Defining 'religion' is not easy. John Hick sees a major division running through the welter of competing definitions:
The major division, as we have already noted, is between religious and naturalistic definitions. According to the former, religion (or a particular religious tradition) centres upon an awareness of and response to a reality that transcends ourselves and our world, whether the 'direction' of transcendence be beyond or within or both. Such definitions presuppose the reality of the intentional object of religious thought and experience; and they are broader or narrower according as this object is characterised more generally, for example as a cosmic power, or more specifically, for example as a personal God. Naturalistic definitions on the other hand describe religion as a purely human activity of state of mind. Such definitions have been phenomenological, psychological and sociological. (An Interpretation of Religion, Yale 1989, p. 3, footnotes omitted, bolding added.)
There is certainly a difference between a religious approach to religion and a naturalistic approach. And Hick is right that it is a major difference. But I suggest that the bolded passage needs correction. It is not that religious definitions of religion presuppose the reality of the intentional object of religious thought and experience, it is rather that they do not foreclose on the possibility of the reality of the intentional object.
When I study religion 'religiously,' what I do is take seriously religion's claim to be about something transcendent of our ordinary experience. Thus when I study Christianity religiously, I take seriously its claim to be a divine revelation both in and through its Scripture and in and through the person of Jesus of Nazareth. I hold myself open to the possibility of divine revelation. But this is not to say that I presuppose the reality of the triune God. I simply do not rule out the possibility of the existence and self-revelation of this God in the manner of the naturalist who, from the outset, assumes that religion is and can be nothing but a natural phenomenon, and therefore cannot be revelatory of anything beyond nature.
To put it another way, when I study religion 'religiously' as opposed to 'naturalistically,' I take seriously its claim to be in some measure true. I don't view it merely as just another natural or cultural phenomenon, which of course it also is. To study religion religiously is to study it in a manner analogous to the way we study science. I can of course concern myself with the sociology of scientific knowledge, with the psychology of scientists, and even with the phenomenology of scientific experience. But few think of natural science as merely a cultural artifact or a psychological product. For most of us, much of (current) science is either true, or a very good approximation to truth. We take physics, for example, to be revelatory of something beyond man and his physics, namely, the natural world which is what it is and has the properties it has whether or not we exist and whether or not we theorize about it. Few would get it into their heads to try to debunk science by explaining it genetically in terms of human fears and interests, the need for some to develop an arcane language-game whereby they can exclude and dominate others, etc. Few would claim that it is nothing but a human product. It is a human product, of course, but it is more: it is knowledge of the natural world.
To study religion religiously is then like studying science scientifically: it is to take the respective truth-claims seriously. But there is an important difference. In the case of science it does seem to be the case that when we study it scientifically (as opposed to viewing it as just another symbolic form or cultural product) we do presuppose the reality of nature. For one thing, nature is massively given to the senses -- hyperskeptical worries aside. How could we not presuppose it? But "a reality that transcends ourselves and our world," in Hick's phrase, is not given or at least it is not unambiguously given. (Mystical givenness is an ambiguous givenness.) This is why I say that when we study religion religiously we do not presuppose the reality of the intentional object of religious thought and experience; rather, we do not foreclose on the possibility of the reality of the intentional object.