existences.^2 Sometimes this principle is glossed as the view that any pattern of
fundamental properties and relations is possible.^3 However, put this boldly, the
principle is far too strong. Suppose thatx is located at Ris a fundamental relation.^4
Being a numberand being a region of space–timemight well be fundamental
properties. But necessarily no number is located at a region of space–time, and no
region of space–time is located at a number!
It seems that any plausible principle of recombination must be a restricted
principle. The examples just mentioned suggest that when stating a restricted
principle, the notion of an ontological category will be indispensable. In general, it
is appropriate to ask for explanations of why recombination fails in a given case. In
the cases just mentioned, the restriction on recombination shouldn’t be brute but
should ratherflow from the nature of the categories. I will argue that, given the theory
of ontological categories defended, even a powerful principle of recombination (such
as the one alluded to earlier) needn’t have the consequences it might be taken to have.
If I am right, the account of ontological categories defended here is capable of doing
substantive metaphysical work.
The account defended here, which identifies ontological categories with ways of
being, is how plenty of historically important ontologists conceived of the nature of
categories of being, although the account does not match how most contemporary
analytic metaphysicians inchoately conceive them. However, in this case, analytic
metaphysiciansoughtto bend to the weight of tradition: insofar as it is important to
one’s metaphysics to employ the notion of an ontological category, one ought to
conceive of them as ways in which things exist.
In what follows, I offer an account of ontological categories that is largely neutral
on the correct list of ontological categories, that respects and accounts for our pre-
theoretic grasp of categorial differences, and that provides a theoretical role for
ontological categories to play.
4.3 Ontological Categories: Competing Accounts
There are many questions that any account of“ontological category”should be able
to sensibly interpret. Does every object belong to some ontological category? Can an
object belong to more than one ontological category? Can one ontological category
wholly subsume another? Can two ontological categories overlap without one sub-
suming the other? Is there a highest ontological category to which every object
(^2) One of the most prominent champions of (mostly) free recombination is David Lewis (1986: 87–92).
See also Nolan (1996). 3
See, for example, McDaniel (2007) and Saucedo (2011). Note, however, that Saucedo does not put
much weight on this particular formulation of the principle, but rather defends a much more precise
version. 4
That some locative relation is fundamental is defended by Hudson (2005), McDaniel (2007), and
Parsons (2007).