The thesis of social constructionism is that emotions are shaped by culture and society. I build on this insight to show that existing social constructionist views of emotions, while providing valid research methods, overly restrict the scope of the social constructionist agenda. The restriction is due to the ontological assumption that social construction is indissociable from language. In the first part, I describe the details of the influential social constructionist views of Averill and Harré. Drawing on recent theorizing in psychology, I suggest that their fixation on language makes these approaches inadequate to the analysis of the social construction of human emotional experience. In the second part, I extend the argument to other species, suggesting that these social constructionist views are incapable of accommodating the fact, ascertained by primatologists, that animals have cultures, and that part of animal culture concerns the social molding of their emotions. I conclude that a reconstructed social constructionism should be regarded not as inimical to, but as part and parcel of, a nonreductive biology of emotions.