committed to the existence of states of affairs and sets of them and their constitu-
ents.) This neutrality also extends to the claim that sets, wholes, structures, or other
compounds belong to different ontological categories than their constituents.^41
Recall that Westerhoff’s account of ontological categories implies that, for example,
sets and other constructed entities never form ontological categories.
A second noteworthy feature of the view is that it does not imply Sommer’s Law,
but it is consistent with it. Recall that Sommer’s Law is that no two ontological
categories merely overlap. The interpretation of Sommer’s Law in the framework
defended here is that the domains of perfectly natural quantifiers never merely
overlap. There is nothing in the framework that implies this, but one is free to
adopt it if one wishes. (Earlier, we saw that some plausible ontological schemes
have ontological categories that merely overlap.) Relatedly, the view takes no stand
on whether there is a unique topmost ontological category: there is if the unrestricted
quantifier is perfectly natural. But on the specificfirst-order ontology I am attracted
to, the unrestricted quantifier is not perfectly natural. (More on this shortly.) Another
way for there to be one topmost ontological category is for there to be a fundamental
mode of being that includes in its domain everything that enjoys some fundamental
mode of being or other.
Third, the account is consistent with the claim that some objects belong to no
ontological category. An object belongs to no ontological category just in case there is
no natural quantifier that ranges over it. Holes, cracks, and shadows are such objects.
Although they exist (they are within the range of the ordinary English unrestricted
quantifier), they do not“fully”exist (since the unrestricted quantifier is not a natural
quantifier, and no natural quantifier ranges over them). We’ll have more to say about
this in chapter 5.
Fourth, the account does not imply that, if something belongs to an ontological
category, it necessarily belongs to that ontological category, although one could
endorse this additional thesis. Although most plausible ontological schemes are
such that every entity belongs to its ontological category essentially—could a sub-
stance have been a property or a proposition?—not every ontological scheme has this
feature. Recall the scheme defended by Bricker (2001) according to which the
possible and the actual form different ontological categories. But nothing merely
possible is essentially merely possible, and much if not all of what is actual could have
failed to have been actual. Similar remarks apply to Meinongian ontological schemes
according to whichexistentis an ontological category: some of what exists could have
been objects while failing to exist, and some of what merely is an object could
have existed.
Note that other accounts of ontological categories struggle to make sense of
Bricker’s (2001) ontological scheme. For example, identity accounts, which imply
(^41) As noted in section 3.7, I doubt that sets and states of affairs/facts belong to the same ontological
category as their elements or constituents, but perhaps mereological sums do.