The Language of Propositions and Events: Issues in the Syntax and the Semantics of Nominalization

Abstract

A theory of nominalization should specify the relation between noun meaning and verb meaning. At least for some classes of nouns, such a theory should also provide a general and systematic way of deriving noun meanings from verb meanings. This is the case, for example, for event-denoting $ing\\sb{\\rm of}$-Nouns. The meaning of these nouns must be derived by a rule from the meaning of the corresponding verb, since there is evidence that they are not listed in the lexicon. A theory of nominalization should also account for the meaning differences and the distributional differences among various kinds of nominals and that-clauses. Following Zeno Vendler, I argue that some of these differences are accounted for by the distinction between propositional entities and event-like entities, and by the semantic selection properties of predicates. I propose a semantics for English nominalization based on Kratzer's (1987) semantics of situations. The central notion of this semantics is the part of-relation among possible situations. This theory provides a general way of recovering the meaning of event-nouns from the meaning of the corresponding verb. Moreover, it is able to account for the distributional and semantic differences among different kinds of nominals and that-clauses. A compositional approach to the semantics of NPs raises the issue of the argument structure of nouns. I discuss different accounts of the role of the of-phrase and of the by-phrase with event-nouns. I provide some evidence that they have argument status. Italian infinitival NPs provide some cross-linguistic evidence for this conclusion. The semantics for English nominalization I develop serves also as a tool for investigating the difference of interpretation derived (and $ing\\sb{\\rm of}$) NPs show in different contexts. I address the issue of why these NPs are paraphrasable by that-clauses in some context, but not in others.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 82 citations exist upstream.