Though Kuhn and Feyerabend introduced the idea of referential incommensurability, many have found their account problematic. Subsequent developments in the theory of reference which address some of these problems are reviewed here, from the Kripke and Putnam semantics to the Ramsey-Carnap-Lewis account of theoretical terms with recent modifications due to Papineau. However Stich argues that such theories of reference can do no work towards solving issues in the philosophy of science. To overcome this objection, and to remedy some remaining shortcomings in the Ramsey apparatus used, we provide some epistemic conditions for naming. The resulting account yields a less problematic understanding of incommensurability, one that is consistent with realism.