Abstract
At the empirical center of James’s theory of emotion is the prediction that people induced to act emotionally will report feeling the corresponding emotion. While the research more or less inspired by James is complex, it is also large. Reisenzein and Stephan (2014) identify a number of problems in this literature, but we think, on balance, the research supports James’s hypothesis. At a minimum, there are literally hundreds of studies showing that people induced to act as if they felt an emotion, an attitude, or a judgment report those feelings. Critics of James’s theory are left with the task of accounting for these findings.
An important place to begin is with the nature of our task, in contrast to that of the other contributors to this special section. We are not primarily concerned with the origins and evolution of James’s theory, but with the central empirical implication of his theory, that feelings might arise from and be dependent upon behaviors and bodily changes. James was certainly the most famous proponent of this challenge to common sense, and we cite him because of his historical importance, not because we think his proposals are necessarily the best statements of the position. In fact, we were more influenced by the philosopher Gilbert Ryle (1949) and the social psychologists in the cognitive dissonance tradition than by James. This is the reason we prefer the label “self-perception” for our theoretical position (for an extended discussion, see Charles, Bybee, & Thompson, 2011).
What is common to James, Ryle, and many social psychologists is the broad hypothesis that feelings do not cause behaviors (including bodily changes), but rather that feelings follow and are dependent upon behaviors. In our article in this special section, we surveyed the varieties of behaviors and feelings that have been explored by this approach, and made the case that at a minimum there is a large body of evidence that must be explained by any proposed alternative. We are sensitive to Reisenzein and Stephan’s (2014) claim that many of the studies we cite might not be completely supportive of our position. In all fairness, much of the research on facial, bodily, and physiological changes in association with specific emotions is rather mixed. Yet our re-review suggests to us that our interpretations are at least as reasonable as Reisenzein and Stephan’s reading, but of course, interpretations of complex phenomena are always subject to selective perception. Indeed, the occurrence of selective perception, taking place outside of the perceiver’s awareness, has been demonstrated by the kinds of psychologists who view feelings and attitudes as interpretations of behaviors (e.g., Festinger, 1964; Frey, 1986).
In our article, we tried to emphasize that physiological changes are only one of a group of responses that occur when an emotionally inciting event happens. During such an event, the inciting stimulus must be interpreted and offers the person situational cues about their emotion: A fear response is elicited by the bear, due to the knowledge that bears have big teeth and bad tempers. This higher-level appraisal generates a multitude of bodily responses such as facial expressions, postures, and physiological changes. Reisenzein and Stephan (2014) point to evidence that facial expressions and certain physiological responses do not always occur along with an emotional experience, but past research rarely examines all of these bodily changes at the same time. In response to the bear appraisal our hearts beat faster, our palms sweat, our facial expressions shift toward a fear expression, and/or our postures become defensive, any of which might be sufficient to induce an emotional feeling.
Reisenzein and Stephan (2014) suggest that evidence of individual differences in the effects of bodily responses on feelings is a weakness, and it certainly is for James’s version of the theory, since he never seems to have considered this possibility. But in our view, this finding reflects just what research is supposed to do: discover new complexities in phenomena. Self-perception theory neatly proposes a role both for situational cues and for internal and behavioral cues (Bem, 1972), and individual differences in a person’s reliance on each set of cues seem reasonable from this perspective.
Rather than rejecting James’s theory based on mixed evidence, it is also useful to ask what other theory could explain the ability of bodily changes to induce emotional feelings? Beyond the emotion research we reviewed previously in the issue, the literatures on self-perception and attitudes, as well to on the foot-in-the-door procedure, support a similar process to that proposed by James. People use their past behaviors and internal states as cues to their attitudes and their future actions (e.g., Bem, 1972; Burger, 1999). This work offers additional support for a Jamesian and self-perception theory of emotions.
Finally, we believe that many of the questions raised by Reisenzein and Stephan (2014) offer promising avenues for future research. Why are some bodily changes experienced as emotions while others are not? Why do studies often indicate that facial expressions do not co-occur with emotional experiences, but experimental manipulations indicate that facial expressions and bodily movements (particularly when in combination) can initiate emotional experiences? These are interesting questions that need to be empirically tested. Yet just because these questions have yet to be answered does not mean that James’s theory should be discarded. We simply have more to investigate.