Textual metadiscourse in research articles: a marker of national culture or of academic discipline?

Academic writers leave traces of themselves in their writing which may be linked to national as well as disciplinary culture. This paper takes a doubly contrastive approach and investigates writer manifestation in three languages, English, French and Norwegian, and three disciplines, economics, linguistics and medicine, in order to see whether language or discipline is the most important variable governing the pattern of metatext in academic discourse. My corpus consists of 180 refereed research articles within these languages and disciplines. My findings suggest that the language variable is the most important one within economics and linguistics, where English and Norwegian show very similar patterns, using much more metatext than French; within medicine, all three languages display a uniform pattern of little metatext. I conclude that English and Norwegian are both representatives of writer responsible cultures, while French represents a reader responsible culture. As regards discipline, I suggest that since economics and linguistics have a less formalised research article text structure and to some extent create their findings through argumentation in the text, national culture will be more important than it is in medicine, where the IMRD (Introduction-MethodResults-Discussion) structure is globally implemented and the research data to a greater extent are -iven outside the text. (C) 2004 Elsevier B.V.d All rights reserved.