Cognition in the dynamics of emotion
The previous chapter ended in the conclusion that the discussion on the cognitive or noncognitive nature of human emotions must be brought from the level of representations and processes to their interaction and integration in real emotions. This approach focuses on the dynamic character of emotions as processes that have a temporal structure. Emotions are elicited, they unfold as multimodal responses that are regulated in various ways, and they attenuate towards an affective baseline or transform into other emotions. Yet it is not obvious on which part of the emotional response we should focus in defining emotions because an emotion can last from fractions of seconds to several minutes or even days, depending on one's theoretical approach.
Noncognitivists typically understand emotions as short-lived responses that are triggered more or less automatically by perceptually salient cues even if remote cognitive causes sometimes elicit emotions as well. However, cognitive causes are never constituents of emotion for noncognitivists. Instead, emotions involve stereotypic patterns of changes in central and autonomic nervous system activation, action readiness, facial and vocal expressions, and subjective feelings but not for instance instrumental behaviours or regulatory processes. In contrast, cognitivists prefer to analyze emotions in terms of longer time frames. In their view, appraisal processes evolve and interact with other components of emotion during the generation of emotional responses whose regulation is also often seen as an integral part of emotion. The challenge is to determine where an emotion begins and ends on the continuum of emotional processing and what if any role do cognitions have within this temporal frame.
Both cognitive and noncognitive theorists have traditionally assumed that the cognitive or noncognitive status of emotions depends on the nature of the mental states that are involved in the elicitation of emotions. If emotions emerge from either sub- or supraliminal perceptions with nonconceptual content, they are noncognitive, whereas they are cognitive if the constitutive causes of emotions are evaluative judgments or thoughts. Accordingly, the dispute on the nature of emotions has focused on the character of emotion-eliciting mental representations, and the disagreement between cognitivists and noncognitivists has moved one step further and now centres on the correct level of analysis for these representations. Cognitivists advocate functional accounts in which all mental representations that give rise to emotions of the same vernacular type qualify as cognitive, whereas noncognitivists favour algorithmic accounts that highlight differences between eliciting representations and processes. This dispute is familiar from the classic Zajonc-Lazarus debate whose repercussions in contemporary emotion theories were analysed in the previous chapter.
Yet the plausibility of the traditional approach depends on the extent to which initial perceptions or appraisals determine the identity of the emerging emotions as well as their duration, intensity, and expression. Insofar as the "late cognition" that emerges after the initial eliciting representations participates in the shaping of emotional responses and their type, it should perhaps be included in the constitution of emotion. The status of emotion as either cognitive or noncognitive could not then be judged on the basis of its eliciting representations but by an overall contribution of cognition to the phenotype of emotion. This approach is particularly relevant in the context of human emotions that become thoroughly integrated with conceptual information processing in normal ontogenetic development.
The first hurdle to any cognitive account of human emotions concerns the question of whether or not mental causes of emotion, either cognitive or noncognitive, are also constituents of emotion. Emotions come out as noncognitive if their mental causes do not belong to emotion, or if their type-identifying appraisals are noncognitive, whatever cognitive representations are sometimes or even always involved in the regulation of emotions. Jesse Prinz (2004) and Jenefer Robinson (2005) advance these two arguments for noncognitivism. Prinz distinguishes elicitors of emotion from emotional responses and defines emotions in terms of the latter, whereas Robinson suggests that emotions are always elicited by automatic affective appraisals that determine their status as noncognitive states even if these appraisals are invariably followed by cognitive monitoring. I shall first argue against Prinz that elicitors of emotions must be included in the definition of emotion. Then I shall turn this conclusion against Robinson's view about noncognitive elicitors of all human emotions. Robinson turns out to be more ambiguous about the nature of emotions than she proclaims. Her process model associates closely with cognitive process models of emotion that explicate the essential role of cognitive representations in the generation and regulation of emotional responses. I discuss two such models, process-oriented appraisal theories and psychological constructivism, at some length. Finally, I turn the overlap and integration of emotion generative and regulative processes into one more argument for a robust role of cognition in human emotions. However, the journey towards this conclusion begins from the role of mental causes in the constitution of human emotions.