Introduction

I take it as a given that the language we speak both reflects and shapes our conceptualization of the world. Theory changes, including changes in uitiveories, often involve the construction of a new language for describing some domain of phenomena, a language incommensurable with that which preceded it. On this view, sometimes theory learning involves the construction of a new representational resource, a new language not expressible or definable in the terms of the old one (Kuhn 1962, 1982; Feyerabend 1962; Kitcher 1978; Hacking 1993; but see Davidson 1974, and Fodor 1998, on the very coherence of the notion of conceptual change; see Carey 1985, 1988, 1991, for a defense of the notion and its application to theories of cognitive development).

Accepting the existence of genuine conceptual change in the course of cognitive development is tantamount to accepting the Whorfian hypothesis. After Black, scientists used a different language for describing thermal phenomena, one which led them to see phenomena such as heating and cooling, thermal expansion, and thermal equilibrium totally differently from the ways earlier scientists in an Aristotelian or Galilean tradition did (Wiser & Carey 1983); after Levoisier, scientists saw phenomena such as burning, and entities such as chemical compounds and air, totally differently from the ways scientists who held the phlogiston theory did (Kitcher 1978; Kuhn 1982). There is a long history of attempts to decide the debates between continuity and Whorfianism with a priori arguments.