Reference, conflicting intentions, and the Madagascar problem

Abstract On Kripke’s causal-historical view of reference, the reference of a proper name is fixed in a baptism and transmitted through a communication chain, provided that each speaker intends to use the name as prior speakers did. A challenge for this view arises from cases – such as the Madagascar case – in which this condition is met, yet the name nonetheless undergoes a reference shift. We defend a solution that appeals to the independently plausible idea, emphasized in action theory by Bratman, that speakers can have conflicting intentions. We argue that, in such cases, the involved intentions may differ in relative strengths and that these can change over time. When a speaker unknowingly harbors conflicting name-using intentions, a reference shift in their idiolect occurs once a previously non-dominant intention becomes dominant. Deferential relations among speakers may then help propagate this shift within a linguistic community. This account explains the reference shift of “Madagascar” while remaining faithful to the core commitments of the causal-historical view. We also show how the notion of a dominant name-using intention can be explicated in terms of a speaker’s (second-order) dispositions to conform their credences to new information and how the core idea of our proposal helps to illuminate the distinction between semantic reference and speaker’s reference.