Hume's dictum as a guide to ontology

Abstract

In this paper I aim to defend one version at least of Hume's dictum: roughly, the idea that possibility is determined by ontology through something like independent variation. My defence is broadly pragmatic, in the sense that adherence to something like Hume's dictum delivers at least three benefits. The first benefit is that, through Hume's dictum, a physical theory's ontology delimits a range of possibilities, that I call \emph{kinematical possibilities}, which serves as a sufficiently permissive notion of possibility to sustain something like an intensional semantics for its claims, and a sufficiently demanding notion of supervenience to sustain plausible claims of inter-theoretic reduction and theoretical equivalence. The second benefit is that Hume's dictum allows us to work backwards from a range of kinematical possibilities to an ontology. This is especially useful when aiming to glean an interpretation of a physical theory, since often we are more confident that we have arrived at the right space of possibilities than that we have arrived at the right ontology. The third benefit is that Hume's dictum -- at least the version I aim to defend here -- may be applied to physical theories more or less as we find them, and therefore we can practice something resembling ontology without having to force our theories into some Procrustean bed, such as a first-order language.