In the sixth season of HBO’s Game of Thrones, Lancel Lannister demands that Cersei appear before the High Sparrow. She refuses. When a soldier steps forward to take her by force, her enormous bodyguard, the Mountain, intervenes. “Order your man to step aside,” Lancel says, “or there will be violence.” There is a long beat. At last Cersei responds, softly, “I choose violence.”

Agnes Callard chooses violence as well. She defends the value of “vengeance, bloodlust, and limitless violence.” For her, this is what morality is all about. Violence is the cost we pay for goodness.

Any social creature that wasn’t inclined to strike back at threats or acts of harm would be, in a word, a chump.

I think she is right; indeed, hers is the received view in evolutionary approaches to the mind. No serious scholar sees anger as a glitch or an accident. Like every other complex cognitive system, it evolved through natural selection and serves a biological function: it motivates us to defend the interests both of ourselves and of those we care for. Any social creature that wasn’t inclined to strike back at threats or acts of harm would be, in a word, a chump—open to exploitation and cruelty, a loser at survival and reproduction. There is now abundant evidence that punitive impulses are universal in humans, present in young children, and bred in the bone.

This is part of what makes human society possible. Humans are unusual among primates: hundreds of people who have never met can share a ride on an airplane for several hours, say, emerging with all their fingers and toes still attached. As the primatologist Sarah Hrdy points out, the outcome would be very different for a planeload of unacquainted chimps. It is likely that the difference stems from our capacity and inclination for retaliation. We control our violent impulses because we know that others won’t let us get away with them. Ironically, then, what might make our species more social than even our closest evolutionary relatives is that we are more easily pissed off.

Callard is thus on the right track when she speculates, “What if we humans do morality by way of vengeful grudges?” She comes to her conclusion not by way of evolution or game theory, of course, but rather through philosophical reflection and a close reading of Nietzsche, Foucault, and Girard. But this is a pleasing convergence: when people arrive at the same idea from different directions, there is good reason to believe that the idea is right.

So far so good. I am puzzled by another feature of Callard’s argument, though: her repeated claim that being angry isn’t just natural, or universal, or essential to morality—more than that, she says it is rational. It is not that I disagree; I don’t understand what she means.

When we talk about rationality, we often refer to reasoning: drawing conclusions in accordance with the laws of logic. If you think that Socrates is a man and all men are mortal, then it is rational to believe that Socrates is mortal. Sometimes we also talk about not just beliefs but actions as being rational, when they are properly suited to one’s goals. If it is raining outside and you don’t want to get wet, it is rational to carry an umbrella. But I am not sure how any of this extends to feelings. If my partner flirts with someone, is it rational for me to get jealous? Is it rational to be cheered by a sunny day, disgusted at certain sex acts, bored at a faculty meeting, bitter at the success of an enemy? It doesn’t seem that normal standards of rationality apply here.

It doesn’t seem that normal standards of rationality apply here.

Now, if by “rational,” Callard means “working as these systems have evolved to work,” then there is no problem. But some of Callard’s examples aren’t consistent with this interpretation. She suggests that anger is a way of grasping that something has gone wrong in the past and reasons that, since the past doesn’t change, it follows that, “Once you have a reason to be angry, you have a reason to be angry forever.” But anger isn’t just an acknowledgment of a state of affairs; it is also, as Callard herself emphasizes, a motivational state, connected to a desire for vengeance. So, yes, it is rational for me to appreciate that even a tiny wrong never goes away, but this isn’t the same as saying that it is rational for me to want to enact vengeance for it many years later. It probably isn’t.

More generally, it is not clear how Callard’s analysis bears on the question of whether the role anger plays in morality can change over time. It could be that anger was once necessary for the emergence of morality but is obsolete now. After all, a child who learns to count using her fingers will turn into an adult who can count just fine with her hands in her pockets. Or, more plausibly, it could be that some amount of anger is needed for morality to continue to work, but this amount is less, perhaps much less, than we have now.

This last claim is an empirical one, and we have quite a lot of data that bear on it. With the exception of those with severe brain damage, every normal human feels anger. But there is plenty of variation, across individuals as well as societies, in how it is experienced and expressed. Do the easily angered lead better lives than those who are quick to forgive? Do angry people make the best romantic partners? Are cultures of honor, in which male violence against transgressors is a core moral value, the best societies to live in? I think the answer to all of these questions is no, and I wonder whether Callard agrees.

“Emotions are how we humans do morality,” Callard tells us, and this is true—but emotions are not the only way we do morality. There are those who meditate to reduce their anger; there are those who think retributive punishment is rooted in a metaphysical confusion; there are Stoics and Buddhists and utilitarians. Callard is perhaps right that our judgments are inevitably “shrouded by the darkness of our morality system.” But, still, we can argue about morality, revise and defend and extend and challenge our initial prejudices. This too is how humans do morality. After all, what else are we up to right now?