I discuss the problems of giving definitions of linguistic terms especially in contentious fields such as religious studies. Alternatives are considered in the form of ostension and the use of exemplars. Having considered problems with some definitions of divination, I provide a new one that couples a rubric with a list of exemplary types and exemplary questions. The exemplars demarcate the domain to which the rubric should be applied. This definition does not rely on beliefs, deities (the supernatural) or modes of cognition.
