claim) is if they have possible duplicates that are grounded. But here is an argument for
the claim that all duplicates of ungrounded physical facts are ungrounded. That a fact is
ungrounded is not itself a brute fact, but rather is itself grounded in the nature of that
fact. But which aspect of the nature is relevant? Consider an ungrounded physical fact
Fa. That the fact hasaas a component rather than, for example,bis irrelevant—this fact
is ungrounded because the property of beingFis a fundamental property. It is the
qualitative aspect of the fact that is responsible for its being ungrounded. And so it is
necessary that all facts of the formFxare ungrounded if any of them are ungrounded.
Any duplicate of the fact thatFawill contain a constituent that is a duplicate ofFand a
constituent that is a duplicate ofa. But any possible duplicate ofFis itself numerically
identical withF.^43 So any duplicate ofFawill be a fact of the formFx, and hence will be
ungrounded as well. So if the second parallel claim is true, ungrounded physical facts are
notapt to be grounded.
Finally, let’s consider a third parallel claim: ungrounded facts that are apt to be
grounded are instances of the same kind of fact that has possible instances that are
grounded. Here we run in to a kind of generality problem—an ungrounded fact is an
instance of many kinds, but what is the relevant kind that makes it apt to be
grounded rather than inapt to be grounded? My intuitions on this score are very
weak, even with purportedly easier cases. Suppose that the fact that the singleton of
Socrates exists is grounded in the fact that Socrates exists. Is the fact that the empty
set exists grounded or ungrounded?
I guess it is ungrounded. Fine (2012a: 48) says that it is grounded, but grounded in
nothing! (Truly, the nothing has nothed here.) Is the fact that the empty set exists apt
to be grounded though, because it is of the kindfacts about sets, and this kind has
many actual (and hence possible) instances that are grounded? Maybe that’sOKto
say. It is also a member of the kindfacts about abstract objects, which in turn all facts
about essences are members of. But many facts about essences are not apt to be
grounded on Dasgupta’s proposal. It’s not clear how to solve this generality problem.
I think Dasgupta should reject the third parallel claim.
Let me close the discussion of aptitude for grounding by suggesting one route
that Dasgupta doesn’t employ. This route also suggests an interesting connection
between being and essence. Suppose that“xis inapt for grounding”amounts to“it
is a category mistake to claim thatxis grounded or ungrounded.”Recall that in
section 4.5 category mistakes are linked to modes of being. So one way to make sense
of Dasgupta’s proposal is to hold that there are two kinds of facts—facts about
essences and the other facts—and these two kinds of facts do not exist in the same
way. In a more metaphysically perspicuous language, the grounding predicate of that
language is type-restricted: the result of attaching it to an expression referring to an
(^43) Unless some version of trope theory is correct!