Iowa Law Review
77 Iowa L. Rev. 625
LENGTH: 47394 words
SYMPOSIUM: Substantive Corrective Justice
NAME: Richard W. Wright *
* Professor of Law, Illinois Institute of Technology, Chicago-Kent College of Law. I received financial support for this article from the Marshall D. Ewell Research Fund of the Chicago-Kent College of Law. Discussions with Allen Buchanan, Jules Coleman, Stephen Perry, and Ernie Weinrib at an informal conference on moral theories of tort law, hosted by Jules Coleman at Yale in October 1990, helped clarify my thoughts on some of the issues discussed in this article. I also have benefitted from comments by Anita Bernstein, Michael Friedman, Wendy Gordon, Steven Heyman, Richard Warner, and Ernie Weinrib.
... As with the concept of causation, which plays a major role in certain areas of liability, the concept of justice seems to be intuitively comprehended and applied to particular situations in a fairly consistent fashion, without any explicit elaboration of the precise content of the concept. ... Thus, although the concept of Right focuses on the external aspect of the exercise of freedom, it presupposes and is dependent on the internal aspect, which is the foundation of Kant's moral philosophy and is the subject matter of ethics or virtue. ... On its face, this argument contradicts Weinrib's often-repeated insistence that the normativity implicit in law, including distributive justice as well as corrective justice, reflects the Kantian concept of Right -- of negative freedom from external interference rather than positive fulfillment of needs -- which allegedly is entirely independent from Kant's ethics. ... If any interaction results in unjust losses or gains in such holdings, the judge in implementing corrective justice reaffirms and restores the preexisting equality by imposing a duty on the injurer to disgorge any unjust gain and to compensate the injured for any unjust loss, so that each party "has his own":
... Corrective justice protects against interactional injuries or threats of injury to a person's goods regardless of the distributive justice or injustice of the pre-interaction division of goods among the parties to the interaction or among the members of the community as a whole. ...
[*625] I. Introduction
Most legal scholars, lawyers, and members of the public traditionally have assumed that law in general, and liability rules in particular, are and should be based on principles of justice. Yet the principles of justice that are assumed to underlie the law have received very little elaboration. As with [*626] the concept of causation, which plays a major role in certain areas of liability, the concept of justice seems to be intuitively comprehended and applied to particular situations in a fairly consistent fashion, without any explicit elaboration of the precise content of the concept. No significant difficulty occurs until problems arise which are not easily resolved by our unelaborated intuitive comprehension of these concepts, or unless a frontal attack is made on their intelligibility.
Both of these challenges have surfaced in recent years. An increasing number of legal problems test the boundaries of our intuitive comprehension of the concepts of causation and justice. At the same time, frontal attacks have been made on the intelligibility of these concepts by those who see them as impediments to the pursuit of their preferred social objectives. The attackers range from the disciples of efficiency, loss spreading, and other aggregative social welfare objectives at one end of the spectrum 1 to the anarchical antilegalists at the other. 2
Given these challenges, those who believe that the law is or should be society's instrument for achieving justice no longer can rest content on the intuitive, unelaborated concepts of causation and justice. If the challenges are to be confronted and overcome, these fundamental concepts must be concretely defined and elaborated. Building on earlier work by H.L.A. Hart and Tony Honore, I have attempted in a series of articles to define and elaborate the concept of causation and to defend it against its critics. 3 In this Article, I begin a similar undertaking for the interrelated concepts of justice and rights. My primary focus will be the concept of corrective justice. However, I also will address the concept of distributive justice, since I believe that a proper understanding of corrective justice can only be gained by exploring the contrasts and similarities between distributive justice and corrective justice.
[*627] A substantial amount has already been written on the concept of corrective justice, most of it limited to liability in tort. 4 The recent literature contains much valuable analysis and has reestablished corrective justice as a concept that must be acknowledged in discussions of tort liability. Unfortunately, however, the academic participants in those discussions, other than the expositors of corrective justice, usually have given the concept of corrective justice only brief acknowledgement, followed by an even briefer dismissal. 5 This curt treatment, I believe, reflects the failure of the current literature to provide a plausible substantive elaboration of the concept of corrective justice.
The most fully developed and best known modern expositions of corrective justice have been provided by Jules Coleman 6 and Ernest Weinrib. 7 Their work is extensive, sophisticated, and enlightening. Yet it is [*628] usually carried on at a high level of formal abstraction, focusing on the coherence or logic of the conceptual superstructure, with relatively little attention being paid to the doctrines which flesh out that superstructure or the normative principles which provide the foundation for and give shape to the superstructure. Indeed, Weinrib, in his recent writings, rejects the idea that there is any such foundation, claiming that law and justice consist solely of, and are exhausted by, the conceptual superstructure. 8 Given the formalism of their respective accounts of corrective justice, it is not surprising that both Weinrib and Coleman are expositors, but not advocates, of corrective justice. Each of them has indicated that he would have no problem with, and indeed might prefer, replacing tort liability with no-fault compensation schemes. 9
At the other end of the spectrum, the legal pragmatists deny the possibility of giving any substantive or even conceptual account of corrective justice. According to them, corrective justice cannot be elaborated in terms of a set of principles or concepts, since no principle, rule, or concept can be given determinate content. Instead, they claim, corrective justice (or tort law) is simply the name for a process of pragmatic dispute resolution which develops and affirms unelaborated, context-specific moral intuitions or social norms. The principles, rules, and concepts employed by the participants in the dispute resolution process allegedly have no substantive content, but rather are mere rhetorical devices. 10
Other writers have adopted an intermediate position. While believing in the substantive intelligibility of concepts, they have focused their efforts on neither the conceptual superstructure nor the principled foundation of corrective justice. Instead, these scholars have concentrated on the rules and doctrines that flesh out the superstructure and have attempted to articulate some overarching paradigm -- for example, "causation" 11 or [*629] "reciprocity" 12 -- to explain those rules and doctrines. These attempts, although novel and stimulating, have not been successful, 13 due at least in part to their failure to investigate the superstructure and the foundations. 14
I believe that the formal and pragmatic accounts of corrective justice, and indeed of law and justice in general, are dead ends. They fail to describe or explain our current legal practices or to help us explain and justify legal obligation. The only plausible approach, descriptively and normatively, is a principled approach, which begins with our actual legal practices, attempts to discern the fundamental principles of morality and justice that underlie those practices, and then uses those fundamental principles to critique and shape our practices in the changing circumstances of political and social life. 15
In this Article, I lay the groundwork for the elaboration of a principled account of corrective justice by attempting to rescue the traditional concept of corrective justice, as first elaborated by Aristotle, from its formalistic evisceration by Coleman and Weinrib. There are several good reasons for beginning with Aristotle's account. First, Aristotle originally elaborated the concept of justice, including the distinction between distributive justice and corrective justice, and, as in many other areas, his treatment is still considered to be one of the best. 16 Second, contrary to Weinrib's portrayal, [*630] Aristotle was a principlist rather than a formalist (or a pragmatist). 17 Indeed, as far as I am aware, Aristotle's account is the only account of corrective justice that has been developed using the principled approach. Since Aristotle also was one of the world's greatest thinkers, anyone who seeks to develop a principled account of corrective justice can reasonably expect to learn much from a study of Aristotle's account. Third, if Aristotle's account seems both descriptively plausible and normatively attractive today, in the light of our current practices and beliefs, this match will itself constitute significant evidence of the enduring existence of certain fundamental principles of liability and justice.
Before turning to Aristotle's account, I critique, in Parts II and III respectively, Weinrib's and Coleman's accounts of corrective justice. In Part II, I argue that Weinrib erred when he shifted from his initial principled approach to a purely formal or conceptual approach, according to which the conceptual superstructure of the law allegedly has an immanent moral rationality that is completely independent from ethics, politics, or any external purpose. I attempt to demonstrate that Weinrib's formal account is an unstable -- albeit intellectually dazzling -- structure of unsupported assertions and logical contradictions. Weinrib's shift to the formal account is attributable, I believe, to a misreading of Kant's theory of law and justice, which Weinrib interprets as being independent from and prior to Kant's ethical and political theory. Actually, Kant's theory of law and justice presupposes and depends on his ethical theory and constitutes his theory of politics as well as his theory of law and justice. I agree with Weinrib's assertion that the conceptual superstructure of the law reflects a nonaggregative Kantian morality rather than an aggregative utilitarian morality. In contrast to Weinrib, however, I maintain this morality is extrinsic to and provides the foundation for, rather than being immanent in, the conceptual superstructure.
In Part III, I claim that Coleman, despite his explicit antiformalist stance, has developed an account of corrective justice that displays a formalist preoccupation with the conceptual superstructure to the virtual exclusion of any investigation of the normative foundation. However, while Weinrib takes concepts too seriously, Coleman does not take them seriously enough. By asserting an analytic and normative distinction between the grounds of recovery and the grounds of liability and, further, between those grounds and the mode of rectification, Coleman fails to take seriously both the concept of a right (or legitimate claim), which logically entails a correlative duty, and the concept of corrective justice, which is a normative structure of correlative bilateral rights and duties of interacting parties.
By discussing corrective justice in purely formal terms, both Coleman and Weinrib, as the major current expositors of corrective justice, have created the impression that corrective justice is an empty, sterile abstraction with little or no relevance to the real world of human interaction. In Part IV, I attempt to reclaim and restore the traditional Aristotelian concept of corrective justice. I argue that the Aristotelian concept of justice, including distributive justice as well as corrective justice, has a nonaggregative and [*631] egalitarian substantive normative content that largely prefigures Kant's moral theory. In Part V, I draw out some important implications of Aristotle's account for the proper assertion and enforcement of legal justice claims.
Some readers may prefer to concentrate on the portions of this Article that contribute to the affirmative elaboration of corrective justice as a substantive normative structure of correlative bilateral rights and duties of interacting parties. I suggest that such persons read section II.D (on the Kantian normativity of corrective justice and distributive justice), the end of section III.A (on corrective justice as a structure of correlative bilateral rights and duties), and Parts IV and V (on the substantive content and structure of Aristotelian corrective justice and distributive justice).
II. Weinrib's Explicit Formalism
A. The Shifting Attack on Instrumentalism
During the last decade, Ernest Weinrib mounted an increasingly strong attack against instrumental conceptions of law and justice. However, as the decade progressed, his definition of instrumentalism shifted to reflect a fundamental shift in his position on the relationships between law, justice, morality, and politics -- a shift from principlism to an extreme version of formalism. 18
At the beginning of the decade, Weinrib endeavored to identify and explore the basic moral principles underlying liability in tort. 19 He assumed that those persons involved in making or interpreting the law generally should incorporate accepted notions of moral obligation, 20 and he articulated the principlist's understanding of the relationship between law, justice, and morality:
The legal argument for a duty [of easy rescue] is not one of deduction but one of coherence. Recognizing that the central concepts are those of obligation and liberty, the argument looks for the common law's concrete manifestation of society's general intuitions about these concepts. Examining instances of the operation of these concepts, the argument extracts general features from the instances, searching for and shaping a coherent pattern that fits as much of the law as possible while respecting the underlying ethical intuitions. 21
Like many others, Weinrib distinguished "instrumental" conceptions of law, justice, and morality from "noninstrumental" conceptions by whether they incorporated an aggregate social welfare (e.g., utilitarian) or a rights-based (e.g., Kantian) ethics, respectively. Although initially treating [*632] both the instrumental and noninstrumental conceptions as "ethical traditions that inform the common-law system," 22 Weinrib soon rejected the former in favor of the latter, at least in the context of tort law.
However, Weinrib's rejection of the instrumental conceptions was based primarily not on their lack of substantive merit, 23 but rather on their incompatibility with the formal structure of tort law. Tort law, he argued, embodies corrective justice's limited focus on the effects that a discrete human interaction has on the presumed equality of the interacting parties, rather than distributive justice's broader focus, independent of any interaction, on the distribution of some resource among competing claimants according to their relative ranking under some comparative criterion. Instrumental conceptions of tort law, with their aggregative focus which inevitably reaches beyond any particular interaction, are incompatible with tort law's corrective-justice form. 24
The formal, rather than moral, basis of Weinrib's rejection of the instrumental conceptions of tort law was emphasized by his agnosticism, constantly reiterated in subsequent articles, on the choice between tort law and no-fault compensation schemes. According to Weinrib, there is no reason as a matter of justice for choosing the traditional corrective justice tort liability scheme as the primary mode of dealing with injuries inflicted by one person on another rather than a more comprehensive no-fault compensation scheme or some other distributive justice scheme. Corrective justice and distributive justice, he states, are forms or types of justice rather than principles of justice. They are alternative structures of justificatory argument, or modes of legal ordering, that are equally but not concurrently applicable to any external human interaction or relationship, which can be viewed either as an immediate interaction (governed by corrective justice) or a mediated relation (governed by distributive justice). The choice of viewpoint must be made on substantive (presumably political or moral) grounds that are extrinsic to the concept of justice per se. 25
Originally, Weinrib believed that extrinsic substantive principles or policies had to be employed not only to choose between distributive justice and corrective justice as the preferred mode of ordering in any particular situation, but also to fill in the specific content of either form of justice. He argued that, to be compatible with corrective justice's narrow focus on discrete interactions of presumed equals, the content to be imported into tort law must be inspired by Kant's ethics -- that is, it must be nonaggregative [*633] and must treat the parties to an interaction as equal members of the kingdom of ends. 26
Subsequently, however, building on a presumed radical disjunction between law and ethics in Kant's philosophy, Weinrib has insisted that law and justice (which, for Weinrib, are synonymous) are completely autonomous from any external principles or policies, even those inspired by Kant's ethics. 27 For Weinrib, the essence of law is its inner intelligibility or "immanent rationality" -- the internally coherent universal unity of its interdependent essential concepts. The immanent rationality of law, he argues, must take the form of either distributive justice or corrective justice. The legal content of law is exhausted by these forms, each of which in turn is constituted by its interdependent essential concepts. The purpose of law is not any political, ethical, or other external purpose, but rather the complete development of its own immanent rationality. 28 Law is "an end in itself constituting, as it were, its own ideal." 29
Weinrib equates corrective justice with "private law" (private adjudication of external human relations viewed as immediate interactions) and distributive justice with "public law" (judicially enforceable limitations on the legislative or administrative regulation of external human relations viewed as mediated relations). 30 External purposes, he maintains, are totally irrelevant for tort law or any other area of private law, the sole purpose of which is "not (as the instrumentalist thinks) to promote independently justifiable goals, but simply to be tort law [or private law] and therefore to develop itself in conformity with its own intelligible nature." 31 For public law, Weinrib acknowledges, external political purposes must be invoked to select the specific distributive (or mediating) criterion for any particular distribution. However, he asserts that the political aspect of a particular distribution is distinct from its legal aspect, enforceable by judges, which is the requirement that the distribution conform to and be a full development of the immanent rationality constitutive of distributive justice. 32
[*634] Thus, Weinrib's attack on instrumentalism has broadened considerably and reflects a major shift in his definition of instrumentalism. While Weinrib initially equated instrumentalism with aggregative conceptions of the ethical principles that were presumed to underlie tort law, his current definition equates instrumentalism with the claim that any area of law can or should be understood (as law) in terms of any external purpose -- political or ethical, aggregative or nonaggregative -- rather than solely in terms of the immanent rationality constituted by the internally coherent universal unity of its interdependent essential concepts. 33
Weinrib's exposition of the alleged immanent rationality of law proceeds, through a process of increasing abstraction and generalization, from specific areas of law (e.g., tort law), to the distinctions between private law and public law and their allegedly associated conceptions of human relations (immediate and mediated) and modes of legal ordering (corrective justice and distributive justice), and finally to the normativity allegedly immanent in the two types of justice. At each of these stages, Weinrib claims, one can discern an increasingly abstract and general immanent rationality that is exclusively constitutive of law. 34 In the following sections, I describe and criticize Weinrib's argument as it progresses through these various stages.
B. The Alleged Immanent Rationality of Tort Law
Weinrib's analysis in the initial stage proceeds in three steps: (1) identifying a specific area of law, (2) provisionally identifying the essential concepts (and associated doctrines and institutions) of that area of law, and (3) attempting to establish that these elements constitute, through their mutual interdependence, an internally coherent universal unity. 35 For Weinrib, these three steps are simply a retracing of the same circle in greater depth. He reasonably assumes that we share a common, albeit provisional, practical knowledge of distinct areas of law and their essential elements. 36 However, he attributes this common practical understanding to the immanent rationality of each distinct area of law, which is made more explicit as it is made more theoretical. 37
Weinrib's preferred starting point is tort law. Consistent with his theory of immanent rationality, Weinrib does not attempt any general definition or description of tort law. 38 Rather, he identifies tort law, as a [*635] distinct mode of legal ordering, with the essential features that constitute its basic structure or form -- that is, the features that are indispensable to our conception of tort law. 39 Weinrib's list of these essential features has narrowed over time. Initially, he included the defendant's causation of the plaintiff's injury, a bipolar adjudicative remedial procedure in which the plaintiff sues the defendant and -- if successful -- obtains his or her remedy from the defendant, wrongful (faulty) conduct by the defendant, and a harm to the plaintiff that is within the original wrongful risk. 40 However, in his more recent writings, Weinrib limits his list of essential elements to the first two elements, since the last two are not essential to our conception of tort law, but rather are the subjects of long-standing disputes within tort law. 41
The third step of Weinrib's analysis involves demonstrating that these essential elements exhibit or constitute an immanent rationality or intrinsic ordering. In an intrinsic ordering, each element is conceptually dependent on the others, "unintelligible in isolation," and "intelligible only through the integrated whole that they form as an ensemble." 42 Thus, each element "presupposes the relevance of the rest" and "is implicated in every other constituent," so that "the whole of tort law's basic structure is implicit in any individual feature." 43 Some apparent examples (not mentioned by Weinrib) of such "intrinsic orderings" are the mutually implied concepts of wife and husband, or the interdependent concepts of child, mother, and father. Weinrib asserts that the concepts of causation and a bipolar remedial procedure are similarly interdependent and constitute the distinct relationship or ordering that we call tort law. 44 This assertion is implausible on its face.
The concept of causation is the concept of some antecedent condition having an effect on some subsequent condition or, in more technical terms, the concept of the antecedent condition's being a NESS condition: a Necessary Element of a Set of actual antecedent conditions that was Sufficient for the occurrence of the subsequent condition. 45 Even when the concept of causation is restricted, as by Weinrib, to situations in which the antecedent condition is someone's conduct and the subsequent condition is [*636] an adverse effect on someone else's person or property ("the defendant's doing and the plaintiff's suffering the same harm"), 46 the concept of causation is perfectly intelligible by itself; it neither depends on nor implies any legal claim or remedial procedure, much less a bipolar adjudicative remedial procedure.
Weinrib might respond that, by intelligibility, he means intelligibility as an element of a juridical relation -- that is, a relation denoting legal obligation. Whether this response would be consistent with his theory is unclear, since the point of his theory seems to be to arrive at the concept of legal obligation by elaborating the universal unity of some set of interdependent essential concepts, rather than by building in an assumption of legal obligation from the very beginning. 47 Yet, setting this problem aside, there still would be no conceptual connection between the concept of causation and the concept of a bipolar remedial procedure. The concept of a legal remedial procedure would simply be built into the initial qualified definition of intelligibility -- intelligibility as an element of a juridical relation -- rather than being dependent on or implied by the concept of causation.
Furthermore, the concept of causation does not imply that the independently assumed legal remedial procedure must be a bipolar (or an adjudicative) remedial procedure. Causes and effects do not occur as discrete bipolar pairs, but rather as nodes in a vast network of causes and effects, with each node being the effect of an ever-expanding regression of prior nodes and itself being a cause of an ever-expanding progression of subsequent nodes. Each suffering of harm is the result of a multitude of doings; each doing is the cause of a multitude of sufferings of harm. The concept of causation, even when limited to the doing and suffering of the same harm, cannot by itself identify a discrete pair consisting of a single doing and a single suffering.
Conversely, the concept of a bipolar adjudicative remedial procedure, in which the plaintiff sues the defendant and, if successful, obtains his or her remedy from the defendant, neither depends on nor implies the concept of causation. On the one hand, as noted above, the concept of causation by itself cannot delimit a specific bipolar plaintiff-defendant pair. On the other hand, many other concepts or combinations of concepts can delimit such a pair: the richest and the poorest, the tallest and the smallest, the oldest and the youngest, the smallest and the poorest, and so forth. The concept of a legal claim by a plaintiff against a defendant which, if successful, results in the defendant's providing some remedy to the plaintiff, does not in itself imply or exclude any specific substantive basis for the claim. In particular, it does not -- simply as a concept -- require or imply [*637] that the defendant must have caused an injury to the plaintiff. There is nothing in the concept of a claim, per se, which specifies any particular basis for the claim.
Moreover, contrary to Weinrib's assertions, neither the concept of causation nor the concept of a bipolar remedial procedure implies the particular bipolar remedial procedure typically (but not exclusively) followed in tort law -- the defendant's payment of damages to the plaintiff in an amount equal to the harm that the defendant tortiously caused to the plaintiff. 48 Why not, instead, have a fine paid by the defendant to the plaintiff, the plaintiff's imprisonment of the defendant, an apology from the defendant to the plaintiff, or even the plaintiff's payment to the defendant of damages, a fine, loss of liberty, and an apology? None of these remedies is excluded or disfavored as a matter of implication from the concepts of causation or a bipolar remedial procedure.
Weinrib himself has pointed out the open-ended nature of causation and its consequent inability, by itself, to delimit a discrete defendant-plaintiff pair. He has argued forcefully against proponents of liability based merely on causation, such as Richard Epstein, that some additional concepts -- such as wrongful (faulty) conduct by the defendant and harm to the plaintiff that is within the original wrongful risk -- are necessary to limit the infinite regression and progression of causes. 49 Yet, as noted above, these latter two concepts cannot be employed by Weinrib as essential elements of the internal intelligibility of tort law, since they are the subjects of longstanding disputes within tort law, and tort liability often exists in the absence of either or both. Thus, Weinrib is forced by his overall argument into the unhappy position of having to assert what he himself has previously persuasively refuted: that causation and a bipolar remedial procedure together (or each alone, since each allegedly implies the other) are sufficient to capture the internal intelligibility, or essence, of tort law as a distinct mode of legal ordering. 50
Weinrib attempts to paper over this problem. He claims that wrongdoing or fault, harm-within-the-wrongful-risk, and other liability-limiting concepts are, along with causation and a bipolar remedial procedure, interdependent and mutually implied constituents of the intrinsic ordering or immanent rationality of tort law, even though they are not essential to our conception of tort law. 51 This claim involves a logical contradiction. If these concepts are all part of the intrinsic ordering of tort law, so that each is dependent on and implies all the others, then none can exist without each of the others. In particular, the concepts of causation and a bipolar remedial procedure cannot exist in the absence of the other concepts. Thus, if, as Weinrib assumes, the concepts of causation and a bipolar remedial procedure are essential to our conception of tort law, then the other concepts must also be essential to that conception. But this contradicts the [*638] second part of Weinrib's claim. Moreover, if fault and harm-within-the-wrongful-risk are indeed part of the inner coherence of tort law, then tort law is and always has been very incoherent, since there have always been significant areas of liability that dispense with these two requirements.
In any event, none of these concepts are mutually dependent. Contrary to Weinrib's repeated assertions, 52 there is nothing in the concept of causation or a bipolar remedial procedure per se that implies the concepts of volitional conduct, misfeasance, fault, reasonable care, duty, or proximity. Insofar as "doing harm" implies volitional conduct, the "implication" does not come from the concept of causation or a bipolar remedial procedure, but rather from a definition or interpretation of "doing" that assumes volitional conduct. 53 Even so interpreted, "doing harm" implies nothing more than volitional conduct, harm, and a causal connection between the volitional conduct and the harm. 54 There is no conceptual basis for a further implication of misfeasance, fault, lack of reasonable care, breach of duty, or proximity limitations on liability. Weinrib is correct to note that some such limitations are necessary to cut off the infinite regression and progression of causation and to justify imposing legal liability on the doer of the harm. 55 But these limitations are not explicit or implicit in the complex concept of "the defendant's doing and the plaintiff's suffering the same harm." All of Weinrib's assertions to the contrary are unsupported (and insupportable) ipse dixit.
More fundamentally, even the linkage of the defendant's volitional conduct ("doing") and the harm to the plaintiff ("suffering") through causation in the phrase "the defendant's doing and the plaintiff's suffering the same harm" is merely a matter of the conjunction of concepts, rather than their mutual implication. The concept of volitional conduct ("doing") does not conceptually imply any consequent harm ("suffering") to oneself or others. Conversely, "suffering harm" does not per se imply a volitional human cause of the suffering. The concept of "doing harm" does imply a causal connection between the volitional conduct and the harm, but only as [*639] a matter of the conjunction of the concepts of "doing" and "harm," rather than through any implication from the separately intelligible concepts of volitional conduct and harm. Thus, the requirement in tort law that there be a causal connection between the (tortious) volitional conduct of the defendant and the harm the plaintiff has suffered cannot be based on conceptual implication, but rather must be based on normative arguments external to these concepts. Once again, Weinrib's assertions to the contrary are mere ipse dixit. 56
Finally, even if Weinrib had demonstrated the conceptual interdependence of causation and a bipolar remedial procedure or, more encompassingly, the conceptual interdependence of those two concepts and the concepts of "wrongful" conduct by the defendant and harm-within-the-wrongful-risk restrictions on the plaintiff's injury, he still would not have identified or explained tort law as an internally intelligible area of law distinct from other areas of private law (e.g., contract law) or even some areas of public law (e.g., criminal law). As Weinrib himself notes, these other areas of law also contain all of these elements. For example, in contract law the plaintiff must prove that the defendant's "wrong" (breach of contract) caused injury to the plaintiff that was within the foreseeable risks, and the plaintiff, if successful, obtains damages or equitable relief from the defendant through a bipolar adjudicative remedial procedure. 57 Similarly, in one commonly accepted view of criminal law, adopted by Weinrib, the criminal defendant's willful rejection of the rules of social order is wrongful conduct that causes an injury to the dignity and security of each and every member of the state (or to "the state itself"). The plaintiff-victims in a criminal action are all the members of the state (or "the state itself"), represented by the prosecutor, who, if successful, obtains a remedy (a fine or loss of liberty) from the defendant, once again through a bipolar adjudicative remedial procedure. 58
The claim that the various areas of private law all display the same inner coherence is essential to the next stage of Weinrib's argument: that this inner coherence is the form of corrective justice, which unifies all of private law. 59 But if these different areas of private law, and even some areas of public law (e.g., criminal law), all possess the same inner coherence, how can any of them be identified as distinct modes of legal ordering? More specifically, how can tort law "be itself" rather than being, for example, contract law, criminal law, or simply corrective justice? 60 It is not open to Weinrib to declare that although each of these areas of law has the same inner coherence, it also has additional distinguishing features. Under his theory of immanent rationality, any distinguishing feature of an area of law must be part of the inner coherence of that area of law. When Weinrib attempts to unify the various areas of private law through the alleged [*640] immanent rationality of corrective justice, he annihilates the distinctions between those areas of law rather than explaining or illuminating them.
C. Private Law and Public Law and Their Allegedly Immanent Conceptions of Human Relations (Immediate and Mediated) and Modes of Legal Ordering (Corrective Justice and Distributive Justice)
For Weinrib, private law is the adjudication of private disputes and public law is the judicially enforceable limitations on the legislative and administrative direction of the community. He argues that all areas of private law have the same inner coherence, which is constituted by the mutually dependent and implied concepts of causation and a bipolar adjudicative remedial procedure or, more encompassingly, those two concepts as well as the concepts of "wrongdoing" and proximity. Viewed most abstractly and comprehensively, he claims, this inner coherence takes the form of Aristotle's corrective justice, which is the allegedly immanent rationality of external human relations understood as immediate interactions: the (wrongful) doing and suffering of the same harm (within the wrongful risk). Public law, on the other hand, deals with external human relations understood as mediated relations -- those in which the external relations consist of participation in some distribution of goods (broadly construed) among individuals in the society according to some mediating criterion, without regard to any immediate interaction. The allegedly immanent rationality of these mediated relations takes the form of Aristotle's distributive justice. 61
As was noted in the last section of this Article, however, the asserted identity of private law with immediate interactions and corrective justice, and public law with mediated relations and distributive justice, does not hold up. Criminal law, which is usually considered to be a part of public rather than private law, 62 is (as Weinrib himself notes) an expression of corrective justice rather than distributive justice. 63 Many other areas of [*641] public law, including much of the legislative and administrative regulation of safety and pollution, are best understood as prophylactic corrective justice rather than distributive justice. Conversely, some areas of private law or adjudication are best understood as being motivated by distributive justice considerations rather than by corrective justice -- particularly many, if not most, of the doctrines of property law, such as the adverse possession doctrine, the rule against perpetuities and other restrictions on future interests and restraints on alienation, nonwaivable warranties of habitability, and restrictions on servitudes. 64
More importantly, the concepts that are employed in Aristotle's elaborations of corrective justice and distributive justice, respectively, are not conceptually interdependent and thus do not exhibit or constitute any immanent rationality. Aristotle describes justice in its particular sense as the virtue of behaving equably in our external relations with others that involve claims to goods or advantages, broadly conceived. He distinguishes corrective justice from distributive justice by the types of relations to which each applies and by the type of equality required by each. Corrective justice applies to immediate interactions ("transactions") and requires that the parties to an interaction be treated as "arithmetically" or absolutely equal with respect to their entitlements to their holdings, regardless of their relative virtue, wealth, need, or other proposed measure of merit. Distributive justice, on the other hand, applies to mediated relations (distributions) and requires, with respect to their shares in a distribution, that the parties to the distribution be treated as proportionally equal in accord with their relative merit. 65
As Weinrib seems to acknowledge, 66 Aristotle's description of the equality in corrective justice as an "arithmetical proportion" (average or mean) is a metaphorical heuristic device that is used to contrast the absolute equality of the parties in corrective justice with their relative or comparative equality in distributive justice. Literally interpreted, the arithmetic-mean concept of equality would require that after an interaction there be an equality of goods among the parties to the interaction -- that is, that each party have, after the correction or adjustment, the average of the parties' unadjusted post-interaction goods. What Aristotle actually has in mind, however, is the absolute equality of the parties to the interaction, which includes an equality of entitlement to their respective pre-interaction goods. Any gain or loss of goods in an interaction that conflicts with the [*642] interacting parties' absolute equality must be rectified by the responsible party. This result will be achieved through the application of an arithmetic mean to the parties' unadjusted post-interaction goods only if, as Aristotle assumes in his heuristic, their pre-interaction goods were equal and one party's unjust gain from the interaction was equal to the other party's unjust loss. 67
We have already discussed, in the previous section, the lack of conceptual interdependence for the concepts that are conjoined in the concept of immediate interaction as a juridical relation (causation, a bipolar adjudicative remedial procedure, "wrongdoing" by the defendant, and proximate harm to the plaintiff). 68 Similarly, the concepts that are conjoined in the concept of a mediated relation -- persons, goods, and some mediating criterion (by which entities of type X are distributed among entities of type Y, with no predetermined content -- e.g., goods, persons, or whatever -- for X or Y) 69 -- do not depend on or imply one another.
Most importantly, contrary to Weinrib's assertions, there is no conceptual interdependence between the concepts of immediate interaction and mediated relations (or any of their constituent concepts) on the one hand, and the concept of equality in general or the two particular types of equality that are described in Aristotle's elaboration of corrective justice and distributive justice on the other. 70 There is nothing in the concept of an immediate interaction per se that implies or depends on any concept of equality, much less the "arithmetical" or absolute type of equality that Aristotle identifies as the core of corrective justice. An immediate interaction ("the doing and suffering of the same harm") is a descriptive concept that applies to any situation in which the volitional conduct of one party causes an adverse effect on the person or property of another party, regardless of whether the two parties are considered to be equal in any sense. Even if the concept is restricted to those situations in which the volitional conduct is "wrongful" and the adverse effect is proximately related to the original "wrongful" risk (thereby importing, as a matter of definition rather than by conceptual implication, a normative element into the concept of an immediate interaction), the word "wrongful" is merely a place holder for a particular conception of wrongfulness, which may or may not include the idea that the parties to the interaction are equal in some sense.
Similarly, there is nothing in the concept of a mediated relation per se that implies or depends on any concept of equality, much less the proportional type of equality that Aristotle identifies as the core of distributive justice. Contrary to Weinrib's declarations, a mediated relation, understood as the distribution of goods among some group according to some mediating criterion, need not employ Aristotle's notion of proportional equality (whereby persons' shares in the distribution are in the same geometrical proportion as their respective merit) or any other notion of [*643] equality to be intelligible as a distribution (i.e., a mediated relation) rather than a "random" or "haphazard" dispersion. 71 Any mediating criterion will do -- for example, one which ranks each person not equally in accord with merit, but rather exponentially in accord with merit (or any other personal characteristic, such as the number of freckles on the person's body) or unequally in accord with some completely impersonal criterion (such as the prime number drawn by each person from a box containing as many successive prime numbers as there are distributees). 72
Thus, contrary to Weinrib's claims, it is not true that corrective justice and distributive justice are conceptually comprehensive -- that is, that they exhaust the possible modes of ordering of external human relations. 73 Although the concepts of immediate interaction and mediated relations may well be conceptually comprehensive, corrective justice's "arithmetical" or absolute equality and distributive justice's proportional or relative equality are each only one of many conceptual possibilities for ordering immediate interactions and mediated relations.
Nor is it true, as Weinrib asserts, that the concepts of "arithmetical" equality and proportional equality necessarily imply, by their very form as mathematical concepts, the fundamental juridical distinction between the bipolarity of corrective justice and the multipolarity of distributive justice. Weinrib contends that an "arithmetical" or quantitative equality can only exist between two terms, whereas a geometrical proportion can extend to any number of terms. 74 Yet a literal arithmetical "proportion" (mean or average) can encompass any number of terms, while a geometrical proportion (ratio) can encompass only two. Furthermore, an equality of arithmetical proportions, like an equality of geometrical proportions, can extend to any number of terms linked by equal signs. 75 If we shift from the literal interpretation of "arithmetical proportion" to the simple notion of quantitative equality, we find that a quantitative equality between terms on either side of an equal sign, like an equality of ratios, again can be extended to any number of terms linked by equal signs. Moreover, unlike an equality of ratios, a quantitative equality can be maintained while adding another term to each side of the equal sign, splitting up the terms on one or both sides of the equal sign, or adding an amount to one term and subtracting the [*644] identical amount from a different term on the same side of the equal sign. 76
In sum, the concepts of "arithmetical" and "geometrical" equality on the one hand, and the concepts of immediate interaction and mediated relations on the other, are not conceptually interdependent. The linking of these concepts in the two kinds of substantive justice is not attributable to any immanent rationality, but rather must be based on normative premises that are external to these concepts.
D. The Kantian Normativity of Corrective Justice and Distributive Justice
Weinrib agrees that normative premises are implicated in Aristotle's linking of the concepts of "arithmetical" and "geometrical" equality with the concepts of immediate and mediated relations in corrective justice and distributive justice, respectively. However, Weinrib insists that the normative premises, rather than being external grounds for the linking of these concepts, are intrinsic to each of them. More particularly, he claims that, intrinsic to the concepts of immediate interaction and mediated relations, there is a normativity that is reflected in corrective justice's and distributive justice's respective concepts of equality, which in turn presuppose or prefigure the notion of self-determining agency subsequently developed in the Kantian concept of Right. 77
Weinrib declares that to be understood as intrinsically normative or juridical rather than as merely natural or descriptive phenomena, interactions and distributions must be conceived as having an internally presupposed normativity. 78 But this declaration merely restates the fundamental problem that Weinrib faces. We have already noted, in the previous section, that the concepts of immediate interaction (the doing and suffering of harm) and mediated relations (the distribution of goods according to some distributive criterion) do not per se contain any normative aspect or implication, but rather are perfectly intelligible as purely descriptive concepts. 79 Although Weinrib often seems to assume otherwise, the mere need (in his theory) for an immanent normativity does not justify an assumption that such immanent normativity exists.
Moreover, even if we assume that the concepts of interaction and distribution have an intrinsic normativity, from a purely conceptual standpoint there is no reason to infer that the intrinsic normativity incorporates [*645] the concept of equality, much less the particular concepts of absolute and proportional equality that are linked with the concepts of interaction and distribution, respectively, in corrective justice and distributive justice. The normativity reflected in these linkages is not intrinsic to any of the linked concepts, but rather must truly be presupposed as a normativity that is external to the concepts and provides the ground for their linkage.
Weinrib employs both of these unsupported assumptions -- the assumption of an intrinsic normativity and the assumption that the intrinsic normativity incorporates the concept of the absolute equality of the interacting parties -- in his discussions of the allegedly immanent normativity of corrective justice. He notes that since the normativity is (presumed to be) intrinsic, it must have the same unmediated form as the interaction itself -- that is, it must not employ any comparative criteria. He further assumes that the exclusion of comparative criteria encompasses potential as well as actual comparative criteria, and therefore excludes all internal or personal attributes of the parties to the interaction. Thus, according to Weinrib, the interaction's normativity must refer solely to its external aspects -- the external impingement of one person's activity on the person or property of another -- and has no connection to virtue or ethics. Weinrib claims that this completely external focus, when coupled with the presumed absolute equality of the interacting parties, presupposes the equality of the parties as self-determining agents that is the foundation of the Kantian concept of Right, which allegedly also incorporates a purely external conception of the normativity of human interaction. 80
There are serious problems at every step of this argument. First, there is the unsupported assumption of intrinsic normativity, which provides the basis for the exclusion of comparative criteria. Second, there is the claim that the exclusion of comparative criteria encompasses not only actual interpersonal comparisons, but also any reference to the internal or personal attributes of the interacting parties, which might be used for comparative purposes (even if they are not so used). This claim is based on the supposed conceptual implications of the formal distinction between unmediated and mediated relations:
The interests motivating the parties and the other factors constituting the material conditions of the interaction cannot be decisive for the nature of the interaction as such. If regarded as particular to each party, these factors cannot capture the dimension of interaction in which both actors participate; and if looked upon as shared characteristics, they do not exclude the wider set of persons, among whom they are diffused, and they therefore cannot mark the mutual impingement of two parties in a single transaction. Thus whether my pressing needs are thought of as peculiar to me or shared with others similarly circumstanced, they do not render intelligible a specific claim by me against you. 81
[*646] But if all internal or personal attributes are irrelevant to the normativity of the interaction, how can criminal law consider, as it does, the defendant's volition, state of mind, excuses, and justifications? 82 How can tort law consider, as it does, the defendant's volition, intent, physical capacities, knowledge, special expertise, and necessity-based justifications, or the plaintiff's knowledge, physical and mental capacities, and consent? How can either area of law consider the particular pre-interaction and post-interaction holdings of an interacting party or the difference between them that is claimed to be an injury? Indeed, how can any particular person be identified as one of the interacting parties? Although internal or personal attributes considered by themselves do not "capture the dimension of interaction" and do not exclude the wider set of persons who have the same attributes, neither does Weinrib's presupposed normative keystone -- the capacity to transcend one's particular circumstances that, when developed, constitutes self-determining agency -- which, as Weinrib himself emphasizes, is shared by all persons. To capture the dimension of interaction, we must incorporate the concept of causation. However, as we have seen, the concept of causation by itself also fails to identify a specific plaintiff-defendant pair. 83 To render intelligible a specific claim by one person against another, we must rely on a combination of causation and personal attributes.
In the next step of his argument, Weinrib contends that the supposed necessity of ignoring, or abstracting from, all internal or personal attributes of the interacting parties leaves the capacity of the parties as self-determining agents as the only conceptually possible ground of the equality that is presumed to be intrinsic in the interaction. 84 Yet the capacity for self-determination is itself an internal personal attribute. 85 Moreover, there are numerous alternative conceptions of the presumed immanent equality: for example, the capacity for volitional conduct without self-determination, the capacity to create risks (reciprocal or otherwise) regardless of volition, or, as argued by Weinrib himself, the mere equality of each party's holdings before and after the interaction:
Corrective justice looks to the transaction in which one party has impinged upon another and focuses upon the impingement alone and not on the attributes of the impinging parties. The structure of justification in corrective justice accordingly leaves no room for reference to persons, and looks only to the holdings which have been implicated in the transaction. Since only the impingement is relevant, corrective justice consists in treating these implicated holdings equally and in ignoring all other personal inequalities of [*647] wealth or virtue. Corrective justice thus operates with reference to the equality between thing and thing rather than to a proportion between thing and person. 86
Weinrib construes corrective justice as "an extreme version of interpersonal externality" 87 that "positively excludes consideration of the internal qualities [of the interacting parties]" 88 so that he can equate it with Kant's doctrine of Right, which allegedly also views human interaction from a completely external, abstract perspective and which allegedly is completely distinct from, and conceptually and normatively prior to, Kant's ethical theory. 89 However, Weinrib's interpretation of Kant's moral philosophy turns it upside down. Kant's elaboration of the doctrine of Right presupposes and depends on, rather than being completely independent from and prior to, his ethical theory.
Weinrib begins with a conception of law as a self-contained and autonomous enterprise and attempts to derive an allegedly immanent Kantian morality. More particularly, Weinrib begins with legal concepts such as causation and wrongdoing. He then attempts, through a process of increasing abstraction and generalization, to discern as inherent in such concepts the normative premise of the absolute and equal moral worth of humans as free rational beings. Kant, however, takes precisely the opposite approach. Kant begins with the normative premise of the absolute and equal moral worth of humans as free rational beings. He then develops from that premise his doctrines of Right and virtue and, as part of the doctrine of Right, his argument for the necessity of moving from private Right in the state of nature to public Right in the civil society.
The foundation of Kant's moral philosophy is the idea of free will or freedom, which is the capacity of a rational being to avoid determination by sensible inclinations by subjecting the maxim of any action to the condition of qualifying as universal law. The idea of freedom does not imply completely unrestricted self-determination, but rather self-legislation: self-determination in accordance with universal law. Moral behavior consists in overcoming, through subjecting the maxim of one's actions to the condition of qualifying as universal law, sensible inclinations that are in opposition to the dictates of the moral law. 90 The supreme principle of morality (the [*648] categorical imperative) is "Act on a maxim that can also hold as a universal law." 91
Freedom, and the moral personality constituted by its possession, is an inherently internal attribute of each rational being. The possession of free will or freedom is what gives each rational being moral worth -- an absolute moral worth that is equal for all rational beings:
Man regarded as a person [rather than a mere animal], that is, as the subject of a morally practical reason, is exalted above any price; for as a person (homo noumenon) he is not to be valued merely as a means to the ends of others or even to his own ends, but as an end in himself, that is, he possesses a dignity (an absolute inner worth) by which he exacts respect for himself from all other rational beings in the world. He can measure himself with every other being of this kind and value himself on a footing of equality with them. 92
In the elaboration of his moral philosophy, Kant distinguishes between a doctrine of Right and a doctrine of virtue. The doctrine of Right focuses on the external aspect of the exercise of freedom -- the constraints on external action required for the practical operation of freedom in the external world. The doctrine of virtue, on the other hand, focuses on the internal aspect of the exercise of freedom -- a person's subjecting the maxim of her actions to the condition of qualifying as universal law. 93 The distinction between the external and internal aspects of the exercise of freedom (which itself is inherently internal) explains the differences between the supreme principle of Right and the supreme principle of virtue, each of which is a corollary of the categorical imperative. The supreme principle of Right is "So act externally that the free use of your choice can coexist with the freedom of everyone in accordance with a universal law . . . ." 94 The supreme principle of virtue is "Act in accordance with a maxim of ends that it can be a universal law for everyone to have." 95
The concept of Right is the authorization to obligate another through external coercion in accordance with a universal law of freedom. 96 The concept follows, Kant notes, from the idea of freedom:
Resistance that counteracts the hindering of an effect promotes this effect and is consistent with it. . . . Therefore, if a certain use of freedom is itself a hindrance to freedom in accordance with universal laws (i.e., wrong), coercion that is opposed to this (as a hindering of a hindrance to freedom) is consistent with freedom in accordance with universal laws, that is, it is right. Hence there is [*649] connected with Right by the principle of contradiction an authorization to coerce someone who infringes upon it. 97
Right, being the authorization to obligate another through external coercion, can only affect -- and hence can only apply to -- the external aspect of the exercise of freedom. The internal (ethical or virtuous) aspect of the exercise of freedom -- a person's subjecting the maxim of her actions to the condition of qualifying as universal law -- cannot be coerced by another. Although a person may be coerced into behaving externally so as to further or hinder some end, she cannot be coerced into adopting or rejecting that end as her own. 98
The notion that there is a radical disjunction in Kant's moral philosophy between Right and ethics stems from Kant's insistence on sharply distinguishing duties of Right per se (strict Right) from ethical obligations and, more particularly, from duties of virtue. 99 Duties of Right are those for which an external incentive to comply ("external lawgiving") is possible, while duties of virtue are those for which external lawgiving is not possible. The distinguishing feature of a duty of Right (what Kant calls "strict Right, namely that which is not mingled with anything ethical") 100 is the external incentive for compliance: the potential coercion by another in accordance with a universal law of freedom. The distinguishing feature of an ethical obligation, on the other hand, is the internal incentive of the consciousness of the duty itself. Every duty of Right, being a law of reason, is also an ethical obligation -- that is, a duty that one should comply with due to the consciousness of the duty itself, even if the external incentive is lacking or ineffective. However, Kant emphasizes, the internal incentive of the ethical obligation should not be confused with the external incentive that peculiarly distinguishes a duty of Right per se (strict Right). 101
Moreover, under Kant's classification scheme, not every ethical obligation is a duty of virtue. Duties of virtue, as distinguished from ethical obligations in general, are duties to adopt certain ends. 102 Duties of Right are assumed not to be concerned with ends per se but rather with restrictions on external action in accordance with a universal law of freedom. 103 Thus, the ethical obligation correlative to a duty of Right -- "respect for Right" 104 -- is not a "duty of virtue" in the strict sense, but rather an "obligation of virtue" in the broad sense:
With regard to the distinction between the matter and the form in the principle of duty (between conformity with ends and conformity with law), it should be noted that not every obligation of virtue (obligatio ethica) is a duty of virtue (officium ethicum s. virtutis); [*650] in other words, respect for the law as such does not yet establish an end as a duty, and only such an end is a duty of virtue. Hence there is only one obligation of virtue, whereas there are many duties of virtue; for there are indeed many objects that it is also our duty to have as ends, but there is only one virtuous disposition, the subjective determining ground to fulfill one's duty, which extends to duties of Right as well, although they cannot, because of this, be called duties of virtue. 105
By classifying the ethical obligation correlative to a duty of Right as an "obligation of virtue" rather than a "duty of virtue," Kant is able to treat duties of Right and duties of virtue as mutually exclusive, not only in terms of the incentive for compliance but also in terms of the substance of the duty. A duty of Right is one for which external lawgiving (as well as internal lawgiving) is possible, while a duty of virtue -- in the strict sense -- is a solely ethical duty: a duty for which no external lawgiving is possible. 106
[*651] Although Kant sharply distinguishes duties of Right from duties of virtue, he does not treat Right as being conceptually or normatively prior to ethics or virtue. Rather, the reverse is true. As was stated above, the foundation of Kant's moral philosophy, which encompasses both his doctrine of Right and his doctrine of virtue, is the idea of free will or freedom: the internal capacity to avoid determination by sensible inclinations by subjecting the maxim of one's actions to the condition of qualifying as universal law. 107 Ethics and virtue focus directly on this internal capacity -- the core internal aspect of the exercise of freedom -- while Right focuses on its derivative external aspect. Without the internal aspect the external aspect would be meaningless, morally and otherwise. An entity without internal freedom is not a moral being, whereas a person deprived of all or almost all external freedom -- e.g., a prisoner or slave -- still has internal freedom and hence is a moral being. 108 As Kant states, the mere conformity of an action with Right establishes its legality, but not its morality. Compliance with a duty of Right as a result of external coercion is not an exercise of one's freedom, but rather a mere aversive animal response to externally generated sensible impulses. 109
Thus, although the concept of Right focuses on the external aspect of the exercise of freedom, it presupposes and is dependent on the internal aspect, which is the foundation of Kant's moral philosophy and is the subject matter of ethics or virtue. 110 Weinrib himself acknowledges this when he identifies the Kantian idea of internal freedom or moral personality as the basic normativity that allegedly is immanent in his formalist account of law and justice. 111 In Kant's moral philosophy Right is conceptually and normatively dependent on the morality that grounds ethics, rather than the other way around. Duties of Right are prior to duties of virtue only in the sense that, whether viewed as duties of strict Right or as ethical obligations, the strict duties encompassed by the doctrine of Right (e.g., repaying one's contractual debts) have priority over the broad duties encompassed by the doctrine of virtue (e.g., the duty of beneficence). 112
[*652] When Kant insists that duties of Right not be confused with ethical obligations (obligations or duties of virtue), he is not insisting on a radical separation between these two parts of morality. Rather, he is merely insisting on a firm understanding of the fundamental difference in the incentive for compliance that distinguishes duties of strict Right and law-abiding behavior from ethical obligations and virtuous behavior. The distinguishing feature of strict Right is the external incentive for compliance -- the authorized coercion by another consistent with the freedom of all according to universal laws. One who complies with some duty as a result of this external incentive can be described as upright or law-abiding, but not as moral or virtuous. Virtue is manifested only when the reason for compliance with the duty is not the external incentive of potential legal coercion, but rather the internal incentive constituted by the consciousness of the duty itself. 113
Similarly, when Kant states that Right "has as its object only what is external in actions" or is "completely external," 114 he does not mean to imply, as Weinrib infers, the irrelevance to Right of all internal or personal attributes (which, as noted above, would even include the internal capacity of self-legislating agency), or, more radically, all particularity. 115 He means only to emphasize that the incentive to conform to a duty of Right per se (strict Right) is wholly external and therefore can only affect the external aspect of the exercise of freedom, not its internal aspect. 116 Yet, in deciding what external aspects of the exercise of freedom must be prohibited or limited in conformity with the idea of freedom, it surely will be relevant to consider -- as the law indeed does -- many of the particular, even internal, attributes of the interacting parties that affect the external aspects. 117 Indeed, Kant himself notes that, in imputing responsibility for consequences in both law and morals, we distinguish between intentional and unintentional transgressions of duty and, when a subjective assessment is appropriate, take into account the natural capacities and state of mind of the agent. 118
[*653] The dependence of Right on ethics is most evident in Kant's argument for the necessity of moving from private Right in the state of nature to public (juridical) Right in the civil society. Kant introduces and summarizes the argument, in a passage entitled "General Division of Duties of Right," through a restatement of Ulpian's three precepts:
(1) Be an honorable man (honeste vive). Rightful honor (honestas iuridica) consists in asserting one's worth as a man in relation to others, a duty expressed by the saying, "Do not make yourself a mere means for others but be at the same time an end for them." This duty will be explained later as an obligation resulting from the Right of humanity in our own person (Lex iusti).
(2) Do not wrong anyone (neminem laede) even if, to avoid doing so, you should have to stop associating with others and shun all society (Lex iuridica).
(3) (If you cannot help associating with others), enter into a society with them in which each can keep what is his (suum cuique tribue). If this last formula were translated "Give to each what is his," what it says would be absurd, since one cannot give anyone something he already has. In order to make sense it would have to read: "Enter a condition in which what belongs to each can be secured to him against everyone else" (Lex iustitiae).
So the above three classical formulas [of Ulpian] serve also as principles for dividing the system of duties of Right into internal duties, external duties, and duties that involve the derivation of the latter from the principle of the former by subsumption. 119
In a subsequent passage, entitled "Transition from What is Mine or Yours in a State of Nature to What is Mine or Yours in a Rightful Condition Generally," Kant describes these three divisions of the duties of Right (or justice 120 ), in the context of rights to objects, slightly differently:
With reference to either the possibility or the actuality or the necessity of possession of objects (the matter of choice) in accordance with laws, public justice can be divided into  protective justice (iustitia tutatrix),  justice in men's acquiring from one another (iustitia commutativa) and  distributive justice (iustitia distributiva). In these the law says, first, merely what conduct is intrinsically right [recht] in terms of its form (lex iusti); second, what [objects] are capable of being covered externally by law, in terms of their matter, that is, what way of being in possession is rightful [rechtlich] (lex iuridica); third, what is the decision of a court in a particular case in accordance with the given law under which it falls, that is what is laid down as right [Rechtens] (lex iustitiae). 121
Weinrib interprets these three divisions as three stages in the progressive externalization of free will and practical reason, which does not attain [*654] the complete externalization that constitutes Right until the third stage. According to Weinrib, the first stage (honeste vive, lex iusti, protective justice) focuses on a single actor, does not yet involve any interaction, and does not incorporate any internal or self-directed perspective. Rather, it merely represents the possibility of interaction with others, the possibility of possession of objects, and the need to actively assert oneself in the external world. The second stage (neminem laede, lex iuridica, commutative justice) adds a second actor and an actual interaction, but still is not sufficiently externalized to constitute Right since each actor must judge for herself what is right and this is an impermissible internal element. Only in the third stage (suum cuique tribue, lex iustitiae, distributive justice), when a third actor -- the judge -- is added, is Right made fully explicit through complete externalization. 122
But Kant clearly has something else in mind, especially with respect to the first division. Kant explicitly states that each division is a division of the duties of Right (or justice), rather than a mere preliminary stage in the conceptual elaboration of Right. 123 The first and second divisions (protective justice and commutative justice) constitute private Right, while the third division (distributive justice) constitutes public Right. 124
The first division (protective justice) already presumes an interaction. It asserts the duty of a person to assert her moral worth in interactions with others by, among other things, resisting nonrightful coercion by those others. Viewed not as a duty but as a right, it encompasses what Kant identifies as the only innate Right that belongs originally to every person by virtue of her humanity: "Freedom (independence from being constrained by another's choice)." 125 Inherent in this innate Right is the authorization to use coercion against another to resist or prevent nonrightful aggression by that other against one's person or property. 126 Yet this right of "protective justice" exists only if the protective conduct is "intrinsically right in terms of its form" -- that is, only if the actor has subjectively determined that her use of coercion conforms with the principle of Right. 127 The innate Right of protective justice is essential to the possibility of possession of external things. Through an argument by contradiction, Kant infers as a postulate of practical reason the right to acquire external things through first possession. 128 However, in the state of nature, possession of external things is dependent on, and its extent is determined by, the would-be possessor's ability to control them by defending them against aggression by others. 129
[*655] The second division (commutative or corrective justice) focuses on the effects of a person's actions on others' established possessory rights in external things. Such actions must conform to those others' rights -- that is, they must be consistent in their external effects with the equal absolute moral worth of those others as free rational beings. Again, in the state of nature, the determination of whether one's actions conform to the principle of Right is necessarily internal and subjective.
Kant describes the third division (distributive justice or public Right) narrowly as the duty to enter into civil society, 130 in which the authorization to use coercion against others is transferred (with a few exceptions such as self-defense) from each person to public institutions. According to Kant, public Right as a purely rational concept "contains no further or other duties of men among themselves than can be conceived in the former state [of nature]; the matter of private Right is the same in both. The laws of the condition of public Right, accordingly, have to do only with the rightful form of men's association (constitution), in view of which these laws must necessarily be conceived as public." 131
It is the unilateral, subjective nature of private Right in the state of nature that grounds the third division of the duties of Right. Through the postulate of practical reason, persons in the state of nature have the right to acquire external things through first possession. 132 Yet the rightful possession thereby acquired is provisional rather than conclusive, since no person by her unilateral action can conclusively bind others. Absent the universal consent of all, no one has any better right than any other person to acquire any external thing, and the rightful limits of acquisition cannot [*656] be established. Nevertheless, under the postulate of practical reason, each person in the state of nature has the provisional right to acquire external things through first possession. 133 However, her acquisition will be (provisionally) rightful only if it is intrinsically right in terms of its form -- that is, only if it conforms with her concept of Right. To fully conform with the concept of Right, she must be willing to enter into the civil condition, which is the only condition in which rightful possession, and its limits, can be conclusively secured through the universal will of all. Thus, the necessity of possession of external things for the practical exercise of freedom, when coupled with the requirement that such possession conform with the principle of Right, necessitates entry into the civil condition. 134
Similarly, a person in the state of nature has the innate Right to protect her person and possessions against nonrightful aggression by others. Indeed, as noted above, possession of external things depends upon this protective justice right. 135 Again, her exercise of this protective justice right will be (provisionally) rightful only if it is intrinsically right in terms of its form, which requires at a minimum that she subjectively determine that her use of coercion conforms with the concept of Right. Yet, no matter how much good faith (virtuous respect for Right) she displays in making this determination, in exercising her protective justice right she will be unilaterally imposing her will on others, who may have different subjective concepts of Right, and thus her action will not conform with the principle of Right. In order for her exercise of her protective justice right in the state of nature to be provisionally rightful -- to be intrinsically right in terms of its form -- she must be willing to enter into the civil condition with others with whom she might come into conflict:
However well disposed and law-abiding men might be, . . . before a public lawful condition is established, individual men, peoples, and states can never be secure against violence from one another, since each has its own right to do what seems right and good to it and not to be dependent upon another's opinion about this. So, unless it wants to renounce any concepts of Right, the first thing it has to resolve upon is the principle that it must leave the state of nature, in which each follows its own judgment, unite itself with all others (with which it cannot avoid interacting), [*657] subject itself to a public lawful external coercion, and so enter into a condition in which what is to be recognized as belonging to it is determined by law and is allotted to it by adequate power (not its own but an external power); that is to say, it ought above all else to enter a civil condition. 136
Weinrib asserts that this passage, in which Kant argues the conceptual necessity of public lawful coercion "even if we imagine men to be ever so good-natured and [right-loving]," demonstrates "the conceptual priority of law over ethics." 137 However, the passage demonstrates the reverse. It bases the duty to enter into the condition of public Right (the civil society) on the internal ethical obligation to act in conformance with the concept of Right. A mutual decision by interacting persons to remain in the state of nature, with each person acting unilaterally in accord with his or her own subjective concept of Right, is wrong only from this internal ethical viewpoint:
Given the intention to be and to remain in this state of externally lawless freedom, [formally] men do one another no wrong at all when they feud among themselves; for what holds for one holds also in turn for the other, as if by mutual consent . . . . But in general [materially] they do wrong in the highest degree by wanting to be and to remain in a condition that is not rightful, that is, in which no one is assured of what is his against violence. 138
However, any person in the state of nature has the innate protective justice Right to compel others, with whom she might come into conflict, to enter into the civil condition with her, if they are not willing to enter voluntarily:
No one is bound to refrain from encroaching on what another possesses if the other gives him no equal assurance that he will observe the same restraint toward him. No one, therefore, need wait until he has learned by bitter experience of the other's contrary disposition; for what should bind him to wait till he has suffered a loss before he becomes prudent, when he can quite well perceive within himself the inclination of men generally to lord it [*658] over others as their master . . . ? 139
In sum, Kant's discussion of the three divisions of the duties of Right repeatedly emphasizes that an essential condition for the rightful use of coercion against another is that the person using the coercion must do so with the internal (virtuous) disposition of respect for Right. The use of coercion must be "intrinsically right in terms of its form." 140 Coercion is rightful only when it is employed to counter the nonrightful aggression of another -- that is, only as an instance of protective justice (the first division of the duties of Right). Otherwise, the use of coercion against another is a violation of corrective justice (the second division of the duties of Right), and the person being aggressed against can exercise his protective justice right to use coercion to counter the wrongful use of coercion by the aggressor. The duty to enter into civil society (the third division of the duties of Right), apart from any coercion by another to do so, is an ethical obligation based on full consideration of the implications of the ethical condition for the rightful use of coercion in protective justice. On the other hand, each person in the state of nature has an innate protective justice Right to compel others, with whom she might come into conflict, to enter into civil society with her. Thus, as Kant stated when first describing the three divisions of the duties of Right: "[The] three classical formulas [of Ulpian] serve also as principles for dividing the system of duties of Right into internal duties [assert your own worth: protective justice], external duties [do not wrong anyone: corrective justice], and duties that involve the derivation of the latter from the principle of the former by subsumption [enter into civil society: distributive justice]." 141
One puzzling issue remains, which up to this point I have tried to finesse. How can the first division of the duties of Right (assert your own worth: protective justice) actually be a duty of Right, enforceable through coercion by others, rather than a duty of virtue? Given such a duty, if a person failed to defend himself against nonrightful coercion by another, then others (who? the aggressor?) could use even further coercion to attempt to force him to defend himself against the original coercion. A duty of Right to resist coercion against oneself enforceable through further coercion seems self-defeating, and there is no indication that Kant had any such duty of Right in mind.
Thus, in my discussion I treated the first division as a division of Right -- the innate protective justice Right to use coercion to counter aggression by others -- rather than a duty of Right. The only obligations I discussed in connection with this Right were ethical obligations: the ethical obligation to assert one's own worth by resisting aggression by others, and the ethical obligation to use protective coercion with the virtuous disposition of respect for Right. 142 This treatment is consistent with Kant's discussion of private and public Right. It also is consistent with his description of the "duties of Right" in the first division as internal duties, as [*659] contrasted with the external duties in the second and third divisions. 143 In his earlier lectures on ethics, Kant was even more explicit. He described the first of Ulpian's precepts (honeste vive) as "an ordinary principle of ethics, for here the ground of the impulse to fulfil one's obligation is derived, not from compulsion, but from the inner impulsive ground," and distinguished it from the second and third, which are "principles of legal obligation, for they relate to compulsory duties." 144
A number of additional considerations support the conclusion that the duties in the first division are ethical obligations rather than duties of strict Right. Kant states that the duties in the first division are duties to oneself, based on an obligation from "the Right of humanity in our own person," rather than duties to others. 145 How, then, would anyone else have a right to enforce these duties? Elsewhere, Kant states that duties to oneself are "special" to ethics, as contrasted with duties that ethics "has in common with Right." 146 As was discussed above, Kant treats duties of Right and duties of virtue as mutually exclusive. 147 In his doctrine of virtue, Kant treats all duties to oneself as duties of virtue -- not only imperfect duties such as the duties to develop one's talents and to strive for moral perfection, but also perfect duties such as the duties not to murder or mutilate oneself, not to make unnatural use of one's sexual inclination, and not to lie. 148 In distinguishing the duty not to lie in Right from the same duty in ethics, Kant states: "In the doctrine of Right an intentional untruth is called a lie only if it violates another's right, but in ethics, where no authorization is derived from harmlessness, it is clear of itself that no intentional untruth in the expression of one's thoughts can refuse this harsh name." 149 Kant specifically treats the duty to respect the humanity in oneself, by not being servile or submissive, as a duty of virtue, stating in particular: "Do not let others tread with impunity on your rights." 150
Even if, as some argue, Kantian duties to oneself can be duties of Right, 151 such duties to oneself would have to be "perfect" (strict or narrow) [*660] duties -- duties for which the law can specify "precisely in what way one is to act and how much one is to do by the action" -- rather than "imperfect" (broad or wide) duties -- duties for which only the maxim of the action can be specified. If only the maxim of the action can be specified, rather than the particular required action, the duty cannot be externally enforced. Thus, all imperfect duties can only be duties of virtue; they cannot be duties of Right. 152 The duty to not let others tread with impunity on your rights would seem to be a duty for which only the maxim of the action, rather than any particular protective action, can be specified. Kant groups the duty not to be servile (which encompasses the duty not to let others tread with impunity on your rights) with the supposedly absolute duties not to lie or to be greedy in a chapter covering perfect duties to oneself as a moral being. However, several of the particular duties not to be servile, including, for example, the duty to be thrifty as well as the duty to not let others tread with impunity on your rights, would seem by their very nature to be imperfect rather than perfect duties. 153
I believe that the considerations addressed in the previous four paragraphs provide strong grounds for interpreting the first division of the "duties of Right" as encompassing the innate Right of protective justice and the duty of virtue to not let others tread with impunity on your rights, but not a duty of strict Right to defend oneself and one's property against aggression by others. 154 In any event, even if the first division is indeed a division of the duties of Right, it is one that explicitly relies on ethical [*661] consideration of what it means to be a human being, rather than on consideration of negative impacts on the external exercise of freedom by others. 155 Thus, whichever way it is interpreted, the first division of Kant's duties of Right, which is the foundation of his argument for the necessity of public Right (law in the civil society), is itself bottomed on Kant's broader moral and ethical theory.
In sum, contrary to Weinrib's interpretation, Kant views internal freedom, justice as a virtue (respect for Right), and ethics as preconditions for external freedom, justice as a juridical relation, and law, respectively. 156 In Kant's moral philosophy, it is the equal status of each person as a moral being with "absolute inner worth" that grounds and limits ethics, law, and politics (which Kant equates with the juridical order). 157 The normativity of interactions -- the requirement that each party to the interaction treat herself and every other party to the interaction as an end and not merely as a means -- is not immanent to the concept of interaction per se, but rather is imposed on interactions by the ethical concept of human beings as moral (self-legislating) beings. Thus, Weinrib has it exactly backwards when he states that, in corrective justice, "Each party is defined not by moral worth but by the interactional nexus of an injury inflicted and received" or that "the status of each litigant is entirely a reflection of the transaction that links [*662] him to the other." 158
Weinrib's discussion of the alleged immanent normativity of mediated relations fares no better. He notes that the concept of a mediated relation -- the distribution of goods among persons -- presupposes a distinction between persons and things. He contends that this distinction implies Kant's (ethical) view of a person as being an end and never only a means to an end, which allegedly is violated by a refusal to give a person her geometrically proportionate share in a distribution (thus supposedly making her "available for use according to the distributor's pleasure"). 159 On its face, this argument contradicts Weinrib's often-repeated insistence that the normativity implicit in law, including distributive justice as well as corrective justice, reflects the Kantian concept of Right -- of negative freedom from external interference rather than positive fulfillment of needs -- which allegedly is entirely independent from Kant's ethics. 160
Furthermore, it simply is not true that a conceptual distinction between persons and things implies Kant's ethics, his concept of Right, or the geometrical-proportion concept of equality in distributive justice. 161 At most, the conceptual distinction implies that only things, not persons, are subject to being owned or used by others, and thus that only things, not persons, are subject to distribution. It implies nothing about the distribution of things among persons, not even that the distribution be mediated by some uniform criterion rather than being random or arbitrary. As noted above, even if we add the concept of a mediating criterion, the mediating criterion need not employ any concept of equality, much less the geometrical-proportion concept of equality. 162 Finally, even if we add the geometrical-proportion concept of equality, the particular standard of merit used to calculate the proportions remains completely open conceptually, as Weinrib himself often states. 163 The standard might be maleness, or white skin color, or noble birth, or blood relationship to the dictator, or the number of freckles on one's body, or whatever. Yet none of these standards would reflect the equality of all persons as self-legislating agents that is the core of Kant's ethics and his concept of Right. 164
Moreover, Weinrib views the mediating criterion in distributive justice as being political and thus (in his view) instrumentally aimed at the attainment of some collective or aggregative goal (e.g., maximization of aggregate happiness or wealth). Under this view, the requirement in [*663] distributive justice that each person's share conform to the mediating criterion is not an imposition of noninstrumental constraints on the instrumentalism of politics, as Weinrib states, but rather a strict enforcement of such instrumentalism in direct contradiction of Kant's ethics and concept of Right. 165
Weinrib's attempt to unify all of law, including distributive justice as well as corrective justice, under the allegedly immanent normativity of Kant's concept of Right recreates, at the final stage of his theory, the same contradictions that appeared in the initial stage. 166 Under Weinrib's theory of immanent rationality, in which each element supposedly implies every other element, how can it be that "the single normative presupposition [i.e., Kant's concept of Right] that underlies immediate and mediated interactions has two different manifestations in positive law": distributive justice and corrective justice with their two different concepts of equality? 167 On the other hand, if, as Weinrib often states, there is no overarching form that connects or integrates distributive justice and corrective justice, 168 how can the totality of law -- including distributive justice as well as corrective justice -- be conceived as possessing or expressing the same immanent moral rationality of Kantian Right? Under Weinrib's formalist theory, the form and content of law are identical and are constituted by its immanent moral rationality. 169
Weinrib's abstract conceptualism reaches its zenith in his remarkable assertion that interactions and distributions are mere conceptual constructs that can be interchangeably imposed on any external human relationship:
Because the forms of justice are justificatory structures, their concern is with the coherence of the justifications for legal arrangements, not with the subject matter of those arrangements as brute facts. Accordingly, the effects that one person might have on another cannot be preclassified as belonging to one or the other form. For the formalist the crucial consideration is not what happened but how one is to understand the justificatory structure that is latent in the legal arrangements that might deal with what happened. My injuring you is in itself neither a transaction that calls for corrective justice nor a distribution that falls under distributive justice. . . .
Formalism, accordingly, is not a kind of jurisprudential federalism with different incidents assigned to the jurisdiction of either corrective or distributive authority. Nor does formalism provide a basis for preferring to treat the facts of the world in accordance with one form rather than the other; such preference can come [*664] neither from within either form nor from any overarching form. 170
How, then, as Weinrib states, is a judge "to apply, in the context of a particular episode of adjudication, the form of justice appropriate to it"? 171 According to Weinrib, neither form is inherently appropriate. How can a judge decide to apply tort law rather than Social Security law, except on the basis of instrumental political grounds that Weinrib claims to rule out? If the judge is not to make such decisions, it must be the legislature (or, least plausibly, the litigants, assuming they agree). But if that is so, the legislature's involvement in corrective justice will not be limited, as Weinrib claims, to "a political decision to establish the appropriate institutions of adjudication." 172 Instead, the legislature's involvement will extend to the determination of which episodes will be treated as interactions rather than as distributions, so that the vaunted independence of corrective justice from the instrumentalism of politics is a sham. Weinrib's agnosticism on the choice of form deprives his theory of any normative bite and undermines his attempt to defend his theory against the charge of detachment from the actual world. 173
The problems that plague Weinrib's theory at every step of its development all flow from his extreme formalism -- that is, from his insistence that the content and normativity of law are completely self-contained and independent from ethics, politics, or any external purpose. Much of what he has written, especially on the nonaggregative concepts of law elaborated by Aristotle and Kant, is extremely powerful and valuable when it is stripped of its formalist assumption. Yet the formalist assumption -- that law can and should be understood independently of ethics or politics or any external purpose -- deprives law of any content or structure.
Contrary to Weinrib's assertions, Kant rejected the formalist assumption. In Kant's philosophy, law and politics are derived from and grounded on ethics. As we will see in Part IV, Aristotle also rejected the formalist assumption: for him also, politics, law, and ethics are inextricably linked. We can hope that Weinrib himself, having taken Aristotle and Kant as his guides, will soon follow them in rejecting the formalist assumption, so that his considerable intellectual talents will be refocused on the principlist enterprise that he originally espoused: elaborating the Kantian ethical principles which underlie and give coherence to the law.
[*665] III. Coleman's De Facto Formalism
Unlike Weinrib, Coleman explicitly rejects the formalist assumption. He states that the content and structure of tort law cannot be explained or justified in purely analytic or conceptual terms, but rather depend on foundational normative arguments. 174 Yet his writings on tort law and corrective justice do not defend any foundational normative arguments. 175 Instead, he focuses almost exclusively on the conceptual implications of abstract concepts, such as "wrong," "wrongdoing," "gain," "loss," and "annulment" of "wrongful gains" and "wrongful losses." 176 Thus, despite his antiformalist insistence on the necessity of elaborating the normative foundations that support and shape law's conceptual superstructure, Coleman's published work displays a formalist preoccupation with the conceptual superstructure that, in practice, is even stronger than Weinrib's. Weinrib has devoted considerable effort to elaborating the normative presuppositions that he claims are intrinsic in the conceptual superstructure itself, 177 while Coleman has concentrated on abstract conceptual analysis to the virtual exclusion of any foundational normative arguments.
A. The Pure Annulment Thesis
Prior to his contribution to this symposium, Coleman had articulated an alleged conception of corrective justice which he now labels "the annulment thesis." 178 He claimed that corrective justice is simply the principle that "wrongful" gains and losses must be annulled or rectified. He also contended that the grounds of claims to recovery or recompense, by those who have suffered losses, are analytically distinct from the grounds for imposing liability. According to Coleman, claims to recovery or recompense are based on the suffering of a "wrongful" loss, while impositions of liability are based on the acquisition of "wrongful" gains, and "wrongful" gains can occur in the absence of "wrongful" loss and vice versa. Finally, Coleman asserted that there is a further analytic distinction between the grounds of recovery and liability on the one hand and the "modes of rectification" on the other. Even when there is both "wrongful" gain and "wrongful" loss, and even when the gain is equal to the loss, the rectification of the gain and the loss (as required by corrective justice) need not occur through the institutional form of tort law, whereby the injurer who "wrongfully" gained is liable to the victim who "wrongfully" lost. Instead, the rectification can occur through any number of bifurcated rectificatory schemes -- e.g., imposition of criminal liability on the injurer and compensation of the victim through a no-fault social insurance system. Thus, under the annulment thesis, corrective justice can ground some (but not all) of the [*666] claims to recovery recognized in tort law, but it cannot explain or justify the mode of rectification that is the central feature of tort law (the injurer's liability to the victim), which instead is grounded on efficiency or utility. 179
Although Coleman's elaboration of the annulment thesis was almost entirely conceptual, he did not take some of the principal concepts that he employed sufficiently seriously. This was particularly true for the concept of corrective justice itself, as Coleman now acknowledges. 180 Coleman assumed that the annulment thesis was a noncontroversial description of the minimum content of corrective justice, with the only debate being over the appropriate conception of wrongfulness and the appropriate mode of rectification. 181 Yet, since it's initial elaboration by Aristotle, the concept of corrective justice has been understood not as a requirement that all wrongful gains and losses somehow be annulled, independent of the particular mode of rectification, but rather as asserting, inter alia, that a person who causes an unjust loss to another person or obtains an unjust gain at the other's expense has a duty, owed to that other person, to rectify the unjust loss or gain, and that the other person has a correlative right to have the unjust loss or gain rectified by the person who caused it. 182
Coleman's annulment thesis has been welcomed, in whole or in part, by the opponents of corrective justice, such as the social-compensation theorists 183 and the efficiency theorists, 184 since it eviscerates the traditional [*667] Aristotelian concept, which is inconsistent with their theories. However, the annulment thesis has been rejected by virtually all other expositors of corrective justice. 185 Coleman has stated, and continues to state, that Richard Epstein's theory of strict liability begins with the annulment thesis -- that all unjust losses must somehow be rectified -- and then adds on the requirement that the duty of rectification be placed on the injurer. 186 However, as Coleman once acknowledged, Epstein's theory does not begin with a right to recovery "in the air," independent of any particular mode of rectification, but rather begins with the Aristotelian concept that there is a claim to recovery only against the person who caused the unjust loss, with a correlative duty of that person to repair the victim's loss. 187 Coleman also claims that George Fletcher's reciprocity theory tracks the annulment thesis by uncoupling the victim's right to recovery from the injurer's liability. 188 Yet, although Fletcher nominally distinguishes the victim's right to recovery from the injurer's liability, he actually merely distinguishes the prima facie case (harm caused by nonreciprocal risk creation) from a defense of excuse. The injurer is the only person subject to the right to recovery, so the nominal "right" is unenforceable if a defense of excuse exists. 189
By asserting an analytic distinction between the grounds of recovery and the grounds of liability, Coleman also has failed to take the concept of a right (or a legitimate claim) seriously. A claim, as opposed to a mere statement of need, is always directed to some person (or persons or group). A legitimate claim -- a claim that must be satisfied -- is a claim of entitlement or right directed to that person, which assumes a correlative duty on her part to satisfy the claim. The correlativity of the right and the duty is analytic or conceptual. 190 There is no such thing, as Coleman's annulment thesis assumes, of a right "in the air," good against no one, or a duty "in the air," owed to no one. Since rights and duties are correlative, the normative argument that grounds a right to recovery necessarily also grounds the duty to repair and indicates the appropriate institutional form for the mode of rectification constituted by the correlative right and duty. 191 The severance of these issues in the annulment thesis is not only normatively incoherent, but also analytically confused. The puzzling question is what induced Coleman to adopt this position -- a position to which he still adheres. 192
[*668] To answer this question, we must retrace the winding route that Coleman has taken in the development of his theory. Coleman's initial articles were written to defend compulsory no-fault insurance proposals for dealing with accident costs from charges that, by displacing the tort system (which he called "the fault system"), such proposals sacrificed justice for utility or efficiency. 193 He originally assumed that when one person injures another, the only relevant issues of justice are retribution for blameworthy conduct and compensation for undeserved losses. 194 He further assumed that the tort system improperly conflates these two issues:
An accident law in which liability for accident costs is the penalty for wrongful driving conflates two distinguishable issues: those of retribution and recompense. The fault system may be characterized by the fact that it joins these two issues, since in that system, the retribution for wrongdoing is recompense for injuries that are one's fault. The distinguishing feature of no-fault is the separation of these two concerns. 195
Coleman then argued that the tort system is deficient as a mode of retribution for at least three reasons. First, the standard of negligence in tort law is an objective one that merely considers whether you have met the specified standard of care ("legal fault"), rather than a subjective one that takes into account your personal ability and efforts to meet the standard of care ("moral fault" or "blame"). Second, regardless of the moral culpability of your behavior, liability is imposed only if harm to another results. Third, tort liability normally is for the damages that you tortiously cause, which often bear little relation to the moral culpability of your conduct. Retribution is much more appropriately achieved through criminal sanctions, which generally are imposed regardless of actual harm and are scaled to the moral culpability of the wrongdoer's conduct. On the other hand, Coleman argued, there is no reason to limit compensation for undeserved losses to the victims of losses caused by tortious conduct, as tort law does, rather than extending compensation to all accident victims (or even beyond) through a compulsory no-fault scheme or a social insurance scheme. 196
[*669] According to Coleman, the injurer's faulty conduct provides a reason for imposing liability for the victim's loss on the injurer only when the issue of who must bear the loss is arbitrarily limited to the injurer and the victim, as occurs when the tort system is the victim's exclusive avenue of relief. Coleman relies on the "weak retributivism" argument articulated by Joel Feinberg. Feinberg argues that, in such circumstances only, the causal fault of the injurer (which he assumes to be morally culpable fault, although he leaves open the possibility of extending the argument to objective legal fault) provides a relevant moral reason, based on desert, for shifting the loss from the innocent victim to the faulty injurer: as between the culpable wrongdoer and the innocent victim, the wrongdoer should bear the loss, unless there are other, weightier reasons of justice or utility for leaving the loss on the victim. 197
Coleman gives Feinberg's argument a special twist. Coleman notes that the point of the "weak retributivism" argument cannot be to ensure that the wrongdoer is punished since that could be accomplished better through an independent criminal sanction. He therefore recasts the argument as a principle that innocent persons should not be "penalized" by requiring them to bear losses. Applying this principle to the situation where either the faulty injurer or the innocent victim must bear the loss, he argues that the principle requires shifting the loss to the faulty injurer "to protect the non-faulty, rather than to penalize the faulty." 198 Following Feinberg, Coleman emphasizes that this argument applies only when there is no other avenue for compensating the victim, since the victim's "innocence" is "protected" whenever he is not required to bear the loss, rather than only when the legally faulty injurer bears it. Similarly, Coleman states, the argument is decisive only when there are no other relevant "moral" differences, such as cost-bearing or cost-spreading capacity. 199
Coleman's focus on the distinct goals of retribution and compensation in this initial stage of the development of his theory seems to have been the genesis of his distinction between the grounds of recovery and the grounds of liability. 200 As he initially understood what was going on, the supposed ground for imposing liability for the victim's loss on the injurer in the fault system is retribution, while accident victims have a meritorious claim (on some unspecified ground of distributive justice or utility) for compensation whether or not they are injured through fault. That is, according to Coleman, the fault system itself relies on grounds of liability that are analytically and normatively distinct from the grounds of recovery. Since [*670] the grounds are distinct, they can and should be split from one another and embodied in distinct institutional forms more appropriate to each ground: the criminal law for retributive liability and no-fault or social insurance systems for recompense or recovery, respectively.
However, this argument does not demonstrate an analytic or normative disjunction between grounds of recovery and grounds of liability. To the contrary, it depends on the assumption that distinct grounds of recovery and liability are improperly paired in the fault system. Insofar as the ground of liability is the harm to the social order created by one's conscious flouting of the rules of that social order, the proper form of liability is public retribution or punishment ("exacting reparation from the moral wrongdoer"), 201 and the proper claimant is "the state" acting on behalf of all its citizens, who all are harmed by such behavior. The grounds of recovery match the grounds of liability, and the criminal law system is the appropriate institutional form for the mode of rectification constituted by the correlative bilateral right and duty. On the other hand, insofar as the ground of recovery is a deficiency in holdings as measured by some standard of distributive justice or utility, the proper form of liability is a shifting of resources from those who have an excess in holdings as measured by the same standard, through some legislatively enacted scheme of redistribution. Once again, the grounds of recovery match the grounds of liability, and an administrative process applying the legislated criteria for redistribution is the appropriate institutional form for the mode of rectification constituted by the correlative multilateral rights and duties. 202
Coleman's retribution-versus-compensation argument was also flawed for another reason. As Roger Pilon noted, Coleman was attacking a straw man since few people familiar with the tort system have viewed either retribution or general compensation for loss as one of its primary purposes. 203 Coleman himself soon realized this. In his second article, Coleman restated the arguments from his first article, which were directed solely against the alleged retributive arguments for the fault system. Coleman then stated:
At one time, I was willing to let the argument for no-fault stop there. But I have come to recognize the possibility that our intuition in the state of nature case may not be grounded in fears that X's innocence will be penalized or that Y's wrongdoing won't be, but in something about the nature of compensatory justice itself. Perhaps this [state of nature] example reveals the essential characteristics of compensatory justice which bar assimilating it with either weaker or stronger principles of retributivism, and [*671] which require that in order to be just, compensation for wrongdoing originate with wrongdoers. I believe, however, that we can distinguish compensation from punishment, and compensatory justice from other principles of justice, while denying that compensatory justice prohibits no-fault accident law. 204
Thus, Coleman seems to have revised his principle of compensatory justice. In his first article, the principle seemed to encompass all undeserved losses and provided no reason for imposing liability for losses caused by human activity on the person who caused them. In the second article, however, relying on James Nickel's elaboration of compensatory justice, Coleman describes compensatory justice as "the elimination of [undeserved or otherwise] unjustifiable gains and losses owing to human interaction: what [Nickel] aptly terms 'distortions,'" and he asserts that, "in order to secure the elimination of distortions, [the principle of compensatory justice] supports a system of correlative rights and duties between respective victims and wrongdoers." 205
Despite an initial impression to the contrary, Nickel's and Coleman's principle of compensatory justice is not equivalent to the traditional Aristotelian concept of corrective justice; rather, it is the annulment thesis. For Nickel, compensatory justice is an aspect of distributive justice. Justice is a matter of people having those holdings they are entitled to under the appropriate conception of distributive justice, and compensation is the means by which unjust losses -- downward distortions in distributively just holdings -- are undone. Compensatory justice is the principle whereby faulty injurers who cause such unjust losses are required to rectify those losses. Rectification is not required as a matter of retribution, even in its weak sense, but simply because the faulty injurer has invaded the rights of another by causing a loss in his distributively just holdings. If the unjust loss is rectified in some other way -- e.g., through private or social insurance or no-fault compensation plans -- there is no reason as a matter of justice for requiring the faulty injurer to compensate her victim. If the faulty injurer acted culpably and secured an unjust gain, that distortion should also be rectified, but it can be rectified through the criminal law or other institutions rather than by requiring the faulty injurer to compensate her victim. 206
Nickel's argument proves both too much and too little. If the reason for holding the injurer liable (in the absence of alternative compensation) is the invasion of the rights of another by causing a loss in his distributively just holdings, it would seem to support absolute liability rather than only [*672] fault liability, as Nickel supposes. Conversely, if the point of compensatory justice is to protect and preserve distributively just holdings, why limit it to losses in holdings that result from human action or, even more restrictively, to losses that result from faulty human action, rather than instituting a social insurance scheme for all losses in distributively just holdings? 207 Although Nickel disclaims reliance on the strong or weak retributive arguments, he provides no other reason for limiting compensatory justice claims, which for him are actually distributive justice claims, to losses that result from faulty human action. 208 If, as both Nickel and Coleman assume, causal human fault, and not merely the suffering of a loss in one's distributively just holdings, is a necessary element of the compensatory justice claim, then the claim can only legitimately be made against the faulty injurer since she alone satisfies this necessary element. The liability cannot be coercively shifted to others who do not satisfy this element -- for example, through compulsory social insurance or no-fault schemes. The grounds of liability are coextensive with the grounds of recovery, analytically and normatively.
Coleman's elaboration of the principle of compensatory justice departed from Nickel's elaboration in two important respects. Coleman initially seemed to agree with Nickel that compensatory justice only supports claims to rectification of distortions in distributively just holdings. 209 However, Coleman subsequently asserted that compensatory justice is completely independent from distributive justice: compensatory justice supports claims to recovery for distortions in existing holdings whether or not those holdings are distributively just, as long as they were acquired justly. 210 But, if departure from the distributively just set of holdings is not the criterion for unjust gains and losses under compensatory justice, what is? Coleman has yet to provide a substantive answer to this critical question. He merely asserts that wrongful (undeserved, unwarranted, unjustified, etc.) gains and losses are those caused by wrongful or faulty conduct or the invasion of a right, but he presents no substantive [*673] account of wrongfulness, fault, or rights. 211 He adopts the objective interpretation of "fault" for purposes of both the "weak retributivism" argument and the supposedly distinct compensatory justice argument, and for wrongful gains as well as wrongful losses. The substantive meaning of "fault," however, is left completely open. According to Coleman, it could be defined in terms of efficiency, reciprocity of risk, or any relevant difference. 212
Second, Nickel viewed the victim, in the absence of some other method of compensation, as having a compensatory justice claim against the faulty injurer who caused the unjust loss regardless of whether the injurer acquired an unjust gain. However, Coleman asserted that the victim would have a compensatory justice claim against the faulty injurer if and only if the injurer acquired an unjust gain as a result of the interaction. In the absence of unjust gain by the injurer, the victim who suffered an unjust loss would have a compensatory justice right to recovery, but not -- as a matter of compensatory justice at least -- against the injurer (or anyone else). 213
Coleman developed this curious concept of a right "in the air," good against no one, through a two-step argument. 214 In the first step, he began by assuming, in accord with Nickel, the correlativity of the victim's claim to recovery and the injurer's duty to repair under the principle of compensatory justice. 215 He then noted that duties need not always be discharged by the person who is bound by the duty. For example, unless the obligation is specified in such a way that it can only be discharged by the obligor, contractual duties or debts can be discharged by someone else on behalf of the obligor. The debt owed to society by a criminal, on the other hand, can only be discharged by the criminal. Thus, Coleman concluded, the question of whether compulsory social insurance schemes or no-fault plans are consistent with the correlative rights and duties of compensatory justice depends on whether compensatory justice duties to repair are more like contractual duties, which can be discharged on the injurer's behalf by others, or the criminal's debt of repayment, which must be discharged by [*674] the injurer herself. 216
But Coleman's conclusion does not follow. Even if (as is the case) a duty to repair in tort is like a contractual duty in that it can be discharged on the injurer's behalf by another person, that other person cannot be compelled to discharge the injurer's duty in the absence of a prior voluntary contractual agreement to do so. All of the examples of permissible discharges of a duty by a person other than the obligor cited by Coleman involve that other person's voluntary discharge of the duty, or, weaker yet, merely a voluntary gift to cover the loss that was not intended to (and hence will not) discharge the obligor's duty. 217 Moreover, contrary to Coleman's assumption, the mere fact that contract and tort duties to repair can be discharged on the obligor's behalf by others -- i.e., that compensation pursuant to these duties need not "originate with" the obligor -- does not imply that the obligor has no duty. 218 The obligor remains subject to the duty unless and until it is completely discharged by the obligor herself or by others on her behalf. Coleman's examples merely support voluntary participation in social insurance or no-fault compensation plans. They do not support compulsory participation in such plans or the abolition of the injurer's duty to repair under the tort system. 219
[*675] In the second step of his argument, Coleman concluded that a faulty injurer has no compensatory justice duty to repair the unjust loss caused by her fault unless she also gained from her fault. The logic involved in this second step is not clear, but it seems to include the following four assumptions: (1) no compensatory justice duty to rectify exists if the duty is capable of being discharged by others; (2) the duty to rectify unjust losses through compensation can be discharged by others; (3) the duty to rectify unjust gains through disgorgement can only be discharged by the person who has the unjust gain; and (4) the only institutional method available for eliminating an unjust gain is by imposing the duty to repair the victim's loss on the injurer. That is, liability for the victim's loss is placed on the injurer not to rectify the victim's unjust loss, but rather to rectify the injurer's unjust gain. 220
We have already dealt with the defects in the first assumption, which comes from the first step of Coleman's argument. The fourth assumption also is defective because there are at least two other methods for obtaining disgorgement of the unjust gain, each of which is better than imposing liability for the victim's loss on the injurer. As Nickel recognized, one method is the compelled disgorgement of the unjust gain in addition to punishment of the wrongful behavior in a criminal proceeding, as frequently occurs with stolen or forfeited property. 221 The other method is an independent civil action for disgorgement brought by the person at whose expense the unjust gain was realized. 222 Each of these methods correctly distinguishes the issue of unjust gain, with its correlative rights and duties, from the distinct issue of unjust loss, with its distinct set of correlative rights and duties. Coleman's solution -- imposing liability for the victim's loss (or is it liability for the gain? or the larger or smaller of the two? or both?) on the injurer as a means of annulling the unjust gain -- is normatively incoherent and analytically confused (as the preceding parenthetical should suggest).
However, as a result of the above argument, Coleman had committed himself to a principle of compensatory justice whereby victims have compensatory justice rights to recovery if and only if they suffer a loss as a [*676] result of faulty conduct, and injurers have compensatory justice duties to repair, owed to the particular victim, if and only if they obtain some gain as a result of the same faulty conduct. Initially, Coleman thought that injurers would have a compensatory justice duty to provide reparation to the victim under this principle only with respect to certain intentional torts, such as fraud and theft, for which the unjust gain generally would be equal to the unjust loss. Believing that negligent actors generally do not gain from their negligence, he stated that victims of negligence, having suffered unjust losses, would have a compensatory justice right to recovery, but the negligent injurer, having obtained no gain, would not have a compensatory justice duty to repair. 223
Critics pointed out that negligence generally results in gains for the injurer -- savings in avoided costs of care -- as well as losses for the victim. 224 Coleman adjusted his principle of compensatory justice 225 to require that, for the injurer's duty to repair to be owed to the victim, the gains must be dependent on the loss. 226 However, this adjustment was not supported by any part of his theory. Critics also pointed out that in cases in which there seemed to be no loss, there always was at least a loss in security due to the exposure to the negligent risk. 227 Coleman need not have, but did, acknowledge such risk exposure as being, in principle, a loss. Yet Coleman now recognized that unjust gains can be annulled through means other than tort liability. He therefore argued that even though there might be valid grounds for injurer liability (i.e., unjust gain) as well as valid grounds for the victim's right to recovery (i.e., unjust loss), the mode of annulling unjust gains and unjust losses always remains open -- the injurer need not compensate the victim. 228 At this point, Coleman's theory became the pure annulment thesis: in every case, the grounds for recovery allegedly are distinct from the grounds for liability, and the mode of rectification allegedly remains open to utilitarian policy.
[*677] Looming large in the background, however, is the major impediment to the annulment thesis and utilitarian policy: the analytic and normative implications of rights. Once Coleman acknowledges that a victim has a right to recovery, based on a loss caused by the injurer's wrong or fault, the analytic connection between right and duty would seem to require the recognition of a duty to repair by the injurer, based also on the injurer's faulty causation of the loss. 229 Coleman himself has often referred to the logical correlativity of rights and duties. 230 As we have seen, he tries to avoid the logical correlativity principle by arguing that although duties to repair correlative to the particular rights to recovery do exist, the fact that they can be discharged voluntarily on the obligor's behalf by someone other than the obligor implies that they also can be displaced by compulsory social insurance or no-fault schemes. The implication, as we saw, is not valid. 231
Coleman also has argued that the particular mode of rectification constituted by the tort system cannot be implied merely from a corrective justice claim to recovery or recompense, since there are other modes of rectification that also implement such claims -- for example, contract law and the law of restitution. 232 But Coleman confuses the bilateral corrective justice mode of rectification that is common to all these areas of law and distinguishes them from his annulment thesis -- the right to rectification that the plaintiff has against the defendant and the correlative duty of rectification that the defendant owes to the plaintiff -- with the appropriate institutional form. Moreover, there is no great difficulty in distinguishing the appropriate institutional forms for the voluntary agreements encompassed in contracts on the one hand and the harmful involuntary interactions encompassed in torts on the other. The normative foundation for the primary rights that undergird each area of law ordinarily will indicate the appropriate institutional form and the appropriate remedy.
As the last sentence indicates, the rights of rectification discussed in this section are themselves secondary rights, brought into play by the failure of the injurer to respect primary rights against certain interferences with our persons and property. Coleman objects that the invasion of a primary right need not logically imply a right to compensation. What is needed, he says, is a normative rather than a logical argument to get from the invasion of a primary right to compensation as an appropriate remedy. 233 In this he is no doubt correct, at least at one level. The primary right itself must be based on some normative argument. But, as Coleman [*678] and Jody Kraus have argued in a very useful article, primary rights are neither absolute libertarian fortresses nor empty vessels with no specific content. Their content includes not only the particular protected interests, but also the permissible and impermissible means of affecting those interests and, going further than Coleman and Kraus, the appropriate remedies for failure to abide by the specified means. 234 Given such robust primary rights, all or almost all of the structure of rights and remedies can be analytically deduced from the primary rights. 235 At the very least, even with a minimal content specified for a particular primary right, there must be some secondary right, as Coleman and Kraus themselves state, 236 and, whatever that secondary right may be, it must be against the person or persons who breached the primary right rather than against anyone else.
In sum, Coleman's annulment thesis failed as an alleged conception of corrective justice because he failed to take corrective justice seriously as a structure of correlative bilateral rights and duties. It remains to be seen whether this deficiency is remedied in his new, "mixed" conception of corrective justice.
B. The Mixed Conception
In his contribution to this symposium, Coleman claims to have seen the light. He acknowledges that the annulment thesis is, at best, a conception of distributive justice rather than corrective justice. 237 He [*679] acknowledges, at least nominally, that a central feature of the concept of corrective justice is the bilateral correlativity of the victim's right to recovery and the injurer's duty to repair, which together constitute the mode of rectification. 238 He recognizes that unjust gains and unjust losses raise distinct issues of justice that should not be artificially forced together as under the annulment thesis. 239
However, after this promising beginning, there is an almost complete regression. Coleman gradually reintroduces all the major distinctions that are central to the annulment thesis. By the end of his article, Coleman has resurrected the distinction between the grounds of recovery and the grounds of liability, as well as the distinction between the grounds of recovery and liability on the one hand and the mode of rectification on the other.
Coleman continues to argue that the ground of recovery is the suffering of a wrongful loss. 240 However, according to Coleman, the element of wrong or wrongdoing is relevant only to the victim's claim to recovery; it has nothing to do with the injurer's duty to repair, which is instead based on the injurer's causal agency:
The wrong or wrongdoing the injurer has [done] is relevant because it renders the losses caused wrongful. Without wrong or wrongdoing, there may be losses, but they are not wrongful. The duty to correct the losses derives not from the agent's having done wrong . . ., but from the losses being in an appropriate sense the agent's responsibility. They are consequences of agency: the agent's causal powers. . . .
The argument has two parts. First, the losses are the concern of corrective justice only if they are wrongful. They are wrongful if they result from wrongs or wrongdoings. The wrong does not ground the duty to repair. Rather, it grounds the claim that the losses are wrongful and thus within the ambit of corrective justice. Secondly, the duty to repair those wrongful losses is grounded not in the fact that they are the result of wrongdoing, but in the fact that the losses are the injurer's responsibility, the result of the [*680] injurer's agency. The duty to repair a loss under corrective justice is grounded in the injurer's connection to that loss. They are, in a suitable sense, the injurer's responsibility; they are the injurer's, and therefore, are the injurer's to repair. 241
Thus, Coleman continues to adhere to the distinction between the grounds of recovery and the grounds of liability. The victim's claim to recovery is based on the suffering of a wrongful loss. It does not matter who caused the wrongful loss, only that it be wrongful. The claim to recovery, considered by itself, is a claim "in the air," directed at no particular person or persons. The injurer's duty to repair, on the other hand, is supposedly independent of any wrongfulness and is based solely on his causal agency: he is responsible, "in a suitable sense," for all the outcomes of his volitional conduct. But this apparently must also be taken, by itself, as a duty "in the air," owed to no particular person or persons. As in the pure annulment thesis, the "claim" and the "duty" in the mixed conception are not determinate or complete until they are brought together.
What brings the claim and the duty together? According to Coleman, it is the "principle of corrective justice," which "imposes on wrongdoers the duty to repair the wrongful losses their conduct occasions." 242 Although Coleman describes this "principle of corrective justice" as "an element of an underlying foundational theory of rights," 243 he does not explain or justify either its inclusion in the mixed conception or its normativeness. Instead, as he presents it, it is simply a formalist deus ex machina, arbitrarily introduced to link the otherwise free-floating claim to recovery and duty to repair.
Assuming for the moment that Coleman's "principle of corrective justice" succeeds in linking the victim's right to recovery and the injurer's duty to repair in the correlative bilateral manner that Coleman now admits is essential to the concept of corrective justice, it would seem that the mode of rectification would be set. The injurer now owes the duty to repair to the victim, and the victim has the right to recover from the injurer. But appearances can be deceiving. Coleman once more asserts the distinction between the grounds of recovery and liability on the one hand and the mode of rectification on the other. He once more argues that since the injurer's duty can be discharged voluntarily by someone other than the injurer, it also can be shifted involuntarily to others, for example, to the [*681] taxpayers. Once more, the argument fails. 244
But Coleman's mixed conception faces a much more serious problem. Although Coleman apparently believes otherwise, the mixed conception does not provide the particularized linking of injurer and victim that is central to the concept of corrective justice. The duty to repair is based on, and applies to, all the consequences of an agent's volitional action: Coleman explicitly states that the wrongfulness of the action is irrelevant. The duty to repair is indeterminate or incomplete until it is linked with claims to recovery. Claims to recovery encompass all wrongful losses, without regard to who caused them. Thus, when claims to recovery are linked with duties to repair by Coleman's "principle of corrective justice," the agent's duty to repair applies to all wrongful losses that were contributed to by the agent's volitional action, whether or not the agent herself behaved wrongfully. That is, Coleman's "principle of corrective justice" is inaccurately stated. To correctly reflect the linking of the claims to recovery and the duties to repair as he has specified them, it should be stated as the principle that "imposes on agents the duty to repair the wrongful or nonwrongful losses their conduct occasions, but only if those losses were wrongfully caused by someone [anyone]." 245 It is agents or volitional causers, not wrongdoers, who are subject to the duty to repair. Their duty to repair extends to all wrongfully caused losses to which their volitional behavior contributed, whether or not they themselves behaved wrongfully.
It gets worse. Contrary to what Coleman occasionally has asserted, 246 causes include omissions as well as affirmative acts. 247 Thus the duty to repair, being based solely on volitional causation, extends not only to losses that you affirmatively cause, but also to those that you fail to prevent -- that is, it extends to all losses. The end result of the mixed conception, with its duty to repair based solely on agency, is that each and every one of us has a duty to repair all wrongful losses suffered by anyone. The mixed conception is thus identical to Coleman's current interpretation of the annulment thesis. 248 The only difference between Coleman's original interpretation of the annulment thesis and his mixed conception is that, in the mixed conception, the injurer's causal agency occupies the role that was filled by the injurer's wrongful gain in the pure annulment thesis. But that is a huge difference. The wrongful gain limitation on injurer liability was extremely restrictive. The causal agency limitation on injurer liability is no restriction at all.
Coleman was right to jettison wrongful gain as the ground of liability for tortiously caused losses. But it is surprising that he would replace it with causal agency. He himself has frequently argued that although volitional causation may be necessary for moral and legal responsibility, it is not sufficient, and that responsibility in the sense of authorship -- mere physical [*682] causation -- should not be confused with responsibility in the sense of moral and legal obligation. 249 Coleman seems to have been persuaded to change his mind on this issue by Stephen Perry, who identifies the notion of "outcome responsibility" -- normative responsibility for the effects of one's volitional conduct -- as one of the two fundamental elements underlying moral and legal obligation in his developing theory of tort liability. 250 Perry, in turn, cites Tony Honore's initial development of this notion. 251 The causal agency argument is not a Kantian one, which focuses on the implications and requirements of being a moral (self-legislating) being in a world of other such beings. Rather, the argument is that outcome responsibility is essential to our status and identity as persons; that we have to accept and live with the consequences of our actions (and inactions); and that, for either or both reasons, normative implications follow merely from the fact that we have acted and produced consequences, whether foreseeable or unforeseeable. 252
All of this seems to me, as it once did to Coleman, to confuse authorship or responsibility in the sense of mere physical causation with responsibility in the normative sense. Inescapably, the consequences of my volitional actions are "mine" in the sense of mere physical causation. But, by itself, this empirical fact implies nothing normative. No one's continuing status, history, or character as a person is threatened in any way by the lack of absolute moral or legal responsibility for the consequences of our actions. Moreover, as Honore recognizes, our omissions as well as our affirmative acts are volitional and produce consequences through our failure to prevent or produce other consequences. 253 If outcome responsibility makes us normatively responsible for all the consequences of our volitional acts and omissions, then, once again, all of us are responsible for everything.
[*683] Coleman's mixed conception, like his annulment thesis, fails as a plausible interpretation of corrective justice because it fails to take corrective justice seriously as a structure of correlative bilateral rights and duties. More fundamentally, Coleman fails to take the concept of a right (or legitimate claim) seriously. Rights to rectification logically imply duties to rectify. Thus, contrary to Coleman's basic premises, the grounds of recovery are not distinct from the grounds of liability, and the mode of rectification is not distinct from the grounds of recovery and liability. Rather, the normative argument that grounds a right to rectification also grounds the duty to rectify, and the mode of rectification is constituted by the correlative bilateral right and duty.
However, Coleman correctly emphasizes (contrary to Weinrib's position) that the substantive content of the rights and duties in corrective justice must be grounded on normative premises that are extrinsic to the concepts of right, wrong, duty, and causation. Coleman has not yet elaborated any such normative premises (although they may be contained in his forthcoming book). Indeed, Coleman seems to view the concept of corrective justice as purely formal and devoid of any particular substantive content. In the next Part, I attempt to demonstrate that this view is mistaken by reclaiming and restoring the traditional Aristotelian concept of corrective justice.
IV. Aristotle's Principled Account of Law and Justice
A. Aristotle's Principlism
In Aristotle's writings, law, morality, justice, and politics are all intimately and inextricably related to one another. He describes politics as "the most authoritative art" and "the master art" because its object is the ascertainment and attainment of the good for each and every citizen of the state. 254 The highest good for humans is activity in accord with a rational principle and in accord with complete virtue over one's life. 255 Justice in its "wide" sense is the manifestation of such complete virtue in our relations with others. 256 Justice in its particular sense -- the sense in which it commonly is used in our time -- is the manifestation in our relations with others [*684] of the specific virtue of behaving properly with respect to claims to goods or advantages, broadly conceived. 257 Law is the instrument by which the state, through politics, seeks to achieve justice by "commanding some acts [the virtuous ones] and forbidding others [the vicious ones]." 258 "The rightly-framed law does this rightly, and the hastily conceived one less well." 259 Insofar as the law is rightly framed, the lawful person is the completely virtuous person -- the person who is just in the wide sense, and hence also in the particular sense. 260
Aristotle's conception of the good is nonaggregative and egalitarian. In his view, the characteristic of a human being that distinguishes it from plants and animals is that a human being lives its life in accord with a rational principle. 261 The highest good or happiness for a human being is "activity of soul" (activity in accord with a rational principle) in accord with complete virtue in a complete life 262 rather than "some plain and obvious thing, like pleasure, wealth, or honour." 263 This highest good or happiness is intrinsic to the individual herself rather than being something external, such as wealth or honor. 264 Those who treat pleasure or enjoyment as the good are "vulgar," and a life aimed at such is "a life suitable to beasts." 265 The good or virtuous person needs a certain moderate amount of property "for it is impossible, or not easy, to do noble acts without the proper equipment." 266 Yet the acquisition of wealth or property is not a good in itself, but rather is "undertaken under compulsion" as a necessary means to an end. 267 It is properly aimed at and limited by what is needed for a virtuous life, and it is justly censured when it is undertaken for its own sake, as a good in itself. 268
The goal of politics is the attainment of the good for each and every citizen of the state. 269 Since "man is by a nature a political [or social] animal," who cannot be self-sufficient (fully realized) as a human being in isolation from others, but rather can achieve such self-sufficiency only in a voluntary community of friends, "the chief end, both of individuals and states," is the attainment of the common good of the citizens of the state [*685] through the creation and maintenance of such a community. 270 Aristotle writes:
[A] state is a community of families and aggregations of families in well-being, for the sake of the perfect and self-sufficing life. Such a community can only be established among those who live in the same place and intermarry. Hence there arise in cities family connexions, brotherhoods, common sacrifices, amusements which draw men together. But these are created by friendship, for to choose to live together is friendship. The end of the state is the good life, and these are the means towards it. And the state is the union of families and villages in a perfect and self-sufficing life, by which we mean a happy and honourable life.
Our conclusion, then, is that political society exists for the sake of noble actions, and not of living together. 271
It is only in this sense -- the necessity of the state for the good (perfect or self-sufficing) life of each of its citizens -- that the state is "prior to the individual." 272 Aristotle rejects Plato's conception of the state as an organic unity. He instead views the state as a diverse plurality or community of equals constituted by its citizens, with the diverse plurality being necessary for the self-sufficiency of the state and its citizens. 273 The common good is nothing more nor less than the good of each and every citizen, which is a life of activity in accord with a rational principle and in accord with complete virtue. 274 Thus, the common good that is the end of politics is justice in its wide sense (the manifestation of complete virtue in our relations with others), 275 and "governments which have a regard for the common interest are constituted in accordance with strict principles of justice." 276
Aristotle emphasizes the intimate connections between politics, justice, and law by describing justice as expressed in the law as "political justice" and distinguishing it from what is ideally or "naturally" just: "As to what we [*686] establish for ourselves and practice, that is thereby just, and we call it just according to law. Natural justice . . . is better than legal. But what we are in search of is political justice. Now the politically just is the legal, not the natural." 277 However, the legally or politically just includes, albeit imperfectly and only to the extent that the law is rightly framed, what is naturally just as well as what is just only in a strictly legal or conventional sense:
Of political justice part is natural, part legal, -- natural, that which everywhere has the same force and does not exist by people's thinking this or that; legal, that which is originally indifferent, but when it has been laid down is not indifferent, e.g., that a prisoner's ransom shall be a mina, or that a goat and not two sheep shall be sacrificed, and again all the laws that are passed for particular cases . . . . 278
Thus, Aristotle's position is totally at odds with Weinrib's. For Weinrib, law as justice is radically separated from ethics and politics; it has no purpose other than "to be itself." 279 For Aristotle, on the other hand, law is the instrument by which politics seeks to assure that each citizen is completely virtuous, and justice is the manifestation of that complete virtue in citizens' relations with one another. 280
In sum, Aristotle is not a formalist. He also is not a pragmatist. Rather, he adopts the basic tenets of the principlist. He believes that there are fundamental principles of morality and justice, which are reflected imperfectly in existing law despite its appearance of constant change: "Do not suppose that, if things change owing to our use, there is not therefore a natural justice; because there is. For that which continues for the most part can plainly be seen to be naturally just." 281 Aristotle recognizes that the written law must always be incomplete. He insists that to fill in the details [*687] or gaps, one must turn to the underlying principles (the "universal" or "natural" law). 282 To fail to do so would be to substitute the partial, lawless rule of man (or woman) for the impartial rule of law:
The laws speak only in general terms . . . . Hence it is clear that a government acting [solely] according to written laws is plainly not the best. Yet surely the ruler cannot dispense with the general principle which exists in law; and that is a better ruler which is free from passion than that in which it is innate. Whereas the law is passionless, passion must always sway the heart of man. 283
The proper application of the fundamental principles to a particular situation requires the virtue of practical wisdom, which depends, inter alia, on experience, so that the young ordinarily do not make good legislators, judges, or jurors. 284 But practical wisdom is not the same as the pragmatist's "situation sense," which is not connected to any set of fundamental principles. Practical wisdom is connected to, and aimed at the attainment of, the fundamental principle(s) that constitute the end of the particular virtue -- e.g., the principles of distributive or corrective justice: "virtue makes the aim right, and practical wisdom the things leading to it." 285 Hence it is important to be as clear as possible about the fundamental principles:
Now of first principles we see some by induction, some by perception, some by a certain habituation, and others too in other ways. But each set of principles we must try to investigate in the natural way, and we must take pains to determine them correctly, since they have a great influence on what follows. For the beginning is thought to be more than half of the whole, and many of the questions we ask are cleared up by it. 286
For Aristotle as well as other principlists, the "natural way" to ascertain the fundamental principles is to begin with an analysis of our existing practices and concepts, and to make sure that the principles that are developed are consistent with those practices and concepts, as critically understood and clarified. Aristotle states:
Let us not fail to notice, however, that there is a difference between arguments from and those to the first principles. . . . For, [*688] while we must begin with what is familiar, things are so in two ways -- some to us, some without qualification. Presumably, then, we must begin with things familiar to us. . . . For the facts are the starting-point . . . . 287
We must consider [the principle], however, in the light not only of our conclusion and our premises, but also of what is commonly said about it; for with a true view all the facts harmonize, but with a false one they soon clash. 288
With these admonitions in mind, we turn now to Aristotle's elaboration of the principles of justice.
B. The Concept of Justice
Aristotle, analyzing the concept of justice as it was employed in his time, distinguished a general sense and a particular sense. The general or "wide" sense of justice, less familiar but still accessible to us today, is complete virtue, or excellence in disposition, as manifested in our relations with others. 289 The narrower, particular sense of justice -- upon which we, following Aristotle, will focus -- is restricted to that part of Aristotle's wide sense of justice that involves claims to goods, broadly conceived:
That there is such a [particular sense of justice] is indicated by the fact that while the man who exhibits in action . . . the other forms of wickedness acts unjustly [in the wide sense] but not graspingly (e.g., the man who throws away his shield through cowardice or speaks harshly through bad temper or fails to help a friend with money through meanness), when a man acts graspingly he often exhibits none of these vices, -- no, nor all [of them] together, but certainly wickedness of some kind (for we blame him) and injustice. There is, then, another kind of injustice which is a part of injustice in the wide sense . . . . Evidently, therefore, there is apart from injustice in the wide sense another, particular, [sense of] injustice which shares the name and nature of the first, . . . for the force of both lies in a relation to others but the [particular sense] is concerned with honour or money or safety -- or that which includes all of these, if we had a single name for it -- and its motive is the pleasure that arises from gain; while the [wide sense] is concerned with all the objects with which the good man is concerned. 290
Insofar as our relations with others are concerned, justice in its wide sense, being complete virtue as manifested in our relations with others, is coextensive with morality. Moreover, for Aristotle, justice in its wide sense constitutes the bulk of rightly framed law. The ascertainment and attainment by all of the good -- activity in accord with a rational principle and in accord with complete virtue over one's life -- is the task of the "master art" [*689] of politics and its instrument, the law. The law requires us to practice virtue and avoid vice in all our relations with others by commanding some acts and forbidding others. The person who is just in the wide sense -- i.e., who demonstrates in his relations with others all the virtues, including justice in its particular sense -- is a lawful person, and vice versa. 291
Aristotle identifies virtue as a disposition, in one's passions and actions, to choose the "equal," "intermediate," or "mean" (Aristotle uses these terms interchangeably) between excess and deficiency -- neither too much nor too little. This equal, intermediate, or mean will vary relative to the particular circumstances and person or persons involved, insofar as they differ with respect to the characteristics relevant to the particular virtue. 292 Justice in its wide sense is the disposition to choose the equal, intermediate, or mean between excess and deficiency in all one's relations with others. 293
Aristotle's elaboration of the concept of justice, even in its wide sense, captures its essential elements as they continue to be understood today. In sum, the concept of justice is generally understood (1) as applying only to people's dealings or relationships with others, (2) that involve or give rise to legal duties and rights, and (3) as requiring that such duties and rights be structured so that all are treated equally (or differently) insofar as they are equal (or different) with respect to the relevant characteristic. 294
Justice in its particular sense is the disposition to choose the equal, intermediate, or mean between excess and deficiency in one's relations with others that involve claims to goods. 295 Contrary to Weinrib's assertions, neither the external focus on relations with others nor the focus on goods prevents justice in its particular sense from being a virtue in the same sense as all the other virtues. 296 Many of the virtues and vices in addition to justice or injustice in its particular sense inherently involve relations with others. 297 And at least one virtue in addition to justice in its particular sense inherently involves relations with others having to do with goods. This virtue is liberality (or generosity). The distinction between the virtue of justice in its particular sense and the virtue of liberality is that the virtue of [*690] justice focuses on one's dealings with others' goods, 298 while the virtue of liberality focuses on one's dealing with one's own goods. 299
For the virtue of justice in its particular sense, Aristotle describes the excess in disposition as the vice of being grasping -- the disposition to take what properly belongs to others. 300 The corresponding deficiency in disposition can be described as the vice of being yielding -- the disposition to allow others to take what is properly yours, which Kant describes as a lack of Rightful honor. 301 Aristotle states:
It is plain that just action is intermediate between acting unjustly and being unjustly treated; for the one is to have too much and the other to have too little. Justice is a kind of mean, but not in the same way as the other virtues, but because it relates to an intermediate amount, while injustice relates to the extremes. And justice is that in virtue of which the just man is said to be a doer, by choice, of that which is just . . . . Injustice on the other hand is similarly related to the unjust, which is excess and defect, viz. because it is productive of excess and defect . . . . In the unjust act to have too little is to be unjustly treated; to have too much is to act unjustly. 302
While both the disposition to be grasping and the disposition to be yielding are defects in a person's character and, in that sense, vices, only the former when manifested in our relations with others is a vice in the sense of particular injustice. Virtue and vice exist only when there is choice. 303 Persons often choose to act unjustly by behaving graspingly in relation to others, and persons sometimes consent to be harmed or suffer yieldingly what is unjust, yet no one, Aristotle notes, chooses to be unjustly treated. 304 Thus, while both acting unjustly and being unjustly treated are bad, "acting unjustly is the worse, for it involves vice [if done deliberately] and is blameworthy . . . while being unjustly treated does not involve vice and [*691] injustice [on the part of the person being unjustly treated]." 305 More generally, acting yieldingly is an instance of injuring yourself, and you cannot be unjust to yourself, except metaphorically, for a variety of reasons (some better than others): no one chooses to be unjustly treated; gains and losses that are consented to are not unjust; there is no net gain or loss since (notionally) they cancel each other out; and justice by definition involves dealings or relations with others rather than oneself. 306
C. Substantive Justice: Corrective Justice and Distributive Justice
We also owe to Aristotle the formal elaboration of the distinction between the two kinds of substantive justice in its particular sense. 307 The first kind of justice discussed by Aristotle is distributive justice, which is concerned with the "distributions of honour or money or the other things that fall to be divided among those who have a share in the [government]." 308 To be just, such distributions must satisfy the criterion of equality understood in terms of geometrical proportion -- an equality of ratios. All individuals in the political community are measured against some distributive criterion (e.g., merit or need 309 ), and goods or advantages are allocated to different individuals in the same proportion as their respective measurements. For example, if the distributive criterion is merit, and Ann is twice as meritorious as Bob, Ann should have twice as much of the goods or advantages in the community as Bob. 310
The other kind of justice is corrective (or rectificatory) justice, which is concerned with interactions between persons that affect those persons' existing holdings or stocks of goods, broadly construed. 311 The equality of corrective justice differs significantly from that of distributive justice. Persons involved in an interaction are considered to be equal, for purposes [*692] of such interaction, no matter how unequal they may be in terms of merit or need or any other comparative criterion:
[Corrective justice] has a different specific character from [distributive justice]. For the justice which distributes common possessions is always in accordance with the kind of proportion mentioned above [geometrical proportion based on comparative measurements of merit or need, etc.]. . . . But the justice in interactions is a sort of equality indeed, and injustice a sort of inequality; not according to [geometrical] proportion, however, but according to arithmetical proportion. For it makes no difference whether a good man has defrauded a bad man or a bad man a good one, nor whether it is a good or a bad man that has committed adultery; the law looks only to the distinctive character of the injury, and treats the parties as equal, if one is in the wrong and the other is being wronged, and if one inflicted injury and the other has received it. 312
The equality of the parties in corrective justice -- justice in interactions -- includes a presumed equality of entitlement to their respective pre-interaction holdings. 313 If any interaction results in unjust losses or gains in such holdings, the judge in implementing corrective justice reaffirms and restores the preexisting equality by imposing a duty on the injurer to disgorge any unjust gain and to compensate the injured for any unjust loss, so that each party "has his own":
Therefore, this kind of injustice being an inequality, the judge tries to equalize it; for in the case also in which one has received and the other has inflicted a wound, or one has slain and the other been slain, the suffering and the action have been unequally distributed; but the judge tries to equalize things by means of the penalty, taking away from the gain of the assailant. For the term 'gain' is applied generally to such cases, even if it be not a term appropriate to certain cases, e.g. to the person who inflicts a wound -- and 'loss' to the sufferer; at all events when the suffering has been estimated, the one is called loss and the other gain. . . . Corrective justice will be the intermediate between loss and gain. . . . Now the judge restores equality; it is as though there were a line divided into unequal parts; and he took away that by which the greater segment exceeds the half, and added it to the smaller segment. And when the whole has been equally divided, then they say they have their own -- when they have got what is equal. . . . The equal is intermediate between the greater and the lesser according to arithmetical proportion [that is, as their arithmetic mean]. . . . These names, both loss and gain, have come from voluntary exchange; for to have more than one's own is called [*693] gaining, and to have less than one's original share is called losing, e.g. in buying and selling and in all other matters in which the law has left people free to make their own terms; but when they get neither more nor less but just what belongs to themselves, they say that they have their own and that they neither lose nor gain.
Therefore the just is intermediate between a sort of gain and a sort of loss, viz. those which are involuntary; it consists in having an equal amount before and after the interaction. 314
The references, in this passage and elsewhere, to having "one's own," "one's original share," or "an equal amount before and after the interaction" should not be read as requiring that one's post-interaction holdings be equal to one's pre-interaction holdings for each and every interaction. First, only unjust losses or gains from interactions are to be rectified. 315 For example, Aristotle clearly did not mean to preclude fully voluntary exchanges or gifts. 316 Second, the injurer's duty to repair an unjust loss may result in her having less holdings after the reparation than she had before the interaction, and the victim's right to disgorgement of an unjust gain may result in his having more holdings after the disgorgement than he had before the interaction. Such effects on holdings as a result of reparations and disgorgements that are themselves required by corrective justice are obviously consistent with corrective justice. 317
A similar caveat applies to Aristotle's use of the concept of "arithmetical proportion" (i.e., arithmetic mean) to elaborate the interpersonal equality of corrective justice and to distinguish it from the interpersonal equality of distributive justice, which, as we have seen, takes the form of a geometrical proportion. 318 Analogizing corrective justice to arithmetical proportion facilitates a simple, formal contrast to distributive justice as geometrical proportion. But, as with the mathematical treatments favored by modern law and economics scholars, it does so at a substantial cost in clarity, and it depends on simplifying assumptions that deprive it of descriptive validity. For heuristic purposes (to allow the analogy to arithmetical [*694] proportion), Aristotle in the quoted passage treats any interaction that results in an unjust loss as notionally involving both unjust gain and unjust loss, with the unjust gain being equal to the unjust loss. Then, since the parties' pre-interaction holdings are also presumed to be equal, the judge restores those pre-interaction holdings by applying an arithmetic mean (average) to their post-interaction holdings. 319
Aristotle himself indicates that his analogy to arithmetical proportion, and its underlying assumption of an unjust gain equal to the unjust loss, should be read only as a metaphorical heuristic device. 320 Note, for example, in the quoted passage, his caveat about the general applicability of the term "gain" and his "as though" qualification to the discussion of corrective justice in terms of arithmetical proportion. 321 Immediately afterward, he moves from the heuristic, which relies on the presumed equality of both the pre-interaction holdings and the unjust gain and loss, to the more general situation of unjust gain or loss:
For when something is subtracted from one of two equals and added to the other, the other is in excess by these two . . . . It therefore exceeds the intermediate by one, and the intermediate exceeds by one that from which something was taken. By this, then, we shall recognize both what we must subtract from that which has more, and what we must add to that which has less; we must add to the latter that by which the intermediate exceeds it, and subtract from the greatest that by which it exceeds the intermediate. . . . To have more than one's own is called gaining, and to have less than one's original share is called losing . . . ; but when they get neither more nor less but just what belongs to themselves, they say that they have their own and that they neither lose nor gain.
Therefore the just is intermediate between a sort of gain and a sort of loss, viz. those which are involuntary; it consists in having an equal amount before and after the interaction. 322
In sum, the equal, intermediate, or mean in corrective justice is not the arithmetical mean or average of the parties' post-interaction holdings, but rather consists in "having one's own" or "having an equal amount before [*695] and after the interaction," whereas to come out of the interaction with more or less than "one's own" is unjust gain or loss, respectively. 323
But what constitutes "one's own"? Which gains and losses are unjust? What constitutes acting unjustly and being unjustly treated? Which particular actions exhibit the vice of acting graspingly or yieldingly? Are these equivalent questions? To explore these issues, we must discuss Aristotle's distinctions between unjust persons, acts of injustice (acting unjustly), and unjust holdings or injuries.
According to Aristotle, a person acts unjustly, and hence ordinarily deserves moral blame, only if he voluntarily causes an unjust holding (as described below 324 ). Aristotle states:
Whether an act is or is not one of injustice (or of justice) is determined by its voluntariness or involuntariness; for when it is voluntary it is blamed, and at the same time is then an act of injustice . . . . By the voluntary I mean . . . any of the things in a man's own power which he does with knowledge, i.e., not in ignorance either of the person acted on or of the instrument used or of the end that will be attained (e.g. whom he is striking, with what, and to what end), each such act being done not incidentally [i.e., accidentally] nor under compulsion [i.e., nonvolitionally] (e.g. if you take my hand and strike someone else with it, I do not act voluntarily; for the act was not in my power). 325
The "compulsion" that prevents an act from being voluntary is lack of volition or agency. Compulsion in the more usual sense of duress or necessity (i.e., to save life or more valuable property) does not prevent an act from being voluntary and hence unjust with respect to the losses caused to others, but may, at least in situations of necessity, prevent it from being considered blameworthy. 326 The ignorance that prevents an act from being voluntary is ignorance of the "particular circumstances of the action and the objects with which it is concerned," rather than ignorance of what is right or wrong. 327 Moreover, Aristotle treats acts as being voluntary, despite ignorance of the particular circumstances, if the actor is responsible for his ignorance:
Indeed, we punish a man for his very ignorance, if he is thought responsible for the ignorance, as when penalties are doubled in the case of drunkenness; for the moving principle is in the man himself, since he has the power of not getting drunk and his getting drunk was the cause of his ignorance. And we punish those who are ignorant of anything in the laws that they ought to [*696] know and that is not difficult, and so too in the case of anything else that they are thought to be ignorant of through carelessness; we assume that it is in their power not to be ignorant, since they have the power of taking care. 328
To be an unjust (wicked) person -- a person with the vice of being grasping -- a person must cause the unjust holding not only voluntarily, but also by prior deliberation or choice rather than on the spur of the moment as the result of some passion, thereby reflecting an unjust disposition or state of character: 329
When [a person] acts with knowledge but not after deliberation, it is an act of injustice -- e.g. the acts due to anger or to other passions necessary or natural to man; for when men do such harmful and mistaken acts they act unjustly, and the acts are acts of injustice, but this does not imply that the doers are unjust or wicked; for the injury is not due to vice. But when a man acts from choice, he is an unjust man and a vicious man. 330
At the other end of the spectrum, Aristotle states that a person may not be an unjust person, due to lack of prior deliberation, nor act unjustly, due to lack of actual or imputed knowledge of the particulars, yet still through involuntary ignorance produce an unjust holding or injury. Such a person may be said to act unjustly not in a blameworthy way, but only "in an incidental [accidental] way":
[A] man acts unjustly or justly whenever he does such acts voluntarily; when involuntarily, he acts neither unjustly nor justly except in an incidental way; for he does things which happen to be just or unjust. . . . So that there will be things that are unjust but not yet acts of injustice, if voluntariness be not present as well. 331
Similarly, when a person causes an unjust result to another without acting unjustly (i.e., through involuntary ignorance), the other who suffers the unjust result does so without being unjustly treated, except in an incidental sense:
One might raise this question also, whether every one who has suffered what is unjust is being unjustly treated, or on the other hand it is with suffering as with acting. In both it is possible to partake of justice incidentally, and similarly (it is plain) of injustice; for to do what is unjust is not the same as to act unjustly, nor to suffer what is unjust to be treated unjustly, and similarly in the case of acting justly and being justly treated; for it is impossible to be unjustly treated if the other does not act unjustly, or justly treated unless he acts justly. 332
[*697] Aristotle describes the injuries in interaction that result from involuntary ignorance as mistakes in a broad sense, and subdivides these into mistakes in a narrow sense and misadventures. Mistakes in the narrow sense are injuries that were involuntary but reasonably expectable, while misadventures are injuries that were involuntary and contrary to reasonable expectation:
Thus there are three kinds of injury in interactions; those done in ignorance are mistakes when the person acted on, the act, the instrument, or the end [result] is other than the agent supposed; the agent thought either that he was not hitting anyone or that he was not hitting with this missile or not hitting this person or to this end, but a result followed other than that which he thought likely (e.g. he threw not with intent to wound but only to prick), or the person hit or the missile was other than he supposed.  Now when the injury takes place contrary to reasonable expectation, it is a misadventure.  When it is not contrary to reasonable expectation but does not imply vice, it is a mistake (for a man makes a mistake when ignorance originates in him, but is the victim of accident [misadventure] when its origin lies outside him).  When he acts with knowledge but not after deliberation, it is an act of injustice -- e.g. the acts due to anger or to other passions necessary or natural to man; for when men do such harmful and mistaken acts they act unjustly, and the acts are acts of injustice, but this does not imply that the doers are unjust or wicked; for the injury is not due to vice. But when a man acts from choice, he is an unjust man and a vicious man. 333
These three types of injuries in interaction correspond roughly to the traditional categories of tortious injuries. Injuries due to acts of injustice are voluntary (intentional or subjectively negligent) injuries. 334 Mistakes seem to be injuries due to objective negligence. They are injuries that were reasonably expectable given typical knowledge, but which were not expected by the actor due to his own involuntary ignorance. Misadventures are injuries that were not reasonably expectable, not because of the actor's own involuntary ignorance, but due to a general lack of knowledge that "lies outside" the actor. Thus, liability for misadventure would seem to correspond to traditional strict liability. 335
Although it is not entirely clear, Aristotle apparently treats each of these injuries in interaction -- whether due to acts of injustice, mistakes, or [*698] misadventures -- as unjust losses, requiring rectification. 336 As we have seen, he clearly treats losses caused by unjust acts as unjust losses. He also states that there are unjust losses that are not due to unjust acts, but rather occur through involuntary ignorance, so that mistakes, at least, must be unjust losses. Finally, he does not attempt to distinguish misadventures from mistakes as not being unjust losses. Instead, he describes all three types of loss as injuries in interaction. Thus, it is at least a plausible reading of Aristotle that all three of the injuries in interaction are unjust losses, which must be rectified. 337
By recognizing some losses as unjust even though they did not result from the vice of deliberately acting graspingly, or even from acting unjustly (except in the incidental sense), does Aristotle's account of corrective justice as a virtue inconsistently invoke a concept of injustice distinct from the manifestation of vice (lack of virtue)? I think not. In these situations, the vice of injustice is manifested not in the conduct that originally causes the unjust loss, but rather in the deliberate choice, by the responsible person(s), not to rectify the unjust loss once it has occurred. 338 It is true, however, that a sense of unjust distinct from the manifestation of vice is needed to identify when such unjust losses have occurred. (As we will see, that sense of unjust is provided by corrective justice's criterion of equality.) 339
Since they are voluntary (intentional or subjectively negligent), injuries caused by acts of injustice ordinarily constitute wrongdoing and are morally blameworthy, thus rendering the injurer liable to punishment in addition to rectification of the unjust loss. 340 In certain cases of duress, or especially in cases of necessity, such acts might not be morally blameworthy, and hence not deserving of punishment. But they still would be acts of injustice with respect to the injury inflicted on the victim, and would result in the victim's being unjustly treated unless the injury were rectified. 341 [*699] Likewise, injuries due to mistakes and (apparently) at least some misadventures are unjust losses that the injurer must rectify. But the mistakes and misadventures, being due to involuntary ignorance rather than being done with actual or imputed knowledge of the consequences, do not ordinarily constitute wrongdoing and thus should be forgiven rather than punished. Aristotle states:
Since virtue is concerned with passions and actions, and on voluntary passions and actions praise and blame are bestowed, on those that are involuntary forgiveness, and sometimes also pity, to distinguish the voluntary and the involuntary is presumably necessary for those who are studying virtue and useful also for legislators with a view to the assigning both of honours and of punishments. 342
Although Aristotle does not explicitly address the issue, the only apparent reason for requiring disgorgement of any gain from an interaction -- in addition to rectification of any unjust loss -- is punishment. If this is indeed the case, unjust gains apparently result only from wrongdoing (acts of injustice).
Yet, as discussed above, losses may be treated as unjust (i.e., as requiring rectification) not only when they result from wrongdoing, but also when they result from involuntary ignorance (i.e., mistake or misadventure). But if wrongdoing is not the criterion that distinguishes unjust losses from losses per se, what is? The answer is suggested by Aristotle's seemingly paradoxical statement that even deliberate causation of loss may not be an unjust action (nor cause an unjust loss): "If a man harms another by choice, he acts unjustly; and these are the acts of injustice which imply that the doer is an unjust man, provided that the act violates proportion or equality." 343 The requirement that the act violate proportion or equality evidently refers to the criteria of equality for distributive and corrective justice, respectively. Even a deliberately caused loss or gain that is consistent with these criteria is not unjust. This suggests that these criteria are the criteria for which we have been searching -- the criteria for identifying unjust gains and losses.
What are these criteria? For distributive justice, the general nature of the criterion is clear, although its particular content is a subject of much debate. The criterion of equality in distributive justice is geometrical proportion -- an equality of ratios. All individuals in the political community are measured against some distributive standard (e.g., merit or need), and goods and advantages are allocated to different individuals in the same [*700] proportion as their respective measurements. 344 For a holding -- even if deliberately caused -- to be unjust as a matter of distributive justice, it must vary from the person's proportionate share under the criterion of geometrical proportion. Conversely, a person who, without her consent, has less than her proportionate share of holdings or goods has a distributive justice right to that share, whether or not anyone has deliberately, voluntarily, mistakenly, or otherwise caused her to have less than her proportionate share. 345
For corrective justice, the general nature of the criterion is less apparent. The arithmetical proportion concept is merely a metaphorical heuristic device, 346 and the description of the equality in corrective justice as the intermediate between being grasping and being yielding in our interactions with others, reinterpreted as avoiding the causation of unjust gains or losses in such interactions, merely initiated our effort to determine what makes gains and losses unjust. 347 We seem to have come full circle.
Yet Aristotle has not left us without any guidance. He states that neither a gain nor a loss is unjust if the person from whom the gain came or the person who suffered the loss, respectively, freely and voluntarily consented to such gain or loss. 348 He also indicates that an unconsented-to loss resulting from an interaction is unjust, and hence must be rectified, if it was caused voluntarily (intentionally or by a failure to foresee what was subjectively foreseeable) or by mistake (a failure to foresee what was objectively foreseeable), and that at least some unconsented-to losses caused by misadventure (an accident that was not reasonably foreseeable) may be unjust. 349 Finally, Aristotle's account suggests that gains achieved at the expense of another without that other's consent are unjust (and hence must be rectified, in addition to the rectification of any unjust loss) only if they result from wrongdoing (blameworthy voluntary actions). 350
This is as far as Aristotle takes us in his specific discussions of corrective justice. He has taken us further than many people think. Even at this point, I think it is fair to conclude that Aristotle's elaboration of corrective justice is consistent with, and indeed seems to have the same general contours as, our current notions of just criminal and civil liability. But we can go further. 351
Aristotle states that corrective justice has a very different character than distributive justice. More specifically, unlike distributive justice, corrective justice does not use interpersonal comparisons or rankings to [*701] implement a relative or proportional equality among the parties to the interaction. Rather, when assessing the impact of an interaction on the holdings of the parties to the interaction, those parties are considered to be absolutely equal, no matter how unequal they may be in terms of merit or need or any other comparative criterion. 352 But this means that all comparative criteria for determining unjust gains and losses in corrective justice -- including utilitarian, efficiency, and all other aggregative criteria -- are excluded from consideration. This is a powerful substantive implication from Aristotle's account.
This implication is confirmed and broadened by Aristotle's general discussion of the relationship between justice and the good. Aristotle's conception of the good is nonaggregative and egalitarian. He emphatically rejects conceptions of the good that are based on wealth, pleasure, or enjoyment, which are the values that are to be maximized in efficiency and utilitarian theories. Instead, he elaborates a conception of the good that is intrinsic to the individual herself: full realization of one's humanity through activity in accord with a rational principle and in accord with complete virtue over one's life. The goal of politics and justice is the attainment of this common good for each and every citizen of the state, which he describes as a community of free and equal individuals. 353
The most important function of the state, Aristotle declares, is the "power of deciding what is for the public interest [i.e., distributive justice], and what is just in men's dealings with one another [i.e., corrective justice]." 354 Distributive justice deals with the comparative or "public" issues. Corrective justice, on the other hand, deals with private interactions, and it requires that those interactions be consistent with the absolute equality of the parties to the interaction, which is reflected, inter alia, in the parties' equality of entitlement to their respective holdings. Together, distributive justice and corrective justice seek to assure the attainment of the good by each citizen by providing him or her with a proportionately equal share of the needed resources and advantages (distributive justice) and by safeguarding his or her existing holdings from actions by others that are inconsistent with the interacting parties' absolute equality (corrective justice). 355
[*702] Since justice is concerned with the attainment of the good, the absolute equality of the parties in corrective justice must be conceived as an absolute moral equality. Given all of the above, the absolute moral equality of the parties in corrective justice seems to prefigure (through ethical presuppositions rather than formalist conceptual implication) Kant's foundational assumption of the "absolute moral worth" of each individual. 356 Moreover, Aristotle's rejection of wealth, pleasure, and enjoyment per se as components of the good, and his emphasis on the attainment of the good by each and every citizen, seem to preclude efficiency-based or utility-based conceptions of distributive justice as well as corrective justice. Thus, the Aristotelian concepts of corrective justice and distributive justice are not empty formalist shells. They have a powerful nonaggregative and egalitarian substantive ethical content.
V. The Structure of Justice
A. Distinguishing Corrective Justice Claims from Distributive Justice Claims
The distinctions that Aristotle draws between the two different kinds of substantive justice -- distributive justice and corrective justice -- have important implications not only for the content of justice claims (as argued in the previous section), but also for the proper assertion and enforcement of justice claims.
Drawing on the discussion in Part IV, we can distinguish corrective justice claims from distributive justice claims by their respective domains of application, goods encompassed, criteria of equality, and persons covered. Corrective justice claims are grounded in an individual interaction. They encompass the goods possessed by the parties to the interaction. The criterion of equality is that the interaction cannot be inconsistent with the absolute moral equality of the parties to the interaction. The persons potentially covered by corrective justice claims are limited to the parties to the interaction. Distributive justice claims, on the other hand, are independent of individual interactions. They are based solely on a person's status as a member of the political community. They encompass all the goods that exist in the community. The criterion of equality is that these goods must be distributed among the members of the community in proportion to their relative ranking under some (as yet unspecified) distributive criterion. The persons potentially covered by distributive justice claims are all the members of the community.
As Weinrib states, corrective justice claims are always bilateral, although not in the strict sense that Weinrib may intend (i.e., that for any particular injury there is always a single injurer and a single injured). 357 They are claims by one party to an interaction that some other party to the interaction has, as a result of the interaction, adversely affected or threatened the first party's holdings by behavior that is inconsistent with the [*703] absolute moral equality of the parties. When there are multiple parties involved in the interaction, the injured party may have claims against more than one of the other parties, and there may be counterclaims and cross-claims, but each corrective justice claim is distinct and bilateral in that it asserts a correlative right-duty relationship between two specific parties to the interaction that is independent of any other corrective justice claims that may have arisen from the particular interaction.
It should be noted, however, since there has been much misunderstanding on the issue, 358 that corrective justice does not always or even usually require that a corrective justice duty be discharged only by the party who is subject to that duty. Such a requirement would seem to apply only when the appropriate mode of rectification is punishment. 359 When the appropriate mode of rectification is compensation for the unjust loss, rather than or in addition to punishment, corrective justice merely establishes the duty of the party who caused the unjust loss to see to it that the required compensation occurs. 360 There is nothing in corrective justice which prevents that duty from being discharged voluntarily, on behalf of the party with the duty, by someone else -- e.g., that party's insurer 361 or rich aunt. 362 Nor is there any problem from the standpoint of corrective justice if the person who discharges the duty spreads the cost of discharging that duty to others through voluntary market processes -- e.g., by raising the prices of its products.
[*704] The departure from corrective justice occurs only when someone is involuntarily required to discharge the duty of another, or when the duty is initially placed on someone other than the party who should bear the duty as a matter of corrective justice, as occurs in compulsory no-fault compensation schemes. In such situations, corrective justice is violated in two different ways: first, in the failure to impose the duty on the party who should bear the duty as a matter of corrective justice, and, second, in the imposition of the duty on someone who has no obligation as a matter of corrective justice (nor often as a matter of distributive justice). Coleman's attempt to distinguish the grounds of recovery from the grounds of liability violates corrective justice in precisely these two ways. As I argued above, claims of any sort logically require not only that there be a person making the claim, but also a person or persons against whom the claim is being made. Corrective justice claims are grounded in individual bilateral interaction. The party whose goods are adversely affected by the behavior of another that is inconsistent with the parties' absolute moral equality has a claim to rectification. The only person to whom such a claim can be directed is the other party to the interaction, who caused the adverse effect through such behavior. As a matter of corrective justice, there is no basis for a claim against anyone else. 363
A claim by the injured person against others who were not parties to the interaction, as contemplated by Coleman, would have to be based on distributive justice rather than corrective justice, and hence would have to be based on the deficiency, if any, in the injured person's proportionate [*705] share of all the goods in the community, rather than on the specific injury or loss occasioned by the interaction. Distributive justice claims are multilateral rather than bilateral. To determine the goods to which a person is entitled as a matter of distributive justice, we must know both the total amount of goods that exist in the community and the person's relative ranking according to the distributive criterion in comparison with all others in the community. Those persons who have too little under the distributive criterion have distributive justice claims against all those who have too much. But there is no basis for anyone who has too little to single out a specific person who has too much, rather than any other who has too much, to satisfy her distributive justice claim, or to prefer herself against others who also have too little. Rather, proper implementation of distributive justice requires concurrent assessments against all those who have too much and disbursements to all those who have too little.
A bilateral distributive justice claim would be possible if and only if there was only one person who had too much and only one person who had too little. Absent those conditions, any unilateral attempt by a person who has too little to obtain part or all of her deficiency from another person who has too much would not be a proper implementation of the principle of distributive justice. As a matter of distributive justice, she who has too little has no better claim than anyone else who has too little against he who has too much, and he who has too much has no greater obligation than anyone else who has too much to she who has too little. Allowing her unilateral claim would allow her to be preferred over all others who have too little and would allow him to be disadvantaged compared to all others who have too much. Such unequal treatment cannot be supported as a matter of distributive justice. Indeed, her unjustified unilateral attempt to satisfy her deficiency from his goods would be a violation of corrective justice. 364
B. The Independence of Corrective Justice and Distributive Justice
On one view, corrective justice is simply a corollary of distributive justice: it merely serves to correct departures from the distributively just allocation of goods in the community and has no independent purpose or significance. Under this view, there is really only one kind of particular justice -- distributive justice. 365 On another view, corrective justice is independent [*706] of and qualitatively distinct from distributive justice. Corrective justice protects against interactional injuries or threats of injury to a person's goods regardless of the distributive justice or injustice of the pre-interaction division of goods among the parties to the interaction or among the members of the community as a whole. 366
From what has been said above, the second view clearly is the correct one. Aristotle himself stresses that corrective justice is a kind of justice that is qualitatively distinct from distributive justice, and that the respective criteria of equality in each of the two kinds of justice are fundamentally different. Distributive justice employs a criterion of relative equality, according to which goods are distributed among the members of the community in proportion to how they measure, in comparison with one another, on some distributive criterion such as merit or need. Corrective justice, on the other hand, employs a criterion of absolute moral equality, according to which the parties to an interaction (and their entitlements to their current goods) are treated as equal, regardless of how much they might vary from one another under the appropriate distributive criterion or any other comparative criterion. 367
Moreover, as we noted in the previous section, and as Aristotle states, 368 corrective justice claims arise from interactions and are limited to the parties to an interaction, while distributive justice claims are independent of interactions and apply to all members of the community. Corrective justice claims are bilateral, while distributive justice claims are multilateral. Thus, the structure of corrective justice claims is inappropriate for achieving distributive justice, and vice versa.
In sum, corrective justice is a kind of justice completely distinct from distributive justice. It recognizes and enforces rights to one's existing goods that are completely independent from the rights to goods that are the subject of distributive justice. This distinction is clearly reflected in our current law, which recognizes a right to rectification (including, in the absence of strict necessity, punishment for the unjust gain as well as compensation for the unjust loss) when a needy person takes from a well-off person, 369 and protects even goods acquired by theft from subsequent theft by others. 370
[*707] C. The Coexistence and Compatibility of Corrective Justice and Distributive Justice
Weinrib asserts that distributive justice and corrective justice are mutually exclusive: they are alternative rather than complementary justificatory principles, which cannot coherently coexist in the same body of law nor simultaneously apply in any particular situation. 371 Yet Aristotle seems to have thought of distributive justice (just distribution of the goods of a political community among the members of that community) and corrective justice (protection of one's existing goods against wrongdoing, mistake, and misadventure) as coexisting and compatible objectives. As noted above, their domains of application are distinct. Corrective justice claims arise from interactions and apply bilaterally to the parties to the interaction. Distributive justice claims are much broader in scope: they arise solely from a person's status as a member of the political community and apply multilaterally to all the members of that community, independent of any interactions.
Proper claims of corrective or distributive justice that are properly implemented cannot themselves be considered unjust or as giving rise to injustice as a matter of either distributive or corrective justice. For example, a person who properly exercises her corrective justice right to defend herself against the unprovoked aggression of another is not thereby subject to a corrective justice claim by the aggressor, even though she injured the aggressor deliberately and without the aggressor's consent. More generally, the proper implementation of the rectification required by corrective justice cannot itself be considered a violation of corrective justice, although it is done deliberately and without the consent of the person upon whom the duty of rectification is imposed. 372
Similarly, a properly implemented distributive justice claim cannot be considered unjust as a matter of corrective justice, even though its implementation, as in the case of corrective justice, involves a deliberate and unconsented to adverse effect on the existing goods of the persons subject to the distributive justice claim. But it is important to remember that a properly implemented distributive justice claim is multilateral rather than bilateral. It is not proper, as a matter of distributive justice, for any one person who has too little to make a unilateral claim against only one or a few of the persons who have too much. Rather, distributive justice ideally requires that assessments be made against all who have too much and that disbursements be made to all who have too little, so that no one who has either too little or too much is treated unequally in comparison with others who have too little or too much, respectively. 373
[*708] Finally, contrary to Weinrib's assertion, distributive justice claims and corrective justice claims may not only coexist in the same body of law, but also may apply simultaneously in the same situation. A good example is public takings of private property under the exercise of eminent domain, 374 or private takings of private servitudes for public benefit as permitted under nuisance law. 375 In each of these situations, the distributive justice claim is not that the person whose property is being taken has too many goods overall or, even if she does, that she is the only person who has too many goods. Rather, the distributive justice claim is that her property is required to accomplish some distributive objective -- either redistribution of goods to those who have too little, 376 or increasing the total amount of goods in society so that everyone's distributive share can be increased 377 -- and that due to the unique suitability of her property for this objective, an unconsented-to taking rather than a voluntary sale is required to prevent her from seeking excessive compensation for her property and thus appropriating an undue share of the public benefit from the project to herself.
Yet, since the distributive justice claim against her is of this special sort, and is not based on her having too much or, at least, being the only one who has too much, the taking of her property without full compensation cannot be justified as a matter of distributive justice. Rather than the distributive justice objective being implemented, as in the case of proper taxation, by assessing all those who have too much and distributing the proceeds to all those who have too little, or assessing all equally who are initially distributively equal and will benefit equally from the project, she would be singled out, with no distributive justification, to shoulder involuntarily the costs of the redistribution. Being unjustified (absent full compensation) as a matter of distributive justice, the deliberate taking of her property without her consent would be a violation of corrective justice. Thus, in these situations, the just implementation of the distributive justice claim depends on recognition and enforcement of the corrective justice rights of the person whose property is being taken.
[*709] D. The Institutions of Justice
Weinrib also equates corrective justice with private law, and distributive justice with public law. 378 As we will see in this section, substantial areas of corrective justice are properly implemented by public law rather than private law. Conversely, although implementation of the preferred distributive criterion is -- or should be -- exclusively a function of public law, major portions of some areas of private law properly are motivated by distributive justice considerations.
Proper formulation and implementation of distributive justice claims to material goods ideally requires knowledge of the total amount of such goods in the community as well as the comparative ranking of each and every member of the community under the preferred distributive criterion. Given these informational needs, as well as the ad hoc invocation of judicial authority by litigants and the limited number of parties subject to the jurisdiction of the court in any particular legal action, the proper formulation and implementation of such claims, in even a very rough fashion, is obviously well beyond the capacity of the courts. Only the legislature or its administrative delegees have the capacity to distribute the material goods of society in accordance with the preferred distributive criterion. 379
However, the courts can enforce distributive justice claims to political goods such as voting rights, civil liberties, and nondiscriminatory treatment. They can review legislative programs that are intended to achieve distributive justice to insure that such programs demonstrate at least a minimal level of rationality and are properly administered. Moreover, although the courts do not have the capacity to distribute material goods in accordance with the preferred distributive criterion, they can limit the property rights of current property owners in order to prevent these owners from unduly restricting the acquisition and use of property by others in the future. Much of the common law of property, which is enforced through private litigation, has this limited distributive justice objective. Consider, for example, the adverse possession doctrine and the limitations on future interests, restraints on alienation, and servitudes. 380
[*710] Corrective justice has a much narrower domain than distributive justice. Corrective justice is only concerned with the effects of interactions on the goods of the parties to the interaction, and it only requires that such effects be consistent with the absolute moral equality of the parties to the interaction. For corrective justice, unlike distributive justice, no community-wide totalling of goods and comparative ranking of persons is required. Hence courts as well as the legislature have the capacity to formulate the principles of corrective justice. Indeed, courts ordinarily would seem to be better suited to the task. They can more readily take into account and learn from the concrete detail and variety of actual experience, and the ad hoc and limited nature of their jurisdiction should make them less likely than the legislature (or its administrative delegees) to confuse corrective justice issues with distributive justice issues. Finally, the courts' ability, not shared by the legislature, to focus on the details of a particular interaction makes them -- or some administrative equivalent -- indispensable to the implementation of corrective justice.
However, when the goal is widespread ex ante prevention of corrective justice violations, rather than ex post rectification or specific ex ante prevention, the ad hoc and limited nature of the courts' jurisdiction once again makes the courts, even with respect to corrective justice, less suitable institutions than the legislature and its administrative delegees. This is especially true when the widespread regulatory effort raises important distributive justice issues as well as the corrective justice issues.
In some areas of corrective justice, such as tort and contract law, private enforcement of judicially developed rules and doctrines has been the dominant mode of implementation, although the legislatures have taken an increasingly active role in modifying the rules and doctrines. In other areas, such as criminal law and environmental and safety regulation, public enforcement of legislatively articulated rules and doctrines has been the dominant mode of implementation, although the courts have played a significant role in developing the principles of criminal liability. 381
In this Article, I have attempted to rescue the traditional Aristotelian concept of corrective justice from its formalist evisceration in recent scholarship. Despite a possible appearance to the contrary, this Article has not been an indulgence in academic antiquarianism, 382 nor has it been a [*711] mere attempt to give "a respectable (preferably ancient Greek) intellectual pedigree" to a self-preferred "noninstrumental theory of tort." 383 Rather, as indicated at the beginning of this Article, it has been an attempt to reclaim, restore, and elaborate the traditional substantive concept of corrective justice at a time when the idea of justice is the target of pervasive and sustained attacks by efficiency theorists, social welfarists, and antilegalists, 384 whose skepticism regarding justice and rights already has escaped the confines of the academy and threatens to cause significant injury to individuals and the social order. 385 The hope was that by returning to Aristotle's seminal elaboration of the concept, we could recapture its substantive content, and that its substantive content would remain descriptively plausible and normatively attractive -- even compelling -- today.
The hope has been at least partially borne out. At the core of Aristotle's concept of justice is the normative premise that the good to which law and politics should be directed is not the meaningless pursuit of wealth or pleasure, or any aggregative goal, but rather the realization by each human being of his or her humanity through a life of activity in accord with a rational principle and in accord with complete virtue. Embodied in this concept of the good is the idea of the absolute moral worth of each human being as a free and equal member of the community, with an equal entitlement to the resources and security needed to realize its humanity. These normative premises resonate in Kant's ethical and legal theory, as well as in a wide variety of modern moral and political theories. 386 They also, as indicated in very summary fashion in this Article, seem to be consistent with the broad outlines of modern as well as ancient law. It remains to be established, through more detailed investigation, whether these normative premises indeed provide the foundation for modern law. It would be surprising if they did not, but demonstrating that they do must be left to other articles.
For related research and practice materials, see the following legal topics:
GovernmentsCourtsJudgesGovernmentsLegislationTypes of StatutesReal Property LawEstatesFuture InterestsInvalid Restraints & Rule Against Perpetuities
n1. E.g., American Law Institute, 1 Reporters' Study on Enterprise Responsibility for Personal Injury: The Institutional Framework 23-52 (1991) [hereinafter ALI Reporters' Study]; Guido Calabresi, The Costs of Accidents 6 n.8, 291-318 (1970); Guido Calabresi, Concerning Cause and the Law of Torts: An Essay for Harry Kalven, Jr., 43 U. Chi. L. Rev. 69 (1975); William M. Landes & Richard A. Posner, Causation in Tort Law: An Economic Approach, 12 J. Legal Stud. 109 (1983); Peter H. Schuck, Two Causation Conundrums: Mass Exposures and Social Causes, 1 Cts., Health Sci. & L. 305 (1991); Stephen D. Sugarman, Doing Away with Tort Law, 73 Cal. L. Rev. 555, 558-59, 603-11 (1985).
n2. E.g., Roberto Mangabeira Unger, Knowledge and Politics (1975); Robert W. Gordon, Unfreezing Legal Reality: Critical Approaches to Law, 15 Fla. St. U. L. Rev. 195 (1987); Morton J. Horwitz, The Doctrine of Objective Causation, in The Politics of Law 201 (David Kairys ed., 1982); Mark Kelman, The Necessary Myth of Objective Causation Judgments in Liberal Political Theory, 63 Chi.-Kent L. Rev. 579 (1987); Mark Tushnet, Critical Legal Studies: An Introduction to Its Origins and Underpinnings, 36 J. Legal Educ. 505 (1986).
n3. See H.L.A. Hart & Tony Honore, Causation in the Law 111-17, 122-25, 128-29, 235-53 (2d ed. 1985); Richard W. Wright, Actual Causation vs. Probabilistic Linkage: The Bane of Economic Analysis, 14 J. Legal Stud. 435 (1985); Richard W. Wright, Causation in Tort Law, 73 Cal. L. Rev. 1735 (1985) [hereinafter Wright, Causation]; Richard W. Wright, The Efficiency Theory of Causation and Responsibility: Unscientific Formalism and False Semantics, 63 Chi.-Kent L. Rev. 553 (1987); Richard W. Wright, Causation, Responsibility, Risk, Probability, Naked Statistics, and Proof: Pruning the Bramble Bush by Clarifying the Concepts, 73 Iowa L. Rev. 1001 (1988) [hereinafter Wright, Bramble Bush].
n4. See infra notes 6-14. Corrective justice underlies not only tort liability, but also criminal and contractual liability, as well as most environmental and safety regulation. See infra notes 380-81 and text accompanying notes 57-58, 62-64, 381. Some articles which discuss corrective justice in areas other than tort include Peter Benson, Abstract Right and the Possibility of a Nondistributive Conception of Contract: Hegel and Contemporary Contract Theory, 10 Cardozo L. Rev. 1077 (1989); James Gordley, Equality in Exchange, 69 Cal. L. Rev. 1587 (1981); Wendy J. Gordon, Intellectual Property and the Restitutionary Impulse, 78 Va. L. Rev. 149 (1992); John Jeffries, Jr., Compensation for Constitutional Torts: Reflections on the Significance of Fault, 88 Mich. L. Rev. 82 (1989); John Jeffries, Jr., Damages for Constitutional Violations: The Relation of Risk to Injury in Constitutional Torts, 75 Va. L. Rev. 1461 (1989); Sheldon Nahmod, Constitutional Damages and Corrective Justice: A Different View, 76 Va. L. Rev. 997 (1990).
n5. See, e.g., 1 ALI Reporters' Study, supra note 1, at 24-25, 50; 2 ALI Reporters' Study 79-80 & n.61; Sugarman, supra note 1, at 603-09.
n6. See Jules L. Coleman, On the Moral Argument for the Fault System, 71 J. Phil. 473 (1974) [hereinafter Coleman, Fault]; Jules L. Coleman, Justice and the Argument for No-Fault, 3 Soc. Theory & Prac. 161 (1975) [hereinafter Coleman, No-Fault]; Jules L. Coleman, Justice and Reciprocity in Tort Theory, 14 U. W. Ont. L. Rev. 105 (1975) [hereinafter Coleman, Reciprocity]; Jules L. Coleman, On the Morality of Strict Tort Liability, 18 Wm. & Mary L. Rev. 259 (1976) [hereinafter Coleman, Strict Liability]; Jules L. Coleman, Reply to Pilon, 59 Personalist 307 (1978); Jules L. Coleman, Mental Abnormality, Personal Responsibility and Tort Liability, in Mental Illness: Law and Public Policy 107 (Baruch A. Brody & H. Tristram Engelhardt, Jr. eds., 1980) [hereinafter Coleman, Mental Abnormality]; Jules L. Coleman, Corrective Justice and Wrongful Gain, 11 J. Legal Stud. 421 (1982) [hereinafter Coleman, Wrongful Gain]; Jules L. Coleman, Moral Theories of Torts: Their Scope and Limits, 1 Law & Phil. 371 (1982) (Part I), 2 Law & Phil. 5 (1983) (Part II) [hereinafter Coleman, Moral Theories: Part I/II]; Jules L. Coleman & Jody Kraus, Rethinking the Theory of Legal Rights, 95 Yale L.J. 1335 (1986); Jules L. Coleman, Property, Wrongfulness and the Duty to Compensate, 63 Chi.-Kent L. Rev. 451 (1987) [hereinafter Coleman, Property]; Jules L. Coleman, The Structure of Tort Law, 97 Yale L.J. 1233, 1248-53 (1988) [hereinafter Coleman, Structure]; Jules L. Coleman, Tort Law and the Demands of Corrective Justice, 67 Ind. L.J. 349 (1992) [hereinafter Coleman, Demands]; Jules L. Coleman, The Mixed Conception of Corrective Justice, 77 Iowa L. Rev. 427 (1992) [hereinafter Coleman, Mixed Conception].
n7. See Ernest J. Weinrib, Aristotle's Forms of Justice, in Justice, Law and Method in Plato and Aristotle 133 (Spiro Panagiotou ed., 1987) [hereinafter Weinrib, Forms]; Ernest J. Weinrib, The Intelligibility of the Rule of Law, in The Rule of Law: Ideal or Ideology 59 (Allan C. Hutchinson & Patrick Monahan eds., 1987); Ernest J. Weinrib, Toward a Moral Theory of Negligence Law, 2 Law & Phil. 37 (1983) [hereinafter Weinrib, Negligence]; Ernest J. Weinrib, The Insurance Justification and Private Law, 14 J. Legal Stud. 681 (1985) [hereinafter Weinrib, Insurance]; Ernest J. Weinrib, Causation and Wrongdoing, 63 Chi.-Kent L. Rev. 407 (1987) [hereinafter Weinrib, Causation]; Ernest J. Weinrib, Law as a Kantian Idea of Reason, 87 Colum. L. Rev. 472 (1987) [hereinafter Weinrib, Kant]; Ernest J. Weinrib, Legal Formalism: On the Immanent Rationality of Law, 97 Yale L.J. 949 (1988) [hereinafter Weinrib, Formalism]; Ernest J. Weinrib, Liberty, Community, and Corrective Justice, 1 Can. J.L. & Jur. 3 (1988) [hereinafter Weinrib, Community]; Ernest J. Weinrib, Adjudication and Public Values: Fiss's Critique of Corrective Justice, 39 U. Toronto L.J. 1 (1989) [hereinafter Weinrib, Public Values]; Ernest J. Weinrib, Right and Advantage in Private Law, 10 Cardozo L. Rev. 1283 (1989) [hereinafter Weinrib, Hegel]; Ernest J. Weinrib, The Special Morality of Tort Law, 34 McGill L.J. 403 (1989) [hereinafter Weinrib, Special Morality]; Ernest J. Weinrib, Understanding Tort Law, 23 Val. U. L. Rev. 485 (1989) [hereinafter Weinrib, Tort Law]; Ernest J. Weinrib, Corrective Justice, 77 Iowa L. Rev. 403 (1992).
n8. See infra text accompanying notes 27-29.
n9. See infra text accompanying notes 25, 193-96, 204, 216, 244.
n10. See Steven D. Smith, Rhetoric and Rationality in the Law of Negligence, 69 Minn. L. Rev. 277 (1984); Steven D. Smith, The Critics and the "Crisis": A Reassessment of Current Conceptions of Tort Law, 72 Cornell L. Rev. 765 (1987); Catherine Pierce Wells, Tort Law as Corrective Justice: A Pragmatic Justification for Jury Adjudication, 88 Mich. L. Rev. 2348 (1990); cf. J.M. Balkin, The Rhetoric of Responsibility, 76 Va. L. Rev. 197 (1990).
n11. See Richard A. Epstein, A Theory of Strict Liability, 2 J. Legal Stud. 151 (1973); Richard A. Epstein, Defenses and Subsequent Pleas in a System of Strict Liability, 3 J. Legal Stud. 165 (1974); Richard A. Epstein, Causation and Corrective Justice: A Reply to Two Critics, 8 J. Legal Stud. 477 (1979); Richard A. Epstein, Intentional Harms, 4 J. Legal Stud. 391 (1975); Richard A. Epstein, Nuisance Law: Corrective Justice and its Utilitarian Constraints, 8 J. Legal Stud. 49 (1979); Richard A. Epstein, Causation -- In Context: An Afterword, 63 Chi.-Kent L. Rev. 653 (1987).
n12. See George P. Fletcher, Fairness and Utility in Tort Theory, 85 Harv. L. Rev. 537 (1972) [hereinafter Fletcher, Fairness]. See also George P. Fletcher, The Search for Synthesis in Tort Theory, 2 Law & Phil. 63 (1983); cf. George P. Fletcher, The Right and the Reasonable, 98 Harv. L. Rev. 949 (1985).
n13. For criticisms of Epstein's approach, see Coleman, Moral Theories: Part I, supra note 6, at 378-83; Izhak Englard, The System Builders: A Critical Appraisal of Modern American Tort Theory, 9 J. Legal Stud. 27, 57-63 (1980); Stephen R. Perry, The Impossibility of General Strict Liability, 1 Can. J.L. & Jur. 147 (1988); Gary T. Schwartz, The Vitality of Negligence and the Ethics of Strict Liability, 15 Ga. L. Rev. 963, 978-1005 (1981); Weinrib, Causation, supra note 7, at 416-29; Wright, Bramble Bush, supra note 3, at 1015-18; Wright, Causation, supra note 3, at 1750-58. For criticisms of Fletcher's approach, see Coleman, Moral Theories: Part I, supra, at 383-90; Coleman, Reciprocity, supra note 6; Englard, supra, at 63-68; Schwartz, supra, at 978-90.
n14. For other writings on corrective justice, see Compensatory Justice: Nomos XXXIII (John W. Chapman ed., 1991); John Finnis, Natural Law and Natural Rights 177-84 (1980); Neil MacCormick, The Obligation of Reparation, in Legal Right and Social Democracy 212-31 (1984); Larry A. Alexander, Causation and Corrective Justice: Does Tort Law Make Sense?, 6 Law & Phil. 1 (1987); Kathryn R. Heidt, Corrective Justice from Aristotle to Second Order Liability: Who Should Pay When the Culpable Cannot?, 47 Wash. & Lee L. Rev. 347 (1990); Heidi M. Hurd, Correcting Injustice to Corrective Justice, 67 Notre Dame L. Rev. 51 (1991); James W. Nickel, Justice in Compensation, 18 Wm. & Mary L. Rev. 379 (1976); Stephen R. Perry, Comment on Coleman: Corrective Justice, 67 Ind. L.J. 381 (1992)[hereinafter Perry, Comment]; Stephen R. Perry, The Moral Foundations of Tort Law, 77 Iowa L. Rev. 449 (1992) [hereinafter Perry, Moral Foundations]; Richard A. Posner, The Concept of Corrective Justice in Recent Theories of Tort Law, 10 J. Legal Stud. 187 (1981); Christopher H. Schroeder, Corrective Justice and Liability for Increasing Risks, 37 UCLA L. Rev. 439 (1990) [hereinafter Schroeder, Risks]; Schroeder, Corrective Justice, Liability for Risks, and Tort Law, 38 UCLA L. Rev. 143 (1990); Kenneth W. Simons, Corrective Justice and Liability for Risk Creation: A Comment, 38 UCLA L. Rev. 113 (1990).
n15. See Ronald Dworkin, "Natural" Law Revisited, 39 Fla. L. Rev. 165 (1982).
n16. See George C. Christie, Jurisprudence 2-3 (1973).
n17. See infra text accompanying notes 254-88.
n18. Weinrib himself notes the extremism of his version of formalism. See Weinrib, Formalism, supra note 7, at 1013.
n19. See Ernest J. Weinrib, The Case for a Duty to Rescue, 90 Yale L.J. 247 (1980) [hereinafter Weinrib, Rescue].
n20. See id. at 263 ("The role of the common-law judge centrally involves making moral duties into legal ones.").
n21. Id. at 275.
n22. Id. at 251; see id. at 258-67, 279-93. Weinrib hinted at his preference for the noninstrumental rights-based theories. See id. at 285-87; cf. Ernest J. Weinrib, Utilitarianism, Economics, and Legal Theory, 30 U. Toronto L.J. 307 (1980) (arguing that Richard Posner's wealth-maximization efficiency theory is subject to the criticisms of utilitarianism, including its incompatibility with the concept of rights) [hereinafter Weinrib, Utilitarianism].
n23. For Weinrib's substantive arguments, see Weinrib, Negligence, supra note 7, at 44, 54-55 & n.30; Weinrib, Utilitarianism, supra note 22.
n24. See Weinrib, Negligence, supra note 7, at 38-48.
n25. See id. at 39-40; Weinrib, Corrective Justice, supra note 7, at 412, 414-15, 417-18; Weinrib, Formalism, supra note 7, at 973-74, 979 & n.66, 981, 984-85, 987-88; Weinrib, Forms, supra note 7, at 138; Weinrib, Insurance, supra note 7, at 684-87; Weinrib, Special Morality, supra note 7, at 412; Weinrib, Tort Law, supra note 7, at 494-95.
n26. See Weinrib, Negligence, supra note 7, at 40-43, 49-54.
n27. See Weinrib, Kant, supra note 7, at 473-74, 501-03. The presumed radical disjunction between law and ethics in Kant's philosophy does not exist. See infra text accompanying notes 87-158.
n28. See Weinrib, Formalism, supra note 7, at 952-66, 977-83; Weinrib, Kant, supra note 7, at 473-74; Weinrib, Public Values, supra note 7, at 9, 11-14; cf. Weinrib, Special Morality, supra note 7, at 410-13 (immanent rationality of tort law).
n29. Weinrib, Formalism, supra note 7, at 956.
n30. See infra text accompanying note 61.
n31. Weinrib, Tort Law, supra note 7, at 491; see id. at 525-26; Weinrib, Formalism, supra note 7, at 992-95; Weinrib, Forms, supra note 7, at 151 ("In realizing itself as an articulation of corrective justice, private law fulfils the only function it can have, which is simply to be private law."); Weinrib, Insurance, supra note 7, at 686 ("the only function of the law of torts is to be the law of torts"); Weinrib, Public Values, supra note 7, at 12-13 ("the end of tort law is tort law").
n32. See Weinrib, Formalism, supra note 7, at 988-92, 995; Weinrib, Forms, supra note 7, at 140-41; infra text accompanying notes 159-65.
n33. See Weinrib, Causation, supra note 7, at 409-10, 444-45; Weinrib, Formalism, supra note 7, at 964-66; Weinrib, Tort Law, supra note 7, at 487-91, 493, 495-96.
n34. See Weinrib, Formalism, supra note 7, at 966-80, 995-99.
n35. See id. at 966-71; Weinrib, Tort Law, supra note 7, at 491-95.
n36. See Weinrib, Tort Law, supra note 7, at 490-92, 497 n. 28.
n37. See Weinrib, Formalism, supra note 7, at 974-75.
n38. Any such attempt is hazardous. See W. Page Keeton et al., Prosser and Keeton on the Law of Torts ¤ 1, at 1-5 (5th ed. 1984) [hereinafter Prosser & Keeton]. However, tort law (rather than torts per se) can be described roughly as that area of law which governs law suits, other than those seeking to enforce promises made by the defendant, in which a private plaintiff (including the government in its capacity as a property owner) seeks damages from or equitable relief against a defendant whose behavior has caused or may cause injury to the person or property of the plaintiff.
n39. See Weinrib, Special Morality, supra note 7, at 406; Weinrib, Tort Law, supra note 7, at 491-94.
n40. See Weinrib, Causation, supra note 7, at 409-10, 429-30, 439-44; Weinrib, Community, supra note 7, at 4-5, 11-12; Weinrib, Formalism, supra note 7, at 966-67, 969-70, 978; Weinrib, Insurance, supra note 7, at 683.
n41. See Prosser & Keeton, supra note 38, ¤¤ 43-44, at 280-319, ¤¤ 75-79, at 534-60; Weinrib, Tort Law, supra note 7, at 494-95, 514-15; cf. Weinrib, Negligence, supra note 7, at 40, 57-62 (noting strict liability and fault liability as competing bases of tort liability). Weinrib has even narrowed the list to causation alone, stating that "causation is constitutive of tort law as an identifiable field." See Weinrib, Special Morality, supra note 7, at 407-08.
n42. Weinrib, Tort Law, supra note 7, at 495.
n43. Id. at 495, 496, 511.
n44. See id. at 495-97, 511-14.
n45. See Wright, Bramble Bush, supra note 3, at 1018-23; Wright, Causation, supra note 3, at 1788-1803.
n46. Weinrib, Special Morality, supra note 7, at 407-08; see Weinrib, Causation, supra note 7, at 415; Weinrib, Public Values, supra note 7, at 9-10, 13-14; Weinrib, Tort Law, supra note 7, at 494, 511.
n47. See Weinrib, Formalism, supra note 7, at 957: "Since the point of my entire exposition of formalism is to present an affirmative conception of the juridical, I can indicate the significance of the term [juridical] at this preliminary [sic] only negatively."
n48. See Weinrib, Tort Law, supra note 7, at 511-12, 513.
n49. See Weinrib, Causation, supra note 7, at 416-20; Weinrib, Community, supra note 7, at 15-16; Weinrib, Tort Law, supra note 7, at 524.
n50. See supra note 41 and accompanying text.
n51. See Weinrib, Tort Law, supra note 7, at 495, 514-15.
n52. See Weinrib, Special Morality, supra note 7, at 406-10; Weinrib, Tort Law, supra note 7, at 514-24.
n53. Cf. Weinrib, Causation, supra note 7, at 417 (noting that the tort requirement of volitional conduct cannot be explained by causal concepts or arguments); Weinrib, Negligence, supra note 7, at 58-59 (same).
n54. Weinrib assumes that the concepts of causation, volitional conduct (doing), and misfeasance apply only to affirmative acts, not to omissions. See Weinrib, Causation, supra note 7, at 417, 430; Weinrib, Formalism, supra note 7, at 978-79; Weinrib, Tort Law, supra note 7, at 516-17; infra note 61. This is incorrect. Any necessary or NESS condition for the occurrence of some effect is a cause of that effect, whether the condition is an affirmative or a negative one. See Wright, Bramble Bush, supra note 3, at 1016 & n.83, 1019. For example, scientists are well aware that it is critical for many physical or chemical processes not only that certain elements be present, but that others be absent, and both types of conditions are treated as part of the total cause of the particular effect. The law also correctly treats omissions as well as affirmative acts as causes, as volitional, and as possible misfeasance. See, e.g., Rogers v. Board of Road Comm'rs, 30 N.W.2d 358 (Mich. 1948) (failure to remove items from land after license or privilege expires constitutes trespass); Prosser & Keeton, supra note 38, ¤ 8, at 34 & n.6, ¤ 56, at 373-77.
n55. See supra note 49 and accompanying text.
n56. See Weinrib, Tort Law, supra note 7, at 512, 515, 518.
n57. See Weinrib, Formalism, supra note 7, at 978, 983 n.74.
n58. See id. at 982 n.73.
n59. See infra text accompanying notes 61-76.
n60. See supra note 31 and accompanying text.
n61. See Weinrib, Community, supra note 7, at 3-5; Weinrib, Formalism, supra note 7, at 976-80 & n.61, 982-83, 989-90, 994-95; Weinrib, Insurance, supra note 7, at 683-87. Although doing can include omissions as well as affirmative acts, and suffering often occurs to active as well as passive parties, Weinrib construes doing and suffering as activity and passivity, respectively. Because of this narrow construction, he is unable to treat the obligation of restitution, in cases involving the unintended conferral of benefits (e.g., mistaken overpayments), as being based on the actual doing and suffering of harm, since the unintended beneficiary is the passive party while the mistaken benefactor is the active party. Once one understands that volitional omissions as well as affirmative acts count as doings, there is no problem in characterizing the unintended beneficiary's failure to return the unintended benefit as the doing of harm to the mistaken benefactor. Weinrib relies instead on a string of non sequiturs to connect the obligation of restitution to the concept (rather than the actuality) of doing and suffering harm. See Weinrib, Formalism, supra, at 978-79; cf. supra note 54 (omissions as causes).
n62. See Weinrib, Community, supra note 7, at 3 n.3 (contract law, tort law, and restitution, but not criminal law, included in list of areas of private law); Weinrib, Formalism, supra note 7, at 982 & n.73 (noting the complexity of trying to fit criminal law into his distinction between private and public law); id. at 972 (suggesting that the alleged internal coherence of tort law "can be found throughout private law (and perhaps beyond)").
n63. See supra text accompanying note 58.
n64. See infra notes 380-81 and text accompanying notes 378-81.
n65. See Weinrib, Formalism, supra note 7, at 977-81; infra text accompanying notes 95, 307-14.
n66. See Weinrib, Corrective Justice, supra note 7, at 408 & nn. 20 & 21; Weinrib, Formalism, supra note 7, at 980-81 & n.71, 983; Weinrib, Forms, supra note 7, at 135-36 & n.9, 137, 139. But see Weinrib, Corrective Justice, supra, at 404 ("fairness as a norm is inseparable from equality as a mathematical function"); id. at 414 & n.40 (invoking Aquinas's literal interpretations); Weinrib, Formalism, supra, at 980 & n.69 (describing Aquinas's literal interpretations of arithmetical proportion and geometrical proportion as the "distinct notions of equality" employed in corrective justice and distributive justice, respectively); Weinrib, Forms, supra, at 146 ("[corrective justice's] arithmetic proportion yields a single sum whereas [distributive justice's] geometric proportion expresses an infinitely continuable ratio"); infra text accompanying notes 74-76.
n67. See infra note 319 and text accompanying notes 18-23, 352-53.
n68. See supra text accompanying notes 44-56.
n69. See Weinrib, Formalism, supra note 7, at 981, 982, 983.
n70. See Weinrib, Formalism, supra note 7, at 979-82.
n71. See id. at 981, 990, 991.
n72. A prime number is a whole number that is not divisible, without a remainder, by any number other than itself or the number one -- i.e., 1, 2, 3, 5, 7, 11, etc.
n73. See id. at 977, 979, 982-85, 1010; Weinrib, Corrective Justice, supra note 7, at 416; Weinrib, Forms, supra note 7, at 146-47.
n74. See Weinrib, Community, supra note 7, at 5; Weinrib, Corrective Justice, supra note 7, at 407, 415-16; Weinrib, Formalism, supra note 7, at 981, 983-84; Weinrib, Forms, supra note 7, at 140, 146.
n75. For example, (6 + 2)/2 = (5 + 4 + 3)/3 = (2 + 5 + 5)/3 = 4 is an equality of arithmetical proportions (averages), while 12/3 = 8/2 = 4 is an equality of geometrical proportions (ratios). Moreover, as these examples illustrate, a geometrical proportion as well as an arithmetical proportion yields a single number (in these examples, the number four). Compare Weinrib's contrary implication, supra note 66.
n76. For example, quantitative equality is maintained, despite the addition or splitting of terms, when expanding a = a to a + b = a + b or to a = a/2 + a/2, respectively. Similarly, when c is taken from a in a + b so that we have (a-c) + b, the original quantity a + b can be maintained or restored by adding c to b rather than to a, so that we end up with (a-c) + (b + c) rather than (a-c + c) + b.
n77. See Weinrib, Causation, supra note 7, at 448-50; Weinrib, Formalism, supra note 7, at 953-55, 990, 995-99; Weinrib, Forms, supra note 7, at 136-40 & n.18; Weinrib, Special Morality, supra note 7, at 408, 409, 410-11. Weinrib's most recent writings also refer to Hegel's concept of Right. See Weinrib, Corrective Justice, supra note 7, at 422-25; Weinrib, Hegel, supra note 7.
n78. See Weinrib, Formalism, supra note 7, at 990, 996; Weinrib, Special Morality, supra note 7, at 408-09.
n79. See supra text accompanying notes 70-72.
n80. See Weinrib, Community, supra note 7, at 7-8; Weinrib, Corrective Justice, supra note 7, at 421-25; Weinrib, Formalism, supra note 7, at 996-98; Weinrib, Forms, supra note 7, at 134-37, 139, 149-50; Weinrib, Special Morality, supra note 7, at 408-11.
n81. Weinrib, Forms, supra note 7, at 149.
n82. See Weinrib, Special Morality, supra note 7, at 405-06 (treating recognition of excuses as incompatible with the immanent rationality of tort law, but not of criminal law, although both tort and criminal law are assumed elsewhere to be expressions of corrective justice and therefore should share the same immanent rationality).
n83. See supra text accompanying notes 44-49.
n84. See Weinrib, Community, supra note 7, at 8; Weinrib, Formalism, supra note 7, at 997-98; Weinrib, Special Morality, supra note 7, at 409, 411.
n85. See infra text accompanying notes 90-92.
n86. Weinrib, Forms, supra note 7, at 139.
n87. Weinrib, Formalism, supra note 7, at 997.
n88. Weinrib, Forms, supra note 7, at 136.
n89. See Weinrib, Community, supra note 7, at 8, 12, 16-17; Weinrib, Corrective Justice, supra note 7, at 421-23; Weinrib, Formalism, supra note 7, at 996-98 & n.100; Weinrib, Forms, supra note 7, at 135-37, 139, 140 n.18, 149; Weinrib, Kant, supra note 7, at 473-74, 487-91, 495-97, 501-03. Relying on this assumed distinction, Weinrib attempts to infer the Kantian concept of Right directly from the alleged lack of any determinate conception of moral or political good in Aristotle's elaboration of the concepts of corrective justice and distributive justice. See Weinrib, Formalism, supra, at 996. Yet how could anything as rigorous as Kantian Right be inferred merely from the absence of any determinate conception of the moral or political good?
n90. See Immanuel Kant, The Metaphysics of Morals * 213-14, 221-23, 225-27, 379-80 & n., 383, 394, 397, 405 (Mary Gregor trans., Cambridge Univ. Press 1991) (1797) [hereinafter Kant, Morals].
n91. Id. at * 225-26.
n92. Id. at * 434-35; see id. at * 223, 237-38.
n93. See id. at * 218-20, 379-80, 396-97, 406.
n94. Id. at * 231.
n95. Id. at * 395.
n96. See id. at * 231-32, 239.
n97. Id. at * 231.
n98. See id. at * 219-20, 239, 381.
n99. As Kant notes, in classical philosophy "ethics" encompassed all of moral philosophy, while in modern moral philosophy "ethics" is reserved for duties that are not subject to any external lawgiving -- i.e., obligations of virtue. See id. at * 379.
n100. Id. at * 232.
n101. See id. at * 214, 218-21, 231-32, 239, 394-95.
n102. See id. at * 239, 380-81, 394-96, 410.
n103. See id. at * 375, 381, 396.
n104. See id. at * 390-91.
n105. Id. at * 410; see id. at * 383 ("To every ethical obligation there corresponds the concept of virtue, but not all ethical duties are thereby duties of virtue. Those duties that have to do not so much with a certain end (matter, object of choice) as merely with what is formal in the moral determination of the will (e.g., that an action in conformity with duty must also be done from duty) are not duties of virtue. Only an end that is also a duty can be called a duty of virtue."); id. at * 394-95. As Mary Gregor points out, Kant's distinction between the one obligation of virtue (the virtuous disposition to conform to duty from a sense of duty) and the many duties of virtue (as particular ends) is problematic, since one of the principal duties of virtue specified by Kant is one's own moral perfection, which includes respect for duty in general, including duties of Right. See id. at * 387, 390-91, 446; id. at 22-23 (Gregor's Introduction).
n106. See id. at * 239, 379, 381, 394; Warner Wick, Introduction, in Immanuel Kant, The Metaphysical Principles of Virtue xlii-xliii (James Ellington trans., Bobbs-Merrill Co. 1964) (1797). The distinction between ethical obligations in the broad sense and duties of virtue in the strict sense is illustrated by Kant's discussion of contractual obligation:
Ethics commands that I still fulfill a contract I have entered into, even though the other party could not coerce me to do so; but it takes the law (pacta sunt servanda) and the duty corresponding to it from the doctrine of Right, as already given there. . . . For if this were not the case, and if the lawgiving itself were not juridical so that the duty arising from it was not really a duty of Right (as distinguished from a duty of virtue), then faithful performance (in keeping with promises made in a contract) would be put in the same class with actions of benevolence and the obligation to them, and this must not happen. It is no duty of virtue to keep one's promises but a duty of Right, to the performance of which one can be coerced. But it is still a virtuous action (a proof of virtue) to do it even where no coercion may be applied. The doctrine of Right and the doctrine of virtue [ethical obligation in general] are therefore distinguished not so much by their different duties as by the difference in their lawgiving, which connects one incentive or the other with the law.
Ethical lawgiving (even if the duties might be external) is that which cannot be external; juridical lawgiving is that which can also be external. So it is an external duty to keep a promise made in a contract; but the command to do this merely because it is a duty, without regard for any other incentive, belongs to internal lawgiving alone. So the obligation is assigned to ethics not because the duty is of a particular kind (a particular kind of action to which one is bound) -- for there are external duties in ethics as well as in Right -- but rather because the lawgiving in this case is an internal one and can have no external lawgiver. . . . Ethics has its special duties as well (e.g., duties to oneself), but it also has duties in common with Right; what it does not have in common with Right is only the kind of obligation.
Kant, Morals, supra note 90, at * 219-20; see infra note 110.
n107. See supra text accompanying notes 90-95.
n108. See Kant, Morals, supra note 90, at * 331, 399-401, 405, 417-18.
n109. See id. at * 213-14, 218-19, 225-27.
n110. See id. at * 239 ("We know our own freedom (from which all moral laws, and so all rights as well as duties proceed) only through the moral [categorical] imperative, which is a proposition commanding duty, from which the capacity for putting others under obligation, that is, the concept of a right, can afterward be explicated."); John Ladd, Introduction, in Immanuel Kant, The Metaphysical Elements of Justice ix-xi (John Ladd trans., Bobbs-Merrill Co. 1965) (1797); Wick, supra note 106, at xliv-xlv. Kant himself derives the duty to keep one's promises, which is a duty of Right rather than a duty of virtue, directly from the categorical imperative. See Immanuel Kant, Foundations of the Metaphysics of Morals * 421-24, 428-31 (Lewis Beck trans., Bobbs-Merrill Co. 1959) (1785); supra note 106.
n111. See Weinrib, Community, supra note 7, at 6-7, 8; Weinrib, Formalism, supra note 7, at 997-98; Weinrib, Kant, supra note 7, at 481-91.
n112. See Immanuel Kant, Lectures on Ethics 50-51, 193-95, 211 (Louis Infield trans., Hackett Pub. Co. 1980) (1930) [hereinafter Kant, Lectures]. But cf. id. at 211 ("At the same time the duties [to others] dictated by right or generosity are inferior to the duties we owe to ourselves.").
n113. See Kant, Morals, supra note 90, at * 218-22, 232, 390-91.
n114. Id. at * 232.
n115. See supra text accompanying notes 80-83. The problems raised earlier reappear here. If, as Weinrib states, the Kantian and Hegelian concepts of Right abstract from all particularity, then how can Right be concerned with the particularity of injury? For a more complete discussion of this problem, see Perry, Moral Foundations, supra note 14, at 480-88. Moreover, if we consider only the external effects on others, and not any internal or personal attributes, are we not driven to absolute causal liability? Weinrib argues that absolute liability would be incompatible with freedom of action. However, absolute liability would not prohibit action, but rather merely require compensation for the external effects of one's free action. Cf. Weinrib, Community, supra note 7, at 14-15 (attempting to rebut this counterargument with an inapposite construal of liability as a tax).
n116. See supra text accompanying note 98.
n117. See supra text accompanying note 83; cf. H.L.A. Hart, The Concept of Law 168-69, 173-75 (1961) (criticizing the attempt to distinguish morality from law by a dichotomous focus on internal versus external conditions).
n118. See Kant, Morals, supra note 90, at * 224, 227-28. Kant states that these concepts are common to both the doctrine of Right and the doctrine of virtue. See id. at * 222.
n119. Id. at * 236-37 (translator's note omitted).
n120. See id. at * 224: "What is right in accordance with external laws [which are obligatory laws for which there can be an external lawgiving -- that is, duties of Right] is called just (iustum); what is not, unjust (iniustum)."
n121. Id. at * 306; see id. at * 267.
n122. See Weinrib, Community, supra note 7, at 6-7, 8, 12; Weinrib, Kant, supra note 7, at 491-500.
n123. Indeed, Kant notes, if this were not so there could be no Right, public or private. See infra note 133.
n124. See Kant, Morals, supra note 90, at * 306.
n125. See id. at * 237.
n126. See supra text accompanying note 97.
n127. See supra text accompanying note 121; Kant, Morals, supra note 90, at * 253, 255-57, 312.
n128. See Kant, Morals, supra note 90, at * 246-47, 250, 262-63.
n129. See id. at * 250-53, 263-65, 269.
n130. Kant's elaboration of the doctrine of Right, like Aristotle's Politics, focuses on the institutional aspect of distributive justice: the creation and distribution of political offices and power. Kant sidesteps the more inclusive concept of distributive justice set forth in Aristotle's Ethics, which encompasses material resources as well as political power, by reading the precept "give to each what is his [due]" as a tautology. See supra text accompanying note 119; infra text accompanying notes 308-10. Yet Kant did not entirely exclude the broader distributional issues. His discussions of the problem of original acquisition are sensitive to these issues. See infra text accompanying notes 133-34. He also notes that the duty of beneficence -- an imperfect duty of virtue with regard to unilateral action -- implicates Right, since beneficence is necessitated by the nonrightful creation of inequalities of wealth, through excessive acquisitions by individuals or unjust distributions by government. See Kant, Morals, supra note 90, at * 454; Kant, Lectures, supra note 112, at 192-95. In his discussion of public Right, Kant notes the right of the people to tax the wealthy to provide for the needy. See Kant, Morals, supra note 90, at * 326.
n131. Kant, Morals, supra note 90, at * 306; see id. at * 256, 312-13. Kant recognizes that, in addition to the a priori principles of private Right (natural law or, more accurately, natural Right), which are binding even in the absence of any external lawgiving, public Right also will contain substantial elements of conventional or positive law (such as traffic coordination rules and the formal actions required to indicate a binding promise or a transfer of property), which come into being as a result of external lawgiving by the universal will of the civil society. See id. at * 224, 237. Furthermore, Kant notes that, due to a need for secure rights that can be strictly and objectively enforced, the judicial elaboration of public Right in the civil society may vary from private Right in the state of nature. See id. at * 297-305. Kant's emphasis on the mathematical rigor and determinacy of strict Right contrasts sharply with Weinrib's embrace of indeterminacy and historical contextuality in his theory of law. Compare id. at * 233-35, 390, 411 with Weinrib, Formalism, supra note 7, at 1008-12; Weinrib, Kant, supra note 7, at 505-07.
n132. See supra note 128.
n133. See Kant, Morals, supra note 90, at * 256-57, 261-64, 266-67. Indeed, Kant notes, "If no acquisition were recognized as rightful even in a provisional way prior to entering the civil condition, the civil condition itself would be impossible. For in terms of their form, laws concerning what is mine or yours in the state of nature contain the same thing that they prescribe in the civil condition, insofar as the civil condition is thought of by pure rational concepts alone. The difference is only that the civil condition provides the conditions under which these laws are put into effect (in keeping with distributive justice). So if external objects were not even provisionally mine or yours in the state of nature, there would also be no duties of Right with regard to them and therefore no command to leave the state of nature." Id. at * 312-13.
n134. See id. at * 255-57, 264, 268, 312. Thus, the possibility of rightful possession depends on protective justice, the actuality of rightful possession is governed by corrective justice, and the necessity of rightful possession necessitates distributive justice (entry into the civil condition). See supra text accompanying note 121.
n135. See supra text accompanying notes 125-29.
n136. Kant, Morals, supra note 90, at * 312; see id. at * 257.
n137. See Weinrib, Kant, supra note 7, at 503.
n138. Kant, Morals, supra note 90, at * 307-08. Kant adds in a footnote:
This distinction between what is merely formally wrong and what is also materially wrong has many applications in the doctrine of Right. An enemy who, instead of honorably carrying out his surrender agreement with the garrison of a besieged fortress, mistreats them as they march out or otherwise breaks the agreement, cannot complain of being wronged if his opponent plays the same trick on him when he can. But in general they do wrong in the highest degree, because they take away any validity from the concept of Right itself and hand everything over to savage violence, as if by law, and so subvert the Right of men as such.
Id. at * 308 n.. In Kant's moral philosophy, the formal aspect of duty in the doctrines of Right or virtue is the required conformity with the principle of external freedom (consistency with a like freedom for everyone else) or internal freedom (conformance with duty from duty), respectively. The material aspect of duty is the end or object of choice, which distinguishes duties of virtue. See id. at * 380, 383, 394-95, 398, 410; supra note 105.
n139. Id. at * 307; see id. at * 256, 312.
n140. See supra text accompanying note 121.
n141. See supra text accompanying note 119.
n142. See supra text accompanying notes 125-29, 140.
n143. See supra text accompanying note 141.
n144. See Kant, Lectures, supra note 112, at 49-50.
n145. Kant, Morals, supra note 90, at * 240; see supra text accompanying note 119.
n146. See supra note 106; see also Kant, Lectures, supra note 112, at 117: "My duty towards myself cannot be treated juridically; the law touches only our relations with other men; I have no legal obligations towards myself; and whatever I do to myself I do to a consenting party; I cannot commit an act of injustice against myself."
n147. See supra text accompanying notes 102-06; Kant, Morals, supra note 90, at 26, 31 n.18 (Gregor's Introduction); Immanuel Kant, The Metaphysical Elements of justice * 236 n.7 (translator's note) (John Ladd trans., Bobbs-Merrill Co. 1965) (1797).
n148. See Kant, Morals, supra note 90, at * 386-87, 390-93, 410, 419-26, 429-30, 444-47.
n149. Id. at * 429; see id. at * 238 n. ("The only kind of untruth we want to call a lie, in the sense bearing upon rights, is one that directly infringes upon another's right . . ."); id. at * 430 ("Lying (in the ethical sense of the word), intentional untruth as such, need not be harmful to others in order to be repudiated; for it would then be a violation of the rights of others.")
n150. Id. at * 436; see id. at * 420, 434-36, 460-61.
n151. See, e.g., J.M. Finnis, Legal Enforcement of "Duties to Oneself": Kant v. Neo-Kantians, 87 Colum. L. Rev. 433 (1987). Finnis relies on two passages. The first is Kant's explicit labeling of the first division, which encompasses perfect duties to oneself based on the Right of humanity in one's own person, as a division of the duties of Right. The second is Kant's discussion of the appropriate punishment in Right for bestiality, which Finnis interprets to be sexual intercourse with a nonhuman animal. See id. at 447-49 & n.68, discussing Kant, Morals, supra note 90, at * 240, 363. However, Kant does not define "bestiality," which could simply mean deviant sex between humans. See 2 The Oxford English Dictionary 142 (2d ed. 1989) ("bestial" and "bestiality"). Only a few pages earlier, Kant seems to have had the latter meaning in mind. He states that a man and a woman cannot engage in sexual intercourse outside of marriage "without both renouncing their personalities (in carnal or bestial cohabitation)." Kant, Morals, supra, at * 359. This passage is a comment on Kant's prior discussion of the duty of Right not to violate the Right of humanity in one's person by engaging in sexual relations with another outside of marriage. See id. at * 277-78; cf. id. at * 270 (referring to the Right of humanity in one's person as preventing one from disposing of oneself as one pleases). In the doctrine of virtue, Kant states: "In the doctrine of Right it was shown that man cannot make use of another person to get [sexual] pleasure apart from a special limitation by a [marriage] contract establishing the right." Id. at * 424 (emphasis in original). He distinguishes the prohibition in Right of such "mere debasing" of the humanity in one's person from the prohibition in the doctrine of virtue of the "defiling" of oneself through unnatural lustful "use (and so misuse) of one's sexual attribute," which is an action so "loathsome" "that it is considered indecent even to call this vice [presumably masturbation] by its proper name." Id. at * 424-25. Thus, Kant seems to require, consistent with his general statements on Right, that a duty of Right to oneself involve an interaction with another (that degrades the humanity of both parties). Finnis's attempt to explain away Kant's restricting the concept of Right to "the external and indeed practical relation of one person to another," by treating the external enforcer of the alleged noninteractional duty of Right as the "other" with whom one is interacting, seems strained. See Finnis, supra, at 450-51, discussing Kant, Morals, supra, at * 230.
n152. See Kant, Morals, supra note 90, at * 390, 411.
n153. See id. at * 421, 428, 437.
n154. Note that the particular duties of Right to oneself that are discussed or mentioned by Kant, see supra note 151, all can easily be specified as perfect duties, while the duty not to let others tread with impunity on your rights cannot.
n155. See supra note 151. Finnis interprets duties of Right to oneself in this way. However, I think he goes too far. He would read all of Kantian morality into the phrase "universal law" in the supreme principle of Right, thus apparently making any external action that violates a duty of virtue -- even an imperfect duty -- also a violation of a duty of Right. See Finnis, supra note 151, at 449-50.
n156. Although an external, impartial judge seems to be indispensable to justice as a juridical relation, it is inapposite to justice as a virtue. On the other hand, the external judge's proper fulfillment of her juridical role depends on her having internalized justice as a virtue: the substantive content of justice as a juridical relation is the same as the substantive content of justice as a virtue. Both Weinrib and Heyman miss these subtleties by conflating justice as a virtue with justice as a juridical relation. Weinrib asserts that the concept of corrective justice or Right abstracts from all internal attributes, including virtue, and hence conceptually implies an external judge. Weinrib, Forms, supra note 7, at 149-50; Weinrib, Kant, supra note 7, at 491, 496-97. Conversely, Heyman claims that corrective justice, at least in Aristotle's account, is a virtue primarily or paradigmatically not of the participants in an interaction, but rather of the judge, which supposedly implies that corrective justice is primarily or paradigmatically a matter of public virtue rather than private right or private virtue. See Steven J. Heyman, Aristotle on Political Justice, 77 Iowa L. Rev. 851, 858-59 (1992). For Aristotle's elaboration of justice as a virtue of the participants in an interaction, see infra text accompanying notes 254-76, 289-306, 324-42, 353-56.
n157. See supra text accompanying notes 90-95. For Kant's elaboration of the purpose of law and politics as guaranteeing respect for each person's absolute moral worth in external, practical relations, see Kant, Morals, supra note 90, at * 311-18; Hans Reiss, Introduction, in Kant's Political Writings 18-23 (Hans Reiss ed. & H.B. Nisbet trans., 1970); supra text accompanying notes 119-41. The linkage of law and ethics in Kant's moral philosophy calls into question Weinrib's implicit assumption that Kant would reject a juridical duty of easy rescue. See Weinrib, Community, supra note 7, at 6-7; Weinrib, Formalism, supra note 7, at 998-99; Weinrib, Kant, supra note 7, at 488-89. Kant's refusal to translate the ethical duty of general beneficence into a juridical duty is based not on any alleged radical disjunction between law and ethics, but rather on the broad (imperfect) nature of the duty of general beneficence. See Kant, Morals, supra, at * 390, 393, 452-54. As Weinrib himself once argued, insofar as a limited duty of easy rescue can be articulated as a strict rather than a broad duty to others, it would seem to be enforceable as a juridical duty under Kant's moral philosophy. See Weinrib, Rescue, supra note 19, at 287-92.
n158. See Weinrib, Forms, supra note 7, at 136.
n159. See Weinrib, Formalism, supra note 7, at 990-91; Weinrib, Kant, supra note 7, at 484-88.
n160. See supra note 89 and accompanying text.
n161. Weinrib also has attempted to infer the absolute equality of the interacting parties in corrective justice from the concept of property alone. See Weinrib, Causation, supra note 7, at 425-29, criticized in Coleman, Property, supra note 6, at 456-60.
n162. See supra text accompanying notes 71-72.
n163. See, e.g., Weinrib, Formalism, supra note 7, at 988-90, 998; Weinrib, Forms, supra note 7, at 140-41.
n164. As Weinrib notes, Kant himself says little about distributive justice in the broader (material resources) sense. See Weinrib, Kant, supra note 7, at 496 n.79; supra note 130.
n165. See Weinrib, Corrective Justice, supra note 7, at 417; Weinrib, Formalism, supra note 7, at 965, 988-93; Weinrib, Forms, supra note 7, at 140-42.
n166. See supra text accompanying notes 57-60.
n167. Weinrib, Formalism, supra note 7, at 999; see id. at 996-99.
n168. See id. at 983-85; Weinrib, Forms, supra note 7, at 146.
n169. See Weinrib, Formalism, supra note 7, at 953-61.
n170. See id. at 985 (footnote omitted); supra text accompanying note 25.
n171. See Weinrib, Formalism, supra note 7, at 987.
n172. See id. at 992.
n173. See id. at 1000-08 (where Weinrib's defense is based on the assumption that interactions and distributions are distinguishable independently of the adjudicative process, which is limited to elaborating their immanent rationality as interactions or distributions).
n174. See Coleman, Mixed Conception, supra note 6, at 427 & n.5, 437; Coleman, Property, supra note 6, at 460-61; Coleman, Wrongful Gain, supra note 6, at 425-27; Coleman & Kraus, supra note 6, at 1342-43.
n175. But cf. infra note 199 and text accompanying notes 197-99, 252 (discussing Coleman's references to "weak retributivism" and "outcome responsibility").
n176. See infra note 211 and text accompanying notes 178-79, 211-28.
n177. See supra text accompanying notes 77-173.
n178. Coleman, Mixed Conception, supra note 6, at 429.
n179. See Coleman, Demands, supra note 6, at 351-57, 361-62 & n.12, 364-66; Coleman, Moral Theories: Part II, supra note 6, at 6-7, 10-14, 35-36; Coleman, Property, supra note 6, at 460-61, 463-65; Coleman, Structure, supra note 6, at 1248-50; Coleman, Wrongful Gain, supra note 6, at 422-27. But see Wrongful Gain, supra, at 425 n.10 (criminal penalty to annul wrong does not annul wrongful gain).
n180. See infra text accompanying notes 237-39.
n181. See Coleman, Wrongful Gain, supra note 6, at 431-32, 434-36. For further discussion of this assumption, see Perry, Comment, supra note 14, at 389-91.
n182. See MacCormick, supra note 14, at 216-17, 219, 225-26; infra text accompanying notes 311-23, 357-64.
n183. See, e.g., Sugarman, supra note 1, at 608-09.
n184. See, e.g., Posner, supra note 14, at 197-99. Posner at one time claimed that the traditional Aristotelian concept of corrective justice is not only compatible with, but also required by the efficiency theory of liability. Id. at 201. But, in making this claim, Posner did what he chided others for. See id. at 192, 194-95, 198, 200. He misread, downplayed, or rejected essential parts of Aristotle's concept. See id. at 190-91, 200 (treating corrective justice as a method of correcting departures from distributive justice, rather than as an independent principle of justice with a distinct domain and a distinct criterion); id. at 190-91, 192, 203 (treating Aristotle's distinction between the absolute-equality criterion of corrective justice and the proportional-equality criterion of distributive justice as "procedural" rather than substantive); id. at 191, 194, 202 (treating Aristotle's elaboration of corrective justice as encompassing correlative rectificatory rights and duties between particular victims and their respective injurers as "[historical] background rather than as a central principle of corrective justice"); id. at 197, 201, 206 (stating that Aristotle did not explain the purpose of corrective justice, which allegedly is maximizing social wealth, whereas Aristotle explicitly rejected wealth-maximization as a good and articulated a nonaggregative purpose for politics, justice, and law); cf. infra text accompanying notes 254-76, 311-23, 343-70 (exposition of Aristotle's theory of justice). More recently, Posner seems to have abandoned the notion that corrective justice and efficiency are identical. See Posner, What Has Pragmatism to Offer Law?, 63 S. Cal. L. Rev. 1653, 1657, 1662-63 (1990) (dismissing Aristotelian corrective justice as backward looking and allegedly not directed toward any goal, and principled approaches to tort and criminal law as metaphysical and formal, while extolling instrumentalism's forward-looking posture).
n185. See, e.g., MacCormick, supra note 14, at 216-17, 219, 225-26; Hurd, supra note 14, at 56, 93-95; Perry, Moral Foundations, supra note 14, at 449-50; Weinrib, Negligence, supra note 7, at 38-39. But see Schroeder, Risks, supra note 14, discussed infra note 363.
n186. See Coleman, Mixed Conception, supra note 6, at 436; Coleman, Wrongful Gain, supra note 6, at 431-32, 435-36.
n187. See Coleman, Moral Theories: Part I, supra note 6, at 379, 383.
n188. See id. at 383-84; Coleman, Demands, supra note 6, at 351; Coleman, Reply to Pilon, supra note 6, at 311 & n.25.
n189. See Fletcher, Fairness, supra note 12, at 541 & n.15, 553-54. Fletcher suggests that those with such unenforceable rights "might have a claim of priority in a social insurance scheme." Id. at 554.
n190. See Karl Llewellyn, The Bramble Bush 85-86 (2d. ed. 1951).
n191. See MacCormick, supra note 14, at 218-19, 223.
n192. See infra text accompanying notes 240-44.
n193. See Coleman, Fault, supra note 6; Coleman, No-Fault, supra note 6. Coleman has maintained a "pretheoretical commitment" to defending no-fault insurance schemes from charges of injustice. See Coleman, Mixed Conception, supra note 6, at 428 n.8, 444.
n194. See Coleman, Fault, supra note 6, at 487:
Though there may indeed be a principle of justice endorsing penalties for wrongdoers (retributive) and another requiring compensation for faultless victims (compensatory), there is no principle of justice requiring that faultless victims be compensated by precisely those persons who have been at fault in causing them harm.
See also id. at 474-75, 484, 488 (distinguishing between punishing wrongdoers and compensating those who suffer accidental losses); infra note 195.
n195. Coleman, No-Fault, supra note 6, at 172; see id. at 167-69 (assuming that the prevalent moral argument for the "fault system" is a retributive argument that assumes morally blameworthy conduct by injurers); Coleman, Fault, supra note 6, at 473-74 (same).
n196. See Coleman, Fault, supra note 6, at 474-88; Coleman, No-Fault, supra note 6, at 162-72.
n197. See Joel Feinberg, Sua Culpa, in Doing & Deserving 216-21 (1970).
n198. Coleman, Fault, supra note 6, at 489.
n199. See id. at 488-90; Coleman, Mental Abnormality, supra note 6, at 118-21, 123-24; Coleman, No-Fault, supra note 6, at 173-74 & n.18; Coleman, Strict Liability, supra note 6, at 283-84. The notion that shifting the loss from the innocent victim to the faulty injurer is necessary to "protect the victim's innocence" is more than a little opaque. Whatever moral content it has would seem to be totally drained when the notion of fault is expanded by Coleman to include objective and not just subjective fault, especially when the standard of objective fault is left totally open and can encompass any criteria, including efficiency criteria. See infra text accompanying notes 211-12.
n200. See Coleman, Fault, supra note 6, at 487-88; supra note 194.
n201. See Coleman, Fault, supra note 6, at 474-75, 488.
n202. See infra text accompanying notes 357-64, 378-81.
n203. See Roger Pilon, Justice and No-Fault Insurance, 57 The Personalist 82, 84-86 (1976); see also MacCormick, supra note 14, at 214-18, 221-24. In his reply to Pilon, Coleman claimed that he had not intended to mount an affirmative defense of no-fault schemes or to reject the fault system in his first article, but rather had only intended to reject the retributive arguments for the fault system without addressing other principles of justice that might support the system. Coleman, Reply to Pilon, supra note 6, at 308, 310. But see, e.g., supra notes 194 & 195 and accompanying text; infra text accompanying note 204.
n204. Coleman, No-Fault, supra note 6, at 174.
n205. Id. at 174; see id. at 180 n.19; Coleman, Reply to Pilon, supra note 6, at 312.
n206. See Nickel, supra note 14, at 381-83, 385-88. Nickel asserts that there is no right to rectification of losses that are caused to distributively unjust holdings, no matter how wrongfully or unjustifiably those losses were caused. See id. at 387-88. In a similar vein, Coleman asserts that there is no right to rectification of losses that are caused to unjustly acquired holdings. See Coleman, Demands, supra note 6, at 369; Coleman, Reply to Pilon, supra note 6, at 312. Both of these positions are contrary to well-established law, which is supported by the Aristotelian concept of corrective justice. See infra text accompanying notes 365-70.
n207. The rationale for a losses-caused-by-human-action limitation might lie in a libertarian conception of distributive justice, which limits distributive justice to protecting holdings or entitlements, originally acquired through first possession, from any and all nonconsensual impingements by others. See Robert Nozick, Anarchy, State, and Utopia 149-82 (1974), cited in Nickel, supra note 14, at 384. However, under this conception distributive justice is displaced by an absolute-entitlement version of corrective justice, rather than the reverse effect that seems to occur under Nickel's version of the annulment thesis.
n208. As Stephen Perry has noted, the same criticisms apply to Coleman's similar attempts to limit compensatory justice claims to losses caused by faulty human action. See Coleman, Demands, supra note 6, at 370-72; Coleman, Mental Abnormality, supra note 6, at 127 & n.13; Perry, Comment, supra note 14, at 397-98.
n209. Although Coleman did not explicitly reiterate this point, it was central to Nickel's elaboration, which Coleman explicitly relied on. If Coleman had disagreed with this central feature of Nickel's elaboration, one would have expected him to say so.
n210. See Coleman, Mental Abnormality, supra note 6, at 132 n.10; Coleman, Mixed Conception, supra note 6, at 428-29; Coleman, Moral Theories: Part II, supra note 6, at 6-7; Coleman, Reply to Pilon, supra note 6, at 312; supra note 206.
n211. See Coleman, Demands, supra note 6, at 364 & n.14, 369-78; Coleman, Moral Theories: Part II, supra note 6, at 7-11, 13-19, 26-29; Coleman, Property, supra note 6, at 463-65; Coleman, Wrongful Gain, supra note 6, at 423-24, 427-28.
n212. See Coleman, Demands, supra note 6, at 370-71; Coleman, Mental Abnormality, supra note 6, at 118-26; Coleman, Moral Theories: Part II, supra note 6, at 9-11, 15; cf. id. at 30 (criticizing Fletcher's nonreciprocal-risk criterion of tortiousness because, "in order to be annulled, corrective justice requires of a loss that it be morally distinguishable from noncompensable losses") (emphasis in original).
n213. See Coleman, No-Fault, supra note 6, at 177-78. Coleman might have been influenced by a misreading of Aristotle in requiring that, for the injurer to be liable to the victim, there must have been unjust gain as well as unjust loss. Weinrib reads Aristotle as stating such a requirement, and struggles to satisfy it. See Weinrib, Corrective Justice, supra note 7, at 408-11, 415, 424-25; Weinrib, Negligence, supra note 7, at 53-54, 58-59. For an analysis of Aristotle's discussion of unjust gains and unjust losses, see infra text accompanying notes 311-50.
n214. Coleman now claims that under the annulment thesis the claim to recovery is good against everyone, as a matter of distributive justice rather than compensatory or corrective justice. But this is not a plausible interpretation of the annulment thesis. See infra note 237.
n215. See supra text accompanying note 205.
n216. See Coleman, Demands, supra note 6, at 367-68; Coleman, No-Fault, supra note 6, at 176-77; Coleman, Strict Liability, supra note 6, at 263-64, 286 n.71; Coleman, Wrongful Gain, supra note 6, at 440.
n217. See Coleman, Fault, supra note 6, at 487-88; Coleman, No-Fault, supra note 6, at 176-77. Coleman himself has noted the injustice of compelling someone to be an insurer for the fault of another. See Coleman, Strict Liability, supra note 6, at 284-85. Coleman argues that in instances where one buys insurance to cover injuries to oneself and obtains compensation through such insurance, "one's claim to compensation is against one's insurance company, not against the wrongdoer." Coleman, Fault, supra, at 487; see Coleman, No-Fault, supra, at 173. This is incorrect. The claim for compensation lies against both the insurer and the wrongdoer pursuant to the correlative rights and duties of contract and tort, respectively. Unless there is a subrogation clause in the insurance contract whereby the insurer is allowed to indemnify itself from the compensation owed by the wrongdoer, the victim is entitled to recover from both the insurer and the wrongdoer. Such "double recovery" has been contracted for by the insured. The abolition of such double recovery through abrogation of the "collateral source" rule is both a denial of the insured victim's rights and a windfall to the wrongdoer, and thus doubly offends against corrective justice.
n218. See Coleman, No-Fault, supra note 6, at 173, 174, 177. Coleman is one of many who have incorrectly assumed that the traditional Aristotelian concept of corrective justice, with its emphasis on the correlativity of the victim's right to recovery and the injurer's duty to repair, has been undermined by private insurance contracts and other voluntary arrangements whereby the compensation "originates with" someone other than the injurer. See infra note 358 and text accompanying notes 358-63. For further discussion of Coleman's confusion of the alleged "direct defendant payment" requirement with the correlativity of right and duty in corrective justice, see Perry, Comment, supra note 14, at 386-88.
n219. Coleman and others also have assumed that the doctrine of vicarious liability disregards the injurer's duty to repair in corrective justice. See Coleman, Fault, supra note 6, at 487; Coleman, Reply to Pilon, supra note 6, at 311; Sugarman, supra note 1, at 603. Yet vicarious liability does not eliminate the tortious employee's duty to repair. Rather, it recognizes an additional duty to repair on the part of the tortious employee's employer, which is secondary to the employee's primary duty. If the employer compensates the victim, the employer then has an indemnification claim against the employee. Properly understood, the employer's vicarious liability is based on the employer's own corrective justice responsibility for injuries that are tortiously inflicted in pursuance of the employer's objectives.
n220. See Coleman, No-Fault, supra note 6, at 177-78; Coleman, Wrongful Gain, supra note 6, at 440. Although Coleman does not seem to notice the discrepancy, this argument is the reverse of what he repeatedly states elsewhere regarding the compensatory justice rationale for the injurer's liability. See, e.g., Coleman, Strict Liability, supra note 6, at 262:
Under a compensatory rationale, faulty injurers are liable for their victims' losses, not to insure that those at fault are punished, but to see to it that fault-free victims are compensated for what are undeserved or otherwise unjustifiable losses. On this view, principles and ideals of compensatory justice rather than those of retributive justice are the touchstone of a system of liability rules.
On its face, the quoted "compensatory rationale" is indistinguishable from the "weak retributivism" rationale, and it reverses the emphasis of the annulment-of-unjust-gains rationale for injurer liability, which focuses on the injurer's unjust gain rather than the victim's unjust loss.
n221. See Nickel, supra note 14, at 386-87.
n222. Disgorgement of unjust gains should not be equated with the institution or remedy of restitution, since most instances of restitution involve reparation of unjust losses rather than disgorgement of unjust gains. See E. Allan Farnsworth, Your Loss or My Gain? The Dilemma of the Disgorgement Principle in Breach of Contract, 94 Yale L.J. 1339 (1985).
n223. See Coleman, No-Fault, supra note 6, at 177-78; Coleman, Reply to Pilon, supra note 6, at 312-13.
n224. See, e.g., Posner, supra note 14, at 197-98.
n225. At this point in the development of his theory, Coleman, with a little nudging from Posner, relabeled his theory as a theory of corrective justice rather than compensatory justice. See Coleman, Wrongful Gain, supra note 6, at 421-23; Posner, supra note 14, at 197; cf. Coleman, Strict Liability, supra note 6, at 261-62, 263-64 (equating corrective justice with retributive justice and distinguishing it from compensatory justice). Posner no doubt was delighted, given the subsequent confusion cast over the concept of corrective justice and the openness of Coleman's version to utilitarian manipulation. Yet, as Coleman now acknowledges, his theory was never a theory of corrective justice, but rather continued to emphasize open-ended, policy-oriented compensation rationales, especially with respect to the mode of rectification. See supra text accompanying note 179; infra note 237 and accompanying text. I therefore have continued to apply the "compensatory" label to Coleman's theory.
n226. See Coleman, Moral Theories: Part II, supra note 6, at 9-11; Coleman, Wrongful Gain, supra note 6, at 424-25, 437.
n227. See Weinrib, Causation, supra note 7, at 436-37.
n228. See Coleman, Moral Theories: Part II, supra note 6, at 14; Coleman, Property, supra note 6, at 465, 468; Coleman, Structure, supra note 6, at 1248-50; Coleman, Wrongful Gain, supra note 6, at 429 n.16, 437-38.
N229. See supra text accompanying notes 190-92.
n230. See Coleman, Fault, supra note 6, at 487; Coleman, Mental Abnormality, supra note 6, at 124-25; Coleman, Moral Theories: Part II, supra note 6, at 33; Coleman, Wrongful Gain, supra note 6, at 426 n.13. I find it difficult to understand Coleman's objection, in the last cited source, to the application of the logical correlativity principle to the right to recompense that he himself specifies earlier on the same page. Only in his most recent articles does Coleman explicitly reject the logical correlativity principle. See Coleman, Demands, supra note 6, at 368; Coleman, Mixed Conception, supra note 6, at 437.
n231. See supra text accompanying notes 214-19.
n232. Coleman, Moral Theories: Part II, supra note 6, at 12-13.
n233. See Coleman, Demands, supra note 6, at 368; Coleman, Mixed Conception, supra note 6, at 437; Coleman, Moral Theories: Part II, supra note 6, at 33-34.
n234. See Coleman & Kraus, supra note 6, at 1340-47, 1354-55; Coleman, Demands, supra note 6, at 372-78; see also Perry, Comment, supra note 14, at 402-08 & n.85. In his most recent articles, Coleman seems to abandon the semi-robust account of rights that he and Kraus articulated and to adopt instead the empty vessel account that he and Kraus criticized. See Coleman, Demands, supra note 6, at 368; Coleman, Mixed Conception, supra note 6, at 437.
n235. For an illuminating sketch of the content and structure of the primary and secondary rights of corrective justice, see MacCormick, supra note 14. In his contribution to this symposium, Coleman asserts that identifying corrective justice with this bilateral structure of primary and secondary rights is mistaken since the structure can be described without "mentioning corrective justice at all." Coleman, Mixed Conception, supra note 6, at 436-37; see Coleman, Demands, supra note 6, at 368. This assertion, it seems to me, confuses labels with substance.
n236. See Coleman & Kraus, supra note 6, at 1342. Again, Coleman seems to reverse ground by rejecting this conclusion in his most recent articles. See Coleman, Demands, supra note 6, at 368; Coleman, Mixed Conception, supra note 6, at 437.
n237. See Coleman, Mixed Conception, supra note 6, at 432-35, 437-38. Coleman accepts Perry's interpretation of the annulment thesis as a conception of distributive justice rather than corrective justice, in the sense that it supposedly assumes some sort of general social obligation regarding the annulment of unjust gains and unjust losses. See id.; Perry, Comment, supra note 14, at 396-98; Perry, Moral Foundations, supra note 14, at 449-50, 473. Actually, Coleman now interprets the annulment thesis, and distributive justice in general, as giving each member of society a particularized reason, or duty, to annul all unjust gains or losses in existing distributions. Coleman, Mixed Conception, supra, at 432, 434. But neither Coleman's original interpretation nor his current interpretation of the annulment thesis is a plausible conception of distributive justice. First, as Coleman repeatedly has emphasized, the annulment thesis requires annulment of "distortions" in existing distributions regardless of whether the existing distributions are distributively just. See supra text accompanying note 210. Second, at least as originally elaborated, the annulment thesis per se (the "principle of compensatory justice") creates no obligations, in anyone, to repair losses; it merely states that they should be annulled, thereby creating rights to recovery "in the air." A central premise of the annulment thesis is
that the right to recovery does not become determinate or complete until a particular mode of rectification is specified. But specification of the particular mode of rectification occurs outside the annulment thesis rather than within it. See Coleman, Demands, supra note 6, at 365-66; Coleman, Structure, supra note 6, at 1248-50; supra text accompanying notes 179, 213, 228; infra text following note 241. Third, distributive justice rights and duties do not necessarily apply to everyone and are not particularized one-to-one duties, as Coleman assumes in his current interpretation of the annulment thesis. Distributive justice claims are multilateral claims by all those who have too little against all those who have too much under the relevant distributive criterion. See infra text accompanying note 364.
n238. See Coleman, Mixed Conception, supra note 6, at 433.
n239. See Coleman, Demands, supra note 6, at 358; Coleman, Mixed Conception, supra note 6, at 429 & n.11.
n240. See Coleman, Mixed Conception, supra note 6, at 442-43. Although Coleman again acknowledges the critical importance of providing a substantive account of "wrong" and "wrongful," he again fails to provide such an account. The concept of wrong or wrongful is still formal, rather than substantive, and apparently remains open to economic interpretation. See id. at 429-30 & n.11, 444.
n241. See id. at 442-43 (footnootes omitted). The quote in the text reflects Coleman's initial statement of his argument. In response to objections raised by myself and Weinrib, Coleman modified the first paragraph by adding the words "as such" where the first ellipsis appears and the sentence "They are his fault" where the second ellipsis appears. Coleman also added a footnote stating, in effect, that an agent has a duty to repair a loss only if it was actually and proximately caused by her wrongful behavior. See id. & n.23. Yet Coleman's unrevised second paragraph seems to directly contradict the new language in the first paragraph. If the new language governs, Coleman apparently has finally abandoned the distinction between the grounds of recovery and the grounds of liability, since both recovery and liability are now based on a loss to P actually and proximately caused by D's wrongful behavior. However, Coleman still clings to the distinction between the grounds of recovery and liability on the one hand and the mode of rectification on the other. See supra text accompanying note 244.
n242. Id. at 442 (italics removed); see id. at 437.
n243. Id. at 437.
n244. See id. at 443-44; supra text accompanying notes 214-19.
n245. Compare the principle as stated by Coleman, supra text accompanying note 242.
n246. Coleman, Moral Theories: Part I, supra note 6, at 380-81; Coleman, Property, supra note 6, at 453; Coleman, Wrongful Gain, supra note 6, at 431.
n247. See supra note 54; infra text accompanying note 253.
n248. See supra note 237.
n249. See Coleman, Mental Abnormality, supra note 6, at 127-28; Coleman, Moral Theories: Part I, supra note 6, at 379-81, 385-86; Coleman, Strict Liability, supra note 6, at 277; Coleman, Wrongful Gain, supra note 6, at 431. See also Wright, Bramble Bush, supra note 3, at 1011-14; Wright, Causation, supra note 3, at 1741-58.
n250. In contrast to Coleman, Perry does not attempt to base the duty to repair solely on causal agency. Indeed, Perry has strongly and consistently criticized theories of "general strict liability" that purport to base liability solely on causal agency. See Perry, Comment, supra note 14, at 401-02, 404; Stephen R. Perry, The Impossibility of General Strict Liability, 2 Can. J.L. & Jur. 147 (1988). Perry employs the notion of outcome responsibility -- limited by proximity considerations -- to identify a restricted group of candidates for bearing the cost of the injury, with the choice among these candidates being based on a "localized distributive justice" comparison of their relative fault. See Perry, Comment, supra, at 399-40; Perry, Moral Foundations, supra note 14, at 467-68, 488-93, 496-99, 502-08, 512-14. Perry analogizes this two-pronged approach to Coleman's "weak retributivism" argument for imposing liability on the faulty injurer when the question of who is to bear the loss is institutionally restricted to the injurer and the victim. See Perry, Comment, supra, at 400 n.73; Perry, Moral Foundations, supra, at 467-68; supra text accompanying notes 197-99. Perry believes that his two-pronged approach resolves the problems that he identifies in either prong taken by itself. I disagree, but I will not attempt to explain the grounds of my disagreement in this Article.
n251. See Tony Honore, Responsibility and Luck, 104 Law Q. Rev. 530 (1988).
n252. See Coleman, Mixed Conception, supra note 6, at 434, 438, 443 & n.24; Honore, supra note 251, at 531, 540, 543-45; Perry, Moral Foundations, supra note 14, at 488-93, 497.
n253. See Honore, supra note 251, at 540.
n254. Aristotle, Nicomachean Ethics I.2 at 1094a18-b12 (W.D. Ross & J.O. Urmson trans.), in 2 The Complete Works of Aristotle (Jonathan Barnes ed., Revised Oxford Translation, Princeton Univ. Press 1984) [hereinafter Aristotle, Ethics]; see Aristotle, Politics I.1 at 1252a1-6, I.2 at 1252b28-30 (B. Jowett trans.), in 2 The Complete Works of Aristotle (Jonathan Barnes ed., Revised Oxford Translation, Princeton Univ. Press 1984). Indeed, the attainment of the good by its citizens is the reason for the existence of the state -- i.e., of civil government. Aristotle, Politics, supra, III.9 at 1280a31-b13.
n255. Aristotle, Ethics, supra note 254, I.7 at 1098a16-19. In the revised Oxford translations of Aristotle's works, which I use in this Article, the word "excellence" is used rather than the word "virtue" that is used in other translations. See, e.g., Aristotle, Nicomachean Ethics bk. I, ch. vii, v. 15-16, at 33 (H. Rackham trans., Loeb Classical Library, Harvard Univ. Press, 2d ed. 1934) [hereinafter Aristotle, Ethics (Rackham trans.)]. Believing "virtue" better conveys Aristotle's meaning to a current reader, I will substitute, without specific indication through brackets, "virtue" for "excellence" wherever the latter appears in the revised Oxford translations.
n256. Aristotle, Ethics, supra note 254, V.1 at 1129b27-1130a13.
n257. Id. V.1 at 1129b3-11, V.2 at 1130a14-b5; see infra text accompanying notes 289-306.
n258. Aristotle, Ethics, supra note 254, V.1 at 1129b14-24; see id. I.1 at 1094b4-7, I.13 at 1102a5-10. Aristotle states: "Practically the majority of the acts commanded by the law are those which are prescribed from the point of view of virtue taken as a whole; for the law bids us practise every virtue and forbids us to practise any vice." Id. V.2 at 1130b22-27.
n259. Id. V.1 at 1129b25-26.
n260. Id. V.1 at 1129a32-b1, 1129b12-14.
n261. See id. I.7 at 1097b30-1098a13.
n262. See id. I.7 at 1098a14-19, I.9 at 1099b25-1100a5.
n263. Id. I.4 at 1095a14-23; see id. I.8 at 1099a12-16.
n264. See id. I.5 at 1095b22-31, I.8 at 1098b11-19.
n265. Id. I.5 at 1095b14-22.
n266. Id. I.8 at 1099a32-1099b6; see id. I.10 at 1101a14-16.
n267. See id. I.5 at 1096a6-8.
n268. See id. I.10 at 1100b8-11, IX.8 at 1168b12-28; Aristotle, Politics, supra note 254, I.8 at 1257a26-32, I.10 at 1258a38-b2, I.11 at 1258b10-11, VII.1 at 1323a36-1324a1.
n269. See Aristotle, Ethics, supra note 254, I.2 at 1094a26-b12; infra text accompanying notes 273-74.
n270. See Aristotle, Politics, supra note 254, I.2 at 1252b28-1253a18; III.6 at 1278b15-25; Aristotle, Ethics, supra note 254, I.7 at 1097b8-11, VIII.9 at 1159b25-1160a14, IX.9 at 1169b3-21.
n271. Aristotle, Politics, supra note 254, III.9 at 1280b33-1281a4.
n272. See id. I.2 at 1253a19-39.
n273. See id. II.2 at 1261a14-b15, III.1 at 1274b39-41; VII.8 at 1328b15-22, VII.9 at 1328a37-b22; Martha Craven Nussbaum, Shame, Separateness, and Political Unity: Aristotle's Criticism of Plato, in Essays on Aristotle's Ethics 395, 395-97, 415-27 (Amelie Oksenberg Rorty ed., 1980).
n274. See Aristotle, Politics, supra note 254, III.7 at 1279a27-31 & 1279b4-10, III.13 at 1283b40-1284a2, IV.11 at 1295a25-b1, VII.1 at 1323b39-1324a1, VII.2 at 1324a6-8 & 1324a22-23, VII.3 at 1325b14-32, VII.9 at 1329a21-23.
n275. See id. III.11 at 1282b14-16; cf. id. I.2 at 1253a15-18 ("And it is a characteristic of man that he alone has any sense of good and evil, of just and unjust, and the like, and the association of living beings who have this sense makes a family and a state.").
n276. Id. III.6 at 1279a17-21; see Aristotle, Ethics, supra note 254, V.6 at 1134a24-30; cf. Kant, Morals, supra note 90, at * 318 ("By the well-being of a state is understood . . . that condition in which its constitution conforms most fully to principles of Right; it is that condition which reason, by a categorical imperative, makes it obligatory for us to strive after.").
n277. Aristotle, Magna Moralia I.33 at 1195a4-7 (St. G. Stock trans.), in 2 The Complete Works of Aristotle (Jonathan Barnes ed., Revised Oxford Translation, Princeton Univ. Press 1984); see id. I.33 at 1194b27-29; Aristotle, Ethics, supra note 254, V.6 at 1134a24-b1 (distinguishing between "what is just without qualification" and "political justice" or "legal justice," which is justice as accomplished through the rule of law); Aristotle, Rhetoric I.10 at 1368b8-9 (W. Rhys Roberts trans.), in 2 The Complete Works of Aristotle (Jonathan Barnes ed., Revised Oxford Translation, Princeton Univ. Press 1984) (distinguishing between "general law" ["all those unwritten principles which are supposed to be acknowledged everywhere"] and "special law" ["that written law which regulates the life of a particular community"]); id. I.13 at 1373b1-9 (distinguishing between "universal law" ["the law of nature" or "natural justice"] and "particular law" ["that which each community lays down and applies to its own members: this is partly written and partly unwritten"]).
n278. Aristotle, Ethics, supra note 254, V.7 at 1134b18-22; see supra note 277.
n279. See supra text accompanying notes 27-33.
n280. See Heyman, supra note 156. While Heyman and I agree that, for Aristotle, law, justice, ethics, and politics are intimately related, we disagree on the nature of the relationship. Heyman apparently reads Aristotle as making the requirements of justice depend on legislation, and identifies a good of the community distinct from and prior to the good of its individual members. I, on the other hand, read Aristotle as recognizing principles of natural justice toward which the political justice of legislation and adjudication aspires, and as equating the "common good" of the community with the good of each and every one of its individual members. See supra note 277; infra note 282 and accompanying notes 278-85.
n281. Aristotle, Magna Moralia, supra note 277, I.33 at 1195a1-7; see Aristotle, Ethics, supra note 254, V.7 at 1790-91.
n282. Aristotle describes such resort to fundamental principles to fill in the details or gaps as the exercise of equity, as opposed to the strictly legal justice of the written law. However, he states that equity is itself a kind of justice -- indeed, a better kind, since it resorts directly to the fundamental principles of justice and morality. See Aristotle, Ethics, supra note 254, V.10; Aristotle, Rhetoric, supra note 277, I.13 at 1374a25-b23. He treats the unwritten "natural" law as well as the written law as part of justice and law in the broad sense. See supra note 277. He states that "laws, when good [just], should be supreme," but acknowledges that laws may be just or unjust depending on the particular constitution of the state. See Aristotle, Politics, supra note 254, III.11 at 1282b1-12 (emphasis added). And he describes judgment as "the right discrimination of the equitable." See Aristotle, Ethics, supra, VI.11 at 1143a19-24.
n283. Aristotle, Politics, supra note 254, III.15 at 1286a10-20; see Aristotle, Ethics, supra note 254, V.6 at 1134a29-b7.
n284. Aristotle, Ethics, supra note 254, VI.11, VI.8 at 1142a11-20, V.2 at 1094b28-1095a11.
n285. Id. VI.12 at 1144a8-9.
n286. Id. I.7 at 1098b3-8.
n287. Id. I.2 at 1095a31-b46.
n288. Id. I.8 at 1098b9-11.
n289. Id. V.1 at 1129b27-1130a13.
n290. Id. V.2 at 1130a16-b5.
n291. See supra text accompanying notes 254-80.
n292. Aristotle, Ethics, supra note 254, II.5 at 1106a3-6, II.6 at 1106a25-1107a8.
n293. Id. V.1 at 1129a9-11, 1130a9-13.
n294. See, e.g., Finnis, supra note 14, at 161-63; Hart, supra note 117, at 154-55.
n295. Aristotle, Ethics, supra note 254, V.1 at 1129b2-11, V.2 at 1130a32-b5 (quoted supra text accompanying note 290), V.3 at 1131a10-23.
n296. See Weinrib, Corrective Justice, supra note 7, at 405-06, 413, 421; Weinrib, Forms, supra note 7, at 133-36. Moreover, although Weinrib implies otherwise, Aristotle uses the concept of equality, along with the terms "intermediate" and "mean," when discussing not only justice in its particular sense, but also justice in its wide sense (complete virtue). See Weinrib, Corrective Justice, supra, at 406; supra text accompanying notes 292-93.
n297. See Aristotle, Ethics, supra note 254, IV.4 (ambition and lack of ambition as the excess and deficiency, respectively, in the disposition to seek honour from others), IV.6 (contentiousness and obsequiousness as the excess and deficiency, respectively, in the disposition to get along with others), IV.7 (boastfulness and mock-modesty as the excess and deficiency, respectively, in the disposition to seek a good reputation in other's eyes), VIII & IX (friendship); infra note 299 (liberality).
n298. See Aristotle, Rhetoric, supra note 277, I.9 at 1366b8-11: "Justice is the virtue through which everybody enjoys his own possessions in accordance with the law; its opposite is injustice, through which men enjoy the possessions of others in defiance of the law."
n299. See Aristotle, Ethics, supra note 254, IV.1 (prodigality and meanness as the excess and deficiency, respectively, and liberality as the virtuous mean in one's disposition to share one's goods or wealth with others).
n300. Id. V.1 at 1129b2-11, V.2 at 1130a16-21 (quoted supra text accompanying note 290).
n301. See supra text accompanying notes 119, 142-50. The "being yielding" conception of deficiency is not explicit in Aristotle's account of particular justice, perhaps because being yielding -- unlike being grasping -- is not unjust. See infra text accompanying notes 303-06. However, the conception is consistent with his account of justice in its particular sense, and also with his description of other vices. See Aristotle, Ethics, supra note 254, IV.3 (undue humbleness as the deficiency in one's disposition to value one's own worth), IV.5 (slavishly "putting up with insults" as the deficiency in one's disposition to get angry); cf. id. III.11 at 1119a6-11 (Aristotle, believing that no human fails to appreciate pleasure, asserts that there can be no vice of deficiency in one's disposition to fulfill one's appetites). My elaboration of the vice of being yielding has benefitted from discussions with Steven Heyman.
n302. Aristotle, Ethics, supra note 254, V.5 at 1133b30-1134a16.
n303. See supra text accompanying note 292; infra text accompanying note 330.
n304. Aristotle, Ethics, supra note 254, V.9 at 1136b5-6.
n305. Id. V.11 at 1138a9-35.
n306. See id. V.6 at 1134b9-12, V.9 at 1136a30-b14, V.11; cf. id. V.11 at 1138a9-14 ("He who through anger voluntarily stabs himself does this contrary to right reason, and this the law does not allow; therefore he is acting unjustly. But towards whom? Surely towards the state, not towards himself. For he suffers voluntarily, but no one is voluntarily treated unjustly. This is also the reason why the state punishes; a certain loss of civil rights attaches to the man who destroys himself, on the ground that he is treating the state unjustly.").
n307. In the rest of this Article, I will be focusing on justice in its particular sense, and, unless otherwise noted, the term "justice" should be read as referring to justice in its particular sense.
n308. Aristotle, Ethics, supra note 254, V.2 at 1130b30-33. I have substituted "government" for "constitution." See Aristotle, Politics, supra note 254, III.6 at 1279a26 ("The words constitution and government have the same meaning.").
n309. Aristotle presumed that "all men agree that what is just in distribution must be according to merit in some sense, though they do not all specify the same sort of merit, but democrats identify it with the status of freeman, supporters of oligarchy with wealth (or with noble birth), and supporters of aristocracy with virtue." Aristotle, Ethics, supra note 254, V.3 at 1131a24-28.
n310. Id. V.3 at 1131a29-b15.
n311. Aristotle refers to "transactions" rather than to "interactions," but contrary to modern usage treats involuntary as well as voluntary interactions as "transactions." See id. V.3 at 1131a1-9, V.4 at 1131b25-26. Believing "interaction" better conveys Aristotle's meaning to a current reader, I substitute, without specific indication through brackets, "interaction" for "transaction" wherever the latter appears in the revised Oxford translations.
n312. Id. V.4 at 1131b26-1132a6.
n313. The equality of entitlement does not assume an absolute entitlement, which would be invaded by any adverse effect. The content of the entitlement depends, inter alia, on the specification of the permissible and impermissible means of affecting the interests that are protected by the entitlement. See supra text accompanying note 234.
n314. Aristotle, Ethics, supra note 254, V.4 at 1132a7-33, 1132b12-20.
n315. See infra text accompanying notes 333-56.
n316. See, e.g., Aristotle, Ethics, supra note 254, V.9 at 1136b9-14 (gifts), V.5 at 1132b31-1133b28 (voluntary exchanges). Aristotle notes that voluntary exchanges depend on a form of reciprocity, according to which goods of equal value are exchanged. For the practical purposes of such exchange only, value is measured by demand and is rendered commensurable through the convention of money. However, Aristotle emphasizes, the monetized demand value of goods should not be taken as their real value, nor as the worth of the people who produce such goods. Id. V.5 at 1133b6-21. Nor should the practical necessity of reciprocity of monetized demand value in voluntary exchanges be confused with either distributive or corrective justice: "reciprocity fits neither distributive nor rectificatory [corrective] justice." Id. V.5 at 1132b24. Exchanges that fail to satisfy such reciprocity will occur only if one party to the exchange takes advantage of the other party -- i.e., fails to treat the other party as an equal. It is the failure to treat the other party as an equal, rather than the lack of reciprocity per se, which raises the issue of corrective justice. Marlena Corcoran, it seems to me, both misreads and overemphasizes this part of Aristotle's discussion of justice. See Marlena G. Corcoran, Aristotle's Poetic Justice, 77 Iowa L. Rev. 837 (1992).
n317. See infra note 372 and accompanying text.
n318. See supra text accompanying note 312.
n319. Let A equal the parties' respective (presumed equal) pre-interaction holdings, and B equal the unjust loss or (presumed equal) unjust gain. Then the post-interaction holdings are A+B and A-B, respectively, and the arithmetic mean (average) of the post-interaction holdings is A+B+A-B divided by two, which is A. The arithmetic mean is implemented by taking B from the injurer and transferring it to the injured.
n320. Contra Perry, Moral Foundations, supra note 14, at 453-57 (interpreting Aristotle's argument as a limited restitutionary argument that is applicable only when there is an unjust gain identical to the unjust loss). But see id. at 454 n.21, 457-58 (acknowledging that this might not be the best interpretation of Aristotle's argument). See also supra note 213 and accompanying text (discussing Coleman's and Weinrib's linking of liability for unjust losses to the injurer's unjust gain).
n321. See supra text accompanying note 314; Aristotle, The Nicomachean Ethics V.C.4 at 1132a n.2 (W.D. Ross trans., Oxford Univ. Press 1925) (explaining Aristotle's discussion of arithmetical proportion as a heuristic device).
n322. Aristotle, Ethics, supra note 254, V.4 at 1132a33-b15.
n323. Again, the quoted phrases should not be read too literally. See supra text accompanying notes 315-17.
n324. See infra text accompanying notes 333-56.
n325. Aristotle, Ethics, supra note 254, V.8 at 1135a19-28; see id. III.1 (discussion of voluntariness as affected by compulsion or ignorance).
n326. See id. III.1 at 1109b35-1110b16, V.8 at 1135b6-9.
n327. Id. III.1 at 1110b28-1111a2; see Aristotle, Rhetoric, supra note 277, I.10 at 1368b20-23 ("The ambitious man does wrong for the sake of honour, . . . the stupid man because he has misguided notions of right and wrong.").
n328. Aristotle, Ethics, supra note 254, III.5 at 1113b29-1114a3.
n329. Id. II.4, V.9 at 1137a5-26.
n330. Id. V.8 at 1135b19-25; see id. V.6 at 1134a16-23; cf. id. III.2 (distinguishing choice from voluntary acts).
n331. Id. V.8 at 1135a16-22.
n332. Id. V.9 at 1136a23-30.
n333. Id. V.8 at 1135b11-25.
n334. See supra text accompanying notes 324-28.
n335. But not to absolute liability, since Aristotle distinguishes unjust gains and losses from gains and losses per se. See infra text accompanying notes 340-56. Liability is absolute when the only requirement for liability is causation of the loss. Liability is strict when something more than mere causation of the loss, but less than negligence, is required for liability. For example, there is strict liability for damages caused by the ultrahazardous aspect of ultrahazardous activities. Ultrahazardous activities are activities that are likely to cause substantial injury if an escape or loss of control should occur, regardless of the care taken to avoid such escape or loss of control. See Prosser & Keeton, supra note 38, ¤ 78, at 545, 554-56.
n336. Contra Coleman, Wrongful Gain, supra note 6, at 436; Perry, Moral Foundations, supra note 14, at 453, 454; Posner, supra note 14, at 190, 200. Each of these authors reads Aristotle as limiting corrective justice to deliberately caused injuries.
n337. Rackham so interprets Aristotle:
The three sorts of injury are [atuchema, hamartema, and adikema]. The second term is introduced first, in its wider sense of a mistake which leads to an offence against some one else (the word connotes both things). It is then subdivided into two: [atuchema], accident or misadventure, an offence due to mistake and not reasonably to be expected, and [hamartema] in the narrow sense, a similar offence that ought to have been foreseen. The third term, [adikema], a wrong, is subdivided into wrongs done in a passion, which do not prove wickedness, and wrongs done deliberately, which do.
Aristotle, Ethics (Rackham trans.), supra note 255, bk. V, ch. viii, at 300-01 n.b; see infra note 342.
n338. Cf. Aristotle, Ethics, supra note 254, V.9 at 1136b26-1137a4 (making this point with respect to the same issue in distributive justice).
n339. See infra text accompanying notes 343-56.
n340. See Aristotle, Rhetoric, supra note 277, I.14 at 1374b32-34 ("legal punishment and chastisement are the proper cure [for wrongdoing]"); id. I.13 at 1373b27-30 ("Being wronged must consist in having an injury done to you by someone who does it voluntarily. In order to be wronged, a man must suffer actual harm and suffer it involuntarily [without his consent]."); supra text accompanying note 325.
n341. See supra text accompanying note 326.
n342. Aristotle, Ethics, supra note 254, II.1 at 1109a30-34; see id. V.8 at 1136a5-9; Aristotle, Rhetoric, supra note 277, I.13 at 1374b4-9:
Equity must be applied to forgivable actions; and it must make us distinguish between wrongdoings on the one hand, and mistakes, or misfortunes [misadventures], on the other. (A misfortune is an act, not due to wickedness, that has unexpected results; a mistake is an act, also not due to turpitude, that has results that might have been expected; a wrongdoing has results that might have been expected, but is due to turpitude.)
n343. Id. V.8 at 1136a1-3 (emphasis shifted).
n344. See supra text accompanying notes 308-10.
n345. See Aristotle, Ethics, supra note 254, V.9 at 1136b15-1137a4; supra text accompanying note 343.
n346. See supra text accompanying notes 318-23.
n347. See supra text accompanying notes 300-02, 311-15 and following note 323.
n348. See Aristotle, Ethics, supra note 254, V.9 at 1136a30-b14; supra note 306, 340.
n349. See supra text accompanying notes 331-37.
n350. See supra text accompanying and following notes 340-42.
n351. Cf. Aristotle, Ethics, supra note 254, I.7 at 1098a21-24 ("But it would seem that anyone is capable of carrying on and articulating what has once been well outlined, and that time is a good discoverer or partner in such a work").
n352. See supra text accompanying notes 311-12.
n353. See supra text accompanying notes 261-76; Aristotle, Ethics, supra note 254, V.6 at 1134a26-30, 1134b14-15:
Political justice . . . is found among men who share their life with a view to self-sufficiency, men who are free and either proportionately or arithmetically [absolutely] equal, so that between those who do not fulfil this condition there is no political justice but justice in a special sense and by analogy. For justice exists only between men whose mutual relations are governed by law; . . . and these as we saw are people who have an equal share in ruling and being ruled.
n354. Aristotle, Politics, supra note 254, VII.8 1328b13-14.
n355. Cf. Aristotle, Rhetoric, supra note 277, I.5 at 1360b13-17 ("We may define happiness as prosperity combined with virtue; or as independence of life; or as the secure enjoyment of the maximum of pleasure; or as a good condition of property and body, together with the power of guarding one's property and body and making use of them. That happiness is one or more of these things, pretty well everybody agrees.").
n356. See supra text accompanying notes 92, 121-31, 156-58.
n357. See supra text accompanying note 74.
n358. This misunderstanding is especially evident in the writings of the social welfare theorists, who rely upon it to argue that corrective justice is archaic and a misguided popular illusion given the ultimate allocation of costs under our current liability regimes. See, e.g., 1 ALI Reporters' Study, supra note 1, at 24-25; sources cited in Gary T. Schwartz, The Ethics and the Economics of Tort Liability Insurance, 75 Cornell L. Rev. 313, 313 n.2 (1990).
n359. See supra text accompanying and following notes 216, 340-42.
n360. The only expositor of corrective justice who might disagree with this position is Weinrib. Some of his writings strongly imply that the compensation to the injured party must come directly from the person who caused the injury. See Weinrib, Corrective Justice, supra note 7, at 409-11; Weinrib, Formalism, supra note 7, at 978, 980-81, 982; Weinrib, Negligence, supra note 7, at 38; Weinrib, Tort Law, supra note 7, at 494, 511-12. Moreover, Weinrib seems to assume that liability in corrective justice depends on an unjust gain by the injurer correlative to the victim's unjust loss, which can only be eliminated by requiring the injurer to compensate the victim. See supra note 213.
n361. See MacCormick, supra note 14, at 217. Gary Schwartz skillfully demonstrates that liability insurance is not contrary to, and may even facilitate the achievement of, principles of fair or just liability. Yet even Schwartz seems to confuse issues of duty -- who is properly held liable -- with questions related to the appropriate discharge of that duty. He assumes that "corrective justice" as currently articulated requires that the person held liable discharge the duty herself, through her own funds, as opposed to "compensatory justice," which would only require that the person held liable see to it that the injured person is compensated. Schwartz, supra note 358, at 321-36. Schwartz acknowledges that what he terms "corrective justice" is narrower than Aristotle's conception, which, he states, would include what he calls "compensatory justice" and "retributive justice" as well as "corrective justice." Id. at 331 n.82.
n362. If two or more parties to the interaction each have independent corrective justice duties to rectify the injured party's loss (i.e., joint tortfeasors, including employers and their employees), and one of the liable parties fully rectifies the loss, thereby discharging the other liable parties' duties to the injured party as well as her own, she may have a corrective-justice-based claim for indemnification or contribution against the other liable parties.
n363. See supra text accompanying notes 190-92, 215-19 and following note 208. Christopher Schroeder's argument that, in theory, corrective justice could be satisfied by a "tortious behavior" fund is subject to the same objections. Under Schroeder's scheme, those who behave tortiously would pay into a common fund in accord with the ex ante expected damages of their tortious conduct, whether or not they cause any injury, and victims injured by tortious conduct would recover from the common fund. See Schroeder, Risks, supra note 14, at 450, 459-69. This scheme not only is inconsistent with the correlative bilateral rights and duties that are a central feature of corrective justice, but also ignores the logical correlativity of rights and duties per se. Although the victim's right to recovery is grounded on tortiously caused injury, the victim's claim is made not against the person who tortiously caused the injury, but against all who have behaved tortiously, whether or not they caused any injury.
Schroeder relies in part on Kant's moral theory, which he construes as being concerned not with ex post consequences, but only with subjectively knowable ex ante risks. See id. at 452-55. But Schroeder is incompletely describing Kant's doctrine of virtue, which, for purposes of assessing moral merit or blame, focuses on the individual's subjective capacity and effort in attempting to ascertain and meet the objective requirements of the categorical imperative. See Kant, Morals, supra note 90, at * 214, 225-26, 228, 382 n., 389, 390-91, 392-93, 394, 401, 404-05, 446-47, 463. Schroeder overlooks Kant's doctrine of justice or Right, which uses those objective requirements to assess the individual's moral and legal responsibility for the adverse effects that he or she causes to others. See id. at * 214, 223-25, 227-32, 312; supra text accompanying notes 93-141.
Schroeder also states that "virtually every area of common-law decisionmaking dealing with matters of individual accountability for actions employs certain excusing doctrines: mistake, ignorance of fact, coercion, undue influence, insanity, fraud, and the like." Schroeder, Risks, supra, at 459. While most of these excuses are relevant to moral blame in criminal law and moral responsibility based on voluntary agreement in contract law, most of them are irrelevant to moral responsibility in tort for one's causation of injury to others without their consent. See Prosser & Keeton, supra note 38, ¤ 17, at 110-11, ¤¤ 20-21, at 130-32, ¤ 22, at 137-38, ¤ 31, at 169, ¤ 32, at 173-84, ¤¤ 134-35, at 1071-75; supra text accompanying notes 324-42.
n364. See MacCormick, supra note 14, at 217; Gordley, supra note 4, at 1591; infra text accompanying note 369.
n365. See, e.g., Ronald Dworkin, Law's Empire 297-309 (1986); John Rawls, A Theory of Justice 10-11 (1971); Heidt, supra note 14, at 350-53; Nickel, supra note 14, at 381-83, 386-88; Posner, supra note 14, at 190-91, 192 & n.22, 196, 200-01; David Rosenberg, Damage Scheduling in Mass Exposure Cases, 1 Cts., Health Sci. & L. 335, 348 n.14 (1991); Jeremy Waldron, Criticizing the Economic Analysis of Law, 99 Yale L.J. 1441, 1450-53 (1990) (book review). Stephen Perry's position is not clear to me. On the one hand, he assumes there is "at least some measure of independence between corrective and distributive justice." Perry, Moral Foundations, supra note 14, at 451. On the other hand, he seems to assume that the holdings protected by corrective justice must themselves be distributively just. See id. at 453, 454 n.21, 455, 513. And he states that "there is no hard and fast dividing line between corrective and distributive justice"; that corrective justice is a "localized distributive inquiry, as constrained by outcome responsibility"; and that this "localized distributive justice [perhaps] can be replaced by more general distributive schemes, like a compulsory no-fault insurance plan, without violating any fundamental moral rights." Id. at 512-13; see supra note 250.
n366. See, e.g., Coleman, Mixed Conception, supra note 6, at 428-29; Coleman, Moral Theories: Part II, supra note 6, at 6-7; Gordley, supra note 4, at 1589-92; MacCormick, supra note 14, at 217; Weinrib, Formalism, supra note 7, at 983-84.
n367. See supra text accompanying notes 308-12.
n368. Id.; see Aristotle, Ethics, supra note 254, V.2 at 1130a30-1131b9.
n369. E.g., State v. Moe, 24 P.2d 638 (Wash. 1933); London Borough of Southwark v. Williams,  2 All E.R. 175 (C.A.).
n370. E.g., Anderson v. Gouldberg, 53 N.W. 636 (Minn. 1892); Tapscott v. Cobbs, 52 Va. (11 Gratt.) 172 (1854).
n371. See supra text accompanying note 25; accord, Alexander, supra note 14, at 6-11.
n372. See Aristotle, Ethics, supra note 254, V.5 at 1132b28-29 ("if an official has inflicted a wound, he should not be wounded in return"), V.11 at 1138a21-23 ("unjust action is voluntary and done by choice, and is prior (for the man who because he has suffered does the same in return is not thought to act unjustly)"); cf. supra text accompanying note 97 (Kant's concept of Right).
n373. See supra text accompanying note 364.
n374. E.g., Housing Auth. v. Midkiff, 467 U.S. 229 (1984); Poletown Neighborhood Council v. City of Detroit, 304 N.W.2d 455 (Mich. 1981).
n375. E.g., Boomer v. Atlantic Cement Co., 257 N.E.2d 870 (N.Y. 1970). Although the opinions of the New York Court of Appeals in Boomer may give the impression that there was little or no public benefit from the cement plant, the majority, in refusing to grant an injunction against the nuisance and limiting the plaintiff to a damages remedy, noted the large number of jobs provided by the plant as well as the huge investment in the plant. Id. at 873 n.*. The lower courts all stressed the major public benefits provided by the plant, including not only the jobs provided directly by the plant, but also the boost to the local economy and the major tax revenues generated. See Boomer v. Atlantic Cement Co., 287 N.Y.S.2d 112, 114 (1967); aff'd, 294 N.Y.S.2d 452, 453 (1968).
n376. See, e.g., Housing Auth. v. Midkiff, 467 U.S. 229 (1984).
n377. See, e.g., Boomer v. Atlantic Cement Co., 257 N.E.2d 870 (N.Y. 1970); Poletown Neighborhood Council v. City of Detroit, 304 N.W.2d 455 (Mich. 1981).
n378. See supra text accompanying note 61.
n379. These considerations suggest that in private takings cases such as Boomer, a court should be reluctant to uphold the taking, even with full compensation, without some indication of legislative affirmation of the private project's desirability because of its public benefit.
n380. Implied warranties of habitability may have either a corrective justice basis (protecting bargained for expectations) or a redistributive basis (redistributing wealth from wealthier landlords to poorer tenants). Insofar as the latter is the purpose, the courts can be charged with having misconceived the distributive justice claim as well as having failed to recognize the limits of their capacity to achieve redistribution. Why should landlords rather than the community at large bear the burden of redistribution? And how effective is this court-devised attempt to achieve redistribution through ad hoc court action likely to be? Similar issues arise with rent-control regulations, which may, depending on the circumstances, either have a corrective justice rationale of preventing unfair advantage-taking or a (defective) redistributive rationale of providing subsidies to tenants at the expense of their landlords (or at the expense of the tenants of uncontrolled units). See Pennell v. City of San Jose, 485 U.S. 1, 15, 19-24 (1988) (Scalia, J., concurring and dissenting).
n381. The nature conservation or preservation component of environmental law, it seems to me, is based on distributive justice, while the pollution control component is based primarily on corrective justice -- the prevention of injuries through interactions. The particular shape that pollution-control measures take is often influenced by distributive justice concerns, especially concerns about the effect that particular types of regulation may have on the economic health and competitiveness of different industries and communities. See, e.g., Bruce A. Ackerman & William T. Hassler, Clean Coal/Dirty Air or How the Clean Air Act Became a Multibillion-Dollar Bail-Out for High-Sulfur Coal Producers and What Should Be Done about It (1981).
n382. Cf. Posner, supra note 14, at 188 (noting that, given the powerful intuitive appeal of the concept of corrective justice, an exploration of its meaning and origins should not be viewed as "an essay in antiquarianism").
n383. See Simons, supra note 14, at 128.
n384. See supra text accompanying notes 1-2.
n385. See, e.g., 1 ALI Reporters' Study, supra note 1, at 24-25, 50; 2 ALI Reporters' Study, supra note 1, at 79-80 & n.61.
n386. See, e.g., Kenneth Baynes, The Normative Grounds of Social Criticism: Kant, Rawls, and Habermas (1991); Charles Fried, An Anatomy of Values 41-42, 97-101 (1970); C.B. Macpherson, Liberal-Democracy and Property, in Property: Mainstream and Critical Positions 199 (C.B. Macpherson ed., 1978); Loren E. Lomasky, Persons, Rights, and the Moral Community (1987); Karl Marx & Friedrich Engels, The Communist Manifesto (1848); Robert Nozick, Anarchy, State, and Utopia 48-51 (1974); John Rawls, A Theory of Justice 408, 504-12 (1971); Jeffrey Reiman, Justice and Modern Moral Philosophy (1990); Richard Warner, Freedom, Enjoyment, and Happiness (1987); Ronald Dworkin, In Defense of Equality, 1 Soc. Phil. & Pol. 24, 25-29 (1983); Gordon, supra note 2, at 197; Martha C. Nussbaum, Comments, in Symposium on Classical Philosophy and the American Constitutional Order, 66 Chi.-Kent L. Rev. 213 (1990).
Send comments to: Andreas Teuber
Last Modified: 01/16/05
Copyright © The President and Fellows of Harvard College