The effects of corpus use on L2 collocation learning

Abstract

In data-driven learning (DDL) settings, few studies have examined the specific effects of corpus use on language learning, especially in the study of collocations. This study examined the effects of corpus use on L2 collocation learning. It remains unclear how corpora can contribute to learning collocations and what types of information learners can access and process from these corpora. This study aimed to investigate the effects of corpus use on learning L2 collocations, specifically how information extracted from corpus data assisted learners in retaining and producing collocations. The effects of corpus use were examined by comparison with the effects of dictionary use. Fifty-five intermediate English as a foreign language (EFL) learners from a university in Tokyo looked up verbnoun collocations by consulting either the corpus or the dictionaries, noted related information, and wrote short essays on the topics related to the collocations. Pre- and post-tests were conducted to assess learning. It was found that corpus use promoted a greater collocation output than dictionary use because of the many example sentences and frequency of co-occurrence words. The corpus users focused on word combinations and looked up high-frequency collocations, while the dictionary users focused on meaning, then looked up and memorized the collocations. Findings suggest that teachers need to consider the information that learners access and process when using corpora and the specific effects of corpus use on L2 learning for more effective vocabulary teaching in DDL settings.