Introduction: How to Do Things with Emotions
1. According to the June 2020 Pew Research Center Report, 71 percent of Americans felt angry when asked to think about the state of the nation. Two-thirds felt fearful. Anger and fear seed each other, creating a potentially self-sustaining loop.
2. What about hope? The year 2020 gave us the COVID-19 pandemic and renewed, widespread #BlackLivesMatter activism across America in the aftermath of George Floyd’s murder by the Minneapolis police. #BlackLivesMatter anger is principled, entirely justified, and, finally, a multiracial movement. It is as if the Spanish flu pandemic of 1918, the Great Depression of 1929, and the social and cultural upheaval of 1967–70 occurred at the same time. The June 2020 Pew Research Center Report revealed that 46 percent of Americans were hopeful about the state of the country, with slightly more hope among blacks and Hispanics than among whites. That said, during 2020, there were disturbing developments alongside positive ones, as former president Trump fired up tribal political anger in citizen militias who love freedom, Second Amendment rights to bear arms, and law and order. Political leaders commonly model, express, and state norms for justified political anger and, what is different, for violence.
3. The June 2020 Pew Research Center Report, which tracked anger and fear about the state of the United States, also measured pride. These data are from June 16–22, 2020, thus well into the COVID-19 crisis and after the murder of George Floyd. Eighty-four percent of Americans were not proud. Although “not proud” does not equal “ashamed,” there is a significant relation. There are good reasons to be ashamed about the state of the nation: there is the shame of being led by a president who embodied and advanced white supremacy, the sycophancy of political representatives at every level, and the shame of the legacies of racism, sexism, and queerphobia and their current poisonous, sometimes intentional, forms.
4. R. M. Seyfarth and D. L. Cheney (2003) argue that vocal expressions in animals as various as frogs, birds, mongooses, and vervet monkeys are strategic and entirely under voluntary control. Like the human smile, which normally occurs only with an audience, so, too, horny frogs only make mating calls when a potential mate is nearby, and they strategically vocalize in ways designed to throw predatory bats off accurately locating them.
5. There are also sources in the phenomenological tradition that emphasize both the active side of emotions and their bodily components. It is this phenomenological lineage that yields the “4-e” view of mind: mental activity is “enactive,” “embodied,” “extended,” and “emotional/affective.” My confidence that emotions are doings, or should be conceived as having a significant actional aspect, is inspired by these pragmatist and phenomenological sources, as well as by the work of my dear friend and intellectual mate, Batja Mesquita, whose work in cultural psychology shows the power of this idea.
6. The schema models emotions as wide states. One could (many do) hypothesize that emotions have narrow profiles that can be characterized in terms of shared qualitative feeling or the qualitative feeling of an emotion + its typical somatic profile across instances. There are smart affective scientists on both sides of this debate. Lisa Feldman Barrett (2017) and Joseph LeDoux (2019) think that this narrow project is not progressive, because one doesn’t find facial or neural regularities across cases, whereas Paul Ekman (2007), Jaak Panksepp and Lucy Biven (2012), Kent Berridge (2018, 2019a, 2019b), and Stephen Asma and Rami Gabriel (2019) think there are such regularities that can be used to characterize emotions narrowly. There are philosophical views on the emotions that divide in terms of which aspect of my wide schema they privilege. There are perceptual theories, feeling theories, judgment theories, conative or motivation theories, and action theories. My view is that perception, feeling, cognition, and action each play a role in making emotional episodes what they are. Emotions are things we do or are for-the-sake-of doing-things, and doings are complex things that, in the case of emotions, require perceptions, feelings, judgments, dispositions to act, and actions. The emphasis on the whole episode might make it look as if my view is a conative, motivational, or action theory. It isn’t. My view starts with the idea that the psychobiological function of emotions is to have me do something—something to release the emotion in the case of negative emotions or something to sustain, enhance, or build the emotion in the case of positive emotions. When an element is missing, e.g., there is no robust somatic/phenomenal feeling or no carry-through, we do not need to say it is not an emotional episode, although we could; instead, we say it is not a paradigm case, atypical to some degree.
7. Moods—such as feeling bored or experiencing free-floating anxiety—often have no content; they are not about anything. Emotions always have an intentional object or content.
8. The motto “emotions are things we do” awaits refinement. There are many things we do that are not emotions or standard examples of emotions. We do means-end reasoning, we spell words, climb mountains, ride bikes, order merchandise on the internet, do long division, empty the trash, dig ditches, write speeches, and garden. One might think that in real life most of these doings are suffused with feelings—emotions, moods, various admixtures of pleasant and unpleasant experiences—and thus that specifying these acts as physical or cognitive doings absent emotions is an analytic abstraction, not the way the doing actually is or unfolds. I think this. But this raises the question of whether, say, writing a short speech consists of multiple acts: various compositional acts, hemmings and hawings, visions and revisions, positive and negative reactions to these acts from both the poses of self-as-author, self-as-audience, and self-as-critic, which then serve as evidence (by serving up cognitive-emotional affordances) on next steps in the editing process, culminating perhaps—after metaphorical blood, sweat, and tears—in the act of acceptance, a sense of satisfaction that the speech is good. I think so. Last point: I recommend a certain nominalism about the general kind of emotion. Let each linguistic community—including the community of expert emotions scientists—have authority to say what emotions there are, what features of emotions (from my wide schema) they use to individuate emotions, and let them explain, first pass, what psychological generalizations hold among emotions.
9. In addition to the sheer size of an immigrant group as a causal agent that could engender emotional adjustments in the dominant culture, an indigenous or exogenous power majority can also have similar effects. So elite, upper-class folks with wealth, or occupying armies and colonists who control markets, can demand for reasons of power—not for reasons of their normative quality—practices that accommodate, or even that are, their emotional norms and practices.
10. Perceptual theories of emotions (Tappolet 2016) make too much of the perceptual aspects of emotions and underestimate the other aspects, as do other kinds of theories (feeling theories, or judgment theories, conative or motivation theories), which also have the downside of overemphasizing one aspect of emotions.
11. Gallup (2019).
12. In fact, even “anger” in English isn’t a good translation of “anger.” This sounds paradoxical, but think about it. There are a million words for anger: “irritation,” “wrath,” “fury,” and “pissed off.” And then there is all the work that adjectives do: “intense anger,” “mild anger,” and so on.
13. In Women, Fire, and Dangerous Things (1990), George Lakoff points out that we can track common meanings inside languages, but we should not underestimate the importance of idiolects, or individual cognitive models that contain the idiosyncratic meaning rules that individuals can have for terms in a shared language.
14. There is an appreciative critique of Jackson et al. (2019) by Asifa Majid (Science 366 [6472]: 1444–45) that expresses the concern that the original study worked by holding the underlying emotion concepts as stable and universal (“as Platonic ideals”) when they might actually be highly variable. This is a legitimate concern, but if it is true that this study holds fixed some conceptual essences that mark different emotions, and there really are not any such conceptual essences, this would suggest that variation is even more robust, deeper, and widespread than the study shows, not less so. It would mean that there is cultural variability in the causes, effects, and feelings used to define even basic emotions.
15. Do speakers of Nakh-Daghestanian languages not experience grief and anxiety? We don’t know from the linguistic evidence, which indicates that they lack words for these emotions. There are many cultures with a limited color vocabulary who can nonetheless easily distinguish colors and shades as well as we do (Lakoff 1990).
16. These points about learning terms for private mental states are familiar from Ludwig Wittgenstein, Gilbert Ryle, W.V.O. Quine, and B. F. Skinner. In Verbal Behavior, Skinner explains: “There are at least four ways in which a reinforcing community with no access to a private stimulus may generate verbal behavior with respect to it: (1) A common public accompaniment of the private stimulus which eventually controls the response may be used.… One teaches a child to say That hurts … by making reinforcement contingent upon certain public accompaniments of painful stimuli (a smart blow, damage to tissue, and so on).… (2) A commoner practice is to use some collateral response to the private stimulus.… The community reinforces the response My tooth aches when it observes such collateral behavior as holding the hand to the jaw, executing certain facial expressions, or groaning in certain temporal patterns.… (3) A third possibility is that the community may not appeal to private stimuli at all; it may reinforce a response in connection with a public stimulus only to have the response transferred to a private event by virtue of common properties, as in metaphorical or metonymic extension. It has often been pointed out that most of the vocabulary of emotions is metaphorical in nature. When we describe internal states as ‘agitated,’ ‘depressed,’ or ‘ebullient,’ certain geometrical, temporal, and intensive properties have produced a metaphorical extension of the responses.… (4) When a response is descriptive of the speaker’s own behavior, there is a fourth possible way in which private stimuli may acquire control. The original contingency may be based on externally observable behavior of the organism, even though this stimulates the speaker and the community in different ways. If the behavior is now reduced in magnitude or scale, a point will be reached at which the private stimuli survive [imagine one suppresses the temptation to express anger (my example)] although the public stimuli vanish. In other words, behavior may be executed so weakly or so incompletely that it fails to be seen by another person, although it is still strong enough to stimulate the behaver himself” (1957, 131–33).
17. Batja Mesquita (2022) distinguishes between MINE and OURS conceptions of emotions. MINE emotions are conceived narrowly and as inside persons; whereas OURS emotions are conceived widely as functional syndromes involving typical causes + feelings + dispositions to act + act. She says OURS emotions are understood as occurring between people. WEIRD people conceive emotions in the MINE way. Mesquita thinks, as I do, that OURS better describes the metaphysics of emotions across WEIRD and non-WEIRD societies. It is just that WEIRD folk like to talk more about the personal inner aspects or properties of emotions than non-WEIRD folk.
18. Cause and content (what the emotion is about) can come apart. I might be annoyed or angry at Mary for asking me where the keys are while I am working. Mary asking me where the keys are is the content, that is, what I am angry about. But the cause or part of the cause might be that I didn’t eat breakfast.
19. In my experience, many people insist that there must be something very specific that makes an emotion the emotion it is, and that because the causes and effects differ so much, it must be the invariant features of the emotion itself that make getting angry at the cars in front of me and angry at my internet server, both cases of anger. The only candidate for sameness of anger, given that causes and content differ, is identity of inner phenomenal feeling. But this requirement, that some narrow invariant feature is required to specify a kind, is unwarranted. Compare to the kind “tennis stroke.” What is a tennis stroke? Where is the tennis stroke? A tennis stroke is a swing with a racket that hits (or aims) to hit a tennis ball within certain boundaries and over a net of a certain height (three feet at center; three feet six inches at posts). The trajectory, speed, and spin of the ball vary a lot coming and going. There are different kinds of strokes: serves, volleys, drop shots, lobs, etc. A good stroke involves certain positions of the entire body, which are responsive to the trajectory, speed, and spin of the ball. Body position, arm position, and core position vary across every stroke. The point is, we have a perfectly good idea of what a tennis stroke is, and where it is, without there being any aspect or dimension of a tennis stroke that is the same across strokes.
20. The contestations in the last kind of case involve complex matters of both ethics and social ontology: Is a person partly constituted by contingent facts about their lineage and thus responsible for actions of the people to whom she is connected or not?
21. Seneca, no pacifist, favored nonangry, efficient armies. Angry armies torture, rape, and pillage. Nonangry armies don’t. They just win necessary battles.
22. In a brilliant paper from 1956 simply titled “Emotions,” Errol Bedford insists both that emotion words are theoretical terms that do not name inner feelings, and that the principal functions of emotional expressions are “judicial not informative” (298). This might be an overstatement since often one says how one feels precisely to inform others how one feels. But Bedford is onto something. He notes that there is “overlap between the lists of emotions, and the lists of virtues and vices that are given by philosophers” (ibid., 294). The “judicial function” is thus one of “moral criticism,” whereby one admits or denies responsibility, assigns blame, makes excuses, or invites an audience to share in a positive or negative evaluation of another. Forty years after Bedford’s essay, William Reddy recommended that we think of statements about emotions as a special kind of speech act that he dubs “emotives” (1997). We might think that saying “I am so pissed off” is just a report of what emotional state we are in. But it is rarely such a simple descriptive report of my psychology (what speech act theorists call a “constative”). It has performative aspects akin to statements like “I baptize you” and “I accept that,” but it displays additional features that prototypical performatives do not have. In saying “I am so pissed off,” I invite my audience to sympathize with me and the object of my anger to fear me. Reddy writes, “Emotives are influenced by and alter what they ‘refer’ to. Thus, emotives are similar to performatives (and differ from constatives) in that emotions do things in the world. Emotives are themselves instruments for directly changing, building, hiding, intensifying emotions” (ibid., 331).
23. My view is not that the language of emotions, and norms and scripts for doing emotions, are required for feelings. A few emotion researchers seem to take that view (Barrett 2017; LeDoux 2019). My view is that we regulate emotional feelings and emotional expression by way of norms that, in the human case, are typically conveyed in the language of caretakers and are commonly enacted and reinforced in public symbolic forms in schools, religious instruction, and ritual performance.
Chapter One: Anger and Morals
1. Another disposition that I think has moved, or shows signs of moving, from one column to the other is something in the vicinity of greed and acquisitiveness. Greed and acquisitiveness are moving from the vice column to the virtue column. Greed is a deadly sin, but it can be rebranded in terms of entrepreneurial energy and a quest for achievement and creative innovation that can seem to be the epitome of virtue in individualistic, neoliberal quarters. Aristotle’s virtues of “great-souledness” and magnanimity created this particular opening for conceiving of wealth acquisition as a virtue, or as enabled by virtue.
2. The Dalai Lama writes that the Tibetan words zhe sdang can be translated as “anger” or “hatred” in English. He then says: “I feel that it should be translated as ‘hatred’ because ‘anger,’ as it is understood in English, can be positive in very special circumstances. These occur when anger is motivated by compassion or when it acts as an impetus or a catalyst for a positive action. In such rare circumstances anger can be positive whereas hatred can never be positive” (1999, 7).
3. Callard’s On Anger (2020) is quite brilliant. And I am one of her targets. She distinguishes between the realistic project of inhibiting and moderating anger and a fantastical project that claims that anger can be purified and has completely innocent forms. Callard defends grudges and revenge not because they are sometimes satisfying—they are—but for a much deeper reason. She thinks that morality is an accommodation to certain dark facts about human nature, including that anger norms are in fact a solution to our natural tendency to meet violence (which is guaranteed to occur) with vengeance, setting off never-ending vengeance. Anger is vengeance domesticated, still a bit bloodthirsty but contained enough so that it invites less vengeance than its undomesticated parent. Anger is thus less corrupt than vengeance but still corrupt, still mean and hurtful, and necessarily so because we are not good enough to do without it.
4. Aquinas, Summa Theologica, I–II.46.6 ad 2; 47.1–2, follows Aristotle, defining virtuous ira as involving the desire to make another suffer. But he explains this in terms of a just retribution. Furthermore, the aim is not simple revenge or pain-passing. There is hope for the future, including hope for the good of the other.
5. Pankaj Mishra calls ours the Age of Anger (2017), and he sees extreme and constant anger across the developed world, and across the political spectrum, on the left and the right. Bageant (2007), Rasmussen and Schoen (2010), Cramer (2016), and Hochschild (2016) are fine ethnographies about twenty-first-century anger and resentment in America. Traister (2018) and Chemaly (2018) are defenses of anger for the sake of social justice.
6. Calling someone “a scumbag” or saying “fuck you” are angry speech acts designed to insult and injure. Such cruel speech is a visible part of the situation that I think is problematic. Some profanity-studded speech might be judged not to be the Queen’s English and thus to be crude—“That was a fucking great cricket match”—but it is not remotely cruel or intended to hurt another.
7. Anger is often met with anger. But anger is not always triggered by anger. Fear is a common cause of anger, as are disloyalty, the experience of being disrespected, and being overburdened.
8. Amia Srinivasan (2016) defends political anger against Nussbaum’s critique. See Cherry and Flanagan (2019) for an edited collection of papers that defend several kinds of moral anger. Macalester Bell (2013) is a study of contempt, a relative of anger. Philip Fisher (2003) is a deep literary, philosophical, and psychological analysis and defense of anger and other vehement passions. See also Agnes Callard (2020), where anger, including grudging and revenge anger, are defended against Martha Nussbaum and me.
9. Thanks especially to Tal Brewer for helping me on this point.
10. Pettigrove (2012) is good on antianger sources in Hume and Butler; Nussbaum (1994) is good on Stoic sources. Judith Shklar (1982) thinks that post-Enlightenment, we put cruelty first as the worst political vice. If Shklar is right, then cruel anger should be something that, at a minimum, we ought to feel ambivalent about.
11. There is excellent work on consciousness (Godfrey-Smith 2016; Tye 2017) that argues for robust phenomenal consciousness among honeybees, crabs, fish, birds, and octopuses. As for the emotions, fear seems common among fish; pleasure, anger, and grief seem common among birds.
12. There is the “epiphenomenalist suspicion” about anger and other basic emotions (Flanagan 1992). The epiphenomenalist says that the complex emotion we call “anger” is some sort of evolutionary side effect, epiphenomenon, spandrel, or exaptation of some core behavioral circuit. According to the epiphenomenalist, the emotion evolved secondarily, as a freebie, on top of the behavioral circuit that does the lion’s share of fitness-enhancing work, and it serves no function. The feeling of anger at most registers to the organism that it has detected a threat or an obstacle, and that a survival circuit has been activated. The feeling might seem to motivate the angry response. It doesn’t. The psychological version of the epiphenomenalist suspicion has been tested empirically for anger and other basic emotions (Panksepp and Biven 2012), and it is false. Eliminate the phenomenal feeling of anger in an organism and you don’t see the angry, fitness-protecting response.
13. My friend and PhD student Robert Bingle has convinced me that in fact Seneca is only interested in extirpating ira, which is a unique type of out-of-control anger. See Bingle (forthcoming). If this is right, then “extirpating anger root and branch” is not intended even by the Stoics to mean extirpating all varieties of anger.
14. David P. Barash and Judith Eve Lipton (2011) are good on the evolution, biology, and psychology of both payback and pain-passing anger. Like Carol Tavris (1989), who roundly criticizes the ventilationist view of anger, Barash and Lipton caution against the idea that it is good to release anger cathartically, as in anger management training where one takes out anger on a punching bag or a pillow. The evidence is that this increases rather than decreases the amount of anger that allegedly needs venting.
15. In The Therapy of Desire, Martha Nussbaum tells a story she heard secondhand about Elie Wiesel that can be read as an example of impersonal rage at the heavens. There is no particular person or group of persons that the rage aims to hurt or pay back: “Wiesel was a child in one of the Nazi death camps. On the day the Allied forces arrived, the first member of the liberating army he saw was a very large black officer. Walking into the camp and seeing what there was to be seen, this man began to curse, shouting at the top of his voice. As Wiesel watched, he went on shouting and cursing for a very long time. And the child Wiesel thought, watching him, now humanity has come back. Now, with that anger, humanity has come back” (1994, 403). This case is sometimes used in arguments to the effect that anger is necessary to accomplish certain moral effects or that it is the best way to accomplish such effects. See Becker (1998), Sherman (2005), and McRae (2015). I think that we can easily imagine the officer in the Wiesel case becoming overwhelmed by sorrow rather than anger, and a contagion of tears spreading among his fellow soldiers. This might have had the same humanizing effect on young Wiesel as the angry episode, as well as the further advantage of not scaring the other kids, who might have been less able than Wiesel to read the shouting and cursing rage as perhaps it was intended (Flanagan 2017).
16. Besides blue anger, anger mixed with or born of sadness, there are also red anger and yellow anger. Red anger is bloodthirsty, vengeful. Yellow anger is the reactive anger of a frightened soul, someone who is cornered, possibly a coward, possibly someone with no other recourse than to return terror for terror.
17. P. Russell (2020) is a smart, popular essay on ways in which some varieties of righteous anger for moral ends embed cruelty and sadism. Moral grandstanding (Tosi and Warmke 2020) often enacts a passive-aggressive form of anger. The grandstander prides themselves on being seen as a hypervigilant, moral connoisseur who shines a bright and humiliating light on the transgressor.
18. Anger plays a significant role in morality for children of Abraham and Aristotle. Anger is used to teach a normative order that contains norms that regulate anger itself, and that keep it from becoming ruthless, sadistic, or frenzied. Stratton (1923) is an important psychological treatise on this topic. He divides religions into three types: (1) “The Irate and Martial Religions,” Judaism, Zoroastrianism, and Islam; (2) “The Unangry Religions,” Taoism, Vishnuism, Buddhism, and Jainism; and (3) “The Religions of Anger-Supported Love,” Confucianism and Christianity.
19. Gibbard (1990) intends that the anger-guilt schema capture a recognizable and powerful vein in our moral psychology and ethical thinking—the same vein that Bernard Williams calls “the morality system” (1985, 51)—which trades in the language of duty and obligation.
20. One could have this view that moralities differ in terms of what moral emotions they primarily engage, but still think that the anger-guilt schema is a good heuristic for us to identify moralities that are not built on anger and guilt. We look for the ways in which the members of another culture deal with behaviors and areas of life, family, sex, resource distribution, and interpersonal conflict that we would govern with anger and guilt norms, and see first, whether they have norms governing these areas, and then next see what emotions do the work of normative governance, which we would do by way of anger and guilt. We inquire what those norms are and which emotions play the role for them in normative governance that anger and guilt play for us.
Chapter Two: Anger across Cultures
1. This thirty-item inventory does not reveal exactly where the cleavage lies between statistical norms, conventional norms, and ideal norms for anger. It also doesn’t mark very clearly what is included in the local concept of anger as understood by cultural psychologists and anthropologists, whether, for example, the concept translated by the word “anger” includes “as if” anger, instrumental anger, primarily felt anger, or primarily angry behavior, impersonal anger, payback, pain-passing anger, or all or only some of these. To say anything very precise or helpful about such matters would require a kind of fine-grained, multidisciplinary, thick description that isn’t yet available. Michele M. Moody-Adams (1997) is wise on the need to exercise care in the use of anthropology to make philosophical points. Andrew Beatty (2019) is also terrific on the complexities of translating emotion terms, emphasizing the complex ways emotions are embedded in an entire form of life, with an idiosyncratic history, religion, and so on.
2. It is not clear from the ethnographies whether the Minangkabau view all anger as vicious, or whether distinctions are made among spheres, whether, for example, personal anger is always wrong because it violates the aims of love and respect, but that anger is allowed in commercial relations when there is cheating, or at the political level when there is grave injustice.
3. The contents of the two moralities cannot be entirely identical even on this assumption, which tries to keep the focus on the socializing emotions. The reason is that there will be different substantive norms about the socializing norms themselves. Furthermore, the dominant socializing emotions will create the moral equivalent of a dominant type of weather—stormy, cloudy, windy, calm, sunny—that will color all aspects of life.
4. David Wong has many wise things to say about the plurality of forms of life that are above a certain threshold of decency, and what is possible by way of critique when this is so (1984, 2006).
5. Sbuddhism is also, for similar reasons, compatible with punishment, war, and violence.
6. Shaming and humiliating are associated with shame. But there is no necessary connection. Shaming and humiliating are excellent vehicles for both revenge and pain-passing varieties of anger.
7. On game theoretic grounds one could imagine that the Bara are vulnerable to rapid moral change, if we assume that they have even a small preference for doling out less pain, all else being equal. If the Bara have this preference and then receive word, and accept it, that some other methods of moral socialization exist that are less painful physically and/or psychologically than their angry methods—e.g., shaming, recite after me, gold stars for good behavior, etc.—are also equally effective in securing their normative order, and are no more expensive timewise, and so on, then they should want to change. The discovery of other methods that inflict less pain might lead to changes in their practices. Similarly, imagine that the Minangkabau have misgivings for some reason or other about shaming and then learn that Jews and Catholics have used guilt rather than shame effectively to maintain the moral order. Moral epidemiology might predict swift migration to guilt-inducing norms if such a thought took hold. See Skyrms (2003) and Bicchieri (2016).
8. A major reason Buddhists give for why anger is the most destructive emotion is that it is incompatible with—perhaps psychologically and performatively inconsistent with—bodhicitta, the awakening of compassion, the virtue that desires to alleviate the suffering of all sentient beings. Anger, at least one familiar kind, aims to produce suffering.
9. The degree to which the disagreement between individualists and nonindividualists is metaphysical, or whether it is mostly normative, pragmatic, and forensic, is very complicated. All parties could agree that there are individuals but believe that well-being and responsibility are not individualistic. Henrich (2020) argues that WEIRD and non-WEIRD peoples differ concerning a host of dimensions associated with individualism, and disagree about whether the locus of blame and responsibility for bad character is an individual or a social unit (a family, village, nation-state).
10. Burdened virtues can burden in different ways. If an oppressed group is required, e.g., to not get angry even though they have reason to get angry, they are burdened. Srinivasan (2018) calls this “affective injustice.” If an oppressed group is asked, e.g., to be angry in a way that would be considered wrong for others—like breaking the law in order to accomplish the goal of liberation—they are burdened in a different way. In the Bara case, I am imagining that they all feel burdened because they have to treat the youth so harshly. But they conceive it as necessary for a good life. Humans are, as Xunzi and Kant both said, “crooked wood,” so one does what needs to be done to straighten us out.
Chapter Three: Anger and Flourishing
1. Annas et al. (2016) is a collection of papers that defends the alternative view that there are, as it were, objective harms caused by certain child-rearing practices.
2. Seneca, in De Ira (1928), also gives some psychobiological reasons why anger is a bad socializing emotion, as well as bad for mature adults. It is naturally unruly. It fusses on the outskirts of things. You hurt me just now by mild inconsiderateness and suddenly your failure to rake the leaves last weekend or failure to promptly empty the dishwasher last Thanksgiving is in play. Angry soldiers rape and pillage; good soldiers simply win battles.
3. Dear reader, do not let “no-self” tie you up in knots. Most Buddhists accept that there are persons with personalities. The doctrine of no-self (anatta; anatman) is a denial of atman, an idea distinctive of and internal to orthodox 500 BCE Brahmanism, where high-caste priests claimed to have a divine part, atman, that linked them to, or made them part of, Brahman, the ultimate divine source of everything. No-self was intended initially to say “no” (an) to the self-puffery and ego of the high priests of what we now call Hinduism. It was never intended as a response to any familiar Western belief about the self, persons, or souls.
4. The majority script could refer to a numerical majority script, or, as likely, to a power majority script, to the script of the most powerful who authorize themselves to speak about what “we” think and how “we” do things.
5. Glen Pettigrove (2012) explores threads in our own Western tradition—e.g., in Hume—that give arguments for meekness and against payback anger, and Nussbaum (2015, 2016) works to retrieve sources in Aristotle, the Stoics, and the practical work of Nelson Mandela and Mahatma Gandhi for the same idea. Arguments against payback and pain-passing anger seem to be increasing at a time when arguments against retributivism in the law are also increasing, and for many of the same reasons. If both arguments win at the same time, I’d like to think it has a lot to do with better psychological theorizing and better metaphysical thinking about causation and agency. This would then be a case where wide reflective equilibrium procedures—but not superwide ones—were successful.
6. What Charles Taylor calls an “unforced consensus” is similar to what Rawls calls an “overlapping consensus.” The atheists, hedonists, Lutherans, and Catholics disagree about many metaphysical and moral truths, but they overlap on the belief that the state should allow freedom of religion. Seeking an unforced or overlapping consensus is not the same as seeking reflective equilibrium, narrow or wide. This is because one might get everyone to agree that we shouldn’t eat pork. For A, this is agreed because it is prohibited by Yahweh in Hebrew Scripture; for B, it is agreed because it is prohibited by Allah in the Quran; for C, it is agreed because they are a vegetarian. This is a case of both unforced and overlapping consensus, but the reasons are not mutually consistent (as required by narrow reflective equilibrium), and, in addition, they disagree about the facts (as required by wide reflective equilibrium).
7. I am assuming that almost all such comparisons will be piecemeal or segmental, about particular moral beliefs, ideals, or practices, or small sets of these. See Ruth Chang’s (2002, 2016) work on choices that involve options that are “on a par,” and the sort of enacted commitment for one’s life and the world to be different involved making “hard choices.”
Chapter Four: Generic Shame
1. The language of autonomy (good) and heteronomy (bad) is from Kant. It expresses an anxiety common among individualists. Markus and Kitayama (1991) is the classic paper in psychology on cultural differences in self-conception as they relate to conceiving the self as an autonomous, self-creating being or a communally created and sustained being.
2. I repeat that I do not take a position on one debate inside shame research. On one side, there are those who think shame is a basic emotion; on the other side, there are those who think it is nonbasic and assembled from some other basic emotions, possibly fear, anxiety, and sadness. It doesn’t make a difference to my normative arguments in favor of shame whether it is a sui generis basic emotion or quickly assembled from other basic emotions.
3. Psychoanalytically oriented object relations theories, which I discuss in chapter 5, also posit a kind of primitive shame or protoshame. But they conceive protoshame as the response of the universal narcissistic child inside us all who discovers its lack of omnipotence in early childhood and carries the shame of this discovery for the rest of its life. This view is not normally expressed in a way that is inconsistent with the theory of evolution. But it doesn’t incorporate evolutionary thinking either, and is not likely to be endorsed by its proponents.
4. The Sznycer et al. study (2018) is especially interesting for two reasons: it claims to find evidence across fifteen cultures for invariance in the neurocognitive shame system, and it explains this invariance in terms of strong selection pressures in favor of a system that assists in quickly acquiring local normative knowledge as protection against ostracism. Humans are made for doing, and doers need to be good predictors of two kinds of outcomes: whether the doing achieves its aims and whether there will be uptake by compatriots. A proleptic system anticipates and raises objections to itself about action plans—ideally at lightning speed—before they are raised by external forces. Shame is unpleasant, so one wants to be good at avoiding it, which requires being a quick study of social norms, both conventional and moral.
5. Corey Maley and Gilbert Harman (2019) developed the interesting idea that feeling guilty comes in varieties that can incorporate different emotions. In their view, feelings are a higher-level category that includes or incorporates emotions. So if I feel guilty that I didn’t exercise today, it might incorporate the emotion of anger at myself. If I feel guilty that I left the dishes in the sink, the underlying emotion might be fear that my partner will be angry. These two cases involve doing something wrong, possibly a resolution not to do it again, and thus are well described as guilt for those reasons. But the underlying emotion is self-directed anger in the first case and fear in the second. This same point might apply to shame. If shame is, as one theory argues, assembled from sadness, fear, and anxiety, then there might be different kinds constituted by different ratios: shame that is 80 percent anxious shame; shame that is 75 percent fear shame. I am more interested in the content of shame than its inner psychological or phenomenological features, so I’ll leave this interesting possibility to one side for others to think about.
6. Using examples from literature to splendid effect, John Kekes argues that there is no general answer to the question of whether shame is good. I agree. It depends on the “evaluative framework” the shame is designed to protect. Even if shame protects good values in an overall good evaluative framework, Kekes thinks that “fear of shame” is normally not the best option among those available for defending and sustaining the normative order. “There are better ways of protecting our self-respect than fearing shame” (2019).
7. Kekes recommends reading Aristotle’s cryptic remark that “shamelessness—not to be ashamed of doing base actions—is bad, but [it does not follow] that it is good to be ashamed of doing such actions” as similar to saying that “having a heart attack is bad … but it doesn’t follow that fearing a heart attack is a good way to heart health” (2019). According to Kekes, what works better than fear of shame to keep one’s eyes on the prize is a settled sense of ideals, a reflectively endorsed evaluative framework, self-knowledge, a proper sense of duty and obligation, and skillfulness at diverting unworthy motives. This is what I call “the mature sense of shame.”
8. Mengzi: With Selections from Traditional Commentaries, trans. Bryan W. Van Norden (Indianapolis, IN: Hackett, 2008), 2A6.
9. The idea that shame is linked to the possibility that God might repudiate a person altogether seems very close to the view that shame involves a negative judgment about the global self. Catholic orthodoxy thinks hell exists, but there is no consensus view about whether it has any occupants. God is all-loving, God loves sinners, God is forgiving, God is full of grace, and no person is wholly irredeemable, wholly worth repudiating.
10. Krista Thomason (2018) provides a helpful conceptual analysis of shame as it might be understood from an entirely secular philosophical perspective. Like Thomason, but for mostly different reasons, I claim shame is, or can be, good.
11. The orthodox story unfolds this way: One desire the narcissistic child in all of us has is to possess the opposite sex parent sexually. Once the child suffers the universal narcissistic wound and experiences primitive shame upon discovering it is not omnipotent, it resolves its desire to sexually possess its opposite sex parent in a compromised way. The girl child sees that she can gain the father, or more likely, someone relevantly like the father, by modeling the mother’s ego ideals. The boy child has a more complicated developmental task. He experiences both sexual desire for the mother, and at the same time fear that the father will castrate him if he acts on it (even in fantasy). He takes his sexual sights off the mother and sets them on eventually finding a partner who is relevantly like the mother by modeling the father (who once wooed the mother by his charms). But modeling the father, who the boy child fears will castrate him, results in internalizing his stern superego.
12. Caelainn Hogan’s Republic of Shame: Stories from Ireland’s Institutions for “Fallen Women” (2019) is the tragic tale of the twentieth-century Irish convent homes for unwed mothers. It is also a story that explains the whys and wherefores of what she calls “the shame-industrial complex” in terms of normative holism. These homes make sense only in the context of a perfect storm of the entirely sexist shame attached to female sexuality and the additional shame of unwed pregnancy, combined with the shame of poverty and religious superstition that required secret baptisms of newborns before they were, once more shamefully, adopted out in a country where such adoption was illegal. The evil here was due to the inhumane and sexist values shame served, not to shame itself. Remember: shame is a general purpose emotion that involves feeling bad about some action or aspect of oneself; it is used to promote normative conformity. Shame is indifferent to whether it serves good or bad values. In the Irish homes for unwed mothers, it served bad ones. The point is that what can seem as if it is a qualitative assessment of the value of shame is usually not. It is an assessment of the values shame serves in a particular sociomoral ecology.
Chapter Five: The Science of Shame
1. Psychoanalytic object-relations theory conceives the project of psychic development in early childhood as being primarily taken up with the child learning that it is not the only thing (object) there is. The task is to be able to differentiate between objects of various sorts—for example, self and other, caretakers with their own wishes and desires, pets, transitional objects like teddy bears, and inanimate objects—and to develop the kinds of relations with these different objects that we expect of mature persons. The theory starts with Sigmund Freud. Nussbaum’s rendition comes from a synthetic reading of a range of thinkers from Freud to John Bowlby, W.R.D. Fairbain, Otto Kernberg, Melanie Klein, Hans Kohut, Margaret Mahler, Andrew Morrison, Daniel Stern, Silvan Tomkins, and Donald Winnicott.
2. Richard Wollheim, author of The Thread of Life, is another philosopher impressed by psychoanalysis. Wollheim proposes that guilt serves the superego, whereas shame serves the ego ideal, where the distinction is roughly as follows: The superego is one’s conscience, an internal figure (in classical Freudian theory, it is the internalized voice of a frightening and potentially punitive father) that “aims to control the inner and outer life of the person who houses it … [by] commanding, forbidding, cajoling, threatening, rewarding, [and] punishing” (1984, 200). The phenomenologies of shame and guilt then differ as follows: “The essence of shame, or what is reverberatory about it throughout the psychology, lies in the look, in the disparaging or reproving regard, whereas the essence of guilt lies in the voice, in the spoken command or rebuke” (ibid.). Like Bernard Williams’s perspective, Wollheim’s is helpful as far as giving shame a second look goes. Wollheim sees shame as the more mature emotion: it emerges after guilt and is less behavioristic, and it is more aligned with the philosophical projects of knowing the self and modifying the self.
3. Freud (1905) thought that there were some such species memories. For example, he explained penis envy in girls who had never seen an actual penis in terms of such a memory.
4. John Deigh, who has been working on and thinking about emotions generally and shame in particular for a long time, wrote an incredibly sensitive review essay (2006) of Nussbaum’s Hiding from Humanity (2004). Deigh distinguishes between tutored and untutored emotions, and questions Nussbaum’s views of the nature of shame in its untutored form. He concludes that Nussbaum’s critiques of the use of shame in the law turn mostly on conservative content of certain legal practices that endorse shaming and humiliation, not on negative features of shame itself.
5. Sexist shaming—for example, slut shaming—is wrong because it abides a sexual double standard and shames women for being sexual beings. It isn’t something women are prone to because of anything about the nature of shame itself, or because they are cursed (like men, according to the view on offer) with “I am a failure” thoughts/feelings. It is because there are social messages that convey to women that their sexuality is something to feel ashamed about. The problem, as always, is with the values shame is called upon to protect, not with shame as such.
6. James (1892, 334–35).
7. Lewis explains that all her patients had suffered “traumatic childhood events” (1971, 12). In most cases, they returned to therapy after successful analysis, blaming “the tenacity of childhood hang-ups for permanently malforming their character” (ibid., 13). Lewis sees shame implicated in three ways for these patients: shame for the childhood trauma, shame for their inability to get over the childhood trauma, and, finally, for transference shame they projected onto the analyst. Lewis explained this last type or variety of shame as due to the fact that the analysand shares with the analyst in dream interpretation and free association all sorts of unseemly thoughts and desires, and is, upon reflection, ashamed of having done so. This creates a “feeling trap,” where shameful feelings are activated and intensified, with no easy way out from the heightened experience of the shame.
8. Maley and Harman (2019, 22–23). See also Giner-Sorolla, Piazza, and Espinosa (2011) for questions about TOSCA’s validity.
Chapter Six: Shame across Cultures
1. The situation is similar with happiness, sadness, fear, and anger. These emotions also abide normative rules in the sense that they are apt or fitting, just in case it makes sense to be in these states, given what they are about. Happiness is apt if you have the winning lottery ticket, but not if you incorrectly think you do. Fear is apt if that really is a snake and not a belt or rope. Anger, like shame and guilt, is more subject to sociomoral discipline than happiness and fear. Thus, anger is apt if there is a justice violation and the anger does not aim to simply or only pass pain or gain revenge. My inclination is to say that, among these various emotions, anger may be the one that in itself, by its very nature, is a cause of trouble. In fact, I am inclined to say of anger, but not shame, what Nussbaum says of shame and disgust: anger is “especially likely to be normatively distorted and thus unreliable as a guide to public practice, because of features of its specific internal structure” (2004, 13). My thought is that anger, as a general response to not getting one’s way, is so undisciplined and rapacious that, as Seneca thought, domesticating it is super difficult.
2. If one imagines that the shameless artists are also amoralists and libertines, one can start to see how neither shame nor guilt would play much of a role in their lives. Then again, edgy artists are usually shameless only in certain zones of life—sexual, artistic, lifestyle—but would not do many shameful things, like be disloyal to friends and loved ones, break laws, rape, pillage, kill, and so on. It is hard to imagine anyone who lacks a sense of shame altogether.
3. As I will say shortly, a mature sense of shame involves a regimen of “inspect then avow.” That is, the person with a mature sense of shame is not overly confident about either their worldview or particular beliefs and value commitments within it. They will be reflective enough to have thought about their values, and if challenged will engage in further reflection.
4. I grew up in a large family in a community of large families. Playing with siblings, and with large groups of kids from a bunch of families, was the stuff of everyday life. So, organizing softball and kick-ball games with fifteen kids from the neighborhood was a regular occurrence. Everyone knew the responsibilities of older kids to include the younger kids in such games, make them feel welcome and supported, and so on. We policed “bad sports” by way of something in the vicinity of shame/guilt, invoking norms about how the game “is supposed to be played,” positions to be allotted, disputes settled, and so on. It was all very elaborate. My hypothesis is that we had all learned at home some prototype rules about how to play with, share with, and take turns with siblings of different ages. This was the basic schema. Then we exported and extended the prototype to games among all the kids, and we adopted and enacted the modes of enforcing norms by adopting and adapting the methods our parents used with us at home to teach us how to play with siblings.
5. Andrew Beatty (2019, 87–89) writes of the Javanese that their shame (isin) is tied mostly to social violations or accidents that look like bad manners (e.g., showing up at a party to which one was not invited), but not to moral violations.
6. Nietzsche (2006) is brilliant at developing the association of guilt with debt, as well as the image of an internalized creditor who is a sadistic part of the self that drools at the prospect of extracting payment in a painful way from other hated parts of itself.
7. The Catholic Church of my youth recommended feeling ashamed (or guilty) for being in certain situations or places that were referred to as “near occasions of sin.” My friend Jeanne Lorio grew up in Louisiana. The after-party for her prom was held at a milkshake shop, which happened to share the front of a building with the local house of prostitution, which was in the back. Jeanne’s girlfriend, on that particular double date, thought she ought to confess to being at the milkshake shop, given that its proximity to the house of prostitution made it a “near occasion of sin.”
8. Quinn (2005) observes that the emotion of pride is a major socializing emotion in middle-class America. This seems right. American children are encouraged to stand out, to have high self-esteem and self-confidence. Many have noticed that the currency of pride, especially when it suffers a steep inflation rate—trophies for showing up at life—cheapens the value of work on oneself, or for the common good, and instead rewards noticeability, which is enhanced by networking, standing out from the crowd on social media, and the like. The worry then is that the pride is detached from the types of activities that would make it pride-for-being-engaged in genuinely worthwhile activities (Lukianoff and Haidt 2018).
9. This failure on translatability is predicted from the discussion of the semantics of emotion terms in the introduction, and from the discussion of conceptual, normative, and network holism in chapters 4–6. In contrasting and comparing how emotions are done in different sociomoral ecologies, we look for family resemblances, never conceptual identity or linguistic synonymy.
10. Nussbaum (2004) provides wise arguments against using shaming in the law, which is entirely compatible with teaching people to be ashamed for law breaking.
11. Gibbard (1990) contrasts our anger-guilt morality with a disdain-shame morality. I’m skeptical that a shame morality would need to use disdain or contempt on the social enforcement side of things. Confucian ethics gives pride of place to shame, but doesn’t encourage disdain or contempt as the social enforcement mechanism.
12. This is not TOSCA shame-proneness, which involves the kind of shame that is associated with low self-esteem and judges the whole self to be the cause of all failures or missteps. Minangkabau shame-proneness is just a kind of natural shyness or modesty.
13. One other point: I am skeptical that the extreme painfulness of WEIRD shame needs explaining. The reason is that shame’s alleged extreme painfulness is cooked up; it is a finding about the kind of shame (TOSCA shame) where the entire self is judged as shameworthy. This kind of shame has to be extremely painful; it is existentially undermining of an entire person’s being, and it is always entirely undeserved. But it is also not normal shame, which takes actions or aspects of the self as its evaluative object, and needn’t have any global self-undermining features.
Chapter Seven: The Mature Sense of Shame
1. Bryan Van Norden examines the possibility that conventional shame and moral shame are entirely different kinds of shame, wondering “is it just homonymy” that “shame” is used in both phrases? He concludes that shame expresses the same concept in both places. “We are social animals who are rightly sensitive to the opinions of others, in the first instance, caretakers. A humble person maintains this sensitivity” (2007, 266–67). I agree wholeheartedly.
2. In this book, I do not take up the important topic of collective shame. I am, however, very interested in the topic. Two observations: (1) There are lessons from genocide studies that might teach us about the way shame can protect a sense of solidarity and common humanity (Morgan 2008; Maibom 2010). Allied soldiers who liberated prisoners from Nazi concentration camps felt shame. Why? They were liberators not perpetrators. The concentration camp survivors who were liberated also felt shame. Why? They were 100 percent victims of evil. The best answer to both questions invokes multiple causes. In the first case, there is a sense of shared humanity and thus feelings of solidarity with the survivors, as well as the horrifying thought that fellow humans—and thus possibly oneself—were capable of this horror. In the case of the survivors, there was the shame at being seen in emaciated and degraded condition, shame about surviving when others, including loved ones they wanted desperately to protect, died; there was sorrow for their liberators being exposed to what the survivors knew but wished for no one else to know (the way a loving parent wishes that their child will never learn certain difficult truths), and possibly much more. (2) We live in a time when many people find it important to identify themselves with a people, with a lineage. If one does this as part of the process of constituting oneself as a person of a certain sort, and considers oneself to be made as one is, in certain favorable ways, based on ancestors, then one should also take responsibility for harms one’s ancestors committed. European Americans like myself who identify as such and have pride in their ancestry share the shame of what those ancestors did to the original people and to chattel slaves. We owe reparations to peoples in those lineages. The argument is conditional. If you conceive yourself as having a historical identity constituted in part by a lineage (ethnic, racial, linguistic, geographical), then you have some responsibility for what your lineage wrought. One could reject the conditional obligation by denying that one has an identity that involves ties to a people or a lineage, or, what might come to the same thing, by asserting that each person is an autonomous individual only responsible for what they do or did in their lifetime (end of story). In the other direction, one can justify a sense of collective responsibility by way of nonindividualistic metaphysics of the self and without mentioning ethnic or any other kind of identity, but instead some kind of what P. J. Ivanhoe calls “the oneness hypothesis,” where one is in solidarity with all of a nation or all of humanity or, at the limit, the cosmos itself (see Markus and Kitayama 1991; Ivanhoe 2017; Ivanhoe et al. 2018).
3. I am still thinking about how best to describe the mature sense of shame in both psychological and metaphysical terms. I do not think it is easy to capture the sense of shame in terms of the three main types of principled moral theories—deontology, consequentialism, or contract theories. Nor is it easy to capture the sense of shame in terms of character or virtue theories. Principle-based theories emphasize the principles one ought to abide by and tend to be neutral or silent on which socializing emotions should enforce the principles. Virtue theories, meanwhile, normally conceive each virtue as having an affective component, which is required for its proper execution. Shame is given an important role in Confucianism and Buddhism, but not so much in neo-Aristotelian virtue ethics. There are historical reasons for this. Aristotle thought shame was only a quasi-virtue. By the time there was an Aristotelian revival in the medieval West, it was in the hands of Christian and Arab philosophers, who officially preferred, broadly speaking, the emotions of anger, fear, and guilt to shame for enforcing morality. I do not think that the sense of shame is a dedicated mental faculty, innate or acquired. Nor is it a particular virtue. I tend to think of it as akin to a master virtue or supervisory disposition. In this respect, the mature sense of shame has features in common with Aristotelian phronesis and Confucian zhi. But the disposition, as I am conceiving it, has both a more affective character than either phronesis or zhi and also powerful deontic features: it tells the person whose sense of shame it is, that this is not the sort of thing that one ought to do. If the deontic standard is that this is not the sort of thing a truly good person does, then the critic of shame might say, “See, once again shame has shown its global nature; it says ‘I am a failure,’ or ‘I am bad.’ ” Not at all. Almost no one is perfectly good or ideally good. When one does what one ought to be ashamed of (even if totally undetected), one feels shame about being insufficiently attentive to the norms and best practices expected by the community that one depends on for flourishing. The kind of shameful act—Is it financially greedy? Does it display an indifference to the truth? Does it violate a promise?—indicates to the agent what aspects of their character deserve attention and work. To be sure, the self-work is about you, but it is almost never about all of you or every aspect of yourself. If people are prone to getting confused about this point, then it is the responsibility of the moral community to clear up the confusion. See Jiminez (2020), Krishnamurthy (2022), Lebron (2013).
4. For further reading on Confucian shame, see Kwong-loi Shun (2001), Van Norden (2007), Geaney (2004), Seok (2017a, 2017b), and Barrett (2014).