Complexities of the attributive use of definite descriptions

Abstract

Keith S. Donnellan’s article “Reference and Definite Descriptions” (1966) was written with the intention of showing that the theories of Bertrand Russell and Peter F. Strawson equally fail to capture the actual linguistic use of definite descriptions. With this end, Donnellan dwells on the distinction between the attributive and referential uses. However, the situation is complicated for Donnellan in light of the fact that many conceivable uses of definite descriptions do not seem to fit into either category. There are uses that should be understood as deferential rather than attributive, just as there are uses in which various possibilities of referential failure committed by speakers result in situations in which the description simply does not account for the variety of possibilities, that is, cases in which one cannot speak of either a referential or an attributive use. The reason for this seems to be that there is hardly a “pure” attributive use. The descriptions used in common examples usually contain referential elements, such as nouns and indexicals. When these are misused, the inadequacy of Donnellan’s distinction is exposed. The second thesis presented here is that the attributive uses envisaged in Donnellan’s classic paper are characterized by their distinctive inferencial nature, with the consequence that the attributive use is generally less natural and plausible
than the referential use.