Disciplinary influence on lexical bundle functions: reassessing the paradigm factor

Many lexical bundle studies have explored how bundle functions are used to shape arguments and construct knowledge in research articles (ras) in hard and soft disciplines, but these cross-disciplinary studies tend to overlook the impact of research paradigms and ra sections. Specifically, previous investigations have not differentiated between qualitative, quantitative and mixed-methods research, which makes it impossible to evaluate whether any observed differences are related to the research paradigm, discipline or both. Furthermore, very few previous contrastive studies have explored the distribution in the in-text sections. To address these two issues, this study compares ras in medicine and psychology representing a single paradigm (i.e., quantitative experimental research) to determine how exactly discipline may influence the distribution of bundle functions in ras as a whole and by section. Unlike the previous studies which find a heavier reliance on research-orientated bundles in hard disciplines and a greater focus on text-orientated bundles in soft disciplines, this study finds no significant difference in research-orientated bundles and more frequent use of text-orientated bundles in medicine than in psychology. This study argues that a more critical eye towards discipline is needed in cross-disciplinary studies on academic writing. Furthermore, this study recommends that corpus linguists examine ras at the level of in-text sections to present a clearer picture of discourse organisation.