Move analysis is a text analytical approach first developed by John Swales (1981) to investigate the underlying generic structure of research articles (RAs) in terms of moves-and-steps for pedagogical purposes. A widely shared aspiration of move analysts has been to identify the linguistic features characterizing the various RA moves not only in English, but also across languages. One shortcoming blocking this advancement is the lack of multilingual corpora fully annotated for their specific communicative functions in a coordinated and reliable manner. In this paper, we describe and discuss a methodology for analysing the various RA sections for their generic structure up from the step level in two languages and across a wide range of disciplines, using the discussion section as a test case for illustrating that methodology. Among the topics treated are establishing criteria for choosing a suitable sample of comparable RA discussions across the two languages, designing a model for annotating the section’s moves and steps, creating an accessible computer-assisted coding scheme, achieving good levels of inter-rater reliability, and obtaining validation from expert informants and writers. In essence, this is a methodology paper offered as a working model for other EAP researchers undertaking similar analyses in future.
