The Corpus of Academic Learner English (CALE)

Abstract

Learner corpora present an option to inform, supplement and advance the way language proficiency is operationalized and assessed, and may also be used in data-driven approaches to the assessment of writing proficiency that are largely independent of human rating. The aim of this contribution is twofold: first, to introduce a new Language-for-Specific-Purposes learner corpus, the Corpus of Academic Learner English (CALE), currently being compiled for the study of academic learner writing; and second, to illustrate how the CALE is useful in a text-centered, corpus-driven approach to the assessment of academic writing to achieve a higher degree of reliability in assessing language proficiency.