes for the same content. The main concern of this chapter is to see (a) whether Kuhn is committed to a dualism of scheme and content, and if so, what exactly it looks like; (b) how such a dualism leads to conceptual relativism; and finally, (c) how Davidson's general criticism of scheme-content dualism could apply to Kuhn's brand of dualism.

There is no doubt in Davidson's mind that Kuhn is guilty of scheme-content dualism. Commenting on Kuhn's view that "scientists operating in different scientific tradition (within different `paradigms') `work in different worlds'," Davidson says that "Kuhn . . . wants us to think of different observers of the same world who come to it with incommensurable systems of concepts" (TI, 187). According to Davidson, the latter "suggests . . . a dualism of total scheme (or language) and uninterpreted content" (ibid., my emphases). Here Davidson is very careful to say that Kuhn's view suggests a particular form of dualism which involves uninterpreted content. `Incommensurable' is, of course, Kuhn and Feyerabend's word for `not inter-translatable'. The neutral content waiting to be organized is supplied by nature.

But is Davidson correct when he comes close to attributing dualism to Kuhn? Immediately before this, Davidson quotes a long passage from Kuhn which seems to suggest that Kuhn denies the existence of neutral or uninterpreted content (TI, 190). This is expressed even more clearly when he repudiates the Western epistemological tradition that posits fixed and neutral data which are susceptible to different interpretations (SSR, 126). Hence, Davidson's attribution of a particular form of scheme-content dualism to Kuhn warrants consideration. The following argues that indeed Kuhn is guilty of such a dualism, but that it is a form of dualism much more akin to that which Davidson finds in Quine than to the one depicted above.

QUINE'S SCHEME-CONTENT DUALISM

That the so-called uninterpreted content is not a necessary ingredient of dualism is clear to Davidson: "The scheme-content division can survive even in an environment that shuns . . . the assumption that there can be thoughts or experiences that are free of theory" (MS, 161). Instead, according to him, "what matters [to scheme-content dualism] is that there should be an ultimate source of evidence whose character can be wholly specified without reference to what it is evidence for" (ibid., 162). This kind of dualism is supposed to be found in Quine. Let us go into some of the details.

Davidson takes Quine's dualism of scheme and content to be strongly suggested in Quine's following words:²

We can investigate the world, and man as a part of it, and thus find out what cues he could have of what goes on around him. Subtracting his cues from his world view, we get men's net contribution as the difference. This difference marks the extent of men's sovereignty--the domain with which he can revise theory while saving the data (WO, 5).

World view or theory is what Davidson calls scheme, while cues or data are content (MS, 162; MTE, 69). What are cues or data, and what is the relation between scheme and this content so understood, are made clear in Quine's "Epistemology Naturalized" (hereafter EN). The main task of naturalized epistemology is the study of the "relation between the meager input and the torrential output" (EN, 83), the stimulations of one's sensory receptors being the inputs (EN, 75), while the output is a theory or "a description of the three-dimensional external world and its history" (EN, 83).

What remains to be seen is the relation between content and scheme. We may glean from Quine's writings the following answers":

(1) the sensory input or content "supports the very physical theory that I am accepting" (QEC, 24; my emphasis):

(2) "society teaches us our physicalistic language by training us to associate various physicalistic sentences directly, in multifarious ways, with irritations of our sensory surfaces" (PR, 253);

(3) "The stimulation of [one's] sensory receptors is all the evidence anybody has had to go on, ultimately, in arriving at his picture of the world." (EN, 75; my emphasis).

Here (2) is concerned with language acquisition, whereas (1) and (3) have to do with theories and their evidence.

Because sensory stimulation by itself is not propositional it can cause one to assent to certain observation sentences (QEC, 25; EN, 85; VITD, 40), which in turn may serve as evidence. The same point is expressed clearly when Quine says that "our only source of information about the external world is through the impact of light rays and molecules upon our sensory surfaces" (NNK, 68).

But conceptual relativism still is not in sight, whereas our purpose here is to search for a form of dualism in Quine's talk of subtracting cues from the world view. Davidson's construal of this talk is, to repeat, that "what matters is that there should be an ultimate source of evidence whose character can be wholly specified without reference to what it is evidence for," or to what goes on around us. It is doubtful that the evidence can be so specified. The difficulty may be glimpsed from Quine's note that "the motivating insight, viz. that we can know external things only through impacts at our nerve endings, is itself based on our general knowledge of the ways of physical objects" (WO, 2). To be prudent we should avoid attributing without reference to what goes on around us.

One way to take Quine's dualism may go like this. We derive our world view from patterns of stimulation in the sense that the latter cause our web of beliefs about the world. But the same patterns of stimulation may occasion quite different world views, though these cues constitute our only source of information about the external world. If, as Quine has it, subtracting these cues from our world view gives us our net contribution, then the cues are data or what are not contributed by us. We may thus take Quine's dualism to be the view that from the same data we may derive quite different world views. In so far as the derived world views might be dramatically different, conceptual relativism would follow.

Another possible construal of Quine's dualism would take `saving the data' as the only constraint for any acceptable world view. That is, one world view is as good as another so long as they are equally good in helping us "to foresee and control the triggering of our sensory receptors in the light of previous triggering of our sensory receptors" (TTP, 1). But this may not lead to the dualism in Quine for which we have been searching unless it is assumed that as world view builders we might come up with dramatically different world views.

Two things should be noted about our way of taking Quine's dualism. First, contrary to Davidson, our version of Quine's dualism is independent of whether there is a way of specifying the data that is independent of what goes on around us. Second, Quine's dualism does not lead automatically to conceptual relativism unless there be added some premise concerning our "conceptual sovereignty". Hence, by itself scheme-content dualism may not be what is to be blamed.

KUHN'S SCHEME-CONTENT DUALISM

We have noted before that Kuhn's scheme-content dualism is more akin to Quine's than to the one depicted by Davidson. It is time to support this claim. First we want to show that Kuhn denies the existence of the so-called uninterpreted content. What we find in Kuhn that comes closest to the so-called uninterpreted content is what he calls pure percepts (SSR, 127). But pure percepts are not so pure, for if they exist their existence presupposes "a paradigm, taken either from a current scientific theory or from some fraction of everyday discourse" (ibid). If pure percepts are not immune from some sort of theoretical underpinning, Kuhn suspects "that scientists are right . . . when they treat oxygen and pendulums . . . as the fundamental ingredients of their immediate experience" (ibid., 127-28). But oxygen and pendulums definitely are not uninterpreted content.

But if there is no uninterpreted content, what else would play the role of content in the dualism? The answer is that stimuli which impinge on us will be the content. In talking about how people who experience communication breakdown because of incommensurable viewpoints may yet have some recourse, Kuhn says that

the stimuli that impinge upon them are the same. So is their general neural apparatus, however differently programmed. Furthermore, except in a small, if all-important, area of experience even their neural programming must be very nearly the same, for they share a history, except the immediate past (Postscript, 201).

A differently programmed neural apparatus may process "certain stimuli differently," thus making us see "different things or the same things differently" (RMC, 276). And if the members of two groups have systematically different sensations on receipt of the same stimuli, "they do in some sense live in different worlds" (Postscript, 193).

According to Kuhn the neural process "is subject to change both through further education and through the discovery of misfits with the environment" (ibid., 196). If we use the word `leading' to cover the two cases then we may simply say that neural process can be changed through learning.

We can readily see from our brief introduction of Kuhn that his position is indeed very similar to that of Quine. For both, retinal imprints or, more generally, stimuli that impinge upon us "are elaborate constructs," and hence not uninterpreted content.

What remains to see is whether they share a view on the connection between sensory stimulation and our world view. We have noted that Quine takes the relation between the two to be that sensory stimulation causes us to assent to certain sentences, that is, causes us to have certain beliefs. Kuhn on the other hand views the relation in terms of our neural apparatus as processors so that we "are processing stimuli" (RMC, 276). This does not show that Quine and Kuhn see the relation differently. For Quine may accept the metaphor that our neural apparatuses are processors and also without contradiction claim that stimuli cause us to have certain beliefs about the world.

How would Kuhn's scheme-content dualism lead to conceptual relativism? What makes us see different things or see things differently and eventually make us dwellers of different worlds are our differently programmed neural apparatus. The reason for the possibility of conceptual relativism would then seem to be the same for both Kuhn and Quine: it all hinges on the possibility that different people might be induced by patterns of stimulation to come up with widely different world views.

Quine's scheme-content dualism as depicted above is obviously a chapter in naturalized epistemology. If so, why does Davidson think that it "cannot be made intelligible and defensible," that it is a dogma that ought to be discarded? (TI, 189) Since Kuhn's dualism is similar to that of Quine, the same question may be asked about Kuhn. What we want to know, at least in outline, is how it would apply to Kuhn or Quine.

WHAT DAVIDSON WOULD SAY ABOUT

KUHN'S (OR QUINE'S) DUALISM?

We have noted that by itself Kuhn's or Quine's dualism is not something to be blamed, for by itself it does not necessarily lead to conceptual relativism. One way that Kuhn's dualism would lead to conceptual relativism is to assume that differently programmed neural apparatus might yield dramatically different world pictures. If Davidson takes his criticism of scheme-content dualism to be applicable to the present case, it would follow that, supposing he would accept Kuhn's depiction of the causal relation between stimuli and our world views, he would deny the above assumption because, no matter how neural apparatus be differently programmed, they can, as a matter of fact, yield only more or less similar world views? But why would this be so; is it a fact about us as human-machines?

Certainly, when Davidson criticizes Kuhn's dualism, he is not making a point concerning matters of fact. But in order to say what actually he is getting at, we should first note that we used to talk about neural apparatus yielding world views. For our present purpose, however, we may think instead of neural apparatus as a mechanism which takes stimuli as input and spoken sentences as output. These spoken sentences are supposed to express a person's beliefs; when taken collectively, they may be regarded as expressing his world view. What spoken sentences a neural apparatus may have as its output is an empirical fact; no one has the right to decide by a priori means which spoken sentences do or do not belong to the output. But whether the output as a collection of spoken sentences constitutes an interpretable whole is not simply an empirical matter.

Thus, Davidson's criticism comes down to this: according to his methodology of interpretation, world views cannot be dramatically different.³ Kuhn's dualism therefor turns out to be no exception to Davidson's general criticism.⁴

CONCLUDING REMARKS

We have shown that some sort of scheme-content dualism can indeed be found in both Kuhn and Quine, but that it has a quite different form from what Davidson takes it to be. The applicability of Davidson's general criticism of scheme-content dualism to it has also been shown.

What needs to be clarified, but what we do not go into here, is how we should take `scheme-content dualism by itself' and what are some of the extra factors that conspire to lead to conceptual dualism. This point may be relevant also to a proper understanding of Davidson general view about conceptual relativism.

1. For abbreviation of titles of articles or books, see Reference.

2. See MTE, 69.

3. For Davidson's methodology of interpretation, one may see various articles in TI.

4. Davidson's general criticism of scheme-content dualism is most completely stated in "On the Very Idea of a Conceptual Scheme", which appears in TI.

Davidson, D.

[TI] Inquiries into Truth and Interpretation (Oxford: Clarendon Press, 1984).

[MS] "The Myth of the Subjective" in Relativism: Interpretation and Confrontation, ed. Michael Krausz (Notre Dame: University of Notre Dame Press, 1989).

[MTE] "Meaning, Truth and Evidence" in Perspectives on Quine, eds., Robert B. Barrett and Roger F. Gibson (Mass.: Basil Blackwell, 1990).

Kuhn, T.S.

[SSR] The Structure of Scientific Revolution (2nd ed., enlarged; Chicago: The University of Chicago Press, 1970).

[Postscript] "Postscript-1969" in T.S. Kuhn, The Structure of Scientific Revolution.

[RMC] "Reflections on My Critics" in I. Lakatos and A. Musgrave, eds., Criticism and the Growth of Knowledge (London: Cambridge University Press, 1970).

Quine, W.V.

[WO] Word and Object (Mass.: The MIT Press, 1960).

[EN] "Epistemology Naturalized" in W.V. Quine, Ontological Relativity and Other Essays (New York: Columbia University Press, 1969).

[NNK] "The Nature of Natural Knowledge" in Samuel Guttenplan, ed., Mind and Language (Oxford: Clarenden Press, 1975).

[PR] "Posits and Reality" in W.V. Quine, The Ways of Paradox and Other Essays (rev. and enl.; Cambridge, Mass.: Harvard University Press, 1976).

[QEC] "Empirical Content" in W.V. Quine, Theories and Things (Cambridge, Mass.: Harvard University Press, 1981).

[VITD] "On the Very Idea of a Third Dogma" in W.V. Quine. Theories and Things.

[TTP] "Things and Their Place in Theories" in W.V. Quine, Theories and Things.