In this article I suggest how theories of emotion construction may inform the study of cognition-emotion interactions. To do so, I adopt the two main concepts core affect and emotions as categories: Core affect, one's current affective state, which is defined by the two dimensions pleasure and arousal, is an inherent part of any conscious experience. Specific emotions are understood as categories including highly diverse exemplars. I argue that (1) affective states can and should not be differentiated from cognitive states, and that (2) specific emotions may follow the same principles as other biological or more general categories. I review some empirical evidence in support of these ideas and show avenues for future research.