Felix Benzant

John Stuart Mill once claimed that all names are names of something, whether real or imaginary. A similar position is held by advocates of a somewhat unified collection of theses often called the Direct Reference Theory of Proper Names (DRT). One of the core tenets of DRT is the idea that proper names have no semantic role other than to refer. A longstanding alternative, associated with Frege, Russell, Quine, and others, takes proper names to be semantically tied to descriptive content, or else to be analyzable by means of descriptions and quantificational structure. These two broad traditions—referential and descriptivist—have, in different ways, long ruled the roost. But despite their traditional and widespread popularity, there are familiar problems associated with each. My aim is to show that neither theory adequately captures ordinary talk involving proper names, either because (i) there are sentences whose proper interpretations resist these familiar analyses, (ii) they incorrectly assimilate meaning to reference, or (iii) they treat logical regimentation as though it were a theory of meaning. These and other problems seem to turn on their shared commitment to the idea that proper names have a single, unifying semantic role. And, as I shall argue, this is incorrect. In place of these accounts, I propose a sketch of a view on which proper names have various semantic functions. Sometimes they are used as vehicles of direct reference, as in cases of ostension, and other times they are used for purposes other than referring at all. This view takes meaning to be associated with cognitive, descriptive content. In short, and contrary to received wisdom, it takes reference and meaning to have less in common with each other than is often appreciated—or so I argue. In the process, I also examine some of the implications of these considerations with regard to particular discussions in metaphysics, namely, ontology and meta-ontology.