Paul Bloom asks, “Do angry people make the best romantic partners?” “Best” is a tricky word here, but speaking only for myself, I would not think that the “best” partner for me would be someone who never became angry or even someone who became angry less frequently than my husband does.

I am not perfect. Sometimes I am insufficiently loving, appreciative, or attentive; sometimes I do not try hard enough to overcome my faults; and sometimes I behave in ways that are outright disrespectful. Because he loves me, my husband doesn’t just observe or notice my disrespect, he directly undergoes or experiences it. My disrespect hits him where it hurts, and his anger hits back, and hurts me. If he became less sensitized to how I act, or if I became less sensitive to his anger, I would not see that as an improvement in our relationship. Rather, I would think that we had come to matter less to one another.

I maintain that there is something unhealthy about even the healthiest—that is, most successful—cases of anger.

One reason why we want romantic partners in the first place is that we need help to become the people we want to be—we are not, already, “best.” When you have a deep connection with someone, their anger allows you to outsource some of your striving: your partner’s anger is a mechanism of your aspiration. Daryl Cameron and Victoria Spring are right to note that making use of this mechanism has real psychological costs. Nonetheless, I want my husband to be willing to shoulder such costs on my behalf. I see his willingness to devote some of himself to regulating me—to dedicate a piece of his own psychological real estate to combating my faults and vices—as a measure of his love.

Of course this system can go haywire, as in cases of abuse. And, if activated too often, it is liable to generate a deadening hardness: the barrier between a high-conflict couple can become so thick that it can be pierced only by escalating to an extreme level of nastiness. But those are the failure cases, and they shouldn’t distract us from the existence of success cases. In fact, as Jesse Prinz emphasizes, the very fact that there are characteristic ways for anger to go wrong suggests that there is something to the idea that it can go right.

So, I disagree with Bloom’s conception of anger as a sickness, best eradicated, or—if that should prove impossible—lessened. Instead I join Oded Na’aman in thinking that anger is part of a healthy human life, though I also continue to maintain that there is something unhealthy about even the healthiest—that is, most successful—cases of anger.

As an analogy, consider a fever. When you are feverish, you are not healthy, in that you cannot engage in your normal, productive functions. Something is wrong with you; your body is awry; fever is a form of sickness. But a fever is also a healthy immune response to the presence in your body of some kind of infection. If I lost my susceptibility for fever, that would be a sign that things had gone very wrong—it would be a way of getting sicker, not healthier. So, while it is clearly not “best” to be weak, feverish, sleepy, and incapacitated, it might nonetheless be the best option on the table. A fever is a healthy way of being sick. I have much the same view about anger—although here “health” becomes a metaphor, and it is best to speak more directly of whether anger is rational, or the degree to which it is a genuine response to moral reasons.

Bloom questions whether the language of reasons is even appropriate in this context—do people get angry “for reasons” and can someone be “rationally” angry? I would note that we do tend to speak in both of these ways. In fact anger is one of the contexts where we take “reasons-speak” to be not only warranted but required. Consider some perfectly ordinary scenarios where this demand isn’t made. We usually take ourselves to have reasons for what we are doing, but sometimes when someone asks me, “Why are you tapping your fingers like that?” or “Why did you suddenly start running?” I face no strenuous objection if all I have to offer by way of a response is, “No particular reason.” Likewise, I sometimes know why I feel sad, but at other times I just feel sad “for no reason.” In stark contrast, no one considers it acceptable to feel angry at a person for no reason. We may not be able (or wish) to articulate the reason, but we feel called upon—compelled—to try. When it comes to practical reasons—theoretical reasons, such as those involved in syllogistic argumentation, are a different matter—I can think of no attitude more implicated in the practice of giving and receiving them than anger. This is why Rachel Achs is absolutely right to emphasize the “communicative” function of anger.

As a counterpoint to Bloom, who thinks that I make anger sound too rational, consider Na’aman’s argument that anger is in fact more rational—or at any rate less grudging and vengeful—than I claim it to be. I argued that if we have a reason to be angry, that reason never goes away or changes. Na’aman claims, to the contrary, that subsequent events can change the significance of a past action in such a way as to eliminate a reason for being angry. His examples of such significance-changers are apology and regret, which make it the case that my reason for being angry—your betrayal—“was addressed.” But he does not say how apology and regret address someone’s anger over a betrayal. One can, I admit, imagine a funny case in which my anger is not aroused by your betrayal of me—I’m fine with you betraying me!—but simply by the fact that that betrayal has gone unregretted and unapologized for. Once you regret and apologize, you have indeed “addressed” my grievance, and I no longer have any reason to be angry with you. But usually what I’m angry about is not the absence of measures to rectify a wrong but the presence of wrongdoing. And my claim was: because that doesn’t change, it cannot be “addressed.”

Barbara Herman argues along lines similar to Na’aman’s when she proposes that there can be backward causation in the moral domain, and that this is “no more difficult to understand than coming to see a past event as a first step in an ongoing process.” But consider an example of the latter: I see you laying bricks on the ground, thinking you are creating an artwork, and then later I realize you were laying the foundation of a house. I can say, “When I first assessed those bricks as an artwork, I was mistaken, because I didn’t realize they were part of an ongoing process.” The parallel is then: when I became angry at you for betraying me, I was mistaken, because I didn’t realize that was part of an ongoing process that would end with you apologizing.

This doesn’t seem right, and it suggests that backward moral causation is, indeed, much more difficult to understand than coming to see a past event as a first step in an ongoing process.

Those who would rescue forgiveness must face the eternal anger argument head on—and it is a formidable foe.

My point here is not quite the one Elizabeth Bruenig takes me to be making—that holding on to anger is in one’s self-interest—but rather that holding on to anger is inscribed in the logic of anger. Bruenig herself gestures at this logic with her lovely description of anger’s point of view as a “kingdom of damage.” I am arguing that forgiveness has no location in that kingdom, even in cases where one might be selfishly benefitted by forgiving. It is not uncommon for someone to be unwilling to let go of anger even when the incentives for doing so are very great. She doesn’t care about those rewards, or being happy, or the overall satisfaction of her preferences; she cares about one thing—the moral fact that, if I am right, continues to constitute a reason for her to be angry.

I do not deny that there might be some way to articulate what it might mean to change the (significance of the) moral facts, but I think that successfully doing so is much trickier than Herman or Na’aman present it to be. Those who would rescue forgiveness must face the eternal anger argument head on—and it is a formidable foe.

Still, I admit that it is one thing to claim that the reasons for anger are eternal, and another to insist that they underwrite vengeance. Na’aman contends that I overstate the role of vengeance in anger, particularly in the case of those we are closest to. But I would note that we are reluctant to describe practices we ourselves are involved in as “vengeance.” For instance, consider criminal justice. We do not feel that the wrongdoer has been held accountable unless he suffers for what he has done. We make his bad our good, and that is vengeance, even if we prefer to call it “retributive punishment.” Likewise, I would call it “vengeance” when you are fighting with your spouse and deliberately say things you know will hurt them; or when you “punish” them by leaving dirty dishes for them to do; or demand that they perform whatever tasks—cleaning the living room, dealing with the car registration—you know they most hate. “Passive aggression” is one of the terms we use for small acts of vengeance that we prefer not to call “vengeance.” I admit there is an approach to marital disputes that is entirely forward-looking, productive, and cheerful: “Can we do things differently in the future?” But note that such a Cheerfully Productive Spouse prescinds from the project of holding her spouse accountable—for that (backward-looking) project cannot, I maintain, be separated from wanting them to suffer in some way.

What is the force of acknowledging that our anger, even at its healthiest, is nonetheless still sick? If you can’t be good in a bad world, then those whose anger is fully justified—the oppressed, the disenfranchised, and those who crusade angrily on their behalf—are tainted by moral imperfection. Desmond Jagmohan and Myisha Cherry are concerned that in pointing this out, my aim is to blame, judge, or censure such people, or to suggest that it would be “best”—to reprise Bloom’s word—for them to be morally purer than they are.

In fact I agree with Cherry (herself echoing Amia Srinivasan) that “even if anger at racial injustice is counterproductive, it is still appropriate.” And my emphasis on the connection between anger and holding accountable helps explain why this is true. As I see it, our concern with morality runs deeper than it being something we cheerfully hope to bring about: we are attached to it, and this attachment underwrites the importance of (backward-looking) considerations of accountability. Like Cherry, I believe this can be traced to the emotion of love.

What is the force of acknowledging that our anger, even at its healthiest, is nonetheless still sick?

One can’t love a principle. When someone loves justice or equality, her love is, in the first instance, directed at the people—friends, associates, neighbors, fellow citizens, fellow humans—in whom those relations are, or should be, embodied. Love is a form of attachment, and therefore an avenue of vulnerability. Those who crusade for justice and equality by way of love open themselves up to being damaged and wounded in the face of injustice and inequality. Such people make a profound sacrifice, and we cannot be properly grateful to them for it without acknowledging that the wound—which takes the form of a fever of grudging vengeance—is real.