Defining Global Supervenience

What does it mean that certain properties globally supervene on others? The paper criticises the now standard way of spelling out the notion in terms of 1-1 correlations between world-domains and proposes a modification that escapes the difficulties. The new definition can secure the additional benefit of resisting an argument to the effect that global supervenience is theoretically dispensable. The rough and ready slogan of supervenience is that A-properties supervene on B-properties just in case there cannot be a difference in A-properties without a difference in B-properties. As Jaegwon Kim (1984) pointed out the slogan may be understood in at least three ways. Correspondingly, he distinguished three kinds of supervenience, a weak and a strong form of individual supervenience and global supervenience. Individual supervenience deals in property distributions across individuals, global supervenience in property distributions across whole possible worlds. Roughly, A-properties globally supervene on B-properties just in case whenever worlds agree in their distribution of B-properties, they also agree in their distribution of A-properties. But when do worlds agree in their distribution of some kind of properties? Philosophers have favoured different analyses of this ingredient notion of global supervenience. In section 1 I give a brief sketch of these elucidations and the resulting notions of global supervenience. In section 2 I argue that the analysis favoured in the current literature is inadequate as a foundation for satisfactory global supervenience notions. Section 3 proposes a better understanding. Section 4 shows how the resulting notions of global supervenience can secure the additional benefit of resisting an argument to the effect that global supervenience is theoretically dispensable. 1 Global Supervenience Sometimes the world-wide distribution of one family of properties B fixes the world-wide distribution of another family of properties A, even when