I. Introduction
We are sometimes required to make up for what we did. It is something we are told from the beginning of our moral education: if you do something bad to others and do not make up for it, those others may have a legitimate grievance against you. And if an authority regulates the relationship between you and them—if you are a child under the control of teachers at school, for example, or parents in the playground—those others can legitimately complain to the authority and demand that they enforce your correction.
Philosophers have commented widely on this received moral wisdom. For instance, Aristotle's Nicomachean Ethics is credited with first identifying an entire branch of justice—corrective justice—the core content of which is the appropriate reparation of wrongs. Moreover, the regulation of reparations is one of the main branches of private law.Footnote 1
So we have the basic understanding of a corrective duty—a duty to make up for a wrong that somebody has inflicted on another—which pervades a host of social practices. But what does it mean to inflict a wrong on another—as opposed, say, to commit a generic moral wrong that may not wrong anyone specifically? And what does it mean to correct a wrong?
The answer to the second question may appear trivial: as both Aristotle and ordinary moral wisdom agree, correcting a wrong means repairing the damages one has created. That means making sure that the general condition in which victims find themselves—their financial situation, possession of goods, honor, and so on—is brought back to where it was, roughly, before the wrong, or to where it would be were it not for the occurrence of the wrong.
As some in the literature have already noticed, however, the reduction of corrective duties to duties of damage reparation is as common as it is pernicious.Footnote 2 First, it does not explain why certain acts count as wrongs against others and trigger demands for corrections. Second, it gives the wrongful party an overly limited set of options to correct the wrong. What if repairing the harm is beyond the ability of the wrongful actor? Are wrongful parties then kept in perpetual debt to the victim?
We are back to the question about what it means for an agent to wrong another. One tradition, inspired by Kantian ethics, has answered by reference to rights. I wrong another when I violate their entitlement either to some spaces that belong to them—including their body—or to a sphere of choice over which they alone have authority. More recently, other theorists have argued that there is nothing special about a duty to correct—it is nothing other than a continuation in other terms of the specific duty wrongdoers breached when they wronged the victim. Now that satisfying that duty's primary injunction has become impossible, wrongful parties can still comply with the duty in a somewhat oblique but ultimately satisfactory manner by ensuring that the adverse effects of one's wronging on the victim's wellbeing are either cancelled or at least mitigated.Footnote 3
Both approaches are, as I will show, unsatisfactory. The Kantian approach forces us to adopt a conception of rights that is either too abstract and uninformative or too dependent on the vagaries of positive law. On the other hand, the continuation approach wrongly assumes that the most paradigmatic cases of corrections are attempts at doing what one should have done before.
I will argue instead that what stands in need of correction is a failure to accord adequate concern to the victim's status as a moral agent. It is because wrongdoers have failed to treat others’ moral status as a source of stringent, other-regarding reasons for action that they ought now to correct. I call this the moral neglect account.
The account I offer can identify the conceptual ground of corrective duties—that is, the ultimate reason in virtue of which those who wrong others are liable to corrective duties. Unlike previous accounts, moreover, it can help fill out the content of corrective duties. Specifically, the moral neglect account holds that corrective duties share one general aim: to counterbalance moral neglect. As moral neglect consists of a failure to treat others as a source of stringent constraints on one's action, wrongdoers can only correct their wronging if they demonstrate to the victim that they take their status as a moral peer seriously enough to let it determine, at least in part, the way they act; that is the point of corrections.
But what about the damages the victim suffered because of the wronging? Surely, what victims need—above and beyond a demonstration of respect—is some form of support that may help them enjoy the same level of wellbeing they enjoyed before the damage. And legal practices usually associated with corrective justice—such as the entire doctrine of liability for damages in tort law—attempt to do exactly that. Here is where my account leads to a kind of conceptual revolution.Footnote 4 The victim's complaint that their wellbeing was impacted negatively by a wrongful act, I argue in the final section, gains strength and plausibility if put in terms of distributive, rather than corrective, justice. It is unfair, distributively, that the victim has to shoulder the adverse effects of an injustice they did nothing to bring about. This does not entail, I conclude, that corrective justice has no role to play in deciding who should pay to redress the damage; however, it does entail that corrective justice cannot give us all the answers in response to that essentially distributive question.
The argument proceeds as follows. In Sections II, III, and IV, I present three alternative accounts of corrective duties: the restorative paradigm we first retrieve in Aristotle; the Kantian rights-based view; and the continuity approach. I introduce the moral neglect account in Section V and demonstrate how, unlike previous accounts, it gives us the answers we were looking for regarding both the ground and the content of corrective duties. Finally, in Section VI, I recapitulate the relationship between corrective and distributive justice in harm redress.
II. Corrections as Reparations: Bring Back Things to How They Were
I start with an example borrowed from the literature.
Mick Jagger. Driving along in my Ford Focus, I negligently crash into Mick Jagger, breaking his leg and totaling his new Rolls Royce. I am ruined. Mick Jagger has to cancel his South American tour with the Rolling Stones, causing him a loss of, oh, let's say £2 million. On top of that, damages for the destruction of his Rolls will net him another £500,000 or so.Footnote 5
Something is puzzling in the scope of the obligation. Why should I be required to repay the Rolls Royce when it was Jagger's decision to drive a luxury car? Isn't driving a luxury car morally equivalent to other activities that impose risks on the perpetrator, such as parachuting or bungee jumping? Certainly, if I have been negligent in the scenario, I have not been more negligent than if I had crashed into Roy Jagger (a distant cousin), a pensioner with no significant work commitments who was about to have his old car wrecked. Yet repairing the entire cost of the accident, even when it is as ridiculously high as in this case, is the liability the law imposes on me as the negligent actor, at least under most jurisdictions.Footnote 6
The suggestion that corrective duties have to do, at the bottom, with reparations goes back to Aristotle's presentation of corrective justice as the “mean between loss and gain.”Footnote 7 “What the judge does,” Aristotle says, “is restore equality. It is as if a line were divided into two unequal parts; the judge takes away that by which the greater segment exceeds the half of the line and adds it to the lesser segment.”Footnote 8
There is something illuminated by the line analogy, and it is an aspect of corrective justice that tort law scholars are particularly eager to emphasize, mainly to resist the tendency, typical of the law and economics movement, to see legal duties of reparations as sensitive to requirements of social efficiency.Footnote 9 Whatever demands corrective justice places on individuals, they will be grounded in the relationship such individuals have with the specific other they have wronged. Corrective duties are a subset of directed duties—the duties we generally owe to others—as they are, in a way, “doubly directed”: not only are they owed to specific individuals but they are owed from one individual to another.Footnote 10 As such, they isolate an independent “order of right”Footnote 11 within which the deliberation of one agent is partly captured by the other; one agent (the right-holder) can require that the other (the duty-bearer) act in a certain way, and can further and unilaterally decide to waive such requirements or to forgive their breach.
What Aristotle's view helps to flag is the bipolarity or correlativity of corrective duties.Footnote 12 It offers us an account of the form of corrective justice, if by that we mean the most abstract characterization of the domain that corrective justice has jurisdiction over and how it construes the relationship between the parties. No other agent partakes in this relationship because the injustice suffered by the victim is equivalent to the injustice perpetrated by the wrongdoer; they are one and the same thing. Because the injustice is confined within the wrongdoer–victim relationship, the corrective duty cannot extend beyond the two participants.
Beyond offering a criterion that we can use to isolate the abstract form of corrective justice, the correlativity model stops being helpful. Going back to the Jagger example, the correlativity model tells us that the perpetrator alone is responsible for the original injustice and liable to corrective duties. But what is the original injustice? Here, “correlativity” gives a very limited answer. It tells, specifically, that the wrongdoer did something wrong that they could have avoided doing—if it were not so, we could not say that they had disrupted an order of justice previously existing between them and the victim. This is important as it shows that, if we want the law to be sensitive to corrective justice, we cannot treat the tortfeasor as somebody who found herself in the wrong place at the wrong time and is now asked to pay the damage just because it is the most socially efficient thing to do.Footnote 13 The Aristotelian account gives us, in sum, one necessary criterion for the identification of the agent liable to corrective duties: they must be someone who acted wrongfully towards the victim. But this still tells us nothing about either what the wrongdoer did that they should not have done or what they should do now to correct the wrong.
One could advance the claim that, since the original injustice disrupted an “order” existing between victim and wrongdoer, it is now the wrongdoer's responsibility to restore that order. But accepting this purely restorative understanding of corrective duties leads to two implausible conclusions.
The first concerns the necessity of returning to the status quo. What is so special about the state in which we find ourselves before we are wronged that the wrongdoers owe us to restore it whenever they wrong us? In general, we are not entitled, legally or morally, to the level of welfare or the economic condition we occupy at a particular time. Only the most radical versions of luck egalitarianism support the idea that somebody ought to be compensated every time they are affected negatively by brute luck. And the law does not say that. If I lose consciousness while driving and crash my car against a tree, it is my money—or the money of my insurance—I will need to repair the car. Yet there is a significant parallelism, to which I shall return in the final section, between the case where I suffer a damage out of brute bad luck and the one where I am damaged because others wronged me. In neither case am I responsible for what happened to me. Yet only in cases of wronging do we take the unfortunate to be entitled to demand that the status quo be restored.
Consider, moreover, material damages that are too big to repair for the specific wrongdoer who caused them. Insurances are a necessarily imperfect solution to this problem for the double reason that people cannot pay monstrous insurance fees to pay for all consequences of their actions and that insurance companies need to place limits on their coverage if they do not want to go bust too easily. What is the wrongdoer required to do if it is outside of their possibilities to repair the damage? Are they kept in perpetual debt towards the victim, to the point where the latter may feel entitled to treat all their resources and maybe their body as collateral that they can now freely dispose of? The Aristotelian account cannot answer these questions. Nor, as I will show in the following sections, can any account that assumes we can deploy the conceptual resources of corrective justice, and them alone, to solve what is in essence a distributive question.
Let me summarize the results so far. The kernel of truth in the Aristotelian model is the idea that corrective justice is an inherently relational domain and that its requirements exhibit a relational form: any meaningful correction is correction from one agent who has wronged another to the victim of the wrong. But this formal model is uninformative regarding two fundamental questions surrounding corrective justice. One concerns the ground of corrective duties. I am here using “ground” in the technical, albeit still deeply disputed, sense of contemporary metaphysics, which refers to the factor that explains more than any other why something occurs.Footnote 14 Because of its focus on the pure form of corrective justice—on what happens, in other words, when corrective justice is already activated—the Aristotelian account somehow presupposes an answer, without offering one, to the question about what grounds wronging and corrective duties. The other question the Aristotelian model fails to answer concerns the content of the corrective duty. The idea according to which corrective duties are reducible to duties of status quo restoration cannot itself be derived from the correlativity model; moreover, we have independent reasons to question its plausibility. In the next two sections, I will consider one account that seems promising regarding the ground and one regarding the content question.
III. Rights, Wrongs, and Corrections: The Kantian Account
One tradition that attempts to answer the ground question takes inspiration from Kantian ethics. It affirms that the ground of corrections is the violation of some right of the victim or, more specifically, of the exclusive entitlement one agent has over their life which falls under the rubric of the “universal right to freedom.” Kant's Universal Principle of Right holds that “if … my action or my condition generally can coexist with the freedom of everyone in accordance with a universal law, whoever hinders me in it does me wrong; for this hindrance (resistance) cannot coexist with freedom in accordance with universal law.”Footnote 15
How can I understand, however, whether what I am doing is compatible with the freedom of others? In his magisterial interpretation of Kant's doctrine of right, Arthur Ripstein answers this question starting from the distinction between persons and things: “A person is a being capable of setting his or her own purposes, while a thing is something that can be used in pursuit of purposes.”Footnote 16 As independence consists in the ability to both set oneself purposes and use one's means to try to realize them, according to Ripstein, I can interfere with another's independence and thus wrong them, “either by drawing that person into purposes that she has not chosen or by depriving her of her means.”Footnote 17
It seems we have an answer to the problem of determining the difference between an act that wrongs others, such as driving negligently and crashing into another's car, and an act that merely affects another negatively, such as beating another in a fair competition. When another uses their means legitimately but in a way that affects you negatively, “they don't interfere with your independence … They just change the world in ways that make your means useless for the particular purpose you would have set.”Footnote 18
One initial problem we might have with this account is that it gives us a somewhat awkward redescription of the Jagger scenario. When I crash into Jagger's car, I am not putting Jagger's car to my use any more than if I had crashed into it to save a pedestrian. So if there is a morally salient difference between the two scenarios, it cannot be explained in terms of appropriation of another's means. Similarly, I use another's means—this time more literally—both if I gatecrash into another's house out of petty curiosity or do so out of necessity, as in Joel Feinberg's example of the hiker forced to enter a cabin to take shelter during a snowstorm.Footnote 19
One may respond by accepting that the two cases are morally alike or, more plausibly, that a general theory of rights and wrongs can treat the two cases alike even if morality at large is bound to recognize differences.Footnote 20 A more sophisticated response (in line with Ripstein's more recent reflections) is the following: I put others’ means to my use not only when I generically appropriate them, but also when I subject them to unreasonable risk. In Private Wrongs, Ripstein indeed distinguishes between torts based on use (such as trespass), which are regulated by strict liability, and torts based on damage, where a person uses their own possession in a way that poses more than trivial risks on other persons and their belongings.Footnote 21
So far so good with the ground question. What about the content question? In answering it, Kantian authors surprisingly fall back on the reparatory paradigm. This is where the crucial problems begin as the move seems to contradict the very idea that wronging means violating another's rights. Whether a violation of rights produces damages, and to what extent, is generally independent of the degree of the violation.Footnote 22 Let's keep assuming, for instance, that one good way to describe what happened between Jagger and me is that I have violated his right to independence by subjecting his car to unreasonable risk. It would be bizarre to claim that, by causing £500,000 worth of damage, I have violated Jagger's right doubly than if I had merely produced £250,000 worth of damage. Even more absurd would be to claim that I have violated another's right more by crashing into a Rolls Royce than by crashing into a cheaper car. What is missing is an explanation of why violating somebody's right to freedom makes the violator liable not merely to some form of (yet to be specified) correction, but to the reparation of all the adverse consequences of their violation, including those that depend largely on sheer luck.
Another problem concerns the definition of what is exclusively mine to use. Kantian authors agree that we cannot derive a conception of the specific rightful entitlements that individuals possess from the abstract conception of a right. Ernest Weinrib is adamant about this:
The juridical conception of corrective justice does not proceed by postulating a conception of agency and then deriving the theory of private law from it. Rather, the juridical conception always works backward from the doctrines and institutions of private law to the most pervasive abstractions implicit in it.Footnote 23
If Weinrib is right that the Kantian (“juridical”) conception of corrective duties presupposes a specific distribution of entitlements in private law, it seems the same conception has no resources to criticize any specific articulation of corrective legal duties so long as such articulation conforms to the very abstract idea that what calls for correction is an interference of some individual with some other individual's use of their means or the choice of their purposes. Many legal systems can conform to this abstract model, including those in which some individuals own nothing beyond their bodies.
Admittedly, some commentators hold that the Kantian conception of right can only be consistent with a distribution of property in which at least the kind of poverty that would make some people dependent on others is eliminated.Footnote 24 Hence, the Kantian state, although not one of distributive equality, is one where the distribution of private property is carefully monitored to avoid at least the most odious forms of poverty. But is this enough to rescue the idea, transmitted from the Aristotelian to the Kantian model, that wronging entitles victims to the restoration of the status quo?
I think not, for two reasons. One is that it is unclear why those fortunate enough to own costly objects such as a luxury car are entitled to complete redress in case of damage to their property, calculated considering the pre-damage market value, rather than a fixed sum that corresponds to the average reparation costs for that category of object (which means, for Jagger, a sum corresponding to the average reparation costs for that type of car accident, even if that sum is less than what is necessary to repair a Rolls Royce).Footnote 25
One answer that may appear obvious is that whoever acquires the ownership of an item is permanently entitled to its market value. But this response is either trivial and uninformative or false. If it is understood to mean that, when someone acquires the ownership of an item, they also acquire the right to sell the item in exchange for whatever amount of money the market deems appropriate, it is a triviality that adds nothing to the present discussion. Because of course Jagger can still sell the car—or what remains of it—after the accident and is always entitled to the car's market value in this trivial sense. However, if it is understood to mean that whoever acquires the ownership of something is entitled to the item's market value “in good status,” that is obviously false. If a hailstorm destroys Jagger's car's windshield, the market value of the Rolls Royce is diminished but Jagger has no right to the pre-damage market value.
The second question left open is why the wrongdoer specifically is liable to the entire redress of the damage.Footnote 26 This is a point I only mention here and develop fully in the final section. If we are concerned about bringing back the welfare condition of the victim to the same level it was before the damage, why not distribute reparative costs across various actors in society, including some that may not have had anything to do with the wronging? Demanding that the wrongdoer alone contributes to the reparation of the damage is peculiarly inefficient, considering they might be hard to identify or, as I noticed above, they might not have the resources to compensate. It is also bound to have pernicious distributive effects. Even assuming that the system of property does not render some people dependent on others, it is still the case that, in an imperfectly egalitarian system, individuals will command hugely different amounts of resources. In such a condition, making the wrongdoer pay for the reparation of the damage up to the return to the status quo is bound to reinforce, rather than merely crystallize, distributive differences. Damage reparation is at the same time a cheap option for the better off, who may well choose to damage and repay as the most effective cost-reductive policy,Footnote 27 but it can be an almost impossible burden for others. A system of corrective justice that led to these results may be internally coherent but is unattractive from the perspective of someone interested in justice in general.Footnote 28
The Kantian account is unfit to answer these worries about the relationship between corrective justice and justice in general because it merely reiterates the ideal of reparations that informs most systems of damage redress in modern law (and, arguably, folk morality), the one according to which “any human act that causes a damage to another, obliges the person whose fault it is to repair it.”Footnote 29 But that ideal is precisely what cannot be assumed to be correct in a critical analysis of corrective practices of the law.Footnote 30
IV. The Continuity Approach: It's Never Too Late to Do the Right Thing
Although I have argued that the Kantian account of corrective duties is unsatisfactory, it is evident which question it attempts to solve. The Kantian account is, ultimately, an attempt to explain the ground of corrective duties. The next account, on the other hand, focuses primarily on what it means for moral agents to discharge corrective duties.
The account in question—the continuity approach—can be schematized in this way:Footnote 31
P1. Every time you wrong someone, you have contravened a duty.
P2. Duties produce (or are equivalent to) protected reasons for action—that is, first-order reasons to do something combined with a second-order reason that requires the duty-bearer not to act following other, potentially countervailing, reasons.
P3. Your breach of a duty cannot defeat the duty's protected reasons of conformity.
P4. The conformity principle: “if one cannot conform to reason completely one should come as close to complete conformity as possible.”Footnote 32
C. Whenever you wrong someone, the reasons you have to correct the wrong are no other than the same reasons that justified your having the duty. “The normal reason why one has an obligation to pay for the losses that one wrongfully occasioned (i.e. that one occasioned in breach of obligation) is that this constitutes the best still-available conformity with, or satisfaction of, the reasons why one had that obligation.”Footnote 33
Premise 2 is a description of how genuine duties operate. Protected reasons work as a “systematic combination of a reason to perform the act … required … and an exclusionary reason not to act for certain reasons” that one would be normally allowed to weigh against the reason to perform.Footnote 34 In the way I presented Premise 2 above, I remain agnostic on whether genuine duties just are protected reasons or whether they, instead, produce protected reasons. I am not sure anything practical would follow from this dispute.
Premise 3 follows naturally from premise 2: if duties generate (or are) protected reasons for action, such reasons cannot be cancelled through an act of will of the duty-bearer. Premise 4 seems hard to reject and is in harmony with the previous two. If the protected nature of duties implies that duties do not lose their normative strength when ignored, it also follows that duty-bearers are still subject to the normative pull of the duties they failed to discharge. So long as one can do anything to conform their action to a duty, they should do that. Better conforming very partially to a duty than not conforming at all.
The trouble with the continuity thesis is the passage from the idea that all cases of wronging correspond to violations of duties (P1—which I accept) to the idea that corrective duties are no other than an evolution of the same duties that the wrongdoer violated. To see what the problem is, I will start with a case where the continuity approach works. I borrow it from John Gardner's original defense of the continuity thesis.
I promise to take my children to the beach today, but an emergency intervenes and I renege on the deal. Let's say I was amply justified in doing so … I am now bound, without having to make a further promise, to take them to the beach at the next suitable opportunity (if there is one). Suppose a suitable opportunity is tomorrow. Am I bound to take them to the beach tomorrow for reasons that are entirely different from the reasons that I had to take them to the beach today? Surely not. Footnote 35
I agree with Gardner's reasoning in the italicized part of the passage. If I have a reason to take my children to the beach tomorrow, those are probably reasons of conformity with the original promissory duty. By undertaking a promise, I do not simply assure the performance of an act; I also signal that I intend to render my agency an instrument to the wellbeing of another person.Footnote 36 Hence, even if I cannot perform the specific act I promised, the fact that I promised still implies I owe it to the promisee to do something good to them—something near the originally promised act.
But now take again the interaction between me and Jagger. Here, there is nothing I have promised Jagger. The duty I breached when I crashed into his car is merely a duty to be considerate towards others when driving. Now that I have breached that duty, is there any residual reason of conformity?
At least under a certain description, it seems to me, the duty of care I owed to Jagger went extinct when I have breached it in the same way that a duty not to kill someone disappears when one completes the killing. That may create some discomfort. Is it then true, against Premise 3 above, that I can cancel the normative strength of duties by breaching them? The answer depends on the level of specificity in the description of a duty. Of course, nobody can extinguish the general duty not to kill others by killing someone, but Oedipus does extinguish his duty not to kill his father Laius when he kills him. In our less macabre case, I do not even extinguish the duty of care I owe to Jagger when I act negligently as I have ample opportunity to act diligently towards him in my future, but I do extinguish the duty-token “act diligently towards Jagger when driving on the Harrow Lane on Monday, October 4, 2021, at 10:25.”
If duties, at least at a certain level of specificity, go extinct when they are breached, we ought to reconsider Premise 3 in the argument leading to continuity. Even though I cannot eliminate the normative pull of duties through an act of will, I can sometimes make duties extinct by breaching them. And, by doing that, I also cancel reasons of conformity. That, incidentally, is what makes violations of duties such a tragic phenomenon. Oedipus can spend the rest of his life atoning for the patricide he unknowingly committed. Still, there is nothing he can do to conform now to the original duty not to kill his father. The same is true for my interaction with Jagger. When I drive past someone on the street, I acquire a duty of care towards them insofar as they are “persons who are so closely and directly affected by my act that I ought reasonably to have them in contemplation.”Footnote 37 But once I have driven past them, regardless of whether I have been considerate, that duty disappears; the reasons I previously had to comply with the duty have lost their normative pull. Why? Because, try as I might, I cannot conform to them any longer, however partially. Which means that, if I do cause an accident negligently and later discharge my corrective duty—whatever it is—I am not conforming partially with reasons derived from the duty of care in the same way as Oedipus's atonement does not constitute partial conformity with reasons derived from the duty not to kill his father. As a quip on the title of this section, it is true that it is never too late to do the right thing but, often, it is too late to do that right thing one should have done.
Why do advocates of continuity insist, then, that corrective duties are explainable in terms of partial conformity with the originally breached duty? One possibility is that they do so because they consider continuity not with reasons of conformity but with reasons grounding duties. To see this distinction, consider that all duties have, in addition to some normative strength, an explanation for why one should do as the duty says. In some passages, supporters of the continuity thesis seem to suggest that continuity applies to reasons grounding the duty. Gardner is most explicit here:
According to the continuity thesis, the further reasons why I had my primary duty are also still in play. At least some of them went at least partly unconformed to when I failed to perform my primary duty. Just as they shaped my primary duty, so they now shape my fallback secondary duty. Only when we know why I had my primary duty, in other words, can we work out what would count as the next best thing to do, now that it is too late to perform my primary duty.Footnote 38
All the italicized parts of the sentence refer to grounding reasons. If that is indeed the best interpretation of the continuity thesis, then the entire thesis is guilty of misrepresenting what it means to conform to a duty.Footnote 39 Merely attending to the reasons grounding a duty, without complying with the requirement, is not partial conformity with the duty; it is no conformity at all.
Consider this rather clichéd example. The reasons that ground the duty to stop at a red signal are numerous: we could cite concern for others’ safety, respect for an overall successful coordination scheme of which I avail myself, and reasons I generally have not to set a bad example. Suppose I am driving in the middle of the night and encounter a red signal. I look around carefully and, only when I am certain I am completely alone and will neither bump into other cars nor set a bad example, I step on the gas pedal. I have attended to the reasons grounding the duty; insofar as I have been considerate about not putting others in danger and not setting a bad example, I have paid adequate respect to such reasons. But I have still disrespected the duty; I have treated it as a mere recommendation to be balanced against competing considerations, instead of a fact about the world capable, because of its exclusionary nature, of determining, and not simply influencing, my conduct.
I conclude, therefore, that the continuity thesis fails to answer the content question convincingly. The duty to correct is not, in its most paradigmatic occurrence, a duty to conform as much as possible to the duty one has breached. We need to go back to the ground question.
V. Adequate Concern, Moral Neglect, and the Case for Corrective Duties
A. Wronging as Neglecting: The General Idea
One thing the continuity thesis seems right about is the idea that, if a person acquires corrective duties, it must be because of something they have done that they should have avoided doing. “Obligation in, obligation out,” in Gardner's words.Footnote 40 Otherwise, there is nothing that calls for correction from that particular person. Footnote 41 Take pure moral luck cases, such as Bernard Williams's example of the lorry driver who kills a child through no fault (or negligence) of his own.Footnote 42 We can agree with Williams that agent-regret is an appropriate emotive reaction to the involvement of one's action in the killing of an innocent. But the lorry driver does not acquire a corrective duty; there is nothing he did that he now needs to correct as there is nothing he should have done differently that would have prevented the child's death.
Here is a suggestion: whenever I am required to correct an act, it is always because I have failed to accord to others the adequate concern that was required from me on the occasion. I wrong others and I am required to correct when I act without paying sufficient respect to the fact that other moral agents are as deserving of this form of consideration as I am.
To add some flesh to this very schematic idea, let me start by integrating it within a contractualist framework. Contractualism understands morality in relational terms; according to it, “our concern with right and wrong is based on a concern that our action be justifiable to others on grounds that they could not reasonably reject insofar as they share this concern.”Footnote 43 The underlying commitment is that, in virtue of their equality of status, moral agents have the authority to place demands on others and to generate expectations for which they can hold others accountable.Footnote 44 Under this conception, “the interests of each person matter and matter equally—not as sources of the kind of satisfaction that we are attempting to maximize impartially, but rather as a potential basis of interpersonal claims against us.”Footnote 45
It follows from this general picture of what it means to behave justly that “one person wronging another … requires that the wrongdoer has, without adequate excuse or justification, violated certain legitimate expectations with which the wronged party was entitled, in virtue of her value as a person.”Footnote 46 What is the most basic expectation for which we can hold others accountable, such that, whenever we see it frustrated, we are justified in requiring that something must be done to make up for it? My suggestion is that the sole expectation that can play such a distinctive role is the expectation of adequate concern. Whatever else we may demand from others, we first demand that they treat our status as a moral equal as a source of stringent, other-regarding reasons for action. Wronging others corresponds to a form of neglect: I refuse to have my conduct shaped, at least in part, by the fact that you are there as a fellow moral agent.
I need to specify what “concern” is ultimately about. Is adequate concern a state of mind, as such possibly inaccessible to others? Or is it, instead, a trait we can reasonably attribute to acts, at least partly independently of the agent's state of mind? If we followed the second route, some acts would qualify as wronging even if the agent had not intended any malice or disregard towards the victim.
We can put the distinction in more precise terms. On the one hand, we have a subjective conception of “concern for others” under which the agent who wrongs others must be subjectively aware that their conduct will either harm others or put them at risk unreasonably.Footnote 47 The subjectively neglectful agent is aware that there are alternative courses of action that, though not excessively burdensome for them, will have a less risky impact on others. That is the mental state typical, for instance, of the normally considerate driver who believes that “just one time” she can overtake the car in front of hers before a blind curve because “today is not the day a car will arrive from the other side.” On the other hand, we have an objective conception of “concern for others,” whereby we do not need to know the mental state of the agent to determine whether they wronged others; at most, knowledge of the mental state can serve to justify or excuse prima facie wrongful conduct. Again following the same example, imagine that the driver was distracted and did not see the signal of an approaching blind curve; in that case, they were genuinely unaware of the risk they were creating. Under the objective understanding of “concern of others,” however, all of this is irrelevant; the act itself qualifies as a prima facie wrong because it is the kind of act that displays disregard for others. The mental state of the agent can still function as a justifying or excusing factor; if the driver can prove, for instance, that they were subject to a sudden and unpredictable urge, their conduct is at least excused whereas, if it turned out that they overtook the car because they were forced to do so by an evil kidnapper who was threatening to kill their child if they did not do so, then their conduct would be justified.
The objective approach is preferable from the contractualist perspective I am following here. The point is not whether the person was subjectively aware of an unjustifiable imposition of risk on others, but whether the amount of concern for others expressed by their conduct can be justified to whoever suffers its consequences, provided such people share the same commitment to take others’ moral status seriously. Very often, someone may be unaware that their conduct subjects others to unjustifiable risk precisely because their subjective evaluation of the risk is biased due to an unreasonably high appreciation of one's ability or the importance of one's activity. In the first case, one is downplaying the risk of one's conduct; the typical example is the car driver who cannot even fathom the possibility of their driving causing an accident. In the second, one is attributing excessive relevance to the needs of oneself or one's dears at the expense of others; one example is the overconcerned parent who believes a scratch on their child's knee is a good reason to rush to the hospital, disregarding traffic norms.
Neither the overconfident driver nor the overconcerned parent can justify their choice to do what they did before fellow moral agents. They cannot claim that they were showing adequate concern for the moral status of others. The overconfident driver may be subjectively unaware of the risk they are imposing on others but, assuming they had been exposed to the average amount of information to which drivers are generally exposed (about the importance, say, of speed limits or safety distance or other such norms), they should know that their conduct, regardless of their personal convictions about risks, expresses neglect for the needs and interests of others.Footnote 48 By choosing to perform that act when other options were available, the overconfident driver shows that their deliberation is peculiarly sensitive to reasons of self-interest and self-relevance (which brought them to assign an excessive trust to their capacities) and not so much to reasons of concern for the safety of others.Footnote 49 The same is true for the unconcerned parent whose deliberation is sensitive to reasons of concern for others, but only when those others are near and dear.
Of course, both agents can present evidence proving that their conduct was either the best thing to do in the dire circumstances in which they operated—in which case their action is justified—or it was particularly burdensome for them to act otherwise—which would excuse their action. But none of this matters in the identification of their conduct as a prima facie wrong against others. Understanding whether an agent's deliberation is sufficiently responsive to others’ status as a moral peer requires analyzing the reasons for their actions and those that were instead dismissed in their thinking about what to do, rather than the presence of malice or disregard in their mind.
How do we move from the recognition that someone has behaved in a way that does not demonstrate adequate concern to the idea of corrective duties? Why, in other words, do people who wrong others acquire corrective duties towards those they have neglected? The neglect account has the resources to answer this question. I have just argued that an agent wrongs another when their deliberation is insufficiently sensitive to others’ equality of status. A thesis commonly accepted about the expressive character of action is that an act can possess a meaning and be understood by other moral agents to express a message, independently of the actor's communicative intention.Footnote 50 We can then agree with both Pamela Hieronymi and Jeffrey Helmreich that the implicit message in wronging “says, in effect, that you can be treated in this way, and that such treatment is acceptable.”Footnote 51
We care about these messages because, as deeply social creatures, we attribute importance to our status in interpersonal relationships and how others behave in relation to it; we can hardly dismiss the neglect shown by others as an irrelevant factor in evaluating how our life is faring.Footnote 52 We also particularly care about these deprecating messages precisely insofar as we recognize others as moral equals: as Seana Shiffrin recently observed, “[I]f I see you as a distinct individual, as a moral agent and as a moral equal … how could I not reasonably care what you think about all sorts of matters, including about me?”Footnote 53
The neglect account thus not only offers a response to the ground question but can also guide us towards answering the content question. The demand for corrections is, in the most general sense, a demand that those who have wronged us do something to demonstrate that they do take our existence seriously, thus counterbalancing moral neglect. We want them to show us that they are capable of letting their conduct be conditioned by the fact that we are their moral equals. This can be done by undertaking expressive acts that, within a community of sense, are canonically interpreted as demonstrating one's recognition of the constraints placed on one's action by the presence of others.Footnote 54 Apologies are, in this regard, the prototype of corrective actions: they carry with them the message that people not only assume responsibility for the wrong, but also distance themselves from it and implicitly commit not to fall into it again.Footnote 55
Now the account has been shown to respond to the ground and content question, I end this section by rebutting possible objections and specifying it further.
B. The Neglect Account: Further Comments
Let's start with the scope of the idea of wronging. So far, I have mostly used examples of behavior that could qualify as either negligent or reckless, but that does not presuppose a deliberate intention to harm. In fact, I have excluded the option that someone must intend malice or disregard for others to wrong them. So is the account I have presented only a conception of reckless or negligent wronging? Not so.
I am confident that the idea that wronging corresponds to a neglect of the proper demands placed by another's moral status on one's conduct can also cover cases of wrongful actions where the wrongdoer deliberately intended to harm or disrespect the victim.Footnote 56 Admittedly, the terminology of “neglect” might feel awkward here. Actors that deliberately disrespect others are not, in the ordinary language sense of the word, neglectful; in fact, others play a prominent role in their reasoning. But the idea of adequate concern is to be understood in qualitative, not quantitative terms. It is not sufficient for adequate concern that some agents “thought enough” about others before acting; the concern shown (to reiterate, in the act, not the mental state) must also be sufficiently sensitive to the fact that others, like ourselves, possess a moral status that entitles them to place demands and expectations on our actions.
Take, for instance, a case of manipulation. Michael manipulates Paula into thinking that her job perspectives are grim in her current position and that her boss is mistreating her, with the sole purpose of having her quit and taking her position. In one sense, Michael is not neglectful of Paula; he is very well aware of her relevance in his master plan. But Michael is not letting his conduct be properly conditioned by the fact that Paula is a moral agent who can place normative requests on his action, one of which is that he should not lie to her for his self-interested purposes.
What about the least deliberate faulty conduct recognized in the law, negligence proper? A number of authors working mainly in criminal law have disputed the idea that negligence proper—distinguished from recklessness—may be a type of wrongful conduct.Footnote 57 The implication for the law of torts is that negligent liability is, in reality, a form of strict liability, a conclusion that, according to Heidi Hurd, should embarrass both law and economics and rights-based torts theorists.Footnote 58
Let's take negligence proper to be a failure, on the part of an agent, to conform to a standard of conduct that is both adequate, because acting otherwise would have imposed unreasonable risks on others, and within the agent's competency.Footnote 59 There are two ways an agent can fail to conform.Footnote 60 One is the case where one made a reasonable effort and still failed because, even though conformity was within their competency, it was particularly burdensome for them. The second is when the agent did not even make an effort. The second case is still not the same as recklessness; the agent who did not make an effort to conform to the standard, unlike the reckless actor, might have been subjectively unaware of the risky feature of their conduct—perhaps because, as in the example above, they cannot even fathom the possibility that their action may harm others.
Through the neglect account, we can see that the second kind of negligence—that is, negligence understood as not making an effort to respect a reasonable standard of conduct—is both a moral failure and a wrong perpetrated against the potential victim. Conforming to a standard of conduct can be burdensome for the agents involved; negligent behavior is then a sort of advantageous conduct, for no other reason at least because it leaves their mind unoccupied when it should be fully focused on trying not to put others at risk. The negligent actor's interests—in doing what she was supposed to do faster or less attentively than in the way recommended by the standard—are taken by them to be more worthy of consideration than the interests of others in not being subject to unnecessary risks.
Not all cases of attribution of liability to negligent actors in the law are examples of negligence of the second kind. Sometimes, agents are treated as negligent in the law even when it is unclear whether they had the capacity to conform to the standard of conduct,Footnote 61 or when it cannot be demonstrated that they did not make a reasonable effort. This need not concern us. It might be that the definition of negligence in the law is ambiguous regarding whether one agent did their best to conform to the standard. If so, we have to accept the conclusion that some cases of negligent liability in the law are, in reality, cases of strict liability. But, if I am right here, the conclusion need not be universalized; there is at least one type of negligence, corresponding to the subjectively beneficial conduct of disregarding a reasonable standard of conduct, that does entail wronging others. Moreover, again anticipating the content of the next section, we do not need to presuppose that the aim of corrective justice is telling us who should redress the damage; in some cases, demanding from someone that they repair a damage they are not at fault for causing may still be the best thing to do, although not for reasons of corrective justice.Footnote 62
Let me offer three final advantages of the proposal. First, my account maintains the intuitive connection between wronging, corrections, and the reactive attitude of resentment. More specifically, it has no difficulty explaining why we resent others for not giving us adequate concern, even when this does not damage us. In the article that inaugurated the discussion on reactive attitudes, P.F. Strawson speaks of “some degree of goodwill or regard” that we demand from others who stand in certain relationships with us, including “as chance parties to an enormous range of transactions and encounters.”Footnote 63 Because we legitimately expect from others that they treat our moral status as a source of stringent reasons for action, we can resent them when they fail to do so.
A second advantage of my account is that, unlike recent proposals that tend to dissociate wrongs from rights violations, it conforms to the “obligation, obligation out” principle without falling back on the Kantian position that all actions calling for corrections are akin to rights violations.Footnote 64 And that is, paradoxically, for a reason we can derive from Kant himself when he writes, “There is connected to right by the principle of contradiction an authorization to coerce someone who infringes upon it.”Footnote 65 Consider an analogy with gratitude. According to any conceivable conception of interpersonal morality, I sometimes ought to be grateful to others. Yet nobody can be coerced to demonstrate gratitude. Similarly, the expectation to receive adequate concern cannot generally be a ground for coercion. Borrowing a distinction from Adrienne Martin, adequate concern is a “normative expectation” rather than a “normative demand”: “Such an expectation is normative—it is a holding to a norm, rather than a prediction—but it is not a claim of either moral or legal right.”Footnote 66
A final advantage of the proposal is that it can help solve some puzzles about the normative treatment of harms and risks. Some authors find it puzzling that we should morally condemn cases of “pure” risk—actions that put others at risk but that, due to fortuitous circumstances, do not harm anyone.Footnote 67 There is a discussion in the literature about whether such cases of pure risk can be redescribed as harming; according to some, that would make our moral condemnation of pure risk easier.Footnote 68 But, if the definition of wronging I have offered is correct, neither harms nor risks qualify, per se, as wronging acts as we can both harm others and put them at risk without wronging them. We harm others without wronging them, for instance, in some cases of self-defense, whenever our decision to harm does not derive from attributing an excessive weight to our life and limb compared with those of others.Footnote 69 And we subject others to risk without wronging them whenever, from the ex-ante perspective from which we deliberate, we can prove that any alternative course of action will produce similar or worse results for a similar number of potential victims.Footnote 70 Hence, the discussion about whether pure risks are harms is normatively inert. What matters, instead, is considering which cases of risk impositions and which eventuated harms constitute cases of wronging. The reference to adequate concern, I submit, can help us answer that normatively relevant question.
One may raise the following objection. If wronging others means not attributing to them an adequate level of concern and corrective duties aim at counterbalancing moral neglect, does this imply that a wrongdoer acquires corrective duties whether or not the risks they create produce injuries? This conclusion may be counterintuitive and definitely contrasts with legal practices, especially in the law of torts. But, with some due caveats, it is the only one I feel we should accept.Footnote 71
Let me set up the caveats that apply here. First, many (probably most) unjustifiable risk impositions that do not produce injuries are epistemically inaccessible. The wrongdoer may be unaware they are imposing a risk and so may be the potential victims. In these cases, the corrective duty does not emerge due to epistemic limitations; nobody knows that some agent has been subject to unjustifiable risk. Second, even in some cases where the wrongdoer (and maybe some third party), but not the victim, is aware of the risk imposition, the risk-imposer may be justified in not discharging the corrective duty—as long as the risk imposition in question is not a demonstration of grave moral neglect—mainly not to make our moral interactions too complicated. Imagine someone who almost crashes into your car out of distraction and, instead of going about their business for the rest of the day, follows you for miles to express her sincere apologies. We would consider it, I guess, awkward behavior. We would probably say of such a person that, in attempting to do the right thing, she is demonstrating a commitment to morality that seems absolute and risks alienating her from morally fallible creatures.Footnote 72 But, at the same time, we would not consider a mouthed apology in the same scenario inappropriate. And we would not condemn an agent who has undergone such a risk and demanded an apology.Footnote 73 This illustrates that our moral practices are consistent with the idea that those who place unjustifiable risks on others are liable to corrective duties even when the risk does not generate actual harm.
Am I seriously saying that the content of the corrective duty is the same whether or not the risk leads to a harm? Some authors sympathetic to the idea that wronging others merely presupposes the violation of a directed duty, regardless of the effects this has on the victim, have nonetheless argued that the wrongdoer is made liable to duties to repair only when such adverse effects eventuate.Footnote 74 The position is probably the closest in the literature to what I am affirming here, but I see a tension in it. I am not specifically persuaded that one can easily defend the view that, “Unless … the breach of duty leads to recognized injury, no claim arises.”Footnote 75 Of course, from the victim's perspective, the fact that there was no injury makes a relevant difference (but not so relevant to annul all complaints in case the injury does not occur). From the perspective of the wrongdoer, however, it does seem unfair, at least prima facie, that two agents who demonstrated the same kind of lack of concern for others are subject to two dramatically different types of liability (namely, no liability at all vs. liability to potentially very costly reparative costs) depending on factors entirely outside their control.Footnote 76
That is why I prefer putting the matter more radically. If a wrong has turned into an injury, the wrongdoer is provided with a very good option to correct the wrong that would not otherwise have existed, namely contributing to the compensation of the damage and the restoration of the victim's wellbeing. That might be the best demonstration of the wrongdoer's will to correct and one successful way of discharging her corrective duties. But we have no reason to consider the restoration of the victim's pre-damage wellbeing as the sole or even the most paradigmatic way for the wrongdoer to discharge their corrective duty; it is but one option among many others that may be peculiarly appropriate when a damage has come about.
If I maintain, as I have done throughout, that in case of a wrongfully originated damage the corrective duty of the wrongdoer is not just a duty to repair, is my account then silent about the appropriate entitlements that victims of wrongful damages, as opposed to victims of wrongs in general, can legitimately claim? This is the puzzle I address in the final section.
VI. Wrongs, Damages, and the Proper Scope of Distributive and Corrective Justice
Isn't corrective justice concerned, at least in part, with the appropriate repayment of damages? My answer here is unequivocal: if the focus is on damages and how they altered the victim's wellbeing, as opposed to the wronging itself and how it neglected the victim's status as a moral equal, we find ourselves in the realm of distributive, and not corrective, justice. Let me explain why.
Many cases of wronging put the victim in a position where they are in dire need of help. Sometimes, our commitment to corrective justice as the standard resource to address cases of wronging leads to denying the victim the help they should have when they most need it. Consider this case.Footnote 77 In 1980, a civilian airplane with 81 people on board flying from Bologna to Palermo was intercepted by a missile and shot down. All passengers and crew members died. Because it was impossible then (and now) to determine which military force fired the missile, the private company operating the flight was forced to shoulder the costs of the accident and eventually went bankrupt. We do not need to be fans of capitalist entrepreneurs to feel sorry for the company's owner. On the contrary, we should treat the case for what it is: one of grave distributive injustice.
Egalitarian theories of distributive justice are often confronted with the problem of finding a role for personal responsibility. Luck egalitarianism addresses the problem in what is arguably the most drastic way: all unequal distributions that are attributable to individual choices (and, for some authors, option luck) are in virtue of that more acceptable than they would be if they were just the product of luck (or, according to some, brute luck only).Footnote 78 But the same attention to reconciling equality and responsibility can be found in Rawls's insistence that the basic structure of each society, if organized around the principles of justice as fairness, guarantees that citizens “accept responsibility for their ends” and “take charge of their lives” by adapting “their conception of the good to their expected fair share of primary goods.”Footnote 79 One of the core purposes of a just basic structure for Rawls is to guarantee that citizens can pursue their conception of the good without interfering with others attempting to do the same.Footnote 80 When such conditions are in place, citizens can accept responsibility for their choices and, more specifically, can be made liable for at least some costs of those choices.
Consider again the case of the airline company. Had the company contributed to the plane's fall—for example, due to its loose compliance with security rules—we would consider its disgrace a case of bad option luck (for luck egalitarians) or the outcome of a choice undertaken within the fair background conditions of a just basic structure. But we know this is not the case; the company was not responsible in any way for the accident. Why should it be forced to shoulder its costs, then—especially when such costs are so severe that they can compromise its survival?
This point generalizes to all cases of wronging. If our focus is on what victims of wrongs lost or suffered, and on the injustice of it, as opposed to how such victims were treated by the wrongdoer, our concern is fundamentally distributive. In the same way as the question about who should pay for someone's appendix removal—the person herself, through their money and insurance, or society at large through taxation—is distributive, so is the question about who should pay for Jagger's Rolls Royce or who should pay compensation for the lost lives and property in the plane crash.
The Kantian-Aristotelian insight according to which, in the case of wronging, the question about the distribution of costs is confined to the wrongdoer and the victim confuses two separate questions. One is whether wrongdoers owe their victims something; that is the genuinely corrective question. The separate question is who should bear the costs of the accident.
I acknowledge that what I am proposing here is a kind of conceptual revolution regarding the domain of application of the concepts corrective justice and distributive justice.Footnote 81 But it is a conceptual revolution that I feel is necessary because, as exemplified by the airline case, an exclusive focus on corrective justice as the conceptual repertoire from which to derive principles for the distribution of responsibilities to repair will create unfairness.
Does this conceptual revolution entail that wrongdoers should never repay the costs of their damages? Not at all, and for two different reasons. The first is that, if we think that desert is one criterion to take into account in distributive matters, we may consider the fact that the wrongdoer is responsible for the accident as a prima facie reason to let the costs fall on them. But, unless we implausibly take desert to be the sole criterion of distributive justice, we will have to balance that against other factors.
Second, there are specific cases of wronging for which reparation does appear as the paradigmatic way for wrongdoers to counterbalance moral neglect. The case where this is most evident is probably theft. By stealing, I demonstrate my lack of concern for another person in their specific role as an owner of external things. Because of this, there is no other way I can demonstrate my respect for that person's moral status than by returning what is theirs. Another case where a passage of resources from wrongdoer to victim seems mandated by corrective justice is where the wrong was specifically undertaken to derive a profit. A company that has breached safety regulations to lower costs is subordinating others’ lives and safety to the pursuit of its profit. In case an accident eventuates, one way the company may demonstrate respect for the victims’ moral status is to cede to them all profits it has accrued by breaching the rules. This shows that corrective justice is not altogether silent regarding the issue of who should pay to repair when some agents have been harmed but has a much more limited role to play than is usually assumed.
VII. Conclusion
I began by writing that we have a general understanding of a corrective duty and I have shown throughout that our pre-theoretical understanding of corrections is incomplete and contradictory. Sometimes we may think that corrective duties aim at repairing a wrong—making the victim's status as close as possible to what it would have been were it not for the occurrence of the wrong. But, I argued, reducing corrective duties to duties of reparation is hard to square with the ideas that victims underwent a wrong and not just a setback of their interests and that, in addition, the size of the wrong and the size of the victim's harm may differ remarkably.
Previous philosophical attempts to make sense of corrective duties have also encountered difficulties. The Kantian idea that the sole ground of correction is the violation of somebody's right is distinctly unhelpful in telling us how to correct the violation. The more recent attempt at showing that corrective duties are no other than ways of complying with an original obligation after its breach does not make sense of some paradigmatic examples of corrections.
To some extent, my account is also, like the previous two, reductionist; I do identify a single duty (adequate concern) whose breach is the sole ground of legitimate claims for corrections. But understanding what calls for corrections as a slight to one's status as a moral agent, rather than a material damage, expands the range of corrective options. This is how it should be, considering the diversity of ways in which we wrong each other, from the most minute and easy to dismiss to the ones that almost seem to take away any form of self-respect. And considering further the variety of mechanisms our social practices already include—and the ones we can engineer—to make it up to others for what we did.
Open access
