Are Complex ‘That’ Phrases Devices of Direct Reference?*

Since the seminal work of David Kaplan @1977#, the orthodox view of complex ‘that’phrases ~e.g. ‘that man drinking a martini’—henceforth I shall refer to these as ‘that’ phrases! has been that they are devices of direct reference. According to this view, the propositional contribution of a ‘that’ phrase as it occurs in a context is an individual. This individual is picked out by the character of the ‘that’ phrase in question, where character is ~or at least is represented by! a function from context to content or propositional contribution. The character of the ‘that’ phrase in turn is determined by the demonstration or accompanying intention associated with the ‘that’ phrase, together with the descriptive material combined with ‘that’ to form the ‘that’ phrase. Though there may be some disagreements concerning the details of the proper directly referential story and though there have been some recent challenges to this orthodoxy, I think it is fair to say that most philosophers with any view on the matter subscribe to the orthodoxy. Despite this consensus, there is a variety of data that suggests that ‘that’phrases are not directly referential, but are in fact quantifier phrases. In the present work, I intend to discuss this data and, to a first approximation, formulate a quantificational semantics for ‘that’ phrases. Those who espouse a directly referential semantics for ‘that’ phrases tend to focus on certain very particular uses of such phrases. They tend to consider only those uses in which a ‘that’ phrase is employed, along with a demonstration, to “talk about” something or someone in the ~physical! context of utterance. Though the direct reference account is plausible as applied to such uses, there are other uses of ‘that’ phrases for which the account seems problematic. To begin with, there are uses of ‘that’ phrases in which they not accompanied by any demonstration, need not be used to talk about something present in the NOÛS 33:2 ~1999! 155–182