II. Questions

I begin with some examples of Gideon Rosen’s. If there is strict liability in the moral domain, I should think that at least one of these cases would provide us with an instance of it. The first case I’ll call Damaged Goods: “Phyllis accidentally (that is, blamelessly) breaks a plate in a crowded shop.” According to Rosen, “There is obviously a sense in which Phyllis is morally responsible for what she’s done. It falls to her to compensate the shopkeeper, to apologize, and so on,” even though “there’s another sense in which she is not morally responsible: she’s not blameworthy.”^{Footnote 3}

Similar examples abound. Here’s one I’ll call Wine-Oh! While at an office party, Laura accidentally spills red wine on her boss’s expensive white dress. If Rosen’s assessment of Damaged Goods is correct, a similar assessment would seem to apply here; Laura should apologize and pay for the dry cleaning, even if she isn’t even partly to blame for what happened.^{Footnote 4}

An importantly different case cited by Rosen as an instance of what I’m calling strict moral liability is Bad Dog in which “your dog escapes by an improbable route and bites a neighbor, despite your having taken every reasonable precaution to keep him in the house.” About this case Rosen says, “there is a sense in which you are responsible for the injury. He’s your dog, after all, and if someone has to bear the costs of his malignity, it should be you. You should apologize and pay the doctor’s bill.” However, Rosen goes on to acknowledge that, since you took every reasonable precaution to prevent something like this from happening, “there is another sense in which you are not morally responsible for the injury. It’s not your fault; you are not culpable.”^{Footnote 5}

The injurer in each of the preceding three cases is said to be strictly morally liable insofar as she has an obligation to compensate the victim or make reparation, despite not being blameworthy for the injury or the behavior from which it resulted. But, as I hope to make clear, it isn’t obvious that these judgments are correct. Indeed, viewed in a certain light, the claim that blameless agents can be liable for harm resulting from their behavior is initially rather puzzling. And even if you disagree with me about that, I hope you’ll agree that the subsequent discussion brings to light some important questions about the normative foundations of strict moral liability.

Start with Damaged Goods. Why should Phyllis have to pay for the broken plate in that case? Why shouldn’t it instead be figured into the cost of doing business? Accidents do happen. Given that Phyllis isn’t the least bit culpable for breaking the plate, it’s initially difficult to see how her breaking it is relevantly different from the plate being blown off the shelf by a gust of wind. In that case it would be absurd to suppose that Phyllis is obligated to compensate the shopkeeper for the damage. Of course, it’s also true that Phyllis isn’t the agent of the damage in that case, whereas in the actual case she is. But—and this is a crucial question—why should that matter? Why does the fact that she is an agent of the damage make a difference if she isn’t even partly to blame for the damage or for the particular instance of agency from which it resulted?

Think about it this way: if Phyllis had been culpable for breaking the plate, then, other things being equal, there would be no question whether she is obligated to compensate the shopkeeper— she clearly would be.^{Footnote 6} And, if the plate had instead been blown off the shelf by a gust of wind, here too there would be no question whether Phyllis, a random customer who had nothing to do with the breaking of the plate, is obligated to compensate the shopkeeper—she clearly would not be. The actual case, though, is somewhere in between these two. Like the first case but unlike the second, Phyllis’s agency is among the proximal causes of the damage to the shopkeeper’s wares. But like the second and unlike the first, Phyllis isn’t even partly to blame for what happened. So, if Phyllis is strictly liable for the damage, this must have something to do with the fact that her agency is involved in its production, insofar as such involvement is the main similarity between the original version of the case and the version in which she clearly has a compensatory obligation. But what exactly is it about being an agent of harm that could render someone liable absent culpability?

Similar remarks apply to Wine-Oh! Given that Laura isn’t to blame for spilling the wine, it’s unclear why she should have to foot the cleaning bill or buy her boss a new outfit. Granted, she is an agent of the harm, but, again, why should that matter absent culpability on her part? Why shouldn’t the cleaning bill be figured into the vicissitudes of office parties? Accidents and all that.

Questions like these arise in connection with Bad Dog, too, though the agent’s contribution in that case is further removed from the harm than in the previous two cases. Granted, it was your dog that bit the neighbor, but why should that matter if what happened wasn’t even partly your fault? You aren’t liable for the harm other dogs do, except in those cases in which you’re somehow to blame for the harm. Why not say the same thing when it’s your dog? What’s so special about it being yours that renders you liable for what it does even in the absence of culpability on your part?

It’s regrettable, of course, when innocent people suffer misfortunes. But given that the agents featured in these cases aren’t culpable for the relevant harms, it’s initially difficult to see why they should have to bear the burden of making restitution. We’re all subject to bad luck. If an agent isn’t to blame for another’s misfortunes, why should she be especially morally obligated to redress them? In the absence of culpability, why not let losses lie where they fall, just as we would in cases in which the victim’s misfortune results from non-agential causes? No doubt any satisfactory answer to this question will have something to do with the fact that injurer’s activity is among the causes of the damage, but it’s not immediately clear what that something is.

Two additional observations complicate matters further. First, victims can be partly causally responsible for the injury as well, since their agency may also be among the contributing factors. To see this, consider Wine-Oh! again. The boss’s decision in that case to strike up a conversation with Laura may be among the causes of the conspicuous wine stain on the dress. If it is, then the fact that Laura contributed causally to the damage won’t be enough to render her liable, since it does nothing to distinguish her contribution to the injury from the contribution of the victim. Those who believe in strict moral liability and who want to ground it in facts about the injurer’s causal contribution to the harm will thus have to maintain that the injurer’s liability in such cases is grounded not simply in the fact that she contributed to the harm, but rather in the specific contribution she made. But then we’ll want to know what it is about the injurer’s contribution in particular that renders her, and not the victim, liable in the absence of culpability.

Second, there are cases in which a person seemingly has no obligation to redress an injury resulting from his behavior. Case in point, Wrestlers: a collegiate wrestler executes a single-leg takedown on his teammate in practice, a common technique both athletes have drilled a thousand times, only this time something goes horribly wrong. The teammate’s leg gets caught at an awkward angle resulting in a torn ACL, an injury sure to end what looked to be a promising athletic career. Neither wrestler is at fault, let’s suppose. In particular, the wrestler executing the takedown wasn’t negligent or unduly aggressive—he wasn’t showboating or horsing around or trying to prove his dominance or anything like that. It’s just one of those tragic accidents for which no one is to blame. In this case, although the wrestler causally contributed to his teammate’s injury, he arguably doesn’t have an obligation to redress the injury. He seemingly isn’t obligated to help with his teammate’s medical expenses or to compensate the teammate for lost opportunities as he would be if he had been to blame for the injury. But if that’s right, we’ll want to know what separates cases like this from those to which Rosen draws our attention? What is it about the particular causal contribution of the agents in those cases that could render them liable absent blameworthiness?

III. Equitably Distributing Harm

Any attempt to make sense out of strict moral liability that fails to address questions like these is seriously inadequate. To see this, consider the following position, which defends strict moral liability on the grounds that it urges a more equitable distribution of undeserved harms.

Imagine a case of assault in which the assailant isn’t blameworthy for the attack or the resulting harm to the victim (for example, because, through no fault of her own, the assailant was insane). In a case like that, if the victim bears both the physical and financial burden of the assault, he is forced to bear a double burden, which seems doubly unfair. Not only must he suffer the undeserved physical and emotional trauma resulting from the assault, he must bear the financial burden as well. Somewhat more equitable, it would seem, is if the assailant were to take on some of the burden herself by, say, footing the victim’s medical bills or by helping with his convalescence in some other way. The unfairness of the situation would then be distributed more evenly, insofar as the assailant removes what would otherwise be a double undeserved burden borne by her innocent victim. It’s still unfair that the assailant, who by hypothesis is blameless, must make sacrifices to offset the effects of her behavior. But isn’t that better than for her equally blameless victim to have to shoulder those consequences alone? Strict moral liability thus seems to point toward a more equitable distribution of undeserved harm.

This defense of strict moral liability suggests a justification for Rosen’s judgment about Bad Dog. If your neighbor has to pay the doctor’s bill herself, she is stuck bearing a double undeserved burden. Not only must she bear the physical pain of the attack, she must bear the financial costs as well. Somewhat more equitable, it would seem, is for you, the dog’s owner, to take on some of that burden yourself. You can’t bear your neighbor’s pain, obviously, but you can take care of the expenses she incurred as a result of being bitten by your dog, which is preferable, it could be argued, to your equally blameless neighbor having to shoulder the harmful consequences of your dog’s misbehavior all by her lonesome. Here, too, strict moral liability seems to encourage a more equitable distribution of undeserved harms and related burdens.

The position fares less well in other cases though. Take Damaged Goods, for example. If Phyllis pays for or replaces the broken plate, she bears the brunt of the burden. The shopkeeper suffers no setbacks to her own interests in that case other than perhaps the minor inconveniences associated with having to order new merchandise. In cases like this, then, strict moral liability doesn’t point toward a more equitable distribution of undeserved harm, since it implies that the blameless injurer is obligated to bear the brunt of the bad consequences of her injurious behavior. Even worse, consider a case in which a poor man blamelessly breaks an irreplaceable heirloom belonging to his wealthy employer. Demanding that the poor man compensate his employer doesn’t seem to equitably redistribute the burden the employer must bear as much as it does place a new burden on someone else—the poor employee—who doesn’t deserve it and can ill afford to bear it. If the broken heirloom is truly irreplaceable, any form of compensation the employee might provide will do little if anything to offset the burden his employer must bear, but it will place a new and undeserved burden on the employee. Here too strict moral liability fails to provide for a more equitable distribution of undeserved harm. Indeed, in cases like this, not only does strict moral liability not equitably redistribute the harms resulting from injurious behavior, it actually seems to increase the number of undeserved burdens people are obligated to bear.

Response: “strict moral liability is limited to cases like Bad Dog in which making restitution would serve the aim of redistributing undeserved harms more equitably.” Whatever its merits, this response significantly limits the range of cases in which agents are strictly morally liable. It seems to exclude Damaged Goods and Wine-Oh! in particular. The response will thus be unavailable to anyone who thinks that agents in those sorts of cases are strictly morally liable.

There is, moreover, a more pressing difficulty with the suggestion that strict moral liability is grounded in concerns about the equitable distribution of undeserved harm, namely, its inability to explain why it’s the injurer’s responsibility to make restitution.^{Footnote 7} Any suitably situated bystander or other uninvolved party would do. Why should the injurer be especially obligated to help alleviate the victim’s burden? The view presently at issue provides no guidance here and so fails to identify an adequate normative basis for strict moral liability. Resolving this difficulty requires identifying a relevant feature of the injurer’s particular causal contribution to the harm that renders him, rather than a bystander or the victim, morally liable in the absence of culpability.

IV. Causation and Liability

So, what might that feature be? One suggestion is that it’s the injurer’s having initiated a causal chain leading to the harm. Some theorists maintain that “remedial duties are attributed to an agent in the absence of fault if this agent is at the incipient end of a causal chain of events linking her to the harm incurred by another agent.”^{Footnote 8} This would explain why it’s the injurer who is liable for the damages and not a suitably situated bystander or other uninvolved party. It would also often distinguish the injurer’s contribution from that of the victim (often, though not always, as we’ll see). The suggestion thus has something going for it. It’s nevertheless unacceptable for several reasons.

An initial difficulty is that it’s unclear how to individuate causal chains so as to identify their point of origin. If a causes b, b causes c, c causes d, and d causes e, where e is the harm to the victim and d the act of the injurer that contributed to the harm, does the causal chain start at d, or does it go further back to include the causes of the injurer’s behavior? Or further yet, to the causes of the causes of the injurer’s behavior? And how are we to decide the matter? Obviously, if the causal chain stretches far enough back into the injurer’s past (prior to his birth, for example), we won’t be able to say that the injurer “stands at the incipient end” of that causal sequence. The issue becomes all the more complicated when we realize that there are typically multiple independent causally relevant factors that jointly contribute to producing a given harm, all of which have their own distinct and complicated causal histories.

Perhaps it could be argued in response that an agent initiates a causal chain, in the sense relevant to questions of liability, when an action of hers is the first instance of agency in a causal chain terminating in some harm to an innocent victim. But this too is untenable for a number of reasons, not least of which is that there are bound to be numerous instances of agency in the etiology of an injury, many of which are antecedent to the relevant contribution of the injurer.

We might also wonder whether the suggestion at issue simply pushes the question back a level. For what, after all, is the normative significance of initiating a causal chain terminating in harm to another person if the initiating agent isn’t morally to blame for initiating the relevant sequence of events or for the injury in which it terminated? In the absence of a compelling answer to this question, it’s unclear why initiating a causal chain that results in harm to someone else would ground an obligation to make reparation in the absence of culpability.

It’s also unclear whether the suggestion distinguishes in every case the contributions of injurers and victims. In cases in which both the victim and the injurer contribute causally to the harm, there often seems to be no principled reason to privilege the injurer’s contribution as having initiated the pertinent causal sequence. To illustrate, recall that the boss’s decision in Wine-Oh! to strike up a conversation with Laura may be among the causes of the stain on her (the boss’s) dress. Why doesn’t her causal contribution to the incident count as initiating the relevant sequence of events? She initiated the conversation, so why not the causal chain? And, obviously, if it was Laura’s boss and not Laura who initiated the relevant causal sequence, we won’t yet have identified anything about Laura’s contribution to the damage that would distinguish it from that of her boss.

The suggestion that “remedial duties are attributed to an agent in the absence of fault if this agent is at the incipient end of a causal chain of events linking her to the harm incurred by another agent” also seems to run afoul of cases like Wrestlers in which the injurer, while perhaps having initiated the relevant sequence of events, seemingly has no substantial remedial duties to the victim. Unless defenders of strict moral liability are willing to embrace some fairly sweeping conclusions about the extent of our reparatory and compensatory obligations, simply having initiated a chain of events terminating in harm to others isn’t sufficient to ground such obligations.

A variant of this objection arises in certain cases in which the causal chain linking one agent to harm suffered by another is extremely attenuated. There are cases in which the causal chain linking one person’s action to a subsequent harm is so long, complex, or unusual that it becomes extremely counterintuitive to suppose that the person is liable for the resulting harm.^{Footnote 9} Here’s a case I’ll call Morning Joe that illustrates the point. I turn on the coffee pot one morning, which causes a freak, unforeseeable power surge, which causes a breaker on the street light outside to explode, which startles my neighbor, causing him to drop his coffee mug on his wife’s foot, which causes her to yell out in agony, “my foot, my foot,” which histrionics irritate her cantankerous grandfather, Joe, so much that he has a coronary right then and there. My action in this case initiates a chain of events linking me to harm suffered by another, but I surely don’t thereby incur a compensatory obligation. Again, simply being at the incipient end of a causal chain terminating in harm to someone else is insufficient to ground an obligation to redress the harm.

V. Wrong-Forfeits-First

Let’s stipulate that “x wrongs y” if and only if x infringes on y’s rights or violates some other duty or obligation to y. Some theorists contend that remedial obligations are grounded not simply in the fact that the injurer causally contributed to the injury, but in the more specific fact that the injurer did so in a way that wronged the victim.^{Footnote 10} Call their view wrong-forfeits-first. Because it’s possible to blamelessly wrong others, wrong-forfeits-first supports some judgments of strict moral liability, including, arguably, Rosen’s judgment about cases like Damaged Goods and Wine-Oh! The shop owner in Damaged Goods has a right against Phyllis, a right Phyllis infringes, that Phyllis not damage the merchandise. Phyllis has therefore wronged the shop owner. Wrong-forfeits-first thus implies that Phyllis must bear the cost of the broken plate, despite the fact that she isn’t to blame for breaking it. Similarly, in Wine-Oh!, Laura’s boss has a right against Laura, a right upon which Laura infringes, that she not damage the boss’s dress. Laura has thus wronged her boss and so, according to wrong-forfeits-first, must bear the cost of cleaning or replacing the dress.

Wrong-forfeits-first may not justify all judgments of strict moral liability, however. It isn’t clear whether it justifies Rosen’s judgment about Bad Dog, for instance, as it’s far from obvious in that case that you’ve wronged the neighbor who was bitten by your dog. What exactly did you do that wronged her? Own a dog? Had you been negligent in a way that contributed to the injury, things would be different. But that’s not the case; you did exactly what you ought to have done to prevent the incident. It’s thus difficult to see how you’ve wronged your neighbor in this case.

Proponents of wrong-forfeits-first might be willing to accept this result. They might concede that not all judgments of strict liability in the moral domain survive critical reflection. Perhaps Rosen’s judgment about Bad Dog is an example. Alternatively, they might insist that there is no single normative foundation for strict moral liability. Perhaps different instances of the phenomenon are grounded in different ethical considerations depending on the situation. This would certainly be an intellectually less satisfying result, but we can’t dismiss it out of hand.

In any event, wrong-forfeits-first merits further attention if for no other reason than it suggests a normative foundation for at least some judgments of strict moral liability while also potentially explaining much of what is initially puzzling about the phenomenon. It identifies a difference between injurers and uninvolved bystanders that potentially explains why it’s the injurer’s responsibility to make restitution and not the bystander’s: the injurer, though blameless, has nevertheless wronged the victim, whereas the bystander hasn’t. That same fact also distinguishes the injurer’s contribution to the harm from whatever contribution the innocent victim may have made. And, finally, the view is consistent with, and potentially explains, our observation that some blameless injurers aren’t strictly liable for injuries to which they’ve contributed. To see this, consider Wrestlers again. The wrestler in that case contributed to his teammate’s injury, but in doing so didn’t violate any obligation to his teammate, nor did he infringe on his teammate’s rights. He therefore didn’t wrong his teammate, which means that wrong-forfeits-first doesn’t imply that he is obligated to make reparation. Wrong-forfeits-first is therefore consistent with the intuitive judgment that the wrestler has no compensatory obligation to his teammate. If proponents of wrong-forfeits-first were willing to take on the additional claim that the only suitable ground for remedial obligations is that one has wronged the victim, their view would also explain why the wrestler has no obligation to compensate his injured teammate, namely, because he didn’t wrong him.

Still, why think wrong-forfeit-first is true? Why might causing harm in a way that infringes on the victim’s rights or that otherwise constitutes a failure of duty toward the victim ground an obligation to make restitution in the absence of culpability on the part of the injurer?

One possibility has to do with respect for rights. Some proponents of wrong-forfeits-first suggest that we have an obligation to respect the rights of others, and that this in turn requires that we make reparation in the event that we infringe on the rights of others. It follows from these claims that we have an obligation to make reparation to those whose rights we have infringed.^{Footnote 11}

I take for granted that we ought to respect the rights of others. What might not be so obvious is that respecting people’s rights requires that we make reparation in the event that we infringe those rights. Perhaps, though, there’s something to the idea. Consider Damaged Goods again. Phyllis infringes on the shop owner’s rights. But the extent to which Phyllis does so depends on whether she compensates the owner. Having broken the plate, she can’t avoid infringing on the owner’s right to some extent. She can, however, mitigate the extent to which she does so by paying for the damage, thereby preventing much of the harm to the owner and her business interests that would otherwise result. For Phyllis to refuse to do so would seem to exacerbate the infringement of the owner’s rights insofar as it would allow (barring outside intervention) the full set of harmful consequences of the infringement to obtain. Respect for the shop owner’s rights would therefore seem to require that Phyllis make restitution by compensating the owner for the damaged plate.

A slightly different motivation for wrong-forfeits-first can be derived from John Gardner’s continuity thesis, according to which the obligation to make restitution “is a rational echo of the primary obligation [the obligation that x violated in harming y], for it exists to serve, so far as may still be done, the reasons for the primary obligation that was not performed when its performance was due.”^{Footnote 12} Once we have harmed another person in a way that violates an obligation we had toward the person, that original obligation can no longer be fulfilled. But, according to Gardner, we can still satisfy at least some of the reasons that generated the obligation in the first place. “Those reasons,” he explains, “are still with us awaiting satisfaction and since they cannot now be satisfied by performance of [the original] obligation, they call for satisfaction in some other way. They call for next-best satisfaction, the closest to full satisfaction that is still available.” Acting in accordance with the reasons in the next-best way constitutes reparation. Thus, Gardner concludes that “[t]he normal reason why one has an obligation to pay for the losses that one wrongfully occasioned,” according to the continuity thesis, “is that this constitutes the best still-available conformity with, or satisfaction of, the reasons why one had that obligation.^{Footnote 13}

Damaged Goods can again serve as an illustration. Phyllis had an obligation not to damage the shop owner’s wares. Part of the reason for this, no doubt, is that damaging the goods would needlessly set back the owner’s business interests. Having now broken the plate, Phyllis can no longer fulfill that original obligation. She can, however, still satisfy (some of) the reasons that generated that obligation, as she can still prevent the setback of the owner’s business interests by paying for the broken plate. She is therefore obligated to do just that, other things being equal.

The two accounts of reparatory obligation just adumbrated strike me as initially quite compelling, and, as we have seen, can be used to support wrong-forfeits-first. There is, however, a serious worry about the view that needs to be addressed. Wrong-forfeits-first can seem appealing when applied to cases like Damaged Goods or Wine-Oh! in which the injury is fairly minor and the injurer can repair the damage at comparatively little cost. Cases involving more severe harms, however, cast a considerably less favorable light on the view. Consider, for example, a case in which a troublemaker surreptitiously spikes your morning coffee with an aggression-enhancing drug that produces in you an irresistible desire to pummel the next innocent person you see. The troublemaker did this in the hopes that it would result in you assaulting an innocent victim. I round the corner and you proceed to beat me to within an inch of my life. You’ve wronged me in this case, though you’re clearly not blameworthy for doing so. Wrong-forfeits-first thus implies that you must bear the cost of reparation, which in this case would amount to tens of thousands of dollars in medical bills, an unbearable burden for anyone but the extremely wealthy. But that seems excessive, doesn’t it? Must you really go to such extreme lengths to make reparation given that you aren’t even partly to blame for my injuries? I should think not.

We can bring out the worry further by noting a connection between obligations and demands. If you have a moral obligation to do something for me, it would typically be reasonable for me to demand that you fulfill that obligation. But if I know all the facts of the case, it seems entirely unreasonable of me to demand that you fork over tens of thousands of dollars, which suggests that you aren’t, in fact, liable in this case for all the damage you caused.

Note too that there is another agent in this case who, let’s assume, is culpable for my suffering, namely, the troublemaker who gave you the aggression-enhancing drug. If anyone has an obligation to make restitution in this case, it’s him. It’s his fault you assaulted me, so he’s the one who should have to bear the cost of making restitution, not you. This conclusion is supported by the fact that, while it wouldn’t be reasonable of me, knowing the facts of the case, to expect you to cover my medical expenses, it would be reasonable for me to demand that the troublemaker do so. Contrary to what wrong-forfeits-first implies, then, the fact that you’ve wronged me in this case doesn’t ipso facto generate an obligation for you to make restitution.^{Footnote 14}

Proponents of wrong-forfeits-first might try to avoid this difficulty by restricting their view to primary agents of harm, where, very roughly, a primary agent of harm is one who initiates a causal sequence leading to a harm. (This would, in effect, be to combine wrong-forfeits-first with the suggestion considered in Section IV about initiating causal chains that lead to harm.) Working out a more detailed account of what makes someone a primary agent of harm doesn’t promise to be easy, but I’ll assume for now that we have a good enough intuitive grip on the idea to proceed. For example, it seems clear that the troublemaker who slipped you the aggression-enhancing drug would be a primary agent of harm. According to the qualified version of the view, it wouldn’t be you but rather this other person who must make restitution, which seems exactly right.

It’s doubtful, though, whether this restricted version of wrong-forfeits-first is really an improvement on the original position. As I noted earlier, it’s unclear why being the primary agent of harm is normatively significant. What if the troublemaker too had been blameless for his transgression? In that case, you’re both in a similar boat; you’ve both wronged me but aren’t to blame for doing so. So why should the fact that the troublemaker is the primary agent of harm render him liable in the absence of culpability? And even if there were a compelling answer to that question, the initial difficulty really hasn’t been dealt with. To see this, suppose you beat me to within an inch of my life not because some troublemaker drugged you but because, through no one’s fault, you went insane. In this version of the case, you are the primary agent of harm. But given that you aren’t blameworthy for what you did, it still seems excessive to suppose that you’re obligated to make full restitution. Again, if I knew all the pertinent details of the case, it would be unreasonable to expect you to shell out tens of thousands of dollars for my medical expenses.

Other examples pose similar difficulties for the restricted version of wrong-forfeits-first. Suppose we take the “eye-for-an-eye” idea literally for a moment. Imagine that eyes are transplantable, that Jack wrongfully pokes out Jill’s, and that there are no spare eyes available. In that case, if Jack were culpable for poking out Jill’s eyes, then I should think he would be obligated to give Jill his beautiful baby blues, and, given the connection between obligations and demands, that it would be reasonable for Jill to demand that Jack fulfill this obligation. But suppose Jack is completely blameless for what happened to Jill. (Perhaps he too was insane at the time of the incident, through no fault of his own.) Would Jack still be obligated to sacrifice his vision for the sake of restoring Jill’s? Wrong-forfeits-first implies that he would, his complete lack of blameworthiness notwithstanding. This is a hard teaching indeed.

If you think this last case too outré, consider a homelier one in which “a poor man [blamelessly] breaks a rich man’s expensive vase.” We may imagine that, “it would cost the poor man years of unremitting toil to pay for the damage at market prices, yet for the rich man the loss is not of great importance and it is replaceable without too much difficulty.”^{Footnote 15} Does the poor man in this example really owe his wealthy victim full repayment, given that he isn’t even partly to blame for what happened? Must he really spend years of unremitting toil to pay for the damage? In its current form, wrong-forfeits-first implies—perversely, it seems to me—that he does.

Richard Swinburne suggests that this difficulty can be circumvented if we understand the obligation to make restitution “as the requirement that [the injurer] pay the same proportion of his resources as the victim has lost of his.”^{Footnote 16} Call this the “proportionality understanding” of the obligation. The proportionality understanding yields the right result in this case, as it would make the burden of making restitution bearable for the poor man. Because it would cost the rich man only a small proportion of his resources to replace the vase, the poor man, according to the proportionality understanding, is only obligated to contribute a correspondingly small proportion of his resources to redress the harm. In other cases, however, the proportionality understanding does little to ameliorate the troubling implications of wrong-forfeits-first.

Suppose that the rich man’s vase was one of a kind, a gift given by his late and much beloved father, irreplaceable both in terms of its craftsmanship and the place it held in the rich man’s affections. Alternatively, we may suppose that the vase was just extremely expensive and that the rich man isn’t that wealthy, so that it really would cost him an enormous proportion of his financial resources to replace the thing. In either case, wrong-forfeits-first implies that the poor man really must spend “years of unremitting toil” to compensate the rich man, and this is so even when we understand the obligation as the requirement that the poor man pay the same proportion of his resources as the rich man lost of his. Consider, too, the implications of this understanding of wrong-forfeits-first for the story about Jack and Jill. The proportionality understanding still has the troubling implication that Jack must give up his eyes so that Jill can regain her vision. The proportionality understanding thus provides little solace to defenders of wrong-forfeits first.

The same is true of what we might call the pro tanto understanding of wrong-forfeits-first, according to which the obligation to make restitution is pro tanto rather than all-things- considered.^{Footnote 17} This is by far the most plausible reading of wrong-forfeits-first. Thus understood, the view is somewhat less demanding insofar as it allows the obligation to make restitution to be overridden by competing ethical considerations. Ultimately, though, this understanding of the view doesn’t avoid the sorts of worries I’ve been advancing, since we could undoubtedly construct versions of the preceding cases in which we would be hard pressed to identify considerations that would mitigate or override the obligation to make restitution specified by wrong-forfeits-first.

It’s doubtful that we’ll be able to find a version of wrong-forfeits-first that doesn’t have the implications to which I’ve drawn our attention. In light of this, I suggest that the best move for proponents of the view is to embrace those implications. Let me briefly sketch one strategy they might pursue for making this response a bit more palatable than it might initially seem.

Sometimes we get extremely morally unlucky in that we find ourselves, through no fault of our own, in circumstances in which doing the right thing demands nontrivial sacrifices on our part. To illustrate the point, consider a variation on a familiar case. Delilah is walking to her wedding when she sees a child drowning in a shallow pond. She realizes that she has an all-things- considered obligation to wade in and rescue the child even though it will be the ruin of her beautiful wedding gown, which, in addition to being worth a pretty penny, also has significant sentimental value, as it’s the very same dress her mother wore when she married Delilah’s father. Delilah, it seems, has had some really bad moral luck in this case insofar as she finds herself, through no fault of her own, in a situation in which doing the morally right thing demands a fairly hefty financial and emotional sacrifice on her part. Unfortunately, though, that’s life. Doing the right thing is sometimes pretty easy, requiring little if any sacrifice on our part. Other times, though, it’s just really damn hard, requiring us to make extensive sacrifices to our interests.

The fact that we sometimes unluckily find ourselves in situations in which doing the right thing requires nontrivial and even extraordinary sacrifice on our part may go some way to ameliorating the troubling implications of wrong-forfeits-first. We can perhaps view the sorts of cases we have been considering as instances of this sort of bad moral fortune. Proponents of wrong- forfeits-first could argue that, like Delilah, you, Phyllis, Laura, Jack, and the poor man have all been unfortunate enough to find yourselves in a situation in which, through no fault of your own, you must make sacrifices in order to do what’s morally required of you. It’s unfair that you must do so, but the unfairness, they could argue, is of a not unfamiliar sort in the ethical life.^{Footnote 18}

Wrong-forfeits-first is a hard pill to swallow, as I’ve been at pains to illustrate. But, as I’ve also attempted to show, it isn’t obviously indefensible either. So, what to make of the view? For my own part, I’m not entirely sure. Perhaps we should choke it down in the end, but if you’re unable to do so, it’s worth noting that there is at least one other route to strict moral liability that we’ve yet to consider and which seems to me worth considering. Let’s take a brief look at it.

VI. Apologies and Liability

Consider the following case. Overwhelmed by pressures at home and work, you forget your best friend’s birthday. After carefully examining your conscience, you come to believe—correctly let’s assume—that you are entirely blameless in the matter. Your forgetfulness wasn’t a result of lack of due regard for your friend; you were simply overwhelmed by the unexpected and unusually taxing demands on your time and attention (though this isn’t immediately obvious to others, your friend included). It seems clear that you nevertheless owe your friend an apology and that you should offer to make it up to her somehow, your blamelessness notwithstanding. A failure to apologize would display a lack of concern for your friend and would, for that reason, pose a further threat to the relationship, in addition to whatever damage your forgetfulness has already done. An apology is thus called for, despite the fact that you aren’t blameworthy for what you did.^{Footnote 19}

As cases like this illustrate, when we injure others in a way that evinces, or could reasonably be suspected to evince, ill will or lack of due regard, we have an obligation to apologize for the injury even if we aren’t to blame for it. This obligation, it seems, is grounded in the value of repairing and maintaining relationships that are damaged or threatened by our injurious behavior.^{Footnote 20} Apologizing in such cases is essential to protecting the relationship from further harm (for example, by reassuring the victim of the injurer’s good will going forward), and will, in many cases at least, go some way to repairing whatever damage to the relationship may already have been done.^{Footnote 21} But how, exactly, do apologies do this? The answer, I suggest, lies in their communicative nature. Apologies, as I understand them, are essentially communicative acts whereby we express regret/remorse for our behavior and any harm resulting from it, repudiate whatever unjustified ill will or lack of due regard was manifested in, or might have been suggested by, our behavior, and reaffirm our commitment to the victim and to the values at stake. Call the regret, repudiation, and reaffirmation at issue the “apologetic attitudes.” What a successful apology does is to communicate those attitudes to others, most notably to those we’ve offended or otherwise injured.

The precise apologetic attitudes an injurer endeavors to communicate when apologizing will differ depending on whether the injury manifested an objectionable quality of will on the part of the injurer. If it did, then the repudiation and reaffirmation communicated by a sincere apology will amount to repentance and the apology an expression thereof. But not all apologies are an expression of repentance. The case in which you forget your best friend’s birthday again provides a helpful illustration. Because you had no ill will or lack of due regard for your friend, you seemingly have nothing of which to repent (though you do, of course, have something to regret, namely, forgetting your friend’s birthday). Repentance involves an intention to change for the better, but, given that there was no ill will or lack of due regard on your part for your friend, no moral change on your part is necessary. You nevertheless have an obligation to apologize, as we have seen. What your apology would communicate, presumably, isn’t repentance, since repentance seems out of the question here, but rather a rejection of whatever objectionable attitudes and values might have been suggested by your behavior, together with a reaffirmation of your commitment to the friendship. We might summarize these points as follows: while all apologies express a repudiation of objectionable attitudes and values, not all repudiations amount to repentance.^{Footnote 22}

Successful apologies, it seems clear, often require more than simply saying the words, “I’m sorry.” This will be especially true in cases in which the injurer and victim aren’t well acquainted or in which the harm is quite serious. Talk is cheap. It’s all too easy to say, “I apologize,” without really meaning it. But how can other people, especially victims, know whether a verbal apology is sincere? The answer is that the injurer must do something else at his own expense to substantiate the apology. This signals better than anything else that the apology is genuine.^{Footnote 23}

But what might this further act be? A plausible suggestion is that the injurer must do enough toward making restitution to demonstrate the sincerity of his apology. There are, of course, other ways a person might provide evidence of his sincerity. Various forms of self-flagellation might do the trick, for example. But that would likely undercut the apology in another way by placing too much emphasis on the injurer and not enough on the victim, indicating that the injurer’s focus is misplaced. Making reparation, by contrast, suggests that the injurer is appropriately focused on the victim. It provides evidence (defeasible evidence, to be sure) not only of the sincerity of his apology, but also that he is sufficiently concerned about the well-being of the victim.

According to the view presently at issue, injurers often have an obligation to apologize to their victims for the harm they’ve caused and to communicate the sincerity of their apology by providing the victims with some sort of restitution, and this is so regardless of whether the injurers are at all blameworthy. If that’s right, it should be clear that injurers will sometimes be strictly liable for at least a portion of the damage resulting from their injurious behavior. To see this more clearly, let’s apply these ideas to the three examples with which we began our discussion.

According to the view of apologies on offer, Phyllis should apologize to the shopkeeper for breaking the plate if for no other reason than to disavow any ill will or lack of due regard for the shopkeeper that Phyllis’s behavior might have suggested, and offering to pay for the broken plate is the most obvious way for Phyllis to demonstrate the sincerity of her apology. A similar line of reasoning supports the conclusion that, in Wine-Oh!, Laura should pay for her boss’s dry cleaning or, if the dress is ruined, buy the boss a replacement. Laura owes the hostess an apology, not because she’s blameworthy, but rather as a means of repudiating any lack of due regard that might have been suggested by the incident, and, again, making reparation (for instance, by having the dress cleaned) is the most obvious way for her to demonstrate the sincerity of her apology.

Matters are less clear in Bad Dog, but here too I think the argument from apologies supports the conclusion that you should at least offer to pay the neighbor’s doctor bill. Although you took every reasonable precaution to prevent your dog from biting others, your neighbor probably won’t be aware of this, and any attempt on your part to establish your innocence is liable to come off as unsympathetic and self-serving. Apologizing to your neighbor and demonstrating the sincerity of your apology by offering to pay for the doctor’s bill is perhaps the only way in a case like this for you to communicate that, in fact, you have sufficient regard for your neighbor’s well-being.

Not only does this view of apologies provide a rationale for common judgments of strict moral liability, it also suggests compelling answers to our other questions about the phenomenon. It explains why it’s the injurer’s responsibility to make restitution and not a suitably situated bystander: whereas the injurer did something that evinced, or might reasonably be suspected of evincing, an objectionable quality of will and has thereby done something to damage or threaten the relationship with the victim, the bystander hasn’t. The injurer thus finds himself with an obligation to apologize, whereas the bystander doesn’t. These same facts about the injurer’s contribution to the injury also distinguishes her contribution from whatever causal role the victim may have played in the incident. Finally, the argument suggests an explanation for why a person sometimes has no obligation to redress an injury resulting from his behavior. In cases like Wrestlers or Morning Joe where it’s clear to everyone that the injurious behavior didn’t evince, and couldn’t reasonably be suspected of evincing, an objectionable quality of will, there is no threat to the relationship between the injurer and victim, and thus no apology is called for.

A feature of the argument from apologies that I’ve yet to highlight also enables it to avoid the troubling implications of wrong-forfeits-first. I’ve suggested that injurers often have an obligation to apologize for the harm they’ve caused, and that they must do enough toward making restitution to demonstrate the sincerity of their apology. But what counts as “enough”? The answer will surely vary along a number of dimensions, including the relationship of the injurer to the victim and the nature and severity of the harm. In cases in which the harm is minimal, “enough” may amount to full reparation or the equivalent in compensation. If I knock over your coffee, that’s presumably a minor inconvenience for you, and the least I could do to substantiate my apology, and certainly the most I’m required to do, is to buy you a fresh cup of Joe. A minor injury such as this thus requires full restitution, but only because restitution requires such a minor gesture on the part of the injurer that anything less would seem to demonstrate that the injurer’s apology isn’t sincere. If I apologize for knocking over your coffee but refuse to get you a fresh cup, my refusal will undoubtedly undercut my apology insofar as it suggests that I’m not really sorry for what I did. In more serious cases, however, the injurer will often be able to communicate the sincerity of his apology without making full restitution. In those cases, “enough” may fall well shy of what it would take to restore the status quo ante or fully compensate the victim. Whether the injurer has any further liability in such cases will depend on whether he is culpable for the injury in question.

This feature of the argument from apologies enables it to avoid the troubling implications of wrong-forfeits-first. We can see this by returning to some of the problematic cases discussed in Section V. Given that the poor man from our earlier example isn’t to blame for breaking the rich man’s expensive vase, he needn’t spend the years of unremitting toil it would take to fully compensate the rich man for his loss; he need only do enough to demonstrate that he is truly sorry for what he did. This will presumably require some sacrifice on his part, for otherwise we might not think his apology sincere. But presumably it won’t require replacing the vase at market prices. Similar remarks apply to the case in which Jack blamelessly pokes out Jill’s eyes, and to the case in which, owing to insanity for which you aren’t to blame, you beat me to within an inch of my life. In both cases, the injurer should apologize and offer something by way of reparation to substantiate the apology. But given that the offender isn’t even partly to blame for the harm or the behavior from which it resulted, he or she need only do enough to show that the apology is, in fact, sincere, which, given the severity of the harm involved in these scenarios, will presumably fall significantly short of what it would take to fully compensate the victims for the horrible injuries they have suffered.

This isn’t to say, however, that the ethic of apologies and reparation just adumbrated won’t prove demanding at times. On the contrary, it certainly will, for the greater the harm, the more that will be required of the injurer to authenticate his apology, other things being equal. If you knock over my coffee, buying me a new cup will suffice to show that you’re sorry. But if you beat me to within an inch of my life, things will of course be very different. In that case, simply buying me a cup of coffee won’t cut it and, indeed, would likely undermine your apology, as it would seem to trivialize the harm I’ve suffered, suggesting that you don’t appreciate the gravity of what you did or perhaps just don’t care. No, to substantiate your apology in that case, much more is required of you. How much? That’s a difficult question to which I won’t venture an answer and which, as I suggested earlier, will likely vary depending on any number of other factors. No doubt it will require more than a trivial gesture on your part, but presumably it will fall short of full restitution. The important thing to see for present purposes is that this limited form of strict moral liability makes substantial demands of us while avoiding the extremes of wrong-forfeits-first.

VII. Conclusion

Other people aren’t generally obligated to redress my misfortunes unless they are to blame for them. Yet believers in strict moral liability claim that we are, or can be, obligated to redress, to some extent at least, injuries to which we have contributed regardless of whether we are blameworthy for those injuries or our contribution to them. But why should that be so? What could ground an obligation to make restitution in the absence of culpability on the part of the injurer?

I’ve canvassed a number of possible answers to this question. Among the most promising is wrong-forfeits-first, according to which the obligation to make restitution is grounded in the fact that, in injuring the victim, the injurer has somehow wronged the victim. Since we can wrong others without being culpable for doing so, this view entails the possibility of strict moral liability. Wrong-forfeits-first follows from several extant accounts of reparative obligations and, if successful, explains much of what is initially puzzling about strict moral liability. A downside of the view, however, is its extreme demandingness, though, as we’ve seen, it isn’t entirely obvious that this renders the view unacceptable. But even if we reject the likes of wrong-forfeits-first, I’ve argued we should still acknowledge a limited form of strict moral liability emerging from the obligation to apologize for injuring others in ways that could potentially damage our relationships with them. This view, too, addresses the main puzzles about strict moral liability, but does so in a way that avoids some of the difficulties facing wrong-forfeits-first. Readers who find themselves moved by those difficulties may therefore find the view more agreeable than wrong-forfeits-first.