A corpus-based study of collocations in Chinese-to-English consecutive interpreting

Abstract

n consecutive interpreting, interpreters, under the constraints of alternate processing of two languages, the mental lexicon of source and target language, and a limited capacity of working memory, are expected to produce speeches that conform to target language conventions. Collocations, habitual co-occurrences of words for more than purely grammatical reasons, are indicators of the naturalness of language use. Seeking to understand features of collocations in second-language interpreted speeches and possible factors leading to the features, this study investigates the use of collocations by Chinesespeaking trainee interpreters in Chinese-to-English consecutive interpreting. Collocations were extracted and compared across two self-built corpora: a corpus of interpreted speeches by trainee interpreters at mock conferences and a corpus of spontaneous speeches by first-language speakers of English. The frequency, diversity, complexity, and naturalness of ten types of collocations in interpreted speeches were analysed with reference to English conventions. To understand underlying reasons for typical collocation features of interpreted speeches, this study than compared interpreted speeches with source speeches. Results show that while trainee interpreters used the ten types of collocations in different manners, interpreted speeches are generally less conventional than first-language English speeches, characterised by a less frequent, less diverse, and less complex use of collocations. Trainee interpreters also produced collocations that do not habitually occur in English. Drawing on models of late bilinguals’ mental lexicon and language control in interpreting, this study suggests that these deviations from English conventions partially stem from collocations stored in interpreters' mental lexicon and the interference from Chinese expressions. Additionally, interpreters may use collocations in a way that helps them avoid overloading their working memory. This study underscores the importance of collocations in interpreting research and interpreter training. It contributes to existing theories of consecutive interpreting by emphasising interpreting as a context in which bilingual processing occurs. Furthermore, the corpus-based method used in this study offers a promising approach to assessing naturalness of interpreted speeches.