Suppose that you are angry on Tuesday because I stole from you on Monday. Suppose that on Wednesday I return what I stole; I compensate you for any disadvantage occasioned by your not having had it for two days; I offer additional gifts to show my good will; I apologize for my theft as a moment of weakness; and, finally, I promise never to do it again. Suppose, in addition, that you believe my apology is sincere and that I will keep my promise.

Could it be rational for you to be just as angry on Thursday as you were on Tuesday? Moreover, could it be rational for you to conceive of a plan to steal from me in turn? And what if you don’t stop at one theft: could it be rational for you to go on to steal from me again, and again, and again?

Though initial anger might be reasonable, we tend to view unending disproportionate revenge as paradigmatically irrational.

Though your initial anger at me might have been reasonable, we tend to view a policy of unending disproportionate revenge as paradigmatically irrational. Eventually we should move on, we are told, or let it go, or transmute our desire for revenge into a healthier or more respectable feeling. This idea has given rise to a debate among academic philosophers about the value of anger. Should we valorize it in terms of the righteous indignation of that initial response? Or should we vilify it in terms of the grudge-bearing vengeance of the unending one?

I am going to explain how that debate goes, but I am not going to try to resolve it. Instead, I am going to peel it away to reveal a secret that lies behind it: we have been debating the wrong issue. The real debate concerns the three questions about anger and rationality in my second paragraph, which are not rhetorical, and to which the answer might well be: yes, yes, and yes.


First, the academic debate. In one corner, we have those who think that we would have a morally better world if we could eradicate anger entirely. This tradition has its roots in ancient Stoicism and Buddhism. The first-century Roman philosopher and statesman Seneca wrote that anger is a form of madness; he authored a whole treatise—De Ira, the title of this volume—about how to manage its ill effects. The eighth-century Indian philosopher and monk Śāntideva enjoined those wishing to travel the road of enlightenment to eliminate even the smallest seeds of anger, on the grounds that the full-blown emotion can only cause harm.

In the contemporary world, the philosopher Martha Nussbaum draws on Seneca and the Stoic tradition to argue that anger is an intrinsically mistaken attitude, since it is infected with a backward-looking “payback wish” that is vengeful and destructive. The correct response to any setback or injustice, in her view, is forward-looking: preventing similar events from occurring in the future. In a similar vein, Owen Flanagan, who draws on both Śāntideva’s Buddhism and a Confucian-inflected metaphysics, sees anger as an intrinsically hostile attitude, one that falsely presupposes a self-centered metaphysics of individuals who possess “intentions to be cruel, and to do harm or evil.”

Nussbaum draws Seneca and the Stoic tradition to argue that anger is infected with a “payback wish” that is vengeful and destructive.

In the other corner of the debate stand those who conceive of anger—up to a point—as an essential and valuable part of one’s moral repertoire: anger is what sensitizes us to injustice and motivates us to uphold justice. By being angry with me on Tuesday, the day after I stole, you create the system and demand the terms under which I must acquiesce and “make things right” on Wednesday.

This pro-anger position has its roots in Aristotle’s view that the (well-trained) passions are what allow the “eye of the soul” to perceive moral value, and finds its fullest expression in the British moral sentimentalists of the seventeenth and eighteenth centuries. The Earl of Shaftesbury, Frances Hutcheson, David Hume, and Adam Smith all held that our feelings are precisely what sensitize us to moral considerations.

Later, Peter Strawson’s watershed paper “Freedom and Resentment” (1960) injected new life into the pro-anger cause by making emotions the fundamental mechanism of moral accountability. Strawson develops Smith’s insight that our status as moral creatures rests on the fact that we care—at an emotional level—what we think of one another. Strawson understands negative emotions in the anger family as paradigmatic expressions of moral assessment. Anger treats its target as someone capable of recognizing that she has done wrong and is to be contrasted with the indifference or calculating carefulness by which we might react to someone we see no hope of reintegrating into the moral community.

Strawson’s continued influence is visible in the work of contemporary philosophers such as R. Jay Wallace, Jesse Prinz, Allan Gibbard, Pamela Hieronymi, and Jean Hampton. Though differing in their conclusions and many of the steps along the way, all begin from the sentimentalist assumption that emotions lie at the bottom of our practices of holding one another morally responsible. Emotions are how we humans do morality.

For sentimentalist philosophers, emotions are how we humans do morality.

But are these two camps—the Stoics versus the sentimentalists—really diametrically opposed? Each must respond to the data that motivate the other, and when they do so, they make some surprising moves toward reconciliation.

Consider the data of the anti-anger side. There are at least two big drawbacks of anger, they note: first, the tendency to cling to one’s anger, bearing a grudge deaf to any reasonable voice of reconciliation, apology, or restitution; second, the tendency to exact (often disproportionate) revenge. The fans of anger carve these phenomena off as pathologies, not essentially associated with anger. They use special words such as “indignation” and “resentment” to refer to anger purified of such impulses. Purified anger, they say, protests wrongdoing but is free of vengeful impulses and is immediately responsive to reasons to give up one’s anger. (In this technical terminology, “resentment” is typically used to mark protesting on one’s own behalf, whereas “indignation” is for protesting on behalf of another.) This move—carve away the dark side—is remarkably similar to the move the enemies of anger make when confronted with what we might call the “moral side” of anger.

Both Flanagan and Nussbaum, for instance, acknowledge that one who fails to react to grievous wrongdoing runs the risk of acquiescing in evil. They grant the importance of a moral sensibility that would lead a person to object to being treated with disrespect, but they hold that such a response is possible without anger proper. Flanagan uses the word “righteous indignation” to cover “judgment that such-and-such state of affairs is grievously wrong, the wrong ought to be righted, and a powerful emotional disposition to want to participate in righting the wrong without being angry.” Nussbaum speaks of “transition anger,” which is not so much anger as “quasi-anger”: “the entire content of one’s emotion is, ‘How outrageous! Something must be done about this.’”

Notice what has happened: what started out as a battle over anger ends with everyone agreeing to avoid using that word. Instead, both sides prefer to segregate the “moral side” of anger (Tuesday’s anger, which takes the form of rational and justified protest at injustice) from the “dark side” (Thursday’s anger, which takes the form of irrational grudges and unjustifiable vengeance). It does not matter whether we follow the Strawsonians and call this moral side “indignation/resentment,” or whether we use Nussbaum and Flanagan’s terminology of “transition anger” or “righteous indignation.”

Now, when philosophers fail to disagree about any question of substance, you know someone is hiding something. In this case, I believe the pseudo-war has distracted us—and the combatants themselves—from the contentiousness of an assumption being made on all sides. Everyone assumes that we can retain the moral side of anger while distancing ourselves from paradigmatically irrational phenomena such as grudges and vengeance. But what if this is not the case? What if we humans do morality by way of vengeful grudges?

What started out as a battle over anger ends with everyone agreeing to avoid using that word.

It is a fact of life that human beings have a direct emotional vulnerability to how we are treated, so when you wrong someone, you inflict on them the distinctive pain of unjust treatment. This moral sensibility on their part is included in the very meaning of what it is to “wrong someone”: part of why wronging people is unjust is that they notice it.

It is also a fact of life that people tend to draw grudge-bearing and vengeful conclusions from premises involving genuinely moral facts about injustice and wrongdoing. I believe that we should not be too quick to pathologize this inference or dismiss it as a psychological tic. I will offer two arguments—the Argument for Grudges and the Argument for Revenge—that link those premises to those conclusions, suggesting that the reasoning in question is, in fact, valid. If we can’t purify morality, we can’t purify anger.


Let’s go back to our original example. Is it rational for you to remain angry with me on Thursday, after all my hard work to restore justice? Both neo-Stoics such as Nussbaum and neo-sentimentalists such as Strawson would say no.

They would say that you ought to take into account how the wrongdoing that prompted your anger has been addressed, via restitution, compensation, apology, and a promise. I have made amends for my wrongdoing in every possible way; if you continue to be as angry as you were, it must be, they would argue, because you are being irrationally insensitive to those amends.

The tendency to cling to anger through apologies and recompense, for years sometimes and to the detriment of all parties concerned, is routinely dismissed as irrational. It is often supposed, specifically, that not “letting go” of one’s anger must indicate a perverse pleasure in that anger. (Thus, poet Robert Burns in 1790: “Whare sits our sulky sullen dame / Gathering her brows like gathering storm / Nursing her wrath to keep it warm.”)

There are reasons to remain angry. And they are not hard to find: they are the same reasons as the reasons to get angry in the first place.

But this idea ignores the fact that there are reasons to remain angry. And the reasons are not hard to find: they are the same reasons as the reasons to get angry in the first place. Apologies, restitution, and all the rest do nothing to cancel or alter the fact that I stole, nor the fact that I ought not to have stolen. Those facts were your reasons to be angry. Since they are not changed by my forms of redress—apology, compensation, what have you—then you still have, after the deployment of these amends, the very same reasons to be angry. Anger, after all, is not a desire to fix something but a way of grasping the fact that it is broken. You are angry about something that is now in the past, and there is nothing to be done about that. What I did will always diverge from what I ought to have done, no matter what I do next.

There are, of course, many nonrational ways your anger might come to an end: you could die, develop amnesia, or it could just fizzle out over time. Suppose one day, out of nowhere, you simply decide to set your anger aside, and you succeed. We might judge that decision to be in some sense “rational”—who wants to go through their whole life angry?—but not in the sense that your reason for anger has been addressed. It hasn’t been, and it won’t ever be. Once you have a reason to be angry, you have a reason to be angry forever. This is the Argument for Grudges.

Now for the Argument for Revenge. Your desire for revenge, like your holding a permanent grudge, is typically taken to be irrational and unjustified. But this conclusion is typically a product of the assumption that revenge aims to solve the problem of anger, once and for all, by balancing out or undoing the wrong done. Once we let go of this assumption that anger can be undone—you have a reason to be angry forever, after all—it is not hard to produce an argument in favor of revenge. We should not find this possibility surprising; it would be strange if one of the oldest and most universal human practices did not have a rationalizing explanation. The Argument for Revenge is simply that revenge is how we hold one another morally responsible.

When I steal from you, you see me as responsible for a serious gap between the way the world is and the way it ought to be; there is a perspectival opposition between us. You see my action as morally unacceptable, and you experience that unacceptability as a pain, a harm. But I, who did it, evidently saw it as a perfectly fine thing to do, having judged the action to be a good thing for me.

There is an uncanny intimacy to anger.

Assuming that I understood that what I was taking was yours, and that I was not acting under some kind of duress—ignorance and compulsion are mitigating factors—my theft indicates that I see the world in value-terms opposed to yours. Your “bad” is my “good.” If you are to hold me accountable for this, instead of letting me off the hook, you will make this (accidental, adventitious) opposition a principle and rule for our interactions.

Revenge allows you to turn the principle of my action into a rule for your conduct toward me: you make my bad your good. This is the opposite of trying to undo or reverse my action. You hold me accountable by holding onto my theft, refusing to forget it, turning its one-off opposition between our interests into a rule to which I am now subject. You do not let me “live it down,” instead you force my own thinking down my throat. Seeing me as accountable for what I have done means treating my action as a principle governing our interactions. Angry people sometimes describe their vengeance as “teaching someone a lesson,” and this is quite literally true: you make my wronging of you into a general principle and then “educate” me by imposing it on me.

Educating me in this way is not easy on you: making my evil your good has psychological costs, among them the fact that you divert yourself away from what would otherwise be good for you. You must remodel your psychological landscape into one devoted to regulating mine. This explains the uncanny intimacy of anger: though you can’t stand to be near me, it is also true that no one could be closer to you than me. I have infiltrated the patterns of your thought; I have my fingers on your heartstrings; I have even been put in charge of your sense perception: you see traces of me everywhere you look. You complain about me to anyone who will listen, and when no one will listen you shout at a mental effigy of me. I’ve colonized your fantasy life. Holding me responsible involves an embrace, albeit an adversarial one. Anger feels exactly as you would expect, if it were true that my moral accountability was a matter of your seeing what’s good for you in terms of what’s bad for me.

Again, as with grudges, the point is not that, all things considered, one should take revenge. One may take other factors besides anger into account in governing one’s behavior. But insofar as one acts from anger, one pursues what is good for oneself by doing what is bad for another. This is perfectly rational, justified, and intelligible. Polemarchus, in Plato’s Republic, expressed the hostile logic of anger: justice gives benefits to friends and harms to enemies.

The Argument for Revenge is simply that revenge is how we hold one another morally responsible.

These two arguments—the Argument for Grudges and the Argument for Revenge—suggest it is not so easy to separate the idea that anger is a moral sense from the thought that we should hold on to grudges, or to embrace anger as a mechanism of moral accountability without endorsing vengeance.

I do not claim that these arguments make an open-and-shut case; objections are certainly possible, and a full defense of the validity of these forms of reasoning would be a big project. My aim here has merely been to show that there is a case to be made for the conclusion that grudges and vengeance are perfectly rational—and that such a case is not an overly complicated one. The arguments I have offered are simple and intuitive, qualities that make their neglect in the philosophical debates—in the form of the unquestioned assumption, on both sides, that grudges and vengeance are irrational—all the more striking.

Striking, but not inexplicable. For if we put the two arguments together, the result is that someone who is angry never has a reason to sever the link between the other’s evil and her own good. Perhaps the simple explanation for the neglect of these arguments is that we do not want to acknowledge the possibility that morally righteous anger provides rational grounds for limitless violence.


While it may seem, then, that the Stoics and sentimentalists are radically opposed, they share more than meets the eye. In particular, they share confidence—misplaced, I think—in a certain project of conceptual analysis. This project aims to identify a purified form of moral response, one incorporating all of the virtues and none of the vices of anger. I am not the first to argue that such a project is quixotic. Some version of my point can be found in a number of thinkers who approach questions of morality from a more historical and anthropological angle. Friedrich Nietzsche, Michel Foucault, and René Girard have all argued that the darkest sides of anger—vengeance, bloodlust, and limitless violence—are baked into the very idea of morality.

We do not want to acknowledge the possibility that morally righteous anger provides rational grounds for limitless violence.

Nietzsche’s Genealogy of Morals (1887) traces our present approach to morality to a turn away from a prehistorical ethic based on nobility and strength. The crucial sentiment guiding the new morality is ressentiment—a form of anger—felt by those previously oppressed and enslaved. What emerges is a “slave morality [that] from the outset says ‘No’ to what is ‘outside,’ what is ‘different,’ what is ‘not itself ’; and this ‘No’ is its creative deed.” The negative or reactive morality we have ended up with foregrounds the concepts of guilt, conscience, promises, and duty. Nietzsche says these concepts are “soaked in blood thoroughly and for a long time. And might one not add that, fundamentally, this world has never since lost a certain odor of blood and torture?”

Girard’s Violence and the Sacred (1972), a work of religious anthropology, discusses the role of human and animal sacrifice in the curbing of violence. Girard begins from the observation that every form of human community is threatened by one basic problem: once one act of violence happens, it threatens to set off a chain reaction of limitless retaliatory violence. According to Girard, what drives this chain reaction is nothing other than the moral horror at violence: “The obligation never to shed blood cannot be distinguished from the obligation to exact vengeance on those who shed it . . . it is precisely because they detest violence that men make a duty of vengeance.” Girard’s book contends that phenomena as widely varied as ancient scapegoating, Greek tragedy, and the sexual norms governing the nuclear family are all attempts to respond to this basic problem of the containment of violence.

Finally, Michel Foucault’s Discipline and Punish (1975) analyzes the shift from punishment by public torture and execution to punishment by imprisonment. Foucault’s thesis is that although these reforms were couched in the eighteenth-century language of human rights, their aim was to turn punishment into a focused attack on the prisoner’s human rights: “From being an art of unbearable sensations, punishment has become an economy of suspended rights.” Foucault then builds outward from the prison and argues that we can see the values of our society inscribed in the methodology of forcible restraint that characterizes such social artifacts as schools, examinations, timetables, and professional careers. The way we value freedom, autonomy, self-determination, and human rights is by taking those things away from people at every turn.

All three of these thinkers remain hugely influential despite having had the empirical details of their argumentation called into question by scholars from a variety of fields. I want to suggest that one reason for their enduring and even cult-like appeal is that they make a compelling and deep philosophical point that floats free of the particular historical-anthropological terms in which it is couched. What do these views have in common, after all? Nietzsche says we have built our whole morality out of resentful bloodlust; Girard says that violence and the opposition to violence are one; Foucault says that punishment is crime. The common denominator is the observation that human morality has a tendency to turn in on itself. Being a good person means, at times, being willing to do bad things.

Nietzsche, Foucault, and Girard have all argued that the darkest sides of anger—vengeance, bloodlust, and limitless violence—are baked into the very idea of morality.

I have offered reasons for thinking that the “dark side of morality” these three thinkers see mirrored in various social institutions derives ultimately from the logic of moral responsiveness: the morally correct way to respond to immorality is to do things—cling to anger, exact vengeance—that are in some way immoral.

If we abandon the anthropological distance and admit that we are the humans we are describing, grappling with this insight should produce nothing short of a crisis. We cannot climb outside of our own moral theory in order to assess it as bankrupt or broken; we must rely on it for the very terms of assessment. Of the three, Nietzsche comes the closest to facing this crisis, though even he often hides behind the suggestion that words such as “health” and “strength” offer him some kind of alternative footing. But who wants a society that is healthier or stronger unless those words are meant in a moralized sense—which is to say, a sense already shrouded by the darkness of our morality system? There is no magic trick that lets us climb outside our own normative skin.


Anger implicates all of us in moral corruption, then. Well, almost all of us. There is a certain Stoic so extreme that his position is represented by neither Nussbaum nor Flanagan, nor any modern thinker I know of. This Extreme Stoic sees emotions as having no role in morality; in order to achieve this complete emotional detachment, he places no value on anything the world can remove from himself, including his children, his life, and his freedom from physical torture. Extreme Stoics take inspiration from Socrates, who claimed that a good man could not be harmed, and who correspondingly denied that the Athenians were harming him when they put him to death for crimes he did not commit. Socrates died anger-free.

Most of us are neither willing nor able to achieve the kind  of detachment that this immunity from wrath requires. When people commit injustice against us, we feel it: our blood boils. At that point, we have to decide how much we want to fight to quell our anger, how much effort we are going to put into repressing and suppressing that upswell of rage. The answer is rarely none. While we do not want to let our anger get away from us and drive us to its logical, eternally vengeful conclusion, if we quash it with too heavy a hand, we lose self-respect and, more generally, our moral footing. Inhibiting any and all anger in the face of genuine wrongdoing is acquiescing in evil. So, we are regularly faced with the complicated question of how much anger to permit ourselves under a given set of circumstances.

But notice that, if the arguments I have offered here are correct, this question is equivalent to asking: How much immorality should we permit ourselves? The realistic project of inhibiting anger must be distinguished from the idle fantasy of purifying it. We  can use a word such as “indignation” or “transition anger” to postulate a feeling that righteously protests wrongdoing without any hint of eternality or vengeance—but the item to which that word refers is a philosopher’s fiction. The multiplication of kinds and flavors and species and names for anger is designed to distract us from the crisis at the heart of anger, which is that affective response to injustice clings to the taste of blood.

The morally correct way to respond to immorality is to do things—cling to anger, exact vengeance—that are in some way immoral.

I believe that, when faced with injustice, we should sometimes get somewhat angry. Such anger is not “pure” and entails submitting oneself to (some degree of) moral corruption, but the alternative, acquiescence, is often even worse. The point I want to emphasize, however, is this: just because the moral corruption of anger is our best option doesn’t mean it is not corruption.

The consequences of acknowledging this point are sobering: victims of injustice are not as innocent as we would like to believe. Either these victims are morally compromised by the vengeful and grudge-bearing character of their anger, or they are morally compromised by acquiescence. Long-term oppression of a group of people amounts to long-term moral damage to that group. When it comes to racism, sexism, homophobia, anti-Semitism, ableism, classism, religious discrimination, anti-neurodiversity, elitism of any stripe, this argument entails that the oppressors have made the oppressed morally worse people. Of course, oppressing people is also bad for your soul, but we do not need to be reminded of that; we are accustomed to the thought that wronging others makes you a bad person. My point is: so does being wronged, even if to a lesser degree.

I moved to the United States from Hungary when I was five years old, but I still spent my childhood summers there, at Lake Balaton. Across the street from my grandparents’ house there was  a resort popular among East Germans. I could not enter the resort area—it was surrounded by a fence—but one summer, when I was around ten years old, I befriended a girl around my age who was vacationing there. We had no common language, but we communicated by way of a marching game: we played soldiers and invented a complicated militaristic dance to which we would add moves day by day. We marched side by side, separated by the fence—until the day I was caught by my grandmother.

My grandmother was a concentration camp survivor, so she was horrified by what she saw: her granddaughter, marching with one of Them. I tried to explain that we were only playing a game, but to her it was clear: I was collaborating with the enemy. I argued that her prejudice against the German girl was no different from the Germans’ prejudice against us. My protest only made her angrier, and I was forbidden from ever approaching the girl again.

But how innocent was my game, really? All four of my grandparents, in fact, had survived concentration camps; all of them lost almost everyone they knew in the Holocaust. My grandmother denied that the Holocaust was the greatest tragedy of her life, giving that honor to the fact that her first child, my uncle, was born with cerebral palsy. But she even blamed that on the Nazis, perhaps not without reason: there are many stories of birth injury in the first generation of children born to women who had suffered malnutrition and other forms of abuse in concentration camps. (My other grandmother’s first baby was stillborn.)

Just because the moral corruption of anger is our best option doesn’t mean it is not corruption.

My parents decided to leave Hungary when the synagogue on our block was blown up. (After that, the Jews in the area went to my grandmother’s house to pray, secretly.) When we arrived in New York City, my parents pulled me out of public school after I was beaten up for wearing a necklace with a Star of David. They could not afford private school, but Orthodox Jewish elementary schools were willing to accept my sister and me for free, as charity cases. Why? Because the Holocaust, of course—which, at those schools, was its own subject, alongside English, math, and science. Before high school, I hardly wrote a poem or short story that was not in some way about the Holocaust.

Anti-anti-Semitism was so much the theme of my childhood that it is simply impossible to believe I accidentally fell into playing soldiers with a German girl. I wasn’t innocent. But my grandmother wasn’t innocent, either: she was full of anger. Innocence was not a possibility.


Nietzsche, Foucault, and Girard contributed to a strand of cultural criticism often invoked in support of attitudes of cynicism, misanthropy, and pessimism about the human condition. They are seen as radicals. In my view, however, all three are to be faulted for their timidity. It is striking the degree to which each writer held himself at a safe anthropological distance from the dark side of morality he so accurately described. If they had stepped inside their own theories, they would have immediately drawn the simple, devastating conclusion that it is impossible for humans—you and me and the three of them included—to respond rightly to being treated wrongly. We can’t be good in a bad world.