In the literature on the emotions, one of the prevailing theories is the so-called James-Lange theory. This text tries to make use of some observations made by Wittgenstein regarding the grammar of psychological terms to shed light on some confusions that arise from the basic definition of this theory. First, I construe the category of a «naturalization of emotions research programme», following Lakatos' methodology, which unites Descartes's, James's and Prinz's theories of emotions under a common core. Then, I identify the problem of naturalizing intentionality as one that is derived directly from that common core. With this in mind, I attend to some of Wittgenstein's recommendations to study the grammar of the questions for the objects and the intentionality of emotions, showing that the answers provided by that research programme are unsatisfactory and that they do not respect the rules of certain uses of language.
