Metadiscourse in Academic English Texts: A Corpus-driven probe into British Academic Written English Corpus

Abstract

This paper reports on a research performed in the field of corpus linguistics on metadiscourse features in the British Academic Written English Corpus. For this purpose, the British Academic Written English Corpus, which is freely available and contains 6,968,089 words, was selected as the data resource of the study. The taxonomy of metadiscourse features compiled by Hyland was used as the theoretical framework and the R program was used as the statistical software. The whole corpus was analyzed. As the data can show, the interactive metadiscourse features were more prevalent than the interactional metadiscourse features. In the interactive category, transitions and endophoric markers were used more than other ones; whereas, in the interactional category, hedges and boosters were the predominant metadiscourse features. The prevalence nature of interactive metadiscourse features can add support to the idea that writers were more interested in organizing discourse rather than conducting interaction. The findings of this research can have useful implications for researchers in such fields as contrastive analysis, text linguistics and corpus-based studies.