in which we addressed whether naturalness and degree of being are the same
phenomenon under distinct guises. My inclination in this case is that the guise of
comparative reality is the more perspicuous guise, since comparative reality is also
the semantic value of a polyadic quantifier.
This discussion so far has operated on the assumption that we are working only
with one relation of ontological superiority, namely, that of comparative reality (or
degree of being). Dropping this assumption yields a more complicated but also more
attractive picture of grounding. One important claim of this book is that there is
more than one kind of ontological superiority. The way in which an attribute or
mode is grounded in its substantial bearer is not the way in which a shadow is
grounded in its host. This is not simply because the connective relations in question
differ; they do, but that understates the difference. In the case of a mode and a
substance, the connective relationship just is a relationship of ontological superiority:
when a mode is instantiated by a substance, that modeexists inthe substance.
(Recall section 2.4.2.) Holes do not exist in their hosts in the way that a mode exists
in a substance. (Recall section 5.3.) Simply saying that both holes and modes are
grounded in substances covers up a vast difference in the how of it.^35
That’s my deeper reservation about strategy three. But it also points to a need to
revise strategy two. Rather than contextually defining grounding as the disjunction of
conjunctions of the more real than relation with a suitable connective relation, we
should think of grounding as the disjunction of conjunctions of some relation of
ontological superiority coupled with a suitable connecting relation (that might in
some cases be identical with a relation of ontological superiority). And even this
probably is not sophisticated enough; probably it is better to call this disjunctive
relationGand then define grounding as the transitive closure of G.^36 The“logic”of
ground still might turn out quite messy, or it might be rather streamlined—what it
looks like will depend largely on what the correct metaphysics is, which after all is
what supplies us with the information about which connecting relations are suitable.
Given this picture of grounding, is it the case that ifxgroundsy, thenxgroundsy
in every world in which they both exist? The answer is probably“no.”I assume that if
xis ontologically superior toy, then this is the case in every world in which they are
found. (We will reassess this assumption in section 9.7.) But the relevant connecting
(^35) I think much of the grounding literature is guilty of this kind of covering up. See, e.g., Schaffer (2009:
375 – 6), who puts holes and singletons on the same“great chain of being.”Similarly, Koslicki (2012b:
206 – 11) distinguishes different relations of what she calls“ontological dependence,”but lumps the kind of
dependence of a hole on its host under the same umbrella as the kind of dependence a trope has to its
bearer. On the other hand, Koslicki (2015: 329–39) defends a kind of lightweight grounding pluralism that
is sensitive to differences in how a coarse-grained relation of grounding is implemented. 36
There’s an interesting question about what to say if grounding is a contrastive relation, as is defended
by Schaffer (2012). We could posit a contrastive relation of comparative reality, but that feels to me
intuitively implausible. That said, I am not convinced by the case Schaffer (2012) presents for“going
contrastive”; see Javier-Castellanos (2014) for criticism.