Cause identifiability and the causative alternation in English: A corpus-based analysis

Cause identifiability and the causative alternation in English: A corpus-based analysis

Many verbs in English show causative and noncausative uses. The goal of this paper is to verify whether the frequencies of the causative and noncausative uses of alternating verbs are modulated by the identifiability of the ultimate cause of the event in a given discourse context. On the basis of a corpus study of 12 change-of-state verbs, it will be shown that verbs that are predominantly used as a causative tend to have a cause argument with a higher degree of identifiability, while verbs that are more frequently used as a noncausative tend to have a cause argument with a lower degree of identifiability. Based on this evidence, I argue that cause identifiability in context is a crucial factor that more accurately characterizes verb-specific preferences for one variant over the other as well as the corpus frequency distribution of the alternation variants than other lexical semantic factors discussed in the literature.