Reply to Horwich

Citation preview

Reply to Horwich Jerrold J. Katz Philosophical Issues, Vol. 4, Naturalism and Normativity. (1993), pp. 159-166. Stable URL: http://links.jstor.org/sici?sici=1533-6077%281993%294%3C159%3ARTH%3E2.0.CO%3B2-R Philosophical Issues is currently published by Ridgeview Publishing Company.

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/rpc.html. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

JSTOR is an independent not-for-profit organization dedicated to and preserving a digital archive of scholarly journals. For more information regarding JSTOR, please contact [email protected].

http://www.jstor.org Fri May 18 08:52:51 2007

PHILOSOPHICAL ISSUES, 4 Naturalism and Normativity, 1993

Reply to Horwich Jerrold J. Katz

1. Horwich focuses on MM's discussion of Wittgenstein's late philosophy, criticizing it from his own conception of what is really a t the heart of Wittgenstein's thinking. I will argue that Horwich's conception is not a plausible interpretation of the late philosophy, and that his defense of Wittgenstein does not succeed. Horwich does not deny that Wittgenstein says the naturalistic sounding things on which I based my interpretation of him as a naturalist, but dismisses it because I rely on the methods of standard textual scholarship t o obtain "an objectively satisfactory interpretation" (p. 157). He claims that Wittgenstein's late writings are too messy and tentative for such methods t o be feasible. Being aware of the problems concerning the texts that Wittgenstein did not himself prepare for the public, I stayed away from them as much as possible, basing my interpretation on P I (and in a few places on the Tractatus). In the rare cases where I mentioned another source, I avoided passages with the dubious status about which Horwich complains. More importantly, Horwich operates on a double standard. On the one hand, he dismisses my interpretation on the grounds t h a t , with Wittgenstein, scholarly business as usual faces an "almost insuperable difficulty" (p. 157) which, for all practical purposes, precludes

textual interpretation. On the other hand, Horwich puts forth his own interpretation that "[tlhe real foundation of Wittgenstein's philosophy [is] his metaphilosophy" (p. 154). But how does he overcome the "almost insuperable difficulty"? Moreover, if he eschews exegesis, how can he expect t o convince anyone that his interpretation is correct? Without documentation, interpretation degenerates into arbitrary opinion. Whereof one cannot document, thereof one cannot interpret. 2. There are several reasons to doubt Horwich's interpretation of Wittgenstein's late philosophy. For one, the view that Wittgenstein's metaphilosophy is the real foundation for his late philosophy does not square with the way Wittgenstein develops his argument in PI. The metaphilosophy -basically P I sections 109-133- evolves out of his prior discussion of meaning, starting with his attempt to clear up the confusion of meaning with reference a t the very beginning of PI. For another, Horwich's interpretation is too uncharitable to Wittgenstein to be plausible. If the metaphilosophy were really the foundation of everything else, Wittgenstein would be asking us t o accept his radical doctrine about philosophy and its attendant claims about meaning, language, rules, knowledge, etc. without providing any reason. If Horwich were right, and P I were based on what amounts t o a stipulated metaphilosophy, the entire edifice of Wittgenstein's late philosophy could be dismissed in favor of an equally arbitrary stipulation. If, in constrast to Horwich's interpretation, we see Wittgenstein's argument for the metaphilosophy as developing out of and based on the criticisms of theories of meaning that occupy him in all the sections preceding those which announce the metaphilosophy, then the earlier criticisms of theories of meaning make an elaborate case for the metaphilosophy in which reasons are presented every step of the way. Since the metaphilosophy has arguments behind it, the critic must provide better opposing arguments. The metaphilosophy cannot be dismissed as a mere stipulation, as it can on Horwich's interpretation. Worse, Horwich's interpretation of Wittgenstein's metaphilosophy is not coherent. He rightly says that "the heart" of it is the "idea that philosophical questions derive from confusion rather than ignorance, [etc.]" (p. 155). Horwich goes on t o claim, "Nothing about meaning is presupposed" (p. 156). This claim, which is a consequence of Horwich's view that Wittgenstein's late philosophy rests on his metaphilosophy, is incompatible with Wittgenstein's idea about philosophical questions because that idea makes no sense unless quite

a lot about meaning is presupposed. What do the confusions underlying philosophical questions stem from if not false pictures about meaning, e.g., that the meaning of a word is the object for which it stands? In reference to that picture, Wittgenstein says, "this general notion of the meaning of a word surrounds the workings of language with a haze which makes clear vision impossible". Further, the ignorance which Wittgenstein is at such pains to show philosophical questions do not derive from is ignorance of hidden senses, as his two principal targets Frege and his former Tractatus self thought. Moreover, references to meaning in critical metatheoretic passages like the last one in P I 120 have to be understood as harking back to the notion of meanings as linguistic objects which Wittgenstein has been unfavorably comparing with use (starting with section 1). Finally, Wittgenstein's doctrine that metaphysical statements about the essence of knowledge, being, etc. are "piece[s] of plain nonsense" (enunciated, for example, in P I 116-119) presupposes an account of sense. Without the account in the earlier sections, we can make no sense of his term "nonsense". The point is clear. Without presupposing the earlier discussion of what meaning is and is not, P I 109-133 is too hopelessly vague to even count as a metaphilosophy, but, presupposing them, Horwich's claim about the role of the metaphilosophy in Wittgenstein's late philosophy is implausible. 3. Horwich misunderstands my discussion of Wittgenstein. He takes me to be saying that Wittgenstein claims "the meaning of a word is its use". But MM (p. 31) explicitly rejects that interpretation, noting P I 43 where Wittgenstein says that that equation only holds for "a large class of cases". Horwich also attributes the 'meaning is use' equation to Wittgenstein, not realizing that it would be inconsistent for Wittgenstein to accept it (as is observed in MM, p. 31). For, the main point of Wittgenstein's criticism of traditional accounts of meaning is that "there is no one thing common which makes us use the same word for all" ( P I 65). My central argument against Wittgenstein does not, as Horwich thinks, take Wittgenstein to be arguing that his account of meaning follows from "the falsity of the various systematic theories he criticizes" (p. 154). Rather than taking Wittgenstein's arguments against "systematic theories of meaning" to be an attempt to assemble premisses for such a deduction, I take them to be part of an effort to show that facts about the meaning of words in a language should be seen as natural facts, reflecting the practices which govern that game-like activity. In large part, the effort takes the form of "bring[ing] words back from their metaphysical to their everyday use" ( P I 116).

In arguing that Wittgenstein's criticisms of theories of meaning are largely tailored to Fregean theories, and hence do not eliminate the new intensionalism based on the proto-theory stemming from (D), I a m opposing Wittgenstein's effort to establish a naturalistic perspective on language. I want to make the case for an alternative to naturalism. My arguments that the new intensionalism survives Wittgenstein's criticisms are one component of that case. The second component is the observation that Wittgenstein assumes without justification that the basic facts for deciding questions of meaning in natural languages are facts about use. A third component is the introduction of another assumption about what the basic facts in decisions about meaning are, namely, facts about the sense properties and relations of expression types (e.g., that "bank" is ambiguous, that "sister" and "female sibling" are synonymous, that "blind" and "sighted" are antonymous, and that "free gift" is redundant). I argue that the proto-theory accounts for such semantic facts about expression types and that the new intensionalism can account for the semantic facts about their tokens on the basis of that semantic theory together with an account of pragmatic reasoning like Grice's. I then show how to interpret the semantic theory Platonistically, thereby providing a conception of the basic facts about the meaning of expression types on which they are non-natural facts, reflecting the structure of abstract objects. As explained in my replies to Boghossian and Zemach, the new intensionalism plus Platonism offers a way out of such otherwise powerful arguments as Wittgenstein's rule following argument. If I've made my case for an alternative to Wittgenstein's perspective on language, then the Wittgensteinian pressure is off metaphysics. As an illustration, consider Wittgenstein's argument in P I 65-67. Here the pressure to abandon the doctrine of universals comes from the examples that Wittgenstein uses to show that there is nothing common to the things to which we apply the word "game" but merely a family resemblance among them. Given the alternative perspective, that pressure disappears. As I show in MM (pp. 107-115), given a Platonistic interpretation of the semantic theory, the new intensionalism, and a standard account of pragmatic reasoning, we can account for Wittgenstein's examples without having to abandon the doctrine of universals. Contra Horwich, I did not claim that Wittgenstein deduces the meaningless of metaphysics from "the use conception of meaning". According to MM, Wittgenstein tries to exhibit the meaningless of metaphysics on the basis of his assumption that facts about the meaning of words are natural facts reflecting linguistic practice.

Rather than a single deduction, there are painstaking exhibitions of such meaninglessness taking various forms. One is t o employ descriptions of the everyday use of words like "knowledge", "being", L L ~ b j e ~etc. t " , to reveal the nonsense in their metaphysical uses. Nor, contra Horwich, do I claim that Wittgenstein deduces naturalism from "the use conception of meaning". On my view of his naturalism, it wouldn't make sense to do so. According to MM (pp. 2-3), the naturalism in PI was part of Wittgenstein's position in the Tractatus, hence part of his philosophical outlook even before he developed his "use conception of meaning". Moreover, in various places where naturalism is mentioned, it is clear that the category of natural fact is broader than the category of linguistic fact. For example, in Culture and Value (University of Chicago Press, 1984, p. 44e), Wittgenstein says that either something rests on beliefs about natural processes or else "No fact justifies it. None can give it any support". 4. Horwich claims that there is no conflict between scientific semantics and Wittgenstein's notion (according t o Horwich) that "the meaning of a word consists in its use". His claim seems surprising a t first because we think of science as synonymous with the construction of theories that explain phenomena in terms of hidden entities like subatomic particles. Wittgenstein is famous for his hostility t o theories, hidden entities, and explanation in connection with linguistic phenomena. But Horwich's claim becomes unsurprising when he explains his notion of scientific semantics. Horwich sees scientific semantics simply as a "systematization of use", thereby making his claim a trival truth. Horwich fails to see the conflict between scientific semantics and Wittgenstein's notion that "the meaning of a word consists in its use". This is because Horwich takes it for granted that a systematization of our ordinary, pretheoretical notion of meaning either would not become theory construction or, if it did, would in the process give the term "meaning" "a new, technical meaning". This assumption ignores both the argument in MM (pp. 52f) that such systematization turns into theory construction and the argument in MM (pp. 31-33) that the theoretical concept of meaning explains what meanings in natural language are in the way that the theoretical concept of number or matter explains the pretheoretic notions of number and matter. Accordingly, Horwich ignores MM's real challenge to Wittgenstein's late philosophy. MM challenges its claim that the philosophically fundamental thing about the meaning of words is their use. MM's line of argument, as indicated above, is that the best overall

account of semantic phenomena in natural language takes abstract meanings as fundamental. Developing a scientific theory of semantics in the sense customary in generative linguistics, it explains the use of words derivatively in terms of such meanings and a standard pragmatics. The challenge is that, not only does such a theory survive Wittgenstein's ~rit~icisms in P I , but it does a better job of systematizing all the semantic phenomena of natural language, from the sense properties and relations of expression types to the sense properties and relations of their tokens. Because of Horwich's limited conception of scientific semantics, he thinks there need be no conflict between Wittgenstein's conception of meaning and mine. He says: For Wittgenstein's bit of conceptual analysis -that "the meaning of X" means "the use of XV- does not preclude a systematic, semantic theory, ci la Katz (p. 158). As already noted, Wittgenstein does not and cannot consistently accept the equation of meaning with use that Horwich attributes to him. More importantly, Wittgenstein cannot accept a semantic theory a la anybody because a theory is much more than a systematic description. In particular, Wittgenstein cannot accept my semantic theory, because its explanations in terms of abstract meanings, its decompositional analysis of underlying sense structure, and its promise of scientific discoveries about the nature of language contradict his purely descriptive approach to meaning. 5. Horwich puts forth the surprising view that Wittgenstein does not deny Platonism. If there is one uncontroversial thesis about twentieth century philosophy, it is that Wittgenstein rejects abstract objects in mathematics, logic, and language. Wittgenstein's naturalism is unequivocally opposed to objects which have no spatial or temporal location, are causally inert, and incorporeal. His approach to language is unequivocally against meanings which are perceptually hidden, exist independently of us, and hence stand apart from our form of life and its social practices. His conception of philosophy is unequivocally against essences or universals which are the subject of discoveries of significance in science and philosophy. In Remarks on the Foundations of Mathematics (Basil Blackwell, 1956), Wittgenstein frequently and consistently rejects mathematical Platonism. He has a diatribe against "mathematical alchemy" (RFM, iv. 16). He says, "The mathematician is an inventor, not a discoverer" (RFM, I, p. 167). Similarly, "I am trying to say something like this: Even if the mathematical proposition seems to point to a reality outside itself, still it only expresses acceptance of a new

measure (of reality)" (RFM, 11, p. 30). Further, "[mathematical proof] makes new connections.. . (It does not establish that they are there; they do not exist until it makes them.)" (RFM, 11, p. 31). Despite such evidence, Horwich thinks that Wittgenstein accepts a "Harmless Light-weight Platonism" which is merely a trivial truth. W h a t Wittgenstein really wants to reject, according to Horwich, is "Problematic Heavy-duty Platonism". The latter position is harmless Platonism plus the further claim that "numbers [be] thought of as similar to physical objects". As already indicated, Wittgenstein would not regard the philosophical doctrine denoted by "Harmless Light-weight Platonism" as either harmless, trivial, or true. Moreover, "Problematic Heavyduty Platonism" does not denote a position distinct from Harmless Light-weight Platonism because the claim that numbers be thought of as similar to physical objects adds nothing. To see this, let us examine the similarity of abstracts objects to physical objects. It's true that abstract objects are similar to physical objects in being objects, in existing independently of us, in not being mind-dependent, and in being the subject of our discoveries, but those features hold of abstract objects in any Platonism, including Harmless Light-weight Platonism. The only thing left is sheer physicalness, but, since nothing with spatio/temporal properties and causal relations can be an abstract object, abstract objects cannot be similar to physical objects in being physical. Because the analogy adds nothing, "Problematic Heavy-duty Platonism" is a vacuous name. 6. Horwich disparages my epistemological views by saying, "Katz supposes that our knowledge of arithmetic is the product of something called 'intuition' -a somewhat mysterious faculty that somehow puts us in tune with abstract objects" (p. 158). Why mysterious? Not because the faculty is supposed to be one that enables us to make causal contact with abstract objects. I explicitly reject that conception of intuition in Language and Other Abstract Objects (pp. 200-216). It is true that we do not know how intuition works, but if that were enough to merit the label "mysterious", we would also have to apply it to perception, reason, imagination, and all our other mental faculties. Finally, in the aforementioned book, I give an account of intuituion. That account was sketchy, but Horwich does not even consider it.' Calling intuition "mysterious" might be passed off just as epater le Platoniste. However, that leaves the impression that the critic of Pla'I have tried again in T h e Knowledge and Nature of Abstract Objects (in preparation).

tonism is in a position to cast stones. Some critics of Platonism may be, but not a Wittgensteinian conventionalist. Benacerraf's "Mathematical Truth" (Journal of Philosophy, 1973, pp. 661-680) points out that conventionalism too has a major problem about mathematical truth to solve, namely, what distinguishes those cases in which stipulating conventions insures truth from those in which it does not? Conventionalists need to answer that question before they disparage an ontology that does as well as Platonism in providing a semantics for number theoretic sentences.