Besides again the fundamental issue about the number of "emotions", what this genre of research fails to understand is that you cannot reliably deduce an "emoti

Besides again the fundamental issue about the number of "emotions", what this genre of research fails to understand is that you cannot reliably deduce an "emotion" from something someone writes on the Internet (i.e., construct validity). And I have my doubts also about detecting "emotions" from face expressions. Maybe we might learn something from dramaturgy or some related wild field?