Abstract

The notion of referential competence (Marconi 1999) has been introduced to account for our capacity to apply lexical items, such as ‘cat’ or ‘dog’, to things like cats or dogs. What such a capacity should consist of in the cases of concrete terms seems to be clear enough. The same does not seems to hold for expressions like ‘thirst’ or ‘fear, however: how we should account for the referential competence for terms of sensations and emotions seems not quite trivial. Some embodied cognition theorists (cf. Dellantonio and Pastore 2017) have gone internalist: according to them, our capacity to apply those terms amounts to our capacity to recognize internal states typically associated with them. So, to accurately apply ‘fear’ would amount to recognize bodily states typically associated with that word (tachycardia, muscle tensions etc.). The aim of this article is to show how that internalist view seems to be just partial. Indeed, that view does capture the intuition according to which for one to apply ‘fear’ she has to have the capacity to proprioceive and recognize the sensations of fear. Internalist account does not seem to take enough in consideration Wittgenstein private language argument, however, for if understanding ‘fear’ coincides with knowing how to recognize certain bodily sensations, we would not be able to distinguish correct and un-correct cases of application of that very word. A dual solution which maintains the internal (proprioceptive) criterion while not giving up the external (perceptive) one, we will argue, seems to be in this sense a more promising solution.