Abstract

Interrogatives appear to present a problem for the received view in semantic theory that the meaning of a natural language sentence is given by its truth conditions. For although meaningful, interrogatives are neither true nor false. If interrogatives are assigned truth conditions, then their meaning is incorrectly specified, but if they are not assigned truth conditions, then their meaning is not specified at all. ;One response to this problem is to explain it away and assign interrogatives suitable truth conditions despite their intuitive lack of a truth value. We critically discuss two such proposals--Donald Davidson's paratactic analysis of non-indicatives and the standard performative analysis--arguing that both are unsatisfactory. A second response is to abandon truth-conditional semantics entirely, but such drastic measures are unwarranted. A third response, and the one pursed in this dissertation, is to selectively revise the truth-conditional semantic framework. The recursive specification of truth conditions is retained for indicative sentences, but the semantics of interrogatives is given in terms of their function in discourse. ;According to the discourse prompt analysis advanced below, interrogative sentences indicate that the next appropriate step in the discourse is a true possible answer form, where the range of these possible answer forms is determined by the grammatical structure and content of the interrogative. By way of illustration, we develop a formal discourse prompt theory for a fragment of English that includes both indicatives and interrogatives. We defend the adequacy of this account by arguing that it satisfies independently motivated constraints on semantic theories. We then discuss, in some detail, how the relevant possible answer forms may be specified for a range of English interrogatives. Finally, we explore the relation between the discourse prompt analysis of interrogatives and pragmatic theory. ;We conclude that the discourse prompt analysis is a well-motivated, intuitively satisfying account of what interrogatives mean, and that integrating this analysis with a truth-conditional account of indicatives results in a better overall theory of meaning