Integrating `expert' and `interlanguage' computer corpora findings on causality: discoveries for teachers and students

Abstract Corpus linguistic techniques involve the use of concordancing software which identifies a KWIC (key-word-in-context) and thus provides empirical evidence on the lexico-grammatical patterning of certain items. Such corpus-based findings have either been used to inform materials design or used as input for CALL activities. However, to date, most of the work in this field has relied upon exploitation of native-speaker corpora. This paper proposes that a comparison of the findings from both learner and native-speaker corpora would also have important pedagogical applications. However, this in turn, raises the question as to how corpus-based findings can best be exploited in the ESP classroom.