The metaphysical status of categories

We are reminded by recent postings that categories were significant in Grice's writings about metaphysics.

Here, on the road to, "The City of Eternal Truth" some importance should be attached to exactly what kinds of metaphysics are at stake, for some, perhaps even most, risk being local and perhaps even arbitrary.
The word absolute is good for the inner core of true metaphysics.

If we look for the absolute in metaphysics, and in Grice's discussions of metaphysics in particular, it seems to me moot whether Aristotle's categories are good contenders, and it is my purpose in this post to share my reservations on that count.

The study of categories in Aristotle's metaphysics undoubtedly is a kind of metaphysics.  I am inclined to use the description "exegetical metaphysics".
And if Strawson looks for categories in ordinary language then that sounds like his "descriptive metaphysics".  Both of these are at worst just some kind of relative metaphysics, rather than belonging to that absolute core.  In the first case relative to particular philosophers or philosophical schools, in the second relative to some particular language.  In both cases, to move forward to the absolute one might look for arguments that the resulting metaphysics is not merely accidental.

In Aristotle the categories play an important role in distinguishing the essential from the accidental.  The do so in a quite definite way, true intracategorial predications are necessary, true intercategorial predications are accidental.
This is an ingenious way of obtaining a more subtle analysis of necessity, than we find in Plato, but it doesn't look convincing in the light of more recent semantic thinking.  From that point of view the same predication could be either necessary or accidental, you cannot establish which just by reference to categories.
You need the meaning.

If they did succeed then there would be grounds for considering the categories to be metaphysical we need some reason to believe that they are not arbitrary but fundamental, and we need that case to justify the particular collection of categories nominated.

One way of looking for that might have been through Ryle's notion of "category mistake".  The idea here is that predications can fail to be true in a more radical way than simply being false.  If they constitute a category error then it is tempting to say that they are neither true nor false (I don't know whether Ryle did say that).  From this perspective this does not look good for Grice, if he wants to avoid truth value gaps.

But if category errors are falsehoods, (and if we accept that the categories cannot succeed in capturing the essential/categorial distinction as argued above) then its not clear why we need the categories at all, let alone why they would be absolute.

To a programmer the system of categories looks like a prototype of a programming language type system.  In that case we do have the situation that type-incorrect programs or expression do lack a value, which is a truth value gap for a boolean expression.  But we also have the case that type-systems are dispensable (in favour of type-free programming languages), and by this analogy we might consider categories disposable in favour of a category-free ontology,

The tendency of these considerations, if we were to take them seriously, would be to suggest that the absolute is as hard to pin down as we might have expected it to be,

Would Grice have cared about this distinction?


RBJ