517297 Buffalo Law Review 62.2_R2.ps 451 Corrective Justice as Making Amends ERIK ENCARNACION† INTRODUCTION Many tort theorists try to explain tort law in terms of corrective justice.1 Formulations vary, but traditional accounts of corrective justice hold roughly that one person who wrongfully injures another has a duty to repair the injury or offset the losses resulting from that injury.2 Suppose that Alice negligently breaks Bob’s wrist. According to † Ph.D. Candidate, University of Southern California, School of Philosophy; J.D., Columbia Law School. The author would like to thank Rima Basu, Sharon Lloyd, Andrei Marmor, Caleb Perl, Scott Soames, Gabriel Uzquiano-Cruz, and Gary Watson for their comments on earlier versions of what eventually became Parts I through III of this article. Special thanks are owed to Steve Bero, Greg Keating, Elizabeth Lee, and Aness Webster, each of whom gave invaluable and detailed feedback on the whole article. Finally, very special thanks are owed to Scott Hershovitz, whose detailed criticisms substantially improved this article—or so the author hopes. This article was written with support from the University of Southern California’s Provost Fellowship. 1. See, e.g., JULES L. COLEMAN, THE PRACTICE OF PRINCIPLE 13-24 (2001) [hereinafter COLEMAN, PRACTICE OF PRINCIPLE]; JULES L. COLEMAN, RISKS AND WRONGS 361 (1992) [hereinafter COLEMAN, RISKS AND WRONGS] (“A loss falls within the ambit of corrective justice only if it is wrongful. . . . Corrective justice responds to such losses by imposing on individuals a duty to repair them.”); ERNEST J. WEINRIB, CORRECTIVE JUSTICE 9-37 (2012) [hereinafter WEINRIB, CORRECTIVE JUSTICE]; ERNEST J. WEINRIB, THE IDEA OF PRIVATE LAW 56-83 (1995) [hereinafter WEINRIB, THE IDEA OF PRIVATE LAW]; Martin Stone, The Significance of Doing and Suffering, in PHILOSOPHY AND THE LAW OF TORTS 131, 137-38 (Gerald J. Postema ed., 2001) [hereinafter Stone, The Significance of Doing and Suffering]. 2. See COLEMAN, PRACTICE OF PRINCIPLE, supra note 1, at 15 (“[I]ndividuals who are responsible for the wrongful losses of others have a duty to repair the losses”) (italics omitted); Martin Stone, On the Idea of Private Law, 9 CAN. J.L. & JURISPRUDENCE 235, 253 (1996) [hereinafter Stone, On the Idea of Private Law] (construing corrective justice in terms of one person’s being “answerable for the harmful effects of her actions on another”). Zipursky suggests the following formulation: “One who causes a wrongful injury to another is obligated to compensate the other for the injury caused.” Benjamin C. Zipursky, Civil Recourse, Not Corrective Justice, 91 GEO. L.J. 695, 700 (2003) [hereinafter Zipursky, Civil Recourse]. 452 BUFFALO LAW REVIEW [Vol. 62 traditional corrective justice theory, Alice incurs an obligation to compensate Bob for the losses associated with repairing the wrist, and Bob immediately obtains a correlative right to receive compensation from Alice. Corrective justice theorists claim that some moral principle of corrective justice explains why Alice and Bob stand in this bilateral relation to each other, morally speaking, and moreover, why tort law appears to embody or reflect this relation.3 In recent years, John C.P. Goldberg and Benjamin C. Zipursky, the leading proponents of civil recourse theory, have objected forcefully to corrective justice theory.4 They deny that corrective justice adequately accounts for important features of tort law.5 In particular, they claim that corrective justice cannot explain (1) the diversity of remedies beyond compensatory damages available in tort (such as injunctive relief and punitive damages); (2) “substantive standing” doctrines that prevent certain plaintiffs from obtaining relief (even in cases where they have been wrongfully injured by the defendants); and (3) the structural fact that a legal duty to repair the victim’s losses does not arise at the moment a tort occurs (even though corrective 3. This raises questions about what it means to “explain” a “central” feature of tort law. These questions are explored in COLEMAN, PRACTICE OF PRINCIPLE, supra note 1, at 25-63 (distinguishing several kinds of explanatory theories and arguing that economic accounts are not persuasive when understood as any of these kinds) and Zipursky, Civil Recourse, supra note 2, at 703-09 (discussing various interpretations of the bipolarity argument and Zipursky’s own version of “pragmatic conceptualism”—a methodology for interpreting legal institutions). 4. See, e.g., John C.P. Goldberg, Two Conceptions of Tort Damages: Fair v. Full Compensation, 55 DEPAUL L. REV. 435, 436 (2006) [hereinafter Goldberg, Two Conceptions] (aiming to “complicate” the apparently tight link between remedies in tort law and the aspiration to make plaintiffs “whole”); John C.P. Goldberg & Benjamin C. Zipursky, Torts as Wrongs, 88 TEX. L. REV. 917, 953-54 (2010) [hereinafter Goldberg & Zipursky, Torts as Wrongs] (claiming the explanatory superiority of their own wrongs-and-recourse model of tort law to theories of corrective justice); Zipursky, Civil Recourse, supra note 2, at 709-24. 5. John C.P. Goldberg, Twentieth-Century Tort Theory, 91 GEO. L.J. 513, 576 (2003) [hereinafter Goldberg, Twentieth-Century Tort Theory] (criticizing corrective justice for, among other things, rendering “the well-established institution of punitive damages problematic”); Zipursky, Civil Recourse, supra note 2, at 709-10. 2014] MAKING AMENDS 453 justice seemingly predicts that such a correlative duty would arise simultaneously with the right to reparations).6 Collectively, these claims purportedly show that corrective justice cannot explain or justify important features of tort law.7 This paper outlines a new conception of corrective justice capable of responding to these attacks.8 To see what motivates the conception, we will begin with some background. Part I explains the three main objections to traditional corrective justice theories advanced by civil recourse theorists. Part II sets out a recent attempt by Scott Hershovitz to revise corrective justice theory in response to these objections.9 Unfortunately, Hershovitz’s theory—the “getting-even” conception of corrective justice—faces a new set of difficulties, which are discussed in Part III. As we will see, the main difficulty is that Hershovitz’s theory blurs the 6. See, e.g., Zipursky, Civil Recourse, supra note 2, at 709-24; John C.P. Goldberg, Wrongs Without Recourse: A Comment on Jason Solomon’s Judging Plaintiffs, 61 VAND. L. REV. EN BANC 9, 13 (2008) [hereinafter Goldberg, Wrongs Without Recourse], http://www.vanderbiltlawreview.org/content/articles/2008/11/ Goldberg-61-Vand-L-Rev-En-Banc-9.pdf. 7. This would also be somewhat ironic because corrective justice theory’s crowning achievements include undermining influential economic explanations of tort law. For the relevant economic explanations, see generally GUIDO CALABRESI, THE COSTS OF ACCIDENTS (1970); WILLIAM M. LANDES & RICHARD A. POSNER, THE ECONOMIC STRUCTURE OF TORT LAW (1987); STEVEN SHAVELL, ECONOMIC ANALYSIS OF ACCIDENT LAW (1987); Richard A. Posner, A Theory of Negligence, 1 J. LEGAL STUD. 29 (1972). The arguments against these approaches appear in numerous places. See, e.g., COLEMAN, PRACTICE OF PRINCIPLE, supra note 1, at 13-24; Stone, The Significance of Doing and Suffering, supra note 1, at 142- 52; Ernest J. Weinrib, Understanding Tort Law, 23 VAL. U. L. REV. 485, 503-10 (1989) [hereinafter Weinrib, Understanding Tort Law]. 8. See Scott Hershovitz, Harry Potter and the Trouble with Tort Theory, 63 STAN. L. REV. 67, 69 (2010) [hereinafter Hershovitz, Harry Potter] (“[T]o generate an adequate corrective justice account of tort, we must revise our understanding of what corrective justice is.”). Another way defenders of corrective justice might respond is by disputing the purported importance of the recalcitrant data that civil recourse theorists rely on to motivate their attacks. A detailed version of this reply can be found in WEINRIB, CORRECTIVE JUSTICE, supra note 1, at 97 (“Punitive damages are inconsistent with corrective justice for reasons both of structure and of content.”); John Gardner, Torts and Other Wrongs, 39 FLA. ST. U. L. REV. 43, 53-54 (2011) [hereinafter Gardner, Torts and Other Wrongs] (emphasizing that the only form of relief awarded “as of right” in tort law is compensatory damages). 9. Scott Hershovitz, Corrective Justice for Civil Recourse Theorists, 39 FLA. ST. U. L. REV. 107, 126 (2011) [hereinafter Hershovitz, Corrective Justice]. 454 BUFFALO LAW REVIEW [Vol. 62 line between retributive and corrective justice and hence distorts rather than illuminates tort law. The point of presenting and criticizing Hershovitz’s views is not simply to expose the shortcomings of a particular theory. Scrutinizing the limitations of Hershovitz’s approach will help us avoid pitfalls en route to an alternative. Part IV sketches a new account, which will be called the “making amends” theory of corrective justice. Roughly put, the principle of corrective justice that will be defended holds that individuals who are responsible for the wrongs that happen to others have a duty to make amends to them unless the victims of those wrongs do not want them to. Tort law, in turn, should be understood as a public institution that aims to facilitate the amends-making process by mitigating certain recurring problems that occur in the informal amends- making process, while protecting the morally important interests of victims in controlling aspects of that process. Reparative relief should be understood, moreover, as the default mode of making amends. There is a lot to unpack in these claims. To preview, notice two differences between the new approach to corrective justice and traditional, repair-based conceptions: first, rather than a duty of repair, the principle contains a duty to make amends. Because the manner in which one makes amends is more flexible than how one repairs wrongful losses, this allows us to respond more readily to objection (1) that corrective justice cannot account for the variety of remedies available in tort. As we will see, making amends may require reparations, but need not. Making amends may require more or less depending on the circumstances. The second salient difference is the inclusion of an “unless” clause. Most statements of the principles of corrective justice overlook the fact that there are limits on the duties of wrongdoers to respond to their victims. One overlooked limit is that sometimes victims do not want to interact with the wrongdoer and do not want them to try to make amends. Understanding this limit, as well as problems that routinely crop up in informal amends-making processes, will be important in answering the structural objections proffered by civil recourse theorists (mentioned above at (2) and (3)). The story is complicated but, again, we will see that the reason that tort law’s basic normative structure differs from that of 2014] MAKING AMENDS 455 corrective justice is that tort law aims to overcome practical obstacles to the amends-making process while also serving to protect the victims’ morally important interests in controlling aspects of that process. Explaining all of this in greater depth will be the task of Part IV. After sketching out the making-amends conception of corrective justice and how it can be used to explain tort law, the making-amends conception will be tested against various objections in Part V. As we will see, the new account stands up to a range of objections including the civil recourse critique. We will conclude by observing that instrumentalist appeals to corrective justice can help explain and justify why that legal structure is the way it is and does not need to carve tort law at the joints in all ways. I. THE CIVIL RECOURSE CRITIQUE OF CORRECTIVE JUSTICE Corrective justice theorists try to explain tort law’s key features in terms of moral principles of corrective justice. There is no canonical statement of these principles. For our purposes we will treat Jules Coleman’s statement as representative: “individuals who are responsible for the wrongful losses of others have a duty to repair the losses.”10 To be sure, not all corrective justice theorists endorse this formulation. And it raises several worries, some of which are peculiar to his view.11 But for the purposes of fixing Goldberg and Zipursky’s target, the principle will do just fine, since the civil recourse critique applies to a wide array of formulations including Coleman’s. With the target in place, let us turn to the tripartite critique of traditional corrective justice theory posed by the civil recourse theorists. Let us call the objections: the remedies objection, the “substantive standing” objection, and 10. COLEMAN, PRACTICE OF PRINCIPLE, supra note 1, at 15. 11. For example, how do we square the principle with strict liability doctrines? Those doctrines seemingly allow liability without wrongdoing. But see, e.g., Gregory C. Keating, Property Rights and Tortious Wrongs in Vincent v. Lake Erie, 5 ISSUES IN LEGAL SCHOLARSHIP 3-4 (2005) (discussing views holding that the relevant wrong in strict liability cases is the failure to make voluntary reparations). 456 BUFFALO LAW REVIEW [Vol. 62 the no-legal-duty objection. Understanding them will be crucial for understanding what follows. A. The Remedies Objection The first objection claims that corrective justice theories cannot explain why courts regularly award non- compensatory relief.12 The objection runs as follows. According to corrective justice theory, defendants have a duty to repair wrongful losses. But duties to repair are compensatory in nature: satisfying one’s duty of repair involves identifying the costs wrongfully incurred by the victim as a result of the wrongdoing and trying to offset those losses or undo the wrong to the extent possible.13 For example, if one is responsible for breaking another’s wrist, the responsible party should at least pay for the losses associated with repairing the wrist, including paying for related medical costs. And it is a well-known commonplace that tort law does, in fact, award compensatory damages to plaintiffs in precisely this way. One Aristotelian metaphor often used to capture this commonplace is that damages awards in tort aim to make victims “whole.”14 The alleged problem, however, is that many remedies besides compensatory relief are available in tort suits.15 12. See Zipursky, Civil Recourse, supra note 2, at 710. 13. See Goldberg, Twentieth-Century Tort Theory, supra note 5, at 573-74. 14. E.g., Che v. Mass. Bay Transp. Auth., 342 F.3d 31, 43 (1st Cir. 2003) (“We have recognized that reinstatement is an important remedy because it ‘most efficiently’ advances the goals of Title VII by making plaintiffs whole while also deterring future discriminatory conduct by employers.”); De Lude v. Rimek, 115 N.E.2d 561, 564-65 (Ill. App. Ct. 1953) (“The controlling principle is that where compensation is the objective of the law, recovery is limited to the damages sustained, and any payments made by MacNevin to the end of making plaintiffs whole must be deducted from the recovery in this action.”); Senn v. Manchester Bank of St. Louis, 583 S.W.2d 119, 135 (Mo. 1979) (claiming that there was authority for having “substituted money for the land itself as a means of making plaintiffs whole”). 15. Zipursky, Civil Recourse, supra note 2, at 710 (“The problem is that courts do many things in tort law once they have decided that the defendant committed a tort upon the plaintiff; the imposition of liability for the wrongful injury created by the defendant is simply one of many remedies granted—a particular form of compensatory damages.”). 2014] MAKING AMENDS 457 Punitive damages, nominal damages, and injunctions are regularly awarded.16 And often we cannot fairly construe these forms of relief in terms of reparations. Punitive damages seemingly aim to punish, not repair.17 Nominal damages have nothing to do with reparative relief.18 And because many forms of injunctive relief aim primarily to prevent or force future conduct, injunctive relief is hard to square with compensatory damages concerned with past wrongdoings.19 The fact that tort law regularly awards these forms of relief seems aberrational when viewed through the lens of traditional corrective justice theory. But they are not aberrational, according to Goldberg and Zipursky.20 They are central features of tort practice, and any explanatory theory of tort law (like corrective justice) that fails to account for them, or that treats them as ancillary or extraordinary, counts as a major shortcoming in that theory.21 The remedies objection points toward a deeper worry. The existence of diverse remedies highlights a fact that corrective justice theories have difficulty grappling with: tort law separates the question of whether a plaintiff has a right to action from the question of the appropriate nature of the remedy that should apply.22 This is why a diversity of remedies is available in tort.23 But corrective justice illicitly unites wrongdoing with a particular type of relief (compensatory remedies) because corrective justice implies 16. Id. at 711. 17. See id. at 750 (claiming that punitive damages “are actually seeking to vindicate their rights by inflicting a sanction on the defendant”) (emphasis added). 18. Id. at 711 (“By definition, these types of damages [such as punitive and nominal damages] do not concern responsibility for the loss created.”). 19. See id. at 711 (“Corrective justice theory is similarly unable to explain why a variety of injunctive remedies are available.”). 20. See id. at 711-13. 21. See id. 22. See id. at 711-12. 23. Id. (“The diversity of remedies indicates that the issue of whether there is a right of action in tort is distinct from the issue of what the remedy should be.”). 458 BUFFALO LAW REVIEW [Vol. 62 that a finding of wrongdoing immediately triggers a duty to repair.24 The law, however, contains no “direct link between the notion of a right of action and the imposition of liability” in compensatory form.25 A finding of a legal wrongdoing does not automatically trigger a particular form of remedy. We will see this point again from a different angle when we address the no-legal-duty objection. B. The “Substantive Standing” Objection The second objection claims that corrective justice allegedly fails to account for so-called “substantive standing” requirements, which appear in most, if not all, causes of action recognized in tort law.26 According to these requirements, not only must a plaintiff show that she has suffered injuries resulting from the defendant’s tortious conduct, a plaintiff must also stand in the right kind of status, as a victim, in relation to the defendant’s wrongdoing in order to properly state a tort claim.27 In other words, “[a] plaintiff may recover against a defendant for a tort only if the defendant’s conduct was tortious relative to the plaintiff.”28 The point is subtle. Perhaps the best way to flesh it out is through examples. Zipursky provides several of them in his seminal article, Rights, Wrongs, and Recourse in the Law of Torts.29 He points out that, in raising a defamation claim, a plaintiff cannot win unless she can show that “she herself 24. Id. at 712. Zipursky states: [I]t is possible for the plaintiff to have a right of action in tort without reaching the question of whether [the] defendant has a duty of repair. So, although the commission of a tort by the defendant gives rise to a right to some sort of remedy in the plaintiff, the existence of this right to a remedy cannot be dependent upon the plaintiff being the owner of a loss and therefore the beneficiary of the defendant’s duty of repair. Id. 25. Id. at 713. 26. Benjamin C. Zipursky, Rights, Wrongs, and Recourse in the Law of Torts, 51 VAND. L. REV. 1, 4 (1998) [hereinafter Zipursky, Rights, Wrongs]. 27. Zipursky, Civil Recourse, supra note 2, at 714. 28. Id. (emphasis added). 29. Zipursky, Rights, Wrongs, supra note 26, at 17-19. 2014] MAKING AMENDS 459 was defamed.”30 This requirement is reflected in the so-called “of and concerning” element of the defamation tort, which holds that defamation plaintiffs must show that the allegedly defamatory statements are “of and concerning” the plaintiffs.31 This is so even if the plaintiff can show that the defendant made a defamatory statement that injured her foreseeably.32 Likewise, plaintiffs alleging fraud must allege more than an injury flowing from the defendant’s intentional deception; they must further allege that they themselves relied on the deception.33 To illustrate the point in the law of negligence, Zipursky cites Cardozo’s famous opinion in Palsgraf v. Long Island Railroad.34 In Palsgraf, Palsgraf’s injury was caused by the wrongful conduct of the railroad’s employee, yet the court concluded that the railroad owed no duty to her in particular, and as a result, she could not recover from the railroad.35 Showing that an injury flowed from wrongful conduct of others is not sufficient to justify recovery. All of these cases illustrate, according to Zipursky, that “tort law declines to impose liability on defendants in favor of the bearers of . . . wrongful losses” without establishing that the plaintiff bears the right relation as a victim of the defendant’s tortious conduct.36 This is the hallmark of tort law’s substantive standing requirements. To see how these observations are supposed to undermine traditional corrective justice theories, Zipursky claims that any attempt to explain these ubiquitous “substantive standing requirements” in terms of corrective 30. Id. at 17. 31. Id. 32. Id. 33. Id. at 18-19. 34. Zipursky, Civil Recourse, supra note 2, at 715 (citing Palsgraf v. Long Island R.R., 162 N.E. 99, 99 (N.Y. 1928)). 35. Palsgraf, 162 N.E. at 99. Cardozo declined to see the issue (of whether Mrs. Palsgraf could recover) in terms of causation—proximate or otherwise. Id. at 101 (“The law of causation, remote or proximate, is thus foreign to the case before us. The question of liability is always anterior to the question of the measure of the consequences that go with liability.”). 36. Zipursky, Civil Recourse, supra note 2, at 714. 460 BUFFALO LAW REVIEW [Vol. 62 justice faces a dilemma.37 To the extent that corrective justice theorists try to explain tort law in terms of genuine principles of corrective justice that contain plausible conceptions of moral responsibility, Zipursky claims that those conceptions of moral responsibility would impose liability in many cases in which tort law would not, given the restrictions imposed by substantive standing requirements.38 This is the first horn. As for the second horn, Zipursky claims that, to the extent that some proposed principles of corrective justice could explain substantive standing requirements, they no longer reflect or contain independently plausible conceptions of moral responsibility.39 To see this dilemma at work, consider Jules Coleman’s proposed principle: “individuals who are responsible for the wrongful losses of others have a duty to repair the losses.”40 On one interpretation, Zipursky would be prepared to acknowledge that this is a plausible moral principle. This “plausible” interpretation depends on a particular way of understanding what it means for X to be responsible for Y’s wrongful losses. If this means that X is responsible for Y’s losses only if those losses are the reasonably foreseeable results of X’s conduct, then Zipursky is indeed willing to 37. See id. at 716-18 (criticizing the work of Stephen Perry (which states, in Zipursky’s view, a plausible conception of moral responsibility for outcomes but one that does not fit actual legal practice) and Jules Coleman (which, Zipursky seemingly claims, might make for a better fit with actual legal practice but which no longer accords with any plausible view of moral responsibility)). 38. See id. at 717. Zipursky states: Remarkably, the doctrines that fall under the rubric of substantive standing do not impose a duty of repair upon defendants even for reasonably foreseeable injuries caused by wrongful conduct. Thus, parents who are traumatized when a surgeon’s negligence on the operating table disfigures their child will not be able to recover from the surgeon for this trauma—even though our tort law now views emotional trauma as sufficiently real to be compensable, even though it regards the surgeon’s negligence as a legal wrong, and even though the emotional impact on parents of having their child disfigured is surely foreseeable. Id. 39. See id. at 718 (“If ‘fault’ is a placeholder for a nexus-requirement that already happens to exist in tort law, then it is an illusion that responsibility has been accounted for in terms of fault.”). 40. COLEMAN, PRACTICE OF PRINCIPLE, supra note 1, at 15. 2014] MAKING AMENDS 461 concede that Coleman’s principle counts as a plausible formulation of a genuine moral principle.41 The principle is plausible because, according to Zipursky, we simply are morally responsible for the foreseeable results of our conduct.42 Corrective justice, on this interpretation, refers to a true principle of moral responsibility. But this moral plausibility comes at a price, since the principle can no longer account for substantive standing requirements in tort law. To see why, Zipursky points out that substantive standing requirements often deny recovery even in cases where a plaintiff’s injury is a reasonably foreseeable consequence of tortious conduct.43 This is why, according to Zipursky, “parents who are traumatized when a surgeon’s negligence on the operating table disfigures their child will not be able to recover from the surgeon for this trauma”—even though the suffering of parents in such circumstances is reasonably foreseeable.44 This is also why plaintiffs alleging defamation must show not only that they were injured as a foreseeable result of the defendant’s defamatory statement; they must also show that the statement was “of and concerning” them.45 The first interpretation of Coleman’s principle,46 even though it renders the principle plausible as a genuine moral principle, would imply that we ought to impose liability in many cases 41. Zipursky, Civil Recourse, supra note 2, at 717 (acknowledging that “[t]he constraint of foreseeability in our non-legal practices of blame ascription reflects something deep in our notion of responsibility”). 42. Id. Actually, Zipursky associates this interpretation with the work of Stephen Perry on outcome responsibility, not Jules Coleman’s principle. See id. at 716-17 (citing Stephen R. Perry, The Moral Foundations of Tort Law, 77 IOWA L. REV. 449, 503-12 (1992)). But the same dilemma is supposed to arise for any plausible principle of corrective justice: either it is a true moral principle (and it is both over- and under-inclusive of the scope of actually recognized torts like defamation), or it is not over- and under-inclusive, in which case it no longer seems plausible as a true moral principle. So the present conflation of Perry’s and Coleman’s different views on corrective justice does not matter for purposes of the critique. 43. See Zipursky, Civil Recourse, supra note 2, at 717. 44. Id. 45. Zipursky, Rights, Wrongs, supra note 26, at 17. 46. See COLEMAN, PRACTICE OF PRINCIPLE, supra note 1, at 15. 462 BUFFALO LAW REVIEW [Vol. 62 where tort law does not. In short, substantive standing requirements sometimes prevent plaintiffs from recovering from defendants even though those defendants are morally responsible for the plaintiffs’ wrongful losses. The first interpretation falls on the first horn of the dilemma—that is, true moral principles of corrective justice exceed the reach of actually recognized torts. There is another interpretation of Coleman’s principle arguably capable of explaining why tort law contains substantive standing requirements. On this interpretation, advanced by Coleman himself, we should understand X’s being responsible for Y’s losses as implying X is at fault for Y’s losses, in the sense of being “within the scope of the risks that make the aspect of [X’s] conduct at fault.”47 How are we to understand fault here? The full answer is complicated, but at a minimum, the notion of fault somehow automatically incorporates “the kind of nexus between conduct and injury that tort law actually reflects.”48 In response to this proposal, Zipursky admits that construing the relation of being responsible for an outcome as Coleman suggests might account for all the relevant substantive standing elements.49 But Zipursky thinks that this answer also comes at a price.50 Coleman’s answer can explain substantive standing requirements but no longer seems plausible as a genuine moral principle. The concept of fault that Coleman invokes simply does not track ordinary moral notions of fault. Tort law sometimes refuses to find defendants liable—again, due to substantive standing requirements—even though the injuries did arguably fall within the “nexus” linking the conduct and injury, cases in which the defendant plainly appears “at fault.” But if Coleman’s concept of fault does not track our ordinary moral concept of fault, this means that Coleman’s version operates as an empty “placeholder” for the standing requirements 47. COLEMAN, RISKS AND WRONGS, supra note 1, at 346. 48. Zipursky, Civil Recourse, supra note 2, at 717. 49. See id. at 718. 50. Id. 2014] MAKING AMENDS 463 already found in tort law.51 And the same argument applies to any proposed specification of the being-responsible-for relation, whether couched in terms of fault or not. To the extent the being-responsible-for relation already presupposes or incorporates substantive standing doctrines found in tort law, the relation seems far less plausible as a genuinely moral one as opposed to being an empty vessel standing in for legally recognized standing requirements.52 To summarize: corrective justice theories aim to explain tort law in terms of genuine moral principles. But moral principles of corrective justice cannot explain substantive standing requirements because the principles imply more liability than those requirements allow, to the extent those principles are plausible as genuine moral principles.53 Adjusting these theories in order to accommodate substantive standing requirements simply renders those principles implausible as genuine moral principles of corrective justice. Theorists of corrective justice are thus gored on the horns of a dilemma. C. The No-Legal-Duty Objection This brings us to the final objection to corrective justice, which focuses on the timing of the duty to repair.54 Traditional, Aristotelian versions of corrective justice hold that a victim’s right to compensation arises immediately when one person wrongs another, and that the wrongdoer 51. Id. 52. Zipursky also claims that, even if we were to accept this understanding of responsibility as it appears in the principle of corrective justice, the law still imposes tort liability in cases where the principle of corrective justice would not. He claims, for example, some people can recover for “consequential damages” not part of the “predicate injury” itself. See id. at 718. This response appears to underscore the diversity-of-remedies issue and therefore does not include it in the discussion above. That is, if corrective justice theorists expanded their conception of the duty of repair to include more than simply reparations, this response does not seem to have much bite. 53. The law also imposes liability where morality arguably would not, such as cases of harmless, accidental trespassing. Id. at 726-27 (mentioning trespass as a case where tort liability would attach but moral blameworthiness would not). 54. See id. at 718-21. 464 BUFFALO LAW REVIEW [Vol. 62 immediately incurs a correlative duty to compensate the victim.55 Coleman’s principle is representative of the Aristotelian tradition: “individuals who are responsible for the wrongful losses of others have a duty to repair the losses.”56 One would naturally expect tort law to reflect this structure if tort law were explicable in terms of principles like Coleman’s. That is, if tort law’s basic structure were a legal correlate to the moral principle of corrective justice, a tortfeasor would have an affirmative legal duty to repair that would arise the moment she commits a tort against a victim. But tort law does not work this way. Goldberg articulates the point succinctly: Notice, however, that the conversion of the moral duty of repair into a legal duty does not happen through the tort system unless and until the victim decides to press a claim against the defendant. In other words, if the defendant is going to be made to heed his duty of repair, it will only be by virtue of the law’s having empowered the victim to demand of the defendant that he make amends for the wrong done. . . . Corrective justice theory thus fails to capture accurately the terms on which tort links a victim to a person who has victimized her.57 In other words: if X breaches his duty of care towards Y and wrongfully causes Y’s injuries as a result, nothing in tort law imposes on X a duty to repair Y’s injuries at the very moment the injury occurs. At most, X is liable to Y, but this does not yet mean X incurs a legal obligation to compensate Y. Civil recourse theorists thus maintain a sharp distinction between a legal liability and an affirmative obligation to pay.58 Corrective justice implies the latter; tort law holds that only the former occurs when someone wrongfully injures another. The duty of repair, in short, is not automatic in the way that the principle of corrective justice suggests. 55. See, e.g., Ernest J. Weinrib, Toward a Moral Theory of Negligence Law, 2 LAW & PHIL. 37, 38 (1983). 56. Id. 57. Goldberg, Wrongs Without Recourse, supra note 6, at 13 (emphasis added). 58. Zipursky, Civil Recourse, supra note 2, at 720-21. 2014] MAKING AMENDS 465 To bolster this claim, Zipursky compares contractual liability with that of liability in tort law.59 Courts award prejudgment interest to plaintiffs that win breach-of-contract claims on the theory that defendants are obligated to pay plaintiffs at the time specified by the original contract.60 Prejudgment interest is justified on the grounds that the defendant’s subsequent delay of payment, due to litigation, further deprived the plaintiff of the time-value of the money the defendant was obligated to pay at the moment the contract was breached. In cases arising from breaches of contract, liable parties are deemed liable as of the moment of breach, and hence, the breaching party is also responsible for the victim’s lost interest.61 But prejudgment interest is not normally awarded in torts cases.62 Why? According to civil recourse theorists, a defendant in a torts case, unlike a defendant in a breach-of-contract case, is not legally obligated to compensate the plaintiff at the moment that the defendant injures the plaintiff.63 The duty to pay arises, if at all, only after a finding of liability.64 But corrective justice seemingly predicts the opposite result. Not only are defendants liable at the moment they wrongfully cause the plaintiffs’ injuries, but, according to corrective justice, they should be liable for reparative relief, most likely in the form of compensatory damages. So we should expect that prejudgment interest would begin to accrue the moment the wrongdoing causes the plaintiff’s injury. Since civil recourse theorists claim tort law does not work this way, corrective justice apparently fails to account for a major feature of tort law’s basic normative structure. 59. See id. at 719. 60. Id. (explaining that “one who fails to pay under a contract will incur prejudgment interest because payment is owed at the time the contract specifies for performance, not at the time a court reaches a judgment”). 61. See Hershovitz, Corrective Justice, supra note 9, at 109. 62. Id.; Zipursky, Civil Recourse, supra note 2, at 719-20. 63. See Hershovitz, Corrective Justice, supra note 9, at 109. 64. See id. ���� ���������� � �� � � ���� ��� �� ��! +!���<+�!�!< �><�!���>< * � 4$���������$���������� �� �������� �� � � `���� ��`����� ����`������ �`��� ���� ����£ �������� �� ����������� � �� � `���� � � #\�����\� ����X�`��`������ 65 *``�����\ �� � �� ����� `����`���� �����`� �``��� � �� � ��`��~ �� � ����\����\ �� ��~� ����� #\��� ����X ��� � � ����\���� 66 �� ����� ���� ��� ��~� #�� \��� ������ � � ��� ���� ����\�� �� ����������� �� \�� ���� X67 ��`���� � � `��`��� �� \�����\ ���� �� `������ �� ���� ����£^� ����������� �� `����`���� �����`� ��� ���� ���� �� �� ���� `���������\ �� ������ � � � ���� #\�����\ ����X �~����� �����\� 68 ��� ���� ����£ `�������� � � ���� �� �������� ~��� ������� �� ���� �� �� `������� � � ������ ���`��`�� �� � �` ��� ������ ������� �� �������\ ����\����� ��� ���� �� 69 �� ���� ��� ���~���� `������� ��~ ��� ����� 70 ������� � �� � �� ��� �������� ��� � �� ����� ����� ��~ ��~� ~���� 71 <�� ������� � ��� ��~���~� ����� ��� ������ ��~ `�� ��� ����� ��`�� ��~ ������~� ��~� \��� ���� ��� ����� 72 *``�����\ �� ���� ����£� ����� �� � � ������� �� ��`���� ��~ #����X �� #������X ��� �� �� �� �� � ��� �� �@ �� {{} ���� ����£ �� �� ����������� ����`���� �� ����������� `����`���� �����`� ���� ��@ �� {{��{� ���\���\ � �� `����`���� �����`� �������� �� �``�~���� � � �~��������� ��~���� ������\ ����\��� �`�� �� �� ������ ~����\ ��`��~� � ���� �����`� �� �� ��� � �� ���� ����£ � ���� � �� �� ��� ������` ������ � �� ����� *�� ��\ �� �� ��� ������� � �� ���� ����£ ��``����� � �� ����� ��`���� �� � � `���� ��`����� ����`����� �� ���� � ��\� ��~��� �� � �@ �� {�� �� � �@ �} >­�>$� !<+@��� �����><*$¨ ��� ��� �� {|}|� ��������\ #�� \�� ���� ���� �X �� #�� ���� ���^� �����\� ����� �� ��������� ��\������X� �| ���� ����£� %����������������������������� |� �� {�{ �#>�� ���\~���� ����� � �� �� ����� �� \�� ������ ���� ��� ����`�� X� �� � �@ �� {{}�{| ����`�����\ � � ��~���������� `���� �{ � �@ �� � �@ ��{��� ��� �K�����;� ���� � �� \�������� ������� ��~^� �����������\ ���� 73 ����� ~��� �� � ���� ��`���� ��~ ��� ����� ���� ��� ��� ���� �� 74 ���� ����£ ���� ��~ ��� ����� �� ���������� ��~� �� � � ��\�������� ����`���� ������ �� ������ ������� ���� ��� ��` �� �� � � ����� � ��\ �� ����`� �� � ��� �� ����� ��� ��~ ��� ����� �� �� ����� ����� ~��� ������~ ��~� �`� �� � �� `���� � � ������~��`� `�~�� �� � � ���~ �� � ����` �`�� �� � �` ����� ��`����� ��~�� ��\ ���� #��^�� ����X �� #��^�� ������ X75 �� ~�\ � `��� � �� ���� �� ����` �`� � ������������ �#� �������� 76 ���� ����£ ������� � �� \�����\ ���� �� ���#��������� ��� ��`� � �� \�����\ ���� ���� ������� ��` � ����` �`� 77 @��� ��� ����` �`�� � �� �� � ��`�������� �� �������� ��������� �� ��������\ ��~ ��� ����� ���� ��� ��` �� �� ������� �� #������7� �������������`�\����� `��������� � �� ~��� �� ��������� �� ����� ��� � � ������~����� �� ��``��� 78 � �� �� � ����` �`� ���� � �������`�~��� �� ~�����\� ��``���� ������� ��� ���� �� � �� �� �� ����`���� ������ � � ������`� #� ��� �������`� ��� ������ ��� �����X ��� ���� �� �� �� ��`�� ���� � �� �� � � ����`���� �� ��� ������ �� ����`���� ~�����\�� 79 @�������� � �� �� �����^� ��`�������� �� �������� ��``������� ����� ��� ����� ��� ��~ �� �� ��`� �� ���� ��� ��` �� ���~��� ��`�\����� `��������� ~��� �� ��������� � ��� `��������� ��`����� ��� ��� ��� ��~���� ��� � � ��������\� • ����!��������@���~� ����\����� ~��� ������� ��� ���~ ��� ����� �� � ��`��� ����� �� � � `��� �� ��~ ��� ����� � � ����\����� �� � �����������\ ���� ��� � ������������ �~���� �� �� � ������@��� {{| �� � ������@ �� ������@ �� � ��� ��@��� {�} �� �@ �� {{| ��� ~��� �� ����` �`��� ��� \�������� � @ *����<� �>� �> �> ���<+� ���� �>$�� �� > ��~��� ª ������ ����® ��� � �� �� {|��� �} � ���� ����£� %������������������ ����� ���� |� �� {{| �| � �@ 468 BUFFALO LAW REVIEW [Vol. 62 money (we suppose).80 But if Jerry and Tom did not owe each other anything, it would not make sense to say that Jerry had the ability to declare Tom even with him. �� A Proportional and Appropriate Act. Another condition is that Tom must have done something for Jerry that is somehow roughly proportional to the debt he owed.81 If Tom’s favor to Jerry were extremely small as compared to the amount owed, then it would not make sense for Jerry to call Tom even.82 One might think that the better characterization of Jerry’s waiver of the debt would be as a gift to Tom— though Hershovitz himself does not say this explicitly.83 But this does not mean that the value of the favor must equal that of the debt.84 The act must also be appropriate, insofar as it’s reasonable to assume that “there are moral limits on what sorts of things people can properly regard as proportional.”85 An act that involves intentionally making Jerry worse off would obviously not do. �� Acceptance. Hershovitz notes that, ideally, Tom and Jerry “are not even until they jointly decide to regard one another as even.”86 Tom and Jerry ideally must be persuaded that they are in fact even with each other. Hershovitz uses these insights to revise our understanding of corrective justice, and in turn, tort law. In the context of corrective justice, rather than a creditor declaring his debtor even, it is the victim of wrongdoing who obtains a power to declare the wrongdoer even once the victim has been wronged. Once we recognize that a victim of 80. Id. at 118. 81. Id. at 120 (“[P]rior to the declaration Tom must have done something which could plausibly count as grounds for [being declared even].”). 82. Id. (“Tom’s action must be proportional to the debt. As I said before, if Tom’s debt was significant and his assistance slight, Tom would likely resist Jerry’s declaration, but even if he acquiesced, Tom and Jerry would not be even.”). 83. Id. 84. Id. at 118 (denying that Jerry’s declaration of evenness means “that he has judged that Tom’s services are worth approximately the debt owed”). 85. Id. at 120 (“In addition to ruling out cases where Tom has done nothing, we can rule out cases in which the act in question harmed Jerry.”). 86. Id. at 119 (emphasis added). 2014] MAKING AMENDS 469 wrongdoing has this power, a new picture of corrective justice emerges. According to Hershovitz, corrective justice occurs when the victim and wrongdoer are validly declared even with each other.87 And, as we have seen, whether this declaration is valid depends on numerous background conditions. Thus, we can understand part of the process of corrective justice as making sure that these various background conditions occur. To illustrate, suppose that rather than a pre-existing debt to Jerry, Tom wrongfully injured Jerry’s leg.88 Again, many factors determine whether Jerry’s declaration of evenness succeeds. There must be a declaration of the kind, “we’re even,” Tom must actually do something for Jerry, and Jerry ordinarily must accept the act as adequate. Whether Tom’s act is adequate will in turn depend on a number of other factors. Suppose Jerry and Tom are friends and the injury was accidental.89 Perhaps a sincere apology might suffice to support a declaration of evenness.90 Suppose they are not friends but that Tom apologizes and offers compensation.91 Maybe this will suffice. But now suppose that Tom intentionally and maliciously injured Jerry.92 A mere apology and offer of compensation might not be enough to support Jerry’s declaration of evenness.93 In any event, corrective justice is fundamentally about securing a declaration of evenness between wrongdoer and victim. And to do so the wrongdoer incurs an obligation to do something for the victim in the aftermath of the wrongdoing, even 87. See id. at 118-20. 88. See id. at 121. 89. See id. (“It is hard to say without knowing something about Tom and Jerry’s relationship, and also just what Tom did. If Tom and Jerry are friends, and Tom was merely negligent in breaking Jerry’s leg, it is easy to imagine that a sincere apology will suffice.”). 90. According to Hershovitz, corrective justice potentially requires apologies. See id. But others disagree. Id. at 112 (“Many philosophers would parse this differently. They would say that returning the ball is a matter of corrective justice, but apologizing is not, even though it may be morally obligatory.”). 91. See id. at 121. 92. See id. 93. See id. 470 BUFFALO LAW REVIEW [Vol. 62 though the precise remedy is not specified. This is the getting-even conception in a nutshell. Notice how Hershovitz’s analysis contrasts with the Aristotelian’s conception. An Aristotelian interpretation of Tom and Jerry’s situation is straightforward: Tom would have to do what he can to help repair Jerry’s leg, which typically means compensating Jerry to offset financial losses resulting from his injury. Nothing more, nothing less. Anything else that we could ask Tom to do for Jerry would not be in the domain of corrective justice. In addition to exposing traditional theories of corrective justice to the remedies objection, Hershovitz claims that requiring reparations is a problematic feature of traditional accounts, since efforts to fully undo the injuries that flow from wrongdoing are bound to fail; after all, at a minimum, we cannot undo the time lost and energy expended in trying to fix things.94 B. Getting Even Through Tort Law As we have seen, Hershovitz thinks that corrective justice is about a victim’s getting even with a wrongdoer, and in turn, that parties are “even” with each other only if there has been a valid declaration of evenness declaring them even with each other. A declaration of evenness, moreover, is valid only when certain prerequisites have been satisfied. There must be, for example, a proportional and appropriate act by the wrongdoer that supports the declaration. And ideally, though not necessarily, both parties must accept the declaration of evenness. We still seem a long way from providing an account that helps illuminate tort law. After all, getting even seemingly depends on innumerable facts besides obtaining an adequately supported, joint declaration of evenness; tort law, by contrast, purports to dole out corrective justice while taking into account a much more limited set of variables. 94. See id. at 117 (“We cannot undo what we have done. No matter how hard we wish that we could turn back time when a trigger is pulled or a driver hits a child, we cannot.”); see also Scott Hershovitz, What Does Tort Law Do? What Can it Do?, 47 VAL. U. L. REV. 99, 110 (2012) [hereinafter Hershovitz, What Does Tort Law Do?] (“We can’t ‘reverse the wrongful transaction’ for someone who has been raped, or slandered, or falsely imprisoned.”). 2014] MAKING AMENDS 471 What’s more, people file lawsuits when they cannot secure joint settlements outside of the court system. But these are the right kinds of observations to make to understand the relationship between the getting-even account and tort law. The felicity conditions that must be satisfied for parties to get even with each other are, according to Hershovitz, complex and “nuanced.”95 And one person seeking a declaration of evenness may not get it from the other person. Nor is it surprising that parties often disagree sharply about what conditions must be satisfied for them to be even. And even if they agree about those conditions, they still might disagree about whether those conditions have been satisfied. Tom might think an apology suffices; Jerry might think the apology insincere or want something more. And all of these problems presuppose that both parties are willing to sit down and negotiate. What if Tom declines to negotiate? Absent a successful negotiation, and in certain societies that do not have elaborate legal systems, Hershovitz notes that one unilateral option for getting even is for the injured party to seek revenge.96 Here Hershovitz embraces the ordinary implication of the phrase “getting even” by pointing to old Nordic methods of dispute resolution as historical precedent.97 He details a case in which Norwegian merchants chop off an Icelandic man’s hand.98 The merchants are confronted by other Icelanders and asked to pay a large amount of money.99 When the Norwegians decline, the Icelanders respond by threatening to chop the hand off one of the other Norwegians.100 The Norwegians back down, agreeing to pay the requested price.101 95. Hershovitz, Corrective Justice, supra note 9, at 121. 96. See id. at 123. 97. Id. 98. Id. 99. Id. 100. Id. 101. Id. 472 BUFFALO LAW REVIEW [Vol. 62 This brutal practice—indeed, any practice of seeking or threatening revenge—seems destined to generate cycles of violence.102 To prevent violence from spiraling out of control, Nordic peoples selected an oddman—an impartial arbiter who tried to persuade disputing parties to agree to a resolution that both parties could regard as fair.103 Hershovitz describes selecting an oddman as a way of “outsourcing the performative necessary to get the parties even, and their work was subject to a familiar felicity condition”—i.e., getting parties to agree that they were even.104 And here is where the point becomes important for purposes of modern tort practice. The law steps in as the modern analogue to oddmen, adjudicating disputes only when the parties cannot themselves iron out how to get even with each other on their own: We prohibit private violence as a response to wrongdoing, but we maintain the institution of the oddman, in the form of judges and juries. When parties cannot negotiate their way their way back to even, we offer a judicial failsafe—compulsory process, followed by garnishment and attachment. A wrongdoer who will not bargain can be haled into court and forced to submit to a jury’s judgment as to what will render him even with his victim. Though courts are fond of saying that the plaintiff should be made whole, that is not in fact what juries are asked to do. They are typically instructed to award “fair and reasonable” compensation for a plaintiff’s injury, and in the cases where the wrongdoing is willful and wanton, they may go beyond, and award punitive damages too.105 So how do our courts achieve corrective justice? The same way that oddmen achieved corrective justice, by persuading us (including, ideally, the parties to the dispute) that the parties are even: “Whether courts succeed in doing justice depends on whether people regard the remedies awarded as sufficient to render prevailing plaintiffs even.”106 In some cases, compensatory damages will suffice, in others punitive 102. See, e.g., id. at 124. 103. Id. 104. Id. 105. Id. at 125. 106. Id. 2014] MAKING AMENDS 473 damages are enough, and still others, injunctive relief will be necessary.107 In sum, Hershovitz thinks getting even does a better job of representing what tort actually involves than ideas traditionally associated with corrective justice, such as repairing injuries or making plaintiffs whole.108 As we will see below, Hershovitz thinks that the getting-even conception resists the civil recourse broadside.109 Moreover, Hershovitz claims that the getting-even picture marks an improvement on the Aristotelian picture.110 After all, corrective justice aspires to do what it simply cannot do—i.e., putting victims back to where they were prior to being injured by repairing or annulling losses.111 But corrective justice is not primarily about repairing or annulling losses—it is about private parties getting even with each other, and failing that, “[giving] people who have been wronged an opportunity to get even” by invoking a nonviolent system able to impose evenness on them.112 One unilateral way someone can get even for wrongdoing is by taking revenge;113 another unilateral option is by filing a lawsuit. The success of the tort system depends on its being seen as imposing reasonable terms of evenness among disputants, including by imposing remedies that might require more than simply affording compensatory damages.114 107. Id. at 127. 108. Id. at 126. 109. Id. 110. Id. 111. Id. 112. Id. at 127. 113. Hershovitz, What Does Tort Law Do?, supra note 94, at 116-17 (claiming that “[t]he virtue in revenge was that it provided victims a way to get even with wrongdoers unilaterally” and that “tort suit plays a similar role to revenge”). 114. See Hershovitz, Corrective Justice, supra note 9, at 125 (“Whether courts succeed in doing justice depends on whether people regard the remedies awarded as sufficient to render prevailing plaintiffs even. That is, it depends on whether their performance is persuasive.”). 474 BUFFALO LAW REVIEW [Vol. 62 III. EVALUATING THE GETTING-EVEN CONCEPTION Having reviewed the three main objections to corrective justice posed by civil recourse theorists and Hershovitz’s attempted rehabilitation of corrective justice in terms of getting even, let us evaluate the getting-even conception. This Part begins by assessing whether Hershovitz’s theory accommodates the tripartite objection posed by Zipursky and Goldberg. Although Hershovitz’s account does provide the resources for adequate responses (once supplemented by additional explanation), we will see that there are strong independent reasons to reject his theory. This will help pave the way for a better approach to corrective justice by helping us avoid pitfalls. A. Does Getting Even Avoid the Remedies Objection? Recall the first civil recourse objection. It observed that tort law allows a broad range of remedies besides compensatory damages, even though only compensatory damages make sense in light of the Aristotelian’s duty of repair.115 Hershovitz’s conception, by contrast, is far more flexible because, on his view, getting even—or for parties to be persuaded that they are even—might require more than simply compensating.116 A maliciously calculated attack might require, for example, punitive or exemplary damages in addition to compensatory relief.117 A similar story might be told for injunctive relief: a serial trespasser might have a restraining order imposed on him.118 This might be required to get even. Notice that the kind of response Hershovitz provides is not unique to the getting-even approach. Hershovitz aims to characterize a duty that is consistent with a broader range of remedies beyond compensatory relief. But Hershovitz recognizes implicitly that the only way to do this is by increasing the level of generality at which the relevant duty is characterized. Out goes the relatively narrow duty of 115. See discussion supra Part I.A. 116. See Hershovitz, Corrective Justice, supra note 9, at 127. 117. See id. 118. See id. 2014] MAKING AMENDS 475 repair. In comes the much broader duty to reach (get?) even. And surely Hershovitz is correct to suggest that the broader characterization is a virtue since it is consistent with a broader range of remedies. But we must also acknowledge that increasing the level of generality has costs. Traditional corrective justice provides some relatively determinate and principled standards for evaluating remedies in tort law. To the extent that a jury renders awards far beyond repair-based, compensatory relief (where punitive damages are unavailable), courts are instructed to rein in those awards.119 And as John Gardner observes, compensatory relief remains the only relief available “as of right” in tort cases.120 Increasing the level of abstraction renders this feature of remedial practice in tort law mysterious. Why should compensatory relief be the default form of relief in tort law, and why are other forms so often characterized as “extraordinary”?121 Ironically, then, increasing the level of generality allows Hershovitz to accommodate more forms of relief (and thus avoid the remedies objection), but at the apparent price of being able to explain or justify fewer details about tort’s remedial 119. Absent that, lawyers typically have solid grounds for appeal. See, e.g., Zipursky, Civil Recourse, supra note 2, at 749. 120. See Gardner, Torts and Other Wrongs, supra note 8, at 62. Gardner also examines in detail theoretical implications of private actions based in equity for civil recourse theory and argues that equity-based claims differ fundamentally from torts in that the former gives remedial priority to restitution and injunctive relief, while the latter gives priority to reparations. See id. at 59. 121. For statements regarding the extraordinary nature of injunctive relief, see, for example, United States v. City of Philadelphia, 644 F.2d 187, 191 n.1 (3d Cir. 1980) (describing injunctions as an “extraordinary” form of relief); Lewis v. Spagnolo, 710 N.E.2d 798, 815 (Ill. 1999) (“A mandatory injunction is an extraordinary remedy which may be granted when a plaintiff establishes that his remedy at law is inadequate and that he will suffer irreparable harm without the injunctive relief.”) (emphasis added); State ex rel. Fenske v. McGovern, 464 N.E.2d 525, 528 (Ohio 1984) (describing injunctive relief as an “extraordinary remedy”). For statements regarding the extraordinary nature of punitive damages, see, for example, Linthicum v. Nationwide Life Ins. Co., 723 P.2d 675, 680 (Ariz. 1986) (calling punitive damages an “extraordinary civil remedy”); Rodriguez v. Suzuki Motor Corp., 936 S.W.2d 104, 110 (Mo. 1996) (“Punitive damages thus are like other cases requiring the clear and convincing standard of proof: the remedy is so extraordinary or harsh that it should be applied only sparingly.”) (emphasis added). 476 BUFFALO LAW REVIEW [Vol. 62 practices, such as the fact that compensatory damages are the default mode of relief.122 That is a cost Hershovitz willingly incurs.123 Accordingly, we should conclude that the getting-even conception does avoid the first objection as advertised. There is a general lesson here, which applies to any attempt to defend corrective justice by increasing the level of generality at which the relevant duty is described. That is, increasing the level of generality will face the prima facie problem of being able to explain fewer features of tort law’s remedial practices. Any account of corrective justice that de- emphasizes repair will have to explain why this is not too tough a pill to swallow. B. Does Getting Even Avoid the Substantive Standing Objection? Hershovitz claims that the getting-even approach also avoids the substantive standing objection, which argued that corrective justice could not account for tort law’s substantive standing elements.124 The idea here was that Palsgraf, for example, needed to do more than simply prove that the railroad company committed a wrongdoing causally connected to her suffering harm; she needed to show that the 122. This does not mean, however, that Hershovitz—or anyone else who prefers the increasing-the-abstraction strategy—has no response available. It is possible that traditional corrective justice theories view the default practice of providing compensatory relief as far too parochial. One could imagine a tort system, quite like ours, in which there were no kinds of remedies afforded as of right, and according to which the question of appropriate remedies were far more up for grabs. The fact that compensatory relief is afforded as of right, or deemed the most important way to achieve corrective justice is, on this view, a highly contingent artifact of the Anglo-American legal culture worthy of no special consideration. One potential difficulty with this view is that much of tort practice is contingent, yet we still make judgments about the relative importance of these contingent practices that we try to nevertheless accommodate in our explanatory theories. 123. And, as we will see in Part V, the making-amends approach also incurs this apparent cost. See infra Part V.D.1. But it will also be explained why this is not too tough a pill to swallow. See infra Part V.D.1. 124. Hershovitz, Corrective Justice, supra note 9, at 108; see also discussion supra Part I.B. 2014] MAKING AMENDS 477 company “wronged her.”125 This criticism does not touch the getting-even account, according to Hershovitz, since the point of tort law is not (as per the Aristotelian conception) to make a person whole, but rather to provide people opportunities to get even.126 Substantive standing requirements determine who “[has] cause to get even” by “pick[ing] out those who have reason to resent, and not simply regret, a tortfeasor’s behavior.”127 Hershovitz thus thinks that his account steers clear of this objection in a way that the Aristotelian conception does not: “The point is to give people who have been wronged an opportunity to get even, and one has not been wronged unless one’s rights were violated. The standing requirements in tort law attempt to pick out people who have cause to get even.”128 This is certainly part of the story. But only part: it is unclear why an Aristotelian could not avail himself of precisely the same response. Yet recall that Zipursky thought that this kind of response was inadequate.129 A dilemma purportedly befalls traditional views of corrective justice.130 On the one hand, principles of corrective justice would impose liability in cases where tort law will not. This is because, according to Zipursky, principles of corrective justice lack the moral equivalent to the substantive standing requirements contained in tort doctrine: a person who is foreseeably injured as a result of another’s negligence has a genuine moral claim for reparations against the injurer without having to establish anything else.131 But it does not accurately describe the basic structure of tort claims, since all tort claims further require that plaintiffs establish substantive standing. On the other hand, principles of corrective justice that succeed in accommodating substantive standing requirements no longer 125. Hershovitz, Corrective Justice, supra note 9, at 108. 126. Id. at 126. 127. Id. at 127. 128. Id. 129. See supra Part I.B. 130. See supra Part I.B. 131. See supra Part I.B. 478 BUFFALO LAW REVIEW [Vol. 62 appear plausible as moral principles according to Zipursky.132 Instead, they contain technical terms like “responsibility” that look like moral concepts, but which are only empty vessels that contain whatever features tort law happens to have, including substantive standing requirements.133 A more complete response to Zipursky involves rejecting his assumption that the scope of liability suggested by moral principles of corrective justice must map perfectly onto the scope of legal liability imposed by the substantive rules of tort law. Corrective justice theorists need not be committed to this position, nor does it seem a reasonable requirement that corrective justice reflect such a mapping. To see why, suppose, for example, that a newspaper reporter negligently reports a story about a government official in a way that harms his reputation severely. Morally speaking, corrective justice might suggest that the reporter ought to do something to compensate the official for the harms he has suffered as a result of the negligence. But the fact that there might be no legal liability in these circumstances may reflect accommodations to, say, the First Amendment. That is, there are fine-grained limitations on liability that no theory of corrective justice could plausibly be faulted for not “predicting,” such as First Amendment constraints on legal liability for defamation.134 After all, we should not expect a theory of corrective justice to provide a complete theory of social or political justice or, more narrowly, a theory of free speech. Goldberg and Zipursky’s complaint boils down, it seems, to the commonplace that the scope of legally actionable wrongs is not coextensive with those that are morally wrong, which is hardly a surprise, since as a general matter what is immoral is often not unlawful. In the same vein, corrective justice theory does not come with a pre-packaged theory of actionable social wrongs. 132. See supra Part I.B. 133. See supra Part I.B. 134. N.Y. Times Co. v. Sullivan, 376 U.S. 254, 279-80 (1964) (requiring a showing not of mere negligence, but of “actual malice”—i.e., “with knowledge that [the defamatory statement] was false or with reckless disregard of whether it was false or not” in order for a public official to maintain a defamation claim relating to his official conduct). 2014] MAKING AMENDS 479 Corrective justice tells us what to do in the aftermath of wrongdoing, but it does not tell us what those wrongs are. Courts and legislatures define what wrongs are actionable in courts, and in any event, it is obvious that not every moral wrongdoing is legally actionable, or should be. Rules of substantive standing are in the business of helping us determine who has been wronged and separating them from those who have simply suffered the unwelcome side effects of those wrongdoings. In short, criticizing corrective justice for failing to account for substantive standing requirements amounts to criticizing it for failing to account for doctrinal guides that help define, albeit imperfectly, who has been wronged. C. Does Getting Even Avoid the No-Legal-Duty Objection? Now recall the no-legal-duty objection, which is that corrective justice suggests, incorrectly, that defendants should owe a legal duty of repair that arises immediately upon wrongfully injuring victims, even though tort law does not recognize any such immediate duty of repair.135 To bolster this point, Zipursky claimed that tort law typically does not award prejudgment interest, which he sees as evidence that no legal obligation to repair or remediate arises the moment that tortious conduct occurs.136 Hershovitz claims that his conception has the resources to respond to this objection.137 But his response is incomplete and problematic in some ways. With his getting-even conception in tow, Hershovitz gives the following response: corrective justice imposes a duty on the wrongdoer to reach evenness with the victim, but the question of what is required for the parties to reach a state of evenness is open ended and potentially subject to negotiation 135. See supra Part I.C. 136. See Zipursky, Civil Recourse, supra note 2, at 719-20 (claiming tort law is different than a situation where “one who fails to pay under a contract will incur prejudgment interest because payment is owed at the time the contract specifies for performance, not at the time a court reaches a judgment”). 137. See Hershovitz, Corrective Justice, supra note 9, at 109. 480 BUFFALO LAW REVIEW [Vol. 62 with the wrongdoer.138 That is, “[u]ntil a court declares that a defendant must pay damages, the defendant is not under a duty to do it.”139 This is supposed to explain why, in most tort cases, prejudgment interest is not awarded.140 During the entire prejudgment period before the verdict, the content of the wrongdoer’s duty to make things even is open-ended.141 And since the duty to reach evenness does not automatically include a duty to compensate the victim at the moment of wrongdoing, prejudgment interest is not awarded.142 After all, there might not have been a duty to make things right by compensating at all. There are two problems with this response. The first is that it ducks the main objection. The issue of prejudgment interest was simply one piece of evidence marshaled to bolster a more fundamental, structural objection by Goldberg and Zipursky: that tort law recognizes no duty at all that arises and attaches immediately the moment a tort occurs, even though all corrective justice theories in the Aristotelian tradition do recognize that the duty of repair attaches at the moment the victim suffers wrongful losses.143 And in response to this deeper objection, Hershovitz has no response at all. This is because, for Hershovitz, like the Aristotelians, wrongdoers have duties to reach evenness that arise and attach to them the moment they wrongfully injure another.144 138. Id. at 128. 139. Id. 140. Zipursky, Civil Recourse, supra note 2, at 719 (claiming that “one who fails to pay under a contract will incur prejudgment interest because payment is owed at the time the contract specifies for performance, not at the time a court reaches a judgment”). 141. See Hershovitz, Corrective Justice, supra note 9, at 128. 142. See id. at 127-28. 143. Zipursky, Civil Recourse, supra note 2, at 720; see also Goldberg, Wrongs Without Recourse, supra note 6, at 13 (noting that “the conversion of the moral duty of repair into a legal duty does not happen through the tort system unless and until the victim decides to press a claim against the defendant”). 144. See Hershovitz, Corrective Justice, supra note 9, at 128 (“The duty that arises at the moment of wrongdoing is imperfect, or open-ended; a wrongdoer must take corrective action sufficient to support a declaration that the parties are even.”) (emphasis added). 2014] MAKING AMENDS 481 Hershovitz thus sides with the Aristotelians over the civil recourse theorists when it comes to the question of when, morally speaking, a duty arises from the point of view of corrective justice.145 Hershovitz departs from the Aristotelians, however, in describing the duty as “imperfect” when the wrongdoing occurs.146 The precise content of the wrongdoer’s duty to the victim arises at the outcome of a process; it is not determined automatically at the moment when the wrongful conduct took place.147 Thus, the actual conduct required of the wrongdoer to make things even with the victim does not arise until the end of this negotiation process. But notice that this still is not wholly responsive. The response fails to explain the fact that those who impose wrongdoings on victims are still not under an affirmative legal obligation to do anything about those wrongs unless the victims prevail in lawsuits (or at least settle). Hershovitz still claims, along with the traditional corrective justice theorists, that a duty to get even still arises the moment the wrongdoing occurs. And this is precisely what Goldberg and Zipursky deny.148 Corrective justice theorists, Hershovitz included, have a better response at their disposal, which is to reject Goldberg and Zipursky’s assumption that corrective justice implies that there ought to be a legal duty whenever the moral duty of repair arises.149 This is not a very charitable assumption. Once again, a better response—though by no means decisive—will simply point out that there are other considerations in play whenever we are considering whether to legally protect a moral right in general. We have moral duties to keep promises and not to lie, other things being equal, yet this does not mean there should be affirmative legal obligations to refrain from lying and keep promises in 145. Id. 146. See id. 147. See id. at 109, 128. 148. See supra Part I.C. 149. I am indebted to Scott Hershovitz for allowing me to see this point more clearly. 482 BUFFALO LAW REVIEW [Vol. 62 the same way we have affirmative legal obligations to pay taxes. A more robust response along these lines will be offered later on.150 So the first claim was that Hershovitz does not seem to provide a direct response to the no-legal-duty objection. A second problem with Hershovitz’s discussion of the objection contests his claims regarding prejudgment interest. In a nutshell, despite Hershovitz’s claim to the contrary, his theory cannot automatically rule out awards of prejudgment interest in tort cases.151 That is, the getting-even conception cannot rule out the possibility that, at least in some tort cases, prejudgment interest should be required in order to make the parties even. We can see this in two ways. First, Hershovitz seems committed to holding that civil actions arising out of contract disputes would seemingly count as torts.152 But once Hershovitz concedes that civil actions arising from contract disputes are torts, he can no longer maintain that all actions arising in tort rule out the award of prejudgment interest. After all, contract law explicitly allows for the award of prejudgment interest.153 Second, setting aside contract claims, nothing in Hershovitz’s theory explains why 150. See infra Part V.C. 151. Hershovitz, Corrective Justice, supra note 9, at 128. 152. At least there is no indication that he disagrees with the civil recourse theorists on this point. See John C.P. Goldberg & Benjamin C. Zipursky, Civil Recourse Revisited, 39 FLA. ST. U. L. REV. 341, 349 (2011) [hereinafter Goldberg & Zipursky, Civil Recourse Revisited] (calling civil recourse “a broad concept” capable of encompassing causes of action arising from breach of contract, as well as “other domains of private law”). 153. See, e.g., CAL. CIV. CODE § 3287(b) (West 2014) (“Every person who is entitled under any judgment to receive damages based upon a cause of action in contract where the claim was unliquidated, may also recover interest thereon from a date prior to the entry of judgment as the court may, in its discretion, fix, but in no event earlier than the date the action was filed.”). Tellingly, California starts the prejudgment-interest clock by reference to the date when the victim makes the wrongdoer aware of the precise amount of compensatory relief sought—not necessarily when the lawsuit was filed. See Levy-Zetner Co. v. S. Pac. Transp. Co., 142 Cal. Rptr. 1, 25 (Cal. Ct. App. 1977) (explaining that “prejudgment interest runs from the date when the damages are of a nature to be certain or capable of being made certain by calculation and when the exact sum due to the plaintiff is made known to the defendant”). 2014] MAKING AMENDS 483 prejudgment interest should not be routinely awarded in cases involving non-contractual torts. Suppose that, as one commentator claims, “prejudgment interest is now widely considered necessary to ensure full compensation to the plaintiff and to prevent unjust enrichment of the defendant.”154 If this is correct, then Hershovitz should allow that getting even might similarly require awarding prejudgment interest in even run-of-the-mill torts cases. Indeed, several jurisdictions appear to do just this: Massachusetts awards prejudgment interest in personal injury actions, with interest accruing from the moment the lawsuit is filed.155 California courts explicitly hold that, in determining whether to award prejudgment interest, “the key distinguishing factor was not . . . whether the cause of action arose in tort or contract, but rather whether the damages were readily ascertainable.”156 The point is not merely that some jurisdictions depart from Goldberg and Zipursky’s view that prejudgment interest is not ordinarily awarded to successful torts plaintiffs (though, as we have seen, a cursory search shows that there are more than simply a few dissenting jurisdictions).157 Rather, the point is that Hershovitz’s conception should not rule out, a priori, the possibility that getting even could reasonably require prejudgment interest in such cases, given how open-ended the getting-even approach presents itself. 154. Michael S. Knoll, A Primer on Prejudgment Interest, 75 TEX. L. REV. 293, 298 (1996). 155. MASS. GEN. LAWS ch. 231, § 6B (2013) (“In any action in which a verdict is rendered or a finding made or an order for judgment made for pecuniary damages for personal injuries to the plaintiff or for consequential damages, or for damage to property, there shall be added by the clerk of court to the amount of damages interest thereon at the rate of twelve per cent per annum from the date of commencement of the action even though such interest brings the amount of the verdict or finding beyond the maximum liability imposed by law.”), available at http://www.malegislature.gov/Laws/GeneralLaws/PartIII/TitleII/ Chapter231/Section6B. 156. Wisper Corp. v. Cal. Commerce Bank, 57 Cal. Rptr. 2d 141, 146 (Cal. Dist. Ct. App. 1996) (emphasis added) (quoting Levy-Zentner Co., 142 Cal. Rptr. at 23); see also CAL. CIV. CODE § 3288 (West 2014) (“In an action for the breach of an obligation not arising from contract, and in every case of oppression, fraud, or malice, interest may be given, in the discretion of the jury.”) (emphasis added). 157. See, e.g., CAL. CIV. CODE § 3288; MASS. GEN. LAWS ch. 231, § 6B. 484 BUFFALO LAW REVIEW [Vol. 62 Once again, we should take a step back and consider some general lessons. First, any theory positing that wrongdoers have duties of redress—which arise and attach to them the moment they wrong others—must contend with the no-legal-duty objection. They will have to explain why tortfeasors face no legal obligation towards their victims until they are held liable (or settle before the imposition of liability). Merely increasing the level of generality at which the duty is described—e.g., as a duty to rectify instead of a duty to repair—fails to provide the needed explanation, since increasing the level of generality at which the duty is described does not address the limitations on that duty—e.g., when that duty attaches and under what conditions. Nor will increasing the level of generality explain why prejudgment interest is not ordinarily awarded in torts. Once we allow that there will be a greater range of remedies by replacing a duty of repair with, say a duty to reach evenness, this immediately opens up the door for awards of prejudgment interest in tort. But, with respect to the last point, this is potentially a virtue: at least some prominent jurisdictions award prejudgment interest in cases arising from non-contractual torts.158 This seems like a heartening prediction of any corrective justice theory that seeks to re- characterize the duty of repair in more general terms. And these jurisdictions also arguably provide countervailing evidence that undermines support for the no-legal-duty objection. Let us assume for the sake of argument that Hershovitz has the resources to overcome the tripartite critique. In the remaining subsections, we will see independent reasons why we should not accept his theory. D. Can Tort Law Help Us Get Even? One worry with Hershovitz’s view is that no amount of relief offered by the tort system could render the victim “even” with the wrongdoer.159 Would we tell a parent who successfully sues a drunk driver for killing her children that, after winning a verdict, she is even? Or would it make sense 158. See, e.g., CAL. CIV. CODE § 3288; MASS. GEN. LAWS ch. 231, § 6B. 159. Hershovitz, Corrective Justice, supra note 9, at 111. 2014] MAKING AMENDS 485 to say to the rape victim who successfully sued her rapist that she is even? Maybe in the eyes of the law. But suppose these victims denied that they were even. Would they be wrong? Hardly: denying that they are “even” would make just as much moral sense as if they were to deny that they’d been made “whole.” Indeed, Hershovitz spends considerable time arguing, on similar grounds, that the Aristotelian make- whole metaphor makes no sense—and it is not obvious that Hershovitz’s get-even metaphor does any better.160 The upshot is that evenness might be just as misleading as the Aristotelian conception and for similar reasons. Often evenness is impossible; and to the extent that the law holds out the hope of evenness, it promises the impossible. In reply, Hershovitz might point out that his view at least holds a comparative advantage over the Aristotelian picture. The Aristotelian conception promises the impossible insofar as it is never possible, strictly speaking, to fully undo the losses attributable to wrongdoings.161 There will always be lost time, for instance, that cannot be given back to the victim. By contrast, the Nordic cultures discussed earlier believed that it was possible for victims to get even with wrongdoers.162 But it is not at all clear why these other cultures’ beliefs about the possibility of “getting even” are any more credible than the judgments of our current legal culture, which holds that it is possible to satisfy one’s duty of repair.163 In short, the claimed comparative advantage is dubious. 160. See id. at 110-17 (arguing against the Aristotelian conception of corrective justice on the grounds that it is impossible to truly ever make a victim “whole” again or to return the victim to the position he or she would have occupied but for the injury). 161. Id.; see also supra Part II.B. 162. See supra Part II.B. 163. See Hershovitz, Corrective Justice, supra note 9, at 116 (claiming that “discussions of corrective justice are chock full of qualifiers”); see, e.g., STATE OF CONNECTICUT, CIVIL JURY INSTRUCTION 3.4-1 (modified Apr. 5, 2012), available at http://www.jud.ct.gov/ji/Civil/part3/3.4-1.htm (“You must attempt to put the plaintiff in the same position, as far as money can do it, that (he/she) would have been in had the defendant not been negligent.”) (emphasis added). 486 BUFFALO LAW REVIEW [Vol. 62 E. Does Getting Even Involve Justice (Fit for Tort Law)? 1. An Initial Objection. Another concern with Hershovitz’s conception of corrective justice is that it ends up licensing potentially horrific acts of vengeance—acts that seem morally troubling to say the least. To see how, recall that, according to Hershovitz, one unilateral way a victim can get even with the wrongdoer is by exacting revenge on a wrongdoer—perhaps even violent revenge.164 Another exemplar of getting even includes Nordic threats of dismemberment and the lex talionis—the “eye for an eye”— retaliatory conception of justice.165 Hershovitz sees these not as cases in which corrective justice has failed, but rather he sees them as successful instances of corrective justice in action. The fact that the getting-even conception licenses both revenge and the lex talionis may seem troubling. The lex talionis, in particular, is not a morally attractive account of justice. Consider one quick argument against the view. An eye-for-an-eye conception of justice implies that it is just to torture torturers or rape rapists. But doing these things seems deeply unjust. Assuming that no plausible conception of justice should legitimize such morally horrifying acts, we should not endorse the lex talionis. The fact that Hershovitz’s getting-even theory legitimizes revenge and the lex talionis, at least in certain circumstances, should make us worry.166 164. Hershovitz, Corrective Justice, supra note 9, at 122 (noting that getting even “in the classic sense” means that someone “can seek revenge”). 165. Id. at 123. Hershovitz states: The Icelanders in the story did not regard taking an eye for an eye as the recipe for corrective justice, akin to the Aristotelian proposal that wrongful transactions should be reversed. For them, an eye for an eye was a failsafe. It provided a way of doing justice unilaterally, which was important in a world without tort law. But an eye for an eye was also what economists call a penalty default rule, encouraging wrongdoers to take their victim’s claims seriously. Id. 166. Although Jeremy Waldron does not endorse this argument against the lex talionis, he states it succinctly in his Lex Talionis, 34 ARIZ. L. REV. 25 (1992). He said: 2014] MAKING AMENDS 487 But this argument is surely too quick. Hershovitz would reject the idea that the getting-even conception, under all circumstances, straightforwardly endorses lex talionis or revenge as appropriate modes of justice. Recall that there were constraints on whether two parties could be rendered even with each other.167 Some of these restrictions were moral constraints.168 And whether it makes sense, morally speaking, to unilaterally seek revenge or seek the protection of the lex talionis depends on facts about the society in which a victim seeks justice. In a lawless society, perhaps unilateral acts of revenge are legitimate options for achieving justice; in primitive societies where the lex talionis is the only option, maybe private, unilateral revenge will no longer be available—instead, the right thing to do to get even is to find an oddman.169 And maybe in modern societies, getting even cannot be accomplished through either the lex talionis or unilateral acts of revenge, given the other options available— like the tort system. So even though Hershovitz uses revenge and the lex talionis to illustrate the getting-even conception, he has room to claim that his conception need not license, say, raping rapists or torturing torturers—or any other morally horrifying act of retaliation. His view is flexible enough to allow this response. Though the principle retains its attraction for defenders of capital punishment (“a life for a life”), people think they can discredit it almost immediately by asking “What penalty is to be imposed on the rapist, according to this principle?” Amidst the general hilarity that follows, the speaker is able to put LT [lex talionis] quietly to one side and move on, as he thinks, to some more plausible version of retributivism. Id. at 25. Jeffrey Reiman offers a version of this argument, allowing that the lex talionis indeed states a form of justice, but nevertheless insisting that some of its implications are deeply immoral. See Jeffrey H. Reiman, Justice, Civilization, and the Death Penalty: Answering van den Haag, 14 PHIL. & PUB. AFF. 115, 134-35 (1985) (arguing that, even though the lex talionis “proves the justice of beating assaulters, raping rapists, and torturing torturers . . . it would not be right for us to beat assaulters, rape rapists, or torture torturers, even though it were their just deserts—and even if this were the only way to make them suffer as much as they had made their victims suffer”). 167. See supra Part II.A. 168. See supra Part II.A. 169. See supra Part II.B. 488 BUFFALO LAW REVIEW [Vol. 62 But his view is not endlessly flexible—otherwise it would be vacuous. He is still committed to the view that revenge, the lex talionis, and tort suits have the same point or goal.170 He states this explicitly, and it is hard to overstate the importance of this claim for Hershovitz’s view. Indeed, the claim that the point that revenge and tort law have the same aim—to allow victims to get even with wrongdoers—is what makes Hershovitz’s claim so novel and provocative. But as we will see in the next subsection, it is precisely this view that makes Hershovitz’s theory so problematic as an explanatory account of tort law. 2. Explanatory Difficulties for Getting Even. The idea that the lex talionis and tort law have the same purpose—to allow a victim of wrongdoing to get even with the wrongdoer—undermines Hershovitz’s theory as an explanatory theory of tort law, insofar as it distorts rather than illuminates key features of tort law. To see why, notice that the lex talionis and revenge are typically regarded as practices that aim to punish wrongdoers.171 Jeremy Waldron, for example, interprets the lex talionis as a view about punishment.172 And in a widely anthologized work on the death penalty, Jeffrey Reiman goes to great length to clarify and understand the lex talionis as a theory of retributive, not corrective, justice.173 Suffice it to say that the lex talionis is widely understood as, at best, a theory of justice fit for punishment—i.e., a theory of retributive justice. Importantly, if we accept this natural understanding of the lex talionis as a practice that aims to punish, and acts of revenge as acts of punishment, then this raises problems for the getting-even theory as an explanatory theory of tort law in at least two important ways. 170. Hershovitz, What Does Tort Law Do?, supra note 94, at 117 (“A tort suit is not an act of revenge. But it aims to do the same thing that people taking revenge aim to do. That is, a tort suit aims to render wrongdoer and victim even in respect of the wrong.”). 171. See, e.g., Waldron, supra note 166, at 25. 172. See id. (calling the lex talionis an “approach to punishment”). 173. See Reiman, supra note 166, at 119 (“The lex talionis is a version of retributivism.”). 2014] MAKING AMENDS 489 The first way that the getting-even conception distorts tort law is by rendering mysterious the way tort law awards damages. More precisely, if the goal of tort law, revenge, and the lex talionis are all the same, and if that aim is punishment of wrongdoers, then tort law’s default remedy should be punitive rather than compensatory damages.174 True, tort law does permit punitive damages. But punitive damages are not the default mode of redress in tort law.175 Punitive damages are viewed as extraordinary and rare remedies to be afforded in extreme cases only—they are certainly not afforded “as of right.”176 But Hershovitz’s account renders this fact strange. At the very least, his account suggests that punitive relief should not be viewed as so “extraordinary” if tort law aims to afford victims the opportunity to get even with wrongdoers. Secondly, and relatedly, if the goal of tort law is the same as revenge or the lex talionis, then tort law seems somewhat redundant. After all, we already have a system of law devoted to meting out retributive justice—namely, criminal law. True, there might be different ways of securing retributive justice based on the perceived seriousness of the wrongdoing and the entity charged with enforcement. As to the latter, prosecutors are charged with pursuing criminal justice in the name of the public, whereas the victim pursues justice in tort law on her own behalf. Still, understanding corrective justice as simply another way of pursuing retributive justice blurs the distinction between tort law and criminal law.177 And if 174. I thank Greg Keating for pressing me to develop this point further. 175. See supra Part I.A. 176. See Gardner, Torts and Other Wrongs, supra note 8, at 53-55. 177. See, e.g., WEINRIB, CORRECTIVE JUSTICE, supra note 1, at 172-73. Weinrib states: Corrective justice rectifies injustices that operate on the parties in a transactionally specific way . . . linking two specific parties through the injury . . . . Punishment is different. It is state action that inflicts an adverse consequence on the wrongdoer without restoring the right of a wronged party. When the state punishes, it acts not to rectify a wrong that is transactionally specific to the plaintiff and the defendant, but to vindicate its own standing as the public guarantor of rightful order. Id. 490 BUFFALO LAW REVIEW [Vol. 62 one of the chief aims of theorizing about tort law is to illuminate how tort law is a distinctive branch of law, then the fact that Hershovitz’s approach blurs the boundaries counts as a demerit to his view as an explanatory theory. Tort law would be, arguably, criminal law by other means. 178 Of course, neither of these explanatory problems is very problematic if Hershovitz presents a revisionist, rather than explanatory, account of tort law. But if he seeks to explain many of tort law’s main features in terms of getting even, construing tort law as sharing the same aim as revenge or the lex talionis is more confusing than illuminating. And this is because, it seems, these practices do not seem to have the same aim. Despite these shortcomings, the getting-even theory paves the way to a broader conception of corrective justice that will be called the “making amends” approach. Let us turn to this conception below. IV. THE MAKING AMENDS APPROACH The previous section posed problems for Hershovitz’s getting-even account of corrective justice. But his account nevertheless yields important insights. He correctly shows, for example, that to accommodate the remedies objection, we will need to describe the relevant duty at a higher level of abstraction. After all, some remedies are not in the business of repair; some are punitive and others are nominal, and the duty must be described broadly enough to capture these facts. Instead of a duty to repair, the new approach describes the duty relevant to corrective justice as the duty to make amends. This duty is part of a principle that holds, roughly, that corrective justice occurs when and only when a person responsible for wrongs that befall others (the “victims”) 178. The same objection has been made against civil recourse theory—i.e., the positive theory of tort law put forward by Goldberg and Zipursky, which has received little attention in this paper. See, e.g., Jane Stapleton, Evaluating Goldberg and Zipursky’s Civil Recourse Theory, 75 FORDHAM L. REV. 1529, 1559- 60 (2006). See generally Gardner, Torts and Other Wrongs, supra note 8 (arguing at length that civil recourse theory fails to illuminate how tort law is distinguishable from several other areas of private law). 2014] MAKING AMENDS 491 makes amends to those victims. By focusing on making amends, we can make substantial progress in formulating an account of corrective justice that helps explain the structure of tort law and rebuts the other criticisms of the classical views encountered in Part I.179 A. Making Amends: A Preliminary Sketch 1. Wrongs and Responsibility. Someone who is responsible for a wrong that befalls another, a victim, has a duty to make amends to that victim. The first thing to notice about this claim is that it incorporates some notion of responsibility: X’s making amends to Y makes sense only if X is responsible for a wrong that has occured to Y. Suppose that Bill is struck by lightning. Now suppose that, upon hearing the news, Albert tries to make amends for Bill’s lightning- induced harms. This makes no sense unless Albert is somehow responsible for the harms. In general, a person can make amends to another only if that person is in some sense responsible for an unwelcome event that happens to another person.180 This raises two further questions. First, what kind of unwelcome events count as a wrong, and second, what kind of action makes a person responsible for the wrongs that befall another person? As to the first question, recall that corrective justice concerns the norms that govern in the aftermath of wrongdoing, but it does not supply, nor should 179. At this point, it is worth quickly mentioning the methodology used in investigating the idea of making amends. The author follows more or less a standard philosophical approach in coming up with his account of the concept of making amends. The approach involves proposing explicit principles or criteria that govern our understanding of a given concept, and which adequately reflect or capture certain “truisms” or platitudes about the concept that should be relatively uncontroversial. For more detailed discussion of an approach in the same spirit, see SCOTT J. SHAPIRO, LEGALITY 13-22 (2011). 180. Notice, by the way, that this observation immediately rules out the version of the Tom and Jerry case in which Tom and Jerry declare themselves even with each other with respect to a pre-existing debt between them. Hershovitz, Corrective Justice, supra note 9, at 118-19. This did not seem like a case involving corrective justice in any event; the fact that the making amends view fails to account for it should be regarded as a benefit of the view. Making amends applies only in cases where someone has wronged another. 492 BUFFALO LAW REVIEW [Vol. 62 it supply, a complete theory of what counts as morally wrong, or what ought to be legally recognized as a wrong. So the fact that this article does not aim to provide a full-fledged theory of wrongs should come as no surprise. That said, we can still say a little about the formal nature of wrongs presupposed by the making amends theory. There is, for example, one important limitation on how broadly we should understand the wrongs relevant to making amends, which is that making amends involves, as Goldberg and Zipursky point out, “relational wrongs.”181 The idea is that some wrongs injure a particular victim, as opposed to being wrongs as such.182 The process of making amends presupposes, as pointed out in the previous paragraph, that the wrong connects a victim with someone who is in some sense responsible for the wrong. Driving drunk is a wrongdoing even if there is no victim. Involuntary manslaughter is a relational wrongdoing given that describing the act necessarily involves mentioning a victim. Making amends necessarily involves the responsible party making amends to someone.183 Accordingly, the wrongs relevant in the making amends process are always relational wrongs, relating a responsible party to an identifiable victim.184 Turning to the second question—concerning the relation of being responsible for that corrective justice presupposes— this arguably is something that a full-fledged theory of corrective justice must grapple with. Indeed, many corrective 181. Goldberg & Zipursky, Torts as Wrongs, supra note 4, at 945-53; see also Zipursky, Rights, Wrongs, supra note 26, at 60-64. 182. Goldberg & Zipursky, Torts as Wrongs, supra note 4, at 945-46. 183. See id. 184. See id. Linda Radzik endorses the conception of making amends according to which it seems that wrongdoers have a duty to make amends for non-relational wrongs. See LINDA RADZIK, MAKING AMENDS 135-39 (2009). Radzik deserves recognition as one of the few, if not only, contemporary philosophers who have focused squarely on the topic of making amends. But I do not accept all of her analysis. 2014] MAKING AMENDS 493 justice theorists have discussed the topic at length.185 And surely the notion (or notions) of responsibility embedded in our practices of making amends is likewise an important topic. One worry is that the concept of responsibility papers over the real criteria determining whether one may fairly be held accountable. Is one “responsible” for all the harms that foreseeably injure others as a result of one’s misconduct? If so, this might be problematic if the principle aims to explain tort law, since very often a person is not liable despite being “responsible” for the harmful consequences of their action in this sense.186 Moreover, regardless of how the concept of responsibility is specified, the moral principle proposed here poses problems because we should not hold those who truly lack capacity morally responsible for their actions. Expecting a truly insane person to have the capacity to make amends would seem unreasonable. Yet tort law does not recognize an insanity defense.187 This article will not take up these important issues. For present purposes, we will not attempt to articulate or rely on any particular theory of responsibility in order to make sense of corrective justice. The limited goal of this article will be to articulate and defend a more expansive conception of corrective justice that is not vulnerable to the objections of 185. See, e.g., COLEMAN, RISKS AND WRONGS, supra note 1, at 197-385; Stephen R. Perry, Responsibility for Outcomes, Risk, and the Law of Torts, in PHILOSOPHY AND THE LAW OF TORTS 131-82 (Gerald J. Postema ed., 2001). 186. See supra Part I.B. (discussing the dilemma that played a crucial role in the substantive standing objection). 187. James Goudkamp, Insanity as a Tort Defense, 31 OXFORD J. L. STUD. 727, 727-28 (2011) (recognizing that “[a]ll of the major common law jurisdictions withhold insanity as an answer to liability in tort” but proceeding to argue against this practice). Courts do, however, make accommodations for defendants lacking certain capacities. Blind people, for example, are not held to the same standard of care applicable to defendants with fully functioning eyesight. See ARTHUR BEST & DAVID W. BARNES, BASIC TORT LAW: CASES, STATUTES, AND PROBLEMS 112 (2d ed. 2007). Best and Barnes state: Exceptions to the reasonable person standard developed when the individual whose conduct was alleged to have been negligent suffered from some physical impairment, such as blindness, deafness, or lameness. Courts also found it necessary, as a practical matter, to depart considerably from the objective standard when dealing with children’s behavior. Id. 494 BUFFALO LAW REVIEW [Vol. 62 Goldberg and Zipursky. And this can be done, as we will see, without signing on to a fully fleshed out theory of responsibility. 2. Conciliatory Gestures. The claim so far is that those who are responsible for wrongs that befall victims have a duty to make amends to those victims. Now let us consider the nature of making amends itself. How do those responsible for wrongs make amends? Making amends requires, we claim, a certain kind of conciliatory gesture by the responsible party toward the victim. To unpack this claim, let us begin with the gesture requirement. That is, in order to make amends, some sort of act performed for the victim is required. To see why, consider Albert and Bill. Suppose that Albert is responsible for wrongful conduct that injures Bill, and that Albert claims to have made amends to Bill. Now suppose Charlie asks Albert how Albert made amends. In reply, Albert says he felt guilty about the wrongdoing and recognized that he should not have done it, but confesses that he did nothing for Bill. Albert, for instance, never apologized to Bill. In fact, Albert and Bill have had no contact with each other since the wrongdoing occurred; Bill does not even know that Albert felt guilty about the wrongdoing. In these circumstances, Albert’s claim to have made amends to Bill is false. Albert is making a mistake about the nature of making amends. Feeling guilty about a wrongdoing and wishing it had never happened does not suffice to make amends.188 These are appropriate attitudes to have in response to a wrongdoing. But the proper vehicle for expressing these attitudes is some sort of gesture performed for the victim. As for the kind of gesture: making amends, like Hershovitz’s notion of getting even, has a built-in flexibility. There is no detailed recipe on how to make amends in any given case. But this flexibility has limits. Not any gesture will do. The gesture should be, as just mentioned, conciliatory (rather than, say, insulting). The very word used to describe 188. Radzik points out that certain theological conceptions of making amends come close to holding that sincere repentance is sufficient to atone for one’s sins. Linda Radzik, Making Amends, 41 AM. PHIL. Q. 141, 142-44 (2004). Radzik and I agree that, for worldly purposes, private atonement is not enough. See id. 2014] MAKING AMENDS 495 the gesture, “conciliatory,” suggests that the wrongdoer has a duty to reconcile or at least seek a reconciliation of sorts with the victim. There is some truth here. But we should be careful not to over-emphasize the notion of reconciliation. Conciliatory gestures need not aim to reconcile victim and wrongdoer, and a conciliatory gesture can succeed in making amends without achieving any recognizable reconciliation. Very often victim and wrongdoer have no prior relationship to restore and no relationship other than one that exists between victim and wrongdoer. Hence, what makes an act conciliatory in the relevant sense is not that it aims to restore a prior relationship. Instead, a gesture counts as conciliatory in the intended sense if it suggests or expresses regret on the part of the wrongdoer to the victim that the wrongdoing had occurred. This last observation—that the act must be understood to signal regret by the wrongdoer that the injury occurred— is a potential source of confusion. It may be taken to suggest that a wrongdoer must subjectively feel regret for the wrongdoing in order for the gesture to count as conciliatory. And ideally, a wrongdoer should feel some regret. But the idea of a conciliatory gesture contemplated here is wholly objective. To illustrate, suppose that Hatfield wrongs McCoy in some way and thereby incurs a duty to make amends to McCoy. Hatfield knows that the right thing to do is to make amends by, say, apologizing and offering up some fresh vegetables from his garden. Now suppose that, because Hatfield and McCoy are mortal enemies, Hatfield cannot bring himself to regret the wrongdoing. Hatfield’s dour attitude does not prevent Hatfield from fulfilling his obligation to make amends. All that matters is that the gesture be reasonably recognizable as one that signals regret, even if, subjectively, wrongdoers like Hatfield cannot bring themselves to actually experience that regret. Hatfield can fulfill his moral obligation to make amends even if he does so grudgingly, just like he can satisfy his legal obligation to pay his taxes even if he does so grudgingly. So how do we know whether the gesture recognizably signals regret, given that wrongdoers need not actually feel any regret? The act should be a reasonable and adequate response by those responsible for the wrongful losses of others. We say reasonable and adequate because conciliatory �|�� ���������� � �� � � ���� ��� \������� ~�\ � �� ���������� ��� � ���� ���������� �� ����� ����� *� �����\� �� ��~��� ������ � ���������� �������� �� � ����\����\ 189 �������� �����\��� ��� ��~��� ������ ~��� ����`���� ��� ����� �� �����\� �� ��� ����\ \���� � � �������� �� � � ��~ ���� 190 ��~���~�� ~����\ �~���� ���� ������� ��������� ~�����\ ������ ���`�� �� �����\ ��� � � ��~�\� ���� �� ���^� ����\��� `����`� �������� ����������� ��� ��~��� ��� ��� �� ������~��\ � �� ���� �� \������ *�� ��~���~��� ��� �����`������ ������ ����\����\�� ~����\ �~���� �������� ���~�����\ �� ����� ~��� 191 � � ~��� �������� ����� �� � �� ~����\ �~���� ~�\ � ������� � ����� ���\� �� ����������� ���������� � �` ��� ��������� ��� �~��\ �� �� � ��\�� � � ������ �� � � ����\ 192 �� �� ����� ��� ��\ ~����\ �~���� ���~� �� ��`���� ~��� ��� ����� � �� ������������ ��~� ��� ����� � � � ���� #~����\ �~����X �� �� ������~��� ��� ~����\ {}| �`��� ���� ����£ ��`�\��£�� � �� \������� ������ `�~������������` �� �����\����~�\ � �� � ���� �� ~����\ �~���� ��� `�~�� `���� �� ����������\ ~����\ �~���� ��� `����`���� �����`� �������� ����£� ����7�_������ ���������� }� �� |� ��� � ����� �����`���� `�������� � �� ����������� �� ���\� ���� ��@ �#�*����� ���~ ���� ����������� ��� �����\� ��� �� ����� �� `������ �� ��� ���`��`� �� ~����\ �~���� �� ~������� `�~��������� �� � � �������� �� ���� �� ������� ������� �� � �� �� ����������� ��� �����\� ��� ���� ~������ �� `����`���� �����`� X� {|� �����\ �~���� �� � ���~����� � �~� �� ����`�������`����� ���`��`��� � �` ���� ~����\ �~���� �� ����������� ~��� � �� ��~��� ��������\ `�~��������� ����� �@�@� �*¯!@�!<� ������� ������� ��� ����� 4$��� ��� ������7� ��������� �£����� ��\����������`� ������� ��\� ����� ������� ���� {�� ��{�� {|{ � ������@ �#$�~�~��� � �� ��� `��~�� ��` �� ����� ������\� ������ ~�\ � ���� �� \� �� `���� ��� ���� � ����� ~��� � ��^� ���� �� ~����\ �~���� �� ���� X� {|� � � >����� !�\��� ��`������� �>!�� ����`����� � � � ���� #~����\ �~����X ��� �� ��� #������������� ������������ ������������ `�~���������� �������`���� X >­�>$� !<+@��� �����><*$¨� ����� ���� �}� �� �}� <���`� ����� � �� � � ����� ������ �� ����`����� ��� #~����\ �~����X ��`���� #�����������X� #`�~���������X� ��� #���������� X �@ *�� ��\ �� � ���� �� `������ ��� �� `������ ��`������� ����������� ��� ���� ���~����� ���� � �� ����� �``����� ���`�������� �� `�~���� ~���� � ���~���� � � >!� ���� ������� ��~� ������`� � �� ~����\ �~���� ����������� �������� ~��� � �� ��~��� � ������ ���~ �� ����������� 2014] MAKING AMENDS 497 reparations.193 This suggests that the duty of repair just is the duty to make amends. But we should resist conflating these two ideas. For one thing, the phrase “making amends” often amounts to little more than loose talk when it occurs in the literature on corrective justice.194 More importantly, our ordinary understanding of making amends seems a much broader moral notion and includes a greater variety of responses to wrongdoing than the narrower notion of a duty to repair. But this does not mean that we ought to jettison the Aristotelian’s duty of repair. The notion of conciliatory gestures contemplated here provides a useful lens through which to view traditional corrective justice theory. Traditional corrective justice theories go wrong insofar as they claim that the duty of repair is the one and only duty associated with corrective justice.195 But the relevant duty is broader than a reparative one. It is a duty to make amends, one essential ingredient of which is a conciliatory gesture. In turn, a gesture counts as conciliatory only if it is recognizable as suggesting some regret that the wrong took place. As it turns out, however, there is often no better way of expressing or suggesting regret than by trying to actually undo the damage done as a result of wrongdoing—at least to a reasonable extent. If A accidentally pushes B to the ground, there is no better way to express regret for this occurrence than for A to try to help B up off the ground. The mistake is to think that this is the only way to signal regret. 193. See, e.g., ARTHUR RIPSTEIN, EQUALITY, RESPONSIBILITY, AND THE LAW 104 (1999) (describing one view as holding that “if I injure someone through my carelessness, and should have foreseen that injury, I can be forced to make appropriate amends, as judged by a court of law”) (emphasis added); Jules L. Coleman, Epilogue to Risks and Wrongs: Second Edition 30 (Yale Law Sch. Pub. Law Working Paper No. 218), available at http://papers.ssrn.com/ sol3/papers.cfm?abstract_id=1679554 (referring to a duty to “make amends or repair”) (emphasis added). Civil recourse theorists do the same. See Goldberg, Wrongs Without Recourse, supra note 6, at 13 (“In other words, if the defendant is going to be made to heed his duty of repair, it will only be by virtue of the law’s having empowered the victim to demand of the defendant that he make amends for the wrong done.”) (emphasis added). 194. See sources cited supra note 193. 195. See, e.g., Hershovitz, Corrective Justice, supra note 9, at 116-17 (describing Aristotelian theories). 498 BUFFALO LAW REVIEW [Vol. 62 We are very close to returning to tort law. But let us take stock by articulating some principles that capture the making-amends conception of corrective justice. Consider a first approximation, called the Making Amends (First Pass) or the First Pass principle: Making Amends (First Pass): Individuals who have relationally wronged others (victims) have a duty to make amends to those victims. This principle captures the observations made so far about making amends. Making amends to person X is something that one must do if one has wronged X. The notion of wronging others, however, must be viewed both expansively and restrictively. Understanding the notion of wronging others expansively means not only that X wrongs Y in run-of-the-mill cases in which X’s direct and wrongful conduct injures Y, but also in cases where we can reasonably impute to X responsibility for wrongs that befall Y. By emphasizing relational wrongs, the principle should be understood to restrict the class of wrongdoings; it is not enough for a wrongdoing to exist “in the air” so to speak. The wrong must be a wrong to another person. We might make these points more explicit by offering the following revised principle, sacrificing concision for precision: Making Amends (Second Pass): Individuals who have relationally wronged others (victims), or are otherwise responsible for the wrongs that befall victims, have a duty to make amends to those victims. The new, italicized clause makes explicit the point just mentioned. Either we read X wronging Y broadly to include X being responsible for (in some unspecified sense) Y’s injuries, or if that seems to stretch too far the notion of one person wronging another, then we offer instead the Second Pass principle. The wrongs in any event must still be relational wrongs that render some individuals victims. Given that making amends just is performing a reasonable and adequate conciliatory gesture under appropriate circumstances, we might make the principle even more explicit: Making Amends (Third Pass): Individuals who have wronged others (victims), or are otherwise responsible for the 2014] MAKING AMENDS 499 wrongs that befall victims, have a duty to perform a reasonable and adequate conciliatory gesture to those victims. The duty to make amends, on this explication, has been satisfied once the responsible party performs a reasonable and adequate conciliatory gesture to the victim or victims. This does not require, as already noted, any full reconciliation. Full “reconciliation” between victim and wrongdoer—whatever that may entail—is not a prerequisite for making amends. Nor does it, as already discussed, require any particular feelings of remorse or guilt on behalf of the wrongdoer. The gesture must, however, be reasonably recognizable as expressing regret that the wrong occurred. In turn, this requires that the gesture be adequate or proportional in relation to the gravity of the wrongdoing that has occurred. So far we have roughly sketched an account of the duty to make amends, the circumstances under which it is triggered, and what making amends actually involves (a reasonable and adequate conciliatory gesture to the victims of the wrongs by those who own the wrong). And we have tried to explicate these ideas. The resulting principle of making amends—the Third Pass—is less concise but somewhat more precise than the First Pass principle. But the Third Pass still omits an important dimension of making amends, which is that there are limits on the duty to make amends. 3. Limits on Making Amends. The duty to make amends is a duty. And like most moral duties there are limits that apply. The moral duty to make amends does not apply, in other words, in any and all circumstances. To see why, notice that it often makes very little moral sense for someone who is responsible for another’s wrong to try to make amends. Sometimes this is a consequence of what we observed in the preceding paragraph: making amends requires the responsible party to have some kind of interaction with the person wronged, in the form of a conciliatory gesture. And depending on the nature of the wrong, and the burdens on the victim, an attempt by the wrongdoer to make amends 500 BUFFALO LAW REVIEW [Vol. 62 might simply add insult to injury.196 As Linda Radzik correctly notes, encouraging a child molester to try to patch things up with his victims is ill-advised and morally dubious—especially if the molester’s idea of making amends includes private, face-to-face interactions.197 The victim may want no further interaction with his injurer whatsoever. And this observation is not a special or extreme case. Some people may simply wish to move on with their lives and do not wish to revisit a painful past. And the process of making amends can be quite taxing on the victim. To avoid adding insult to injury, or if not insult then unwanted burdens, ideally victims should have some control over the amends-making process regarding whether, when, and how the process of making amends should go forward. With this in mind, let us revise the making amends principle: Making Amends (Fourth Pass): Individuals who have wronged others (victims), or are otherwise responsible for wrongs that befall victims, have a duty to perform a reasonable and adequate conciliatory gesture to those victims, only if those victims want the wrongdoer (or responsible party) to make amends. The “only if” clause here tries to capture the idea that it’s the victim’s prerogative to determine whether amends- making proceeds. The injurer’s duty to make amends to the victim is contingent on the victim’s wanting the wrongdoer to make amends. 196. I am indebted to Linda Radzik’s work for emphasizing this point. See Radzik, supra note 188, at 146; see also RADZIK, supra note 184, at 84. Radzik does not appear to suggest, however, that the duty to make amends—or “atone” in her words—can be turned off by the interests of the victim. She instead suggests that, when making amends is unwelcome by the victim, the wrongdoer should redirect that energy towards reconciling with the community at large. See RADZIK, supra note 184, at 84 (claiming that, even in cases where the wrongdoer should not make personal gestures to the victim, “[t]here is still important work for atonement to do . . . to achieve reconciliation with the broader community”). This seems a good idea, morally, but does not amount to making amends with the victim, which is the conception of amends-making that I focus on, and which is the conception of the amends-making process relevant to tort law. 197. Id. 2014] MAKING AMENDS 501 4. Obstacles to Making Amends. We noticed that one’s duty to make amends to a victim is contingent on the victim’s desires and that this places the wrongdoer in a tricky position if he genuinely wishes to make amends. Although figuring out what the victim wants is often easy to work out, sometimes it is not. To figure out what the victim wants, maybe the wrongdoer should simply ask. Many times this makes sense. But even asking might be an unwelcome gesture. Recall that the victim of molestation may want no interaction with the wrongdoer, and even asking may contravene the victim’s wishes. But this is just one possible obstacle in the amends- making process, and perhaps a small one. The most obvious impediment is disagreement. The injurer might disagree with the victim as to whether he is responsible for a wrongdoing. And even if the injurer concedes wrongdoing, she might disagree with the victim over what conciliatory gesture will suffice. The victim might want something other than what the injurer is willing to provide. Other obstacles abound. There are epistemic problems. The victim might not know who the injurer is. This is often the case during hit-and-runs. Likewise, the injurer might not know who the victim is. A factory that leaks toxic chemicals may not find out for many years that a cluster of birth defects is attributable to that spill, and on the flipside, those who suffer from the defects may not have a clue what caused them. Or the injurer might simply not know he has committed a wrong, as is the case with an innocent trespass over land that the trespasser mistakenly thought was public. How do we overcome these and other obstacles and difficulties in the amends-making process? No more amendments to the moral amends-making principle will be offered. Perhaps further refinement is possible, but these problems are likely to persist despite them. These are practical and epistemic difficulties that are not resolvable by specification of a moral principle. They are problems that arise in relation to efforts to enforce or realize that moral principle by either the injurer or the victim. To illustrate, suppose we have a perfect specification of the moral principle of making amends. Once we are told that the principle has an “escape” clause—that our duties to make 502 BUFFALO LAW REVIEW [Vol. 62 amends are contingent on whether others want us to make amends—wrongdoers may shy away from making amends by convincing themselves that everyone will be better off if everyone simply avoids the amends-making process. “If the victims really thought about it,” wrongdoers tell themselves, “they would realize that dredging up old wrongs does nobody any good.” And this thought is not entirely without reason. Economists have a label for this idea—a “sunk cost”—and economically rational agents never take sunk costs into consideration.198 Sometimes the best response to a wrong might be for everyone to just move on. This shows that, even if we manage to arrive at a perfect formulation of the duty to make amends, practical problems of application of that principle still arise. The fact that the moral process of making amends faces so many obstacles may not be reassuring. But these difficulties are important for understanding the relationship that tort law bears to making amends. This sketch of the informal, moral process of making amends leaves many questions unanswered. Are sincere apologies by wrongdoers vital? Is making amends tantamount to a kind of atonement? Although the author is inclined to answer ‘no’ to these questions, no arguments on these points will be offered here. It should be noted that Linda Radzik is one of the few to face these issues head on in her work.199 My own analysis of making amends does not appear to conflict with the broad contours of her view. By the same token, we should not endorse everything Radzik claims about the morality of making amends either. Radzik presupposes far more than we should be prepared to accept. For one thing, her views are shaped by a particular expressive theory of the nature of moral wrongdoing, a theory that arguably overgeneralizes based on cases of intentional wrongdoings and that will therefore have trouble capturing cases of negligence that are the bread and butter of tort 198. But see R. Preston McAffee et al., Do Sunk Costs Matter?, 48 ECON. INQUIRY 323-36 (2010) (describing sunk costs as costs that cannot be recovered, describing the conventional economic perspective that holds that rational actors ignore sunk costs in making decisions, and arguing within an economic framework that reflecting on sunk costs may be rational under certain constraints). 199. See RADZIK, supra note 184; see also Radzik, supra note 188, at 146. 2014] MAKING AMENDS 503 lawsuits.200 Making amends, however, is broader than that and can also capture our morally appropriate responses to accidents.201 Also worrisome is Radzik’s claim that the goal of making amends is reconciliation between wrongdoers and their victims, which implies that making amends fails whenever it falls short of actual reconciliation.202 Although we agree that making amends involves a conciliatory gesture, we were also careful to distance this concept from the idea of reconciliation, especially given that strangers who may have no interest in reconciling can still make amends. A full defense of these views will not be offered here. But hopefully the preliminary sketch outlined above will play a useful role in helping to explain tort law, a project that follows below. B. Making Amends Through Tort Law The previous subsection analyzed the duty to make amends in terms of what triggers the duty, some of its limits, and certain practical obstacles to making amends. The present proposal in this subsection is that tort law aims to facilitate the amends-making process by mitigating these obstacles, while protecting victims’ morally important interests in controlling certain aspects of the amends-making process. The making-amends conception of tort law—as opposed to those provided by pragmatic conceptualists like Coleman and Zipursky203—is thus an instrumentalist explanation. As such, any mismatch between the structure of the moral principle of making amends and normative structure of tort law is not necessarily a detriment to the explanation on offer. The goal of the following explanation is to show how any lack of “fit” between moral principles of 200. RADZIK, supra note 184, at 76-80. 201. Section B below will take on the question of the nature of wrongs presupposed in tort law in greater detail. 202. See generally RADZIK, supra note 184. Wrongdoers, it seems, can successfully make amends without any reconciliation with the victim by performing a reasonable and adequate conciliatory gesture. But this claim will not be defended here. 203. COLEMAN, PRACTICE OF PRINCIPLE, supra note 1 at 25-63; Zipursky, Civil Recourse, supra note 2, at 703-09; Benjamin C. Zipursky, Pragmatic Conceptualism, 6 LEGAL THEORY 457, 457-85 (Jules L. Coleman et al. eds., 2000) [hereinafter Zipursky, Pragmatic Conceptualism]. 504 BUFFALO LAW REVIEW [Vol. 62 making amends and tort law’s basic normative structure can be explained in terms of the law’s attempt to overcome certain practical obstacles to the amends-making process. Some mismatch is to be expected. Perhaps the best way to see how tort law mitigates obstacles to achieving corrective justice— or making amends—is by considering the obstacles that prevent people from making amends and how the institution of tort law is designed to correct for those failures. 1. The Problem of Disagreement. As already noted, probably the most obvious obstacle to making amends is disagreement. Sometimes the wrongdoer denies being responsible for any wrong. Other times he denies committing a wrong with respect to the putative victim. And even if there is no disagreement about whether an injurious wrong has occurred, or whether the putative injurer has the right relationship of ownership toward the wrong, there still may be disagreement about what kind of conciliatory gesture is adequate. Tort law addresses all of these issues. It defines the conduct and injuries that qualify as actionable wrongs.204 True, the law does not make actionable all injurious wrongdoings. A threshold of importance must be met.205 The law does not generally concern itself, for example, with de minimis wrongs.206 Furthermore, because tort law aims to facilitate the amends-making process, and because making amends is something that a wrongdoer does with respect to the person he has wronged, tort law also has features that ensure that the putative victim is in fact a genuine victim of the wrongdoer’s wrongful conduct. This point is an important one that we will return to when we discuss the substantive standing requirement. As we will see, the point of these requirements is at least in part to ensure that the litigants 204. John Gardner, What is Tort Law For? Part 1. The Place of Corrective Justice, 30 LAW & PHIL. 1, 50 (2011) [hereinafter Gardner, What is Tort Law For?] (“Private law can (and may be needed to) make such obligations more determinate than they would be in their raw moral form, but it is not needed to bring them into existence in the first place.”). 205. See, e.g., Ringgold v. Black Entm’t Television, Inc., 126 F.3d 70, 74 (2d Cir. 1997). 206. Id. (“The legal maxim ‘de minimis non curat lex’ (sometimes rendered, ‘the law does not concern itself with trifles’) insulates from liability those who cause insignificant violations of the rights of others.”). 2014] MAKING AMENDS 505 stand in the proper amends-making relationship that tort law aims to facilitate. Why this is important will also be discussed in greater detail below. Another source of disagreement relates to conciliatory gestures. Although a reasonable and adequate conciliatory gesture is required, what counts as reasonable and adequate might be a source of disagreement. Tort law helps smooth over these difficulties as well by providing a set of off-the- rack remedies, the most important being the default rule of compensatory damages.207 Different jurisdictions differ on the details, but the civil recourse theorists are correct to point out that the kinds of remedies involve at least compensatory and punitive damages, as well as many forms of injunctive relief.208 In summary, tort law settles disagreements about what conduct constitutes a wrong in an important class of cases and disagreements about who qualifies as a genuine victim of those wrongdoings. Tort law accomplishes this by providing off-the-rack categories of relational wrongs to potential plaintiffs and a menu of potential remedies. 2. Victim Control. There is another important moral problem on the horizon, relating to victims’ morally important interests in exerting some control over the amends-making process—i.e., whether the wrongdoer should make amends and how that process takes place. Consider Carol and Will. Carol has wronged Will, but Will simply does not want to have any further contact with Carol. Will, in short, does not want Carol to make amends. On the amends- making view, this means Carol has no duty to make amends with Will. But suppose that Carol, for whatever reason, still wants to make some conciliatory gesture to Will. Maybe she feels guilty or believes (incorrectly) that she still has to make it up to Will in some way. In any event, by unilaterally deciding to approach Will, Carol is making an unwelcome gesture and depriving Will of some degree of control over the amends-making process. This is so regardless of whether Carol’s gesture ultimately succeeds in making things better, 207. See supra Part I (discussing views of civil recourse theorists Goldberg and Zipursky). 208. See discussion supra Part I. 506 BUFFALO LAW REVIEW [Vol. 62 all things considered. This situation reflects the problem of protecting the victim’s morally important interest in exerting some control over the amends-making process. In particular, the case of Carol and Will concerns primarily the victim’s morally important interest in controlling whether the amends-making process takes place at all. The case is also an example of bilateral struggle for control over the amends-making process between injurer and victim. Carol, in a sense, wrests Will’s right to determine whether or how the amends-making process takes place. But usurpations of control can occur in multilateral situations among the putative “victims” of wrongs. The threat in such cases is not that the responsible party will set the terms of engagement, thereby disregarding the victim’s interests, but rather that other self-proclaimed “victims” will behave in ways that effectively undermine those interests. To see how this can happen, consider a wrongdoing affecting three people: Allison, Bob, and Carrie. Allison and Bob are married. Allison discovers, however, that Bob is having an affair. Allison is distraught. Carrie cannot stand to see her friend Allison so upset, so she takes it upon herself to confront Bob and demands that Bob pack up his belongings and move out of Allison’s house. Carrie—as Allison’s friend— is in a sense a foreseeable “victim” of Bob’s infidelity since she must suffer along with Allison and help her get through this trying time. But there is a sense in which Bob can rightly claim that this is none of Carrie’s business, that she lacks moral standing to confront him, and that she also lacks the standing to make the demand that she is making. Bob did not cheat on her, after all. The wrongdoing is primarily a matter between Allison and Bob—not Carrie. We may not sympathize with Bob’s claims. But let us not forget that Allison also has reason to be perturbed by Carrie’s conduct, no matter how well intentioned it was. By taking it upon herself to demand that Bob move out, Carrie has overstepped her bounds and usurped Allison’s morally important interests in controlling how the amends-making process take place (if at all). By insisting that Bob move out, Carrie has usurped Allison’s judgment as to whether and how the amends-making process plays out with Bob. What if Allison wants to try to make the marriage work despite Bob’s 2014] MAKING AMENDS 507 affair? Carrie has usurped, in effect, Allison’s judgment on the matter. Now suppose that Allison agrees with Carrie that Bob should move out. Surely, Allison should be consulted on how to execute Bob’s moving out and the manner by which it proceeds. So Carrie has interfered with not only Allison’s morally important interests in controlling whether the amends-making process takes place; by demanding that Bob move out, she has also encroached on her important interests in controlling how this takes place, if at all. Tort law slices through these Gordian knots. Consider first the question of how to reinforce the victim’s morally important interests in control. Tort law protects these interests by allocating to victims the right to file lawsuits against wrongdoers. By allocating the right to demand wrongdoers to make amends, victims are given control over the initiation of the amends-making process, the timing of that initiation (within defined limits), the claims asserted against the wrongdoer, as well as defining the remedies sought against the wrongdoer. To be sure, this control is far from absolute; the remedies and claims brought in courts of law are much more limited than those available in our moral lives. Nor does tort law do anything to prevent the Carols of the world from unilaterally trying to make amends informally, against the wishes of the victim; the wrongdoers have liberty interests in interacting with others that the law may not wish to obstruct unnecessarily. But tort law does not afford wrongdoers an avenue for coming forward to make amends. Victims, not wrongdoers, have the primary power to initiate—and hence control—the amends-making process in tort law. Allocating the right to initiate litigation to victims vindicates this interest in initiating the amends-making process, as well as in controlling both how the amends- making process takes place and the nature of the conciliatory gesture demanded. 3. Epistemic Problems. Allocating the right to initiate litigation to victims solves epistemic problems as well. These problems are closely tied together with some of the problems previously considered, but it is helpful to re-state them in epistemic terms. Allocating the right to victims lets the wrongdoer know, once victims come forward: (1) that there is someone who feels very seriously wronged (one does not initiate litigation, one hopes, without actual serious wrongs); 508 BUFFALO LAW REVIEW [Vol. 62 (2) what the plaintiff wants; (3) how to make amends (the amount demanded, or other remedies); and (4) that the victim does want the injurer to make amends. This explains why the duty to make amends is triggered upon demand rather than at the moment of wrongdoing. The moral duty is triggered upon the moment of wrongdoing, assuming the victim wants amends to be made. But this feature of the moral principle—the fact that the duty is contingent on what the victim wants—is fraught with epistemic uncertainty. Tort law reduces this uncertainty by stipulating that the amends-making process does not commence until the victim explicitly says so in the form of a demand. And the reason why victims, rather than wrongdoers, are allocated the responsibility to initiate amends-making processes is a moral one: victims have morally important interests in controlling aspects of the amends-making process, especially with respect to whether there will be any such process at all. Tort law eliminates a lot of the guesswork created by informal amends-making procedures, as well as the moral principle of making amends itself. So much for the rough sketch of how making amends relates to tort law. Tort law aims to mitigate the obstacles facing the informal process of making amends and does so by formalizing and standardizing the process in a way that protects victims’ morally important interests in controlling aspects of that process. Saying that tort law aims to mitigate obstacles, however, does not mean it always succeeds. This is obvious. Making available a formal mechanism for making amends may induce victims to rush to the courthouse in an effort to extract payments rather than exhaust their informal amends- making options—options that may sometimes do a better job satisfying everyone involved.209 But, as we will see in Part VI, the amends-making account still promises to make sense of many of the institutional features of tort law that traditional corrective justice accounts have difficulty explaining. And— unlike Hershovitz’s account—the amends-making account takes as its touchstone principle an attractive principle of 209. See generally ROBERT C. ELLICKSON, ORDER WITHOUT LAW: HOW NEIGHBORS SETTLE DISPUTES (1991) (describing informal and relatively low-cost norms of dispute resolution among ranchers in California). 2014] MAKING AMENDS 509 making amends, one that remains attractive regardless of whether we label it a theory of “corrective justice.” 4. A Legal Principle of Making Amends. We have seen some problems surrounding the moral process of making amends, problems that we might usefully label the circumstances of making amends.210 Tort law aims to overcome these circumstances and facilitate those who want to pursue claims through a formalized version of that process. In light of these foregoing observations, we are in a position to articulate a principle of making amends that resembles the basic structure of tort law. Consider the following principle of tort law: Making Amends (Legal Version). Individuals who have wronged others in ways recognized by law, or are otherwise responsible for wrongs recognized by law that befall victims, have a legal duty to perform a reasonable conciliatory gesture as recognized by law to those victims, only when the victim makes a legal demand of the wrongdoer (or responsible party) to make amends. The important part is the italicized addition “only” clause at the end, which replaces the desire-based exception to the general rule that wrongdoers have a moral duty to make amends. The biggest difference between the moral version of the principle and the legal version is that it specifies exactly when the legal duty to make amends becomes operative: upon demand by the victim. 210. Hershovitz mentions the “circumstances of corrective justice”—but we mean different things. Hershovitz, Corrective Justice, supra note 9, at 117. He is concerned, primarily, with the problem that corrective justice faces in all cases: that a wrong cannot be undone. See id. We are concerned primarily, and instead, with the many obstacles that often conspire to make it practically difficult for parties to make amends with each other. Tort law aims, according to the making- amends conception, to correct for these obstacles and smooth the path towards making amends. For more on the origins and similar uses of this “circumstances of” locution, see SHAPIRO, supra note 179, at 170, 420 n.11 (attributing the phrase “circumstances of justice” to David Hume and John Rawls, and adopting his own “circumstances of legality” that “obtain whenever a community has numerous and serious moral problems whose solutions are complex, contentious, or arbitrary”). 510 BUFFALO LAW REVIEW [Vol. 62 V. OBJECTIONS, OLD AND NEW Now that the broad outlines of the making-amends approach are in place, let us consider how the approach help answers the various objections already on the table. Then we’ll consider some new objections. A. The Remedies Objection The duty to make amends answers the remedies objection. Rather than a duty of repair, there is a duty to make amends. Notice that this is the feature of the principle that allows us to respond to the remedies objection. As noted in Part IV, the notion of making amends, like getting even, is more open-ended than the notion of repair, which seems distinctively interested in setting things back the way they were before the injury occurred, or at least as close to that position as practicably possible.211 Making amends does not concern itself, primarily, with repairing losses. Making amends is primarily about the person responsible for the wrong making a reasonable and adequate conciliatory gesture to victims. The way that this is accomplished is not necessarily through offers to pay reparations—though in many cases, this is the best way of expressing regret. Sometimes an apology and an offer to pay for the neighbor’s damaged yard will be enough.212 As a result, and as we observed in the prior subsection, the ways by which one can 211. To return to Tom and Jerry, when Tom negligently breaks Jerry’s leg, it makes sense to say that Tom should make amends to Jerry in consideration of the loss. See Hershovitz, Corrective Justice, supra note 9, at 110-19. But it does not make sense to say that a repayment of monetary debt owed between Tom and Jerry involves a case of making amends. The existence of a debt does not involve a wrong (unless the debt is overdue); but the negligently caused broken leg does plausibly involve a wrongdoing. Hershovitz wanted to iron over the obvious differences between Tom and Jerry in the debt case and in the broken leg case, respectively; making amends highlights the differences. See id. 212. The neighbor may accept my apology, appreciate and acknowledge the offer, yet decline it. Amends have been made. But on the Aristotelian account, this is a failure of corrective justice because I have not repaired the damage done; the neighbor has, at best, simply waived his right that I pay full reparations. This does not change the fact, however, that Aristotelian corrective justice has not been done on its own terms. See, e.g., Hershovitz, Corrective Justice, supra note 9, at 117 (discussing Aristotelian theory). 2014] MAKING AMENDS 511 make amends includes, but are not limited to reparations. The fact that other forms of relief—like nominal, punitive, and injunctive relief—are available in tort law should not be surprising from the perspective of making amends. B. The Substantive Standing Objection Recall the substantive standing objection.213 Goldberg and Zipursky point out that causes of action arising in tort all contain a substantive standing requirement that in effect limits the class of plaintiffs to only those who can show that they stand in a sufficiently close relationship to the defendants’ wrongdoing at issue.214 They must show that defendants accountable for that wrongdoing have wronged them, not simply that they have suffered an injury as a result of wrongful conduct.215 The distinction is a subtle one that is recognized in law but not easily captured in terms of plausible moral principles. This creates a dilemma. If a proposed principle of corrective justice contains a plausible principle of moral responsibility, then the principle probably fails to account for tort law’s substantive standing requirements. But if the principle can account for them, then the principle probably does not count as a plausible notion of responsibility and simply parrots the law’s requirements without providing independent explanatory power.216 The amends-making account helps to respond to this objection. On the amends-making account, many of the institutional features of torts can be thought of as solving epistemic, practical, and moral problems that commonly arise trying to make amends informally or as a result of trying to enforce moral duties to make amends.217 In a (cumbersome) slogan, the institution of tort law aims to facilitate the amends-making process while protecting morally important interests of the victim. But notice that doing so will inevitably require having some way of identifying a victim whose interests should be protected by 213. See supra Part I.B. 214. See supra Part I.B. 215. See supra Part I.B. 216. See supra Part I.B. 217. See supra Part IV. 512 BUFFALO LAW REVIEW [Vol. 62 that institution. In light of this inevitability, so-called “substantive standing” rules can be viewed as rules for identifying genuine victims, with morally important interests in initiating and controlling the amends-making process, from those who have merely been collaterally injured by the defendant’s wrongdoing. This maneuver echoes Hershovitz’s to the extent that it views substantive standing requirements as restrictions on who may legitimately claim to have been wronged by the defendant.218 But the present explanation goes further. Apart from being more explicit about being an instrumentalist or functionalist response, we now see precisely why the response—if understood correctly—avoids the dilemma. We should concede that moral liability for wrongs often does not track legal liability for torts. That is, the response concedes the point made by Goldberg and Zipursky: that tort law often holds defendants legally liable even though they should not be morally liable, and sometimes defendants should be held morally liable even if they legally should not.219 And they are also correct in explaining that this mismatch is very often attributable to substantive standing requirements.220 But this concession is far from fatal. Goldberg and Zipursky are mistaken to suggest that there is simply no moral analogue to the substantive standing requirement.221 Indeed, not only is there one (see examples infra Part V.D.3), the amends-making account explains why the mismatch between moral and legal liability nevertheless persists: morally speaking, drawing a line between primary victims— with morally important interests in controlling the amends- making process—from those with much weaker claims, can be quite difficult and fraught with disagreement and controversy. Given this potential for disagreement, tort law slices through the Gordian knot by introducing standards for discerning genuine victims, in the form of substantive standing requirements. There is still a mismatch, however, given that law provides blunt instruments; the mismatch is 218. See supra Part III.B. 219. See supra Part I.B. 220. See supra Part I.B. 221. See supra Part I.B. 2014] MAKING AMENDS 513 predictable. But corrective justice still does justificatory work indirectly: it explains why the institutional features of tort law are what they are and tort law’s mission statement, but it does not explain tort law by providing a perfect normative map by which we can navigate the main features of tort law. Indeed, if we treat moral principles of corrective justice as a map in this way, we wind up pretty disoriented. Civil recourse theorists are right about this. But their dilemma, it turns out, is a false one, because we can cite the practical problems that arise in the context of applying moral principles to explain the role of substantive standing requirements, while explaining why moral and legal liability will not be co-extensive: substantive standing requirements are very, very blunt instruments used to separate genuine victims of wrongdoing from those who have merely suffered unwelcome fallout.222 C. The No-Legal-Duty Objection So far we have seen that the making-amends account provides the resources to respond to the remedies and substantive standing objections. The third objection also poses no difficulty. Indeed, the answer has already been telegraphed. Recall Goldberg’s statement of the objection, which is basically that moral principles of corrective justice fail to cut tort law at the joints: Notice, however, that the conversion of the moral duty of repair into a legal duty does not happen through the tort system unless and until the victim decides to press a claim against the defendant. In other words, if the defendant is going to be made to heed his duty of repair, it will only be by virtue of the law’s having empowered the victim to demand of the defendant that he make amends for the wrong done. . . . Corrective justice theory thus fails to capture accurately the terms on which tort links a victim to a person who has victimized her.223 We are now in a position to see what is correct about this statement and what is not. To be sure, principles of corrective justice that impose automatic duties of repair on wrongdoers 222. See supra Part IV.B. 223. Goldberg, Wrongs Without Recourse, supra note 6, at 13 (emphasis added). 514 BUFFALO LAW REVIEW [Vol. 62 the moment they wrong others fail to cut tort law at the joints. These duties, after all, are not sufficiently sensitive to the victim’s interests. But the Fourth Pass principle does not impose this kind of duty.224 The principle instead specifies that the victim must want the wrongdoer to make amends to trigger the duty to do so.225 But this means that the moral duty, even if automatic, depends on the victim. Victims who do not want wrongdoers to make amends deactivate wrongdoers’ duties to do so. The possibility that duties of repair (or, on the amends-making view, duties to make amends) pop in and out of existence mitigates the mismatch between the law’s structure and the structure of the relevant moral principle. The moral duty to make amends, just like the legal duty, is contingent on the victim in some way. This does not mean, however, that the amends-making conception is out of the woods yet. The moral duty to make amends, after all, is contingent on a mental state of a victim— i.e., victims must want wrongdoers to make amends to activate that duty. What if the victim’s desires change minute to minute? Does that suggest that the wrongdoer’s duty flips on and off like a light switch? Seemingly so. Meanwhile, the legal duty to make amends is triggered by the plaintiff’s conduct. What conduct suffices may vary by jurisdiction.226 But this thought/conduct disparity between the moral and legal principles prima facie shows that the no- legal-duty objection still has some bite—at least on its face. After all, the triggering conditions for those duties do not perfectly align. If they did, then the legal duty to make amends would be triggered the moment that the victim 224. See Part IV.A.3. 225. See Part IV.A.3. 226. Sometimes the filing of the lawsuit is required; other times, the victim need only put the alleged wrongdoer on notice of the demand and when a precise damages calculation can be made. Compare N.J. CT. C.P.R. 4:42-11(b) (starting the clock for prejudgment interest in tort claims “from the date of the institution of the action or from a date 6 months after the date the cause of action arises”), with Levy-Zetner Co. v. S. Pac. Transp. Co., 142 Cal. Rptr. 1, 25 (Cal. Ct. App. 1977) (explaining that, for torts suits in California, “prejudgment interest runs from the date when the damages are of a nature to be certain or capable of being made certain by calculation and when the exact sum due to the plaintiff is made known to the defendant”). 2014] MAKING AMENDS 515 wanted the wrongdoer to make amends, and this is clearly not how it works in tort law. But once we understand the nature of the making- amends explanation, the worry about mismatch goes away. Recall that our understanding of tort law is a functionalist one. Tort law aims to facilitate the amends-making process by removing common obstacles to that process, as it occurs informally. For this kind of explanation to succeed, alleged mismatches between the moral and legal processes must be plausibly explained by showing how those “mismatches” help remove those obstacles. And this is exactly the kind of explanation provided. Tort law allocates the right to initiate litigation—the right to initiate the legal amends-making process—to victims. And tort law does this to solve epistemic problems (e.g., the victim’s identity, the nature of the alleged wrongdoing, and what amends they want made), and the problem perhaps most overlooked: ensuring vindication of the morally important interest that victims have in controlling whether and how the amends-making process takes place. So tort law is explained in terms of corrective justice. The explanation, however, is a functionalist one: that tort law aims to facilitate the amends-making process involving a certain class of serious wrongs.227 Mismatches between the moral structure of making amends and the legal structure of tort practice are not necessarily inadequacies of the explanation or data points that the explanation failed to capture. Rather, these mismatches must be explained in light of the overall purpose of having an institution of tort law. The point of the institution is to overcome many of the difficulties arising from the informal, vague, and ambiguous moral structure of the making amends principle and the process it 227. See, e.g., Gardner, What is Tort Law For?, supra note 204, at 19. Gardner states: A legal norm cannot play its partly constitutive role in relation to a moral norm unless it also has some instrumental role to play in relation to the same moral norm, unless conformity with the legal norm would help to secure conformity with the moral norm of which the legal norm is supposed to be partly constitutive. Id. 516 BUFFALO LAW REVIEW [Vol. 62 gives rise to. These mismatches, in many cases, can be explained in terms trying to overcome these difficulties.228 We should not expect, therefore, that tort law perfectly align with all the amorphous contours of the informal, moral process of making amends. Indeed, if the morality of making amends mapped so cleanly on to the law, we might wonder why we even need the institution of law to get in the business of making amends at all.229 D. Other Objections The making-amends conception of corrective justice faces challenging questions about its adequacy as an explanation and justification for tort law, as well as about its status as a genuine moral principle. These questions extend beyond the challenges to corrective justice theory posed by Goldberg,230 228. The methodology presupposed here—that legal institutions arise to address problems that cannot be solved by informal moral norms or customs—is far from idiosyncratic. Perhaps the best known example of this methodology comes to us from H.L.A. Hart’s suggestion that law in general aims to overcome the various inefficiencies and uncertainties that would be faced by a “pre-legal” society governed exclusively by informal customs. H.L.A. HART, THE CONCEPT OF LAW 91 (1961). This tradition is carried on today in, among other places, SHAPIRO, supra note 179, at 170-71 (positing that legal institutions in general aim to correct for deficiencies in nonlegal forms of social planning). John Gardner discusses a similar methodological perspective in connection with tort law, specifically, en route to defending instrumentalism in tort theory against non-instrumentalists Jules Coleman and Ernest Weinrib. See Gardner, What is Tort Law For?, supra note 204, at 19 (“[T]o fulfill its morally constitutive role, tort law’s norm of corrective justice must be evaluated as an instrument. It must be evaluated as an instrument of improved conformity with the very moral norm [of corrective justice] that it helps to constitute.”) (first emphasis in original; second emphasis added). 229. Here is another way to put the point. The explanation offered here does not expect a one-to-one match between the basic structures. Indeed, the reason why we even need something like the institution of tort law is because of the various latent defects arising from reliance on first-order moral norms that are vague, ambiguous, and in any event difficult to apply. Put differently, the reasons we have tort law is to correct for epistemic, practical, and moral problems that arise from the amends-making process. Tort law corrects for the defects of corrective justice. So it would be surprising if there were a perfect match between the normative structure of the moral norm and the legal one, given that the structure of the moral norm is part of the problem. 230. See supra Part I. 2014] MAKING AMENDS 517 Hershovitz,231 and Zipursky.232 Although not every objection can be considered, let us address some salient ones. 1. Can We Make Amends Through Tort Law? Here is one objection: in an ideal case, a wrongdoer would come forward and try to make amends to the victim for the wrongdoing and would do so by offering a reasonable and adequate conciliatory gesture, which the victim would accept. But lawsuits seem far from the ideal. After all, there is a sense of “conciliatory” behavior that may require the wrongdoer to actually regret his wrongdoing, and which suggests that the wrongdoer must personally acknowledge his conduct as wrong. But if tort law aims to facilitate the amends-making process, and if amends-making implies conciliatory gestures, then tort law must be doing a horrible job since tort cases do not police these subjective feelings of contrition. Nor do tort suits require explicit acknowledgment of wrongdoing by the defendant. In short, tort law does not look like it really is geared towards making amends after all. This objection rests on a misunderstanding. As explained earlier, conciliatory gestures do not require the wrongdoer to maintain any particular attitudes, even though ideally the wrongdoer would regret wrongdoing.233 In the relevant sense, we should understand conciliatory gestures as objectively signaling regret and acknowledgement of wrongdoing. It is perfectly possible within the framework outlined above that someone can be found liable and forced to make a conciliatory gesture to the victim, even though the wrongdoer does so grudgingly. Recall Hatfield, who offers up, say, fresh produce because he must, even though he does so grudgingly.234 Or consider bickering siblings who are forced by their parents to shake hands and hug to signify the end of the dispute, even though neither sibling particularly wants to do so. Conciliatory gestures needed to satisfy the duty to make amends are objective, not subjective in nature. If this analysis is correct, then the upshot for tort law is clear: tort law can force parties to make amends by forcing 231. See Hershovitz, Corrective Justice, supra note 9, at 110-17. 232. See supra Part I. 233. See Part IV.A.2. 234. See Part IV.A.2. 518 BUFFALO LAW REVIEW [Vol. 62 wrongdoers to offer up reasonable and adequate conciliatory gestures to victims. The fact that the litigants may loathe each other during and after the litigation and the fact that the wrongdoer never feels regret, or even acknowledges having done anything wrong, does not necessarily mean that the wrongdoer has failed to satisfy his obligation to make amends. And it should not be surprising that people can satisfy their obligations without necessarily accepting the legitimacy of those obligations, just as someone who rejects the legitimacy of federal income taxes may wind up satisfying his legitimate tax obligations by paying them. 2. The Distinctiveness Objection. Another challenge goes as follows. In seeking to capture a broader array of remedies, which the making-amends account surely does, the amends- making conception arguably loses sight of what is (allegedly) important and distinctive about tort law, which is its apparent emphasis on repair.235 This duty gives guidance as to the forms of redress appropriate and the amount of compensation that is appropriate.236 But arguably the making-amends conception fails to provide comparable guidance.237 On the making-amends conception, the fact that tort law awards compensatory damages as of right seems contingent or accidental. To the contrary, a traditionalist might insist, the fact that reparations are the default form of redress reflects something normatively important that is far from simply an accident of history. Only Aristotelian views come close to explaining why reparations make any normative sense.238 This objection has some bite. To some extent, the greater indeterminacy in the making-amends conception with respect to forms of redress is simply the theoretical cost of being able to account for a greater variety of remedies. And 235. See Gardner, Torts and Other Wrongs, supra note 8, at 59; Stapleton, supra note 178, at 1559-60; Part IV.A.2. 236. See Gardner, Torts and Other Wrongs, supra note 8, at 59. 237. Jules Coleman & Gabriel Mendlow, Theories of Tort Law, STANFORD ENCYCLOPEDIA OF PHILOSOPHY, http://plato.stanford.edu/entries/tort-theories/ (claiming that civil recourse theory gives “little guidance as to what sort of redress is appropriate”). 238. See supra Part I.C. 2014] MAKING AMENDS 519 this is a cost that the making-amends conception—along with the getting-even conception—is willing to incur. To the extent that the making-amends view feels pressured to explain why the default rule of redress is the default rule of compensation, the view seems destined to either go on offense—like Hershovitz does, who claims that notion of “repair” is itself misleading239—or emphasize that repair- based explanations incur their own costs, such as the difficulty in explaining the variety of remedies and damages available beyond purely reparative ones. The whole exercise feels like trying to put queen-sized sheets on a king-sized mattress. But perhaps the fact that courts award compensatory relief by default does not, on reflection, represent any particularly deep moral truth. As noted earlier, very few, if any, gestures better signal regret than compensatory relief designed to undo damage to the extent possible.240 Again, if A accidentally pushes B to the floor, then the most obvious way to signal regret is by attempting to help B to his feet. So it is no surprise that tort law’s predominant method of redress takes the form of undoing harms done via monetary or injunctive relief. The mistake, however, is to see reparative relief as the sine qua non of corrective justice. We can scale up this point by imagining cultures where, say, apologizing is a response morally on par with providing adequate compensation.241 We must be careful not to mistake the culture-independent, abstract moral principle of corrective justice—which is fundamentally about making amends—with particular entrenched forms by which we make amends. Separating the two is not easy work; and, by 239. See supra Part III.A. 240. See supra Part IV.A.2 (explaining that conventional reparative relief might be viewed as simply the best way to signal regret, objectively speaking). 241. We might not even have to imagine this society. See Jon O. Haley, Comment: The Implications of Apology, 20 LAW & SOC’Y REV. 499, 500-01 (1986) (claiming that the refusal of one corporation to provide an apology to Japanese defendants “held up a settlement for months”). For more on the role of apologies and compensation in Japanese legal culture, see Hiroshi Wagatsuma & Arthur Rosett, The Implications of Apology: Law and Culture in Japan and the United States, 20 LAW & SOC’Y REV. 461 (1986). 520 BUFFALO LAW REVIEW [Vol. 62 explicating the concept of making amends, we risk committing the same sin. But it is worth trying. Here is another way to reconcile Aristotelian conceptions of corrective justice with the idea of making amends. Perhaps, after all is said and done, traditional corrective justice theories have indeed articulated one genuine repair- based principle of justice. Suppose further that they succeed in showing that the core of tort law reflects this principle. It still does not follow that the Aristotelian conception exhausts the field of corrective justice, or that making amends has no work left to do in illuminating private legal practices. Aristotelians might simply be mistaking a particular species with the broader genus. Corrective justice in this broader sense of making amends might be the goal of private litigation more generally. And if this is correct, then much of the problem might be a problem of labeling—one that is not particularly interesting. Making amends might be the fundamental category pertaining to the problem of what wrongdoers must do for their victims in the aftermath of their wrongdoings, while corrective justice specifies the default manner by which amends are made to victims. Let us not rule out the possibility of reconciliation between Aristotelians and those (like Hershovitz and the present author) seeking to expand the scope of remedies with which Aristotelians are preoccupied. 3. Does Making Amends Presuppose a Concept of Repair? Another objection is that making amends simply presupposes the concept of repair, since making amends typically suggests trying to repair a frayed relationship.242 If so, this objection would spell trouble because it (1) threatens to reintroduce the problematic concept of repair; (2) shows that making amends itself presupposes some version of the Aristotelian, repair- based view of corrective justice; and (3) shows that making amends might not be so well suited to illuminate tort practice after all, since tort suits are frequently populated by strangers who had no prior relationship before the suit. So if there is no prior relationship, what is there to “mend”? The short answer to this is that the amends-making account of making amends presupposes neither a pre- 242. See Radzik, supra note 188, at 146-47. 2014] MAKING AMENDS 521 existing relationship (prior to the wrong that unites the victim and wrongdoer) nor any robust concept of repair. On the amends-making account, making amends does not presuppose that a pre-existing relationship exists between wrongdoer and victim. To be sure, there is a sense in which we all bear a moral relationship with respect to one another, a relationship that we do not hold toward, say, crab grass. But if this abstract, moral sense of “relationship” is the relevant one, then we should be skeptical that this is the kind of relationship that ever needs repair. It is an unbreakable relationship between persons that makes us accountable to one another invariably. On the other hand, if a weaker sense of “relationship” is in play—such as the relationship between lovers that can be broken—then we should resist the claim that making amends necessarily is in the business of repairing these more fragile relationships. It is too easy to think of cases in which the victim, although wanting and demanding amends, wants nothing further to do with the wrongdoer besides those amends—and can hold this attitude without denying that the moral relationship that holds between persons can ever be frayed.243 4. The Arbitration Objection. The Arbitration Objection begins by observing that alternative dispute resolution (ADR), such as arbitration or mediation, might do a better job at getting parties to make amends than tort law.244 Three considerations might bolster this view: (1) making amends and ADR both allow for a broader range of remedies than tort law;245 (2) ADR is (arguably) more likely to result in a reasonable and conciliatory gesture than tort law; and (3) the substantive rules that govern ADR and the process of making amends are far more flexible than those found in substantive 243. There is clearly more to this objection—as well as more to say in response. Much of Linda Radzik’s recent book is devoted to defending a conception of making amends that sees the goal of the process as repairing relationships. See generally RADZIK, supra note 184. 244. See, e.g., Thomas O. Main, ADR: The New Equity, 74 U. CIN. L. REV. 329, 362 (2005). 245. See, e.g., id. (noting that remedies available in alternative dispute resolution (ADR) include not only “monetary and injunctive remedies” but can also “promote certain behavior, restructure relationships, and impose outcomes beyond the legal and practical reach of courts”). 522 BUFFALO LAW REVIEW [Vol. 62 tort law and civil procedure.246 Given these observations, a natural thought is that the amends-making conception of corrective justice helps make sense of ADR, but not tort law. None of these observations is worrisome. Recall that the making-amends approach seeks to explain tort law, but it does not seek to explain it by showing how the normative structure of corrective justice maps cleanly onto the normative structure of tort law. The explanation is functionalist in a way that explains why there will be some structural mismatches between the two normative domains. The approach predicts that legal norms aim to correct for certain shortcomings in the ordinary moral process of making amends. But to the extent that the ordinary moral process works for people or some other more formal process of private arbitration provides an adequate substitute, so much the better: tort law need not get involved. We can see how this observation helps answer the objection. Tort law does not aspire to be the only formal mechanism that facilitates the amends-making process; it is the formal mechanism of last resort. To the extent that individuals can make amends without arbitrators or mediators, they might avail themselves of a much broader range of potential remedies, as well as address a much wider range of wrongs. To the extent individuals cannot make amends on their own, other, formal and private grievance procedures might be available. The range of remediable wrongs and remedies narrows, however, once these individuals try to take their grievances to private arbitrators. And by the time the victim and wrongdoer reach the court 246. See id. at 366. Main notes the “substantive flexibility” of certain forms of ADR: In the more voluntary and less structured forms of ADR, such as mediation, where the ultimate authority belongs to the participants themselves, the parties (perhaps with the benefit of a third party facilitator) can fashion a unique solution that will work for them without being strictly governed by precedent. Thus, “mediators do not ‘judge’; they aid the parties in ending a dispute.” ADR is attractive to some, then, because of the system’s promise of “better” results that serve “the real needs of the participants or society.” These results may or may not “follow the law,” and it arguably does not matter because of the parties’ voluntary acquiescence to the resolution. Id. 2014] MAKING AMENDS 523 system, the range narrows even further. So the objection is in a sense correct: ADR does prima facie promise to achieve corrective justice—and maybe even better than tort law— given the broader range of remediable wrongs and remedies potentially available. But that does not mean that the making-amends view fails to explain tort law. It just shows that the nature of the explanation is an instrumentalist one. Tort law steps in as the institution of corrective justice of last resort, to correct for failures of the informal process of making amends, as well as formal (but nonlegal) institutions like private arbitration or mediation. Looking at tort law this way, moreover, has a considerable benefit in squaring corrective justice with another commonly observed fact about civil litigation: most cases settle, and moreover, public courts actively encourage settlement by litigants.247 This feature of tort litigation might be viewed as worrisome from the perspective of traditional conceptions of corrective justice, which hold that corrective justice is accomplished only upon awarding full reparative compensation by the wrongdoer to the victim.248 After all, widespread settlement raises the possibility that, at nearly any stage of litigation, a plaintiff may agree with a defendant’s offer to settle at values well below what they might be entitled to as a matter of compensatory relief. From the perspective of repair-based accounts of corrective justice, any time a plaintiff entitled to compensatory damages settles for something below full compensatory relief; this is a prima facie injustice. Settlement, in these circumstances, fails to achieve corrective justice of the Aristotelian variety. 247. It is widely acknowledged that most lawsuits in the United States settle, and that the legal system contains various mechanisms that serve to encourage settlement rather than public adjudication. For a thorough but critical review of these facts about U.S. legal practice, see Marc Galanter & Mia Cahill, “Most Cases Settle”: Judicial Promotion and Regulation of Settlements, 46 STAN. L. REV. 1339 (1994). 248. JULES L. COLEMAN, MARKETS, MORALS AND THE LAW 203 (1988) [hereinafter COLEMAN, MARKETS] (“If justice requires giving each party his or her due . . . his or her desert, then settlements almost never satisfy this principle of justice. One concern, then, is the extent to which parties can legitimately or justifiably negotiate around the requirements of justice, and the extent to which legal institutions that encourage their doing so . . . are desirable and defensible.”). 524 BUFFALO LAW REVIEW [Vol. 62 True, encouraging settlement might be explained on grounds of efficiency and cost. Courts are costly institutions with heavy caseloads. So it makes sense that they concern themselves with live disputes only.249 Surely, however, if tort law embodies the principle of corrective justice and seeks to ensure that corrective justice is done, why should courts encourage settlement, even in cases where they think plaintiffs might be entitled to full compensation?250 Why then do judges regularly approve of settlements that fall well lower than compensatory relief, even in cases where the defendants appear to have tortiously injured the plaintiffs?251 And why do courts lack the authority to more regularly revisit settlements that are inadequate? The law seems to discourage revisiting settlements and binding arbitration, even in cases in which it is fairly clear that the results of those processes fall short of full compensation.252 None of these mysteries is difficult to explain, however, on accounts of corrective justice that do not regard full compensatory relief as the sine qua non of corrective justice. The default rule in some jurisdictions is that courts simply do not get involved in settlements.253 True, they sometimes police the outer boundaries of decency in approving these settlements, and hence are at least nominally required to 249. See, e.g., People v. Lybarger, 700 P.2d 910, 915 (Colo. 1985) (“Courts exist for the purpose of deciding live disputes involving ‘flesh-and-blood’ legal problems with ‘data relevant and adequate to an informed judgment.’”) (quoting New York v. Ferber, 458 U.S. 747, 768 (1981) (internal quotation marks omitted)). 250. See COLEMAN, MARKETS, supra note 248, at 203. 251. Id. at 204-05. 252. On arbitration, see, e.g., Duferco Int’l Steel Trading v. T. Klaveness Shipping A/S, 333 F.3d 383, 388 (2d Cir. 2003) (“It is well established that courts must grant an arbitration panel’s decision great deference.”). The circumstances under which federal courts are permitted to revisit arbitration decisions are severely limited by statute. See, e.g., Stolt-Nielsen SA v. AnimalFeeds Int’l Corp., 548 F.3d 85, 90-91 (2d Cir. 2008) (citing the Federal Arbitration Act, 9 U.S.C. § 1 et seq. (2006)). 253. See, e.g., Estate of Sa’adoon v. Prince, 660 F. Supp. 2d 723, 724-25 (E.D. Va. 2009) (“Federal courts are not vested with a general power to review and approve settlements of suits between private parties. . . . It is fair to say, then, that federal courts as a general matter do not review and approve settlements of lawsuits between private parties.”). But see id. at 725 (listing exceptions). 2014] MAKING AMENDS 525 determine whether they are reasonable.254 But they police to ensure reasonableness—not necessarily to ensure that plaintiffs who might prevail receive full compensatory relief.255 The making amends account explains why: corrective justice requires a reasonable and adequate conciliatory gesture to the victim—and victims are typically in the best position to determine what gesture counts as reasonable and adequate. CONCLUSION Corrective justice theorists seek to explain tort law’s key concepts in terms of principles of corrective justice, which are typically understood to incorporate duties of repair. But these accounts face seemingly powerful criticisms from civil recourse theorists, who point out important features of tort law’s normative structure that corrective justice theory has difficulty explaining. They propose rejecting corrective justice theory. Scott Hershovitz responds by arguing that, rather than abandoning corrective justice as an explanation of tort law, we should alter our understanding of corrective justice itself. This paper continues the dialectic. We saw how Hershovitz’s understanding of corrective justice incompletely responds to these criticisms and threatens to distort rather than illuminate tort law. And we have tried to provide a new account of corrective justice both plausible in its own right 254. A federal judge in the Southern District of New York, made headlines when he declined to approve a settlement between the Securities and Exchange Commission and Citigroup Global Markets. See Edward Wyatt, Judge Rejects an S.E.C. Deal with Citigroup, N.Y. TIMES, Nov. 29, 2011, at A1. This grabbed attention in part because courts rarely disapprove of settlements involving the agency. The court found that the proposed settlement, according to which Citigroup admitted no wrongdoing, was “neither fair, nor reasonable, nor adequate, nor in the public interest.” SEC v. Citigroup Global Mkts., 827 F. Supp. 2d 328, 332 (S.D.N.Y. 2011). The case is on appeal. See SEC v. Citigroup Global Mkts. Inc., 673 F.3d 158 (2d Cir. 2012). 255. See, e.g., Lynn’s Food Stores, Inc. v. Dep’t of Labor, 679 F.2d 1350, 1353, 1355 (11th Cir. 1982) (explaining that, even in the context of claims pursued under the Fair Labor Standards Act (FLSA), which requires federal courts to closely scrutinize settlements between employees and employers, the overarching goal of these judicial evaluations is to ensure that they represent “a fair and reasonable resulution [sic] of a bona fide dispute over FLSA provisions”). 526 BUFFALO LAW REVIEW [Vol. 62 and able to rebut the objections against traditional, repair- based conceptions. In reorienting our understanding of corrective justice, we relied on some of Hershovitz’s insights. He is correct, for example, to recognize that responding to one of the objections—the “remedies” objection—requires describing the core duty of corrective justice in more general terms. Out went the duty of repair, and in came the more fundamental, and more abstract, duty to make amends. The other objections attacked the ability of corrective justice to explain the basic normative structure of tort law, by focusing on so-called “substantive standing” requirements and the fact that no duty of redress arises at the moment a tort occurs. Because these objections are structural in nature, they require a structural response. To that end, we conjectured that the goal of tort law is to facilitate the amends-making process in a way that protects the morally important interests of victims to control how that process proceeds, if at all. This instrumentalist approach to tort theory helped to explain the structural mismatch. We also examined the issue of victim control, which has been almost entirely overlooked in the literature on corrective justice. But it should not be. Considering victims’ interests in control helps to provide a more nuanced understanding of corrective justice, as well as a largely untapped explanatory resource wholly consistent with corrective justice. It allows us to explain, for example, why victims are allocated rights of action in tort, as opposed to imposing affirmative legal obligations on wrongdoers to come forward in the aftermath of their wrongdoings: allocating that obligation on wrongdoers would overlook the moral importance of affording the right of victims to simply forego amends. And imposing obligations on wrongdoers threatens to divest victims of the important interest they have in initiating the amends-making process, including having the first crack and making and shaping their allegations. But there has been a tacit concession to Goldberg and Zipursky throughout this paper that I have tried to make explicit. They argue that corrective justice theories do not always cut tort law at the joints. The amends-making account does not either. But the account does explain why this should not be worrisome. This is because, in saying that tort law aims to facilitate the informal amends-making process by 2014] MAKING AMENDS 527 correcting for certain problems inherent in that process, one need not be committed to claiming that the normative structure of tort law and corrective justice map onto each other perfectly and in every respect. Indeed, because law is a blunt and formal instrument that tries to make the rights and responsibilities of parties more explicit than they would otherwise be in the subtle and informal world of morality, we should expect there to be an imperfect mapping from one domain to the other. The making-amends account gives us some way to predict where the mismatch will happen legitimately. This article remains a sketch. Making good on the promises of the making-amends conception of corrective justice and tort law will require further work. One issue that needs to be addressed is to see whether the making-amends conception is compatible with strict liability doctrines. This is an important area of tort law that has notoriously posed difficulties for corrective justice theories. Another important project is to show how, if at all, this very abstract theory has normative bite. Does it leave tort law just as it is or does it help us resolve some enduring controversies about tort law? Does the making-amends conception favor some default rules over others? There is also work to be done in fleshing out a conception of responsibility that best coheres with the making-amends conception. In any event, we should share Scott Hershovitz’s optimism that tort theorists and theorists of private law more generally will benefit from expanding our understanding of corrective justice. Doing so will not only help illuminate tort law, but will also help us to reconcile corrective justice and civil recourse theory. They are natural complements, not competitors.256 But this is also a project for another occasion. 256. Others seek a similar reconciliation. See, e.g., Arthur Ripstein, Civil Recourse and Separation of Wrongs and Remedies, 39 FLA. ST. U. L. REV. 163, 198- 203 (2011).