Differences between the two approaches examined so far – the essentialist and the functionalist – make evident that, on the one hand, there are those who think that we can study religion by identifying private traits not readily available to our normal senses and, on the other, there are those who think that, like all other aspects of human behavior, the things we call religions have public features that perform observable functions. Given that we're trying to determine how best to study religion as an element of human culture, the distinction between the private traits and public actions deserves our attention.

When it comes to defining religion, there are thus two common approaches: one inductively browses through the group of things already known to be religions, looking for an essentially shared feature; and the other sets about looking for the common function performed in each religious practice or institution. If one takes the former route, then objects are defined by some key feature that is internal to them, more often than not some non-empirical feature judged to be sui generis (that is, self-caused, one of a kind, unique).