Descriptions and uniqueness

Abstract

This paper argues against the Russellian theory of definite descrip- tions. In contending that this theory is inadequate, I am certainly not alone; in the past fifty years it has seen a vast number of attacks. What distinguishes my positive proposal from most – although by no means all – other accounts in the literature is how far it goes in agreeing with the Russellians. For I don’t contest the claim that definite descriptions can be identified with devices of quantification; I believe that the Russellian theory is mistaken only in picking out the quantificational device.1 The paper divides into three sections. In the first, I introduce my semantic proposal accompanied by certain considerations that give it its initial appeal; I also briefly counter two immediate objections. In the second, the semantic proposal is supplemented by a pragmatic one. In the third, I offer arguments for the resulting view; I follow the Russellians in claiming that a pure quantificational view of descriptions is preferable to views according to which descriptions sometimes refer, and then argue that among the pure quantificational views my theory fares better than the Russellian alternative.