Expressivity and Multidimensional Semantics

Abstract A multidimensional semantics is a semantic system that assigns, simultaneously, more than one kind of meaning to an expression, which can be kept more or less separated from each other. This corresponds to one of the key insights from Kaplan’s paper that kicked off the study of expressive meaning in formal semantics: for many natural language expressions, we need to calculate both descriptive (truth-conditional) and expressive (use-conditional) meaning content while restricting how the truth-conditional and use-conditional meaning dimensions interact with each other. This chapter provides an overview of three semantic systems that have implemented a multidimensional semantics for expressive meaning: LCI and its expansion modifications and LCI+ and LTU. It focuses on the core design decisions that each of these systems makes, especially with respect to how compositionally is conceptualized and how different kinds of expressives, like shunting or mixed expressives, can be analyzed.