A Study of change-of-state verbs and Frame Semantics: Focusing on the Break verb class

Abstract

Levin’s classification has often been criticized for relying primarily on native-speaker intuitions in mapping meaning to syntax. This classification does not sufficiently account for actual usage frequency, discourse context, or the possibility of cross-classification arising from polysemy. In response, this study reanalyzes the Break verb class by applying Fillmore’s (1982, 1985) framework of Frame Semantics. By integrating the FrameNet-based Frame Element (FE) analysis into Levin’s classification, this study highlights the strengths of Levin’s intuition-driven framework while addressing its limitations by incorporating corpus data, discourse sensitivity, and polysemy. In conclusion, combining Levin’s event-structural perspective with Fillmore’s frame-based model enriches both the theoretical and empirical accounts of the semantics of change-of-state verbs and suggests the potential for expanding this research into a more context-sensitive, usage-based approach.