Necessity, The A Priori, And The Standard Meter

This article critically examines Saul Kripke’s (1972) argument for the separability of necessary truths from truths known a priori, focusing on his criticism of the standard meter case presented by Wittgenstein (1968). It attempts to show that Kripke’s argument is unworkable on any of several readings. Wittgenstein’s own broadly conventionalist account of necessary truth is then considered in the light of the standard meter example. Saul Kripke’s Naming and Necessity (1972) advances a series of theses in metaphysics and the philosophy of language which pose potentially serious problems for many conventionalist accounts of necessary truth. In particular, Kripke’s famous conclusion that necessary truths can sometimes be known a posteriori and that some contingent truths may be known a priori is incompatible with treating statements of necessary truth as disguised expressions of linguistic norms which, being products of stipulation, must be known a priori. I take the later Wittgenstein’s account of necessary truth to be conventionalist in at least this broad sense. As such, the arguments and illustrations used by Kripke to defend his separation of what is necessary from what is a priori would, if sound, undermine a key element of Wittgenstein’s account of necessary truth. I will here demonstrate why this is, and offer a defense of Wittgenstein’s position against Kripke. 1. THE STANDARD METER Suppose that we stipulate that some object, say a particular bar of metal, is to be the standard for some linear measurement, such as one meter. Given that we have selected a bar, and stipulated that it is the standard for one meter, can we not say that it is necessarily the case that the standard bar is one meter long? Not according to either Kripke or Wittgenstein, albeit for very different reasons. For Kripke, the claim that the standard meter is one meter long is not necessarily true, even if the bar is defined to be the standard for something’s being one meter long. For Wittgenstein, it simply cannot be meaningfully claimed that the standard meter bar is or is not one meter long. The supporting arguments that each philosopher provides Synthese 121: 291–307, 1999. © 2000 Kluwer Academic Publishers. Printed in the Netherlands.