Reference Magnetism

Abstract

The influential response of David Lewis to the rule-following problem (posthumously described as "reference magnetism") is described. Three distinct approaches are traced to Lewis's seminal papers. The first treats the world's structure as metaphysically providing resources that supplement what individuals have to determine reference. Our words (concepts) have determinate reference beyond the psychological and neurophysiological resources of any individual, and beyond what any community of such individuals can manage on its own. The second treats interpreters of natural languages as required by semantic theory (along with background scientific practice) to impose determinate reference on the terms of those languages, along with natural kinds as the relata of the kind terms of those languages. The third treats the a priori constitutive imposition of natural kinds to be required by the Moorean facts of determinate reference and by semantic theories (that presuppose determinate reference) being "the only game in town." I show that none of these approaches to reference magnetism work. Straight solutions to the rule-following paradox try to preserve classical correspondence metaphysics between our assertions and the world. This can be by attempting to show that Kripke's meaning sceptic hasn't noticed resources we actually have for making our references to things in the world determinate, or resources we have for fixing definitively that we intend to count instead of to quount. Or it can be by supplementing in some way an individual's otherwise insufficient capacities to manage this by resources located outside that individual. Kripke's meaning sceptic claims that the needed resources (the grounding facts) aren't to be found in an individual's psychology or neurophysiology. One family of straight solutions-that I criticized in Sect. 2.5-respond to this challenge by locating the additional needed resources in the community of speakers that the individual belongs to, or in a community of speakers that an interpreter of that individual belongs to. A different family of straight solutions-ones associated most prominently with the work of David Lewis-locates those resources either in the world itself, or