A computational approach to collectivity and distributivity in Turkish quantificational sentences

Abstract

This dissertation discusses the nature of quantificational sentences with the quantifiers her “every” and bütün “all” in Turkish and investigates how quantificational sentences get interpreted in terms of collectivity / distributivity. Departing from previous approaches, the study proposes to examine the role each constituent of a sentence attributes to the interpretation of a sentence and claims that the distributive or collective interpretation of a sentence is a function of the meanings of its constituents pertaining to the notion of collectivity and of the way they are combined. The model proposed in the dissertation puts forward a computational interpretation mechanism referred to as the Modified Plus Principle which operates on the feature of [alphaCOLL] that every constituent in a structure is assumed to bring from the lexicon. This interpretation mechanism enables us to argue that lexical items associated with the [-COLL] feature such as the distributive predicates, the quantifier her “every” and singular nouns are potentially have a more influential role than the ones associated with the [+COLL] feature such as the collective predicates, the quantifier bütün “all” and plural nouns in interpreting a sentence as either collective or distributive.