From beginning, scientific study emotions has been encumbered by inflexible rules and assumptions about Right Way to conduct research. Most commonly, it was understood that right way to study emotions was to measure one or more bodily processes: sympathetic nervous system activation, muscular tension, facial expression, vocal expression, respiration, skin conductance, and brain wave activity have all been recommended as best indices emotion, and, for more than 50 years, most experimental studies emotion have used at least one these methods (Woodworth & Schlosberg, 1954). William James's (1890/ 1950) emphasis on muscular and visceral feedback undoubtedly played a major role in this concern with physical manifestations emotion (Ellsworth, 1994), but its roots are far more ancient, tracing dichotomy between reason and passion that has dominated discussions emotion since beginning Western philosophy (Solomon, 1993). For centuries, scholars have argued that emotion is not of in same way that reason is; instead, emotion is of body in a way that reason is not. In experimental psychology, this attitude translated into a preoccupation with physiological and other bodily measures so stultifying in its effects that, by mid-20th century, research on emotions was one dullest enterprises in psychology. It took Arnold (1960), Schachter and Singer (1962), and Lazarus himself (e.g., Lazarus, Averill, & Opton, 1970) to bring mind into world emotions and revitalize field. In his target article, Lazarus once again reminds us how pernicious distinction between rationality and irrationality has been for study emotion. The term rational typically serves more as a value judgment than a scientific concept, and I think Lazarus is right when he suggests that our thinking would be clearer if we abandoned it. Of course, experimental psychologists in their shiny laboratories were not only ones who were struggling to understand emotions. Almost simultaneously, and quite independently,' Freud and his followers explored complexities emotional life. Like experimentalists, Freudians knew one right way to study emotions, and it was psychoanalysis. On surface, psychoanalytic method seems opposite brass-instrument technology experimentalists: The patient simply talks, trying to say whatever comes to mind, what emotions feel like, and what they mean to him or her, and therapist listens. But therapist does not listen in a straightforward manner, does not take patient's account his or her emotional experience at face value; therapist is listening not to story recounted, but to hidden story, story repressed emotions inaccessible to patient' s conscious mind. Thus, although Freudians and experimentalists disagreed about right way to understand emotions, they agreed that serious consideration conscious experience emotion, as reported by subject, was not right way. Lazarus seemingly shares this dim view when he expresses doubts about the validity appraisals when they are assessed by self-report methods. Yet, self-report methods are central to current appraisal models emotion,' including Lazarus's own cognitive-motivational-relational theory (Lazarus, 1991), and I would argue that present lively renaissance research on emotion stems partly from scientists' new interest in people's conscious experience and their new willingness to take what people say about their own emotions seriously. Nonverbal and physiological measures can be enormously useful, but clear-cut nonverbal signs emotions are uncommon and often less finely nuanced than language. And, for questions about people's appraisals or interpretations a situation-for example, their attributions responsibility for a pleasant or unpleasant event-nonverbal measures are even more problematical guides. I think emotion researchers have barely begun to learn how to use valuable tool self-report, barely begun to think interesting questions to ask respondents, and I would hate to abandon it so soon. Of course, Lazarus is not recommending that we abandon self-report measures, only that we be cautious
Stored only when both papers are in the corpus — 10 citations exist upstream.