that objects belong to their ontologically categories essentially, cannot. And, if one is
suspicious of“absolute”de remodal properties, one might be suspicious of identity
accounts as well. For example, suppose one is attracted to a kind ofcounterpart
theoryaccording to which objects in the actual world havede remodal features in
virtue of being similar to objects in other possible worlds. One prominent champion
of counterpart theory, David Lewis (1986: 251), seems to believe that in some
contexts it is true to say you couldn’t have been an angel while in other contexts it
is true to say that you could have been an angel. An angel is a very different kind of
thing from a human being. Why then not allow for a context in which it is true to say
that you could have been, for example, a property or a set? (How much freedom does
counterpart theory allow?) However, someone who holds a counterpart theory of this
sort might still hesitate to say that what ontological category I actually belong to
differs from context to context. And so a counterpart theorist might reject an identity
account of ontological categories.
Fifth, the account defended here is consistent with the claim that individuals of
different ontological categories can be alike with respect to every perfectly natural
property. On Bricker’s (2001, 2006) ontological scheme, some merely possible objects
areduplicatesof actual objects even though they belong to different ontological
categories. (Bricker and I adhere to Lewis’s (1986: 61–2) view that things are
duplicates just in case there is a one–one correspondence between their parts
that preserves natural properties and relations.) This additional thesis is not ruled
out by the ontological scheme defended here. Were ontological categories to be
identified with natural properties, this additional thesis would be inconsistent. The
best that someone attracted to Bricker’s scheme could say is that it is possible
for objects from different ontological categories to be alike with respect to all
naturalqualitativeproperties.^42 But this puts quite a lot of weight on the difference
between the qualitative and the non-qualitative—perhaps more weight than this
distinction can bear.
Finally, because the account of ontological categories defended here does not
simply defineontological categoryin terms of failures of intersubstitutability in states
of affairs, one can appeal (without threat of circularity) to the notion of an onto-
logical category to explain why and how the principle of recombination must be
restricted. In the next section, we will look at how one might use the framework
defended to formulate and justify a restricted principle of recombination. Similarly,
because the account of an ontological category does not appeal to the notion of an
essence, it might be of use in explaining why things have the essences they have. The
connection between essences and modes of being will be explored extensively in
chapter 9.
(^42) In fact, this is what Bricker (2006) ends up saying.