Since the early days of cultural anthropology, religion has been a topic of interest. Sir Edward Burnett Tylor (1832–1917) is considered as the founding father of the anthropological study of religion. He saw religion as a way to understand the unexplainable (Kottak 1996:260). Nowadays the phenomena religion is seen as a cultural universal (van Beek 1982: 3)(Kottak 1996: 260)(Morris: 1). But the concept of religion arguably not. In the nineteenth century French encyclopedists introduced the concept, that etymological can be traced back to the Latin `religare´, meaning to tie back (Encarta World English Dictionary 1999: 1587). The anthropologist, in contrast with for instance theologians, do not ask whether there is divine truth in religion, but look at the content of the religion (van Baal & van Beek 1985: 1). Max Weber refused to define religion (Morris: 69) and that might indeed be the wisest thing to do. But I will take the risk of burning my fingers by looking at some anthropological definitions.
In classical anthropology the definitions tend to focus on religion in `traditional´ societies. In this they put emphasis on the interaction with supernatural entities (Spiro 96)(van Beek 1985)(Kottak 1996). Although subject of debate, religion in my view is very well possible without any supernatural beings, as Durkheim has debated in the case of Buddhism. The most important point of Durkheim is that religion can be seen as something sacred or as he puts it:
a unified set of beliefs and practices relative to sacred things, that is to say, things set apart and forbidden, — beliefs and practices which unite [into] one single moral community, all those who adhere to them (Durkheim  1964: 37 cited in Morris 1987).
What also is worth noting is that he takes religion as being both believe and practice. Furthermore Durkheim sees religion specific as something collective, while magic would be typified by individual practice. With his emphasis on a community as the basis of religion, the spiritual practice of what often is referred to as shamanism would for instance not be considered a religion .
Another definition is by Clifford Geertz. He defines religion as
(1) a system of symbols which acts to (2) establish powerful, pervasive, and long-lasting moods and motivations in men by (3) formulating conceptions of a general order of existence and (4) clothing these conceptions with such an aura of factuality that (5) the moods and motivations seem uniquely realistic (Geertz 1985: 4).
Implicitly Geertz suggest that the religious experience is not true (an aura of factuality, seem…realistic), which seems a biased view. Durkheim takes a more neutral approach: All [religions] are true in their own fashion. It is the task of the anthropologist to understand these fashions. Van Baal & van Beek add as a critical point to Geertz´ definition that it lacks specificity. It wouldn’t contain a `directly observable, formal characteristic which is universally applicable as a means of identifying the religious´ (van Baal & van Beek 1985: 3) In other words, it doesn’t leave the anthropologist much to study. Hereafter they define religion themselves as:
all explicit and implicit notions and ideas, accepted as true, which relate to a reality which cannot be verified empirically (van Baal & van Beek 1985: 3)
Here, Geertz’ rich symbols, moods, motivations and conceptions have been reduced to only notions and ideas. However, there is a similarity between the two definitions. In contrast with Durkheim both Geertz and van Baal & van Beek take non-empirically perceivable objects (notions and ideas; a system) as the core of religion instead of also having an eye for the practice. However, not only that what is believed but also that what is done is an essential part of religious tradition. Some people might refer to their religion as mainly being the actions they are taking while others might emphasis the beliefs. Hence, a truly anthropological definition would embrace both possibilities.
Radcliffe-Brown puts more emphasis on the social aspect of religion, according to Morris:
“We should see religious beliefs and observances as a part of a complex system by which human beings live together in an orderly fashion. We should look, he maintains, at the social functions of religion, that is, the contribution that it makes to the formation and maintenance of a social order” (Morris 1987: 127)
In other words religion itself cannot be studied, at least not by a social scientist. Only expressions of religion can be observed.
So, summarizing, a definition in the line of the given arguments should have eye for the cognitive as well as the performative nature of religion. It should somehow incorporate the dimensions in which it is expressed and take awareness of the fact that it has something to do with the spiritual. With would lead to:
A set of beliefs and/or actions to regulate and approach reality, expressed in: (a) doctrine, (b) philosophy, © myth, (d) symbol (e) ethic, (f) ritual, (g) matter, (h) experience and (i) social organisation, in some way related to spiritual qualities, phenomena or entities.
Guido Verboom, 2002
(revised 2015: language and style)