Corpus Linguistics: a data mining methodology or a branch of Linguistics?

Abstract

The present paper problematizes empirical research based on corpus, in the light of the role played by Corpus Linguistics. The article opens with a discussion of the place of Corpus Linguistics itself, i.e., whether it can be considered part of Linguistics proper or whether it is no more than a methodology utilizing the computer in the investigation of linguistic phenomena. Without favoring either position, the discussion focuses on practical examples of corpus-based and corpus-driven approaches carried out in recent studies of both Portuguese and English as a Foreign Language. The work argues in this way that the status of a “branch of Linguistics” or linguistics “data mining methodology” may be established at the onset of the analysis of digital data. Whichever the role opted for, the results yielded make important contributions to the ways we view linguistic phenomena.