James Milford

In 2008, Graff Fara presented relative-sameness semantics, a semantics for a first-order modal and temporal language with the explicit aim of being able to render true certain contingent/temporary identity claims (relative to certain contexts). Graff Fara achieves this aim by abandoning a straightforward analysis of de re modal/temporal claims in terms of identity. Instead, such a claim is analyzed in terms of her relative-sameness relations (which need not be the identity relation), with the relevant relative-sameness relations in play determined by context and individual terms occurring in the claim. However, there is a significant problem for relative-sameness semantics concerning (i) iterated modality/temporality, and (ii) existence that renders it an inappropriate account of our modal and temporal talk. In this paper, I present that problem and a solution to it.

Lewis (The Journal of Philosophy, 65(5), 113–126, 1968) attempts to provide an account of modal talk in terms of the resources of counterpart theory, a first-order theory that eschews transworld identity. First, a regimentation of natural language modal claims into sentences of a formal first-order modal language L is assumed. Second, a translation scheme from L-sentences to sentences of the language of the theory is provided. According to Hazen (The Journal of Philosophy, 76(6), 319–338, 1979) and Fara & Williamson (Mind, 114(453), 1–30, 2005), the account cannot handle certain natural language modal claims involving a notion of actuality. The challenge has two parts. First, in order to handle such claims, the initial formal modal language that natural language modal claims are regimented into must extend L with something like an actuality operator. Second, certain ways that Lewis’ translation scheme for L might be extended to accommodate an actuality operator are unacceptable. Meyer (Mind, 122(485), 27–42, 2013) attempts to defend Lewis’ approach. First, Meyer holds that in order to handle such claims, the formal modal language L$$^*$$ ∗ that we initially regiment our natural language claims into need not contain an actuality operator. Instead, we can make do with other resources. Next, Meyer provides an alternative translation scheme from L$$^*$$ ∗ -sentences to sentences of an enriched language of counterpart theory. Unfortunately, Meyer’s approach fails to provide an appropriate counterpart theoretic account of natural language modal claims. In this paper, I demonstrate that failure.