People working in various scholarly disciplines have been divided into two camps: the lumpers and the splitters. Lumpers create relatively broad categories and splitters create more narrow categories. Both create categories, classifications, or taxonomies, however, because that is what scholars or scientists do.
For instance, scholars of religion might speak of Christianity as a single religion, which is a case of lumping, or they can speak of Catholicism and Protestantism, which is a case of splitting. But categories of Protestantism and Catholicism are also broad categories. There are many Protestant “sects” that might be called religions in their own right. And how many different groups or kinds of Baptist “sects” are there? There are also many kinds of Catholicism. We can also speak of the broad category called religion and then classify the major world religions–Christianity, Judaism, Islam, Buddhism, Hinduism. But what makes a religion “major”? And more importantly, what do these major religions have in common? Is there any basis for putting Buddhism and Islam, for example, in a single category? Hinduism as a religion was invented in the 19th century to unify the many different religious practices and beliefs native to India—in opposition to Western influences.
Even within a tight category of religion (e.g., Southern Baptists) each group of “worshippers” is doing something different. Two Southern Baptist churches on the same street in the same town might approach their religion quite differently. Studying the uniqueness of a single church might be more interesting than finding out what that church has in common with other churches.
When we speak of Christianity as a religion, we follow a genetic approach in saying that all the different forms of Christianity trace themselves back to the life and teachings of Jesus Christ or the New Testament. Likewise, the various forms of Buddhism are traceable back the life and teachings of the Buddha, and Islam is traceable to Muhammad and the Koran. Biologists also use a genetic approach when classifying genera and species; they look for common ancestors.
We can look at this issue in terms of the realism versus nominalism debate. Realism maintains, for instance, that the biological classification system–life, domain, kingdom, phylum, class, order, family, genus, species–represents the true or approximately true organization of life; there is a natural order. A nominalist, in contrast, would say these are all just words or concepts, although the words are useful (or essential) for scientific work.
As discussed elsewhere on this blog, Deleuze avoided the conventional nominalism/realism split by arguing that concepts (e.g., homo sapiens) mark differences in kind rather than merely a difference in degree. A concept, moreover, is an ontological category rather an epistemological category. That is to say, for Deleuze, a concept is not a mere intellectual representation but a real thing (an intensive multiplicity); it is something that marks (or makes) a difference–specifically, a difference in kind, not just a difference in degree. A concept is real because it does something.