A great deal has been written about ‘would’ counterfactuals of causal dependence. Comparatively little has been said regarding ‘would’ counterfactuals of ontological dependence. The standard Lewis-Stalnaker semantics is inadequate for handling such counterfactuals. That is because some of these counterfactuals are counterpossibles, and the standard Lewis-Stalnaker semantics trivializes for counterpossibles. Fortunately, there is a straightforward extension of the Lewis-Stalnaker semantics available that handles counterpossibles: simply take Lewis's closeness relation that orders possible worlds and unleash it across impossible worlds. To apply the extended semantics, an account of the closeness relation for counterpossibles is needed. In this article, I offer a strategy for evaluating ‘would’ counterfactuals of ontological dependence that understands closeness between worlds in terms of the metaphysical concept of grounding.
