How does natural language quantify?

Abstract

It has traditionally been assumed that Natural Language uses explicit quantifier expressions (such as and most, and a) for the purpose of quantification. We argue that expressions of the first type are comparatively rare in real world Natural Language sentences, and that the latter (articles) cannot be considered straightforward quantifiers in the first place. However, practically all applications of Natural Language Processing require sentences to be quantified unambiguously. We list a few possible (syntactical, semantical, and pragmatical) sources of implicit quantificational information in Natural Language; they combine in sometimes intricate ways to give a sentence a (more or less) unambiguous quantification.