The self-consciousness argument why Tooley's criticisms fail

Philosophical Studies

Volumn 105, Issue 3, 2001, Pages 281-307

  Beaeer, George ^a  

^a UNIVERSITY OF COLORADO   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 52549126256     PISSN: 00318116     EISSN: 15730883     Source Type: Journal    
DOI: 10.1023/a:1010358801138     Document Type: Article

References (32)

1. "Self-consciousness," The Philosophical Review, vol. 106, 1997, pp. 69-117. In the paper I criticize two versions of ontological functionalism - the familiar Ramsified version and the Language-of-Thought version. In the present context only my arguments against the Ramsified version are under attack, so I will confine my remarks exclusively to those arguments and the criticisms of them.
2. Hereafter, unless otherwise indicated, page numbers and note numbers will be those in "Self-consciousness." 2 Philosophical Studies, vol. 105, 2001, pp. 251-279.
3. For example, during comments and discussion at nearly every presentation of the paper. See also Mark McCullagh, "Functionalism and Self-Consciousness," Mind and Language, vol. 15, pp. 500-510.
4. It is in fact just a concept-theoretic variant of the sort of definition contemplated at the end of note 21 in "Self-consciousness."
5. P. 103, "An Argument for the Identity Theory," in Philosophical Papers, Volume I, New York: Oxford University Press, 1983, pp. 99-107.
6. P. 59, "The Mind-body Problem," in The Mind-Body Problem: A Guide to the Current Debate, Richard Warner and Tadeusz Szubka (eds.), Oxford: Basil Blackwell, 1994, pp. 55-60.
7. As David Lewis emphasizes, "We must assume that all occurrences of T-terms in the postulate of T are purely referential, open to existential generalization and to substitution by Leibniz's law. We need not assume, however, that the language of T is an extensional language." (p. 80, "How to Define Theoretical Terms," in Philosophical Papers, Volume I, New York: Oxford University Press, 1983, pp. 78-95). Now since the embedded psychological predicates in principle IP express acquaintable properties (see section 5), they do occur referentially according to the traditional doctrine of acquaintance. Hence, the Lewis style recipe directs functionalists to replace embedded occurrences of such psychological predicates with predicate variables. (Note that the first step in Lewis's recipe needs to be omitted in the case of embedded predicates. Why? Because if, following Lewis, one were at this first step to replace them with copulas and primitive property names, an intensionality error would arise: the resulting principle would not be a truth of psychology.)
8. For similar reasons, Putnam's recipe would be effectively the same (section VI, "On properties," in Nicholas Rescher et al., eds., Essays in honor of Carl G. Hempel, Dordrecht: D. Reidel, 1970). Likewise for other ontological functionalists in the Putnam tradition.
9. Another possibility is that there is no sequence of first-order realizations R satisfying A. If so, the right-hand side of the definition would be null and therefore would certainly not define the relation of being self-consciously aware. One familiar response to this argument is to "bite the bullet," that is, to hold that propositions involving such first-order realizations really are typical objects of the self-consciousness relation. See my "The Mind-Body Problem" (forthcoming), for an explanation of why this response is mistaken. Note that in the argument I am talking about the property of being in pain, which is denoted by the canonical gerundive phrase 'being in pain.' All ontological functionalists - "Australian" (e.g., Lewis, et al.) as well as "American" - are committed to identifying this property, not with a first-order realization, but with a second-order functional property (see, e.g., Lewis, ibid.; see also my "Ramsification and Intensionality".)
10. 1x)'. But the latter is exactly the Ramsification of ℘ used in the Self-consciousness Argument.
11. For simplicity I am assuming here that 'self-attribute' is the fifth mental predicate in the psychological theory A.
12. def x thinks that he is F. But, to be correct, this definition requires that the embedded occurrence of 'F' on the right-hand side is an externally quantifiable predicate variable ranging over properties. So in this setting the intensionality objection is dead in its tracks. Some people might hold that we self-attribute concepts, but this is implausible on its face: you self-attribute the attribute of being in pain.
13. Remember, in this present paper we are confining ourselves to Ramsified formulations of ontological functionalism; for simplicity we are supposing that the language-of-thought version is off limits, though that position is dealt with in "Self-consciousness."
14. Some Russellians hold that, for every complex property, there is a necessarily equivalent simple property. Others disagree, holding that all simple properties are basic properties.
15. See David Kaplan, "How to Russell a Frege-Church," The Journal of Philosophy, vol. 72, 1975, pp. 716-29.
16. Also my Quality and Concept, 1982, Oxford and my
17. "A Solution to Frege's Puzzle," Philosophical Perspectives, vol. 7, 1993, pp. 17-61.
18. Someone might try to challenge the claim in the text on the ground that properties are mind-independent entities whereas concepts are mind-dependent. But, not only would this challenge be out of step with Frege's anti-psychologism, a certain modal argument shows that this view of concepts cannot be right. See my "Universals," The Journal of Philosophy, vol. 90, 1993, pp. 5-32.
19. This point draws attention to another slip in MT's argument about intensional logic, namely, its invalid inference from the conclusion that some propositions need to be analyzed in terms of concepts (vs. properties), to the further conclusion that all propositions need to be analyzed that way.
20. Thus, unlike the original functional definitions, which were formulated in the predicative logical framework of "ramified type theory" (see Putnam, ibid.), these definitions are formulated in an impredicative type-free logical setting.
21. Toward this end, the underlying psychological theory A might contain iterated-attitude clauses such as 'It is possible for someone to think that he is thinking something', and the predicate variables in A might be restricted to natural (or basic) properties. Of course, the thesis in the text would hold if A implicitly defines the standard mental properties and so is unqiuely satisfied by them. For in this case, no first-order physical realizations of these properties would also satisfy A.
22. An exception to this general tendency is the proposal made by Mark McCullagh, ibid.
23. This goal has been articulated by many people: Putnam, Lewis, Shoemaker, Block, Jackson, and others.
24. Some Ramsifying functionalists believe that the psychological theory upon which their functionalist definitions are to be based will be a priori. For these people, the resulting definitions will also be a priori. But other Ramsifying functionalists believe that the relevant psychological theory will be an a posteriori scientific theory, and so they believe that the associated definitions will be a posteriori scientific definitions. Both sorts of functionalist, however, accept this account.
25. 2. But in a molecule-for-molecule description D of our being y, there obviously would be no clause with this logical form; chemical descriptions simply are not like that.
26. 4 to the proposition that he is in pain!) In short, our functionalists have no choice but to reject Thesis III. But, taken together, Thesis I and Thesis II, entail Thesis III. Therefore, functionalists have no choice but to reject either Thesis I or Thesis II. But, as we saw above, faced with this choice, functionalists must reject Thesis II. Thus, the ploy of trying to shift upward the biological-psychological boundary is seen to be futile.
27. Furthermore, the antecedent of principle ℘ is that x is not only in pain but also engaging in introspection. This of course means conscious introspection (is there any other kind?). Surely beings satisfying these two conditions are selfconsciously aware that they are in pain (at least if relevant qualifiers are adjoined). MT's example leaves out the requirement that x be engaging in conscious introspection.
28. Of course, this problem generalizes: functional definitions of just about any psychological concept could be blocked by similar families of examples. Incidentally, I have heard the following objection to the inclusion of principles like ℘ (this is not MT's objection). Given that creatures of the sort described by MT are possible, the associated principles would not describe an essential property. Hence, such principles should not be included in the theory upon which functional definitions are based. The error, of course, is that such principles are conditionals: if..., then .... A great many (most?) essential properties are associated with such (metaphysically necessary) conditionals.
29. Here are some examples (i) For propositions p concerned only with phenomenal qualities, if x is thinking p and engaging in introspection, then x will think that he is thinking p. (ii) If x is thinking p and engaging in introspection, then in normal psychological conditions x would be more likely to think that he is thinking p than to think that he is not thinking p. (iii) If x thinks p and considers the question whether he thinks p, then he will think that he thinks p. (The latter principle leads to the usual result but requires a slightly more complicated version of the Self-consciousness Argument. See my forthcoming book The Integrity of Mind.)
30. See pp. 254 ff., "Psychophysical and Theoretical Identifications," Papers in Metaphysics and Epistemology, Cambridge: Cambridge University Press, 1999, pp. 248-261. Lewis's proposal rests on Dana Scott's treatment of vacuous descriptions, which I find extremely artificial and unintuitive.
31. Of course, it is understood here that, e.g., "Electrons are the things to which our term 'electron' applies" is not a genuine theory in physicsl And this point generalizes to theoretical terms in other natural sciences.
32. I am assuming here that the difficulty in MT's definition mentioned in section 4 has somehow been solved. At several points in his disucssion MT seems to suggest that all the clauses in the theory upon which a Ramsified definition is based would turn out to be 'analytic' if the definition were correct. But there are counterexamples to this claim (e.g. purely existential clauses). It is the case, however, that the claim holds for certain forms of conditionals (of which ℘ is an illustration).