Wallace on Propositional Attitudes

Citation preview

Wallace on Propositional Attitudes Robert C. Stalnaker The Journal of Philosophy, Vol. 66, No. 22. (Nov. 20, 1969), pp. 803-806. Stable URL: http://links.jstor.org/sici?sici=0022-362X%2819691120%2966%3A22%3C803%3AWOPA%3E2.0.CO%3B2-I The Journal of Philosophy is currently published by Journal of Philosophy, Inc..

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/jphil.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 Wed Apr 18 07:25:03 2007

803

WALLACE ON PROPOSITIONAL ATTITUDES

belief thesis. In both cases, however, his attack has forced us to clarify what we mean by justification. In particular, we have gained the important insight that the justification for accepting a proposition is not always transmissible to propositions that it entails.10 I. THALBERG

Universiy of Illinois at Chicago Circle WALLACE ON PROPOSITIONAL ATTITUDES

F

ROM a set of seemingly innocuous premises, John Wallace has derived the conclusion that the ideally rational man . wants every proposition to be true that he believes to be true.* It is indeed surprising that so strong a conclusion can be got from his assumptions; nevertheless it is clear which premise causes the problem. T h e sixth premise-the one Wallace defends only with reservations-seems to be clearly false. T h e premise is as follows: "(6) T h e ideally rational agent wants b to F if he wants a to F and believes that a = b-this for any pair of singular terms a and b and any open sentence F (146). A counterexample is as follows: Eldridge wants the President of the U.S. to be black and believes that the President of the U.S. is Nixon, but he does not want Nixon to be black. He might, I suppose; but it surely does not follow that he does. It does not follow because the way in which Eldridge envisions his want being satisfied is for someone else-perhaps himself-to be the referent of the description 'the President of the U.S.' Even though he believes the President to be Nixon, he wants the President not to be Nixon. These being rational wants and beliefs, and Eldridge being ex hypothesi a rational man, it cannot follow that he wants Nixon not to be Nixon, since that is a contradiction. T h e counterexample seems obvious enough, especially when one sees how the premise is used in the argument. What is the basis of Wallace's defense of it? T w o considerations count against giving u p premise (6). One: i n a large number of cases, it gives plausible results. Someone who believes Linus to be Peter's father and who wanted Linus to come i n second but did not want Peter's father to come i n second would be unreasonable. There seems to be a general rule for replacing believed equals lo Many of the ideas in this paper developed from an essay ("A Mistake about Justified Belief") that I read at the x ~ v t hInternational Congress of Philosophy, Vienna, September 3, 1968. * "Propositional Attitudes and Identity," this JOURNAL, LXVI,6 (Mar. 27, 1969): 145-152. Page references to this article are given in parentheses following quotations.

804

THE JOURNAL OF PHILOSOPHY

for equals i n desired contents; if (6) overstates the rule, there is still a problem of stating it correctly. Two: in English, the placing of the infinitive construction in the natural English expression of desired contents seems to mark exactly where one can and cannot replace believed equals for equals. Singular terms that have occurrences outside the infinitive may be replaced; they have a purely referential role in the desired content. Singular terms that occur only within the informative may not be replaced; they have a predicative role i n the [plays] desired content. . . These facts suggest that the infinitive a systematic logical role in English that is not reflected in current logical theories (15 1).

.

...

In the remainder of this note, I shall argue that the qualifications and distinctions suggested can be made in the context of current logical theory of attitude and c0ntent.l Consider the example about Linus and Peter's father: what principle is it that renders that set of beliefs and desires unreasonable? There are, I think, at least two very different answers to this question. First, for the most part we do our wanting in a context in which some things are presupposed to be fixed and unalterable. We may wish that such things had been, or would be, different, but we do not want them to be different. The distinction between what is accepted as already fixed and what is thought to be open determines rational wants in the following way: the ideally rational agent wants anything that is necessary, given the context of unalterable truth, in order to fulfill any of his other wants. That is, the rational man wants it to be the case that q if q is a consequence of any other want in conjunction with the set of unalterable truths.2 On this principle the pair of wants concerning Linus and Peter's father is unreasonable, not because the identity statement is believed, but because it is believed to be fixed-not in doubt or subject to change. What is believed may differ from what is accepted as fixed, since one may believe that something is true, but think that it might not remain so, or one may believe that something will be the case, but that it is not yet decided for sure. I By "current logical theory" I have in mind the semantical theories of modal logic and propositional attitudes as developed by Kripke, Hintkka, Scott, Montague, and others. 2 On this principle, everything that is accepted as fixed is "desired" vacuously. So if we replaced 'believed' by 'accepted as fixed' in Wallace's premises, the paradox would be embraced rather than avoided. It would be odd to say that someone wanted it to be the case that q merely because q was inevitable, since want statements are used to make distinctions among the open possibilities. T o avoid the oddness, we may say that something is wanted in the strict sense if it is (a) wanted in the broad sense defined above, and (b) is not accepted as unalterable.

WALLACE ON PROPOSITIONAL ATTITUDES

805

This explanation of the unreasonableness of the pair of wants has nothing to do specifically with identity or with substitution inside or outside of infinitives. It has to do instead with the fact that wanting, like intending and being obliged, but unlike wishing, is closely tied conceptually to what is possible. But one can change the example so as to avoid this connection, and thus require another explanation. "One who believes Linus to be Peter's father, and who wished that Peter's father had come in first, but did not wish that Linus had come in first is unreasonable." This is not so clear as the wanting example, but there is perhaps a little unreasonableness to be explained. If so, then the first explanation will not work, since the wishing statements are clearly counterfactual. Open possibilities are irrelevant. T h e conflict between the two wishes does have to do, I think, with identity, and with whether or not the singular terms are in purely referential position. "a wants b to F" is ambiguous. On the one hand, it may mean roughly that a wants "b F's" to be true; on the other hand, it may mean that a wants the person who is in fact b to F-in other words, that b is wanted by a to F. The latter reading is the referentially transparent sense, exactly analogous to the transparent belief contexts discussed by Q ~ i n eIn . ~ the transparent sense, the wishes that Peter's father come in first and that Linus come in first are identical, not because Linus is believed to be Peter's father, but because he is Peter's father. In the opaque sense, the wishes are independent since the propositions are independent. Hence the conflict arises only if the wish statements are read transparen tly. It may be that, in English, the distinction between referential and nonreferential occurrences of singular terms is sometimes indicated by placement in infinitive and subjunctive constructions. Current logical theory dealing with propositional attitudes does not have subjunctive mood or infinitive forms, but it does have a way of representing the distinction. I t is simply a matter of scope, or of order of application of operators. We may first construct the complex property "wanted by a to F" and then ascribe it to b, or we may instead first construct the proposition Fb and then apply the propositional operator "wanted by a" to it. With extensional operators and predicates, these two procedures will result in the same complex proposition, but with intensional ones they may be different. 3 W. V. 0.Quine, "Quantifiers and Propositional Attitudes," this JOURNAL, LIII, 5 (Mar. 1, 1956): 177-187.

806

THE JOURNAL OF PHILOSOPHY

Using an abstraction operator, the difference can be formally represented. ('2Gx' expresses the property of being G. 'x^Gx(c)'is read "c has the property of being G." W, is the wanting operator, "a wants it to be the case that.") "a wants b to F" in the opaque sense is formalized by 'W,Fb' or, equivalently, by 'W,AFx(b)', "a wants b to F" in the transparent sense is formalized by 'AW,Fx(b)'. In the first formula, the term b is within the scope of the want operator; in the second it is not.4 So we may say that there are two possibly relevant explanations for the inference from "a wants b to F" to "a wants c to F in a context where b is believed identical with c. First, if the want statements are taken in the opaque sense and if the identity statement b = c is presupposed to be unalterably true, then the inference holds simply because the content of the second want is a consequence of the content of the first, together with the contents that are presupposed unalterably true. Second, if the want statements are taken in the transparent sense and if the identity statement is true (whether it is believed or not is beside the point), then the inference holds because of the substitutivity of identicals in transparent contexts. T h e plausibility of Wallace's premise (6), I hold, derived from one or both of these principles. But premise (6) itself is false no matter which way the ambiguity in the want statement is taken. The example about Eldridge shows that it fails on the opaque reading (this is the reading needed for the derivation of the paradox). It is equally clear that it fails on the transparent reading, since believed identity propositions may be false.6 Neither of the two revised principles is sufficient to derive the paradoxical conclusion. ROBERT C. STALNAKER

University of Illinois at Urbana ON WHAT THERE NEED NOT BE

0

N E of Quine's dominant themes is the avoidance of unnec-

essary entities. The ideal of such an ontological program is to finish with an ontology that is both plausible and acceptable. Thus the standards for the repudiation of putative entities must be sufficiently liberal to avoid including more entities than one 4For a formal development and philosophical exposition of this device, see R. Stalnaker and R. Thomason, "Abstraction in First-order Modal Logic," Theoria, xxx~v,3 (1968): 208-207, and R. Thomason and R. Stalnaker, "Modality and Reference," Nods, rr, 4 (December 1968): 359-372. 6 If a rational man believes that b = c and believes that he wants b to F (transparent sense), then he will believe that he wants c to F. But want statements, read transparently, are statements we can be mistaken about.