|
The Publicity of Thought and Language Daniel Laurier
|
|
The Publicity of Thought and Language Daniel Laurier
|
The sceptical problem of Kripkenstein pertains to both the notions of content of thought and linguistic meaning in such a way that if the sceptical solution allowed us to conclude that language is essentially public, then we should also be able to conclude that thought is essentially public. But, when addressing the question of the way in which one could, under this hypothesis, reach the conclusion that thought is essentially public, there would seem to be two possible types of answers. The first one is that this follows from the fact that language is a necessary condition of thought, thus: there is no thought without language, but there can be no language without there being more than one speaker, hence there can be no thought without there being more than one thinker. The second answer (which does not exclude the first) is that we should then be able to formulate a version of the sceptical solution which applies directly to the question of knowing under which conditions one is justified in judging that someone has a certain thought, and that that thought is correct. But if an answer of this second type were possible, it would perhaps no longer be necessary to rely on the sceptical solution in order to conclude that language is public, for in all likelihood, this conclusion would follow from the fact that thought is public, together with the idea that thought is a necessary condition of language, thus: there is no language without thought, but there can be no thought without there being more than one thinker, hence there can be no language without there being more than one speaker. Hence, there seems to be at least three different ways in which one could try to reach the two desired conclusions. However, the foregoing remarks hide a few difficulties that can partly be disclosed by stating more precisely what the publicity of language and thought is supposed to consists in. For the case of thought, one can distinguish at least the following four theses: (1) strong publicity of thought: if an agent has the capacity to think T, then there is another agent who also has the capacity to think T, (2) moderate publicity of thought: if an agent X has the capacity to think T, then it is possible that another agent also has the capacity to think T (or, more accurately, that X is not alone in having the capacity to think T), (3) weak publicity of thought: if an agent X has the capacity to think, then there is another agent who has the capacity of having at least one of the thoughts that X is able to have, (4) anatomicity of thought: if an agent has the capacity to think, then there is another agent who also has the capacity to think. And for the case of language we have the four corresponding theses: (5) strong publicity of language: if an agent speaks a language L, then there is another agent who also speaks L, (6) moderate publicity of language: if an agent X speaks a language L, then it is possible that another agent also speaks L (or, more accurately, that X is not alone in speaking L), (7) weak publicity of language: if an agent X speaks a language, then there is another agent who speaks at least one of the languages spoken by X, (8) anatomicity of language: if an agent speaks a language, then there is another agent who also speaks a language. It will be noted that the weak publicity thesis (i.e., (3) or (7)) presents the advantage of not excluding (as strong publicity does) the existence of a private language or thought, while the moderate publicity thesis (i.e., (2) or (6)) presents the advantage of excluding (as weak publicity does not) the possibility of a private language or thought being necessarily private. Let us now return to the three strategies that could lead us to the desired conclusions (in one form or another). Two of these rely on a specific thesis regarding the relation between thought and language. The first (call it "the L-to-T strategy") aims to show that thought is public (anatomic) by relying on a prior proof of the publicity (anatomicity) of language and of the thesis that thought depends on (and thus implies) language. The second (call it "the T-to-L strategy") aims to show that language is public (anatomic) by relying on a prior proof of the publicity (anatomicity) of thought and of the thesis that language depends on (and thus implies) thought. Since the dependence of thought on language is more controversial than the dependence of language on thought, and since the T-to-L strategy claims to dispense with the first thesis, while the L-to-T strategy claims only to dispense with the second, the T-to-L strategy seems at first glance more interesting. Yet, this appearance is misleading. Depending on the content that one wishes to confer to the publicity/anatomicity thesis, the L-to-T strategy proposes one of the following four (schematic) arguments: (9) from strong publicity of language to strong publicity of thought: (5) If an agent speaks a language L, then there is another agent who also speaks L. If an agent has the capacity to think, then he speaks a language (i.e., thought weakly depends on language). (1) Therefore, (1) if an agent has the capacity to think T, then there is another agent who also has the capacity to think T. (10) from moderate publicity of language to moderate publicity of thought: (6) If an agent X speaks a language L, then it is possible that another agent also speaks L (or, more accurately, that X is not alone in speaking L). If an agent has the capacity to think, then he speaks a language. Therefore, (2) if an agent X has the capacity to think T, then it is possible that another agent also has the capacity to think T (or, more accurately, that X is not alone in having the capacity to think T). (11) from weak publicity of language to weak publicity of thought: (7) If an agent X speaks a language, then there is another agent who speaks at least one of the languages spoken by X. If an agent has the capacity to think, then he speaks a language. Therefore, (3) if an agent X has the capacity to think, then there is another agent who has the capacity of having at least one of the thoughts that X is able to have. (12) from anatomicity of language to anatomicity of thought: (8) If an agent speaks a language, then there is another agent who also speaks a language. If an agent has the capacity to think, then he speaks a language. Therefore, (4) if an agent has the capacity to think, then there is another agent who also has the capacity to think. Quite clearly, (12) cannot be accepted unless one supposes that an agent cannot speak a language without having the capacity to think, hence this argument (as well as the other (stronger) three) implicitly relies on the premise that language depends on thought. The next thing to observe is that this additional premise does not suffice to turn (9) into a valid argument, and this for two reasons having to do with the way in which one should understand the idea that language or thought depend on one another. The two premises of (9) immediately imply that if an agent has the capacity to think T, then each language spoken by that agent is spoken by at least one other agent. From this, the additional premise that an agent speaks a language only if he has the capacity to think allows one to conclude that all the agents speaking one of the languages of the first agent have the capacity to think, but not that everything that the first agent has the capacity to think can be thought by at least one of these agents. To reach this conclusion, some stronger versions of the premise that thought depends on language and of the (implicit) premise that language depends on thought are needed. These are provided, respectively, by saying that (13) if an agent has the capacity to think T, then he speaks a language containing a statement meaning T (i.e., thought strongly depends on language), and that (14) if an agent speaks a language containing a statement meaning T, then he has the capacity to think T (i.e., language strongly depends on thought). (2) Without premise (13), one might find an agent having the capacity of having thoughts he cannot express in any language; in that case, the fact that each language spoken by this agent is also spoken by another would not guarantee that everything that can be thought by this agent can also be thought by another, even while supposing that he must be able to think everything that is expressible in a language he masters (i.e., even while admitting that language strongly depends on thought). While the necessity of adding (14) confirms that the L-to-T strategy does not dispense with the thesis that language depends on thought, it doesn't introduce any further difficulty, insofar as it does't seem more problematic to suppose that language strongly depends on thought than to suppose that it weakly depends on it. But this is not the case for (13), since it seems more problematic to suppose that thought strongly depends on language than to suppose that it weakly depends on it. Yet the L-to-T strategy is still available to someone who wants to hold that thought is moderately or weakly public, for in that case, the arguments proposed by the L-to-T strategy would be (10) or (11) rather than (9). The case of (10) is similar to that of (9), insofar as the addition of (14) does not suffice in making it acceptable, while the addition of both (13) and (14) does. Yet it seems that a weaker premise than (13), and one also easier to defend, could do the job, for example a premise such as (15) if an agent has the capacity to think T, then it is possible that he speaks a language containing a statement meaning T (i.e., thought moderately depends on language). (3) On the other hand, it suffices to add (14) to the two explicit premises of (11) to be able to conclude not only that if an agent X has the capacity to think, then there is another agent having the capacity to think at least one of the thoughts that X is able to think, but also that if an agent X has the capacity to think, then there is another agent having the capacity to think all the thoughts expressible in the languages he shares with X. Hence, this argument does without the hypothesis that thought strongly depends on language (= (13)). (4) To summarize: no version of the L-to-T strategy really dispenses with the hypothesis that language (weakly) depends on thought, and some of them (i.e., (9)-(10)) require more than the hypothesis that thought (weakly) depends on language. This suggests that the T-to-L strategy could also need a premise stating that thought depends on language, and not simply that language depends on thought, which would eliminate one of the advantages it seemed to have over the L-to-T strategy, namely, the advantage of not having the burden of establishing that thought depends on language. This strategy proposes one of the following four arguments: (16) from strong publicity of thought to strong publicity of language: (1) If an agent has the capacity to think T, then there is another agent who also has the capacity to think T. If an agent speaks a language, then he has the capacity to think (i.e., language weakly depends on thought). Therefore, (5) if an agent speaks a language L, then there is another agent who also speaks L. (17) from moderate publicity of thought to moderate publicity of language: (2) If an agent X has the capacity to think T, then it is possible that another agent also has the capacity to think T (or, more accurately, that X is not alone in having the capacity to think T). If an agent speaks a language, then he has the capacity to think. Therefore, (6) if an agent X speaks a language L, then it is possible that another agent also speaks L (or, more accurately, that X is not alone in speaking L). (18) from weak publicity of thought to weak publicity of language: (3) If an agent X has the capacity to think, then there is another agent who has the capacity of having at least one of the thoughts that X is able to have. If an agent speaks a language, then he has the capacity to think. Therefore, (7) if an agent X speaks a language, then there is another agent who speaks at least one of the languages spoken by X. (19) from anatomicity of thought to anatomicity of language: (4) If an agent has the capacity to think, then there is another agent who also has the capacity to think. If an agent speaks a language, then he has the capacity to think. Therefore, (8) if an agent speaks a language, then there is another agent who also speaks a language. It should be clear that (19) is acceptable only if one supposes that an agent cannot have the capacity to think if he does not speak a language, and hence that this argument (again, as well as the other (stronger) three) implicitly relies on the premise that thought depends on language. On the other hand, argument (16) is not conclusive, and remains so even when (13) is added and the second premise is replaced by (14). In other words, it does not suffice to admit that thought is strongly public and that language and thought are strongly interdependent to reach the conclusion that language is strongly public. This argument yields, at most, the conclusion that if an agent speaks a language L, then there is another agent who speaks a language at least as rich as L, i.e., a language in which everything that can be expressed in L can also be expressed. This reveals an interesting, and somewhat surprising, asymmetry between the T-to-L and L-to-T strategies. The case of (17) is similar to that of (16), and manifests the same asymmetry. If adding (13) and (14) does not suffice to make (16) acceptable, then a fortiori it does not suffice (any more than would the addition of (15) and (14)) to make (17) acceptable. Finally, just as (13) played no role in argument (11), (14) seemingly plays no role in argument (18). Moreover, adding (13) leads at most to the conclusion that if an agent speaks a language L, then there is another agent who speaks a language containing a statement meaning the same thing as at least one statement of L. (5) These observations certainly do not allow us to claim that it is impossible to conclude that language is public/anatomic from the idea that thought is public/anatomic, but they may upset the initial impression that the T-to-L strategy is less demanding, or more promising, than the L-to-T strategy. The T-to-L strategy purports to dispense with the much debated thesis that thought depends on language, and to rely exclusively on the almost trivial thesis that language depends on thought, in a manner that reduces the difficulty to that of finding a plausible way of establishing the publicity of thought without presupposing that of language. But, if the preceding remarks are correct, there is some reason to believe that this advantage would remain illusory even if one could find a way of reaching the desired conclusions. It would consequently seem that one can show neither that the publicity/anatomicity of thought follows from that of language, nor that the publicity/anatomicity of language follows from that of thought, without presupposing in both cases that language and thought are interdependent. Furthermore, it would easily be aknowledged that essentially the same conclusions hold when theses (1)-(8) are replaced by stronger and more realistic theses such as the following: (1') if an agent X has the capacity to think T, then there is another agent who also has the capacity to think T and to attribute T to X, (2') if an agent X has the capacity to think T, then it is possible that another agent also has the capacity to think T and to attribute T to X, (3') if an agent X has the capacity to think, then there is another agent who has the capacity to have at least one of the thoughts that X is able to have and to attribute this thought to X, (4') if an agent X has the capacity to think, then there is another agent who also has the capacity to think and to attribute this capacity to X, (5') if an agent X speaks a language L, then there is another agent who also speaks L and has the capacity to attribute L to X, (6') if an agent X speaks a language L, then it is possible that another agent also speaks L and has the capacity to attribute L to X, (7') if an agent X speaks a language, then there is another agent who speaks at least one of the languages spoken by X and has the capacity to attribute this language to X, (8') if an agent X speaks a language, then there is another agent who also speaks a language and has the capacity to attribute the mastery of language to X. But I obviously don't have the space to argue for this here. (6)
|
Notes (1) A strong dependence thesis will be formulated in a moment. (2) In using a conditional, here and elsewhere, to express the idea that a thing depends on another, I do not mean that the dependence relation is reducible to that material implication, but only that material implication is at least a necessary condition for the existence of a dependence relation. (3) Argument (10) then becomes:
(4) It would not even be affected by the addition of this hypothesis, which shows that the idea of the strong dependence of thought on language is compatible with the possibility that there are private languages and thoughts. Moreover, since one could replace the first premise of (11) by the first premise of (9) (i.e., (7) by (5)) without damage to the argument, one also sees that there could be private thoughts even if there could not be any private language, on the sole condition that thought does not strongly depend on language (i.e., on the condition that (13) is false). (5) To these difficulties concerning the T-to-L strategy, one can of course add the fact that it is very hard to see how one could, in the first place, contemplate establishing the publicity/anatomicity of thought without showing or presupposing that thought depends on language. (6) I owe thanks to Paul Bernier, Dave Davies, Martin Montminy, Alain Myette, Michel Seymour, Richard Vallée and Alain Voizard for comments on earlier versions of this paper, and to Robert Waldron for translating it from the French. References Boghossian, P. A. (1989) 'The Rule-Following Considerations'. Mind 98: 507-549. Davidson, D. (1975) 'Thought and Talk'. In Davidson, D. (1984a: 155-170). ____ (1982) 'Rational Animals'. Dialectica 36: 317-327. ____ (1984a) Inquiries into Truth and Interpretation. Oxford: Oxford U. Press. ____ (1984b) 'First-Person Authority'. Dialectica 38: 100-111. ____ (1992) 'The Second Person'. In French, P. A. et al. (ed.) (1992: 255-267). ____ (1994) 'The Social Aspect of Language'. In McGuinness, B. & G. Oliveri (ed.) (1994: 1-16). Fodor, J. A. & E. Lepore (1992) Holism. Oxford: Blackwell. French, P. A. et al. (ed.) (1992) Midwest Studies in Philosophy, Vol. 17: The Wittgenstein Legacy. Notre Dame: Notre Dame U. Press. Heil, J. (1992) The Nature of True Minds. Cambridge: Cambridge U. Press. Kripke, S. (1982) Wittgenstein on Rules and Private Language. Oxford: Blackwell. McGuinness, B. & G. Oliveri (ed.) (1994) The Philosophy of Michael Dummett. Dordrecht/Boston: Kluwer.
|
|
Paideia logo design by Janet L. Olson. |
|
