proponents of tropes do not accept facts as well into their metaphysics.^22 Let us
consider a fact-free trope view. So instead of invoking a physical fact that grounds a
mental fact, the proponent of tropes would directly invoke the relevant physical
property tropePand the mental property tropeM, and the specific grounding
tropeGthat relates them. We can now consider what groundsG, and note that
similar answers as earlier are available. PerhapsPalone. PerhapsPandMcollectively
groundG. Perhaps the essence ofMalone groundsG. But note that whichever of
these answers we select, there is no further room for grounding tropes to be ultimate.
The upshot is this: the possible range of answers to the question of what grounds
grounding depends on whether grounding is a universal or there are tropes of ground-
ing. We can’t approach the question of the grounds of grounding in a“metaphysically
neutral”way.
There is onefinal option worth considering. The proceeding discussion assumed
that grounding belonged to the same ontological category as other properties or
relations. But perhaps this is not so; perhaps grounding belongs to asui generis
ontological category. It does seem to be an entity unlike all the others in the
ontological system of Schaffer: it is the structuring agent rather than a part of what
is structured. It might still be right to call grounding a“relation,”but perhaps
“relation”is said in many ways. In short, there might be a categorial divide among
the properties and relations, just as there is a categorial divide among the entities.^23
Suppose this is right, and that grounding does not share a mode of being with other
properties or relations. Then one might say that grounding is notapt to be grounded
in the following sense: it is a category mistake to self-ascribe grounding. Recall that in
section 4.5 we discussed a view in which sets can sensibly be said to have members,
even the empty set, but at the fundamental level one can’t even express that
individuals have members. In the case of ground, entities of other ontological
categories can sensibly be said to be grounded or ungrounded; but with respect to
the category that ground belongs to, the type restriction rules this out. On this view,
what it means to say that grounding is not even“apt to be grounded”is that it is a
category mistake to ask whether ground is grounded.^24 (A similar view could also be
developed concerning groundingfactsas well; in section 9.4 we will discuss a view
like this concerning facts about strict essence.) This is not a view I favor; insofar as
I accept a relation of entity grounding, I view it as a being by courtesy rather than as
(^22) That said, there’s no inconsistency in doing so and it might be theoretically helpful to do so. And
perhaps facts can be identi 23 fied with suitable constructions constituted by tropes.
In personal conversation, Schaffer has suggested to me that hefinds this picture appealing. In his
words, he is interested in exploring the idea that the fundamental categorial divide is between what he calls
“entities”and“principles.”Then we have fundamental and derivative entities, where fundamental entities
are ungrounded entities. But grounding rules are principles and not entities at all. Note that Schaffer wishes
to remain neutral on whether categories are best construed as modes of being. 24
The importance of the notion ofaptitudefor grounding is stressed in Dasgupta (2015, forthcoming).