Research Group
in Analytic Philosophy

Abstract

While (relativized) contents correspond to intensions, so-called "Austinian propositions" correspond to relativized contents as applied to some particular circumstance of evaluation (see Recanati 2007). We could expect the two notions to play complementary roles in an account of indexical belief: contents play a modal role (they are contingent or necessary), while Austinian propositions play a semantic role (they are true or false). Against the skeptics who claim that essential indexicality was a myth, I argue that the distinction between contents and Austinian propositions – which is only implicit in Lewis's (1979) framework – helps to get a clearer take on the meta-problem of the essential indexical. I think that Lewis is almost right on the nature of the problem, but that some distinctions remained to be made. I argue for the following claims. (1) The problem of the essential indexical arises only for an internalist doctrine of general propositions. (2) The cases of Lingens and the two gods are not just "boring" Frege-cases. (3) Unlike what happens in the knowledge argument against physicalism, they are not either "interesting" Frege-cases in which the epistemic gap can never be closed. Instead they are intermediary cases in which one has all the relevant objective knowledge plus an indexical concept used to refer, but no indexical knowledge. (4) The same two steps of epistemic progress distinguished by Nida-Rümelin (1998) in the knowledge argument must be recognized in indexical discoveries too. (5) To pose the case of the two gods, Lewis needs a notion of "elusive omniscience" somewhat parallel to his account of elusive knowledge in Lewis (1996). (6) Stalnaker (1981, 2008) is actually correct that there is no intra-world ignorance, I think, yet given the doctrine of general propositions, sometimes distinctions between entire worlds which cannot be made with propositions can be made with self-ascribed (and interesting) properties. (7) It is Austinian propositions, not just relativized contents, which exclude possibilities and thereby allow us to draw distinctions in modal space. (8) Contrary to what is commonly assumed, the contrast between de se and de re belief has nothing to do with the problem of the essential indexical arising for the doctrine of (general) propositions. The latter problem arose from a contrast between indexical and nonindexical beliefs, while the former problem arises from a contrast which is only internal to indexical beliefs: de se beliefs have special epistemic features within the class of indexical beliefs.