Abstract
The central contention of this article is that the classificatory scheme of contemporary affective science, with its traditional categories of emotion, anger, fear, and so on, is no longer suitable to the needs of affective science. Unlike psychological constructionists, who have urged the transition from a discrete to a dimensional approach in the study of affective phenomena, I argue that we can stick to a discrete approach as long as we accept that traditional emotion categories will have to be transformed in order to do any scientific work. I conclude by articulating some general rules for turning traditional emotion categories into suitable scientific tools.