Can definite descriptions be used as proper names?

Referential use of definite descriptions in Donnellan’s sense can be accounted for in semantic and pragmatic terms. One of the semantic accounts was outlined by Kaplan and elaborated in detail by Marti. According to their theory, when a definite description is used referentially, (1) it is used as a throw-away proper name, and (2) its attributive semantic content loses any semantic relevance. (In this theory, proper names are treated as directly referring devices.) The paper presents a criticism of this view. I argue that in typical cases of referential use, the attributive semantic content of a definite description is pragmatically relevant, and because of this, only subsequent use can show its semantical irrelevance. But the subsequent use is possible only for stable names, not for throwaway names. Thus, (2) can only be grounded by evidence that (1) excludes, which undermines the empirical basis of the theory in question. I conclude that this theory has no advantage over the pragmatic account of referential use, and that the pragmatic account is preferable for reasons of parsimony.