Questions about the connotations of proper names have produced much philosophical discussion. In this paper, I push a plausible Fregean argument to its logical conclusion. I argue that there are important cases where this argument holds, but where Frege failed to apply it, and that its correct application leads to a new view of the connotations of proper names. The view I present accords with the insights behind this argument and also incorporates some insights of Kripke and Donnellan. Thus, my view provides an alternative to both major types of views of names, but retains some advantages of each.
Much of my discussion relies on the following Fregean argument. Let 'TI' and 'T2' be coreferential singular terms.
-
Suppose there is at least one propositional attitude B such that it is possible for someone to have B that T1 is T1 without having B that T1 is T2, or vice versa.
-
The proposition that TI is TI is not identical with the proposition that T1 is T2.
-
The sentences 'T1 is T 1' and 'T1 is T 2' express different propositions. 4. 'T 1 ' and '7"2' differ in connotation. 5. Therefore, if there is at least one propositional attitude B such that it is possible for someone to have B that T1 is T1 without having B that T1 is T2, or vice versa, then 'TI' and 'T 2 ' differ in connotation.
Reasoning in accord with this argument is common, 1 but some discussion of it may be useful. One justification for the proposition individuation condition of step 2 is as follows. Leibniz's Law holds that for any property F and any entities x and y, if x is identical with y, then x has F iffy has F. Therefore, if the proposition that T1 is Ta is identical with the proposition