Résumé en anglais

This paper has a methodological purpose, as we are aiming to show practices of accounting research designing. In that heuristic, we are basing our argument on Burrell’s and Morgan’s (1979), Feyerabend’s (1975), Quattrone’s (2000, 2004b) and Lowe’s (2004a, b) epistemo-methodological writings and consider accounting research a comprehensive coherent whole in which methodology choices must be consistent with ontological assumptions revealed in research questions and influencing epistemological stances. We evidence our claim through the bottom-up in-depth study of a research stream characterised by a form of homogeneity and revealing various designs though. We found it in works on linkages between accounting and religion, all publications on the subject focusing on the Church of England or the Victorian Synod Church of Australia, and arriving at opposed conclusions. Indeed, two bodies of literature emerge, one concluding on semantic dichotomies between accounting and religion, and another viewing accounting as a religious practice. Thence, we argue the difference lies in the intertwinement of research question formulation with ontological assumptions, epistemological stances and methodology choices.