Keeping in mind the relationship--suggested from the outset of this site by Mary Douglas's comments on "soil" and "dirt"--between classifier, the system of classification, and that which is being classified, we can see why a number of contemporary scholars have found the essentialist approach to be unproductive inasmuch as it presumes a common identity, or essence, to underlie a thing's many varied manifestations--the presumption that motivated an earlier scholarly movement known as the Phenomenology of Religion (e.g., see the Dutch scholar, Gerardus van der Leeuw's classic 1933 work, Religion in Essence and Manifestation). Classification is now seen by some to be an inherently and inescapably political activity. So, just as studies of the politics of scholarship have recently appeared throughout the human sciences, so too the study of religion is being re-conceived as a site constituted by choice and practical interests rather than one based on sympathetic spiritual insight (see, for instance, the work of William Arnal, Talal Asad, Bruce Lincoln, and Tomoko Masuzawa).

But as we have seen, functionalist and family resemblance approaches to defining religion can also be judged to be insufficient. Does that mean that, as so many scholars before us have concluded, religion cannot be defined--that, as someone like Rudolf Otto might have concluded, due to its complexity and subjectivity, it can only experienced and insufficiently expressed? If so, then how can one study it rather than just feel it? Is the academic study of religion even possible? Is the answer to the question posed by this site--"What is the Study of Religion?"--that the study of religion is an impossibility?

If identifying the shortcomings of various approaches to definition prompt readers to throw their hands up in frustration--as if they were awaiting the correct definition to be revealed to them--then they may have missed the point of Mary Douglas's work on classification. Take the cognitive scientist and linguist, George Lakoff, who is also well known for his work on classification. In a book entitled Women, Fire, and Dangerous Things: What Categories Reveal About the Mind, Lakoff cites the philosopher of science, Stephen Jay Gould's discussion of the difficulties in settling on how best to classify that animal commonly known as a "zebra." Biologists, we learn, generally classifying living things either cladistically (that is, cladists classify biological organisms in terms of their shared, and thus evolutionarily inherited traits) or phenetically (that is, pheneticists classify based on biological organisms in terms of their shared form, function, and role). Depending on which of these two different systems of classification the biologist uses, the three species of zebra end up being classified in rather different ways, with one of the three having more in common--as judged by the cladistic system--with horses than with the other two species of zebra.*

The moral of this story of classification? If we presume, as Lakoff suggests, that there is one and only one proper way to classify the items of the world (an assumption he thinks we commonly make, but which might be rather mistaken), then clearly, when it comes to classifying those black and white striped creatures found in Africa, and in zoos throughout the world, we've got a big problem on our hands. For we now need to come up with a way to judge which classification system is right. But to do that, we need to employ yet another classification system, with internal criteria of its own, to judge "correct" from "incorrect." If we understand language itself to be a classification system (i.e., What gets to count as a letter, a word, a sentence, a text? What gets to count as a correct meaning?), then we find ourselves stuck in the midst of classifications systems not of our making, without which we might not be able to form a thought, much less get on with living our lives, get from point A to B, let alone make sense of the worlds in which we find ourselves. Perhaps, Lakoff says, this folk view of classification--an approach that assumes that there is some right system, if we could only find it--is itself our problem. What if, as Mary Douglas's work suggests, all we have are classifiers inventing and using systems that assist them to achieve practical ends of importance to them--like shelving books in a manner that makes them easily retrievable or grouping people in a manner that enables the distribution (or withholding) of goods? What if classification systems are seen as human products, tools, that we make, use, fine tune, and, sometimes, discard. If that was the case, then would the difficulties of each approach to defining religion prompt us to give up the search for a definition altogether? Or, instead, would these difficulties each be seen merely as the flaws that all products inevitably have?