1. One confuses between a definition and a hypothesis. Definitions tell us how we use a word and which objects it refers to. In the case of the term ‘religion’, a definition should merely address this referential task: which things are we referring to, when we use this word. It cannot give us any understanding of the structure or properties of religion and does not have empirical consequences (much as defining any other word, say ‘species’ or ‘psychosis’, does not have empirical consequences). It merely points out the phenomena that are to be theorized. The result of this second process of theorizing is a hypothesis, which characterizes the properties or structure of a phenomenon.
2. When one confuses between defining a word and theorizing a phenomenon, as you do, one enters a dead-end: it is as though the task of a definition is to classify objects and decide whether or not they belong to a particular category. But this generates disputes that cannot be settled. Imagine one scholar who defines ‘religion’ as a belief in supernatural and non-empirically verifiable beings. Another points out that ancient Greek religion does not involve any supernatural beings (since Greek theoi belong to the natural world) and that ‘religion’ requires another definition. The first author then says that they are non-empirically verifiable entities and, therefore, ancient Greek traditions do belong to the category of religion. But then the second points out that many particles predicted by theoretical physics are not empirically verifiable; hence, theoretical physics would also be a religion according to this definition. Our second author could then point out that we need a definition focusing on family resemblances or that ‘religion’ should be defined as ‘a set of methods for spiritual uplift’ for everything religious falls under this umbrella. But then another guy comes and points out that taking heroin is also a method for spiritual uplift. So is this also religion? Etcetera.
3. This type of exchange illustrates the problem: when one confuses the task of defining ‘religion’ (by pointing out its referent) with the task of classification, one ends up in disputes as interminable as disputes about taste. Scholars are busy inventing definitions that should accommodate their intuition that all cultures have religion or that some tradition is ‘not really a religion’. The resulting definitions are inadequate and increasingly vague. Consequently, we now face a cemetery of discarded definitions of ‘religion’.
4. When Balu suggests that religion is an explanatorily intelligible account of the cosmos and itself, he is not providing a definition but a hypothesis. Of course, as a first step, the term ‘religion’ needs to be defined so that we know which object(s) should be theorized. We could say that ‘religion’ has no referents in the world, but this is an epistemic decision that cannot be taken before developing a theory. Instead, Balu suggests the following: we provide an ostensive definition of ‘religion’ by pointing out an instance of how we use the word. Minimally, ‘religion’ refers to Christianity, since Christianity has used the word to refer to itself and to certain others. Only in that sense, Christianity is taken as a prototypical example of what we refer to when we use the word ‘religion’.
5. Balu then characterizes some of the constraints that any hypothesis about religion should live up to. First, Christianity has used the term ‘religion’ to refer to itself and to certain others as its religious rivals: Islam, Judaism, Roman religio, Hindu traditions … This has resulted in a particular set of descriptions invoking the term ‘religion’. Second, Islam and Judaism not only recognized themselves in such descriptions, but also provided similar descriptions of each other and of Christianity. Moreover, they described the same others (Roman religio in the case of Judaism, Hindu traditions in the case of Islam) as religious rivals. Thus, using whatever word they may have used, they share a basic notion of religion with Christianity. Third, the followers of Roman religio and the Hindu traditions recognized neither themselves nor Islam and Christianity in these descriptions. In fact, they showed a remarkable indifference and blindness towards the idea that all of these were religious rivals.
6. Any hypothesis about religion should be able to account for this dual set of historical facts. That is, it should explain as to why the Semitic religions recognized each other and the same others as religious rivals and why the ‘pagan’ traditions failed to recognize themselves and the relevant others in these descriptions. It is not a good idea to develop the required hypothesis about religion through an inductive approach: that is, we cannot study the enormous body of descriptions that Christianity, Islam, Judaism, and the ‘pagan’ traditions have given of each other and then try to infer a hypothesis that accounts for all these descriptions.
7. Instead, Balu takes a very different route: that of formulating a hypothesis and deriving its logical and empirical consequences so as to test whether the hypothesis is able to account for past phenomena and predict new facts. Now, since this hypothesis does not rely on induction (inferring some common structure of religion by culling over the available body of descriptions), it is not constructed by accepting any ‘Christian’ or ‘Abrahamic’ concept or definition of religion. Rather than being founded in any such concept or definition, it is formulated independently. Consequently, its relation to the available body of ‘Abrahamic’ and ‘pagan’ conceptions of religion is entirely different. These function as a test or constraint for the hypothesis: it should account for both sets of conceptions as manifested in their historical encounters. And it does.
Once you get these points, you will see that what you write stems from a basic ignorance of Balu’s work.
- On the Indian Notion of Enlightenment: Reflections Based on Experience
- Caste system I