How best are we to understand the significance-for a proper account of meaning and propositional attitude content-of the sorts of considerations to which Frege appeals in order to individuate Gedanken, considerations that reveal differences in the cognitive significance of different sentences? Providing anything close to a satisfactory account of what underlies such differences proves to be surprisingly difficult. We all know, for example, that the exchange of coreferring proper names in a sentence will typically change the cognitive significance of that sentence for speakers of the language. But why should that be? It is not much help to be told that this possibility arises because a speaker may not believe that the object to which she is referring with the one name is the same as the object to which she is referring with the other. How, after all, are we to understand that possibility? The problem here is, at bottom, the same as before. Nor, in the end, does it help much to say that two such names, while alike in reference, must differ in their Fregean senses. That is, at best, to propose a strategy for solving the problem. The strategy takes it for granted that the content of a sentence is compositionally determined by the contents of the words that make it up. It then postulates that associated with the relevant two names are different cognitively relevant values or senses, each of which makes a distinct compositional contribution to determining the cognitive significance of the sentences in which they appear. But without a plausible substantive account of the notion of sense, the suggestion that sentences of the relevant sort differ in cognitive significance because the counterpart names differ in sense is not yet an explanation, and little more than an empty promise.