Nominalization, Prediction and Type Containment

摘要

Type disciplines have featured prominently in formal approaches to naturallanguage. The paper examines type-free theories of properties, in order to accommodate the difficulties raised by nominalization and self-application. Historically, type disciplines for languages have developed in close association with intended models for interpretation. The proposals the authors shall make can also be construed in this way, inasmuch as they were inspired in part by the notion of a Frege structure, which is intended to provide a consistent formulation of Frege's logical notion of set. A system of types provides a classificatory scheme for the domain and range of functors. The type of an expression determines the domain in which that expression receives an interpretation.