A frame-semantic analysis of five English verbs evoking the Theft frame

Abstract

An important problem in lexical semantics is the explanation of how verbal meaning interacts with the syntactic realization of arguments. Levin (1993) recognizes the relation between syntax and semantics in her classification of English verbs, in which similar syntactic behavior among verbs is assumed to reflect shared meaning components. However, her classes do not accurately predict the verbs’ semantic and syntactic properties. Other researchers (Taylor 1996, Boas 2008) argue for the inclusion of detailed encyclopedic meaning in explanations of syntactic behavior. Frame Semantics provides the necessary tools for fine-grained analyses of the syntax-semantics interface because it offers a rigorous method for the description of meaning and documents syntactic information about argument realization from corpus data.

This report uses concepts from Frame Semantics and data from its practical application, FrameNet (http://framenet.icsi.berkeley.edu), to assess the importance of fine-grained verbal meaning for argument realization by comparing the verbs embezzle, pilfer, shoplift, snatch and steal. Each verb construes the general semantics of the Theft frame differently, emphasizing or specifying individual participants in the event (frame elements). They also exhibit subtle differences in whether and how these frame elements are syntactically realized. In linking their syntax to their semantics, I show that the verbs’ syntactic distribution may be influenced by aspects of meaning such as their degree of descriptivity, the detailed specification of certain frame elements, and their occurrence as LUs in different frames.