OXFORD LEGAL P HI LOSO P HY

Series Editors: TIMOTHY ENDICOTT , JOHN GARDNER ,

and LESLIE GREEN

Corrective Justice
 OXFOR D LEGAL P HI LOSO P HY
Series Editors
Timothy Endicott, John Gardner, and Leslie Green

Oxford Legal Philosophy publishes the best new work in philosophically-

oriented legal theory. It commissions and solicits monographs in all
branches of the subject, including works on philosophical issues in
all areas of public and private law, and in the national, transnational,
and international realms; studies of the nature of law, legal institutions,
and legal reasoning; treatments of problems in political morality as
they bear on law; and explorations in the nature and development
of legal philosophy itself. The series represents diverse traditions of
thought but always with an emphasis on rigour and originality. It sets
the standard in contemporary jurisprudence.
Corrective Justice

Ernest J. Weinrib

1
3
Great Clarendon Street, Oxford, ox2 6dp,
United Kingdom
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide. Oxford is a registered trade mark of
Oxford University Press in the UK and in certain other countries
© E. Weinrib, 2012
The moral rights of the author have been asserted
First Edition published in 2012
Impression: 1
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by licence or under terms agreed with the appropriate reprographics
rights organization. Enquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above
You must not circulate this work in any other form
and you must impose this same condition on any acquirer
Crown copyright material is reproduced under Class Licence
Number C01P0000148 with the permission of OPSI
and the Queen’s Printer for Scotland
British Library Cataloguing in Publication Data
Data available
Library of Congress Cataloging in Publication Data
Library of Congress Control Number: 2012942995
ISBN 978–0–19–966064–3
Printed in Great Britain by
Clays Ltd, St Ives plc
Links to third party websites are provided by Oxford in good faith and
for information only. Oxford disclaims any responsibility for the materials
contained in any third party website referenced in this work.
To Lorraine
This page intentionally left blank
Series Editors’ Preface

Ernest Weinrib is best-known for his 1995 monograph The Idea of

Private Law. Nobody who works in the theory of private law could
fail to be gripped by it or to be influenced by it. Although the main
ideas set out in that book had been explored by Weinrib in a series of
earlier articles, it was their synoptic presentation in book form that
had the greatest impact on the subject. For those who disagree with
its approach as much as for those who agree, The Idea of Private Law
inaugurated a new era in private law theory, marked by a new philo-
sophical energy amongst private law scholars and a revitalized interest
in private law categories amongst legal theorists.
In this new book, Weinrib collects important writings from the
period since The Idea of Private Law (with significant revisions and
additions). As well as nuancing and reinforcing his original claims,
Weinrib demonstrates their extensive reach. While the emphasis in
The Idea of Private Law was primarily upon the law of torts, and the
law relating to reparative damages for negligence in particular, this
new book gives extensive attention to other remedies (injunctions,
gain-based awards, etc.) and to other causes of action (breach of
contract, unjust enrichment). Nor do its horizons end with the pri-
vate law of municipal legal systems; in the last third of this book,
Weinrib shows us how fertile his distinctive patterns of thought can
be in other settings, such as in Jewish law, in the political theory of
the welfare state, and in the theory and practice of legal education.
Weinrib is one of the leading legal theorists of our era. His ideas, for all
their declared classical and Enlightenment origins, are breathtakingly
viii series editors’ preface

original and far-reaching in their implications.We are proud to include

this important new work in Oxford Legal Philosophy.

Timothy Endicott
John Gardner
Leslie Green
June 2012
Acknowledgments

Much of this book reworks, sometimes substantially, material that has

appeared in a variety of journals and edited volumes: “Correctively
Unjust Enrichment,” in Philosophical Foundations of the Law of Unjust
Enrichment, ed. Rob Chambers, Charles Mitchell, and James Penner
(Oxford University Press, 2009), 31–53; “Two Conceptions of Rem-
edies,” in Justifying Private Law Remedies, ed. Charles Rickett (Hart
Publishing, 2008), 3–32; “Can Law Survive Legal Education?” (2007)
60 Vanderbilt Law Review 401–38; “The Disintegration of Duty,” in
Exploring Tort Law, ed. Stuart Madden (Cambridge University Press,
2005), 143–86; “Planting Another’s Field: Unrequested Improvements
under Jewish Law,” in Understanding Unjust Enrichment, ed. Jason
W. Neyers, Mitchell McInnes, and Stephen G. A. Pitel (Hart Publish-
ing, 2004), 221–46: “Punishment and Disgorgement as Contract Rem-
edies,” (2003) 78 Chicago-Kent Law Review 55–103; “Poverty and
Property in Kant’s System of Rights,” (2003) 78 Notre Dame Law
Review 795–828; “Correlativity, Personality and the Emerging Con-
sensus on Corrective Justice” (2001) 2 Theoretical Inquiries in Law
107–59; “Restitutionary Damages as Corrective Justice,” (2000) 1 The-
oretical Inquiries in Law 3–37. I am grateful to the publishers of these
articles for permission to make use of them for this volume.
I have been very fortunate in working for the last four decades at
the Faculty of Law, University of Toronto, which has become a major
centre for the theory of private law. There I have enjoyed the constant
benefit of thoughtful and stimulating colleagues, enthusiastic but
skeptical students, and a series of supportive deans. A draft of this book
was the subject of the weekly discussions of the Faculty’s Law and
x acknowledgments

Philosophy Discussion Group during the 2010–11 school year. The

book has been much improved by the scrutiny its ideas received from
the participants in the group’s discussions and from other colleagues:
Peter Benson, Alan Brudner, Simone Chambers, Bruce Chapman,
Abraham Drassinower, Harry Fox, Louis-Philippe Hodgson, Amnon
Lehavi, Mark Migotti, Benny Porat, Arthur Ripstein, Hamish Stewart,
and Jacob Weinrib.
My elaboration of the Kantian themes in this book benefited from
the fact that at the same time my colleague Arthur Ripstein was
working on Force and Freedom, his path-breaking study of Kant’s legal
and political philosophy. I was the happy beneficiary of numerous
protracted conversations with Arthur about the details and implica-
tions of a Kantian approach to legal philosophy. Similarly, I owe par-
ticular thanks to Jacob Weinrib, whose acuity as an interpreter of
Kant’s legal philosophy compelled me to greater accuracy in present-
ing Kantian ideas.
My greatest debt is to my wife Lorraine, to whom this book is
dedicated. Her least important contribution to my life, but the one
most relevant to this book, is that, initially as a constitutional lawyer
appearing before appellate courts and subsequently as a law professor,
she made me aware, and constantly reminded me by her example, of
the adamantine necessity for coherence in legal thinking.
Contents

Introduction 1
1. Correlativity and Personality 9
2. The Disintegration of Duty 38
3. Remedies 81
4. Gain-based Damages 117
5. Punishment and Disgorgement as Contract Remedies 148
6. Unjust Enrichment 185
7. Incontrovertible Benefit in Jewish Law 230
8. Poverty and Property in Kant’s System of Rights 263
9. Can Law Survive Legal Education? 297
Conclusion 334
Index 345
This page intentionally left blank
 Introduction

It has been said of Martin Buber that the fundamental insight under-
lying all of his thought is that “primary reality lies neither in the sub-
ject nor the object, ie, in the poles of the relation, but in the relation
itself.”1 A similar distinction between the poles of a relation and the
relation itself—and a similar claim that the poles can be understood
only through the relation between them—is the theme of this book.
For Buber, the relation in question was the dialogic I–Thou relation,
in which the related entities realize themselves by confronting each
other as ends in themselves. The relationship with which this book
deals is that of parties linked through the norms of private law, and
thus through the possibility that one may be found liable to the
other.
The theme of this book may seem very distant from Buber’s con-
cerns. Unlike the warm self-realization of the I–Thou relation and the
encounter with God as the eternal Thou, private law is a domain of
contested claims and institutional coercion. Insistence on one’s rights
under private law is often, as Hegel put it, “a fitting accompaniment of
a cold heart and restricted sympathies.”2 Yet, though private law gives
rise to a different kind of relationship than those that Buber postulat-
ed, its relational character is nonetheless crucial to its operation and
intelligibility. A finding of liability always relates a particular plaintiff
to a particular defendant. Moreover, in sophisticated legal systems
liability is supported by reasons that attempt to set out (not always

1
Manfred Vogel, “The Concept of Responsibility in the Thought of Martin Buber,”
(1970) 63 Harv. Theological Rev. 159.
2
Hegel’s Philosophy of Right, tr. T. M. Knox (1952), s. 37R.
2 corrective justice

felicitously, of course) why the norm being applied is fair to both par-
ties and how it participates in an ensemble of norms that treats their
relationship as a coherent normative unit. Given the relational nature
of liability, these reasons must themselves be relational. And given the
aspiration to fairness and coherence, the reasons must treat the parties
as equal persons whom the law does not subject to normatively arbi-
trary demands. Accordingly, private law gives rise to a question that
parallels the issues that animated Buber: what does it mean for private
law to be expressive of a relationship in which plaintiff and defendant
each stand to the other as ends in themselves?
Corrective justice is the term given to the relational structure of
reasoning in private law. This term has a venerable history, having first
been formulated by Aristotle and then continuously discussed and
refined in the two and a half millennia since.3 It conceptualizes the
parties as the active and passive poles of the same injustice (as the doer
and the sufferer of injury in Aristotle’s account). It directs us, accord-
ingly, to normative considerations that pertain not to either (or even
to each) of the poles taken on its own, but to the relationship as such.
The material in this book presents what these relationally norma-
tive considerations are and how they work across various bases of
liability. Over the last few decades corrective justice has become well
entrenched in the theory of tort law. In this book, however, the theor-
etical issues raised by tort law, though present, are not dominant. The
book includes treatment of the areas of contract law, unjust enrich-
ment, restitution, and the law of remedies. It also explores the signifi-
cance of corrective justice for the comparative study of law (using a
specific set of issues from Jewish law), for legal education, and for con-
sidering the connection between property and the state’s obligation to
the poor. It thus presents corrective justice as crucial to understanding
both the normative character and the intellectual significance of the
phenomenon of liability in its many forms.
The key to the structure of reasoning denoted by corrective justice
is, as I shall argue in the opening chapter, the idea of correlativity.
Under this idea the reasons for liability treat the parties’ relationship as
a bipolar unit in which each party’s normative position is intelligible
only in the light of the other’s. The significance of correlatively struc-
tured reasoning is that the justification for regarding something as an
injustice, and consequently for holding the defendant liable to the

3
Izhak Englard, Corrective and Distributive Justice: From Aristotle to Modern Times (2009).
 introduction 3

plaintiff, is the same from both sides. Because such reasoning deals
with the relationship as such and not with its poles considered inde-
pendently, it simultaneously explains both why the plaintiff wins and
why the defendant loses.
The advantages of viewing private law as actualizing this structure of
reasoning are many. First, because the structure focuses on the relation-
ship between the parties rather than on the parties taken separately, it is
fair to both of them. Considerations unilaterally favorable or unfavor-
able to either of the parties play no role. Neither party can, accordingly,
rightly complain of being sacrificed to forward the interests of the
other. Indeed, reasoning within this structure does not refer to their
individual interests as such, but rather to the relational implications of
their interaction. Second, the reasoning is not composed of a hodge-
podge of considerations applying to the parties individually and then
somehow traded off against one another. Rather it is directed toward
treating the relationship as a unified whole, thereby producing reasons
that coherently interlock with one another. Third, these virtues of fair-
ness and coherence provide a standpoint for criticism that is internal to
the relationship and can therefore not be dismissed as irrelevant to the
controversy between the parties or beyond the institutional compe-
tence of the court. Fourth, because any sophisticated legal system
aspires to the fair and coherent treatment of the disputes before it, this
standpoint for criticism is internal to the law, in the sense of being fully
consonant with the law’s own aspirations and with the doctrinal and
institutional arrangements that reflect those aspirations.
The goal of this book, then, is to exhibit how this reasoning works
with reference to the common law, to criticize examples of legal doc-
trine that fail to adhere to the corrective justice structure, and to draw
out the implications of this structure for such activities as legal educa-
tion and comparative law. Thematic is the idea that reasoning within
private law features a distinctive mode of legal thinking and discourse
that reflects the relational nature of liability.
Corrective justice illuminates the structure of legal reasoning by
working back to the content of private law from the adjudicative and
remedial features of liability. Adjudication involves a claim by a par-
ticular plaintiff against a particular defendant. If the plaintiff succeeds,
the remedy awarded requires the defendant to pay damages or give
specific relief to the plaintiff . Both the adjudicative and the remedial
stages of the parties’ controversy thus link them in a bipolar relation-
ship. Corrective justice is the theoretical notion through which the
4 corrective justice

implications of this adjudicative and remedial bipolarity are discerned

in the structure of the norms.The underlying contention is that unless
the norms that define the injustice themselves have a bipolar structure,
the practice of correcting this injustice through bipolar adjudicative
and remedial processes would make no sense. Properly understood,
therefore, corrective justice is not exclusively about the remedy’s role
in rectifying what the plaintiff has suffered at the defendant’s hands.
Rather, it is also about the structure of norms that such bipolar recti-
fication presupposes.
The structure of corrective justice bears on the content of the rea-
soning in private law through two mutually complementary move-
ments of thought. The first is the negative idea that corrective justice
disqualifies any reasoning inconsistent with its relational structure.
Excluded on this ground are considerations, such as a party’s deep
pocket or insurability against loss, that refer to the position of only
one of the parties. Similarly excluded are instrumental considerations
such as the promotion of economic efficiency, for although these may
refer to both of the parties, they relate the two parties not to each
other but to the goal that both parties serve.
The positive idea complementary to this is that reasoning about
liability operates through concepts that are themselves correlative. At
the point of liability, the relevant correlative concepts are those of
right and correlative duty. In private law every right implies that others
are under a duty not to infringe it, and no duty stands free of its cor-
responding right. As I argue in the opening chapter, what is presup-
posed in the rights and duties of private law is the conception of the
person as a free being who has the capacity of setting his or her own
purposes. In the light of this conception of the person, rights and their
correlative duties function as the juridical markers of the freedom of
the parties in relation to each other.
Much of the book is devoted to explicating the way in which rights
so conceived figure in various legal contexts. This involves relating the
highly abstract conceptual apparatus of corrective justice to the
detailed exposition of legal doctrine. The point of blending theory
and doctrine in this manner is to show that corrective justice, far from
being an empty formal category as some have asserted, provides a dis-
tinctive approach to controversial issues across the whole range of
liability.
The book’s exploration of corrective justice proceeds largely by
illustration rather than by the methodical unfolding of corrective
 introduction 5

justice’s nature and implications. Except for the theoretical handling

of the concepts of correlativity and personality in chapter 1, the book
describes a series of particular problems and issues, and then brings
corrective justice to bear on them.4 Because the aim of the present
book is to display the resources that a theory of corrective justice
brings to the treatment of specific issues, it seemed appropriate simply
to move from issue to issue, and then to draw some of the themes
together in the Conclusion. Almost all the issues discussed here have
been matters of legal or academic debate over the last decades. To that
extent, the book attempts to formulate the contribution that an
ancient idea may make to reflection about current controversies.
By connecting rights and their correlative duties to the freedom of
self-determining activity, the book presents an unmistakably Kantian
picture of private law. It should, accordingly, occasion no surprise that
the book makes reference to Kant’s ideas. However, because Kant is so
formidable a figure and his vocabulary so forbidding, it might be help-
ful to specify the limited purpose for which Kant is invoked.
Despite reference to Kant’s ideas of private law in these pages, the
book pays no attention to Kant’s own paramount philosophical ambi-
tion of expounding the metaphysics of legal obligation. Kant’s treat-
ment of the basic categories of private law comes in his work The
Doctrine of Right, which is the first part of his Metaphysics of Morals.
Metaphysics for Kant comprehends “all that can be known a priori,”—
that is, independently of experience; the metaphysics of morals “con-
tains the principles which in an a priori fashion determine and make
necessary all our actions.”5 The Doctrine of Right, accordingly, is an
exposition of the principles of interaction between free beings that
hold a priori. However important Kant’s metaphysical claim about law
might be to students of his philosophy, it is of no interest to lawyers
and plays no role in the exposition of corrective justice in this book.
The use to which this book puts Kant’s ideas is juridical, not meta-
physical. Kant is of interest because of the light he casts on the basic
concepts of contemporary private law. Kant’s writing powerfully elu-
cidates what it means to think of private law systematically in terms of
rights. These rights are not merely conclusions attached to the oper-
ations of positive law, but features of interaction that, by virtue of their

4
A methodical exposition of the theory of corrective justice can be found in my previous
book, The Idea of Private Law (1995).
5
Immanuel Kant, Critique of Pure Reason, tr. Norman Kemp Smith (1929), A841, B869.
6 corrective justice

content, impose obligations on others that are both normatively justi-

fied as markers of reciprocal freedom and legitimately enforceable by
legal institutions. In developing his account of these rights, Kant also
erects a barrier against widespread but questionable assumptions that
are still current today. Among these are the still oft-repeated dogmas
that, because rights (especially property rights) presuppose a distribu-
tion, corrective justice ultimately rests on a distributive foundation, or
that promisees of contractual performance do not have a juridical
entitlement to expectation damages. Of particular importance for this
book’s treatment of contract law and unjust enrichment is Kant’s
explanation of the legally well-entrenched distinction between in rem
and in personam rights. The most explicitly Kantian chapter is the
interpretive effort in chapter 8 to unravel Kant’s conceptions of prop-
erty and public support for the poor. The chapter attempts to explain
as clearly as possible the non-distributive foundation of property in
Kant’s thought and the connection between property and redistribu-
tion. It fits into the overall theme of this book by presenting an
account of property that conforms to corrective justice and by situat-
ing corrective justice within the legal order as a whole. This interpre-
tive chapter aside, the book makes use of Kantian material only to the
extent that it casts light on particular problems in understanding pri-
vate law and can be formulated without reliance on Kant’s metaphys-
ical conceptions.
At the end of the day, one might ask: understood as the actualiza-
tion of corrective justice, what purpose or purposes does private law
serve? The answer to this question depends, of course, on what the
question means.
One possibility is that the question seeks to determine whether
private law (or not insignificant parts of it) should be replaced by dis-
tributive schemes like worker’s compensation or no-fault automobile
insurance. In other words, can private law be justified against alterna-
tive legal arrangements? When the question is understood in this way,
two broad points should be noted. First, the corrective justice approach
to private law, taken on its own, implies nothing about this issue one
way or the other, because the acceptance of corrective justice does not
entail rejection of distributive justice. These two forms of justice are
different ways of ordering legal relations. Accordingly, a given slice of
social life, such as workplace injuries or automobile accidents can be
coherently treated under either form of justice. One can consistently
affirm that, so long as a given class of injuries falls under private law, it
 introduction 7

should be dealt with in accordance with the coherent conception of

private law that corrective justice provides, and yet believe that those
injuries would be better dealt with through a distributive scheme.
Second, the choice between the two forms of justice for dealing with
particular injuries involves consideration of the role of both corrective
justice and distributive justice within a contextually rich account
of the legal character of the state. The issue between corrective and
distributive justice cannot be determined on the basis of corrective
justice alone, for no comparison is possible if only one of the com-
paranda is in view. All that a corrective justice approach to private law
can do on its own is to show the infirmity of the argument that private
law should be replaced because it is inevitably incoherent.6 That is
something, but is not dispositive of the question as a whole. Revealing
the inadequacy of one argument does not preclude the existence of
other and perhaps better ones. And yet no progress on this issue from
a theoretical perspective can be made at all unless the relational nature
of private law is properly appreciated.
A second possibility is that the question about the purpose of pri-
vate law is directed to private law in itself, rather than to the choice
between private law and what might replace it. Then the answer from
the corrective justice perspective is straightforward. The point of pri-
vate law is to subject the interactions between one person and another
to a system of coherent norms that is fair to both. It does this by
viewing the parties as free beings who interact with each other as
holders of rights (to physical integrity, to property, to contractual per-
formance, and so on) that are the juridical manifestations of their free-
dom. These rights are secured through the adjudicative and remedial
processes of coercive legal institutions operating in accordance with
corrective justice’s relational conception of public reason.
Conceived in this way, private law is an inherently normative phe-
nomenon. The corrective justice approach seeks to work out (or to
understand how the law works out) the principles, concepts, and rules
suitable to the relationship between the parties that the adjudicative
and remedial processes presuppose. In exhibiting the structure for this
enterprise, corrective justice indicates the distinct kind of practical
reasoning that supports fair and coherent determinations of liability.
Because the law itself strives to treat the relationship between the

6
Marc A. Franklin, “Replacing the Negligence Lottery: Compensation and Selective
Reimbursement,” (1967) 53 Va. L. Rev. 774.
8 corrective justice

parties fairly and coherently, it is worth taking the reasoning present

in the legal material seriously on its own terms. As will be apparent
from this book, this is not to say that every reason that courts offer
actually conforms to corrective justice. Rather, it indicates the struc-
ture that legal reasoning has to have if that reasoning is to work out
the implications of the private law’s own institutional character and
normative aspirations. That is why, as I put it years ago in words that
scandalized some readers, “the purpose of private law is to be private
law.”7
7
Ernest J. Weinrib, The Idea of Private Law (1995), 5.
 1
Correlativity and Personality

1. Introduction
The point of liability is to undo the injustice that the plaintiff suffers
at the defendant’s hand. Over the last several decades the importance
of this simple and obvious idea—what Aristotle termed “corrective
justice”—has been increasingly recognized. This chapter articulates its
structure and content.
Among English-speaking scholars, the recent history of corrective
justice has been one of eclipse and rediscovery. Formulated by Aristo-
tle in antiquity, nurtured by the scholastic tradition in the Middle Ages,
and then reworked by modern philosophers of natural right, correct-
ive justice had long been a staple of legal theory. However, by the
second half of the twentieth century, instrumental conceptions of law
had largely supplanted it. Having been displaced by policy analysis and
its concomitant intellectual disciplines, the characteristic concepts and
underlying assumptions of corrective justice were no longer part of
the common intellectual inheritance of academic lawyers. In the late
twentieth century interest in corrective justice revived, primarily
through the emergence of tort theory out of the clash between eco-
nomic analysis and its opponents.1 Because corrective justice focuses
on the normative relationship between the parties, it readily appealed
to theorists who viewed tort law as a repository of moral reasoning
about responsibility for injury rather than as a device for promoting
economic goals.

1
Izhak Englard, “The System Builders: A Critical Appraisal of Modern American Tort
Theory,” (1980) 9 J. Legal Stud. 27.
10 corrective justice

The organizing idea of corrective justice is that of correlativity.

Under this idea, liability reflects the conclusion that the defendant and
the plaintiff have respectively done and suffered the same injustice.
Correlativity structures this injustice: the elements of liability can be
explicated only in terms of concepts whose normative force applies
simultaneously to both parties. Liability thus involves a conception of
fairness that recognizes the equal normative status of the two parties
and treats their normative positions as mirror images of each other.
Corrective justice is exposed to two potential misconceptions. The
first misconception is that corrective justice is a substantive rather than
a structural principle. A substantive principle directly presents a pro-
posed content for legal doctrine. This content may have a limited
scope that comprehends certain bases of liability but not others.
Viewed as a substantive principle, corrective justice might be thought,
for instance, to illuminate only particular doctrines of tort law but
neither tort liability generally nor other bases of liability such as con-
tract law or unjust enrichment. In contrast, a structural principle refers
to a pattern of argument to which the content (whatever it is) of pri-
vate law should conform. Viewed as structural principle, corrective
justice applies throughout the whole law of obligations, on the
grounds that correlativity is built into the very structure of liability as
a normative phenomenon directly linking plaintiff and defendant. The
second misconception is that corrective justice is so abstract an idea
that it contributes little to the specification of the content of private
law, or that it is concerned only with the remedial mechanism through
which wrongs, whatever they are, are corrected. This ignores the close
connection between corrective justice and a robust conception of
rights. These rights and the duties correlative to them constitute the
content of private law across the various bases of liability.
In this chapter I address both of these potential misconceptions. In
response to the first, I examine the normative significance of correla-
tivity as a pervasive feature of liability. In response to the second,
I trace the conceptual relationship between the correlativity of cor-
rective justice and liability as a regime of rights. In particular, I focus
on the idea of personality as the abstraction through which the nature
of these rights is understood.
I thereby present a juridical conception of corrective justice. The
conception is juridical in the sense that it reflects, though at an abstract
level, the justifications internal to liability in private law, treating them
as normative in their own terms rather than as the disguised surro-
 correlativity and personality 11

gates for extrinsically justifiable social goals. The juridical conception

views the determination of liability as a distinctive domain of practical
reason that subjects the interaction between the plaintiff and the
defendant to a coherent ordering that is fair to both of them. Because
legal argument attests to the law’s self-reflective engagement with its
own fairness and coherence, the principles and concepts already
present to private law can provisionally be regarded as constituents of
that ordering. Thus, in its endeavor to exhibit the normative ideas
interior to a fair and coherent regime of liability, the juridical concep-
tion of corrective justice draws on the law’s own efforts.
Fundamental though it is, correlativity is not the only component
of the juridical conception of corrective justice. That conception also
features a distinct notion of the person that philosophers in the nat-
ural right tradition have termed “personality.” Personality in this con-
text is not a psychological but a normative idea: it refers not to the
pattern of an individual’s behavioral characteristics, but to a presup-
position about imputability and entitlement that is implicit in the
rights and duties of private law. This presupposition is that, as partici-
pants in a regime of liability, the parties are viewed as purposive
beings who are not under duties to act for any purposes in particular,
no matter how meritorious.This capacity for purposive action under-
lies the rights and duties that are its juridical manifestations. In not
requiring action for any particular purpose, personality reflects the
structure of the law of obligations as a system of negative duties of
non-interference with the rights of others. This does not mean that
so circumscribed a notion of duty is exhaustive of one’s obligations
in all moral contexts. Rather, personality encapsulates a normative
standpoint that is indigenous to private law.
In the juridical conception of corrective justice, correlativity and
personality are complementary ideas. They are the mutually entailed
parts of a single conception but they highlight different aspects of it.
Just as correlativity is the most abstract representation of the terms on
which the parties interact in private law, so personality is the most
abstract representation of the parties themselves as interacting beings.
And just as correlativity exhibits the structure of the justifications that
pertain to private law, so personality articulates the presupposition that
informs the content of those justifications. Correlativity and personal-
ity pass over the same theoretical ground from different directions.
Accordingly, one would have expected that the increasing academ-
ic recognition of correlativity would be accompanied by a similar
12 corrective justice

recognition of personality. That, however, has not occurred. This is

because of the apprehension that personality, with its roots in the
natural rights philosophies of Kant and Hegel, implies both a philo-
sophical claim about the truth of Kant’s or Hegel’s conception of
rational agency2 and a methodological claim that the theory of pri-
vate law is derived from a more comprehensive philosophical pro-
gram. They reject personality because they reject what they take to
be these further implications.
In this chapter I point out that these further implications do not
follow from the presence of personality within the juridical concep-
tion of corrective justice. Personality articulates at the high level of
abstraction what is implicit in private law as a regime of rights and
their correlative negative duties. Personality’s status within the juridi-
cal conception is no different from that of correlativity. With respect
to both personality and correlativity, the juridical conception operates
by working back from the principles and concepts of private law to
the most general ideas latent within them. Thus the juridical concep-
tion regards corrective justice neither as deriving from nor as imply-
ing the philosophical truth of Kant’s or Hegel’s conception of rational
agency. Of course, this does not exclude reference to Kantian and
Hegelian texts and ideas as a source of insight about the nature of pri-
vate law when understood as corrective justice. Nor does it exclude
the possibilities of deriving corrective justice from, or establishing the
truth of, rational agency; consideration of those possibilities, however,
would require philosophical argument that lies beyond the bounds of
the theory of private law and does not affect it. What the juridical
conception of corrective justice asserts is merely that correlativity and
personality are implicit in private law as complementary ideas, so that
accepting one of them is inconsistent with rejecting the other.

2. The complementary abstractions

The juridical conception of corrective justice is the centerpiece of a
theory of liability. The object of the theory is to understand liability as
a distinct and familiar normative practice, in which the law assesses
2
The term rational agency in this connection refers to Kant’s notion that the will ration-
ally generates rules for its own conduct or Hegel’s notion that the will realizes itself by will-
ing, through the various stages of normative experience, a content that is appropriate to its
own rational form.
 correlativity and personality 13

and responds to the claim that a plaintiff makes against a defendant.

Considered as a normative practice, liability includes both the legal
consequences that a court might impose and the grounds that justify
those legal consequences.
Within this practice justification has a pervasive role. The rules,
concepts, and principles that figure in the assessment of the plaintiff ’s
claim are the ingredients and the products of a justificatory process.
Moreover, the institutions of liability are designed for the presentation
of these justifications and for giving effect to the conclusions that may
be drawn from them. Consequently, the normative significance of a
finding of liability depends on the cogency of the justifications that
support it.
The juridical conception of corrective justice takes the justificatory
ambitions of this practice seriously by focusing on its internal norma-
tive dimension. The juridical conception repudiates the idea that the
justifications that figure in private law derive from goals that are desir-
able independently of the role that they can coherently play in a lia-
bility regime. Instead, its eye remains fixed on the practice itself, on
the institutional structure through which it unfolds, and on the rea-
soning through which it expresses its distinctive mode of justification.
The juridical conception of corrective justice thus honors the law’s
reasoning as a good faith attempt—sometimes successful and some-
times not—to make the exercise of official power the product of an
internally coherent ensemble of justificatory considerations.
The aim of the juridical conception is to disclose the structure and
the normative presuppositions of the law’s internal processes of justifi-
cation. It takes the doctrinal and institutional features that are charac-
teristic of a regime of liability, and asks what must be presupposed
about them and about their interconnection if the law is to be (as it
claims) a coherent justificatory enterprise. The answer lies in identify-
ing the most abstract unifying conceptions implicit in the doctrinal
and institutional arrangements of private law. Thus the juridical con-
ception of corrective justice purports to bring to the surface ideas that
are latent in liability as a normative practice.
Within the juridical conception the movement of thought is from
the particulars of liability to its most abstract characterization, thus
carrying to its extreme the tendency to abstract that marks legal think-
ing. Although the events that give rise to a legal relationship are par-
ticular, the law treats these events in terms of categories. The
particularities of the events are legally relevant only inasmuch as they
14 corrective justice

instantiate a category applicable to the legal relationship to which they

give rise. Just as legal thinking views particulars in the light of these
categories, so the juridical conception of corrective justice abstracts
further from these categories to the barest and most general ideas
underlying the law’s construction of the parties’ relationship.
The juridical conception of corrective justice gives voice to the
internal structure of a liability regime by presenting its doctrinal and
institutional features as the specifications of its most pervasive and
general characteristics. If these characteristics can be understood as
expressing a set of unifying and complementary ideas, the liability
regime will be seen as coherent to the extent of its participation in
those ideas. When presented abstractly, these ideas afford an unclut-
tered view of the nature of liability, because they pertain to liability as
such without being confined to any particular doctrine. Their very
abstractness brings into view the systematic connections that might
obtain among the considerations adduced to support the manifold
features of liability. Moreover, they provide a critical perspective intern-
al to the law, because justifications that do not fit within these unify-
ing ideas are problematic from the standpoint of liability itself.
These ideas emerge from the notion that liability is justified when a
certain kind of connection obtains between the parties. This descrip-
tion of liability indicates that a theory of liability must comprehend
two general features. The first is the nature of the connection between
the parties: what is it that allows the law to single out two specific
parties from all the people in the world and link them as plaintiff and
defendant? The second is the nature of the parties: in view of the
diversity of human interests and characteristics, what conception of
the parties is normatively capable of serving as the basis of the defend-
ant’s liability to the plaintiff? The unifying ideas implicit in liability
are the answers, formulated in the most abstract terms, to these two
questions.
The conceptions of the parties and of the connection between
them are mutually complementary. In dealing with liability, we are
interested in the parties only because of the normative connection
between them; and we are interested in that connection only because
the parties are normatively capable of association in terms of liability.
Accordingly, the parties must be conceptualized in a way that makes
liability the necessary mode of connecting them; and the connection
between the parties must be conceptualized in a way that makes ne-
cessary a certain construal of what, from the standpoint of liability, is
 correlativity and personality 15

normatively significant about them. These two unifying ideas are thus
the same idea presented under different aspects. Indeed, if they were
not the same idea, the legal phenomenon to which they apply would
have to be understood not as a unity but as a pluralism of at least
two independent ideas. This conclusion would defeat the aspirations
of the juridical conception by indicating that liability—and the justi-
ficatory considerations that underlie it—was incapable of being
understood as an internally coherent whole.
In the juridical conception of corrective justice, the two comple-
mentary ideas are correlativity and personality. Correlativity, which
was first highlighted in Aristotle’s account of corrective justice,3 is the
abstract formulation of the connection that exists between the parties
in a regime of liability. Personality, which was most fully articulated in
the philosophical tradition of natural right that culminated in Kant
and Hegel,4 presents in similarly abstract terms what is normatively
significant about the interacting parties for purposes of liability.
Although the two ideas are ultimately congruent, they start from dif-
ferent aspects of liability. Correlativity abstracts from the institutional
framework of the plaintiff–defendant lawsuit and inquires into the
structure of the justifications that coherently fit into this institutional
framework. Personality abstracts from the doctrinal framework of
rights and duties in order to exhibit the content of private law justifi-
cation in its most general form; it then extends its attention to the
institutions that give coherent effect to that doctrinal framework.
Together, correlativity and personality are the interlocking foundation
stones of a theory of liability.

3. Correlativity
Aristotle presents corrective and distributive justice as two contrasting
forms of justice. Corrective justice, which deals with voluntary and
involuntary transactions (today’s contracts and torts), focuses on
whether one party has done and the other has suffered a transactional

3
Aristotle, Nicomachean Ethics,V, 4.
4
Immanuel Kant, The Metaphysics of Morals, tr. Mary Gregor, in The Cambridge Edition of
the Works of Immanuel Kant: Practical Philosophy (1996), [6:242]. Numbers in square brackets
refer to the pagination found in the standard German edition of this work and reproduced
in the margins of The Cambridge Edition. G. W. F. Hegel, Philosophy of Right, tr. T. M. Knox
(1952), ss. 34–40.
16 corrective justice

injustice. Distributive justice deals with the distribution of whatever is

divisible (Aristotle mentions honors and goods) among the partici-
pants in a political community. For Aristotle, justice in both these
forms relates one person to another according to a conception of
equality or fairness (the Greek to ison connotes both). Injustice arises
in the absence of equality, when one person has too much or too little
relative to another.
The two forms differ, however, in the way they construe equality.
Distributive justice divides a benefit or burden in accordance with
some criterion that compares the participants’ merit relative to one
another. Distributive justice therefore embodies a proportional equal-
ity, in which all participants in the distribution receive their shares
according to their respective merits under the criterion in question.
Corrective justice, in contrast, features the maintenance and restor-
ation of the notional equality with which the parties enter the trans-
action. This equality consists in persons having what lawfully belongs
to them. Injustice occurs when, relative to this baseline, one party real-
izes a gain and the other a corresponding loss. The law corrects this
injustice when it re-establishes the initial equality by depriving one
party of the gain and restoring it to the other party. Aristotle likens the
parties’ initial positions to two equal lines.5 The injustice upsets that
equality by adding to one line a segment detached from the other.
The correction removes that segment from the lengthened line and
returns it to the shortened one.The result is a restoration of the origin-
al equality of the two lines.
As its name indicates, corrective justice has a rectificatory function.
By correcting the injustice that the defendant has inflicted on the
plaintiff, corrective justice asserts a connection between the remedy
and the wrong. From the perspective of corrective justice, a court does
not treat the situation being adjudicated as a morally neutral given
and then ask what is the best course for the future, all things consid-
ered. Rather, because the court aims to correct the injustice done by
one party to the other, the remedy responds to the injustice and
endeavors, so far as possible, to undo it.
Aristotle’s account makes it clear that this rectification operates cor-
relatively on both parties. A remedy directed to only one of the parties
does not conform to corrective justice. For the court merely to take
away the defendant’s wrongful gain does not suffice, because then the

5
Aristotle, above n. 3, at 1132b6.
 correlativity and personality 17

plaintiff is still left suffering a wrongful loss. Nor does it suffice for the
court merely to replenish the plaintiff ’s loss, for then the defendant
will be left holding his or her wrongful gain. The remedy consists in
simultaneously removing the defendant’s excess and making good the
plaintiff ’s deficiency. Justice is thereby achieved for both parties
through a single operation in which the plaintiff recovers precisely what
the defendant is made to surrender.
From these two features of the corrective justice remedy—that it
responds to the injustice and is correlatively structured—a third fol-
lows. A correlatively structured remedy responds to and undoes an
injustice only if that injustice is itself correlatively structured. In bring-
ing an action against the defendant the plaintiff is asserting that they
are connected as doer and sufferer of the same injustice. As is evi-
denced by the judgment’s simultaneous correction of both sides of the
injustice, what the defendant has done and what the defendant has suf-
fered are not independent items. Rather, they are the active and pas-
sive poles of the same injustice, so that what the defendant has done
counts as an injustice only because of what the plaintiff has suffered,
and vice versa. The law then rectifies this injustice by reversing its
active and passive poles, so that the doer of injustice becomes the suf-
ferer of the law’s remedy. Only because the injustice is the same from
both sides does the remedy treat the parties as correlatively situated.
Thus throughout the transaction, from the occurrence of the injustice
to its rectification, each party’s position is normatively significant only
through the position of the other, which is the mirror image of it.
The idea that correlativity informs the injustice, as well as its recti-
fication, is a central insight of the juridical conception of corrective
justice. This insight points to the kind of justifications that are appro-
priate for determinations of liability.To think of something as an injust-
ice is not to refer to a brute event but to make a normative ascription.
The correlativity of the injustice is, therefore, the correlativity of the
normative considerations that underlie that ascription. Because the
defendant, if liable, has committed the same injustice that the plaintiff
has suffered, the reason the plaintiff wins ought to be the same as the
reason the defendant loses. Thus in specifying the nature of the injust-
ice, the only normative factors to be considered significant are those
that apply equally to both parties. A factor that applies to only one of
the parties—for example, the defendant’s having a deep pocket or
being in a position to distribute losses broadly—is an inappropriate
justification for liability because it is inconsistent with the correlative
18 corrective justice

nature of the liability. Accordingly corrective justice not only rectifies

injustice in transactions; by structuring the justificatory considerations
relevant to transactions, it is also regulative of the notion of injustice
that is applicable to them.
Thus, correlativity is the structural idea that underlies the most
obvious and general feature of liability, that the liability of the defend-
ant is always a liability to the plaintiff. Liability consists in a legal rela-
tionship between two parties each of whose position is intelligible
only in the light of the other’s. In holding the defendant liable to the
plaintiff, the court is making not two separate judgments (one that
awards something to the plaintiff and the other that coincidentally
takes the same from the defendant), but a single judgment that embra-
ces both parties in their interrelationship. The defendant cannot be
thought of as liable without reference to a plaintiff in whose favor
such liability runs. Similarly, the plaintiff ’s entitlement exists only in
and through the defendant’s correlative obligation. The court’s finding
of liability is the response to an injustice that, accordingly, has the
same correlative shape as liability itself.
Justifications for holding someone liable that exhibit the parties as
the doer and sufferer of the same injustice render the law both coher-
ent and fair. Legal reasoning composed of such justifications treat the
parties’ relationship as a normative unity that embraces them both
rather than as a hodgepodge of factors separately relevant only to one
or the other of them. With this justificatory coherence comes fairness
as between the parties. A justification that fails to match the correla-
tive structure of the parties’ relationship necessarily favors one of the
parties at the expense of the other, thereby failing to be fair from the
standpoint of both. In contrast, by insisting that the normative consid-
erations applicable to liability reflect the parties’ correlative situation,
corrective justice construes private law as setting terms for the parties’
interaction that take account of their mutual relationship and are
therefore fair to both of them.
Aristotle’s original account contrasts the correlativity of corrective
justice with the categorically different structure of distributive justice.
Corrective justice links the doer and sufferer of an injustice in terms
of their correlative positions. Distributive justice, on the other hand,
deals with the sharing of a benefit or burden; it involves comparing
the potential parties to the distribution in terms of a distributive criter-
ion. Instead of linking one party to another as doer and sufferer, dis-
tributive justice links all parties to the benefit or burden they all share.
 correlativity and personality 19

The categorical distinction between correlativity and comparison is

certified by the difference between the numbers of parties that each
admits. Corrective justice links two parties and no more, because a
relationship of correlativity is necessarily bipolar. Distributive justice
admits any number of parties, because in principle no limit exists for
the number of persons who can be compared and among whom
something can be divided.
The consequence of Aristotle’s contrast between corrective and
distributive justice is that no distributive consideration can serve as a
justification for holding one person liable to another. The correlative
structure of liability entails the irrelevance of any factor that is nor-
matively significant only because of its possible role in a distributive
comparison. For purposes of justifying a determination of liability,
corrective justice is independent of distributive justice.
Accordingly, the idea of correlativity brings out the interior struc-
ture of justification within the connection that liability forges between
a particular plaintiff to a particular defendant. In considering liability
we might wonder: why is the plaintiff entitled to recover from this
defendant rather than from someone more evil, or why is the defend-
ant held liable to this particular plaintiff rather than to someone more
needy? The correlative nature of liability shows that such questions
are misplaced. Evil and need are moral categories that may well figure
in other normative contexts, but they are not pertinent to liability. It
may make sense as a matter of distributive justice, for instance, to
divide benefits or burdens on the basis of a comparison of relative vir-
tue or need. Virtue and need, however, do not connect any two
particular persons as correlatively situated.
The irrelevance of virtue and need as justificatory considerations
for liability exemplifies a wider principle that applies to anything
whose normative significance is not correlatively structured. In insist-
ing that correlativity is the general organizing feature for justifications
of liability, corrective justice requires practical reason to adopt and
maintain a special posture, or to be shifted, as it were, into a special
normative gear,6 to which every aspect of liability must conform.
For the law of obligations, the overarching justificatory categories
expressive of correlativity are those of the plaintiff ’s right and the

6
The metaphors of posture and gear are adopted from the illuminating discussion of cor-
relativity by Michael Thompson, “What Is It to Wrong Someone? A Puzzle about Justice,” in
Reason and Value, ed. R. J. Wallace et al. (2004), 333, 346.
20 corrective justice

defendant’s corresponding duty not to interfere with that right. Unlike

need and virtue, a right is an intrinsically relational idea that immedi-
ately signifies the existence of a duty correlative to it. Right and duty
are correlated when the plaintiff ’s right is the basis of the defendant’s
duty and, conversely, when the scope of the duty includes the absten-
tion from the kind of right-infringement that the plaintiff suffered.
Under those circumstances the reasons that justify the protection of
the plaintiff ’s right are the same as the reasons that justify the exist-
ence of the defendant’s duty.
Negligence law provides a paradigmatic example of the operation
of this correlativity in the common law. For the defendant to be held
liable, it is not enough that the defendant’s negligent act resulted in
harm to the plaintiff . The harm has to be to an interest that has the
status of a right, and the defendant’s action has to be wrongful with
respect to that right.7 In Justice Cardozo’s words, “What the plaintiff
must show is ‘a wrong’ to herself; i.e., a violation of her right, and not
merely a wrong to someone else, nor conduct ‘wrongful’ because
unsocial but not ‘a wrong’ to any one.”8 Under the condition stated by
Cardozo, freedom from the injury of which the plaintiff is complain-
ing is both the content of the plaintiff ’s right and the object of the
defendant’s duty. Then correlativity obtains, because the parties are the
doer and sufferer of the same injustice, and the reason for the plain-
tiff ’s entitlement to win the lawsuit would be the same as the reason
for the defendant’s liability to lose it.
The identity of the content of the right and the object of the duty
informs what it means to do and suffer the same injustice. One suffers
an injustice when one’s right is infringed. One does an injustice when
one breaches one’s duty not to infringe another’s right.This doing and
suffering come together in the same injustice when the duty breached
is correlative to the very right that was infringed. Given the correla-
tive positions of the parties, imputing the breach of such a duty to the
defendant justifies his or her liability to the plaintiff . For the juridical
conception of corrective justice, then, an injustice is an inconsistency
with the plaintiff ’s right that is imputable to the defendant.
The various branches of civil liability work out the circumstances
for such imputability. As Cardozo’s statement illustrates, the concepts
and principles of tort liability set out the conditions under which the

7
See below, chapter 2.
8
Palsgraf v. Long Island Railroad Co., 162 N.E. 99 (N.Y.C.A., 1928), at 100.
 correlativity and personality 21

defendant’s conduct counts as a wrongful infringement of the plain-

tiff ’s right. Similarly, the central doctrines of the law of contract (offer
and acceptance, consideration, unconscionability, and expectation
damages) allow the parties, through the mutuality of their conduct, to
establish in the plaintiff a right to the defendant’s performance of the
promised act.9 And the law dealing with unjust enrichment at anoth-
er’s expense gives legal specificity to the plaintiff ’s right to the defend-
ant’s restitution of a gratuitous benefit that was neither given nor
accepted donatively.10

4. Personality
The juridical conception of corrective justice abstracts to the most
pervasive organizing ideas of private law. Correlativity is one such
organizing idea. It is the abstraction that presents the many particular
relationships between parties in private law—and the justifications
applicable to those relationships—in terms that match the bipolar
nature of liability. By attending to correlativity, the juridical concep-
tion of corrective justice concludes that, first, to the extent that private
law is fair and coherent, the justifications underlying liability evince a
correlative structure and, second, right and duty are the normative
categories that conform to this structure. The question that then arises
is whether the rights and duties of private law in turn fall under a par-
allel abstraction. In other words, can the normative content of the
parties’ positions, as well as the correlative structure of their relation-
ship, be illuminated through some abstraction that coherently organ-
izes the manifold of liability?
It goes without saying that this abstraction must be fully consistent
with the conception of liability as correlative, because the content of
the rights and duties has to fit the correlative structure that they
instantiate. This desideratum means that rights cannot be conceived
merely as shorthand references to components of the plaintiff ’s wel-
fare, for then rights will ultimately be as unexpressive of correlativity
as welfare itself. From the standpoint of corrective justice, the advance

9
On the nature of contractual right, see below chapter 5. On the relationship among the
central doctrines of contract law, see Peter Benson, “The Unity of Contract Law,” in The
Theory of Contract Law: New Essays, ed. Benson (2001), 118.
10
Below, chapter 6.
22 corrective justice

achieved by introducing the notion of right and correlative duty

would be undone by the non-correlative conception of the content
of those rights as being based on welfare.
Thus, under the juridical conception of corrective justice, rights are
not normatively significant for private law simply by virtue of the fact
that they enhance the plaintiff ’s welfare. Of course, having a right
contributes to a person’s welfare by protecting some interest from
wrongful interference. And it is also true, of course, that private law
responds to the infringing a right by measuring diminutions of the
right-holder’s welfare under the legally recognized heads of damages.
That, however, does not mean that rights are synonymous with aspects
of welfare or that their normative significance is to be understood in
terms of it. In the law’s contemplation, the increase in welfare through
having a right and the decrease through the infringement of a right
are the consequences rather than the grounds of the right. That is why
(as in cases of negligently caused economic loss)11 a decrease in wel-
fare that does not violate the plaintiff ’s rights is not actionable, and
conversely (as in cases of nominal damages) a violation of a right that
does not decrease the plaintiff ’s welfare is. Welfare serves only the sec-
ondary function, only injustice has occurred and a remedy is being
considered, of quantifying the injury done to the object of the right.
The reason that rights matter for private law lies elsewhere.
The insignificance of welfare as such is also evident on the duty
side. Tort law, for instance, is a regime of negative duties that mandate
non-interference with the rights to which they are correlative. If the
rights themselves represented aspects of welfare, one would expect
that at least certain aspects (human life itself, for instance) would be so
important that, at least in an extreme circumstance, persons would be
under a positive duty to promote or preserve them. As the law’s dis-
tinction between misfeasance and nonfeasance indicates, this is not
the case.The promotion of another’s welfare, even the saving of anoth-
er’s life in an emergency, counts merely as the bestowal of a benefit
and therefore not obligatory.The fact that private law deems no aspect
of welfare important enough to ground a positive obligation to for-
ward it indicates that welfare as such is not what is significant about
the rights that the law recognizes.
11
Peter Benson, “The Basis for Excluding Liability for Economic Loss in Tort Law,” in
Philosophical Foundations of Tort Law, ed. David Owen (1995), 427 (analyzing tort law’s reluc-
tance to recognize economic loss as a basis of liability in the absence of an infringement of a
right).
 correlativity and personality 23

The upshot of these considerations is that welfare does not lie at

the heart of the abstraction that might capture the organizing idea
implicit in the rights and duties of private law. What does?
Correlativity can serve as a model for what an answer to this ques-
tion involves. As an abstraction aimed at illuminating the nature of
liability, correlativity presents no novel claims. Nor does it reflect some
obscure or hitherto unnoticed consequence of private law that,
through virtuoso analysis, is worked up to disclose a mysterious but
supposedly comprehensive truth.12 The significance of correlativity is
not in the heavens but before our very noses, inescapably present
whenever we consider liability. In its reference to correlativity correct-
ive justice merely takes the most manifest aspect of private law—that
liability links plaintiff and defendant—and works out its theoretical
implications. This it does by extrapolating from the plethora of private
law causes of action to the feature that is pervasive in and presupposed
by all of them. Similarly, in articulating the abstraction pertinent to
the content of rights and duties, corrective justice merely looks for
what is pervasively present in particular rights and duties.
What is present is the parties’ capacity for purposiveness, which
assumes legal significance when externalized. On the duty side, this is
evident in the trite doctrine of tort law that the defendant cannot be
liable in the absence of an “act,” defined as an external manifestation
of the volition.13 Mere physical movement is irrelevant to liability;
there must be the outward expression of an inwardly determined pur-
pose. However, although tort law insists that the wrong originates in
an exercise of purposiveness, it does not condition liability on the fail-
ure to act for a particular purpose, however meritorious. As the non-
feasance doctrine attests, tort law makes obligatory no particular
purposes, not even purposes that would preserve another’s most fun-
damental interests. Thus the indispensable presupposition of the
defendant’s having breached a duty is the sheer purposiveness of the
defendant’s action, rather than a conception of some set of particular
purposes that should have guided the defendant’s conduct.
The same picture appears in connection with rights. The acquisi-
tion or transfer of a right involves the exercise, in a legally cognizable
manner, of the acquirer’s or the transferor’s purposiveness toward the
subject matter of the right. As in the case of acts that violate rights, an

12
In this respect, contrast the economic analysis of law.
13
Restatement (Second) of Torts, s. 2 (1965).
24 corrective justice

external manifestation of the volition is necessary if the law is to

ascribe consequences to what one has done. Moreover, the legal basis
for the thing’s being used in the exercise of the owner’s purposive
capacity is what acquisition creates (and transfer terminates). Accord-
ingly, the law regards a right as the power to treat something as sub-
ject to one’s will as a consequence of an antecedent connection that
one’s will has established with the thing in question. Yet the law
regards as irrelevant the specific purpose that motivates the acquisi-
tion, transfer, or use. Nor does it require that the right, once acquired,
be used for any particular (and arguably laudable) purposes, such as
to increase the utility of all or to maximize wealth or to produce an
equality of resources. Of course, the acquisition, transfer, and use of
one’s entitlements is fueled by one’s particular needs, interests, and
desires, but the law pays these no heed when determining the entitle-
ments’ validity. The law responds merely to the external indicia of
an exercise of purposiveness, rather than to a schedule of required or
desirable purposes.
Thus, in the acquisition, transfer, use, and violation of a right, pri-
vate law presupposes the exercise of one’s capacity for purposive
activity without specifying any set of necessary purposes. In this
respect, a liability regime differs from other normative domains.
While morality, for example, also presupposes purposiveness in the
moral actor, it aspires to specify particular purposes essential to living
a moral life. Similarly, distributive justice requires the identification,
at least at some level of generality, of the particular purposes at which
a distribution ought to aim. In contrast, the presupposition of purpos-
iveness without regard to particular purposes is specific to a regime
of liability.
As the basis of the private law’s attention to the parties, purposive-
ness without regard to particular purposes defines the conception of
the person that underlies liability. This conception is what the natural
right tradition called “personality.”14 Just as the ancient Roman legal
texts use “person” when discussing the indicia of one’s legal standing,
so personality refers to the capacity for purposive agency that forms
the basis for the capacity for rights and duties in private law. Personal-
ity encapsulates the normative standpoint from which private law has
to view the parties if it is to regard them as having its rights and being
subject to its duties.

14
See especially Hegel, above n. 4.
 correlativity and personality 25

The rights and duties in a coherent liability regime specify the

manifestations of personality in the parties’ legal relationship. Because
personality signifies the capacity for purposiveness without regard to
particular purposes, no obligation exists to exercise this capacity
toward any particular end. Any duties that reflect personality are there-
fore the negative correlates of rights. These rights arise insofar as the
capacity for purposive agency is not merely an inward attribute but
achieves external existence in social interactions through its exercise
by or embodiment in an agent. Among these rights are the right to
the integrity of one’s body’s as the organ of purposive activity, the
right to property in things appropriately connected to an external
manifestation of the proprietor’s volition, and the right to contractual
performance in accordance with the mutually consensual exercises of
the parties’ purposiveness. The existence of these rights gives rise to
correlative duties of non-interference whose content and application
depend on the nature of the right. Moreover, these rights and duties
are actualized through a set of judicial institutions that endows them
with a determinate shape, makes public the mode of reasoning that
accords with what is presupposed in them, and undoes the conse-
quences of conduct inconsistent with them.
Personality is the abstraction that captures the conception of the
person presupposed in the justifications appropriate to the doer–
sufferer relationship. The correlativity of doing and suffering means
that welfare as such, because it is not correlatively structured, does not
serve as the basis of liability; accordingly, the agency of the parties can-
not be conceived in terms of the importance to them of their welfare.
Instead, the parties’ interaction presupposes a conception of the person
in which doing and suffering are each normatively significant because
of their relationship to the other. Personality offers such a conception.
From the standpoint of the doer of injustice, personality as the cap-
acity for purposiveness contains the indispensable conditions for the
ascription of responsibility for the effects of one’s action. From the
standpoint of the sufferer, personality is the basis of the rights that
mark out a sphere that others must treat as inviolate. Injustice occurs
when this purposiveness is actualized on both sides of this relationship,
through the right infringed and the correlative duty breached. Within
the parties’ interaction personality thus operates in a correlative
mode.
Personality illuminates the injustice done and suffered as the
infringement of a right. Not every harm or disadvantage counts as an
26 corrective justice

injustice for purposes of corrective justice. The mere doing and suffer-
ing of something that adversely affects the sufferer is not in itself sig-
nificant. What is done and suffered must be an injustice, with the
specific kind of injustice being the infringement of a right. Personality
models this kind of injustice by presenting a conception of the parties
for whose normative status welfare does not matter except insofar as it
forms the content of a right. For persons conceived in this way, being
better or worse off does not in itself constitute an injustice. In this
respect, the idea of personality replicates at an abstract level the law’s
own distinction between damnum and iniuria.
Thus, personality is the conception of the parties formulated at a
high degree of abstraction. This conception organizes and unifies the
justificatory considerations that pertain to liability. It organizes them
by exhibiting the content of rights and duties as expressive of the
capacity for purposiveness that they presuppose. It unifies them by
exhibiting the single abstraction implicit in private law’s multiplicity
of rights and duties. Personality can then function as a conception to
which the legal reasoning that underlies the various rights and duties
must conform if they are to fulfill the possibilities for coherence that
are latent within them.
Its attention to personality brings the juridical conception of cor-
rective justice into contact with the natural right philosophies of Kant
and Hegel. These philosophies locate personality in private law and
then, in their different ways, reconstruct the entire normative realm
on the basis of conceptions of rational agency for which personality is
the indispensable first stage. The juridical conception draws on the
notion of personality articulated in these philosophies. One might
conclude that the juridical conception of corrective justice thereby
purports to derive the theory of private law from a supposedly correct
philosophical account of rational agency. That conclusion would be a
mistake for two reasons.
First, the juridical conception of corrective justice does not pro-
ceed by postulating a conception of agency and then deriving the
theory of private law from it. Rather, the juridical conception always
works backward from the doctrines and institutions of private law to
the most pervasive abstractions implicit in it. These abstractions (cor-
relativity and personality) then serve to illuminate the justifications
that figure in a fair and coherent private law. The argument moves
from private law as a normative practice to its presuppositions, which
then serve as vehicles of criticism and intelligibility that are internal
 correlativity and personality 27

to the practice. Personality matters to the theory of private law not

because personality is the source from which that theory is derived,
but because personality is latent in the normative practice that the
theory aims to comprehend.15
Second, the juridical conception does not depend on the correct-
ness of any philosophical account of rational agency. The juridical
conception of corrective justice is concerned only with the normative
perspective specific to private law, not with rational agency as such. To
be sure, in their accounts of rational agency Kant and Hegel formu-
lated the notion of personality and analyzed its significance for private
law. Their expositions of the rights and duties of corrective justice are
exemplary in their systematic elaboration of the relationship among
various of the concepts that figure in a regime of liability. Because
Kant and Hegel articulate their differing accounts under the most rig-
orous constraints of coherence, they provide a repository of insights,
often ignored by English-speaking scholars, about the nature of coher-
ence within legal relationships and about the role of specific doctrines
in achieving that coherence. However, their treatments of personality
as an aspect of rational agency would have been nugatory even for
their own purposes unless personality was indeed immanent in the
conceptual structure of private law. It is this immanence that the
juridical conception highlights.The Kantian and Hegelian accounts of
private law are available to the juridical conception not because they
reflect the correct view of rational agency (though one or the other
of them may do so), but because they bring out the internal connec-
tions integral to a coherent understanding of liability.
Of particular interest to the judicial conception of corrective just-
ice are the ways in which Kant and Hegel connect personality to the
various kinds of rights in private law. Contemporary scholars often
assume that the rights presupposed by corrective justice are themselves
a product of distributive justice. On this view, corrective justice is
alleged to rest on a distributive basis, so that the normative autonomy
that corrective justice claims for private law is undermined right from
the beginning.16 Each in his own way, the two great thinkers of the

15
For an example of the mistaken view that the juridical conception works by derivation,
see Jules Coleman, Risks and Wrongs (1992), 478, n. 1. For a description of the movement of
thought in the juridical conception as a process of working back from the juristic experience
of private law, see Ernest J. Weinrib, The Idea of Private Law (1995), 19.
16
For example, Hanoch Dagan, “The Distributive Foundations of Corrective Justice,”
(1999) 98 Mich. L. Rev. 138.
28 corrective justice

natural right tradition provide an antidote to this assumption. The

fundamental rights of private law, including property and contract, get
entrenched in their accounts not through an exercise of distributive
justice but through an exploration of the normative implications of
personality within a regime of corrective justice. Their presentations
are closely tied to philosophical premises that, as such, are not part of
the juridical conception of corrective justice. Nonetheless, their
accounts of these rights are available to the juridical conception to the
extent that they reflect fundamental features of the private law.17
Within the juridical conception these accounts can then be used not
only to disabuse scholars of the notion that the rights of private law
are inescapably distributive, but also to assist in drawing out what, from
the standpoint of corrective justice, are the theoretical implications of
the features in question. In this respect, too, the Kantian and Hegelian
accounts matter to the juridical conception of corrective justice not
because it is derived from them but because they are serviceable to its
theoretical purposes.
Thus, although the presence of personality within the juridical con-
ception takes its inspiration from the Kantian and Hegelian elucida-
tions of rational agency, the juridical conception itself has no stake in
rational agency. Rational agency is constitutive of the entire normative
order of values. The juridical conception of corrective justice, in con-
trast, concerns itself only with values that reflect the distinctive nature
of justification in private law. It therefore views the capacity for purpo-
siveness simply as the moral power that is implicit in the rights and
duties of private law. No more ambitious claim is necessary for its pur-
poses. It may well be (as Kant and Hegel thought) that personality so
conceived is after all an aspect of rational agency and that rational
agency is indeed constitutive of the entire order of normative values.
Those conclusions, however, require a further philosophical argument,
which, even if successful, would not itself be part of the juridical con-
ception. In other words, the cogency or truth of rational agency is a
matter for philosophy and not a matter for the theory of private law.18

17
For instance, Kant purports to show that contractual obligation arises as a synthetic a
priori judgment under the concept of causality.This account of contract is of interest for the
juridical conception of corrective justice not because it contains a correct philosophical
argument—an issue irrelevant to that conception—but because it makes intelligible the
deep-seated legal distinction between in rem and in personam rights. See below chapter 5.
18
On the distinction between theory and philosophy, see John Rawls, “The Independ-
ence of Moral Theory,” in Collected Papers (1999), 286–302.
 correlativity and personality 29

In sum, then, the juridical conception of corrective justice offers

the following picture of liability. Liability is the normative practice
through which the law responds to the doing and suffering of an
injustice. Implicit in this practice are two ideas that are the most
abstract representations of the practical reasoning specific to justifica-
tion within private law. The juridical conception arrives at these ideas
by working back to them from the most general features of liability,
including the bipolarity of litigation, the distinction between nonfea-
sance and misfeasance, and the role of rights and correlative duties.
The first idea, correlativity, represents the connection between the
parties as doer and sufferer of an injustice. Correlativity structures the
fair and coherent terms on which the parties interact and to which
the justifications that pertain to their interaction ought to conform.
When regarded correlatively, doing and suffering form a single nor-
mative unit in which the injustice imputed to the doer is the same as
the injustice done to the sufferer. The second idea, personality, repre-
sents the parties who interact on the basis of correlativity. Personality
signifies the capacity for purposive action without regard for particu-
lar purposes.This capacity is implicit in the rights and duties of private
law.
These two ideas are complementary, each implied by the other.
Correlativity is the most abstract conception of the structure that con-
nects rights to duties in private law; personality is the most abstract
conception of the content of those rights and duties. Correlativity
represents the interrelation of the parties; personality represents the
parties in their interrelation. For each of them liability is the response
to the doing and suffering of an injustice, but they highlight a differ-
ent aspect of it: correlativity highlights the connection of doing to
suffering, and personality highlights the nature of the injustice as an
inconsistency, imputable to the defendant, with a right of the plaintiff .
Together they exhibit the doing and suffering of injustice as an intern-
ally unified and normatively coherent phenomenon.

5. Stone on external validation

Among contemporary tort theorists, no one has articulated the dis-
quiet occasioned by the idea of personality with greater clarity than
Martin Stone. Stone combines a familiarity with the classic expos-
itions of natural right, a reflectiveness about the nature of tort theory,
30 corrective justice

and an underlying sympathy with the juridical conception of correct-

ive justice. Taking over the idea of correlativity, Stone sees the doing
and suffering of the same wrong as the defining theme of corrective
justice. However, he rejects personality on the ground that the place
that the juridical conception assigns to it rests on a mistake about
the theoretical significance of corrective justice.
Stone understands corrective justice as describing “an abstract
framework for arguments concerning the terms on which one person
is responsible for the harmful effects of her actions on another.”19 The
theoretical significance of corrective justice is that it makes us aware
of the contours of a practice in which a distinctive sort of reason is in
play; by grasping this sort of reason one understands the contours of
the practice.20 Corrective justice exhibits “a characteristic sort of rea-
son already captured in the ongoing activity of argument and judg-
ment directed towards the situation to which modern liability rules
are a judicially evolved response.”21 The idea of correlativity consti-
tutes “the abstract framework” for the sort of reason in question, by
showing that requisite kind of reason is present when the grounds for
saying that the defendant has done wrong are the same as the grounds
for saying that the plaintiff has suffered wrong.
So far, Stone and the juridical conception of corrective justice are
ad idem. For the juridical conception, tort law is an ongoing norma-
tive practice characterized by a distinctive exercise of practical reason.
The task that the juridical conception sets for itself is to exhibit at an
abstract level the organizing ideas internal to that exercise of practical
reason. Correlativity, formulated in precisely the terms that Stone uses,
is one of those ideas.
Stone, however, thinks that by introducing personality the juridical
conception goes off the rails. Personality is an idea formulated by Kant
and Hegel as an aspect of rational agency. The purpose of invoking it
within the juridical conception, Stone claims, is to provide a ground-
ing from which corrective justice can be derived. Personality func-
tions to satisfy a demand for the external validation of corrective
justice by pointing to a ground—rational agency—that is available in
advance of the legal practice that is derived from it. This demand is

19
Martin Stone, “On the Idea of Private Law,” (1996) 9 Canadian J. of L. and Juris. 235,
at 253.
20
Ibid., 258.
21
Ibid., 259.
 correlativity and personality 31

misplaced, Stone points out, because no external validation of correct-

ive justice is called for. One’s interest in corrective justice arises only
when one has accepted the notion that tort law expresses genuine
reasons; corrective justice then brings one to a better understanding of
what it is that one has accepted. Corrective justice does this by being
the abstract framework that exhibits, in the words quoted above, “a
characteristic sort of reason already captured in the ongoing activity.”
In this way the relationship between corrective justice and tort law
“traces a circle.”22 However, this very circularity might engender the
baseless but unnerving suspicion that one has failed to make contact
with genuine matters of reason. In the effort to remedy this non-
existent deficiency, someone within this circle might suppose that
“there might be some set of considerations to which one could appeal
that would rationally compel anyone to enter the circle.”23 The intro-
duction of personality, Stone suggests, is the unnecessary response to
this unnecessary thought.
Stone allows that personality, although in his view external, is not
reductive.24 The paradigmatic instance of a reductive validation is eco-
nomic analysis. This is because economic analysis reduces the kind of
reasons that figure in determinations of liability to something else.
Instead of seeing tort law through a characterization of what it is, eco-
nomic analysis sees it in terms of goals (wealth maximization, market
deterrence, and so on) that are desirable independently of it. Econom-
ic analysis thus deflects attention from practical reason as operative
within tort law to some other exercise of practical reason. The jurid-
ical conception of corrective justice does not do this. The account it
produces matches the kind of reasoning that figures within the nor-
mative practice that it is an account of.
What makes personality external, in Stone’s view, is that it validates
corrective justice through derivation. His thought seems to be that
arguments purporting to validate corrective justice in this way go
beyond what is necessary to understand liability as the kind of norma-
tive practice that it is. Such arguments are external even if they are
non-reductive, because they invoke considerations that are not them-
selves part of the self-contained intelligibility of tort law. Corrective
justice “goes beyond” the distinctive sort of reason found in tort law

22
Ibid., 263.
23
Ibid.
24
Ibid., 264.
32 corrective justice

“only in being more general and reflective.”25 Validation, however,

goes beyond tort law in a different way. It locates the outside source of
the normativity that feeds into corrective justice and from which cor-
rective justice can be “derived” in a subsequent intellectual operation.
Stone concludes that once one overcomes the longing to derive
corrective justice from rational agency, personality ceases to matter.The
sole reason he sees for introducing personality into the juridical con-
ception is to satisfy a mistaken feeling that only by breaking out of the
circularity of corrective justice can one establish its rational credentials.
This reconstruction of the motivation for including personality
within the juridical conception of corrective justice is suspect for a
number of reasons. First, the juridical conception of corrective justice
has no embarrassment about circularity. The juridical conception can
be summed up in the brazenly circular proposition that the only pur-
pose of private law is to be private law.26 The juridical conception
treats liability as a self-contained normative practice that it seeks to
understand in terms of its internal unity. Circularity it regards as a vir-
tue. To step outside the circle in search of a source from which to
derive what is inside it risks leaving unintelligible the starting point
on which the rest depends.27
Second, the juridical conception of corrective justice already has a
means of validation and does not need the one that Stone ascribes to
it. What validates the juridical conception is its success in representing
at an abstract level the coherence of practical reason as it operates to
determine liability. This in turn requires that the complementary ideas
of correlativity and personality cohere with each other and that
together they make explicit how the injustice connects the doer and
the sufferer as parties to a coherent normative relationship. Coherence
implies a self-contained circle of mutual reference and support among
the elements of what coheres. It validates what is within this circle by
pointing not outward to some transcendent ideal but inward to the
harmonious interrelationship of its constituents. In contrast to the
notion of validation that Stone criticizes, this notion of validation does
not go beyond what is necessary for understanding liability as the kind
of normative practice it is.

25
Ibid., 265.
26
Weinrib, The Idea of Private Law, 5.
27
Ernest J. Weinrib, “Legal Formalism: On the Immanent Rationality of Law,” (1988) 97
Yale L.J. 949, at 974.
 correlativity and personality 33

Contrary to what Stone suggests, personality does not function as

the means of externally validating the juridical conception of correct-
ive justice. Its status is no different from that of correlativity, which
Stone himself accepts as central. Personality only articulates the con-
ception of agency presupposed in a regime of liability. It is merely
private law’s own notion of entitlement and obligation distilled to its
most abstract formulation, just as correlativity is private law’s own
notion of the nexus between the parties similarly distilled to its most
abstract formulation. As it does with correlativity, the juridical con-
ception reaches personality by working backward from the features of
liability through reflection on juristic experience, not by working for-
ward to private law from a validating notion of rational agency. The
significance of personality within the juridical conception is that it
exhibits the conception of the person appropriate to liability as a spe-
cific mode of association. Personality is as much inside the circle traced
by corrective justice as is correlativity.
By in effect arguing for the elimination of personality from the
juridical conception of corrective justice, Stone may also be making a
further point. Perhaps he is also suggesting that personality has no
necessary place even within the circle. From his detailed account of
negligence law solely in terms of doing and suffering,28 one might
infer that Stone thinks that correlativity alone is sufficient to exhibit
the distinctive sort of reason that animates tort law. This might explain
his preoccupation with the supposed link between personality and
external validation. Being unable to discern a role for personality on
the inside, he assigns it the spurious task of validating corrective just-
ice from the outside. Conversely, because he views personality solely
as an aspect of rational agency, he does not see its significance as the
moral power presupposed in a regime of liability.
On this reading of Stone, his criticism of the juridical conception is
that one of its two complementary ideas is superfluous. All the work
can be done by correlativity. By not mentioning personality in his
treatment of tort doctrine, he might be taken as challenging the jurid-
ical conception to tell him what he is missing.
I have already noted that, within the juridical conception of cor-
rective justice, correlativity exhibits the structure that justifications
must have if they are fairly and coherently to connect the plaintiff to

28
Stone, “The Significance of Doing and Suffering,” in Philosophy and the Law of Torts, ed.
Gerald J. Postema (2001), 131.
34 corrective justice

the defendant. Accordingly, it disqualifies considerations (such as wel-

fare) that do not have the requisite structure, and insists on consider-
ations (such as rights and their correlative duties) that do. Indeed, unless
rights were available, corrective justice could not be actualized in
practice, because private law would be unable to comply with its
structural standard. But while correlativity needs rights, it gives no
positive indication of their nature and content. Personality performs
this function. Drawing on the law’s doctrines concerning the role of
purposiveness in the acquisition and transfer of rights, the juridical
conception of corrective justice postulates personality as the abstrac-
tion that brings out the connection between the content of any per-
son’s right and the external manifestation of that person’s volition.
The significance of personality for representing the nature and con-
tent of rights provides the answer to the supposed inquiry as to what
Stone is missing by his single-minded focus on correlativity. Stone,
in common with the juridical conception, sees corrective justice as
providing a “more general and reflective”29 understanding of the dis-
tinctive kind of reason found in tort law. Hence his interest in correla-
tivity. However, tort law’s distinctive kind of reason also presupposes
the existence of rights. The idea of correlativity shows, negatively, that
the rights of private law cannot coherently be based on non-correla-
tive notions, but it does not positively indicate the nature and content
of these rights. Unless Stone, contrary to his theoretical aims, is will-
ing to forgo a more general and reflective understanding of these
rights, he will have to appeal to some such idea as personality. This is
because personality (to use Stone’s language about correlativity) “sup-
ports—by continuing in a more abstract way—the sort of practical
thinking instinct in the law’s everyday elaboration.”30 The idea of per-
sonality, in other words, represents precisely the kind of explanation
that Stone seeks.
Perhaps the reason that contemporary scholars miss the significance
of personality is that they come to corrective justice through tort the-
ory. A tort theorist naturally focuses on the correction of the wrong
rather than on the nature of the right. It then becomes comparatively
easy to see that the structure of the correction must mirror the struc-
ture of the wrong, and that consequently, because the former involves
correlativity, the latter must also. Just as tort law assumes the existence

29
Above n. 19.
30
Stone, “The Significance of Doing and Suffering,” above n. 28, 168.
 correlativity and personality 35

of rights but concentrates on specifying what constitutes an infringe-

ment of those rights, so tort theorists preoccupy themselves with the
role of wrongdoing while ignoring the significance of rights or taking
them for granted. Then the abstraction that reflects the nature of the
wrongdoing assumes greater salience than the abstraction that reflects
the nature of the infringed right.
This observation suggests a wider issue. It is the case that con-
temporary tort theory has been crucial to the renewed interest in cor-
rective justice. However, that form of justice is not exemplified solely by
tortious wrongdoing. Correlativity and personality are abstractions
drawn from the normative relationship of plaintiff and defendant within
a liability regime. Accordingly, all grounds of obligation in private law,
including the law of contract and the law of unjust enrichment, actual-
ize a correlatively structured and rights-based form of justice. Correc-
tive justice thus opens the door to a more comprehensive understanding
of the normative character of private law. Only when we aim at that
more comprehensive understanding will contemporary scholarship have
paid adequate attention to the insight that Aristotle first formulated.

6. Corrective justice as theoretical framework

A corrective justice approach to the understanding of private law
attempts to discern the normative character of liability as a familiar
practice within which justification has a pervasive role. Corrective
justice takes the justificatory ambitions of this practice seriously by
focusing on the law’s internal normative dimension. With its eye fixed
on the institutions through which the practice unfolds and on the
mode of justification that is distinctive to a liability regime, it focuses
on the structure of the relationship between the two parties. Because
the liability of the defendant is always a liability to the plaintiff, cor-
relativity ranks as the most abstract formulation of that relationship.
And because the rights and duties through which this correlativity
operates express the purposiveness of the plaintiff and the defendant
without mandating particular purposes, personality ranks as the most
abstract formulation of the parties to that relationship. Correlativity
and personality are thus the complementary ideas that constitute the
juridical conception of corrective justice.
These complementary ideas then become the key to understanding
and assessing private law’s concepts, principles, and doctrines.The only
36 corrective justice

considerations that conform to corrective justice are those that apply

(or figure within a larger ensemble of considerations that applies) cor-
relatively to both parties. Thus, in its regulative function corrective
justice holds the practice of liability to the normative implications of
liability’s own correlative structure.
Understood in this way corrective justice is a normative structure
that is fair to both parties. Fairness in this context refers not to the
determination of one’s appropriate share as a matter of distributive
justice, but to reasons for liability that, by treating the parties as cor-
relatively situated, favor neither of them at the expense of the other.
Because the fair terms of a bilateral interaction cannot be set on a
unilateral basis, considerations whose justificatory force extends only
to one party are inadmissible. Corrective justice therefore requires jus-
tifications for liability that pertain to the relationship between the
parties rather than to one or the other of the parties in isolation.
Furthermore, corrective justice so conceived is an integrating idea
that discloses the nature of coherence within a system of liability. One
aspect of its integrating power is that, because the injustice—and there-
fore the justification for holding the defendant liable—is identical from
the standpoint of both parties, corrective justice presents the finding of
liability not as the consequence of a miscellany of disconnected factors
applicable to one party or the other, but as the outcome of reasoning
coherently applicable to the relationship as a whole. Being correlative
to each other, the doing and the suffering of injustice form a single
juridical sequence in which each party participates only through the
presence of the other. Accordingly, all the elements of liability must
themselves constitute a coherent ensemble that expresses the unity of
the parties’ legal relationship. A second integrating aspect is that correc-
tive justice connects the injustice and its rectification. Under corrective
justice, the setting of the remedy is not an opportunity for the court to
do what is best for the future, all things considered. Rather, because the
rectification consists simply in undoing, to the extent possible, the very
injustice that has been committed, the remedy is a response only to the
factors that make up the injustice. The nature of the right infringed
determines the nature of the remedy to be awarded. A third integrating
feature is that, since all the relationships of private law are subject to
the demands of correlativity, the coherence of legal justification can
operate not only within any given relationship, but also systemically
across relationships, encompassing all the grounds of liability in their
interconnection.
 correlativity and personality 37

Any sophisticated system of liability aspires to realize the values of

fairness and coherence. Corrective justice provides the most abstract
representation of the meaning of fairness and coherence with respect
to liability. Even when legal doctrine is unfair or incoherent and
therefore ought to be changed, corrective justice constitutes the
immanent critical standpoint informing the law’s effort to work itself
pure.Thus, corrective justice as a theoretical construct and liability as a
familiar normative practice are reciprocally illuminating. Corrective
justice is the structure of justification implicit in the practice; and to
the extent that it is fair and coherent, the practice, through its doc-
trines, institutions, and modes of discourse, is the specific realization of
corrective justice in a functioning system of liability.
By attending to the normativeness implicit in liability as a familiar
practice, corrective justice directs us away from contemporary conse-
quentialist and reductionist understandings of law. Corrective justice
repudiates the idea that the justifications that figure in private law
derive from goals that are desirable independently of the role that they
can coherently play in a liability regime that is fair to both parties.
Instead, corrective justice attempts to bring to the surface the struc-
ture of normative thought latent in the institutions and concepts of a
fair and coherent regime of liability. Corrective justice thus allows one
to discern in private law the possibility of vindicating the notion that
law is, in Aquinas’ striking phrase, “a certain ordering of reason.”31

31
Thomas Aquinas, Summa Theologica, I–II, Q. 90, A. 4 (“quaedam rationis ordinatio”).
 2
The Disintegration of Duty

1. The general conception of duty

In his great judgment in Donoghue v. Stevenson1 Lord Atkin, before
articulating the neighbor principle, noted a deficiency in the judicial
treatment of duties of care in negligence. Only rarely had the com-
mon law formulated “statements of general application defining the
relations between the parties giving rise to a duty of care.”2 Instead,
proceeding on a case-by-case basis, the courts had concerned them-
selves with the particular kind of relationship before them on any
occasion and had therefore focused on the specific status of one or
other of the parties, “whether manufacturer, salesman or landlord, cus-
tomer, tenant, stranger, and so on.”3 Thus, the only way to determine
whether a duty of care existed was to see whether the case could be
referred to some “particular species”4 that the law had already recog-
nized. He continued:

And yet the duty which is common to all cases where liability is established
must logically be based upon some element common to all the cases in
which it is found to exist . . . . [T]here must be, and is, some general concep-
tion of relations giving rise to a duty of care, of which the particular cases
found in the books are but instances.5

1
[1932] A.C. 562 (H.L.).
2
Ibid., at 579.
3
Ibid.
4
Ibid., at 580.
5
Ibid.
 the disintegration of duty 39

The neighbor principle6 was Lord Atkin’s attempt to set out this general
conception.
The general conception of the duty of care—its theoretical basis, its
structural constituents, its more recent disintegration back into par-
ticular duties, and the need to recapture what a general conception of
duty implies—is the subject of the present chapter.
Lord Atkin regarded the existence of a general conception as a
necessity (“there must be, and is, some general conception”). Although
he said little about the nature of this necessity, he presumably had in
mind something like this: the common law, by its own internal logic
and dynamism, cannot treat the particular instances of duty as a cha-
otic miscellany of disparate and independent norms.7 Duties of care
are constituents of a normative system. The coherence of such a sys-
tem requires that all duties of care should be thematically unified
through the same underlying principle. The general conception of
duty reflects the common aspect that each particular duty must have if
it is to be systematically related to every other particular duty. The
necessity to which Lord Atkin refers is a juridical one: the general
conception of duty is an implication of the internal coherence required
by the law’s systematic nature.
In recent years the sense of juridical necessity apparent in Lord
Atkin’s judgment has waned. Lord Atkin’s own formulation of the
general conception in terms of a duty to avoid foreseeable harm to a
neighbor, path-breaking as it was, has been recognized not to provide
a practical test. Moreover, the very idea of a general conception is
sometimes thought to be superfluous given the casuistic nature of
common law reasoning.8 Even in courts that accept the need for an
overarching framework for the duty of care, the general conception
takes the form of multi-staged formulae that are verbally comprehen-
sive without necessarily being juridically coherent.9 The widely

6
“The rule that you are to love your neighbour becomes in law, you must not injure your
neighbour; and the lawyer’s question, Who is my neighbour? receives a restricted reply.You
must take reasonable care to avoid acts or omissions that you can reasonably foresee are likely
to injure your neighbour. Who, then, in law is my neighbour? The answer seems to be per-
sons who are so closely and directly affected by my act that I ought to have them in contem-
plation as being so affected when I am directing my mind to the acts or omissions which are
called into question.” Ibid.
7
This point was made by Deane J. in Stevens v. Brodribb Sawmilling, [1986] 160 C.L.R.
16, at 51 (H.C. Aust.).
8
For an example, see Dawson J. in Hill v. Van Erp [1997] 188 C.L.R. 159, 177 (H.C. Aust.).
9
Below, section 4.
40 corrective justice

accepted idea that duty is a matter of “policy” (whatever that might

mean) has led to a distaste for the abstract practical reasoning that
undergirds a general conception of duty.
The result has been a disintegration of duty in two related senses.
First, because each kind of duty reflects the particular set of policies
deemed appropriate to it or the particular constellation of casuistic
considerations from which it emerges, the preoccupation with particu-
lar species of duties, which Lord Atkin decried, has returned. Duties
are differentiated, usually according to the kinds of harms with which
they deal (physical injury, economic loss, psychiatric injury, and so on),
without attention to (in Lord Atkin’s words) “the element common to
all cases in which [a duty] is found to exist” and therefore without
awareness of the strands that weave all duties of care into a coherent
system in which each duty illuminates, and is illuminated by, all the
others. Second, not only has the whole ensemble of duties disinte-
grated into a collection of particular kinds of duties, but the very idea
of a duty has disintegrated. Instead of the duty of care being an intern-
ally coherent normative category, it has been fragmented into the
separate factors that determine the duty’s ground and limits. Hence,
the reasoning in support of a duty is marked by ad hoc compromises
among these separate factors rather than by an elucidation, in the con-
text of a particular case, of the conception under which the defend-
ant’s act and the plaintiff ’s injury form a unified normative sequence.
These two senses of disintegration go together, because the absence of
coherence within the notion of duty renders otiose the necessity for
coherence that Lord Atkin postulated among particular duties.
This disintegration of duty has undermined the most notable
achievement of negligence law in the twentieth century. However
indeterminate Lord Atkin’s own formulation of the general concep-
tion of duty, other leading negligence cases also took up and advanced
the striving for coherence to which he gave voice. In contrast, the
more recent disintegration of duty manifests a failure to develop tort
law in a normatively coherent way. If I am right about this, a principal
task for negligence law in its ongoing evolution is to grope its way
back to the conception of coherence that was implicit in Lord Atkin’s
celebrated judgment.
In this chapter, I want to discuss these developments from the stand-
point of corrective justice. This standpoint is especially germane,
because corrective justice is the theoretical notion that accounts for
whatever coherence private law might have. Corrective justice ties
 the disintegration of duty 41

coherence to the justifications that inform private law’s characteristic

concepts. Legal doctrine is viewed as coherent only to the extent that
its underlying justifications are coherent. These justifications, in turn,
are coherent only if their force as reasons is congruent with—rather
than artificially truncated by—the structure of the relationship between
the parties. Because liability relates the defendant as doer of the injust-
ice to the plaintiff as sufferer of the same injustice, a finding of liability
is coherent when the reasons for considering the defendant to have
done an injustice are the same as the reasons for considering the plain-
tiff to have suffered that injustice. As was argued in chapter 1, the
normative basis of such coherence is a right of the plaintiff that imposes
a correlative duty on the defendant.
Seen in the light of corrective justice, Lord Atkin’s comments can
be understood as follows. The notion that the law of negligence is a
coherent system of justification precludes particular duties of care
from being regarded as isolated norms. Rather, the particular duties
are systemically related to one another because they participate in a
shared general conception of duty. To play its role as a necessary
determinant of coherence across particular duties, the general con-
ception of duty must itself be an expression of the very coherence
that it imparts. For “the relation between the parties” (as Lord Atkin
calls it)10 to be coherent, that relation has to be regarded as correla-
tively structured through the defendant’s doing and the plaintiff ’s suf-
fering of the same wrong. This in turn has two complementary aspects.
On the one hand, the general conception of duty must be part of an
integrated ensemble of concepts that allows the sequence from the
defendant’s doing of the negligent act to the plaintiff ’s suffering of
the injury to be understood as a normatively coherent unit in which
the injustice done and the injustice suffered are the same. On the
other hand, the general conception of duty considered in itself must
be formed from normative elements that reflect the correlativity of
the defendant’s duty to the plaintiff ’s right. When these two aspects
are present, the general conception of duty then embodies the cor-
relative structure of justification that renders any particular duty
coherent both with the other members of the ensemble of negli-
gence concepts and with other particular duties. There “must be”
such a general conception (as Lord Atkin insists) because otherwise

10
Above, n. 2.
42 corrective justice

the law of negligence would be incoherent—a possibility that he

assumes the law cannot acknowledge.
Taking corrective justice as the theoretical notion underlying Lord
Atkin’s insistence on a general conception of duty, this chapter dis-
cusses the disintegration of duty in the following steps. Section 2 is
concerned with the first of the aspects of coherence mentioned
above—that is, with how duty fits with other negligence concepts
(failure to exercise reasonable care, factual causation, and proximate
cause) to connect the defendant’s act to the plaintiff ’s injury in a nor-
matively coherent way. Accordingly, the section outlines the role of
the landmark cases of the twentieth century in treating the injustice
done by the defendant as identical to the injustice suffered by the
plaintiff. Section 3 deals with the second aspect of coherence; it sets
out the internal structure of the duty of care—that is, what its con-
stituents must be if it is to reflect a coherent conception of the doing
and suffering of a wrong. In so doing, the section sets out the general
conception of the duty of care that juridical coherence requires. The
duty of care will thereby be exhibited as having a definite structure
toward which the legal reasoning of any particular case ought to be
oriented. The presence of such a structure suggests that it is mere dog-
matism to maintain, as tort scholars often do, that duty is nothing but
“a shorthand statement of a conclusion, rather than an aid to analy-
sis.”11 Section 4 examines the two-stage test for negligence that has
been used in Canada and elsewhere. My argument is that this intern-
ally fractured test, as applied by the Supreme Court of Canada, pro-
vides a conspicuous example of the disintegration of duty and,
accordingly, is inadequate for the development of a coherent jurispru-
dence of negligence. Finally, section 5 discusses the meaning and rele-
vance of the much-invoked “policy” for the determination of the duty
of care. The disintegration of duty is the consequence of thinking that
duty is a matter of policy, and that policy, in turn, refers to the various
independent goals that liability might serve. On this view each par-
ticular kind of duty represents the balance of goals, in themselves
diverse and competing, that is peculiar to it. However, another notion
of policy refers to the exercise of practical judgment in elucidating
what the general conception of duty might mean in particular cir-
cumstances. The general conception provides not (as has often been

11
W. Page Keeton et al., Prosser and Keeton on the Law of Torts, 5th ed. (1984), 358.
 the disintegration of duty 43

assumed) a test of duty, but a structure of thinking that is actualized in

legal reasoning through the casuistic assessment of facts or comparison
of cases or through the elucidation of its particular normative features
in the overall context of a legal system that values coherence. This sec-
ond notion of policy is, I suggest, not only compatible with but indeed
required by the general conception of duty.

2. The place of duty among the negligence

concepts
How, then, does negligence doctrine treat the defendant and the plain-
tiff as the correlatively situated sufferer and doer of the same wrong?
Negligence concerns the plaintiff ’s being wrongfully injured through
the defendant’s creation of an unreasonable risk. If negligence liability
is to be a coherent normative phenomenon, the injury and the risk-
creation have to be understood as the constituents of a single wrong
that is elucidated through an integrated ensemble of legal concepts. In
this way, the differing experiences of the parties as doer and sufferer
and the temporal gap between the doing and the suffering are sub-
sumed under a unified set of normative categories that render the
wrong done identical to the wrong suffered.
Coherence requires that the injustice relate act to injury and vice
versa. Precluded are definitions of the injustice between the parties in
terms that pertain to one of them alone. As negligence law recognizes,
the injustice does not consist merely in the unreasonably created risk
considered in itself; that would one-sidedly focus on the defendant’s
wrongful action and entail liability for unreasonable risk-creation even
without damage. Nor does it consist in the harmful effect considered
in itself; that would one-sidedly focus on the plaintiff ’s injury and
entail strict liability, rather than negligence.12 Nor, again, does the
injustice consist in the combination of wrongful action and injury
each considered one-sidedly; that would mean that, although the
defendant has committed a wrong and the plaintiff has suffered one,
these are two different wrongs, each resting on its own foundation.

12
For a recent analysis of negligence either in terms of risk caused or harm suffered
without attending to the possible unity of the sequence as a whole, see Heidi M. Hurd and
Michael S. Moore, “Negligence in the Air,” (2002) 3 Theoretical Inquiries in Law 333,
at 348.
44 corrective justice

Rather, the injustice embraces the entire span from the act to injury:
the defendant’s act is viewed as a wrong because of its potential to
cause this kind of injury, and the plaintiff ’s injury is viewed as a wrong
because its potential occurrence is a reason for considering the defend-
ant’s act wrongful. Then the sequence from act to injury forms the
single wrong that the defendant can be said to have done and the
plaintiff to have suffered.
To be coherent, tort doctrine elaborates legal concepts that treat
the defendant’s act and its effect on the plaintiff as an integrated
sequence in which there is a single injustice that is the same for both
parties. In legal terms, this sequence begins with the defendant’s breach
of the standard of reasonable care and ends in the factual causation of
injury. However, the sequence can be regarded as integrated only if its
two termini operate not as atomistic elements that the law simply
adds together, but as constituents of liability that, for purposes of tort
law, each derive their significance from the other. Hence, the unrea-
sonableness of the risk created by the defendant must lie in the pos-
sible occurrence of the kind of injury that the plaintiff suffered. This
way of relating the negligent act to the injury makes the injustice of
unreasonable risk-creation the same for both parties.
The signal achievement of negligence law in the twentieth century
was to develop the concepts of negligence analysis in a way that coher-
ently links the unreasonable risk to the harm suffered. Duty and prox-
imate cause are crucial components in this linkage. These concepts
connect wrongdoing and injury by describing the wrongful risk in
terms of the range of the potential victims and consequences through
which the risk is to be understood as wrongful. Duty connects the
defendant as a wrongdoer to the plaintiff as a member of the class of
persons wrongfully put at risk. Similarly, proximate cause connects the
defendant’s negligence to the plaintiff ’s suffering of the kind of injury
or accident the risk of which rendered the defendant’s act wrongful.
Because both duty and proximate cause are requirements of liability,
the defendant is not held legally responsible for the materialization of
a harm that is not within the set of possibilities that supply a reason
for exercising due care. When, however, the injury suffered by the
plaintiff is to a member of the class of persons that the defendant
wrongfully put at risk and is the kind of injury or accident that due
care is supposed to avoid, then the wrongfulness of both the defend-
ant’s action and the plaintiff ’s injury is referable to the same sort of
risk. Under those circumstances, the sequence from the defendant’s
 the disintegration of duty 45

creation of an unreasonable risk to the materialization of that risk in

injury to the plaintiff constitutes the same injustice for both parties.
The leading twentieth-century cases on duty and proximate cause
gave legal expression to this conception of negligence liability. Three
developments were particularly noteworthy. The first, of course, was
Donoghue v. Stevenson13 (and its American predecessor MacPherson v.
Buick Motor Co.).14 In place of a fragmented collection of duties that
varied according to the particular social and contractual relationships
between the parties, Donoghue established that the duty of care flowed
from the defendant’s risk-creating action as such and from its reason-
ably foreseeable effect on those who ought to have been within the
defendant’s contemplation.The general conception of the duty of care
formulated in this case thereby connected the defendant as the creator
of an unreasonable risk and the plaintiff as a person whose endanger-
ment made the risk unreasonable.
A more explicit and complete development of the same idea
appeared in Cardozo’s judgment in Palsgraf v. Long Island Railroad Co.15
Cardozo held that there could be no liability unless the defendant’s
breach of duty was a wrong in relation to the plaintiff. Because in that
case the defendant’s conduct was not wrongful toward the plaintiff
(although it was arguably wrongful toward someone else), the defend-
ant was held not to be under a duty with respect to the plaintiff ’s loss.
Cardozo’s explicit purpose was to construe the wrongfulness of
negligence in a way that was specifically appropriate to tort law. He
noted “the shifting meanings of such words as ‘wrong’ and ‘wrong-
ful.’”16 What the plaintiff had to show was a wrong that was “a viola-
tion of her own right, and not merely a wrong to some one else, nor
conduct ‘wrongful’ because unsocial but not ‘a wrong’ to any one.”17
To detach the notion of wrong from a beach of duty owing to the
plaintiff “is to ignore the fundamental difference between tort and
crime.”18 For Cardozo, tort liability—in contrast to other kinds of
societal judgments about culpability—featured a wrong done by the
defendant in relation to the plaintiff . Accordingly, the very reason for
thinking that the defendant acted wrongfully also had to be the reason

13
Above n. 1.
14
111 N.E. 1050 (N.Y.C.A, 1916).
15
162 N.E. 99 (N.Y.C.A, 1928).
16
Ibid., at 100.
17
Ibid.
18
Ibid., at 101.
46 corrective justice

for thinking that the injury suffered by the plaintiff was wrongful. The
duty of care in negligence law was to be understood as rendering the
normative significance of the wrong identical for both parties. “Negli-
gence . . . is thus a term of relation.”19
Cardozo’s outstanding contribution was to align the relational nature
of tortious wrongdoing with the relational significance of unreasonable
risk. In the words of the opinion, “The risk reasonably to be perceived
defines the duty to be obeyed, and risk imports relation.”20 As a way of
referring to the harmful potentialities inherent in a given act, risk is
not intelligible in abstraction from a set of perils and a set of persons
imperiled. A negligent act releases a set of harmful possibilities that due
care should have avoided.The plaintiff cannot recover unless the injury
she suffered actualizes a possibility within this set. The imperiling of
the foreseeably affected class of persons is the reason for considering
the defendant’s act negligent; it also must be the reason for thinking
that the plaintiff has been wronged. Only if the plaintiff is among that
class does the reason for thinking of the defendant’s action as wrongful
pertain to the plaintiff. Because in Palsgraf the prospect of the plaintiff ’s
injury was not a reason for considering the defendant’s action negli-
gent, the defendant was not under a duty toward her.
Cardozo’s treatment of duty forges a normative connection between
the defendant’s action and the plaintiff ’s injury. The connection is
normative, because it is based on the reason for considering an act to
be wrongful. Given that the structure of the relationship between the
parties is one of correlativity, that reason must simultaneously provide
a ground both for holding the defendant liable and for holding the
plaintiff entitled to recover. Under Cardozo’s analysis of the relational
quality of unreasonable risk, the duty of care makes the same reason
pertain to both parties.
In contrast, Andrews’s dissenting judgment in Palsgraf connects the
parties in a historical but not a normative way. The connection is
merely historical, because the element of fault required for the defend-
ant’s liability to the plaintiff is satisfied by the fact that the defendant’s
negligence is the historical antecedent of the plaintiff ’s injury. In
Andrews’s view, the injury suffered can be the basis of the plaintiff ’s
recovery even though the wrong was relative to a third party rather
than the defendant. Unanswered is the question of why the merely

19
Ibid.
20
Ibid., at 101.
 the disintegration of duty 47

historical connection between the defendant’s negligent act and the

plaintiff ’s injury should justify liability on the basis of fault. It is no
answer to say that this negligence caused the injury; that answer, by
transforming cause into the determinant of the plaintiff ’s recovery,
should also apply to causally effective action that is innocent (and thus
as nonculpable relative to the victim as the defendant’s conduct in
Palsgraf ). Negligence liability would then collapse into strict liability.
Another difficulty flows from Andrews’s postion. Ever since the
devastating critique of strict liability by Oliver Wendell Holmes, no
one has been able to sustain the position that liability should be based
on causation.21 The particular problem that Holmes identified is that
causation is unable to generate its own limits, so as to preclude a
regression to causes however distant. Andrews is not insensitive to this
problem. He insists that for the defendant to be liable, the negligence
must be not merely a cause, but the proximate cause of the damage.
However, having rejected the notion that the parties are normatively
connected (and thus liability is limited) through the idea of unreason-
able risk, he has no coherent conceptual framework within which
judgments about the proximateness of causation can be situated. In
Andrews’s own account the determination of proximate cause turns
out to be an arbitrary exercise in practical politics, intelligible more
through rhetoric and metaphor than through legal analysis.22
A more coherent conception of proximate cause than the one
Andrews put forward emerged from yet another major development
of tort jurisprudence in the twentieth century, the Privy Council’s
decision in the Wagon Mound case.23 The case held the requirement of
proximate cause unsatisfied when the defendant negligently exposed
the plaintiff to the risk of one kind of injury, but the plaintiff suffered
an injury of a different kind. This decision made proximate cause run
parallel to Cardozo’s conception of duty.24 Just as Palsgraf required that
21
O.W. Holmes, The Common Law (1881), Lecture III. For a criticism of Richard Epstein’s
more recent attempt to vindicate strict liability, see Ernest J. Weinrib, The Idea of Private Law
(1995), 172–77.
22
Above n. 15, at 103.
23
Overseas Tankship (UK) v. Morts Dock and Engineering (Wagon Mound No. 1) [1961]
A.C. 388 (P.C.).
24
The connection of Wagon Mound and Palsgraf is sharply put in a leading text on tort law
in the United States: “The decision is the logical aftermath of Cardozo’s decision in the
Palsgraf case, since there is an obvious absurdity in holding that one who can foresee some
harm to A is liable to consequences to A which he cannot foresee, but is not liable for similar
consequences to B.” Page Keeton et al., Prosser and Keaton on the Law of Torts, 296. A similar
point was made in the Wagon Mound opinion; see above n. 23, at 425.
48 corrective justice

the plaintiff be within the class of persons unreasonably imperiled, so

Wagon Mound required that the injury or accident be of the sort that
renders the defendant’s risk-creation unreasonable. Thus, with respect
both to the person injured and to the injury or accident, the harm
that occurred actualized the possibilities for danger that it was negli-
gent for the defendant to have created. Duty and proximate cause
both functioned to make the wrongfulness of what the defendant did
the same as the wrongfulness of what the plaintiff suffered.
When this complex of leading cases is considered as a whole, the
main categories of negligence liability—duty, breach of duty, prox-
imate cause, and factual cause—form a coherent set that traces the
sequence from the defendant’s negligent act to the plaintiff ’s injury.
Breach of duty and factual causation are the termini of this sequence,
with the former referring to the defendant’s creation of unreasonable
risk and the latter to the materialization of risk in injury to the plain-
tiff . Duty and proximate cause integrate these termini into a norma-
tively coherent unit by characterizing in terms of the same
unreasonable risk the wrongfulness of what the defendant did and of
what the plaintiff suffered.The result is that the duty of care is a mem-
ber of an interconnected ensemble of concepts through which the
parties to a finding of negligence can be understood as doer and suf-
ferer of the same injustice.

3. Right and correlative duty

Having situated the duty of care within the ensemble of concepts that
allows negligence liability to be understood as a normatively coherent
phenomenon, I want in this section to focus more specifically on the
duty itself. What are its constituents and how do they give expression,
in accordance with corrective justice, to the correlativity of right and
duty? The answer to this question indicates the internal structure of
the duty of care, thus revealing (in Lord Atkin’s words) “the general
conception of relations giving rise to a duty of care, of which the par-
ticular cases found in the books are but instances.”25
Cardozo’s judgment denying liability in Palsgraf provides the most
explicit judicial elucidation of the correlativity of right and duty in
the negligence context. Cardozo observes that liability requires that

25
Above, n. 1.
 the disintegration of duty 49

“the plaintiff sues in her own right for a wrong personal to her, and
not as the vicarious beneficiary of a breach of duty to another. A dif-
ferent conclusion would involve us, and swiftly too, in a maze of
contradictions.”26 In Palsgraf the defendant arguably created an unrea-
sonable risk to the third party’s package but was being sued for a
wrongful infringement of the plaintiff ’s right to bodily integrity. The
right that the defendant unreasonably put at risk was therefore differ-
ent from the right whose wrongful infringement was the basis of the
plaintiff ’s complaint. The duty to the person carrying the package was
thus not correlative to the plaintiff ’s right in her bodily integrity. Lia-
bility to the plaintiff would mean that the defendant was being held
liable for the infringement of a right that its employee did not wrong-
fully imperil, so that the award of damages would then not represent
the wrong that the defendant did. The contradiction is that under
these circumstances liability to the plaintiff would be inconsistent
with the nature of tort law as a mechanism for obligating defendants
to make reparation for the rights that they have wrongfully injured.
As Cardozo points out, Palsgraf featured a particularly striking
example of “the maze of contradictions.” Palsgraf was a case in which
the defendant’s breach of duty and the plaintiff ’s complaint involved a
diversity both in the nature of the rights (property and personal integ-
rity) and in the holders of the rights (the passenger carrying the pack-
age and the injured plaintiff). Cardozo correctly observes that the
same contradiction would obtain even if the rights were of the same
order but the person foreseeably endangered was different from the
person actually injured.27 He also suggests, as Wagon Mound (No.1)
later decided, that the contradiction would also be present if the dis-
crepancy between the defendant’s breach and the plaintiff ’s injury
were to rights of different orders.28 In all these instances there would
be no liability because the plaintiff would have been injured with
respect to a right that was not the basis of the duty that the defendant
breached.
When negligence law is conceived in terms of the correlativity of
right and duty, the issue of the duty of care is composed of two con-
stituents. First, the interest of the plaintiff that is protected against the
defendant’s conduct must have the status of a right as against the

26
Palsgraf v. Long Island Railroad Co., above n. 15, at 100.
27
Ibid.
28
Ibid., at 101.
50 corrective justice

defendant. Second, the duty breached must be correlative to that right.

These two constituents are the complementary aspects of a unified
conception of the duty of care, because a right both implies and is
required by the correlative structure of liability. A right implies correla-
tivity because a right always entails the existence of a corresponding
duty. A right is required by correlativity, because (along with its corres-
ponding duty) it is the only normative concept that has the correla-
tive structure inherent in a regime of liability. Thus, the notions of
right and correlative duty together form a unified general conception
of the duty of care.
Cardozo’s treatment of duty in the Palsgraf case gives paradigmatic
legal expression to this conception of the duty of care. For Cardozo,
the principal issue presented by the case was that of correlativity.
Because in that case the defendant’s conduct was not wrongful toward
the plaintiff (although it was arguably wrongful toward someone else),
the defendant was held not to be under a duty with respect to the
plaintiff ’s loss. Although Cardozo’s focus was on the absence of cor-
relativity between the duty breached and the injury, Cardozo also
insists that the duty breached has to be correlative to a right: “Negli-
gence is not actionable unless it involves the invasion of a legally pro-
tected interest, the violation of a right.”29 Because “the commission of
a wrong imports the violation of a right,”30 the plaintiff is precluded
from recovering unless the defendant’s conduct is a wrong in relation
to that right. Hence, “[w]hat the plaintiff must show is a wrong to
herself, i.e., a violation of her own right.”31 That the plaintiff had a
right to her bodily security was not disputed in the case, but the
defendant’s action was not wrongful relative to that right. Thus,
Cardozo affirms that the existence of a right in the plaintiff is presup-
posed in the requirement that the duty breached by the defendant be
a wrong relative to her.
Accordingly, Cardozo’s opinion presents the two interrelated func-
tions of the inquiry into the defendant’s duty. The first function is to
establish whether the plaintiff ’s damaged interest has the status of a
right, because it is only to a right that the defendant’s duty can be
correlative. The second function is to establish whether correlativity
obtains in the case at hand—that is, whether the defendant breached a

29
Ibid., at 99.
30
Ibid., at 101.
31
Ibid., at 100.
 the disintegration of duty 51

duty correlative to that right by creating an unreasonable risk to per-

sons such as the plaintiff . When these two functions are brought
together, the question of duty produces a structure of inquiry geared
to ascertaining whether the parties can plausibly be regarded as the
doer and sufferer of the same injustice. Action by the defendant that is
incompatible with the plaintiff ’s right marks the injustice of the
defendant’s conduct, and that injustice is the same for both parties.
Injury to a right differs from the suffering of harm or loss. A right
immediately signifies the existence of a correlative duty; harm or loss
does not. Neither the harm as something suffered by the plaintiff nor
the process of suffering it at the defendant’s hand establishes a link
between the parties that is at once correlative and juridically norma-
tive. Being harmed is merely a fact—that the harmed person is now
less advantageously situated than before—that in itself has no correla-
tive normative significance. And the plaintiff ’s suffering of harm at the
defendant’s hand links the parties historically through the link between
consequence and cause or grammatically as object and subject of the
same verb, rather than normatively as the doer and sufferer of an
injustice. Because one can be harmed without being wronged, harm
considered in itself is not a notion from which one can, within the
correlative structure of negligence law, impute a wrong. Harm matters
only inasmuch as it stands under a right, for only when the duty
breached by the defendant is correlative to the plaintiff ’s right do the
parties occupy correlative normative positions. Accordingly, if the loss
of which the plaintiff is complaining is not the subject matter of a
right or if the defendant’s conduct is not wrongful with respect to that
right, then the defendant is not under a duty of care to the plaintiff.
The supposed absence of a right accounts for some of the situations
where the common law does not recognize (or earlier had been reluc-
tant to recognize) the existence of a duty. In situations of nonfeasance,
for example, the entitlement claimed is not merely to one’s own phys-
ical integrity—which ex hypothesi the defendant has not endangered—
but to the defendant’s positive assistance. Under the common law,
however, one has no general right to be benefited by another. Simi-
larly, the perceived absence of a right in the plaintiff also may explain,
at least in part, the law’s slowness to recognize negligence liability
for psychiatric and prenatal injury. In the case of psychiatric injury,
the psychological interest was perhaps regarded as too speculative or
insubstantial to count as part of one’s right to physical integrity.
Over the past century a stronger appreciation of the psychological
52 corrective justice

aspect of physical integrity has rightly led to a steady broadening of

liability and to a growing dissatisfaction with arbitrary restrictions.32
In the case of prenatal injuries, the common law position that legal
personality begins at birth allowed them not to be viewed as viola-
tions of the child’s rights. This fallacious view has now been repudi-
ated: even if the plaintiff was not in existence at the time of the
negligent act, that act can subsequently materialize in a wrongful
infringement of the plaintiff ’s right to physical integrity.
The role of rights has also been crucial for the traditional approach
to liability for economic loss.33 Traditionally the common law restrict-
ed liability to situations in which the economic loss represented the
quantified value of a right belonging to the plaintiff and negligently
injured by the defendant. On this basis the plaintiff could recover for
“financial damage . . . consequential on foreseeable physical injury or
damage to property”34—that is, for lost earnings because of physical
injury or for the lost economic value of property negligently damaged
by the defendant. In such cases the basis of liability was the plaintiff ’s
right to physical integrity or to the exclusive use of the property.
Concomitant to these rights—indeed, part of their meaning—was a
correlative obligation on others (including the defendant) not to
wrongfully interfere. The economic loss was merely the monetary
measure of the right’s infringement. Conversely, if the economic loss
was independent of a right that ran between the parties, the plaintiff
could not recover for it. The law thereby recognized that a set of eco-
nomic advantages enjoyed or anticipated by the plaintiff was not in
itself the subject matter of a right. Although such economic advan-
tages make the plaintiff better off than he or she would be without
them, they have no inherently correlative significance for anyone else;
the defendant, therefore, was not liable for them. Negligence law
expressed this conclusion about these free-standing economic losses
by saying either that no duty of care existed toward someone whose

32
The treatment of psychiatric injury in England, where “control mechanisms” function
“as more or less arbitrary conditions” to restrict liability (White v. Chief Constable of South
Yorkshire Police [1999] 2 A.C. 455, at 502 (H.L.)) has been particularly unfortunate; see the
criticism of the English approach in Tame v. New South Wales; Annetts v. Australian Stations
Pty [2002] 191 A.L.R. 449 (H.C. Aust.).
33
On economic loss, see the important article by Peter Benson, “The Basis for Excluding
Liability for Economic Loss in Tort,” in Philosophical Foundations of Tort Law, ed. David Owen
(1995), 427.
34
Spartan Steel & Alloys v. Martin [1972] 3 All E.R. 557, at 571 (C.A., per Lawton L.J.).
 the disintegration of duty 53

proprietary right was not injured or that such losses, being the result
of an injury to someone else’s rights, were too remote.35
The traditionally restrictive approach to economic loss, then,
excludes liability except as compensation for wrongful injury to the
plaintiff ’s right. In the standard instances of lost income or diminished
value, the right in question is, respectively, a right to bodily integrity
and a right to property or possession, where the entitlement against
the defendant is merely an aspect of the plaintiff ’s exclusive entitle-
ment against the whole world. These rights precede the plaintiff ’s
interaction with any specific defendant. With respect to them the
defendant is under a duty to act with reasonable care because the
whole world is.36
During the twentieth century another kind of right gained recog-
nition in the context of economic losses resulting from negligent mis-
representation.37 This was a right based on justified detrimental

35
The former formulation is exemplified by Weller v. Foot and Mouth Disease Research
Institute [1966] 1 Q.B. 569 (Q.B.D.), the latter by Connecticut Mutual Life Insurance v. New
York and New Hampshire Railroad [1856] Conn. 265 (S.C.). The paradigmatic situation, of
which these cases are examples, is that of the plaintiff who operates a business that depends
on some facility (e.g., a bridge, pipeline, electrical cable) that the plaintiff does not own but
that is negligently damaged by the defendant, causing the plaintiff economic loss because
business operations cannot proceed as normal. One cannot plausibly argue that, within the
systemic logic of the law, the plaintiff has a right to the free-standing economic gain that he
or she was prevented from realizing. Because the parties are strangers, there is no possibility
of an in personam right. Nor can the plaintiff have an in rem right to the prospective economic
gain. First, the intentional diversion of the gain by a competitor is permissible; but it would
be a very odd in rem right if someone could rightfully interfere with it intentionally but not
negligently. Second, the plaintiff can have a right to an external thing only if the plaintiff has
acquired that thing; but there is no mode of acquisition for the prospective economic gain
from the existence of another’s facility. Indeed, the classic case of acquisition, Pierson v. Post,
3 Caines (N.Y.) 175 (1805) repudiates the notion that the prospect of a gain in itself creates a
right. (I am indebted to Abraham Drassinower for discussion of these points.) And so the
plaintiff ’s argument must be based on the suffering of the economic loss even though the
plaintiff had no right to what was lost. This is the argument that the common law tradition-
ally rejected using the language of duty or remoteness.
36
The question of whether in any particular case the plaintiff had a right as against the
defendant is of course subject to legal argument, even in the case of the proprietary and pos-
sessory rights that are paradigmatic in this context; see, for example, Courtenay v. Knutson
[1957] 26 D.L.R. (2d) 768 (B.C.S.C.) (liability to the bailee of a barge). An interesting variant
is Perre v. Apand [1999] 164 A.L.R. 606 (H.C. Aust.), where the defendant, by supplying
diseased seed to one potato grower, caused potatoes of the plaintiff, a neighboring potato
grower, to be embargoed even though the plaintiff ’s crops were not infected. The defendant
was held liable for the plaintiff ’s economic loss; although the defendant’s land was not con-
taminated, the defendant’s negligence caused it to be treated as if it were.
37
Glanzer v. Shepard, 135 N.E. 275 (N.Y.C.A., 1921); Hedley Byrne v. Heller [1964] A.C.
465 (H.L.). See the exemplary treatment by Benson, above n. 33, at 450–54.
54 corrective justice

reliance, created through the interaction of the parties and giving the
plaintiff an entitlement as against the defendant specifically, to recover
the economic loss flowing from the defendant’s reliance-inviting con-
duct. Although the plaintiff has no right against the world for eco-
nomic loss as such, in situations of justified detrimental reliance the
plaintiff recovers for economic loss because of the special relationship
that arose between the parties. The relationship is special in that, given
the circumstances in which the misrepresentation took place, the
defendant can reasonably be regarded as having invited the plaintiff to
rely on it for a particular transaction or kind of transaction, and thus as
having voluntarily assumed responsibility for the loss that results from
that transaction.38 After leading the plaintiff reasonably to rely on the
representation or on other reliance-inviting conduct for the kind of
transaction in question, the defendant cannot fairly disclaim responsi-
bility for the consequences. What places the loss within the scope of
the defendant’s duty is the imputation to the defendant, based on his
or her knowledge of the purpose to be served by the representation,
of an express or implied invitation to the plaintiff (or to the class of
persons that includes the plaintiff) to rely on the representation for
that kind of purpose.39 This purpose, known by the defendant and
detrimentally acted upon by the plaintiff, links the defendant’s act to
the plaintiff ’s loss by making the prospect of the loss the reason for
considering the act to have been negligent. Using the analogy of con-
tract,40 one might say that by making the representation the defendant
has offered the plaintiff information that purports to be reliable for
the purpose of a particular kind of transaction, and that the plaintiff ,
by detrimentally acting on this information, has accepted it as reliable
for the purpose for which it was offered. Therefore, to the extent of
the plaintiff ’s detrimental reliance, tort law views the plaintiff ’s pre-
existing economic situation as an entitlement that runs against the
defendant. The basis of the entitlement—the invitation to rely for a
particular (kind of ) purpose—also defines the scope of the duty

38
The fact that the entitlement is created through the interaction of the parties and is thus
personal to them rather than good against the whole world puts the parties into what Lord
Devlin called “a relationship equivalent to contract”; Hedley Byrne v. Heller, above n. 37,
at 530.
39
For example, in Haig v. Bamford [1976] 72 D.L.R. (3d) 68 (S.C.C.), the defendant knew
the purpose for which it was asked to prepare an audited financial statement. Similarly, in
Hedley Byrne v. Heller, above n. 37, the defendant knew the purpose of the plaintiff ’s inquiry
about its client’s credit-worthiness.
40
See above n. 38.
 the disintegration of duty 55

correlative to it. Accordingly, detrimental reliance that falls outside the

purpose for which the representation was made does not lead to lia-
bility even if the reliance that the representation in fact occasions is
reasonably foreseeable.41 Because it is created by justified detrimental
reliance, the entitlement in question applies only as between the par-
ties; it does not count as a right held by the plaintiff against the world
as a whole. Recognition of such an entitlement thereby leaves intact
the traditional restrictions on recovery for economic loss, which reflect
the idea that economic advantage (and the prospect of impairing it)
does not as such generate duties in everyone else to avoid causing loss.
In the traditional treatment of liability for economic loss, no diffi-
culty arises about what some of the more recent cases regard as cru-
cial, the avoidance (in Cardozo’s famous words) of “liability in an
indeterminate amount for an indeterminate time to an indeterminate
class.”42 It is of course true that, given the interdependence of eco-
nomic activity within society, economic losses to some will foresee-
ably lead to further economic losses to others, which in turn will
foreseeably lead to still further economic losses, and so on. But the
indeterminacy of losses is problematic only if liability is a response to
the suffering of a loss. Implicit within the traditional treatment of eco-
nomic loss, however, is the notion that tort law concerns itself not
with losses as such but with injuries to rights, so that economic losses
matter only as quantifications of those injuries. The plaintiff ’s right,
therefore, both grounds and limits the defendant’s liability.
One should, therefore, appreciate the significance of references,
such as Cardozo’s, to indeterminate liability in the traditional treat-
ment of economic loss. The point is not that liability for economic
loss as such exists but has to be cut off to avoid indeterminacy. Rather,
the prospect of indeterminacy is adduced to indicate the implausibil-
ity of supposing that such liability exists to begin with. As Cardozo
himself said, the indeterminate nature of the supposed liability
“enkindle[s] doubt whether a flaw may not exist in the implication of
a duty that exposes to these consequences.”43
In recent years some jurisdictions (Australia and Canada among
them)44 have abandoned the traditional framework of liability for eco-

41
Caparo Industries v. Dickman [1990] 1 All E.R. 568 (H.L.).
42
Ultramares v. Touche, 174 N.E. 441 (N.Y.C.A., 1931), at 444.
43
Ibid.; see also Weller v. Foot and Mouth Disease Research Institute, above n. 35.
44
Caltex Oil (Australia) v. The Dredge “Willemstad” [1976] 11 Aust. L.R. 227 (H.C.);
Canadian National Railway v. North Pacific Steamship [1992] 91 D.L.R. (4th) 289 (S.C.C.).
56 corrective justice

nomic loss. The basis of the defendant’s liability has shifted from injur-
ing a right to inflicting a loss. Loss differs from right in that it lacks
both the distinct legal content and the correlative significance that
together impose coherent limits on liability. Yet, as is universally
acknowledged, some limit must be formulated because the inter-
dependence of economic interests makes it intolerable for liability to
follow merely from the foreseeability of creating an economic loss.
Accordingly, the duty of care imposed on the defendant by the fore-
seeability of the plaintiff ’s loss is made subject to the limitation of a
policy-based notion of “proximity.”
This development has been accompanied by a transformation of
the role of Cardozo’s famous phrase about indeterminate liability.
Almost invariably invoked in the judgments on economic loss, it is no
longer taken as an indication that the basis of negligence liability must
be sought elsewhere than in the foreseeable loss. Instead, it has become
the prelude to attempts to mark the border to which liability based on
foreseeable loss is to expand. The prospect of indeterminate liability
prompts the court to assess whether, in the circumstances of the par-
ticular case, the plaintiff ’s loss was (despite its foreseeability) insuffi-
ciently proximate to the defendant’s negligence. That foreseeability of
loss is the basis of liability is taken for granted; the focus is on the cri-
terion for limiting liability.45
However, formulating the limiting conditions that satisfy the
requirement of proximity in a principled way has turned out to be
troublesome. This is hardly surprising. Unlike the notions of right and
correlative duty, loss and proximity do not form a unified juridical
conception. Proximity limits foreseeability by considerations that are
not intrinsic to it. On the one hand, foreseeable losses are initially
regarded as possibly worthy of attracting liability simply by reason of
their being foreseeable losses. On the other hand, such losses are not
ultimately regarded as worthy of attracting liability except through
the additional presence of proximity factors that artificially limit the
reach of liability. Because the same normative considerations do not
both ground and limit liability, a constant tension arises between the
unrelated normative impulses that respectively support and confine
the defendant’s duty. Having abandoned the inner coherence of hav-
ing restrictions on liability that are conceptually indigenous to the

45
Bruce P. Feldthusen, “Liability for Pure Economic Loss: Yes, But Why?” (1999) 28 West.
Aust. L. Rev. 84.
 the disintegration of duty 57

correlativity of right and duty, the courts struggle to formulate new

and artificial restrictions on the excessively broad liability that would
be engendered by the mere foreseeability of loss.46 In this respect, the
new approach to liability for economic loss represents a jurispruden-
tial decline.
How, finally, does correlativity illuminate Lord Atkin’s general con-
ception of the duty of care? Adapting the biblical precept that one is
to love one’s neighbor, Lord Atkin famously characterized the “neigh-
bor” to whom the duty of care is owed as “persons so closely and
directly affected by my act that I ought to have them in mind as being
so affected.”47 In the correlative conception of liability, the obligation
on the defendant to have these persons in mind arises from their pos-
session of a right that runs against the defendant and thus places the
defendant under an obligation not injure it. An act “closely and directly”
affects this right when the act’s tendency to endanger the right is a
reason for considering the act negligent. The duty that results is one
that draws its justification from its correlativity with the plaintiff ’s
right. Such a duty defines the “relationship between the parties” in
terms that are at once legal and general. The legal dimension lies in
demanding not the beneficence of neighborly love, which Lord Atkin
considered excessively broad for a legal system, but the more restricted
avoidance of injury.The general dimension lies in construing this legal
notion of injury as expressing the most general normative categories
applicable to the relationship between the parties in private law, that
of right and correlative duty. Because correlativity requires that the
content of the right be the same as the object of the duty, the plain-
tiff ’s suffering is unjust for the same reason that the defendant’s action

46
Among the markers of proximity that have been suggested are: knowledge that the
defendant has or ought to have of the specific individual(s) likely to suffer economic loss
(Stevenson J. in Canadian National Railway, above n. 44); Mason J. in Caltex Oil, above n. 44;
knowledge that the defendant has or ought to have of the class of first line victims (McHugh
J. in Perre, above n. 36); salient features about the defendant’s actual or constructive knowledge
of the prospective injury, about the nature of the detriment and about the nature of the dam-
ages claimed (Stephen J. in Caltex Oil, above n. 44); the relationship between the parties, physi-
cal propinquity, assumed or imposed obligations and close causal connection (McLachlin J. in
Canadian National Railway, above n. 44); the pre-existence of contractual arrangements
between the parties (Martel Building v. Canada [2002] S.C.R. 860). Acknowledging that their
treatments of the problem of indeterminate recovery are themselves quite indeterminate,
Stephen J. and McLachlin J. expect that over time the judicial decisions will crystallize the
characteristic features of various situations for future guidance. One can be skeptical, however,
that the positive law will provide legal certainty if it lacks conceptual coherence.
47
Above, n. 1.
58 corrective justice

is. This conception of the duty of care thus captures the coherence
presupposed in Lord Atkin’s insistence that, aside from particular
duties, there must be a general conception of duty that all particular
duties instantiate.

4. The two-stage test for the duty of care

As the previous sections indicate, one can discern two broad strategies
for elucidating the duty of care. One strategy, represented pre-eminently
by the landmark cases of the twentieth century, coherently integrates
the duty of care both into the entire ensemble of negligence concepts
and into the correlativity of right and duty. Under this strategy the
plaintiff ’s right both grounds and limits the defendant’s liability. An
alternative strategy, evident in Andrews’s dissenting opinion in Palsgraf
and in the more recent economic loss cases, allows the foreseeability
of loss initially to expand the scope of liability, which is then limited
by some other mechanism. In Andrews’s opinion the danger to any-
one creates a duty to whoever is in fact injured, but proximate cause
excludes certain consequences from the expanded liability that would
otherwise ensue. In the more recent economic loss cases, the foresee-
ability of loss triggers a duty subject to considerations of proximity
that address the problem of indeterminacy. Under this strategy the
limit emerges from a factor independent of the one that creates the
initial possibility of liability.
In the last decades of the twentieth century several common law
jurisdictions developed a comprehensive version of this latter strategy.
Starting from the idea that Lord Atkin’s neighbor principle “ought to
apply unless there is some justification or valid explanation for its
exclusion,”48 the House of Lords worked out a two-stage test for the
duty of care.49 This test was adopted elsewhere and, although subse-
quently abandoned in England and Australia, continues to be used in a
number of jurisdictions.50 The two-stage test required the courts to
consider, first, whether the parties had a relationship as neighbors that
was sufficient to give rise to a prima facie duty and, then, whether

48
Home Office v. Dorset Yacht [1970] A.C. 1004, at 1027 (H.L., per Lord Reid).
49
Anns v. Merton London Borough Council [1978] A.C. 728 (H.L.).
50
Daniel More, “The Boundaries of Negligence,” (2003) 4 Theoretical Inquiries in Law
339, 343–45.
 the disintegration of duty 59

there were any considerations that ought to negative or limit the duty.
Foreseeability of loss was a significant component of the first stage,
opening the road to a broad liability that the second stage could
narrow.
Perhaps the most steadfast champion of the two-stage test has been
the Supreme Court of Canada, which has continued to apply it even
after its repudiation by the House of Lords. In its current Canadian
formulation the two-stage test goes as follows:

In order to decide whether or not a private law duty of care existed, two
questions must be asked:
1. is there a sufficiently close relationship between the parties . . . so that
in the reasonable contemplation of the [defendant], carelessness on its part
might cause damage to [the plaintiff]? If so,
2. are there any considerations that ought to negative or limit (a) the
scope of the duty (b) the class of persons to whom it is owed or (c) the
damages to which the breach of it may give rise?51

The first stage of this formulation incorporates the traditional notion

of reasonable foreseeability to establish a prima facie duty. At this
stage the judge attempts to discern “whether, as a matter of simple
justice, the defendant may be said to have had an obligation to be
mindful of the plaintiff ’s interests.”52 The second stage allows that
prima facie duty to be circumscribed or cancelled because of the
presence of “policy concerns that are extrinsic to simple justice but
that are nevertheless fundamentally important.”53
The two-stage test has radically altered negligence law in Canada.
Negligence analysis no longer consists in scrutinizing the parties’ rela-
tionship in the light of a coherent series of concepts. The mode of
argument that underpinned the great doctrinal achievements of the
twentieth century has been abandoned. Instead of examining whether
the materialization of the risk created by the defendant is an injustice
to the plaintiff, the test directs the court to a mélange of justice and
policy considerations. This momentous change has had several ques-
tionable features. In this section I outline these features, then illustrate
the operation of some of them by comparing a Canadian case with its

51
Kamloops v. Nielsen [1984] 10 D.L.R. (4th) 641, at 662 (S.C.C.).
52
Hercules Managements v. Ernst & Young [1997] 146 D.L.R. (4th) 577, at 591 (S.C.C.).
53
Ibid.
60 corrective justice

English counterpart, and finally discuss a recent Canadian case that,

although reaffirming and supposedly clarifying the two-stage test, may
mark the court’s initial step back to a more coherent jurisprudence.
Application of the two-stage test has the following characteristics.
First, policy considerations relevant to the second stage have the power
to override the conclusions about justice reached in the first stage.
These policy considerations are uncontrolled by the relationship
between the parties and indeed may be beyond the court’s institu-
tional competence to judge. A plaintiff can therefore be denied com-
pensation on the basis of policy considerations that, while one-sidedly
pertinent to the defendant or to persons carrying on a similar activity,
have no normative bearing on the position of the plaintiff as the suf-
ferer of an injustice. From the plaintiff ’s point of view, the denial of
recovery, operating (as the court says) extrinsically to simple justice,
amounts to the judicial confiscation of what was rightly due to the
plaintiff in order to subsidize policy objectives unilaterally favorable to
the defendant and those similarly situated.
Second, even as policy analysis the second stage is one-sided. It refers
only to policy considerations that negative liability, not to those that
might confirm liability. Under the court’s formulation, the plaintiff ’s
claim for compensation is entirely constituted by the first stage; the
second stage is devoted to factors favorable to defendants. Although the
court occasionally gestures in the direction of a policy adverse to
the defendant, it rarely engages either in an extended examination
of that policy or in a rigorous comparison of the competing policy
considerations.54 Indeed, to do so would expose a further difficulty: that,
54
In defense of the court on this point, Justice Major said that:
this criticism is too narrow in view of what was said in Hercules [above, n. 52]. In that case LaForest
J. considered not only indeterminate liability, which is a policy consideration that negates liability,
but also deterrence, which is a policy consideration that favours liability. Although the concern for
indeterminate liability won the day, it is clear that the policy analysis was not completely one-sided.
Major, “Anns and the Law of Negligence,” in The Continuing Legal Education Society of
British Columbia, Torts—2001, posted 12 April 2001 at <http://www.cle.bc.ca/CLE/
Analysis/Collection/01-5123601-anns>. In Hercules the court’s entire treatment of deter-
rence was as follows:
Certain authors have argued that imposing broad duties of care on auditors would give rise to sig-
nificant social and economic benefits so far as the spectre of tort liability would act as an incentive
to auditors to produce accurate (i.e., non-negligent) reports. [References]. I would agree that deter-
rence of negligent conduct is an important policy consideration with respect to auditors’ liability.
Nonetheless I am of the view that, in the final analysis, it is outweighed by the socially undesirable
consequences to which the imposition of indeterminate liability on auditors might lead.
Hercules, above n. 52, 593. It seems extravagant to characterize this desultory mention of
deterrence as showing that deterrence was “considered” in the “policy analysis.”
 the disintegration of duty 61

given the heterogeneity of possible policy considerations, a rigorous

comparison would require the elaboration and application of some
metric of social gains and losses—a task beyond judicial competence.
How, for instance, is a court to know whether, in a case of negligent
misrepresentation by auditors, the benefits of deterring carelessness are
outweighed by the disadvantages of indeterminate liability?55 It is there-
fore hardly surprising that the policy considerations that interest the
court are those that, as the wording of the test indicates, “ought to nega-
tive or limit” the duty.56
Third, the relationship between plaintiff and defendant is fragment-
ed not only by the very presence at the second stage of policy con-
cerns extrinsic to that relationship, but also by the disjunction between
the justice and policy considerations of the two stages. This disjunc-
tion, in turn, requires judges to balance categorically different consid-
erations, in order to determine whether in a given case the policy
considerations are more important than the justice considerations that
they can displace. How is this balancing of incommensurables to be
done? In effect, the two-stage test puts into circulation two different
normative currencies between which no rate of exchange exists.
Fourth, the two-stage test transfigures the notion of foreseeability
itself. When considered within the framework of injustice done and
suffered, foreseeability is an intrinsically correlative notion through
which the law constructs the identical nature of the injustice on both
sides by linking the plaintiff ’s injury to the reason for characterizing
the defendant’s action as wrongful. Accordingly, foreseeability is intern-
ally limited by the scope of the wrongfulness to which it refers. In
contrast, under the two-stage test, foreseeability constitutes a “rela-
tively low threshold”57 for recognizing a prima facie duty, which is

55
Hercules, above n. 52, at 593.
56
For example, in Dobson v. Dobson [1999] 174 D.L.R. 1, at 31 (S.C.C.) the court held that
a mother was not liable for the prenatal injuries that she caused her own child by her negligent
driving. Although injury was foreseeable under the first stage, the court negatived liability on
policy grounds in order to safeguard the pregnant woman’s autonomy and privacy. However,
the court also rejected the suggestion that the existence of a mandatory automobile insurance
regime justified liability on policy grounds. Thus, a judicially enunciated policy prevented the
victim of negligent driving from gaining access to insurance proceeds whose availability had
been legislated. Elsewhere, in Canadian National Railway v. Norsk Pacific Steamship, above n.
44, the court had given extensive attention to insurance factors in determining whether to
deny liability, but insurance is apparently not available as a policy factor that can support liability
at the second stage of the two-stage test.
57
Ingles v. City of Toronto [2000] 183 D.L.R. (4th) 193, at 202 (S.C.C.); Ryan v. City of
Victoria [1999] 168 D.L.R. (4th) 513, at 525 (S.C.C.). In these cases the court treats proximity
62 corrective justice

then extrinsically limited by policy considerations. Any reasonably

prospective damage counts as being foreseeable under the first stage of
the test, without inquiry into why the defendant’s action should be
characterized as a wrongful infringement of the plaintiff ’s right.58
Fifth, even taken on its own the first stage is internally fractured.
Although the formulation of the first stage refers to the existence of
“a sufficiently close connection between the parties,” that connection
is analyzed not (as coherence would require) in terms of a reason for
characterizing the defendant’s action as a wrong relative to the plain-
tiff ’s right, but as a combination of two factors, foreseeability (with its
low threshold) and proximity. Thus, in its first stage the two-stage test
takes over the tensions, noted in the discussion of economic loss in
the previous section of this essay, that result from ascribing to proxim-
ity a policy-based limiting function; the second stage then piles on top
of proximity yet another layer of policy considerations that also have a
limiting function. Within this complex structure, “proximity” is mere-
ly a conclusory label that, as the court frankly observed, states a result
without itself providing a principled basis for liability.59 It stands for a
variety of case-specific factors that go to the determination of wheth-
er it is fair and just, having regard to the relationship between the par-
ties, to impose a prima facie duty of care.60 Where the duty of care has
already been recognized by law, proximity is the term under which

as yielding a low threshold that is synonymous with reasonable foreseeability, not as imposing
an additional restriction. Ryan formulates the first stage as follows:
In order to establish a prima facie duty of care, it must be shown that a relationship of proximity
existed between the parties such that it was reasonably foreseeable that a careless act by the [defend-
ant] could result in injury to the [plaintiff].
In Ingles the formulation is substantially the same, except that “would” replaces “could.”
58
As Robert Keeton observed:
[P]utting the crucial question in terms of whether the injuries were foreseeable . . . carries the
misleading implication that the scope of legal responsibility extends to every consequence that is
foreseeable as a possibility in any degree . . . The crucial standard is better expressed as the question
whether all her injuries were within those risks by reason of which the defendant’s conduct was
characterized as negligence.
Robert Keeton, Legal Cause in the Law of Torts (1963), 55. A graphic expression of how low
the threshold of foreseeability can be appears in Modbury Triangle Shopping Centre v. Anzil
[2000] 176 A.L.R. 411, at 436 (H.C. Aust., per Hayne J.):
In almost every case in which a plaintiff suffers damage it is foreseeable that, if reasonable care is
not taken, harm may follow. As Dixon CJ said in argument in Chapman v. Hearse, ‘I cannot
understand why any event that does happen is not foreseeable by a person of sufficient imagination
and intelligence.’ Foresight of harm is not sufficient to show that a duty of care exists.
59
Hercules, above n. 52, at 588.
60
Cooper v. Hobart [2001] 206 D.L.R. (4th) 193, at 204 (S.C.C.).
 the disintegration of duty 63

one subsumes the various factors that differentiate the different cat-
egories of liability, such as physical and proprietary harms, negligent
misrepresentation, certain cases of economic loss, the municipality’s
obligation to prospective purchasers to inspect new housing develop-
ments, the duty of public authorities who have undertaken a policy of
road maintenance to execute the maintenance with due care, and so
on.61 In this way proximity thematizes not the unifying principle
underlying the general conception of duty, but the disparateness of
particular kinds of duty.
In short, the introduction of the two-stage test has transformed Cana-
dian negligence law into an enquiry into one-sided policy consider-
ations at the ultimate stage that are extrinsic to justice between the
parties and that are mysteriously balanced against a first stage that com-
bines an excessively expansive notion of foreseeability with the invoca-
tion, under the term “proximity,” of a miscellany of limiting
case-specific factors. This ramshackle enquiry, composed of mutually
alien parts that labor to contain the specter of unlimited liability that it
itself lets loose, is hardly conducive to the elaboration of coherent and
principled justifications for liability. The test represents a high point for
the disintegration of duty. It conceives of the notion of duty as inter-
nally fragmented between and within its stages. It also conceives of the
duties themselves as particular species each of which represents its own
specific considerations of policy and proximity. All that remains of Lord
Atkin’s notion of a general conception is the comprehensive verbal
umbrella that applies to, but does not coherently unify, these different
duties.
A contrasting pair of cases illustrates some of these themes. In the
past few years the highest courts in England and then in Canada have
dealt with the liability of auditors for negligently preparing the annual
report of a corporation’s accounts. As is well known, investors in the
stock market buy and sell on the basis of the information in these reports.
Previously, accountants had been held liable for the reliance losses
caused by negligence in a report they knew was prepared for the guid-
ance of a specific class of investors with respect to a specific class of
transactions.62 The question that now arose was whether investors gen-
erally could recover for their admittedly foreseeable reliance on state-
ments in a report prepared for the corporation’s annual meeting. Both

61
Ibid., at 205.
62
Haig v. Bamford [1976] 72 D.L.R. 68 (S.C.C.).
64 corrective justice

the English and the Canadian courts answered in the negative, but they
used different modes of reasoning.
For the House of Lords,63 the crucial issue was whether the requisite
normative link existed between the defendant’s negligence and the plain-
tiff ’s loss. Foreseeability of the loss could not ground liability, because the
duty of care could not be considered in abstraction from the kind of
damage which the defendant must avoid causing. In this case the audit
was presented to fulfill a statutory obligation aimed at the informed exer-
cise by the corporation’s stakeholders of their powers of corporate gov-
ernance. Because the plaintiff ’s loss was not connected to the purpose of
the audit, the defendant owed no duty of care with respect to that loss.
The Canadian case64 followed the English decision in result but
transformed the structure of its thought. Applying the two-stage test,
the Supreme Court of Canada determined that a prima facie duty of
care arose because investor reliance on the audit was both reasonable
and foreseeable. This duty, however, was negatived by the undesirable
social consequences of the indeterminate liability generated by so
broad a conception of foreseeability. Among these consequences were
the increased insurance premiums, the higher costs faced by account-
ants, the opportunity costs in time spent on litigation rather than on
generating accounting revenue, reduction in the availability in
accounting services as marginal firms are driven to the wall, and
increased costs for consumers. Looking to the purpose of the auditor’s
report, as was done in the English case, was, “in reality, nothing more
that a means by which to circumscribe—for reasons of policy—the
scope of the representor’s potentially unlimited liability.”65
The contrast between these two cases is stark.The English judgment
straightforwardly applied the mode of reasoning that had been set out
in the classic twentieth-century cases adapted to negligent representa-
tion. The House of Lords examined whether the investor suffered the
kind of loss that lies within the scope of the auditor’s duty, which, in
turn, was defined and limited by the purpose for which the audit was
required. Given that such audits are not prepared for the guidance of
decisions to buy or sell shares, the defendant could not be viewed as
having assumed responsibility for the plaintiff ’s losses. Because these
investment transactions fell outside the range of the defendant’s duty,

63
Caparo, above n. 41.
64
Hercules, above n. 52.
65
Ibid., at 590.
 the disintegration of duty 65

the plaintiff ’s loss, despite being the foreseeable outcome of a negli-

gently prepared report, did not count as the suffering of a wrong at the
defendant’s hands. In this judgment the purpose of the audit functions
as the normative idea through which the court considers the connec-
tion between parties. The reasoning is relational throughout, and liabil-
ity is denied because the plaintiff ’s loss is not normatively correlated to
the defendant’s negligence.66
The Canadian case transforms this relational reasoning into a policy-
based restriction on liability. The two-stage test, so the court claims,
allows us to recognize the relational criteria “for what they really are—
policy-based means by which to curtail liability.”67 Reliance that is rea-
sonable and foreseeable, now no longer situated within the framework
of doing and suffering the same injustice, creates a prima facie duty
that is “potentially infinite.”68 To solve this problem of its own creation,
the court curtails the scope of the duty by reference to policy factors.
Hence, what the court calls the simple justice of the plaintiff ’s claim
yields to the need to preserve the availability of accounting services in
Canada. The court does not explain why justice is to be sacrificed to
the need for accounting services, or why the policy to maintain
accounting services outweighs the policy of deterrence, or how the
court knows that the current level and pricing of accounting services
in Canada is optimal. Instead of offering reasons for thinking that the
defendant did not wrong the plaintiff, the court indulges in specula-
tions beyond its competence about the undesirable consequences of
liability for the providers and consumers of accounting services.69
These particular differences in the two judgments reflect a wider
contrast between the nature of the justifications that they employ. The

66
The following passage from the English case is a striking judicial formulation of the
notion that the sequence from negligence to injury forms a single normative unit:
[A] postulated duty of care must be stated with reference to the kind of damage that the plaintiff
has suffered and in reference to the plaintiff or the class of which the plaintiff is a member . . . His
duty of care is a thing written on the wind unless damage is caused by the breach of that duty;
there is no actionable negligence unless duty, breach and consequential damage coincide . . . ; for
the purpose of determining liability in a given case, each element can be defined only in terms of
the others.
Caparo, above n. 41, at 599 per Lord Oliver, quoting Brennan J. in Council of the Shire of
Sutherland v. Heyman [1985] 60 Aust. L. Rep. 1, at 48 (H.C. Aust.).
67
Hercules, above n. 52, at 592.
68
Ibid., at 591.
69
For similar speculations as to whether auditor liability for foreseeable loss is in the pub-
lic interest, see Esanda Finance Corporation v. Peat Marwick Hungerfords (Reg) [1997] 142
A.L.R. 750, at 782 (H.C. Aust., per McHugh J.).
66 corrective justice

English judgment assumes that the justifications relevant to liability

embrace both parties as the doer and the sufferer of the same injustice.
Because of liability’s correlative significance for both parties, such jus-
tifications are required if a judgment is to provide coherent reasons for
considering whether one party is liable to the other. Conversely, a
court that decided issues of liability without reference to such justifi-
cations would fail to treat the parties fairly in relation to each other.
The reasoning in the Canadian judgment provides an example of this
failure. Its expansive conception of foreseeability at the first stage
focuses on the prospective damage that might result from the defend-
ant’s action without articulating the reason for regarding that damage
as a wrongful injury to the plaintiff ’s right. Similarly, its restrictive elu-
cidation of policy at the second stage deals only with the effect of
liability on accountants and is thus unrelated to the plaintiff ’s claim to
have suffered a wrong. As a result the judgment contains a series of
unintegrated considerations that are divorced from the articulation of
the wrong that might link the parties to the action. Such consider-
ations are inherently incapable of fairly and coherently determining
whether the defendant should be held liable to the plaintiff .
More recently the Supreme Court of Canada has reaffirmed the
two-stage test, but in terms that might open a path back to a more
coherent approach to the duty issue. In Cooper v. Hobart70 the plaintiff
was an investor who was suing for money lost by advancing funds to a
registered mortgage broker who had used the funds for unauthorized
purposes. The investor sued the statutory regulator, the Registrar of
Mortgage Brokers, claiming that her losses would have been avoided
or diminished had the Registrar acted more promptly in suspending
the broker or notifying investors that the broker was under investiga-
tion. The court held that under the two-stage test the Registrar did
not owe the investor a duty of care even though the investor might
well have been able to show that her losses were the reasonably fore-
seeable consequence of the defendant’s alleged negligence.
The crucial determination was that the proximity required by the
first stage was lacking. The Registrar’s powers and duties were entirely
the creation of statute, and an analysis of the statute showed that the
Registrar’s duties were owed to the public rather than to individual
investors. The point of the regulatory scheme was to ensure efficiency

70
Cooper v. Hobart, above n. 60; see also the companion case Edwards v. The Law Soci-
ety of Upper Canada [2001] 206 D.L.R. (4th) 211 (S.C.C.).
 the disintegration of duty 67

in the mortgage marketplace. This required the Registrar to consider

not only the private interests of individual investors but also the pub-
lic interests in access to capital through mortgage financing and in
public confidence in the system as a whole. Consequently, the statute
could not be construed to impose on the Registrar a tort duty of care
that is specific to individual investors. Such a duty would be inconsist-
ent with the overall regulatory purposes evident in the powers granted
under the statute. The statutory purpose thus defined the notion of
proximity applicable to the relationship of the parties. The court’s
conclusion was that even if foreseeability was present, proximity was
not. Therefore, the search for a duty of care did not survive the first
stage. However, for the sake of completeness, the court proceeded to
the second stage, where it found that even if a prima facie duty had
been established, it would have been negated for overriding policy
reasons.
Even as the court reaffirms the two-stage test, a significant feature
of this decision is the comparative effacement of the second stage in
the overall architecture of the court’s thought.The second stage remains
as the point at which one examines “residual policy considerations”
that concern not the relationship between the parties but the effect of
postulating a duty on the legal system and society more generally.71
However, in contrast to the court’s previous jurisprudence, which had
often treated the second stage as decisive, the second stage has now
become less prominent.
There are three indications of this. First, the court explicitly says
that the second stage will never arise if the duty of care has fallen
within a recognized category of liability. Moreover, in this case (where
the court indicated that overriding policy reasons would have negated
a prima facie duty, even had one been found at the first stage) the
considerations mentioned as relevant did little independent work.
Most of the “residual policy considerations” centered on the Regis-
trar’s exercise of a public function entailing discretionary decisions
that balance public and private interest. The policy considerations of
the second stage were therefore not different in kind from the consid-
erations that went to proximity at the first stage.Third, the court treat-
ed the purpose of the statutory provisions as relevant to the elucidation
of proximity in the first stage, in striking contrast to its earlier treat-
ment of the statutory requirement of an audit in the negligent mis-

71
Cooper v. Hobart, above n. 60, at 206.
68 corrective justice

representation case mentioned earlier. The effect of using the statute

to inform proximity is to bring the analysis close to the parallel Eng-
lish case, in which the duty was defined and limited by the purpose
for which the statute mandated an audit. The court’s application of
the two-stage test in Cooper v. Hobart can, accordingly, perhaps be
regarded as implicitly repudiating its earlier reduction of relational
criteria to policy factors that are extrinsic to the parties’ relationship.
Accordingly, the following picture emerges from Cooper v. Hobart
about the second stage.The second stage is now to have only a restrict-
ed application; considerations that appear at the second stage may rep-
licate the contents of proximity at the first stage; and factors
previously considered at the second stage as extrinsic to the relation-
ship are now to be analyzed as relational at the first stage. Perhaps, as
its previous content is poured out of the second stage into the first
and what remains is seen as redundant, Canadian courts will pay less
and less attention to the second stage. Presumably, a duty found at the
first stage will rarely be negated at the second stage.72 If so, the new
case will be seen as the first step in the atrophy of the two-stage test.
What about the first stage? Cooper v. Hobart contains a welcome
emphasis on the relational nature of the considerations that govern
the first stage.73 Perhaps this attention to the relational aspect of the
duty issue will be further strengthened if the second stage does indeed
atrophy, and the reasoning about duty is effectively confined to the
examination of what is foreseeable and proximate. For the moment,
however, the court has not yet developed a fully adequate view of
what it means for considerations to be relational. Instead of being
understood as a coherent and integrated whole, the first stage is seen
as an amalgam of foreseeability plus proximity, with proximity itself
embodying a collection of specific indicia that vary with the particu-
lar relationship in question. Perhaps its newly announced sensitivity to
the relational quality of the first stage will eventually lead the court to
two salutary realizations. The first is that proximity cannot capture
what is normatively significant about the relationship between the
parties so long as it is regarded simply as something that is added to an
expansive notion of foreseeability from the outside in order to restrict

72
An example of this rare occurrence is R. v. Imperial Tobacco Canada [2011] 3 S.C.R.
45, in which the court held that policy reasons at the second stage negated a duty of care
regarding negligent misrepresentations on which the plaintiffs’ reliance was invited.
73
“The proximity analysis involved in the first stage of the Anns test focuses on factors aris-
ing from the relationship between the plaintiff and the defendant.” Cooper, above n. 60, at 203.
 the disintegration of duty 69

it. Rather (if these terms are to be used) proximity should be under-
stood to reveal the restricted meaning that forseeability itself has in
the negligence context—that is, that foreseeability is a way of inquir-
ing into the risks by reference to which the defendant’s action is char-
acterized as negligent. The second is that the relational quality that the
court now highlights has to be expressed in normative categories that
are themselves relational. Accordingly, behind the particular duties
must stand a general conception of duty governed by the correlativity
of right and duty—that is, by a normative framework whose elements
are intrinsically related to each other. Then the duty issue will again
take its place in the coherent ensemble of concepts that treats the
defendant’s creation of the unreasonable risk and the plaintiff ’s suffer-
ing from the risk’s materialization as falling under the same correla-
tively structured wrong.

5. Two notions of policy

This final section will focus on the connection between duty and
policy. Throughout the common law world the notion that the for-
mulation of the duty of care involves a determination of policy is
accepted almost as a truism. In his tort judgments Lord Denning often
gave voice to this supposed truism. For instance, whether correction
officers owe a duty to persons whose property might be damaged by
the escape of borstal trainees is, he said, “at bottom a matter of public
policy that we as judges must resolve.”74 Similarly, in dealing with lia-
bility for economic loss he observed that “whenever the courts draw a
line to mark out the bounds of duty, they do so as a matter of policy.”75
The argument in this section is that the link with policy, at least as it is
commonly understood, has played a major role in the disintegration
of duty and yet rests on a misconception.76 The misconception involves
running together two distinct notions of what policy means.

74
Dorset Yacht v. Home Office [1969] 2 All E.R. 564, at 567 (C.A.).
75
Spartan Steel & Alloys Ltd. v. Martin [1972] 3 All E.R. 557, at 561 (C.A.); see also Lamb v.
London Borough of Camden [1981] Q.B. 625, at 636 (C.A.).
76
The difficulties in the invocation of policy have been illuminatingly discussed in essays
by Martin Stone; see Stone, “Focusing the Law: What Legal Interpretation is Not,” in Law
and Interpretation: Essays in Legal Philosophy, ed. Andrei Marmor (1995), 31, 72–84; Stone,
“Formalism,” in The Oxford Handbook of Jurisprudence and Philosophy of Law, ed. Jules Cole-
man and Scott Shapiro (2002), 166, 187–204.
70 corrective justice

The first notion is the common view that policy involves articulat-
ing some independently desirable goal(s) and then dealing with a par-
ticular tort case in a way that forwards these goals or, if they are in
tension, balances some against others to produce a result that is desir-
able overall. The goals are independent both in the sense that they rest
on justifications that are independent of tort law, to which they are
then applied, and that they are independent of one another, so that
they may represent incompatible normative impulses that need to be
balanced. For instance, a favorite policy of Lord Denning was that
losses should be widely distributed, because it is easier for many to
bear comparatively small losses than for a single person to bear a com-
paratively heavy one.77 The justification for this policy is independent
of tort law, in that the policy states a normatively appealing way to
deal with any sort of loss, not with a loss merely caused by tortious
wrongdoing. Indeed, tort law imposes an artificial limit on its oper-
ation by restricting the distribution of losses to the insurance pool
brought into play by tort law’s initial allocation of the loss.78 More-
over, when this policy favors the defendant, it may come into conflict
with other policies, such as deterrence, that would be forwarded by
liability. Hence, a determination that takes account of all the heteroge-
neous policies can be regarded as involving a process of balancing
those that favor the plaintiff against those that favor the defendant.
The invocation of such independent policies entails the disintegra-
tion of duty as a systematic and coherent concept. Given the hetero-
geneity of the available policies and their different weightings in the
balancing process, a systematically unified conception of duty based
on (in Lord Atkin’s words) “the element common to all cases in which
[a duty] is found to exist”79 is out of the question. The variety of pol-
icies and the shifting balance among them leaves no place for a com-
mon element on which the various duties (again in Lord Atkin’s
words) “must logically be based.”80 In these circumstances there can
only be different specific kinds of duty, with each kind representing
the particular policies or the particular balance among policies that
are recognized as decisive in situations of that sort. Moreover, the con-
ception of duty is inwardly fragmented into the various policies that

77
Spartan Steel & Alloys v. Martin, above n. 75, at 564; Lamb v. London Borough of
Camden, above n. 75, at 637 (C.A.).
78
Weinrib, The Idea of Private Law, 36–38.
79
Above n. 1.
80
Ibid.
 the disintegration of duty 71

favor one party or the other. The duty issue is therefore seen as the
locus not for defining the wrong identically from the standpoint of
both parties, but for forwarding or balancing policies that rest on con-
siderations that apply differently to each of them.
Were the duty issue necessarily to involve policies of this sort, the
general conception of duty as coherently linking the parties as doer
and sufferer of the same wrong would be a chimera. That such policies
are crucial to determining the duty of care is often taken for granted.
In Cooper v. Hobart, for instance, the Supreme Court of Canada approved
the proposition that a decision about duty “is in fact a conclusion
embracing within it, and yet concealing the identity of, the several
considerations of policy, and the balancing of interests which have led
the court to decide that a duty is owed.”81 Thus, the court regards it as
self-evident that the duty issue requires the consideration of a multi-
plicity of policies that represent interests to be balanced.
But is this really self-evident? It is notable that in Cooper v. Hobart
itself the court, after proclaiming the necessity of balancing, seemed at
the first stage to resolve the proximity issue (which it declares to be an
issue of policy) without doing any balancing whatsoever. The judg-
ment did not compare the interests of the investors with the interests
of the Registrar of Mortgage Brokers, putting each set of interests
into notional pans whose contents were calibrated to some notional
measurement of weight, and seeing which pan notionally sank.82
Rather, it analyzed the statute under which the Registrar of Mortgage
Brokers operated to show that the Registrar’s duty was of a public
nature and thus not owed to specific individuals. The court merely
drew out what it thought, rightly or wrongly, was implicit in the stat-
utory scheme within which the Registrar functioned. The exercise in
question was not one of balancing policies or interests but of specify-
ing the nature of the Registrar’s duty through analysis of the institu-
tional framework created by the statute.
The comment about policy in Cooper v. Hobart exemplifies a peculiar
inversion.The comment posits the existence of a familiar practice about
which unsophisticated observers might make certain assumptions, and
then treats those assumptions as an illusion that can be dispelled by

81
Cooper v. Hobart, above n. 60, at 202.
82
The procedure envisaged would perhaps be reminiscent of the timeless spoof on bal-
ancing in Aristophanes’ Frogs, where Dionysus in the underworld judges between Aeschylus
and Euripides by employing a set of scales that balances the ponderous verse of the former
against the fluffy verse of the latter.
72 corrective justice

pointing to what is “in fact” the case. The practice in question is the
giving of reasons for holding that a duty of care exists, and the illusory
assumption is that this practice does not involve attention to “the sev-
eral considerations of policy and the balancing of interests.” The com-
ment then purports to dispel this illusion by insisting that determinations
of duty “in fact” do require the embracing of multiple policies and the
balancing of interests. But at the point of application this effort at
enlightenment is immediately undermined by reasoning that exempli-
fies the very illusion that was supposed to have been dispelled. The
court’s consideration of the actual duty at stake suggests that the com-
ment about policy has gotten it backward. It turns out that the illusory
assumption, rendered familiar over the years by constant scholarly and
judicial repetition in the aftermath of legal realism, is that the duty issue
necessarily involves the balancing of the interests represented by differ-
ent policies. The illusion can be dispelled by attention to what “in fact”
happens in cases that ignore this supposed necessity.
There is, however, a second notion of policy that is both required
by and consistent with the conception of tort liability as a response to
the doing and suffering of a wrong. Consider Lord Diplock’s judg-
ment in Home Office v. Dorset Yacht,83 where the issue was whether cor-
rection officers were under a duty of care to a yacht owner whose
vessel was damaged when borstal boys under their supervision were
negligently given an opportunity to escape from an island on which
they were working. In terms of the framework for duty suggested
above in section 3, this called for a determination of whether the offic-
ers’ negligent behavior breached a duty that was correlative to the
plaintiff ’s undisputed right in the boat. Lord Diplock began by signal-
ing his agreement with Lord Denning’s view that issue was “at bot-
tom a matter of public policy that we, as judges, must resolve.”84
However, the significance that Lord Diplock attached to public policy
was different from the one that appears in many of Lord Denning’s
judgments. Whereas Lord Denning often equated policy with inde-
pendent goals such as loss-spreading, Lord Diplock understood the
reference to the judges’ role in resolving matters of policy to be an
invitation to explore the specifically judicial function of casuistically
developing the law. The task for Lord Diplock was not to identify and
balance independent goals, but to elucidate the meaning of Lord

83
[1970] A.C. 1004, at 1057 (H.L.).
84
Ibid., at 1658.
 the disintegration of duty 73

Atkin’s general conception of duty in the circumstances of the present

case. This required a number of steps: first, an identification of the rele-
vant characteristics, as informed by the general conception, of the
kinds of conduct and relationships that have been held to give rise to
a duty; second, a comparison, again influenced by the general concep-
tion, of the characteristics of the situation he was considering with the
characteristics of other situations where a duty had been found; and,
third, in a novel case in which these sets of characteristics from differ-
ent situations were not congruent, an evaluation, still under the guid-
ance of the general conception, of the significance of the differences
and of the substitutability of a present characteristic for a missing one.
The general conception of the duty of care thereby constituted the
standpoint from which the characteristics of various situations were
selected, compared, and evaluated.
In Dorset Yacht, this reference to policy—that is, to the process of
casuistic judgment under the general conception of duty—worked as
follows. Previous cases had held the correction authority liable for
negligence that resulted in one detainee’s injuring another.85 Lord
Diplock considered it a “rational extension”86 of the principle in those
cases to substitute for the custodian’s right to control the physical
proximity between detainees the knowledge that the custodian had or
ought to have had of the particular risk to which, because of the phys-
ical proximity of its property, the plaintiff would be exposed by the
defendant’s negligence. The general conception of the duty of care,
which Lord Atkin formulated as owed to persons so closely and
directly affected by the act that the actor ought to have them in con-
templation, led Lord Diplock to differentiate between the particular
risk of damage consequent on escape, which affected only those in
the vicinity, and the general risk of suffering from criminal activity, to
which all members of the public are exposed. It was the former that
constituted the unreasonable risk created by the defendant’s conduct,
and that therefore rendered the plaintiff a person so closely and directly
affected that the prospect of this damage ought to have been in the
defendant’s contemplation. Because the escape was from an island and
could not be attempted without watercraft, the owners of boats
moored in the vicinity were within the class to whom a duty of care

85
Ellis v. Home Office [1953] 2 All E.R. 149 (C.A.); D’Arcy v. Prison Commissioners,The
Times, 17 November 1955.
86
Dorset Yacht, above n. 83, at 1071.
74 corrective justice

was owed.The damage to the plaintiff ’s yacht was, therefore, the materi-
alization of a risk that was unreasonable because of the prospect of this
kind of damage. Thus, the parties were the doer and sufferer of the
same wrong.
This second notion of policy reflects the existence of scope for
judgment in the determination of a duty of care. The duty of care
does not operate in a mechanistic or syllogistic fashion. This is espe-
cially the case when particular kinds of duties or duties in particular
cases are regarded as instantiations of a general conception of duty.
The general conception is, after all, a conception and not a recipe or
even a “test.” The general conception does not state a specific formula
from which one can immediately discern whether a duty is present in
any and every particular case; rather, it brings out what the idea of
duty must be if the law of negligence is coherently to link the defend-
ant’s negligent act and the plaintiff ’s injury. The very generality of the
conception means both that it is not defined by reference to any par-
ticular situation in which a duty is found and that it informs all such
particular situations. It therefore requires to be related to its instantia-
tions through an exercise of judgment, the point of which is to exhib-
it what, in the view of the person making the judgment, the duty of
care means in the circumstances of a particular case. Accordingly, the
general conception of duty does not render superfluous this exercise
of judgment but guides it, indicating that the judgment is to be direct-
ed toward the existence of a right and of negligence with respect to
that right. In this way the exercise of judgment is an operation of
practical reason that plays itself out within the ensemble of concepts
that the law constructs for considering whether the defendant has
done and the plaintiff has suffered the same injustice.
Lord Diplock’s opinion is outstanding for providing an account by a
great judge of how this process of judgment might be described. He
does not regard the general conception of the duty as a “test” that is
applied externally to the facts, like a touchstone to gold; that would “mis-
use as a universal”87 what Lord Diplock regards instead as a useful guide
to the relevant characteristics. Rather, because the process of judgment is
supposed to bring together the general conception of duty and the
determination of a particular duty, it operates simultaneously from both
ends. On the one hand, it attends to the fact situation at hand and to the
history of judicial determinations in analogous fact situations. On the

87
Ibid., at 1060.
 the disintegration of duty 75

other hand, it elucidates the relevance of particular facts and similarities

by reference to the general conception of duty that Lord Atkin articu-
lated. As Lord Diplock says, the judge starts by identifying “the relevant
characteristics common to the kinds of conduct and relationships
between the parties which are involved in the case for decision and to
the kinds of conduct and relationships which have been held in previous
decisions of the courts to give rise to a duty of care,”88 but the judge
must “know what he is looking for; and this involves his approaching his
analysis with some general conception of conduct and relationships
which ought to give rise to a duty of care.”89 The result is the fusion of
general and particular in a judgment about whether the defendant in the
case at hand breached a duty owed to the plaintiff.
Under the second notion of policy the scope for judgment need not
involve, as it did in Dorset Yacht, a casuistic comparison of the character-
istics that give or have given rise to a duty of care. It can also involve
the elucidation through legal argument of the issues of law that pertain
to the relationship of the parties as doer and sufferer of the same wrong.
Among the legal issues that may require elaboration in any given case
are the nature of the plaintiff ’s right (for example, under what circum-
stances, if at all, does the plaintiff have a right to security from psychi-
atric injury?),90 the nature of correlative wrongdoing by the defendant
(for example, is the distributor of a product under a duty of care with
respect to its safety?),91 and the relevance of the connection between
the supposed duty of care and other juridical considerations affecting
the parties. Thus, problems in determining the duty of care “may con-
cern the need to preserve the coherence of other legal principles, or of
a statutory scheme which governs certain conduct and relationships.”92
An example is the long-standing controversy, now exemplified in the

88
Ibid., at 1058.
89
Ibid. One can contrast Lord Diplock’s view that casuistic analysis proceeds under a
general conception with that of Dawson J. in Hill v.Van Erp, above n. 8, at 177:
Reasoning by analogy from decided cases by processes of induction and deduction, informed by
rather than divorced from policy considerations, is not, in my view, dependent for its validity on
those cases sharing an underlying conceptual consistency. It is really only dependent on the fact
that something more than reasonable foreseeability is required to establish a duty of care and that
what is sufficient or necessary in one case is a guide to what is sufficient or necessary in another.
However, it is hard to see how one case can serve as a guide to another unless there is implicit
some common standpoint that informs the comparison between them.
90
Above n. 32. See also above n. 36.
91
Watson v. Buckley and Osborne, Garrett and Co. Ltd (Ogee Ltd.) [1940] 1 All E.R. 174
(K.B.).
92
Sullivan v. Moody [2001] 183 A.L.R. 404, at 415 (H.C. Aust., per Hayne J.).
76 corrective justice

varying treatments of economic loss arising out of a defect of quality,

about the relationship between contract and the duty of care in negli-
gence.93 Even if the particular parties are not bound to each other by
contract, the nature and limits of contractual obligation may nonethe-
less have the implications for the existence of a tort duty of care
between them. For instance, the argument may be made in cases of
defect in quality, that apart from contract one has no right to an object
of a certain quality and that, therefore, no duty of care regarding qual-
ity arises in tort. Contemporary courts disagree as to the success of this
argument,94 but, whether successful or not, it is an argument that per-
tains to the relationship between the parties without invoking any
independent policy. This is because the parties to a tort suit are related
to each other as legal persons—that is, as parties whose legal relation-
ships are expressive of the systematic coherence of the entire law of
obligations. Accordingly, they do not interact juridically apart from
the whole ensemble of intertwined legal concepts and principles that
governs their participation in the law’s systematic nature.95

93
Murphy v. Brentwood District Council [1990] 2 All E.R. 908 (H.L.); Winnipeg Con-
dominium Corp. No. 36 v. Bird Construction [1995] 121 D.L.R. (4th) 193 (S.C.C.); Bryan v.
Maloney [1995] 128 Aust. L.R. 163 (H.C.).
94
Contrast the Murphy and Winnipeg Condominium cases, above n. 93.
95
The Supreme Court of Canada in Cooper v. Hobart, above n. 60, also held that “differ-
ent types of policy considerations are involved at each of the two stages” of the two-stage
test; ibid., at 202. At the first stage policy goes to the definition of proximity in the circum-
stances of the case; it focuses on factors that arise out of the relationship between the parties.
The court describes this enquiry as follows (ibid., at 204):
Defining the relationship may involve looking at expectations, representations, reliance, and the
property or other interests involved. Essentially, these are factors that allow us to evaluate the close-
ness of the relationship between the plaintiff and the defendant and to determine whether it is just
and fair to impose a duty of care upon the defendant.
Then at the second stage “the question remains whether there are residual policy consider-
ations outside the relationship of the parties that may negative the imposition of a duty of
care.” Ibid., at 203.
This distinction between relational and extra-relational policy considerations is welcome,
and one can expect, if the second stage recedes in significance (as suggested above in section
4 of this chapter), Canadian courts in the future will put greater effort into elucidating the
relational aspect of policy. The court, however, still has a confused conception of what the
distinction really is. First, it seems to view both kinds of policy as involving a balancing of
interests that legal reasoning embraces and yet conceals; above at n. 81. Moreover, factors that
are properly relational it views as extra-relational. For example, it regards the effect of recog-
nizing a duty of care on other legal obligations as an extra-relational policy consideration;
ibid., at 206. Similarly, it regards as extra-relational the question of whether recognition of a
duty of care would “create the spectre of unlimited liability to an unlimited class” (ibid.),
whereas, because the liability should be limited by the scope of the right to which the duty
is correlative, this properly belongs to the relational analysis.
 the disintegration of duty 77

The second notion of policy is merely a way of signaling the pres-

ence of a conception of the duty of care that becomes significant for
particular cases through the exercise of practical judgment. Such an
exercise may involve the casuistic consideration and comparison of
cases in the light of the conception of duty that they instantiate, or it
may involve a process of legal argument that elucidates the right and
the correlativity of the wrong and that coherently integrates the con-
ception of duty with the other norms in play in the circumstances of
the case. Recourse to practical judgment is concomitant to the inher-
ent generality and abstractness of the legal concepts, including the
duty of care, out of which the relationship between the parties is
juridically constructed. Policy in this sense differs from the first notion
of policy suggested earlier, which referred to independent goals out-
side the relationship and to the balancing of interests where these goals
are in tension. The practical judgment involved in casuistic analysis
and legal argument does not actualize goals extrinsic to the parties’
relationship as doer and sufferer of a wrong, but, rather, explicates the
legal meaning of that relationship in its particular circumstances. Policy
in this sense is not only consistent with but also required by the gen-
eral conception of duty. For the very generality of that conception
necessitates its being related to the particular case by an exercise of
practical judgment. Through practical judgment the indeterminacy
of the general conception of duty becomes determinate for the case
at hand.
Thus, inasmuch as the general conception of duty is a constituent
of the coherent legal relationship between the doer and the sufferer of
the same wrong, it is only the first notion of policy but not the second
that is inimical to it. Only the first notion of policy effaces the coher-
ence of the parties’ relationship in the name of external goals that
favor (and may require balancing between) the interests of one party
or the other. The second notion, by contrast, far from effacing that
coherence, posits the exercise of practical judgment that renders it
effective for a particular case.
Whence arises the mistaken idea that the duty issue, being (in Lord
Denning’s words) “at bottom a matter of public policy”96 requires
recourse to the first notion of policy? Perhaps the answer is that this
idea is part of the damnosa hereditas of instrumentalist legal thinking.
Confronted by legal concepts that are indeterminate, that is, that do

96
Above n. 75
78 corrective justice

not immediately dispose of the particular issue at hand but require a

further operation of legal argument or casuistic reasoning, instrumen-
talists assume that two alternatives are available: either these concepts
deductively produce legal certainty, or they are merely the rhetorical
cover for the identification and balancing of external goals. Because
no deductive framework is available, all that remains is reasoning in
terms of external goals. Having realized that the legal material does
not allow judges to be conceived as automata devoid of freedom, they
assume that the only way to exercise this freedom is to choose and
balance goals.97
This conclusion is mistaken for two related reasons. First, it poses a
false choice between deduction and instrumentalism. Deduction is
not the exclusive—or even a very important—mode of reasoning
internal to the determination of liability. It may well be the case that
no interesting legal question can be approached deductively, with the
major premise being provided by an unambiguous statement of the
law, the minor premise by an unambiguous recital of the facts, and an
instant and unshakeable conclusion emerging from the subsumption
of the latter under the former. But to think that the absence of deduc-
tion leads inexorably to the necessity for identifying and balancing
independent policies is to assume that deduction is the only move
internal to the elucidation of legal relationships. Ignored is the possi-
bility of the kind of reasoning included under the second notion of
policy, where the judge either compares the relevant characteristics of
one case with other cases that instantiate the same general conception
or elucidates the meaning of the conception in question in a way that
coherently construes both the legal relationship between the parties
and the whole ensemble of legal concepts.
Second, the instrumentalist conclusion moves too quickly from the
indeterminacy of the general conception of duty to the external goals

97
In “Privilege, Malice and Intent,” (1894) 8 Harv. L. Rev. 1, Oliver Wendell Holmes
provides a classic example of this approach. In Holmes’s view, adjudication involves decisions
about questions of policy—that is, legislative questions concerning relative advantages to the
community of liability and no liability. These questions have to be addressed by comparing
the gain from permitting the impugned act with the loss that the act inflicts. Judges shy away
from acknowledging that this is the true ground of their decisions, because “the moment you
leave the path of merely logical deduction, you lose the illusion of certainty which makes
legal reasoning seem like mathematics. But certainty is only an illusion, nevertheless.” Ibid.,
at 7. Accordingly, judges present their decisions not as grounded in legislative policy but as
“hollow deductions from empty general propositions” or as unexplained postulates about
what constitutes a wrong.
 the disintegration of duty 79

that are supposed to ground the decision in a particular case. The fact
that the general conception of duty does not immediately determine
particular cases merely indicates the existence of scope for practical
judgment. It does not imply that the general conception is without
meaning, a mere mirage that vanishes when one focuses on it from
close in, leaving an empty space that can be filled up by whatever the
judge thinks is a good idea. Rather, in leaving scope for practical judg-
ment, the general conception indicates what the judgment must be a
judgment about. What Lord Atkin himself postulated was “a general
conception of relations giving rise to a duty of care.”98 Accordingly, the
exercise of practical judgment through which this general conception
is brought home to a particular case involves reasoning that is rela-
tional, not reasoning about goals that are independent of the relation-
ship. Such relational reasoning is precisely what is encompassed in the
second notion of policy. The first notion of policy, in contrast, cannot
determine the meaning of the general conception of duty in a particu-
lar case, because it does not address itself to that conception. Instead,
by inquiring into the independent goals that might be forwarded by
decisions about liability, it offers an answer to a question that the law
of negligence does not ask, while ignoring the question that it does
ask.

6. Conclusion
Lord Atkin’s judgment in Donoghue v. Stevenson is one of the great
monuments of the modern law of negligence. Sweeping aside the
received idea that negligence law was comprised of a miscellany of
particular duties, he suggested that there must be a general conception
of duty based on the prospective injury to others from unreasonable
risk-creation. This general conception implied coherence both among
the particular duties and within the conception of duty itself. The
argument in this chapter has been that this idea of coherence requires
that the parties to a negligence action be understood as the doer and
sufferer of a single wrong, and that the wrong must be seen as an
integrated sequence in which prospect of the plaintiff ’s injury is a rea-
son for considering the defendant’s act negligent. The leading cases
of the twentieth century on the duty of care and on proximate cause

98
Above n. 1 (emphasis added).
80 corrective justice

developed this conception of a coherent civil wrong. By avoiding

basing liability on one-sided justifications, this approach to the law of
negligence was normatively coherent, consistent with the judicial role
and judicial competence, and fair to both parties. In contrast, more
recent cases, especially those that employ the two-stage test, are appeal-
ing to a notion of policy in which the promotion and balancing of
external and independent goals is leading the law back to the disinte-
gration of duty that Lord Atkin repudiated. There is, however, another
notion of policy that, avoiding reference to external and independent
goals, elucidates the relational significance of the wrong through an
exercise of practical judgment. This second notion of policy merely
reflects the fact that the coherence of its underlying justifications is
itself the supreme policy of the law of obligations.
 3
Remedies

1. Two conceptions of remedies

In this chapter I explore two ways of conceptualizing the relationship
between the basis of the defendant’s liability and the remedy—that is,
between what Peter Birks termed the causative event and the
response.1 In the first way, originally formulated in Aristotle’s account
of corrective justice and later elaborated in the philosophical tradition
of natural right, the causative event is the reason for the remedial
response. In the second way, paradigmatically set out in Kelsen’s pure
theory of law, the causative event is the condition of the remedial
response. For each of these two ways the causative event enters into
the practical reasoning about the response in a different manner. In
the first, the causative event is a condition of the response because it is
the reason for it. In the second the causative event is the reason for the
response because it is a condition of it.
Consider first the following observations from Aristotle’s account
of corrective justice about the function of the judge in a civil action:

When people have a dispute, they have recourse to the judge. To go to

the judge is to go to what is just, for the judge means to be, as it were,
justness ensouled . . . The judge restores equality. As though there were a
line segmented into two unequal segments, he takes away as much as
the larger segment is greater than half the line and adds it to the smaller
segment. And when the whole has been divided into two parts, only

1
Peter Birks, Unjust Enrichment, 2nd ed. (2005), 21.
82 corrective justice

then, when they take what is equal, do they say that each has what is his
own.2

Aristotle represents what properly belongs to each of the disputing

parties as an equal segment of a line. The injustice—the causative
event—consists of the defendant’s having taken part of the segment
that properly belongs to the plaintiff , thereby destroying the underly-
ing equality. The judge remedies this injustice by reattaching to the
plaintiff ’s segment the amount by which the defendant’s part exceeds
the half-line that each should have.
In this representation, the causative event is the reason for the par-
ticular response. What the defendant has done to the plaintiff deter-
mines what the judge requires the defendant to do for the plaintiff.
The defendant is now obligated to return what the defendant unjustly
took from the plaintiff . Because justice between the parties obtains
when the line is equally divided between them, the disturbance of the
equality counts as an injustice, which the judge undoes by restoring
the initial equality. Just as the causative event for liability consists in
the defendant’s taking a segment of the plaintiff ’s line, so the remedy
is the retaking of that segment from the defendant and reattaching it
to the plaintiff ’s part of the line. If one were to ask Aristotle’s judge
why he redivided the line in this way, he would answer that this was
the only just response to the defendant’s action.
One can contrast Aristotle’s conception of the remedy with the
one formulated by Hans Kelsen. For Kelsen, the legal order stipulates
the conditions under which certain coercive acts function as sanctions
that react against illegal acts or omissions. What counts as a wrong or
delict is an act or omission that the legal order makes the condition of
the coercive act; conversely, what counts as the sanction is the coercive
consequence that the legal order attaches to that act or omission.Thus,
the relationship between causative event and remedy is solely that of
condition and consequence: “Given, as condition, behaviour opposite
that which the norm establishes, then a coercive act is to be forth-
coming as consequence.”3
Aristotle and Kelsen link the causative event to the remedy in
entirely different ways. For Aristotle, the causative event is a reason for

2
Aristotle, Nicomachean Ethics,V, 1132a19–29.
3
Hans Kelsen, Introduction to the Problems of Legal Theory, tr. Bonnie Litschewski Paulson
and Stanley L. Paulson (1997), 30.
 remedies 83

the remedy because it is an injustice that the judge, as justness ensouled,

must reverse. For Kelsen, the notion that “a moral value element is
immanent in the concepts of delict and sanction”4 is untenable,
because only the positive legal order imbues an act or omission with
the character of a delict. To put the point another way: for Aristotle,
the causative event is the condition of a remedy because of the kind
of event it is, whereas for Kelsen, the event counts as causative only
because it is the condition of a remedy. In Kelsen’s words:

[A] definite action or refrainment is not—as traditional jurisprudence

assumes—connected with the coercive act because this action or refrain-
ment is a delict, but a definite act or refrainment is a delict because it is
connected with a coercive act, that is, with a sanction as its consequence.
No immanent quality, no relation to a meta-legal natural or divine norm is
a reason for qualifying a specific human behaviour as a delict; but only and
exclusively the fact that the positive legal order has made this behaviour the
condition of a coercive act—of a sanction.5

This juxtaposition of Aristotle and Kelsen brings out the contrast

between relating the causative event to the remedy as a reason or as a
condition. Aristotle and Kelsen are led to these differing conceptions
by the differences in their projects. Aristotle’s interest is in presenting
the form of justice that is immanent in the relationship between the
parties in private law. Integral to this form of justice is the idea that
the direct relationship of the parties characterizes both the causative
event and the remedy—indeed, that it characterizes the remedy
because it characterizes the causative event.Thus, the reason for think-
ing that the defendant’s act is an injustice to the plaintiff is also a rea-
son for thinking that the remedy that corrects the injustice has to have
the same relational structure as the injustice. Kelsen, in contrast, is
concerned not with justice but with the posited nature of law. Because
a norm can be legally valid even if it is thought to be unjust, the con-
nection between the causative event and the remedy must be under-
stood in terms of condition and consequence with respect to the
coercion that the positive law mandates. Whereas Aristotle views the
relationship of causative event to remedy as immanent to the structure
of justice between the parties, Kelsen regards that relationship as

4
Hans Kelsen, Pure Theory of Law, tr. Max Knight (1967), 111.
5
Ibid.
84 corrective justice

exhibited through “an analysis of the immanent meaning of the legal

order”6 as a system of positive law. The injustice that is paramount for
Aristotle is therefore irrelevant to Kelsen.
The sections of this chapter that follow develop some of the impli-
cations of each of these conceptions of remedies. I first elaborate, in
section 2, corrective justice’s conception of the relationship between
the injustice and the remedy. Corrective justice integrates the injustice
and its rectification by construing the latter as undoing the former.
The injustice is not an occasion for a court to do what is best, all
things considered, given the present situation of the parties. Rather,
even after it has occurred, the injustice remains the decisive feature in
the parties’ relationship, because the injustice to be corrected deter-
mines the available range of remedies that can correct it.What is right-
fully the plaintiff ’s is the subject matter both of the right and of the
remedy, the right entailing a duty of non-interference, the remedy a
duty of restoration or reparation. Because what is rightfully the plain-
tiff ’s remains constant throughout, the remedy is the continuation of
the right; together they make up a single unbroken juridical sequence.
In postulating so intimate a relationship between right and remedy,
corrective justice merely draws out what the law takes for granted.
Long ago Learned Hand formulated this relationship as a truism when
he characterized a remedy as “an obligation destined to stand in place
of the plaintiff ’s right, and be, as nearly as possible, equivalent to him
for his rights.”7 My goal is to explicate the sense in which the remedy
is equivalent to, limited by, and continuous with the injured right.
I then turn to the conception of the remedy as merely conditioned
by the causative event. Kelsen’s account of this conception is part of
his elucidation of what it is for a norm to have the form of positive
law regardless of the norm’s content. This indifference to content
means that one cannot extrapolate from his account anything that
would address the lawyer’s interest in what the legal system should
look like. In this respect there is an asymmetry between Kelsen’s
account and Aristotle’s. For, as we saw in Aristotle’s treatment of the
segmented line, Aristotle thinks that the remedy has to match the
bipolar structure of the injustice that it corrects; accordingly, his
approach reproduces, although at a very high level of abstraction, the

6
Ibid., 112.
7
Learned Hand, “Restitution or Unjust Enrichment,” (1897) 11 Harv. L. Rev. 240, at 256.
 remedies 85

lawyer’s concern for the normative intelligibility of the law’s content.

For Kelsen, in contrast, the law is a posited order that can have any
content. Kelsen allows that actual legal systems do have remedies that
purport to right the wrong by requiring the defendant to repair the
damage illegally inflicted upon the plaintiff ,8 but he considers these to
exemplify a kind of logical error in which the way a person behaves (a
statement of what is) is mistakenly taken to contradict a norm (an
statement of what ought to be).9 In Kelsen’s view, from the standpoint
of positive law the illegal act means nothing more than that the sanc-
tion ought to be forthcoming. Inasmuch as the sanction is an oper-
ation of the legal order, the illegal act should therefore be understood
not as the negation of law but as its condition.
From the lawyer’s standpoint Kelsen’s observations may suggest an
unpromising framework for considering the role of remedies within a
system of private law. Sophisticated legal systems, after all, are not
indifferent to their own content. Rather, they view that content as
something that makes at least incipient sense from a suitably defined
moral point of view. Within private law, events are causative of liability
because they are thought to work some sort of injustice toward the
plaintiff. Nonetheless, torn from its context within Kelsen’s pure the-
ory, the idea of the causative event as a condition of the remedy is no
stranger to contemporary discussions of remedies. It surfaces when-
ever the remedy is seen not as normatively continuous with what
makes the causative event an injustice, but as the locus of an inde-
pendent enquiry. By breaking the normative connection that might
exist between the remedy and the causative event, such an enquiry
treats the latter as a condition but not a reason for the former.
To the extent that remedies exemplify this conception, they ignore
or go beyond the injustice that calls them into being. Such remedies
are not concerned—or at least not concerned only—with respond-
ing to whatever aspect of the parties’ interaction forms the basis for
liability. In the language of the old remoteness cases, what goes to
culpability is viewed as distinguishable from what goes to compensa-
tion (or to other remedial impositions).10 This disjunction between
the injustice of the causative event and the remedial response creates

8
Above n. 4, 109.
9
Ibid., 113.
10
The distinction between what “goes to culpability” and what “goes to compensation”
was current in the law of negligence until overruled in Overseas Tankship (UK) v. Morts
Dock & Engineering (The Wagon Mound, No. 1) [1961] A.C. 388 (P.C.).
86 corrective justice

a tension in which the injustice occasions the remedy without

grounding it. On the one hand, the causative event is seen as some
sort of injustice that requires a remedy; yet on the other, the remedy’s
operation is independent of the reason for thinking that the causative
event was an injustice to begin with. Thus, so far as the remedy is
concerned, the injustice of the causative event is both indispensable
and superfluous.
Section 3 explores the problematic nature of this conception by
reference to a conspicuous example of it, the indemnified injunction.
This kind of injunction allows the plaintiff to bring a nuisance to an
end only if the defendant is compensated for costs imposed by the
injunction. Although absent from the traditional corpus of remedies
and rarely invoked in practice, the indemnified injunction has received
extensive attention in the economic analysis of law. It is sometimes
even regarded as suggesting a new “grand theory of remedies”:11 by
breaking the “sensible convention” of “not asking those in the right
to pay when they are wronged,”12 it is supposed to contribute to a
more unified theory of remedies that combines rights, remedies, and
bargains around those rights and remedies. Whatever its virtues on
that score, the remedy operates independently of the injustice that lies
at the root of the determination of liability, as its breach of the “sens-
ible convention” shows. Indeed, the remedy obscures the very nature
of the injustice that gives birth to it. In these respects it is characteris-
tic of the conception of the remedy as merely conditioned by the
injustice.
Finally section 4 aligns the distinction between these two concep-
tions with the recent contrast made by remedies scholars between
monistic and dualistic approaches to remedies. The focus within cor-
rective justice on the normative ground of the remedy may cast light
on the tension—present in both monism and dualism—between the
centrality of the infringed right to the remedy and the introduction of
new considerations at the remedial stage. Corrective justice reconciles
this tension by situating the right within the system of rights, thereby
allowing systemic considerations to be introduced at the remedial
stage. In this way, I shall suggest, the corrective justice conception may
narrow the gap between these two approaches.

11
Saul Levmore, “Unifying Remedies: Property Rules, Liability Rules, and Startling
Rules,” (1997) 106 Yale L.J., 2149.
12
Ibid., at 2150.
 remedies 87

2. The corrective justice conception of remedies

On what grounds does corrective justice posit a continuity of right
and remedy?13 This continuity flows from the mutually complemen-
tary ways in which corrective justice conceives of the structure and
the content of the private law relationship.14 The structure consists in
the parties being correlatively situated as doer and sufferer of the same
injustice. The content consists in the plaintiff ’s having a right and the
defendant’s being under a correlative duty, so that injustice occurs on
the defendant’s breach of a duty correlative to the plaintiff ’s right. The
continuity of the remedy reflects the persistence of this structure and
content in the aftermath of the injustice.
This structure and content go to the reasons for holding a particu-
lar defendant liable to a particular plaintiff. As a matter of structure,
the normative considerations that appropriately govern finding of lia-
bility are those that implicate both parties in their relationship. As a
matter of content, these considerations presuppose that the injury is to
something to which the plaintiff has a right and with respect to which
the defendant is under a correlative duty. Being juridical manifesta-
tions of the parties’ self-determining freedom with respect to each
other, right and duty are the ingredients, and not merely the conclu-
sions, of legal argument about the terms of the parties’ interaction.
The task of private law is to work out the meaning of these rights and
duties so as to make them coherent with one another, reflective of the
idea of self-determining freedom, and applicable to the myriad con-
crete situations of human interaction.
In correcting an injustice, the remedy has the same correlative
structure as the relationship itself, because a relational injustice cannot
be corrected non-relationally. Accordingly, the remedy operates simul-
taneously against the defendant and in favor of the plaintiff. In an
award of damages, for instance, the plaintiff is entitled to receive the
very sum that the defendant is obligated to pay. If the law took money
13
The continuity of right and remedy is explicit in German jurisprudence as the Rechtsfort-
setzungsgedanke, the idea that “the injured right lives on in a claim for damages”; Walter van
Gervan et al., Common Law of Europe Casebooks: Tort Law (2000), 753. The standard German
legal textbook treats the idea of continuity as one aspect of—and therefore less comprehensive
than—the idea of compensation (the Ausgleichsgedanke), because it views consequential dam-
ages as falling outside the idea of continuity; see Karl Larenz, Lehrbuch des Schuldrechts, Band I
(Algemeiner Teil), 14th ed. (1987), 424. The implication of my argument in this chapter is that,
from the theoretical perspective, continuity is the more fundamental idea.
14
Above, chapter 1.
88 corrective justice

from the defendant without giving it to the plaintiff, the injustice suf-
fered by the plaintiff would remain uncorrected. Similarly, if the law
gave money to the plaintiff without taking it from the defendant, the
injustice done by the defendant would remain uncorrected. And even
if the law took money from the defendant and gave an equivalent
amount of money to the plaintiff in separate operations (say, by requir-
ing payment into one government fund and out of another), the
injustice as something done by the defendant to the plaintiff—and
therefore as being of relational significance between them—would
still remain uncorrected. Structurally, the remedy is the mirror image
of the injustice. Both feature the same movement from one pole of
the relationship to the other, so that, to the extent possible, the rela-
tionship ends up as free of injustice as it was at the beginning.
The correction maintains not only the structure but also the con-
tent—the right and the correlative duty—of the parties’ relationship.
What is correctively just about a private law relationship is the absence
of breaches of any duty correlative to another’s right. Conversely,
injustice lies in an inconsistency with the plaintiff ’s right that is imput-
able to the defendant. The point of the remedy is to eliminate this
inconsistency. In this progression from justice to injustice and back
again, the same right (and, of course, the same correlative duty) is the
focus of the law’s attention. The right survives the injustice and con-
tinues into the remedy, which is nothing other than the judicially
crystallized post-injustice shape of the right.
Now one might think that identifying the remedy with the pre-
injustice right (and its correlative duty) overstates the closeness of the
connection between them. Suppose that the defendant has tortiously
destroyed an object belonging to the plaintiff and now has to pay the
plaintiff a sum equal to the object’s value. Before the destruction the
defendant was under a duty to abstain from exposing the object to an
unreasonable risk. After the destruction the defendant cannot be under
this duty, because the object no longer exists.The action now required
of the defendant is not abstention from creating an unreasonable risk,
but transfer to the plaintiff of a certain sum of money. A duty man-
dates a specific action, and if the specific actions mandated are differ-
ent, so are the duties.15

15
“[O]bligations . . . are individuated according to the actions that they make obligatory,”
J. Gardner, “What is Tort Law For? Part 1: The Place of Corrective Justice,” (2011) 30 Law
and Philosophy 1, 35.
 remedies 89

A similar argument can be made on the rights side. A right gives its
holder the freedom to act within its bounds. Yet the actions permitted
before the injustice may differ from those permitted after the injustice.
For example, my right to bodily integrity cannot be alienated, but it
may be possible for me, within restrictions set out by the positive law, to
assign the damages claim that arises from the injury.16 The fact that after
the injustice one has the freedom to do actions unavailable previously
indicates (so the argument would go) that the different freedoms reflect
different rights rather than the continuation of the same right.
That the variety of actions prohibited or permitted attests to a vari-
ety of duties and rights is an appealing but misleading notion. It is not
the case that if the specific actions mandated by the law are different,
so are the legal duties. Different actions can be required by a single
duty and a single action can be required by different duties. An ex-
ample of the latter is that the same specific action may be required
both contractually and delictually. As for the former, suppose that the
defendant, being under a duty of care as a bailee with respect to an
object belonging to the plaintiff , was obliged both to keep his car
locked as he transported the object and to water the object regularly.
The law would regard these two different actions as different ways of
fulfilling the same legal duty, not as the fulfillment of two different
duties.The fact that there are innumerable ways in which a duty could
be breached does not mean that each possible breach is the breach of
a different duty.17
A legal duty takes its character from the legal category that informs
it, not from the specific action that it prohibits or requires. The same
action required as a matter of both contract law and tort law is gov-
erned concurrently by two duties, one for each possible ground of
liability. In my example of the bailment, the legal duty is that of a
bailee, not that of a person who waters an object or transports it in a
locked car.
Considered as a theoretical issue, the relation between right and
remedy engages a still higher level of generality. Theory is concerned

16
I owe this example to Lionel Smith.
17
The distinction between a duty and a required specific action tracks Kant’s obscure
distinction between an obligation (“the necessity of a free action under a categorical impera-
tive of reason”) and a duty “that action to which someone is bound. It is therefore the matter
of the obligation.” Kant adds that “there can be one and the same duty (as to the action,
although we can be bound to it in different ways.” Immanuel Kant, The Metaphysics of Morals,
tr. Mary Gregor (1996), [6:222].
90 corrective justice

not with particular grounds of liability and their respective remedies,

but with the nature of liability as such and the corresponding concep-
tion of a remedy. As noted above, under corrective justice the injustice
that gives rise to liability is an inconsistency with the plaintiff ’s right
that is imputable to the defendant. At its most general, having a right
in private law means that the right-holder is normatively so connect-
ed to the object of the right that another person is under a duty not
to interfere with that object.18 The legal system lays down the grounds
for acquiring and holding rights of various sorts—offer and accept-
ance for contract, animus donandi and factum donandi for gift, and so on.
As long as these grounds obtain, the relationship of right and duty
continues regardless of what the defendant has done to the object of
the right. Only actions consistent with the holder’s right can termin-
ate this normative connection, as when property is alienated or a con-
tract is discharged by performance. Conversely, the right (and the duty
correlative to it) always survives an injustice, which by definition is an
inconsistency with the right.
Accordingly, the defendant who, in breach of her duty, destroys an
object belonging to the plaintiff does not thereby destroy the plain-
tiff ’s right to the object. The plaintiff remains linked to the defendant
through a right that pertains to the object as an undamaged thing.
Although the defendant’s wrong has modified the physical condition
of the object embodying the plaintiff ’s right, the right remains intact
as the normative marker of the relationship between them with respect
to that object. Even if the object no longer exists as a physical entity,
the parties continue to be related to each other through the object’s
normative connection to the plaintiff and the consequent duty on the
defendant to act in conformity with that connection. Instead of being
embodied in the object itself, the right and its correlative duty with
respect to the object now take the form of an entitlement and have
the defendant furnish the plaintiff with its value.
The survival of the right means that its correlative duty also sur-
vives. The defendant’s breach of duty did not of course bring to an
end the duty with respect to the plaintiff ’s right, for, if it did, the
duty—absurdly—would have been discharged by its breach. To be
sure, the specific action required of the defendant has been trans-
formed by the defendant’s tort. Just as the plaintiff ’s right is no longer

18
“That is rightfully mine (meum iuris) with which I am so connected that another’s use
of it without my consent would wrong me.” Ibid., at [6:245].
 remedies 91

embodied in the specific object, which has been destroyed, but in an

entitlement to receive the object’s equivalent from the defendant, so
the defendant’s duty is no longer to abstain from its destruction, which
has already taken place, but to provide the plaintiff with the object’s
equivalent. The specific action that the duty requires is different, but
the defendant is not under a different duty. This is because, from a
juridical point of view, what determines the nature of the duty is not
the specific action that the duty requires but the right to which the
duty is correlative. And what determines the right is the appropriate
normative connection between the object of the right and the person
holding it. So long as that connection persists, the right and correla-
tive duty with respect to the object remain.
Thus, the right and its correlative duty continue to exist with dif-
ferent specific content before and after the injustice. Underlying the
succession of specific characteristics of the right and its correlative
duty is the relationship that the parties have through the plaintiff ’s
connection with the object of the right. That relationship remains
identical throughout the metamorphosis that the defendant’s injustice
has wrought in the object of the right. To put it in familiar philosoph-
ical terms, the diachronic identity of the right is merely a juridical
exemplification of the category of substance as that which persists
through change: during the legal relationship the existence of the
right remains constant, but the way in which the right exists changes.19
Just as a person has different characteristics at different times of life
while yet remaining the same person, so a right and its correlative
duty have different characteristics at different points in their existence
while yet remaining the same right and duty. Juridically, the parties
step twice into, or rather stand continuously in, the same river.
Blackstone summed up the relation between right and remedy by
stating that remedies “redress the party injured, by either restoring to
him his right, if possible, or by giving him an equivalent.”20 Black-
stone’s formulation is a paradigmatic expression of corrective justice.
It implies three theses. The first is the thesis of identity, that the plain-

19
“In all appearances, the permanent is the object itself, that is, substance as phenomenon;
everything, on the other hand, which changes or can change belongs only to the way in
which substance or substances exist, and therefore to their determinations. I find that in all
ages, not only philosophers, but even the common understanding, have recognized this per-
manence as a substratum of all appearance, and always assume it to be indubitable.” Kant,
Critique of Pure Reason, tr. Norman Kemp Smith (1929), A184, B227.
20
Blackstone, IV Commentaries, 9.
92 corrective justice

tiff ’s injured right and the right restored by the defendant are the
same right or its equivalent. One cannot regard a right as being
restored if it is other than the one that the defendant wronged. The
second is the thesis of limitation, that the remedy restores only the
plaintiff ’s right and does not give the plaintiff more than that right (or
its equivalent). Thus, the reason for creating liability also limits it.21
The third is the thesis of continuity, that the plaintiff ’s right survives
the injury intact and continues to be the normative marker of the
parties’ relationship. Because the right continues to exist, plaintiffs can
justly apply to courts for the restoration of what remains rightfully
theirs.22
These three theses are interrelated. Rights could not be enjoyed as
domains of freedom unless the law secured them against wrongs by
requiring wrongdoers to restore what they have injured (the identity
thesis). However, because the relationship between the parties is one
of equal freedom, the plaintiff ’s freedom does not entitle the court to
coerce the defendant into providing the plaintiff with a windfall over
and above the restored right, for that, in turn, would be inconsistent
with the defendant’s freedom (the limitation thesis).23 With the ideas
of injury and restoration in place, one might wonder how the tem-
poral gap between them is normatively bridged. For one might suppose

21
Compare Warren A. Seavey, “Mr. Justice Cardozo and the Law of Torts,” (1939) 39
Colum. L. Rev. 20, 34.
22
Blackstone’s terminology of restoring the plaintiff ’s right is not entirely felicitous, as it
participates in the ambiguity of right as both something that a plaintiff has and a normative
status that attaches to something that the plaintiff has. One should not think that the very
description of the remedy as the restoration of a right shows that the plaintiff did not have
what the remedy restores. Kant draws attention to this terminological imprecision in his
discussion of external right. In Kantian terms, ownership involves possessing an object intel-
lectually rather than empirically, because the essence of ownership is that it persists even
when the owner is not in physical possession of the thing owned. For this reason, he writes,
“it is not appropriate to speak of possessing a right to this or that object but rather of pos-
sessing it merely rightfully; for a right is already an intellectual possessing of an object and it
would make no sense to speak of possessing a possess[ing].” Kant, above n. 17, [6:249] (transla-
tion slightly modified). In the same way here, the wrong is a deprivation of what is rightfully
the plaintiff ’s and the remedy restores to the plaintiff what is rightfully hers. For an illuminat-
ing treatment of remedies from a Kantian perspective, see Arthur Ripstein, “As If It Had
Never Happened,” (2007) 48 Wm. and Mary L. Rev. 1957.
23
As Kant observed in his comment about tort law:
I cannot acquire a right against another through a deed of his that is contrary to right (facto iniusto
alterius); for even if he has wronged me and I have a right to demand compensation from him, by
this I will still only preserve what is mine undiminished but will not acquire more than what I
previously had.
Kant, above n. 17, [6:271].
 remedies 93

that the occurrence of the injury puts an end to the plaintiff ’s right,
leaving the plaintiff without a basis for claiming what he no longer
has. Perhaps all that the plaintiff can expect is an apology for the mis-
fortune that the defendant caused.24 The continuity thesis holds, in
reply, that even after the injury the plaintiff continues to have the
right to what was wrongly injured. From the normative point of view,
no gap in the plaintiff ’s right-holding exists between the injury and
the remedy.
One might object that this account of corrective justice implausibly
effaces the significance of the wrong, because it entails that one is dis-
charging one’s duty to refrain from wrongdoing ex ante by compensat-
ing the victim ex post.25 This is not the case.The continuity of right and
remedy means that the same relationship of right and duty continues
through a sequence of stages that, on the duty side, require different
specific actions. A sequence is not a smorgasbord from which the
defendant can mix and match.What counts as the discharge of the duty
in any given stage is determined by the actions that the duty calls for at
that stage, not at a previous or subsequent one. Accordingly, the defend-
ant cannot satisfy the duty as it existed at one stage by performing the
action called for at a subsequent stage. As a juridical instantiation of the
category of substance,26 the right and its correlative duty persist through
change; they do not remain unchanged. Just as my being a more mature
version of the person I was as a child does not now require me to enroll
in kindergarten, so the defendant who has committed an injustice can
no longer satisfy his duty in its original form. The injustice committed
earlier remains an injustice. The remedy vindicates the plaintiff ’s right
by restoring what is rightfully his, thereby affirming rather than denying
that the injustice occurred. Because the defendant has not complied
with a duty owed to the plaintiff, the duty continues to exist in a new
form that requires the performance appropriate to this new stage of the
parties’ relationship.27

24
Stephen R. Perry, Loss, Agency, and Responsibility for Outcomes: Three Conceptions of Cor-
rective Justice, in Tort Theory, ed. Ken Cooper-Stephenson and Elaine Gibson (1993), 24.
25
Benjamin C. Zipursky, “Rights, Wrongs, and Recourse in the Law of Torts,” (1998) 51
Vand. L. Rev. 1, 74.
26
Above n. 19.
27
Nor is the corrective justice account affected by the Austinian distinction between a
primary duty and the secondary duty that arises out of a violation of a primary duty; John
Austin, Lectures on Jurisprudence, 5th ed. (1885), 764. That distinction merely sets out different
stages in the parties’ relationship; it does not address the nature of the normative connection
between them.
94 corrective justice

By regarding the remedy as a continuation of the injured right,

corrective justice provides a unifying framework for understand-
ing and assessing the various remedies that the law makes available.
Corrective justice is a normative regime of rights and their correla-
tive duties. In awarding a remedy the law aims to remove the incon-
sistency with the plaintiff ’s rights by having the defendant restore
what is rightfully the plaintiff ’s. The diversity of the remedies reflects
the different ways of impairing and restoring what is rightfully the
plaintiff ’s.
Restoration of the plaintiff ’s right can take two forms: the qualita-
tive and the quantitative. The qualitative form restores to the plaintiff
the very thing that is the subject matter of the right, thereby allowing
the plaintiff to have and enjoy “its specific qualitative character.”28 In
such cases the law gives specific relief, such as specific delivery of a
unique or unusual chattel, specific performance of a contractual obli-
gation, or an injunction against a private nuisance or trespass. The
quantitative form restores to the plaintiff, through an award of dam-
ages, the monetary equivalent of the injury. One of the tasks of the
law of remedies, of course, is to work out which of these forms of
restoring the plaintiff ’s right is available in what circumstances—an
issue that different jurisdictions handle in different ways. Nonetheless,
in accordance with corrective justice, both forms of restoration exem-
plify the continuity of right and remedy.
Now, one should not think that the availability of injunctive relief is
inconsistent with corrective justice on the grounds that an injunction
against future wrongdoing applies when there is yet no wrong to cor-
rect. Under corrective justice, the private law relationship is correla-
tively structured by the plaintiff ’s right and the defendant’s duty. The
remedy instantiates that structure by vindicating the plaintiff ’s right
against the defendant’s breach of the correlative duty. What matters is
not the temporal relation between the injustice and the remedy, but
the structure of the injustice and the consequent structure of the rem-
edy. For instance, if (as I have argued elsewhere)29 the norms against
nuisances instantiate corrective justice, then so do the injunctions that
prevent nuisances.Thus, corrective justice operates not only by requir-
ing the defendant to repair a wrong once it has occurred, but also by

28
Hegel, Philosophy of Right, tr. T. M. Knox (1952), s. 98R.
29
Ernest Weinrib, The Idea of Private Law (1995), 190–96.
 remedies 95

granting the plaintiff an injunction that prevents the defendant from

extending the wrong into the future.30
The continuity of right and remedy also holds for the various kinds
of damages that figure in the quantitative form of restoration: substi-
tutive damages, nominal damages, consequential damages, gain-based
damages, and aggravated damages. Let us briefly consider each of these
in turn.
First, the right-holder is entitled to the physical integrity of the
thing that forms the subject matter of the right.31 Correspondingly,
others are under a duty not to wrongfully interfere with that physical
integrity. When such interference occurs, the right-holder’s entitle-
ment to an intact thing continues as against the wrongdoer. The
wrongdoer then has a correlative duty to transfer the sum of money
that leaves the right-holder with the equivalent of the thing’s value in
its intact state. Such damages have been termed “substitutive,” in that
they are awarded to the plaintiff as a substitute for the right that the
defendant infringed.32
Second, when a right is infringed without impairing the physical
condition of the object, a court awards nominal damages. The avail-
ability of nominal damages is the remedial affirmation that private law
vindicates rights and does not merely repair losses. Just as no liability
follows when a loss is not the wrongful infringement of a right, so
conversely can a defendant be held liable, and required to pay nominal
damages, for a wrong to the plaintiff ’s right that does not occasion a
loss. The obligation to pay nominal damages is the continuation of the
defendant’s duty not to interfere with the plaintiff ’s right even when
no loss results from such interference.
Third, the right-holder’s entitlement to have the thing physically
intact carries with it an entitlement to use the thing in its intact con-
dition for his or her purposes. Accordingly, a wrongful interference
with the thing’s physical integrity may wrongfully interfere with the
use, actual or prospective, to which the right-holder is putting or is
likely to put the intact thing. The entitlement to use the intact thing
imports a correlative duty not to wrongfully interfere with such use.
This duty finds its remedial continuation in what the law terms

30
Ibid., 144.
31
For purposes of exposition I assume a wrong with respect to a corporeal object. The
argument would not essentially change for non-corporeal objects, though it would be refor-
mulated to accord with the non-corporeal nature of the subject matter of the right.
32
Robert Stevens, Torts and Rights (2007), 60
96 corrective justice

“consequential” damages—that is, the monetary sum equivalent to

the worth of the use of which the defendant wrongly deprived the
right-holder.
Fourth, the right-holder has an exclusive entitlement to deal with
the thing owned, and can realize the thing’s value by charging for its
use or by selling it. The gain from the use or the sale is as much the
right-holder’s as is the thing itself. Accordingly, the right-holder can
claim restitution of such a gain from a wrongdoer who made it
through a use or a sale that was unauthorized. This award of gain-
based damages (and its historical antecedent in “waiver of tort”) is the
continuation of the right-holder’s entitlement to the thing’s value.33
Fifth, the common law recognizes that a wrongdoer may not only
have injured the object of the right, but may also have done this so
high-handedly as to injure the dignity of the right-holder. To com-
pensate for such injuries to dignity the law awards aggravated dam-
ages, for a court may “take into account the motives and conduct of
the defendant when they aggravate the injury done to the plaintiff.”34
This form of damages reflects the connection between the object of
the right and the dignity that the law ascribes to the holder of the
right. As a system of rights, the law presupposes a distinction between
persons, (entities imbued with the dignity that attends the capacity for
rights) and things (entities devoid of that dignity). The dignity that
comes from the right-holder’s connection to the object of the right is
as much within the entitlement of the right-holder as the object of
the right itself. Accordingly, the law awards additional damages, which
it regards as compensatory, for a wrong committed in a way that
imparts injury to the right-holder’s dignity over and above the injury
done to the object of the right itself. Such damages are the continu-
ation of the dignity inherent in being the holder of a right.35
Corrective justice also provides a standpoint for criticizing the con-
troversial practice of awarding punitive damages. This critical stand-
point is the consequence of the theoretical aspirations of the corrective
justice approach. Corrective justice explicates the internal structure
and presuppositions of the private law relationship as found in sophis-
ticated legal systems, in order to present at a high level of abstraction

33
Below, chapter 4.
34
Rookes v. Barnard [1964] A.C. 1129, 1221 (H.L., per Lord Devlin).
35
On aggravated damages as reparation for injury to dignity, see Allan Beever, “The
Structure of Aggravated and Exemplary Damages,” (2003) 23 Oxford J. Legal Stud. 87; John
Murphy, “The Nature and Domain of Aggravated Damages,” (2010) 69 Cambridge L.J. 353.
 remedies 97

what it means for private law to be fair and coherent on its own
terms. Because sophisticated systems of private law strive—not always
with success, of course—to be fair and coherent, they are composed
of norms that might exhibit the specific meaning of corrective justice
for a particular legal system or legal tradition. Conversely, corrective
justice provides an internal standpoint for the criticism of norms that
are not consonant with a liability regime’s own aspiration to fairness
and coherence. Accordingly, if the corrective justice arguments against
the punitive damages are sound, then corrective justice has fulfilled its
theoretical function of providing the internal standpoint for identify-
ing unfair or incoherent doctrine.
Punitive damages are inconsistent with corrective justice for rea-
sons both of structure and of content. So far as structure is concerned,
corrective justice requires that the normative considerations applic-
able to the relationship between defendant and plaintiff reflect the
parties’ correlative standing as sufferer of and doer of the same injus-
tice. Accordingly, it excludes considerations that refer to one of the
parties without encompassing the correlative situation of the other.
The standard justifications for punitive damages—deterrence and ret-
ribution—are one-sided considerations that focus not relationally on
the parties as doer and sufferer of the same injustice, but unilaterally
on the defendant (and anyone else who might be similarly situated) as
doer. The place of such considerations is not private law but criminal
law, because criminal law is concerned not with whether the accused
has injured someone’s particular right, but with whether the accused
has acted inconsistently with the existence of a regime of rights in
general.36 In effect, punitive damages function as a defendant-financed
reward for acting as a private prosecutor while subjecting the defend-
ant to punishment without the protections of the criminal law.
So far as content is concerned, punitive damages are inconsistent
with the role of rights in corrective justice. Punitive damages do not
restore to plaintiffs what is rightfully theirs, but instead give them a
windfall. Punitive damages based on deterrence and retribution thus
violate what I earlier termed the limitation thesis, that the remedy
should only restore the plaintiff ’s right and not give the plaintiff more
than that right (or its equivalent).
Thus the corrective justice account, through the robust role that it
assigns to rights and their correlative duties, provides a unifying frame-

36
Below, chapter 5, section 4.
98 corrective justice

work for understanding both the relationship between right and rem-
edy and the range of different remedies that the law makes available.
All appropriate remedies reflect the reason for liability, that the injus-
tice imputed to the defendant is inconsistent with the plaintiff ’s right.
The different kinds of damages reflect the various kinds of entitle-
ment that a right gives, including an entitlement to the intactness of
the object of the right, to its use and value as an intact object, to its
inviolability even in the absence of loss, and to the dignity that attach-
es to the right-holder. And the distinction between monetary damages
and specific remedies such as injunctions reflects the different ways in
which the injured right can be restored.

3. The causative event as a condition of the remedy

I now turn to the alternate conception of remedies, in which the
causative event functions as a condition and not as a reason. As was
mentioned at the outset, I want to focus on a conspicuous if rather
unusual example: the indemnified injunction. Under an indemnified
injunction the plaintiff is entitled to an injunction provided that the
defendant is compensated for the damage that the injunction causes.
This remedy was not known to exist until 1972. In that year it was
coincidentally employed in the nuisance case of Spur Industries v. Del
E. Webb Development Co.37 and suggested in a celebrated article by
Guido Calabresi and Douglas Melamed that has revolutionized aca-
demic thinking about remedies in the United States.38
Because compensating the defendant is manifestly at odds with the
supposedly wrongful behavior that occasions the remedy, the indem-
nified injunction is an extreme example of this alternate conception
of remedies. When an indemnified injunction is awarded, the causa-
tive event has to be understood not as the reason that grounds the
remedy as a matter of justice but as the occasion that triggers the
operation of a normatively independent remedial policy. This gap
between response and the normative ground of the causative event
means that the remedy obscures the nature of the injustice being rem-
edied. Indeed, from the standpoint of corrective justice, such a remedy

37
Spur Industries v. Del E. Webb Development Co., 494 P. 2d 700 (Arizona S.C., 1972)
38
Guido Calabresi and A. Douglas Melamed, “Property Rules, Liability Rules, and In-
alienability: One View of the Cathedral,” (1972) 85 Harv. L. Rev. 1089.
 remedies 99

works a new injustice to one or the other of the parties: it requires the
defendant to do either too much (if the defendant’s behavior was not
really a wrong) or too little (if it was). In any event, the extreme nature
of this instance makes it an apt illustration of what this alternate con-
ception entails.

A. Spur v. Del E.Webb

Consider first the nuisance case.39 The defendant was the operator of a
cattle feedlot built in what had long been an agricultural district. The
plaintiff was a real estate developer who had purchased land in the
neighborhood of the feedlot in order to develop an urban area. As the
development grew, it came closer and closer to the cattle feedlot, until
the feedlot’s smells caused the developer to encounter sales resistance
from prospective purchasers and persistent complaints from past pur-
chasers.The developer sued in nuisance.The court held that the devel-
oper was entitled to an injunction. It also held, however, that having
brought into the previously agricultural area the population that
makes the granting of an injunction necessary, the plaintiff should also
indemnify the defendant for the damages it would sustain by having
to move or cease operations under the injunction.
The striking feature of this remedy is that it makes it hard to dis-
cern the injustice of, or even to identify, the causative event. Indeed,
because the remedy has two parts, the injunction for which the
defendant is liable and the indemnification that the plaintiff provides
in return for the injunction, it is not even easy to single out the party
responsible for the causative event. Nor is it easy to identify the victim
of the wrong, especially since the court took into account the inter-
ests not only of the parties but also of the residents in the develop-
ment, who purchased their properties from the developer. On the facts
of this case four possibilities suggest themselves: (1) the feedlot oper-
ator wronged the developer; (2) the feedlot operator wronged the
residents; (3) the developer wronged the feedlot operator; (4) the
developer wronged the residents. None of these possibilities, however,
is completely satisfactory.
The first possibility, matching the roles of defendant and plaintiff
in the case, is that the feedlot operator committed a wrong against
the developer. On this possibility, which accords with the classic

39
Above n. 37.
100 corrective justice

understanding of nuisance law,40 the presence of the cattle feedlot

prior to the activity of the developer does not exonerate the operator
of the feedlot from what was otherwise a nuisance. The defendant
cannot restrict the plaintiff from developing its land, and once the
land is developed, the plaintiff ’s rights in the use of that land are pro-
tected by the law of nuisance. This protection includes the availability
of an injunction to bring to an end the defendant’s interference with
the plaintiff ’s rights. But then it is hard to account for requiring the
plaintiff to indemnify the defendant for the injunction. It seems odd
that the victim of a wrong should have to indemnify the wrongdoer
for exercising the remedy that the law grants to bring about the ces-
sation of that very wrong.41
The second possibility, that it was the residents rather than the
developer who were the victims of a wrong done by the feedlot oper-
ator, looks to the fact that the plaintiff can secure freedom from the
smells only by purchasing that freedom. One might infer from this
that the defendant did not really commit a wrong against the plaintiff
in carrying on with its feedlot operations despite the harm that the
smells inflicted on the approaching development. The court adopted
this approach. It referred to cases of “coming to the nuisance” under
which a use that is prior in time does not constitute a nuisance at all
against parties who subsequently enter the neighborhood with incon-
sistent uses.42 Had only the developer been harmed and not also the
purchasers of the houses, it would have dismissed the case on this
ground.43 The plaintiff ’s entitlement to relief was based on the damage
“to the people who have been encouraged to purchase homes” in the
development.44 Although the remedy is placed in the hands of the devel-
oper, the wrong was done to the residents. On this reasoning the court
granted an injunction to a party that was not the victim of a wrong on
the expectation that the remedy would be pursued for the benefit of
the real victims. This means that, contrary to the ruling idea of tort
law,45 the plaintiff was awarded a remedy on the basis of a wrong done
to someone else.
40
Sturges v. Bridgman [1879] 11 Ch. D. 852 (C.A).
41
This is the criticism of the court’s treatment by Holohan J., dissenting in the subsequent
case Spur Feeding Company v. Superior Court of Maricopa County, 505 P 2d 1377 (Arizona
S.C., 1973).
42
Spur v. Webb, above n. 37, at 707.
43
Ibid., at 707.
44
Ibid., at 708.
45
Palsgraf v. Long Island Railroad Co., 162 N.E. 99 (N.Y. C. A., 1928).
 remedies 101

The third possibility is that the developer, although entitled to an

injunction, actually committed some sort of wrong against the feedlot
operator. The court remarks that in bringing the development into
proximity with the cattle feedlot the plaintiff is not “blameless.”46 The
defendant’s business was a lawful one, and the plaintiff was the cause
of foreseeable detriment to it. The indemnification is appropriate, the
court ruled, in a case in which the developer “has, with foreseeability,
brought into a previously agricultural . . . area the population which
makes necessary the granting of an injunction against a lawful business
and for which the business has no adequate relief.”47 It is as if the
court was postulating a notional wrong that consists in populating an
area in a way that would foreseeably lead to an injunction against a
pre-existingly lawful business. One problem with this is the implaus-
ibility of thinking that building on one’s own property and then sell-
ing what one has built constitutes a wrong. Another problem is that
an inherent element in such a wrong would be the entitlement of the
supposed wrongdoer to an injunction, the securing of which would
simultaneously be the assertion of a right and the commission of a
wrong against the same person.
A fourth possibility is that the developer committed a wrong not
against the feedlot operator but against the residents who were
exposed to the feedlot’s smells. As noted in connection with the sec-
ond possibility, the court regarded the residents as the true victims of
the nuisance, and it awarded the developer the injunction only
because of the harm that they suffered. Moreover, as noted in con-
nection with the third possibility, the court also regarded the devel-
oper rather than the cattle feedlot operator as the true cause of the
feedlot’s having to move. Putting these two points together, one might
surmise that the developer was primarily responsible for the nuisance
suffered by the residents.
That the court thought so is indicated by subsequent litigation. At
the time of the case brought by the developer another action in
nuisance brought by several hundred residents was also pending
against the feedlot operator. The feedlot operator filed a third-party
complaint against the developer, so as to be able to obtain indemnity
from the developer for the damages for which it might be liable to
the residents. The court dismissed the developer’s contention that

46
Spur v. Webb, above n. 37, at 707.
47
Ibid.
102 corrective justice

the claim between the developer and the feedlot operator, having
been definitively settled by the litigation between them, was res
judicata. Consequently, the court concluded, “the [feedlot operator]
is entitled to have litigated the conduct of [the developer] as to each
of the [residents] and to have the question of indemnity litigated as
to each of them.”48 This decision raised the prospect that the devel-
oper would have to compensate the feedlot operator for whatever
remedy the residents might be awarded. This too seems an odd pros-
pect. The entitlement that the residents had against the developer
concerning the smells to which they would be exposed in the new
homes they had purchased was a matter of contract.49 If they had no
such entitlement through their contracts, it seems strange that they
could secure one by the indirect route of suing the feedlot owner in
nuisance for harms for which the developer would ultimately have
to pay.
From the consideration of these four possibilities emerges the dif-
ficulty of specifying the normative ground of the causative event to
which the indemnified injunction is the response. The wrong is not
the reason for the remedy, because no viable wrong can be established
that would match so curiously structured a remedy. The court states
the circumstances in which this remedy is available—the developer’s
bringing in of a population that makes the injunction foreseeably
necessary—but these circumstances are merely the conditions that
trigger the remedy without providing a normatively coherent reason
for it.
Because a sophisticated system of private law aspires to make an
intelligible connection between the causative event and the response,
the following difficulty arises. On the one hand, the production of
smells by the feedlot operator is considered to be a wrong of some
sort because otherwise a court would have no occasion to formulate
any remedy at all. On the other hand, once the remedy is formulated,
no matching wrong can be located. The remedy is occasioned by a
wrong about whose nature the remedy provides no testimony. Like a
door swinging without the hinges that attach it to anything, the rem-
edy is both conditioned by and normatively independent of the events
that give rise to it.

48
Spur Feeding Company, above n. 41, at 1379.
49
The developer apparently advertised that purchasers would be able to enjoy the out-
door living that the properties afforded; Spur v. Webb, above n. 37, at 705.
 remedies 103

B. The Calabresi–Melamed framework

That the award of an indemnified injunction obscures the injustice of
the causative event is also evident from the analysis of Calabresi and
Melamed.50 In suggesting a unified remedial framework that encom-
passes both property law and tort law, they disclose the economic basis
for a new way, corresponding to Spur v. Webb, of addressing the legal
issues posed by pollution. Their analysis, however, is indifferent to the
normative character of the entitlements involved.
Calabresi and Melamed argue that the indemnified injunction is
necessary to complete the set of remedial possibilities applicable to
nuisance claims. Take the conflict between a polluter who is claiming
an entitlement to pollute and the victim who is claiming an entitle-
ment to be free from pollution. Under the traditional remedial frame-
work the victim’s entitlement can be protected in two ways. An
injunction protects the entitlement by what Calabresi and Melamed
call a “property rule,” so that in order to continue polluting the pol-
luter must purchase the entitlement from the victim at a value to
which the victim agrees. An award of damages protects the victim’s
entitlement by what they call a “liability rule,” so that the polluter pays
the victim the value of the entitlement as objectively determined by a
court.51 However, if the victim’s claim cannot successfully be made,
the entitlement to pollute remains in the hands of the polluter. The
effect of this is that the polluter’s entitlement is protected by a prop-
erty rule, because then the victim will have to purchase freedom from
pollution by buying out the polluter at the polluter’s price. Thus,
under the traditional framework, the victim’s entitlement is protected
by both a property rule and a liability rule, whereas the polluter’s enti-
tlement is protected by a property rule only. Stating the entitlement-
protecting rules in this way shows, they argue, that something is

50
Above n. 38.
51
Richard Epstein gives a lucid definition of these rules and the difference between them:
A property right gives the individual the right to keep the entitlement unless and until he choos-
es to part with it voluntarily. Property rights are, in this sense, made absolute because the owner-
ship of some asset confers sole and exclusive power on a given individual to determine whether
to retain or part with an asset on whatever terms he sees fit. In contrast, a liability rule denies the
holder of the asset the power to exclude others or, indeed, to keep the asset for himself. Rather,
under the standard definition he is helpless to resist the efforts by some other individual to take the
thing on payment of its fair value, as objectively determined by some neutral party.
Richard Epstein, “A Clear View of the Cathedral: The Dominance of Property Rules,”
(1997) 106 Yale L.J. 2091.
104 corrective justice

missing: a liability rule to protect the polluter’s entitlement.52 Under

such a rule the victim could stop the polluter from polluting but
would have to compensate the polluter on the basis of an objectively
determined value. The contemporaneously decided Spur case illus-
trates this. There the developer could force the cattle feedlot operator
to cease producing the smells only by indemnifying the feedlot oper-
ator for the disruption of its activities. In the terminology of Calabresi–
Melamed, Spur protected the feedlot operator’s entitlement with a
liability rule.
The purpose of this additional rule—Rule 4 as it has come to be
known—is to render the parties’ positions completely symmetrical.
The defect of the traditional framework is that the victim’s entitle-
ment, but not the polluter’s, is protected by a liability rule. Under the
Calabresi–Melamed framework both the property and the liability
rules that apply to the victim have counterparts that apply to the pol-
luter.53
The achievement of such symmetry is integral to the economic
analysis that informs the Calabresi–Melamed framework. Economic
analysis illuminates how the law assists the movement of resources to
those who value them most. In the absence of transactions costs this
movement is achieved through bargaining, which determines the fate
of the resource regardless of the party to whom the law initially
assigned it.54 Because the parties bargain against the background of
their respective opportunity costs rather than of the one’s rights against
the other, each is symmetrically the potential cause of the other’s dep-
rivation. Legal determinations matter only when transactions costs
disrupt the possibility of bargaining. Inasmuch as the point of such

52
Above n. 38, at 1116.
53
Even if one is attracted, as Calabresi and Melamed are, to the elegance of a remedial
framework in which every possibility on the victim’s side had a counterpart on the polluter’s
side, it does not follow that a liability rule is missing on the polluter’s side. Instead of adding
a liability rule that protects the polluter’s entitlement, one could also subtract the liability rule
that protects the victim’s entitlement, leaving the victim with the protection of a property
rule only. Something like this is the position in the Commonwealth jurisdictions, where the
plaintiff in a successful nuisance suit is entitled to an injunction except in very limited cir-
cumstances; Shelfer v. City of London Electric Lighting Co. [1895] 1 Ch. 287 (C.A.). For a
long time this was also the position in U.S. jurisdictions (see Whalen v. Union Bag and Paper
Co., 101 N.E. 805 (N.Y.C.A., 1916), until the change effected by Boomer v. Atlantic Cement
Co., 257 N.E. 2d 870 (N.Y.C.A., 1970). This would prevent the oddity of an “entitlement”
that others could expropriate at will provided they gave the owner its value; see the strong
comments of Idington J. in Canada Paper Co. v. Brown, 63 S.C.R. 248 (1922).
54
Ronald Coase, “The Problem of Social Cost,” (1960) 3 J. Law & Economics 1.
 remedies 105

determinations is to move the resource to its most valued use—that is,

to replicate the result of bargaining in the absence of a process of bar-
gaining—the law has to have available a set of remedies that situates
the parties as symmetrically as they would have been as bargainers.
Rule 4 accomplishes this symmetry for liability rules, thereby com-
plementing the symmetry of property rules already present in the pre-
vailing framework. With Rule 4 in place the polluter’s entitlement is
liable to be bought out by the victim, just as the prevailing framework
allows the victim’s entitlement to be bought out by the polluter. The
result is that the remedy, be it a property rule or a liability rule, can be
applied against the one party or the other, depending on the nature of
the transactions costs.
The symmetry thus achieved deals with effects. It focuses on the
point at which the court determines whether the defendant ceases to
pollute, pays damages, is absolved from liability, or is indemnified for
having to cease polluting. The premise is that whatever kind of effect
the law can impose on the polluter for the protection of the victim’s
entitlement, it should also be able to impose on the victim for the
protection of the polluter’s. If the law protects the victim’s entitlement
with a property rule, it should also protect the polluter’s entitlement
with a property rule, as in fact it does under the traditional framework.
If the law protects the victim’s entitlement with a liability rule, it
should also protect the polluter’s entitlement with a liability rule, as it
would be able to do under Rule 4.
The oddity of such symmetry of effect is that it operates despite the
asymmetry in the normative character of the parties’ entitlements. If
the victim triumphs, the entitlement protected by the award of an
injunction or of damages is, in Hohfeldian terms, a claim right.55 In
the law of private nuisance this is a right that the possessor of land has
to be free from “an inconvenience materially interfering with the
ordinary comfort physically of human existence.”56 The recognition of
this right in the victim implies a correlative duty of non-interference
for the polluter. The injunction or damages are then supposed to
restore the victim to the enjoyment (or to its monetary equivalent) of
the right with which the polluter was duty-bound not to interfere. In
contrast, however, the polluter’s success in staving off this suit does not

55
Wesley N. Hohfeld, “Some Fundamental Legal Conceptions as Applied in Judicial Rea-
soning,” (1913) 23 Yale L.J. 16, at 32.
56
Walter v. Selfe, 4 DeG & S 315 (1851), at 322.
106 corrective justice

show that the polluter has a claim right against the victim. Nor is the
victim under a correlative duty to continue to suffer from the pollu-
tion. If, for instance, the victim encased the property exposed to the
pollution in a dome, so that the victim could no longer smell the pol-
luting odors, no duty to the polluter would be violated.57 The polluter’s
entitlement is merely a liberty58—that is, the recognition that no right
of the victim is being infringed and that therefore the polluter is under
no correlative duty to the victim. Accordingly, whereas the victim’s
entitlement is a claim right correlative to which is a duty on the pol-
luter, the polluter’s entitlement is to a liberty correlative to which is
not a duty on the victim but the absence of a claim right. The term
“entitlement” used by Calabresi and Melamed is therefore misleading.
Its application to both the victim and the polluter within a symmetri-
cal framework of protection masks the difference between a claim
right and a liberty—and thus also the difference in their correlatives
between being and not being under a duty.
By positing that a liability rule is symmetrically applicable to the
entitlements of the polluter and the victim, the Calabresi–Melamed
framework flattens the normative landscape in two ways. First, the
award of damages against the polluter is no longer seen as the response
to the violation of a duty correlative to the victim’s right. Applying
the notion of duty to the polluter’s conduct would destroy the sym-
metry with the victim’s payment—clearly not a matter of duty but
rather a condition for getting an injunction—under the indemnified
injunction. Instead, the damage award is regarded merely as giving the
polluter an option to purchase the victim’s entitlement at an object-
ively established price.59 The polluter’s violation of the victim’s right is
thereby treated as an allowable choice, rather than as a wrong. Second,
in giving the victim a parallel option to purchase the polluter’s enti-
tlement in a case like Spur, the indemnified injunction distorts the
victim’s right. Instead of vindicating the victim’s right against the pol-
luter’s wrongful conduct, the injunction operates only as the prelude

57
For a similar point, see Henry Smith, “Exclusion and Property Rules in the Law of
Nuisance,” (2004) 90 Va. L. Rev. 965, at 1012.
58
In Hohfeld’s terminology, a “privilege”; above n. 55, at 33.
59
Daniel Friedmann, “Rights and Remedies,” in Comparative Remedies for Breach of Con-
tract, ed. Nili Cohen and Ewan McKendrick (2005), 3, at 7. One of the results of the Calabresi–
Melamed analysis is the growth of a complex literature treating liability rules as options. See
Madeline Morris, “The Structure of Entitlements,” (1993) 78 Cornell L. Rev. 822, at 851–56;
Ian Ayres and Paul M. Goldbart, “Optimal Delegation and Decoupling in the Design of
liability Rules,” (2001) 100 Mich. L. Rev. 1; Ian Ayres, Optional Law (2005).
 remedies 107

for the victim’s option to purchase the polluter’s abstention from pol-
luting. In this way the Calabresi–Melamed framework of symmetrical
liability rules substitutes parallel options for the correlative categories
of right and duty.
In sum, then, Calabresi and Melamed put forward a remedial frame-
work of symmetrical effects, even though these effects protect entitle-
ments of differing normative characters. This is inconsistent with the
idea that the point of the remedy is to match the injustice that moti-
vates it. In the Calabresi–Melamed framework the response is based
not on the normative character of the causative event, which it ignores
and obscures, but on the efficiency gains that result from movement of
resources to their most valued uses. As is typical of the second concep-
tion of remedies, the Calabresi–Melamed analysis treats the causative
event merely as the preliminary to a response that is independent of
the reason for considering the event an injustice.

4. Monism and dualism

In recent years commentators on the law of remedies have distin-
guished two approaches to the relationship between the causative
event and the response.60 The “monist” integrates the right and the
remedy, treating the remedy as the mirror image or reflex of the
right—in Peter Birks’s words, “the same thing as the right, looked at
from the other end.”61 The “dualist” separates the right from the rem-
edy, postulating that the court in determining the remedy chooses
from the basket of all potential remedies the context-specific one
that is most appropriate in the circumstances. How does this distinc-
tion between monism and dualism relate to the reason and condition
distinction?
Obviously these two sets of distinctions overlap, with the corrective
justice conception being monistic and the other conception dualistic.
The former conception treats the remedy as merely the continuation
of the plaintiff ’s right in circumstances in which the defendant has, or

60
G. Hammond, “Rethinking Remedies: The Changing Conception of the Relationship
between Legal and Equitable Remedies,” in Remedies: Issues and Perspectives, ed. Jeffrey Ber-
ryman (1991), 87, 90–91; Michael Tilbury, “Remedies and the Classification of Obligations,”
in The Law of Obligations and Boundaries, ed. A. Robertson (2004), 11, 17–24.
61
Peter Birks,“Definition and Division: A Meditation on Institutes 3.13,” in The Classifica-
tion of Obligations, ed. Peter Birks (1997), 1, 24.
108 corrective justice

has done, something inconsistent with that right. In this sense the
remedy is, as Birks said, the right “looked at from the other end.”62
Conversely, the dualistic approach allows for the possibility, character-
istic of the latter conception, that “the remedy granted in any case
may serve purposes unrelated to the reason(s) for the imposition of
the liability in the first place.”63 The principal difference between of the
two sets of distinctions is that the reason–condition distinction oper-
ates closer to the normative basis of liability. Its central concern is to
conceptualize the relationship between right and remedy in terms of the
reasons for thinking that justice requires that the defendant be held
liable to the plaintiff. In contrast, the debate between monism and
dualism focuses on adherence to or rejection of precedent, history,
and established categories.64
However, corrective justice may assist in narrowing, or at least illu-
minating, the gap between dualism and monism. By presenting the
remedy closer to its normative ground within a conceptual frame-
work that is more abstracted from the particularities of precedent,
corrective justice affords a more unobstructed vista of the issues that
divide monism and dualism. In particular, corrective justice may point
the way to dealing with a tension that exists in both monism and
dualism, and thus to reconciling monism and dualism themselves.
Consider two features of dualism. The first feature is the dualist
emphasis on the fact that the law brings into play at the remedial stage
certain considerations that were absent earlier. This is especially the
case with equitable remedies, where the determination of the remedy
is leavened by various grounds for exercising discretion. In the dualist
view, the introduction of these new considerations at the remedial
stage makes it implausible to regard the remedy simply as the right
viewed from the other end. The second feature is that the dualist does
not posit the complete separation of the right from the remedy. When
recommending the availability of all possible remedies, the dualist does
not have in mind all the ingenious remedies that the fertile human
mind can excogitate—for example, the indemnified injunction and its
numerous notional relatives65—however disconnected they are from

62
Ibid.
63
Tilbury, above n. 60, at 19.
64
J. D. Davies, “Duties of Confidence and Loyalty,” (1990) Lloyd’s Maritime & Commer-
cial L.Q., 4, at 5.
65
Saul Levmore lists sixteen remedies available within the Calabresi–Melamed frame-
work; see Levmore, above n. 11, at 2173.
 remedies 109

the normative ground of the cause of action. Even a dualist recognizes

the “unbreakable relationship” between obligation and remedy, there-
by acknowledging that “the nature of the obligation breached is the
starting point and generally the most important factor (while not nec-
essarily the only one) determining the appropriate remedy in any par-
ticular case.”66 Thus, the dualist maintains that the remedial stage
involves considerations that are both different from and yet unbreak-
ably related to the infringed right. How is the combination of these
apparently incompatible features to be understood?
The monist position faces a parallel challenge. The monist wants to
underline the second of these two features, the adamantine nature of
the connection between right and remedy. Yet the first feature, that
the remedy may be determined by considerations not present in the
specification of the right and its infringement, seems also undeniable
in the law’s remedial practice. If the monist position is to be plausible,
it too must find a way to combine these two features despite the
apparent tension between them.
So far as corrective justice is concerned, the monist position that the
remedy is “the same thing as the right, looked at from the other end”67
has to be understood in a particular way. Under corrective justice the
identity of right and remedy means that they form a single continuous
normative sequence, so that the right both grounds and limits the rem-
edy. The function of the remedy is to remove the inconsistency with
the plaintiff ’s right that is imputable to the defendant. The plaintiff ’s
right, therefore, is the reason both for the defendant’s duty to abstain
from doing or having something inconsistent with it and for the defend-
ant’s duty to eliminate any such inconsistency should it occur. These
duties, being correlative to the same right, are conceptually the same
duty, though the duty has a different content before and after the injus-
tice. The right and its correlative duty thus persist through the causative
event despite the change that the causative event works in the particular
thing to which the plaintiff has a right and in the particular action that
the defendant has a duty to perform. The identity of right and remedy,
in other words, is not one of content but of normative ground. When
one gazes at the right from the other end, one looks at the same thing
even though what one may see is different.

66
Tilbury, above n. 60, at 26; cf. also David Wright, “Wrong and Remedy: A Sticky Rela-
tionship,” (2001) Sing. J. Legal Stud. 300.
67
Above n. 61.
110 corrective justice

One kind of circumstance that may manifest a change in the con-

tent of the right at the remedial end is trivial. As was noted above in
section 2, the causative event may have had effects in time or space
that render the specific subject matter of the right irretrievable. For
example, the plaintiff may have had a right to a specific thing that was
destroyed or to contractual performance at a specific time that has
passed. Nonetheless the right, as a marker of the normative relation-
ship between the parties with respect to the specific thing or to the
contractual performance at a specific time, remains intact. The court
then requires the defendant to give the plaintiff a quantitative or quali-
tative equivalent. Although not physically or temporally identical with
the original content of the right, this equivalent is from the normative
standpoint nonetheless “the same thing as the right, looked at from
the other end.”68
More germane to the present discussion is another kind of circum-
stance, in which the law at the remedial stage refers to fresh consider-
ations that pertain not to the physical or temporal landscape but to
the normative one. These fresh considerations arise out of the rela-
tionship between the plaintiff ’s right and the entire system of rights.
In this connection, three points have to be kept in mind.
First, rights are necessarily systematic. A right does not merely give
a particular right-holder an entitlement to a particular thing. Rather,
it exists normatively within a comprehensive system of rights. Rights
are the juridical manifestations of the freedom of self-determining
beings who interact with one another as equals. Accordingly, a person
claiming a right necessarily acknowledges that others also have rights
with which the claimant’s right must systematically coexist.69 There is
no particular right divorced from its situation within a system of
rights.
Second, rights have to be enforced by a court—that is, by a disin-
terested and impartial public authority. Because a claim of right
involves the acknowledgment that others also have such claims within
the system of rights, the obligations generated by any particular right
can be treated as binding only insofar as there exists a public authority
that interprets and enforces all the obligations contained in all the
rights considered as a totality.70 This authority does not give effect to

68
Ibid.
69
Kant, above n. 17, [6:255–56].
70
Ibid.
 remedies 111

the unilateral will of any particular right-holder or look at any right

in isolation. Rather, it impartially and disinterestedly brings out the
normative implications of the system of rights and applies them to
each right within the system. Accordingly, a claim of right implies a
judicial role in interpreting and enforcing the claim. The claim is,
therefore, qualified both by the limits of judicial competence and by
the exercise of judgment in accordance with public reason. In this
way the court functions as guardian of the system of rights, preserving
the rights of all in their juridical relationships with one another.
Third, the special responsibility of the court as the guardian of the
system of rights becomes particularly salient at the remedial stage. The
remedy is the point at which the coercive enforcement of the plain-
tiff ’s right directly impacts the defendant. It is at this point that a court
must be particularly sensitive not only to the plaintiff as the claimant
of the right, but also to its own institutional role within the system of
rights and to the defendant as a participant in that system.
Thus, within corrective justice the infringed right that is the reason
for the remedy has its being within a system of rights whose meaning
is determined by a court exercising its judgment within its institutional
competence. In being “the same thing as the right looked at from the
other end”71 as the monist claims, the remedy reflects the right consid-
ered not in isolation but as a component of the system of rights. This
does not change the right at the remedial stage from what it was ini-
tially: the particular right always had normative significance—that is, it
generated a duty that the defendant had to treat as binding—only
because it belonged to the system of rights.
Seen in this light, the dualist contention that the remedy may “serve
purposes unrelated to the reason(s) for the imposition of liability in
the first place,”72 can also be understood in a way that brings it within
the corrective justice conception of remedies. If the reason for impos-
ing liability refers to the infringed right taken in isolation, the remedy
may appear to be unrelated to it. However, the moral significance of a
right precludes taking the right in isolation from the system of rights as
a whole. The system of rights—what Kant called “the sum of the con-
ditions under which the choice of one can be united with the choice
of another in accordance with a universal law of freedom”73—forms an

71
Above n. 61.
72
Above n. 63.
73
Kant, above n. 17, [6:231].
112 corrective justice

objective normative order of which the right of the plaintiff is a par-

ticularized crystallization.74 Thus, so long as it is intelligible within the
system of rights, the remedy is not unrelated to the reason for impos-
ing liability, even if it embodies considerations that did not figure in
the imposition of liability in the first place.
Within a functioning system of law, in other words, systemic con-
siderations are always implicit in the right even as initially regarded. In
the determination of the remedy, however, the court has to exercise its
coercive authority so as not to affect the defendant in a manner incon-
sistent with the system of rights and therefore with the defendant’s
status in the community of interacting rights-holders.The court there-
fore makes explicit the systemic considerations appropriate to the cir-
cumstances at hand. This difference, between the implicit and the
explicit, accounts for the appearance of new considerations at the
remedial stage, yet in a way that preserves the centrality of the right, as
monists and dualists agree.
One example of the role of systemic considerations about rights,
drawn again from the law of nuisance, will have to suffice. As is well
known, in Commonwealth jurisdictions the successful plaintiff in a
private nuisance suit is entitled, with rare exceptions, to an injunction.
This injunction follows from the nature of the plaintiff ’s right. Because
the wrong of nuisance is an interference with the plaintiff ’s use and
enjoyment of property, the injunction forbidding the offending activ-
ity restores to the plaintiff the very thing of which the plaintiff was
wrongfully deprived. Under the corrective justice conception of rem-
edies, the granting of an injunction is straightforward.
But what about the rare exceptions where the plaintiff can get no
more than damages for a continuing nuisance? The case of Shelfer v.
City of London Electric Lighting Co.75 provides the classic formulation of
the circumstances in which a court may in its discretion substitute
damages for an injunction. According to the judgment, four factors
must all be present: a small injury to the plaintiff ’s rights, an injury
capable of being estimated in money, adequate compensation by a
74
The notion of “objective normative order” is drawn from German constitutional law
(see the Luth decision, Decisions of the Bundesverfassungsgericht—Federal Constitutional Court—
Federal Republic of Germany, vol. 2/I: Freedom of Speech (1998), 1–20) where the same issue
arises of the relationship between a particular right and the system of rights. For treatment
of this issue from the constitutional perspective see, Lorraine E. Weinrib, “The Postwar
Paradigm and American Exceptionalism,” in The Migration of Constitutional Ideas, ed. Sujit
Choudhry (2006). I am grateful to Lorraine Weinrib for discussion of this aspect.
75
[1895] 1 Ch. 287, at 222 (C.A.).
 remedies 113

small monetary payment, and the oppressiveness to the defendant of

an injunction. Under these circumstances the plaintiff is not restored
to the use and enjoyment of the property, which is the plaintiff ’s right,
but must instead accept the monetary equivalent of the wrongful
injury. To that extent, as the dualist would claim, the remedy does not
match the plaintiff ’s particular right but is rather the product of con-
siderations present only at the remedial stage.
These considerations represent a specific conception of remedial
fairness. On the one hand the plaintiff ’s injury is small, monetizable,
and capable of being adequately compensated by a small monetary
payment, so that the difference between damages and an injunction is
minimal from the plaintiff ’s perspective. On the other hand, the
injunction would be oppressive to the defendant, because it would
inflict on the defendant a significant hardship from which the plaintiff
would derive no substantial benefit. Under these circumstances the
only point of seeking an injunction would be to damage the defend-
ant rather than to promote any legitimate interest of the plaintiff .
Accordingly, by not awarding an injunction the court prevents the
plaintiff from using the judicial process to harm the defendant through
a remedy that would not materially benefit the plaintiff . In the lan-
guage of the civil law, the substitution of damages for an injunction
prevents the plaintiff from abusing his or her right.76
The notion of abuse of rights is often regarded as introducing into
the law considerations of social or moral good that are alien to the
idea of rights.77 A right provides an area of untrammeled freedom for
the right-holder. The notion of abuse of right seems to contain the
contradictory impulses of recognizing the right while yet limiting the

76
A similar explanation would apply to the requirement at common law that the plaintiff
mitigate his or her damages. It would also apply to some of the situations, exemplified by
Patel v. Ali [1984] Ch. 283 (C.A.), in which hardship to the defendant prevents the court from
issuing a decree of specific performance of a contract. Similar are situations of laches, in
which the prejudice suffered by the defendant as a result of the plaintiff ’s delay in seeking
specific performance is the basis for denying the remedy. These remedial considerations
illustrate Tony Weir’s observation that “we are not supposed to have any doctrine of abuse of
rights (but it is only the doctrine which is lacking).” Weir, “The Staggering March of Neg-
ligence,” in The Law of Obligations: Essays in Celebration of John Fleming, ed. Peter Cane and
Jane Stapleton (1998), 99, at 124. See also Joseph M. Perillo, “Abuse of Rights: A Pervasive
Legal Concept,” (1995) 27 Pac. L.J. 37.
77
S. Herman, “Classical Social Theories and the Doctrine of ‘Abuse of Rights’,” (1977) 37
La. L. Rev. 747; W. T. Tete, “Tort Roots and the Ramifications of the Obligations Revision,”
(1986) 32 Loy. L. Rev. 47, 68–72 (decribing Josserand’s conception of abuse of right).
114 corrective justice

right-holder’s freedom. Hence arises the impression that the norma-

tive basis of the limit is external to the rights perspective.
This impression, however, is mistaken. The idea that the law should
not legitimize the infliction on another of gratuitous harm is fully
consonant with the normative presuppositions of a system of rights.
Participants in the system of rights are conceived as persons with a
self-determining capacity for purposive action in their relations with
one another. Within this system all persons pursue their self-chosen
purposes, subject only to the constraint that their actions be capable of
coexisting with the purposiveness of others. This requires that one
pursue one’s purposes as ends that one is trying to achieve for oneself,
not as an obstacle against what someone else is trying to achieve.
Actions for the sake of creating mutual obstacles against the actions of
others cannot systemically coexist.
As juridical manifestations of self-determining freedom, rights pro-
vide the space within which all the right-holders may pursue ends of
their own. Such ends are consistent with the self-determining free-
dom of others only if the point of pursuing them is independent of
the adverse effect on someone else. When all act to pursue ends of
their own in this sense, they all rank equally as persons whose activ-
ities can coexist within the system of rights. Conversely, if the freedom
to perform an act merely to frustrate the purposes of another were
legitimate, rights would be transformed from markers of mutual free-
dom to instruments of subordination. Accordingly, it would be incon-
sistent with what is normatively presupposed in the system of rights
to allow a right to operate in a way that would harm another without
promoting (in the language of the civilians)78 a “serious and legitimate
interest” of the right-holder.
In awarding damages in lieu of an injunction to the victim of a
nuisance, the court is acting as guardian of the system of rights, keep-
ing the remedial consequences of the infringement of a particular
right aligned with what is normatively presupposed in the system as a
whole. The plaintiff is ordinarily entitled to an injunction, as that
would restore to the plaintiff the use and enjoyment of the property
that makes up the content of the plaintiff ’s right. However, through its
determination of the remedy the court prevents the plaintiff ’s right

78
Antonio Gambaro, “Abuse of Rights in the Civil Law Tradition,” in Aequitas and Equity:
Equity in Civil Law and Mixed Jurisdictions, ed. Alfredo Mordechai Rabello (1997), 632, at 637;
A. N.Yiannopoulos, “Abuse of Right in Louisiana,” ibid., 690, at 700.
 remedies 115

from operating in a way that prejudices the defendant without for-

warding an interest of the plaintiff . To grant an injunction under the
conditions enumerated in Shelfer would be inconsistent with the nor-
mative grounding of rights in the coexistence of the parties’ self-
determining freedom. This latter consideration is systemically implicit
in all rights, including the particular right infringed by the defendant.
Thus, the award of damages rather than an injunction treats the
infringed right not as free-standing but as situated within the system
of rights that is required for the right’s normative significance.
Similar observations about the systemic aspect of rights can be
made about other circumstances in which new normative consider-
ations are introduced at the remedial stage. Especially in exercising
their discretion to grant specific performance, courts often invoke
such factors as the difficulty of constant curial supervision, the futility
of the remedy, the absence of mutuality, the plaintiff ’s failure to do
equity, and the reluctance to enforce a decree of personal service.
These factors refer to substantive and institutional aspects normatively
presupposed in the system of rights, from the maintenance of the
transactional equality of the parties to the avoidance of the court’s
exceeding its institutional competence. To be sure, the application of
these factors to particular circumstances is controversial, as is evi-
denced, for instance, by the contention that the difficulty of judicial
supervision has been overstated.79 Nonetheless, however such factors
ought to figure in particular circumstances, the point for present pur-
poses is that, within the corrective justice conception of remedies,
attention to the system of rights rather than merely to a given right in
isolation opens up the normative space in which such factors can
coherently be considered.
If this is so, the corrective justice conception of remedies may be
able to reconcile, at least to some extent, the apparently competing
insights of monism and dualism. From the monist perspective, the
notion that the remedy is just the right looked at from the other end
is maintained, but with the understanding that because a particular
right participates in a system of rights, systemic considerations may
modulate the remedy. From the dualist perspective, the systemic char-
acter of rights may introduce new considerations at the remedial stage
while preserving the central importance of the infringed right and the
remedy’s unbreakable connection with it.

79
Jeffrey Berryman, The Law of Equitable Remedies (2000), 175–85.
116 corrective justice

5. Conclusion
In presenting the two conceptions of remedies, this chapter has ultim-
ately been concerned with how remedies might figure within a nor-
matively coherent system of private law. The corrective justice
conception maintains this coherence in two ways. First, it integrates
the causative event and the response by treating the injustice as an
inconsistency with the plaintiff ’s right that the remedy is supposed to
eliminate. The remedy is thus continuous with the right, reflecting its
structure and content. In contrast, the conception of remedies as
merely conditioned by causative events fragments private law by mak-
ing the remedy the locus of considerations that are independent of
the injustice of the causative event. Second, corrective justice treats
rights not as isolated entitlements but as members of a system of rights,
so that the remedy conforms to what is normatively presupposed in
the system as a whole. Accordingly, to be continuous with the right,
the remedy also has to be continuous with the right’s systemic aspects.
Only by attending to these two dimensions of juridical coherence can
a judge, in determining the remedy, live up to Aristotle’s observation
that “to go to the judge is to go to what is just, for the judge means to
be, as it were, justness ensouled.”80
80
Above n. 2.
 4
Gain-based Damages

1. Introduction
On what basis can damages for tortious conduct be measured by the
defendant’s gain rather than the plaintiff ’s loss? This question recently
has received increasing attention for reasons that are not hard to see.
Gain-based damages for torts implicate fundamental issues in our con-
ception of private law. On the one hand, they open up the possibility
of a more nuanced assessment of damages both by extending the long-
established jurisprudence of waiver of tort and by linking tortious
liability to the newly invigorated interest in restitutionary liability. On
the other hand, they present an intellectual puzzle. If tort law is con-
cerned with wrongful injury to the plaintiff, special arguments are
required to explain why, as a matter of justice, the remedy should refer
to the gains of the defendant. The reparation of injury seems satisfied
by compensating the plaintiff for his or her loss. To place into the
plaintiff ’s hands the defendant’s gain in excess of that loss seems to
confer a windfall.
My immediate excuse for revisiting this topic is to draw attention to
the relevance of inquiring into the plaintiff ’s entitlement to damages
measured by the defendant’s gain. Many of the current treatments of
gain-based damages for torts focus on the defendant’s desert in the
aftermath of wrongdoing or on the social good that can be achieved
by compelling the disgorgement of the wrongdoer’s gain. Hence com-
mentators appeal to the idea that one should not profit from a wrong,1
that disgorgement of wrongful gain is an effective deterrence for

1
Andrew Burrows, The Law of Restitution, 2nd ed. (2002), 455.
118 corrective justice

potential wrongdoers,2 or that gain-based damages are directed toward

the protection of legal facilities in whose integrity the community has
an interest.3 However, the injustice or social inexpediency of the
defendant’s retention of the gain indicates only the party from whom
the gain should be taken, not the party to whom it should be awarded.
Thus, such accounts fail to provide a reason for the law to transfer the
defendant’s gain to the plaintiff, of all people.4 If the basic difficulty
with an award of gain-based damages is the supposed windfall of the
plaintiff, an adequate treatment must show either that the award is jus-
tified despite being a windfall or that the award, where appropriate, is
not really a windfall but damages that the plaintiff may of right demand
from the defendant. The latter is the strategy that I will essay here.
More broadly, my aim in this chapter is to situate gain-based dam-
ages within the theoretical framework of corrective justice. Like any
remedy, an award of gain-based damages presupposes a conception of
the injustice that it remedies. Because corrective justice views damages
as undoing an injustice, it is particularly sensitive to the connection
between the remedy that the plaintiff can claim and the injustice that
is imputed to the defendant. The significance of that connection for
gain-based damages is the subject of this chapter.
Corrective justice embodies a notion both of the relationship of
the remedy to the injustice that it remedies and of the relationship
between the parties to that injustice. The two parties are correlatively
situated as the doer and sufferer of an injustice that is itself undone by
the corresponding remedy. Correlativity is inherent in the notion of
liability, which treats the injustice done by the former as the very injust-
ice suffered by the latter. Correlativity is also inherent in the idea of
damages, which treats the plaintiff as entitled to receive the very sum
that the defendant is obligated to pay. These instances of correlativity

2
Peter Cane, “Exceptional Measures of Damages: In Search of a Principle,” in Wrongs and
Remedies in the Twenty-First Century, ed. Peter Birks (1996), 301.
3
I. M. Jackman, “Restitution for Wrongs,” (1989) 48 Cambridge L.J. 302. By “legal facil-
ities” Jackman means “private property, relations of trust and confidence, and (with some
qualification) contracts,” which “require protection against those who seek to take the bene-
fits of an institution without the burdens thereof.” Ibid., 302.
4
Burrows, above n. 1, 480, shows some sensitivity to this problem. He accordingly sug-
gests that gain-based damages might be available in situations where the idea that one should
not profit from one’s wrong can be supplemented by additional reasons for restitution, such
as the need to protect facilitative institutions and to deter cynical wrongdoing.The difficulty
is that these additional reasons no more single out the plaintiff than the consideration they
supplement. It is hard to see how a multiplicity of reasons for restitution that are not plaintiff-
specific generate a recovery that is plaintiff-specific.
 gain-based damages 119

are mirror images of each other, with the plaintiff ’s entitlement to

damages from the defendant reflecting the plaintiff ’s entitlement to be
free from suffering injustice at the defendant’s hands.
Because the remedy mirrors the correlative structure of the injust-
ice, corrective justice disqualifies accounts of gain-based damages that
focus solely on the wrongdoer. Among the accounts excluded on this
basis are those that, under the banner of punishment and deterrence,
focus on the past or future actions of defendants, thereby treating the
plaintiff merely as a convenient conduit of social consequences rather
as someone to whom damages are owed to correct the wrong suf-
fered. Of course, this does not mean that explanations of gain-based
damages must be oriented to the plaintiff as some are now oriented to
the defendant. Corrective justice rejects all one-sided accounts, regard-
less of the particular side singled out. Rather, the parties must be seen
as related through the injustice in such a way that the plaintiff can
demand such damages as of right from the defendant. Accordingly, the
justification for awarding such damages must include not only the
reason for making the defendant pay but also the reason for entitling
the plaintiff to receive them; indeed, the same reason must apply on
both sides. Only then is the correlativity that marks the injustice car-
ried forward into the account of the remedy.5
The corrective justice framework, then, makes salient the need to
account for the plaintiff ’s entitlement to gain-based damages as a
response to the defendant’s having something, or having done some-
thing, inconsistent with the plaintiff ’s right. This chapter attempts to
satisfy this need in the following steps. Section 2 deals with the radical
proposal that would allow the plaintiff gain-based damages for any
wrongful gain. The problem with this proposal, that it does not ade-
quately link the damages to the normative quality of the wrongful act,

5
The basic idea of this article is similar to the one animating the analysis of the Eingriff-
skondiktion in German law since the writings of Wilburg and von Caemmerer, that the avail-
ability of restitution for an interference with the plaintiff ’s right depends on the scope and
purpose that the legal order attributes to the right. See B. S. Markesinis et al., The German
Law of Obligations, vol. 1, The Law of Contracts and Restitution: A Comparative Introduction
(1997), 744–45.This congruence of approach is hardly surprising, since the German approach
proceeds from the assumption that both restitution and delict are informed by Aristotle’s
notion of corrective justice; see E. von Caemmerer, Bereicherung und unerlaubte Handlung, in
Festschrift fuer Ernst Rabel, Band I (1954), 333, 335. Wilburg’s comment that restitution in the
Eingriffskondiktion “grows organically out of the mother-right, so to speak of itself ” (quoted
in John Dawson, Indirect Enrichment, in Ius Privatum Gentium (Rheinstein Festschrift), vol. 2
(1969), 789, 798) is a graphic way of portraying the intimate connection that corrective just-
ice posits between the injustice and the remedy.
120 corrective justice

leads in section 3 to an examination of the more traditional connec-

tion of gain-based damages with dealings with another’s property. The
virtue of the focus on dealings with another’s property is that the idea
of property includes within the proprietor’s entitlement the potential
gains from the property’s use or alienation. Section 4 generalizes from
the property cases to the conclusion that gain-based damages ought to
be available only insofar as they correspond to a constituent element
in the wrong that the defendant has done to the plaintiff. Section 5
discusses the relevance of the willfulness of the wrong insofar as it
entails a denial of the plaintiff ’s right, not because of notions of pun-
ishment or deterrence. The final section deals with situations where,
although the plaintiff ’s interest is not, strictly speaking, a proprietary
one, the analysis of the property cases nonetheless applies.

2. The Goff–Jones principle

Requiring tortfeasors to disgorge their wrongful gains is an intuitively
appealing idea. The tortfeasor’s moral claim to retain the fruit of his or
her own wrongdoing is a weak one. Permitting a wrongdoer to enjoy
the benefits of the wrong might seem an additional victimization of
the wronged party. Moreover, through the old “waiver of tort” cases,
the law has long recognized the principle of disgorgement for certain
torts, and the task of distinguishing torts that allow for disgorgement
from those that do not has proved a difficult one. Indeed one might
think that task misguided, on the ground that the morally relevant
feature of the wrongdoer’s gain is the commission of the wrong, not
the kind of wrong committed.
Presumably these considerations lie behind the radical proposal of
Lord Goff and Professor Jones that the victim of a tort should be
allowed restitution of all wrongful gains. They formulate their pro-
posal as follows: “If it can be demonstrated that the tortfeasor has
gained a benefit and that benefit would not have been gained but for
the tort, he should be required to make restitution.”6 This suggestion
has the advantage of simplicity, for it offers a comprehensive principle
that obviates the need to distinguish between wrongs that do and
wrongs that do not admit of gain-based damages.

6
Lord Goff of Chieveley and Gareth Jones, The Law of Restitution, 7th ed. (2007),
36-006.
 gain-based damages 121

Despite its simplicity and its intuitive appeal, the Goff–Jones prin-
ciple has not been accepted by the courts. In the view of Professor
Jones, this is because courts “fear the great unknown.”7 I would like to
suggest a more charitable reason, arising out of an ambiguity in the
notion of wrongful gain.
When we think of wrongful gains for restitutionary purposes, pre-
cisely how are the ideas of wrongfulness and gain connected? One
possibility is that a gain is wrongful because of its history; that is, a
gain is wrongful if it is the consequence of a wrongful act. Rather
than pointing to a feature of the gain itself, “wrongful” is used to indi-
cate that wrongful conduct by the defendant is a historical antecedent
of the defendant’s gain. The wrongdoing that underlies ascription of
wrongfulness stands to the gain as cause to effect. “Wrongful gain,”
then, could be understood as shorthand for the more accurate descrip-
tion “gain resulting from a wrongful act.”
The other and more restrictive possibility is that we call a gain
“wrongful” by virtue of its inherent normative quality. Here the sig-
nificance of the wrongfulness is not merely that it produces the gain,
but that it survives into the gain and informs it. The gain’s origin in
wrong is a necessary condition for the gain’s having this normative
quality, but something further is required. For the gain to take on the
normative quality of wrongfulness, it must be the materialization of a
possibility—the opportunity to gain—that rightfully belonged to the
plaintiff. Because it is an incident of the plaintiff ’s entitlement that the
defendant has wrongfully infringed, the gain is not merely the result
of a wrongful act, but is the continuing embodiment of the injustice
between the parties.
The Goff–Jones principle rests on the assumption that the relevant
understanding of wrongful gain is the historical one. Using language
reminiscent of the factual causation test in negligence law (“the bene-
fit would not have been gained but for the tort”), they formulate the
relation between wrongfulness and gain solely in terms of cause and
effect. In their view the fact that the plaintiff has been wronged by the
defendant, plus the fact that the defendant has consequently realized a
benefit, add up to the liability of the defendant to surrender the bene-
fit to the plaintiff .
Why might one be reluctant, as Jones notes that courts are, to
accept a principle that bases liability on the historical rather than the

7
Gareth Jones, Restitution in Public and Private Law (1991), 77.
122 corrective justice

normative connection between wrongdoing and gain? I suggest that

it is because the Goff–Jones principle challenges the internal coher-
ence of private law. By this I mean not that the principle measures
the plaintiff ’s injury by the defendant’s gain (the waiver of tort cases
show that this is not in itself unacceptable), but that the connection it
posits between what the wrongdoer has done and what the victim
recovers is at odds with the principle underlying the law’s treatment
of wrongful loss.
The phenomenon of compensatory damages for wrongful loss is
the counterpart to gain-based damages for wrongful gain. In wrongful
loss the same two conceptions of wrongfulness present themselves. A
loss is wrongful by reference to its history if the occurrence of the loss
was the result of the defendant’s wrong. A loss is wrongful by refer-
ence to its normative quality if the potential for such loss is a reason
for considering the defendant’s conduct to have been wrongful in the
first place.
The cases on duty and proximate cause in negligence illustrate the
difference between these two ways of connecting wrongfulness and
loss. In the famous case about duty, Palsgraf v. Long Island Railroad,8 the
plaintiff was injured by an act that negligently imperiled the property
of someone else. The plaintiff ’s loss was wrongful in the historical
sense, in that one of the historical antecedents of the loss was a wrong-
ful act by the defendant. However, because the plaintiff was beyond
the ambit of reasonably foreseeable injury, the prospect of her being
harmed was not a reason for thinking that the defendant’s conduct
was wrongful.The conduct was thus not a wrong relative to her. Simi-
larly, the Wagon Mound case holds that the requirement of proximate
cause is not satisfied when the defendant negligently exposes the
plaintiff to the risk of one kind of injury, but the plaintiff suffers an
injury of a different kind.9 Such losses are historically connected to
the wrongful conduct, in that they would not have occurred without
it, but they do not partake of the conduct’s wrongful quality since
they are not the losses by virtue of which the conduct is regarded as
wrongful.
These doctrines show that in the context of compensatory damages
for negligence, liability exists only when the connection between the
wrongfulness and the loss is normative and not merely historical. That

8
248 N.Y. 339, 162 N.E. 99 (1928).
9
The Wagon Mound, No.1 [1961] A.C. 388 (P.C.).
 gain-based damages 123

the negligent conduct is factually caused by the loss is not sufficient.

To be recoverable, the loss must be within the risk the creation of
which rendered the defendant’s act unreasonable.
From the perspective of corrective justice, negligence law has good
reason for insisting, through the doctrines of duty and proximate cause,
that a loss should be considered wrongful by virtue of its normative
quality rather than merely its history.10 In negligence law, wrongdoing
consists in the creation of unreasonable risk. When the plaintiff ’s loss
is within the ambit of the very risk that renders the defendant’s con-
duct wrongful, the parties stand to each other as the active and passive
poles of the same injustice. Because freedom from this kind of loss is
both the content of the plaintiff ’s right and the object of the defend-
ant’s duty, the parties are normatively linked through the wrongfulness
of the defendant’s risk-creation. Liability then obligates the defendant
to eliminate the loss wrongfully imposed on the plaintiff, and thus to
restore (to the extent that monetary damages can) the freedom from
loss which was the plaintiff ’s original entitlement. In contrast, when
the plaintiff ’s loss, although caused by the defendant’s wrongdoing, is
not within the ambit of what makes it wrongful, the defendant’s con-
duct cannot be said to be wrongful with respect to that plaintiff ’s loss.
Because the parties are then not related to each other as doer and suf-
ferer of an injustice, the plaintiff lacks the normative standing to call
upon the defendant to make good the loss.
Negligence law’s treatment of wrongful loss undermines the Goff–
Jones proposal concerning wrongful gain. To highlight the parallel
between the two, one may say (echoing the Goff–Jones formulation)
that negligence law rejects the principle that the tortfeasor must pay
compensation “if it can be shown that the victim suffered a loss and
that loss would not have been suffered but for the tort.” The compen-
satory principle that negligence law rejects has the same structure as
the gain-based principle that Goff and Jones propose. Both principles
use factual causation as a sufficient condition for the damage award.
The only difference between them is that whereas the compensatory
principle deals with loss and compensation, the Goff–Jones principle
deals with gain and restitution. Since it is the significance of factual
causation that is at issue in each, the fact that one deals with compen-
sation for loss and the other with restitution for gain is unimportant. If
the law has good reason for rejecting factual causation as a sufficient

10
Above, chapter 2, section 2.
124 corrective justice

condition of liability for wrongfully caused loss, then it also has good
reason to reject it as the test for wrongfully caused gain. Accepting the
Goff–Jones principle would introduce the inconsistency of allowing
factual causation to be sufficient for restitution when it has been found
to be insufficient for compensation.
The corrective justice analysis of compensation for wrongful loss
applies, mutatis mutandis, to restitution for wrongful gain. If the wrong-
fulness consists in creating the prospect of a loss (as, let us assume for
the moment, is the case with negligence), the fact that the defendant
has realized a gain as well adds nothing to the plaintiff ’s case. Because
the gain lies beyond the wrong done to the plaintiff, the plaintiff suf-
fers no injustice through the existence of the gain. The parties do and
suffer injustice only with respect to the loss, not the gain; the gain
remains external to their relationship.
Accordingly, from the standpoint of corrective justice, factual caus-
ation no more suffices for liability on the gain side than it does on the
loss side. What matters is not the historical connection of gain to
wrong, but rather the nature of the wrong as an inconsistency with
the plaintiff ’s right and whether the gain partakes of that inconsisten-
cy. Gain-based damages are justified when the defendant’s gain is of
something that lies within the right of the plaintiff and is therefore
integral to the continuing relationship of the parties as the doer and
sufferer of an injustice. Then the gain stands not merely as the sequel
to the wrong but as its present embodiment, and the plaintiff is as
entitled to the gain as he or she was to the defendant’s abstention from
the wrong that produced it. A gain that thus embodies the injustice
done by the defendant to the plaintiff immediately implies restitution
of that gain.
The Goff–Jones principle has the twin virtues of simplicity of for-
mulation and comprehensiveness of application, but it also has the
corresponding vices. On the one hand, the proposed principle obvi-
ates the need to distinguish among wrongs by using the simple test of
whether the wrong factually caused the gain. On the other hand, the
principle is insensitive to the limited significance of factual causation
and to the need to forge a normative link between the wrong and the
gain. Given that the restitution of the gain depends on the gain’s nor-
mative quality, and that quality varies with the nature of the wrong,
there is no alternative to the difficult task of distinguishing between
the wrongs that do and the wrongs that do not admit of the restitu-
tion of their resulting gains.
 gain-based damages 125

Aside from highlighting the necessity for a normative connection

between the defendant’s wrong and the plaintiff ’s entitlement to the
gain, the parallel between wrongful loss and wrongful gain suggests a
more particular point: gain-based damages are especially appropriate
when property rights are violated. The reason that negligence law
requires the defendant to compensate the plaintiff for wrongful loss is
that such loss is the materialization of an adverse possibility—the
unreasonably created risk of harm—to which the defendant ought
not rightfully to have exposed the plaintiff. Because the defendant
then inflicts a loss from which the plaintiff is entitled to be immune,
the loss constitutes an injustice between the parties that an award of
compensatory damages reverses. Similarly, gain-based damages should
be available when the defendant’s gain is the materialization of a favor-
able possibility—the opportunity to gain—that rightfully belonged to
the plaintiff. Then the gain to be nullified by the award of gain-based
damages represents an injustice both committed by the defendant and
suffered by the plaintiff . Since a proprietary right includes the oppor-
tunity to gain from what one owns, one may plausibly regard the
defendant’s dealings with the plaintiff ’s property as an occasion
for gain-based damages. That, at any rate, is the argument of the next
section.

3. Dealing with another’s property

As has often been noted, the misappropriation of another’s property is
the paradigmatic example of an event that gives rise to gain-based
damages.11 Because property rights give proprietors the exclusive right
to deal with the thing owned, including the right to profit from such
dealings, gains resulting from the misappropriation of property are
necessarily subject to restitution. Gains from dealings in property are
as much within the entitlement of the proprietor as the property
itself.
The disgorgement of these proprietary gains fits readily within the
correlativity of corrective justice. Property consists simultaneously in a
right of the proprietor and in a correlative duty on others to respect

11
See especially Daniel Friedmann,“Restitution of Benefits Obtained through the Appro-
priation of Property or the Commission of a Wrong,” (1980) 80 Colum. L. Rev. 504; von
Caemmerer, above n. 5, 353.
126 corrective justice

that right. Just as the owner’s right to set the terms on which property
is used or transferred implies a correlative duty on others to abstain
from using or selling it, so the owner’s right to the profits from the use
or transfer of the property imports a correlative duty on others to
abstain from such profits. This correlativity of the proprietor’s right
and the wrongdoer’s duty means that the realization of an unauthor-
ized gain is an injustice as between them. The gain is the continuing
embodiment of this injustice, and the injustice is undone when the
gain is restored to the owner of the object from which the gain
accrued.
Gain-based damages for dealing with another’s property mirror the
wrong and illuminate its nature. The law’s focus on the benefits of
ownership at the remedial stage presupposes the defendant’s intention
to act on the owned object at the stage of wrongdoing. In appropriat-
ing the benefits from using or alienating the object, the defendant
implicitly asserts the ownership that alone would entitle the defendant
to those benefits. Gain-based damages reverse the wrong by showing,
through the return of the benefits, that the law considers the defend-
ant’s implicit assertion of ownership to be a nullity whose conse-
quences are to be undone. The remedy is conditioned, therefore, not
merely on the defendant’s realization of a benefit but on the defend-
ant’s having treated the object as if it were his or her own. One treats
an object in this way when one so directs one’s attention to the object
that its use or alienation can be regarded as an execution of one’s pur-
poses. In contrast, action that inadvertently produces an effect on the
object does not qualify as an expression of one’s will with respect to
the object, and so is not the basis for gain-based damages. Thus, gain-
based damages are available for intentional torts against property and
not for harm to property that results from negligence.
From the perspective of corrective justice, gain-based damages for
proprietary wrongs are an entitlement of the proprietor, and not
merely a mechanism for protecting the integrity of property as a
facilitative institution.12 The argument for referring to property as
a facilitative institution is that, because damages measured by the
defendant’s gain do not reflect an injury to the plaintiff personally,
they must be justified by pointing to an institutional harm. The premise
of this argument is false.The fact that the damages are gain-oriented does
not exclude their reflecting an injury to the plaintiff personally. One’s

12
Jackman, above note 3.
 gain-based damages 127

rights provide the baseline for measuring injury. If those rights include the
possibility of gain, then the defendant’s gain measures the extent of the
plaintiff ’s injury. The relevance of property is not that it is a facilitative
institution, but that it connects the parties in such a way as to make the
object owned—and thus the gain that dealings in that object can pro-
duce—the locus of a right and a correlative duty.
Two broad categories of dealings in property can give rise to profits
that the wrongdoer is obligated to disgorge.The defendant either might
purport to alienate that which belongs to the plaintiff, or might benefit
by putting it to an unauthorized use. In either instance the plaintiff
ought to be allowed damages measured by the defendant’s gain.
The disgorgement of gains has long been uncontroversial in con-
nection with purported alienations of property. The old waiver of tort
cases allowed recovery of the proceeds realized from a thief ’s sale of
the owner’s goods, even if the proceeds exceeded the goods’ market
value.To accomplish this within the framework of an action in assump-
sit for money had and received, the courts implied a contract to repay,
ascribing to the thief a fictitious relationship of agency. Although such
reasoning led to well-known difficulties and is no longer necessary, its
basic normative impulse was sound. The language of agency expressed
the implications of property. The idea behind the ascription of agency
was that the only legal basis for selling what belongs to another is that
the seller is acting as the owner’s agent and therefore holds the pro-
ceeds on the owner’s behalf.
The jurisprudence concerning gains realized through the use rather
than the purported alienation of another’s property is more complex.
The notorious case of Phillips v. Homfray13 is often taken to indicate
that such gains are not recoverable.14 In that case the plaintiff claimed
wayleave rent for the use of underground passages through which cer-
tain minerals had secretly been conveyed. Because of the death of the
defendants, the established doctrine that actio personalis moritur cum per-
sona barred an action for tort. Accordingly, the plaintiff based his claim
not on his own wrongful loss but on the defendant’s wrongful gain
through the free and unauthorized use of the passageways.The English
Court of Appeal dismissed the claim. In the court’s view, one could

13
[1883] 24 Ch. D. 439 (C.A.).
14
But see William Swadling, “Phillips v Homfray (1883),” in Landmark Cases in the Law of
Restitution, ed. Charles Mitchell and Paul Mitchell (2006), arguing that the usual interpreta-
tion is mistaken.
128 corrective justice

recover only for wrongful gains in the form of property or the pro-
ceeds or value of property withdrawn from the plaintiff and added to
the estate of the defendant. Here the defendant’s profit in using the
plaintiff ’s passageways consisted merely in the saving of an expense,
not in bringing into the estate any additional property, or proceeds or
value of property, belonging to the plaintiff .
The notion that the use of another’s property is not a benefit is
now universally reprobated by restitution scholars. Even under the
doctrine of the day, the unauthorized use of another’s land was com-
pensable by the assessment of a wayleave rent, as is shown by other
episodes in the protracted litigation of this plaintiff ’s claim.15 More-
over, from the gain-based standpoint it seems odd to distinguish
between the enrichment that the estate would have achieved posi-
tively by having its assets swollen by the proceeds or value of property,
and the enrichment it achieved negatively by not having its assets
diminished by the payment of a wayleave rent.
Perhaps the best that can be said about the case is that it deals with
a narrow point that arose out of the positive law of the time. On the
death of the tortfeasor, the actio personalis rule necessitated a distinc-
tion between actions to repair a wrongful loss, which were barred, and
actions to recover a gain, which were allowed. In cases of the unauthor-
ized use of another’s property, the action often can be conceptualized
either way, since the use, in and of itself, is both a benefit to the
defendant and a loss by the plaintiff of an opportunity to exploit a
potentially profitable asset.16 In the Phillips case the tortfeasor, by using
the passageways without authorization, was spared the expense of
negotiating for the plaintiff ’s consent or making arrangements that
would avoid the need to trespass on the plaintiff ’s property. This, how-
ever, was merely “a negative benefit . . . acquired by saving himself the
expense of doing his duty.”17 There are few breaches of duty that could

15
W. M. C. Gummow, “Unjust Enrichment, Restitution, and Proprietary Remedies,” in
Essays on Restitution, ed. P. D. Finn (1990), 47.
16
Cf. Strand Electric and Engineering v. Brisford Entertainments [1952] 2 Q.B. 246
(C.A.), where two members of the court analyzed the claim in terms of loss and one judge
analyzed it in terms of gain. See also Robert J. Sharpe and S. M.Waddams, “Damages for Lost
Opportunity to Bargain,” (1982) 2 Oxford J. of Legal Stud. 290 for the argument that dam-
ages for the use of another’s property are compensatory rather than gain-based, in that they
compensate for the deprivation of the amount for which the plaintiff would have bargained
away the right.
17
Phillips v. Homfray, 24 Ch. D. at 465 (Lord Justice Bowen uses this language to explain
one of the cases on which he is relying).
 gain-based damages 129

not be avoided by undertaking some expense. All interferences with

property, for instance, can be seen either as wronging the owner or as
sparing the wrongdoer the expense of purchasing the owner’s con-
sent. If the actio personalis rule barred actions for the former but allowed
actions for the latter, its scope would be nugatory. The saving of the
expense of doing one’s duty, in other words, was not a benefit distinct
from the wrongfulness of the loss, but was rather that wrongfulness
itself, formulated in terms of a benefit. The court, anxious to preserve
the integrity of the actio personalis rule (in the words of Lord Justice
Bowen, “[i]t is part of the law, and while so, ought not to be frittered
away”18), in effect held that the rule caught any gain that was indistin-
guishable from a wrongful loss.
Be that as it may, in emphasizing the importance of property or its
proceeds or value, the Phillips case in effect distinguishes, with respect
to the availability of gain-based damages, benefits realized through use
from benefits realized through alienation. This distinction should have
no further vitality. If the key to the recovery of gain-based damages is
that the defendant has dealt with the plaintiff ’s property, the availabil-
ity of such damages should not depend on whether the dealing took
the form of a use or an alienation. It is true, as noted above, that the
damages in cases of the unauthorized use of property often can be
seen to be loss-based, because the defendant’s use of the plaintiff ’s
asset deprives the plaintiff of the opportunity to realize profits through
it. Some cases, however—as where the defendant uses an object that
has been stored19 or that is part of a discontinued business20—are
harder to construe as involving a loss, since the defendant’s unauthor-
ized use cannot realistically be said to have deprived the plaintiff of
money that the plaintiff would otherwise have had. The damages in
such cases may more easily be regarded as based on the gain that the
defendant realized by not paying for the use. Whether seen as quanti-
fying a wrongful loss or a wrongful gain, gain-based damages for
unauthorized use are justified because they restore the value of the
owner’s pre-existing right. Indeed, what ultimately matters is not the
semantic exercise of designating the damages as gain-based or loss-
based, but the normative task of connecting the damages to the
infringed right.

18
Ibid., at 456.
19
Olwell v. Nye and Nissen, 173 P.2d. 652 (Wash. 1946).
20
Penarth Dock Engineering v. Pounds [1963] 1 Lloyd’s List L. Rep. 359 (Q.B.D.).
130 corrective justice

The difference between unauthorized use and purported alienation

goes not to the availability of gain-based damages but to the way in
which they usually are computed. In the case of unauthorized use, the
measure of the damages is the value of the use; in the case of alien-
ation, the plaintiff can choose either the value of the thing alienated
or the price the defendant received. This difference implies no princi-
ple that would bar gain-based damages in cases of use, but merely
reflects the contingency that, in contrast to alienation, unauthorized
use does not necessarily involve the defendant in a further exchange
of the plaintiff ’s property. Were such an exchange to take place (for
instance, if the defendant charged a fee purportedly to license the use
to a third party), the plaintiff presumably could choose to have the
defendant disgorge this fee.21
From the perspective of corrective justice, two aspects of these
damages call for comment: the role of the notion of value and the
option of the plaintiff to insist on disgorgement if there has been an
exchange.
The notion of value fits into corrective justice in the following way.
Corrective justice deals with interacting parties correlatively as doer
and sufferer of an injustice. Inasmuch as it governs interaction, correct-
ive justice applies to parties who impinge upon each other by acting
on particular things in the world pursuant to their specific needs and
wants. But inasmuch as it embraces the two parties as correlatively
situated, corrective justice abstracts to a common standpoint from the
particularity of these things and from the specificity of these needs
and wants. Value is the economic notion that fulfills this abstracting
function.22
By making objects quantitatively comparable, value enables cor-
rective justice to apply to property despite the heterogeneity of per-
sons’ specific needs and wants. Because corrective justice is concerned
with the correlativity of the doing and suffering of an injustice, it
does not regard something owned merely as a particular thing used
to minister to the specific needs and wants of the particular person
who owns it. Its interest is in the thing as a factor in the juridical
relationship between the owner and others.Value provides the means
for quantitatively comparing the owner’s wants to other wants and
the owner’s thing to other things, thereby enabling the possible uses

21
Cf. Edwards v. Lee’s Administrator, 96 S.W. 2d.1028 (Ky. Ct. App. 1936).
22
On value, see below chapter 6, section 2.
 gain-based damages 131

of a thing to figure in the interactional framework of a juridical rela-

tionship.
When a property right is violated, corrective justice requires the
wrongdoer to undo the wrong perpetrated against the proprietor.
Neither the specific wants satisfied by the defendant’s wrongful use
nor the specific wants frustrated by the unavailability to the plaintiff of
the thing used can be reversed in their specificity. Value, however,
presents the quantitative equivalent of that use or that thing from a
standpoint that both parties share. Thus, seen in the light of the value
of the thing taken or used, the defendant’s infringement of the plain-
tiff ’s right becomes capable of a remedy in accordance with corrective
justice. By awarding the value of the use, a court reverses the injustice
that consists in the use. Similarly, by awarding the value of the thing
alienated, a court reverses the injustice that consists in the alienation.
Also in accordance with corrective justice is the owner’s option to
recover the wrongdoer’s gains from selling the thing at a higher price
than the market. Because value arises from specific needs and wants, it
includes the possibility of a purchaser who is willing to pay more than
the market price. This possibility is as fully within the owner’s entitle-
ment as the thing itself and its value. For such a possibility to be the
owner’s, the payment that happens to realize the possibility must also
be the owner’s. Otherwise, the owner would have a possibility that is
juridically incapable of fulfillment—which is no possibility at all. A
lawyer may justify this conclusion by saying that it does not lie in the
mouth of the wrongdoer to deny that the owner could have made the
sale. What such a formulation points to, however, is not the empirical
likelihood that the owner would have made this sale, but the irrele-
vance of who made it given that ownership carries with it an entitle-
ment to the proceeds. Of course, the recovery of the proceeds remains
merely an option that the owner need not exercise when the proceeds
are less than the market value. This option is the continuation of every
owner’s entitlement either to retain the thing owned and its value or
to dispose of it for the price that a willing buyer will pay. Thus, con-
sistently with corrective justice, the plaintiff ’s option replicates at the
remedial stage the content of the plaintiff ’s substantive right.
In sum, gain-based damages are justified where there are dealings
in—use or alienation of—another’s property. Because property
includes the possibility of gains, the plaintiff may as of right reclaim a
gain realized through the defendant’s use or alienation of the property.
Contrary to the suggestion of Phillips v. Homfray, the distinction
132 corrective justice

between use and alienation makes no difference in principle for the

availability of gain-based damages. These damages quantify the value
of the plaintiff ’s right, so that when the defendant encroaches on this
right, gain-based damages respond to and undo the injustice between
the parties.

4. Indirect benefits
In denying the plaintiff ’s claim to sue for the unauthorized use of his
passageways in Phillips v. Homfray, Lord Justice Bowen emphasized that
the gain realized through the wrongdoer’s trespass was an indirect or
negative benefit that consisted in the expense saved from not paying a
wayleave rent. The significance of this for Lord Justice Bowen was that
the gain from the wrong did not increase, but merely avoided decreas-
ing, the wrongdoer’s estate. As noted above, this distinction makes no
sense from a gain-based standpoint. Nonetheless, I suggest that Lord
Justice Bowen was struggling—not without insight—with the neces-
sity to distinguish recoverable from irrecoverable gains. This necessity
is present even if one acknowledges, as Lord Justice Bowen did not,
that the plaintiff should be able to recover the defendant’s gain from
the wrongful use of property. For then the question arises: what counts
as the relevant benefit for purposes of assessing the gain-based dam-
ages?
The case of Olwell v. Nye and Nissen23 is an apt illustration of the
need to be attentive to this issue. After selling his egg-packing business
to the defendant, the plaintiff stored his egg-washing machine in an
adjacent space. Subsequently, without the plaintiff ’s knowledge or
consent, the defendant took the machine out of storage and began to
use it.When the plaintiff discovered this, he offered to sell the machine
to the defendant, but the negotiations fell through. The plaintiff then
sued. The Washington Supreme Court held, following the waiver of
tort cases, that since the defendant had benefited from his wrong, the
plaintiff could elect to sue for restitution. Although the case involved
unauthorized use of another’s property, Phillips was effectively rejected
in favor of the proposition from the Restatement of Restitution that
“[a] person confers a benefit on another . . . not only where he adds to
the property of another, but also where he saves the other from

23
Olwell v. Nye, 173 P.2d 652 (Wash. 1946).
 gain-based damages 133

expense or loss.”24 The court then approved an assessment based on

the expense that the defendant would have incurred had the eggs
been washed by hand during the period that the machine was in use.
The court’s reasoning seems attractive. An action for loss-based
damages runs into the difficulty that the defendant’s conduct neither
damaged the machine nor, since it was in storage, deprived the plain-
tiff of the income that it might generate. Gain-based damages, in con-
trast, allow reference to the defendant’s enrichment regardless of what
the plaintiff lost. The enrichment includes the saving of expense. Here
the defendant was saved the expense of having the eggs washed by
hand. Therefore the damage award is for the cost of having the eggs
washed by hand.
Yet something has gone wrong in the court’s analysis. In the nego-
tiations for the sale of the machine, the plaintiff had asked for $600
and the defendant had counter-offered $25. The court, after consider-
ing the number of hours it would take to wash the eggs and the hour-
ly wages of the washers, awarded $900. Remarkably, the plaintiff
recovered 50 per cent more from the use of the machine than the
highest price he wanted from its sale.25
The court committed two interrelated conceptual errors. First, it
rejected the correct formulation of the benefit. The defendant had
argued that the damages “should be based on the use or rental value of
the machine.” The court thought this measure of damages was unsuit-
able because it was loss-based and the plaintiff had suffered no loss
from the defendant’s use of a machine that was not in the stream of
commerce. However, as we have seen, damages geared to the value of
the use can be gain-based as well as loss-based. One usually may char-
acterize the avoidance of rental costs indifferently as an expense saved
by the defendant or as income lost by the plaintiff. Even if on the
Olwell facts the loss-based characterization is implausible, the gain-
based characterization remains pertinent, since the opportunity for
gain is within the owner’s entitlement. Second, in focusing on the
expense of hand-washing the eggs, the court accepted an incorrect
formulation of the benefit. It may well be the case that without the
machine the defendant would have had the eggs washed by hand. But

24
§ 1(b), at 12.
25
The full facts of the case were even more egregious. The expense of hand-washing the
eggs came to $1,560, but that amount turned out to be more than the plaintiff had claimed,
and therefore was reduced on appeal.
134 corrective justice

that is no concern of the plaintiff. The plaintiff ’s only interest in the

defendant’s egg-washing operation is in the use of this particular
machine, not in how the defendant would have operated his business
without it. Therefore, the court should have based the calculation of
damages on the value of the use of the machine.
Corrective justice illuminates the court’s errors. Corrective justice
requires that the remedy undo the inconsistency on the part of the
defendant with the plaintiff ’s right. The role of damages in Olwell is to
make good the breach of the defendant’s duty not to violate the plain-
tiff ’s right. Because the remedy should reflect the duty correlative to
the plaintiff ’s right, basing the damages on the cost of hand-washing
the eggs implies that the defendant was under an obligation to the
plaintiff to wash the eggs by hand. This is absurd. The plaintiff had a
right in the machine but no right to have the defendant hand-wash
the eggs. The only relevant duty that the defendant owed the plaintiff
was not to use the machine. Accordingly the damages should have
been set at the value of the use as reflected by the rental value of the
machine.
One should observe that this criticism of calculating the damages
by reference to the cost of hand-washing the eggs involves a concep-
tual point, not an empirical one. The criticism does not suppose that
the defendant was in fact unlikely to have had the eggs hand-washed.
The point is rather that, whatever the alternatives to using the
machine and however probable their employment, none of them
forms the basis for calculating the damages, because in principle none
is directly relevant to the injustice between the parties. Assume the
situation most favorable to the plaintiff , that in the absence of the
machine there was no alternative to hand-washing the eggs if the
defendant was to stay in business. Such an absence of alternatives pre-
sumably would have increased both the value of the machine and the
value of the machine’s use. The basis of the gain-based damages, how-
ever, would still be the value of the use (increased by the necessity for
hand-washing without it) rather than the cost of the hand-washing as
such.
These deficiencies of the Olwell case illustrate a general point. Gain-
based damages, like other remedies in private law, must correct the
injustice that the plaintiff has suffered at the defendant’s hand. Such
damages, accordingly, must correspond to the elements constitutive of
the juridical relationship between the parties. Factors absent from the
law’s conceptualization of the defendant’s duty to the plaintiff, and
 gain-based damages 135

therefore external to that relationship between the parties, ought not

to be the basis for the calculation of damages. If, as in Olwell, the
injustice consists in the unauthorized use of the plaintiff ’s property,
the damages are to be calculated with reference to the value of the
use. Since the alternatives to that use are external to the juridical rela-
tionship and their performance is not a duty owed to the plaintiff, the
benefits of the savings from not having recourse to those alternatives
also are not owed to the plaintiff.
This analysis suggests that Lord Justice Bowen’s reference in Phillips
v. Homfray to indirect benefit is not completely off the mark. He was
aware that the saving of expense does not in and of itself constitute a
benefit that invariably gives rise to gain-based damages. However, he
formulated indirect benefits in terms of a contrast with benefits that
increase the wrongdoer’s assets. He ought to have formulated these
benefits in terms of a contrast with benefits, including those involving
the saving of an expense, that are within the ambit of the plaintiff ’s
entitlement.
Understood in this way, the notion of indirect benefits applies to
other situations besides the unauthorized use of property. The law of
nuisance, for example, imposes an obligation not to interfere with the
use and enjoyment of another’s property. Take the example of a pol-
luter who fails to install anti-pollution equipment costing x and causes
discomfort assessed at y, and assume x is greater than y. The plaintiff
cannot recover x. The defendant is under a duty to avoid causing the
plaintiff the discomfort, which is quantified at y. How the plaintiff
achieves this end is no business of the defendant. Of course, if the
defendant had made expenditure x, the plaintiff would not have suf-
fered the discomfort; but the tort consists in causing the discomfort,
not in saving the expense that would have avoided the discomfort.26
The same applies to negligence law, where the defendant is under a
duty not to create an unreasonable risk. Here recovery of the defend-
ant’s benefit is almost unheard of, even though the defendant might
have avoided the unlawful risk by undertaking certain expenditures.
Assume that the risk, which caused the plaintiff damages of y, could
have been avoided had the defendant expended x on precautions, and
that x is greater than y. Again, the plaintiff does not recover x. It is true
that the defendant realized a benefit by not taking the precautions and
that the injury would not have occurred had those precautions been

26
Cf. Kirk v. Todd [1882] 21 Ch. D. 484 (C.A.).
136 corrective justice

taken. The duty owed, however, was to avoid imposing the risk, not to
undertake the expenditures.
Distinguishing the duty from the contingent means to avoid violat-
ing it fits within corrective justice in three related ways. First, correct-
ive justice highlights the juridical connection between the plaintiff
and the defendant. Because the juridical connection is forged through
the correlativity of the plaintiff ’s right and the defendant’s duty, the
remedy is determined by the precise contours of the right and its cor-
relative duty. The steps that the defendant can take to avoid violating
the duty owed to the plaintiff do not in themselves juridically connect
the defendant to the plaintiff . They are merely factual possibilities that
pertain to the defendant’s situation.
Second, in emphasizing the plaintiff ’s right and the defendant’s cor-
relative duty, corrective justice sets its face against consequentialist
understandings of private law. The consequentialist assumes that what
matters is the state of affairs that exists at the end of the day. Accord-
ingly, from the consequentialist standpoint there is no reason to distin-
guish the defendant’s duty from actions that produce the situation that
will obtain if the duty is discharged. Corrective justice, in contrast,
focuses on the normative dynamics internal to the interaction between
the right-holder and others. What matters is not the situation at the
end of the day, but whether the defendant acted wrongfully with
respect to the plaintiff ’s right.
Third, the duties of corrective justice result from the moral capacity
of rights to put others under obligations. The fact that these are duties
relative to rights rather than to other possible normative categories
(for example, maximizing human welfare or fostering virtue) means
that these duties are negative in nature: they function as prohibitions
against wrongful interferences with another’s rights rather than as
positive commands to do particular acts, even if those acts promote
another’s good. The common law reflects the essentially negative
nature of private law duties by denying the existence of duties in situ-
ations of nonfeasance;27 in such situations the actor’s behavior, how-
ever morally reprehensible, merely fails to promote another’s interest
but does not interfere with another’s rights. Accordingly, tort and other
private law duties are defined negatively in terms of non-interference

27
This statement is subject to exceptions that are not relevant here and that, in any case,
do not undermine the principle at stake. See Ernest J.Weinrib, The Idea of Private Law (1995),
at 153–54.
 gain-based damages 137

with rights rather than positively in terms of particular acts that are
obligatory.28 In view of this, a requirement in the examples discussed
here to do a particular act, such as having eggs hand-washed or install-
ing antipollution equipment or undertaking the burden of precau-
tions, would be inconsistent with the structure of private law norms.
Thus, the line between wronging the plaintiff and failing to do
something that would result in the plaintiff ’s not being wronged, far
from being adventitious, is both essential to the definition of duty
within corrective justice and confirmed by the organization of the
common law.
The key, then, to the availability of gain-based damages lies in align-
ing the remedy with the injustice it corrects. Through the assessment
of damages, the law transforms the plaintiff ’s right and the defendant’s
correlative duty not to interfere with that right into their monetary
equivalent. Thus, what the defendant owes the plaintiff at the remedial
stage of their relationship corresponds to what the defendant owed
the plaintiff at the stage of conduct. One must distinguish, however,
between the duty itself and the measures that would avoid a breach of
the duty. Only the former defines the legal relationship between the
parties; the latter are merely the contingent ways to prevent trans-
forming that relationship into a wrongful one. The defendant may
owe the plaintiff a duty not to convert the plaintiff ’s egg-washing
machine or not to commit a nuisance or not to create an unreasonable
risk; and it may well be true that hand-washing the eggs or purchasing
antipollution equipment or undertaking precautions would in the cir-
cumstances have avoided the breach of these duties; and it may also be
true that the defendant is better off for not having done these actions.
But these latter actions were not constitutive of the duties incumbent
on the defendant by virtue of the plaintiff ’s right. Consequently, what
the defendant gained by not performing these actions also is not with-
in the plaintiff ’s right.
The damages for the value of using another’s property stand on a
different footing. The reason that the court in the Olwell case should
have assessed damages at the rental value of the egg-washing machine
is not because renting the machine was obligatory on the defendant.
The defendant’s only duty—and the duty that was breached—was not

28
Even contractual obligations are not an exception: the promisor is obligated to perform
the particular acts specified by the contract only because the contract has made the perform-
ance of those acts a right of the promisee.
138 corrective justice

to use the machine. Damages set at the rental value do not quantify a
contingent means of preventing the wrong from occurring. Rather,
they quantify the wrong that did occur. By owning the machine, the
plaintiff is also entitled to the value that could be realized from using
it. Since that value is an incident of the plaintiff ’s proprietary right, it
is also an element in the duty correlative to that right.
These reflections provide theoretical support for what many com-
mentators have observed, that dealing with another’s property is the
paradigmatic case for gain-based damages. From the standpoint of
corrective justice, dealing with another’s property is paradigmatic
because the idea of property weaves the plaintiff ’s entitlement to gain
into the fabric of the juridical relationship with the defendant. In
other situations the gains that a wrongdoer might realize are at best
indirect, forming not an element of the plaintiff ’s right but a benefit
realized from the non-performance of the duty correlative to it.29

5. Innocent and willful wrongdoing

I mentioned at the outset of this chapter that, because it focuses on the
bipolar relationship between the parties, corrective justice is unrecep-
tive to justifications like punishment and deterrence that consider the
defendant independently of the plaintiff. To some, this starting point
may appear peremptorily to cut off a promising line of exploration,
that gain-based damages respond to deliberate or outrageous conduct.
Such conduct is intuitively offensive to our moral sensibilities, and
ideas like punishment and deterrence seem particularly well suited to
the analysis of its legal consequences. Conversely, my emphasis on

29
The idea that a gain-based award responds to an interference with the plaintiff ’s propri-
etary right receives its logical development in the German Eingriffskondiktion, which von
Caemmerer argued should be classed with such property-protecting devices as the vindicatio;
above n. 5, at 353. Strictly speaking, what matters for the Eingriffskondiktion is not the wrong-
fulness of the defendant’s interference but the scope and purpose of the plaintiff ’s right, and
thus whether the defendant’s gain was within that scope and purpose; Markesinis et al., above
n. 5, 745. The common law, in contrast, generally protects property not directly but through
the law of torts. Accordingly, for the common law the issue of gain-based damages is usually
thought to arise in the context of restitution for wrongs.The decisive question should, none-
theless, remain whether the defendant’s gain rightfully belongs to the plaintiff as an incident
of the plaintiff ’s property. This may be the case even in situations in which no tort claim
against the defendant is available; see Daniel Friedmann, “Restitution for Wrongs: the Basis
for Liability,” in Restitution Past, Present and Future: Essays in Honour of Gareth Jones, ed. W. R.
Cornish et al. (1998), 133, 133–38.
 gain-based damages 139

rights and their correlative duties may seem less felicitous, since the
rights of the plaintiff do not seem to be affected additionally by the
deliberateness with which the defendant violated them.
To a certain extent the law reinforces these doubts about my dis-
missal of punishment and deterrence.The law presents instances where
the deliberateness or innocence of the wrongdoing affects the plain-
tiff ’s remedy. One set of instances (discussed in this section) concerns
dealings in another’s property, where gain-based damages are uncontro-
versial but the extent of the damages depends on whether the wrong-
doing was innocent or deliberate. Another set of instances (discussed
in the following section) concerns wrongs that are not proprietary, so
that gain-based damages seem unavailable on the approach suggested
above, but where additional damages are nonetheless awarded on a
gain-based or punitive basis. Can such instances be comprehended
within corrective justice?
An example of the law’s differentiating innocent from willful
wrongdoing is the defendant’s wrongful removal from the plaintiff ’s
realty of some valuable resource, such as timber or minerals, that the
plaintiff intends to exploit. The severance enhances the value of the
resource, and the general (though not invariable) rule is that the plain-
tiff recovers the enhanced value.30 Such recovery is a straightforward
application of gain-based damages to dealings in another’s property,
and poses no special theoretical problem. Of particular interest, how-
ever, is the relevance of the trespasser’s willfulness to the question of
whether the defendant is credited with the expense of severing the
resource or of otherwise making it more marketable. An innocent
trespasser (one who mistakenly thought that taking the resource was
not a violation of the plaintiff ’s right) is allowed to deduct such
expenses; a willful trespasser is not.
The higher damages that willfulness attracts, however, do not sig-
nal the presence of punitive considerations in the law’s treatment
of the trespasser.31 The law can be explained through the standard

30
A leading House of Lords case, however, Livingstone v. Rawyards Coal, 5 App. Cas. 25
(1880), awarded the owner of land from which coal was taken only the royalty value of the
coal while it was in the ground. The court emphasized the special features of the case: both
parties thought that the trespasser had the right to take the coal, and the owner’s plot was so
small and so completely surrounded by the trespasser’s holdings, that only the trespasser
could have extracted the coal.
31
Cf. James Edelman, Gain-Based Damages: Contract, Tort, Equity and Intellectual Property
(2002), 137–38, who claims that these cases exemplify what he calls “disgorgement damages,”
which, in his classification, are always based on deterrence (ibid., 83).
140 corrective justice

conceptual tools in the law of unjust enrichment—itself consistent

with corrective justice.32
From the standpoint of the law of unjust enrichment the rules con-
cerning severed resources make sense. As noted above in section 3, the
proprietor has an entitlement to any increase in the value of what is
owned.With respect to the enhanced value, the innocent and the will-
ful trespasser stand on the same footing, since even the fact that the
trespass was innocent does not create for the trespasser a proprietary
interest. Their situations are different, however, with respect to the
expense of severing the resource. By undertaking this expense, the
willful trespasser improved property known to belong to someone
else. Under the law of unjust enrichment, the improver of property
known to belong to another is considered to have willingly taken the
risk of losing the value of the benefit conferred on the owner.33 The
owner’s retention of this benefit is, therefore, not an injustice. Accord-
ingly, the owner who reclaims the property does not have to credit
the willful trespasser with the expenditure that improved it. In con-
trast, the expenditure by the trespasser who mistakenly thinks that he
or she is entitled to sever the resource cannot be construed as a volun-
tary risk-taking. Rather, by innocently anticipating the plaintiff ’s own
intended severance and exploitation of the resource, the expenditure
can be considered an incontrovertible benefit—that is, a non-gratuit-
ous enrichment that, because of the practical inevitability of the sever-
ance, can be returned without unduly interfering with the proprietor’s
autonomy.34 Accordingly, the proprietor cannot claim from the inno-
cent trespasser the enhanced value clear of the expenditure necessary
to produce it.35
These rules disclose no trace of a punitive impulse.36 They merely
apply, in the context of increased value through trespass, the usual
considerations at play in the restitution of unrequested benefits under
the principle of unjust enrichment.The key questions are whether the

32
See below, chapter 6.
33
Peter Birks, An Introduction to the Law of Restitution (1989), 102–3.
34
Ibid., at 116–24. Birks suggests that the argument of incontrovertible benefit is available
to any defendant sued for damages, because then the claim turns the improvement into
money and prevents the defendant from subjectively devaluing the benefit (ibid., at 122). If
this suggestion is correct, the plaintiff in a severance case will have to credit the defendant’s
expenses even if the plaintiff did not intend to exploit the resource.
35
Hugh Evander Willis, “Measure of Damages when Property is Wrongfully Taken by a
Private Individual,” (1908–09) 22 Harv. L. Rev. 419, 425–26.
36
But see Lord Diplock’s observation in Broome v. Cassell [1972] A.C. 1027, 1129.
 gain-based damages 141

benefit conferred by the trespasser plausibly can be regarded as gratuit-

ous, and, if it cannot, whether restitution would be consistent with the
owner’s proprietary right. Because one cannot unilaterally create an
obligation by marking another person out for a benefit for which one
expects recompense, the law understands the willful trespasser’s know-
ing improvement of another’s property as the manifestation of a dona-
tive intent. Consequently, justice between the parties allows the
proprietor to keep what has thus been given gratuitously. In contrast,
the innocent trespasser lacks a donative intent, and can recover the
value of the benefit if the benefit to the plaintiff is incontrovertible in
light of the plaintiff ’s anticipated exploitation of the resource. Then
the plaintiff cannot retain what the defendant did not intend to give.
In both situations the law works out the circumstances under which
the enrichment of one right-holder at the expense of the other con-
stitutes the doing and suffering of an injustice. The idea of punish-
ment, with its one-sided focus on the defendant, is entirely absent.
Indeed, from an explanatory standpoint the corrective justice
account is superior to the invocation of punishment or deterrence
with respect to the treatment of the expenditures of the willful tres-
passer. A notable feature of the treatment of willful trespass is that the
trespasser is disqualified from deducting the expenses, whatever they
are. No attempt is made to calibrate the unrecouped costs to the tres-
passer’s desert. From a punitive standpoint, this is odd, since the
amount of the supposed punishment is not necessarily related to the
trespasser’s culpability.37 From a deterrence perspective also, this
appears difficult, though perhaps some elaborate economic story could
be told of why, despite the overall gain in utility through the creation
of enhanced value, neither disgorgement of the net gain nor the impos-
ition of an additional penalty greater or less than the amount of the
expense would create the proper incentives. Corrective justice avoids
these difficulties because it looks to the normative implications of the
parties’ interaction without orienting that interaction to any external
end that might be forwarded by punishment or deterrence. Corrective
justice calls for no assessment of culpability apart from the fact of will-
fulness, because that fact in itself allows the implication that the
defendant’s conduct is donative, given his awareness of the plaintiff ’s
right. Similarly, corrective justice is not concerned with deterring
potential trespassers, but with making the remedy correspond to the

37
Dan B. Dobbs, Law of Remedies, 2nd ed. (1993), 511.
142 corrective justice

wrong. For corrective justice the expenses as such, regardless of their

relationship to any other policy, pertain to the interaction of the par-
ties, and so the remedial consequence of the wrong—the absence of
the trespasser’s entitlement to recoup them—can attach to the expens-
es as such.

6. Property-like rights
So far I have emphasized the significance of property for a corrective
justice approach to gain-based damages for tort. Since such an
approach highlights the correlativity of the injustice done by the
defendant and the injustice suffered by the plaintiff , it conditions lia-
bility on the requirement that the defendant’s conduct be wrongful
with respect to the plaintiff ’s right. If the plaintiff is to recover gain-
based damages, that right must include an entitlement to the profit
from whatever embodies the right. Proprietary rights contain this
entitlement.
Strictly speaking, a proprietary right has two features. First, a pro-
prietary right can be asserted against the world, and therefore the right
carries with it a correlative duty, incumbent on everyone else, not to
interfere. The proprietor’s entitlement to the profit from what is
owned derives from the power to determine the object’s use, includ-
ing the conditions under which it can be alienated, to the exclusion
of everyone else. Since the proprietor must agree to the terms on
which the object can enter the stream of commerce, the proprietor
also owns whatever can be realized through use or alienation.
Second, the subject matter of a proprietary right has to be capable
of being acquired and alienated. For example, an incident of a person’s
bodily integrity is not the subject of a proprietary right. One’s body is
not what one owns but what one is; it is the organism through which
humans as self-conscious and purposive beings express themselves in
the world. One does not come to be entitled to one’s body by any act
of acquisition, and one cannot alienate it to someone else. The right
to one’s body is so intimately connected to the person whose body it
is that it lacks the moral possibility of being externalized and passing
into the possession of someone else. Similar considerations apply to
other aspects of one’s dignity—to what Hegel compendiously termed
“those goods, or rather substantive characteristics, which constitute
my own private personality and the universal essence of my self-
 gain-based damages 143

consciousness.”38 Such interests in physical integrity and dignity are, of

course, legally protected with the status of rights, but they are not
considered to be rights of a proprietary kind.
Under certain circumstances, gain-based damages are justified even
if either or both of these features are absent. Although gain-based
damages do not then respond to the violation of what is strictly speak-
ing a proprietary right, the relationship between the parties can give
rise to an interest sufficiently property-like to allow this kind of
award.39 These relational property-like wrongs can be grouped into
two broad categories. In the first category, a gain-based remedy
emerges from the pre-existing relationship between the parties, so that
the remedy is available only against the defendant and not against the
whole world. In the second, the remedy emerges as a response to the
defendant’s particular conduct.
The violation of a fiduciary duty is the paradigmatic example of the
situation in which a gain-based remedy can emerge from the objective
nature of the relationship. From the perspective of corrective justice, a
fiduciary relationship reflects the Kantian idea that private law as a sys-
tem of rights supposes persons to be ends in themselves rather than
means to the ends of others.40 Accordingly, a relationship such as that
between fiduciary and beneficiary, the legal structure of which makes
one person’s interests entirely subject to another’s discretion, must have
as one of its incidents the duty of loyalty owed by the latter to the former.
The fiduciary’s duty of loyalty then becomes for purposes of this rela-
tionship an entitlement of the beneficiary. Since the meaning of this
duty of loyalty is that the fiduciary cannot profit from the relationship,

38
G. W. F. Hegel, Philosophy of Right, tr. T. M. Knox (1952), § 66. Cf. Immanuel Kant, The
Metaphysics of Morals, tr. Mary Gregor (1991), [6:237] (describing innate right as something
“belonging to every man by virtue of his humanity”).
39
The danger, of course, is that the possibility of calling something “property-like” may
appear to provide a convenient black box into which to stuff the residual instances that do
not fit what a property-based approach requires. However, I think that the property-like
aspects of these situations are salient enough to bring them within the approach I have sug-
gested. Peter Birks writes that “[i]t is not helpful . . . to say that gain-based damages should
always be available for a ‘proprietary’ tort. The difficult questions will merely be transferred
to the definition of property.” Peter Birks, Civil Wrongs: A New World (1990–1), 98. While the
fear is well founded, the truth is that, as with every interesting legal issue, difficult questions
are unavoidable. The basic issue is whether property is the appropriate concept, not whether
the concept is completely determinate in its application. See Weinrib, above n. 5, at 222–27.
In any case, property seems to be a more manageable criterion than the one Birks proposed,
namely that one ask whether there is sufficient justification for giving the plaintiff a windfall
and for tolerating the suppression of economic activity.
40
Kant, above n. 38, at [6:236].
144 corrective justice

gains can be regarded as the material embodiment of the breach of

duty—what the fiduciary has, as it were, sold out the duty for—and the
beneficiary is as entitled to these profits as he or she was to the duty for
which they were exchanged. Courts occasionally refer to the opportu-
nity to profit from the relationship as the beneficiary’s “property”41
though, because it entails the right to exclude only the fiduciary and
not the whole world, it “is not property in the strict sense.”42 Seen in
this light, the fiduciary’s liability to disgorge profits is not an example of
a policy of deterrence impacting the relationship from the outside, but
is rather the remedial consequence that reflects the nature of the obliga-
tion owed by the fiduciary to the beneficiary.
The other property-like wrongs are those characterized by action of
the defendant that implicitly or explicitly treats the plaintiff ’s right as
an asset whose value the defendant can appropriate.43 For things that
can be acquired and alienated, value belongs to the owner as an aspect
of property; as noted above in section 3, the wrongdoer who deals with
such things can be liable for the market value or for the realized gains.
In the case of property-like wrongs, the same liability is available for
wrongs done with respect to things, like physical integrity, that cannot
be acquired and alienated. Compared to the victim of a true propri-
etary wrong, the plaintiff is not placed in a worse position by virtue of
the fact that the right was too intimately connected with the plaintiff ’s
being and dignity even to rank as proprietary. Because the defendant
acted with knowledge of the plaintiff ’s right and with the intent to
appropriate its value, the law ascribes a proprietary quality to the right
so far as the relationship between the defendant and the plaintiff is
concerned. In doing this, the law merely holds the defendant to the
implications of his or her own conduct. Since the plaintiff ’s right was
treated as a commodity whose value was available to the defendant, the
plaintiff is allowed to recapture the gain that was realized through it.
Thus, once the wrong is construed as a property-like one for purposes
of the parties’ relationship, the plaintiff has available the gain-based
damages that attend a dealing with another’s property.
Consider the example posed by Professor Birks of the thug hired to
beat someone up.44 The thug cannot resist a claim by the victim of

41
See, e.g., Boardman v. Phipps [1967] 2 A.C. 46, 107, 115 (H.L.).
42
Ibid., at 102.
43
See Peter Benson, “The Basis for Excluding Liability for Economic Loss in Tort Law,”
in Philosophical Foundtions of Tort Law, ed. David Owen (1995), 427, 457.
44
Birks, above n. 33, at 319.
 gain-based damages 145

the beating for gain-based damages on the basis that the wrong was a
bodily injury and thus too closely connected to the victim to count as
property. In these circumstances the battery is part of a process of illegit-
imate commodification. Having treated the plaintiff ’s bodily integrity
as an item that the thug was in effect selling for a price, the thug
cannot take refuge in the argument that bodily integrity is really an
inalienable pearl beyond price. Although bodily integrity is not in
itself property, the thug’s relationship to the defendant’s bodily integ-
rity has become property-like through the thug’s conduct. Conse-
quently, the thug is liable for gain-based damages, as he would be for
any dealing with another’s property.
The idea that special remedial consequences attend the defendant’s
profiteering from another’s right finds its most general expression in
Lord Devlin’s second category of punitive damages, which applies to
conduct calculated by the defendant to make a profit that may exceed
the compensation payable to the plaintiff .45 Lord Devlin ascribed to
this category of damages the admonitory function of teaching the
wrongdoer that tort does not pay, rather than the restitutionary func-
tion of causing the disgorgement of the gain. Nonetheless, given that
disgorgement is a way of preventing the wrongdoer from profiting,
the two functions are not easily kept separate.46 Lord Devlin himself
noted the connection between this category of punitive damages and
the misappropriation of property when, instancing defamation, he
remarked that “no man should be allowed to sell another man’s repu-
tation for profit.”47 Indeed, gain-based damages have been recom-
mended as a way of introducing greater specificity into the punitive
idea.48 And even in jurisdictions that do not follow the approach to
punitive damages set out by Lord Devlin, courts may factor gain-based
considerations into their punitive awards.49
Despite their similarity, gain-based damages differ from punitive
damages in their conception of the plaintiff ’s role. Commenting on

45
Rookes v. Barnard [1964] A.C. 1129, 1227 (H.L.).
46
In Broome v. Cassell [1972] A.C.1027, 1129 (H.L.), Lord Diplock notes the analogy of
this category of punitive damages to the restitution of an enrichment.
47
Ibid.
48
Jeff Berryman, “The Case for Gain-based Damages over Punitive Damages: Teaching
the Wrongdoer that Tort Does Not Pay,” (1994) 73 Canadian Bar Rev. 320; John Glover,
“Gain-based Principles in Tort: Wrongful User of Property and the Exemplary Measure of
Damages,” (1992) 18 Monash U. L. Rev. 169.
49
Austin v. Rescon Construction [1989] 57 D.L.R. (4th) 591 (B.C.C.A.); Huff v. Price
[1990] 76 D.L.R. (4th) 138 (B.C.C.A.).
146 corrective justice

this category of punitive damages, Lord Diplock once observed that,

because their purpose is merely to prevent the wrongdoer from
obtaining a reward for his wrongdoing, “the plaintiff is the accidental
beneficiary of a rule of law based on public policy rather than on the
reparation of private wrongs.”50 Punitive damages are oriented toward
striking the gain from the hand of the defendant; the plaintiff is noth-
ing but the contingent recipient of a windfall. Gain-based damages, in
contrast, allow us to focus on the juridical relationship between the
parties. Even in the property-like cases, we can work through the
implications of the defendant’s wrongful treatment of the plaintiff to a
remedy that reverses that wrong. Just as the plaintiff has a right not to
be wronged, so the plaintiff has an entitlement to the damages that
undo that wrong.

7. Conclusion
The reawakening of interest in restitution has given gain-based dam-
ages a new salience. Now that the significance of restitution for pri-
vate law is widely recognized, the question arises: what is the
conceptual framework within which gain-based damages are to be
understood?
To this question, corrective justice supplies an old answer. Correct-
ive justice treats the defendant as the doer and the plaintiff as the suf-
ferer of the same injustice. Corrective justice therefore highlights the
correlativity of right and duty that characterizes the norms connect-
ing the parties. From the perspective of corrective justice, the point of
a legal remedy is to undo that injustice, and so the remedy must mir-
ror the structure of the injustice. Corrective justice thereby ties both
the defendant to the plaintiff and the remedy to the injustice. Under
this approach, gain-based damages are available when the potential for
gain is an incident of the right that the wrongdoer violated. Hence
the paradigmatic case for the availability of gain-based damages is the
defendant’s violation of a property (or property-like) right held by the
plaintiff.
With this understanding of gain-based damages comes a repudi-
ation of the notion that they are occasions for the promotion of social

50
McCarey v. Associated Newspapers Ltd. (No. 2) [1965] 2 Q.B. 86, 107 (C.A.) (Diplock
L.J.).
 gain-based damages 147

purposes extrinsic to the relationship between the parties. Purposes

such as punishment or deterrence (or broader purposes such as the pro-
motion of economic efficiency or of other goals), even if they other-
wise seem desirable, cannot be accommodated to the correlative
nature of private law justifications and, therefore, cannot explain the
most characteristic and pervasive features of private law. Thus, in this
context as in others, corrective justice breaks free of the instrumental-
ist modes of explanation that over the last decades have so brilliantly
obscured private law.
 5
Punishment and Disgorgement
as Contract Remedies

1. Introduction
For corrective justice the remedy corrects the injustice suffered by the
plaintiff at the defendant’s hands. This chapter examines the implica-
tions of this simple statement for contract damages. The focus will be
on two kinds of damage award for breach of contract: punitive dam-
ages and damages that require the disgorgement of gains. The fact that
over the last decades these two kinds of damage award have received
notable elaboration by the highest courts in Canada, England, and
Israel1 attests to the continuing relevance of the issues that they raise.
In private law the idea that compensation is an appropriate remedy is
generally accepted.The award of compensation reflects the plaintiff ’s entitle-
ment to recover at least the loss that the defendant’s wrongful act has caused.
More problematic is the issue of whether compensation is also the limit
of what the plaintiff can be awarded. Damages that go beyond compensa-
tion and aim at punishment or disgorgement operate in circumscribed
situations and are subject to special, often controversial, justifications.
Contract law, however, poses a special difficulty. Here the very
notion of compensation is uncertain and its primacy disputed. The
standard measure of damages for breach of contract is the expectation
measure, which puts the plaintiff in the position in which the plaintiff
would have been had the contract not been breached. In their classic
article on contract damages Fuller and Perdue denied that this measure,
which reflected the value of something that the promisee did not yet

1
Whiten v. Pilot Insurance Co. [2002] 209 D.L.R. (4th 257 (S.C.C.)); Attorney-General v.
Blake [2000] 4 All E.R. 385 (H.L.); Adras Building Material v. Harlow & Jones [1995] Resti-
tution L.R. 235 (Supreme Court of Israel, 1988).
 punishment and disgorgement as contract remedies 149

have, was compensatory.2 Expectation damages, they suggested, might

better be viewed as having the quasi-criminal purpose of penalizing
the promisor for breaching the contract.3 This suggestion raises the
possibility that a punitive impulse is present even in the most routine
award of contract damages. On this view, truly compensatory assess-
ments of contract damages are comparatively rare, whereas non-
compensatory damages merely extend and make more explicit the
non-compensatory policies already pervasive in contract damages.
At the heart of these issues about remedies lies the more fundamen-
tal issue: what is the nature of the right to contractual performance? So
far as corrective justice is concerned, the remedy is merely the contin-
uation at the remedial stage of the correlativity of right and duty that
defines the parties’ relationship. Accordingly, the first step to specifying
the plaintiff ’s remedy against the defendant is to identify the right that
contract law gives the plaintiff and the correlative duty that it lays upon
the defendant. I turn to this in section 2, presenting a contrast between
the function that Fuller and Perdue assign to the contract remedy and
Kant’s now largely forgotten treatment of contractual right. The Kant-
ian account also casts light, as I contend in section 3, on the inaptness
of requiring the disgorgement of gains resulting from contract breach,
despite the superficial attractiveness of preventing wrongdoers from
profiting from their wrongs. In section 4, I turn to punitive damages,
addressing first the preliminary question of how corrective justice and
punishment—and the institutions devoted to them—coexist and are
differentiated in a legal order based on rights. Finally in section 5, I dis-
cuss the difficulties that emerge from the elaborate but ultimately
unsatisfying recent attempt in Canada to work out a coherent treat-
ment of punitive damages for contract breach.

2. Contractual right
What, then, is the nature of a contractual right and how does an award
of damages undo the violation of that right?4 The basic rule of contract

2
L. L. Fuller and William R. Perdue, “The Reliance Interest in Contract Damages,” (1936)
46 Yale L.J. 52, 373.
3
Ibid., 61.
4
The most detailed contemporary application of corrective justice to contact law is found
in the work of Peter Benson; see especially Benson, “The Unity of Contract Law,” in The
Theory of Contract Law: New Essays, ed. Peter Benson (2001), 118.
150 corrective justice

damages is that damages are awarded on the expectation measure: the

plaintiff is to be put in the position that the plaintiff would have been
in had the contract not been breached. Is there the internal connection
that corrective justice requires between the injury to the promisee’s
contractual right and what the award of expectation damages restores?
In other words, are expectation damages consistent with corrective
justice?
This question was the starting point of the celebrated article on
contract damages by Fuller and Perdue, who answered it in the nega-
tive.5 In their view the purpose of corrective justice is “the mainten-
ance of an equilibrium of goods among members of society.”6 This
the law accomplishes by awarding compensatory damages “to heal a
disturbed status quo.”7 In the contracts context, corrective justice can
be seen to be at work in the protection accorded to the restitution
and reliance interests; the equilibrium of goods represented by the sta-
tus quo has been disturbed in the former both by a gain for the
defendant and an identical loss for the plaintiff, and in the latter by a
loss for the plaintiff. Expectation damages, they argue, are different.
Such damages protect a future expectancy—“something [the plaintiff]
never had”—rather than a loss already suffered. “[T]his seems on the
face of things a queer kind of ‘compensation.’”8 And so they contend
that “[i]n passing from compensation for change of position to com-
pensation for loss of expectancy we pass . . . from the realm of correct-
ive to that of distributive justice.”9
Having discarded corrective justice, Fuller and Perdue then locate
the rationale for expectation damages in considerations of policy.They
suggest that expectation damages are an effective means of protecting
the reliance interest. Expectation damages function not only as com-
pensation for reliance losses (for reliance can consist in loss of the
opportunity to enter other contracts) but also as a quasi-penal prophy-
laxis against breaches of contract that occasion reliance losses. More-
over, expectation damages promote and facilitate business agreements,
which in turn stimulate economic activity, especially within a credit

5
Above n. 3. My treatment of the article by Fuller and Perdue has been much influenced
by Peter Benson,“Contract,” in A Companion to Philosophy of Law and Legal Theory, ed. Dennis
Patterson, 2nd ed. (2011), 29, 30–34.
6
Above n. 2, 56
7
Ibid.
8
Ibid., 53.
9
Ibid., 56.
 punishment and disgorgement as contract remedies 151

economy. In this way expectation damages attest, in their view, to the

intertwining of legal institutions and the economic system.
The Fuller–Perdue account of expectation damages stands at the
confluence of two conclusions. The first is that corrective justice,
although appropriate for rectifying the gains and losses associated with
the restitution and the reliance interests, is inapplicable to the award of
expectation damages. The second is that expectation damages are to
be justified in terms of remedial policies concerning the indirect pro-
tection of the reliance interest and the promotion of commerce in a
credit economy. These two conclusions are related. Having rejected
the applicability to expectation damages of corrective justice, which
internally connects the injustice to the remedy, Fuller and Perdue have
recourse to considerations of remedial policy that present such dam-
ages as instrumental to the desirable social goals of protecting reliance
and facilitating business agreements.
The basic presupposition of this account is that corrective justice
does not operate in the absence of a disturbance of the status quo’s
equilibrium of goods among members of society. Unless there is a
loss (as occurs with detrimental reliance) or a gain that corresponds
to a loss (as when the restitution interest is in play), an award of
damages cannot be construed as the working of corrective justice.
Expectation damages, Fuller and Perdue argue, are not truly com-
pensatory: by breaching the promise, the defendant merely with-
draws a future good without inflicting a present loss. Only when the
plaintiff relies on the prospect of receiving this good and thereby
puts the future to some present detrimental use does the plaintiff
suffer a loss that grounds a claim for compensation.10 Of course, by
awarding expectation damages the law signals its willingness to treat
the promise as creating something of present value. But one cannot
deduce the justification of expectation damages as compensating for
the loss of a present value from their mere existence. Apart from
policies like the protection of reliance and the promotion of com-
merce (so Fuller and Perdue claim), there is no argument, independ-
ent of a circular appeal to the consequences that the law attaches to
a breach, for regarding the promise as creating a present right in the
expectancy.11 In and of itself, they assume, a contractual undertaking
does not suffice to do so.

10
Ibid., 59.
11
Ibid., 59–60.
152 corrective justice

Crucial to this reasoning is the idea that contract itself does not
transfer the subject matter of the contract to the promisee. If contract
did so, expectation damages would lose their mystery: given that the
subject matter of the contract would belong to the promisee, its value
would of course determine the level of compensation owed when the
promisor withholds it through breach. Since Roman times, however,
the law has distinguished between contract and conveyance.12 At com-
mon law, only specific kinds of contracts, such as contracts of sale,
effect an immediate transfer of title.Thus an agreement to sell (as con-
trasted with a contract of sale) gives the purchaser not a property
interest in the object to be sold, but only the expectation of owning
such an object in the future. Yet if the vendor breaches, the purchaser
is nonetheless, under the rule of expectation damages, entitled to
the object’s value. This seems strange. Usually one’s entitlement to the
value of something stems from one’s ownership of the thing that has
that value. The rule of expectation damages thus presents the paradox
that the law, by requiring that the promisor make good the value
withheld through the breach of the contract, treats the promisee as
entitled to the object’s present value even though it does not yet regard
the promisee as owner of the object itself.
To resolve this paradox, it is worth considering Immanuel Kant’s
account of the distinction between in rem and in personam rights. This
account provides a response to the kind of position subsequently put
forward by Fuller and Perdue.13 In Kant’s understanding, law is a sys-
tem of universal reciprocal freedom that includes the freedom to
acquire rights to what is external to the interacting parties as self-
determining agents. Such external rights mark a relationship between
the right-holder and the object of the right that imposes a correlative
duty on others. The different kinds of external rights reflect the cat-
egories of the understanding that deal with relations. Accordingly, the
distinction between rights in rem and rights in personam expresses
juridically the epistemological distinction between the relational cat-
egories of substance and causality respectively.14 An in rem right relates
the holder of the right to substance—that is, to something that all
others are obligated to leave intact; an in personam right relates the

12
Barry Nicholas, An Introduction to the Study of Roman Law (1962), 103.
13
The difficulty to which Kant was responding was not dissimilar to that raised by Fuller
and Perdue; see Helge Dedek, “A Particle of Freedom: Natural Law Thought and the Kant-
ian Theory of Transfer by Contract,” (2012) 25 Can. J. L. & Juris L, 313.
14
Immanuel Kant, The Metaphysics of Morals, tr. Mary Gregor (1996), [6:247].
 punishment and disgorgement as contract remedies 153

holder of the right to a causality—that is, to an act that a promisor is

obligated to perform because of the relationship with the promisee. A
relation to substance is necessary for an in rem right to be good against
the whole world; a relation to causality is necessary for an in personam
right to be good against a particular person.
Kant is explicit about the nature of a contractual right. It is not a
right to the subject matter of the contract. Nor is it a right to the situ-
ation that would result from the performance of the promised act.
Rather, it is a right merely to the performance of that act, to what
Kant calls “another’s choice to perform a specific deed.”15 Kant for-
mulates this important conclusion as follows:

By a contract I acquire something external. But what is it that I acquire?

Since it only the causality of another’s choice with respect to the perform-
ance he has promised me, what I acquire directly by a contract is not an
external thing but rather his deed, by which that thing is brought under my
control so that I make it mine.—By a contract I therefore acquire another’s
promise (not what he promised), and yet something is added to my exter-
nal belongings; I have become enriched (locupletior) by acquiring an active
obligation on the freedom and means of the other.16

Thus, Kant continues, what the promisee acquires through a contract

is not a right to a thing but a right against the specific person obli-
gated to perform the requisite act.17
This Kantian account of contractual entitlement provides a basis in
corrective justice for the expectation measure of damages. By breach-

15
Ibid., [6:402].
16
Ibid., [6:424].
17
In Kant’s view, the acquisition of a thing by means of a contract involves two conceptual
steps: the contract that makes a certain act (delivery of the thing) obligatory, and then the
delivery that accomplishes the transfer of property by putting the promisee into possession
of the thing (ibid., [6:424–26]). While it is true that the contract to deliver something makes
the promisee’s acquisition of the subject matter of the contract a “rightfully necessary result
of it” (ibid., [6:432], emphasis in original), that result is the consequence of the promisor’s
discharge of the obligation, not “a part of the contract” (ibid.)—that is, not constitutive of
the obligation itself. Kant here is following a principle of Roman law that survived in
Germany, that the contract of sale (emptio venditio) does not itself transfer property; that hap-
pens only through a subsequent conveyance, such as delivery (traditio); see Fritz Schultz,
Classical Roman Law (1951), 526–33. In holding this view, Kant implicitly disagreed with
Grotius and Pufendorf, both of whom rejected the principle of Roman law, that contract
does not convey property; see Hugo Grotius, De Jure Belli ac Pacis Libri Tres, 2 vols., tr. Francis
W. Kelsey (1925), vol. 2, 308–9; Samuel Pufendorf, De Jure Naturae et Gentium Libri Octo, 2
vols., tr. C. H. and W. A. Oldfather (1934), vol. 2, 610–11.
154 corrective justice

ing the contract, the defendant unjustly deprives the plaintiff of the
performance to which the plaintiff is entitled. The law undoes that
injustice by restoring to the plaintiff either the specific performance
that has been lost or the value of that performance. This value, in turn,
reflects the value of the subject matter of the contract. Hence the
plaintiff is entitled to damages that put the plaintiff into the position
that the plaintiff would have been in had the contract not been
breached. This is so not because the plaintiff has acquired an entitle-
ment to (in Kant’s formulation) “what was promised,” that is, the thing
that was the subject matter of the contract, but rather because the
plaintiff has acquired the promise itself, that is, the act that the defend-
ant is obligated to perform. The value of the thing promised is merely
a way of measuring the value of the promise itself. The plaintiff has
not acquired the thing promised “directly” (as Kant notes),18 but the
thing figures indirectly in the plaintiff ’s entitlement because the entitle-
ment’s value reflects the value of the thing.
This account resolves the Fuller–Perdue perplexity, that expectation
damages seem to be “a queer kind of compensation” in that they give
the plaintiff something that the plaintiff never had.19 It is true that the
plaintiff never had the thing promised; its loss is therefore not some-
thing for which the plaintiff can rightly claim compensation. But the
plaintiff did have an entitlement to the performance itself; it is for the
infringement of this entitlement that expectation damages compen-
sate. Kant thereby answers the question that implicitly troubles Fuller
and Perdue, “How can the law treat the plaintiff as entitled to the
thing’s value if the plaintiff is not entitled to the thing?” The plaintiff
turns out to be entitled to the thing’s value because that value deter-
mines the value of the performance to which the plaintiff is entitled.
Both the Kantian account and the Fuller–Perdue critique of expecta-
tion damages presuppose the disjunction between contractual per-
formance and ownership of the subject matter of the contract. But
this very feature of contract that is problematic for Fuller and Perdue
is what for Kant characterizes contract as a distinct kind of right.
The two accounts employ different conceptions of what is involved
in providing a justification for the rule about expectation damages.
For Fuller and Perdue justification consists in identifying the social
purposes that the rule serves. Indeed, they regard this conception of

18
Kant, above n. 14, [6:424].
19
Above n. 8.
 punishment and disgorgement as contract remedies 155

justification as so well established that it has achieved a pervasive trite-

ness.20 Therefore, once they dismiss the suggestion that expectation
damages maintain the equilibrium of goods among members of soci-
ety, they are free to rummage through the repertoire of social pur-
poses until they alight on the protection of reliance and the promotion
of commerce. Kant, in contrast, working in the tradition of corrective
justice, does not justify legal concepts by reference to a social purpose,
because that would involve the law’s treating the parties as means
rather than as agents who interact as ends in themselves. Instead, Kant
views justification as immanent in a system of rights. Because a system
of rights requires that the action of one self-determining person be
consistent with the freedom of another, a rule is justified simply inas-
much as it manifests this consistency. Conversely, a restriction of self-
determining activity for any reason except consistency with the
freedom of others (for example, a refusal to give legal recognition to
contractual entitlement) would eo ipso be unjustified.Thus, once a rule
can be understood as the juridical manifestation of self-determining
agency in one person’s interaction with another, no further work
remains for justification to do. The rule is justified by virtue of its
expressing the self-determining freedom of the interacting parties.
This freedom forms the baseline from which deviations count as
injustices. And then the undoing of such injustices in accordance with
corrective justice partakes of the normativeness immanent in the sys-
tem of rights and duties as a whole.
A virtue of the Kantian account of contractual entitlement is that it
is consonant with the compensatory function that law itself implicitly
assigns to expectation damages. In awarding the plaintiff the value of
what the contract would have given, the law treats promisees as enti-
tled to the expectancy that breach deprives them of. In contrast, a
long and complicated narrative—which Fuller and Perdue attempted
to provide—is needed to divert expectation damages from their osten-
sibly compensatory role to the remedial policies identified by Fuller
and Perdue. Although the classification of interests that Fuller and Per-
due offered has taken hold, their account of expectation damages and
the reconceptualization of contract law that this account entails have,
on the whole, had little effect on the law.21 The interest in securing

20
Above n. 2, at 52.
21
Daniel Friedmann, “The Performance Interest in Contract Damages,” (1995) 111 Law
Q.R. 628, 646–54.
156 corrective justice

the promised performance or its equivalent remains “the only pure

contractual interest.”22 This interest in performance as the distinctive
feature of contractual entitlement is the focus of Kant’s attention. In
providing a theoretical account that allows us to understand expecta-
tion damages for what they purport to be—that is, as compensation to
the plaintiff for an injustice suffered at the defendant’s hands—Kant’s
treatment of contract exemplifies the commitment of corrective just-
ice to understand the basic structure of private law in the law’s own
terms.

3. The disgorgement of gains from breach

Whereas expectation damages, whatever their proper theoretical basis,
are well established in the law, awards based on the plaintiff ’s gains are
more controversial. The question of whether the promisee is entitled
to what the promisor has gained from the breach has been called
“devilishly difficult.”23 Favoring gain-based awards are strong ethical
intuitions that promises should be kept and that those who breach
their contracts should not profit from their wrongs. On the other
hand, the difficulty in working out the applications of a new gain-
based principle reinforces the suspicion that the traditional approach
may be justified after all.
In many contract situations gain-based awards lie at the margins of
the law’s traditional compensatory framework. One such situation
occurs when the defendant’s gains may be used as a means of measur-
ing the plaintiff ’s losses. For example, when the defendant competes
with a plaintiff to whom the defendant has given an exclusive license
to sell or manufacture a certain commodity, the usual approach is to
treat the defendant’s gain as evidence of the profit that the plaintiff lost
through the breach.24 Another such situation occurs when defective
performance saves the promisor an expenditure without ultimately
causing the promisee a further loss.Then the promisee can deduct what
the promisor saved from the agreed price in order to bring the pay-
ment into line with what the promisee received, thus preventing what

22
Ibid., 629.
23
Andrew Burrows, “No Restitutionary Damages for Breach of Contract,” (1993) Lloyd’s
Maritime & Commercial L.Q. 453.
24
John Dawson, “Restitution or Damages?,” (1959) 20 Ohio St. L.J. 175, 189.
 punishment and disgorgement as contract remedies 157

turns out to be an overpayment.25 Yet another such situation occurs

when the promisor builds in breach of a negative covenant with the
promisee but without causing the promisee financial loss. The promis-
ee’s entitlement to receive, in lieu of an injunction, the amount that
reasonably would have had to be paid for securing a relaxation of the
covenant can be interpreted either as gain-based or as compensatory.26
A fourth such situation occurs when the breach of contract is also the
breach of a fiduciary duty, when the aggrieved fiduciary can secure an
accounting of the principal’s profits, which would have been unavail-
able from a mere breach of contract.27 A fifth such situation occurs when
a purchaser breaches a provision of the contract of sale that limits the
price at which the item can be resold.28 Disgorgement to the original
seller of the purchaser’s excess profit on resale can perhaps be justified
on the ground that, so far as the purchaser is concerned, the seller
retained the value of the item above the price limit.
It is tempting to regard such miscellaneous instances of gain-based
recovery not as particular applications of traditional categories, but as
the scattered embers of a general conception of gain-based damages,
to be collected into a new and explicit principle of disgorgement for
breach of contract. An appealing analogy from the law of torts beck-
ons. For centuries the owner of an object that the defendant convert-
ed and sold has been able, by “waiving the tort,” to recover the
proceeds of the sale from the defendant, even when this would give
the owner more than the value of the lost object.29 The gain-based
award that is controversial for breach of contract is universally accept-
ed for the misappropriation of property. This difference, one might
suppose, is entirely a product of history rather than reason. For why
should profiting from another’s contractual right be treated less severe-
ly than profiting from another’s proprietary right?

25
Samson and Samson v. Proctor [1975] 1 N.Z.L.R. 655 (S.C.) (builder, in breach of
building contract, puts insufficient steel reinforcing into house which is sold at a price undi-
minished by the defect; court holds that a deduction from the contract price is not a depar-
ture from the fundamental principle of compensation). See also Blake, above n. 1, at 398.
26
Wrotham Park Estate v. Parkside Homes [1974] 1 Weekly L. Rep. 798. (Ch. D.); see the
different interpretations of this case in Blake, above n. 1, at 395–97, 410.
27
E. Allan Farnsworth, “Your Loss or My Gain? The Dilemma of the Disgorgement Prin-
ciple in Breach of Contract,” (1985) 94 Yale L.J. 1339, 1354–60.
28
British Motor Trade Association v. Gilbert [1951] 2 All E.R. 641 (Ch. D.).
29
Lamine v. Dorrell, 92 Eng. Rep. 303 (1705). See Daniel Friedmann, “Restitution of
Benefits Obtained through the Appropriation of Property or the Commission of a Wrong,”
(1980) 80 Colum. L. Rev. 504; Graham Virgo, The Principles of the Law of Restitution, 2nd ed.
(2006), 454.
158 corrective justice

In recent years two important cases, one from Israel and the other
from England, have provided the most extensive discussions favoring
the disgorgement of gains from contract breach. In Adras Building
Material v. Harlow & Jones30 the defendant had agreed to sell steel to the
plaintiff, but when the price of steel spiked, the defendant instead sold
the steel stored for that purpose to a third party. Because the plaintiff
did not purchase substitute steel at a higher price before the market
receded to its former level, no loss was proved. The Supreme Court of
Israel awarded the plaintiff the gain that the defendant realized by sell-
ing its steel to the third party above the contract price. In Attorney-
General v. Blake31 a former employee of the intelligence service, who
had been convicted of spying and had escaped from prison, breached
his contract of employment with the Crown by publishing his mem-
oirs. Although he was not a fiduciary and the published information
was no longer confidential, the House of Lords held that the Crown
was entitled to the money owed to him by the publisher, on the
ground that in the circumstances the Crown had a legitimate interest
in preventing him from profiting from his breach of contract.
The basis of disgorgement in such cases is the sentiment that one
should not profit from one’s wrongdoing. This sentiment has obvious
moral resonance. It treats the breach of contract as a wrong—that is, as
an act that the promisor was morally obliged not to commit. By strik-
ing the gains of contract-breach from the hand of the promisor, dis-
gorgement gives teeth to the long-standing principle that promises
are to be observed (pacta sunt servanda).
In this respect disgorgement is at odds with the notion of efficient
breach. Efficient breach, a dominant idea in the economic approach to
contract theory, postulates that a contract breach from which the
promisor gains more than the value of the promisee’s expectancy is
economically efficient.32 By allowing the promisor to gain more than

30
Above, n. 1; see Daniel Friedmann, “Restitution of Profits Gained by Party in Breach of
Contract,” (1988) 104 L.Q.R. 383.
31
Above, n. 1.The parallel case in the United States, Snepp v. United States, 100 S. Ct. 763
(1980), differs in that the promisor in Snepp was held to be a fiduciary. On the other hand, it
is hard to resist the impression that, in ordering the disgorgement, the court in Blake was
aiming not merely at the promisor’s breach of contract in publishing his memoirs, but at the
traitorous activities that gave him the notoriety that made his memoirs profitable—an aspect
not present in Snepp.
32
A clear formulation of this much-discussed notion appears in Richard A. Posner, Eco-
nomic Analysis of Law, 5th ed. (1998), 131. See also David W. Barnes, “The Anatomy of Con-
tract Damages and Efficient Breach Theory,” (1988) 6 S. Cal. Interdisc. L.J. 397.
 punishment and disgorgement as contract remedies 159

would be sufficient to redress the promisee’s loss, the breach moves the
subject matter of the contract to its most valued use. In this way,
the self-interested preferences of the parties tend to the production of
the greatest social good. From the economic point of view, therefore,
no reason exists for the law to discourage such a breach. Conversely,
requiring the promisor to disgorge gains made through the breach dis-
courages the promisor from engaging in this wealth-maximizing step.
How do these matters stand from the perspective of corrective just-
ice? Corrective justice of course has no more interest in the promotion
of efficiency than it has in the promotion of any other goal extrinsic to
the interaction of the parties as the doer and sufferer of an injustice.
Indeed, the theory of efficient breach conceptualizes the breach of con-
tract not as an injustice to the promisee, but as an option available to the
promisor within the system of incentives that the law makes available
for the forwarding of efficiency. The breach of contract is simply a way
of channeling resources to their most valued use; the normative status of
a contract as imposing an obligation on the promisor plays no role. In
contrast, corrective justice shares with the disgorgement principle the
supposition that breach of contract is a wrong.
On the other hand, disgorgement involves the following difficulty
from the standpoint of corrective justice. The fact that the promisor
has profited from committing a wrong appears to supply an intuitively
plausible reason for requiring the promisor to surrender the gain, but
not for transferring that gain to the promisee. The taint that attaches
to the promisor’s gain by the wrongful manner of its acquisition does
not in itself make the promisee rather than someone else the justified
recipient of that gain. To be sure, the gain was realized through a
breach of contract with the promisee, but the question that corrective
justice raises is whether this breach establishes not merely the histor-
ical origin of the gain—its cause in fact, to use tort terminology—but
also the normative connection between the gain and the promisee’s
entitlement to it.33 This normative connection is present only when
the gain represents something to which the promisee had a right of
which he or she was deprived by the promisor’s wrongful act. Only
then would the gain be a constituent of the rights and correlative
duties obtaining between the parties, and only then would the award
of the gain render unto the promisee what was the promisee’s. It is not

33
On the difference between historical and normative connection, see above, chapter 4,
section 2.
160 corrective justice

immediately apparent, however, on what grounds the gain can be

considered an entitlement of the plaintiff.
For an award of damages to be consistent with corrective justice
not only must the breach of the contract be a wrong to the promisee
but the damages must be the measure of that wrong. Otherwise, there
is no reason for the damages to be awarded specifically to the prom-
isee for the injustice suffered. The Kantian account of contractual
right, which highlights the promisee’s entitlement to the promisor’s
performance, shows why a breach of contract can be viewed as a
wrong to the promisee within the framework of corrective justice.
That account, however, carries us no further than the promisee’s entitle-
ment to compensation for the loss of the value of the performance.
Even though the gain realized by the promisor resulted from the
wrong to the promisee, it does not seem to be part of that wrong.
Accordingly, from the standpoint of corrective justice, it is not
enough for the proponent of disgorgement to dismiss the economic
theory of efficient breach as indifferent to the normative dimension of
contract law. If that normative dimension is to be fully respected, it is
also necessary to indicate the positive ground for thinking that the
gain is within the promisee’s entitlement as quantifying the wrong
that the promisee has suffered. The fact that efficient breach is incom-
patible with disgorgement does not entail the conclusion that dis-
gorgement follows from the rejection of efficient breach.
Neither the Israeli nor the English disgorgement case succeeds in
showing the promisee’s particular entitlement to the gain that the
promisor is made to disgorge. In the Adras case Justice Barak explicitly
referred to the theory of efficient breach, rejecting it with the follow-
ing observations:

Moreover, it seems to me that the economic approach does not give enough
weight to considerations which cannot be measured in economic terms.
The law of contract is not only meant to increase economic efficiency but
also to enable society to lead a proper life. Contracts are there to be per-
formed, whether or not damages to be payable on breach, an approach by
which we encourage people to keep their promises. Promise keeping is the
basis of our life as a society and a nation.34

34
Adras, above n. 1, at 272. Similarly Justice S. Levin, at 241, acerbically remarked that “the
approach of the economic school of law ignores in cases like this the fact that we are dealing
with people with moral feelings and not with robots.”
 punishment and disgorgement as contract remedies 161

With these words Justice Barak rightly emphasized that contract has a
normative dimension that the economic approach ignores. However,
the vindication of the morality of promise-keeping against the amor-
ality of economically efficient breach is insufficient to ground a legal
entitlement in the promisee to the promisor’s gains. What is needed to
sustain the decision is reference not to the social morality of promis-
ing, but to contract as a juridical regime of rights and correlative duties
that renders the promisor’s breach (and, in particular, the realization of
profit through breach) an injustice to the promisee. Justice Barak’s
allusion to the role of promise-keeping in the proper life of society
makes it seem that the promisor’s profiting from the breach was not
specifically a wrong against the promisee, but more generally a sub-
version of the collective effort to preserve promise-keeping as the
basis on which social and national life rests. This view of the profit in
turn leaves unexplained why the promisee is entitled to recover for
what was a wrong against society as a whole.
Blake, the English case concerning the profits from the former spy’s
memoirs, has a parallel difficulty. In ordering disgorgement Lord
Nicholls, while acknowledging that disgorgement is an exceptional
remedy not subject to fixed rules, offers as a general guide “whether
the plaintiff has a legitimate interest in preventing the defendant’s
profit making activity and, hence, in depriving him of his profit.”35
The legitimate interest included the need to preserve the trust of
informants and to uphold the morale of secret service officers, appar-
ently even with respect to information that was no longer confiden-
tial.These considerations, however, do not address the question of why
the plaintiff was entitled to the profits, even assuming the defendant
was not. This absence of any entitlement to the gain on the part of the
plaintiff is perhaps why the dissenting judge, Lord Hobhouse, pointing
out that the defendant’s gain was not made at the plaintiff ’s expense,
stigmatized the claim as being of an “essentially punitive nature.”36
The reasoning in both Adras and Blake is directed to the pursuit of
social goals rather than to the justice intrinsic in the parties’ interac-
tion. The plaintiffs in these cases are awarded the profits realized from
the defendants’ breaches of contract not because the plaintiffs can
show their respective entitlements to these profits, but because they
are conveniently situated for assisting in the accomplishment of cer-

35
Blake, above, n.1, at 398.
36
Ibid., 407.
162 corrective justice

tain social goals. In Adras the goal is to encourage the socially import-
ant practice of keeping promises. In Blake the goal is to forward the
effective functioning of the secret service. These considerations focus
on the desirability of preventing the defendants from keeping what
they might gain from breaching their contracts. The goals as such are
indifferent to the question of who might get the profits thus struck
from the defendants’ hands. Instead of treating the gain as the locus of
an injustice done by the defendant and suffered by the plaintiff , the
reasoning points one-sidedly to the inadmissibility of the defendant’s
profit. The position of the plaintiffs is adventitious; they are connected
to their respective defendants through their contractual entitlements
even though the profits that they are awarded are not themselves
constituents of those entitlements. In this respect the reasoning in the
cases is incompatible with the correlative structure of corrective
justice.37
As was indicated in the previous chapter, disgorgement to the plain-
tiff of profits made by the defendant is the appropriate response, so far
as corrective justice is concerned, to the unauthorized alienation by
the defendant of something to which the plaintiff had a proprietary
right. By virtue of ownership the owner is entitled to all the profits
that accrue from the alienation of what is owned. Just as the owner’s
exclusive right to the object implies a duty on others to abstain from
it, so the owner’s right to the profits that accrue from its alienation
imports a correlative duty in others to abstain from such profits or, if
there was a failure to abstain, to yield these profits to the owner. The
profits are the owner’s as surely as the object that produced them. The
correlativity of the owner’s right and the wrongdoer’s duty means that
the wrongful gain is an injustice as between them. The injustice
embodied in this gain is undone when the gain is restored to the
owner of the object from which the gain accrued. The proprietary
nature of what was alienated makes the realization of profits a

37
Once the exercise is conceived instrumentally as the forwarding of extrinsic goals, the
choice of certain goals at the expense of others becomes significant. What, for instance,
makes the promotion of promissory good faith more important than the promotion of eco-
nomic efficiency (as the Supreme Court of Israel assumes)? Hanoch Dagan, for example, has
suggested that both are important and the profits should be split to reflect this; see Dagan,
“Restitutionary Damages for Breach of Contract: An Exercise in Private Law Theory,” (2000)
1 Theoretical Inquiries in Law 115, at 151. This natural consequence of instrumental analysis
would produce the incoherence of two considerations (efficiency and promissory good
faith) each artificially limiting the reach of the other. On the problematic normative struc-
ture of such incoherence, see Weinrib, The Idea of Private Law (1995), 32–44.
 punishment and disgorgement as contract remedies 163

correlatively structured wrong that accounts for the role of both par-
ties in the remedy.
This suggests that in the contracts context the promisee should be
awarded the promisor’s profits if the breach of contract can be con-
strued as the alienation of what belongs to the promisee. The Kantian
account of contractual entitlement, however, reveals the implausibility
of regarding the breach as the wrongful alienation of the promisee’s
property. The difficulty is to identify the property alienated through
breach. There are only two possibilities. In the first of these there is no
property; in the second there may be property in some sense, but there
is no alienation.
The first possibility is that the promisee owns the object promised
in the contract. However, on the Kantian view the promisee is entitled
to performance but does not have property in the object of the con-
tract. In the Adras case, for instance, the contract to ship a certain
quantity of steel to the promisee did not in itself transform any of the
defendant’s steel into the property of the plaintiff. The plaintiff, there-
fore, should have had no claim to the profits made by the sale to the
third party, even if the steel would otherwise have been shipped in
fulfillment of the contract. To the contrary: the defendant simply sold
its own steel to the third party and was accordingly entitled to the
profits from the sale by virtue of its ownership of what it sold.
The other possibility concerns the contractual performance itself,
which, in the Kantian account, constitutes the promisee’s contractual
entitlement and promisor’s correlative obligation.To be sure, the entitle-
ment to contractual performance can be treated as a species of prop-
erty for certain purposes (for example, classification as a chose in
action, assignment, constitutional protection).38 However, the promi-
sor’s breach cannot plausibly be regarded as a purported alienation of
this entitlement, so as to give the promisee a claim to the profits in the
promisor’s hand. The Adras case is illustrative. The promisor’s obliga-
tion was to perform a certain act—that is, to deliver steel to the prom-
isee. The breach consisted in not delivering the steel, which had been
38
For example, in a discussion of contract damages Daniel Friedmann suggests that,
because contractual rights enjoy the constitutional protection of property for purposes of the
takings clause of the constitution of the United States, property includes contractual rights;
Friedmann, “The Efficient Breach Fallacy,” (1989) 18 J. Legal Stud. 1, 16. However, one can-
not conclude that because the state compensates for the taking of contractual rights, the
promisor should be liable to non-compensatory damages for breaching a contract.The com-
pensatory norms of constitutional law are irrelevant to the private law’s treatment of non-
compensatory awards.
164 corrective justice

sold to a third party. This breach can hardly be construed as the prom-
isor’s alienation of something that can be conceptualized as “the deliv-
ery of steel as owed to the promisee.” What it sold to the third party
was the steel that would have fulfilled its contract with the promisee;
it did not sell the act of delivery that was owed by the promisor to the
promisee. Nor did it alienate the promisee’s entitlement. The third
party contracted with respect to the steel, not with respect to the
promisee’s entitlement. Nor, finally, did the promisor alienate the obli-
gation correlative to the promisee’s entitlement; by buying the steel,
the third party did not become obligated, as the promisor was and
remained, to deliver steel to the promisee. The breach of the contract
is not an alienation of something that the promisee owned.
Some have contended that, although sound when the subject mat-
ter of the contract is indefinite, this conclusion does not apply, even
on Kantian grounds, when the subject matter of the contract is spe-
cific or unique.39 If the promisee’s entitlement was to the delivery of a
horse, then the promisor’s sale of his only horse, Bessie, to a third party
is not the alienation of the promisee’s right, because the promisee’s
right did not specifically include Bessie. This is essentially the situation
in Adras, where disgorgement was, accordingly, not justified. If, how-
ever, the promisee’s entitlement was to the delivery of Bessie, then the
act of delivering Bessie to someone else “is identical to the act which
has already been transferred to the [promisee].”40 Selling this act of
delivery to a third party then represents a misappropriation by the
promisor of what belongs to the promisee.The profit thereby obtained
is rightfully the promisee’s.
This argument hinges on the identity of what was promised to the
promisee and what was sold to the third party. Only if this identity
exists can the promisor be construed as selling to the third party what
already belonged to the promisee. It is crucial to the argument’s Kant-
ian character that this identity consists not in the horse delivered (the
horse named “Bessie”) but in the act of delivering the horse; the spe-
cification of the horse as Bessie merely allows the two acts of delivery
to be identified as the same act, so that the second can be regarded as
a misappropriation of the first. However, the argument confines the

39
Peter Benson,“Disgorgement for Breach of Contract and Corrective Justice: An Analysis
in Outline,” in Understanding Unjust Enrichment, ed. Jason Neyers, Mitchell McInnes, and
Stephen G. A. Pitel (2004), 311, at 329; Andrew Botterell, “Contractual Performance, Correct-
ive Justice, and Disgorgement for Breach of Contract,” (2010) 16 Legal Theory 135.
40
Benson, above n. 39.
 punishment and disgorgement as contract remedies 165

specification of the act to the specification of the object transferred.

This limitation is artificial. Once the specification of the required act
is expanded to include all the terms of the respective transactions, the
identity between the two acts disappears. The two transactions may
call for acts of delivery at different times and places. Moreover, the
specification of delivery will inevitably refer to different recipients.
Accordingly, what was sold to the third party was not the act of deliv-
ery that belonged to the promisee, but a different act of delivery
inconsistent with the first one. This new act of delivery made breach
of the contract unavoidable, thereby exposing the promisor to dam-
ages, but it was not the misappropriation of an act of delivery identical
to, and therefore owned by, the promisee.The effect of insisting on the
importance of specification but confining this specification to the
object of the contract (“Bessie”) is to transform—inconsistently with
the Kantian argument—the contractual right to performance into a
right to the thing.
Thus, a breach of contract is not tantamount to the alienation of a
proprietary right. Because a proprietary right imposes a duty of non-
interference on the whole world, it has a juridical significance that is
independent of any particular wrongdoer. This independence means
that the duty correlative to the proprietary right has to be defined in
terms of a particular object that is separate from the indefinite number
of juridical relationships in which it figures. That object is therefore
available for misappropriation in a transaction between one non-pro-
prietor and another, with consequent liability to the owner for the
proceeds of this transaction. In the waiver of tort cases involving the
unauthorized alienation of property, the thing misappropriated is
indeed identical to the content of the owner’s right. Formulated in
terms of Kantian theory, the stable identity of the proprietary right
with respect to the duties it imposes on the whole world brings such
a right under the Kantian category of substance. Breaches of contract
are different. Contractual performance falls under the Kantian cat-
egory of causality—that is, of a relationship of consequence and ground,
because performance is a normative consequence of the nexus
between the two contracting parties. The required performance is
defined by the contract between the parties. The obligation to do the
contracted-for act is particular to that contractual relationship. By
breaching the contract the promisor infringes the promisee’s right but
does not alienate it.The subject matter of the promisee’s entitlement—
which is always the promisor’s act and not the thing required for the
166 corrective justice

act—has not been and cannot be passed on to a third party by the

promisor’s breach. Accordingly, the profits that the promisor has real-
ized from the breach do not come within the entitlement of the
promisee.
In the Blake case, the significance of the difference between con-
tractual and proprietary entitlement was a matter of dispute between
the judges. Lord Hobhouse, dissenting, was in favor of dismissing the
claim for the disgorgement of profits on the ground that “that is a
remedy based on proprietary principles when the necessary propri-
etary rights are absent.” 41 Lord Nicholls, on the other hand, remarked
in allowing the claim that “it is not easy to see why, as between the
parties to a contract, a violation of a party’s contractual rights should
attract a lesser degree of remedy than violation of his property rights.”42
Lord Nicholls’s assumption was that the only difference between a
contractual and a proprietary right is a quantitative one, that the
former obligates only on one person whereas the latter obligates an
indefinite number of persons.43 Lord Nicholls therefore concluded
that “it is not clear why it should be any more permissible to expro-
priate personal rights than it is permissible to expropriate property
rights.” 44
Lord Nicholls’s conclusion is suspect for a number of reasons. First,
given that breach of contract is the violation of a right, the issue is not
one of the relative permissibility of the defendant’s conduct, but of the
remedial response appropriate to that conduct. Secondly, a breach of
contract is not helpfully characterized as an expropriation; breach does
not involve the taking of an object that can exist independently of the
relationship between the parties or the extinction of an entitlement
owing to the plaintiff. Moreover, Lord Nicholls’s conclusion proves
too much: if proprietary and contractual entitlement did not differ,
disgorgement would be the standard remedy when the promisor prof-
itably breaches, rather than the extraordinary one that even Lord
Nicholls treats it as. Most importantly, the difference between propri-
etary and contractual entitlement is not merely quantitative. In Adras,
for instance, the promisee did not have a contractual right that

41
Blake, above n. 1, 410.
42
Ibid., 395.
43
This assumption is developed in the article on which Lord Nicholls expressly draws; see
Lionel Smith, “Disgorgement of the Profits of Breach of Contract, Property, Contract and
‘Efficient Breach,’” (1994–95) 24 Canadian Business L.J., 121, 130–2.
44
Blake, above n. 1, 395.
 punishment and disgorgement as contract remedies 167

consisted in owning the steel with only the promisor being subject to
a correlative obligation, in contrast to a proprietary interest in the
steel that created a correlative obligation for the whole world. The
promisee in Adras did not own the steel at all; all that the promisee
was entitled to was a certain performance. The difference between a
proprietary and a contractual right is qualitative; the former goes to an
object, the latter to an action.45 The result of this is that nothing is
available for the promisor to expropriate or alienate, since these verbs
are inapplicable as descriptions of what the promisor does with respect
to an entitlement that consists in his own actions. The upshot of this is
that, as Lord Hobhouse saw in dissent, disgorgement is an inappropri-
ate remedy for contract breach.
Is this conclusion affected by whether the remedy to which the
promisee is entitled is specific performance? For instance, in contracts
for the sale of land, where the purchaser is entitled to specific per-
formance, courts may describe the vendor as a trustee and the pur-
chaser as the equitable owner, and hold the vendor liable for the
profits realized from reselling the land to the third party at a higher
price.46 One may be tempted to regard such instances as evidence
that, even though disgorgement is not generally a remedy for contract
breach, the availability of specific performance transforms the prom-
isee’s contractual entitlement into a proprietary one so as to allow the
promisee to claim the promisor’s profits from sale. In this way of con-
sidering the promisee’s entitlement, the supposed transformation of
the contractual entitlement into a proprietary one is merely a termin-
ological shorthand for anticipating the availability of specific perform-
ance;47 the conclusion about property is the result of the premise
about the remedy.
This approach to the entitlement is inconsistent with corrective
justice’s conception of the relation of right and remedy. For corrective
justice the right is conceptually prior to the remedy that responds to
the right’s infringement. Of course, if the system of private law is well
ordered, the remedy will reflect the kind of entitlement that the plain-

45
In Kant’s terminology, proprietary and contractual obligations deal with objects of
choice that come within different categories: property deals with substance (an external
thing) and contract deals with causality (another’s choice to perform a specific deed); see
Kant, above n. 14, at [6:402].
46
Timko v. Useful Homes Corporation, 168 A. 824 (N.J. Ch., 1933); Lake v. Bayliss [1974]
1 Weekly Law Rep. 1073 (Ch. D.).
47
Dawson, above n. 24, 186.
168 corrective justice

tiff has. The remedy, however, does not determine the nature of the
underlying right. Whether the entitlement is proprietary or not
depends on the concepts internal to the juridical relationship between
the parties (such as the connection between the alienation of property
and the claim to proceeds). It does not depend on the court’s response
to the defendant’s injustice. The remedy, therefore, cannot transform
into a proprietary right what is not already one before the remedy is
fixed.48
Thus, the general picture that emerges from the present discussion
is as follows. Making the promisee disgorge his or her gains to prevent
profiting from a wrong has intuitive appeal. Corrective justice, how-
ever, requires that the parties be treated as correlatively situated
through the right of the plaintiff and the corresponding duty of the
defendant. One-sided attention to the defendant’s gains does not
reflect this correlativity. If disgorgement for contract breach is to

48
At common law, specific performance is available where damages are inadequate due to
the real or (as in the case of land in traditional contract doctrine) deemed uniqueness of the
subject matter of the contract.The consequence of this uniqueness is that the market cannot
reliably determine the value of what the promisee has lost through the breach. Only by
granting specific performance or by treating the promisee as equitable owner and therefore
entitled to the proceeds can the law ensure that promisor is awarded what he or she has been
unjustly deprived of. Accordingly, cases such as those mentioned above in n. 46 above need
not be considered as examples of disgorgement rather than compensation.
Even the fact that in a given jurisdiction (for example, in Israel) specific performance is
the default remedy should not affect the argument about disgorgement. In the Adras case, the
Supreme Court of Israel thought that the institution of specific performance as the default
remedy for contract breach in Israel marks a fundamental difference from the common law
that makes disgorgement more plausible; see Adras, above n. 1, at 241, 271. This, however, is
not necessarily the case. All that the status of specific performance as the primary remedy
indicates is that the Israeli system takes very seriously the idea that the promisee is, as correct-
ive justice affirms, entitled to performance; it does not necessarily change the subject matter
of the performance into a proprietary right. The temptation to slide from the former to the
latter should be resisted. For example, at p. 271 Justice Barak writes:
The injured party has a right not only to compensation for breach of contract, but also to specific
performance . . . Therefore, under Israeli law, a buyer in a contract of sale is entitled to receive the
subject matter of the sale, and an enrichment of the seller which infringes this right is an unjust
enrichment at the buyer’s expense . . . When there is a contract for the sale of a horse, the buyer has
the right to receive the horse, not damages for non-delivery. If the seller receives a benefit from
selling the horse to a third party, he . . . takes from the buyer a right to which the buyer is entitled.
There is an equivocation here about the right to which the last sentence refers. A right to
receive the horse is not the same as a right to the horse (to which the language of “taking”
might be applicable). The former right is to the performance of an act, the latter is to a par-
ticular thing. The former is contractual, the latter proprietary. The enhanced role of specific
performance does not change the categorical difference between what Kant, above n. 45,
called substance and causality.
 punishment and disgorgement as contract remedies 169

conform to corrective justice, the promisor’s profit must be under-

stood as proceeds from the alienation of the promisee’s property. Given
the Kantian account of contractual entitlement, an alienation of prop-
erty cannot be made out. Disgorgement of the gains from contract
breach awards to the promisee something that the promisee was not
deprived of.
To a large extent this conclusion coincides with the result favored
by the theory of efficient breach. However, this convergence in result
masks a fundamental divide between the two approaches. As its critics
have emphasized,49 efficient breach abstracts from the normative
dimension of contract to the promotion of efficiency. In contrast, cor-
rective justice, by exhibiting the immanently normative structure of
the contract relationship as a nexus of rights and correlative duties, is
normative through and through. Instead of disavowing (as the eco-
nomic approach does) an interest in the duty of contractual perform-
ance, corrective justice maintains that the nature of that duty is
precisely what excludes the notion of disgorgement. Thus, its rejec-
tion of the disgorgement of gains from contract breach comes not in
the pursuit of an external goal like economic efficiency, but as an
internal implication of the very idea of contractual entitlement.

4. Punishment and corrective justice

I now move to the possibility of punitive damages for breach of con-
tract. This possibility has lurked in the background of the preceding
discussion of expectation damages and disgorgement. Once one
assumes that a given head of damages is not compensatory, one is
tempted to ascribe to it a punitive purpose.Thus, Lord Hobhouse, dis-
senting in Blake, maintained that a claim for disgorgement had an
“essentially punitive nature.”50 Similarly, Fuller and Perdue at one
point suggested that, in the absence of a plausible compensatory justi-
fication for them, expectation damages might be viewed as having the
implicit quasi-criminal purpose of penalizing the promisor’s breach
for the sake of protecting the reliance interest.51 Such observations
about implicit purpose raise the question of whether an explicitly

49
See especially Friedmann, above n. 38.
50
Blake, above n. 1, at 407.
51
Above n. 2, 61.
170 corrective justice

punitive component should be added to the damage award. The two

sections that follow address this question by considering from the
standpoint of corrective justice first, the theoretical relationship
between punishment and liability and then, the use of punitive dam-
ages specifically for contract breach.
At the juncture of punishment and liability lies the issue of punitive
damages. Such damages are encased in controversy. Formally unrecog-
nized in the civil law jurisdictions but widely accepted in the com-
mon law world, punitive damages have been especially contentious
over the last several decades. Developments in both the United States
and England have contributed to this. In the United States, the rela-
tively unstructured discretion of the jury to determine damages has led
to concerns that the standards for awarding punitive damages are too
vague and that the awards themselves may be excessive.52 In England a
more fundamental development occurred: the House of Lords, un-
equivocally repudiating punitive damages as anomalous, restricted
their scope to the minimum allowed by precedent, a position in turn
rejected by courts in the old Commonwealth.53
The House of Lords was of the view that almost the only circum-
stance when punitive damages were available at common law in a
dispute between private parties was when the wrongdoer’s conduct
was calculated to make a profit that would exceed the compensation
payable to the victim.54 Lord Reid termed the traditional broader
conception of punitive damages a “form of palm tree justice” and he
characterized the objections to it as “overwhelming.” He explained:

To allow pure punishment in this way contravenes almost every principle

which has been evolved for the protection of offenders. There is no defin-
ition of the offence except that the conduct punished must be oppressive,

52
David G. Owen, “The Moral Foundations of Punitive Damages,” (1989) 40 Ala. L. Rev.
705, 727–38.
53
The House of Lords cases are Rookes v. Barnard [1964] 1 All E.R. 367 (H.L.) and Cas-
sell v. Broome [1972] App. Cas. 1027 (H.L.). The leading cases in the Commonwealth reac-
tion are Uren v. John Fairfax [1966] 117 C.L.R. 118 (H.C. Aust.); Fogg v. McKnight, [1968]
N.Z.L.R. 330 (S.C.); Lamb v. Cotogno [1987] 164 C.L.R. 1 (H.C. Aust.); Vorvis v. Insurance
Corporation of British Columbia [1989] 1 Sup. Ct. Rep. 1085; Whiten, above n. 1 (S.C.C.).
54
The House of Lords also allowed punitive damages where they are authorized by statute
and where there was oppressive, arbitrary, or unconstitutional action by government employ-
ees.These categories are not relevant to the present theme. On the second of these, see Kud-
dus v. Chief Constable of the Leicestershire Constabulary [2000] 2 Weekly Law Rep. 1789
(H.L.).
 punishment and disgorgement as contract remedies 171

high-handed, malicious, wanton, or the like—terms far too vague to be

admitted to any criminal code worthy of the name. There is no limit to
punishment except that it must not be unreasonable. The punishment is
not inflicted by a judge who has experience and at least tries not to be
influenced by emotion: it is inflicted by a jury without experience of
law or punishment and often swayed by consideration which every judge
would put out of his mind. And there is no effective appeal against
sentence.55

Critics of the English approach have responded that the institutional

distinction between criminal law and private law does not dictate so
exclusive an allocation of punishment to the criminal law.56
This controversy raises two fundamental issues. The narrower issue
concerns the role of punishment, as expressed through an award of
punitive damages, within private law. In dealing with this Lord Reid
adverted to the significance of the procedural distinctions between
tort law and criminal law. This in turn implicates the second and
broader issue of determining the nature of the demarcation between
civil and criminal liability. How does corrective justice stand with
respect to these two issues?
The first of these two issues can be briefly treated. That corrective
justice renders punitive damages problematic is obvious on its face.
Corrective justice insists that the normative considerations applicable
to the relationship between defendant and plaintiff reflect the parties’
correlative standing as the sufferer and doer of the same injustice.
Accordingly, it excludes considerations, regardless of how appealing
they otherwise might be, that refer to one of the parties without
encompassing the correlative situation of the other. Punishment is a
one-sided consideration of this sort. Punishment focuses not relation-
ally on the parties as doer and sufferer of the same injustice, but uni-
laterally on the defendant as doer. From a punitive standpoint, we do
not ask what would restore to the plaintiff what he or she was deprived

55
Cassell, above n. 53, at 1087.
56
Nicholas McBride, “Punitive Damages,” in Wrongs and Remedies in the Twenty-First
Century, ed. Peter Birks (1996), 175; Peter Cane, The Anatomy of Tort Law (1997), 118. For an
assessment of the controversy, see Andrew Burrows, “Reforming Exemplary Damages:
Expansion or Abolition,” in Wrongs and Remedies in the Twenty-First Century, ed. Peter Birks
(1996), 153. In England the Law Commission under Burrows’s direction recommended the
expansion of punitive damages; see “Aggravated, Exemplary and Restitutionary Damages,”
Law Comm. No. 427 (1997). Also favoring punitive damages is James Edelman, Gain-Based
Damages: Contract,Tort, Equity and Intellectual Property (2002), 9–21.
172 corrective justice

of by the defendant, but rather what punishment is deserved in view

of the defendant’s behavior. Accordingly, damages at private law that
are the vehicle of punishment are a windfall to the plaintiff because
they do not represent anything that the plaintiff has been wrongly
deprived of. Instead of measuring the plaintiff ’s entitlement, punitive
damages in effect function as a reward for providing the socially useful
service of acting as a private prosecutor.57
The second and broader issue is the difference between civil and
criminal liability. Although corrective justice is a regime of rights and
their correlative duties, the denial of the relevance of punishment to
corrective justice is not a denial of punishment’s place in a right-based
legal order. Instead, the point is that from the standpoint of corrective
justice, this place must be located within criminal law rather than
within the private law relationship. Since corrective justice brings out
the distinctiveness of the private law relationship, it illuminates Lord
Reid’s differentiation between criminal and civil liability and his
insistence that punishment is the concern of the former and not the
latter. From the standpoint of corrective justice, the relationship
between compensation and punishment, and between the legal insti-
tutions of private law and criminal law that correspond to them is, I
suggest, as follows.
Corrective justice rectifies injustices that operate on the parties in a
transactionally specific way. This transactional specificity involves link-
ing two specific parties through the injury by one of them of a right
held by the other. For example, the defendant may have tortiously
injured the plaintiff as a result of acting inconsistently with the plain-
tiff ’s right to physical integrity or to a specific item of property that
the plaintiff owns. Kant’s account of contractual entitlement includes
the promisor’s performance under a valid contract among the particu-
lar rights that a specific person might have. Or the defendant may
have been unjustly enriched by the plaintiff ’s having transferred value
that rightfully ought to be restored. Such unjust losses or gains relate
to rights that a specific plaintiff is entitled to vindicate against a spe-
cific defendant as a matter of corrective justice. When one’s right is
injured, one is entitled to restoration of the right or to its monetary
equivalent.
Punishment is different. It is state action that inflicts an adverse con-
sequence on the wrongdoer without restoring the right of a wronged

57
Whiten, above n. 1, para. 36.
 punishment and disgorgement as contract remedies 173

party. When the state punishes, it acts not to rectify a wrong that is
transactionally specific to the plaintiff and the defendant, but to vindi-
cate its own standing as the public guarantor of rightful order. The
underlying idea is that the actualization of rights requires that they be
publicly guaranteed, so that their establishment, delineation, and
enforcement do not reflect the merely unilateral will of any particular
person.58 Rights cannot be enjoyed in a world lacking public institu-
tions—a condition that Kant termed a “state of nature.” Accordingly,
everyone is obligated to be and to remain subject to what Kant called
public right—that is, a condition in which public institutions lay down
norms that guarantee the rights of all. The consequence of this is that
the state, as the juridical manifestation of public right, has an interest of
its own, distinct from the interest of any particular party, in having its
norms respected. Whereas each person’s interest is to prevent or rectify
an injury to one’s rights, the state’s interest is to prevent a defection
from the condition of public right into the state of nature. Thus arises
the possibility of a wrong that is not transactionally specific: it is rela-
tive to the condition of public right, rather than relative to a particular
plaintiff. Punishment is the response to this kind of wrong.
The role of public right accounts for two defining features of pun-
ishable wrongs. Because the wrong is a defection from public right, its
vindication lies at the hand not of a particular right-holder (as would
be the case with corrective justice), but of the state. Hence the state,
acting through its public prosecutors, has the role of initiating and
carrying the legal process that determines the guilt and consequent
punishment of the accused wrongdoer. Moreover, the intention with
which the act was performed is crucial to assessment of the actor’s
guilt. Because the offense signals that the offender has, at least to the
extent of the offending act, exempted himself from the requirements
of public right, it must be intentional. Otherwise, the offense would
not reflect the reason for the state’s interest in the actor’s behavior.
This accounts for the salience of mens rea in the criminal law.
Thus, intent plays a double role within a system of rights. On the one
hand, the defendant’s intention to injure the plaintiff constitutes the ele-
ment of culpability linking the defendant’s wrongful action to the plain-
tiff ’s wrongful injury. Indeed, if the act is performed with malevolence
sufficient to injure the plaintiff ’s dignity, the plaintiff may be awarded
aggravated damages. This role is transactionally specific and therefore

58
Arthur Ripstein, Force and Freedom: Kant’s Legal and Political Philosophy (2009), 145.
174 corrective justice

within the scope of corrective justice. On the other hand, the wrongful
intent also goes to the offender’s self-exemption from the norms of pub-
lic right. This role is not transactionally specific, because, although the
injustice takes place within a transactional context in which a specific
person may be59 injured, the deliberateness has significance beyond that
injury. Because of this double role, intent can figure both as the mental
element of an intentional tort and as the mens rea of a criminal act.
The difference between wrongs relative to another and wrongs rela-
tive to public right provide the conceptual ground for regarding civil
liability and punishment as distinct. This distinctiveness works in both
directions. The punitive arrangements of criminal law are not ren-
dered superfluous by the existence of a system of civil liability; that
system deals merely with transactionally specific violations of rights
rather than with transactionally non-specific challenges to public
right. Nor can punishment coherently be stuffed into the framework
of civil liability, for then lawsuits that have a transactionally specific
structure, in which a specific plaintiff sues a specific defendant to
restore a particular right or its equivalent, would have to deal with the
transactionally non-specific aspect of the wrong. The result of this
would be an award of punitive damages, which inevitably gives a
windfall to the plaintiff on the basis of considerations that go one-
sidedly to the deliberateness of the defendant’s conduct.
Lord Reid’s repudiation of punitive damages, quoted above,60 picks
up the institutional implications of this incoherence. Because a delib-
erate self-exemption from the regime of public right is more serious
than the infringement of a particular right and because conviction
carries the stigma of criminality, criminal law insists on express defin-
ition of offenses and of possible punishments. Criminal law also
entrenches procedural safeguards (such as the benefit of a more strin-
gent burden of proof) for those who are accused, which would be out
of place in a civil trial because the advantage they would give to
defendants would be incompatible with the notional equality of the
parties as the alleged doer and sufferer of the same injustice. Lord
Reid’s criticism of punitive damages as involving a form of palm tree
justice, as well as his almost complete rejection of punitive damages,

59
I say “may be” rather than “is” because the challenge to the regime of rights does not
necessarily require that someone actually be injured. This, I would suggest, is why there are
attempted crimes but not attempted torts.
60
Above at n. 55.
 punishment and disgorgement as contract remedies 175

thus give legal expression to the conceptual difference between cor-

rective justice and punishment and to the institutional roles that each
has within a rights-based approach to law.

5. Punitive damages in contract

With these general considerations about punitive damages in hand, I
want to turn more specifically to punitive damages in contract law. In
contrast to the situation in tort law, where such damages have histor-
ically been well entrenched and the controversy has been whether the
legal order ought to restrict them, contract law was traditionally hos-
tile toward punitive damages. One can ascribe this hostility to the fear
of disturbing the certainty of commercial dealings by introducing a
damage component that floated free of the value of the contractual
performance to which the parties had agreed. In England this hostility
remains. In the United States, however, the rule against awarding puni-
tive damages for breach of contract has been eroding for the past cen-
tury. Beginning with an exception that punitive damages could be
allowed when the breach was accompanied by a fraudulent act,61 the
present majority position in the United States is that punitive damages
can be awarded if the conduct constituting the breach of contract is
also a tort for which punitive damages can be given, with some state
courts allowing such damages on even more expansive grounds.62
Perhaps the most dramatic recent developments have occurred in
Canada. In 1989 the Supreme Court of Canada, signaling its rejection
of the restrictive English approach to punitive damages, recognized
that punitive damages were available (though it did not in fact award
them) for breach of contract.63 There matters stood until 2002, when
in Whiten v. Pilot Insurance64 the court upheld a punitive damage award

61
Welborn v. Dixon, 39 S.E. 232 (S.C. Sup. Ct., 1904); see Simpson, “Punitive Damages for
Breach of Contract,” (1959) 20 Ohio St. L.J. 284.
62
Dodge,“The Case for Punitive Damages in Contracts,” (1999) 48 Duke L.J. 629, 636–51.
The majority United States position is of limited interest in a discussion of contract damages,
because it merely prevents the fact that there was a breach of contract from precluding the
concurrent tort remedies.
63
Vorvis v. Insurance Corporation of British Columbia [1989] 1 Sup. Ct. Rep. 1085.
64
Whiten, above n. 1. At the same time the court issued a companion judgment applying
Whiten and concluding that, in the circumstances, an award of punitive damages for contract
breach was not justified; Performance Industries Ltd. v. Sylvan Lake Golf & Tennis Club Ltd.
(22 February 2002, S.C.C.).
176 corrective justice

against an insurer who attempted to evade honoring a fire insurance

policy. The plaintiff in Whiten had insured her house with the defend-
ant insurer. When her house burned down, the defendant’s denied her
claim under the policy, alleging without basis that the plaintiff had
committed arson. For two years the defendant persisted in hostile and
groundless opposition to the claim. In the meantime, because the
insurance claim was substantially her sole asset, the plaintiff ’s financial
situation deteriorated. The defendant’s conduct was deliberately
designed (so it was found) to starve the plaintiff into an unfair settle-
ment. The court upheld a jury award that gave the plaintiff not only
the insurance proceeds to which she was entitled under the policy, but
also an additional million dollars as punitive damages.
In coming to this conclusion the court dealt comprehensively with
the issue of punitive damages.The court emphasized that both the deci-
sion to award punitive damages and the determination of the quantum
had to be rational in the light of the objectives of punitive damages.The
amount awarded also had to be proportionate to the accomplishment of
those objectives. While punitive damages were the exception rather
than the rule, they nonetheless could be employed when compensa-
tory damages and the law’s other sanctions were inadequate to achieve
the retribution that the defendant deserved, the deterrence that would
prevent similar conduct in the future, and the denunciation that would
mark the community’s collective condemnation. Retribution, deter-
rence, and denunciation are the objectives of punitive damages; so
long as an award was not so disproportionate as to exceed the bounds
of what is rational for the achievement of these objectives, the award
could stand.
As a result of this development the Canadian jurisprudence now
provides one of the most extensive recent discussions in the common
law world of punitive damages in contract law. Like many Common-
wealth jurisdictions, Canada has definitively repudiated the idea
(which corrective justice supports) that punitive damages have no
place in private law. Instead the court has affirmed that “[p]unishment
is a legitimate objective not only of the criminal law but of the civil
law as well.”65 The question that arises is whether, having dispensed
with the coherence imparted by corrective justice, the court has none-
theless succeeded on some other basis in working out a coherent role
for punitive damages in relation to breach of contract.

65
Whiten, above n. 1, para. 36.
 punishment and disgorgement as contract remedies 177

The threshold issue is what are the circumstances in which punitive

damages may be awarded for breach of contract. On the one hand,
something more than a breach of contract, even if deliberate, is required.
On the other hand, it would have been inappropriate simply to award
the plaintiff in Whiten the insurance proceeds; some account had to be
taken of the abusive manner in which the defendant had dealt with her
claim. The defendant insurer’s high-handed treatment of the plaintiff
made it liable to the payment of a premium over and above the amount
that would have satisfied its contractual obligation had it paid promptly.
What is the legal basis of the obligation to pay this premium?
When this issue originally came up in the earlier case in 1989, the
court had been divided between two alternatives. The minority favored
simply assessing whether the conduct was “deserving of punishment
because of its shockingly harsh, vindictive, reprehensible or malicious
nature.”66 The majority feared that such exclusive reliance on what has
been called “the whole gamut of dyslogistic judicial epithets”67 would
open the door to subjective judgments on the basis of emotive adjectives.
Although the court rejected the restrictive English approach, it was none-
theless sensitive to the institutional concerns that the House of Lords had
articulated. If punitive damages were to be allowed, they had to operate
within a recognizable set of legal constraints. Once it was detached from
the criminal law and imposed as occasion demanded by the institutions of
civil law, punishment required a legally objective form of justification:

It must never be forgotten that when awarded by a judge or a jury, a punish-

ment is imposed on a person by a Court by the operation of the judicial proc-
ess. What is it that is punished? It surely cannot be merely conduct of which
the Court disapproves, however strongly the judge may feel. Punishment may
not be imposed in a civilized community without a justification in law.68

Unwilling to accept the sufficiency of reference to the manner of the

defendant’s conduct, the court formulated an additional substantive
requirement: to be liable for punitive damages for breach of contract,
the defendant must have committed an actionable wrong.69

66
Vorvis, above n. 63, at 1130.
67
Lord Diplock in Cassell, above n. 53, at 1129.
68
Vorvis, above n. 63, at 1105.
69
Ibid. That the requirement of actionable wrong is an addition to the requirement of
reprehensible and high-handed conduct is evident from Vorvis, ibid., at 1108 and from Whiten,
above n. 1, para. 83.
178 corrective justice

This requirement was satisfied in the Whiten case. The insurer had
not only refused to pay the proceeds as due. It had also breached the
contractual duty of good faith it owed to its insured. Although the
duty of good faith and the duty to pay the loss were both contractual,
they were independent of each other. The breach of the duty of good
faith thus constituted the actionable wrong that could trigger an award
of punitive damages.70
This is a laudable attempt to accommodate a punitive function for
private law to the ideal of legality. However, it raises a number of dif-
ficulties that are variants of the same question: why should an accu-
mulation of actionable wrongs lead to punitive damages rather than to
an accumulation of compensatory damages for the various wrongs
suffered?
The court assumed that compensation would be exhausted by
payment of the insurance proceeds and that, therefore, any award
above this amount that was based on the defendant’s obstructionist
processing of the claim had to be punitive.71 The court’s specifica-
tion of the defendant’s breach of its duty of good faith as a further
actionable wrong shows that this assumption was mistaken on the
court’s own reasoning. For if the defendant’s conduct not only
breached the contractual duty to pay the proceeds but also consti-
tuted the further actionable wrong of breaching the defendant’s
good faith obligation as an insurer, then there must be some sum,
however notional, that would provide compensation for that action-
able wrong.
Indeed, this is only one of the alternatively available ways of hav-
ing compensatory assessments take care of what the defendant owed
for its high-handed treatment, over and above its liability to pay the
insurance proceeds. Another way flows from the court’s repeated
characterization of the fire insurance policy as a homeowner’s “peace
of mind” contract.72 If the defendant’s obstruction of the claim
breached its contractual obligation with regard to the plaintiff ’s peace
of mind, then that could have been the subject of a compensatory
assessment. A third way would have been to claim aggravated dam-
ages for the harm that the defendant’s outrageous conduct caused to

70
Whiten, above n. 1, paras. 79–83. As the court noted, by not insisting that the actionable
wrong be tortious, the Canadian position is more expansive than the parallel provision of the
Restatement on the Law of Contracts (Second), 255.
71
Whiten, above n. 1, para. 129.
72
Ibid., paras. 4, 115, 129.
 punishment and disgorgement as contract remedies 179

the plaintiff ’s feelings and sense of self-worth.73 Thus, even though

the defendant should have been made to pay more than the insurance
proceeds it withheld, the court’s assumption that punitive damages
were therefore necessary was based on its ignoring the compensatory
implications of its own description of good faith and peace of mind
as aspects of the contract.74
This failure to make a comprehensive assessment of the compensa-
tory damages leads to another difficulty. In the court’s view, punitive
damages are to be awarded only where compensatory damages are
insufficient to accomplish the punitive purposes.75 The court regards
even compensatory damages as forwarding the punitive objectives of
denunciation, retribution, and deterrence,76 so that the punitive dam-
ages are understood as residual to the compensatory ones from the
standpoint of punishment itself. Punitive damages, therefore, are add-
itional to compensatory damages without being independent of them:
punitive damages are merely the continuation of the aspect of punish-
ment already present in the award of compensation. Thus one cannot
tell whether or to what extent punitive damages are needed to sup-
plement the compensatory damages until all the compensatory dam-
ages are in view. The court regards this sequencing, in which punitive
damages are considered only after compensatory damages are seen to
be insufficient, as an important device for preventing immoderate

73
In Vorvis, above n. 63, the court recognized the possibility of awarding aggravated damages
for breach of contract. See John Swan, “Extended Damages and Vorvis v. Insurance Corpora-
tion of British Columbia,” (1990) 16 Can. Bus. L.J. 213, 216–21 Bruce Chapman, “Punitive
Damages as Aggravated Damages: The Case of Contract,” (1990) 16 Can. Bus. L.J. 269.
74
A compensatory claim along one of these lines would have yielded significantly less
than the punitive damages that the plaintiff received. In the case itself the plaintiff was
awarded approximately $318,000 in insurance proceeds, a similar amount in legal costs, and
$1 million in punitive damages. It is inconceivable that compensation for breach of the duty
of good faith or for infringing her contractual interest in peace of mind or for the aggravated
damage would have amounted to $1 million. The insurer’s duty of good faith in processing
the insurance claim would presumably not have been assessed at more than thrice the value
of the claim itself. Nor would her peace of mind about her home have been assessed at more
than thrice the value of the home itself. Nor would the damage to her sense of self-worth
nave been evaluated at so much more than the maximum that the Canadian courts allow
($100,000 in 1978 Canadian dollars) for non-pecuniary damages in personal injury cases; see
Andrewes v. Grand & Toy Alberta [1978] 83 Dom. L. Rep. (3d) 456 (S.C.C.); Lindal v. Lindal
[1981] 121 Dom. L. Rep. (3d) 263 (S.C.C.). Of course, had the plaintiff been awarded a
smaller compensatory amount rather than the large punitive one, she would not have been
able to complain; from the plaintiff ’s standpoint the punitive award is always a windfall.
75
Above n. 1, paras. 74, 94, 123, 129.
76
Ibid., para. 94.
180 corrective justice

awards.77 The consequence of the court’s insistence that punitive dam-

ages are awarded if and only if compensatory damages are insufficient
should be that punitive damages are unavailable in the absence of a
full compensatory reckoning.78 Thus, the absence in Whiten of a com-
prehensive compensatory assessment undermines, according to the
court’s own reasoning, the appropriateness of the award of punitive
damages that the court approved.
Furthermore, the requirement that punitive damages need to be
triggered by a further actionable wrong is inconsistent with the idea
that punitive damages are to be awarded only if compensatory dam-
ages are insufficient. Under this idea, in seeking punitive damages the
plaintiff in Whiten should have claimed compensatory damages for the
defendant’s breach of its duty of good faith. But then the breach of
that duty could not have served as the further actionable wrong in
accordance with the requirement the court lays down, for that wrong
would have been something for which she was seeking compensation.
She would then need to locate yet another actionable wrong. How-
ever, if that actionable wrong were another breach of contract, it too
would have been something for which she should have claimed the
compensation that would have prevented it from being regarded as a
further actionable wrong, and so on ad infinitum. The point at which
this sequence stops would be the point at which punitive damages
could not be awarded for lack of an independent actionable wrong to
trigger them. It thus turns out that (at least so long as all the actionable
wrongs are breaches of contract, as in Whiten) were the plaintiff to
claim compensatory damages for all actionable wrongs suffered, then
punitive damages could not be awarded. In other words, the require-
ment of a further actionable wrong, in combination with the notion
that punitive damages are given only if compensatory damages are
insufficient, renders impossible the very award that the requirement is
supposed to condition.
Now the idea that punitive damages are awarded only when com-
pensatory damages are insufficient is itself part of a more comprehensive
idea that punitive damages are awarded only if all other penalties, includ-
ing criminal and regulatory sanctions, have been taken into account and

77
Ibid., para. 74.
78
In Whiten, there was no consideration of aggravated damages because the plaintiff did
not claim them; ibid., para. 91. But it is odd that the plaintiff could expose the defendant to
a punitive fine simply by not claiming under a compensatory head of damages.
 punishment and disgorgement as contract remedies 181

found to be inadequate to accomplish the objectives of retribution,

deterrence, and denunciation.79 These other sanctions are relevant but
not a bar to the award of punitive damages. The court explains:

The prescribed fine, for example, may be disproportionately small to the

level of outrage the jury wishes to express. The misconduct in question
may be broader than the misconduct proven in evidence in the criminal or
regulatory proceeding. The legislative judgment fixing the amount of the
potential fine may be based on policy considerations other than pure pun-
ishment.80

Thus, the assessment of contract damages can rank as the final deter-
minant of the sum total of the punitive consequences visited on the
defendant.
This is odd. It will be recalled that the reason for introducing the
requirement of an independent actionable wrong was to prevent
the decision to award punitive damages from being subjective. But on
the issue of the quantum of damages, such subjectivity is allowed (pro-
vided that it is not so egregious as to violate vague notions of ration-
ality and proportionality). Criminal and regulatory proceedings, which
are devoted to punishment, have standard legal constraints on subject-
ivity, such as the absence of common law crimes, the procedural and
evidentiary protections for the accused, and the prohibition of double
jeopardy. But the judge or jury determining the quantum of punitive
damages, operating free of those constraints, can impose an additional
penalty out of a higher feeling of outrage or conclusions reached on a
lower burden of proof or a judgment that the legislated level of pun-
ishment was inadequate. It is little wonder that the dissenting judge in
Whiten warned against “a sort of private criminal law, devoid of all the
procedural and evidentiary constraints which have come to be associ-
ated with the criminal justice system.”81

79
Whiten, above n. 1, para. 123.
80
Ibid.
81
Ibid., para. 158. A notable feature of Whiten is that, despite the court’s elaborate treat-
ment of rationality and proportionality, there is little indication of what placed this specific
punitive damages award of $1 million within the acceptable range. Aside from observing that
the judges below thought that this sum was not unreasonable and that there had been an
analogous increase in the size of the punitive damages award for defamation (ibid., paras.
135–40), the court twice mentioned the fact—which it acknowledged to be irrelevant under
the test of rationality that it was formulating—that the award was less than two times the
total of compensatory damages and legal costs (ibid., paras. 4, 132).
182 corrective justice

Moreover, the court’s own view of punishment makes this cavalier

attitude toward other legal processes more questionable. The court
specifies the objectives of punishment as retribution, deterrence, and
denunciation. These objectives are indeterminate in two ways. First,
there is an indeterminacy about what penalty would achieve any of
these objectives considered on its own. But, second, there is also an
indeterminacy about what would achieve all of them in combination.
This is because the three objectives rely on considerations that are at
least partly inconsistent with each other.82 There can be no single cor-
rect view of what penalties or ranges of penalties would achieve these
three divergent objectives. All that a legal system can hope for is that
its institutions of positive law make determinations that are general,
transparent, authoritative, responsible, and based on the appropriate
specialized expertise and institutional competence. One would think
that the judge or jury in a contracts trial would be institutionally the
least qualified to decide the final amount of the defendant’s punish-
ment. To have the judge or jury determine as part of a contracts case
whether other institutions specifically charged with punishing have
indicated the appropriate outrage or had access to adequate evidence
or legislatively formulated the appropriate punishment does not seem
consistent with a well-ordered legal system.
These reflections bring us back to the criticisms of punitive dam-
ages voiced by the House of Lords, that punitive damages are institu-
tionally misplaced in private law. It should be apparent that the recent
Canadian developments have not satisfactorily obviated those criti-
cisms. Nor, having rejected the approach consistent with corrective
justice, has the Supreme Court of Canada yet established a plausible
approach for punitive damages in the contracts context. Perhaps future
elaboration will alleviate the inadequacies of the court’s present juris-
prudence.83 Or perhaps, with the passage of time, these inadequacies
will be recognized as the inevitable consequence of the incoherence

82
Bruce Chapman and Michael J. Trebilcock, “Punitive Damages: Divergence in Search
of a Rationale,” (1989) 40 Ala. L.R. 741, 779–86 (cited in Whiten, above n. 1, at para. 43). For
example, as Chapman and Trebilcock observe at 797, deterrence would favor and retribution
would oppose grossing up the penalty to reflect the fact that the probability of its enforce-
ment is less than one.
83
In subsequent cases claiming punitive damages for breach of contract, the Supreme
Court of Canada held that the conduct in question was not sufficiently reprehensible to
merit punitive damages (Fidler v. Sun Life Assurance Co. of Canada [2006] 2 S.C.R. 3;
Honda Canada Inc. v. Keays [2008] 2 S.C.R. 362).
 punishment and disgorgement as contract remedies 183

of introducing punitive objectives into the contract law’s framework

of corrective justice.

6. Conclusion
Starting with the basic idea of corrective justice, that the remedy cor-
rects the injustice suffered by the plaintiff at the defendant’s hand, this
chapter has examined the significance of various conceptions of con-
tract damages. Its conclusions can be briefly stated. Despite the con-
tentions of Fuller and Perdue, expectation damages are justified as
compensation for the promisor’s breach of contract in accordance
with corrective justice. Expectation damages represent the value of
the promisor’s performance; the promisee’s entitlement to this per-
formance is illuminated by the Kantian account of contract, which
construes the doing of the contractually required act as the content of
the promisee’s entitlement. Kant’s insistence that “what I acquire
directly by contract is not an external thing but rather his deed”84 also
indicates the deficiency of requiring the promisor to disgorge the
gains from the breach. The plaintiff is entitled to the disgorged gains
only if the gains came from the alienation of the plaintiff ’s property;
the alienation of property, however, is not a concept applicable to the
promisor’s failure to perform a contractually obligatory act. Nor does
it make sense to regard disgorgement (or, as Fuller and Perdue thought,
expectation damages) as punitive in nature, in view of the categorical
distinction between liability and punishment. Moreover, as the Cana-
dian experience shows, even when damages are expressly punitive,
they seem incapable of being coherently integrated into the fabric of
contractual liability.
At the heart of this argument lies the identification of the nature of
the contractual entitlement. So far as corrective justice is concerned,
right and its correlative duty are the legal concepts that mark the
doing and suffering of an injustice. Unless the contractual right is
properly identified, the law’s interest in awarding expectation damages
becomes obscure, as the classic discussion by Fuller and Perdue shows.
Conversely, the nature of the contractual right has implications for
how disgorgement is to be viewed. In particular, the identification of
the contractual entitlement with the performance of an act reveals the

84
Above n. 16.
184 corrective justice

inappositeness of assimilating contract breach to the alienation of

property, and thus also the inappositeness of disgorgement.
From the standpoint of corrective justice, private law is a distinct
form of practical reason, in which justification reflects the correlative
situation of the parties as doer and sufferer of the same injustice. One-
sided considerations, no matter how appealing, such as that the party
in breach should disgorge profits made from its wrong or should be
punished for its malevolent conduct, do not conform to this correla-
tivity. Such considerations can be incorporated only if private law is
willing to countenance unfairness as between the parties and the dis-
turbance of the law’s internal coherence. Perhaps sensing this, the
common law traditionally did not use damage awards to punish the
breaching party or to force disgorgement of the gains from breach. In
recent decades both courts and commentators have been willing to
reconsider. If the argument presented here is correct, the law has to
that extent become more flexible but less just.
 6
Unjust Enrichment

1. The theoretical challenge

“A person who has been unjustly enriched at the expense of another
is required to make restitution to the other.”1 In the past few decades,
this principle of liability has recently become as firmly established in
the common law jurisdictions as it has long been among civil law sys-
tems. Being a relatively new basis of liability, unjust enrichment is now
the most dynamic of all areas of private law. Nonetheless, skeptical
voices continue to be heard. Scholars have contended that unjust
enrichment adds little to the traditional arsenal of private law cat-
egories;2 that the idea of unjust enrichment is either hopelessly circular or
is a conclusion based on unmentioned normative values that do the
real work;3 and that the principle of unjust enrichment submerges
within a common framework types of claims that should be governed
by diverse principles.4 Despite being recognized as never before, unjust
enrichment remains the most embattled of the bases of liability in pri-
vate law.
Two interrelated theoretical puzzles have fueled such expressions of
skepticism. First, at the heart of unjust enrichment lies the mystery of
what makes an enrichment unjust. This question concerns not merely
the positive law, but the normative theory implicit in it, exposing a

1
Restatement of Restitution (1937), s. 1.
2
Steve Hedley, Restitution: Its Division and Ordering (2001).
3
Hanoch Dagan, The Law and Ethics of Restitution (2004), 19.
4
Peter Jaffey, “Two Theories of Unjust Enrichment,” in Understanding Unjust Enrichment, ed.
Jason W. Neyers, Mitchell McInnes, and Stephen P. A. Pitel (2004), 39; Christopher T.Wonnell,
“Replacing the Unitary Principle of Unjust Enrichment,” (1996) 45 Emory L.J. 153.
186 corrective justice

gap through which seep doubts about the nature and scope of the
principle of unjust enrichment. Even proponents of unjust enrich-
ment acknowledge the absence of a viable theory of the unjustness
that grounds this form of liability.5 For many years the development
of unjust enrichment was impeded by the suspicion that, once recog-
nized as a category of liability, it would direct judges away from trad-
itional legal reasoning to the amorphous exercise of legal discretion on
unspecified grounds that vary according to one’s personal sense of
justice.6 How, then, can the unjustness of the enrichment be concep-
tualized in a juridically disciplined manner?
Second, how do the principle’s three elements (that the plaintiff has
been “enriched,” that the enrichment has been “at the defendant’s
expense,” and that the enrichment was “unjust”) as well as the defense
of change of position fit together to form a coherent basis of liability?
Historically, the prime impetus for the development of unjust enrich-
ment has been to bring together various instances of restitutionary
liability that the common law had assigned to separate compartments.
This drive for unity across different kinds of transactions, however,
would be pointless unless the principle provided unity within each
transaction. For if the elements of liability are merely a potpourri of
mutually inconsistent or indifferent considerations, liability would
depend on the particular way in which the various considerations are
balanced and combined in any given case or group of cases. Unjust
enrichment would then provide merely a common label, not a com-
mon pattern of reasoning. Only by combining in a coherent set could
the elements of unjust enrichment impart the unity of an overarching

5
A distinguished commentator on the law of restitution has lamented:
Lurking beneath the surface is an awkward question that needs to be answered by jurists: what is
it that makes a particular enrichment unjust? It is a question which has not been answered in
modern writing on the law of restitution. Indeed, most modern writing on the law of restitution
is notable for its apparent indifference to theoretical issues. What is the notion of justice which
underpins the law and its development? If this area of law is to be restyled the law of unjust enrich-
ment, surely it cannot avoid openly addressing questions that relate to the conception of justice
which underpins the law? Writers avoid the issue . . .
Ewan McKendrick, “Taxonomy: Does it Matter?” in Unjustified Enrichment: Key Issues in Com-
parative Perspective, ed. David Johnston and Reinhard Zimmermann (2002), 627, 654.
6
“[T]he adoption of this concept . . . would clothe judges with a very wide power to
apply what has been described as ‘palm tree justice’ without the benefit of any guidelines. By
what test is a judge to determine what constitutes unjust enrichment? The only test is his
individual perception of what he considered to be unjust.” Martland J., dissenting, in Pettkus
v. Becker [1980] 117 D.L.R. (3d) 257 (S.C.C.). A similar note was sounded decades earlier in
W. S. Holdsworth, “Unjustifiable Enrichment,” (1939) 55 L.Q.R. 37, at 49.
 unjust enrichment 187

principle to the various situations that contemporary scholars of resti-

tution claim fall under it.
Moreover, only when they are coherently unified do the individual
elements themselves have a stable meaning. What counts as an “enrich-
ment” of the defendant cannot be determined independently of what
renders the enrichment “unjust” and “at the expense of the plaintiff.”
The idea of an enrichment presupposes a normatively relevant baseline
against which the enrichment is measured, and what is normatively
relevant has to refer to something to which the appropriate consider-
ations of unjustness can intelligibly apply, and vice versa. The same is
true of what it means for the enrichment to be realized “at the expense
of ” the plaintiff. As with other bases of liability, the meaning of each
element conditions, and is conditioned by, the meaning of all the others.
Each element is intelligible not on its own but through its place in
the principle of unjust enrichment conceived as a unified whole.
The theoretical idea that reflects this concern with the inner unity
of the principle of unjust enrichment, as well as of other principles of
private law, is corrective justice. Indeed, in a recent judgment the
Supreme Court of Canada remarked: “Restitution is a tool of correct-
ive justice. When a transfer of value between two parties is normat-
ively defective, restitution functions to correct that transfer by
restoring their parties to their pre-transfer positions.”7 Such an obser-
vation testifies to the intuitive plausibility of understanding liability
for unjust enrichment as an instantiation of corrective justice. Unjust
enrichment at another’s expense seems to be an obvious example of
an injustice as between the parties, which a finding of liability then
corrects by requiring restoration of the enrichment.
However, it is one thing to assert the intuitive plausibility of con-
necting unjust enrichment to corrective justice, and another to pro-
vide an adequate theoretical account of the connection. Such an
account must show how liability for unjust enrichment manifests the
three interwoven features that make up corrective justice.8 First, cor-
rective justice, reflecting the bipolar nexus of plaintiff and defendant,
signifies a normative structure in which the parties are correlatively
situated. Second, because this structure requires a content that is itself
informed by correlativity, the organizing features of the parties’

7
Kingstreet Investments v. New Brunswick (Department of Finance) [2007] 1 S.C.R. 3,
para. 32.
8
Above, chapter 1.
188 corrective justice

relationship are the plaintiff ’s right and the defendant’s correlative

duty. Third, rights and their correlative duties imply a conception of
the parties as persons who interact with each other as free and equal
agents, without the law’s subordinating either of them to the other.
Accordingly, as an instantiation of corrective justice, liability for unjust
enrichment should exhibit the correlative structure of the parties’
relationship, vindicate the plaintiff ’s right as against the defendant, and
affirm the parties’ freedom and equality. And so the question arises:
how are these features of corrective justice implicit in liability for
unjust enrichment?
In the recent scholarship on unjust enrichment, doubts about correct-
ive justice flow from difficulties that supposedly pertain to the second of
these features, the correlativity of right and duty. On the duty side, the
difficulty rests on the defendant’s being obligated to make restitution
despite his or her passivity in receiving the enrichment. This passivity
means that the recipient cannot plausibly be regarded as a wrongdoer—
that is, as someone who is to be subjected to liability as a consequence
of having breached a duty owed to the plaintiff. One consequence of
this is that liability in unjust enrichment is “strict” or without fault, and
therefore at odds with the corrective justice account of tort law.9 Even
this, however, understates the difficulty. In fact, the recipient has not
merely done no wrong, but has not done anything at all: “there is no
sense in which the defendant is the agent of the plaintiff ’s misfortune.”10
Nor does it seem possible to say that the injustice consists in the act of
withholding restitution.That would be an injustice only if the defendant
were already under an anterior duty to make restitution.This is the duty
that corrective justice is alleged to be incapable of grounding.
A parallel difficulty emerges in connection with the plaintiff ’s right.
Cases of unjust enrichment deal with the transfer of money or goods
or with the provision of services. In none of these cases does the plain-
tiff retain a proprietary right to the subject matter of the enrichment.
Corrective justice postulates that liability vindicates some right of the
plaintiff ’s. In cases of unjust enrichment, however, the enriching trans-
action seems to have wiped away that right, leaving the plaintiff with
nothing to vindicate.11

9
Mitchell McInnes, “The Measure of Restitution,” (2002) 52 U.T.L.J. 163, 186–96.
10
Dennis Klimchuk, “The Structure and Content of the Right to Restitution for Unjust
Enrichment,” (2007) 57 U.T.L.J. 661, 677.
11
A sophisticated statement of this difficulty is found in Klimchuk, ibid.
 unjust enrichment 189

The challenge concerning the plaintiff ’s right moves from the

premise that the unjust enrichment leaves the plaintiff without a pro-
prietary right to the conclusion that the plaintiff has no right at all.
This conclusion does not follow. The plaintiff may well have a non-
proprietary right—that is, a right not in rem but in personam. Such a
right, consisting in an entitlement against a particular defendant rather
than against the world in general, is established through a transaction
in which both parties participate. A paradigmatic instance is the right
to contractual performance, which arises through a course of dealings
(offer, acceptance, and so on) between the parties and runs in favor of
a particular promisee and against the particular promisor. In pointing
to this instance of an in personam right, I am not proposing a revival of
the pernicious view that a claim in unjust enrichment is based on an
implied contract. However, I am suggesting that perhaps the reason
that unjust enrichment could have been assimilated to contract for so
long is that both bases of liability are of a kind that do not rest on a
proprietary claim. Of course, one can make good on this suggestion
only by specifying (as I hope to do) the elements in the parties’ deal-
ings that perform for unjust enrichment the function that such con-
cepts as offer and acceptance do for contract.
The purpose of this chapter is to elucidate unjust enrichment as an
in personam basis of liability that conforms to corrective justice. As in
the case of contract, the parties to liability in unjust enrichment (so I
shall argue) establish the correlative right and duty through the inter-
action in which they both participate. Being interactionally estab-
lished, the in personam right stands in contrast to an in rem right, which
the right-holder has prior to and independently of the defendant’s
wrong. With respect to both in rem rights and in personam rights, cor-
rective justice undoes an injustice that consists in an inconsistency
with the plaintiff ’s right that is imputable to the defendant. But there
is this difference between them. The in rem right imposes a duty on
the whole world; the defendant’s particular duty arises out of mem-
bership in that world. The in personam right imposes a duty specifically
on the defendant; this duty is the product of a right-establishing inter-
action with the plaintiff. To be sure, the in personam right is not a cre-
ation ex nihilo. It arises out of what is rightfully the parties’: the parties
to a contract, for example, have pre-existing rights to what they are
exchanging. However, the juridical effect of the parties’ interaction is
to transform these pre-existing rights into components of a new rela-
tionship of right and duty.
190 corrective justice

How, then, does an unjustly enriching interaction establish a cor-

relative right and duty of restitution? As the Supreme Court of Canada
observed, unjust enrichment occurs “[w]hen a transfer of value
between two parties is normatively defective.”12 My focus is on the
role of value in a corrective justice account of unjust enrichment. I
intend to trace how value starts out as an entitlement of the plaintiff ’s
that is transferred to the defendant in a normatively defective way,
with the result that the value must be retransferred to the plaintiff .

2. The juridical significance of value

What is value and what is involved in its transfer? To address these
questions I start with the account of value in Hegel’s Philosophy of
Right. This account continues a tradition of enquiry about value that
stretches back to Aristotle and was subsequently carried forward by
Marx and others. Hegel’s remarks are particularly illuminating for
issues of liability, because he treats value initially not as an economic
concept but as a juridical one.
Hegel regards value as an incident of property. One is entitled to
the value of something by virtue of one’s ownership of the thing that
is the locus of the value. Hegel’s description of value as an object of
ownership reflects commonplace notions drawn from contract and
tort law, respectively, that the owner of anything alienable is entitled to
realize its value through exchange13 and to be compensated to the
extent of its value in the event of wrongful injury or deprivation.14
The owner of the thing owns the value in the sense that ownership of
the thing carries with it an entitlement to something equivalent when
the thing is exchanged or injured.Value is thus the potentiality that is
actualized through a set of legal operations—exchange and liability—
with respect to things that one owns. Indeed, unless it were possible to
conceive of this potentiality as an entitlement of the thing’s owner, the
transformation of an entitlement to what one owns into an entitle-
ment to what is substituted for it through exchange or liability would
make no sense. The entitlement to value thus marks the continuity
through the process of exchange and the determination of liability of
the owner’s entitlement to the thing owned.
12
Above, n. 1.
13
G. W. F. Hegel, Philosophy of Right, tr. T. M. Knox (1952), s. 77.
14
Ibid., s. 98.
 unjust enrichment 191

After dealing with the right to use one’s property, Hegel makes the
following observations about the notion of value:

A thing in use is a single thing, determined quantitatively and qualitatively

and related to a specific need. But its specific utility, being quantitatively
determinate, is at the same time comparable with [the specific utility] of
other things of like utility. Similarly, the specific need which it satisfies is at
the same time need in general and thus is comparable on its particular side
with other needs, while the thing in virtue of the same considerations is
comparable with things meeting other needs. This, the thing’s universality,
whose simple determinate character arises from the particularity of the
thing, so that it is eo ipso abstracted from the thing’s specific quality, is the
thing’s value, wherein its genuine substantiality becomes determinate and
an object of consciousness. As the full owner of the thing, I am eo ipso
owner of its value as well as of its use.15

Hegel here draws attention to three characteristics of value. First, value

is quantitative. This is apparent from the contrast between value and
use. Use involves reference to the specific qualities that a thing has
that allow it to satisfy the specific needs of the specific person using it.
I can make use of my shoe, for example, because my shoe has certain
qualities: a concave shape into which my foot fits, a flexible material
that will bend as I lift and lower my foot, a slightly curved sole that
facilitates locomotion, and so on. Only with such qualities can the
shoe satisfy the needs of movement and protection that the shoes
serve. In contrast, when I enquire into the value of the shoe, my inter-
est is entirely quantitative: to how many units of something else is the
shoe equivalent? The movement of our attention from use to value is
thus a movement from quality to quantity.
Second, value is relational. As is indicated by the enquiry into the
number of units of something else to which the shoes are equivalent,
value relates a thing to other things by quantitatively comparing them.
The value of my shoes is not at issue so long as these shoes are consid-
ered exclusively on their own. Attention shifts to the shoes’ value
when I compare the shoes, say, to food by wondering how many
loaves of bread the shoes equal.16 Value is thereby a quantitative repre-
sentation of the relation between the shoes and the loaves.

15
Ibid., s. 63.
16
The example is Aristotle’s: Nicomachean Ethics,V, 1133a23.
192 corrective justice

Third, such quantitative comparison presupposes abstraction from

the particular uses to which the compared items are put and from the
particular needs that they serve. Nothing can have value that another
cannot put to some particular use or that does not serve another’s par-
ticular need. When quantitatively compared, however, these diverse
uses and needs implicitly share a common dimension of commensur-
ability as instances of usefulness and need generally. For instance, shoes
and loaves of bread serve the particular needs of ambulatory comfort
and nutrition respectively, which, qua particular needs, are not intrin-
sically related to each other. In their particularity they are incompar-
able. Accordingly, if so many shoes are to be treated as equivalent to so
many loaves, then the needs served by shoes and bread, or the uses
that satisfy these needs, figure in this equivalence only through the
general idea of need or usefulness in which they participate as instan-
ces. Abstraction from the particularity of need and use is the presuppos-
ition of value’s functioning as a medium of quantitative comparability
between qualitatively different things.
These three characteristics of quantity, relation, and abstraction are
reciprocally entailed in the idea of value. The presence of each is
demanded by the presence of the other two.Value’s quantitative char-
acter requires abstraction from one thing’s particular use to the useful-
ness that allows that thing to be quantitatively compared to another. If
the value of shoes is represented by the formula that so many shoes
equal so many loaves, each characteristic of value pauses on a different
aspect of the shoes’ valuation. Quantity looks to the numbers modify-
ing the things being compared. Relation looks to the fact that the
quantification of shoes is here stated comparatively by reference to the
number of loaves that the shoes equal. Abstraction from the particu-
larity of need and use looks to what is presupposed in conceiving of
the relation between two particular items, such as shoes and loaves, in
terms of their quantitative equality.
Value refers to the possibility of exchange and is concretized
through the process of exchange. In exchange the owners of things of
value determine what is to be exchanged for what and how many
units of one thing are to be given for how many units of the other.
They thereby give expression to the relational and quantitative char-
acteristics of value. They do so, however, not merely in fulfillment of
their own particular needs or in anticipation of the particular uses
they will make of what they receive through exchange, but as partici-
pants in a world of value that abstracts from those needs and uses.
 unjust enrichment 193

Because value treats particular needs and uses as instantiations of

need and usefulness in general, the quantitative comparison between
the two things being exchanged is systematically linked to quantita-
tive comparisons between these things and other things. By abstract-
ing from the particularity of need or use, value becomes indifferent to
the particularity of the things being exchanged and can therefore be
expressed through the comparison of anything of value with any other
thing of value. If I wish to exchange a pair of shoes for loaves of bread,
the number of loaves that I receive is a function not merely of the
relationship between shoes and loaves, but also of the relationship
between shoes or loaves and anything else for which they might pos-
sibly be exchanged. Even if I want bread and someone else wants
shoes at the end of the day, we can each reach our desired destinations
through a series of exchanges involving other commodities. The value
of my shoes and the value of another’s bread exist in equilibrium with
the value of other things. The formula that so many shoes equal so
many loaves can be expanded to include the equality of these num-
bers to so many of any other thing of value. Value thus reflects the
possibility of equalizing all things of value with one another, and
equating any given thing of value with all other things of value.17
Consequently, the quantification of equivalents in exchange is
beyond the power of the exchanging parties alone. The value to be
attached to things that they exchange is determined not by either the
party’s subjectivity or by the relationship between their respective
subjectivities. Rather, value reflects the relationship between all pos-
sible exchanging parties with respect to all possible exchangeable
things. Equivalence in exchange is an objective rather than a subjec-
tive idea. In the formula “so many shoes equal so many loaves,” the
quantities do not represent merely what the owners of the shoes and
the loaves, as persons who wish to satisfy their particular needs, accept
from each other. Given the abstracting characteristic of value, the
quantities of shoes exchanged for loaves represent what the shoes and
the bread are objectively worth relative to each other.
Value is thus the medium for measuring whether what was received
is quantitatively equivalent to what was given. The difference between
these two marks the extent to which one party gave the other some-
thing for nothing. Only if there is such a difference does the transac-
tion constitute a transfer of value.

17
I. I. Rubin, Essays on Marx’s Theory of Value (1972), 109.
194 corrective justice

The reason for this is as follows. When dealing with transfers, one
must distinguish between things that have value and value itself. The
transfer to another of a thing that has value does not necessarily mean
that there has also been a transfer of value. Take the example of
exchange. When I exchange a certain quantity of shoes for a certain
quantity of food, I no doubt have transferred something of value, the
shoes, and received something of value in return, the food. But if the
food is of equal value to the shoes, no value has been transferred.
Exchange on such terms features the reciprocal transfer of things of
value but not the transfer of value itself, since it keeps constant the
value to which each party is entitled. Exchange demonstrates that
value “is distinct from the external things which change owners in the
course of the transaction,”18 because in an exchange external things
are transferred but value is not. To be sure, I would not engage in this
exchange unless the food I received was more useful or valuable to
me than the shoes I surrendered. But the value that is expressed in and
through the exchange abstracts from me as a particular person with a
particular preference for this amount of food rather than that amount
of shoes. What matters to value in exchange is not the value to me in
isolation, but value as determined by the intrinsically relational pro-
cess of exchange among those trading shoes and food.
Only to the extent that the transfer is gratuitous—that is, involves
no receipt of equivalent value—does the transfer of a thing of value
become a transfer of value as well. If I transfer shoes but receive in
return nothing or food of less value (like the Homeric hero who fool-
ishly “exchanged gold armor for bronze armor, a hundred oxen’s
worth for nine”),19 then I have transferred not only the shoes as things
of value but value itself. In contrast to what happens in an exchange,
the transaction does not preserve intact the amount of value that I
have, because there is no equivalence of value in what was given and
received. Through this gratuitous transfer the value of what is right-
fully mine has been diminished and the value of what is rightfully the
transferee’s has been increased by the amount of value that has been
transferred without reciprocation. In the language of unjust enrich-
ment, the transferee has been enriched at my expense. This does not
mean, of course, that the transferee is obligated to return the enrich-
ment. That further consequence depends on whether the retention of

18
Hegel, above n. 13, s. 77.
19
Homer, Iliad, vi, 235–36.
 unjust enrichment 195

that enrichment is unjust—that is, whether the transfer occurred

under conditions that generate an obligation to restore the transferred
value.
Being unreciprocated, the transferred value of the shoes thus has a
double aspect. On the one hand it is an incident of the transferee’s
proprietary right in the shoes. As is the case with every owner, the
transferee who becomes the owner of the shoes also thereby becomes
entitled to their value. If the transferee sells the shoes, the transferee
is entitled to keep the value realized through the sale. If the shoes are
tortiously destroyed or converted, the transferee is entitled to receive
from the wrongdoer their equivalent value as compensation. Because
the transferee’s right to the shoes (and the consequent entitlement to
their value) is good against the whole world, I am not differently situ-
ated with respect to this value than is everyone else.
On the other hand, the transferred value in the shoes is also a com-
ponent of the normative relationship, unshared by anyone else,
between me as transferor and the transferee. Even if the transfer of the
shoes (and therefore of their value) is valid from a proprietary stand-
point, the gratuitousness of the transfer raises a distinct issue of justice
between us as parties to the transaction. Because the law assumes that
persons generally act to further their own ends rather than another’s,
it seeks to ensure that I truly intended the transfer to be gratuitous.
And conversely because the law does not allow obligations to be cre-
ated behind another’s back, restitution of the transferred value has to
be consonant with the free will of the transferee. Thus, aside from the
passage of title in the shoes, the question arises whether the circum-
stances of the gratuitous transfer of the value in the shoes are such that
the transferee is under an obligation to restore this value to me. These
circumstances pertain to the relationship between the two of us as
participants in the transfer of value rather than the relationship
between the transferee, as the new owner of the shoes, and everyone
else. Put more technically, although I have lost the in rem right to the
shoes (and thus to their value), one can still ask whether the condi-
tions of transfer were such that I now nonetheless have, as against the
transferee, an in personam right to their value. It is this aspect of the
transferred value that engages the principle of unjust enrichment.
Exchange and transfer of value are thus mutually exclusive notions.
Exchange features a movement of things of value from each party to
the other. It does not, however, feature any movement of value. In
contrast, a transfer of value occurs when one party gives the other
196 corrective justice

something of value, but in return receives nothing or something of

lesser value. This transaction transfers not only a thing of value, but
also value itself, for through this interaction one party loses and the
other party gains value.
This contrast between a transfer of value and a transfer of a thing of
value is the consequence of the inherently relational nature of value.
Characterized as it is by quantitative comparability that abstracts from
the qualitative differences between things of value, value equates an
amount of one thing to an amount of another.Value is not concerned
with anything on its own but with the quantitative relationship
between one thing and another. Whether a transfer of value has
occurred thus depends not on the movement of any single thing of
value, but on whether that movement has been matched by the recip-
rocal movement of some other thing of equivalent value. This recipro-
cal movement is a contingent matter. When it occurs, one has an
exchange in which things of value have been transferred but not value
itself. When it does not occur, one has a transfer of value.

3. The transfer elements of liability

I now turn to the juridical significance of treating unjust enrichment
as involving a transfer of value, understood as the giving of something
(whether objects or labor) for nothing. How is the idea of a transfer of
value actualized through the requirements for liability? These require-
ments reflect the idea to the extent that it provides a structure to
which they conform. They also construct the idea by endowing with
legal specificity what would otherwise be an abstraction. Several points
deserve notice.
First, at the most general level the idea of a transfer of value is
reflected in two of the requirements for liability under the principle of
unjust enrichment, that the defendant be enriched and that the enrich-
ment be at the expense of the plaintiff. Understood as aspects of a
transfer, these two requirements are not mutually independent ele-
ments but the integrated moments of a single bilateral phenomenon.
Unjust enrichment deals not with maintaining the wealth of one party
or another against an increase or decrease in the value of their respect-
ive resources, nor even with a matching increase and decrease in each
party’s wealth, but with a relationship between the parties that can
ground the liability that one of them may have to the other. The idea
 unjust enrichment 197

of a transfer establishes the requisite relationship by pointing to an

enrichment that has moved from the plaintiff to the defendant. Accord-
ingly, the “enrichment” and “expense” mentioned in the principle of
unjust enrichment are terms of mutual relation, each requiring the
other in order to function as constituents of liability. They refer not to
gains and losses simpliciter—that is, to one person being better off and
another person being worse off than before, but to the connection of
each to the other through the giving and receiving of value.
Second, enrichment at the expense of another should be under-
stood as structured by the immediacy of the link between the parties
as transferor and transferee of value. The notion of a transfer thereby
defines the ambit of liability, preventing liability that is either too
restrictive or too expansive. Liability is too restrictive when the plain-
tiff ’s claim is disqualified on the grounds of absence of enrichment
even though value has been transferred to the defendant. An example
is the now fading suggestion that the plaintiff ’s passing on to third
parties of the loss from the transfer excludes recovery of the value
from the transferee.20 Liability is too expansive when the plaintiff ’s
claim is allowed even though the parties are not related as transferor
and transferee. An example is the view that the relationship between
the parties can be indirectly established through the remotely causal
stages of the enrichment’s transmission.21
Third, not every benefit realized from the action of another involves
a movement of value. For value to move, the enriching action must be
directed toward something that is the defendant’s. If the purpose and
intended effect of the action refer only to the plaintiff and the plain-
tiff ’s property, the value remains with the plaintiff even though the
defendant has been advantaged as a result. The absence of liability for
incidental benefits illustrates this. In a typical case of incidental benefit,
the plaintiff acts with reference to what is his or her own property or
in the exercise of his or her own rights but in the process happens to
confer a benefit on a neighbor. Classic examples are the cutting down
of a wood that obscures a neighbor’s prospect or building a wall that
happens to shield a neighbor’s house from windstorms.22 Because the

20
For criticism of the “passing on” defense, see Michael Rush, The Defence of Passing On
(2006).
21
Peter Birks, “‘At the Expense of the Claimant’: Direct and Indirect Enrichment in
English Law,” in Unjust Enrichment: Key Issues in Comparative Perspective, ed. David Johnston
and Reinhard Zimmermann (2002), 493, 518.
22
Ruabon Steamship v. London Assurance [1900] A.C. 6, at 12 (H.L.).
198 corrective justice

work was done not on the defendant’s property but on the plaintiff ’s
property and for the plaintiff ’s own purposes, nothing has occurred
that can be construed as a transfer of value from the plaintiff to the
defendant.23 One can phrase this conclusion in the terms of the prin-
ciple of unjust enrichment by saying that the defendant’s enrichment
has not come at the plaintiff ’s expense.24 What this means is that by
virtue of the labor having been expended on the plaintiff ’s property
and for the plaintiff ’s purposes, the value of the labor has been retained
by the plaintiff and has not passed to the defendant.25
Fourth, conceiving of the enrichment as a transfer of value casts
doubt on the appropriateness of terms like “subjective devaluation”
and “incontrovertible benefit.” In the current treatment of unjust
enrichment, these terms qualify the segment of the analysis that deals
with enrichment. “Subjective devaluation”26 suggests that a benefit
may not qualify as an enrichment if a defendant can plausibly assert
that, despite the benefit’s objective value, he or she subjectively attaches
no value to it. Once enrichment is understood as signaling a transfer of
value, however, subjective devaluation cannot pertain to the determi-
nation of whether there has been an enrichment. Because value
abstracts from the parties’ particularity, neither value nor its transfer is
determined subjectively. Whether a person who gives another some-
thing of value has in return received something of equivalent value is
an objective question, the answer to which is systemically determined
by market exchanges.Value, therefore, cannot be subjectively devalued.
Nor can subjective devaluation be defeated by subjectively revaluing
the benefit on the ground of its incontrovertibility. At bottom, subjec-
tive devaluation is not about the nature of the enrichment, but about
the transferee’s freedom to make his or her own choices.27 This is of
course an important consideration, but it concerns not the existence
of an enrichment but the justness of the defendant’s retaining it.

23
Daniel Friedmann, “Unjust Enrichment, Pursuance of Self-Interest, and the Limits of
Free Riding,” (2003) 36 Loy. L. A. L. Rev. 831, at 845 (“recovery is denied simply because the
nature of the benefit consists of an increase in value without a transfer of property or
labor”).
24
Edinburgh and District Tramways Ltd. v. Courtenay [1909] S.C. 99, at 106.
25
“When a person does something on somebody else’s property, in the mistaken idea, it may
be, that it is his own, then the jactura is obvious enough. He has expended money, or something
else, which has passed into other persons’ property. But here nothing has passed.” Ibid.
26
Peter Birks, An Introduction to the Law of Restitution (1985), 109.
27
Mitchell McInnes, “Enrichment Revisited,” in Understanding Unjust Enrichment, ed.
Jason W. Neyers, Mitchell McInnes, and Stephen P. A. Pitel (2004), 175.
 unjust enrichment 199

Fifth, because one can transfer only that to which one has a right,
the notion of a transfer of value recognizes that the transaction
enriched the defendant with what was initially within the plaintiff ’s
entitlement. Recall Hegel’s observation that “as full owner of the
thing, I am eo ipso owner of its value as well as of its use.”28 The enrich-
ment is at the plaintiff ’s expense not merely because the transaction
had an adverse effect on the plaintiff , but because that effect operates
on value as an incident of what the plaintiff owned on entering the
transaction.29 The plaintiff ’s right to the value at the inception of its
transfer is the precondition of the claim that the value should be
retransferred to the plaintiff once the transfer is shown to be defect-
ive.30
Sixth, although it originates in the plaintiff ’s pre-transfer right to
the thing of value, the plaintiff ’s claim is to the retransfer of the value
independently of that thing. Ownership of the thing of value involves
a relationship between the transferee and the rest of the world; the
claim to the value independently of the thing of value involves a rela-
tionship between the transferee and the transferor. The bilaterality of
the transfer endows the value with a juridical significance that is sep-
arate from the one it has as an incident of a thing of value. This is
because, due to its quantitative, relational, and abstract character, value
is transferred not through the movement of any particular thing from
transferor to transferee, but through the absence of a reciprocal move-
ment of something of equivalent value from the transferee to the
transferor. What is crucial is the giving of something for nothing, not
the title to what was given. This is why the plaintiff ’s lack of a propri-
etary right to the thing that embodied the value does not undermine
the claim to the value’s retransfer.
Seventh, the idea that enrichment at the expense of another denotes
a transfer of value casts light on what renders an enrichment unjust. As

28
Above n. 13.
29
Conversely, the plaintiff cannot complain of the diversion to the defendant of a benefit
that lacked this status—for example, customers’ patronage for which the defendant success-
fully competes; James Gordley, Foundations of Private Law (2006), 425.
30
See Brian Fitzgerald, “Ownership as the Proximity or Privity Principle in Unjust
Enrichment Law,” (1995) 18 U. Q. L. Rev.166, at 172:
If the plaintiff did not own the property in which the value inhered and emanated before the
unjust enriching event, then it is incomprehensible . . . to say that value has been subtracted from P
and transferred to D. Ownership, then, is the starting point of any claim in unjust enrichment
because it tells us who holds value at the start and in turn who can come into court and claim loss
of value.
200 corrective justice

noted at the outset, a perennial apprehension raised by opponents of

unjust enrichment is that “unjust” is too amorphous a qualifier of
enrichment to yield a juridically manageable basis of liability. This
apprehension is unfounded. Once the enrichment is understood as a
transfer of value, the familiar idea of a transfer determines the nature
of the applicable unjustness. A transfer of value is unjust if its condi-
tions are not consistent with the norms of justice that govern transfers
generally. Unjust enrichment is, on this view, concerned not with just-
ice at large, but with the specific and narrowly legal issues of just
transfer. To be sure, the fact that the object of the transfer is value
rather than some particular asset introduces special considerations; just
as value’s being the object of transfer implies a special notion of what
a transfer is—the non-receipt of an equivalent—so it also implies spe-
cial notions of what makes such a transfer normatively defective. Such
special considerations, however, do not make the enquiry into the
justness of the enrichment amorphous. The law’s task remains that of
drawing out the legal implications of justice in transfer, given the
object of the transfer. To this task I now turn.

4. Justness in the transfer of value

What considerations make a gratuitous transfer defective? I propose to
address this question by comparing gifts and unjust enrichment. The
two are similar in that both feature the giving of something for noth-
ing, gift through the requirement of delivery, and unjust enrichment
through the notion of enrichment at another’s expense. They are,
however, dissimilar in several respects. A gift involves the transfer of a
particular object, whereas unjust enrichment involves the transfer of
value. Value is the antithesis of particular objects, in that it abstracts
from their particularity to the quantitative comparison between them.
Moreover, defectiveness in the transfer of a gift prevents a change in
property rights, leaving intact the in rem relationship between the
owner of the object and everyone else, whereas defectiveness in the
transfer of value creates in the transferor an in personam right to have
the transferee restore the value. Defectiveness in the transfer of gifts is
well understood from the long and settled jurisprudence about gifts.
Defectiveness in the transfer of value is more mysterious, and its juris-
prudence more fluid and controversial. Accordingly, it makes sense to
begin with the conception of defectiveness operative for gift, and then
 unjust enrichment 201

focus on how that conception is modified when applied to transfers

of value.
So far as gifts are concerned, justice in transfer requires that the gift
manifest the will of both the donor and the donee. “In order to make
it a gift, it must be made out not only that it was given as a gift but
that it was received as a gift.”31 On the donor’s side, the delivery of the
gift must be accompanied by a donative intent; on the donee’s side,
the gift must be accepted as donatively given. The giving of the gift
for nothing forms the terms of the transfer, which is effective only if
both parties subscribe to them. However, if (for example) the trans-
feror really intended a loan rather than a gift, or if the transferor
intended a gift but the transferee treated it as a loan32 or refused to
accept it, then the transfer is normatively defective, so that the propri-
etary right remains with or reverts to the transferor.33
Donative intent and acceptance are thus the legal concepts through
which justice in transfer expresses the freedom of both of the parties. To
be effective, a transfer must connect the parties in a way that they both
consent to. The donor is entitled to what is his or her own until freely
parting with it. Similarly, the donee is entitled not to be subjected even to
the beneficence of another unless he or she finds such beneficence
acceptable. Of course, one can usually presume acceptance of the confer-
ral of a benefit. However, if the evidence shows that the intended trans-
feree does not accept the transfer as gratuitous, the law does not force a
conferral against the transferee’s will.The transferee may graspingly desire
the gift of something for nothing, and the transferor may benevolently
desire to give something for nothing, but neither of them can impose his
or her will on the other party. They each remain free from subjection to
the purpose of the other unless they accede to that purpose.
Accordingly, in the case of gifts, the unwillingness of either party
renders the transfer defective. A transfer purports to move the owner-
ship of a specific object from the transferor to the transferee. This
movement can be interrupted by a lack of the relevant intent as the
object purports either to exit the transferor’s ownership or to enter
the transferee’s.
I turn now from the gratuitous transfer of a specific object dealt
with by the law of gift to the transfer of value, also gratuitous, dealt

31
Hill v. Wilson, L.R. 8 Ch. App. 888 (1873).
32
The situation in Hill, ibid.
33
J. Hill, “The Role of the Donee’s Consent in the Law of Gift,” (2001) 117 L.Q.R. 127.
202 corrective justice

with by the law of unjust enrichment. Here too donative intent fig-
ures in justice in transfer. If the transferor intended a transfer of value,
that is, intended to benefit the transferee for nothing, then (assuming,
as is almost invariably the case, that the benefit is acceptable to the
transferee) the transfer complies with justice, and cannot be reversed
on the ground of unjust enrichment. It is not unjust for value dona-
tively transferred to be irretrievable.34 In the terminology of the Can-
adian jurisprudence, donative intent counts as a “juristic reason” for the
enrichment.35 Another such juristic reason is that the benefit was con-
ferred in fulfillment of a valid obligation owed to the transferee or in
the settlement of an honest claim made by the transferee, for it is not
inconsistent with justice in transfer for the transferee to retain what
the transferor was obligated to give.36 Thus, when donative intent is
present, the transfer of value conforms to justice in transfer no less
than does the transfer of a specific object through gift.
However, in the absence of donative intent by the transferor, justice
in transfer requires that the transfer of a specific object through gift be
treated differently from the transfer of value. The absence of donative
intent is sufficient to render a gift defective because the failure of the
object rightfully to leave the ownership of the donor keeps intact the
transferor’s entitlement to what was given. For transfers of value, how-
ever, the absence of donative intent is not sufficient to render the
transfer defective. Because a transfer of value involves the transfer of a
thing of value without receiving an equivalently valued thing in
return, the thing that embodies the value—and therefore the value
that is an incident of owning that thing—has already irrevocably

34
Donative intent includes risk-taking—that is, the conferral of a benefit under circum-
stances in which the transferor hopes for remuneration but knows that the transferee is not
obligated to give it.
35
Peter v. Beblow [1993] 1 S.C.R. 980, at 987.
36
In Peter, ibid., McLachlin J. listed the following considerations among those going to
juristic reason:
(i) Did the plaintiff confer the benefit as a valid gift or in pursuance of a valid common law,
equitable or statutory obligation which he or she owed to the defendant?
(ii) Did the plaintiff submit to, or compromise, the defendant’s honest claim?
The presence of a valid obligation is a ground for supposing that, despite the brilliance of the
judgment, Moses v. MacFerlan, 97 E.R. 676 (K.B., 1760) was incorrectly decided, as subse-
quent cases held; see Dublin v. Building and Allied Trade Union [1996] 2 I.L.R.M. 547 (S.C.).
If one excludes the possibility of MacFerlan’s contractual obligation, Moses was in effect a
risk-taker who, by incurring the obligation entailed in endorsing the notes, gave MacFerlan
the benefit of his own potential liability under circumstances in which MacFerlan was not
obligated to abstain from availing himself of that liability.
 unjust enrichment 203

entered the ensemble of the transferee’s entitlements.37 Unlike the

defectiveness in the case of gift, which interrupts the movement of a
specific object from transferee to transferor, defectiveness in the case
of a transfer of value creates a claim that value that has already moved
with the thing that embodies it be restored to the transferor. If the
claim succeeds, it has to be satisfied out of assets that rightfully belong
to the transferee. In contrast to gifts, the transferor’s claim is not that
the transferee never received the value but that, having received for
nothing what was not intended to be gratuitously given, the transferee
is not entitled to retain the value. Consequently, the duty of the trans-
feree to restore value cannot be determined solely by one-sided refer-
ence to the absence of donative intent on the part of the transferor.
In dealing with the defectiveness of a transfer of value, the law has
to observe a strict equality in its treatment of plaintiff and defendant.38
The basis of the plaintiff ’s claim is that, in the absence of donative
intent on her part, the transfer of something for nothing has deprived
her of value that belonged to her without her having freely parted
with it. If the plaintiff ’s lack of donative intent were to suffice for li-
ability, defendants in turn would be deprived of what belongs to them
without their having freely parted with it. The principle of justice in
transfer, that one is entitled to what one owns until one freely parts
with it, would thereby be inconsistently asserted for plaintiffs and
denied for defendants. To vindicate justice in transfer, the law must
apply it equally to both parties.

5. Acceptance
Accordingly, a finding that the defendant cannot retain an enrichment
made at the plaintiff ’s expense has to be based on considerations that
implicate both of the parties in their relationship to each other. In this
connection two ideas, one normative and the other contextual, are
relevant. The normative idea is that, with respect to the beneficial
transfer, the wills of the two parties are so related to each other as to
converge on the reason for not allowing the defendant to retain the
benefit. On the plaintiff ’s side, this reason consists in the absence of an
37
Similarly, in the case of pure services, the transfer of value has left no residue that is
independent of the totality of the transferee’s entitlements.
38
For an exposition of this point, see Abraham Drassinower, “Unrequested Benefits in the
Law of Unjust Enrichment,” (1998) 48 U.T.L.J. 459, at 477.
204 corrective justice

intention (or of an obligation) to give the defendant something for

nothing. On the defendant’s side, the reason consists in the defendant’s
acceptance of the benefit as non-gratuitously given. Acceptance here
refers not to an express affirmation by the defendant but to the inte-
gration of the benefit into the defendant’s purposes. It is present when
the beneficial transfer is consonant with the defendant’s projects, so
that the defendant can plausibly be regarded as satisfied with the non-
donative transfer of value. If the plaintiff did not intend to give some-
thing for nothing, and if the defendant accepted the benefit as not
having been given for nothing, then an obligation to restore the value
arises. The plaintiff cannot retain gratis what was neither given gratis
nor accepted as given gratis.
The contextual idea is that of entanglement. Entanglement occurs
when the benefit can no longer be separated from what the defendant
is otherwise entitled to. The provision of a service is a paradigmatic
example. A benefit, once entangled, can no longer be an independent
object of the defendant’s choice. Nor can the benefit be returned to
the plaintiff in its disentangled state. Because the benefit has merged
into the defendant’s entitlements, the defendant can no more be
required to part with it or its value than she can be required to part
with whatever else she owns. The only circumstance in which this
does not hold is if defendant has accepted the benefit as non-dona-
tively transferred. Then the wills of the parties converge on the non-
donativeness of the transfer, precluding the defendant from retaining it
as if it were a gift.
The classic instance of entanglement appears in a famous statement
by Chief Baron Pollock: “One cleans another’s shoes; what can the
other do but put them on? . . . The benefit of the service could not be
rejected without refusing the property itself.”39 In this graphic exam-
ple, the benefit has been so completely entangled in the recipient’s
property, that the cleaning of the shoes cannot be treated as an object
of choice that is independent of his use of them. If one assumes, as
Chief Baron Pollock clearly does, that the shoes were cleaned without
the owner’s knowledge, to compel the owner to make restitution of
the value of the cleaning would be to hold her liable for owning shoes
whose condition was changed without her consent. Nor can the mere
wearing of the shoes count as an acceptance, for this would derogate
from the pre-existing entitlement of the owner, as having exclusive

39
Taylor v. Laird [1856] 156 E.R. 1203.
 unjust enrichment 205

dominion over the shoes, to owe nothing to anyone else through their
lawful use. Entanglement means not merely that the transferred value
has been absorbed into the totality of the defendant’s entitlements—
this happens to all unjust enrichments—but that within that totality it
cannot be separated from the other components, as here the cleanness
of the shoes cannot be separated from the shoes themselves.
Entanglement sets the context within which acceptance operates to
link the parties’ wills to the non-donative transfer. Once entanglement
occurs, the defendant can be held liable only if the non-donatively
transferred benefit was accepted. In contrast, the situation prior to
entanglement allows for an additional option, at least in theory.
Although acceptance may also be present prior to entanglement, as
where the defendant requests a benefit or acquiesces in its conferral,
this is not the sole circumstance of liability. So long as the benefit
remains disentangled, the defendant can be held liable even if the
defendant’s situation or conduct can be construed as a rejection of the
benefit. Because the rejected benefit is still separately available for
return without affecting what the defendant is otherwise entitled to,
no impediment exists to returning the value that neither party wants
transferred.
Leaving aside the unusual case of rejected benefit, one may say that
there are two obligation-creating conditions for liability. The plaintiff-
oriented condition is that the benefit was non-donatively transferred.
The defendant-oriented condition is that the benefit was accepted as
non-donatively given. Although these obligation-creating conditions
are each oriented to a different party, they share a common focus that
normatively links the plaintiff as transferor of value and the defendant
as transferee of value. Because a transfer of value is the giving of some-
thing for nothing, the common focus is on how the parties stand with
respect to the gratuitousness of what one gave and the other received.
The point of these two conditions is that their joint presence renders
the obligation to restore the transferred value consonant with the free
will of both parties.
These two conditions do not refer to disconnected moments. Rath-
er, to accept the benefit is to accept it as given. In the movement of
the value from the plaintiff to the defendant, the non-gratuitousness
with which the plaintiff transfers the value is completed by the defend-
ant’s acceptance of the transfer as non-gratuitous. Because acceptance
is relevant to liability as a relational phenomenon, the juridically per-
tinent notion of acceptance is of the benefit considered not on its
206 corrective justice

own, but as it has arisen through the parties’ interaction. The absence
of donative intent in the plaintiff is crucial to that interaction. Accord-
ingly, acceptance establishes a relationship not of the defendant to the
benefit, but of the defendant to the plaintiff though the benefit. If the
benefit has been given as a gift, the acceptance of it is as a gift, and no
liability follows. But if the benefit has been non-donatively given, the
consonance of the benefit with the defendant’s will marks an accept-
ance as non-donatively given.
When the benefit is accepted by the defendant on the same non-
donative basis with which the plaintiff gave it, the defendant cannot
justly retain it as if it had been given and accepted as a gift. The two
obligation-creating conditions thereby have an analogous function in
the law of unjust enrichment to that of offer and acceptance in the
law of contract: they link the wills of the parties to each other through
the subject matter of the transaction, so as to establish an in personam
legal relationship between them. In the contractual context, the par-
ties’ wills converge on the contractual performance offered by the
promisor and accepted by the promisee, with the effect of creating a
contract between them. In the unjust enrichment context, the parties’
wills converge on the non-gratuitous transfer of value, with the effect
of creating not a contract, but a right to the retransfer of the value.
Because it connects the defendant’s will to the intention of the plain-
tiff in conferring the benefit, acceptance is a relational idea.
In accordance with their relational significance, the two obliga-
tion-creating conditions have reference to the juridical world of pub-
lic meaning that the two parties share. On the plaintiff ’s side, the
notion of donative intent is an extended one. It goes beyond subject-
ive intent to include situations in which, whatever the transferor’s
subjective intent, the background legal categories justify the imputa-
tion of an intention to bestow a gift. In this extended sense, donative
intent draws on the public meaning that the plaintiff ’s action has in
the relationship between the parties. Imagine, for example, that the
plaintiff makes an unrequested improvement to property that he
knows belongs to another in the hope of being compensated for his
labor. Subjectively, he may have no intention of giving a gift. But
because his action takes place within a legal regime under which, as
he knows or ought to know, only the owner has the right to deter-
mine whether to improve one’s property, the improver can be taken
to know that his action cannot obligate the owner to pay for the
improvement. Accordingly, the law treats his action as the bestowal of
 unjust enrichment 207

a gift. The background legal category of property, which recognizes

in the owner the exclusive power to improve the condition of what
is owned, justifies the law’s viewing the improvement as the expres-
sion of a donative intent. In this example the imputation of donative
intent is based not on what is subjectively within the plaintiff ’s mind,
but on how the plaintiff ’s conduct is to be publicly understood and
categorized in relation to the defendant’s property. Accordingly, one
can conclude that having transferred the value free of any obligation
on the transferee, the improver is a risk-taker with respect to the
hoped-for compensation.40 Conversely, however, if the improver mis-
takenly thinks that the property is his own or that he is improving it
at the owner’s request, donative intent can no longer be imputed to
him. Because the improver is unaware that his improvement was not
authorized, he cannot be held to what is implied by the knowledge
that the power to improve property is exclusively the owner’s. For
the improver who acts out of mistake or ignorance, an obligation-
creating condition is in place.
On the defendant’s side also, acceptance of the enrichment as non-
gratuitously given refers to the absence of donative intent in the trans-
fer of value. It imbues this absence of donative intent on the plaintiff ’s
part with a relational significance by connecting it to an imputed
expression of the defendant’s free will. The non-gratuitousness of the
transfer—the consideration at the heart of unjust enrichment—there-
by embraces and normatively links the parties. If the defendant can be
regarded as having accepted a benefit as non-gratuitously given, then
in fairness the benefit cannot be retained gratis. Nor can the defend-
ant be compelled to give something in exchange for that benefit as if
there were a contract between the parties, since the benefit was not
given as part of an enforceable exchange or with the intention to cre-
ate a contractual relationship. All that the defendant can do is return
the value, so as to avoid keeping as a gift what was neither given nor
accepted as a gift.
Acceptance is thus a relational notion. It refers to what is to be
imputed to the defendant in the light of the plaintiff ’s non-gratuitous
transfer of value. Although it is defendant-oriented, it does not treat the
defendant in isolation from what the plaintiff did. It is not one “unjust
factor” in a list of unjust factors. Nor does it point to a moral failure
consisting in the defendant’s unconscientious receipt of something for

40
Birks, above n. 26, 101–3.
208 corrective justice

which he or she wants to avoid payment.41 Rather, as a member of the

conceptual sequence that unites the transferor and transferee of value
within an obligation-creating relationship, it is a structural feature of
liability for unjust enrichment. Within that relationship the defendant’s
acceptance of the beneficial transfer as non-gratuitous and the plaintiff ’s
lack of donative intent are correlatives.
As with donative intent, the idea of acceptance draws on the public
meaning of the parties’ interaction. What matters is not the defendant’s
inner psychological state, but the judgments and assumptions about
the parties’ interaction that can reasonably be made against the back-
ground of the legal structure in which they operate.42 In particular,
the defendant who receives something for nothing has no reason to
assume that the benefit was given gratuitously. Private law is a legal
regime through which parties act for their own purposes without
subordinating to others their freedom or the means for realizing it.
The law does not presume—and therefore those subject to the law are
not entitled to presume—that someone has chosen to transfer value
gratuitously, thereby surrendering the means for pursuing one’s own
ends. To be sure, a person may on occasion identify another’s interest
with one’s own and therefore confer gratuitous benefits on the other.
However, such donative intent must be established for each particular
case, and not assumed to be the general rule. Except when the enrich-
ment was intended and accepted as a gift, the defendant can be regard-
ed as assuming that no benefit is given gratuitously, even if the
defendant has not turned his mind to this issue.
Acceptance is imputed when the law can reasonably regard the
beneficial transfer as something that forwards or accords with the
defendant’s projects. The imputation of acceptance thereby connects
the law’s construction of the defendant’s will both to the transferred
value and to the terms on which it was transferred. In this context the
will—the capacity to set and pursue one’s own purposes—is a jurid-
ical, not a subjective or psychological notion: what matters is the pur-
pose not as internally formed but as externally pertinent to the
relationship of plaintiff and defendant. Awareness of the benefit and
acting with respect to it are sufficient but not necessary to indicate

41
For “free acceptance” as an unjust factor or as a signal of unconscientious receipt, see
Birks, ibid., 114, 266; Peter Birks, “In Defence of Free Acceptance,” in Essays on the Law of
Restitution, ed. Andrew Burrows (1991), 105.
42
Compare Deane J.’s reference to “a benefit actually or constructively accepted” in Pavey
& Matthews Proprietary Ltd. v. Paul [1986] 162 C.L.R. 221, at 257.
 unjust enrichment 209

acceptance; a benefit can be consonant with the defendant’s purposes

even if these are lacking. Acceptance goes, accordingly, not to the
defendant’s particular psychological state, but to what the law can rea-
sonably impute to the defendant, given the defendant’s purposes and
the law’s background assumptions about the significance of donative
intent. Of course, any defendant might subjectively prefer to keep the
transferred value rather than return it to the transferor. Nonetheless,
by imputing acceptance of the enrichment as non-gratuitously given,
the law indicates its view of how the defendant’s will can reasonably
be regarded as standing with respect to what was received, with the
implication that the defendant has no right to retain it for the service
of his projects as if it had been given gratuitously.
One might think that, because acceptance can occur subsequent to
receipt of the benefit,43 ascribing a structural role to acceptance is
inconsistent with the cause of action being complete upon receipt.
This, however, misapprehends the effect of acceptance. What the
recipient accepts is not the benefit conceived statically at the time that
the acceptance becomes operative, but rather the benefit as transferred.
This includes not only the transfer’s non-donativeness but also its
occurrence at a particular point of time. The acceptance, in other
words, is of the transfer whose subject matter is the benefit, not of the
benefit standing alone, for it is the transfer that links the parties to
each other. Thus, even when the acceptance comes after the benefit’s
receipt, its effect is retrospective to the time of the receipt. This is why
the moment of enrichment can reasonably be regarded as the time
from which judgment interest44 and the statute of limitations45 run.

6. Situations of acceptance
The consonance of the benefit with the defendant’s purposes can arise
in three ways: through action by the defendant with respect to the
benefit, through a specific project of the defendant’s that the benefit
forwards, or through a benefit—money—that fits with any project
that the defendant might have. These different ways form the various
situations in which acceptance can be imputed to the defendant.

43
For example, McDonald v. Coys of Kensington [2004] 1 W.L.R. 2775 (C.A.).
44
Woolwich Equitable Society v. Inland Revenue Commissioners [1993] A.C. 70.
45
Robert Goff and Gareth Jones, The Law of Restitution, 7th ed. (2009), ss. 43-004–006.
210 corrective justice

In the first situation the defendant knows or takes the risk that the
benefit is non-gratuitously given and yet requests it or acquiesces in it
by foregoing the opportunity to refuse it. Some of the best-known
cases in the law of restitution illustrate this: the plaintiff performs a
service for the defendant under an unenforceable contract, which
serves as evidence both of the defendant’s request for the service and
of the plaintiff ’s non-donative intent in providing it;46 or the defend-
ant has been enriched by the plaintiff ’s labor in a quasi-spousal rela-
tionship although he knew or ought to have known that the benefit
was given to him not as a personal gift but as an reflection of the full
integration of their economic well-being;47 or the owner “lies by”
when he knows that another is expending money to improve the
property on the mistaken supposition of his own title.48 The same
holds if after receipt the defendant refuses to restore a non-gratuitous-
ly given benefit that is easily returnable.49 In such cases, the defend-
ant’s action or inaction in the face of the non-gratuitous conferral can
be regarded as an acceptance of those benefits as given without dona-
tive intent.50
By failing to take the opportunity to reject a benefit, one both
expresses one’s free will with respect to it and assumes responsibility

46
Deglman v. Guarantee Trust Co. and Constantineau [1954] S.C.R. 725; Pavey & Matthews,
above n. 42.
47
Pettkus v. Becker [1980] 117 D.L.R. (3d) 257, at 274:
[W]here one person in a relationship tantamount to spousal prejudices herself in the reasonable
expectation of receiving an interest in property and the other person in the relationship freely accepts
benefits conferred by the first person in circumstances where he knows or ought to know of that
reasonable expectation, it would be unjust to allow the recipient of the benefit to retain it.
48
Ramsden v. Dyson [1866] L.R. 1 H.L. 129.
49
McDonald v. Coys of Kensington [2004] 1 W.L.R. 2775 (C.A.); at paras. 37–38, Mance
L.J. describes the defendant’s action in keeping the personalized mark as a choice and as the
exercise of a deliberate preference.
50
It is sometimes thought that this notion of acceptance through inaction is in tension
with the absence of tort liability for omissions; see G. Mead, “Free Acceptance: Some Further
Considerations,” (1989) 105 L.Q.R. 460, 463–64; Andrew Simister, “Unjust Free Acceptance,”
(1997) Lloyd’s Maritime & Commercial L.Q. 103, 118–20. Properly understood, however, the
absence of tort liability is not about inaction as such but about conferring a benefit to which
the recipient has no right; see Peter Benson,“The Basis for Excluding Liability for Economic
Loss in Tort Law,” in Philosophical Foundations of Tort Law, ed. David G. Owen (1995), 427,
447–49. In the unjust enrichment context, the imputed acceptance through the defendant’s
inaction does not reflect a duty to bestow a gratuitous benefit on the plaintiff , but goes rather
to whether the defendant’s non-gratuitously given benefit can be treated as non-gratuitously
accepted. Indeed, there is a deep harmony between the absence of liability for nonfeasance
and the requirement of acceptance: both actualize the parties’ freedom of choice by express-
ing the law’s antipathy to gratuitous benefits that are not gratuitously intended.
 unjust enrichment 211

for the implications of its non-gratuitous nature. Accordingly, by com-

pelling a retransfer of the value, the law is not acting inconsistently
with the defendant’s will; such liability does not fall foul of Lord
Justice Bowen’s famous dictum that “[l]iabilities are not to be forced
upon people behind their backs.”51 Rather, the law is merely follow-
ing through on how the defendant’s will is related to the plaintiff ’s
through the transfer of value. For example, a defendant who is aware
that another is bestowing an apparently gratuitous benefit and does
not intervene to prevent it takes the risk that donative intent is absent.
By allowing the enrichment to occur, the defendant manifests his
volition with respect to it. If it turns out that the plaintiff indeed had
no donative intent (for example, if the plaintiff was improving the
defendant’s land on the mistaken impression that it was his own), then
the defendant’s failure to prevent the transfer can be considered an
acceptance of that benefit as non-gratuitously given.
In the second situation for imputing acceptance, the law treats the
defendant as having accepted the beneficial transfer because, given the
nature of the defendant’s activities and projects, the defendant has no
reason not to accept it. For example, the defendant holds property
destined for a particular use or disposition that is forwarded by the
benefit that the plaintiff non-gratuitously conferred;52 or the plaintiff
discharges an obligation owed by the defendant;53 or a director exer-
cises his skill to the advantage of the corporation although in breach
of his fiduciary duty.54 In such instances the issue is not whether the
defendant as a rational maximizer is better off with the benefit in
some global sense, but whether the benefit forwards the specific pur-
poses implicit in the defendant’s antecedent activities. If it does, then
requiring restitution of the transferred value is, from the public stand-
point of the parties’ relationship, consistent with the defendant’s free
will.
Under the heading of “incontrovertible benefit” this second situ-
ation is conventionally treated as establishing the enrichment rather
than the unjustness of retaining it. In this respect incontrovertible bene-
fit is the counterpart of subjective devaluation. But as with subjective

51
Bowen L.J. in Falcke v. Imperial Insurance Co. [1886] 34 Ch. D. 234 (C.A.).
52
Lac Minerals v. International Corona Resources [1981] 64 D.L.R. (4th) 14 (S.C.C.)
(defendant developed mine and constructed mill on plaintiff ’s mining property); Greenwood
v. Bennett [1973] 1 Q.B. 195 (C.A.) (improvements to a car that was to be sold).
53
Brooks Wharf & Bull Wharf v. Goodman Bros. [1936] 3 All E.R. 696 (C.A.).
54
Phipps v. Boardman [1967] 2 A.C. 46.
212 corrective justice

devaluation, the considerations for postulating an incontrovertible

benefit go not to whether a transfer of value has taken place but to
whether the defendant’s retention of the transferred value is conson-
ant with the parties’ free will. The point of invoking incontrovertible
benefit is to show that imposing an obligation to make restitution
would not violate the defendant’s freedom of choice: “[t]he principle
of incontrovertible benefit . . . exists when freedom of choice as a prob-
lem is absent.”55 If this is the case, it should be situated where it struc-
turally belongs: as an obligation-creating condition pertaining to value
transferred without donative intent.
A difference between these two situations is that in the first, but not
in the second, the entanglement of the benefit with the defendant’s
entitlements is a barrier to recovery. The reason for this is that the two
situations relate the defendant’s will to the benefit received in different
ways. In the first situation, instanced by request or acquiescence, the
defendant’s acceptance has to be referable to the benefit as such without
cutting into the use that the defendant is otherwise entitled to make of
what she owns. Therefore, the defendant’s will has to be specifically
directed to the benefit independent of the defendant’s use of the owned
thing. Once the benefit becomes entangled in the defendant’s entitle-
ments without indication of the defendant’s acceptance, the defendant is
not liable for enjoying the benefit through the use of what belongs to
her. In the second situation, in which the benefit is incontrovertible, the
acceptance is imputed because the benefit forwards the use that the
defendant would otherwise have made. Unlike the first situation, here
the actual or prospective use shows the consonance of the benefit with
the defendant’s projects, and is therefore the reason for regarding the
benefit as accepted. Because the will need not be directed to the benefit
independently of the use, the benefit’s inextricable entanglement with
what the defendant is otherwise entitled to use poses no barrier to lia-
bility.Thus, the distinction between the two situations is that in the first,
acceptance is independent of use and, accordingly, cannot operate on an
entangled benefit, whereas in the second, acceptance occurs through
use, thereby rendering entanglement irrelevant.
Finally, the third situation for imputing acceptance is the non-
donative payment of money. The peculiarity of money is that, as the

55
Regional Municipality of Peel v. Her Majesty the Queen in the Right of Canada
[1992] 98 D.L.R. (4th) 140 (S.C.C.), quoting J. R. M. Gautreau, “When Are Enrichments
Unjust?” (1988–89) 10 Advoc. Q. 258, at 271.
 unjust enrichment 213

universal medium of exchange,56 it forwards any and every specific

purpose that the defendant might have. The payment of money, there-
fore, is an incontrovertible benefit.57 To be sure, in unusual circum-
stances a defendant might reject such a benefit—for example, if the
mistaken payment makes the defendant ineligible for means-based
government services.58 But until spent, money is not entangled in the
defendant’s other entitlements and so can be returned even if rejected.
Thus whether it is accepted or rejected, the defendant is under an
obligation to make restitution.
Aside from such unusual circumstances, money is an incontrovert-
ible benefit. As such it differs in two ways from the non-monetary
instances of incontrovertible benefit. First, acceptance is imputable for
a non-monetary benefit only if, given the benefit’s particular qualities,
it forwards some particular project that the recipient has. In contrast,
except in unusual circumstances, money forwards the recipient’s
projects, whatever they are. Second, whereas in the case on non-mon-
etary benefit the imputed acceptance arising from the benefit’s incon-
trovertibility operates despite being entangled in the defendant’s other
entitlements, money has no such entanglement. However, the receipt
of the money may lead the recipient to spend it on projects that she
would not otherwise have undertaken—that is, to make the extraordin-
ary expenditure that constitutes a change of position.59 Accordingly, a
benefit that subjects the defendant to liability while in the form of
money may, when spent, become entangled in the defendant’s entitle-
ments. The imputation based on money’s being the universal medium
of exchange is then no longer appropriate. As long as the money is
unspent, the defendant’s position is no different than what it would be
in the second situation. Once the defendant changes position by mak-
ing an extraordinary expenditure that entangles the benefit in the
defendant’s entitlements, the defendant’s position is not different from
that of the person with the cleaned shoes in the first situation.
This account requires that change of position through expenditure,
like incontrovertible benefit, should be regarded as going to the

56
B.P. Exploration Co. (Libya) v. Hunt (No. 2) [1979] 1 W.L.R. 783, at 799 (Goff J.).
57
Graham Virgo, The Principles of the Law of Restitution, 2nd ed. (2006), 75.
58
Consider the situation in Ontario (Ministry of Community & Social Services) v. Hen-
son [1987] 28 E.T.R. 121 (Ont. Dist. Ct.), affirmed on appeal (1989) 36 E.T.R. 192 (Ont.
C.A.). I am grateful to Jason Neyers for pointing this out.
59
Rural Municipality of Storthoaks v. Mobil Oil Canada [1975] 55 D.L.R. (3d) 1
(S.C.C.).
214 corrective justice

unjustness, not the existence, of the enrichment. Change of position is

now conventionally explained in terms of enrichment, that the
defendant has been “disenriched.”60 The explanation is not without
difficulty. The consumption of the value is the exchange of the trans-
ferred value for some good to which the consumer attaches a still
greater value. At the general level one may ask why such consumption
negates the legal effect of the original transfer. Nor is the idea of dis-
enrichment unambiguous. If I have spent the mistaken payment on a
trip around the world,61 I am still enriched by what I did with the
money—I now have the recollection of adventure and discovery, the
slides and photographs, the seemingly inexhaustible store of conversa-
tional material—even though the money is no longer in my bank
account. What has occurred is a transformation of the enrichment,
not its disappearance.
A more particular form of this difficulty is that the defense of
change of position sometimes applies even when one can still discern
the enrichment among the defendant’s assets. Say that the defendant,
having received a mistaken payment from the plaintiff , makes the
extraordinary expenditure of throwing out her old shoes and buying a
shining new pair. If the defendant had no notice of the mistake at the
time of the purchase, the defense of change of position is available
even though the defendant seems to remain enriched.62 The argument
to this conclusion on enrichment grounds is that—the defendant’s
choice to buy new shoes having been vitiated by her mistaken belief
about her resources—she can subjectively devalue the shoes.63 This
preserves the relevance of the defense to enrichment, at the cost, oddly,
of saying that the defendant can subjectively devalue what she decided
was worth purchasing.
Even in its own terms this argument, with its reference to the viti-
ation of the defendant’s choice, is about freedom of choice, not enrich-
ment. The impression that the change of position in this instance is
about enrichment is the product of the idea that subjective devalu-

60
Birks, above n. 26, 208; at 261 Birks allows that there may turn out to be “very rare”
examples of unjust-related change of position.
61
Lord Templeman’s example in Lipkin Gorman v. Karpnale [1992] 4 All E.R. 512, at 517.
62
Cf. RBC Dominion Securities v. Dawson [1994] 111 D.L.R. (4th) 230 (Nfld. C.A.). I
have substituted new shoes for new furniture in order to bring the case close to Pollock
C.B.’s famous comment.
63
Peter Birks, “Overview: Tracing, Claiming and Defences,” Laundering and Tracing, ed.
Peter Birks (1995), 289, at 331–32 .
 unjust enrichment 215

ation is also about enrichment. At bottom, however, subjective devalu-

ation is (as is widely acknowledged) really about freedom of choice.
Clarity about the structure of unjust enrichment requires that issues
of freedom of choice be treated as going to the unjustness of retaining
the value rather than to the occurrence of the transfer. This applies
also to change of position through expenditure, which constitutes a
defense not because it disenriches but because it precludes an
obligation-creating condition.64
Accordingly, the legal consequences of a mistaken payment can be
conceptualized as follows. A payment made by mistake lacks donative
intent, thereby fulfilling the first obligation-creating condition. Given
that money forwards any purpose that the recipient has, the recipient
is incontrovertibly benefited and (except in unusual circumstances
that do not affect the final result) has no reason not to be viewed as
accepting it. However, subsequent events can confirm or disconfirm
the acceptance that would have been imputable on payment. Once
the recipient is made aware that the unspent money was given by mis-
take, acceptance of the money as non-gratuitous becomes conclusive
because the recipient now has notice of the very circumstances that
grounded its imputation. Notice to the defendant that the payment
was mistakenly made cuts off the possibility of the defendant’s subse-
quently spending the money in a way that would constitute a change
of position, because the defendant would be spending money that she
knew or should have known she was obligated to return to the plain-
tiff .
Conversely, when the recipient changes position prior to knowing
the payer’s mistake, the circumstances supporting the imputation of
acceptance that was available previously no longer obtain. Change of
position operates as a defense insofar as the spending of the enrich-
ment entangles it in the recipient’s entitlements. If it has been spent to
purchase something separate from the defendant’s other entitlements,
the value remains disentangled; restitution can be made of the second-
hand value of what was purchased.65 If, however, the expenditure has

64
The notion that change of position involves the defendant’s disenrichment is more
apposite to the case of the innocent recipient of a mistaken payment that is subsequently
stolen or lost. There the defense is applicable even if the defendant had knowledge of the
mistake. See National Bank of New Zealand Ltd. v. Waitaki International Processing (NI)
Ltd. [1999] 2 N.Z.L.R. 211, 228–29 (C.A.).
65
This is Lord Templeman’s example of the purchased automobile in Lipkin Gorman v.
Karpnale [1992] 4 All E.R. 512, at 517.
216 corrective justice

consumed the value in such a way that nothing separately ascribable

to it remains, then the enrichment is irrecoverable. Whether the
defendant has spent the money on getting her old shoes cleaned or on
replacing the old shoes with a new pair of clean shoes, her situation is
no different from that person whose shoes are mistakenly cleaned by
another. An obligation to restore such an enrichment would not be
consistent with the defendant’s freedom of choice. Pollock C.B.’s
observation, that the benefit could not be rejected without refusing
the property itself, applies.
The three situations described in this section correspond to the
three ways of aligning the defendant’s projects with the benefit
bestowed by the plaintiff . The first aligns the project with a specific
benefit that the defendant wishes to attain either by request or by
non-rejection. The prospect of the benefit is what causes it to be
incorporated into the defendant’s purposes. The second is the con-
verse of this: it aligns the benefit with a specific project that is other-
wise evident in the defendant’s activities. Because of the defendant’s
particular purposes the benefit that forwards them is regarded as
accepted. The third deals with money as the all-purpose means for
forwarding any project and with the consequences of transforming
the money received through expenditure on a specific project.66 In all
of these situations, the benefit that the plaintiff non-donatively gives
fits the purposes that the defendant pursues. This fit is the basis for
imputing acceptance to the defendant.
One might object that only in the first kind of situation, and not
in situations of incontrovertible benefit, is acceptance genuinely
present. If so, acceptance cannot be a general feature of liability for
unjust enrichment. Acceptance (so the objection would go) involves
acting out of awareness of the existence or potential existence of the
benefit. This is how acceptance operates in cases of request or acqui-
escence. In situations of incontrovertible benefit, in contrast, the
defendant can be held liable even if unaware of the benefit. If the

66
These different ways of aligning the benefit bestowed by the plaintiff and the projects
pursued by the defendant seem to constitute a more or less exhaustive taxonomy. The only
addition necessary for completeness is the converse of the third situation. The third situ-
ation features a possible movement from the universality of money to the specificity of
change of position through expenditure. The converse is a movement from the specificity
of the benefit to a transformation of the benefit into money.This last possibility corresponds
to Birks’s view that a benefit becomes incontrovertible when realized in money; Birks,
above n. 26, 221.
 unjust enrichment 217

point of acceptance is to implicate the defendant’s will so that, in

the words of Lord Justice Bowen’s famous dictum mentioned earlier,
“[l]iabilities are not to be forced upon people behind their backs,”67
the defendant’s ability to choose whether or not to take the benefit
cannot be dispensed with. Acceptances that are constructed or
imputed on the basis of the consonance of the benefit with the
defendant’s projects will not do.
The answer is that in this context the will has a relational, and not
merely an interior, significance. Only on this basis is it appropriate to
corrective justice, which conceives of liability’s norms as interactional,
rather than as unilaterally applicable to one or the other of the two
parties. Accordingly, the role of the benefit in motivating the defend-
ant’s conduct is not the sole relevant factor. What matters is the con-
nection between the defendant’s will and the benefit bestowed by the
plaintiff. This connection can take different forms. There is no reason
why the difference between aligning purpose with benefit (situation
one) and aligning benefit with purpose (situations two and three)
should matter to the plaintiff ’s liability. What is relationally significant
is the idea that includes all three situations: the benefit that the plain-
tiff non-donatively gives fits the purposes that the defendant pursues.
The imputation of acceptance merely expresses the relational signifi-
cance that all the situations share.
As for Lord Justice Bowen’s dictum about not forcing liability
behind a person’s back, the objection proves too much. If the dictum
is understood as applying only to situations in which the defendant
was aware of the benefit, then incontrovertible benefit would not give
rise to liability at all, regardless of whether incontrovertible benefit
was seen as going to enrichment or unjustness. But although doubt is
sometimes cast on the soundness of the doctrine of incontrovertible
benefit on the grounds that it amounts to a forced exchange,68 the
doctrine is now solidly entrenched in the law. Consequently, unless
one thinks that either the dictum or the doctrine of incontrovertible
benefit is wrong, the two have to be brought into harmony. This har-
mony is achieved if one recognizes that, because liability for an incon-
trovertible benefit is based on the defendant’s purposes, it does not
operate behind the defendant’s back.

67
Above n. 51.
68
Andrew. M. Tettenborn, Law of Restitution in England and Ireland, 2nd ed. (1993), 20–22.
218 corrective justice

7. The sequence of elements

The obligation-creating conditions of lack of donative intent and
acceptance echo the conditions for the defectiveness of gift, with the
exception that they are cumulatively necessary for rendering the trans-
fer defective rather than alternatively sufficient. As noted earlier, to
make out a gift, what is given as a gift has to be received as a gift.69
When these two conditions are present, the donee acquires the right
to the gifted object; when either of them is absent, the object remains
the donor’s. Unjust enrichment has aptly been called the law of non-
gifts.70 An obligation to make restitution arises when what is given as
a non-gift is accepted as a non-gift—that is, when the plaintiff ’s unin-
tended transfer of something for nothing is matched by the defend-
ant’s acceptance of what was transferred as something that was not
intended to be given for nothing. This non-gratuitousness on both
sides of the relationship between the transferor and the transferee of
value triggers a reversal of the transfer, because the transferee cannot
retain the value on terms other than those on which it was given and
accepted.
So understood, the elements of liability form a sequence. The first
stage in this sequence is to determine whether the plaintiff gave the
defendant something for nothing—a stage formulated legally as the
defendant’s enrichment at the plaintiff ’s expense and theoretically as a
transfer of value. If something was indeed given for nothing, one then
moves to a series of questions that address the justice of the defend-
ant’s retaining what was given. The first of these questions is whether
the plaintiff intended either a gift or the discharge of an obligation to
the defendant.71 An affirmative answer means that the claim is defeated.

69
Above n. 31.
70
Drassinower, above n. 38, at 478.
71
The obligation at issue must be one that is owed to the defendant and not to some third
party. Because the law’s interest is in the relationship between the transferor and the trans-
feree of value, only the plaintiff ’s discharge of an obligation to the defendant entitles the
defendant to retain the benefit. Accordingly, Owen v. Tate [1976] 1 Q.B. 402 (C.A.) was
wrongly decided. As a favor to a friend, the plaintiff substituted his personal guarantee for
security given by the friend for a bank loan to the defendant, and then unsuccessfully sued
for reimbursement when the security was applied to part of the debt. The plaintiff intended
to benefit his friend, not to give a gift to the defendant. Moreover, since the plaintiff ’s obliga-
tion on the guarantee was to the bank not the defendant, it supplied no reason for the
defendant to be entitled to retain the benefit of having its debt to the bank reduced to the
extent of the plaintiff ’s payment.
 unjust enrichment 219

A negative answer, concluding that the plaintiff gave something for

nothing but had no donative intent, leads to the final question in the
sequence: did the defendant accept the transferred value as non-dona-
tively given? An affirmative answer to this question means that the
defendant cannot justly retain the enrichment and is under an obliga-
tion to restore it to the plaintiff .
The leading Canadian case of Deglman v. Guarantee Trust Co. and
Constantineau72 illustrates this sequence. The plaintiff agreed to per-
form incidental services for his aunt in return for an oral promise that
in her will she would leave him her house. The aunt died without
doing so. The aunt’s promise was unenforceable under the Statute of
Frauds. Nonetheless, the nephew succeeded in getting restitution of
the value of the services provided. The first stage was satisfied because
the performance of the services at the aunt’s request was a transfer of
value. Turning to the question of justice in transfer, one might be
tempted to think that, on the sequence that I have suggested, the claim
should have been dismissed on the ground that this transfer of value
was in fulfillment of a contractual obligation to the aunt, an obligation
that was not voided but merely rendered unenforceable by the statute.
However, for purposes of unjust enrichment, the relevant question is
not whether the plaintiff was obligated to do certain acts, but whether
he was obligated to transfer value—that is, to do those acts without
receiving their quantitative equivalent in return. Although the nephew
was contractually obligated to provide the services, he was not con-
tractually obligated to provide them for nothing, as if he had made a
gratuitous promise under seal. When the aunt died without leaving
the house to the nephew, what the parties intended as an exchange of
services for property was revealed to be a transfer of value, to which
the nephew had not obligated himself. Nor did the plaintiff transfer
the value donatively. Although the unenforceable contract could not
be the basis for compelling performance, it showed that the services
were not provided gratuitously.73 Similarly (although the court did
not expressly make this point), the contract showed that the aunt
accepted the services on the assumption that she was not getting them

72
[1954] S.C.R. 725.
73
“[T]he services were not given gratuitously but on the footing of a contractual relation:
they were meant to be paid for.” Justice Rand, ibid., at 728; compare Justice Deane’s observa-
tion in the parallel Australian case that “it will ordinarily be permissible for the plaintiff to
refer to the unenforceable contract as evidence, but as evidence only, on the question whether
what was done was done gratuitously.” Pavey & Matthews, above, n. 42, at 257.
220 corrective justice

for nothing.74 Because the aunt accepted as non-gratuitous the ser-

vices that the nephew performed on a non-gratuitous basis, her estate
could not retain their value as if they had been given for free.
Compared to the conventional understanding, this description of
the structure of unjust enrichment shifts considerations usually associ-
ated with enrichment (subjective devaluation, incontrovertible benefit,
change of position) to the obligation-creating conditions that make
retention of the enrichment unjust. A consequence of the conven-
tional placement is that the overloading of the “enrichment” slot
within the principle of unjust enrichment involves the emptying of
the “unjust” slot. It is then hardly surprising that the nature of the
unjustness becomes a mystery. Once the “unjust” slot has been refilled
in the way I have suggested, the meaning of unjustness in the context
of unjust enrichment follows from the nature of enrichment at an-
other’s expense as a transfer of value. A transfer of value is the giving
or doing of something for nothing. The issue of justice as between the
parties that arises from a transfer of value centers on the normative
defectiveness of a transfer in which the transferee retains for nothing
what was both given non-gratuitously and accepted as non-gratuit-
ously given. The question of what renders an enrichment unjust is,
therefore, an enquiry not into justice at large but into the justice that
is specific to the transfer of something for nothing.

8. The Kantian conception of the in personam right

Liability for unjust enrichment reflects the plaintiff ’s in personam right
to have the defendant retransfer the value. In this section I explore the
in personam nature of this right, drawing on Kant’s taxonomy of pri-
vate law. As was noted in the first section of this chapter, the identifi-
cation of the plaintiff ’s right in cases of unjust enrichment has aroused
considerable perplexity, especially among those who are skeptical
about the relevance of corrective justice to unjust enrichment.
Although Kant never discussed (and was presumably unaware of )
unjust enrichment as a ground of liability, his account of the in per-
sonam right might assist in dissolving the perplexity.
To what does the plaintiff assert a right in an action for unjust
enrichment? Since liability entails a retransfer of value, two possible

74
Ibid.
 unjust enrichment 221

answers are available, depending on how one thinks of the relationship

between the retransfer and the value. On one view, the plaintiff asserts
a right to the value, with the retransfer being merely the mechanism
for getting it. On the second view, the plaintiff asserts a right to the
retransfer, with the value being merely the object of the retransfer.
On the first view, the plaintiff simply asserts his or her right in the
value as originally owned, on the ground that this right survives the
defective transfer, just as the donor’s right in what is given survives
the defective gift. The action for unjust enrichment would be con-
ceived as a kind of action for conversion or as a kind of vindicatio of a
special asset that consists in the value. It would then turn out that lia-
bility for unjust enrichment refers not to a distinctive principle but to
a novel kind of asset.75
I think that this is not the best way to conceptualize the plaintiff ’s
right.The value returned to the plaintiff is not, as the subject matter of
a right, identical to the value that the plaintiff originally held: the
original value was an incident of the thing of value, whereas the value
returned to the plaintiff is independent of the thing of value, which
now rightfully belongs to the defendant. Moreover, one cannot con-
clude that the defectiveness of the transfer of value means that the
value remains the transferor’s, as in the case of gift. In the gift context,
defectiveness of transfer refers to the role of either party’s will in block-
ing the movement of title in the object of the gift. In contrast, in the
unjust enrichment context, defectiveness of transfer refers to the joint
role of both parties’ wills, precisely in order to take account of the
movement of title in the thing of value. Indeed, if the transferor had a
continuous right to the value, it is hard to see why the transferee’s
acceptance of the benefit should be indispensable to the transferor’s
assertion of that right. In view of this dissimilarity between gift and
unjust enrichment in what makes a transfer defective, one cannot
straightforwardly argue that defectiveness of transfer means that what
was purportedly transferred, whether a specific object or value, con-
tinues to be owned by the transferor.
Instead of looking at value as a static object of ownership that
remains the transferor’s through the defectiveness of the transfer, I sug-
gest that one regard value more dynamically, as the content of a pro-
cess of transfer and retransfer. The significance of the plaintiff ’s
entitlement to the value at the inception of the transaction is that

75
Klimchuk mentions this possibility, above n. 10, at 680.
222 corrective justice

value counts as the possible object of a transfer, and—if the obliga-

tion-creating conditions are present—as the possible object of a
retransfer, thereby making justice in transfer applicable to value. The
normative defectiveness that compels a retransfer itself partakes of the
dynamism of the process of transfer and retransfer, because the transfer-
or’s right to have the value retransferred is established not by reference
to the transferor’s intent alone, but through the convergence of the
parties’ wills on the transfer’s non-gratuitousness.
In unjust enrichment the plaintiff ’s right is not to the value as such,
but to a retransfer of the value. The object of the right is an action by
the defendant, not a thing. The right arises through the will-to-will
relationship of the parties—that is, through the unity of their wills
when the plaintiff ’s unintended transfer of something for nothing is
matched by the defendant’s acceptance of the transfer as an unintend-
ed giving of something for nothing. When this non-gratuitousness
appears on both sides of the relationship between the transferor and
the transferee of value, the law reverses the transfer, because the trans-
feree cannot retain the value on terms other than those on which it
was given and accepted. The reversal of the transfer in response to the
obligation-creating conditions is the plaintiff ’s right and the defend-
ant’s correlative duty.
This right to have the defendant perform an act can be presented in
Kantian terms. The matter of the plaintiff ’s right in the unjust enrich-
ment context is what Kant called a causality of the defendant’s will,
and its form is a ius personale (or, as we would put it, a right in person-
am).76 I now want to set out more fully what this means.
Kant’s conception of the in personam right emerges from two layers
in his classification of rights. The first is the division between innate
and acquired right. Innate right is the right that one has by virtue of
one’s very existence, so that one does not have to do anything to
acquire it.77 One’s physical embodiment is a manifestation of this right.
One does not acquire one’s body through the performance of any act
of acquisition. Indeed, the notion of such an act would be self-contra-
dictory. Because an act involves the use of one’s body, the act of acquir-
ing one’s own body would presuppose that one had a right to one’s
body prior to the act that acquired it. Kant specifies that the only
innate right is one’s freedom—that is, one’s independence from

76
Immanuel Kant, The Metaphysics of Morals, tr. Mary Gregor (1996), [6:259–60].
77
Ibid., at [6:237–38].
 unjust enrichment 223

constraint by the action of another, though this right has several aspects
(bodily integrity, freedom of speech and thought, immunity from
reproach until one has wrongly affected another’s rights, the innate
equality of not being bound by others more that one can in turn bind
them, and so on). The ensemble of the aspects that constitute one’s
innate right comprises what is internally one’s own in one’s relations
with others.
In contrast, acquired rights are rights to objects external to the per-
son, which become one’s own through an appropriate act of acquisi-
tion. Because these objects are distinct from the person and are
acquired through an act of the will, Kant calls them “external objects
of choice.”78 Through the acquisition of an external object of choice
one becomes connected with the object in such a way that another’s
action with respect to it can count as an infringement of one’s rights.
An acquired right is thus a relation between a right-holder and an
external object of choice that places others under a duty to the right-
holder with respect to that object of choice.Whereas there is only one
kind of innate right, there are as many kinds of acquired rights as
there are ways of relating a person to an external object of choice.
Within acquired right, then, a further division can be made into the
kinds of relations that can link a person to an external object of choice.
This division expresses the concepts of the understanding that are
exercised in judgments about relation. Kant holds that there are three
such concepts: substance, causality, and community.79 Substance is that
which subsists through the variations of the properties that inhere in
it, as an apple subsists through its alterations in color as it ripens and
then decays. Causality is unidirectional determination of a conse-
quence by its ground. Community is the reciprocal determination of
the different parts of a whole. Consequently, every right that connects
a person to an external object of choice must be a right either to a
substance or to a causality or to a community.80
First, substance is the object of a right in rem. A right in rem is a prop-
erty right, which by connecting the right-holder to a substance puts all
others under a correlative duty to abstain from that substance. Because
it is a right to a substance, the entitlement of the right-holder remains
intact whatever changes are undergone by the object of the right; if I

78
Ibid., at [6:247].
79
Immanuel Kant, Critique of Pure Reason, tr. Norman Kemp Smith (1929), A80/B106.
80
Kant, above n. 76, at [6:247].
224 corrective justice

own the apple, it remains mine even when it turns from green to red
and from red to brown. A property right is good against the whole
world because it presupposes a general will of all under which everyone
recognizes the legitimacy of anyone else’s rightful acquisition.81 Acquisi-
tion restricts the freedom of others with respect to the acquired object;
it therefore cannot be achieved simply by the acquirer’s unilateral action
on that object. Accordingly, a notionally universal consent to the system
of property rights is required for rightful acquisition under the category
of substance.Thus, a property right imposes a correlative duty on every-
one to adhere to the general will recognizing a system of property
rights; the generality of the duty merely reflects the generality of the
will that any particular property right presupposes.
Next, causality is the object of a right in personam. Contract is the
paradigmatic example of a right to a causality.82 What the promisee
acquires through contract is the right to the promisor’s performance of
a particular deed—that is, a right to the causality of the promisor’s will.
The capacity to determine performance of this deed becomes part of
the promisee’s patrimony. In Kant’s words, the promisee acquires “an
active obligation on the freedom and means of the other.”83 Thus, the
contract provides the rightful basis for the unidirectional determination
of the promisor’s will in fulfillment of the promisee’s right. Such a right
does not arise through the initiative of either party alone, as that would
be inconsistent with the freedom of the other. Rather, the contractual
right is jointly established by both parties through contract-forming
steps—the promisor’s making and the promisee’s acceptance of the
promise84—that express their united will. Unlike the general will of all
that is presupposed in proprietary rights, the united will that establishes
a right to the causality of another’s deed creates a duty that runs between
the two parties whose wills are united. Accordingly, the distinction
between an in rem and an in personam right lies not solely in the number
of persons that they respectively bind85 but in the different relational
categories of substance and causality that they respectively instantiate.

81
Ibid., at [6:256, 261].
82
Ibid., at [6:271–74].
83
Ibid., at [6:274].
84
These are Kant’s terms (ibid., at [6:272]) for what common lawyers call offer and
acceptance.
85
Wesley Newcomb Hohfeld, “Fundamental Legal Conceptions as Applied in Judicial
Reasoning,” (1971) 26 Yale L.J. 710, at 718 (distinguishing between “paucital” rights in per-
sonam and “multital” rights in rem, with the latter made up of a great number of the
former).
 unjust enrichment 225

Finally, community: the object of this kind of right is another per-

son’s status, insofar as one has the right to make arrangements about
that person.86 Among Kant’s examples are relationships between
spouses and relationships between parents and children. In these rela-
tionships there is a reciprocal determination of rights and duties. The
parent, for instance, has a right to manage and develop the child, but
the parent is simultaneously under a duty to care for the child until
the child matures. The right and the duty are not independent of each
other, but are the mutually entailed aspects of the same relationship. A
contemporary example of such status-based reciprocal determination
is the fiduciary relationship; the fiduciary has the right to make
arrangements about another, but because of that very right is subject
to the obligation not to profit from the relationship or to allow inter-
est to conflict with duty.
Kant regarded these three categories as exhausting our under-
standing of relation. When these categories are applied to law, a right
in what is external to oneself places others under a correlative duty
by being either an in rem right to a particular substance, an in per-
sonam right to cause the other to perform a deed, or a status right of
a community within which rights and duties are reciprocally deter-
mining.
Within this taxonomy, liability for unjust enrichment is an instance
of the plaintiff ’s in personam right to a causality of the defendant’s will.
The causality in question—the deed whose performance is the con-
tent of the plaintiff ’s right—is the defendant’s retransfer to the plain-
tiff of the value. This right is established through the unity of the
parties’ wills with respect to the non-gratuitousness of the original
transfer. Just as the promisor’s making and the promisee’s acceptance
of the promise establish a right to contractual performance, so the
plaintiff ’s non-donative transfer of value and the defendant’s accept-
ance of the value as non-donatively given establish the plaintiff ’s right
to the value’s retransfer.
This understanding of unjust enrichment views liability under the
category of causality, not substance. The basis of the liability is not
that the plaintiff has retained ownership in the transferred value,
but that the plaintiff has acquired a right to have the defendant retrans-
fer the value.

86
Kant, above n. 76, at [6:259, 276–84].
226 corrective justice

Value plays a double role for external objects of choice. On the one
hand, value inheres in a thing of value considered as a substance, so
that the transfer of the thing of value is also a transfer of the value that
inheres in it. Unjust enrichment is indifferent to this value, which can
be recovered only through the assertion of a property right to the
thing of value. On the other hand, value is the content of a process of
transfer in which something is given for nothing. This transfer gives
rise to a relationship particular to the two parties to it. When neither
of them treats the value as the content of a transfer—that is, when the
transferor does not intend to make value the content of a transfer and
when the transferee accepts the value as not having been the content
of a transfer—then the law undoes the process of transfer by requiring
restitution of the enrichment. The point of the liability is not that
value, as an attribute inhering in the substance of what was given to
the defendant, did not pass. It did pass along with the substance in
which it inheres. Rather, the basis of liability is that the process of giv-
ing something for nothing was intended by neither party and there-
fore has to be reversed. The convergence of the parties’ wills with
respect to the non-gratuitousness of the transfer establishes the trans-
feror’s right to a retransfer, which is a causality of the transferee’s will.
In the light of the Kantian categories of relation, one can under-
stand as follows the opening words of the Restatement of Restitution,
that “[a] person who has been unjustly enriched at the expense of
another is required to make restitution to the other.”87 Enrichment at
the expense of another refers to the transfer of value. The unjustness
refers to the non-donative terms on which the plaintiff has given and
the defendant has accepted this transferred value. This non-donative-
ness on both sides signifies the relationship of will to will that estab-
lishes the plaintiff ’s in personam right (and the defendant’s correlative
duty) to the retransfer of the value. The Restatement calls the duty to
retransfer the value a “require[ment] to make restitution.” Making res-
titution is the performance that constitutes the content of the in per-
sonam right as a causality of another’s will.88

87
Restatement of Restitution, above n. 1.
88
Compare Ross B. Grantham and Charles E. F. Rickett, Enrichment and Restitution in New
Zealand (2000), 470:
In a case of subtractive enrichment, where the relevant right is defined in terms of the principle
of restorable enrichment, the primary right must be the right to restitution. Where a restorable
enrichment occurs, the plaintiff ’s primary right is to restitution from the defendant.
 unjust enrichment 227

9. Corrective justice revisited

I mentioned at the outset that corrective justice has three interwoven
features: the correlative structure of the parties’ relationship, the pres-
ence of a right and a correlative duty, and the conception of the par-
ties as free and equal persons. In this section, I sum up by noting how
the account of unjust enrichment that I have offered conforms to all
these features of corrective justice.
The challenge to corrective justice in the unjust enrichment con-
text revolves around supposed difficulty of formulating the right and
the correlative duty. In response, I have suggested that the norma-
tively defective transfer of value relates the parties as will to will
regarding the gratuitousness of the transfer, so as to establish in the
plaintiff an in personam right (and to impose a correlative duty on
the defendant) to have the value retransferred. Liability signifies that
the defendant’s retention of the value in the face of the two obliga-
tion-creating conditions that render the transfer defective is incon-
sistent with this right. Such retention makes the defendant and the
plaintiff the active and passive poles, respectively, of an injustice
between them.
In this conception of liability the parties are conceived as free and
equal persons. The point of liability is to assure that the transfer and
retransfer of value is in accordance with the parties’ freedom of will.
Hence the law construes as normatively defective a transfer of value in
which the transferor did not intend to give something for nothing
and in which the transferee accepted what was transferred as not hav-
ing been given for nothing. These obligation-creating conditions treat
the parties as equals by insisting that both of them are entitled to what
rightfully belongs to them until they freely part with it.
Moreover, both the notion of a transfer and the circumstances
under which the retention of what was transferred is unjust display
the requisite correlativity. First, the requirement that the defendant’s
enrichment has to be at the plaintiff ’s expense constructs the transac-
tion between the parties as a transfer of value from the plaintiff to the
defendant. The notion of a transfer links the parties to each other by
situating them in correlative positions as transferor and transferee. Li-
ability undoes this transfer by obligating the defendant to restore the
value to the plaintiff.Transfer thus marks out the two particular parties
to the legal relationship by establishing between them the nexus of
value given and received.
228 corrective justice

Second, the notion of unjustness with respect to the transfer is also

correlatively structured.Whether the transferee’s retention of the value
is unjust depends on how the parties’ wills are related to each other
through the gratuitousness of the transfer. The transferor’s giving
of value without donative intent and the transferee’s acceptance of
that value as non-donatively given triggers the obligation to restore
the value, because the transferee cannot retain for nothing a benefit
that was neither given nor accepted as gratuitous. Thus, the enquiry
into whether the enrichment was unjust situates the parties correla-
tively as transferor and transferee of what was not gratuitously trans-
ferred. Consequently, the interaction between the parties as the
non-gratuitous transferor and transferee of value establishes a correla-
tive right and duty of restitution.
The notion of a transfer and the circumstances under which the
retention of what was transferred is unjust are conceptually linked.
The unifying thread lies in the idea that a transfer of value consists in
giving something to another for nothing. Only because what is trans-
ferred is value—that is, something for which the transferor receives
nothing in exchange—does the issue of the injustice of its retention
revolve around whether the value was given and received non-dona-
tively. Thus, the correlativity both of the transfer and of its normative
defectiveness come together in an integrated ensemble of liability-
creating elements.
The correlative situation of the parties is also observed through the
way that liability works to remedy the lack of justice in transfer. Just as
the unjustness of the enrichment lies in normative defectiveness of
the transfer of value, so the correction of this injustice consists in the
retransfer of the value. The retransfer that rectifies the injustice has the
same subject matter, the same correlative structure, and the same cor-
relatively situated parties as did the defective transfer.
The consequence of these considerations is that liability for unjust
enrichment fully conforms to corrective justice. This should astonish
no one. Corrective justice is nothing more than a theoretical account
of the obvious normative link between the parties within a regime of
liability. This link, of course, is present in unjust enrichment no less
than in other areas of liability.
Indeed, corrective justice has always been implicit in the revival of
interest in unjust enrichment throughout the common law world.
Decades ago the formulators of the first Restatement of Restitution
justified their project on the ground that unjust enrichment unifies a
 unjust enrichment 229

variety of doctrines under a legal concept that works justice between

the parties.89 Corrective justice is merely the theoretical construct that
reflects this justification for the development of the law of unjust
enrichment. Corrective justice does this by capturing the fairness and
coherence of justice between the parties and by providing the unify-
ing structure under which various doctrines can be understood as
instantiations of a single principle. To the extent that the development
of the law of unjust enrichment has remained true to the Restate-
ment’s aspirations, an elucidation of the corrective justice of this basis
of liability does no more than confirm in theory what everybody
already assumes in practice.
89
Warren Seavey and Austin Scott, “Restitution,” (1938) 54 L.Q.R. 29.
 7
Incontrovertible Benefit
in Jewish Law

1. The case of the planted trees

At some point in the early third century of the Common Era, a man
in Babylonia went into another’s field and, without the owner’s per-
mission, planted trees there. The question then arose: under Jewish law
was the owner liable for this unsolicited improvement to his property?
The case was brought before Rav, the pre-eminent Jewish jurist of the
time. The Talmud gives the following account of the proceedings:1

A man came before Rav. Rav said to the owner of the field, ‘Go and make
an assessment for him.’ The owner said, ‘I do not want the trees.’ Rav said,
‘Go and make an assessment for him, and he shall have the lower hand’
[that is, on the standard interpretation, the improver shall be entitled to the
lesser of his expenses and what the owner would pay to have the trees
planted]. The owner said, ‘I do not want the trees.’ Subsequently, Rav saw
that the owner had built a fence around the field and was guarding it. Rav
said to the owner, ‘You have revealed your view that you are pleased with
the trees. Go and make an assessment for him, and he shall have the upper
hand.’

This ancient incident brings together features familiar to modern stu-

dents of the law of restitution. On one side is the improver, who claims
remuneration for the benefit, albeit unrequested, of the planted trees.

1
Babylonian Talmud, Baba Mezia, 101a (throughout this chapter the translations are my
own). On Rav, see Ephraim E. Urbach, The Halakha: Its Sources and Development (1986),
295–302.
 incontrovertible benefit in jewish law 231

On the other side is the owner, who (anticipating the modern notion
of subjective devaluation)2 repeatedly denies that this is a benefit that
he wants. Rav, adjudicating the dispute, indicates the measure of the
owner’s payment with various formulations, but makes a decisive rul-
ing only when the owner’s behavior shows that he was satisfied with
the trees after all. Pervading the whole account is familiar tension
between the owner’s freedom to determine the use of his own prop-
erty and the prevention of enrichment at the improver’s expense.
Rav’s treatment of the planting of these trees represents one of the
fundamental building blocks of the Jewish law’s jurisprudence about
unrequested improvements.This jurisprudence is extraordinarily com-
plex and sophisticated. The dictum about the common law—that it
has been “fined and refined by an infinite number of Grave and
Learned Men”3—is even more apposite to the development of Jewish
law. But because its long history has been accompanied by wide geo-
graphical dispersion and a largely decentralized structure of legal
authority, legal doctrine has often been fluid and evolving within the
stable framework provided by the Talmud, the system’s basic text.
Accordingly, although the jurisprudence of unrequested improve-
ments originates in the brief Talmudic segment that centers on Rav’s
case, centuries of commentaries, responsa, and codifications have pro-
duced varied understandings of the legal elements of the problem and
different suggestions of how those elements are to be combined. To
examine, or even refer to, all the possibilities is beyond the scope of
this chapter. I hope, instead, to highlight issues that run parallel to
those found in the common law.
This chapter thus contributes to the burgeoning comparative liter-
ature that the revival of restitution in the common law world has
stimulated.4 From the standpoint of the common law, this literature
has an obvious attraction. Although the modern common law of

2
Peter Birks, An Introduction to the Law of Restitution (1985), 109–16.
3
Thomas Hobbes, A Dialogue between a Philosopher and a Student of the Common Laws of
England, ed. Joseph Cropsey (1971), 55.
4
Especial attention has been paid to German law. See, for example, Thomas Krebs, Resti-
tution at the Crossroads: A Comparative Study (2001); Gerhard Dannemann, “Unjust Enrich-
ment by Transfer: Some Comparative Remarks,” (2001) 79 Texas L. Rev., 1837; B. S.
Markesinis, W. Lorenz, and G. Dannemann, The German Law of Obligations, vol. I, The Law of
Contracts and Restitution: A Comparative Introduction (1997), 710–816; Reinhard Zimmermann,
“Unjustified Enrichment: The Modern Civilian Approach,” (1995) 15 Oxford J. Legal Stud.
403; Reinhard Zimmermann, The Law of Obligations: Roman Foundations of the Civilian Trad-
ition (1990), 834–901.
232 corrective justice

restitution has antecedents that stretch back several centuries, only

within the last decades have scholars and courts made a sustained
effort to develop a set of distinct principles of unjust enrichment.
Attention to the sophisticated older European traditions about unjust
enrichment not only exposes further possibilities of analysis, but also
contributes to the intellectual self-consciousness necessary for pro-
ductive reflection about unjust enrichment as a juridical concept.
However, scholars of restitution have had little opportunity to con-
sider Jewish law, as is understandable given its obscurity and inacces-
sibility. Nonetheless, as the episode involving Rav indicates, issues of
unjust enrichment have engaged the attention of the leading figures
of the Jewish legal tradition for almost two millennia. This makes Jew-
ish law the locus for the world’s oldest uninterrupted and continuing
discussion of unjust enrichment.
Although the Jewish law of unrequested improvements developed
without apparent attention to Aristotle’s conception of corrective just-
ice, my exposition of Rav’s case is nonetheless informed by that con-
ception.5 Corrective justice, with its insistence on the correlative
structure of the private law relationship, is a theoretical idea for under-
standing the fairness and coherence of private law. It is not itself a
notion explicitly invoked in the formulation of legal doctrine. How-
ever, inasmuch as a sophisticated regime of private law—as Jewish law
assuredly is—aspires to deal fairly with both parties and to work out
coherent doctrine, one should not be surprised if its norms are intel-
ligible in the light of corrective justice.
Corrective justice also serves as a framework for the comparison of
legal doctrine across legal systems. Scholars of comparative law are
divided about whether to emphasize the commonalities or the differ-
ences between legal systems.6 Corrective justice accommodates both
commonality and difference. At the level of commonality it directs
attention to the correlative structure of the parties’ relationship, and
thus to the most pervasive characteristic of legal doctrine insofar as it
is fair and coherent. At the level of difference, it acknowledges the
diversity of the ways in which legal systems construct and actualize

5
For a brief treatment from an economic perspective contending that Jewish law con-
verges with efficiency, see Aaron Levine, Free Enterprise and Jewish Law: Aspects of Jewish Busi-
ness Ethics (1980), 78–83.
6
Daniel Visser, “Unjustified Enrichment in Comparative Perspective,” in The Oxford
Handbook of Comparative Law, ed. Matthias Reimann and Reinhold Zimmermann (2007),
969, 972–73.
 incontrovertible benefit in jewish law 233

correlatively structured relationships.7 As a theoretical idea, corrective

justice orients us within the conceptual space of fair and coherent
legal doctrine. It thereby alerts us to considerations that are inappro-
priate to a private law relationship because they are so structured that
they could not possibly come within that space. However, corrective
justice does not usurp the role of specific legal systems in working out
the doctrines that are intelligible in its light. This allows the compara-
tive examination of the different ways in which different legal systems
formulate the doctrinal requirements of a common conceptual frame-
work.
My especial focus is on the interplay between the improver’s claim
to remuneration for the benefit bestowed and the owner’s freedom
from having to pay for an unwanted improvement. From the stand-
point of the idea of unjust enrichment, this interplay raises the issue of
whether the owner’s enrichment at the improver’s expense is unjust.
Within the framework of corrective justice, which conceives of unjust
enrichment as involving a transfer of value that the transferee is obli-
gated to restore, the lack of justice in the enrichment refers not to
injustice at large but to injustice regarding the transfer of value.8 Inas-
much as value is transferred only if the transferor has given something
for nothing, the absence of justice in transfer revolves around how the
parties’ wills are related to each other through the gratuitousness of
the transfer. Accordingly, in determining whether an enrichment was
unjust, two issues regarding the justice of the transfer arise. The first is
whether the transferor intended to give the value gratuitously. If the
transferor had donative intent, justice in transfer is achieved.The trans-
ferred value cannot be recalled by claiming that the enrichment was
unjust, because there is no injustice in the transferee’s retention of
what the transferor willingly gave him. If, however, the transferor did
not intend to give the value gratuitously, a second issue arises: did the
transferee accept the value as non-gratuitously given? An affirmative
answer triggers the obligation to restore the value, because the trans-
feree cannot retain for nothing a benefit that he accepted as non-gra-
tuitously given. Thus, the enquiry into whether the enrichment was
unjust situates the parties correlatively as transferor and transferee of
value that was not gratuitously transferred.

7
Ernest J. Weinrib, The Idea of Private Law (1995), 222–29.
8
Above, chapter 6.
234 corrective justice

The transferor’s lack of donative intent and the transferee’s accept-

ance are structural components of liability for unjust enrichment
within the corrective justice framework. As a functioning system,
however, the law of unjust enrichment must render this framework
specific and concrete. From what is the transferor’s donative intent to
be inferred? Under what circumstances can acceptance of the benefit
be imputed to the transferee? Different systems of law address the
issues raised by these questions in different ways. The consequence is
that Jewish law and the common law may differ in the way each han-
dles unrequested improvements and yet, despite the absence of his-
torical connection between the two systems, may implicitly share a
framework of thought that conforms to corrective justice. Such com-
monality would merely reflect the aspiration of each to fairness and
coherence in their respective constructions of their regimes of private
law.
In its treatment of unrequested improvements Jewish law pays
considerable attention to the aspect of acceptance. Indeed, as I hope
to show, the history of Jewish law in this context is the history of
the shifting conceptions of acceptance. Speaking broadly and ignor-
ing the details for the moment, one can say that the treatment of the
issue in Jewish law will seem familiar to someone conversant with
the common law. The common law employs two notions of accept-
ance. Acceptance of a benefit can be imputed to the recipient either
because the recipient has acted in a way that reveals an acceptance
(for example, by acquiescence) or because, given the recipient’s
activities and projects, he has no reason not to accept it (cases of
“incontrovertible benefit”). Parallels to both notions of acceptance
appear in Jewish law. In the incident of the planted trees, for instance,
the fencing and guarding were taken by Rav to be acts that revealed
an acceptance even though the owner denied that he wanted the
trees. More importantly, Jewish law crafted different conceptions of
incontrovertible benefit, based on a working and reworking of the
Talmudic material. These various conceptions are the subject of this
chapter.
This chapter, accordingly, proceeds in the following stages. Sec-
tion 2 explains the different measures of remuneration (“having the
upper hand” and “having the lower hand”) to which, in the opin-
ion of subsequent commentators, the Talmudic account of Rav’s
case refers. These different measures are tied to the suitability of the
property to the activity of the improver. As section 3 then outlines,
 incontrovertible benefit in jewish law 235

the notion of suitability was the basis of the earliest conception in

Jewish law of what we would term “incontrovertible benefit.” The
basic idea was that the owner of a field that was suitable for plant-
ing trees could be compelled to pay for them on the higher meas-
ure because the owner would not be averse to having the field
brought to its optimal use. This idea depended on interpreting Rav’s
case as involving a field that was not suitable for trees. Section 4
outlines the collapse of this interpretation of Rav’s case in favor of
the view that, regardless of whether the field was or was not suit-
able for planting trees, owners retained their liberty to reject the
improvement. Nonetheless, simultaneous with this collapse, a dif-
ferent basis for incontrovertible benefit arose, from which modern
commentators derived two different conceptions of the conditions
under which the owner could not reject an improvement. Section 5
discusses these different conceptions. Finally, section 6 offers some
brief concluding reflections.

2. The measures of remuneration

Rav’s case appears in a section of the Talmud, extending less than
twenty lines, that deals with unrequested improvements. The section
discusses two situations in which the improver acts for the owner’s
benefit but without the owner’s permission. In the first, the improver
plants trees in another’s field, and the Talmud discusses the quantum
that the owner is to pay for this improvement. In the second, the
improver rebuilds another’s dilapidated structure, and the Talmud dis-
cusses the improver’s right to remove his materials. Rav’s case is the
final element in the discussion of the first of these situations.
The Talmud’s conclusion in the first situation is that the amount to
be paid by the owner depends on whether the field was “suitable for
the planting of trees.”9 Where trees are a more profitable use of the
field than the crops that otherwise would be there, the improver is
entitled to a higher level of remuneration. Rav formulated the differ-
ent levels of remuneration in terms of whether the improver had “the
upper hand” or “the lower hand.”10 What precisely he meant by this
was a matter of dispute among subsequent commentators.

9
Babylonian Talmud, Baba Mezia, 101a.
10
Ibid.
236 corrective justice

By the Middle Ages the most accepted view was as follows.11 The
practice was that persons who were employed by others to plant on
their behalf were paid a proportion, determined by local custom, of
the appreciation in the yield produced by their efforts. The unrequest-
ed improver, of course, had not been employed by the owner. None-
theless, if the field was suitable for trees, the improver got either his
expenses or the customary share of the yield, whichever was greater.
By being entitled to the more advantageous of these alternative meas-
ures of remuneration, the improver “had the upper hand.”
The reason for this treatment of a field that was suitable for trees is
that the planting of trees brought the field to its optimal use. Accord-
ingly, the improver did what the owner would have done in any case,
and therefore the owner can be treated as if he wanted the trees plant-
ed. To arrive at the improver’s remuneration, “one assesses how much
a man would give to have this field planted.”12 Such an owner would
have been willing to allot to the improver a share of the yield in
accordance with the usual practice of the locality. Moreover, if the
planter’s expenditures exceeded his prospective share of the yield, the
owner would have at least reimbursed those expenditures; otherwise
the trees would not have been planted, because the improver would
not have agreed to do it at a loss.13 Therefore, once one treats the
owner as desiring the improvement, the improver becomes entitled to

11
Rashi (Rabbi Solomon Yitzhaki, France, 11th century) on “gilita adaatech deniha lach,”
Babylonian Talmud, Baba Mezia 101a; Ramban (Rabbi Moses ben Nahman, Spain, 13th
century), Milhamot HaShem on Baba Mezia 101a; Rosh (Rabbi Asher ben Yehiel, Spain, 14th
century) on Baba Mezia, chapter 8, 22. This view was described by Rashba, who disagreed
with it (see below n. 17), as held by most of the commentators; Rashba (Rabbi Solomon ben
Abraham Adret, Spain, 13th century), Hiddushei HaRashba on Baba Mezia 101a. Rabbi Joshua
ben Alexander HaKohen Falk (Poland, 16th–17th century), Sefer Meirat Einayim, on Shulhan
Aruch 375, n. 7 summed up the view as follows: “Know that according to the opinion of
Rashi and the Rosh in several places that ‘he has the upper hand’ means that if the appreci-
ation exceeds the expenditure he takes part of the appreciation like the other planters in the
city, and if the expenditure exceeds the appreciation, he takes all the expenditure even though
the owner got no benefit from it.” There were many controversies concerning the details of
this and similar approaches. What distinguishes these approaches from the minority view
mentioned below at 239 is that they involve a comparison of expenditure and appreciation.
Encyclopedia Talmudit, v. 23 s.v. “Yored lenichsei haveiro shelo midaato,” chapter 2, gives a cata-
logue of the various interpretations.
12
Samuel’s formulation of the improver’s remuneration in Babylonian Talmud, Baba
Mezia 101a.
13
Falk, Sefer Meirat Einayim, above n. 11, observes about Rav’s award of the upper hand to
the improver: “If he had not planted the field, the owner himself would have planted it and
expended this amount on it.”
 incontrovertible benefit in jewish law 237

the expenses or the planter’s customary share of the appreciation,

whichever is the greater.
If the field is not suitable for planting trees, the situation is different.
Although the owner has benefited, the trees do not represent the opti-
mal use of the field, so that the reason for treating the owner as want-
ing the improvement falls away. All that remains is the benefit itself,
which is valued as the lesser of the cost of creating it and the appreci-
ation that accrues from it. On the one hand, the value transferred from
improver to owner is the value of the efforts expended in improving
the property. On the other hand, the improver’s expenditure does not
enrich the owner beyond the appreciation in the yield; indeed, if the
improver could charge the owner for expenses that exceed the value
that his efforts added to the yield, he would be impoverishing rather
than enriching the owner. Accordingly, the increase in the yield’s value
functions as a ceiling in the calculation of the quantum of the benefit
received by the owner from the improver. Hence, the classic explan-
ation of what it means for the improver to have the lower hand is that
“if the appreciation is greater than the expenditure, he gets the
expenditure, and if the expenditure is greater than the appreciation he
gets no more than the appreciation.”14
For improvements to non-agricultural properties such as buildings,
where the notion of a yield was not relevant, the notional comparison
of expenditure and appreciation worked in a slightly different way.
Having the lower hand gave the improver the lesser of the expend-
iture and the increased value of the property.15 However, the improver
who had the upper hand was entitled to what the owner would have
paid to have the work done, even if this exceeded the increase in the
value of the property. The difference between the lower and the upper
hand is that in the former the appreciation set the upper limit of the
improver’s remuneration, whereas in the latter improvers were entitled
to the cost of the improvement without limit.16
Thus, the accepted view of the contrast between the improver’s
having the lower hand and having the upper hand involves a differ-
ence in the principle on which the remuneration is assessed. When

14
Rashi on “yado al hatahtona,” Babylonian Talmud, Baba Mezia 101a.
15
The same rule is mentioned by the Roman jurist Celsus in his treatment of inadvertent
improvements; Justinian, Digest, 6.1.38 (Celsus).
16
Alfasi (Rabbi Isaac Alfasi, Morocco, 11th century) Sefer HaHalachot on Baba Mezia 101a;
Rabbi Yosef Haviva (Spain, 15th century), Nimukei Yosef, on Alfasi, Sefer HaHalachot on Baba
Mezia 101a.
238 corrective justice

the improver has the upper hand, the assessment is quasi-contractual.

Because the improvement moves the property to its optimal use and is
thereby equated to one that the owner desires, the assessment is based
on a reconstruction of what the owner would have agreed to pay an
improver to achieve the desired improvement. In contrast, when the
improver has the lower hand, the assessment is restitutionary. Because
the planting of trees benefits the owner without moving the property
to its optimal use, the confidence in the owner’s desire for the improve-
ment is absent. What matters then is not what the owner would have
agreed to pay, but rather the value of the benefit that was transferred
to the owner through the improver’s efforts.
Some commentators in the Middle Ages found the accepted inter-
pretation of the upper hand implausible regardless of whether the
expenditure or the appreciation was greater.17 If the share of the
appreciation was greater than the improver’s expenses, the accepted
view, by giving him the customary share of the appreciation, treated
him like a person who had been hired to plant the trees. But this, so
the objection went, treated a non-consensual transaction as if it were a
consensual one. Moreover, giving the improver more than he expend-
ed meant that what the improver received exceeded the benefit that
was attributable to him.18 On the other hand, awarding the improver
his expenses, no matter how large they were, even if those expenses
exceeded the appreciation in the yield’s value, would also entail hav-
ing the owner pay for more than he benefited. The most that could be
awarded to the improver is the expenses up to the value of the appre-
ciation, since anything above that is a loss that the improver inflicted
on himself.19 The accepted view of the upper hand, in other words,
remunerated the improver on a contractual measure despite the non-
existence of a contract, while failing properly to measure the benefit
that was the basis of the improver’s claim.

17
Rezah (Rabbi Zerahia Halevi Gerondi, France, 12th century), Hamaor Hagadol, on
Alfasi, Sefer HaHalachot on Baba Mezia 101a; Rashba, Hiddushei HaRashba on Baba Mezia
101a; see also Rabbi Yosef Karo (Israel, 16th century), Beit Yosef, Hoshen Mishpat 375, 3.
18
As Rashba puts it, above n. 11, “[on the standard view] the owner gives him what the
planters of the city get, meaning, even more than the expenditure; this view is surprising, for
on what basis will the owner give the improver more than the latter has benefited him?”
Rashba is presumably assuming that the enrichment that the improver can claim consists
only in the amount that quantifies his efforts, not in a share of the yield’s appreciation, since
the yield belongs to the owner unless he freely parts with it.
19
Rezah, above n. 17.
 incontrovertible benefit in jewish law 239

Instead these commentators suggested a different view of the con-

trast between having the upper and the lower hand. What mattered
for them was not the comparison of expenditure and appreciation,
but different ways of measuring the expenditure:

The meaning of ‘he has the upper hand’ is as in the superior kind of hir-
ing, when a man says to his fellow, ‘Build on this land of mine, or plant this
field of mine, so that I myself won’t have to bother with it,’ for this cer-
tainly is of conspicuous benefit to him. And the meaning of ‘he has the
lower hand’ is as with the inferior kind of hiring, when the inferior work-
ers treat it cheaply.20

The owner for whose field the planting of trees is the optimal use can
be presumed to want the work done and to be willing to hire a more
able contractor and pay him at a high rate to have the planting prop-
erly executed. The owner benefits by being spared the trouble of
attending to this desired project himself; in particular he does not have
to bother with bringing in various workmen to attend to the various
stages of the work.21 In contrast, the owner for whose field the plant-
ing of trees is not optimal would be satisfied to have it done at the
minimal cost using the cheapest labor. The benefit consists simply in
having someone put trees where there were none before.Thus, accord-
ing to this view the benefit received by the owner varies with the
kind of field he has. With respect to both kinds of field, the analysis is
oriented to the enrichment that accrued to the owner, and the amount
of the remuneration is conceptualized in what we would consider to
be restitutionary terms.
We can now return to Rav’s case and set it into its Talmudic con-
text. The Talmud introduces the case to show Rav’s view of the remu-
neration to be paid to the improver. Immediately before the Talmud’s
account of the incident, the Talmud mentions an apparent dispute
between Rav and his contemporary Samuel with respect to the unso-
licited planting of trees. Rav had said that the planter has the lower
hand, whereas Samuel had said that the planter receives what the

20
Ibid.
21
As Ritva (Rabbi Yom Toi ben Abraham Ishvili, Spain, 14th century) explained in gloss-
ing Rezah’s idea, “One estimates how much a person would be willing to pay to someone
who will undertake to do this as a single project, so that the owners will not have to bother
with it by arranging for workmen to come and go; for a person would gladly pay a lot of
money for this.” Ritva, Hiddushei HaRitva on Baba Mezia 101a.
240 corrective justice

owner would have been willing to pay to have the field planted. The
Talmud then cites an opinion that these sages do not disagree; their
stated views simply apply to different kinds of fields: Rav’s statement
applies to a field not suitable for planting, whereas Samuel’s statement
applies to a field suitable for planting.That Rav does not disagree with
Samuel is inferred from the incident that came before him, where
Rav envisages two measures of remuneration, the lower hand that the
Talmud had previously attributed to him and the upper hand that is
equivalent to the view attributed to Samuel. In dealing with the tree-
planting, Rav makes three interventions. He first orders remuneration
but without specifying its measure, to which the owner replies that he
does not want the improvement. He then orders remuneration with
the improver having the lower hand, which is the measure appropriate
to a field that is not suitable for planting. The owner then repeats his
insistence that he does not want the improvement. Finally, when Rav
notices that the owner, by fencing and guarding the trees, has demon-
strated that he wants them despite his earlier denials, Rav tells him to
remunerate the improver with the improver having the upper hand—
that is, at the higher level appropriate to a field suitable for planting.

3. Incontrovertible benefit
The special significance of a field that is suitable for planting is remi-
niscent of the common law’s notion of an incontrovertible benefit. An
incontrovertible benefit is one that would not have been declined
even if the owner would have had the opportunity to choose.22 For
the determination of whether an improvement is incontrovertible, the
nature of the improved property and the necessary or optimal use of it
are relevant.23 Jewish law takes the suitability of a field for the planting
of trees as paradigmatic of optimal use; what is necessary to produce
this optimal use then becomes the measure of the improver’s remu-
neration, as the party who “has the upper hand.”
To equate the field’s suitability for trees with the incontrovertibility
of the benefit at common law would, however, be premature. The

22
Regional Municipality of Peel v. Her Majesty the Queen in Right of Canada [1993] 98
D.L.R. (4th) 140, 156 (S.C.C.).
23
LAC Minerals v. International Corona Resources Ltd. [1989] 61 D.L.R. (4th) 14, 53
(S.C.C., per LaForest J.):“on the assumption that Corona had acquired the Williams property,
it would of necessity have had to develop the mine.”
 incontrovertible benefit in jewish law 241

incontrovertibility of the benefit goes to the owner’s liability to make

restitution of an unrequested benefit. In contrast, our discussion of the
kinds of field has gone not to the owner’s liability but to the measure-
ment of the improver’s remuneration. To this point the upper hand
and the lower hand function merely as default rules for quantifying
what the improver receives for the improvement. Whether the owner
can be legally compelled to pay is another issue.
In the Jewish legal literature this issue arises in the following way.
The short Talmudic section on unrequested improvements deals with
two problems. The first is the remuneration of the person who plants
trees without the owner’s permission. The second is whether a person
who reconstructs a dilapidated building without the owner’s position
can change his mind and remove his materials. The answer that the
Talmud gives is that the improver can remove building materials from
a structure but not trees from a field.24 There are two reasons for this.
The first, applicable only to the land of Israel, is that the special value
of settling the land would be undermined by removing the trees. The
second is that because trees are nourished by the earth, removing them
would weaken the owner’s soil. But what if it is the owner, not the
improver, who wants the trees or the building materials removed? Can
the owner reject the improvement by telling the improver to take his
materials and go? The Talmudic text does not explicitly deal with this
issue.
In the absence of explicit treatment, post-Talmudic commentators
looked to the implications of the Rav story. Two features of this story
attracted their attention. First, Rav told the owner to “go and make an
assessment for him and he shall have the lower hand.”25 This mention
of the lower hand indicates that the field in question was not suitable
for planting trees. Second, when the owner then repeated his state-
ment that he did not want the improvement, Rav did nothing in the
face of this apparent defiance until the owner revealed his true senti-
ments by fencing and guarding the trees. From Rav’s failure to compel
the owner to obey him, commentators concluded that, so far as that
particular field was concerned, the owner was within his rights to
refuse the improvement, at least until his conduct contradicted his
professed rejection.

24
Babylonian Talmud, Baba Mezia 101b.
25
Above n. 1.
242 corrective justice

Having thus determined that the field was not suitable for planting
trees and that no obligation to pay arose from an improvement to
such a field, the commentators reconstructed the various stages in the
Rav incident as follows.26 When the case was brought to Rav initially,
he did not know whether the field was suitable for planting or not.
He accordingly required an assessment without indicating whether
the improver was to have the upper or the lower hand. When the
owner declared that he did not want the trees, Rav inferred that the
owner was unwilling to pay on the higher measure for having trees
planted in a field that was not suitable for that use. Rav therefore told
him that he should pay only on the lower measure, as was appropriate
for a field not suitable for trees. The owner then repeated his assertion
that he did not want the trees. Rav interpreted this as implicitly
requiring the improver to remove the trees. Thus, Rav’s inaction in
the face of the owner’s statement led to the conclusion that the owner
of a field that was unsuitable for trees could require the improver to
remove the improvement. Subsequently, by fencing in and guarding
the trees, the owner revealed that he did want them after all. This
expression of the owner’s desire allowed the field to be treated as if it
was one that was suitable for planting. Rav accordingly ordered him
to pay for the trees on the higher measure.
Accordingly, on this interpretation, the owner could refuse to accept
the trees if the field was not suitable for that use. Although the trees
might well be a benefit from an objective point of view, the freedom
of the owner of such a field to assert that he preferred not to have
them was untrammeled.
The converse, however, also obtained. This interpretation empha-
sized that the field in Rav’s case was not suitable for planting. Had the
field been suitable for that purpose, the owner would not have been
able to refuse the trees and would have had to pay for them on the
higher measure. Just as Rav finally compelled the owner who fenced
and guarded the trees to pay, with the improver having the upper
hand, so the owner of a suitable field could be forced to accept and
pay on the higher measure for trees planted without his authoriza-
tion.
It is worth emphasizing the radical nature of this conclusion from
the standpoint of the common law. Common lawyers are familiar

26
Ramban, Hiddushei HaRamban on Baba Mezia 101a; Rashba, above n. 11; Nimukei Yosef,
above n. 16; Ritva, above n. 21.
 incontrovertible benefit in jewish law 243

with—and often troubled by—cases where the owner is made to pay

for an improvement by someone who mistakenly thought he owned
the object improved.27 This interpretation of Rav’s case, however,
imposes on the owner an obligation to remunerate the improver who
knowingly plants trees in another’s field, provided that the field is suit-
able for planting. The improver in Rav’s case did not make a mistake
of title or of any other kind.28
This contrast between the common law and Jewish law reflects differ-
ing premises about the volunteer.The common law views a person who
improves property that he or she knows belongs to another as a volun-
teer who is making a gift. The improver’s expectation of remuneration
merely indicates that in bestowing this gift, the improver is also taking
the risk that the owner will pay for it. The common law sees no reason
to reverse the gift or reallocate the risk through an award of restitution.29
Because the improver knows that only owners have the right to deter-
mine the condition of what they own, the improver is taken to know
that the improvement cannot obligate the owner to pay for it. Hence the
improvement counts as nothing more than a donative act. However,
when the benefit is conferred by mistake, the argument that the improv-
er was acting with donative intent evaporates. Then the issue becomes
whether it is truly a violation of the recipient’s freedom to compel pay-
ment and thus to treat the recipient as accepting the benefit.When con-
sidered as an incontrovertible benefit, for example, the improvement is
one where—given the nature of the property, its necessary or optimal
uses, or the owner’s plans for it—the owner can be viewed as having
every reason consistent with his or her own autonomy to accept it.30

27
Greenwood v. Bennett [1973] 1 Q.B. 195 (C.A.); Gidney v. Shank [1995] 5 W.W.R. 385
(Man. Q.B.), reversed [1996] 2 W.W.R. 383 (Man. C.A.); Matthews, “Freedom, Unrequested
Improvements, and Lord Denning,” (1981) Cambridge L.J. 340.
28
Moreover, the liability obtains even where Jewish law sees no difficulty in detaching the
improvement from the improved property. The Talmud itself indicates that building materials
can be detached from a reconstructed building, at least at the instance of the builder. Nonethe-
less, on the argument to this point, if the ruined building had been suitable for reconstruction,
the owner would be legally compelled to remunerate the stranger for restoring the building.
Compare the suggestion of Richard Sutton that the owner should owe restitution if he or she
could have allowed the improver to remove the improvement; Sutton, “What Should be Done
for Mistaken Improvers?,” in Essays on Restitution, ed. Paul Finn (1996), 252–54.
29
Birks, above n. 2, 101–3; Graham Virgo, The Principles of the Law of Restitution, 2nd ed.
(2006), 39–40.
30
Incontrovertible benefit is usually understood as negating the possibility of subjective
devaluation by the defendant and thus going to the existence of the enrichment; Birks, above
n. 2, 116. My account here varies from this understanding for reasons set out in chapter 6.
244 corrective justice

In contrast, Jewish law does not assume that the improver who
knowingly acts on another’s property does so with donative intent.
“We do not say that simply because he went down into another’s
field, his intention was just to give a gift.”31 The general principle rele-
vant to all unrequested benefits is that “if any person does an action
or benefit for another, one cannot say to him that ‘you did it for me
gratuitously because I did not ask you to do it,’ but one must give him
his remuneration.”32 The person who planted the trees did so in order
to get paid by the owner.33 Jewish law treats this as a purpose worth
giving effect to. Rather than disqualifying him as a volunteer or an
officious intermeddler, the improver’s knowing operation on another’s
property is the basis of his claim, which is inescapable (on the present
interpretation of the Rav story) when the property is suitable for the
improvement.
In this context both the common law and Jewish law are individu-
alistic, but they exhibit different conceptions of individualism.34 For
the common law, the individualism resides in the institution of prop-
erty and in the juridical construction of the social understandings to
which property gives rise. Because the improver can be taken to know
that only owners have the right to change the condition of their prop-
erty, the improver is understood as giving a gift and, accordingly, is not
eligible for restitution. For Jewish law, the individualism resides in giv-
ing effect to the improver’s non-altruistic motivation. In the absence
of evidence of a donative intent, improvers are treated as forwarding
their own self-interested purposes.
From the standpoint of the common law, the approach of Jewish
law has a paradoxical implication. Whereas for the common law the
improver’s knowledge of another’s ownership weakens the improver’s
case, for Jewish law this knowledge strengthens it.The improver’s claim
to be remunerated on the higher measure presupposes that the
improver has acted to improve another’s land, for only if the improv-
er’s intention was to have another pay for his work can he be paid
what (in Samuel’s words) “a man would give to have this field

31
Rashba, Hiddushei HaRashba on Nedarim 33b.
32
Rema (Rabbi Moses Isserles, Poland, 16th century), Shulhan Aruch on Hoshen Mishpat
264, 4.
33
Ramban, Milhamot HaShem on Baba Batra 4b.
34
For a provocative comparison of Jewish and American law in another restitutionary
context, see Hanoch Dagan, Unjust Enrichment: A Study of Private Law and Public Values (1997),
50–57, 109–29.
 incontrovertible benefit in jewish law 245

planted.”35 Accordingly, if he mistakenly thought that he was working

his own land, he can be remunerated only on the lower measure, even
if the field is one that was suitable for planting trees.36
The absence in Jewish law of the common law notion of a volun-
teer exposed owners to the possibility that, if their property was suit-
able for the improvement in question, they would be compelled to
pay the improver for effecting it. How could this apparent indiffer-
ence to the autonomy of owners to determine the use of their own
property be justified? This question did not present itself explicitly; for
those who interpreted the Rav story to lead to this result, the Tal-
mudic origin of the rule was justification enough. Nonetheless, three
general observations are in order.
First, the premise was that, if the field was suitable for planting, the
improver was doing something that the owner wanted done. “The
person who knowingly improves another’s field thinks that the own-
ers are pleased with it.”37 Conversely, the owner has no liability for the
improvement if he has warned the improver not to do the work. A
ruling to this effect in the thirteenth century included the comment
that this was “in order to prevent everybody from forcing someone
else to plant and to build.”38 Owners could protect their autonomy
over their property by telling improvers that they did not want the
contemplated improvement.
Second, some commentators based the higher measure of remu-
neration on a notion of acquiescence. “Because the field was suitable
for planting and the owners knew and kept quiet, it is certain that
they were pleased with his work, and he is like a person who made
the improvement with their knowledge.”39 When this explanation of

35
Above, n. 12.
36
Ramban, Hiddushei HaRamban on Baba Mezia 40a; Rivash (Rabbi Isaac ben Sheshet, Spain,
14th century), Responsa of the Rivash, 515.This resembles the rule in Roman law, that if I improved
another’s property thinking it was my own, I have no actio negotiorum gestorum for reimbursement,
“because I did not intend to obligate anyone to me”; Digest, 10.3.14.1 (Paulus).
37
Rivash, Responsa of the Rivash, 515.
38
Rashba, Responsa of the Rashba, 4, 54.
39
Ritva, above n. 21; Nimukei Yosef, above n. 16 (cf. also Nimukei Yosef on Alfasi, Sefer
HaHalachot on Baba Batra 4b); Rosh on Baba Batra 7. For a different view of the significance
of the owner’s silence, see Shalom Albeck, Dinei Hamamonot Batalmud (1976), 193. Albeck
suggests that the silence operates as a retrospective consent once the owner discovers the
improvement. This implies the effectiveness of a protest made against an improvement dis-
covered after being completed, which seems inconsistent with the view of Ritva and Nimukei
Yosef that the owner can be compelled to accept the improvement if the land is suitable for
planting trees.
246 corrective justice

the improver having the upper hand is combined with the owner’s
inability to refuse the improvement, the picture that emerges is not
unlike the recent suggestion of a common law doctrine of free accept-
ance.40 On the one hand, the improver acts (and is assumed by Jewish
law to act) not gratuitously but with the expectation of payment. On
the other hand, the owner, knowing this and allowing the work to
proceed without protest, cannot subsequently treat the improvement
as undesired, especially since the improvement is consonant with the
optimal use of the property. This notion of acquiescence can be
regarded as an extension of the owner’s power to warn away the
potential improver: because at any moment the owner’s protest puts
an end to the potential improver’s claim, omission to make the protest
can be construed as free acceptance of the improvement.41
Third, because liability is not confined to situations where “the
owners knew and kept quiet,”42 a more comprehensive basis for liabil-
ity must be postulated. The common law notion of incontrovertible
benefit is a formulation, applicable to Jewish law, of this more com-
prehensive basis. Given that liability depends on the status of the land
as suitable for planting trees, the improvement must be regarded as a
benefit that the improver has no reason not to accept. By bringing the
land to its optimal use, the improver has done what the owner wants
done and, accordingly, what the owner cannot repudiate when done
by another. It is not merely that the improver has made the owner
better off in some general way; if that were all that were necessary,
owners on this reading of the Rav story would be liable—as they are
not—for improvements even to land that was not suitable for planting
trees. Rather, the land’s suitability for trees frames the legal construc-
tion of what owners can be assumed to want. By differentiating their
wealth into specific pieces of property that have particular characteris-
tics, owners can be regarded as signaling the terms on which they are
prepared to accept improvements. Thus, although one cannot con-
clude that they are willing to remunerate an improver for producing

40
Birks, above n. 2, 277–79; Peter Birks, “In Defence of Free Acceptance,” in Essays in the
Law of Restitution, ed. Andrew Burrows (1991), 105.
41
Writing several centuries later than the authorities mentioned in n. 39 above and with-
out referring to those texts, Maharit (Rabbi Yosef of Trani, Turkey, 17th century), Responsa of
the Maharit, I, 106, denied that an owner could incur liability for an unsolicited improvement
by knowing about it and keeping quiet. He contended that, unlike fencing and guarding,
silence does not constitute a manifestation of the owner’s view that he is pleased with the
improvement.
42
Above, n. 39.
 incontrovertible benefit in jewish law 247

general increases in their wealth (such increases might take forms

incompatible with owners’ specific projects), having property of a cer-
tain sort can be taken to indicate their willingness to have this prop-
erty developed in accordance with its optimal use. An improvement
consonant with the specific nature of a piece of property is a benefit
that the law assumes the owner has no reason to reject. On this view,
the nature of the property itself indicates what the owner of the prop-
erty wants and is willing to pay for, thereby supposedly reconciling
the owner’s freedom of choice with the improver’s entitlement to
remuneration.
An example of such reasoning appears in a responsum of the early
seventeenth century.43 The author is commenting on the view of an
earlier authority that, whereas the owner of a field suitable for plant-
ing trees could be compelled to accept the improvement, the owner
of a dilapidated structure that was suitable for building could not.44
He finds the argument in the earlier authority convoluted,45 but he
suggests that the distinction between an improved field and an
improved structure might be supported in a more straightforward way:

Even the commentators who differed about whether we listen to the

owner of a field suitable for planting when he says ‘Take your trees and go’
can acknowledge that we do listen to the owner of a house even though it
is suitable for rebuilding. And the reason for this is that in the case of a field
suitable for planting trees it is well known that a field of trees is worth
more than a field of grain, and an unplanted field is available for being

43
Maharit, above n. 41.
44
Nimukei Yosef, above n. 16.
45
The convoluted argument in the Nimukei Yosef is based on three elements. The first is
that the implication of Rav’s case is that the owner of a field suitable for planting trees cannot
refuse the improvement but the owner of a non-suitable field can.The second is the Talmud’s
explicit statement at Baba Mezia 101b that the improver can remove his materials in the case
of a building but not in the case of a field.The third is that the early post-Talmudic Babylon-
ian academies were of the view that there was reciprocity between the owner and the
improver, in that the owner can have the materials removed in any situation in which the
improver can have them removed. From this the Nimukei Yosef argued that it followed from
the first of these considerations that issue of removal mentioned by the Talmud applied only
to properties that were suitable for the improvements (since we know from Rav’s case that
owners could have improvements removed from properties that were not suitable); that it
further followed from the second and third of these considerations that the owner can have
the improvement removed only in the case of suitable buildings but not of suitable fields; and
that it further followed from the first of these considerations, that such removal could apply
even to buildings that were suitable for rebuilding, but not to fields that were suitable for
planting trees.
248 corrective justice

planted with trees, and because the improvers have made the field more
valuable, one needs to give restitution for this surplus. But in the case of
courtyards and fields, even if they are suitable for building, nonetheless not
every person is ready to squander his wealth and to busy himself with
buildings, which impoverish their owners, as the Talmud says, ‘Repair and
you will not have to rebuild.’46

The difference is that improving a field by planting trees in it is an

unequivocal benefit, but improving a dilapidated structure involves a
commitment to continuing efforts and expenditures that the owner
might not want. Fields thus signal the acceptability of improvements
in a way that buildings do not.
In sum, Rav’s case was initially interpreted as allowing the owner of
the field to repudiate the improvement only if the field was not suit-
able for the planting of trees. It was because Rav was dealing with a
field of this sort that the owner could with impunity declare that he
did not want the improvement, even after Rav had told him to pay on
the lower measure. The owner, however, would have been compelled
to remunerate the improver if the field had been suitable for trees. In
this liability of the owner one can discern the Jewish equivalent of the
common law notion of incontrovertible benefit.

4. The demise and rebirth of incontrovertible

benefit
Although this interpretation of Rav’s case attempts to bring the
improver’s claim for remuneration into harmony with the owner’s
freedom of choice, the truth is that the optimal use of the field can
serve only as a rough surrogate for the desires of the owner. It is easy
to imagine cases in which planting trees in a field used for grains
interferes with the owner’s autonomy, even though it increases the
owner’s wealth. The owner may be leading a life that is interwoven
with the field’s unimproved state: he may want to feed himself and his
family with grain grown by his own hands, or he may be knowledgeable
about growing grain but not about managing trees, or he may simply

46
The reference is to Babylonian Talmud,Yevamot 63a: “Rav Papa said: ‘. . . Plug the hole
and you will not have to repair; repair and you will not have to rebuild; for whoever engages
in building becomes poor.’”
 incontrovertible benefit in jewish law 249

be unable to afford the improvement.47 Thus, although non-optimal,

the particular use may satisfactorily match his particular projects,
which would be disrupted by the obligation to accept the improve-
ment imposed by a stranger. If the owner could not reject the
improvement, “it would turn out that any person could compel
another to transform his field into an orchard and his courtyard into
buildings.”48
In the fourteenth century, Rosh, a leading authority in Jewish law,
proposed a different interpretation of Rav’s case that avoided this
result.49 As we have seen, the accepted view until then was that Rav
was dealing with a field that was not suitable for planting, as is indi-
cated by his ordering the owner to pay on the lower measure; the fact
that Rav did not react to the owner’s assertion that he nonetheless did
not want that improvement led to the inference that the owner of a
field not suitable for trees could not be compelled to pay the improver.
Rosh contended that this reading of the incident was mistaken.
Rosh made two interpretive arguments. The first argument noted
that Rav twice told the owner to pay, not specifying the measure of
remuneration the first time and specifying the lower measure the sec-
ond time. If the field was one that was not suitable for planting, the
first order would have implied remuneration on the lower measure,
thus making the second order a pointless repetition of the first. In
Rosh’s view, the field in question was one that was indeed suitable for
planting trees, and the owner had the right, even for such a field, to
refuse the improvement. When Rav told the owner to pay on the
lower measure, he was not implying that the field was not suitable for
planting. He was merely saying that “if the owner acquiesces in the
improvement, let him pay the improver so much.”50 In other words,
Rosh thought that Rav was merely proposing a settlement that might
appeal to both parties: to the owner, because he would be getting the
improvement for which his field was suited at a lower price than if he
employed someone to plant the trees, and to the improver because he
would get something for the trees instead of having to remove them.

47
These examples are taken from Rabbi Abraham Karelitz (1878–1953), Hazon Ish, Hoshen
HaMishpat, Baba Batra 2, 3.
48
Maharit, above n. 41 and Rabbi Yosef Karo, Beit Yosef, above n. 17, 375, 2 (citing the
students of Rashba).
49
Rosh (Rabbi Asher ben Yehiel, Spain, 14th century), Baba Mezia 8, 22.
50
Ibid.
250 corrective justice

Rosh’s second argument noted that after Rav saw that the owner
had fenced in and guarded the trees, Rav compelled the owner to pay
on the higher measure. This disposition creates a puzzle if one thinks
that the field in question was not suitable for planting. Remuneration
for planting a non-suitable field should have been on the lower rather
than the higher measure. To be sure, the owner’s actions can be taken
to show that his earlier denial that he wanted the trees was untruthful;
but, given that the level of remuneration depends on the status of the
land, it is hard to see why the fact that he really does want the trees
should change the amount he has to pay for them. If, however, the
field was suitable for planting trees (as Rosh thought), then the mys-
tery about the award on the higher measure disappears.
With this transformation of the story, any semblance of liability for
an incontrovertible benefit falls away from it. Rav’s inaction after the
owner repeated that he did not want the improvement shows that an
owner could not be compelled to accept an improvement even if that
improvement accorded with the optimal use of the property. All that
the difference between the upper and the lower hand does is mark the
different default measures of remuneration

if the owner is willing to keep the seedlings in his field. But if he is not
willing to keep them, he can tell the improver to take his seedlings and go
even if the field is suitable for planting trees, for he can say that ‘as far as I
am concerned, it is more satisfactory for me to have a field of grain’. . . It
makes no sense at all for the owner of the field to have to pay the improver
as a hired planter when he does not want the seedlings.51

In the story the owner ends up paying on the higher measure, but this
is because, having shown through his actions that he wanted to keep
the trees, he became liable to pay at the appropriate default level. If he
really had not wanted the trees, he could have insisted that the improv-
er remove them even though his field was suitable for planting. Rosh
thus sees the story as reflecting the unbridled freedom of the owner to
determine the use of his own property. In Rosh’s interpretation of the
case, what is paramount is the autonomy of the owner, not the status
of the land.
Although Rosh’s interpretation of Rav’s case became the accepted
one, it turned out not to be the end of the notion of incontrovertible

51
Ibid.
 incontrovertible benefit in jewish law 251

benefit in Jewish law. For simultaneously with shutting off this avenue
involving the planting of fields, Rosh opened or kept open another
involving the building of structures.
Because of the Talmud’s omission to deal with the matter expressly,
disagreement had arisen among Rosh’s predecessors about the right
of the owner to refuse an improvement that consisted in the building
or rebuilding of a structure, even if this structure was the land’s opti-
mal use.52 This right to refuse the improvement, if it existed, would be
expressed by the owner’s telling the improver to remove his materials,
thus restoring both parties to the position they were in before the
improver’s activities. One of the strands in this disagreement was the
view that the owner had no such right, because building materials
lose part of their value by being incorporated into a structure. Accord-
ingly, by ordering the improver to remove his materials, the owner
would be harming the improver rather than merely restoring the sta-
tus quo.53
This disagreement stands in the background of Rosh’s treatment of
the owner’s right to refuse the improvement. So far as fields were con-
cerned, Rosh championed the view that even if the field was suitable
for planting trees, owners did not have to keep the trees if they did
not want them. Structures, however, were different from trees. In the
case of trees, the Talmud had ruled that the improver could not remove
his trees, because once they received nutriments from the earth, their

52
For one facet of this disagreement, see above n. 44.
53
Ravad (Rabbi Abraham ben David, France, 12th century), Comments on Alfasi, Sefer
HaHalachot on Baba Mezia 101a, in Shitah Mekubetzet on Baba Mezia 101a. The Talmud had
dealt expressly only with the instance of the improver who wanted to remove his materials;
it had been silent about whether the owner could order the materials removed. The rule
stated in the Talmud was that the improver could remove his materials in the case of a struc-
ture but not in the case of a field; above, n. 24. The early post-Talmudic Babylonian acad-
emies were of the view that whatever rule about removal applied to the improver also applied
to the owner, with the result that the owner of the structure could order the improver to
remove the materials; but see above n. 45. Ravad contested this view. The controversy is
summed up by Ravad’s follower Meiri (Rabbi Menahem ben Solomon, France, 13th cen-
tury) as follows:
The Talmud does not mention what the rule is if the owner of the land tells the improver to take
his wood and stones. The heads of the academies agreed to treat the improver and the owner
equally, and because the improver can say ‘I am taking my wood and stones,’ so the owner can say
‘Take your trees and stones.’ Nonetheless, the greatest of the commentators disagree with this,
because in the latter case there is a great loss. When the improver says ‘I am taking my wood and
stones,’ we listen to him because he is waiving his loss. But in the case of an owner who says ‘Take
your wood’, it is appropriate to say that we do not listen to him but we allow the improver to
occupy the structure until the owner reimburses his expenses or gives him what we assess for
him.
252 corrective justice

removal would weaken the owner’s soil.54 This ruling does not apply
to the owner’s requesting the removal of the trees, because the owner
can decide to accept the weakening of his own soil. Nor is this weak-
ening something about which the improver can complain; the improv-
er gets his trees back unimpaired. However, the situation is reversed if
the improvement involves building materials. In this case it is the
improver who suffers from the depreciation of his materials. If the
improver wishes to remove the materials and accept this loss, the
owner cannot complain—which perhaps accounts for the Talmud’s
ruling that improvers can remove building materials but not trees. But
the owner cannot impose a loss on the improver by requiring the
materials to be removed.
Accordingly, Rosh ruled that although owners could not be
forced to accept trees, they could be forced to accept structures.
However, instead of merely adopting the conclusion that the owner
could not require the improver to suffer a loss by removing the
building materials, Rosh also outlined the conditions under which
this conclusion applied. Rosh was insistent that the consideration of
the harm to the improver not be at the expense of harm to the
owner. He therefore asserted that the owner is barred from having
the improvement removed only if the improvement did not cause
him a loss. Thus, commenting on the view that the owner causes a
loss by requiring removal of the building materials from a rebuilt
ruin, he remarked:

This consideration makes sense where the owner was not using the ruin
and he had his own place, because it is appropriate to build a structure like
this when it does not impair the owner’s livelihood, because [if the owner
then requests removal of the materials] he is really seeking a pretext to
inflict loss on the improver, and so we do not listen to him.55

On the surface Rosh’s reasoning seems delictual rather than restitu-

tionary. His reason for preventing the owner from having the mater ials
removed is that the owner cannot gratuitously inflict loss on the improver,56

54
Above n. 24.
55
Rosh, Baba Mezia 8, 23.
56
Rosh’s statement that the owner seeks a pretext to inflict loss on the improver is pre-
sumably related to the recognition in Jewish law of a doctrine of abuse of rights. On this
doctrine, see Aaron Kirschenbaum, Equity in Jewish Law: Formalism and Flexibility in Jewish
Civil Law (1991), 185–252.
 incontrovertible benefit in jewish law 253

rather than that the improver is entitled to have the owner give restitution
for a benefit. Nonetheless, in this context the delictual cannot be dis-
entangled from the restitutionary. Rosh’s formulation sets out what he
regards as the conditions in which the owner is not adversely affected
by—and therefore cannot complain about—the improvement. Under
these conditions, the owner’s freedom to determine the use of his
own property no longer obtains, and the owner must pay for the
improvement on the higher measure. Although the reasoning is not
explicitly restitutionary, it has a restitutionary dimension.
This restitutionary dimension reflects the principle implicit in
Rosh’s formulation. Rosh’s view is that owners can be compelled to
keep the improvement when it does not leave them worse off than
they would otherwise have been. An owner who is left worse off by
the improvement could hardly be described as “seeking a pretext to
inflict loss on the improver”; by having the materials removed in those
circumstances all that the owner would be doing is avoiding his or her
own loss. In other words, only if the improvement can be regarded as
a benefit that the owner has no reason to reject does the preclusion
against inflicting loss on the improver get off the ground. In this way,
incontrovertible benefit, which was expelled in Rosh’s account of
trees, returns in his treatment of buildings.

5. Two conceptions of incontrovertible benefit

But what counts as an incontrovertible benefit in Rosh’s line of rea-
soning? Given that Rosh holds that planting trees in a suitable field
does not obligate the owner to pay for the improvement, incontro-
vertible benefit can no longer refer to the optimal use of the property.
What sort of benefit, then, is implicit in Rosh’s thinking? Modern
authorities who have addressed this question have offered a restrictive
and an expansive suggestion.
The restrictive suggestion appears in the codification of Rabbi
Yehiel Michal HaLevi Epstein.57 The owner could not refuse a benefit

where the court sees that it is necessary for the builder to build in this place
according to the owner’s circumstances and the circumstances of the city,

57
Rabbi Yehiel Michal HaLevi Epstein (Poland, 1829–1908), Aruch HaShulhan, Hoshen
Mishpat (1892), 375, 10–16.
254 corrective justice

and the owner himself would have built there, and the builder built it
properly, in a way that the owner himself would not have improved
upon.58

The benefit must be accepted only when the improver merely antici-
pates what the owner necessarily and inevitably would have done.
This consideration is independent of the optimal use of the property.
An owner can tell the improver to remove trees planted even in a field
suitable for trees, because it is always possible that the owner prefers to
grow grain.59 The fact that the field is suitable for trees does not make
trees necessary. As long as the owner can plausibly point to some other
use for the property than the one exemplified by the improvement,
the improvement can be rejected. Epstein thus allows the improver to
interfere with the owner’s freedom to determine the use of the prop-
erty only when that interference matches what the owner would do
in any case.
Epstein illustrates the distinction between necessary and unneces-
sary improvements with a responsum authored by Rosh himself. This
instructive responsum is worth quoting almost in full:60

Reuben owned some houses, but travelled away from his city. Simeon came
and lived in them and saw that the house was tottering and about to col-
lapse. He reconstructed it and reinforced it and saved it from the danger of
falling and plastered and paneled it. When Reuben returned, he wanted to
evict Simeon from his house, saying that his initial entry was unauthorized.
Does Reuben have to reimburse Simeon . . . ? Answer: Because it was tot-
tering and close to collapsing, Reuben cannot evict him until he reim-
burses all the expenditures that Simeon made to reinforce the house and
save it from collapsing, but what Simeon built that was not necessary, like
making rooms and plastering and paneling, he did for his own benefit, and

58
Ibid., 11. By “the circumstances of the owner” Epstein means that the owner has to be
able to afford the improvement. If the owner does not have the resources to pay for the
improvement, the owner can require the improver either to remove the materials (ibid., 12)
or to buy the property at its pre-improved value (ibid., 14–15). By “the circumstances of the
city” Epstein means that the building has to be conform to the “useful building appropriate
to that courtyard as is the custom of that place.” Maimonides, (Egypt, 12th century) Mishneh
Torah, Laws of Robbery and Loss, 10, 6; Rabbi Joseph Karo, Shulhan Aruch, Hoshen Mishpat,
375, 7.
59
Epstein, above n. 57.
60
The responsum is available in the work of Rosh’s son, Rabbi Jacob ben Asher, Tur
Hoshen Mishpat, 375.
 incontrovertible benefit in jewish law 255

Reuben does not have to pay him for that but he tells him ‘Take your
wood and your stones,’ and this assessment will be according to builders
who will determine what Simeon had to expend to avoid the house’s col-
lapse, and they will assess him, and when Reuben gives this to Simeon,
then Simeon will vacate Reuben’s house.
The questioner asked further, Let our rabbi teach us: why do we not
take into account the usefulness of the construction apart from the danger
of collapse, especially since the houses had previously been plastered, and
beneath the plaster the wall was mouldy and tottering, and Simeon had to
destroy the plaster in order to fix the wall, and he fixed it to its previous
state, and he fixed gates and windows and leaky roofs and locks? The
answer: For any construction that is not to deal with the danger that the
building could not continue to exist and remain standing and be kept from
collapsing out of decay, why should we obligate the owner to pay? He can
say, ‘I do not want to spend my money on it, because the house could con-
tinue to exist without this construction. If you built it for your benefit, take
your wood and your stones.’

This responsum draws a sharp distinction between the existence and

the condition of the improved object. One cannot be sure that the
owner wants a change in an object’s condition, even if the improve-
ment restores the object to the previous condition from which it
deteriorated. The fact that the house had previously been plastered
and that the plaster had to be removed to save the house from collaps-
ing does not mean that the improver can charge the owner for replas-
tering. The owner is therefore free to insist that the improver undo
this aspect of the renovation, even though, since plaster cannot be
reused, this causes the improver a loss. However, the owner can be
compelled to pay for those aspects of the renovation that preserved
the house. This benefit is incontrovertible, so that it would be abusive
for the owner to make its conferral the reason for harming the
improver. As Epstein remarks in glossing the responsum:

The court determines whether it was necessary for the owner to do this,
and makes an assessment for him with the improver having the upper hand,
because although the improvement was without authorization, the improv-
er nonetheless conferred a benefit on him, and so why should he suffer a
loss?61

61
Epstein, Aruch HaShulhan, 375, 12.
256 corrective justice

Rosh, and Epstein following him, think that one can safely ascribe to
owners a desire to have their properties preserved, so that then they
have no legitimate reason to harm improvers by compelling the
removal of the materials. Because the owner necessarily realizes a bene-
fit from the preservation of the property, the improver is entitled to be
reimbursed for the expenses entailed in achieving that end. This
responsum shows that, although in the case of structures Rosh did not
nullify the notion of incontrovertible benefit (as he did in the case of
trees), he narrowed the basis for the improver’s claim to circumstances
in which the interference with the owner’s freedom of choice is not
significant, because the owner is merely paying for work that in any
case had to be done if the structure was to be saved from collapsing.
So much for Epstein’s narrow construction of incontrovertible bene-
fit. In contrast stands the more expansive view of Rabbi Abraham
Karelitz.62 Karelitz’s argument is that the owner must accept and
remunerate the improver at the higher level for any improvement that
increases the value of the property, unless the owner has a bona fide
reason for not accepting it. Whereas Epstein restricts what common
lawyers would think of as the benefit’s incontrovertibility to what was
necessary in the circumstances, Karelitz broadens it to include what-
ever adds value without genuine inconvenience to the owner.
Karelitz’s view of the circumstances under which the improvement
must be accepted arises from a reconceptualization of what it means
to say that a field is suitable for planting trees. A field is considered
suitable for planting not because trees are its optimal use, but because
a field with trees is worth more than a field with grain. In his view, the
difference between a plantable and non-plantable field is simply a
matter of the objectively higher economic value of the former. If the
improvement increases the value of the field, the owner is assumed to
want the improvement and must remunerate the improver at the
higher rate. Similarly, if the owner reveals that he actually wants the
improvement—as was the case with the owner who fenced and
guarded the trees in the dispute that came before Rav—then the
owner must pay on the higher measure even if improvement does not
add value to the property. What is decisive is the benefit to the owner,
whether that benefit is manifested in the increased value of the prop-
erty or in conduct revealing the owner’s desire for the improvement
even though it does not increase the property’s value.

62
Rabbi Abraham Karelitz (1878–1953), Hazon Ish, Hoshen HaMishpat, Baba Batra 2, 3.
 incontrovertible benefit in jewish law 257

To this notion of benefit objectively manifested through value or

conduct, Karelitz adds an important qualification: an owner who has a
bona fide reason for not wanting an improvement that increases the
property’s value can refuse to pay and can have the materials removed.
This qualification too involves an objective inquiry into the circum-
stances and motivations of the owner. Karelitz would have had little
sympathy for the suggestion that the possibility of the owner’s subject-
ive devaluation of the benefit immunizes the owner from liability.63
For Karelitz Jewish law does not, and never did, attach any signifi-
cance to the owner’s rejection of the benefit on the strength of “pri-
vate reasons of his own.”64 The freedom of the owner to determine
the use of his or her property is not absolute; it must reflect a plausible
reason for rejecting a benefit that increases the value of the property.
“Everything is according to what appears to the eye of the judge, as to
whether the increased value is truly not to the benefit of the owner,
or whether the owner is just saying so to put the improver off.”65
Karelitz reinterprets even the dispute between Rosh and his predeces-
sors about whether the owner of a plantable field had to accept the
trees as involving not a difference of legal principle but a distinction
on the facts. Rosh’s predecessors, who held that the owner of a field
suitable for planting could be compelled to pay for the improvement,
were merely referring to a situation where

the owner has no excuse or reason for why he would not want to plant the
field, and we see his response ‘Take your seedlings’ as being merely for the
sake of angering the improver, as if to say ‘Neither I nor you will get any-
thing,’ or as evading payment; but if we see that it is actually the truth that
he does not want the seedlings and he wants them uprooted, then we treat
him as having field that is not suitable for planting.66

Similarly, when Rosh allows the owner to reject the improvements, he

is dealing with a situation in which the owner has a genuine reason
for preferring his land in an unimproved state.
Karelitz accordingly regards Rosh’s statement about the rebuilt
ruins67 as illustrative of these principles. It will be recalled that Rosh

63
Birks, above n. 2.
64
Karelitz, above n. 62, second para.
65
Ibid., Baba Kama 22, 6.
66
Ibid., Baba Batra 2, 3.
67
Above n. 55.
258 corrective justice

held that, although the owner generally could have the improvement
removed, the owner of a ruin rebuilt without authorization could not
order the improver to remove his materials if he was really seeking a
pretext to inflict loss on the improver. Karelitz noted that Rosh indi-
cated two circumstances that presented genuine reasons for not want-
ing the rebuilt structure. The first was that the owner was already
using the ruin for something else, so that the pre-existing use could
be taken as a manifestation of his genuine desires concerning the
property. The second was that the rebuilding impaired his livelihood,
which Karelitz interpreted as meaning that the owner could not
afford to pay for the improvement. If, however, despite such consid-
erations, the owner moved into the rebuilt ruin and began to live
there, this conduct could be taken as a manifestation of his acceptance
of the benefit, triggering an obligation to remunerate the improver
on the higher measure.68
In sum, Jewish law refers to three different conceptions of what
common lawyers regard as incontrovertible benefit. The earliest is that
the owner can be compelled to accept an improvement that moves
the property to its optimal use. This conception was destroyed by
Rosh, who held that, regardless of the nature of the property, the
owner is at liberty to order the improver to remove the improvement.
In qualifying this with the observation that the right to order the
removal of building materials cannot be turned into a pretext to cause
loss to the improver, Rosh allowed incontrovertible benefit to be
reborn. One form of this rebirth was the stringent view that the owner
was barred from rejecting the improvement only if the improvement
was necessary. The other form was the more liberal view that the
owner had to accept any improvement that increased the value of the
property, except if the owner had a genuine reason for rejecting it.

68
Karelitz, above n. 62, at 2, 6. Karelitz’s view has one additional complexity that I wish
merely to mention for the sake of completeness. Having reconceptualized the suitability for
planting trees in terms of an increase in the property’s value, Karelitz was faced with the dif-
ficulty of explaining why the Talmud and the legal tradition to this point distinguished
between fields that were or were not suitable for planting trees. For on Karelitz’s view what
matters ultimately is not the kind of field it is but whether the owner receives a benefit
manifested either through value or conduct. Karelitz’s solution to this difficulty was that even
if the owner did not want the improvement and had genuine reason to reject it, the owner
might nonetheless be willing to put up with it.Then whether the field was suitable for plant-
ing trees (that is, in Karelitz’s view, whether the trees increased the value of the land) would
determine the rate of the improver’s remuneration; Karelitz, above n. 62, at 3, 4.
 incontrovertible benefit in jewish law 259

6. Conclusion
In this chapter I have traced the main lines in the development of the
Jewish law governing unrequested improvements.The point of depart-
ure for this development is the story in the Talmud of Rav’s dealings
with the owner whose field was planted with trees. From this story
emerge subsequent discussions of the difference between having the
upper and the lower hand, of the significance of an improvement that
puts the property to its optimal use, and of the conditions under which
the owner could be compelled to accept the benefit.
At first blush, the Talmud’s account of the trees planted in an-
other’s field and the jurisprudence that flows from it might, from the
standpoint of the common law, seem peculiar on a number of
grounds. The most important of these is that throughout the Tal-
mudic passage Rav seeks, with eventual success, to have the owner
pay for the improvement. In the eye of the common law, this solici-
tude for the improver seems misplaced. By planting trees in what he
knew was another’s field, the improver was the most unappealing of
restitutionary claimants, a mere volunteer or officious intermeddler.
Conversely, the owner should surely be able to determine for him-
self whether to plant trees in his own field. Right from the begin-
ning Rav’s assumption that the owner should be assessed for the
trees seems eccentric.
A second peculiarity is that Jewish law gives the knowing improver
a preferential position over an innocent one. Only the improver who
knows that the property being improved belongs to another can claim
remuneration on the higher measure, which under some circumstanc-
es gives the improver a share of the yield and under other circum-
stances allows the improver to recover expenditures in excess of the
property’s increased value. This is because the basis of such remuner-
ation is what the owner would have paid to have the improvement, a
quasi-contractual measure that presupposes that the improver is ren-
dering a service to someone else rather than merely being mistaken
about the extent of one’s own ownership.
A third peculiarity is that situations in which the materials for the
improvement (the trees in the field example, the wood and stones in
the building example) can be removed are paradigmatic for the dis-
cussion of unsolicited improvements. English law encapsulates its con-
cerns in Chief Baron Pollock’s famous question, “One cleans another’s
260 corrective justice

shoes; what can the other do but put them on?”,69 because it assumes
that the benefit has been irretrievably entangled in the owner’s prop-
erty. Jewish law, in contrast, expresses the owner’s rejection of the bene-
fit through the owner’s telling the improver to “take your materials
and go.” In part this reflects the idea that even after the materials have
been affixed to the owner’s land or structure, the owner acquires prop-
erty in them only on signaling acceptance of the improvement by
offering to pay.70 More deeply, however, it reflects a commitment to
restoring the parties to their positions before the improvement, or at
least to preventing the owner from ostensibly rejecting the benefit
while continuing to enjoy it. When the Jewish jurists turned their
attention to unremovable benefits (the classic example was dyeing
someone else’s wool), they adopted an approach similar to the one
that governed removable ones: the owner had to pay unless there was
reason to suppose that the improvement was not in fact a benefit.71
Perhaps from the perspective of systems other than the common
law, some of these peculiarities might not seem all that eccentric after
all. The possibility of removing an enrichment for which the owner is
not liable was mentioned by the Roman jurist Celsus in the second
century, is present in contemporary German law, and has been sug-
gested for the common law as well.72
Be that as it may, the differences between Jewish law and the com-
mon law mask an important similarity. Although the particular moves
about unrequested improvements in each system diverge, these moves
respond to the same challenge. When dealing with unrequested bene-
fits, any rational system of private law must reconcile the owner’s
freedom to determine the use of his or her property with the improv-
er’s claim that the owner should not be unjustly enriched at the
improver’s expense. As noted in section 1 of this chapter, this in turn
requires attention to two issues.The first issue is whether the improve-
ment is a gift from the improver to the owner; if it is, the improver has
no reason to complain that the owner’s use of it is unjust. The second
issue is whether, even assuming that the improvement was not a gift,

69
Taylor v. Laird [1856] 156 E.R. 1203.
70
Rashba, above n. 11.
71
Rabbi Jacob Lorbeerbaum (Poland, 1760–1832), Netivot HaMishpat on Shulhan Aruch
375; Karelitz, above n. 62, at 2, 6.
72
Dirk A.Verse, “Improvements and Enrichment: A Comparative Analysis,” (1998) Resti-
tution L. Rev. 85, at 88, 102–3.
 incontrovertible benefit in jewish law 261

acceptance of the improvement can reasonably be imputed to the

owner; if it can, there is no injustice in compelling the owner to pay
the improver. The first of these issues directs us to consider whether
the improver has acted with donative intent; the second directs us to
consider the conditions under which an unrequested benefit would
nonetheless be consistent with the autonomy of owners with respect
to their own property. Taken together, these two issues vindicate the
corrective justice conception of private law by insisting that, in the trans-
fer of a benefit from the improver to the owner, both its bestowal and
its receipt are the expression of the free will of the parties.
The law’s treatment of these two issues, although of course based
on the specific events of the transaction in question, is a matter not of
fact but of juristic construction. It involves not merely ascertaining
what happened but working out the relevant legal categories and
ascribing meaning in their light to the conduct of the parties. Differ-
ent legal systems, while addressing the same issues, can nonetheless
reasonably differ in their construction of the legal categories, or in the
meaning they ascribe to the parties’ conduct, or in their understand-
ing of the relationship between conduct and categories.
The contrasting attitudes toward donative intent in the common
law and in Jewish law provide a dramatic illustration of this. Both the
common law and Jewish law concern themselves with whether a
knowing unsolicited improvement is to be construed as a gift, but
they elaborate different answers. Because the common law imputes to
the parties the awareness that only the owner can determine the use
of what is owned, the common law includes the taking of a risk of
being remunerated within its conception of the improver’s donative
intent, with the result that volunteering or intermeddling becomes a
fatal obstacle to the improver’s claim. Jewish law does not regard the
improver as a risk-taker but looks more single-mindedly at the intent
or the presumed intent of the improver, who is therefore assumed not
to have acted with donative intent. As a result, the only substantial
barrier in Jewish law against liability for unrequested improvements is
the second issue of determining the conditions under which accept-
ance of the improvement can reasonably be imputed to the owner.
It is, accordingly, hardly surprising that the Jewish law, over the
eighteen centuries of its recorded discussions of unrequested improve-
ments, has on this issue elaborated the variety of views outlined in this
chapter. These discussions direct attention either to the optimal use of
the property being improved, or to the necessity for the improvement,
262 corrective justice

or to the genuineness of the owner’s reason for not welcoming the

increase in the property’s value. The common law, in contrast, both
because of the relative youth of its law of restitution and because of
the filtering effect of the issue of volunteering, is only at the begin-
ning of a similar elaboration of its parallel idea of incontrovertible
benefit. However, with the intense interest that the law of restitution
is now enjoying in the common law world, we common lawyers can
perhaps already sympathize with the observations expressed by the
distinguished author of a sixteenth-century responsum about unre-
quested improvements: “I do not have time to go on at length about
these matters, because there are many controversies and the questioner
is pressing.”73

73
Maharashdam (Rabbi Samuel de Medina, Salonika, 16th century), Responsa of the
Maharashdam, Hoshen Mishpat 227.
 8
Poverty and Property in
Kant’s System of Rights

1. Kant on the public duty to support the poor

In a passage from the Doctrine of Right that is particularly enigmatic
even by his own high standards, Kant announces the state’s right to
tax in order to fulfill a public duty to support the poor.1 The passage
raises fundamental issues about the interpretation of Kant’s legal
philosophy, about the connection between private law and public law,
and about the conceptual resources available to a system of rights for
dealing with poverty. Kant, however, says almost nothing about the
basis of this duty. In his Preface to the Doctrine of Right Kant had
remarked that he had elaborated the latter sections less thoroughly
than the earlier ones, because inferences from the earlier sections
could easily be made.2 The sparseness of his treatment of the duty to
support the poor exemplifies this less thorough elaboration, but—as is
evident from the perplexity to which the passage has given rise—not
the ease of inference.
One might suppose that state taxation to support the poor involves
a clash between distributive and corrective justice. Whether the state
should satisfy the basic needs of its citizens is, of course, a standard
issue of distributive justice.This recourse to distributive justice requires
the state to use its taxing powers to take something that would other-
wise remain within the private resources of those taxed. In a well-
ordered state these resources reflect proprietary rights worked out and
protected by private law within a conception of corrective justice.

1
For the text of the passage, see below nn. 7 and 9.
2
Immanuel Kant, The Metaphysics of Morals, tr. Mary Gregor (1996), [6:209].
264 corrective justice

Thus, the state’s support of the poor, one might think, accomplishes
distributive justice at the expense of citizens’ corrective justice entitle-
ments.
This supposed clash between distributive and corrective justice
leads to the temptation to eliminate one form of justice in favor of the
other. Contemporary legal and political thinking shows this tempta-
tion operating in both directions. Those opposed to the state’s distrib-
utive operations claim, in effect, that corrective justice is all the justice
that there is.3 On this view, justice is fully satisfied by the private law
notions that recognize entitlements to property and personal integrity,
allow for the voluntary transfers through contract and gift, and protect
rights through the law of contract, torts, and unjust enrichment. These
notions themselves are interpreted as embodying a mode of practical
reason distinctive to private law in that it works justice between indi-
vidual parties without reference to any distributive purposes. “Dis-
tributive justice” can be regarded merely as a euphemistic term that
camouflages the injustice of the state’s treating individuals and their
entitlements as means to collective ends. This primacy of corrective
justice honors private law entitlements while renouncing the exist-
ence of a state obligation to satisfy citizens’ basic needs.
On the other hand, those who favor the state’s distributive role may
be tempted to regard the working of distributive justice as norma-
tively fundamental. The doctrines of private law then become nothing
more than special operations of distributive justice. On this view,
property can then be seen simply as the residue remaining after the
state’s distributive activity rather than as a locus of independent nor-
mative significance.4 Liability rules also, whether dealing with con-
tracts, torts, or unjust enrichment are regarded as justified to the extent
that they embody distributive moves.5 State support for the poor is
then merely one distributive operation among many. Abandoned or
explained away is the distinctive significance of the private law con-
cepts as the legal manifestations of corrective justice.

3
The leading representative of this view in recent years was Robert Nozick (though he
would not have described his position in the terminology I use in this paragraph); see Robert
Nozick, Anarchy, State, and Utopia (1974), 149–82.
4
For a recent example of this, see Liam Murphy and Thomas Nagel, The Myth of Owner-
ship: Justice and Taxes (2002), 74–75, 173–77.
5
Notable examples for contract, torts, and unjust enrichment, respectively, are: Anthony
Kronman, “Contract Law and Distributive Justice,” (1980) 89 Yale L.J. 472; Gregory Keating,
“Reasonableness and Rationality in Negligence Theory,” (1996) 48 Stan. L. Rev. 311; Hanoch
Dagan, The Law and Ethics of Restitution (2004).
 poverty and property in kant’s system of rights 265

Kant’s remarks on the state’s right to tax in order to fulfill a public

duty to the poor indicate that he does not share these one-sided views
of justice. As a philosopher working within the tradition of natural
right—indeed, as perhaps its greatest expositor—Kant gives a detailed
non-distributive account of the principal features of private law, espe-
cially of property and contract. Developing corrective justice in terms
of his own metaphysics of morals, Kant portrays private law as a sys-
tem of rights whose most general categories give juridical expression
to the coexistence of one person’s action with another’s freedom
under a universal law.Yet despite his affirmation that private law entitle-
ments, understood non-distributively, are the necessary components
of a free society, Kant nonetheless holds that there is a public obliga-
tion (and not merely a liberty) to support the poor. He thus seems to
regard this aspect of distributive justice as compatible with corrective
justice, with the state being duty-bound to actualize both. Neither of
the temptations that characterize certain contemporary approaches to
law attracts him.
However, the question that arises is whether Kant is entitled to
the view about the alleviation of poverty that he professes. Kant’s
view of property is at least as extreme as the most extreme of today’s
libertarians.6 How on his view can the state function both as the
guarantor of purely non-distributive property rights and as the pub-
lic authority that levies taxes in order to fulfill a public duty to sup-
port the poor? This question is all the more serious because Kant is
a systematic philosopher for whom obligation signifies necessity, so
that the duty to support the poor that he posits must somehow arise
out of, and not merely be consistent with, his non-distributive
account of rights. Furthermore, for Kant rights are the juridical vin-
dications of freedom that the state coercively protects against
infringement; coercion for the benefit of anyone, including the poor,
seems inadmissible within the Kantian framework. Kant offers almost
nothing resembling an argument in support of the duty he announces.
Nor does he explain how this duty is to be integrated into his aus-
tere system of rights.
In the crucial passage, appearing in his section on public right in
the Doctrine of Right, Kant describes the state’s right to tax in order to
fulfill its duty to the poor in these terms:

6
Kant, for instance, has nothing like the Lockean proviso that limits property rights for
Nozick; see Nozick, above n. 3, at 178–82.
266 corrective justice

To the supreme commander there belongs indirectly, that is, insofar as he has
taken over the duty of the people, the right to impose taxes on the people
for its own preservation, such as taxes to support organizations providing
for the poor, foundling homes, and church organizations, usually called charita-
ble or pious institutions.7

Because for Kant a right is always connected to the authorization to

use coercion,8 Kant goes on to specify that the state’s support of the
poor should be achieved by coercive public taxation and not merely
by voluntary contributions. He explains the basis of the right to tax as
follows:

The general will of the people has united itself into a society that is to
maintain itself perpetually; and for this end it has submitted itself to the
internal authority of the state in order to maintain those members of
the society who are unable to maintain themselves. For reasons of state the
government is therefore authorized to constrain the wealthy to provide the
means of sustenance for those who are unable to provide for even their
most necessary natural needs. The wealthy have acquired an obligation to
the commonwealth, since they owe their existence to an act of submitting
to its protection and care, which they need in order to live; on this obliga-
tion the state now bases its right to contribute what is theirs to maintaining
their fellow citizens.9

No reader of Kant’s legal philosophy can fail to be struck by the appar-

ent oddity of these paragraphs. Kant’s legal philosophy is an elucida-
tion of the concept of Right—that is, of “the sum of conditions under
which the choice of one can be united with the choice of another in
accordance with a universal law of freedom.”10 In introducing the
concept of Right, Kant notes that “it does not signify the relationship
of one’s choice to the mere wish (hence also to the mere need) of
others, as in actions of beneficence . . .”11 The consequence of this
abstraction from “mere need” is a complex of proprietary, contractual,
and domestic rights which place others under correlative duties of
non-interference, “for anyone can be free as long as I do not impair

7
Kant, above n. 2, [6:328].
8
Ibid., [6:231].
9
Ibid., [6:326].
10
Ibid., [6:230].
11
Ibid.
 poverty and property in kant’s system of rights 267

his freedom by my external action, even though I am quite indifferent

to his freedom.”12 Yet when outlining the rights of government in the
quoted paragraphs, Kant introduces—seemingly out of the blue—a
positive duty, which government takes over from the people, to sup-
port those “unable to provide for even their most necessary natural
needs.” As Jeffrie Murphy remarks, “it is very difficult to see what
Kant is up to.”13
Kant’s legal philosophy is so parsimonious and its architecture so
austere that little leeway is available in dealing with a perplexity of this
sort. Kant’s adamantine boundary between right and ethics—the
former dealing with externally coercible duties, the latter with unco-
ercible duties done for their own sake—prevents recourse to appeal-
ing ideas found in Kant’s writings on ethics. For example, because
Kant does not formulate the duty to support the poor as the reflex of
any correlative right that the poor have, one might be tempted to
regard that duty as somehow connected to the personal duty, postu-
lated by Kant elsewhere,14 to come to another’s aid. However, the duty
to aid is an ethical rather than a juridical one; it therefore cannot be
associated with the coercive taxation authorized for support of the
poor. Kant’s own description of the concept of Right, with its con-
trast between rightful actions and actions of beneficence, confirms
that state support of the poor does not fall under the duty to aid.15
Some commentators have seen Kant’s requirement of support for
the poor as an expression not of benevolence but of political pru-
dence.16 The alleviation of poverty facilitates the state’s survival by

12
Ibid., [6:231].
13
Jeffrie Murphy, Kant:The Philosophy of Right (1970), 145.
14
Above n. 2, [6:393]; Groundwork of the Metaphysics of Morals, Morals, tr. Mary Gregor (1996),
in The Cambridge Edition of the Works of Immanuel Kant: Practical Philosophy, [4:423], [4:430].
15
For an argument that support of the poor exemplifies a duty of beneficence, see Allen
D. Rosen, Kant’s Theory of Justice (1993), 173–208. Rosen draws on the duty to aid another
that Kant sets out in the Groundwork, above n. 14, to conclude (at 201) that:
if no one can rationally will the maxim of never helping others as a law of nature . . . then neither can
an entire people rationally will as a law of political society that the state should allow them to perish
rather than supply their basic needs. The same reason that makes it impossible rationally to will the
maxim of never helping others as a law of nature also makes it impossible rationally to consent to a
law of political society that would permit the state to ignore the basic needs of its citizens.
However, it by no means follows from the notion that everyone is obligated to help another
that everyone is also obligated to give the state the power to coerce such help.
16
Mark LeBar, “Kant on Welfare,” (1999) 29 Canadian J. Phil. 225; Wolfgang Kersting,
“Kant’s Concept of the State,” in Essays on Kant’s Political Philosophy, ed. Howard Williams
(1992), 143, 164; Bruce Aune, Kant’s Theory of Morals (1979), 157; Mary Gregor, Laws of Free-
dom (1963), 36.
268 corrective justice

promoting the state’s strength and stability against internal disorder

and external attack. Kant elsewhere indeed seems to authorize the
state to legislate on this basis for the happiness and prosperity of its
citizens.17 However, the acknowledged instrumentalism of such legis-
lation18 fits awkwardly into Kant’s exposition in the paragraphs quoted
above. In these paragraphs the relief of poverty is viewed not as some-
thing from which the state might contingently benefit, but as a duty
of the people that the state assumes. Like all duties in Kant, this duty
presumably reflects a normative necessity rather than a prudential
option.19
In this chapter I want to develop a different possibility.20 My
contention is that, far from being inconsistent with the internal
logic of Kantian right, the state’s duty to support the poor is the
inexorable outcome of that logic. Kant includes support of the
poor as an “effect with regard to rights that follows from the nature
of a civil union.”21 The civil union results from the transition to
public right from the property regime in the state of nature. Kant’s
theory of property rights necessitates not only this transition but
also—as its consequence—the people’s duty to the poor. Just as for
Kant the movement from property in the state of nature to the
public right of a civil union is obligatory, so the state’s support of
the poor is an obligatory consequence of that movement. Were the
state under no such obligation, the legitimacy of the civil union
that replaces the state of nature would itself be impugned. On this
reading of Kant, the very idea of private property implies the state’s
right to tax property owners in order to discharge a public duty to
relieve poverty.

17
Kant, “On the Common Saying: That May Be Correct in Theory but Is of No Use in
Practice,” in The Cambridge Edition of the Works of Immanuel Kant: Political Philosophy, tr. Mary
Gregor (1996), [8:298].
18
Ibid.: “If the supreme power gives laws that are directed chiefly to happiness (the pros-
perity of the citizens, increased population and the like), this is not done as the end for which
a civil constitution is established but merely as a means for securing a rightful condition, espe-
cially against a people’s external enemies.”
19
Kant, above n. 2, [6:222] says that duty, “that action to which someone is bound,” is “the
matter of obligation,” and he defines obligation as “the necessity of a free action under a
categorical imperative of reason.”
20
For brief suggestions of an approach similar to the one set out here, see Mary Gregor,
“Kant on Welfare Legislation,” (1985) 6 Logos 49, 55; Leslie Mulholland, Kant’s System of
Rights (1990), 317, 39 (see below n. 81). For a related interpretation see Arthur Ripstein, Force
and Freedom: Kant’s Legal and Political Philosophy (2009), 267–99.
21
Kant, above n. 2, at [6:318].
 poverty and property in kant’s system of rights 269

Interpreted in this way, Kant’s account casts light on two aspects of

corrective justice. The first concerns the place of property within a
corrective justice theory of private law. It is easy to assume that prop-
erty is problematic for corrective justice: because private law presup-
poses an antecedent distribution of property, corrective justice itself
seems to rest on a foundation of distributive justice.22 Kant, however,
rejects the assumption that the justification for property lies in an
antecedent distribution. For Kant, as the next section of this chapter
explains, property arises from the freedom to act in relation to others
in accordance with self-chosen purposes. This freedom characterizes
the conception of the person—what was termed “personality” in
chapter 1—that corrective justice presupposes. Property so conceived
completely conforms to corrective justice. In this account property
does not arise even notionally through distributive justice—that is,
through a determination by a common authority to parcel goods out
on the basis of a distributive criterion. To the contrary, common
authority itself becomes possible only as a consequence of the right-
fulness of property.
The second aspect on which Kant’s account casts light concerns
the relationship between corrective and distributive justice. As noted
above in chapter 1, these two forms of justice connote categorically
different structures of justification. Neither of them can integrate the
other within it. Given the legitimacy of both forms of justice, what
(aside from the sheer positivity of laws within a single jurisdiction) is
the nature of the unity of a legal order within which both are present?
Clearly, it cannot be the unity of the same justificatory structure
applying to all its legal arrangements. Implicit in the Kantian account
as interpreted in this chapter is a different answer: it is unity by appro-
priate sequence.23 Starting from the underlying conception of person-
ality, Kant traces a series of conceptual steps each of which
presupposes but complements the preceding one until the full spec-
trum of rightful legal arrangements is exhibited. The underlying unity
is provided by the idea of what self-determining persons require in
order to realize their freedom through law. Under this notion of unity
neither corrective justice nor distributive justice trumps the other, nor
is either derived from the other. Rather each finds its appropriate

22
Hanoch Dagan, “The Distributive Foundation of Corrective Justice,” (1999) 98 Mich.
L. Rev. 138.
23
This phrase is taken from John Rawls, Political Liberalism (1993), 259–62.
270 corrective justice

place in the conceptual sequence that actualizes the reciprocal free-

dom of all.
In accordance with this conceptual ordering, property generates
the distributive justice that consists in the alleviation of poverty
through taxation. Far from being a self-sufficient and free-standing
institution of justice, property requires redistribution to the poor for
its own legitimacy. Thus Kant transcends the categorical difference
between corrective and distributive justice while preserving and elu-
cidating the distinct roles that each plays in a free society. In this chap-
ter I attempt to reconstruct the argument, implicit in his theory of law
but not articulated by Kant himself, that underlies this remarkable
conclusion.

2. Kant’s account of property

Kant’s account of property in the Doctrine of Right features a concep-
tual progression that starts from the innate right to freedom and cul-
minates in the establishment of property as an institution of positive
law.24 Kant exhibits the phases of this progression as implicit in the
relationship of free persons under the conditions of human existence.
Because property is consistent with the freedom of all, it is rightly
secured and protected by the law’s coercive powers.
This progression has three phases, which Kant presents from a var-
iety of standpoints as befits their structural importance. Sometimes he
describes these phases in terms of the categories of modality (the pos-
sibility, the actuality, and the necessity of possessing objects).25 Some-
times, he refers to them as divisions of justice (iustitia tutatrix, iustitia
commutativa, iustitia distributiva).26 Sometimes he refers to the division

24
A terminological note: The word “property” that I use in this section is not strictly
accurate from a Kantian perspective. Kant is dealing with the more general idea of that he
calls “having something external as one’s own” or “external mine and thine” [6:245]. What
we think of as property, namely, things that have an in rem legal status, is just part of what Kant
is referring to. Kant’s “external mine and thine” also includes in personam contractual entitle-
ments and entitlements within the status relationships of the household [6:247–48]. Perhaps
“ownership” might be a more appropriate English term for capturing Kant’s capacious
notion of what one has as one’s own, although I doubt that English-speakers would differ-
entiate between ownership and property in this way. (Kant himself uses the word “owner-
ship” (Eigentum) for what we would call “property” [6:270].) I have simply used the word
“property” throughout, despite its inaccuracy.
25
Kant, above n. 2, [6:306].
26
Ibid., [6:267], [6:306].
 poverty and property in kant’s system of rights 271

of duties that accompanies the divisions of justice.27 Sometimes he

refers to these phases in terms of form and matter.28 Sometimes he
calls them different variations of right (what is intrinsically right, what
is rightful, what is laid down as right)29 or different kinds of laws of
justice (lex iusti, lex iuridica, lex iustitiae).30
However the phases are referred to, the progression through them
exhibits a dialectical structure of argument.31 In the first phase Kant
starts with the universal principle of Right, which mandates the coex-
istence of one person’s action with another’s freedom under a universal
law, and notes the juridical relationship analytically contained within
that principle. This juridical relationship does not include property in
external things, but it does encompass certain “authorizations” such as
equality and non-dependence,32 which are normative attributes impli-
cit within the universal principle of Right and therefore ascribable to
the parties at this phase. In the second phase he extends this initial
argument on the ground that having something external as one’s own,
although not analytically contained in the universal principle of Right
marks a connection to external things that matches the capacity for
choice characteristic of self-determining action. This extension, how-
ever, is problematic, because although ownership of external things is
now permissible, it is not yet put into effect under conditions conson-
ant with the authorizations articulated in the first phase. The second
phase, accordingly, is merely provisional. The problems it raises are
resolved at the third phase, where the conditions of acquisition take a
form that is fully consistent with what was analytically contained in
the universal principle of Right. As Kant puts it with unfortunate opa-
city when he lists the threefold division of duties, the duties of the third
phase “involve the derivation of the [duties of the second phase] from
the principle of the [duties of the first phase] by subsumption.”33
Although presented in a sequence, these three phases are conceptual,
not temporal. Kant is not offering a philosophical reconstruction of
the historical evolution of property. Rather, the three phases represent

27
Ibid., [6:236].
28
Ibid., [6:306].
29
Ibid., [6:267], [6:306].
30
Ibid., [6:236–37], [6:267], [6:306].
31
Kant, ibid., [6:255] notes that the concept of having something external involves “an
antinomy of propositions concerning [its] possibility” that forces reason into “an unavoidable
dialectic.”
32
Ibid., [6:237–38].
33
Ibid., [6:237].
272 corrective justice

aspects that together are constitutive of property in the juridical rela-

tionships of free persons (e.g., that external things can be acquired
through acts of will, that property does not require actual possession,
that property rights are enforceable, and so on), but presented in an
ordering that purports to show property’s normative necessity within a
system of rights. The three phases comprise an articulated unity: each
phase proceeds with its distinct mode of argumentation (the first is
analytic, the second is synthetic, the third works by subsumption), but
the account of property stands or falls on the totality of the three
phases taken together. Kant himself presents property as absent at the
first phase and as problematic at the second. If these phases were con-
sidered independently, the argument would not get off the ground or
would collapse as soon as it did so. Nor does the third phase stand
alone either; its role is to incorporate what is necessary to reconcile the
second phase to what is analytically contained in the first one. The
result is that the institutions of public law that emerge at the third
phase determine and guarantee the property entitlements that are the
product of the second phase in a way that expresses the normative sig-
nificance of the principle of right that initiated the first phase.
The first phase features the innate right to freedom. The innate
right to freedom consists in the independence of one’s actions from
constraint by the actions of another, insofar as such independence is
consistent with the freedom of everyone else.34 This right stands in an
analytic relationship with the universal principle of Right, which
requires that one person’s action be able to coexist with the freedom
of everyone under a universal law. Formulating freedom as an innate
right adds nothing to what the universal principle already contains; it
merely isolates a constituent element of, and represents what is already
involved in thinking about, that principle.
The innate right is “the only original right belonging to every man
by virtue of his humanity.”35 This right is innate because every person
has it simply by virtue of his or her existence. Similarly, it is original
because it arises independently of any act that would establish it.
Because my innate right is not mine by virtue of some act of acquisi-
tion, it is what is internally mine, in contrast to what is externally
mine, which must always be acquired.36 What is internally mine is my

34
Ibid.
35
Ibid.
36
Ibid.
 poverty and property in kant’s system of rights 273

freedom37—that is, my capacity to act in the execution of the pur-

poses I form as a self-determining being.
For human beings the paradigmatic manifestation of what is inter-
nally mine is the body, the physical organism through which the per-
son expresses his or her freedom as a self-determining being.38 By
mandating actions that can coexist with the freedom of all, the uni-
versal principle of Right signals its application to the actions of self-
determining agents. In the case of human beings, self-determining
activity takes place through the body. Because the body is an “insep-
arable unity of members in a person,”39 interference with any part of
another’s body is a wrong against that person’s freedom. This right
with respect to one’s own body is innate. It arises not through the
performance of an act of acquisition (indeed, no such act is conceiv-
able because the body itself is what would have to perform it), but
simply by virtue of one’s being born. Thus, the body is the primary
locus of what Kant calls the “right of humanity in our own person.”40
The occupation by a person’s body of a particular space is an exer-
cise of this right: “All men are originally (i.e., prior to any act of choice
that establishes a right) in a possession of land that is in conformity
with right, that is, they have a right to be wherever nature or chance
(apart from their will) has placed them.” 41 Given the finitude of the
earth’s surface, the occupation of space carries with it the possibility of
persons coming into contact with one another.42 Such contacts are
governed by the universal principle of Right. Because no one can
interfere with the body of anyone else, a person who occupies a par-
ticular space excludes all other persons from that space.
In this phase, where one’s only right is the innate right of humanity
in one’s own person, property as the entitlement to something distinct
from the person’s body does not exist. Of course, a person may come

37
Ibid., [6:248] (“what is internally mine (freedom)”), [6:250] (“what is internally mine
(my freedom)”).
38
Kant does not say this explicitly, but it is clear from what he does say about the innate
right; see Leslie Mulholland, above n. 20, 214. In addition to the passages Mulholland cites,
one can adduce the significance of the right in our own person for legitimate sexual relations
(Kant, above n. 2, [6:277], [6:278]) and Kant’s reference to “my innate right to security of the
person” in his Vorarbeiten, in Immanuel Kant, Gesammelte Schriften, XXIII (1910), 287.
39
Kant, above n. 2, [6:278].
40
Ibid., [6:236], [6:239]. Other aspects of the innate right that Kant mentions are the
infant’s right to parental support (ibid., [6:280]), the right to a good reputation (ibid., [6:295]),
and the right to one’s religious beliefs (ibid., [6:327]).
41
Ibid., [6:262].
42
Ibid.
274 corrective justice

into physical possession of some external object. I might (to use Kant’s
examples)43 hold an apple or lie on the earth. But someone who
wrested the apple away from me or pushed me off the land on which
I was lying would be wronging me with respect to my body, not my prop-
erty. By disturbing the disposition of my fingers as they grasped the apple
or of my physical frame as it rested on the earth, the wrongdoer would
be acting inconsistently with my innate right to occupy a particular
space, rather than infringing a right that I have in the apple or in the
resting place as such. The interference would be with what is intern-
ally, not externally, mine.
Property goes beyond innate right by treating the person as entitled
to an external thing even when it is not in the person’s physical pos-
session. Innate right prohibits another’s interference with an external
thing only insofar as such interference would simultaneously be an
interference with my body as something internally mine. Property, in
contrast, entails treating the thing as externally mine, so that the apple
I was holding remains mine even when I set it down, and similarly the
land upon which I was lying remains mine even when I have moved
elsewhere. Under a property regime anyone who interferes with what
is mine wrongs me despite the fact that my body is not immediately
affected.
The extension of the scope of rights to include what is externally
mine is the second phase of Kant’s account of property. Kant intro-
duces what he calls “the postulate of practical reason with regard to
rights,” under which “it is possible to have any external object of my
choice as mine.”44 This postulate asserts both the possibility of owning
the external objects of a person’s will and the existence of a duty of
justice to act towards others in recognition of that possibility.45
The postulate is based on the notion that external objects of choice
have to be conceived in a way that corresponds to the choosing sub-
ject. Under the concept of right, what is relevant are not the particular
purposes that choosing subjects pursue through their interactions with
each other, but rather their purposiveness as choosing subjects what-
ever their particular purposes. A contract is valid, for instance, because
it expresses the purposiveness of both contracting parties, rather than
because of the particular purpose that either party has in mind. Thus,

43
Ibid., [6:247].
44
Ibid., [6:246].
45
Ibid., [6:252].
 poverty and property in kant’s system of rights 275

from the juridical perspective freedom is a formal concept that refers

to the capacity for choice rather than directly to the content of par-
ticular choices.46 It thereby abstracts from the wants and needs that
fuel such particular choices. External objects of choice are the objects
of choice so conceived. Accordingly, an external object of choice can-
not get its juridical status merely from the particular properties that
are engaged by particular uses that satisfy particular wants or needs of
a particular choosing subject, as when the apple is held in the hand
and then consumed. Rather, it must be possible for an object of choice
to lie within the choosing subject’s capacity for use even when no
particular use is being made of it.
Indeed, if an external object of choice could not rightfully be with-
in a choosing subject’s power, the latter’s freedom would be restricted
beyond what the universal principle of Right required. That principle,
under which the action of one person must coexist with the freedom
of another, makes the freedom of another—the other’s innate right of
humanity in her own person—the limit of one’s action. Accordingly,
everything that is distinct from what a person has an innate right to is
available as an external object of choice. Nothing intrinsic to such an
object of choice can put it beyond the rightful power of the choosing
subject to exercise control over it even when not in physical posses-
sion of it. In this way the postulate of practical reason with respect to
rights makes it possible to have something external as one’s own, thus
extending the scope of rights beyond the body-bounded regime of
innate right.
Yet the rights made possible by the postulate are problematic.
Although they are consistent with the regime of innate rights in one
respect, they are inconsistent with it in another. They are consistent
with it by allowing persons to exercise their freedom by controlling
external objects of choice, which are not aspects of innate right, in a
way that matches the concept of choice operative within innate right.
They are, however, inconsistent with it because their actualization
does not treat the parties involved as innately equal. Because in the
first phase everyone has an innate right that everyone else is obligated
to abstain from coercing, the participants in a regime composed

46
Ibid., [6:230]: “[I]n this reciprocal relation of choice no account at all is taken of the
matter of choice, that is, of the end that each has in mind with the object he wants; it is not
asked, for example, whether someone who buys goods for me for his own commercial use
will gain by the transaction or not. All that is in question is the form in the relation of choice
on the part of both, insofar as choice is regarded merely as free . . .”
276 corrective justice

exclusively of innate right have innate equality, which Kant defines as

“independence from being bound by others to more than one can in
turn bind them.”47 This equality does not obtain in the actualization
of the external rights allowed by the postulate of practical reason.
Unlike the innate right, external rights are acquired through the per-
formance of an act. All such rights presuppose the possibility of origin-
al acquisition, when the proprietor comes to own something not
owned by anyone. However, the act of original acquisition is the exer-
cise of a unilateral will that puts others under an obligation that they
would not otherwise have. Despite the fact that innate right author-
izes innate equality, the proprietor, by virtue of his or her unilateral
act of acquisition, binds others with respect to the acquired thing
without being reciprocally bound to them. This should be beyond the
rightful power of one person’s unilateral will, for it is inconsistent with
innate equality of all that the acquirer should be able to subordinate
others to his or her purposes.48
This inequality has further consequences. Because an acquired
right, like all Kantian rights, carries with it the power to coerce others
not to violate it, the unilateral act of acquisition that creates the right
also gives the right-holder coercive power. Accordingly, although the
universal principle of Right forbids an act that does not coexist with
the freedom of another, the coercion occasioned by acquisition is pre-
cisely such an act, in that it allows the unilateral will to serve as a coer-
cive law for everyone.49 Moreover, although others also can create
rights through their unilateral will, no one can be sure of how others
interpret the extent of their respective rights or of whether they are
willing to abide by the rights of others. In the interplay of unilateral
wills that results, each person is the judge of his or her own entitle-
ments, doing what seems right and good in his or her own eyes.50
Thus, the actualization of the rights made possible by the postulate
of practical reason creates a conceptual tension, to be resolved in the
subsequent phase, between the unilateralism of the proprietor’s con-
duct and the equality authorized by innate right.The universal princi-
ple of Right, in which innate right is analytically contained, forbids
one person’s coercing the freedom of another. Yet the postulate of
practical reason allows one person coercively to restrict another’s
47
Kant, above n. 2, [6:237].
48
Ibid., [6:264]; cf. [6:259].
49
Ibid., [6:256].
50
Ibid., [6:312].
 poverty and property in kant’s system of rights 277

freedom through unilateral acts that establish proprietary rights to

exclude. Because of this tension, such rights are provisional pending
an additional move that brings them back into conformity with the
equality of innate right. Accordingly, the postulate of practical reason
with respect to rights allows us provisionally to hold the notion of
external property in place until the thought of it can be completed in
a further phase that establishes the conditions under which external
property is conclusively rightful.51
This transformation of provisional rights into conclusive ones
occurs in the third phase of Kant’s account of property. Kant intro-
duces a further postulate, the postulate of public Right, which marks
the transition from the state of nature to the civil condition of law-
governed society. The postulate declares that “[w]hen you cannot
avoid living side by side with all others, you ought to leave the state of
nature and proceed with them into a rightful condition.”52 In this
rightful condition the state provides duly authorized institutions of
adjudication and enforcement. These replace the exercise of private
judgment about controversial claims with the authoritative judgments
of courts that determine the scope of each person’s entitlements
according to what is laid down as right.53 Moreover, the coercion that
secures each person’s rights is no longer private but emanates from a
public lawful regime under which rights are secured by adequate
power external to the contending parties.54
Entry into the civil condition transforms the significance of origin-
al acquisition in two related ways.55 First, as the assertion of a rightful
power to place others under an obligation, acquisition is the preroga-
tive not of any person in particular but of all persons. This means that
everyone not only can acquire but is also under an obligation not to
interfere with what others have acquired. However, in the absence of
reciprocal assurance that others will refrain from interfering with one’s
own acquisitions, adherence to one’s obligations to others runs the
risk of subordinating oneself to others’ purposes.This would be incon-
sistent with the equality of innate right. The civil condition provides

51
Ibid., [6:257] (possession “in anticipation of and preparation for the civil condition . . . is
provisionally rightful possession”), [6:267] (property arises “before the establishment of a civil
condition but with a view to it, that is, provisionally”).
52
Ibid.
53
Ibid., [6:297], [6:313].
54
Ibid.
55
Ibid., [6:256–57].
278 corrective justice

this assurance through state institutions that can compel all to adhere
to the obligations that property creates. This assurance reflects not the
particular will of any person but a general will of all persons omnilat-
erally related to one another through the state in which they all par-
ticipate. All can now both enjoy their property and respect the
property of others on terms of equality.
Second (and more fundamentally), the lack of assurance in the state
of nature reflects the doubtful basis of any particular obligation to
refrain from using what has been originally acquired by another. In
the state of nature this obligation is unilaterally imposed by another’s
act of acquisition, thereby trenching on the freedom of the person
obligated.This difficulty applies no matter how many acquisitions take
place by how many persons. Each acquisition can be considered only
on its own, and therefore each faces the same difficulty, that it unilat-
erally restricts the freedom of others. The plurality of instances makes
no difference to the normative infirmity of every such acquisition. In
contrast, the civil condition involves a notional union of all wills that
transforms the external acquisition of unowned things from a merely
unilateral act on the part of the acquirer to an omnilateral act, to
which everyone as possible owners of property implicitly consents
and whose rights-creating significance everyone acknowledges. Acqui-
sitions are no longer a series of isolated acts. Instead, all acquisitions
are mutually related through a system of property in which all are
reciprocally bound and publicly coerced to respect the property rights
of others.
This second point can be formulated in terms of the equality of
innate right. As already noted, that equality consists in “independence
from being bound by others more than one can in turn bind them.”56
In the state of nature, where every act of original acquisition is con-
sidered on its own, this equality does not obtain; the acquirer binds
others to abstain from the acquired object without being in turn
bound by them. Of course, a second person may bind the first acquir-
er by acquiring a different object, but the difficulty would be repli-
cated, with the first acquirer now in the converse position of being
bound without being able to bind. And so on with further acquisi-
tions by others. It is true that across all the acts of original acquisition
each acquirer binds with respect to what she acquires and is bound
with respect to what others acquire. However, the state of nature

56
Ibid., [6:237].
 poverty and property in kant’s system of rights 279

supplies no conceptual basis for relating all the acquisitions to one

another. They figure in the state of nature simply as a plurality of dis-
crete acquisitions, every one of which is inconsistent with innate
equality. The civil condition transforms this plurality into a totality, in
which the various acquisitions instantiate a single system of property
that applies to all acts of acquisition by all acquirers over time. Within
this totality the equality of innate right is satisfied because all the
acquisitions taken together form a unity within which different
acquirers both bind and are bound.57
Kant affirms that leaving the state of nature and entering the civil
condition is “objectively necessary, that is, necessary as a duty.”58 This
duty follows from the recognition in the second of the three phases
that private property is rightful. The postulate of practical reason with
respect to rights allows the conclusion that having external things as
one’s own is an exercise of freedom that, although not analytically
contained within the universal principle of Right, is formally consist-
ent with it, and therefore capable of putting others under a duty of non-
interference. The duty is only provisional, because in the second phase
of Kant’s account this formal consistency with the universal principle
of Right is not accompanied by conditions that reflect the innate
equality entailed by that principle. The absence of these conditions
does not negate the validity of the duty. Cancelling the second phase
and returning to the first one is therefore not an option. While this
regressive move would restore the original innate equality, it would
nullify the possibility of external freedom that the postulate of practi-
cal reason identified with having something external as one’s own,
thereby leading to infringements of the duty provisionally established
through that possibility. Hence arises the necessity to move forward to
the third phase, in which the impediments to the conclusive operation
of the duty are eliminated. Just as all are under a duty at the second
phase to respect proprietary entitlements, so all are under a duty to
create the conditions under which those proprietary entitlements are
fully rightful. The establishment of institutions of public right
removes—indeed, is the only way to remove—the unilaterality of
judgment, coercion, and acquisition that characterizes the state of
nature. Accordingly, the duty of non-interference with property that

57
“Thus allness or totality is just plurality considered as unity.” Kant, Critique of Pure Reason,
tr. Norman Kemp Smith (1929), B111.
58
Kant, above n.2., [6:264].
280 corrective justice

makes its appearance in the second phase of Kant’s account matures at

the transition to the third phase into a duty to leave the state of nature
and to enter (and force others to enter) the civil condition.
For Kant, the civil condition is formed through a social contract—
what Kant calls the original contract—that unites the will of all under
rightful coercive laws.59 This contract is not a historical event but is
rather the idea of reason in terms of which the legitimacy of the state
can be conceived. Under this contract every individual gives up the
lawless freedom of the state of nature in order to find his or her free-
dom undiminished under law in the rightful condition formed by the
union of all wills. Because the state must live up to the undiminished
freedom of the original contract that legitimates the state’s formation,
the idea of this contract serves as a norm for the internal constitution
of any state.60 Kant regards the public duty to support the poor as a
juridical effect of the move from the state of nature through the orig-
inal contract into the civil condition. As the passage in the first section
of this chapter indicated, he does not explain why he posits such an
effect or why he regards support of the poor as a public duty.61 How-
ever, he presumably thinks that a state that fails to fulfill this duty
would not accord with the idea of the original contract. More specifi-
cally, without such a duty the transition to the civil condition would
not leave freedom undiminished from its starting point in innate
right.

3. Freedom undiminished
Kant’s description of the innate right to freedom indicates why this
might be so. Kant unequivocally connects innate right to what is
needed for self-preservation. Referring to the aspect of innate right
that consists in occupying space, Kant writes: “Through the innate
possession of land I can exclude everyone from using that which is
necessary for sustaining my existence.”62 In accordance with their
innate right, possession by humans of the land on which they find
themselves “is the supreme condition of the possibility of using this
land, as long as this use is absolutely necessary only for purposes of
59
Above n. 2, [6:315–16].
60
Ibid., [6:313].
61
Above n. 9.
62
Vorarbeiten, in Kant, Gesammelte Schriften, XXIII, 286.
 poverty and property in kant’s system of rights 281

sustaining their existence.”63 By excluding others from whatever the

person is physically connected to, such as the very space that the per-
son occupies, innate right allows the use, unaffected by others, of what
is immediately necessary for one’s survival. The occupation of space is
immediately necessary for one’s survival, because, as an embodied
creature, a human being cannot exist without being somewhere. Of
course, one can physically possess items that do not contribute to one’s
subsistence, as when I examine the beauty of a flower that I hold
between my fingers. But nothing can be used for my subsistence with-
out being physically connected to me.
By linking innate right with the sustaining of one’s existence, Kant
is not claiming that one has a positive right to the means of existence
and that others are under a correlative duty to assist in my continued
existence. Such a duty would make their actions subservient to my
needs, contrary to our innate equality. If my continued existence
depends on my consumption of the only available apple, my neigh-
bor’s pulling it off the tree before I do may mean that I die from star-
vation. However, my neighbor has no duty to abstain from seizing the
apple so long as he can do so without physically interfering with me.
My starvation would have to be imputed not to his permissible act of
seizing the apple or to refusing to give it to me, but to my consequent
inability to gain access to the apple without disturbing my neighbor’s
body.
Kant’s argument is rather that the innate right signifies that others
have a negative duty not to interfere with the incidents of that right,
and that one’s existence is therefore immune from deprivation at the
hands of others. This is because one’s mere existence is not a wrong to
anyone (because wrong is imputable only to actions) and accordingly
provides no rightful ground for others to interfere with one’s body
and with whatever one physically possesses. The body, of course, is the
organ through which one maintains one’s physical existence. Similarly,
the objects most immediately implicated in one’s continuing survival
are things that come into contact with the body (for example, the
earth on which one lies, the apple in one’s hand). Accordingly, although
innate right does not include a positive right to survival, it imposes a
duty of non-interference on others.
A regime characterized only by innate right systemically maximizes
the mutual non-dependence of all. In such a regime, everything is

63
Ibid., 318.
282 corrective justice

available for use by everyone, except the space that others occupy and
whatever is in their physical possession. Accumulation is impossible
because no one has external things as one’s own. As long as I occupy a
particular space, nobody can push me out of it. But when I move,
I can occupy any other space not occupied by someone else, gaining
my new space and simultaneously losing the power to prevent others
from occupying the old one. Similarly, as long as the apple is in my
hand, no one can take it from me. But I cannot store it while gather-
ing others, for as soon as the apple is released from my grasp it becomes
available to anyone else who can take it. This general availability of
everything except the space that others occupy and the things that
they physically possess means that my survival cannot directly be
affected by the actions of others.Whatever external things are available
to my neighbors are also available to me. So far as my relationship
with others is concerned, I am (in the terminology of Kant’s author-
izations of innate right) my own master,64 able to act on my own and
without dependence on others for my continued existence.
The non-dependence of one’s existence on others is the hallmark
of juridical relationships restricted to the exercise of innate rights.
Innate right protects each person’s existence on the basis of the innate
equality of all, since everyone was reciprocally bound not to interfere
with the bodily integrity of everyone else. Because one’s body is the
limit of one’s rights, everything not immediately connected to the
body of someone else is available to everyone. The result is that no
one’s survival is dependent on anyone else’s actions.
In this account, the significance of one’s survival is entirely rela-
tional. The gross corporeality of the body and its continued existence
do not demand attention on the ground that life, taken on its own,
forms a basic value in abstraction from one’s interaction with others.65
The standpoint of right is concerned only with the relationship
between one person’s freedom and another’s action. Accordingly, the
body has a juridical significance because as the organ through which
human beings exercise their freedom, it imposes duties on others
whose actions must be capable of coexisting with everyone’s freedom.
Similarly, when innate right is considered as a regime of equality, the
body as the organ of one’s freedom cannot rightfully become the

64
Kant, above n. 2, [6:238].
65
For an example of life treated as a basic value, see John Finnis, Natural Law and Natural
Rights (1980), 86.
 poverty and property in kant’s system of rights 283

means through which that freedom is compromised through subordin-

ation to or dependence on others. Thus, from the standpoint of right,
continued existence matters not because of its unilateral importance
to the person whose existence it is, but because of its role within a
relationship of free, equal, and mutually non-dependent beings.
The inevitable non-dependence that characterizes innate right dis-
appears with the introduction, first provisional and then conclusive, of
external property. Because ownership obtains even in the absence of a
physical connection between the owner and what is owned, the accu-
mulation of external things is now possible. My range of rightful
action is now confined to what might be left over from others’ efforts
at accumulation. The possibility of amassing land makes it conceivable
that, given the finitude of the earth’s surface,66 all the land may be
appropriated by others, leaving me literally with no place to exist
except by leave of someone else. My continued existence may thus
become dependent on the goodwill or sufferance of others, to whom
I might then have to subordinate myself, making myself into a means
for their ends, perhaps becoming their bondsman or slave.67 Moreover,
my inability otherwise to satisfy my basic needs may make me depend-
ent on the generosity of others—that is, on that to which I have no
right. The legitimation of the ownership of external things produces a
juridical regime in which the survival of one person may be depend-
ent on how others dispose of what is rightfully theirs.
This transformation of one’s position relative to others from assured
non-dependence to potential dependence means that in making
acquired rights conclusive, a civil condition entrenches the possibility
of dependence. Compared to relations characterized solely by innate
right, this would, inconsistently with the original contract, diminish
freedom in the relations among persons.
It is worth noting that the criteria for determining whether free-
dom is diminished arise from the conceptual implications of innate
right itself, not from a calculus about the effect of property on collect-
ive or individual welfare. It may well be, as Locke famously observed,
that instituting external property produces an enormous surplus com-
pared to what preceded,68 and that the judgment of whether this

66
Kant, above n. 2, [6:262], [6:311].
67
Locke has a similar apprehension: “But we know that God hath not left one man so to
the mercy of another, that he may starve him if he please,” John Locke, Two Treatises of Govern-
ment, First Treatise, ed. P. Laslett (1965), s. 41.
68
Ibid., s. 37.
284 corrective justice

would be an improvement or a worsening of any individual’s situation

depends on how that individual factors his or her degree of risk aver-
sion into the calculation of the chances of capturing part of the sur-
plus. For Kant, however, the issue is more constrained. Non-dependence
with respect to one’s continued existence is characteristic of innate
right, because such non-dependence is sheltered by the duty not to
interfere with another’s physical integrity and by the consequent avail-
ability of everything to everyone. The standards for judging whether
freedom remains undiminished in a civil condition that protects prop-
erty are provided by the normative ideas authorized by innate right:
not probability and welfare, but innate equality and non-dependence.
Whether freedom is diminished raises not a quantitative issue (“Will I
have less?”), but a relational one (“Will my existence change from not
being to possibly being dependent on others’ actions?”). In dealing
with the undiminished freedom of the original contract, the focus remains
where the universal principle of Right put it from the start: on the
relationship between the freedom of one person and the action of
another.
Thus, the progression from innate right to the state’s guarantee of
all property holdings seems to reach an impasse. On the one hand, this
progression is a normative necessity in which I am obligated to par-
ticipate. On the other hand, innate right at least has the advantage that
no person or aggregate of persons can engross the world’s resources,
shut me off from access to what is necessary for my existence, and
thereby make the exercise of my freedom dependent on the benefi-
cent or exploitative will of another. Entering a civil condition that, by
guaranteeing property, also guarantees the circumstances of possible
dependence diminishes my freedom inconsistently with the original
contract. The civil condition, it seems, is both the fulfilment of my
duty as a free person and the diminishing of my freedom.
The public duty to support the poor breaks this impasse. The
requirement allows freedom to be undiminished in the civil condi-
tion. The sovereign’s assumption of the duty to support the poor
makes up for the possible inaccessibility of the means of sustenance.
The result is that in the civil condition, just as under innate right, no
one’s subsistence is dependent on the actions of others. Everyone can
now participate in the civil condition as one’s own master.
Furthermore, the duty is incumbent on the people (and derivatively
on the sovereign) rather than on any particular person. The institution
of a regime of external property allows for the accumulation of
 poverty and property in kant’s system of rights 285

property; no individual commits a compensable wrong simply by

engaging in this process. Moreover, the prospect of impoverishment is
created by the systemic legitimacy of acquisition, rather than by the
appropriative acts of any particular acquirer. The systemic difficulty that
property poses for innate right is resolved by the collective duty imposed
on the people to provide subsistence as needed.
To be sure, individuals pay the tax. This, however, is not because
they are duty-bound as individuals to support the poor but because
the sovereign is authorized to tax them for a necessary state purpose.
The obligation of the taxpayers is to the state, not to the poor directly,
because the taxpayers whose property is secured by the state are the
beneficiaries of the transition to a civil condition. The incidence of
this tax is based on a notion of reciprocity that flows from the state’s
guarantee of property and, with it, of the proprietors’ means of sur-
vival: because the wealthy “owe their existence to an act of submitting
to the [commonwealth’s] protection, which they need in order to
live,”69 they are obligated to contribute what is theirs to sustain the
existence of those who, because of the property regime, now lack
what they need in order to live.
Kant remarks that the sovereign has the right to tax the people “for
its own preservation,” explaining that “[t]he general will of the people
has united itself into a society that is to maintain itself perpetually”
and that taxation of the wealthy is therefore authorized “for reasons of
state.”70 One might be tempted to think that in these references to the
people’s preservation, society’s self-perpetuation, and reasons of state
Kant is pointing to the right of the sovereign to act prudently in order
to prevent anarchic social unrest.71 This is not, however, the most plau-
sible interpretation. Kant presents support for the poor not as a merely
permissible exercise of prudence but as a requirement of duty, which
the sovereign takes over from the people.
Crucial to the understanding of this passage is the meaning of the
assertion that a society (regarded by Kant as the entity made by the
civil union)72 “is to maintain itself perpetually.” Perpetuity for Kant
does not involve a prediction of the empirical likelihood of social
tranquility or upheaval; perpetuity is rather a normative attribute of
the constitutional order. Something is “called perpetual . . . if it is bound
69
Above, n. 2, [6:326].
70
Ibid.
71
Above n. 16.
72
Above n. 2, [6:307].
286 corrective justice

up with the constitution of the state itself (for a state must be regarded
as perpetual).”73 A form of government is perpetual by virtue of its
being institutionally so structured that it can continuously and for the
indefinite future function in accordance with the norms that animate
it. Perpetuation is thus the temporal dimension of the state as a totality—
that is, as a juridical entity that unites a multitude of human beings
and of norms into a systematic whole operative over time. A rightful
condition strives to be perpetual in this sense. By making freedom its
principle, a rightful condition has a constitution “in which law itself
rules and depends on no particular person.”74 It accordingly has no
natural lifespan but rather endures through an unending process of
juridical self-animation.Through its institutions, such a state “preserves
itself in accordance with the laws of freedom.”75 This, Kant claims, “is
the only constitution of a state that lasts.”76
Society “is to maintain itself perpetually” because the state is under
a duty to have a constitution that accords most completely with free-
dom. As the temporal dimension of the state, perpetuity signifies the
continuous process of making the state most fully conform over time
to the original contract—that is, to the idea of reason “in terms of
which alone we can think of the legitimacy of a state.”77 All institu-
tions of the state are obligated to participate in this process of bring-
ing the state to the fullest possible realization of its own legitimating
grounds. In so doing, these institutions act for the state’s well-being,
which for Kant refers not to the happiness or welfare of the citizens
but to the promotion of the state’s fullest conformity to the principles
of right.78 Thus, fulfilling the obligation to make the state “suited to
the idea of the original contract” leads to the state’s having “the only
constitution . . . that lasts.”79
Accordingly, the terms Kant uses in describing the state’s duty to
support the poor—the preservation of the people, the perpetuation of
society, and taxation for reasons of state—fit together as follows. The
people requires a state for its own preservation not merely because the
state provides protection to its members, but because without a state

73
Ibid., [6:367].
74
Ibid.
75
Ibid., [6:318].
76
Ibid., [6:367].
77
Ibid., [6:315].
78
Ibid., [6:318].
79
Ibid., [6:340–41].
 poverty and property in kant’s system of rights 287

the people simply does not exist as a distinct collective and juridical
entity. The preservation of the people, therefore, entails the perpetu-
ation of the state. Perpetuation in turn involves having the state, through
its institutions, live up to its obligation to maintain and perfect its own
normative character. In acting to that end, these institutions act for
reasons of state, in the sense that they act for the state’s well-being as a
rightful condition that strives fully to conform itself to the grounds of
its own legitimacy. By tying the state’s duty to support the poor (and
its consequent authorization to tax) to the perpetual maintenance of
society, Kant treats that duty as a necessary aspect of the state’s obliga-
tion to maintain its character as condition that accords with the prin-
ciples of right. The state’s support of the poor is, in other words, a
constitutional essential.80
This does not mean that, for Kant, the poor have a right to subsist-
ence. Since a right for Kant is always accompanied by the authoriza-
tion to coerce and the state is the ultimate repository of legitimate
coercive power, Kant can recognize no right against the state. The
poor are supported not because they hold a right but because they are
the beneficiaries of a duty. The sovereign takes over from the people
the duty to support the poor that is an incident of the obligation to
make the state conform to the original contract. 81
The operation of this duty re-establishes the non-dependence that
marked innate right and was threatened by the introduction of private
property. In one’s relations with another, everyone continues to have
the same right to bodily integrity that they had as a matter of innate
right. The availability to everyone of everything that was distinct from

80
On the idea of constitutional essentials, see Rawls, above n. 23, 227–30. Rawls also
considers “a social minimum providing for the basic needs of all citizens” to be a constitu-
tional essential; ibid., 228.
81
Cf. Mulholland, above n. 20, at 317, 395. Mulholland’s very brief but instructive treat-
ment of the problem of the needy notes the connection between innate right and Kant’s
passage about supporting the poor. However, he concludes from this, unnecessarily, that Kant
recognizes a right to welfare. For this LeBar, above n. 16, at 247–48, correctly criticizes him,
demonstrating that no right to welfare can emerge from innate right. LeBar in turn con-
cludes that the defect in Mulholland’s argument means that Kant’s remarks can be under-
stood as referring only to political prudence. Mulholland’s conclusion is understandable in
the sense that in a modern polity the duty could be juridically recognized and enforced only
if it was constitutionally expressed through the explicit or implicit positing of a correlative
right. For examples of the operation of such a right, see Government of the Republic of
South Africa v. Grootboom [2001] (1) SA 46 (Const. Ct. So. Africa); Social Minimum Case,
Judgment of the Federal Constitutional Court of Germany, 9 February 2010 (English sum-
mary at <http://www.bundesverfassungsgericht.de/en/press/bvg10-005en.html>).
288 corrective justice

others’ bodies has been superseded by the public duty to support the
poor. Non-dependence with respect to one’s continued existence is as
well served by the juridical order of the third phase as it was by the
juridical order of the first phase. The danger of being reduced to a
means for others, present in the second phase, has been eliminated by
the public duty to the poor.
That the requirement of sustaining the poor is a duty of the sover-
eign rather than a right of the needy against any particular person
bears on two possible objections to the interpretation I am proposing.
The first objection is that the public duty to support the poor does
not, after all, succeed in reconciling property with innate right. Does
it not merely replace possible dependency on the actions of others
with an equally unsatisfactory dependency on the state? In answer,
one may note that Kant’s discussion of citizenship expressly differenti-
ates between being dependent on others for one’s existence and being
dependent on the state.82 Kant apparently does not consider the rela-
tionship of the poor to the state to involve real dependence. In his
view, one may surmise, dependence involves a relationship with some-
one who, without breaching a duty, can withhold a benefit necessary
for one’s survival. This is the case when someone owes his survival to
the choice of another, because in exercising such a choice no one is
under a duty of beneficence as a matter of right. The state is different.
The poor receive support from the state because it is owed to them as
members of the commonwealth. Because the state is under a duty, it
has no discretion to withhold the support; and having no private
interest of its own,83 it also has no motivation to withhold support.84
The receipt of state support thus does not make the needy subservient
to the will of others.
The second objection is that the duty of support recognizes need,
despite Kant’s earlier indication85 that need is irrelevant to the concept
of right. One should, however, consider the context and significance
of the earlier reference to the irrelevance of need. In expounding the
“concept of Right, insofar as it is related to an obligation corresponding
82
Kant, above n. 2, at [6:314].
83
Ibid., at [6:324] Kant argues that the state cannot have domains of its own.
84
In these respects, the duty of the sovereign stands in contrast to voluntary contributions
for the support of the poor, which Kant, giving the instance of lotteries, criticizes as exploita-
tive.; ibid., at [6:326]. Conversely, Kant also says that begging “is closely akin to robbery”
(ibid., at [6:326]), perhaps because the beggar exercises a kind of emotional coercion on the
sympathies of the donor to cause the donor to surrender what the beggar has no right to.
85
Ibid., at [6:230].
 poverty and property in kant’s system of rights 289

to it,” Kant remarks that “it does not signify the relationship of one’s
choice to the mere wish (hence also to the mere need of the other) as
in actions of beneficence or callousness, but only a relation to the
other’s choice.”86 Kant is affirming that, when the interaction between
persons is examined from a juridical viewpoint, one cannot ascribe a
right—or a duty corresponding to that right—on the basis of the
needs of one of the parties. Kant here is pointing to a conceptual fea-
ture of rights and their correlative duties when one person acts upon
another.87 The duty to support the poor, in contrast, deals not with
action by one person that is inconsistent with the rights of another,
but with the relationship between the individual and the state. To be
sure, the state is under a duty defined in terms of the individual’s needs;
that duty, however, does not arise through its correlativity with the
individual’s right. The state is under this duty not because the indi-
vidual has a right to subsistence, but because a rightful property-pro-
tecting regime, which there is a duty to create, is legitimate only when
the accumulation of property does not render the poor dependent on
the actions of others. The public duty to support the poor is, accord-
ingly, not the response of one person to the need of another, but the
response of the commonwealth to the possible dependency that is
incompatible with the original contract.

4. “Original possession in common”

If, as I have suggested, Kant connected the notion of subsistence under
a property regime to the availability of everything to everyone before
the emergence of property, he was following a well-marked path.
Property theorists in the centuries preceding Kant typically treated
the legal issues surrounding subsistence in terms of a notional residue
from primordial use rights that had been transformed into property
rights. Kant’s account of property includes vocabulary and ideas inher-
ited from the older tradition of natural right. Kant, however, reconfig-
ured these to make them consistent with his distinctive philosophy. In
this section I want to note the originality of Kant’s duty to the poor

86
Ibid.
87
“The concept of Right . . . has to do . . . with the external and indeed practical relation-
ship of one person to another, insofar as their acts, as facts, can have direct or indirect influ-
ence on each other”; ibid., [6:230].
290 corrective justice

by discussing its relationship to the ideas presented by his great pre-

decessors in the natural right tradition, Grotius and Pufendorf. The
contrasting views of these two influential figures crystallized the issues
that dominated subsequent legal theory. Because Kant was undoubt-
edly aware of their treatments of property and drew upon them, the
comparison also reinforces from a historical perspective the argument
that I have so far presented on interpretive grounds.
Like Kant, Grotius posits a connection between a use right in the
state of nature and the satisfaction of basic needs in a developed prop-
erty regime. According to Grotius, humans originally had a common
right over things, in consequence of which “each man could at once
take whatever he wished for his own needs, and could consume what-
ever is capable of being consumed. . . . Whatever each had thus taken
up another could not take from him except by an unjust act.”88 Each
person’s position was similar to that of the theatre-goer, who has a
right to the seat in which he is sitting even though the theatre is a
public place.89 With the loss of innocence and the growth of greed
and refinement, this primitive ownership in common, as well as the
right to immediate use and consumption that it involved, was super-
seded by a social contract that allowed possession to be replaced by
property.90 One of the implicit terms of this contract was that a per-
son in direst need could, under a claim of necessity, revive the right of
the user that was part of the original common ownership.91
Kant’s account, although differing in its details, is similar in its struc-
ture. Kant has the person start off with a right to occupy space with
one’s body (like Grotius’s example of the seat in the theatre) and to
satisfy one’s needs through a right to use and consume with which
others could not justly interfere. This general right to use and con-
sume does not survive the institution of private property. Nonetheless,
for both Kant and Grotius, the original ease of access to resources for
purposes of survival means that the law that emerges from the social
contract must continue to allow for survival. This is the basis both for
Grotius’s right of necessity and for Kant’s duty to support the poor.
Within this shared structure, however, Kant recasts Grotius’s argu-
ment in two significant ways. The first is that Kant explicitly rejects

88
Hugo Grotius, De Jure Belli ac Pacis Libri Tres, 2 vols., tr. Francis W. Kelsey (1925), vol. 2,
186.
89
Ibid.
90
Ibid., 188–90.
91
Ibid., 193.
 poverty and property in kant’s system of rights 291

the notion of a primitive community of common ownership that he

assumes underlies Grotius’s account.92 For Kant the challenge is not to
derive private ownership from common ownership but to understand
the basis of ownership as such. Kant criticizes the notion of a primi-
tive community on the grounds that such a community would itself
have arisen from contract and thus, since the possession it yielded
would be derived from the antecedent possession of the contracting
parties, would cast no light on original acquisition. Moreover, the
legitimacy that arises from primitive possession would be merely con-
tingent on the historicity of the contract that instituted it, rather than
grounded in principle.93
However, although Kant discards Grotius’s conception of a primi-
tive community of common property, he signals that he retains the
structure of Grotius’s argument by replacing it with the “original
possession in common.”94 This original possession in common is
merely the innate right understood in terms of the space that a per-
son occupies. The possession is in common not because the earth is
owned by all, but because the space one occupies can be anywhere
and changes with one’s change of location. Accordingly the earth on
which I now stand and from which, by virtue of my innate right,
I cannot be removed becomes part of your innate right when I move
away and you take my place.95 The commonality of this kind of pos-
session thus refers to the access we all have to the earth due to our
freedom to be wherever someone else is not. By using the term
“original possession in common,” Kant reconfigures the inherited
notion while retaining the inherited terminology. The rightful
response to dire need in the civil condition then emerges from the
original common possession as surely as its analogue in Grotius
emerges from primitive community.

92
Kant, above n. 2, at [6:251], [6:258]. Grotius in fact posited not a primitive community
of common ownership, but merely a right to use. This is made clear in Grotius, Commentary
on the Law of Prize and Booty, ed. van Itttersum (2006), 315, a work that was not rediscovered
until decades after Kant’s death. Kant’s knowledge of Grotius’s view of property was filtered
through Pufendorf and Achenwall, who ascribed the notion of common ownership to him.
See B. Sharon Byrd and Joachim Hruschka, Kant’s Doctrine of Right: A Commentary (2010),
123–26.
93
Kant, above n. 2, at [6:251], [6:258].
94
Ibid., at [6:262].
95
“Each person has an innate right to this place or that place on the earth, i.e., each per-
son is in potential but merely disjunctive general possession of all places on the surface of the
earth.” Above n. 62, at 320.
292 corrective justice

The second way in which Kant recasts Grotius’s argument is by

shifting the mechanism for alleviating the dire need that a property
regime might occasion.Whereas Grotius focuses on the claim of neces-
sity, Kant posits a public duty of welfare. The claim of necessity would
be completely misplaced within Kant’s argument about the poor.
Necessity relates two individuals through the resource that one seizes
from another. However, the need that motivates this seizure is the con-
sequence not of any particular person’s owning the resource, but of the
system of property ownership as a whole. Hence Kant posits a duty on
the state as the guarantor of the entire property regime. Moreover, as
Kant’s classic discussion of necessity indicates,96 a claim of necessity
involves the existence of a wrong, albeit one that is unpunishable by a
court of law. But if the problem presented by innate right was that the
non-dependence of free persons must survive into the civil condition,
necessity would not be the solution. For it would be inconsistent with
freedom to ensure subsistence by the prospect of committing a wrong.
The equality of persons under innate right does not give anyone an
immunity to commit a wrong against another. Moreover, innate right
postulates the blamelessness of a person who commits no wrongful act,
not the culpability of someone who is unpunishable because of the
wrongful act’s necessitous circumstances. In contrast, for purposes of
Kant’s argument the state’s duty to support the poor succeeds where
necessity fails, for the duty operates not through the commission of a
wrong but as an implication of the original contract.97
Grotius was not alone in connecting the satisfaction of basic needs
under a regime of private property with the original right to immedi-
ate use. The same connection reappears in Pufendorf, despite his
extensive criticism of other aspects of Grotius’s account.
Pufendorf rejected Grotius’s conception of an original community
of common ownership. He contrasted what he termed Grotius’s “pos-
itive community” of common ownership in which all things are
owned by everyone, with the “negative community” in which “all
things lay open to all men, and belonged no more to one than to
another.”98 Pufendorf then traced the development of ownership

96
Kant, above n. 2, [6:235–36].
97
This is not to say that in the civil condition necessity is unavailable as a response to
urgent need, but only that necessity does not deal with the problem that I argue is resolved
by the duty to support the poor.
98
Samuel Pufendorf, De Jure Naturae et Gentium Libri Octo, 2 vols., tr. C. H. and W. A.
Oldfather (1934), vol. 2, 537.
 poverty and property in kant’s system of rights 293

through a series of social pacts, the first of which was to the effect that
“whatever one of these things, which were left open to all, and its
fruits, a man had laid his hands upon, with intent to turn it to his uses,
could not be taken from him by another.”99
Kant’s conception of the earliest stage in the development of prop-
erty is more or less identical to Pufendorf ’s. Both Kant’s original pos-
session in common and Pufendorf ’s negative community of all things
imply the same criticism of Grotius, that the task is not to derive pri-
vate from common ownership but to trace the development of own-
ership from non-ownership. To this end both Kant and Pufendorf
situate persons within a regime of rights and their correlative negative
duties. From an early stage—Kant through the innate right and Pufen-
dorf through the initial social pact—both regard interference with
another’s use as an injustice. Their main difference is that whereas
Pufendorf treats the negative community of all things as a historical
stage, Kant declares that “[o]riginal possession in common is, rather, a
practical rational concept which contains a priori the principle in
accordance with which alone men can use a place on the earth in
accordance with the principles of Right.”100
To deal with the problem of subsistence within a fully developed
property regime, Pufendorf indicates that the magistrate, as well as
other persons of means, has a duty to relieve extreme need.101 This
duty is embedded within Pufendorf ’s account of necessity. For Pufen-
dorf, assistance to someone in extreme need is an imperfect duty, cor-
relative to the needy person’s imperfect right to the assistance. The
reason for regarding the duty as imperfect is that, because “it is per-
formed upon a kind of voluntary impulse arising from a man’s good
nature,”102 the person who would benefit from the performance of
the duty has neither the power to compel him to perform nor the
right to be compensated for non-performance.103 Instances of the
breach of imperfect obligations are ingratitude and the failure to give
benefactions, as Pufendorf thinks it unjust to neglect occasions for
beneficence and not to return the favor for benefactions received.104

99
Ibid., 535, see also 537 (“all things should lie open for all, for the promiscuous use of
every man”).
100
Kant, above n. 2, [6:362].
101
Pufendorf, above n. 98, 305.
102
Ibid., 315.
103
Samuel Pufendorf, Elementorum Jurisprudentiae Universalis, tr. W. A. Oldfather (1931), 66.
104
Ibid., 60.
294 corrective justice

In the case of aiding the needy, “[t]he reason why only an imperfect
right is allowed, especially to such things as are owed on the grounds
of humanity, is that thereby a man finds the opportunity to show that
his mind is intent upon voluntarily doing his duty, and at the same
time possesses the means to bind others to him by his kindness.”105
But “when necessity does not admit of any other means to secure his
safety,”106 Pufendorf allows the needy person to exercise self-help.
“Must some poor fellow die of hunger because he cannot overcome
by his prayers the inhumanity of some man of wealth?”107 The result is
that, although aiding the poor is an imperfect obligation, “the urge of
supreme necessity makes it possible for such things to be claimed on
the same ground as those which are owed by a perfect right.”108 The
wealthy person’s breach of duty in failing to be beneficent to the
needy, when combined with the urgency of the needy person’s situ-
ation, gives the needy person the right to take for himself what he
ought voluntarily to have been given. Necessity, as it were, perfects an
otherwise imperfect right.
From Kant’s perspective, this account of necessity is rife with con-
fusion. First, the obligation to assist the poor is based on consider-
ations of beneficence, which are ethical rather than juridical. Moreover,
the duty imposed on the magistrate is not a public one derived from
the political nature of the magistrate’s power, but is merely a reflection
of the magistrate’s opportunity for beneficence. Furthermore, when
necessity takes hold, the benefactor’s imperfect duty gets matched to a
corresponding right that can be treated as perfect, thus destroying the
correlativity of duty and right. Finally, necessity is regarded as trigger-
ing the operation of a right, rather than excusing the commission of
an unpunishable wrong.
For our purposes the basis of the right to subsistence that underlies
Pufendorf ’s account of necessity is of particular interest. For Pufend-
orf, property emerges from the original availability of everything to
everyone through a series of social pacts. These pacts allow for a var-
iety of property arrangements, “provided that they involve no contra-
diction and do not overturn society.”109 But, Pufendorf notes, all
property regimes acknowledge that the necessity of another provides

105
Pufendorf, above n. 98, 305.
106
Pufendorf, above n. 103, 59.
107
Pufendorf, above n. 98, 305.
108
Ibid.
109
Ibid., 537.
 poverty and property in kant’s system of rights 295

an exception to the proprietor’s rights.110 Pufendorf sees necessity as

reviving the right of everyone to everything that characterized the
original negative community in which property rights did not exist.
This is because Pufendorf is of the opinion that a condition to that
effect is implicit in social pacts establishing property: “it is understood
that whenever any man in the division of things renounced his right
to such things as are assigned others he did so with this reservation:
Provided he cannot otherwise secure his own safety.”111 Thus, a per-
son’s desperate need does not create a new right in the property of
others, but triggers the operation of the condition under which every-
one’s right to everything was surrendered, that property would not
endanger survival. Presumably the reason for Pufendorf ’s insistence
on reading this condition into the social contract is that “individuals
would not consent to a system of rights that might require that they
starve.”112
On my suggested reconstruction of Kant’s thought, Kant reconfig-
ures the idea that individuals would not contract into the possibility
of starvation, by relating that possibility to a dependence on the actions
of others. Kant’s conception of right eliminates the ethical elements
that vitiated Pufendorf ’s account of necessity and adds a public duty
to the poor. The requirement of undiminished freedom in the origin-
al contract and the consequent need to reconstruct in the civil condi-
tion the lack of dependence on others that was present in innate
right enable Kant conceptually to connect the duty to support the
poor with each person’s innate right. This connection is consistent
with the natural right tradition represented by Grotius and Pufendorf,
who each in their different ways traced the legal categories for bene-
fiting the poor to the rights that everyone had before the institution
of private property. Kant’s attention to what he termed “original pos-
session in common” indicates his continuity with this tradition. Kant
did not explicitly link innate right in the first phase of his account to
subsistence for the poor in the third phase. But it may well be that this
connection was so well known from the writings of his predecessors
in the natural rights tradition that he regarded it as obvious.

110
Ibid., 538.
111
Ibid.
112
Thomas A. Horne, Property Rights and Poverty: Political Argument in Britain, 1605–1834
(1990), 36. Although Kant lies outside the scope of Horne’s treatment, Horne’s book in its
entirety casts valuable light on the contemporary understanding of the connection between
property and the relief of poverty.
296 corrective justice

Drawing on and arguably perfecting this tradition, Kant purported

to show how instituting private property and taxing to support the
poor were jointly necessary if society was to be a legitimate union of
self-determining agents. His account treats property and taxation to
support the poor as distinct but interconnected. On the one hand
private property and taxation figure in different kinds of juridical
relationship. Property relates one person to another through the cor-
relativity of right and duty; the duty to support the poor relates the
taxpayers to the state and the state to the poor. The consequence of
this distinction is that considerations of poverty have no effect on the
definition and application of property rights. On the other hand,
property and the public duty to support the poor are connected
through a single sequenced argument that extends the reach of the
universal principle of Right while preserving consistency with the
ideas of innate equality and non-dependence that this principle
implies. For Kant, taxation is not theft, and neither is property. On the
contrary, taxation and property are jointly necessary for a civil condi-
tion legitimized by the idea of the original contract. On Kant’s view as I have
reconstructed it, the public duty to support the poor is latent within
private property as a rightful institution. Formulating Kant’s point in
terms of the Aristotelian forms of justice, one may say that the dis-
tributive justice of taxing to support the poor is the conceptual con-
comitant of establishing a legally effective system of corrective justice.
 9
Can Law Survive Legal
Education?

1. Three activities
Legal education exists at the confluence of three activities: the prac-
tice of law, the enterprise of understanding that practice, and the study
of law’s possible understandings within the context of a university.The
first of these, the practice of law, consists of the activities consciously
governed by law, including, for example, lawyers giving legal advice,
citizens contemplating the legality of prospective actions, legislators
creating law within the limits of their jurisdiction, and judges deter-
mining the rights and duties of litigants. It thus comprehends the
entire field of legal institutions, legal doctrine, and legal interaction.
The second activity, the enterprise of understanding law, refers to the
elucidation of the character of this practice. This enterprise seeks to
determine the extent to which the practice’s various characteristics
can be grasped as exhibiting, through the coherence of their interrela-
tionships, some generically determinate character. The third activity,
university study, requires that the student’s reflections about law be
appropriate to an institution devoted to caring for the intellectual
inheritance—the stock of ideas, images, beliefs, skills, and modes of
thinking—that composes the world’s civilization.1
These three activities exercise a reciprocal effect on one another.
On the one hand, the practice of law supplies the materials that are to
be understood through university study. On the other hand, that
practice is transformed by the very enterprise of articulating under-
standings of it. Scholars are not merely the passive recipients of the

1
Michael Oakeshott, “The Study of ‘Politics’ in a University: An Essay in Appropriateness,”
in Rationalism in Politics and Other Essays (1962; new and expanded ed., 1991), 184, 187–94.
298 corrective justice

law’s materials. Rather, their understandings influence the practice by

making practitioners conscious of the possibilities that are implicit in
it.2 When these understandings originate in the universities and are
thus invested with the authority of prestigious institutions of learning,
the practice of law itself can become either (at best) more aware of
law’s distinct voice in the conversation of civilized humanity or
(worse) more prone to succumb to prevailing academic orthodoxies.
The central challenge that has faced legal education since it was
wrested from the legal profession and lodged in the universities3 has
been how to integrate the three activities. The relation between the
practice and the university study of law has proved particularly prob-
lematic. One influential critique of legal education laments the grow-
ing disjunction between them:

The schools should be training ethical practitioners and producing scholar-

ship that judges, legislators, and practitioners can use . . . But many law
schools—especially the so-called ‘elite’ ones—have abandoned their proper
place, by emphasizing abstract theory at the expense of practical scholar-
ship and pedagogy . . . [I]f law schools continue to stray from their principal
mission of professional scholarship and training, the disjunction between
legal education and the legal profession will grow and society will be the
worse for it.4

2
The classic statement of this is Friedrich Carl von Savigny’s comments on the Roman
jurists:
[T]he action of the jurists, appears at first sight a dependent one, receiving its materials from with-
out. However, by their giving to the materials so presented a scientific form which strives to dis-
close and perfect the unity dwelling in them, there arises a new organic life which shapes and
reacts upon the materials themselves, so that from science as such, a new sort of generation of law
incessantly proceeds.
Friedrich Carl von Savigny, System of the Modern Roman Law, tr. William Holloway (1867),
37–38.
3
In Canada this happened relatively recently. The decisive event was the defection of
three of Canada’s leading law professors (Cecil A.Wright, Bora Laskin, and John Willis) from
the law school operated by the Law Society of Upper Canada and their establishing the
modern Faculty of Law at the University of Toronto in 1949.Within a decade, the Law Soci-
ety of Upper Canada surrendered control of legal education to the universities by recogniz-
ing that graduation from a university faculty of law qualified the graduate to enter the
profession without penalty. For a description of this “revolution” in Canadian legal educa-
tion, see C. Ian Kyer and Jerome E. Bickenbach, The Fiercest Debate: Cecil A.Wright, the Bench-
ers, and Legal Education in Ontario 1923–1957 (1987). For a recent discussion, see R. C. B. Risk,
“My Continuing Legal Education,” (2005) 55 U. Toronto L.J. 313.
4
Harry T. Edwards, “The Growing Disjunction between Legal Education and the Legal
Profession,” (1992) 91 Mich. L. Rev. 34, 41. Judge Edwards’s views are extensively discussed
in Symposium, “Legal Education,” (1993) 91 Mich. L. Rev. 1921.
 can law survive legal education? 299

Formulated in these terms, the critique forcefully indicates what its

author thinks is at stake. The practice of law and its university study as
currently constituted are regarded as competitors, such that the uni-
versity’s preoccupation with “theory” operates “at the expense of ”
practical professional concerns. The proper function of the university
study of law, according to this critique, is to produce scholarly work
for the professional consumption of those engaged in the practice of
law. The diagnosis, in effect, is that the practice of law is effaced
through university study, and the remedy suggested is that the latter be
recalled to its “principal mission” of being useful to the former.
This criticism has, I think, a truth that should be recognized, though
my version of this truth is perhaps not what its author intended. A
disjunction between the practice of law and its university study would
indeed be disquieting. This is not, however, because the disjunction
would be a disservice to the legal profession (though it might be), but
rather because it would be a disservice to the university itself.
The university exists as a locus for the study of law not for the sake
of the legal profession, but because law is a component of the intel-
lectual inheritance of civilization. The “principal mission” of universi-
ty study is to care for and develop this inheritance. That the legal
profession should benefit from this through the university’s graduates
and its ideas is all to the good. Moreover, it is both desirable and nec-
essary that those who are most intimately connected with and con-
scious of the workings of law should support its study within the
university—thus manifesting a commitment to the idea that law is
integral to civilized modes of thinking and living. But criticism of the
university study of law should come from a standpoint internal to
university activity itself. Accordingly, the disjunction between legal
education and the legal profession is troubling only if it represents a
failure on the university’s own terms.
The disjunction would be such a failure in the following sense.
University study of any kind must have a definite object; it must be
the study of something. Law is a phenomenon that exists only through
a set of legal doctrines, institutions, and juristic activities. The univer-
sity study of law can therefore be nothing other than the study of the
practice of law. Accordingly, legal education is inevitably concerned
with the activity of “judges, legislators, and practitioners,” not in order
to produce scholarship that they “can use” (though, if they can legit-
imately use it, so much the better), but in order to reflect upon the
meaning and intelligibility of their activity. A disjunction between the
300 corrective justice

practice of law and the university study of law is troubling because it

suggests that the university study of law actually has no object, that it
is the study of nothing, similar to the zoological study of unicorns.
Such study would be a failing from the university’s standpoint, quite
apart from its uselessness to the legal profession.
But what does it mean to say that the university study of law is dis-
jointed from the practice of law? The answer lies in how these two
activities conceptualize the character of law. If the university study of
law expressly or implicitly attributes to law a different character than
that which is presupposed in the practice of law, then one cannot say
that the practice of law is the object of university study. Under those
conditions, the practice of law and the university study of law would
be activities lacking a common interest; the law that the latter studied
would not be the same as the law that the former practiced.
Thus, the difference between the two activities of practice and uni-
versity study has to be mediated through the third activity, that of
understanding the law. For only when that understanding is common
to the law as practiced and as studied is there no disjunction between
legal education and the practice of law.
So formulated, the issue raised by the supposed disjunction between
the legal profession and legal education turns out to be primarily one
of legal theory rather than one of straightforward sociological obser-
vation. Of course, what is discussed in a university differs from what is
discussed in a law office or a judges’ conference. What might link the
two is a conception of how law is to be understood. Those participat-
ing in university life as students, teachers, and scholars regard law as a
significant component of civilization’s intellectual inheritance and
attempt to think through the features implicit in the practice of law
that make that practice worthy of academic attention. The process of
identifying these features and thinking them through requires refer-
ence, implicitly or explicitly, to some understanding of what the prac-
tice of law is. This is an exercise in legal theory, because legal theory
consists of nothing but a self-conscious examination of the range of
possible understandings of law. And so the critic who blames the dis-
junction on too much “abstract theory” necessarily, if ironically, issues
an appeal for further theorizing.
In this chapter, I wish to present more concretely this abstractly
formulated notion of disjunction. My focus is on the way that this
disjunction arises in the university study of private law. A justification
for this focus is that private law, as the enduring bedrock of legal
 can law survive legal education? 301

education, is a primary vehicle for the transmission of conceptions of

legal understanding. More importantly, the simplicity and the restrict-
ed scope of the relationship between the parties allow the disjunction
and its implications to be set out with particular clarity. Section 2 of
this chapter suggests that prevailing instrumentalist approaches within
the legal academy, exemplified by (but not confined to) certain ver-
sions of the economic analysis of law, systemically distort legal prac-
tice. This distortion effaces the characteristic concepts of private law,
ignores the direct relationship between the parties, and assimilates pri-
vate law into public law. In these respects, economic analysis fails to
comprehend private law as the distinctive kind of normative phenom-
enon that it is.
My purpose in making these observations is not to criticize eco-
nomic analysis in particular, but to point to a structural problem that
accompanies an assumption—that law is to be explained instrumentally—
that is widely popular in the academic treatment of law and that yet
separates the university study of law from law’s practice. Economic
analysis thus merely provides the paradigmatic example of an instru-
mentalism that emerges from a distinctly academic enterprise but that
mischaracterizes the legal practice it purports to explain. In section 3,
I will sketch a different mode of legal understanding that both respects
legal practice and affirms private law as a component of our intellec-
tual inheritance that is worth studying in its own terms. Finally, in
section 4, I will trace the implications of this mode of understanding
for the interdisciplinary turn that is a conspicuous feature of contem-
porary legal education.
My goal in this chapter is a modest one. It is easy to read critiques
of present legal education as exhortations to exclude, either through
curricular change or appointment policy, certain kinds of currently
entrenched enquiry.5 My argument here, however, is not about what
to exclude but what to include. By exploring the supposed disjunc-
tion between practice and university study, and by suggesting how
to overcome that disjunction, I want to point to a conception of the
core of legal education, at least for private law, that links the three
activities. This in no way denies the insights of other ways of thinking
about law. Inasmuch as they are about law, however, those insights

5
This is, for instance, the way Sandy Levinson reads Judge Edwards’s critique; Sanford
Levinson, “Judge Edwards’ Indictment of ‘Impractical’ Scholars: The Need for a Bill of Par-
ticulars,” (1993) 91 Mich. L. Rev. 2010.
302 corrective justice

presuppose—and therefore are ancillary to—an understanding of law

that is not disjointed from the practice of law.Thus, my focus is on what
legal education should necessarily deal with, whatever else it deals with.

2. Disjunction: the role of instrumentalist analysis

A. The example of economic analysis
To see the sort of disjunction that I have in mind, consider the notion,
popular among expositors of the economic analysis of law, that eco-
nomic efficiency is the key to understanding tort doctrine. The basic
assumption of this approach is that a defendant should be liable for
failing to guard against an accident only when the cost of precautions
is less than the probable cost of the accident. From the economic
standpoint, the goal of the liability rules of private law is to provide
incentives for cost-justified precautions. Ambitious claims have been
made on behalf of this mode of analysis: economic ideas have been
said to reveal the inner nature,6 implicit design,7 and unifying per-
spective8 for tort law.
This approach constitutes a notable attempt to link the university
study of law to the practice of law. On the one hand it draws on the
insights of economics, the academic discipline that provides a system-
atic understanding of what Hegel called “the infinitely complex criss-
cross movements of reciprocal production and exchange.”9 On the
other hand, it deploys this discipline to explain leading doctrines in
the practice of tort law. The vast academic literature that this attention
to economic efficiency has inspired is one of the most impressive
achievements of contemporary legal scholarship.
One would have thought that an approach that purports to reveal
the inner nature of tort law would be particularly illuminating about
the concepts that pertain to tort law. Negligence liability, for instance,
involves a conjunction of legal concepts, such as duty, proximate cause,
factual cause, and the standard of reasonable care. Such concepts are
fundamental to our understanding of tort liability because they struc-
ture the thinking of those who participate in the practice. Through

6
Richard A. Posner, The Problems of Jurisprudence (1990), 361.
7
William M. Landes and Richard A. Posner, The Economic Structure of Tort Law (1987), vii.
8
Richard A. Posner, Tort Law: Cases and Economic Analysis (1982), 2.
9
G. W. F. Hegel, The Philosophy of Right, tr. T. M. Knox (1952), s. 201.
 can law survive legal education? 303

such concepts, tort law discloses its own normative character, thereby
indicating the terms in which it is to be understood. Revealing the
inner nature of such concepts would (one would expect) disclose how
they function or should function within the reasoning of those
engaged in legal practice. Among the issues that would then be
addressed are: What are the conditions that call each of these concepts
into play? How are they related to each other and do they form a
coherent set? What is the relationship between the abstract formula-
tions of these concepts and the institutional processes of adjudication
that particularize them for specific cases? And are these concepts suit-
able vehicles for the normative considerations that justify or can jus-
tify the determination of liability? Attention to these issues would
involve taking the concepts seriously as objects worth explicating in
their own terms, with a view to examining whether they have or
could have the significance that tort law ascribes to them when it ori-
ents legal practice, as manifested in the reasoning of lawyers and judg-
es, along their lines.
In fact, economic analysis does the opposite. When economic ana-
lysis is presented as the key to understanding tort law, the point of the
analysis is not to take the fundamental concepts seriously as concepts
used in legal practice, but to render them otiose. Economic analysis
has its own stock of ideas that operate without reference to the legal
concepts. The result is that ideas about economic efficiency replace
rather than illuminate the legal concepts. Instead of functioning as
vehicles of thought, the legal concepts are at most labels pinned to
conclusions once economic analysis has done all the work.
Consider two instances, causation and intention. Causation plays a
central role in determinations of liability as a matter of legal practice.
For the economic analysis of tort law, however, causation turns out to
be an idea “that can largely be dispensed with.”10 Given that the pur-
pose of tort law is thought to be the promotion of efficiency, the
defendant will be held liable—and thus deemed to be the cause of an
injury—when such liability will promote the efficient allocation of
resources to safety. Thus, cause does not mark the law’s concern for
the transitivity of the relationship between the defendant’s conduct
and the plaintiff ’s injury, but functions merely as the label that is
attached to the conclusion of a cost–benefit analysis. Because both
parties might have taken precautions, the task for economic analysis is

10
Landes and Posner, above n. 7, at 229.
304 corrective justice

to determine not whether the defendant caused the plaintiff ’s injury

in the conventional legal sense, but which of them could have avoided
the accident more cheaply.
Similarly dispensable is the concept of intention.11 For the eco-
nomic analyst, intention refers not (as it does in the law itself ) to the
actor’s purpose with reference to a wrongful consequence, but to the
connection between the probability of harm and the ease with which
the actor could have avoided it. What “establishes a clear-cut econom-
ic basis for condemning a distinct form of misconduct” is not the
wrongfulness of making another’s injury the object of one’s conduct
but instead the injurer’s low cost of avoidance relative to the social
benefits of the injurer’s activity.12
The economic analysis, in other words, produces a disjunction
between the significance of tort concepts for legal practice and their
significance for academic study. While purporting to offer an account
of legal practice—indeed, while claiming to reveal its inner nature—
the economic approach instead effaces the very concepts that consti-
tute legal reasoning when determining liability within that practice.
In presenting its analysis of concepts like causation and intent, the
economic analyst aims not to illuminate those ideas in their own
terms, but to make them disappear in the face of the analytic power of
economic efficiency. Economic analysis thereby offers a theory that
negates rather than explains the concepts supposedly being analyzed.
The deficiency of this form of scholarship lies not in its presenting
nothing about legal practice that “judges, legislators, and practitioners
can use,”13 but in its presenting nothing about legal practice at all.
There is a second respect in which economic analysis does not
reflect legal practice. Through the process of litigation, the practice of
law directly links the particular plaintiff to the particular defendant.
Liability is thus a relational phenomenon in which the court responds
to the wrong or injustice that the defendant has done to the plaintiff.
This linkage assumes that the same reasons for liability apply simul-
taneously to both parties. In contrast, economic analysis does not treat
the parties as directly connected. Rather, it views them each as subject
to different incentives that somehow happen to be conjoined in a find-
ing of liability. For economic analysis the point of liability is to induce

11
Ibid.
12
Ibid., at 153.
13
Edwards, above n. 4, at 34.
 can law survive legal education? 305

the parties to take cost-justified precautions. These incentives, however,

apply separately to each of them. Awarding damages against a defend-
ant provides defendants with an incentive to act efficiently, “[b]ut that
the damages are paid to the plaintiff is, from the economic standpoint, a
detail.”14 The plaintiff ’s receipt of the damage award reflects a different
group of incentives (such as the need to induce enforcement of the
norm and to prevent prospective victims from pre-empting the pre-
cautions incumbent on actors)15 that do not in themselves entail taking
the money from the actual defendant. Both parties are thereby involved
in the damage award, but for separate reasons. Efficiency might as easily
be served by two different funds, one that receives tort fines from inef-
ficient actors, another that disburses the indicated inducements to vic-
tims. Instead of linking each party to the other, economic analysis
ascribes the presence of both to a combination of incentives independ-
ently applicable to each. Accordingly, liability is the consequence of
one-sided considerations that somehow come together, rather than of
relational considerations that treat the parties as belonging together
because of what the defendant has done to the plaintiff.
This sundering of the parties’ relationship leads economic analysis
to mischaracterize private law in a third way. The fundamental con-
cepts that express the unity of the parties’ relationship make private
law a distinctive mode of legal ordering, with its own discourse, its
own internal organization, and its own normative presuppositions.
Within the legal domain, the distinctiveness of private law allows it to
be contrasted to public law. Private law normatively connects the par-
ties directly to each other, not to the state. Although the state is present
through the machinery of adjudication, the purpose of this machinery
is merely to give authoritative expression to what the relationship
between the parties requires. In contrast, public law is concerned with
the forms and limits of the state’s exercise of power with respect to
those who are subject to it. Whereas private law deals with the rela-
tionships between participants in the community, public law deals
with the relationships between participants and the community as
embodied in its official organs.
By denying the significance of fundamental concepts of private law
and negating the unity of the defendant–plaintiff relationship, eco-
nomic analysis divests private law of the possibility of constituting a

14
Richard A. Posner, Economic Analysis of Law, 2nd ed. (1977), 143 (emphasis in original).
15
Richard A. Posner, Economic Analysis of Law, 6th ed. (2003), 192.
306 corrective justice

distinctive mode of legal ordering. From the economic standpoint,

private law is to be understood as a judicially created and enforced
regime for the taxation and regulation of inefficient activity.16 Courts
act as administrative tribunals that set norms for efficient behavior and
exact fines when those norms are breached. The plaintiff ’s function in
initiating a lawsuit is not to secure redress for wrongful injury but to
claim a bounty for prosecuting inefficient economic activity. Eco-
nomic analysis thus submerges the private nature of tort law in a pub-
lic law of economic regulation.
Thus, the link that economic analysis posits between academic
study and the practice of private law is vitiated by its mistaken charac-
terization of that practice. Instead of illuminating private law, eco-
nomic analysis discards its fundamental concepts, breaks apart its
relationships, and subverts its private nature.The economic analysts are
not so much concerned with understanding private law as with assess-
ing the degree to which its rules coincide with what efficiency
demands. Far from being the focus of their attention, private law is
merely the foreign language into which economic discourse has
somehow been translated.17 The result is a profound disjunction
between the economic analysis of law as a method of university study
and the practice that is being studied.18

B. The dynamic of instrumentalism

My point in making these comments is not to criticize economic
analysis in particular. Rather, in contemporary legal education, eco-
nomic analysis is paradigmatic of the instrumentalist structure of
academic enquiry. What occasions the disjunction with legal prac-
tice is this instrumentalist structure, not economic analysis as such.

16
Richard A. Posner, “A Theory of Negligence,” (1972) 1 J. Legal Stud. 29, 48–49, 51.
17
For law as the translation of economic principle, see Landes and Posner, above n. 7,
at 23; Posner, above n. 6, at 361.
18
Economic analysis may lodge itself within the practice through the influence of eco-
nomic scholarship on judges, who then apply it in their judgments. Compare the observa-
tions of von Savigny, above n. 2. To the extent that this occurs, the disjunction between
academic study and legal practice is lessened. However, in its stead a different and ultimately
more serious problem arises. Because economic analysis cannot coherently reflect the char-
acter of the law, its entry into legal practice sets up irresolvable tensions between the law’s
fundamental concepts and relational structure, on the one hand, and the economic analysis
on the other.Thus, the disjunction between academic study and legal practice is displaced by
a disjunction internal to the legal practice, between the economic analysis and the practice’s
concepts and structure.
 can law survive legal education? 307

Economic analysis is nothing but an instance of a more comprehen-

sive dynamic.
The instrumentalism of economic analysis consists in the interpre-
tation of tort law as forwarding the goal of economic efficiency. As
the disjunction just described indicates, the normative attractiveness
of this goal—what makes it worthy of being considered a goal that
tort law should forward—does not arise out of the law itself, by
reflection, for instance on the fundamental concepts of tort law or on
the nature of the relationship between the parties. Rather, this goal is
thought to be desirable independently of tort law and is then given
to tort law from the outside. Tort law is only an instrument in the
goal’s promotion.
Economic efficiency is merely one of the goals that modern schol-
arship has proposed. These goals come in many varieties, ranging from
the general, such as promoting communal responsibility19 or basic
aspects of the good20 to the more specific, such as alleviating injury.21
All such goals base their appeal on some conception of human welfare
that is considered desirable independently of the law and that the law
should therefore strive to forward.
Regardless of the goal it advances, an instrumentalist analysis of pri-
vate law mischaracterizes its object in the same way that economic
analysis does. An instrumentalist approach makes three errors. First, it
imports outside goals for immanent concepts of private law. Second, it
ignores the relationship between a plaintiff and a defendant. Third, it
wrongly converts all private law into public law.
Instrumentalist approaches substitute for the concepts of private
law the outside notions that are appropriate for the promotion of the
preferred goal. Instead of working out the meaning of the applicable
legal concepts in particular situations, as legal practice requires, the
instrumentalist specifies the mechanisms through which the social
goal might be forwarded in different circumstances. Because the really
important work is done by the apparatus of instrumental reasoning,
the law’s invocation of the standard legal concepts is regarded as a

19
Robert A. Baruch Bush, “Between Two Worlds: The Shift from Individual to Group
Responsibility in the Law of Causation of Injury,” (1986) 33 U.C.L.A. L. Rev. 1473,
1480–502.
20
John Finnis, Natural Law and Natural Rights (1979), 59–75.
21
Marc A. Franklin, “Replacing the Negligence Lottery: Compensation and Selective
Reimbursement,” (1967) 53 Va. L. Rev. 774, 785–88.
308 corrective justice

mere ritual,22 a veil to be pierced by clear-headed analysis,23 or even as

a salutary obfuscation that itself has instrumental value.24
Second, within instrumentalist analysis the plaintiff and defendant
are not directly related to each other. The goals are considered ele-
ments of the social good, and therefore are concerned with the overall
benefit, however construed, to society as a whole, not with the rela-
tionship between two particular parties. Instead of linking the plaintiff
to the defendant who has done her an injustice, instrumentalist ana-
lysis groups each party with those who are, from the standpoint of the
goal in question, similarly situated. For example, the alleviation of
injury, when considered as a goal of tort law, connects the injured
party not to the particular person who has wrongfully caused the
injury, but to other injured persons who have a like claim on the dis-
tribution of society’s resources. Analysis in terms of a goal thus breaks
apart the relationship between the parties, in order to apply the appro-
priate goal to each of them. The result is that reading an independ-
ently desirable goal back into private law creates a dissonance between
the parties’ nexus as a matter of legal practice and the goal’s indiffer-
ence to this nexus within the instrumentalist understanding of law.
When university study accepts the instrumentalist understanding and
develops it, this dissonance appears as a disjunction between university
study and legal practice.
Third, for the instrumentalist, all law is public law.The favored goals
must be selected by the state and inscribed into a schedule of collect-
ively approved aims. The various method of elaborating the commu-
nity’s purposes—adjudication, legislation, administrative regulation,
and so on—are merely the species of the generically single activity of
making the goals a legal reality. The singling out of a particular goal
from among all the possible goals, the balancing of one goal against
competing goals, and the positing of the means for promoting the
chosen goals require legislation by political authority. Norms of pri-
vate law are therefore considered the product of legislative acts, even
when formulated through the adjudicative process.25 Instrumentalism
thereby dissolves the very idea of private law as a distinctive mode of

22
Jerome Frank, “What Courts Do In Fact,” (1931) 26 Ill. L. Rev. 645, 653.
23
Felix S. Cohen, “Transcendental Nonsense and the Functional Approach,” (1935) 35
Colum. L. Rev. 809, 809–12.
24
Guido Calabresi, “Concerning Cause and the Law of Torts: An Essay for Harry Kalven,
Jr.,” (1975) 43 U. Chi. L. Rev. 69, 107.
25
Oliver Wendell Holmes, Jr., “Privilege, Malice and Intent,” (1894) 8 Harv. L. Rev. 1, 3.
 can law survive legal education? 309

legal ordering. Private law turns out to be nothing but public law in
disguise.26
These three features of instrumentalist analysis are intimately con-
nected.The legal concepts (such as causation and intent) are the appar-
atus that the law has elaborated to treat the relationship between the
parties as a single normative unit. The process of determining a
defendant’s liability by working through these concepts is what stamps
private law as a distinctive kind of normative ordering. The concepts,
the relational unity, and distinctiveness of its form of legal ordering are
thus the mutually entailed aspects of private law as a legal practice.
Instrumental analysis distances itself from all of these when it distances
itself from any of them.
To the extent that contemporary legal education revolves around
instrumental understandings, it inevitably separates itself from private
law as a legal practice. Economic analysis is simply exemplary in this
respect. Those who, out of skepticism about or antagonism toward
economic efficiency as a goal, think that legal education should center
on different goals contribute to this disjunction no less than do the
economic analysts themselves. The disjunction is the consequence not
of one particular goal or set of goals rather than another, but of the
very orientation toward goals.
In the face of this disjunction between the instrumentalist under-
standing and the legal practice, two responses are tempting. Each of
these responses leaves the disjunction intact, while submerging one or
the other of the disjoined activities.
The first response is embodied in academic work that expressly dis-
connects the university study of law from legal practice. In private law
this work takes the form of “decoupling” the position of the plaintiff
from that of the defendant. One suggestion, for example, is that the
defendant should pay more and the plaintiff should receive less than
the compensatory amount, with the difference going to cover the
state’s administrative costs.27 Another example is the suggestion that
efficient incentives would be best achieved by arranging that contract
damages be awarded to a third party rather than to the victim of the
breach.28 Such decoupling embraces the disjunction by foregoing the

26
Leon Green, “Tort Law: Public Law in Disguise,” (1959) 38 Tex. L. Rev. 1.
27
A. Mitchell Polinsky and Yeon-Koo Che, “Decoupling Liability: Optimal Incentives for
Care and Litigation,” (1991) 22 RAND J. Econ. 562.
28
Robert D. Cooter and Ariel Porat, “Anti-Insurance,” (2002) 31 J. Legal Stud. 203, 204.
310 corrective justice

aspiration to see the university study of law as an endeavor to under-

stand the practice of law. In terms of instrumentalist scholarship, pro-
posals of this sort represent an advance over the more traditional
project of explaining the law. They are based on the recognition that
the relationship between the parties constricts the free play of instru-
mentalist reasoning. Once one unravels the parties’ relationship, the
limits of legal scholarship are set not by the law as an object of the
enquiry, but by the imagination, ingenuity, and brilliance—all amply
present—of the scholars themselves. This allows a more consistent
presentation of the kind of instrumentalism favored by the particular
scholar. But there is also a parallel disadvantage. Having severed the
link to legal practice, these proposals seem to be nothing more than
dreamy exercises in instrumentalist utopianism, far removed from the
hard-headed contact with the real world that instrumentalists like to
profess.
The second response goes in the opposite direction by emphasizing
the primacy of legal practice. This response is exemplified in the call,
mentioned earlier, for university study to adhere to its “principal mis-
sion” of professional training by producing scholarship that can be
used by legal practitioners.29 Offered in the name of overcoming the
disjunction between legal practice and university study, the suggestion
merely subordinates the latter to the former, raising the question of
why this “mission” would require university study at all rather than a
more direct system of professional training and apprenticeship. After
all, on this conception, what are law professors except legal practition-
ers with more leisure and lower salaries? By connecting the university
study of law with the demands of legal practice rather than with the
purposes of the university, the suggestion dismisses the significance of
any understanding of law that is not coterminous with legal practice
itself. In effect, the university study of law is regarded merely as a para-
site on the practice of law.
These two responses are the consequence of viewing law in instru-
mentalist terms. Instead of attempting to overcome the disjunction
between university study and legal practice that instrumentalism cre-
ates, the responses cut the Gordian knot by accentuating one element
and disregarding the other. The decoupling view has a strong notion
of the university study of law, which, however, turns out to be not
about law but about the possible artifacts of instrumental reasoning.

29
Edwards, above n. 4, at 41.
 can law survive legal education? 311

In contrast, the parasitic view is attentive to legal practice but, given

the open-endedness of instrumentalist analysis, sees little value in uni-
versity study beyond what can be used by legal practitioners. In their
separate ways each responds to the problem of disjunction by giving
up on it.

3. Overcoming the disjunction

A. The character of private law
In this section I want to sketch a possible solution to the problem of
disjunction. In the previous section I argued that, so far as private law
is concerned, the disjunction between the university study of law and
the practice of law is a consequence of the instrumentalist framework
that dominates contemporary legal studies. The instrumentalist frame-
work subjects private law to analyses that inevitably mischaracterize it.
Instrumentalist approaches efface the concepts of private law, fail to
connect the parties directly to each other, and assimilate private law to
public law. Accordingly, a solution to the problem of disjunction
involves rethinking the assumptions that create it.
Two mutually reinforcing moves are involved. Negatively, the
instrumentalist framework is to be rejected, since it is the infelicity
of this framework that generates the problem to begin with. Posi-
tively, the organizing concepts of private law, the direct relationship
between the parties and the distinctiveness of private law as a mode
of legal ordering must be understood as the indicia of the specific
character of private law. The primary task of the university study of
private law—what it should do, whatever else it does—is to enquire
into this character. By so doing, university study both maintains its
continuity with the legal practice, which is its starting point, and yet
goes beyond that practice to disclose its implicit structural and nor-
mative ideas.
Central to the elucidation of the character of private law is the
assumption that private law is (or at least aspires to be) a normatively
coherent practice that can accordingly yield a coherent understand-
ing. Under this assumption the organizing concepts of private law are
harmoniously connected both to one another and to the institutional
structure of private law litigation. This assumption is a necessary start-
ing point for several reasons. First, only on the assumption of the
coherence of legal practice can one make sense of the endeavor to
312 corrective justice

regard private law as a systematically intelligible body of knowledge

that can be amenable to university study. Otherwise, private law would
be assumed to be nothing but a piled aggregate of propositions that
together had no specifiable character. Second, coherence is an internal
value and aspiration of the private law itself, being integral to its rea-
soning and discourse. Thus, the assumption of coherence is not only a
methodological postulate of university study but also a pervasive
premise of legal practice and therefore itself part of the character of
the practice. Third, the practice of private law is a normative phenom-
enon, where the disposition of one person’s claim against another has
to be justified through the use of public reason. The process of justifi-
cation presupposes that the private law’s entire apparatus of justifica-
tion is internally coherent, for if elements of that apparatus pulled in
different directions, then it could not function to justify anything.
The character of private law provides the general framework for
understanding it as a coherent normative phenomenon. Specification
of this character arises by a process of scholarly reflection on, and gen-
eralization from, the law’s particular arrangements (its doctrines, its
institutional structure, its ensemble of concepts, its methods of reason-
ing, and so on). However, the character of private law is not composed
simply of the sum total of these arrangements. Some of these arrange-
ments may achieve that character only imperfectly or deficiently,
because they may not accord with the practice viewed as a coherent
whole. Whether the positive law’s treatment of a particular problem
bears out the assumption of coherence is a contingent matter. Where it
does not, the character of private law then provides the standpoint,
internal to private law as a whole, from which to criticize the particu-
lar legal arrangement. Character is thus an ideal construct that makes
the particulars of the practice of private law intelligible in the light of
the practice’s most general and pervasive features. Understood in terms
of its character, private law is “a unity of particularity and genericity.”30
The task of specifying the character of private law belongs to the
university study of law, but it is rooted in legal practice. Unless univer-
sity study takes legal practice seriously, the enquiry into its character
would, of course, be self-stultifying. No disjunction exists between the
character and the practice of private law. Even when some particular
arrangement of private law is thought to be deficient in the light of

30
Michael Oakeshott, On Human Conduct (1975), 5.
 can law survive legal education? 313

the law’s character, the criticism that emerges is consonant with the
law’s self-critical commitment to “work itself pure,”31 because it
expresses the law’s own striving for internal coherence. The character
of private law is implicit in its practice, but scholarly reflection brings
it into focus, defines and refines it, and presents legal doctrine as its
expression. Indeed, any treatment of private law, whether in the class-
room or in the academic literature, that focuses on the requirements
of coherence helps elucidate the character of private law.
The most ambitious efforts to specify the character of private law
are necessarily exercises in “abstract theory.”32 The theorist strives to
render what is implicit in the law as explicit as possible. To do this one
must have recourse not only to legal material but also to conceptual-
izations, to the philosophical literature, and to modes of discourse and
analysis that, while treating the practice of law seriously in its own
terms, are not themselves part of that practice. This exercise in the-
orizing has to be abstract, in that it abstracts from the brute particular-
ities of legal practice to the more general standpoint inherent in the
specification of the law’s character. One might even say that the more
abstract the better, because the goal is to formulate the most general
framework possible. However, throughout this project of specifying
the character of private law, continuity with legal practice is always
maintained, because otherwise the character specified would not be
the character of anything. In this context, abstract theory is the friend
of legal practice, not its competitor.

B. The stages of elucidation

The elucidation of the character of private law has three stages: attend-
ing to the legal practice, eliciting its inner structure, and enquiring
into that structure’s normative presuppositions.33
The first stage, attending to the legal practice, anchors the elucida-
tion of the law’s character in the practice’s features. As is evidenced by
the operations of those who engage in it with a mastery of its con-
cepts and procedures, the practice of private law is, at least to some
extent, an intelligible activity. Disclosure of the character of private

31
Lord Mansfield’s famous phrase in Omychund v. Barker [1744] 26 Eng. Rep. 15, 23
(K.B.).
32
Edwards, above n. 4, at 34.
33
See Oakeshott, above n. 30, at 1–31 (presenting Oakeshott’s suggestive account of the
engagement of understanding).
314 corrective justice

law is intended to render what is incipiently intelligible about private

law even more intelligible. This exercise requires that one take notice
of the justificatory and institutional apparatus of private law, so that its
features are the subject and the starting point for a more general
account.
Although the character of private law is an ideal construct, its eluci-
dation does not involve the imagining of a Utopia. The practice of
private law is taken as the given object of enquiry, and it is viewed as
it views itself: as a specific kind of normative order that governs human
interactions according to its own distinctive yet coherent conceptions
of fairness and rationality. Subject to confirmation at subsequent stages,
the arrangements of private law can provisionally be regarded as
the indicia, however indistinct, incomplete, or inadequately articulat-
ed, of the kind of normative order that they constitute. The task is not
to excogitate a new and perhaps even superior kind of normative
order, but to disclose the character of this one.
Of especial significance are the concepts of private law and the
institutional linkage of the parties as plaintiff and defendant. Because
elucidating the character of private law involves specifying the most
general framework for understanding it, particular attention should be
paid to aspects of private law that are already general. Every applica-
tion of private law presupposes a legal institution that directly links a
particular plaintiff and a particular defendant. Of course, one can have
a normative order in which this linkage is absent, for example where
injury is dealt with by payment from a state fund; whatever its merits,
such a normative order is not a form of private law. Also general, but
having a scope that is more local, are the concepts of private law. These
are the ideas through which the law requires us to organize our think-
ing about issues of liability. In specifying the character of private law,
we do not seek, as does the instrumentalist, to show that these con-
cepts are otiose, but to understand their role within a coherent con-
ception of liability. Do these concepts, and other determinants of
liability, sustain private law’s claim to an internally coherent rationality,
or must they be adjusted, abandoned, or supplemented for the sake of
that rationality?
To deal with these questions we must move to a second stage in the
elucidation of the character of private law. The second stage seeks to
bring out the inner structure of the arrangements of private law.
Because we are treating private law as a normative phenomenon, our
particular interest is in the structure of the considerations that justify
 can law survive legal education? 315

liability. The legal concepts and the other determinants of liability are
the vehicles for these considerations.The second stage inquires whether
these considerations have a uniform general shape. For if they do, that
shape would reveal the character of the law in which these considera-
tions are decisive.
Crucial to the disclosure of this structure is the institutional nexus
between plaintiff and defendant. Private law works through an adjudi-
cative mechanism by which the plaintiff sues the defendant and, if
successful, is given an award of damages or other relief that the defend-
ant must satisfy. As just noted, this direct linkage between plaintiff and
defendant is the most pervasive feature of private law. If private law is
to be understood as a normatively coherent practice, the justification
for liability in any particular case has to reflect the structure of this
linkage. The institutional framework for the litigation attests to the
fact that the point of liability is to remedy an injustice between the
particular parties. Accordingly, the reason for considering the defend-
ant to have done an injustice to the plaintiff can be coherent only if it
evinces the same direct link as is present in the institutional frame-
work. Justification within private law is thus the expression of a bipo-
lar normativeness that directly links the particular parties within this
institutional framework. The structure of this bipolar normativeness is
that of correlativity, in which the same injustice predicates both the
doing and the suffering.
As the most general description of the structure of the parties’
interaction, correlativity marks the character of private law as a dis-
tinctive normative order. No justification that does not participate in
this character can find a coherent place within private law. Correlativ-
ity accordingly excludes considerations, no matter how appealing, that
focus unilaterally on one or the other of the parties (for example, the
depth of the defendant’s pocket or the plaintiff ’s insurability against
injury). Such one-sidedness was the defect, noted in the previous sec-
tion, of instrumentalist approaches, which break apart the relationship
between the parties by invoking social goals that operate on one or
the other of them and on persons who are similarly situated.
To the extent that they form a coherent unity, the legal concepts
relevant to any particular basis of liability also partake of this correla-
tivity. Such concepts are the markers of a framework of normative
reasoning that operates relationally to connect two particular parties
as the doer and the sufferer of an injustice. The role of the concepts is
to entrench the correlativity of the parties’ situation into the reasoning
316 corrective justice

and discourse of private law. The doing and the suffering of the same
injustice is a single normative sequence that preserves its unity while
moving from one party to the other. The legal concepts pertaining to
each type of injustice are the devices through which legal practice
presents and integrates the moments of that sequence.
In negligence law, for instance, the legal inquiry is broken down
into a complex yet unified set of concepts (duty, breach, proximate
cause, cause in fact). When liability for negligence is being considered,
the unreasonableness of risk created by the defendant is seen in terms
of the probability and the gravity of its effect on others; the duty not
to create the risk is seen in terms of its foreseeable effect on a group
that includes the plaintiff; the definition of the risk through proximate
cause is seen in terms of the kind of effect that leads us to think of the
risk as unreasonable; and the factual causation of injury seen in terms
of the materialization of this risk. Thus, the concepts that constitute
the negligence enquiry trace the sequence that begins with the
defendant’s unreasonable exposure of others to risk and is completed
by the materialization of that risk in injury to the plaintiff. The two
termini of this sequence are linked by the concepts of duty and prox-
imate cause, which keep the plaintiff and the plaintiff ’s injury, respect-
ively, within the risk by reason of which the defendant’s action is
negligent, thereby ensuring that the risk that materialized in the plain-
tiff ’s injury is the same as the risk that the defendant unreasonably
created. Each of the concepts thus refers both to the defendant’s
wrongful act and to that act’s wrongful effects, potential or realized,
on the plaintiff. Together the negligence concepts form an integrated
ensemble that articulates what it is for the same injustice to be done
and suffered when unreasonably created risk matures into injury.34
This treatment of legal concepts can be readily contrasted with that
of economic analysis. Recall the examples of causation and intention
mentioned in section 2. Under the economic approach, factual caus-
ation is largely dispensable, to be replaced by a cost–benefit analysis. In
contrast, when understood as a feature of tort law’s character, factual
causation is simply what it purports to be: the concept that deals with
the materialization of risk into actual injury. A similar observation can
be made about intention. Instead of being twisted (as economic ana-
lysis suggests) into a reference to the connection between the high prob-
ability of harm and the ease of avoidance, intention is, again, just what it

34
Above chapter 2.
 can law survive legal education? 317

purports to be: the concept that makes the execution of the defendant’s
purpose the link between the plaintiff ’s injury and the defendant’s con-
duct. Both causation and intention are concepts that belong to private
law’s bipolar normativeness. Each of them has a single normative sig-
nificance for both parties, and each is an element in an integrated
sequence that directly connects what the defendant has wrongly done
to what the plaintiff has wrongly suffered. Whereas economic analysis,
having pulled the parties apart, is unable to take seriously the legal
concepts that normatively link them, the endeavor to specify the law’s
character allows these concepts to be understood in their own terms
and to play a coherent role in the determination of liability.
Understood in terms of the correlativity of the parties’ positions,
private law is a system of rights and of the duties correlative to them.
Such an understanding maintains continuity with the practice of law.
Both the understanding and the practice treat rights seriously, not as
superfluous proxies for instrumentalist considerations, but as genuinely
normative determinants of correlatively structured liability.
But, one might ask, what exactly are these rights, where do they
“come from,” how are they distinguished from aspects of human well-
being, and how is their normative character to be understood? Such
questions point to a third stage in the understanding of private law.
This stage builds on the other two, the attention to legal practice and
the specification of correlativity as the most general structuring idea
immanent in that practice. At the third stage one enquires into the
normative presuppositions of correlativity and of the notion of rights
that emerges from it.
In structuring the interaction between doer and sufferer, correlativ-
ity presupposes an abstract conception of the interacting parties.With-
in this conception the parties are viewed as exercising the capacity for
purposive action, whatever might be their particular purposes. In the
natural right tradition of legal philosophy, this conception of the
capacity for purposiveness without regard to particular purposes is
known as personality. Personality is to the actor what correlativity is
to the interaction: the most general normative conception immanent
in being a party to a private law relationship. Because personality is
the presupposition of correlativity, personality and correlativity are
complementary conceptions formulated from the perspective of the
actor and the interaction, respectively.35

35
Above chapter 1.
318 corrective justice

Personality illuminates the normative standpoint specific to private

law, including its mutually entailed conceptions of freedom and equal-
ity. In articulating the juridical implications of the parties’ purposive
capacity, private law treats the parties as endowed with self-determin-
ing freedom. The regime of rights and correlative duties is a system of
negative freedom, with one’s rights demarcating spheres of freedom
for the right-holder and limits on freedom for others. Within this
regime the parties are formal equals because, when they are conceived
as purposive beings without regard to their specific purposes, the
sources of substantive inequality, such as differentials in wealth or vir-
tue, become irrelevant.Thus the parties are equal, in that they are both
treated as the loci of self-determining freedom; and they are free, in
that neither is subordinated to the unilateral purposiveness of the
other. Seen in the light of this normative standpoint, the practice of
private law involves the continual elaboration and refinement of how
personality and its attendant notions of freedom and equality figure in
particular determinations of liability.
It might be objected that, given their abstractness and formality,
these conceptions of freedom and equality are not very robust or
appealing. However, this objection, even if it were true, is irrelevant.
The issue is not whether these conceptions are appealing but whether
they are immanent in private law. Recall that the three stages outlined
here have been presented in order to specify the character of legal
practice. The point is not to work out the most attractive conceptions
of freedom or equality as if we were constructing a Utopia from the
ground up, but to derive the most general framework for understand-
ing private law from the structure and presuppositions of its legal rela-
tionships. If the progression from the practice of private law to
correlativity and then to personality is sound, then the conceptions of
freedom and equality that emerge are those that are presupposed in
the practice.
The understanding of private law elucidated through these three
stages differs sharply from the instrumentalist conception discussed
earlier. The instrumentalist conception starts by specifying goals that
are desirable apart from private law and then examines private law in
their light. As a consequence, the instrumentalist misconstrues the
law’s concepts, cuts the direct relationship of the parties, and subsumes
private law under public law. Having started outside private law, the
instrumentalist does not succeed in re-entering it. The result is a dis-
junction between instrumentalist scholarship—including the kind of
 can law survive legal education? 319

legal education it inspires—and the legal practice that is its subject

matter. In the elucidation of the character of private law, no such dis-
junction appears. The process of elucidation starts from within private
law and then considers its legal concepts on their own terms. The
direct connection between plaintiff and defendant, far from being the
embarrassment that it is for the instrumentalist, is the manifestation of
the correlativity that structures the justifications for liability. In this
conception, private law is categorically different from public law. Pri-
vate law can therefore be understood as the juridical realization of a
bipolar normativity in which one purposive being is directly related
to another through a system of rights and their correlative duties.
When private law is understood in terms of its ideal character, no
disjunction exists between two of the three activities with which we
began, the practice of law and the enterprise of understanding that
practice. The two activities are continuous without being congruent.
They are continuous, because the understanding of law is elucidated
through reflection on the structure and presuppositions of the prac-
tice of private law. They are not congruent, however, because the
activity of understanding private law requires a theoretical effort that
works out, according to its own methods and idiom, the most general
conceptual framework immanent in the practice. The ingredients of
that conception—correlativity and personality—deal with legal prac-
tice in its own terms, but they are not themselves explicit in that prac-
tice. Rather, they are the theoretical constructs that illuminate the
character of private law.

C. Legal education
Turning now to the third of our initial three activities, one can see
why private law, understood in this way, is suitable for university study.
The purpose of university study is to care for the intellectual inherit-
ance of civilized life. Private law is a significant and distinctive part of
that inheritance. Private law is the ongoing attempt, actualized through
society’s legal institutions, to submit the direct interaction of one per-
son with another to a system of reason. It involves an immense collect-
ive intellectual effort carried out over centuries and in different
jurisdictions, featuring failures as well as successes, mistaken diversions
as well as majestic triumphs. Its distinctiveness as a normative ordering
lies in its correlative structuring of the parties’ relationship, which makes
the morality of private law categorically different from that of either
320 corrective justice

personal ethics or political action. Private law is thus the forum for a
special mode of thinking, which it is the function of university to
impart.
In the enterprise of legal education, the university study and the
professional training perform complementary functions, but each has
its own focus. Professional training produces familiarity with the
present operation of private law, developing skills based on particular
legal materials and suitable to particular demands. University study,
however, imparts (or should impart) a sense of the intelligibility of
private law as a whole. Its interest is not in particular legal materials
but in the mode of thinking that has produced them. Or, more accu-
rately, its interest is in particular legal materials not for the information
that they convey but for their exemplification of correlatively struc-
tured thinking, reasoning, and discourse.
As part of the university study of law, then, legal education is the
process by which students of the law are initiated into a world com-
posed of this correlatively structured mode of reasoning. Such reason-
ing is something students learn by engaging in it—that is, by being
exposed to and discussing paradigms of it and by being provided with
opportunities to develop their skills at it. This requires serious focus
on legal doctrine not merely as a collection of rules or as a checklist
for lawyers’ dealings with particular situations, but as the crystalliza-
tion of the distinctive mode of reasoning that directly links the defend-
ant’s conduct to the plaintiff ’s entitlement.
Several facets of this engagement with the law are particularly
important. One is attention to the interrelation of the organizing con-
cepts within a given basis of liability, and to the question of whether
these concepts, as presented by the positive law, form a coherent set—
that is, whether they adequately realize the legal relationship’s correla-
tive structure. Another is developing an appreciation of the casuistic
reasoning that gives those concepts specific meaning in the rich var-
iety of particular circumstances. A third is the exploration of the rela-
tionship between content and process, between the substantive
considerations that justify liability, and the adjudicative context in
which these considerations are assessed.
In this enterprise the theoretical constructs of correlativity and per-
sonality play a background role. They bring to the surface the charac-
ter of the practice, so one can be aware of the nature of coherent legal
discourse.The practice, however, proceeds on its own terms. Just as the
practice of law carries on without explicit reference to these constructs,
 can law survive legal education? 321

so students must learn to formulate arguments about liability without

invoking them.The general conceptual structure of private law cannot
serve as a substitute for considering the legal material itself. Correla-
tivity and personality reflect the character of private law reasoning in
general, so far as that reasoning is coherent, but it cannot generate
a complete legal code to be mechanically applied in particular cases.
All that the theoretical constructs can do—and it is enough that they
do so—is orient us toward the requirements of justificatory coher-
ence, and thus assist in eliminating considerations that are incompat-
ible with it.36
Because the ideas of correlativity and personality are general and
abstract, different systems of private law can manifest them in different
ways. Although correlativity and personality are the stable theoretical
constructs implicit in any regime of private law that values and aspires
to its own justificatory coherence, they can have a variable content
that is relative to a society’s particular tradition of positive law, to the
history of its legal responses to given problems, and to the shared social
understandings that obtain at a given time and place. Accordingly, the
comparative study of law across different jurisdictions and historical
periods has a natural place within this conception of legal education.37

36
On orientation as a role of theory, compare the remarks of John Rawls about political
philosophy:
[I]t belongs to reason and reflection (both theoretical and practical) to orient us in the (concep-
tual) space, say, of all possible ends, individual and associational, political and social. Political phi-
losophy, as a work of reason, does this by specifying principles to identify reasonable and rational
ends, and by showing how those ends can cohere within a well-articulated conception of a just
and reasonable society.
John Rawls, Justice as Fairness: A Restatement (2001), 3. In this vein one might say with respect
to private law that legal theory orients us in the conceptual space of all possible justifications
for liability. It does this by specifying the constructs of correlativity and personality in order
to identify the appropriate kind of justifications and by showing how justifications of that
kind can cohere within a well-articulated conception of private law.
37
The importance of comparative legal studies has long been asserted, as is evident from
the following observations of James Bradley Thayer:
[I]t has been wisely said that if a man would know one thing, he must know more than one. And
so our system of law must be compared with others; its characteristics only come out when this is
done . . . If any one would remind himself of the flood of light that may come from such com-
parisons, let him recall the brilliant work of Pollock’s predecessor at Oxford, Sir Henry Maine, in
his great book on Ancient Law.That is the best use of Roman law for us, as a mirror to reflect light
upon our own, a tool to unlock its secrets . . .
Of the values of such comparative studies, and the immense power to lift different subjects of
our law into a clear and animating light, no competent person who has once profited by them can
ever doubt.
James Bradley Thayer, “The Teaching of English Law at Universities,” (1895) 9 Harv. L. Rev.
169, 178–79.
322 corrective justice

Because legal education so conceived focuses on legal practice as a

culture of justification, comparative study involves the comparison not
of differing legal doctrines across systems, but of the justifications
offered for differing doctrines, of the conceptual structures into which
such justifications fit, and of the adequacy of these conceptual struc-
tures to the underlying ideas of correlativity and personality. Such
study can create awareness of the possible latitude for actualizing the
constructs of correlativity and personality within legal practice, and
thus of the particularities and contingencies of one’s own legal system.
It also allows for an appreciation of the persistence of correlativity and
personality in the structure of different legal systems, and thus of the
existence of a distinctive mode of normative reasoning that transcends
the particularities of different systems of private law and that consti-
tutes the condition for the possibility of productive comparison among
them.
Correlativity and personality can also move from the background
to the foreground. When this occurs, the focus shifts from legal prac-
tice to legal philosophy. Correlativity and personality are then used to
connect private law as a normative phenomenon to the corpus of
legal and political philosophy. Particularly relevant is the history of
reflection about correlativity that begins with Aristotle’s treatment of
corrective justice38 and culminates in the accounts of natural right
formulated by Kant and Hegel.39 These texts can then become the
vehicle for considering further regressions from personality as a pre-
supposition of correlativity to what is presupposed in the notion of
personality itself.They thereby present an opportunity to deepen one’s
understanding of private law by opening up a further series of ques-
tions about its normative foundations, its relationship with other kinds
of normative phenomena, and its place within more comprehensive
philosophical systems.
What I have suggested in this section with respect to private law is
a conception of legal education that remains rooted in the practice of
law while focusing on the most general conceptual framework impli-
cit in it. This conception is incompatible with the instrumental
approaches to law that now enjoy primacy. It replaces the instrumen-
talist emphasis on independently justifiable goals with attention to the
internal structure of legal relations and to what must be presupposed

38
Aristotle, Nicomachean Ethics, V, 1131b25–1132b20.
39
Hegel, above n. 9; Immanuel Kant, The Metaphysics of Morals, tr. Mary Gregor (1996).
 can law survive legal education? 323

if legal relations are to be normatively coherent. Its enquiry is internal

in every respect. It is internal to the law in that it purports to make
sense of legal thinking in its own terms. In its focus on coherence, it
enquires into the internal relationship among the components of an
integrated justificatory structure. And in specifying the character of
private law, it identifies theoretical constructs that are internally relat-
ed to one another and to private law. It thereby brings together the
three activities of the practice, the understanding, and the university
study of private law, so that law can indeed survive legal education.

4. The interdisciplinary turn

In this final section, I explore the implications of this chapter’s under-
standing of private law for interdisciplinary study. The rise of interdis-
ciplinary scholarship has been perhaps the most dramatic development
in legal education over the last generation. In place of the previous
emphasis on cases and doctrine, a new paradigm of scholarship and
teaching has arisen, which brilliantly mobilizes the insights of other
disciplines—economics, literature, and philosophy are among the
favorites—to the analysis of legal material. Of course, the use of these
disciplines can be consistent with the character of private law. In the
previous section I suggested that philosophical issues readily emerge
from the elucidation of this character, while cautioning that these
issues deal with the presuppositions of legal discourse without being
explicitly part of it. However, much of the current interdisciplinary
interest is not based on such elucidation of the law’s character, but is
nonetheless rooted in the aspiration to weave the study of law into the
intellectual fabric of university life.40 How does this kind of interdis-
ciplinary work relate to legal education that places the law’s ideal char-
acter at its core?
The popularity of interdisciplinary scholarship is a natural out-
growth of the instrumentalist approach to private law. As evidenced by
its cavalier treatment of the fundamental legal concepts, the instru-
mentalist approach denies that the content of private law arises indi-
genously in accordance with the correlative structure of its legal
relationships. Rather, the law is merely the passive receptacle of what-
ever goals are to be imposed on it from the outside. The study of law,

40
See Menachem Mautner, On Legal Education (2002) (Hebrew).
324 corrective justice

accordingly, is not an autonomous body of learning, but an empty

shell dependent on non-legal disciplines for the validation of the pro-
posed goals. Hence the proliferation of rich interdisciplinary interests
in “Law and . . . ,” with the vital element in the pairing being invariably
the non-legal one. Law provides only the authoritative form into
which the conclusions of non-legal thinking are translated. Law is
considered to have no meaning except that which it absorbs from
other disciplines and enquiries. Indeed the capacity to funnel insights
about law so conceived through alien concepts and terminology is
considered the mark of scholarly detachment and sophistication.
The paradoxical consequence of this basis for interdisciplinary
scholarship is that the interdisciplinary turn is actually an illusion. For
academic work to be truly interdisciplinary, it must engage more than
one discipline. Law, however, is regarded not as a discipline in its own
right with something of its own to contribute to the interdisciplinary
enterprise, but merely as a context for projects from other disciplines.
The resulting study is nothing but the application of a particular non-
legal discipline to examples drawn from the law. The economic ana-
lysis of law, for instance, then turns out to involve not an enlarged
study of law but a restricted study of economics. Given that the legal
side lacks intellectual resources of its own, what motivates the
interdisciplinary scholar is interest in the non-legal discipline. As was
once observed about the parallel phenomenon in literary studies:

[a] scholar with a special interest in . . . economics expresses that interest by

the rhetorical device of putting his favorite study into a causal relationship
with whatever interests him less. Such a method gives one the illusion of
explaining one’s subject while studying it, thus wasting no time.41

The recognition that private law embodies a distinctive mode of

thinking is both a prerequisite of interdisciplinary work about private
law and a determinant of that work’s nature. It is a prerequisite because
without it reference to the insights of another discipline is reductive
rather than interdisciplinary; the other discipline is invoked to show
that the law, at the most, reflects a deformed version of those insights.
It is a determinant of the nature of interdisciplinary work because that
work, of whatever kind it is, has to allow private law the independent
space entailed by that recognition. Interdisciplinary work can then not

41
Northrop Frye, Anatomy of Criticism (1957), 6.
 can law survive legal education? 325

be conceived as the construction of a repository of homogeneous

knowledge, because such homogeneity is inconsistent with the dis-
tinctiveness of the legal mode of thinking. Rather, knowledge has to
be regarded as pluralistic—that is, as organized into categorically dif-
ferent kinds of enquiry each of which (including the study of law) has
its distinct character.
Within the university law school, the point of interdisciplinary
study is to present different perspectives for the understanding of the
transactions governed by the law. Each of these perspectives has its
own validity, rests on its own presuppositions, and operates within its
own disciplinary boundaries. A medical misadventure, for example,
may raise not only issues of liability, but also issues of economics, of
sociology, of political science, of psychology, and so on. Within a law
school the legal perspective has of course a certain contextual primacy,
because, whatever else it does, a law school must impart to its students
a sense of the law’s distinctiveness as a mode of normative discourse.
This distinctiveness excludes other perspectives, but does not deny
their authority within their own spheres. Indeed, exposure to these
other perspectives plays an important role for the study of law for sev-
eral reasons. First, the very contrast between legal and non-legal modes
of enquiry casts light on the law’s distinctive structure and presuppos-
itions (as the law does on theirs). Second, the contrast reveals the place
of law as an intellectual enterprise among other such enterprises, and
that civilized life requires the cooperation and mutual respect of all of
these. Third, an awareness of the contrast induces an appreciation of
the limits of law, and thus a proper sense of humility: although the law
governs all of life, the person who is learned in the law is not there-
fore omniscient.
Accordingly, one can view the interdisciplinary study of law as cre-
ating an academic conversation with different disciplinary voices.42
The object of this conversation is not to have one voice suppress any
of the others but to maintain the individuality of each. When every
voice contributes the insight that derives from its own distinctive
activity, the conversation can enlarge the understanding of its partici-
pants.
It is not always easy to maintain the line between respectful atten-
tion to a different voice and hearing that voice as a deformed version

42
For the metaphor of a conversation, see Oakeshott, above n. 1, at 195–96; Michael
Oakeshott, The Voice of Liberal Learning (2001), 109.
326 corrective justice

of one’s own. Consider, for example, the most celebrated and influen-
tial piece of interdisciplinary legal scholarship of the twentieth cen-
tury, Ronald Coase’s treatment of social cost.43 Coase’s article deals with
what he calls “a technical problem of economic analysis”44 regarding
the harmful effects of one’s actions on others. In the exposition of his
analysis he uses classic court judgments dealing with the law of private
nuisance. It is worth pausing on how he views the relationship between
his economic argument and the judicial illustrations.
On the one hand, Coase is admirably sensitive to the difference
between economic and legal analysis. The economic problem is how
to maximize the value of production.45 The legal problem is how to
determine liability. Coase insists, rightly, that this difference should not
confuse economists about the nature of their problem.46 “The reason-
ing employed by courts in determining legal rights,” he observes, “will
often seem strange to an economist because many of the factors on
which the decision turns are to an economist irrelevant.”47 In decid-
ing the economic problem, certain legal considerations are “about as
relevant as the colour of the judge’s eyes.”48 Economists do one thing
and judges do another. Therefore, economists should not take their
cue from how judges deal with the external effects of the defendant’s
actions.
On the other hand, Coase does not think that the converse obtains.
Although the economists should not be influenced by the judges, he
thinks it desirable for judges, because their decisions directly influence
economic activity, to take economic analysis into account.49 His
assumption is that although the economist’s problem is different from
the judge’s, the judge’s problem is not all that different from the econ-
omist’s. He takes it for granted that the law of nuisance has no charac-
ter of its own beyond the influence it exerts on economic activity, and
that it therefore should be animated by properly formulated economic
considerations. Of course, qua economist Coase has no reason to be
alert to (let alone, to explore) the distinctiveness of the legal mode of
thinking in matters of private law. However, the consequence of his

43
See Ronald H. Coase, “The Problem of Social Cost,” (1960) 3 J.L. & Econ. 1.
44
Oakeshott, above n. 1, at 1.
45
Above n. 43, at 15.
46
Ibid., at 9.
47
Ibid., at 15.
48
Ibid.
49
Ibid., at 19.
 can law survive legal education? 327

justifiable preoccupation with his “technical problem of economic

analysis”50 is that he reads the cases as containing renditions, often
inadequate, of the economic argument.
In Bryant v. Lefever,51 for example, the defendants extended upward
the wall that ran beside the plaintiff ’s chimneys, with the result that
when the plaintiff lit a fire in any of his rooms, the smoke could not
clear but came back into the plaintiff ’s house. What attracts Coase’s
attention is not the court’s ultimate judgment in favor of the defend-
ants (in the absence, as here, of transactions costs, the parties would
bargain toward the efficient result whatever the court decided) but
rather the court’s misapprehension of the causal relationship between
the parties’ activities. In dismissing the plaintiff ’s claim, the court held
that, although the defendants erection of the wall materially interfered
with the plaintiff ’s comfort and thus constituted a nuisance, the nui-
sance was caused not by the defendants act but by the plaintiff ’s light-
ing of the fires. Criticizing this crucial element in the court’s analysis,
Coase observes that it was “fairly clear” that both parties caused the
smoke nuisance, because “[g]iven the fires, there would have been no
smoke without the wall; given the wall, there would have been no
smoke without the fires.”52 Accordingly, “both were responsible and
accordingly both should be forced to include the loss of the amenity
due to smoke as a cost of deciding whether to continue the activity
which gives rise to the smoke”—which in fact is what would happen
because of the possibility of costless bargaining. The court made the
mistake of assuming that the wall was “the given factor,”53 thus
transforming the parties’ joint causation into the self-infliction of
harm by the plaintiff .
This criticism ignores the juridical quality of the court’s reasoning.
The court’s task is to determine liability within a system of rights and
correlative duties. These rights and duties establish the baseline from
which to determine the direction in which causation moves. Caus-
ation is, accordingly, not a natural phenomenon that reflects the fact
that the smoke was the result of the combination of the defendants
wall and the plaintiff ’s fire lighting. Nor is it an economic conclusion
geared to the maximization of the value of production in accordance
with what Coase calls “the beauties of a smoothly operating pricing
50
Ibid., at 1.
51
Bryant v. Lefever [1879] 4 C.P.D. 172 (Eng.).
52
Coase, above n. 43, at 13.
53
Ibid.
328 corrective justice

system.”54 Rather, causation functions here within an argument about

the imputation of liability. Liability requires that the defendants’ actions
have been inconsistent with the rights of the plaintiff and not merely
the exercise of one of their own rights. Here by extending the wall
upwards, the defendants did nothing but occupy the space that they
owned. The legitimacy of occupying space that one owns is insepar-
able from the right involved in ownership. It therefore does not con-
stitute a wrong that can be imputed to the defendants. Moreover,
because the plaintiff ’s harm was the consequence of the smoke’s fail-
ure to clear the defendants’ building, the plaintiff was wronged only if
he had a right to the space over the defendants’ property required for
clearance. That space, however, belonged to the defendants and was
not subject to any right of way in favor of the plaintiff. Of course, the
plaintiff was harmed by the defendants’ action in the sense that he was
worse off than before the wall was built. However, in view of the con-
figuration of rights in this situation, the plaintiff was not wronged.
One can sum up what is problematic about Coase’s criticism of the
court as follows. For Coase the given factor in the court’s analysis was
the wall, because the wall made what is properly a cause of the smoke
into a condition for ascribing causation only to the plaintiff ’s lighting
of the fire. From the court’s perspective, however, it would be more
accurate to say that the given factor was not the wall itself, but the
right to build it. The court’s assumption was simply that within a
coherent system of proprietary rights, one has the right to build on
one’s own property. Instead of viewing the court’s decision as exem-
plifying a properly legal mode of analysis of the relationship between
owners of adjoining properties, Coase reads it as a misstatement of the
obviously reciprocal causation that should underlie the economist’s
attitude toward the maximization of the value of production.
Similar observations can be made about Coase’s famous treatment
of Sturges v. Bridgman.55 The issue in that case was whether activity by
the defendant that would otherwise constitute a nuisance escaped lia-
bility because it temporally preceded the use by the plaintiff with
which it interfered. The defendant was a confectioner who had been
using the back section of his property for decades as a kitchen in
which he pounded his meats. No problem arose as long as the adjoin-
ing property was used as a garden, because the noise did not inconvenience

54
Ibid.
55
Sturges v. Bridgman [1879] 11 Ch. D. 852 (Eng.).
 can law survive legal education? 329

anyone. The plaintiff, however, was a physician who had recently built
a consulting room on the site of the garden, and complained that the
noise of the pounding interfered with his practice. The court held the
defendant liable.
Coase singles out the court’s statement that a different result would
“produce a prejudicial effect upon the development of land for resi-
dential purposes.”56 He points out that so long as market transactions
between the confectioner and the physician were costless, the court’s
decision could have no effect on the allocation of resources. If one
party gained more from the continuation of his activity than the other
lost from the cessation of his, the party that stood to gain more would
strike a bargain that would allow him to continue even if he lost his
case. “The judges’ view that they were settling how land was to be
used,” Coase writes, “would be true only in the case in which the
costs of carrying out the necessary market transactions exceeded the
gain which might be achieved by any rearrangement of rights . . . But
of this the judges seem to have been unaware.”57
Here, too, Coase is hearing in the court decision not the distinctive
voice of legal discourse but an inferior version of his own economist’s
voice. Coase treats the court as attempting to achieve a certain eco-
nomic goal (the development of residential housing) in ignorance of
proper economic reasoning. If, however, one reads the judgment as a
whole and views the offending sentence in its light, a different picture
emerges. The court was concerned not with settling how the land was
to be used, but with determining the conditions under which an
action by the defendant could diminish a right of the plaintiff . The
court’s focus was juridical, not economic.
For the court, the problem with the confectioner’s position was
that it asserted a power unilaterally to restrict another’s right. Under
the law of nuisance the plaintiff had a right to the use and enjoyment
of his property, but this right could not be vindicated in a court of law
so long as his use and enjoyment was unaffected. While the property
was being used as a garden, the physician (or his predecessor) had no
cause of action in nuisance, because the noise from the confectioner’s
pounding did not inconvenience him. The confectioner’s argument
once the physician built the consulting room was that the physician
had lost the right to complain of the nuisance. But how can one

56
Ibid., at 866.
57
Coase, above n. 43, at 10.
330 corrective justice

person’s right be extinguished by the unilateral acts of another? The

confectioner might argue that the long history of meat pounding
shows that the physician (or his predecessor) implicitly acquiesced in
the confectioner’s acquisition of an easement. As the court pointed
out, however, one can acquiesce only in what one can prevent. Until
the physician suffered an inconvenience that allowed him to sue in
nuisance, he could not prevent the meat pounding, and therefore can-
not be taken to have acquiesced in it.
What, then, is the significance of the court’s reference to “the
prejudicial effect upon the development of land for residential pur-
poses,” around which Coase’s treatment of the case revolves?58 The
comment comes after the court considered a hypothetical case,
which it rightly regarded as exactly analogical to the case at hand, of
“a blacksmith’s forge built away from all habitations but to which, in
the course of time, habitations approach.”59 The court disqualified
two possible treatments of this hypothetical case. On the one hand,
it would be unreasonable to extend liability for nuisance to the
period before the habitations approached because that would give the
adjoining landowners a right to sue for an inconvenience that they
have not yet, and may never, suffer. But, the court continued, “it
would be on the other hand in an equal degree unjust, and, from a
public point of view, inexpedient that the use and value of the ad join-
ing land should, for all time and under all circumstances, be restrict-
ed and diminished by reason of the continuance of acts incapable of
physical interruption, and which the law gives no power to pre-
vent.”60 This is the sentiment that the subsequent remark about the
prejudicial effect on residential development encapsulates. It is not
that the court necessarily thinks that the growth of the habitations
mentioned in the hypothetical case is itself desirable: the court is not
attempting to settle land use for a hypothetical case. What is unjust
and inexpedient is that the owners of the land on which habitations
might arise should have their rights prejudiced by actions of the
blacksmith that they cannot physically interrupt or legally prevent.
The “prejudicial effect on the development of land for residential
purposes” is not that there will be too few residences, but that right
of the hypothetical landowners to develop residences protected by

58
Sturges, above n. 55, at 866.
59
Ibid., at 865.
60
Ibid.
 can law survive legal education? 331

the law of nuisance will be abridged by the blacksmith’s unilateral

activity.
Both Bryant and Sturges are exemplary displays of the mode of
thinking distinctive to private law. Each case treats the alleged injust-
ice as one to which the parties must be correlatively situated as doer
and sufferer. The doing and the suffering consist, respectively, in the
infringement, and in being the victim of the infringement, of the right
to use and enjoy one’s property. This right represents a sphere of law-
ful freedom with which others are obligated not to interfere.The cases
enquire into the meaning of this right in the specific circumstances of
building up and of pre-existing activity by the defendant. In both cases
the court’s reasoning provides a sophisticated elaboration in legal
terms of the relationship between the plaintiff ’s right and the defend-
ant’s action.
In contrast, Coase assumes that the courts are addressing issues of
economic rather than juridical thinking. He reads the reference to
causation in Bryant not as working out an imputation of liability
against the background of the parties’ rights, but as overlooking the
obvious causal reciprocity that informs a proper economic analysis.
Similarly, he reads the reference to the prejudicial effect on residential
development in Sturges not as underlining the illegitimacy of the
defendant’s unilaterally restricting the plaintiff ’s right, but as failing to
anticipate the bargaining that would take place when transactions are
costless. In each case the court was not making an economic argu-
ment in ignorance of the economic consequences that Coase describes;
it was making a legal argument about the connection between one
person’s rights and another person’s will that Coase has misappre-
hended.
This misapprehension, of course, does not mean that Coase’s bril-
liant article has no place in a university program of legal education. To
the contrary, it should have a place of great honor as a masterful exhib-
ition of the subtleties of the economic approach to situations that
have attracted the law’s interest. Both law and economics feature
modes of thinking that present systematic and comprehensive under-
standings of human interaction. In the nuisance context economic
thinking, as portrayed by Coase’s analysis, is concerned with harms
and costs, whereas legal thinking is concerned with rights and rem-
edies. The contrast between these two modes of thinking is not only a
matter of genuine intellectual significance in its own right, but also a
considerable resource for imparting to students a sense of the character
332 corrective justice

of private law. For we often begin to understand what something is by

seeing what it is not.
If the interdisciplinary turn in legal education is to bear fruit, it
must be supposed that law has an independent voice that can contrib-
ute to the conversation among the university’s disciplines. Whatever
else it does, legal education is charged with the task of inculcating in
students the capacity to speak in this voice and to understand its dis-
tinctive character. In the conversation among disciplines this voice is
juxtaposed against other voices, so that the significance of each—its
presuppositions, its organizing structures, its way of relating to the par-
ticulars of the world that all share—can be better appreciated. The
idea of a conversation is not that one voice replaces or silences or
dominates the others, but that each puts forth the ideas appropriate to
it, humble in the awareness of its own limits and respectful of the distinct-
iveness of others. For in a conversation, “[i]ts integration is not super-
imposed but springs from the quality of the voices which speak, and
its value lies in the relics it leaves behind in the minds of those who
participate.”61

5. Conclusion
The central theme of this chapter—indeed, of this entire book—has
been that private law is animated by a distinctive mode of thinking
and discourse, marked by the structure of correlativity and informed
by the presupposition of personality. Accordingly, with respect to pri-
vate law, the university study of law, whatever else it does, has the task
of engaging the student in this mode of thinking and discourse. The
disjunction that critics of legal education have noted between univer-
sity study and legal practice is the consequence of understanding law
in instrumental terms and thereby obscuring the law’s distinctive char-
acter. One overcomes this disjunction by attending to the role of cor-
relativity and personality in an understanding of private law that is
faithful to the law’s conception of itself as a normative phenomenon
that strives for justificatory coherence in the relationship between
plaintiff and defendant.
That law embodies a distinctive mode of thinking and discourse is
a venerable idea. One recalls Coke’s response to the assertion by James

61
Oakeshott, above n. 42, 110.
 can law survive legal education? 333

I that, because the law was founded upon reason, he as a person

endowed with reason was as qualified to sit in judgment as were the
judges. Coke replied with a ringing affirmation of the distinctiveness
of reason in the legal context:

[T]rue it was that, God had endowed His Majesty with excellent science
and great endowments of nature; but His Majesty was not learned in the
laws of his realm of England, and causes . . . are not to be decided by natural
reason but by the artificial reason and judgment of the law, which law is an
art which requires long study and experience, before that a man can attain
to cognizance of it.62

In the modern context universities have the responsibility of beginning

the process of “long study and experience” that imparts “cognizance”
of this “artificial reason and judgment.” This is a responsibility in which
recognition of the law’s distinctive character converges with the uni-
versity’s calling to care for law as one of civilization’s pre-eminent
achievements.

62
Prohibitions del Roy (1607) 77 Eng. Rep. 1342, 1343 (K.B.).
 Conclusion

The preceding chapters have illustrated the significance of corrective

justice for various legal contexts: for the grounds of liability, for the
relationship between right and remedy, for legal education, for the
comparative understanding of private law, and for the compatibility of
corrective justice with state support for the poor. The aim has been to
show, through the concrete treatment of a variety of issues, how cor-
rective justice illuminates the structure and content of private law. The
chapters have thereby set out a broad conception of corrective justice.
The ideas presented here stand in contrast to the rather modest
role that others have ascribed to corrective justice. For example,
Richard Craswell, writing from the perspective of economic ana-
lysis, has asserted that “[w]hile corrective justice theory can give us a
way of talking about what to do when the relevant baseline is
infringed, it cannot tell us what baseline ought to be selected as rele-
vant.”1 The attraction of this view for proponents of economic
analysis is that it allows the baseline to be set by considerations of
efficiency, while reducing corrective justice to the task of rectifying
violations of those baselines.2 However, one does not have to be an
economic analyst to regard corrective justice as having solely a remed-
ial function. John Gardner, who has evinced little sympathy for eco-
nomic analysis, has taken the view that corrective justice involves
the allocation of something back to the plaintiff; the consequence of
this is that corrective justice deals not with the norms that govern a

1
Richard Craswell, “Against Fuller and Perdue,” (2000) 67 U. Chi. L. Rev. 99, at 127.
2
This was the strategy in Richard A. Posner, “The Concept of Corrective Justice in
Recent Theories of Tort Law,” (1981) 10 J. Legal Stud. 187.
 conclusion 335

transaction but only with restoring the situation antecedent to the

norm’s violation.3
Common to these views is the limited scope that they postulate for
corrective justice. Corrective justice performs a certain operation: cor-
recting the consequences of a transaction that has gone askew. How-
ever, the basis on which it performs this operation—what it means for
the transaction to go askew—is independent of the operation. In con-
trast, the argument in this book is that the phenomenon of liability,
being correlatively structured, would not coherently correct the injust-
ice between the parties unless the reasons for considering something
an injustice were also correlatively structured. The correlativity of the
correction presupposes correlativity in the basis for the law’s perform-
ing this correction. We can term the notion that corrective justice
covers both the correction and the basis of liability the “juridical con-
ception,” because it exhibits the structure that informs the phenom-
enon of liability in all its aspects, from how the law conceives of injustice
as between the parties to how the law corrects that injustice by award-
ing a remedy. This is in contrast to what might be termed the “remed-
ial conception,” the notion that corrective justice covers only the
correction effected through the remedy.
The claim inherent in the juridical conception, that both the oper-
ation of the remedy and the normative character of the bases of liabil-
ity fall under a single conception of justice, is substantive, not
terminological. The point of this claim is not to stipulate the proper
usage of the words “corrective justice,” but to exhibit the complex of
ideas that yields a comprehensive and coherent understanding of the
private law relationship. To be sure, the juridical conception and the
remedial conception each make terminological assumptions about the
words “corrective justice.” While both regard corrective justice as the
form of justice that deals with corrections rather than distributions,
they differ in their appreciation of the role of correction within that
form. The remedial conception takes corrective justice to be about
the operation that corrects and nothing else. The juridical conception
takes corrective justice to be about the system of norms within which
the operation that corrects makes sense. The task of theory is to deter-
mine not which of these conceptions involves a superior terminological

3
John Gardner, “What Is Tort Law For? Part 1:The Place of Corrective Justice,” (2011) 30
Law and Philosophy 1. An earlier adumbration of Gardner’s views appeared in Gardner, “The
Purity and Priority of Private Law,” (1996) 46 U. Toronto L.J. 459.
336 corrective justice

practice (by consulting a lexicon, for example, or surveying linguistic

usage) but which of them sheds greater light on the normative char-
acter of private law.
Gardner’s criticism of the juridical conception illustrates the futility
of confusing terminology with argument. Drawing on Nozick’s long-
ignored terminology, Gardner regards distributive and corrective just-
ice as distinguishing not between different structures of coherent
justification but between different subject matters. Norms of distribu-
tive justice regulate the allocation of goods among persons, whereas
norms of corrective justice regulate the allocation of goods back from
one person to another. In his view, norms of either sort of justice are
not necessarily just; whether a norm is just depends on soundness of
the particular norm, not on its being a norm of justice.4 Adapting a
famous dictum, one might say that Gardner holds that the existence
of a norm of justice is one thing, its merit or demerit another.
The result of characterizing corrective justice in this way is that
corrective justice has nothing to do with the coherence of the reasons
for liability. Presumably, coherence would count toward soundness,
but that would be a consideration independent of or additional to
corrective justice, not a consideration of corrective justice itself. Even
an unsound reason for allocating something back to the plaintiff is for
Gardner a reason of corrective justice. Furthermore, corrective justice
has nothing to do with matching the reason for the norm breached
by the defendant with the correlative structure of liability. The norm
can be a distributive one that allocated a certain good to the plaintiff ,
and yet triggered the allocation of that good back to the plaintiff as a
matter of corrective justice. Indeed for Gardner, the tort that leads to
the allocation of the good back to the plaintiff is generally not an
injustice at all—the tortious act is itself generally neither an allocation
nor an allocation back—but is the violation of other sorts of moral
norms.5
This leads Gardner to make the following criticism of the juridical
conception. The juridical conception includes the norms of private
law within corrective justice, because the reasons supporting the bases
of liability have to be correlatively structured if a correlatively struc-
tured finding of liability is to function as a coherent correction of
injustice. Gardner objects to this, on the ground that if the norms are

4
Gardner, “What Is Tort Law For?,” above n. 3, 12–15.
5
Ibid., 23.
 conclusion 337

part of corrective justice, they must be correcting something else, and

so on in infinite regress.6 The norms, in other words, are not correct-
ive because they do not correct anything. Nor does a person correct
anything by acting in accordance with them.
To this criticism the reply on behalf of the juridical conception is
that the norms indeed do not correct anything, but that is beside the
point. Under the juridical conception the norms fall under corrective
justice because they belong to the system within which bipolar cor-
rections make sense, not because norms themselves operate to correct.
Of course, the reply is circular, but so is Gardner’s objection. This is
because both the objection and the reply merely rehearse the initial
terminological difference between the two conceptions. Gardner
thinks that, in pointing to the inconsistency of the juridical concep-
tion with what he means by corrective justice, he is making a real
criticism. But all that he is doing is drawing out an implication of his
own terminology.
Over the more than two millennia that the notion of corrective
justice has been in circulation, it has been formulated in many differ-
ent ways.7 From the standpoint of the Aristotelian tradition, within
which corrective justice originated and was nurtured, Gardner’s view,
with its consequence that the tortfeasor does not commit an injustice,
is eccentric. Of course, Gardner is as entitled to his terminology as
others are to theirs. However, the juridical conception cannot be dis-
qualified by terminological fiat. It is no argument against the ideas
that the juridical conception brings together under corrective justice
that Gardner uses the words “corrective justice” in a different way. In
this context as in others, we should think thoughts, not words.
Inherent in the juridical conception is the ambitious claim that a
single set of ideas is decisive for understanding the normative charac-
ter of private law. Whereas the remedial conception confines the rem-
edy to its separate theoretical box, the juridical conception integrates
the remedy and the reasons that would justify awarding it. This inte-
gration is part of an extensive argument that comprehends all the bases
of liability within a single conception that renders the plaintiff–
defendant relationship coherent. On this view corrective justice is
neither an empty abstraction that cannot account for the content of
private law, nor merely a way of referring to the bipolar operation

6
Gardner, “The Purity and Priority of Private Law,” above n. 3, 470.
7
Izhak Englard, Corrective and Distributive Justice: From Aristotle to Modern Times (2009).
338 corrective justice

of the duty to repair, as the remedial conception supposes. Rather,

corrective justice is a comprehensive idea based in the self-determin-
ing freedom of the interacting parties, extending to the entire range of
private law, and operating through the correlativity of right and duty.
This account of the juridical conception of corrective justice weaves
together a series of interrelated and mutually dependent ideas.
First and foremost is the idea that correlativity is the organizing
structural feature of the relationship between the parties. Corrective
justice works out the normative implications of the most general and
pervasive characteristic of liability, that the liability of a particular
defendant is always a liability to a particular plaintiff . This characteris-
tic is evident in the bipolar process of litigation and in the remedy
that obliges the defendant to pay damages or give specific relief to the
plaintiff. Because liability corrects an injustice, the structure of liability
and the structure of the injustice have to be congruent. Accordingly,
the correlativity that provides the structure for the correction presup-
poses correlativity in the reasons supporting the norm whose breach
occasions the correction. For if those reasons are non-correlative, in
that their normative force does not link the particular defendant who
commits the injustice to the particular plaintiff who suffers it, then the
law has no coherent basis for tying the defendant’s liability to the
plaintiff ’s entitlement.
Second, because correlativity pervades the phenomenon of liability
generally, the juridical conception of corrective justice is pertinent to
all the bases of liability in private law, and not merely to particular
ones. Corrective justice is, therefore, not to be identified solely with
tort law or restricted to specific causes of action. Rather, it displays the
structure of legal relationships within a general theory of private law.
This book has illustrated this by providing a corrective justice account
not only of negligence law, but also of property, contract law, and
unjust enrichment.
Third, the correlative structure of the parties’ relationship receives
its content through a system of rights and their correlative duties. As
chapter 1 maintained, the structure and the content are governed by
complementary abstractions. Just as the correlativity of the parties’
normative positions most abstractly represents the structure of their
relationship, so the notion of personality most abstractly represents the
content of that relationship. Personality refers to the self-determining
freedom of acting for purposes that one sets for oneself in one’s rela-
tions with another. It gives a conception of the parties in which what
 conclusion 339

counts for their legal relationship is that their actions are the expres-
sion of purposiveness, rather than of particular obligatory purposes.
Because personality is not the product of a distribution but the con-
ception of agency that underlies the correlatively structured relation-
ships of private law, it is implicit within the juridical conception of
corrective justice.
Fourth, the rights and duties of private law reflect this conception
of personality. Within a given legal relationship, the rights and duties
of the parties articulate a relationship of will to will, in which the
actions of each party coexist with the purposiveness of the other,
regardless of the independent desirability of any particular purpose.
Rights so conceived are not interests that the positive law sanctifies
because of their unilateral importance to the plaintiff , but the jurid-
ical manifestations of reciprocal freedom in the will-to-will relation-
ship between the parties. This conception of rights and duties
accounts (as was noted in chapter 2) for the absence of a tort duty in
circumstances of nonfeasance, no matter how important the plain-
tiff ’s interest might be. It is also crucial for the corrective justice
account of both in rem rights and in personam rights. Kant’s account of
property, described in chapter 8, is illustrative of an argument that
constructs the institution of property out of the normative implica-
tions of freedom in one will’s interaction with other wills, without
either invoking distributive considerations or mandating particular
purposes. Similarly, the discussion of unjust enrichment in chapter 6
suggests that the plaintiff ’s in personam right to the restitution of an
unjust enrichment arises out of the law’s endeavor to assure that the
transfer and retransfer of value accord with the freedom of the will of
the two interacting parties.
Fifth, private law achieves coherence within the relationship
between the parties by linking them through an integrated ensemble
of legal concepts that melds the normative positions of the parties
into a single unified normative unit. Accordingly, as set out in chapter
2, negligence law translates into a series of legal concepts, the progres-
sion from the defendant’s creation of an unreasonable risk to the
materialization of that risk in injury to the plaintiff . The termini of
this progression are the concepts of breach of the standard of reason-
able care and factual causation. However, these two termini do not
operate as atomistic elements, one applying to the defendant and the
other to the plaintiff , that the law simply adds together. Rather, the
law insists that the termini be coherently linked through the concepts
340 corrective justice

of duty and proximate cause. These concepts in turn connect wrong-

doing and injury by describing the wrongful risk in terms of the range
of the potential victims and consequences through which the risk is
to be understood as wrongful. Thus, the risk that materializes into
injury to the plaintiff is the same as the risk that the defendant unrea-
sonably created, and the wrong done in its creation is the same as the
wrong suffered in its materialization. Through this ensemble of con-
cepts the law treats the defendant’s act and its effect on the plaintiff as
an integrated sequence.
The treatment of unjust enrichment in chapter 6 presents a more
detailed illustration of the same idea. In that context, transfer and
unjustness form the two interwoven strands for unjust enrichment.
The transfer strand deals with the movement of a gratuitous benefit
from defendant to plaintiff . The unjustness strand relates the wills of
the two parties through the non-donative making and accepting of
the gratuitous transfer, thereby bringing out the normative implica-
tions of non-donative transfer for their relationship. Each strand is
both correlatively structured and related to the other. Together the
four components of liability—the defendant’s enrichment, the
enrichment’s being at the plaintiff ’s expense, the plaintiff ’s lack of
donative intent, and the defendant’s acceptance of the benefit as
non-donatively given—form an integrated ensemble that states a
coherent basis of liability. Within this ensemble, non-donative
intent and acceptance provide a reason, normatively relevant to
both parties in their interrelationship, to reverse the transfer of a
non-donative benefit.
Sixth, as set out in chapter 3, rights and remedies form a unified
sequence. Within the juristic conception of corrective justice, the cor-
relative operation of the remedy implies a matching correlativity in
the reasons that justify the award of the remedy. The logic of rights
reveals more specifically why the remedy and the ground of liability
are so closely aligned. The injustice that the finding of liability cor-
rects is an inconsistency with some right of the plaintiff that is imput-
able to the defendant. To have a right is to be so connected with the
object of the right that another person is under an obligation not to
interfere with that object. Because that connection is unaffected by
the occurrence of the injustice, the plaintiff ’s right and its correlative
duty survive, though in a different form, and continue to regulate the
relationship. The remedy is just that right and its correlative duty at a
later stage of their existence.
 conclusion 341

Seventh, the idea that the remedy is a continuation of the right goes
hand in hand with the idea that the plaintiff is not entitled to a remedy
that gives more than the right or its equivalent. This accounts for the
opposition of corrective justice to punitive damages, which by defin-
ition go beyond the plaintiff ’s rights.The criticism of punitive damages
for contractual breach in chapter 5 exemplifies this opposition. The
role of the right in limiting and defining the remedy is also at the heart
of the analysis of gain-based damages in chapters 4 and 5. The crucial
point is that one cannot determine the remedy without ascertaining
the precise nature of the right whose infringement is being remedied.
Consequently, in many contexts where gain-based damages have been
proposed or even awarded—for example for breach of contract in the
cases discussed in chapter 5—attention to the nature of the right in
question may reveal that such awards are misplaced.
Eighth, corrective justice understands the particularity of specific
norms as falling under the abstract and general ideas of correlativity
and personality. Bridging the particular and the general are intermedi-
ate concepts, including the conceptual ensembles (such as those just
mentioned regarding negligence law and unjust enrichment) that
integrate the normative positions of the plaintiff and the defendant.
This interplay of the general and the particular enables students of
comparative law profitably to explore the diversity of the particular
norms of different legal systems in the light of the general ideas com-
mon to all legal systems insofar as they are coherent.The Jewish law of
unjust enrichment, described in chapter 7, provides an example.
Although historically unconnected with and uninfluenced by the
common law, Jewish law resembles the common law in constructing
the parties’ relationship by reference to the absence of donative intent
by the plaintiff and the acceptance of the benefit by the defendant.
However, Jewish law diverges from the common law by specifying
donative intent so stringently that it is rarely present and, therefore,
can rarely disqualify the plaintiff ’s claim. As a result, the focus of doc-
trinal elaboration is on working out what constitutes acceptance,
especially what renders benefits incontrovertible. The difference
between the two systems reflects their different specification of shared
concepts. Corrective justice thus comprises the framework of com-
parison, within which the observation of difference is embedded in
the recognition of similarity.
Ninth, as the actualization of corrective justice, private law is a dis-
tinctive normative practice. The complementary abstractions of cor-
342 corrective justice

relativity and personality allow private law to be understood as having

its own relational structure, its own relational set of concepts, and its
own relational mode of reasoning. As a field of study, it is a discipline
unto itself. As a normative practice, it is categorically different
both from personal morality, with its striving for each person’s true
good, and from politics, with its search for independently desirable
social goals. As contended in chapter 9, to understand this disciplinary
and practical autonomy and to gain the competence to participate in
it is an indispensable object of legal education within the modern
university.
To conclude, at the beginning of his great work The Pure Theory of
Law, Hans Kelsen explains that his theory is pure because it aims “to
free the science of law from alien elements.”8 He then laments that
the preceding centuries had uncritically adulterated the science of law
with elements of psychology, sociology, ethics, and political theory. His
pure theory, he continues, undertakes to delimit the cognition of law
against these disciplines because their admixture has obliterated the
limits imposed on the science of law by the nature of its subject mat-
ter. To this end he offers a general theory of positive law.
Kelsen’s comments provoke the query: what would a pure theory
of law look like if it were not positivist? For Kelsen and other positiv-
ists such a theory would be an absurdity. Yet the history of reflection
about law contains notable examples. At the end of his Prolegomena
to the De Jure Belli ac Pacis, for instance, Grotius remarks that he has
treated Jus in a manner analogous to the mathematician’s abstraction
from particularity, and that he has refrained from discussing anything
that belongs to politics, which has its own special mode of enquiry.9
Similarly, Kant’s legal philosophy, with its austere distinction between
the legal and ethical obligation and with its insistence that the positiv-
ity of law must itself be justified as an expression of what is rightful,
can be viewed as a pure theory of the juridical.
Presumably a non-positivist pure theory would focus not on the
validity of law but on the reasons that are properly in play in the
determination of legal controversy. How are these reasons suitable to
the nexus between the parties? What is the normative nature of that
nexus? What does it mean for these reasons to be coherent? Do they

8
Hans Kelsen, The Pure Theory of Law, tr. Max Knight (1967), 1.
9
Hugo Grotius, De Jure Belli ac Pacis Libri Tres, 2 vols., tr. Francis W. Kelsey (1925), vol. 2,
29–30.
 conclusion 343

have a distinctive structure and distinctive presuppositions? What con-

cepts are the appropriate vehicles for this reasoning? Are the consid-
erations that govern liability systemically autonomous from “alien
elements” (as Kelsen calls them) taken from other disciplines? In
addressing these questions, this book has offered purity without
positivism.
This page intentionally left blank
Index

abstract theory, 298, 313 entanglement of, 204–205

acceptance in unjust enrichment, failing to take the opportunity to
203–209, 234 reject, 210
as consonance with defendant’s incontrovertible benefit see
projects, 204, 208, 216 incontrovertible benefit
as relational notion, 205–208 indirect benefits, 132–138
entanglement and, 204–205 rejected benefits, 204, 205
of benefits, 206, 234 transfer from improver to owner, 261
situations of, 209–217 unremovable benefits, 260
statute of limitations and, 209 Birks, Peter, 81, 107, 108, 144
acquiescence, 212, 216, 245–246 Blake, Attorney-General v. Blake, 158, 161,
acquisition, 23, 24, 34, 159, 223, 224, 162, 166, 169
271, 276, 277, 278, 279, 285 bodily integrity, 49, 53, 89, 142, 145,
adjudication, 3, 277, 305 282, 287
Adras Building Material v. Harlow & Bowen, Lord Justice, 129, 132, 135, 211, 217
Jones, 158, 160, 161, 162, 163, 164, breach of contract, disgorgement of
166–167 gains from, 156–169
aggravated damages, 96, 173, 178 breach of duty, 45, 48, 49, 90, 144, 294
alienation breach of imperfect obligations, 293
of property, 165, 169, 183, 184 Bryant v. Lefever, 327, 331
purported alienation, 127, 130, 163 Buber, Martin, 1
Andrews, Justice William, 46–47, 58
Aristotle, 2, 9, 15–16, 18, 19, 81–82, 83, Calabresi, Guido, 98
84, 116, 190, 232, 322 Calabresi-Melamed framework, 103–107
Atkin, Lord, 38, 39, 40, 41, 42, 57, 58, 73, Canada, 63, 65, 148, 149
75, 79, 80 negligence law, 59
Attorney-General v. Blake, 158, 161, 162, punitive damages in contract law,
166, 169 175–176
automobile accidents, 6 Supreme Court of, 42, 59, 64, 66, 71,
175, 182, 187, 190
Barak, Justice Aharon, 160, 161 Cardozo, Justice Benjamin, 20, 45–46,
bargaining, 104–105 47, 48, 49, 50, 55, 56
benefits causality of the will, 152, 153, 165, 222,
aligning defendant’s projects with 223, 224, 225, 226
benefit bestowed by the causation, 44, 47, 48, 121, 123–124, 303,
plaintiff , 216 304, 316, 317, 327–328, 329, 331
defendant’s acceptance of beneficial causative event, 81, 82, 83, 84, 85, 86,
transfer, 211 98–107, 109, 110, 116
346 index

civil condition, 277, 278, 279, 280, 283, remedial conception, 335–338
284, 285, 292, 295, 296 Stone on, 30, 34
civil liability, 20, 172, 174 structural principle, 10
civil union, 268 substantive principle, 10
claim of necessity, 290, 292 theoretical framework, 35–37
Coase, Ronald, 326, 327, 328, 329, correlative duty, right and, 48–58
330, 331 correlativity, 15–21, 23, 29, 30, 33, 57,
coherence, 32, 37, 41, 43, 312, 339 317, 319, 321, 322, 338, 341–342
common law and Cardozo’s treatment of duty in
and acceptance of benefits, 234 the Palsgraf case, 50
donative intent, 261 in gain-based damages, 118
on improver of property, 243 liability and, 10
individualism, 244 and personality, 11, 35
and Jewish law, 243 of right and duty, 188, 296
community, 223, 225 Stone on, 34
comparative law, 232, 341 courts, as guardian of the system of
compensation, 60, 148 rights, 111
compensatory damages, 122, 125, 150, Craswell, Richard, 334
176, 179, 180 criminal law, 97, 171, 172, 173, 174, 177
concept of right, 266, 267, 274, 288 criminal liability, 172
consequential damages, 96
continuity thesis, 92, 93 damages
contract damages, 148, 149–150, 181, aggravated damages, 96, 173, 178
183, 309 award of, 87–88, 305
contract law calculation of, 135
compensation in, 148 compensatory damages, 122, 125, 150,
punitive damages in, 175–183 176, 179, 180
contract remedies, punishment and consequential damages, 96
disgorgement as, 148–184 contract damages, 148, 149–150, 181,
contracts for sale of land, 167, 224 183, 309
contractual entitlement/right, 149–156, expectation damages, 6, 149, 150, 151,
160, 162, 163, 166, 167, 169, 183, 224 152, 154–155, 156, 169, 183
Cooper v. Hobart, 66–68, 71 gain-based damages, 96, 117–147, 341
corrective justice loss-based damages, 133
conception of remedies, 87–98 nominal damages, 95
definition, 2 punitive damages see punitive damages
focus of, 15–16 restitutionary damages, 126
history of, 9 set at rental value, 138
juridical conception of, 10–11, 13–15, substituted for an injunction, 112–113
17, 26–27, 28, 32 substitutive damages, 95
misconceptions, 10 for the value of using another’s
punishment and, 169–175 property, 137
and punitive damages, 171 deduction, 78
rectificatory function, 16 Deglman v. Guarantee Trust Co. and
rectifying injustice, 84, 172 Constantineau, 219
relationship between distributive Denning, Lord, 69, 70, 72
and, 269 deterrence, 65, 117, 138, 139, 141, 144,
relationship between tort law and, 31 147, 176, 179, 181, 182
 index 347

punishment and, 119 exchange, 192–193

and retribution, 97 equivalence in, 193
detrimental reliance, 53–54, 55 and transfer of value, 195
Devlin, Lord, 145 expectation damages, 6, 149, 150, 151,
Diplock, Lord, 72, 73, 74, 75, 146 152, 154–155, 156, 169, 183
disgorgement, 145, 183 external objects of choice, 223, 275
of gains from breach of contract, external property, 277, 283, 284
156–169 external validation, 29–35
distributive justice, 6, 7, 18–19, 24, 27,
36, 263–264, 265, 269, 270, 296, 336 factual cause, 48
Aristotle’s presentation of, 15, 16 fairness, 10, 16, 18, 36, 37, 97, 113, 229,
relationship between corrective 232, 234
justice and, 269 and coherence, 2, 3, 11
divisions of justice, 270–271 fiduciary, 143–144, 225
Doctrine of Right,The (Kant), 5, 263, 265, fines, 181
270 fire insurance policy, 178
doing and suffering, 20, 25, 26, 29, 30, foreseeability, 61–62
33, 36, 41, 42, 43, 65, 72, 130, 141, foreseeable losses, 56
183, 315, 316, 331 freedom, 280–289
donative intent, 202, 203, 206, 207, 261 innate right to, 272–273
Donoghue v. Stevenson, 38, 45, 79 relationship between one person’s
dualism, monism and, 107–115 freedom and another’s action, 282
duty/duties, 48, 122 Fuller, Lon and Perdue,
correlativity of right and, 188 William, 148–155, 169, 183
general conception of, 38–43
and negligence concepts, 43–48 gain-based damages, 96, 117–147, 341
policy and, 69–79 availability of, 137, 146
and proximate cause, 340 dealing with another’s property and,
relative to rights, 136 125–132, 139
right and, 20 justification of, 131, 143
right and correlative duty, 48–58 Olwell v. Nye and Nissen, 133
violation of a fiduciary duty, 143 gains
duty of care, 48 disgorgement of, 127
two-stage test for, 58–69 through using a stored object, 129
through using an object part of a
economic analysis, 31, 302–306, 326, 334 discontinued business, 129
instrumentalism of, 306–311 Gardner, John, 334–337
economic efficiency, 307 gifts, 218, 221, 243, 244
economic loss, 52, 55, 76 unjust enrichment and, 200–203
efficient breach, 158–159, 160, 169 Goff-Jones principle, 120–125
England, 58, 63, 148, 158, 170, 175 Grotius, Hugo, 290, 291, 292, 293,
entanglement, 204–205, 212–213 295, 342
entitlement, used by Calabresi and
Melamed, 106 Hand, Justice Learned, 84
Epstein, Rabbi Yehiel Michal HaLevi, harm, 51
253–254, 255, 256 Hegel, Georg W.F., 1, 12, 15, 26, 27, 28,
equality, 16 30, 142, 190–191, 199, 302, 322
of innate right, 278–279 Hobhouse, Lord, 161, 166, 167, 169
348 index

Holmes, Oliver Wendell, 47 incontrovertible benefit in, 230–262

Home Office v. Dorset Yacht, 72, 73 individualism, 244
House of Lords, 58, 59, 64, 158, 170, upper and lower hand, 230, 234,
177, 182 235–240, 241, 242, 246, 250, 255, 259
juridical conception, 10–15, 17, 20–22,
I-Thou relation, 1 26–35, 335–339
imperfect obligations, 293–294 justice
improvement Kant’s divisions of, 270–271
of building materials, 252 see also corrective justice; distributive
to buildings, 254–255 justice
existence and condition of improved justness in the transfer of value, 200–203
objects, 255
to non-agricultural properties, 237 Kant, Immanuel, 5–6, 12, 15, 26, 27, 30,
in personam right, 6, 152, 153, 189, 195, 111, 173, 287, 289, 290, 296, 342
200, 206, 227, 339 categories of relation, 152–155, 165,
Kantian conception of, 220–226 167, 168, 223–225
in rem right, 6, 152, 153, 189, 195, 200, classification of rights, 222–225
223, 224, 225, 339 conception of the in personam right,
incontrovertible benefit, 140, 141, 198, 220–226
211–213, 216, 217, 220 on contractual entitlement, 152–156,
conceptions of, in Jewish law, 253–258 160, 163, 169, 183
demise and rebirth of, in Jewish law, distinction between in rem and in
248–253 personam rights, 152, 222–225
expansive view of, in Jewish law, 256 innate right to freedom, 143, 222–223,
in Jewish law, 230–262 270–288, 291–295
restrictive view of, in Jewish law, juridical use of, 5–6
253–254 metaphysics, 5
indemnified injunction, 86, 98, 102, original contract, 280, 283, 284, 286,
103, 106, 108 287, 289, 292, 295, 296
indeterminate liability, 55, 56, 61, 64 postulate of practical reason with
indirect benefits, 132–138 respect to rights, 274–277, 279
injunctions, 94–95, 112–115, 157 postulate of public right, 277
see also indemnified injunction private ownership and common
innate right, 222–223, 280–281, 282 ownership, 291
equality of, 278 property, 270–280, 339
to freedom, 272 public duty to support the poor,
property and, 288 263–270, 284–289, 292
instrumentalist analysis, 302–311 right of sovereign to tax the people,
intent, 173–174 285
intention, 304, 316, 317 universal principle of Right, 271–273,
interdisciplinary study of private law, 276
323–332 Karelitz, Rabbi Abraham, 256–257, 258
interference with external things, 274 Kelsen, Hans, 81, 82, 83, 84, 85, 342

Jewish law, 341 legal education, 297–333

appreciation and expenditure, 236–238 loss, 51, 56
donative intent, 261 wrongful loss, 17, 122, 123, 124, 125,
improver knowingly acting on 127, 128, 129
another’s property, 244 loss-based damages, 133
 index 349

Metaphysics of Morals (Kant), 5, 265 Phillips v. Homfray, 127–128, 129, 131,

misappropriation of another’s 132, 135
property, 145 Philosophy of Right (Hegel), 190
mistaken payments, 213, 214, 215 poles of a relation, and the relation, 1
money policy
as incontrovertible benefit, 213 in Cooper v. Hobart case, 71
non-donative payment of, 212–213 and duty, 69–79
monism, and dualism, 107–115 in Home Office v. Dorset Yacht case, 73
morality, 24, 161, 319 two notions of, 70–77
Pollock, Chief Baron, 204, 216, 259
necessity, 294–295 pollution, 103–107
need, 19, 20, 191–193, 266, 288–289, poor, public duty to support the, 6,
290–295 263–270, 280, 284, 288, 296
negative community, 292, 293, 295 positive community of common
negligence, judicial treatment of duties ownership, 292
of care in, 38 positive law, 5, 83, 84, 85, 182, 270, 312,
negligence law, 20, 38–80, 123, 302, 320, 339, 342, 263
316, 339 positivism, 342–343
ensemble of concepts, 40, 41, 43, 48, practice of law, 297
58, 69, 74, 76, 78, 316, 339–340 attending to the, 313
in Canada, 59 disjunction between university study
indirect benefits and, 135 of law and, 302–311
place of duty in, 43–48 overcoming disjunction between the
in terms of doing and suffering, 33 university study of law and, 310,
and wrongful loss, 123 311–323, 332
neighbor principle, 38, 39, 58 prenatal injuries, 52
Nicholls, Lord, 161, 166 private law, 1–2, 3, 5, 22, 32, 117, 171,
nominal damages, 95 184, 232, 300, 305–306, 309
non-dependence of one’s existence, 271, character of, 311–313
281, 282, 283, 284, 287 consequentist understanding of, 136
non-gifts, 218 corrective justice and, 6, 7
nonfeasance, 22, 23, 29, 51, 136, 339 distributive justice and, 7
normative practice, liability as a, 12, 13, elucidation of the character of, 313–319
16, 17, 19, 30, 31, 32, 37, 341, 342 inner structure of, 314–315
nuisance law, 86, 94, 98–105, 112, 114, instrumentalist analysis, 318
135, 137, 326–331 interdisciplinary study, 323–332
Kantian and Hegelian accounts
occupation of space, 280–282 of, 27
Olwell v. Nye and Nissen, 132–134, normative perspective, 27
135, 137 normative presuppositions, 317
original possession in common, 289–296 norms of, 336–337
purpose of, 7–8
Palsgraf v. Long Island Railroad Co., 45–50, restitution for, 146
50, 58, 122 as a system of rights, 265
perpetuity of a society, 285–286 professional training, 320
personality, 24–29, 33, 34, 35, 269, 317–318, profits, 162
319, 321, 322, 338, 342 property, 6, 274
correlativity and, 11–12, 15 in corrective justice theory of private
Stone on, 30–31 law, 269
350 index

property (continued) remuneration in Jewish law, 245

gain-based damages for dealing with for improver for improving another’s
another’s property, 125–132 property, 241, 259
as a facilitative institution, 126 measures of, 235–240
Kant’s account of, 270–280 for planting a non-suitable field, 250
property-like rights, 142–146 Restatement of Restitution, 132,
proprietary rights, 142, 165, 166, 167 226, 228
proximate cause, 44, 45, 47, 48, 79, 122, restitution, 146, 185, 187, 210, 218, 228,
123, 316, 340 230, 232
proximity, 56, 62–63 for private law, 146
psychiatric injury, 51 of value of services, 219
public law, 263, 272, 305, 308 right(s), 20, 22, 23–24, 25, 89,
public right, 173, 174, 265, 268, 277, 279 110, 116
Pufendorf, Samuel, 290, 292–295 abuse of rights, 113
punishment, 147 acquired rights, 222, 223, 276, 283
civil liability and, 174 to contractual performance, 189
and corrective justice, 169–175 and correlative duty, 48–58, 91
objectives of, 182 correlativity of duty and, 188
punitive damages, 96–97, 145–146, 170, and duty, 20, 183
171, 174, 341 enforced by courts, 110
actionable wrongs, 177, 178, 180, 181 external rights, 276
for breach of contract, 148, 175–183 injury to a, 51
House of Lords criticism of, 182 innate right see innate right
Pure Theory of Law,The (Kelsen), 81, 342 of necessity, 290
purposive activity, 24 over things, 290
in personam, 6, 152, 153, 189, 195, 200,
quality, defect in, 76 206, 220–226, 227, 339
postulate of public Right, 277
rational agency, 12, 26, 27, 28, 30, 32, 33 private law, 1, 34
Rav, 230–232, 234–235, 239–242, property-like rights, 142–146
244–246, 248–250, 256, 259 qualitative restoration of plaintiff ’s
Reid, Lord, 170, 171, 172, 174–175 right, 94
relations, 1 quantitative restoration of plaintiff ’s
relationships, 225 right, 94
remedial conception, 335, 337 reciprocal determination of duties
remedies, 81–116 and, 225
causative event as condition in rem, 6, 152, 153, 189, 195, 200, 223,
of, 98–107 224, 225, 339
conceptions of, 81–86 and remedies, 87–98, 109, 167, 340
continuity thesis, 92, 93 Rosh (Rabbi Asher ben Yehiel), 249–258
corrective justice conception Rule 4, 104, 105
of, 87–98
identity thesis, 91–92 sale of land contracts, 167, 224
limitation thesis, 92, 97 services, provision of, 219
monistic and dualistic approaches Shelfer v. City of London Electric Lighting
to, 86 Co., 112, 115
and pre-injustice right, 88 social cost, 326
rights and, 89, 109, 340 society, perpetuity of, 285–286
 index 351

Spur Industries v. Del E.Webb unjust enrichment, 140, 185–229, 232,

Development Co., 98, 99–102, 103, 234, 340, 341
104, 106 acceptance, 203–217, 234
state of nature, 173, 278 at the expense of another, 197
state, the, 173 change of position, 186, 213–216, 220
perpetuation of, 287 defective transfer of value, 187, 190,
Statute of Frauds, 219 200–203, 227
Stone, Martin, 29–35 development of, 186
strict liability, 43, 47 entanglement, 204–5, 212–213
Sturges v. Bridgman, 328, 331 ensemble of concepts, 340
subjective devaluation, 198, 211, 214–215, in personam right, 152, 53, 189, 195,
220, 231, 243, 257 200, 206, 220–227
subjective intent, 206 incidental benefits, 197–198
substance, 223, 224 incontrovertible benefit, 140, 141, 198,
and causality, 152 211–213, 216–217, 220,
substantive principle, 10 230–262
substitutive damages, 95 gifts and, 200–203
system of rights, 111, 114, 115, 116, 263–296 liability for, 228
money payment, 212–216
Talmud, 230, 231, 235, 239, 240, 241, 251, passing on, 197
252, 259 plaintiff and, 188
tax, 263, 285 subjective devaluation, 198, 211,
state’s right to, 265–266 214–215, 220, 231, 243, 257
to support the poor, 296 and transfer of value, 196–200
tort law, 2, 9, 22, 23, 30, 34, 40, 54, 70, unrequested improvements, 231, 234,
117, 302, 303 235, 259
relationship between corrective remuneration of the improver in
justice and, 31 Jewish law, 241
tort liability, 20, 45 use, 191
tort theory, 9, 34–35 usefulness, 192
tortfeasors, 120, 123, 128, 337
transfer of value, 220, 227–228, 233 validation, external validation, 29–35
defectiveness in the, 221 value
justness in, 200–203 abstraction from particular uses, 192
and transfer of a thing of value, 196 characteristics of, 191
transfers, 194 corrective justice and, 130–131
gratuitous, 194, 195 exchange and transfer of, 195–196
and unjustness, 340 and external objects of choice, 226
trespassers, innocent and willful, 138–142 juridical significance of, 190–196
justness in transfer of, 200–203
unauthorized use, and purported movement of, 197
alienation, 130 in Philosophy of Right (Hegel),
United States of America, punitive 190–192
damages, 175 relationality of, 191
university study of law, 297, 298, 299, transfer of, 194, 220
303–323, 332 and use, 191
overcoming disjunction between the violation of a fiduciary duty, 143
practice of law and, 311–323 virtue and need, 19
352 index

Wagon Mound case, 47–48, 49, 122 willful trespass, treatment of, 141
wayleave rent, 127–128, 132 willfulness, 138–142
welfare, 21–26, 34, 136, 183, 184, 196, workplace injuries, 6
292, 307 wrongdoing
Whiten v. Pilot Insurance, 175–176, 177, and injury, 340
178, 180, 181 innocent and willful, 138–142
will, 208, 209 wrongful gains, 120–121