Agnes Callard is right that a retaliatory instinct is often rationally bound up with the way we hold each other morally accountable. But before we join her in bemoaning the impossibility of retaining innocence as a victim of injustice, it is worth scrutinizing the rationale that ties moral responsibility and retribution together. According to the “Argument for Revenge,” retaliating against wrongdoers is a way of “teaching someone a lesson.” But is it always the only way, or the best? 

It seems to me that it depends—on what lesson we wish to teach, what the transgressor already knows, and what sort of admonishment might succeed in getting this particular individual to learn from my reciprocal anger.

Is revenge always the only way, or the best?

Consider some examples. One lesson we might wish to impart on a wrongdoer is not to harm again in the same way. If you steal from me on Monday, I might hope to teach you not to do so again on Tuesday, nor, indeed, to steal from me ever again. Alternatively or additionally, we might hope that the offender comes to understand the wrongness of what she has done: to learn either that her behavior was wrong, what it means for her behavior to have been wrong, which aspects of her behavior made it wrong, and so on. I might wish for you to know, for example, that it is considered disrespectful to flout property rights, or to learn some of the ways that your theft inconvenienced me.

Yet it is far from clear that the best way to teach all these lessons, in every case, is by retaliating against the transgressor. Surely there are instances when alternative methods would be superior? For instance, perhaps in some scenarios a simple request would suffice to prevent the harm from being repeated—and have the added benefit of being much less likely to backfire. Nor will causing a wrongdoer to suffer help her to understand the wrongness of what she has done more deeply in those cases where she already possesses this sort of understanding.

One might defend retaliation against these charges by saying it is essential to convey to a wrongdoer how it felt to be a victim of what she did. Vengeance, on this line of thought, is always required to teach her that.

But this is either a lesson that can’t be taught, or one for which retribution is likewise often unnecessary. If what we wish is for a wrongdoer to know exactly how it felt to experience what she did to us, then we are probably setting ourselves up for disappointment. After all, even retaliating against you with an eye for an eye won’t make you feel what it was like to have the experience of being wronged  as me. If, on the other hand, what we hope for is not complete immersion into our own psyches, but rather just that the people who wrong us come to develop some empathy with us as victims, then, again, retribution may sometimes be a useful tool, but it certainly won’t always be a necessary one. In many cases the people who hurt us are people who have been hurt in the past themselves, and they are fully capable of putting themselves in our shoes by reflecting on their own past experiences. Retaliating may prompt this sort of reflection, but it may also just as well incite defensiveness.

I don’t deny, then, that moral education might sometimes best be served by unleashing retaliatory anger on a wrongdoer. But it seems undeniable that there are plenty of instances in which alternative strategies would serve equally well, if not better, to achieve this end.

If we wish for a wrongdoer to know exactly how it felt, then we are probably setting ourselves up for disappointment.

This analysis makes trouble for Callard’s larger conclusion. If vengeance is really only fruitful in a subset of cases, it seems wrong to conclude that there is reason for revenge to be the way that humans hold one another responsible. At most, it looks like we can say that there is reason for revenge to be a way that some people ca sometimes hold others responsible. This could hardly be described as a situation in which it is impossible for victims of injustice to avoid getting their hands dirty.

There is a way out of this problem, though, because there is a different way to argue that retaliation is the right response to being wronged. Rather than contending that there is something intrinsic to vengeance that makes it especially edifying in any and all cases, we might instead begin with the observation that many humans, as a matter of convention, take revenge to be an appropriate response to wrongdoing. The reason for retaliation would then be the same one that people usually have to enact conventional norms: that conventions have communicative power. Just as one might communicate respect by adhering to certain norms of etiquette, or communicate one’s thoughts by using linguistic conventions, so too does retaliating against those who have offended us communicate something—in this case, that we recognize we have been wronged.

If this is right, then anyone who is wronged does have some reason to retaliate. Avenging oneself is how a person can communicate that she perceives what has been done to her is a wrongdoing. Note that the communicative and educational justifications for vengeance are not necessarily incompatible. Insofar as it is important that this perception of wrongdoing be communicated to the perpetrator, we might still see the justification for vengeance as rooted in the value of educating wrongdoers. But whether or not the wrongdoer really listens to me in a particular instance, or really learns her lesson, may be of minimal importance. Instead, it might matter more to us to convey our perception of how we have been treated to other members of our community, or even just to mark it for ourselves.

Where does this leave us, then, with the question of innocence in the face of injustice? Is it possible, after all, to practice moral accountability without vengeance?

On the one hand, if the norm for responding to wrongdoing is to retaliate, then responding to an objectionable event in some other way isn’t going to as successfully register that event as a wrong. For example, I take it that a powerful reason for discontent with German denazification after World War II was precisely that the absence or lightness of sentencing in many cases was inadequate to communicate the horrific nature of Nazi crimes. (This sort of pressure toward conservatism—reasons not to depart from or try to reform what we already do—arises in any norm-governed communicative context. Deviating from a convention, the worry goes, will fail to have the same communicative effect.)

Is it possible to practice moral accountability without vengeance?

On the other hand, it would seem that we have reason to be more optimistic than Callard is about the prospects of accountability becoming less vengeful. If the main justification for holding people responsible via retribution resides in the communicative power of conventions, then we can expect there to be plenty of instances in which the benefits of developing more humane ways of holding people responsible outweigh whatever value there is in retribution. A conservative might still worry, and rightly so, that violating the convention to retaliate may introduce communicative confusion: if we muddy the semiotic waters by transgressing communicative norms, we may fail to signal what we intend to get across. But absolute clarity doesn’t always have to be our top priority—precisely because the limitations on what we can convey using reformed practices need only be temporary, since the fact that retribution holds so much communicative power stems merely from convention. Once new communicative conventions catch on, they may work just as well, or even better than, the old ones.

The upshot is that, while there is some reason that the affective response to injustice clings to the taste of blood, it doesn’t have to cling quite so tightly.