Presuppositional Indexicals

Kaplanian, two-dimensional theories secure rigidity for indexicals by positing special contexts and semantic mechanisms reserved only for indexicals. The result is a deep and unexplained chasm between expressions that depend on the extra-linguistic context and expressions that depend on the discourse context. Theories that treat indexicals as anaphoric, presuppositional expressions (e.g. Hunter & Asher 2005, Maier 2006 and 2009, Roberts 2002, Zeevat 1999) have the potential to be more minimal and general than Kaplanian, two-dimensional theories—the mechanism of presupposition, unlike that of Kaplanian character, is useful for the semantics of a great many expressions. Maier (2006, 2009), however, has argued that presuppositional theories of indexicals must be supplemented with a twodimensional semantics in order to secure rigidity for indexicals. If this is right, then presuppositional theories of indexicals will suffer from the limitations of Kaplan’s system. This article argues that Maier is not right on this point: presuppositions can completely replace Kaplanian characters. A presuppositional theory can secure rigidity for indexicals without positing two independent dimensions of meaning that can never interact; in particular, it can do so without positing that indexicals have a special kind of meaning that by its nature can never interact with the kind of meaning that Kaplan called ‘content’. The result is a more general, minimal, and flexible theory that better handles the data on indexicals. Kaplanian, two-dimensional theories secure rigidity for indexicals by positing special contexts and semantic mechanisms for indexicals that cannot be extended to treat expressions, like third person pronouns, that are intuitively similar to indexicals but can depend on the discourse context. Theories that treat indexicals as anaphoric, presuppositional expressions (e.g. Hunter & Asher 2005, Maier 2006 & 2009, Roberts 2002, Zeevat 1999) have the potential to be more minimal and general than Kaplanian, two-dimensional theories—the mechanism of presupposition, unlike that of character, is useful for the semantics of a great many expressions. Maier (2006, 2009), however, has argued that presuppositional theories alone cannot secure rigidity for indexicals; a presuppositional theory must be supplemented with a two-dimensional one if it is to treat indexicals. Accordingly, Maier (a) adopts contexts that divide extra-linguistic information (information about utterance events, perception, etc.) and information introduced via discourse and (b) adds an anchoring mechanism for the extra-linguistic level of his contexts so that discourse referents at this level will receive a rigid interpretation. He then goes on to (c) treat dependence on the utterance context and dependence on the discourse context as two different and independent kinds of meaning and (d) force dedication to one kind of meaning (extra-linguistic context dependence) into the semantics of indexicals. If Maier is right that a presuppositional theory must take steps c and d, then we lose the minimality and generality promised by such a theory and are left with the limitations of a Kaplanian one. This article argues that while steps a and b are necessary to secure rigidity for indexicals in a presuppositional theory, steps c and d are not only not necessary, but they are also not supported by the data on indexicals. Steps a and b, together with Hunter & Asher (2005)’s resolution strategies, are sufficient to secure rigidity for indexicals in a presuppositional theory. Adopting these three