Abstract

This paper examines the two leading simulation approaches to mental selfascription, Alvin Goldman’s introspectionist account and Robert Gordon’s nonintrospectionist, “ascent routine” account, with a view to determining their adequacy as accounts of our ordinary self-ascriptions of mental states.I begin by reviewing the features of everyday mental state ascriptions and argue that an adequate account of mental state attribution must be able to account for the salient features of those mental attributions we make by using the sentences of a language we know (section 1). By way of introducing the simulation accounts, I outline the tenets of the ‘Theory’-Theory of mental state ascription and sketch the simulationists’ objections to it (section 2). The specific proposals of Alvin Goldman (section 3) and Robert Gordon to ascent routine simulation approach (section 4) are then examined in detail. I argue that both Goldman’s and Gordon’s approaches to mental self-ascription have serious shortcomings. However, the difficulties facing their respective positions suggest *investigating a third approach to the self-ascription of mental states.