Abstract This paper argues that ontological category systems (OCS) should be assessed against explicit target criteria. To compare the highly diverse OCS, we focus on the epistemic issue of the category problem, not on the ontological one, and examine the system architecture at the bicategorical level. From existing systems we extract mutual exclusiveness , exhaustiveness , and generalness , and add a neglected criterion: recursiveness – the requirement that, at the most general level, the categories also categorize themselves, thereby conceptually closing the system. On this basis we motivate the most defensible number of categories and re-evaluate the traditional divide between universals and particulars in a slightly revised form. The resulting framework clarifies how OCS can remain maximally general while retaining classificatory power and shows why systems that exceed two top-level categories trade generalness for individuation without clear gain.
Norman Meisinger (Mon,) studied this question.