Corpus-based approaches to issues in applied linguistics

The present paper illustrates the use of corpus-based analytical techniques to address a range of issues in applied linguistics. Two main strengths of the corpus-based approach are identified: text corpora provide large databases of naturally-occurring discourse, enabling empirical analyses of the actual patterns of use in a language; and, when coupled with (semi-)automatic computational tools, the corpus-based approach enables analyses of a scope not otherwise feasible.These strengths are illustrated with respect to three areas of applied research: (1) English grammar; (2) lexicography; and (3) ESP and register variation. Throughout, the paper argues for the importance of a variationist perspective, comparing the patterns of structure and use across registers, and it shows how analysis of large corpora provides the empirical foundation for such a perspective.Two general points are discussed in relation to the illustrative analyses. First, corpus-based analyses frequently show that earlier conclusions based on intuitions are inadequate or incorrect-that is, the actual patterns of use in large text corpora often run counter to our expectations based on intuition. Second, corpus-based analyses show that even the notion of core grammar needs qualification, because investigation of the patterns of structure and use in large corpora reveals important, systematic differences across registers at all linguistic levels.