Using Focus to Improve Definition: What Counts in Quantification

Abstract

Szabolcsi (1997a) proposes that some left-peripheral syntactic positions encode not compositional semantic information, but different procedures for the assessment of truth conditions. These procedures are said to be reflected in the range of quantified noun phrases that may appear in certain positions in the pre-verbal field in Hungarian. While one of Szabolcsi’s proposed procedures correctly predicts a monotonicity-based constraint on the quantifiers appearing in certain positions, her other procedure is too vaguely defined to produce useful predictions. I argue that the appropriate procedure in this latter case is the same as the one that produces ‘narrow focus’ interpretations and that the related position that Szabolcsi proposes for quantificational processes is nothing other than the well-known ‘focus position’ of Hungarian. Apparent interpretive differences between the relevant quantificational phrases and other uses of syntactic focus follow naturally from an inferential pragmatic approach to this position. This has important theoretical implications: an inferential analysis of syntactic focus requires (1) a ‘dynamic’, parsing-based view of the relationship between syntax and semantics and (2) a re-alignment of the burden of explanation between linguistically encoded semantics and inferential pragmatics. An analysis of this nature proves to explain the quantifier distribution facts and a number of other syntactic phenomena in an extremely parsimonious fashion.