It might help to read these two pages:
https://www.open.edu/openlearn/hist...studies/studying-religion/content-section-5.1
https://www.open.edu/openlearn/hist...studies/studying-religion/content-section-5.3
The author writes:
"Scholars offer us many different definitions of religion, but these definitions tend to be of two types. The first type is known as a substantive definition: that is, a definition that tells us what kind of thing religion is by pointing to its distinguishing characteristic...We can find an example of a substantive definition of religion in my summary of the definitions found in the Concise Oxford Dictionary. Think again about d. According to this definition, religion is the 'human recognition of superhuman controlling power and especially of a personal God'...
...The selection of a defining characteristic, upon which a substantive definition of religion depends, often reveals prejudice -- perhaps a personal religious (even a denominational) bias or a broad cultural bias. Trying to define religion in terms of one kind of belief -- for example, the belief in one god --- may be understandable within the context of Western Europe, which has been dominated historically by Christianity, but is narrow and inflexible when considering religion as a global phenomenon --- Buddhism is a case in point...
...In order to avoid being too narrow and too rigid, many scholars prefer a different type of definition known as a functional definition. A functional definition concentrates not on what religion is (its beliefs and practices, for example) but on what these beliefs and practices do for the individual and the social group..."
My own opinion is that functional definitions suffer from the same kind of weakness that afflict substantive definitions. They are both looking for a single particular defining essence that all religions (and only religions) share in common, whether it's an item of belief (in the "supernatural" or whatever) or a particular psychological role in people's lives.
I'm skeptical that any single defining characteristic for religion even exists. Religion takes countless forms, includes no end of different ideas and doctrines, and plays many roles in different people's personal and social lives. I'm skeptical whether there is any single essence in all of it.
So, what makes something a 'religion' when other aspects of human life aren't?
That's where I think that the concept of
resemblance comes into play. We call social beliefs and practices 'religious' when they sufficiently resemble (and 'sufficient' will be a matter of opinion) other things that we already think of as religions. That resemblance won't be total and perfect, since that would make any differences between religions impossible. So 'religion A' will share some descriptive and functional characteristics with 'religion B', but not every one.
And crucially, I'm not convinced that there is, or needs to be, any single defining characteristic (whether substantive or functional) that all examples of 'religion' share in common and that only religions possess. All that's needed is that the examples share enough characteristics with enough other things that we already identify as religions that they share the religious
'family resemblance' to a suitable degree (which might be a matter of judgment).
Actually, I think that many of our everyday concepts work in this 'family resemblance' way. Just try to define 'Art' for example. Perhaps the only place we encounter concepts with precise definitions is in the technical vocabulary of science and mathematics. (And I'm not 100% convinced about science.) But that's a matter for another thread. (Wittgenstein moves in this direction in his famous section 80 of this
Philosophical Investigations when he addresses the concept of a 'chair'.)
My own views on how best to define 'religion' might be somewhat similar to those of Ninian Smart in this little essay here:
https://www.open.edu/openlearn/ocw/mod/resource/view.php?id=17576