4. DIALOGIC COMMUNITY IN THE LAW OF NEGLIGENCE

We have been calling the relationship of mutual recognition between free and equal selves dialogic community. Since the conscious grasp of this relation as the source of individual rights underlies the law of negligence, we can resume the interpretive method we employed in the chapter on contract. That is to say, we shall understand negligence law by recognizing in its central doctrines the intersubjective structure characteristic of dia-logic community. Let us first briefly recall the elements of this structure.

An individual right begins as a claim of absolute worth by the person. This claim must be objectively validated, and the right enforced by law is the outcome of the process of validation—the legal recognition of a claim that has gained implicit confirmation prior to enforcement. The person's claim of absolute worth is satisfyingly validated only through the submission to its freedom of another self, for only another self has the requisite alterity and independence to confirm objectively the end-status of another. In contract, the submission of one self to another is centrally effected through promise; apart from contract, it consists in the care exercised by one self for the human welfare of the other. To give an effective or satisfying recognition (i.e., to constitute the other's right to care), however, the self must preserve its independence in its submission, and this is possible only if the other self reciprocally submits to the end-status of the first. In contract, where affirmative duties are narrowly circumscribed by the will of the parties, this reciprocal act of submission can take the relatively simple form of a promise or the performance of some discrete service. However, to gain a right to the generalized care of another, a more complicated exchange must occur if the other's end-status is to be preserved in its submission. In particular, the self can retain its independence in generally caring for another's welfare only if the following two conditions are fulfilled: first, the self must reserve (and gain recognition for) a space for self-oriented action, that is, a space where it can act for its own ends free of any requirement to defer to the other's interests; second, the object (O) of the self's (S) care must reciprocally defer to S's welfare to the same degree as S defers to O. We can call the first condition of a valid right to care the requirement of equal respect; we can call the second the requirement of equal concern.97

Accordingly, the self's worth is objective only within a relationship of mutual regard between selves who are conscious of themselves as ends;

and so legally recognizable rights to another's care are grounded in such a relationship. This means that a right to another's care is the product of a threefold mediation. The self's worth is mediated, first of all, through another self's respect for its liberty to pursue its self-interest as well as through the other's concern for the well-being of the necessary instruments of its doing so; second, it is mediated through the self's reciprocal respect and concern for the other; third, it is mediated through the connection formed by each self's need for the other's freedom—a connection ensuring that the dependence of each on the other's will is compatible with their mutual preservation as ends. Our task now is to interpret the basic doctrines of negligence law as embodiments of this structure.

4.1 The Standard of Care

According to the rule in Vaughan v. Menlove,98 one must compensate another for injuries caused by one's failure to take the care that an ordinarily prudent person would have taken in the circumstances. A full comprehension of this rule requires an understanding of its three components. First, one is liable to compensate a person one has injured only if one was at fault, and fault is understood as a failure to take reasonable care. Second, the standard for determining whether the care one took was reasonable is an objective one; what matters is not whether the agent acted to the best of his judgment but whether he acted as a reasonable person of ordinary prudence would have acted. Third, one must understand the formula for determining how much care the reasonable person would have taken. Here there is controversy in the law, and the approach we take can suggest a solution.

First, the fault requirement. Earlier we saw that liability for trespass to the person and to property is independent of fault. Because possessory rights are established by unilateral action, they presuppose no, or (in the case of nonbodily things) only the barest, reciprocal accommodation to the liberty or welfare of others. Strict liability for trespass means that one has no liberty to touch another or to take his goods without consent; there is no sphere of privileged invasions. However, negligence law protects not a property right strictly understood—not an exclusionary right of sovereignty over one's body and possessions—but a right to another's affirmative care for the material conditions of one's freedom, for one's health, bodily integrity, and possessions. Thus damage or harm is essential to an action for negligence but not to an action for trespass. Moreover, the right to another's affirmative care emerges only from a complex reciprocity of respect and concern between independent selves. My worth is objective in the care exercised by another self for my sake. Yet this submission is effective to validate my worth only if it is the submission of a free self who

regards himself as an end and who preserves himself as such in the relationship. Hence my worth is also mediated by my reciprocal respect for the other. To respect someone as an end, however, is to acknowledge a moral space wherein he is free to act for his own ends and to prefer these ends to mine. Hence my right to another's care presupposes my recognition of his claim to a limited sphere wherein he may act without care for my interests.

Now a regime of strict liability for harm implies the absence of such a sphere of self-regarding liberty. Under strict liability, one has no right to act in ways that happen to injure another. Since, however, all action carries the risk of such injury, strict liability means that I have a right that you be governed in all your actions by concern for my welfare, and you have the same right over me. No doubt there is a mutuality of care here; but it is the mutual care of extreme altruists who, because they claim no worth as independent selves, can neither give nor receive effective confirmation of worth and hence can acquire no valid right to care. By contrast, a fault requirement establishes a reciprocity of care between selves. Such a requirement signifies that each person acknowledges the other's liberty to act for his own ends and careless of the other within bounds compatible with their equal security against harm. Each is liable to compensate the other only if he oversteps these bounds.

So far we have understood why fault—understood as the exceeding of a threshold of reasonable indifference to the other's welfare—is the correct principle of interpersonal liability for causing harm. We have now to consider the specific issue raised in Vaughan v. Menlove —whether the reasonableness of risk-imposing conduct should be determined by the best judgment of which the defendant is capable or by that of the ordinarily prudent person. We will understand the law's decision in favor of the objective standard when we see it as required by the structure of dialogic community.

It will help to recall the facts in Vaughan v. Menlove. The defendant built a hayrick on the edge of his field close to the wood and thatch buildings of his neighbor. He was warned that hayricks are prone to spontaneous combustion and that they posed a danger to the surrounding area. Nevertheless, the defendant decided to "chance it.' Sure enough, the hayrick burst into flames and destroyed the neighboring buildings. The trial judge instructed the jury to find for the plaintiff if they thought that the defendant failed to exercise the care of the ordinarily prudent person. The defendant appealed against this instruction, arguing that he had done enough if he had acted to the best of his judgment and that "he ought not to be responsible for the misfortune of not possessing the highest order of intelligence."99

The application of our usual analysis will show that the objective stan-

dard is the only one supportive of a valid right to care. We have said that the person's end-status is confirmed through the care exercised on its behalf by another self. Yet this dependence on another's will would contradict rather than confirm the absolute worth of the person if it were a dependence on an arbitrary will external and indifferent to its own. The person is dependent on an external will if it is dependent on another's subjective opinion as to whether the care he or she exercised was reasonable. Hence if the plaintiff's right to care depended on the intelligence or other idiosyncratic capacities of the defendant, it would be no right at all. By contrast, the person's dependence on another for self-confirmation is compatible with its absolute worth (and so with a right to care) if it is dependent only through the mediator of a common will ensuring that dependence on another is consonant with autonomy. The idea of the reasonable person of ordinary prudence is this mediator. Each person is dependent for self-confirmation on an exercise of care judged sufficient not by the opinion of the other but by a reason common to both.100

Finally, we have to understand the formula the court applies to determine the standard of ordinary prudence. Here the American and Commonwealth traditions are at odds. Following Judge Learned Hand's test in United States v. Carroll Towing,101 American courts determine whether conduct is negligent by comparing the present cost of a potential accident with the cost to the defendant of avoiding it. The present cost of the accident is calculated by multiplying the expected cost by the probability of the accident's occurrence. Thus, if the expected cost of an accident is L, the probability P, and the cost of avoidance B, the defendant is negligent if he failed to take precautions when B << L × P. If the cost of avoidance is greater than the present cost of the potential accident, no liability is incurred by failing to take the necessary precautions.

The English test, by contrast, takes the cost of avoidance into account only in limited circumstances. In Bolton v. Stone,102 Lord Reid stated that one had a duty to avoid imposing "substantial" risk on one's neighbor. By a substantial risk he meant a risk in excess of that ordinarily adjunct to social interaction in "the crowded conditions of modern life." The magnitude of the risk was a product of the seriousness of the expected harm and the probability of its occurring. If the product of these factors is substantial in the sense of extraordinary, then one must avoid the risk whatever the cost to oneself. In Wagon Mound, No. 2,103 Lord Reid amplified this rule, stating that one must also avoid a risk falling within the range of the ordinary if one can do so at no disadvantage to oneself. For Lord Reid, therefore, the cost of avoidance becomes relevant only within the range of reciprocal risks, and it works only to enlarge the sphere of duty, never to contract it. Which test—the English or the American—is the proper one?

In the discussion of the fault requirement, we saw that a right to care is

generated only from a relationship wherein each self recognizes the other's right to prefer its own interests within bounds compatible with the equal security of both. On the one hand, the relationship must be one between persons who are mutually caring; but on the other, it must be one between mutually caring persons. The substantial cost test of Bolton v. Stone reflects such a relationship: as long as the present cost of the accident (i.e., risk times seriousness) falls within the range each imposes on the other as an incident of everyday social interaction, each may pursue his ends without thought for the safety of others. Only a nonreciprocal cost must be avoided. The amplification in Wagon Mound does not disturb the relationship as one between mutually caring egos, for one must take thought for the other within the range of the ordinary only if doing so involves no cost to oneself; so the principle of self-preference within the range of the ordinary still holds (although one may question whether avoidance is ever costless).

The Learned Hand formula, by contrast, reflects a relationship that either lacks mutuality of care or is not a relation between persons. Since the agent is always permitted to prefer his own interests if the cost of avoidance exceeds the present cost of the accident, the formula permits self-preference even if the risk exceeds the ordinary and so allows scope for liberty greater than that consistent with the equal security of both. Since no person could recognize such a liberty in another without compromising himself as an end (and hence without disqualifying himself as one capable of giving an effective recognition), there can be no right to a liberty of that scope. At the same time, the Learned Hand formula disallows self-preference in some cases within the range of ordinary risks, for it requires one to prefer the other's interest whenever the cost of avoidance is lower than the present cost of the accident. In requiring greater altruism than is necessary for the equal security of both, the formula envisages beings who do not assert themselves as ends and who can thus acquire no validation for their worth from the relationship. On either count, the formula generates no valid right to care.

4.2 To Whom Is a Duty Owed?

In Palsgraf v. Long Island Railroad Co.,104 a guard stationed on a railway platform helped push a passenger into a train after it had begun moving and when the passenger was in danger of falling. In the process, a package containing fireworks dropped from the passenger's grasp and exploded. The force of the explosion knocked down a scale at the far end of the platform, and the scale struck Palsgraf, causing her injuries. Palsgraf sued the railroad. Justice Cardozo, speaking for a majority of the New York Court of Appeals, argued that the guard's conduct, while perhaps negligent in

relation to the holder of the package, did not wrong Mrs. Palsgraf. The guard may have failed to exercise due care, but negligence in the abstract is not a tort, for the plaintiff must show that she in particular had a fight to the defendant's care for her physical security. She could not do this, argued Cardozo, because [n]othing in the situation gave notice that the falling package had in it the potency of peril to persons thus removed."105

Justice Andrews thought otherwise. He argued that a failure to take reasonable care for the security of others was in itself wrongful, because "[e]veryone owes to the world at large the duty of refraining from those acts that may unreasonably threaten the safety of others."106 Someone who violated this public duty was legally responsible to anyone whom he injured whether or not the person injured was someone he could reasonably contemplate as standing within the ambit of the risk.

Andrews's conclusion is surely a non sequitur. From the fact that one has committed an offense against the public welfare it follows that one must pay the penalty for the offense; it does not follow that one has a duty to compensate an injured person. The latter duty follows only from a civil duty of care correlative to the injured person's right to one's solicitude for her. Cardozo believed that a plaintiff had a personal right to the defendant's concern only if she was "within the range of apprehension" as someone likely to be injured by the defendant's act. He did not say why this should be the criterion of the plaintiff's right, but we can perhaps supply the missing reasoning.

We have seen that a fight to another's care is the product of a relationship of mutual concern and respect. My worth is objective in the self-abnegation for my sake of someone whose selfhood I reciprocally confirm by my respect for his liberty and by my concern within limits for his welfare. This relationship may arise contingently through contract, or it may be grasped in thought as a necessary symmetry between relations of deference. In the latter case, my right to another's care is mediated through a concept wherein the parties are so identical that the duty of care of one for the other is simultaneously the other's duty of care for the first. This is the neighbor principle according to which someone whom a reasonably circumspect person would contemplate as likely to be injured by his activity has a fight to the exercise of care on his behalf. The reasonably circumspect person merges all individuals into one; hence the duty of each is matched by a reciprocal duty of all. Given this reciprocity of concern, the neighbor principle generates an authentic right to care.

Justice Andrews's principle, by contrast, fails to ensure mutuality in the parties' relationship. To compel an agent to compensate anyone who happens to be injured by his negligence is to compel his solicitude for a stranger under a rule that fails to impose necessarily reciprocal duties. Whereas plaintiff and defendant are identical in the idea of the

reasonably circumspect person, they remain discrete in the sight of the Andrews principle, for the latter provides no homogenizing standpoint from which the parties' duties toward each other (however asymmetrical in fact) may be regarded as one and the same. The Andrews principle generates one-sided relationships of deference from which, as we have seen, no fight can emerge.107

4.3 Remoteness

The same analysis helps us understand the rule governing the extent of damages for which a negligent defendant is responsible. In this regard, William Prosser once posed an interesting hypothetical:

The defendant, delivering a parcel, drives his truck up a private driveway to the back door of a home. On the way up he notices at the side of the driveway a large paper box or carton, open and visibly empty. Two minutes later, coming down the driveway, he negligently runs over the box. Negligently, because he knows it is there, it may be owned by someone, and it has some small value. In the meantime a two-year-old child, whose presence could not reasonably be anticipated, has concealed himself in the box. Is the defendant liable for the death of the child?108

Prosser thought he was. He was supported at the time by the leading English case of In Re Polemis,109 where it was stated that once a defendant had failed to exercise due care for the plaintiff, he was liable for all harm "directly" caused by his negligent act even if the damage was unrelated to the foreseeable harm he was obliged ex ante to prevent. However, the Polemis case was overruled by the Judicial Committee of the Privy Council in a case known as The Wagon Mound, No. 1.110 There Viscount Simonds stated that one is liable to a plaintiff whom one has negligently injured only for the type of harm that a reasonable person would foresee as likely to result from one's failure to take care. Under the rule in Wagon Mound, the truck driver in Prosser's hypothetical would be liable for damage to the box but not for the death of the child.

Our analysis supports the rule in Wagon Mound. It is not enough to ask whether the plaintiff has a right to the defendant's care on his behalf. One must further inquire as to the kinds of harm the plaintiff may compel the defendant to avoid for his sake. The plaintiff has a right only to as much exertion for his benefit as is consistent with the equal worth of the defendant, for (as we saw) only another end can give the objective confirmation of one's worth that we call a right. Yet in the absence of a contract, liability for all chance harm would one-sidedly subordinate the defendant to the plaintiff, for the defendant would be compelled to submit to the plaintiff under a rule that fails to ensure the plaintiff's matching duty to him. By contrast, a rule limiting recovery to reasonably foreseeable harms

merges the parties in the concept of the reasonably circumspect person and imposes a duty on one that is identical in all respects to that owed by the other.

4.4 Causing Harm

Suppose A breaches a duty of care owed B but the negligence does not materialize in any harm to B. Or suppose A breaches a duty of care owed B and B is harmed but not as a result of A's breach of duty; he would have suffered the same harm even had A not acted negligently. In either case, A has failed to take the care to which B has a right, but by chance the breach of duty has occasioned no harm. Why is B not entitled to nominal damages from A? Why is B not entitled to an injunction against A's negligent conduct?

One might argue that the causing-harm requirement singles out the plaintiff from the public at large as the one to whom a civil remedy is owed.111 He is, after all, the one in whom the unreasonable risk of harm created by the defendant has materialized. But this is not an adequate answer. The neighbor principle already singles out the plaintiff as someone to whom a duty of care is owed. Why is wrongdoing not complete with the breach of this duty? There are certainly ways of enforcing a civil duty of care besides requiring the defendant to compensate for injuries caused. One could, for example, require the defendant to pay a token remedy to any foreseeable and unharmed plaintiff who brought suit; or one could allow the plaintiff to recover from the defendant the cost of any reasonable measures he took to guard against the risk.112

Perhaps the causation requirement singles out the defendant from the world at large as the person responsible for the plaintiff's injury.113 If the defendant did not cause the injury, there is no link between him and the plaintiff's misfortune and so no more reason to lay the burden of compensation on him than on anyone else. Yet this answer assumes what must be explained. If wrongdoing is not complete until harm is caused, then of course the defendant is a wrongdoer only if he caused the harm. But why is the chance causing of harm necessary to wrongdoing? Why is not the defendant who has breached a duty of care owed the plaintiff already a wrongdoer? Can our theory of the foundations of a right to care support the requirement of causing harm as a necessary element of negligent wrongdoing?

The harm requirement was already implicit in the transition from a trespass theory of wrongdoing to a negligence theory. Each of these theories, we saw, presupposes a certain conception of individual rights. Since a wrong is the infringement of a right, the elements of wrongdoing are determined by the manner in which rights are defined. Now in formal right,

the person's rights are proprietary only; they are rights against interference with the sovereignty of one's will over its body and possessions. Such interference is trespass, and the latter is complete even in the absence of harm to the things to which one's dominion extends, for one's sovereignty over a thing may be contradicted whether or not the thing is damaged. The rights generated by formal right are exclusively proprietary because of the abstract foundations of formal right. Where all interests are regarded as subjective, rights will attach only to the end-status of the formal will as embodied in proprietary sovereignty; there will be no explicit recognition of a right against harm to essential interests. Such a right, however, is implied in the will's need for embodiment, and so as long as formal right remains the dominant paradigm, this right will exist somewhat as a bastard child—theoretically disowned but practically acknowledged.

In negligence law, however, all this formalism and dissemblance are superseded. As a result of contract law, the person's absolute worth is known to be real in the mutual submission of independent persons, and this relationship grounds a right to the positive realization of freedom. This positive right is foundational for negligence law. The right underlying negligence law is not simply a right to do as one pleases with one's body or possessions; the plaintiff is compensated for damage even if no challenge to his proprietary sovereignty has occurred. Rather, the right vindicated by negligence law is a right to the well-being of the will's embodiments, a right against harm to interests essential to the objective realization of freedom. The precise scope of this right is no doubt determined in accordance with the concepts of duty, remoteness, and fault reflecting the intersubjective basis of rights. However, these concepts have no status apart from their role in framing the limits of the person's right to compel a stranger to take heed for his welfare. They do not themselves create rights but only structure a right vis-à-vis others to security against harm. Since, however, the underlying right of negligence law is a circumscribed right against harm, there is no wrong until harm is caused.

4.5 Corrective Justice

So far we have dealt with concepts—fault, duty, remoteness, harm—that define the contours of a right to another's care. We have seen that such a right is established only within a dialogic relationship reconciling selfishness and altruism. Each cares for the other's safety up to a point that allows space for their equal liberty to prefer their own ends; and each respects the other's liberty to pursue his ends within bounds consistent with their equal concern. Since a person has a right to care only within the parameters defined by this reciprocity of concern and respect, it follows that he is wronged only when indifference to his safety breaches the condition

of reciprocity—that is, flows from an exercise of liberty greater than that consistent with their equal respect and concern. Such an exercise of liberty is wrong because it asserts one's end-status at the expense of another's and so claims a right to liberty in excess of that capable of being validated through the process of mutual recognition.

With this understanding of the nature of wrong we shall be able to understand the way in which the common law remedies negligent wrongdoing. The remedy takes the form of a payment by defendant to plaintiff of a sum of money sufficient to restore the plaintiff (as far as money can do so) to the level of welfare he or she enjoyed prior to the wrongful transaction. This remedy is supposed to accomplish two things at once: it is supposed to annul the defendant's wrong and to vindicate the plaintiff's right. An adequate explanation of this secondary transaction must resolve the problem we noticed in our critique of Epstein. There we observed that vindication of a victim's right requires a remedy moving from the tortfeasor (i.e., a private-law remedy) only if the victim's loss and the tortfeasor's gain are correlative. Loss and gain are correlative if one person's gain implies another's symmetrical loss and vice versa. An image might help to clarify this relation.

Imagine a teeter-totter on which two riders are sitting at opposite ends of the board. They are initially balancing the board so that each rider is equidistant from the ground. Any upward movement of one rider will necessarily involve a downward movement of the other of an equal number of degrees. The lower rider cannot return to the position of balance unless the upper rider is lowered to that position—that is, unless the upper rider yields to the lower each degree by which his position exceeds the one of balance. Only if the tortfeasor's wrongful gain and the victim's wrongful loss are related in this way does the satisfaction of the victim require that the tortfeasor surrender his gain to him; for only then can we say that the victim's loss is the tortfeasor's gain and vice versa.114 Moreover, the correlativity condition must obtain if the same payment of damages is to annul the wrong and vindicate the right, for only in that case does the gain surrendered by the wrongdoer equal the deficit recovered by the victim. If plaintiff's wrongful loss and defendant's wrongful gain are not correlative, then the defendant's surrender of his gain to the plaintiff either overcompensates or undercompensates the plaintiff, and conversely the compensation of the plaintiff either leaves the defendant with some of his wrongful gains or takes these gains and then some. A perfect match would be a mere coincidence.

The problem, however, is that, while the correlativity of gain and loss is evident in wrongs involving infringements of property, it is not so evident in wrongs involving negligent inflictions of harm. In the former case, the defendant claims an authority over something that rightfully belongs to

the plaintiff. The plaintiff's right can be vindicated only by the defendant's relinquishing to him the authority he wrongfully claimed, either by paying nominal damages or by returning a value equivalent to the one he took. However, where the wrong consists in an infliction of harm without an usurpation of proprietary sovereignty, we could not earlier see the correlativity between plaintiff's loss and defendant's gain. At first sight, the defendant's gain from the transaction equals the wealth he saved by forgoing the required precautions, and this amount bears no necessary connection to the loss suffered by the plaintiff. In view of this asymmetry, one might be attracted to a system of remedial justice that required wrongdoers to disgorge their wrongful gains to a public fund from which reparations would be made to victims of wrongdoing.115 Why does the common law assume that the same sum of money annuls the defendant's wrong and vindicates the plaintiff's right? Why does it assume that remedial justice requires that the negligent tortfeasor compensate the victim?

Having come this far, we can now see that in the case of negligent wrongdoing too, there is a correlativity between plaintiff's loss and defendant's gain. Negligence consists in an infringement of a right to security against harm through an exercise of liberty greater than that which the other can, consistently with his equal end-status, recognize as valid. Just as in trespass, therefore, the wrong involves an assertion of one's end-status at the expense of another's. This means that the wrong suffered by the victim and the wrong committed by the tortfeasor are strictly correlative. The wrong suffered is a debasement of self that presupposes the self-elevation of the wrongdoer; and the self-elevation of the wrongdoer entails the debasement of the victim. The defendant's wrongful gain is not the savings realized in forgoing precautions (he would have been a wrongdoer even had the precautions cost nothing); rather, it is the preeminence vis-à-vis the plaintiff signified by his act, and this preeminence is correlative to the wrongful humbling of the plaintiff. The correlativity of wronging and being wronged means that there can be no satisfaction for the victim unless the wrongdoer gives it and no atonement for the wrongdoer unless he atones to the victim. It means that the victim's fight to redress is a right exclusively in relation to the one who wronged him and that the wrongdoer's duty to make amends is a duty specifically to the person he wronged.116

That the victim's remedial fight and the wrongdoer's remedial duty are exclusively in relation to each other explains the importance in classical tort law of a causal nexus between plaintiff's injury and defendant's negligent conduct.117 Correspondingly, it explains the availability of defenses (such as contributory negligence) involving the claim that, while the plaintiff has been wrongfully injured and the defendant has been negligent in relation to him, nevertheless the defendant is not bound to re-

pair the injury because his negligent conduct did not cause it. The requirement of a causal link between plaintiff's injury and defendant's negligent conduct serves to pair the persons whose wrongful gains and losses are correlative. It excludes from liability to the plaintiff ail those whose wrongdoing is not at the plaintiff's expense; and it denies a fight of recovery from the defendant to all those whose loss is not obversely related to the defendant's gain.

So far we have explained why the victim's remedy must move from the tortfeasor and why the tortfeasor's amends must be made to the victim. We have now to explain the specific form the remedy takes—a sum of money that compensates the victim for his or her loss of welfare.

The remedy that simultaneously annuls the defendant's wrongful gain and actualizes the plaintiff's fight is one that restores the reciprocity of respect and concern as between plaintiff and defendant. Since this reciprocity was disturbed by the wrongdoer's self-exaltation at the plaintiff's expense, it can be restored only through the wrongdoer's unilateral self-humbling before the plaintiff, one just sufficient to relinquish the preeminence acquired by the tort and so to raise the plaintiff to the antecedent position of equality. The payment of damages to the plaintiff is this act of deference. Because the defendant's wrongful gain consisted in his arrogation of an unrealizable self-worth rather than in material profit, one and the same act of deference can yield up the defendant's wrongful gain and repair the plaintiff's wrongful loss.

One might object, however, that this account of corrective justice does not explain the damage award. We can see why the plaintiff's remedy must consist in the wrongdoer's unilateral submission to him. But why must this act take the form of a reparation of the plaintiff's loss of welfare? It would seem that the tortfeasor's amends could as easily take the form of an apology or a payment of token damages for an affront to a dignitary interest.

That the tortfeasor must compensate the victim for his loss of welfare follows from the concrete conception of personal dignity that underlies negligence law. The latter is based on an explicit awareness that the person's end-status is real only in a recognized fight to security for the embodiments of freedom; that the claim of personal dignity is objective not only in a mutuality of respect for liberty (as in trespass and nuisance) but also in a mutuality of concern for welfare. Hence the deference required of the tortfeasor is not simply a submission to the victim's formal self; such a remedy would imply that the person's dignity is radically independent of its worldly circumstances—an implication of formal right that negligence law has forsworn. Rather, the act of deference called for is one that acknowledges what the tort denied, namely, the victim's right to as much concern for his human welfare as is consistent with the equal liberty and security of both. The tortfeasor acknowledges this fight by restoring the

victim to the level of welfare he enjoyed prior to the transaction that wrongfully set him back.