Abstract
The current study investigates why abstracts are written differently in theses and research articles. To this end, 200 abstracts of PhD theses in prestigious American and British universities were compared with 200 research article (RA) abstracts in highly ranked journals in the field of Applied Linguistics. The analysis relied on adaptations of Hyland's (2000) model of the rhetorical structure of abstracts and Hyland's (2005) taxonomy of metadiscourse. The results reveal a number of differences, including a lengthier description of methodology, findings and implications in RA abstracts and lengthier introductions in thesis abstracts. Additionally, thesis abstracts include more transitions, evidentials and hedges while RA abstracts make more use of code-glosses and self-mentions. Other variations are noted with respect to sub-categories of moves and metadiscourse markers. The results are interpreted in light of the differences between RAs and theses as two genres in the scientific community. The interpretations support Koutsantoni's (2006) claim for varied power asymmetries between disciplinary gatekeepers and thesis writers/RA writers, and Kawase's (2015) classification of theses as an educational genre and RAs as a professional genre. The study argues for the need of examining variations between the same sections of theses/RAs in terms of their context. (c) 2018 Elsevier Ltd. All rights reserved.