A report of a person's desire can be true even if its embedded clause underspecifies the content of the desire that makes the report true. It is true that Fiona wants to catch a fish even if she has no desire that is satisfied if she catches a poisoned minnow. Her desire is satisfied only if she catches an edible, meal-sized fish. The content of her desire is more specific than the propositional content of the embedded clause in our true (...) report of her desires. Standard semantic accounts of belief reports require, however, that the embedded clause of a true belief report specify precisely the content of the belief that makes it true. Such accounts of belief reports therefore face what I call "the problem of underspecification" if they are extended to desire reports. Such standard accounts are sometimes refined by requiring that a belief report can be true not only if its subject has a belief with precisely the propositional content specified by its embedded clause, but also only if its subject grasps that content in a particular way. Such refinements do not, however, help to address the problem of underspecification for desire reports. (shrink)

According to James McCawley (1981) and Richard Larson and Gabriel Segal (1995), the following sentence is three-ways ambiguous: -/- Harry wants to be the mayor of Kenai. -/- According to them also, the three-way ambiguity cannot be accommodated on the Russellian view that definite descriptions are quantified noun phrases. In order to capture the three-way ambiguity of the sentence, these authors propose that definite descriptions must be ambiguous: sometimes they are predicate expressions; sometimes they are Russellian quantified noun phrases. After (...) explaining why the McCawley-Larson-Segal solution contains an obvious flaw, I discuss how an effort to correct the flaw brings to light certain puzzles about the individuation of desires, about quantifying in, and about the disambiguation of desire ascriptions. (shrink)

I want to discuss a certain argument for the claim that definite descriptions are ambiguous between a Russellian quantificational interpretation and a predicational interpretation.1 The argument is found in James McCawley’s (1981) book Everything Linguists Have Always Wanted to Know about Logic (but were ashamed to ask). The argument has also been resuscitated by Richard Larson and Gabriel Segal in their more recent (1995) book Knowledge of Meaning.2 If successful, the argument would not only show that descriptions have both quantificational (...) and predicational interpretations, but would also provide confirmation for the commonly held view that the verb ‘to be’ is ambiguous in its interpretation—that it sometimes expresses the ‘‘‘is’ of identity’’ and sometimes the ‘‘‘is’ of predication.’’ But the argument is not successful; it contains an obvious flaw. What’s interesting is that when you try to correct the flaw, certain puzzles arise about the nature of propositional attitudes and the semantics of their ascriptions. (shrink)