Components,Networks,and Their Interactions

Appraisal theories of emotion have been around since the 1980s, but have rarely figured in neuroscientific investigations of emotion, which have generally focused on categorically distinct basic emotions, such as fear, anger, and disgust, or on simple two-dimensional valence/arousal mechanisms. Appraisal theory has perhaps been neglected because it is so much more complex than the idea of a small number of localized emotions or a simple two-dimensional model, but so far these models have not got us very far (Le Doux, 2012), so greater complexity may be necessary. Thus I am very much in agreement with the authors’ mission.

Most modern emotion theories are componential theories, sharing the assumption that emotions are a combination of (a) appraisals of the situation, (b) physiological responses, (c) inclinations to different kinds of action, (d) expressions, and (e) subjective feelings. Sander, Grandjean, and Scherer (2018) endorse this multicomponential view, and propose that each of these five components involves a separate brain network. However, they also propose a fundamental distinction among the components: the distinction between the elicitation of emotion and the emotional response. Physiological changes, expressions, action tendencies, and subjective feeling are all part of the emotional response, but appraisals come first and are the processes that elicit the emotional response in the first place. The elicitation network is further differentiated by its component appraisals.

This framework raises a multitude of questions. Are the five networks really separate? Or perhaps the question is when are they separate? Some theoretical perspectives see higher order processes as the product of increasing differentiation of a simpler underlying state, as in Schachter’s idea that the same generic physiological arousal is differentiated by interpretations of the situation (Schachter & Singer, 1962). Others see them as the product of the imposition of order and organization on a collection of separate processes. The authors (Sander et al., 2018) seem to belong in the second category: Our conscious experience of emotion results from the synchronization of the five underlying components.

It seems unlikely that the four response networks are completely independent of each other, since in any emotion episode they are all triggered by the same set of eliciting appraisals, so presumably Sander et al.’s (2018) theory is that they become less independent—more coherent—in the process of synchronization. This raises a second question: What is the person’s emotional experience when the four response systems have been set in motion but are not yet synchronized? Is there an emotional experience? Is there any subjective experience at all, or does that only emerge when the synchronization is complete?

A third question involves the temporal relationship between the eliciting appraisals and the components. Is it a combination of appraisals that sets the process going, or does each appraisal have its own effect on the response components? Or perhaps these response systems—or some of them—are not caused by the appraisals, but are part of them. A change in the environment initiates an appraisal of novelty, and that appraisal includes a change in heart rate, an orientation towards the stimulus, and both an expression and a feeling of something like interest. If the change is then appraised as trivial, the organism returns to whatever it was thinking or doing before the interruption and the emotion process is aborted. If it is appraised as good or bad, further changes in physiological responses, action tendencies, expressions, and subjective feelings result, and so on as other appraisals follow.

There are a lot of “networks” in this article (Sander et al., 2018)—five for the five components of emotion, and the magic number seven plus or minus two corresponding to the appraisals in the elicitation network—relevance, goal congruence, novelty, certainty, agency, coping potential, and sometimes intrinsic valence and compatibility with personal or social norms, depending on the theorist. As the authors state, with striking understatement, when it comes to brain networks, “the precise underlying neural mechanisms remain to be elucidated” (2018, p. 222). The elicitation “network” would seem to be vastly more complex than the expression “network,” and it is as yet an open question how similar any of the networks corresponding to the components or to the appraisals within the elicitation component might be to each other. Berridge and Robinson (2016), for example, find that “wanting,” related to some form of goal conduciveness, is mediated by a large, robust mesocorticolimbic system involving dopamine projections from the midbrain to the forebrain, while “liking,” perhaps equivalent to intrinsic valence, involves a much smaller and more fragile system of spatially separated but interactive small hedonic hotspots. The authors’ exhortation to researchers to look for brain processes corresponding to the various components and subcomponents is a valuable start, but there is no reason to believe that they will look anything like each other.

Finally, there is the question of the prevalence of emotional experience. Sander et al. argue that “for stimuli to be categorized as ‘emotional stimuli’ they need to have high relevance for . . . survival and well-being” (2018, p. 225). This suggests that emotions are events that punctuate a nonemotional background state, since most events in people’s daily lives have little relevance for survival or long-term well-being. Although it is highly plausible that emotions evolved to deal with stimuli that had high relevance for survival and well-being, it seems to me quite implausible that most of our daily emotional experiences have anything to do with survival or important goals. Either these authors have an extremely broad definition of survival relevance, or they face the task of fitting these commonplace emotional experiences into their theory. An alternative view is that people are always experiencing some ever-changing background stream of milder emotion, like an emotional default network that drifts along until it is interrupted by some more urgent emotional event.

One of the most important functions of science is to raise new questions, and I think Sander et al. have succeeded admirably.