Computer learner corpora, or how can we turn our students' interlanguage into a resource for EFL research and teaching?

Abstract

The essay presents recent developments in the field of learner corpora, with special reference to some written corpora that have been collected in European academic contexts. One particular corpus, "The International Corpus of Learner English" (ICLE), is dealt with in more detail since it allows cross-linguistic comparison among 11 different national subcorpora. The research that has already been carried out on ICLE ranges from ad hoe case studies of mistakes made by specific groups of learners to wider and more systematic investigations of areas of difficulty and of features of "foreign-soundings" across different mother tongue groups, and from the discussion of corpus design criteria to the empirical testing of Second Language Acquisition hypotheses and findings. Using an already existing learner corpus, joining a project in progress or building a new learner corpus appear to be useful and challenging enterprises which may help reconcile research and teaching needs especially in the field of English, which is the most highly required foreign language in European universities.