Polis religion has become the dominant model for the description of ritual activity in ancient Greek communities. Indeed, scholars have invoked polis religion to try to resolve the much-debated question of the definition of magic vs. religion, arguing that particular ‘magical’ practices, and their practitioners, do not belong to ‘collective polis religion.’ However, the relationship to polis religion of a ‘magical’ practice such as the writing of binding spells is surely more ambiguous, as well as of other cult activity relating (...) (in various ways) to the worship of Dionysos. Further examination suggests that defining what it means for ritual activity to be integrated within the schema of polis religion becomes increasingly difficult as we examine the variety of cult organisations and the different levels and types of involvement by the polis. This paper argues that social network theory may be able to overcome these conceptual difficulties. This approach can offer an alternative, more fluid construction of ancient Greek religion, which allows us to take account of coexisting, sometimes overlapping, networks of ritual activities. (shrink)