What about the question of the ground of grounding? Assume that grounding is a
universal. Perhaps one connective relation between properties and relations is the
relation of being a constituent. Perhaps since grounding is a disjunction of conjunc-
tions, each of the conjunctions is a constituent of grounding. Each of these con-
junctions (which are disjuncts of grounding) is more real than grounding, and bears
an appropriate connective relation to ground. On the second strategy, then, it would
be plausible to say that each is a ground of grounding. (This wouldfit the idea that
disjunctive properties are grounded in each of their disjuncts.) Perhaps a similar but
more complicated story could be told about facts of grounding.^32
On the second strategy, whatever work the notion of grounding is called to do, the
notion of comparative reality can do just as well. We therefore have a choice between
two systems, each of which accepts the same plurality of connecting relations. But
one of these systems takes the notionx is at least as real as yas basic and defines
existence simpliciter and grounding in terms of it, while the other system takes both
the notion of existence and the notion of grounding as basic. (As I noted in section 7.6,
Schaffer (2009: 374) explicitly denies that existencecan be defined in terms of
grounding.) I suggest that thefirst system scores better with respect to ideological
parsimony—we needn’t take the notion as primitive once we grasp the recipe for
defining it and settle on an adequate systematic metaphysics—and since it can do the
same work as the second, it is to be preferred. (Compare the argument given here
with that of section 7.6 for the priority of degrees of being over naturalness.)
Suppose we do not accept the second strategy. There is a third strategy available,
which is to deny thatx is at least as real as yis comparable in the sense defined in
section 7.5. Ifx is at least as real as yis not comparable, and we want to capture the
connection between a thing and its modes wholly in terms of the grounding relation,
we can do so. In the case mentioned earlier, one needs to deny that the mode of
redness is equally as real as the mode of blueness, and hold that neither mode is more
real than the other. On this third strategy, they are not related to each other by the at
least as real as relation.^33
The third strategy is a mirror of thefirst strategy. According to thefirst strategy,
the grounding relation induces as much structure as comparative reality. It is illusory
that grounding is richer. On the third strategy, comparative reality has more struc-
ture than we initially thought. On this third strategy, grounding and comparative
reality are the same phenomena under two different guises.^34 This suggests that
“ground”and“comparative reality”are mere notational variants of each other. This
possibility is especially salient in light of the considerations of sections 7.4–7.7,
(^32) Again, Wilson (2014) is relevant.
(^33) On each of these strategies, the grounder is ontologically superior to the grounded. Audi (2012: 102)
rejects this. As I see things, my project needn’t accommodate everything every grounding theorist says
about grounding in order to be successful. 34
Daly (2012: 94) suggests that talk about grounding and degree of reality might just be the same thing
under different names.