Categories

7. Their uses

I want to conclude this entry with a brief look at the uses of ontological categories. Why, philosophers’ desire for neatness in their ontology aside, should we want to construct a system of ontological categories, and what is it good for?

Reference to ontological categories features prominently in philosophical problem-solving, some philosophers might even argue that all philosophical difficulties could be traced back to confusion about ontological categories. While this might be too strong a position, two examples where ontological categories are appealed to in order to solve philosophical problems immediately come to mind.

The first are various paradoxical constructions such as the Russell set (‘the set of all sets that do not contain themselves’; see Paradoxes of set and property) and the liar paradox (‘this sentence is false’; see Paradoxes, epistemic). One common diagnosis is that these constructions violate the hierarchy of types, which is a form of a system of ontological categories (see Theory of types). Type theories come with various restrictions built into them, one of these is the necessity that higher-type entities can only contain lower-type entities. The set of all sets is a higher-type entity that contains members of a lower type (namely the sets it contains) but, being a set, also contains itself, a higher-type entity. Once categorial (in this case, type theoretic) restrictions are taken into account, such constructions are no longer possible.

A second famous example is Carnap’s discussion of ‘illusory problems’ (Scheinprobleme) in metaphysics (see Carnap, R.). Carnap points out that once we are clear about the category the entities referred to in our metaphysical sentences belong to, certain problems will no longer arise. Metaphysical worries about the nature of nothingness, for example, will no longer arise once we realize that there is no individual that the term ‘nothing’ in such statements as ‘there is nothing to be seen here’ or ‘I have nothing to sell you’ picks out. Proper logical analysis of the respective sentences will reveal that the name ‘nothing’ vanishes and its semantic role is carried out by various pieces of logical machinery in the sentence. As nothingness does not belong to any category in our ontology there is also no necessity to provide an account of its nature. One of the main targets of Carnap’s critique here is Martin Heidegger and his metaphysical theorizing about the nature of nothingness (das Nichts). For Carnap, Heidegger is simply misled by language. Had he analysed the logical function of the term ‘nothing’ he would have realized that there is no more need to worry about the nature of nothingness than about the curious status of the fractional children the average man with 2.4 children is supposed to be having.

Finally, despite the fact that ontological categories are a fairly technical concept in theoretical philosophy, the notion has some extratheoretic uses. When designing knowledge bases as part of expert systems that allow fast and automated access to a specific body of knowledge (such as aiding doctors in diagnosis, or prospectors in searching for natural resources) it is essential that the system has some scheme for organizing the information about the various entities encoded in it, a system that must also be sufficiently similar to the way a human expert (who is the recipient of the system’s assistance) classifies it. Such schemes often benefit from being constructed by taking into account the systems of ontological categories formulated within philosophy for purely theoretical purposes (see Guarino 2005; Munn and Smith 2008).