With the return of religion to international and comparative politics, attention has turned to the theoretical and conceptual issues this raises, not least the problematic nature of the category of religion itself. While some have sought to expand how religion is conceptualised to overcome the distortions of Western-centric definitions, others have rejected efforts to identify religion, arguing that we should instead explore the politics of defining religion. In this paper I argue that neither approach is satisfactory: those who reject the concept downplay the historical emergence of distinct clusters of belief and practice that can be named as religion, while those expanding the category mistakenly assume that the inherited, common sense notion of religion as a genus remains valid. I propose instead that scholars of religion in IR and comparative politics should be wary of generalisations about the world religions, and focus instead on mid-range concepts with demonstrable cross-cultural genealogies.
