Categories

4. The demarcation problem: what are ontological categories?

As the interest of philosophers is primarily in a specific kind of categories (the ontological categories) it is first of all important to find a way of telling the ontological from the nonontological ones. Ontology is not simply interested in what there is (it is not a kind of universal science), but in the most general features of what there is. As such one might be tempted to assume that the categories philosophers are interested in are also the most general categories. Systems of categories are often presented in the form of an inverted tree (or a set of trees), with the more general categories at the top. For example, the system of categories described in (Lowe 2001: 181) has the categories of ‘entities’ at the top, which is then divided into ‘universals’ and ‘particulars’, the latter being again split up into ‘abstracta’ and ‘concreta’. We could then simply identify the philosophically weighty categories with the top parts of the tree.

Yet this approach faces the problem of where to make the cut. If a system has a top-level category (say, ‘entity’), we presumably do not want to include just it amongst our categories. But then we need some kind of criterion for telling us how far down the tree we are supposed to go. There are two general strategies for addressing this problem.

The first is to find some additional criterion, such that the first category moving downwards from the tree that no longer fulfils the criterion is the first (and most general) nonontological category. One way in which such a criterion might be formulated is in terms of conceptual containment (for an account along these lines see Katz 1966). One might hold that complex concepts, such as ‘Peter’, can be analysed into simpler ones, such as ‘male’, ‘more than six-foot tall’, ‘having brown hair’, and so on. ‘Male’ can then be again analysed into ‘human’, this into ‘living being’, this into ‘concrete object’, and so on. Once we have reached the end of this chain of analysis we have reached the simple concepts, concepts which can then be deemed to select the set of ontological categories. However, this account does not really solve the problem of systems of categories with one top-level category. Every chain of analysis will terminate in this concept, leaving the set of ontological categories looking rather bare. If every concept we analyse ends up containing the same simple concept that cannot be analysed any further (maybe the concept ‘entity’, or ‘thing’, or ‘object’) and if this is all there is to ontological categories, the study of such categories does not seem of great interest. Rather than taking the detour through conceptual analysis to determine what the ontological categories are we could have achieved the same result by simply identifying the ontological categories with the top-level category, without bothering with the analysis of conceptual containment.

The second strategy is to reject that there is any clear cutoff point at all. This need not turn into a kind of defeatism according to which we do not really know what we are talking about when we refer to ontological categories. It could also mean that we label those categories as ontological that fulfil a specific role in our ontological theory, and that depending on the ontological theory we consider, different categories may play that role. For example, we might want our ontological categories to form a constructive basis of our ontological theory as a whole. Some entities in our ontology might be understood as complexes of or constructs from other entities, and the set of ontological categories may then be identified with those entities that can construct all the other entities. As such ontological categories can be understood as similar to axioms in a formalized theory. Axioms are those sentences in a theory from which all other sentences in the theory can be derived. In the same way, ontological categories would be conceived as that portion of the system of categories that can act as a constructive basis for the entire system. Of course, as there may be more than one axiomatization of a theory there may be more than one set of ontological categories, and as axioms are defined by their role in derivations, not by self-evidence, ontological categories, according to this understanding, would be defined by their relation to constructive relations between categories, not because of an essential nature any of these are supposed to have. The resulting position would therefore be a kind of relativism with respect to ontological categories (Westerhoff 2005).