Philosophical and historical background

Pessoa avoids committing himself to any strict definition of affect, emotion, motivation, and cognition. He dislikes dichotomies and views the differences as a matter of degree rather than kind. And, as an empiricist, he prefers the data to guide definitions (see Ch. 1). But it is useful to at least remind ourselves why many modern functionalist frameworks find it useful to view emotion and cognition as different beasts, though, of course, they have long moved from Plato-like fractionation of the mind into the reasoning, the desiring, and the emotive components, and other simplistic frameworks. So, why would modern functionalist care to distinguish cognition and emotion? Most important, because much of what makes human mind and human behavior in society interesting touches on this difference. Why does the heart seem to have reasons that reason cannot know? Why are we afraid of things that we rationally know are safe and do not like things we rationally should? Why does it seem that some of our decisions seem like mere justifications for emotional preferences? Why can powerful bodily and feeling reactions be elicited by simple stimulus features, in every sensory domain, such as gentle touches, baby screams, rancid smells, or seeing heights? And why can perception, thinking, decision and action radically change depending on our emotional state? These important questions seem harder to ask and answer when one blurs the emotion-cognition distinction, yet they have inspired decades of fascinating research and insightful theorizing (e.g., Haidt Reference Haidt2001; Loewenstein et al. Reference Loewenstein, Weber, Hsee and Welch2001; Schwarz Reference Schwarz, Higgins and Sorrentino1990; Zajonc Reference Zajonc1980).

More conceptually, it just seems useful to view cognition as processes concerned primarily with “representing” and “judging” – transformations of representations that aim to be “truth preserving” and which often take a propositional form. In contrast, it seems useful to reserve terms like affect, emotion, and motivation for processes primarily aimed at getting the organism to “care about” and to “do” something, and recruiting necessary physiological and experiential states to handle its concerns (Frijda Reference Frijda1988; Zajonc Reference Zajonc1980). Critically, some affective processes can involve precognitive mechanisms of sensation and perception, and even completely noncognitive mechanisms, such as global neuromodulation (Panksepp & Biven Reference Panksepp and Biven2012). In fact, some of these mechanisms are so basic that they are shared with species with clearly noncognitive status (e.g., neuromodulation of anxiety-like behaviors in species like crayfish, Fossat et al. Reference Fossat, Bacque-Cazenave, De Deurwaerdere, Delbecque and Cattaert2014). Reflecting these insights, there are several modern philosophical works that emphasize low-level, nonpropositional, perceptual, or embodied components of affect and emotion (e.g., Charland Reference Charland1995; Goldie Reference Goldie2000; Prinz Reference Prinz2006a).

In psychology, the noncognitivist view has always been amply represented, going back to the founding fathers of psychology like Wundt and James, but it found perhaps the most eloquent and passionate expression in the writings of Robert Zajonc (Reference Zajonc1980; Reference Zajonc and Forgas2000). His proposals of “affective primacy,” as well as his notion that “preferences need no inference,” still inspire contemporary research (see Niedenthal et al. Reference Niedenthal, Augustinova and Rychowska2010; Winkielman Reference Winkielman2010). As such, it is perhaps worth visiting some of the claims that affect, emotion, and motivation can be induced with minimal significant perceptual and conceptual processing and work in relative dissociation from the explicit belief system.

Affect with minimal cognition

We've already mentioned that basic affective states (simple, bivalent reactions) can be induced by rudimentary sensation-like processes. As Zajonc (Reference Zajonc1980; Reference Zajonc and Forgas2000) has pointed out, inspired by James, affective states can also be influenced by noncognitive manipulations of bodily states, including peripheral and central somatosensory feedback mechanisms (for a recent review, Winkielman et al. Reference Winkielman, Niedenthal, Wielgosz, Eelen, Kavanagh, Mikulincer, Shaver, Borgida and Bargh2015). Social psychologists have also provided plenty of examples for implicit, or even unconscious contributions to preferences, attitudes, and prejudices, some of which appear to involve simple learning mechanisms and are impervious to rational interventions (Greenwald & Banaji Reference Greenwald and Banaji1995; Winkielman et al. Reference Winkielman, Berridge, Sher, Decety and Cacioppo2011). Interestingly, there is even a class of phenomena where affect appears to result from the simple dynamics of processing. The best known is the “mere exposure effect” – enhancement of liking as a function of sheer stimulus repetition. But enhancement of liking, as measured with a variety of means, can also be obtained by enhancing perceptual clarity, contrast, or reducing visual noise – all low-level perceptual manipulations. The current view on such phenomena holds that there is a link between greater perceptual fluency and positive affect. Some accounts of this link are inferential in nature, but others merely propose that easy and fast dynamics, nonspecific signals of familiarity, and low conflict are fundamentally, perhaps innately, marked as communicating positive states of affairs (for a recent review, see Winkielman et al. Reference Winkielman, Huber, Kavanagh, Schwarz, Gawronski and Strack2012). So, perhaps in the same sense that one does not need to learn or “infer” that sugar tastes good and that injuries are painful, organisms know to “dislike” disfluency and processing conflict. In short, some seemingly “cognitive” phenomena actually illustrate the minimal conditions for affect induction, without much elaborating, structuring, categorizing, or cognitive interpreting needed to explain preferences.

Motivation with minimal cognition

Pessoa proposes that motivational processes are also highly dependent on associated cognition. However, several phenomena highlight the possibility of rudimentary, “subcognitive” influences on motivation. For example, approach-avoidance motivation can be changed by simple manipulations of embodiment, such as direct stimulation of body-related brain areas and actual body position (e.g., Price et al. Reference Price, Peterson and Harmon-Jones2012). They can also be manipulated by direct biological interventions into the underlying brain chemistry, biofeedback, and direct stimulation (for a recent review, see Harmon-Jones et al. Reference Harmon-Jones, Harmon-Jones and Price2013). Further, basic motivational signals and states (including reward signals) can spill over to completely unrelated stimuli, highlighting that they are not tightly bound to any particular cognitive representation and operate with a different dynamics (e.g., Inzlicht & Al-Khindi Reference Inzlicht and Al-Khindi2012; Knutson et al. Reference Knutson, Wimmer, Kuhnen and Winkielman2008; Winkielman et al. Reference Winkielman, Berridge and Wilbarger2005). Though Pessoa nicely highlights the neural and computational sophistication of “reward” processing, this analysis slightly detracts from the fact that on a psychological, “person” level such processing often leads to irrational pursuits and alienated desires (Berridge Reference Berridge, Brocas and Carrillo2003). Though addiction is often taken as a best example of such irrational wanting, psychologically oriented economists have highlighted a wealth of similar phenomena in daily life (Loewenstein Reference Loewenstein2007).

Cognition without affect is powerless

Notice also that taking a “separatist” perspective highlights key aspects of psychological processes that would otherwise be missed. A low-level example is that most animal learning research relies on the use of unconditioned stimuli and most unconditioned stimuli are emotive. This research illustrates that learning rarely occurs without affective input (though this point is not emphasized often enough in this research literature; Panksepp Reference Panksepp2011). A higher-level example is the case of cognitive control – nicely discussed by Pessoa. To remind, cognitive control refers to the mental processes that allow behavior to vary adaptively from moment to moment, with one of its core functions being to inhibit unwanted, yet dominant response tendencies. Often seen as the paragon of higher cognition, recent evidence suggests that cognitive control is often aided by emotion (e.g., Koban & Pourtois Reference Koban and Pourtois2014; Shackman et al. Reference Shackman, Salomons, Slagter, Fox, Winter and Davidson2011), with a recent model suggesting that control is initiated when goal conflicts produce phasic twinges of negative affect that not only focus attention but also energize goal-directed behavior (Inzlicht & Legault Reference Inzlicht, Legault, Forgas and Harmon-Jones2014). Emotional change, according to this view, is at the heart of control, and when emotion is removed by misattributions (Inzlicht & Al-Khindi Reference Inzlicht and Al-Khindi2012), reappraisals (Hobson et al. Reference Hobson, Saunders, Al-Khindi and Inzlicht2014), or using pharmaceutical agents (Bartholow et al. Reference Bartholow, Henry, Lust, Saults and Wood2012), control becomes impotent. One thus gains a deeper understanding of cognitive control when appreciating some of the emotional ingredients that go into it. Such an understanding would not come into relief by labeling all phenomena as cognitive, as has been in fashion lately. In fact, Pessoa does an admirable job rebalancing the neural picture, though perhaps at the cost of blurring some crucial, heuristically useful distinctions.

Restoring the balance

In conclusion, the target article offers a useful framework showing how cognition and emotion work together in the brain, clarifies imprecise understandings of such terms as “low-road,” or “emotional brain,” and highlights the role of emotion in supposedly cognitive functions. In our commentary, we offered some insights from psychology that support thinking about emotion and cognition as slightly different beasts, and some fascinating phenomena that illustrate their struggle. Curiously, in psychology, we currently have a problem of gratuitous and imperialistic cognitivism. For example, a recent analysis of the theorized process variables in the 2011 volume of the primary journal of social psychology (JPSP Sections I and II) found that cognitive explanatory variables were invoked almost 2.5 times as often as emotional explanatory variables, with emotion process variables accounting for less than 23% of all phenomena (Inzlicht et al. Reference Inzlicht, Tritt and Harmon-Jones2013). Given the view that emotion pervades most, if not all, of social life (Zajonc Reference Zajonc and Forgas2000), the finding that emotion was invoked as a process variable in less than 25% of all papers should raise concerns. According to some views, which assume that any transformation of input is cognition (e.g., Lazarus Reference Lazarus1984), there is now a “primacy of cognition,” with the distinctiveness of emotion being practically dismissed, reduced completely to cognition by some writers (e.g. Duncan & Barrett Reference Duncan and Barrett2007). As such, the target article and the excellent book bring a welcome “balance to the force,” without reestablishing naïve dichotomies. We hope that our commentary can inspire some additional appreciation of how emotion, motivation, and cognition interplay and sometimes separate in the mind and behavior.