Spoken lexical chunks used by successful learners at B2 level: forms and functions

Abstract

This study analyses the lexical chunks used in successful spoken language by a sample of learners at Common European Framework of References for Languages (CEFR) B2 level in the speaking component of a four-skills test. The data forms the initial part of a corpus of speaking tests at this level and the intention is to provide practitioners and researchers with evidence in order to more clearly interpret the broad ‘can do’ statements of the CEFR. Results suggest that the most frequent chunks used at this level employ the first thousand most frequent words in the British National Corpus (BNC) spoken lists and are often comparable to frequent chunks used in native speaker corpus data. The findings also show that learners make use of a narrow range of multi-functional chunks rather than employing a range of items for one function. Moreover, some chunks we normally would expect to find at this level were underused. The data has implications for how spoken language, in this case, chunks, could be taught to students at this level and for further research in this area.