Corpus-based Analysis of Collocational Errors 1

Abstract

The present study investigates the use of high-frequency verb collocations in L2 learner English. Although it has been widely accepted that collocations are essential to language use, until recently, the complexity of collocations was mainly hidden. However, developments in the new technology of corpus linguistics and the availability of large electronic texts from spoken and written sources have made possible new perspectives to learner language use. In addition, learner corpus data constitutes a new type of performance data which offers a valuable view of learners’ errors and non-errors as well. This study is thus embedded in a combination of elaborate manual investigation and computer-aided analysis using a concordance program, WordSmith Tools. The results of this study show that errors are not entirely mismatch between the verb and the noun, and other types such as prepositional errors and determiner errors were relatively more frequent among advanced L2 learners. It is also found that the interference of the Korean learners’ L1 plays a role in the use of collocations. Based on the results, some suggestions are made for collocation learning that is helpful for EFL learners.