Learner corpora and second language acquisition

Abstract

Second language acquisition (SLA) research has traditionally relied on elicited experimental data, and it has disfavoured natural language use data. Learner corpus research has the potential to change this but, to date, the research has contributed little to the interpretation of L2 acquisition, and some of the corpora are flawed in design. We analyse the reasons why many SLA researchers are still reticent about using corpora, and how good corpus design and adequate tools to annotate and search corpora can help overcome some of the problems observed. We do so by describing how the ten standard principles used in corpus design (Sinclair 2005) were applied to the design of CEDEL2, a large learner corpus of L1 English – L2 Spanish (Lozano 2009a).