Attention, Emotion, and Evaluative Understanding

Abstract

This paper assesses Michael Brady’s claim that the ‘capture and consumption of attention’ in an emotion facilitates evaluative understanding. It argues that emotional attention is epistemically deleterious on its own, even though it can be beneficial in conjunction with the right epistemic skills and motivations. The paper considers Sartre’s and Solomon’s claim that emotions have purposes, respectively, to circumvent difficulty or maximize self-esteem. While this appeal to purposes is problematic, it suggests a promising alternative conception of how emotions can be teleological. The fact that emotional attention manifests dispositions, which have been habituated by repeated association with pleasure or relief, explains how the emotion can acquire the function of producing pleasure or relief. Hence, the emotion can have ‘non-cognitive function,’ in which its patterns of attention reliably produce beliefs that disregard or distort the truth. Not only is non-cognitive function the default condition of emotion prior to any training, but even those who have successfully trained their emotions often revert back to non-cognitive function when faced with trauma or stress.

1 Emotion and Understanding

Michael Brady (2013) argues that “emotions facilitate or promote evaluative understanding” (p. 118). Understanding, let us suppose, is a variety of epistemic success that encompasses, but is not exhausted by, knowledge. It entails recognizing connections between various elements, and the connections so recognized correspond to real causal or explanatory relations. Evaluative understanding is then “understanding why objects and events have the evaluative properties they do … [It] is a matter of correctly grasping why something is shameful, dangerous, lovable, disappointing” (Brady 2013 , p. 142). That is to say, one with evaluative understanding recognizes real causal and explanatory relations that underlie the evaluative properties things possess.

The reason Brady offers in support of this conclusion is that emotions essentially involve the “capture and consumption of attention,” which refers to the re-orientation and persistence of attention upon the emotion’s intentional object. Define attention as the capacity to highlight or select a target for further cognitive processing (see also Jennings 2012, pp. 536–537; Watzl 2011, p. 847).While other philosophers of emotion have emphasized attention, Brady stands out for defending the idea that attention underwrites a significant, and overall positive, epistemic role for emotions. He accordingly opposes the received philosophical wisdom that emotions, as R. S. Peters says, “cloud and distort judgment” (1962, p. 120).

While I am generally convinced that emotions involve the capture and consumption of attention, there is a kernel of truth to the received wisdom. A person’s attention during an emotion often disregards or distorts the evaluative properties objects actually have. Drawing from the work of Jean-Paul Sartre and Robert Solomon, I argue that the inherent structure of emotion, and not something external or accidental to emotion, obstructs evaluative understanding. The reason, as I shall argue, is that emotional attention manifests dispositions, which are underlying tendencies to some outcome. The fact that all such dispositions are habituated, i.e., shaped by repeated association with pleasure or relief, shows that the capture and consumption of attention can have the function of producing pleasure or relief. In consequence, the emotion has non-cognitive function, which means it is governed by a principle that reliably conflicts with the truth. Not only is non-cognitive function the default condition of emotion prior to any training, but even those who have successfully trained their emotions can revert back to non-cognitive function when faced with trauma or stress.

This more pessimistic approach does not deny that, under the right conditions, emotional attention can be utilized to achieve evaluative understanding. However, “the right conditions” requires the presence of other psychological factors that are external or accidental to emotion. In particular, it requires epistemic skills such as the capacity to reflectively monitor one’s emotions. Thus, Brady’s account may correctly describe the epistemic role of many or most emotions. Where he errs, however, is conflating an achievement with a constitutive feature of emotion.

2 The Capture and Consumption of Attention

To begin, I want to consider what the “capture and consumption” model is and why Brady thinks the model supports the conclusion that emotions yield evaluative understanding. Then, I consider how the possibility of rationalization complicates Brady’s argument.

The “capture and consumption” model holds emotion to be a temporally extended process comprising two distinct phases of attention.¹ The first (so-called “capture”) phase is where one’s attention is re-oriented toward objects and events in the environment. In so doing, the emotion provides an appraisal of the object, and thus, enables the person to act in response.² In this phase, emotions are indiscriminate: they are very fast, reflexive reactions, elicited by a very narrow range of stimuli. For example, fear quickly and reflexively re-directs the individual’s attention upon some potential threat, e.g., a dog approaching very quickly, and then delivers an appraisal of that threat’s danger so that the person may fight or flee as necessary. Thus, the capture phase specifically facilitates action, especially in situations that are immediately pressing and in which the delay caused by deliberation would be too costly.

During the second (so-called “consumption”) phase, however, one’s attention persists upon the new object or event. These “hold sway over us, often making it difficult to disengage our attention and shift focus elsewhere” (Brady 2013, p. 92; see also Brady 2010, 2011). For example, the fearful person’s attention remains upon the dog and its putatively danger-constituting qualities such as claws or teeth. During this time, one’s behavioral reaction is inhibited and various cognitive capacities, including for deliberation and judgment, are activated. The previous “capture” phase may lead to a behavioral response which effectively prevents the person from carefully determining the danger-constituting qualities of the emotion’s object. When this does not happen, however, the emotion recruits the person’s cognitive resources and directs them upon the object, even to the exclusion of other important matters. In this way, the person with the emotion remains in a state of heightened sensitivity to and awareness of the emotion’s object.

According to Brady, the model of emotion as “the capture and consumption of attention” supports the conclusion that emotions facilitate evaluative understanding, because emotions themselves stand in need of justification. They are the sort of mental phenomena for which reasons are necessary (see also Taylor 1975; D’arms and Jacobson 2000). For example, when Ella says that she is jealous of Bill’s new promotion, Ella can legitimately be asked why. What supports her jealousy? Is it the higher pay, or the increased authority? Ella’s reasons can in turn be criticized. If Ella insisted that higher pay was her reason for being jealous, but it turned out that Bill’s new pay was not higher, her emotion would be unfounded. According to Brady, those who experience emotion “often feel the need to discover reasons or evidence in support of our emotional experience in normal circumstances” (2013, p. 86). The person does not simply assume that his initial emotional appraisal is correct and well-founded. Instead, he admits the demand for justification and tries to meet it.

The demand for justification is met specifically by the second “consumption” phase of emotion. The consumption of attention motivates a search for reasons for or against the initial appraisal of the object. The individual looks for and deliberates upon on those properties in the environment, such as the sharpness of the claws or the pleasantness of an increase in pay, that explain the emotion’s initial action-oriented appraisal. Thus, the second phase serves to refine the quick, action-oriented appraisals found in the “capture” phase. It does so by deploying the person’s capacities for deliberation and judgment to answer the question of whether the object really is as the emotion presents it. The emotion thus facilitates evaluative understanding, since the person comes by way of the consumption of attention to recognize real explanatory relations that underlie his emotional appraisals. The emotion not only helps the person to discover whether its object is dangerous or desirable, but also why it is dangerous or desirable. This shows how, according to Brady, “emotions have developed to solve a problem that emotions are themselves responsible for setting up,” viz., that of meeting the demand for justification (2013, p. 109).

However, the same search for reasons often yields rationalizations. Rationalization is a post hoc explanation that construes the explanandum as more rational, moral, or otherwise better than it actually was.³ According to Peter Goldie, emotions are constitutively prone to rationalization. He points out how the search for reasons during an emotion is not in the service of truth. Rather, the initial appraisals function as “idées fixes with which reason has to cohere” (2004, p. 99). Sometimes this search may exaggerate evidence that exists or fabricate evidence altogether. Goldie says “when we are afraid, we tend unknowingly to seek out features of the object of our fear that will justify the fear – features that would otherwise (that is, if we were not afraid) seem relatively harmless” (2004 , pp. 99–100). If Goldie is correct, the capture and consumption of attention does not facilitate evaluative understanding, because it often seeks to vindicate emotion regardless of whether that emotion is actually justified.

Brady admits the possibility of rationalization, but denies that it undermines his conclusion. He says that “the urge to discover reasons that bear on emotional accuracy is often so strong that when individuals cannot discover genuine reasons, they are inclined to invent them.” (2013, p. 88). In short, the person desires to understand to such an extent that he resorts to exaggeration and fabrication. However, Brady maintains that the capture and consumption of attention can be non-rationalizing, so long as the person has already achieved some evaluative understanding. Once this preliminary evaluative understanding is achieved, “there is no psychological pressure or need” to rationalize (2013 , p. 178). Since the person grasps why the object has certain evaluative properties, his desire to understand is satisfied and he no longer has any motive to rationalize.

This rebuttal, however, makes an important concession which undermines Brady’s argument. It concedes that what facilitates evaluative understanding is not strictly the emotion itself, but rather the emotion in conjunction with some prior evaluative understanding. However, the intended conclusion was that emotions have the sort of intrinsic nature that underwrites a positive epistemic role.⁴ That Brady meant this strong conclusion is evidenced by his various statements that emotions “raise questions” about themselves, or alternatively, have the “function” of yielding evaluative understanding and “motivate” it (2013, p. 97, 109, 177).⁵ However, the possibility of rationalization shows that evaluative understanding could be due to psychological factors that accompany some – but not all – instances of emotion in normal circumstances. Emotional attention may only facilitate evaluative understanding for an individual who possesses certain epistemic skills and is suitably motivated to understand why the emotion’s object has various evaluative properties. If these psychological factors do not normally accompany every emotion, perhaps because they are only found in mature, rational individuals, then Brady has ultimately confused a contingent feature of emotion for an essential one.

It is possible to defend Brady’s account with a broadly functionalist approach to the mind. The functionalist defines mental states by the causal role those states play in a complex system. To take a simple example, belief represents facts in the environment. In order to play its role, the mental state requires the cooperation of other parts of the complex system. So, to represent the facts, belief takes information from the system’s perceptual inputs. Belief could not play this role unless the perceptual inputs were operating normally. However, it is certainly possible to have an individual whose perceptual inputs were destroyed or malfunctioning. During that time, the beliefs she forms do not represent facts in the environment, and she fails to form beliefs about facts that she otherwise would. Such a possibility would not justify denying that beliefs essentially have the role of representing the facts in the environment. In general, then, a mental state’s essential role may presuppose the operation of other parts of the complex system, but these parts (or certain operations of those parts) do not necessarily accompany that mental state. Hence, critics could argue, the possibility that many emotions lead to rationalization when not accompanied by prior evaluative understanding and the desire to understand does not impugn Brady’s claim that facilitating evaluative understanding is an essential role of emotion.⁶

In reply, it is correct that the analysis of a mental state in terms of functional roles presupposes other parts of a complex system, parts which do not necessarily accompany the mental state being analyzed. However, such analyses are also implicitly restricted to the “normal operating conditions” of the complex system. The claim that belief’s role is to represent facts in the environment describes what belief actually does within the system when all the other parts are in place and operating as they typically would. So, while it is technically possible for the state to fail to play its role, because other parts are missing or not operating as usual, it is not possible for it to fail to play that role in normal operating conditions. The latter failure would provide conclusive evidence the state’s role had been mischaracterized.

I contend that precisely this sort of mischaracterization besets Brady’s account of emotion. It is a mistake to deny that emotions can yield evaluative understanding in conjunction with a desire to understand and some preliminary evaluative understanding. It is not, or not obviously, a mistake to deny that a desire to understand and some preliminary evaluative understanding are part of the normal operating conditions of the complex system in the same way that perceptual inputs are. Firstly, the desire to understand and the preliminary understanding are plausibly regarded as achievements, which indicate excellence or success that comes through training. By contrast, having normal eyesight and hearing is not an achievement. Secondly, when a person rationalizes, she may run afoul of numerous substantive epistemological norms, but her state is not obviously defective as an emotion. The irrational fear that leads a person to rationalize could still be a paradigmatic case of fear. By contrast, beliefs about the environment that are produced by destroyed or malfunctioning perceptual inputs are defective as beliefs. So, the objection for Brady is that, even in normal circumstances, the emotion need not be accompanied by those psychological factors necessary to yield evaluative understanding.

3 Non-Cognitive Purposes

The upshot of the previous argument is that there is a critical difference between what is inherent to emotion and what contingently accompanies emotion. It very well may be that the capture and consumption of attention can facilitate evaluative understanding in those who want to understand and who have already achieved some measure of evaluative understanding. However, no reason has so far been given to suppose it does so when the person lacks these factors. In fact, emotional attention, on its own, might even counteract evaluative understanding. For the remainder of the paper, I defend the claim that the capture and consumption of attention in emotion has an inherently negative epistemic role. I begin with the views of Jean-Paul Sartre and Robert C. Solomon. In separate works, they each maintain that emotions serve purposes which are non-cognitive. By way of these emotions, the person can disregard or distort the truth. While this view faces its own problems, it suggests a promising direction to defend the received wisdom that emotions cloud or distort judgment.

Sartre and Solomon each assert that emotions are purposive. Solomon claims that “emotions are aimed at changing the world… [they] serve the ends of the subject, and consequently can be explained by … ‘in-order-to’ explanations” (1980, p. 11). Sartre similarly calls emotions “transformations of the world” (1948, p. 58; see also 1981, p. 218). The main idea is that a person has an emotion in order to bring about some desired outcome. For example, a cheating husband may get angry at his wife so that she share some responsibility for the failure of their relationship. Or, a distant acquaintance may feel grief at a death in order to count among the intimate relations of the deceased. These emotions non-accidentally “fit into the structure of the subject’s purposes and intentions” (Solomon 1980, p.16). That is, the teleological explanations are the sort that apply to intentional actions. In fact, according to these authors, emotions are, or comprise, actions.⁷ This claim shows why emotions have ends for the individual, rather than just natural functions, which describe the fitness of some trait for an ancestral group of organisms in their original environment.

The purposes accomplished by emotion are self-interested. Solomon says that emotions “maximiz[e] self-esteem …. [and] personal dignity” (1993, pp. 129, 122). In short, the emotion is explained by the fact that the truth of its appraisal would increase the person’s self-esteem and prevent the lowering of his self-esteem. Self-esteem includes evaluative judgments about one’s positive qualities, but also self-regarding attitudes such as pride, shame, embarrassment, etc. On the other hand, Sartre thinks that emotions circumvent “difficulty” in which one’s desires are frustrated (1948, p. 53). The truth of the emotion’s appraisal would alleviate the difficulty. Neither Solomon’s nor Sartre’s claim entails that the emotion is necessarily about the person who has the emotion. Such an implication would be false, as grief and sympathy illustrate. The unfaithful husband could regard himself more favorably if his wife shared some responsibility for the failure of their marriage. So, while his anger is not, strictly speaking, about himself – it is about his wife and what she has allegedly done – it nevertheless enables him to avoid embarrassment, guilt, or in some other way to serve himself.

Sartre and Solomon each maintain that the emotion’s purpose is achieved by altering the person’s own attitudes, including beliefs, instead of bringing about real changes in the world.⁸ Solomon says “the change that is brought about may be strictly subjective … [and involve] considerable restructuring of one’s views and values”(1977, p. 36). Sartre concurs, saying the emotion “seeks by itself to confer upon the object, and without modifying it in its actual structure, another quality, a lesser existence, or a lesser presence.” (1948, pp. 60–61). This “lesser existence” is merely psychological: “[the emotion] is accompanied by belief. The qualities conferred on objects are taken as true qualities” (1948, p. 73). Consider the phenomenon of sour grapes. George is rejected for a desirable job. Instead of admitting that bad luck or other more qualified candidates prevented him from getting the job, George convinces himself that the job was not desirable in the first place. To facilitate this conviction, George feels happy and relieved. George’s happiness and relief changes how he estimates the value of the job, and hence, his self-esteem is preserved. At no point, however, has any objective feature of the job been altered. Only George’s attitudes have changed.

For this reason, the purposes served by emotions are non-cognitive. Once it has been admitted that emotions facilitate belief revision, and do so for the sake of self-esteem or to circumvent difficulty, it has already been admitted that emotions are governed by a principle that reliably conflicts with truth. That is, emotions are not about the evaluative properties their objects actually possess. Instead, they are about evaluative properties the instantiation or absence of which would increase the person’s self-esteem or circumvent difficulty. In consequence, they can often disregard or distort the evaluative properties things actually possess. The husband’s anger may obscure how nurturing and supportive his wife has been. George’s happiness may obscure what a great job he has missed out on. In these cases, emotion enables one to disregard what is true for the sake of believing what is congenial. Emotions thus facilitate self-deception. Solomon fittingly described such self-deceptions as “the internal politics of emotion” (2003b, pp. 154–155).

While neither Sartre nor Solomon discuss attention, non-cognitive purposes show how the capture and consumption of attention might be at work here. The initial, indiscriminate appraisal in the capture phase, as well as the motivated search for reasons during the consumption phase, would be selective in ways that further the purpose in question. The sorts of evidence the person notices or fails to notice, and the relative salience of this evidence, will non-accidentally tend to support claims whose truth would increase his self-esteem or circumvent difficulty. By the same token, evidence supporting unfavorable claims will be “silenced,” i.e., the person will not notice or consider them further.⁹ For example, while angry, the husband notices his wife’s failings, and overlooks the ways in which she has been conciliatory or kind. So, he may be perfectly sincere in attempting to ascertain the truth. Nevertheless, the data upon which he works are highly biased, like an expert investigator who only sees clues that favor his own hypothesis.

However, there are at least two problems with the claim that emotional attention serves non-cognitive purposes. The first is that it rests upon an implausible analogy between emotions and actions. Emotions are not, like actions, under the person’s direct control, in the sense that one could alter them just by trying or willing to do so (see Adams 1985, p. 7; Gordon 1986, p. 373). Emotional attention in particular is not under the person’s direct control. Tappolet and Faucher argue that there is a difference between attention under the control of the will and attention under the control of stimuli (2002, p. 111). Given current psychological and neurological evidence, they hypothesize that emotions fall in the latter category, saying “as long as we are in the emotional state, we won’t have total voluntary control of shifting and maintaining attention” (2002, p. 128). That is to say, it is difficult, often exceedingly so, to attend to other objects during the course of an emotion. Therefore, it does not make sense to ascribe a purpose to what a person notices or overlooks during an emotion.

The second problem is that it remains unclear whether these non-cognitive purposes are essential features of emotion or features that only accompany emotion in irrational or self-deceived people. Some argue that non-cognitive purposes do not apply to all emotions. Elster accuses Solomon of a “reckless generalization from a few selected cases” (1999, p. 310). Wollheim calls these emotions “malformed,” and denies that “all emotions are formed in this deviant way” (1999, p. 84). These quotes suggest how Brady and other opponents of the received view might explain cases of emotion that inhibit evaluative understanding. They could maintain that the capture and consumption of attention only serves non-cognitive purposes in the presence of some interfering, but non-essential, psychological factor. What the person notices or overlooks would only be selective when influenced by, e.g., irrational desires of the sort found in self-deception. Thus, emotions that serve non-cognitive purposes would be deviations from an otherwise positive epistemic role. When they are not deviant, emotions could make the sort of salutary contribution Brady claims.

4 Emotional Dispositions

My claim in this paper is that the capture and consumption of attention in emotion has an inherently negative epistemic role, even though it can be consulted by an epistemically skilled and suitably motivated individual to understand why objects have various evaluative properties. Sartre and Solomon’s theory of non-cognitive purposes supports this claim by showing how the patterns of attention in emotion are selective. However, this explanation is problematic, since emotional attention is not the sort of thing that can have purposes, and moreover, it is not evident why non-cognitive purposes would be essential rather than contingent to emotion. Due to these problems, I conclude that Sartre and Solomon’s non-cognitive purposes are unworkable as stated. In the remainder of the paper, however, I offer a less problematic way to conceive of a related sort of teleological (‘in-order-to’) explanation of emotion by appealing to the commonly recognized idea of emotional disposition.

Emotional dispositions are mental dispositions whose manifestations consist in, among other things, emotions.¹⁰ Ordinary language contains terms for emotional dispositions, such as lachrymose, vainglorious, envious, etc., but many emotional dispositions lack distinct names because they are too idiosyncratic or localized. For example, somebody might be disposed to guilt over how she treats Sandy and Terence at work or to fear while performing music at the local bar. Even when the dispositions are idiosyncratic or localized, however, it is still possible to make informative, predictive generalizations about which emotions the person will have and when. As Deonna puts it, “people do react emotionally in some pattern-like ways because … [they] are just humans, or because … they are the specific humans they are” (2006, p. 36). Of course, the evidence for such generalizations may be limited, and thus the ascription of an emotional disposition is a tentative empirical hypothesis subject to revision in light of counter-evidence. There nevertheless exist patterns in the person’s emotional responses which such hypotheses capture correctly or incorrectly. An emotional disposition is what underlies each pattern.

Emotional dispositions, like other mental dispositions, are affected by what else happens in the mind. Wollheim says that “most mental dispositions that are not innate originate in mental states…However mental states, as well as generating new dispositions, reinforce existing dispositions” (1984, p. 58). In other words, one’s experiences, perceptions, thoughts, and emotions can establish or modify an emotional disposition. A single significant or traumatizing event is often sufficient (Rorty 1980). For example, the terror of one poor performance on a mathematics exam can establish a disposition to fear exams, mathematics, or school that persists long afterwards. Whether the disposition was originally innate or established through some significant or traumatic event, it has also been modified by habituation. Habituation refers to the mechanism whereby a mental state can be reliably elicited in certain conditions because that same state had been repeatedly accompanied by pleasure or relief from pain, discomfort, tension, or anxiety in relevantly similar conditions.¹¹ Habituation alters the relative likelihood that a disposition will be manifested, and the sort of manifestation it will have.¹² For example, the innate fear of heights has been weakened in a veteran sky-diver by repeated, tolerable exposure to jumping out of airplanes. Alternatively, somebody raised in a punitive society may be habituated to be angry in exactly those conditions where he would have otherwise felt shame or sympathy.

Habituation shows how emotions can occur for the sake of self-interested ends. When an emotional disposition has been habituated, the emotions manifesting that disposition have, during some phase of the person’s past, been repeatedly accompanied by pleasure or relief from pain, discomfort, tension, or anxiety. In consequence, the person now has a tendency to manifest those emotions in conditions relevantly similar to those which in the past brought pleasure and relief. For example, Cynthia had a competitive relationship with siblings, and when this competition was at issue, she felt envy. Her past reactions of envy, let us suppose, relieved the anxiety of being disfavored by her parents and her own sense of inadequacy. Now Cynthia is disposed to envy her competitors whenever she is uncomfortable or anxious about being passed over or viewed negatively. I claim that Cynthia is now envious in order to avoid these results of disfavor and inadequacy. After all, past episodes where she avoided these results through envy explain why Cynthia now has the determinate disposition to envy she does. It cannot be coincidence that she continued to feel envy in relevantly similar conditions, or that envy in those conditions continued to have similar results. If these results are not coincidental, then Cynthia must be envious for their sake.¹³ There is no third alternative.

Such an explanation preserves the spirit – if not quite the letter – of Sartre and Solomon’s theory. It differs in two notable respects however. Firstly, it holds that the ends of the emotion consist in pleasure and relief rather than self-esteem.¹⁴ This difference does not seem problematic, since occasionally taking pleasure in one’s good qualities seems a plausible prerequisite of self-esteem. In fact, this explanation constitutes an improvement insofar as it extends to creatures who lack the capacities for reflection necessary to self-esteem. Secondly, the present explanation is teleological, i.e., it refers to ends, without requiring that emotions are, or comprise, actions. The implausibility of such a requirement was the first problem with non-cognitive purposes. However, if the emotion’s ends are grounded in habituation, the relevant teleology is more appropriately termed “functional,” where function neither means purpose nor natural function nor the causal role some state has within a complex system. Rather emotions have acquired functions, because the ends accrue and are subject to change during the life-span of the individual.

Acquired functions apply especially to the capture and consumption of attention in emotion. Rorty claims that emotional dispositions are frequently “magnetized,” which means that the disposition manifests itself in ways that tend to produce some characteristic outcome (1980, p. 107). For example, irascibility characteristically leads to frustration and fights. Similarly, vainglory leads to pride and boastfulness or preening. Accordingly, Rorty’s idea of magnetization is an endorsement of teleological explanations for emotions. However, she claims that habits of selective attention and interpretation are specifically responsible for producing the characteristic outcome. Irascibility leads to frustration, because it entails “looking for frustrating conditions, perceiving situations as frustrating” (1980, p. 107). Thus, the magnetization of an emotional disposition consists in the fact that one will, everything else equal, notice features of the environment that are relevant to the characteristic outcome and interpret them in ways conducive to that outcome. Furthermore, the patterns of attention and interpretation that render a disposition magnetized are explained by events in the person’s past, including past experiences of pleasures and relief (1980, p. 106). That is to say, when a disposition is magnetized, one’s emotional attention has been habituated to produce the characteristic outcome.

In this way, emotional attention can acquire a non-cognitive function, wherein the person disregards or distorts the sort of evaluative properties objects have. The reason is that the capture and consumption of attention entails, as Brady notes, an “inclination to believe” (2009, p. 425). During the consumption phase, the person engages in a search for reasons in favor of the emotion. He will be “on the look out” for indications that support the initial action-oriented appraisal (2009, p. 425). This search will create a “kind of pressure” in which the person is “tempted” to accept the appraisal (2009, p. 421). Of course, such pressure does not necessarily result in belief, because the person might clearly recognize evidence to the contrary. However, it remains that the attention found in frustration will often cause the person to believe that the object is vexing, dismaying, etc., and similarly the attention found in fear will often cause the person to believe that the object is dangerous. If this causal relationship between attention and belief holds generally, emotional dispositions can be habituated to manifest themselves in patterns of attention that cause pleasant or relieving beliefs. Insofar as emotional attention is essential to the production of these comforting or pleasant beliefs, it has thus acquired a non-cognitive function. These beliefs are not necessarily false, but are governed by a principle that reliably conflicts with the truth.

Here is a brief, but unfortunately common, illustration: dangerous drivers are often disposed to anger when confronted about their driving. Their emotional disposition to anger has acquired a non-cognitive function through habituation. Anger, when the driver has been accused or confronted in the past, tended to produce the belief that the accuser is wrong and that he, the driver, is right. In particular, these circumstances were such that the driver tended to notice the faults and wrong-doings of others, and to overlook his own. As a result, he felt some relief. That is to say, he has not been habituated to be angry when, and only when, he has truly been wronged. Rather, he has been habituated to be angry in order to ignore or disregard his own wrong-doing.

5 Why Non-Cognitive Function is Essential to Emotion

Accordingly, I have argued that certain teleological explanations of emotion are defensible because emotions manifest dispositions which have been habituated. So, what a person attends to, or overlooks, during an emotion can acquire the function of producing pleasure or relief. This often means attending in ways which cause pleasant or relieving beliefs. Accordingly, it is possible to account for non-cognitive ends of emotion akin to what Sartre and Solomon suggested without endorsing their implausible assumptions about emotion and action. However, the outstanding problem is to show why non-cognitive function is essential to emotion rather than being a contingent feature that only accompanies emotion in the irrational and self-deceived.

Critics may argue that habituation, though it establishes the mere possibility that emotions have non-cognitive functions, is far from establishing that non-cognitive function belongs to the emotion inherently. After all, these critics would argue, emotions can exemplify virtues as well, including epistemic virtues. For our purposes, define an epistemically virtuous emotion as one that reliably produces evaluative understanding. As Aristotle says, it is virtuous to have emotions “at the right times, about the right things, toward the right people, for the right end, and in the right way” (1106b21–24).¹⁵ Crucially, these virtuous dispositions are established via habituation, no less than dispositions serving non-cognitive functions. So, the mere fact that emotional dispositions are habituated does not seem on its own to provide any reason to think that emotions obstruct evaluative understanding. In short, habituation is neutral with respect to evaluative understanding, because it can equally well promote or obstruct.

Such critiques ignore the fact that habituation presupposes a cumulative process of psychological development. For humans, that developmental process begins in infancy and occurs throughout the next few decades. Once this fact of development has been admitted, it can no longer be reasonably maintained that the habituation of emotions is neutral with respect to evaluative understanding. Consider the following three claims about human development and emotions:

Firstly, for humans, the capacity for emotion precedes the capacity for propositional reasoning and thought. As Deigh argues, there are primitive emotions, “the liability to which is… an inherited trait whose development, to the extent that it depends on the existence of environmental conditions, depends only on those necessary for meeting biological needs” (2004, p. 10; see also Deigh 1994, 2010). Primitive emotions are exemplified by pre-linguistic infants and non-human animals. It is unreasonable to deny that infants or higher animals can be angry or joyful, although they may not be capable of anger or joy in all the same situations or towards the same objects as adult humans. Primitive emotions, Deigh argues, show that propositional thought and rationality involving propositional thoughts are not necessary to emotion. These rational capacities only arise with the learning of language (2010, p. 27). However, these rational capacities are presumably necessary for evaluative understanding, because evaluative understanding consists in conceptualizing why something has the value properties it does.

Secondly, certain types of emotions derive their identity conditions from events that occurred in the pre-rational period of psychological development. Thus, even in adults, emotions of that type retain primitive features. Gardner offers a psychoanalytic theory of envy. On that theory, envy originates in phantasies, had during infancy, of attacking and destroying the food source (1992, p. 42).¹⁶ These destructive reactions were initially responses to the infant’s perceived sense of vulnerability and dependence upon the food source, and thereby, on the caretaker. As the individual develops, however, the same reaction is repeated with different sorts of objects, all of which are perceived to be external goods upon which she is dependent. The subsequent occasions when the individual has these destructive phantasies and what happened on those occasions explain her disposition, as an adult, to envy jobs, romantic partners, or prestige, etc. So, therefore, envy, as an emotion-type, bears a constitutive relation to mental events during infancy (1992, p. 45).¹⁷ That is to say, what makes some emotion envy, rather than anger or jealousy, is its arising from a primitive destructive impulse towards external goods. Similar developmental accounts could be offered for hatred, gratitude, sadness, guilt, anger, pride, and other emotions. Gardner claims that this type of explanation likely applies to “the class of emotion-kinds that are necessary for any recognizably human psychological constitution” (1992 , p. 48).

Lastly, the achievement of a virtuous emotional disposition is not unassailable or permanent. There is always the standing possibility of regression, that is, “a failure or loss of mastery and maturation and a return to what is developmentally earlier and more primitive” (Stocker and Hegeman 2000, p. 140). Regression is triggered by conditions of stress, trauma, depression, or injury. For example, suppose Jane spent early childhood fearing social situations that involved more than a few people. However, with help, she was able to master these fearful reactions, partly out of a recognition of the good reasons that favor attending (some) larger gatherings. Suppose Jane loses an important relationship. In consequence, her disposition to fear reverts back to its earlier, childish form, and she begins to fear large groups again. Examples like this illustrate how even virtuous dispositions can manifest the sort of emotions that belonged to the conditions of the disposition’s formation. Regressions show, as Burnyeat says, “stages in the development of [the person’s] character he has not completely left behind” (1980, p. 85).

These three claims together establish that every emotional disposition, including those that are epistemically virtuous, is inherently prone to non-cognitive function. Given the possibility of regression, which was the third claim above, every emotional disposition can manifest the sort of emotions that characterized earlier periods of the person’s psychological development. Furthermore, non-cognitive functioning was the guiding principle during these earlier periods. That is to say, whether one was envious, hateful, or angry, etc., was governed entirely or primarily by the pleasure and relief that preceded and followed the emotion, and not by whether the emotion accurately appraised its object. This premise is supported by the first two claims above, namely, that the capacity for emotion predates capacities for discursive reasoning, and that certain types of emotions by their nature belong to those pre-rational periods of psychological development.

Overall, then, the fact that emotional dispositions are habituated provides reason to think that emotional attention has an inherently negative epistemic role. According to the foregoing argument, any emotional disposition that has been habituated either has a stable non-cognitive function or is prone to non-cognitive function in conditions of anxiety or stress. But all emotional dispositions, including innate ones such as the tendency to be disgusted at excrement or to fear steep drops, are habituated, since they can be trained. Therefore, non-cognitive function is the default condition of the emotion, and not some contingent property found only in the irrational or self-deceived. In other words, the fact of psychological development shows that emotions begin with and essentially retain some non-cognitive orientation.¹⁸

Of course, it is reasonable to admit that emotions can be habituated to epistemic virtue, and in these conditions, the capture and consumption of attention can help the person to determine the evaluative properties objects really have. However, any success here must be due significantly to non-emotional skills that counteract the emotion’s inherent non-cognitive tendencies. Karen Jones describes the skill of “reflective self-monitoring,” wherein one ascertains the cognitive reliability of one’s own dispositions, and discounts or re-calibrates them when those dispositions mislead (2003, p. 194). Peter Goldie similarly refers to “monitoring and correcting” one’s emotions (2008, p. 159). If reflective self-monitoring is necessary for an emotion to be epistemically virtuous, then the consequent evaluative understanding would be due to a complex psychological state: emotional attention together with the non-emotional epistemic skill of separating what is credible in that emotion from what misleads.

One critical task in separating what is credible in an emotion from what misleads is to recognize regressions in oneself. In particular, one would have to recognize when one’s patterns of attention have manifested more childish or immature epistemic tendencies. While I cannot confidently say what the signs of such a lapse are, because they may be both subtle and idiosyncratic, it would (if the argument so far has been sound) not be possible to reliably have an emotion “at the right times, about the right things, toward the right people, for the right end, and in the right way” without having somehow already practiced and honed the ability to make these recognitions in oneself.

6 Conclusion

During an emotion, attention shifts to some new object and persists upon it for some time. This pattern of attention manifests dispositions which have been habituated and often predate our capacities for propositional thought. So, by default, what one attends to, or overlooks, in an emotion, will be explained by what in the past conduced to pleasure and relief, rather than the evaluative properties objects actually have. Admittedly, these patterns in attention could, for an epistemically skilled and suitably motivated agent, help yield evaluative understanding. But the credit for such success properly goes to a conjunction of psychological states, including certain epistemic skills, such as reflective self-monitoring, and special motivational states, such as the desire to understand. Without these extrinsic factors, and sometimes even despite them, emotional attention reverts back to non-cognitive function. In this sense, the received wisdom that emotions cloud or obscure judgment is correct.