This paper was originally prepared for a meeting of the Methodology Seminar of the Distant Worlds Graduate School (LMU, Munich) and was meant as an introduction to the use of the category of magic in the study of ancient Greek and Roman religions offered to a group of scholars coming from different disciplines.

The study of ancient magic is complicated by the fact that most of ancient Greek and Latin terms usually translated by “magic” or “magical” were used in different and contradictory ways. Approaches trying to reconcile rather than expose these different meanings can be divided in two large groups: the so-called essentialist approach, exemplified here by the work of H.S. Versnel and the sociological approach, represented here by the work of P. Bourdieu. Against these two approaches, it has also been argued that the modern term “magic” should be abandoned. Against this last position, I will first repeat – as Versnel and others already did – that we cannot represent alien (i.e. foreign or ancient) categories of thought without using our own categories. Finally, I will present Versnel’s methodology, its problems, and the solution that Bourdieu’s notion of the religious field can provide. While not without problems, it gives an idea of what could be gained by tinkering with common-sense notions rather than assuming that their definitions are self-evident.