Questions and Generalized Quantifiers

Abstract

In this paper I present the basic elements of an analysis of matrix interrogatives within the theory of Generalized Quantifiers (Barwise and Cooper 1981, Keenan and Stavi 1986, Keenan and Westerståhl 1994). Current syntactic theories defend a uniform treatment of declarative quantifiers such as every, some and most, and interrogative quantifiers such as which or what. For instance, theories of quantification within the Government and Binding framework and later developments assume that movement of declarative and interrogative quantifiers is subject to the same constraints on well-formedness (subjacency, government, both types of quantifiers bind variables, etc.) at the relevant level of representation. Thus, at least from a syntactic point of view, it seems desirable to give a uniform account of the different types of quantification. Prom a semantic point of view, we also have strong arguments in favor of the proposed connection. It can be easily noticed that interrogative quantifiers have the property of Conservativity, a property satisfied by all declarative determiners (Keenan and Stavi 1986). This sepmantic property explains why sentence (1a) is equivalent to (1b), and sentence (lc) is equivalent to (1d).

I wish to thank Ed Keenan for hours of discussions and extensive comments on several versions of this paper. I would also like to thank Anna Szabolcsi, Irene Heim and the audiences at the Fourth CSLI Workshop on Logic, Language and Computation and the Tenth Amsterdam Colloquium, especially Johan van Benthem, William Ladusaw and Barbara Partee for their comments.