Content Nouns and the Semantics of Question-Embedding

Abstract

In this paper, we focus on the semantics of the first class of attitude verbs, such as know, forget and tell, which can select either a declarative or an interrogative complement. Hereafter, we refer to this class of attitude verbs as RESPONSIVE PREDICATES, following Lahiri (2002). One of the basic issues in the semantics of question-embedding concerns the selectional property of these predicates. Namely, how we can semantically account for their compatibility with both declarative and interrogative complements. This question does not have a straightforward answer if we assume that responsive predicates are unambiguous, and that declarative complements and interrogative complements denote different kinds of objects, i.e., propositions and questions. The standard answer to this question states that the basic denotation of responsive predicates selects for a proposition, which is the meaning of declarative clauses, and assumes some form of reduction from the meaning of embedded interrogatives to propositions (e.g., Karttunen 1977; Groenendijk and Stokhof 1984). However, such an account wrongly predicts that a believetype predicate would be able to embed an interrogative complement if it were not for further stipulations.1 For example, in Groenendijk and Stokhof’s (1984) theory, the intension of an interrogative clause is a propositional concept (i.e., a function from worlds to propositions). Thus, the extension of an interrogative clause is a proposition (corresponding to the true exhaustive