Inferential patterns of generalized quantifiers and their applications to scalar reasoning.

摘要

This thesis studies the inferential patterns of generalized quantifiers (GQs) and their applications to scalar reasoning. In Chapter 1, I provide an overview of the Generalized Quantifier Theory (GQT) and expand the scope of study of this theory to left-oriented GQs. In Chapter 2, I introduce the main aspects of scalar reasoning to be studied in this thesis, which include two main types of scalar reasoning, i.e. scalar entailments (SEs) and scalar implicatures (SIs), and four types of scalar lexical items, i.e. scalar operators, climax construction connectives, subjective quantity operators and lexical items denoting extreme values. Some outstanding problems are also identified. In Chapter 3, I study four main types of quantifier inferences, namely monotonicity inferences, argument structure inferences, opposition inferences and (non-classical) syllogistic inferences. The research findings are summarized in a series of theorems and principles. Special emphasis is put on devising general principles and methods that enable us to derive valid inferential patterns of iterated GQs from the inferential properties of their constituent monadic GQs. In Chapter 4, I apply the major findings worked out in Chapter 3 to resolve the outstanding problems identified in Chapter 2 by developing a formal framework for SEs and SIs. The framework is also used to conduct a cross-linguistic study on scalar lexical items in English and Chinese introduced in Chapter 2. The conditions of use for these lexical items are formulated and the association of SEs and SIs with different types of scalar lexical items is also explored. Chapter 5 discusses the significance of the major findings of this thesis and possible extensions of the study.