Dummett's Criteria for Singular Terms

Abstract

In his excellent book Frege: Philosophy of Language, Michael Dummett proposes five criteria that are supposed to delimit formally the class of singular terms (in Frege's terminology, 'proper names').' I shall argue that Dummett's criteria, even as amended by Crispin Wright and Bob Hale, do not delimit the class of singular terms in English,2 that further refinements of the criteria probably will not help, and that in any event the criteria face other difficulties. For Frege the class of singular terms includes what might be called proper proper names (like 'New York' and 'Socrates'), and most definite descriptions (like 'the longest river in North America', 'that lime tree', 'the famous teacher of Plato', 'the colour red', 'the number nine', 'the weight of the earth', 'the shape of the earth', and 'the equator'). It does not include sentences, predicates, or general terms like 'nothing', 'nobody', 'something', 'everything', 'every sheep', or 'a poet'. Of course these brief remarks are merely intended to give someone proficient in English the general idea. Dummett's criteria are meant to be more formal. Why should it be thought important (to a Fregean at any rate) to delimit formally the class of singular terms? If Michael Dummett and Crispin Wright are correct, then as Wright puts it, for Frege 'the notion of an object is posterior in the order of explanation to that of a singular term'.3 That is, syntax is, in a sense, 'prior to' ontology; to determine what there is, determine first what the syntax is. The objects are simply the referents of the singular terms, and the functions are the referents of the predicates. Predicates can be defined in terms of singular terms and sentences as follows: a predicate is what remains when one or more singular terms has been removed from a sentence. (We assume that we have some means for determining what a sentence is.) Hence everything hinges on the singular terms. On Dummett's understanding of Frege, then, it is crucial to the Fregean programme to characterize the class of singular terms. Besides, the idea that the singular terms might be a guide to what there is has much to recommend it, Quine notwithstanding.