Abstract
The paper combines learner corpus research with contrastive analysis to test the applicability of corpus-driven methods to the study of phraseology in learner academic English. It explores phraseological patterns in English L2 academic texts written by Czech university students in comparison with English L1 novice and expert writing.
Three corpus-driven approaches are employed: frequency lists, keywords and lexical bundles. The results indicate that a combination of corpus-driven methods can indeed serve as an effective starting point for the contrastive study of phraseology, highlighting potential areas of under- and overuse of multi-word patterns in English L2 novice academic texts.
However, in order to give a more comprehensive picture of learner academic English, quantitative methods have to be combined with qualitative contrastive analysis.