Response‐Dependence, Noumenalism, and Ontological Mystery

Let us call representations such as concepts and terms ‘response-dependent’ just in case they are connected in an a priori manner to responses of suitably situated subjects to their referents. To possess the concept RED and master the term ‘red’, on this view, is to possess the concept and master the term that are linked to the responses of normal human beings to objects that appear red to them under normal conditions. Notions of response-dependence trace to John Locke’s (1979) analysis of secondary-quality concepts (or ‘ideas’). On one reading of Locke, the concept RED is the concept of something to which normal human beings respond by perceiving that it is red under normal conditions. Tracing the notion forward, on one reading of Immanuel Kant (1998), all concepts are connected in an a priori manner to responses that normal human beings have to objects that appear to them in particular ways. Because he understands all concepts in terms of how they relate to human beings, Kant has been called the first to give a thoroughgoing ‘anthropocentric’ account of knowledge. For Kant, all concepts are connected in an a priori manner to paradigmatic anthropocentric responses. Recently there has been much interest among analytic philosophers in response-dependence theories. Most theories proposed have been, like Locke’s, local. They concern only some concepts. One prominent response-dependence theory, however, takes its cue not from Locke but from Kant. Philip Pettit (2005b) has proposed a global response-dependence theory, according to which all concepts and terms are linked to human responses. In fact Pettit himself explicitly calls all such representations ‘anthropocentric’ (2005b: 13–7, 53–8). While Pettit introduces his global theory to rebut skepticism about rule-following (1990; 1996: chs. 2, 4; 2005b: essay 1), my interest is not in whether Pettit succeeds in doing so. It is instead in the price that Pettit must pay to be a global response-dependence theorist in the first place. As Michael Smith and Daniel Stoljar (1998) explain, Pettit’s theory entails noumenalism, the thesis that reality possesses an intrinsic nature that remains unknowable. Though Pettit acknowledges that his global response-dependence theory does entail noumenalism, he argues contra Smith and Stoljar that this commitment is not particularly problematic. For Pettit claims that his noumenalism, like Kant’s own, admits of ontological and epistemic construals. Pettit takes Smith and Stoljar to construe noumenalism ontologically, according to which there is some ‘nature’ or ‘realm’ of existence that is separate and inaccessible from the one in which we reside. Pettit himself construes noumenalism epistemically. While there