Emotions, Character, and Associationist Psychology

In: Journal of Moral Philosophy
View More View Less
  • 1
  • | 2

Emotions are pivotal in the manifestation and functioning of character traits. Traits such as virtues and vices involve emotions in diverse but connected ways. Some virtues (justice, generosity, compassion, truthfulness) are exemplified, in important part, by feeling emotions. Others (self-control, perseverance, courage) are exemplified in managing, bypassing, or even eliminating emotions. And one virtue at least (humility) is exemplified in not-feeling a certain range of emotions. Emotions are a kind of perceptual state, namely construal, involving concern or caring (motivation) about something, in which the elements of a situation are organized and understood in terms of their significance or import. Emotional understanding can be morally right or wrong. As such construals, emotions can be morally excellent (the feeling of joy about a rectified injustice) or perverse (envy and contempt of persons). Emotions thus have a logic or grammar that is crucial to their entering into, or being set upon by, or simply not occurring because of, virtues. The virtuous person is attuned, implicitly or reflectively, to this grammar, and that attunement constitutes one of the major dimensions of practical wisdom. An associationist psychology (behaviorist or Humean) attempts to reduce the conceptual and intentional richness of emotions to mere associations or correlations of pleasant or unpleasant “affect” with various things (behaviors, “values”). Such a psychology is fundamentally unfit to represent practical wisdom, and thus the moral life. We sketch an account of the generation and degeneration of character traits using the above conceptual framework and contrasting it with an associationist framework.


Emotions are pivotal in the manifestation and functioning of character traits. Traits such as virtues and vices involve emotions in diverse but connected ways. Some virtues (justice, generosity, compassion, truthfulness) are exemplified, in important part, by feeling emotions. Others (self-control, perseverance, courage) are exemplified in managing, bypassing, or even eliminating emotions. And one virtue at least (humility) is exemplified in not-feeling a certain range of emotions. Emotions are a kind of perceptual state, namely construal, involving concern or caring (motivation) about something, in which the elements of a situation are organized and understood in terms of their significance or import. Emotional understanding can be morally right or wrong. As such construals, emotions can be morally excellent (the feeling of joy about a rectified injustice) or perverse (envy and contempt of persons). Emotions thus have a logic or grammar that is crucial to their entering into, or being set upon by, or simply not occurring because of, virtues. The virtuous person is attuned, implicitly or reflectively, to this grammar, and that attunement constitutes one of the major dimensions of practical wisdom. An associationist psychology (behaviorist or Humean) attempts to reduce the conceptual and intentional richness of emotions to mere associations or correlations of pleasant or unpleasant “affect” with various things (behaviors, “values”). Such a psychology is fundamentally unfit to represent practical wisdom, and thus the moral life. We sketch an account of the generation and degeneration of character traits using the above conceptual framework and contrasting it with an associationist framework.

1 Introduction

This paper’s purpose is twofold: to propose a broad outline of the relations among character traits and emotions, and to consider what kind of psychology is best suited to make sense of the moral life so outlined, with special attention to a virtue that arguably pervades a well-developed moral character, namely, practical wisdom (moral understanding). Like Aristotle and other classical thinkers, we think a viable moral psychology must make sense of the fact that some of the central virtues have objects. The virtue of justice is about justice in people and situations; compassion is about the suffering of creatures and its alleviation; truthfulness is about owning and communicating truth. To make our point that an adequate moral psychology must do justice to this “intentionality”1 of virtues and emotions, we’ll consider a recently published paper that attempts the same or a similar job using a modern associationist psychology after the manner of David Hume or, rather differently, behaviorism. We’ll argue that such a psychology cannot do justice to the intentionality of the moral life.

We begin in section ii by sketching a schema of virtues and vices that reveals the variety of ways emotions (or the lack thereof) manifest character traits. In section iii, we explain the important connection between emotions and moral understanding that underlies the various interconnections between emotions and character. Then, in section iv, we evaluate Charles Starkey’s recent attempt to employ an associationist psychology to explain the role of emotions in the formation and maintenance of character traits. While Starkey offers some valuable insights, he does so in spite of the associationism to which he appeals, not because of it, and because his associationism is not thoroughgoing. We conclude in section v by drawing on our cognitivist account of emotions to outline a more plausible alternative framework for understanding the role of emotions in the generation and degeneration of character traits.

2 Sketch of a Schema of Virtues and Vices

Character traits are stable and reliable dispositions and powers of persons. The class of character traits includes virtues and vices, traits that make their possessors good or bad, well or poorly developed, specimens of humankind. In this paper we will focus mostly on virtues. We distinguish four kinds of virtues, all of which involve or relate to emotions in one way or another. The ways of this involving or relating are diverse. The classes of virtues that we distinguish are 1) substantive and motivational virtues; 2) the virtues of willpower; 3) humility; and 4) practical wisdom.

Substantive and Motivational Virtues

As their name suggests, the substantive and motivational virtues are the heart of moral character. We call virtues like justice, generosity, compassion, and truthfulness “substantive and motivational” because they embody the content of morality, and do so by supplying the distinctly moral kind of motivation. Assuming that the core moral goods are justice, truth, and human and sentient wellbeing, stable concerns for these will be the core virtues and will supply core moral motivation. In their own ways, the second and third types of virtues revolve around such substantive virtues. And the fourth type, practical wisdom, is largely an aspect of the substantive virtues. Central to the substantive and motivational virtues are their defining concerns. A concern is a kind of love, caring, or heartfelt sensitivity to, the importance of something. It is more personal than merely valuing something, if to value something is just to assess it as having a certain positive value. If you are concerned about something, it matters to you. To just people, for example, just states of affairs, just relationships, just actions, and just people matter; they are important to them; they care about them and are interested in promoting them. Concerns, in the present sense, are dispositional, but they give rise in particular circumstances to active or occurrent concerns. These we call desires. For example, in face of an unjust action or state of affairs, a just person, one who cares about justice, will be averse to the injustice and will often feel moved to rectify or oppose it; he will want that injustice to be removed. This is the sense in which the virtue of justice, in addition to being morally substantive, is also motivational: it is in large part a readiness to be moved in appropriate circumstances by considerations of justice.

Justice, then, is defined in part2 by the concern for objective justice.3 Generosity is defined by a non-instrumental concern for other people’s enjoyment, convenience, and advantage. Compassion is a non-instrumental concern for the relief of suffering and harmful defect in fellow humans and other animals. Generosity and compassion can be grouped under one heading, and called benevolence, a concern for the other’s wellbeing. Truthfulness is defined by the concern for truth – for example, the concern to know the truth and the concern that others are not deceived, especially by oneself. In general, the substantive and motivational virtues are centered in concerns for good things and in opposition to things that are bad.4

Concerns are the basis of emotions. We feel anger, gratitude, fear, hope, joy, sadness, and so forth, about things we care about, depending on our perception of the bearing of the circumstances or situation on what we care about. A person who cares about justice will feel anxious or afraid if he sees a significant potential for injustice as in the offing. He will rejoice when injustices are rectified, and feel guilty when he perpetrates an injustice. He will feel hopeful if he sees good prospects for an injustice to be rectified. It is the mark of the substantive and motivational virtues that they are exemplified, not only in actions, but also in emotions of a wide range of types. One way, then, that emotions are related to character traits is this: emotions can manifest or exemplify the substantive and motivational virtues. Emotions based on one of these substantive virtues will have as an orienting object what that virtue is a concern about. Emotions based on the concern for justice, for example, will all be about justice as the particular circumstances giving rise to the emotions impinge on justice.

Virtues of Willpower

Emotions based on other concerns can also be a challenge to moral and other proper functioning. Fear, boredom, anger, envy, jealousy, sadness, and even joy and hope, can stand in opposition to our acting justly, generously, compassionately, and truthfully. In such circumstances we need an ability to transcend the emotion, to check its expression in action, and to mitigate or even expel it.5 This will be one of the functions of such virtues as self-control, courage, patience, and perseverance. These virtues have, thus, a very different relation to emotions than the substantive and motivational virtues. They are not expressed in emotion, but are directed at emotions (among other mental items: for example, urges, appetites, and habits). Unlike the substantive virtues, they are not concerns, but are combinations of powers, abilities, and skills. The actions in which they are expressed have no particular characteristic motive (compare justice or generosity), but can be motivated in a variety of ways, both virtuous and vicious and neither. You can exemplify courage in the pursuit of justice, but you can also show it in saving your classic guitar from a house fire. Insofar as they belong to good moral character, the virtues of willpower constitute a kind of repertoire of auxiliaries to the substantive and motivational virtues.6 The emotions that the virtues of willpower control will be based on different concerns than the substantive and motivational virtues that the virtues of willpower serve. For example, the fear that courage overcomes in defending a colleague from a false accusation will be based, most likely, on a concern for one’s own welfare, while the desire to defend the colleague arises from the concern basic to truthfulness or justice or friendship, or a combination of these. The courageous and just person is able to manage his fear in a way that serves justice.


One kind of concern that is particularly pernicious in human affairs is the concern for what we might call “self-importance.” This is different from taking one’s “self” seriously, as in moral self-examination or self-discipline. The concern for self-importance is the concern that is the heart of such vices as envy, conceit, snobbishness, arrogance, vanity, domination, pretentiousness, self-righteousness, and grandiosity. It is the non-instrumental concern for glory, prestige, standing, and superiority (to others). Like other concerns, the concern for self-importance can be the basis for emotions: anger when somebody diminishes your glory; joy when you are in the limelight and the object of much admiration; anxiety about how well you’re going to come across where your self-importance is at stake; hope for various kinds of prestige; contempt for people who are “below” you; envy of people who have succeeded better than you at garnering glory. That these emotions are based on a bad concern is a very large part of what makes them vicious, just as the excellence of the emotions based in the substantive virtues is largely due to their being so based.

Emotions of vicious pride can be the targets of acts of the virtues of willpower. For example, you can try to control your envy or your snobbery or vanity, and perhaps you can find strategies for lessening your concern for self-importance, and thus weaken all of the “vices of pride” that we have mentioned. In that case, the virtues of willpower will have helped you to acquire the virtue of humility. Ideally, humility is the state of having no concern at all for self-importance (which is not the same as thinking one’s self has no importance). If you have no concern for self-importance, you will have none of the emotions that are based on that concern. So in our view, humility is a virtue with a different relation to emotions than either the substantive and motivational virtues or the virtues of willpower. Unlike the substantive virtues, humility is not exemplified in any emotion or range thereof; and the person who is mature in humility doesn’t deploy the virtues of willpower to correct for the vices of pride. He or she simply lacks the emotions characteristic of those vices, through lacking the concern on which they are based. But she doesn’t merely lack them (she might achieve that through lobotomy), but lacks them in a personality that is otherwise normal – or even better, excellent.7

Practical Wisdom

The moral life is a life of intelligence (not intellectualism) in action, self-assessment, and assessment of others’ actions and characters, with implicit or explicit reference to some adequate conception of a fulfilled, mature, and meaningful human life. The person of practical wisdom understands the moral life, and particularly his or her own moral life. To understand this is to have a fairly clear conception of what makes life good (especially what should be the object of cares and concerns) and what threatens to undermine the goodness of a life (for example, the vices of pride, or unconcern about justice). The substantive virtues, as concerns about this or that dimension of the good, are necessary to practical wisdom. Each of those virtues can be seen as an ability to appreciate a dimension of the good, and they are collectively essential to the understanding of the moral life. Unless and until a person is appropriately concerned for some aspect of the good, she does not fully (appreciatively) understand it. The concerns for the good at the heart of the substantive virtues contribute this necessary evaluative orientation to the moral understanding constitutive of practical wisdom. Practical wisdom is not just deft calculation, but also a sense of the importance of what is important. Practical wisdom, in turn, contributes to each of the substantive virtues the moral understanding of that aspect of the good for human beings that concerns each of these virtues. A love for justice would be blind, disoriented, or stupid to the extent that it was not qualified by an understanding of justice.8 In relation to actions, practical wisdom is manifested in both deliberation and perception.

Deliberation is the activity of weighing alternative courses of action, given the circumstances, and concluding which of the alternatives it is right or best to take. The perception characteristic of practical wisdom, by contrast, is simply seeing that such-and-such is the thing to do. Someone who is mature in practical wisdom will have both of these capacities, but perception functions also in the course of deliberation. The conclusion of a deliberation is ideally perception, an insightful seeing what to do (and why), and perception is also needed to see what the circumstances of the possible action are, and to evaluate each of the deliberated alternatives. Damasio9 has shown that a capacity for something like emotion is necessary for practical deliberation, and it functions, as Charles Starkey10 has also noted, in the perceptual dimension of practical reasoning.

Since the substantive virtues are centrally concerns for what is good, each of them constitutes an aspect of the concern for the good life, the concern that guides practical wisdom and infuses it with motivation. And since emotions are based on concerns while registering the perceived impingement of the circumstances on the objects of those concerns, the emotions expressive of the substantive virtues constitute the moral perceptions of the practically wise person. For example, the compassionate person sees the badness of others’ suffering clearly and accurately (wisely) through her emotion of compassion. Her intelligent concern for others’ wellbeing also enables her to deliberate well (wisely) about the best prospects for remedying their suffering. As she deliberates, she perceives with dissatisfaction potential short-term remedies that might present an obstacle to long-term wellbeing and she perceives with hopefulness the prospects for more promising remedies. The compassionate person may also see possible remedies in the light of justice, feeling discomfort with remedies that may do injustice to the sufferer or the sufferer’s associates or others. Here we see a kind of “unity” or “reciprocity” of the virtues.

We have said that the virtues of willpower and humility are auxiliaries of the substantive and motivational virtues, but we cannot say the same of practical wisdom. Indeed, the virtues of willpower and humility are also auxiliaries of practical wisdom insofar as the virtues of willpower are sources of strength to overcome obstacles in acting wisely and humility is the absence of a “blinding” obstacle – the concern for self-importance – which, when present, impedes and obscures moral vision (i.e., practically wise perception). That the virtues of willpower and humility are auxiliaries of practical wisdom as well as of the substantive and motivational virtues is just what we should expect since, as we suggested above, practical wisdom is an aspect of the substantive and motivational virtues. It is in a way a uniting of the substantive virtues, or perhaps the epistemic aspect of those virtues, because each of the virtues – justice, generosity, compassion, and truthfulness are our examples – contains its own powers of judgment, in the form of deliberation and (largely emotional) perception. The just person must understand justice and so be able to recognize cases of justice and injustice with a fairly high degree of accuracy, and, to be able to act justly, must also have some knowledge of how to implement justice where it is breached or threatened, and to maintain it where it prevails. And one of the things she must know is how to temper justice with compassion and generosity. Thus the wisdom of each of the substantive virtues is in part an understanding of how each virtue goes with the others. We might say that each of the virtues includes an understanding of its status as neighbor of the other virtues. Each has a partial perspective on itself, so to speak, by understanding other virtues that interact with it. And practical wisdom itself is a conjunction and coordination of the wisdoms of the particular virtues.

We have said that the substantive and motivational virtues (justice, generosity, compassion, truthfulness) are exemplified, in important part, by feeling emotions. Others (self-control, perseverance, courage) are exemplified in managing, bypassing, or even eliminating emotions. And one virtue at least (humility) is exemplified in not-feeling a certain range of emotions. We have suggested briefly that the emotions that manifest the substantive and motivational virtues constitute the moral perceptions of the practically wise person and contribute to practically wise deliberation. To clarify further these connections between emotions and the virtues, we must say something more about the nature of emotions themselves.

3 What an Emotion Is and the Phenomenon of Moral Understanding

Having identified a variety of ways in which emotion-dispositions are related to character traits, we now turn to the question of precisely what it is about the nature of emotions that explains these interconnections with character traits. How are they perceptions of moral import in situations, and how can they be expressions of the concerns that are constitutive of the substantive and motivational virtues?

We say that emotions are concern-based construals.11 Construal is a crucial feature of perception. It is the organization of elements into a sense-making whole impression. Paradigm cases are the visual impression of the old lady / young girl gestalt figure as a picture of a young girl, or the auditory impression of a word-string like fish fish fish fish fish as a sentence equivalent to “fish that are fished by other fish themselves fish for yet other fish.” In each case data (the spatially arranged lines of the drawing or the sequence of sounds) is “constructed” or “grasped” into an integrated intelligible whole. It is made sense of. Thus perception itself is a kind of understanding.

Emotion types such as fear, hope, anger, joy, envy, gratitude, contempt, and jealousy are types of construal, each with its distinguishing conceptual structure. To be angry, for example, is to construe someone as a culpable agent of offense; to be afraid is to construe something as a plausible threat to something of value; to feel envy is to construe someone as a succeeding rival in the contest for one’s self-importance; to be jealous is to construe someone as a rival for the exclusive affection of a beloved; and so forth for all the emotion types.12 This is part of the story about emotions; another part is about the concern on which the construal is based. All of the above illustrations can be non-emotional construals if the concern is missing. In fear, you have to care about what you construe as threatened; in anger, you have to care about what is offended against; in envy you have to care about your self-importance. Emotions, then, are conceptually shaped construals of situations where the construal integrates not just the features of the situation, but also a concern on which that integration of features impinges.

We have said that the substantive and motivational virtues are the core of the moral life, that each of them is a concern or caring or love for some aspect of the good and aversion to the corresponding aspect of the bad, that each of the virtues is a kind of intelligence, and that the wisdoms of these virtues taken together constitute practical wisdom. We’ve said that practical wisdom is a kind of broadly moral understanding, and understanding, in general, is an acquired capacity to make, see, and appreciate connections among things – causes and effects, the thematic relatedness of a musical development, the place of a situation in a novel in the overall story, the parts of a gestalt figure, the words of a sentence, and the elements of a moral situation. We’ve also said that each of the substantive and motivational virtues can be expressed or exemplified in a wide range of emotion types, depending on how the situation of the object of concern is construed. We can now see how crucial emotions are to the virtuous person’s understanding of the world and his or her life. Without the concern-based construals, the virtuous person would be without practical wisdom, and so also without virtues. Understanding is crucial to the moral life, and the emotions are crucial to moral understanding. Emotions arising from the substantive virtues are modes and vehicles of moral understanding.

Our account of the relations between virtues and emotions has stressed three features of emotions:

  1. (a)their dual intentionality,
  2. (b)their “logical” or “grammatical” systematicity (interconnectedness), and
  3. (c)their conceptual structure.

All of these features are essential to the way they interact with virtues and vices. As to (a) their dual intentionality: as situational concern-based construals (Roberts 2003, 2013), virtuous emotions integratively target:

  1. 1)the object of the concern (say, justice, truth, or the wellbeing of another person) and
  2. 2)the situation in question (say, that Judith has lyingly convinced the boss that Hal has been harassing her and ought to be fired).

The concern-based construal integrates the value-object with the situational object – in our example, by caring about the situation as a case of injustice or untruth. One such form would be anger: Judith is construed as culpably bringing about an injustice by lying and trying to get Hal fired.

As to (b) emotions’ logical systematicity, notice how other possible emotions may arise as the situation develops or is differentially seen as further elements are brought into view: another emotion might be hope that the boss, as fair-minded, will listen to Sally’s testimony that Judith’s accusation is false. Another might be relief or joy that Judith’s lie has been exposed and Hal has been saved from an injustice. Or, if the boss won’t listen to testimony, frustration and sadness. These emotions are systematically or logically related to one another, in virtue of their clustering around a common concern and their status as different construals of the situation in which the object of that concern (justice) is at stake.

As to (c) their conceptual structure: this systematicity depends on the emotion of each of these types having a conceptual structure of its own that makes it fit to pick out possible features of a situation that impinge on justice: anger about offense, hope about good prospects, relief about good (or lack of bad) that has arisen out of difficulty, joy about the occurrence of a good, frustration about the continuing prevention of a good, and sadness about the loss of a good.

In the just person, this range of emotions expresses her appreciative understanding of the goodness of justice and the badness of injustice. Moreover, as direct perceptual experiences (sense-making construals), her emotions enhance (and, in a way, constitute) her understanding of the various ways in which the complex features of this particular situation bear on justice.

So far, we have seen that emotions relate to virtues in a variety of ways and that these relationships are explained, in large part, by the fact that emotions, as concern-based construals, are modes and vehicles of the moral understanding at the heart of the substantive and motivational virtues. As we will show in the following sections, associationist psychology neglects this epistemic role of emotions and is thus inadequate as a framework for understanding the place of emotions in character formation.

In a recent paper on our theme, Charles Starkey argues that emotion and character are connected, in large part, by the ways emotions affect and enhance perception.13 For a person with a character trait to be able to act or make a judgment from that trait, says Starkey, he or she needs to perceive relevant features of the situation about which he is to judge, or in which he is supposed to act, including their moral import. And because of their influence on perception, emotions facilitate this functioning of character traits. How do emotions affect perception, according to Starkey?

First, emotions enhance our attention. If we are very angry about an injustice, we will probably ruminate on it, keeping it before our minds, to the exclusion of other things that might be going on around us, and we may be especially sensitive to the moral difference between the offender and those on whom he has perpetrated the injustice (194; the illustration is ours). Second, if you’re already angry, you’ll be more likely to notice offenses and “someone who is emotionally aroused by the distress of another person will be more likely to spot someone in distress than someone who is not emotionally aroused” (Starkey, p. 195). Third, when we are emotionally engaged with something, we are more likely to investigate it, in the course of which we see more aspects of it.14 Our fear of an injustice in the offing may make us enter more deeply into the features of the situation and thus come to understand it better (195–6; again, the illustrations are ours). Fourth, our emotions allow us to see what is important to us. When Anna Karenina’s heart leaps with joy upon seeing Vronsky at the train station, she realizes that she is in love with him (our illustration). And “our discomfort at seeing someone else mistreated can provide us with information about our attitude toward fair treatment” (Starkey, 196). Here, obviously, the perception is self-perception, not, as in the other cases, situation-perception.

Starkey points out that these perceptual enhancements are relevant to character traits. He imagines a situation in which a person has an opportunity to help others that may function at the same time as bringing him some kind of advancement relative to others (197). So he may perceive the situation either as an opportunity to help someone in need, or as an opportunity for personal advancement, and Starkey says that the first perception goes with one character trait (say, generosity) while the other goes with a different one (say, competitiveness). He speaks in rather general terms about the person being “aroused” by the opportunity for advancement or the opportunity to help, but we might specify that he feels glad expectation or hope – either to gain personal advancement or to help someone (of course, his motives might be mixed). What is it about emotions that affects perception, and about character traits that such perception is relevant to them?

Starkey cites various empirical studies in support of the features of emotional perception, but our account of the nature of emotions and the motivational virtues explains his observations. Emotions increase our attention to a situation because an emotion is a perception of the situation in terms that highlight a way that some of its features touch one of our concerns. We would add that people are more likely to notice things that belong to categories of things they care about. If you care about color coordination in the decoration of rooms, you will be more likely to perceive when a room is well or poorly decorated, and you’ll be more likely to notice people in distress if you are a compassionate person (care aversively about human suffering). This fact obviously implies that people with the substantive and motivational virtues will be sensitized to situations that involve vicissitudes of the kind of good that they care about (justice, the wellbeing of persons). Emotion about one situation makes us more likely to notice similar features in another situation because, as a construal, the emotion readies our attention conceptually for the features of the second situation. Emotions sometimes make us more likely to investigate the situations they are about (Starkey’s third point) because emotions are based on our caring about something to which the situation as we see it has relevance, and caring is a kind of interest. And lastly, our own emotions can tell us something about what we care about if we properly understand our emotions because emotions are concern-based construals. Anna perceives, and thus understands, her (emotional) perception of Vronsky as excitingly good, as indicating her love for him. Most adults have enough intuitive understanding of how emotions work to realize that when we’re very glad to see somebody, apart from any ulterior motive, we must care about him or her.

All of this explanatory work depends on emotions’ double intentionality, systematicity, and conceptual structure. These features explain, too, how emotions express character traits like justice, compassion, and generosity, and play an important role in the perceptual function of practical wisdom. To follow up Starkey’s example, a person with the trait of personal ambition will feel anticipatory joy on perceiving an opportunity to advance himself relative to others, because he perceives the situation as potentially satisfying his concern for personal advancement. A person with the trait of generosity might see the very same situation as an opportunity to help someone, and her anticipatory joy is based on her concern for others and is a perception of that opportunity as a good. If we think that her perception is wiser than his, it will be because we suppose that generosity is more virtuous than personal ambition.

4 Virtues, Emotions, and Associationist Psychology

Starkey’s own explanation of these phenomena makes heavy use of the notion of association by conditioning, reminiscently of a Humean or behaviorist psychology, yet also shows that he appreciates the necessary role of intentionality in moral psychology. He points out that:

Someone’s discomfort at the prospect of dodging the truth in a conversation can provide information about the cost of doing so (in terms of compromising the person’s values), and in this way provide immediate information about it even before he or she explicitly thinks about the consequences. Emotional responses can work in the other direction as well: the discomfort can be paired with the unpleasant consequences of being truthful. In this latter case the emotion could increase the likelihood of behaving dishonestly (197).

The contrast Starkey highlights here shows how crucial it is for the identification of the trait to know what the discomfort is about. Is the discomfort the noble one that shows the love of truth by execrating one’s own temptation to dodge it? Or is it the far less noble one, based on the concern to avoid whatever unpleasantness or inconvenience to oneself telling the truth might bring on? The capacity to discern the difference between such concerns, as indicated in one’s emotional reaction, is essential to moral self-understanding. This wisdom will be completely lacking if all you know is that you feel uncomfortable when you think about dodging or telling the truth.

Let’s consider a somewhat different case. You feel discomfort when you are inclined to dodge the truth. You want to make sense of this discomfort. You know that you feel it whenever you are inclined to dodge the truth, but you don’t know whether this emotion is one to be morally commended and to cultivate, or to be morally eschewed and eradicated if possible. Now if all you know about the discomfort is that it’s associated with (or paired with – see the second sentence of the quotation) having the option of telling or dodging the truth, you have no answer to the question how it – and you – are to be evaluated. What you need to know is why this prospect brings on this discomfort. And to know that, you need to know what concern of yours this prospect is touching: is it touching a concern that the truth be told, or a concern not to be caught dodging the truth? In other words, what is the discomfort ultimately about? Of the emotion’s dual intentionality, one aspect is clear enough, perhaps:15 the discomfort is about the choice of possibly dodging the truth; but the other aspect is not clear, because you don’t know what you’re really concerned about – being truthful, or not being caught dodging the truth.

If the relation between the discomfort and the possibility of dodging the truth is only that they’re “paired” with one another, then this wisdom about your virtue or lack thereof will not be possible. Moral understanding is stymied. But the discomfort isn’t just “paired with” the differential objects; it’s about them, and that makes for the deep moral difference between the two possible discomforts. Only if that is understood will we understand how the emotion is related to character. Here we see clearly one of the weaknesses of an associationist psychology as a moral psychology of character. It is true that the emotional response is “associated” with one character trait or another, but the crucial question is about the logical character of that association. “Association” is a vague, non-committal term in ordinary parlance, but in the hands of the Humeans and behaviorists it’s ideological, and means “mere association” or merely habitual pairing.

Starkey says that both classical and operant conditioning can help explain how character traits are acquired and maintained. As an example of classical conditioning, he reminds us of the case of Albert,16 who met a white rat at age nine months, and wasn’t afraid of it. Then repeatedly when the rat was presented, a nasty sound that scared Albert was paired with the presentation. After that, Albert was “afraid of the rat.” (The nasty sound is the unconditioned stimulus, and Albert’s fear of it is the unconditioned response. The repeated pairing of the nasty sound with the presentation of the white rat turns the rat into a conditioned stimulus, and Albert’s consequent “fear of the rat” is the conditioned response. This is how classical conditioning works: it starts with a natural or unconditioned stimulus-response pair, repeatedly associates (co-presents) a different stimulus with the natural one until the subject habitually associates the two, upon which the response to the natural stimulus has become a response to the added new one.) Thus Albert acquired an emotion disposition, fear of the rat. Starkey thinks that such classical conditioning can partially explain why some people respond with pleasant emotions to “values” (and, presumably, with unpleasant emotions to “disvalues”). They have acquired a (conditioned) positive emotion disposition to the presentation of a “value” by its having been paired with an unconditioned positive stimulus, and presumably a negative emotion disposition to the presentation of disvalue by its having been regularly paired with an unconditioned aversive stimulus.17

Another kind of conditioning, operant conditioning, can also explain our acquisition and retention of character traits, according to Starkey. Operant conditioning begins with a behavior (an “operation”) that an organism already emits, and then shapes or strengthens it through positive response (reward), or reduces or weakens it through negative response (punishment). Starkey thinks that such conditioning can reinforce not only a person’s actions, but also his motives.

Praising someone’s telling the truth for the sake of being honest can reinforce that intention, as can the positive feelings within that person that can result from their thinking of their intentions as good intentions that are approved by others (201).

Here we see Starkey’s heterodoxy (or inconsistency) with regard to associationism. Notice how he individuates the behavior that is to be conditioned. It is not just telling the truth, but telling-the-truth-for-the-sake-of-being-honest,18 that is the action the praise of which can reinforce the character trait of honesty. In other words, it is not just behavior and not just action, but action defined by a particular intentionality.

Notice, however, that the motive by which the action is conditioned (rewarded) in this example, namely, the concern to be praised (or, perhaps less plausibly, the concern to act out of intentions that are approved by others), is not the motive that the reward is supposed to reinforce, namely, the concern for truth, or the concern to be honest. So there is an explanatory gulf between the conditioning and what is supposed to result from it. Operant conditioning, therefore, cannot be an adequate explanation of the virtuous disposition to act. At most it can lead to telling-the-truth-for-the-sake-of-being-praised, though on strict associationist terms it should only lead to a habitual non-intentional association of truth telling with being praised. Even if praise is effective in leading eventually to a genuinely moral disposition, as indeed it seems to be, operant conditioning seems not to be the right explanation of how it works.

In our opinion, the right story has to be something like the following: the child, in being praised for her honest behavior, enjoys the praise, no doubt, and wants to continue to please. But the praise is an embedded part of the child’s induction into a moral community in which, through moral discourse and through her I-thou communion of minds with morally admirable adults and a concomitant innate potential for transcending the mere concern for praise, she comes to appreciate such abstract objects as truth and justice, and such things as the wellbeing of others, because she sees their human and spiritual import. For example, as an immature member of the community, who tells the truth at first from sub-moral motives, she comes to appreciate some of the excellences of truth telling: the trust that is created, the promoted wellbeing of all concerned, the beauty of truth itself. And so she grows in wisdom. Skinner19 and his ilk didn’t have the problem of the explanatory gap, because they really were only interested in behavior. Starkey, by contrast, wants to account for full-fledged action and, more than that, action that exemplifies virtue and is thus motivated by distinctively moral reasons. So he seems to be in a bind created by his commitment to the intentionality of virtue combined with his commitment to associationist psychological explanation.

If operant conditioning fails to explain the acquisition of the substantive and motivational virtues, classical conditioning is, if possible, even less plausible. Classically conditioned emotions fail to exemplify the moral understanding at the heart of virtuous character. Recall the case of Albert. It isn’t clear what the intentional object of Albert’s conditioned fear is. Is he afraid of the rat? He acquired an aversion not only to the rat, but to other white things. Perhaps he is afraid of whiteness. He’s only nine months old, but if he were more mature, we might think that he’s still afraid of the sound, of which the appearance of the rat and other white things is indicative; that would make some sense. He thinks to himself, as it were, “whenever I see a white rat, there’s almost sure to be that sound that so scares me.” Whereas fearing the rat makes no sense unless Albert sees something threatening in the rat. But if that’s the case, then the conditioning is not the right story about how Albert got the fear (maybe he has learned that the rat might be rabid). Pure classical conditioning doesn’t seem capable of giving the fear an intentional object. But without fear’s defining intentional object (something that threatens what I care about), “fear” doesn’t rise to the status of fear proper. At most it’s a kind of non-rational aversion (perhaps with a physiology and behavior similar to fear).

When I see my fuel gauge on empty in a remote place and react with fear, I’m clearly not afraid of the fuel gauge; I’m afraid of what the fuel gauge indicates (I’m running out of gas), or maybe a consequence of what the fuel gauge indicates (that if I run out of gas I may be stranded in a remote place). This is what makes my fear reasonable (or even wise). If I had just been classically conditioned to feel fear when the fuel gauge reads empty, my “fear” wouldn’t be part of a rational scheme of life. And that seems to be true of Albert. I do indeed “associate” the fuel gauge’s being on empty with running out of gas and being stranded out in the country; but the association is not merely association, as in classical conditioning. Instead, it’s an “association” that reflects understanding. It’s a sense-making association. When I see the fuel gauge in that position, I’m afraid because I know what it means.

This is the truth that motivates Starkey’s insistence on including the virtuous motive in his description of the action that he thinks operant conditioning can reinforce, contributing to the acquisition of the character trait of truthfulness. Telling the truth for the sake of being honest is meant to contrast with telling the truth to retain one’s customers, to avoid being caught in a lie, and so forth. These are all possible intentionalities of the action, and so, on our account, of the agent’s concern. But only telling the truth from a concern that the truth be told will count as expressing the trait of truthfulness.

5 Emotions’ Role in the Acquisition and Extinction of Character Traits

Starkey’s thesis that emotions’ role relative to character traits is to enhance the kind of perception that character traits need for judgment and choice includes the thesis that emotions contribute sensitivity to the importance of “values.” This makes sense inasmuch as moral perception without sensitivity to the importance of, say, justice in a situation would hardly be adequate moral perception. This aspect of Starkey’s thesis seems to him to imply that emotions are capable of being powers of resistance to “axiological entropy,” the gradual erosion of our sense of the importance of “values.”

In the ongoing absence of relevant emotions, a value such as truth would cease to be important to a person, and would thus be unable to drive the person’s perceptions, choices, and actions (199).

His account of this resistance appeals to the mechanisms of conditioning and association.

An emotional reaction to a perceived value can produce the pairing of emotion and value in a way that reflects classical conditioning. We may see someone behaving callously to another person, which arouses indignance on our part directed at the callous act, reinforces within us the sense of the significance of callousness, and sustains the aversion to callous acts (203–4).

If white rats and other white things are repeatedly not paired with the aversive stimulus in Albert’s experience, his conditioning by their earlier pairing will gradually fade and he will no longer respond aversively to white rats and other white things. And similarly, thinks Starkey, if a value like justice is for a long time not paired in a subject’s experience with emotions that enhance the perceptions required for the trait of justice, that subject’s perception of the importance of justice will diminish, and therewith his trait of justice. Even art and literature function in this associationist way: “…art and literature are valuable because of their ability to induce emotions and pair them with values” (204, italics added).

As anticipated in our earlier discussion, we think that this associationist explanation is inadequate because the emotions that a person has in consequence of her trait of justice are not merely associated with the “value” of justice, but take justice as their doubly intentional object by way of their conceptual structures, which relate those emotions to one another in logically systematic ways. Furthermore, the very capacity to be angry about an injustice depends on the subject’s concern about justice. If she didn’t already (dispositionally) love or care about justice, the emotion of anger about injustice wouldn’t arise to get “paired” with the value of justice. The notions of association, pairing, and conditioning are inadequate to the task of explaining the relationship of emotions to the character traits that they express and instantiate. Starkey himself testifies to this truth when he talks about our indignation being “directed at the callous act.” That indignation is not merely “paired with” or “associated with” the (dis)value of callousness.

Our criticism of the associationist explanation leaves untouched Starkey’s thesis that emotions have a role in sustaining character traits and resisting axiological entropy. We would call such an erosion a diminishment of concern for goods such as the wellbeing of others, justice, and truth – the goods the concern for which is at the core of the substantive and motivational virtues. So put in our terms, Starkey’s thesis is that emotions can help to sustain moral concerns. Thus, for example (appealing again to our thesis of emotions’ systematic clustering around a concern), your frequent feelings of anger, gratitude, hope, despair, joy, sorrow, relief, and frustration about justice will tend to keep up your interest in (enthusiasm, concern for) it, and protect you from becoming jaded about it. We will close with some further reflections about how emotions might serve in the formation and sustaining of character traits.

In general, it seems that we often sustain our concerns and interests by actualizing them. A person who finds her enthusiasm for justice waning might decide to act in favor of justice (perhaps she is motivated only by a sense of duty, either to do the justice or to recover her sense of justice). But she finds that in the course of action, she has enlivened her enthusiasm for the object of action, so that, though she was not intrinsically motivated when she undertook the action, she became so in the course of acting.

As we have explained, emotions are concern-based construals. This thesis implies that emotions are instantiations of the concern – specified versions of it. Just as, whenever a person acts from a love for justice, she is putting her concern for justice to work (so to speak), so whenever a person feels an emotion based on the concern for justice, she is actualizing that concern – putting it to work. Emotions can be thought of as concern-filters. The different emotion types filter the concern in different ways. The concern goes through the filter, changes as it goes through, and emerges on the other side, the same concern but now specific to the situation as construed. Roberts calls the concern that emerges the “consequent concern.”20 The virtue of justice is a concern for justice. This concern goes through anger about injustice and emerges as a desire to punish the offender or otherwise rectify the injustice, but the same concern goes through gratitude for another’s contribution to the cause of justice and emerges as a desire to reward the favor (by some token). So, just as it seems plausible that whenever we exercise any of our capacities we are sustaining them, it is plausible that in feeling emotions based on the concern for justice we are sustaining our concern for justice.

Another way that emotions seem to support character traits depends on the systematicity that we have just mentioned. We say that the emotions expressive of the virtue of justice are systematically related perceptions of situations that bear on justice in various ways: anger about justice offended against is the concerned perception of justice as offended against, gratitude about justice aided is the concerned perception of justice as graciously abetted by someone, remorse about an injustice is a perception of oneself as a perpetrator of it, and so on. Because of the systematic relations among emotion types clustering around a single concern as determined by each emotion’s complementary conceptual structure, it is not enough to be disposed to any single emotion type with respect to the good.

In George Eliot’s novel The Mill on the Floss, Tom Tulliver (the 13-year-old brother of the heroine) is very concerned about justice, but the range of his emotion repertoire about it is extremely limited. He has an impressive capacity for anger about injustices done to himself and his family, but seems to have little capacity for feeling hope for their rectification, or joy in their avoidance, or guilt about his own injustices. He also doesn’t have much sensitivity to injustices done to others than himself and his family; so even his anger disposition falls well short of satisfying the requirements of virtue. Thus, though he’s concerned about justice in a way, he doesn’t have the virtue of justice. If a person who at one time had the virtue of justice finds himself narrowing down and becoming obsessed with and angry about injustices done to himself, but insensitive to issues of justice in other lights than the light of anger, he will be suffering from something akin to axiological entropy. And the cure for him will be somehow to broaden and deepen his emotional vision by becoming more attuned to injustices of which he is not the victim and by seeing justice from other emotional angles – that is, in terms of the conceptual structures of a wider range of emotion types.

Further, it’s possible that the “moral” emotions that are elicited by a morally deep novel like those of George Eliot may awaken moral concerns that were dozing in the reader and seldom, if ever, elicited by real-life situations. Such awakening is creditable to the literary and rhetorical powers of the writer, and her ability to involve the reader with the characters in her story. An analogous power to awaken sleepy concerns by eliciting the emotions that depend on them is an excellence of first-rate preaching. The phenomenon of awakening dormant moral emotional sensitivities seems to be in tension with a more extreme (associationist) thesis of axiological extinction by dearth of emotional activation. If the lack of emotional activation predicts that a concern will be completely extinguished, would we not expect it to be impossible to elicit those emotions?

So we think that Starkey is right to say that emotions have a role in sustaining virtues. Here we would speak more specifically about the substantive and motivational virtues and practical wisdom. A somewhat different story probably needs to be told about the role of emotions in sustaining the virtues of willpower and humility. For example, self-reflexive emotions such as guilt, regret, or disappointment or anger toward oneself for failing to act rightly (say, in a case of weakness of will) might motivate one to engage in practices aimed at strengthening one’s willpower. An emotion of confidence (perhaps informed by the sense that one is being aided in the moral struggle by a friend, a recovery group sponsor, or God) might be added to one’s willpower, giving one the strength to overcome temptation, and this practice of self-overcoming might strengthen one’s willpower for the future. In the case of humility, regular experiences of compassion felt toward suffering others or gratitude in response to the blessings in one’s life (construed gratefully as unearned and undeserved) might indirectly enhance humility by further training one’s focus on the needs of others and by ridding one of a sense of entitlement, and the agent might learn to exploit such emotions to such purpose. We think that Starkey’s attraction to associationist explanation abridges what he might have to say about the intentionality of emotions and these virtues. It discourages him from seeing the relevant emotions as instantiations of the virtues and the sustaining power of active emotional exemplification of the virtues. It also discourages him from seeing the systematic conceptual relations among the emotion types and the role that those relations can have in the sustaining or recovery of the virtues.

6 Conclusion

We have shown that at the center of the moral life is a set of character traits whose core is concerns for such objects as truth, justice, and human wellbeing. In the course of a human life these concerns get filtered through (construed in terms of) a diversity of such conceptual schemata as threat (fear, anxiety), good prospects (hope), fulfillment (joy), loss (grief, sadness), culpable fault (anger, indignation, resentment, guilt, remorse), defect (shame), and suffering (compassion), depending on the perceived vicissitudes of the objects of concern. Each of these emotion types is a type of perception and of understanding, a way of making sense of situations as they bear in their different ways on the object of concern. Emotions thus have intentionality, a conceptual structure, and cluster in a conceptually orderly way around some concern. The emotions that express the substantive and motivational virtues and other, analogous, character traits cluster around the dispositional concerns that are at the cores of those traits, actualizing them. The intentionality, conceptual structure, and systematic interconnection of the moral emotions make them crucial for practical wisdom, and make an associationist psychology (Humean or behaviorist) deeply inadequate in regard to the understanding of emotions and their relations to character traits.

The emotions that are characteristic of the substantive and motivational virtues are also important, as are actions characteristic of those virtues, as sustainers of those traits from which they spring. They perform this sustaining function by instantiating the concerns, by their character as forms of insight, and by being loci of engagement with the social world that elicit sustaining (and sometimes discouraging) responses from that world. The concerns at the core of the motivational virtues also beget reflexive concerns, such as the concern to be concerned about justice, and emotions that are based on them, such as regret that one no longer cares for justice as one formerly did, and hope to recover that concern. Such concerns are also the motives for acts of willpower and self-management that correct for deficits in the motivational virtues and contribute to the formation and sustaining of those virtues.21

Robert C. Roberts taught philosophy at Western Kenctucky University, Wheaton College, and Baylor University. He now holds the Chair of Ethics and Emotion Theory in the Jubilee Centre at the University of Birmingham (uk).

Adam C. Pelser is associate professor of philosophy at the United States Air Force Academy in Colorado.


See Pierre Jacob, ‘Intentionality,’ Stanford Encyclopedia of Philosophy, http://plato.stanford.edu/entries/intentionality/.


Another important part is the practical understanding of justice, the refined ability to distinguish justice from injustice in particular cases, and to discern appropriate correction of injustices. See the discussion of practical wisdom below.


For considerably more elaboration, see R. C. Roberts, ‘Justice as an Emotion Disposition’, in John Deigh (ed.), On Emotions: Philosophical Essays (Oxford: Oxford University Press, 2013), pp. 14–28.


On the distinction of virtues that are substantive and motivational, on the one hand, from the virtues of will power, see R. C. Roberts, ‘Will Power and the Virtues’, Philosophical Review 93 (1984), pp. 227–247. On the nature of concerns, see R. C. Roberts, Emotions: An Essay in Aid of Moral Psychology (Cambridge: Cambridge University Press, 2003), pp. 141–151. On generosity, see R. C. Roberts and W. Jay Wood, Intellectual Virtues: An Essay in Regulative Epistemology (Oxford, Oxford University Press, 2007), pp. 286–304. On compassion, see R. C. Roberts, ‘Compassion as an Emotion and as a Virtue’, in Ingolf Dalferth and Andreas Hunziker (eds.), Mitleid (Religion in Philosophy and Theology Vol. 28: Tübingen: Mohr Siebeck, 2007), 119–137.


When the Stoics said that emotions are false judgments that tend to motivate bad actions, they must have had in mind emotions other than the ones that are based in the motivational virtues. See Seneca, On Anger, in John M. Cooper and J. F. Procopé (eds.), Moral and Political Essays, Cambridge University Press, 1995.


See Roberts, ‘Will Power and the Virtues’.


Earlier versions of this view of humility can be found in chapter 9 of Roberts and Wood, Intellectual Virtues, pp. 236–256; R. C. Roberts, ‘The Vice of Pride’, Faith and Philosophy 26 (2009), pp. 119–133; and R. C. Roberts, ‘Learning Intellectual Humility’, in Jason Baehr (ed.), Educating for Intellectual Virtues (London: Routledge, 2015), pp. 184–201.


In Aristotle’s discussion of choice he flags this reciprocity and integration between what we’re calling the substantive virtues and the capacity of thought that he calls practical wisdom. “Since moral virtue is a state of character concerned with choice, and choice is deliberate desire, therefore both the reasoning must be true and the desire right, if the choice is to be good, and the latter must pursue just what the former asserts. … Hence choice is either desiderative reason (orektikos nous) or ratiocinative desire (orexis dianoêtikê), and such an origin of action is a man” (Nicomachean Ethics 6.2.2, 1139a23–26; 6.2.5, 1139b5–6, translated by David Ross and revised and edited by J. L. Ackrill and J. O. Urmson, Oxford University Press, 1980, p. 139).


Antonio Damasio, Descartes’ Error (New York: Harper Collins, 1995). But we have doubts about the “somatic marker” thesis by which he explains the perception. For discussion, see Roberts, Emotions, pp. 155–7.


Charles Starkey, ‘Emotion and Character,’ in Christian B. Miller, R. Michael Furr, Angela Knobel, and William Fleeson (eds.), Character: New Directions from Philosophy, Psychology, and Theology (Oxford: Oxford University Press, 2015), pp. 192–211.


This view is elaborated in R. C. Roberts, Emotions in the Moral Life (Cambridge: Cambridge University Press, 2013), especially chapters 3–5. See also Roberts, Emotions, and R. C. Roberts, ‘What an Emotion Is: A Sketch’, Philosophical Review 97 (1988), pp. 183–209.


These are rough approximate characterizations of the emotion types; please find somewhat more adequate ones in chapter 3 of Roberts, Emotions .


Starkey, ‘Emotion and Character.’


While this may be generally true, fear can also sometimes discourage us from investigating what we dread – e.g. that our child is on drugs or that we have cancer.


We say “perhaps,” because we’re assuming that the discomfort is about the choice, but strictly speaking, assuming even that is not allowed if the discomfort is connected to the situation only by association.


See J.B. Watson and R. Rayner, ‘Conditioned Emotional Reactions,’ Journal of Experimental Psychology 3 (1920), pp. 1–13.


See p. 200. Note the conceptually noncommittal congruence of the notions of condition, stimulus, and positive / negative. An intentionality-oriented psychology, by contrast, will use richer, more concrete, and more human-specific concepts like accustom, learn, concept, concern, love, desire, (intentional) object, tradition, person, I-thou relation, and, instead of positive / negative, all the concrete and mixed-valence emotion-words such as anger, fear, nostalgia, envy, jealousy, resentment, hope, etc.


We might quibble with the specific characterization of the motive. Honest people don’t tell the truth for the sake of being honest, but for the sake of conveying the truth. In other words, the action is not about the agent, but about the truth. It is true, though, that some people tell the truth for the sake of being honest, especially if they are not very honest and are trying to become so. They treat the action as an exercise in honesty.


See B. F. Skinner, Science and Human Behavior (New York: The Macmillan Company, 1953).


See Roberts, Emotions, p. 144.


Thanks to Christian Miller for his insightful and challenging editorial comments on a previous draft.This publication was made possible through the support of two grants from the Templeton Religion Trust, one via the Self, Motivation & Virtue Project, and one via the Beacon Project. The opinions expressed in it are those of the authors and do not necessarily reflect the views of Templeton Religion Trust. The views expressed in this article also do not necessarily reflect the official policies or positions of the us Air Force, the us Department of Defense, or the us government.