A corpus-driven study of lexicalization models of English intransitive verbs.

Abstract

This bottom-up study applied a corpus-driven approach to extract the major lexicalization models of English intransitive verbs (EIVs) through an analysis of their meanings under the Ideal Motion Event Category inspired by Talmy. A 710-photo specialized multimodal corpus of EIVs was constructed for this purpose. Data analysis showed that the major lexicalization models of EIVs include [Motion + Patient], [Motion + Manner], [Motion + Path], [Motion + Result], [Motion + Location], and [Motion + Purpose]. In-depth analysis of these models identified three major possible reasons why EIVs cannot originally take direct objects: the incorporation of [Patient], the inheritance of the intransitive feature, and the internalization of the actions expressed by EIVs. By comparing Chinese learners' most misused EIVs with their corresponding Chinese verbs, the current study provides empirical data to illustrate why transitive misusages of EIVs might occur among Chinese learners. The findings of this study will help English learners and users better apply EIVs.