The issue of what distinguishes systems which have original intentionality from those which do not has been brought into sharp focus by Saul Kripke in his discussion of the sceptical paradox he attributes to Wittgenstein. In this paper I defend a sophisticated version of the dispositionalist account of meaning against the principal objection raised by Kripke in his attack on dispositional views. I argue that the objection put by the sceptic, to the effect that the dispositionalist cannot give a satisfactory account of normativity and mistake, in fact comprises a number of distinct lines of argument, all of which can be satisfactorily answered by the dispositionalist. Two central problems raised by the sceptic consist in explaining (1) the extension of the term which a subject uses, and (2) the fact that the subject intends to use the term with that extension, and is thus justified in her use. I adapt a suggestion of Blackburn‘s, that the extension of a word is fixed by a person‘s extended first-order dispositions to use it, and suggest that there is no coherent possibility of a subject being disposed to make systematic mistakes in connection with many ordinary words; the sceptical problem does not apply in the same way throughout language. It is further argued that an account which appeals to a subject‘s second-order dispositions to maintain a consistent pattern of extended first-order dispositions to use words is able to provide a naturalistic basis to answer the normative questions about justification of use. Three other variants of the mistake objection which are also revealed by a careful examination of Kripke‘s arguments are distinguished and it is shown that the dispositionalist has adequate resources to meet them. The appeal to second-order dispositions provides a principled way of distinguishing between those systems (animal or machine) which have original intentionality, and those which do not.