EXPLORING L AW ’S EMPIRE:

THE JURISPRUDENCE OF
RONALD DWORKIN
Exploring Law’s Empire:
The Jurisprudence of Ronald Dworkin

Edited by
SCOT T HERSHOVITZ

3
 This book has been printed digitally and produced in a standard specification
in order to ensure its continuing availability

1
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
© the several contributors 2006

The moral rights of the author have been asserted

Reprinted 2012

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 book in any other binding or cover
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
Data available
ISBN 978-0-19-954614-5
 Preface

The plan for this book was simple: ask leading scholars in jurisprudence and con-
stitutional law who are sympathetic to Ronald Dworkin’s work to write essays
about it. The invitation was open-ended. Authors were asked to contribute to
Dworkin’s project by offering new arguments for old conclusions, by offering new
arguments for new conclusions consistent with Dworkin’s project, by filling gaps,
or by criticizing Dworkin’s work to show how it could be improved. We hoped
that in putting together a collection of such essays, and by asking Dworkin to
respond to them, we would enrich the discourse to which Dworkin has con-
tributed so much.
With such an open-ended invitation, we were bound to get an eclectic mix of
essays, and that was the point. The scope of Dworkin’s contributions to our
understanding of law is vast. He has made seminal contributions to debates over
doctrinal and theoretical issues in American constitutional law, to debates over
topics such as civil disobedience and the obligation to obey the law, and, of course,
to our understanding of the nature of law and its relationship to morality. And this
list leaves unmentioned much of Dworkin’s groundbreaking work in political and
moral philosophy, which has had a significant impact on legal discourse. We
therefore set authors loose with the charge that they should engage with whatever
aspect of Dworkin’s work they saw fit, not simply to honor it, but to increase our
understanding of it and push it forward. In that way, we hoped to produce a book
that would explore the full richness of Dworkin’s contributions to our under-
standing of law’s empire.
While there is no overall theme to the book that resulted, a number of themes
recur throughout. Justice Breyer’s introductory essay recounts a discussion
Dworkin led at a conference of international judges. Justice Breyer goes on to
describe how Dworkin’s work has influenced the way judges on constitutional
courts around the world conceive of their role in democratic societies.
Dworkin’s constitutional theory is further examined in essays by Chris
Eisgruber, Jim Fleming, and Rebecca Brown. Eisgruber and Fleming address the
role of history in Dworkin’s moral reading. Eisgruber argues that history has a
more significant role in Dworkin’s account than many recognize. But he argues
that the moral reading as Dworkin presents it does not provide a fully satisfactory
account of constitutional interpretation because it does not adequately explain
how American-style constitutionalism, which requires a legislative supermajority
to overrule a constitutional decision, facilitates government based on principles of
justice. Fleming’s essay examines the relationship between Dworkin’s moral read-
ing and both process-oriented and originalist approaches to constitutional inter-
pretation. Brown’s essay explores the influence that Dworkin has had on the
vi Preface

development of constitutional law in the United States. She focuses on the ways in
which Dworkin’s constitutional theory helped scholars come to terms with Brown
v. Board of Education, and she traces Dworkin’s influence on contemporary consti-
tutional law.
Essays by Susan Hurley and Scott Hershovitz transition the book from explor-
ations of Dworkin’s constitutional theory to explorations of his legal theory more
generally. Hurley argues for the somewhat counterintuitive view that the content
of the law is not simply a function of actual cases which courts have decided, but is
also a function of hypothetical cases which courts have not and may never be called
upon to decide.¹ She develops a coherence account of legal reasoning similar to
Dworkin’s, and she addresses objections Kenneth Kress has raised to such
accounts. Hershovitz’s essay tackles a different aspect of legal reasoning—stare
decisis. He argues that accounts of stare decisis that justify the practice of follow-
ing precedent by appealing to the values of efficiency and fairness are unsuccessful.
Stare decisis, he urges, is best understood as a practice which promotes integrity in
adjudication.
Integrity is also a subject of investigation in three further essays. Dale Smith
considers problems in Dworkin’s account of checkerboard solutions, policies that
Dworkin claims are defective due to a lack of integrity. Smith aims to clarify the
ways in which checkerboard solutions are defective. Jeremy Waldron considers
objections the Critical Legal Studies movement raised to Dworkin’s theory of law
as integrity and the adequacy of Dworkin’s response. Waldron suggests that there
is a tension between the role of integrity in Dworkin’s account of legal practice and
his constructivist approach to legal interpretation. Finally, Stephen Perry con-
siders integrity as it relates to the problem of political obligation. His essay clarifies
the relationship between Dworkin’s account of political obligation and his
account of the nature of law.
This volume ends where Dworkin’s career began—with questions about the
nature of law. John Gardner’s essay considers whether law has a distinctive pur-
pose, and the role that such a purpose plays in Dworkin’s account of law. Gardner
argues that some of Dworkin’s commitments in this regard push him towards legal
positivism. Mark Greenberg, on the other hand, argues for a broadly Dworkinian
view of the nature of law, on which value facts are among the determinants of the
content of the law. Greenberg’s approach is novel because it focuses on law’s meta-
physics—on, as he puts it, “how facts make law”—rather than on how judges and
lawyers come to know what the law is. Greenberg’s argument extends across two
essays. The first essay was originally published in Legal Theory;² the second essay,

¹ Hurley’s essay was previously published as Coherence, Hypothetical Cases, and Precendent, 10(2)
OXFORD JOURNAL OF LEGAL STUDIES 221 (1990). It is reprinted here by kind permission of Oxford
University Press.
² Greenberg’s essay was previously published as How Facts Make Law, 10 LEGAL THEORY 157
(2004). It is reprinted here with permission.
 Preface vii

which is published here for the first time, responds to a family of objections to the
argument of the first essay and further develops that argument in the course of
criticizing positivist theories commonly associated with H. L. A. Hart.
The volume concludes with responses from Ronald Dworkin. Dworkin com-
ments on each essay, but focuses his attention on those that challenge views he has
defended. His responses are in each instance illuminating and they contain valu-
able statements of his views across a range of topics.
Each of the contributors to this book played an important role in its produc-
tion, and I am grateful to all of them for the work they took on. Three deserve spe-
cial thanks: Chris Eisgruber, John Gardner, and Mark Greenberg. Chris hosted a
conference at Princeton University in September 2004, at which many of the
essays in this book were presented. Mark and John have been incredibly generous
with their support and encouragement. This project would not have come to
fruition without their help, and I appreciate all that they did. I also appreciate the
support—and patience—of John Louth and Gwen Booth, our editors at Oxford
University Press.
Collectively, our deepest gratitude is reserved for Ronnie. All of us know that
our work would be less insightful and less exciting had he not taught us so much.
We appreciate his willingness to join our project and engage with our essays. We
didn’t set out to write a conventional festschrift, but we hope that in developing
and criticizing Ronnie’s work, we have honored him nonetheless.
SAH
Washington, D.C.
January 2006
 List of Contributors
Stephen Breyer is an Associate Justice of the United States Supreme Court.
Rebecca L. Brown is the Allen Professor of Law at Vanderbilt Law School.
Ronald Dworkin is the Frank Henry Sommer Professor of Law at New York University
School of Law and the Bentham Professor of Jurisprudence at University College London.
Christopher L. Eisgruber is the Provost of Princeton University and the Laurance S.
Rockefeller Professor of Public Affairs in the Woodrow Wilson School and the University
Center for Human Values.
James E. Fleming is the Leonard F. Manning Distinguished Professor of Law at Fordham
University School of Law.
John Gardner is the Professor of Jurisprudence at the University of Oxford.
Mark Greenberg is Acting Professor of Law and Assistant Professor of Philosophy at the
University of California, Los Angeles.
Scott Hershovitz is a faculty-designate at the University of Michigan Law School. He cur-
rently practices law in Washington, D.C.
S. L. Hurley is Professor of Philosophy at the University of Bristol and a Fellow of All Souls
College, Oxford.
Stephen Perry is the John J. O’Brien Professor of Law and Professor of Philosophy at the
University of Pennsylvania Law School.
Dale Smith is a lecturer in the Faculty of Law at Monash University, Australia.
Jeremy Waldron is University Professor at New York University Law School.
 Introduction: The “International”
Constitutional Judge
Stephen Breyer

In July 2000 a small group of jurists from several western nations, mostly judges,
met for a week in Southern France to discuss the judge’s role in our increasingly
similar, interdependent societies. Our focus was upon constitutional judges,
including judges of international courts. We did not try to reach conclusions. We
hoped to learn through conversation.¹
Professor Dworkin led a discussion about the judge’s moral authority, a matter
we had entitled, with some irony, “the secular papacy.” His paper tied the discus-
sion to a more basic view of constitutional law, which he set forth in three theses.²
First, law, particularly constitutional law, inevitably embodies standards that
require a judge to make moral decisions. Administrative law, for example, sets aside
agency actions that are “unreasonable,” “disproportionate,” or an “abuse of power.”
Constitutional law protects “liberty,” including “freedom of speech,” “freedom of
the press,” and “freedom from unreasonable searches and seizures.” Judges cannot
apply these standards in difficult cases without revealing their own moral views, at
least those that concern proper political structure (itself a matter of political moral-
ity). The thought that judges must rely upon their own “political convictions”³
makes judges uncomfortable. But there is no satisfactory alternative.
Secondly, it is not “undemocratic”⁴ for unelected officials such as judges to pos-
sess the power to enforce these standards. This is not because “the people . . . do
not object to” the judicial practice “and, from time to time in different ways
endorse it.”⁵ Nor is it because “democracy compromised” is a “better form of gov-
ernment” (say, because it produces “a more just community by protecting the
rights of minorities”).⁶ Rather, it is because the best “conception” of “democracy”⁷
is a conception under which “majority rule is fair;” and majority rule is “fair” only
when certain conditions are met. Where these conditions are realized, then
democracy in its best sense is realized. Insofar as the judge helps to assure that
those conditions are met, the judge’s work helps to support democracy itself.

¹ See generally, Judges in Contemporary Democracy: An International Conversation

(R. Badinter & S. Breyer eds., 2004). ² Id. at 67–79.
³ Id. at 71. ⁴ Id. at 73. ⁵ Id. ⁶ Id. at 73–74. ⁷ Id. at 74.

Introduction: The “International” Constitutional Judge. Stephen Breyer.

© Oxford University Press 2006. Published 2006 by Oxford University Press.
2 Stephen Breyer

Thirdly, societies increasingly have turned to judges to assure that these con-
ditions are met because they are increasingly uncertain that parliaments or “the
people themselves”⁸ will do so unaided. Before the “short and terrible Twentieth
Century”⁹ many thought that parliaments alone might do so by properly bal-
ancing the “general interest” with “fairness to individuals.”¹⁰ That is to say,
legislatures might “at once” be “the voice and the conscience of the people,”
pursuing “the general interest, but subject to ancient constraints of fairness and
decency to individual citizens.” Some thought this because they held the aristo-
cratic notion that parliament embodied the wisdom of a “political elite”¹¹ with
the best interests of the nation at heart. Others thought this because they held
the romantic notion that parliament embodied the majority’s “decent and hon-
orable general will.”¹² But given a Twentieth Century history that discredited
both notions, societies now seek institutionally to separate the “social acceler-
ator” from the “moral brake.”¹³ And they have given judges ever greater authority
to operate the latter.
What controls the judge when he or she exercises this increased authority? The
control lies not only in the general ends they pursue—assuring the conditions of a
democracy in its best sense, namely, fair majority rule—but also in the specific
means they employ, in particular the process of judicial reasoning. This process
encompasses the judicial “responsibility for articulation,” which constrains judges
to “do nothing that they cannot justify in principle, and to appeal only to prin-
ciples that they thereby undertake to respect in other contexts as well.”¹⁴
These three theses require elaboration, a matter that I leave to Professor
Dworkin and to others to pursue elsewhere. Here, I shall simply note that the
judges at the conference found that the theses, though set forth in summary form,
usefully informed a wide-ranging discussion, touching upon subjects as disparate
as judicial activism, court/media relations, and international human rights tri-
bunals. I shall also point to several aspects of Dworkin’s theses that may help to
explain their broad appeal to practical judges, not themselves legal philosophers,
at work in different national and international settings.
For one thing, Dworkin accurately described the role that reason plays in the
judicial process. He said that judges “owe us . . . an argument that meets two condi-
tions. First they have to believe it; they must offer it in good faith as an argument.
Second, it has to connect what they do in a particular case with something more
general and basic.”¹⁵ He added that “in the end, what argument a judge finds con-
vincing will depend on that judge’s more general attitude and convictions, and no
one can demonstrate to those who do not share these attitudes and convictions that
they are inescapably right. But they must seem right, after open argument and seri-
ous reflection, to those who rely on them.”¹⁶ As a judge, I would say that that is how
it seems to us. And I would add that in a modern era where we must apply old values

⁸ Id. at 75. ⁹ Id. at 77. ¹⁰ Id. ¹¹ Id. at 76. ¹² Id.

¹³ Id. at 70. ¹⁴ Id. at 78. ¹⁵ Id. at 250. ¹⁶ Id. at 250–251.
 Introduction: The “International” Constitutional Judge 3

to new, highly complex, technically-shaped circumstances, the need to base judicial

decisions upon clear reasoning seems more important than ever.
Others at the conference elaborated upon this basic idea. Some pointed out
that the judicial process also provides a forum for serious reasoned argument, not
only in the special judicial sense just mentioned, but also as providing “a place of
debate, of public discussion and decision.” Unlike a modern legislature the mem-
bers of which do not usually “debate” each other (but rather often speak to show
their own side, via media reports, how strongly they support its views), a “judicial
forum” helps to organize “differing fluctuating opinions for and against a particu-
lar matter. It is a place where lawyers representing differing interests publicly have
to confront contrary opinions.”¹⁷ It permits “a continuous exchange of opinions,”
thereby helping “the public debate along.”¹⁸
For another thing, Dworkin’s definition of judges’ constitutional objectives,
namely the identification of conditions essential to a fair democratic process,
imposes a significant constraint upon the constitutional judge’s exercise of lawful
authority. Applying this basic principle, judges should not interpret highly general
constitutional language, such as the word “liberty” in a Constitution, so broadly
that it encompasses “all the values that a decent society should permit”¹⁹ or
answers all the complex questions that society confronts (say, whether nuclear
power is too dangerous to permit a legislature to adopt it). Democracy demands
that legislatures, not courts, make the trade-offs among such conflicting, desirable
objectives as “security and efficiency or environmental amenities.”²⁰ Legislators,
not judges, will ordinarily decide how to rank both desirable objectives and un-
desirable behaviors.
The discussion also showed how these two elements of Dworkin’s theses—his
substantive description of what constitutional judges at a basic level are seeking to
do and his procedural insistence upon the use of reason—could inform the debate
about new “international” or “trans-national” tribunals. Can citizens of different
nations reach agreement about which conditions are necessary for modern
democracy to flourish, or are there too many culturally based disagreements to
maintain confidence in a court with jurisdiction over such matters? To what
extent should we circumscribe jurisdiction to produce agreement, say limiting the
jurisdiction of international human rights courts to the most serious kinds of vio-
lations? Is there an inverse relation between agreement about basic democratic
conditions and insistence upon political control? Where jurisdiction is broad, will
citizens of different nations increasingly insist upon control over the nomination
of some, or all, of the individual judges? How detailed should we make the defini-
tions of individual “international” human rights crimes? Should it be up to legisla-
tures or to courts to decide, for example, whether rape is “torture” or simply a
serious crime? Dworkin’s theses do not answer these questions, but they do pro-
vide a framework that helps to guide us towards proper answers.

¹⁷ Id. at 97. ¹⁸ Id. at 96–97. ¹⁹ Id. at 49–50. ²⁰ Id. at 50.

4 Stephen Breyer

The most important characteristic of Dworkin’s theses and their related

philosophy, however, concerns their relation to the matter of the approach a judge
should take to interpreting the highly general language of the Constitution.
American judges must apply the Constitution in ways that vindicate, in today’s
highly complex society, the basic values that the Founders enunciated more than
200 years ago. To treat the Constitution as if it contains a laundry list of the
Founders’ specific detailed expectations is to doom to extinction its more basic and
general democratic objectives. But to do the contrary, to read the scope and
content of that eighteenth-century language as evolving in light of modern cir-
cumstances risks the judges’ assumption of the legal authority to impose their own
subjective views of what is “good” upon the Nation. Some judges and scholars
believe that, to avoid the latter, judges should approach the Constitution with
what one might call “literalist,” “traditionalist,” or “originalist” attitudes. Are such
approaches necessary to keep subjective judicial decision-making in check?
Dworkin’s views suggest they are not. His theses point to standards, based upon
important general values embodied in a Constitution that establishes, if not the
“best” form, at least a very good form, of democracy. Defining judges’ constitu-
tional objectives as identifying the conditions necessary for a workable democracy
limits the judicial role and guides judges’ discretion. Dworkin’s views also suggest
that through the use of reason a constitutional judge can produce an opinion that
is transparent, permitting an informed reader to criticize the opinion in light of
substantive standards. The result is a system that allows the public, and the judge,
to know when an opinion goes too far, thereby keeping judicial subjectivity in
check. The standards are not totally objective. Indeed, they permit, perhaps some-
times require, judges to invoke their own notions of political morality. But even
then they permit informed public criticism. And that is the point. Criticism is
perhaps the best way to keep subjective judicial decision-making in check. I doubt
that those who advocate literal or “traditionalist” or “originalist” theories can do
any better.
In this sense Dworkin’s theses are highly practical. They cast light on a question
that divides the American judiciary as never before. They, like the bulk of his con-
stitutional thought, have had considerable impact upon American constitutional
law. And I am not surprised they generated stimulating and helpful discussion in
an international forum as well.
 1
Should Constitutional Judges be
Philosophers?
Christopher L. Eisgruber*

I. Introduction

Ronald Dworkin has urged American judges and lawyers to embrace the “moral
reading of the [United States] Constitution.” The moral reading insists that the Bill
of Rights and the Fourteenth Amendment “invoke moral principles about political
decency and justice.”¹ Some critics worry that the moral reading emphasizes
morality too much. They believe that it gives short shrift to text and history and
that it calls upon judges to become philosophers instead of lawyers.² Dworkin, in
return, has contended that his theory allows ample room for historical reasoning.
Yet, when Dworkin interprets the Constitution, the principles that he finds in it are
very abstract ones—such as the principle that “government must treat everyone as
of equal status and with equal concern”—and his analysis of them is thoroughly
philosophical. Historical concerns rarely figure in his arguments.
Why is Dworkin’s “moral reading of the Constitution” so moral—that is, why is it
concerned so much with moral issues and so little with other considerations, includ-
ing historical ones? Can Dworkin’s “moral reading” in fact accommodate historical
argument and other traditional forms of legal reasoning, or does it require constitu-
tional judges to become philosophers? If the latter, is that requirement defensible?
This essay examines these questions. It concludes that the abstract, philosophical
character of Dworkin’s constitutional interpretation is partly—but only partly—a
matter of Dworkin’s intellectual style. The general theoretical parameters of his

* For helpful comments on earlier drafts, I would like to thank Ronald Dworkin, Mark
Greenberg, Lewis Kornhauser, Liam Murphy, Larry Sager, the Faculty and Fellows of Princeton’s
University Center for Human Values, and the NYU Colloquia in Constitutional Theory and in Law,
Philosophy, and Social Theory. The Filomen D’Agostino and Max E. Greenberg Faculty Research
Fund at the New York University School of Law and the Woodrow Wilson School at Princeton
University provided financial support for this research.
¹ R. M. Dworkin, Freedom’s Law: The Moral Reading of the Constitution 2 (1996).
² An example of such criticism is M. McConnell, The Importance of Humility in Judicial Review:
A Comment on Dworkin’s Moral Reading of the Constitution, 65 Ford. L. Rev. 1269 (1997).

Should Constitutional Judges be Philosophers? Christopher L. Eisgruber.

© Oxford University Press 2006. Published 2006 by Oxford University Press.
6 Christopher L. Eisgruber

“moral reading” permit interpreters with different styles to invoke history more
often than Dworkin himself does. This preliminary conclusion is consistent with
Dworkin’s claim that lawyers who want to make historical arguments should do so
within the umbrella of the moral reading rather than in opposition to it.
Dworkin’s emphasis upon abstract moral principle is not, however, wholly
attributable to his intellectual style. This essay’s second conclusion, and its most
important claim, is that Dworkin’s arguments about moral principle and the
Constitution presuppose an unarticulated, controversial theory about the purpose
of written constitutions and super-majoritarian amendment procedures. They pre-
suppose, in particular, a theory that explains why people might adopt a constitu-
tion with super-majoritarian amendment procedures if they aim to accommodate
evolving judgments about abstract standards of justice. Some of Dworkin’s critics
deny that such an explanation is possible. They believe that the Constitution is
hard to amend because its purpose is to entrench past practices against revision on
the basis of subsequent judgments by later generations. Lawyers and judges who
hold this view about the Constitution sometimes assume that they should interpret
ambiguous constitutional provisions in ways that increase the influence of the past
upon the present. They believe that they are thereby faithful to what they take to be
the Constitution’s chief purpose—namely, preventing change. As it turns out,
there exist reasons to reject this past-oriented view of the Constitution’s purpose.
These reasons, however, are not yet part of the case that Dworkin has offered on
behalf of his position, and they depend on considerations different in kind from
those upon which he typically relies (they depend, in particular, upon pragmatic
judgments about political institutions). Unless supplemented by such reasons (or
some comparable theory of the Constitution’s purpose), Dworkin’s theory of
constitutional interpretation is significantly incomplete.

II. History Within the Moral Reading

Almost three decades ago, Dworkin called for “a fusion of constitutional law and
moral philosophy.”³ His subsequent work has carried out that project. When
Dworkin offers views about specific constitutional issues, moral philosophy takes
center stage. For example, when Dworkin analyzes the Equal Protection Clause, he
says that there are only two possible interpretations of it. One possibility is that the
Clause merely requires government to honor the terms of its laws, whatever those
laws may say. If the law prohibits theft, then the police must enforce that law against
anybody who steals, without regard to the race of the culprit or victim. This princi-
ple is a very weak one. It does not preclude the government from writing racially dis-
criminatory terms straight into its laws; it merely prohibits the executive and judicial
branches from discriminating when the legislature has not authorized them to do so.

³ R. M. Dworkin, Taking Rights Seriously 149 (1977).

 Should Constitutional Judges be Philosophers? 7

Dworkin says that precedent forecloses this minimialist interpretation of the Equal
Protection Clause. I doubt that anybody would disagree. But, according to
Dworkin, “[o]nce that much is conceded, . . . then the principle must be something
much more robust, because the only alternative, as a translation of what the framers
actually said in the equal protection clause, is that they declared a principle of quite
breathtaking scope and power: the principle that government must treat everyone
as of equal status and with equal concern.”⁴ And just like that we are off to the
philosophical races, free (if not compelled) to consult liberal political theory in order
to determine the meaning of “equal status and concern.”
Indeed, Dworkin says that, on his approach to the Constitution, it “seems
unlikely that anyone who believes that free and equal citizens would be guaran-
teed a particular individual right will not also think that our Constitution already
contains that right, unless constitutional history has decisively rejected it.”⁵
Moreover, “since liberty and equality overlap in large part,” the Equal Protection
Clause and the Due Process Clause will “each be comprehensive in that same
way”—comprehensive enough, that is, to embrace every right appropriate to free
and equal citizens. One clause—or indeed, either of two clauses—will suffice to
answer every civil liberties need! Dworkin’s argument seems to make not only his-
tory, but much of the constitutional text (including the Bill of Rights, which sin-
gles out specific liberties for constitutional protection), superfluous.
It is thus easy to see why critics accuse Dworkin of eliding the distinction between
constitutional law and moral philosophy. Nevertheless, as James E. Fleming has
correctly argued, Dworkin’s approach to constitutional interpretation can accom-
modate a great deal of historical argument.⁶ To begin with, Dworkin does not
claim that every constitutional provision articulates a moral principle. On the
contrary, he points out that some provisions (he mentions the Third Amendment,
which prohibits the government from quartering troops in homes during peace-
time, and the Presidential Qualifications Clause, which requires that the President
be thirty-five or older) do not.⁷ Moreover, even with regard to provisions that do
state moral principles, Dworkin permits interpreters to use history when deciding
which principle is named by the provision. For example, Dworkin says that when
construing the Equal Protection Clause, judges should remember that
“Congressmen of the victorious nation” were “trying to capture the achievements
and lessons of a terrible war” and so likely enacted an expansive moral principle
rather than a limited one.⁸
In addition, Dworkin maintains that judges owe allegiance to precedent as well
as to constitutional principle. The obligation to fit past decisions limits the ability
of judges to honor their moral convictions about, for example, the principle
embodied in the Equal Protection Clause. Thus, as we have seen, Dworkin con-
tends that precedent prevents judges from reading the Clause so narrowly that it
⁴ Dworkin, Freedom’s Law, supra note 1, at 10. ⁵ Id. at 73.
⁶ J. E. Fleming, Fidelity to Our Imperfect Constitution, 65 Ford. L. Rev. 1335 (1997).
⁷ Dworkin, Freedom’s Law, supra note 1, at 8. ⁸ Id. at 9.
8 Christopher L. Eisgruber

would stipulate “only the relatively weak principle that laws must be enforced in
accordance with their terms, so that legal benefits conferred on everyone . . . must
not be denied, in practice, to anyone.”⁹ He also maintains that precedent prevents
judges from reading the Clause so broadly that it would require equality of wealth.¹⁰
Finally, Dworkin’s theory permits history to play a role internal to moral and
political argument. History might enrich political theory in at least three ways:
with examples that stimulate reflection; with data about institutional perform-
ance and human behavior; and with rhetorical resources that enhance the
persuasive power of moral argument. Dworkin does not discuss connections of
this kind; perhaps he believes that political theorists have little to learn from
history. Yet, that belief, if indeed Dworkin holds it, is not a necessary part of his
view of the Constitution. If we believe that moral principles are best identified and
applied through historical reflection or argument, then Dworkin’s moral reading
of the Constitution will compel us to consult history.
This last possibility bears emphasis. Some influential defenses of historical
argument about the Constitution depend on claims about policy-making in gen-
eral, not about constitutional law in particular.¹¹ These claims, which sometimes
operate under the banner of “pragmatism,” suggest that we should rely on
contextual arguments about institutions and history whenever we make policy
decisions, not just when we are interpreting a written Constitution. Dworkin has
argued vigorously on behalf of abstract moral theory and against pragmatism and
its cousins.¹² Nevertheless, those arguments are distinct from his defense of the
moral reading. If one believes that historical reasoning of some sort is in general
the best way to make practical moral judgments, then one will also believe that
historical reasoning is the key to applying the moral principles identified by
Dworkin’s “moral reading.”

III. What the Moral Reading Entails

If the moral reading can in principle accommodate historical and structural

argument, why is Dworkin’s own application of the theory so heavily dependent
upon moral philosophy? Critics might suppose that, whatever the capacity of
the moral reading to accommodate historical argument, it must somehow stack
the deck against historical and other non-philosophical forms of reasoning.
Dworkin, after all, names the conception “the moral reading,” not “the historical
reading” or “the pragmatic reading.” It appears no accident that, in practice, the
reading recommends something like moral philosophy. Dworkin acknowledges
that many constitutional scholars would prefer a middle ground between his
philosophical approach to the Constitution and the rigid historicism of strictly
⁹ Dworkin, Freedom’s Law, supra note 1, at 9. ¹⁰ Id. at 11.
¹¹ See, e.g., L. Kramer, Fidelity to History—and Through It, 65 Ford. L. Rev. 1627 (1997).
¹² See especially R. M. Dworkin, In Praise of Theory, 29 Ariz. St. L.J. 353 (1997).
 Should Constitutional Judges be Philosophers? 9

originalist interpretive protocols.¹³ Dworkin, however, has taken a strong line

against such approaches, claiming that a third way between the moral reading and
strict originalism is not only unattractive, but entirely unavailable.¹⁴
This surprising claim turns out to be correct, but only because the moral read-
ing’s entailments are more modest than its provocative name would suggest.
Dworkin’s moral reading of the Constitution boils down to two propositions: (1)
the Constitution’s meaning depends upon what the framers intended to say, not
upon how they expected the Constitution to be applied;¹⁵ and (2) when the
framers invoked abstract moral language, they used it in its ordinary sense, as
referring to abstract moral principles.¹⁶ These propositions yield an important
result: as we shall see, they rule out a common form of originalist argument. But
the moral reading makes the Constitution less thoroughly moral than one might
suppose, because it tells us little, if anything, about which moral principles the
framers invoked.
The moral reading’s first proposition depends upon the idea that there is a
difference between the meaning of a concept and our views about how best to
apply the concept.¹⁷ It is easy to construct examples that exhibit the validity of this
distinction. Suppose somebody tells you, “You should respect equality—by which
I mean you should respect equality itself, not just my theory about it. My theory
of equality suggests that affirmative action policies are wholly consistent with the
ideal of equality, but I might be mistaken about that view, and, if so, then you
should oppose affirmative action.”¹⁸ You might find it implausible that anybody
would utter such a speech (or you may think the speech unbearably pompous).
But the speech is not unintelligible: it is perfectly possible for a speaker to distin-
guish between equality itself and her view of equality. She can do so because there
is a difference between what “equality” means and her view (or anybody else’s
view) of how equality applies in particular cases.
The speaker’s views about equality thus do not define the meaning of the
concept she invokes. Her intentions nevertheless matter to the content of her
instructions. In particular, it matters that she intended to invoke the concept we
commonly call “equality” and not some other concept. We would not disrespect
her instructions if we rejected her views about whether affirmative action is

¹³ Dworkin, Freedom’s Law, supra note 1, at 14–15. ¹⁴ Id.

¹⁵ Id. at 13 (“The moral reading insists that the Constitution means what the framers intended to say”).
¹⁶ Id. at 7 (“Many . . . clauses are drafted in exceedingly abstract moral language . . . . According to
the moral reading, these clauses must be understood in the way their language naturally suggests: they
refer to abstract moral principles and incorporate these by reference, as limits on government’s
power”). Id. at 2 (“Most contemporary constitutions declare individual rights in very broad and
abstract language . . . . The moral reading proposes that we all . . . interpret and apply these abstract
clauses on the understanding that they invoke moral principles about political decency and justice”).
¹⁷ On the distinction between the “meaning” and “application” of concepts, see M. D. Greenberg
and H. Litman, The Meaning of Original Meaning, 86 Geo. L.J. 569, 586–591 (1998).
¹⁸ This example, of course, tracks Dworkin’s famous example from Taking Rights Seriously,
where he discusses a father who wants his children to honor the best conception of fairness, which
might differ from his own conception. Dworkin, Taking Rights Seriously, supra note 3, at 134.
10 Christopher L. Eisgruber

consistent with equality, but we would indeed do so if we simply substituted some

other concept—such as “color-blindness” or “libertarianism” or “good health”—
for “equality.” In Dworkin’s terminology, we would then be ignoring what the
speaker intended to say, rather than merely her intentions, hopes, or expectations
about the consequences her instructions would have.
Once we understand the distinction between the meaning of a concept and our
views about how it should be applied, the moral reading’s first proposition quickly
follows. The proposition states that if the framers intended to incorporate a par-
ticular concept in the Constitution (such as “equality”), then the Constitution
requires us to respect that concept, not the framers’ expectations about how the
concept would be applied. It is hard to see why anybody would take the opposite
view. Certainly respect for the framers would not recommend doing so—the
framers, by hypothesis, intended to refer to the concept, not their expectations,
in the Constitution. Nor is there any apparent reason of political justice, unrelated
to respect for the framers, that would justify allowing their expectations to govern
at the expense of the concept that they intended to incorporate into the
Constitution.¹⁹ Indeed, originalists generally concede this point. For example,
Justice Antonin Scalia writes that “it is simply incompatible with democratic
government, or indeed, even with fair government, to have the meaning of a law
determined by what the lawgiver meant, rather than by what the lawgiver
promulgated.”²⁰
The moral reading’s first proposition tells us nothing, however, about whether
the framers actually intended to invoke “equality” or any other moral concept in
the Constitution (it tells us only that if they did so, then the Constitution’s mean-
ing depends upon the meaning of that concept, not upon the framers’ expecta-
tions about how it would be applied). The second proposition, by contrast,
addresses what the framers said. It maintains that they sometimes used abstract
moral concepts in the Constitution, and it insists that they used those concepts in
their ordinary sense—which is to say, they used them to refer to abstract moral
ideas. Dworkin’s argument for this point is simple. He says, “The Framers were
careful statesmen who knew how to use the language they spoke. We cannot make
good sense of their behavior unless we assume that they meant to say what people
who use the words they used would normally mean to say.”²¹
Is that simple argument sufficient to establish Dworkin’s claim? Keith
Whittington contends that it is not. Whittington points out that people sometimes

¹⁹ Professor Michael Moore has argued that neither the framers’ intended meaning nor their
expectations should determine what the Constitution means. M. S. Moore, Justifying the Natural
Law Theory of Constitutional Interpretation, 69 Ford. L. Rev. 2087, 2096–2099 (2001). On Moore’s
view, Dworkin is too much concerned with history and framers’ intention. I will put this interesting
view to one side for purposes of this essay, where my project is to assess the charge that Dworkin has
been too little interested in history.
²⁰ A. Scalia, A Matter of Interpretation: Federal Courts and the Law 17 (A. Guttman,
ed., 1997).
²¹ R. M. Dworkin, The Arduous Virtue of Fidelity: Originalism, Scalia, Tribe, and Nerve, 65 Ford.
L. Rev. 1249, 1253 (1997); see also Dworkin, Freedom’s Law, supra note 1, at 8–9, 76.
 Should Constitutional Judges be Philosophers? 11

use words in unconventional ways, so that meaning depends upon “context.”²² He

suggests, by way of example, that a boss might speak in a shorthand that is under-
stood by his subordinates.²³ Whittington concludes that it is at least partly an
“empirical question” whether the framers used words in their ordinary sense, and
that Dworkin’s theory is unconvincing because it fails to adduce any relevant
historical evidence.²⁴
Whittington is half right: the question whether the framers used words in their
ordinary sense is indeed partly “empirical” and dependent upon “context.”²⁵ But
to characterize the question as empirical is not to say that it depends upon a
detailed excavation of the legislative record or the framers’ expectations. The deci-
sive piece of evidence about the “context” (to use the term Whittington favors) of
the framers’ words is the uncontested fact that the framers were engaged in a
national constitution-making project. That project is different from a conversa-
tion between a boss and his employees, or between other people who know one
another personally and so have developed communicative shorthands. Except in
very rare circumstances, it would be both impractical and unattractive for consti-
tution-makers to use words in any but their ordinary sense. The framers of the
American constitution were a large group—at a minimum, they included all the
ratifiers as well as all the drafters. It is hard to imagine how so many people in so
many places could have communicated effectively if they were not using words in
their ordinary sense. Even aside from this practical problem, principles of fair
notice require that laws either use words in their ordinary sense or at least clearly
indicate departures from it. It would be possible, in principle, to take the latter
route: the framers might have stipulated that they were using certain words in a
special way—as “terms of art,” for example. But we would expect that declaration
to be very prominent, not just in the drafting process, but also in the ratifying con-
ventions, and perhaps in the text of the Constitution itself.
Perhaps Whittington, or somebody else, will be able to find widespread state-
ments of this kind, in which the framers explicitly note that they are assigning
exceptional meanings to words used in the Constitution. But until somebody
comes forward with such statements, the available empirical evidence cuts over-
whelmingly in Dworkin’s favor. There is, as he says, no way to “make good sense of
[the framers’] behavior unless we assume that they meant to say what people who
use the words they used would normally mean to say.”²⁶ Not surprisingly, many
originalists have endorsed something very much like the moral reading’s second
proposition. For example, Judge Robert Bork has written that “[w]hen lawmakers
use words, the law that results is what those words ordinarily mean.”²⁷
²² K. Whittington, Dworkin’s “originalism”: The Role of Intentions in Constitutional Interpretation,
62 Review of Politics 197, 213–214 (2000). ²³ Id. at 215–216.
²⁴ Id. at 214, 216.
²⁵ Dworkin takes the same view. “[W]e must know something about the circumstances in which a
person spoke to have any good idea of what he meant to say in speaking as he did.” Dworkin,
Freedom’s Law, supra note 1, at 8.
²⁶ Dworkin, Arduous Virtue of Fidelity, supra note 21, at 1253.
²⁷ R. H. Bork, The Tempting of America 144 (1997).
12 Christopher L. Eisgruber

The two propositions of the moral reading are thus relatively modest. They not
only might be endorsed, but in fact have been endorsed, by prominent originalists
such as Scalia and Bork. They more or less echo views articulated almost two cen-
turies ago by John Marshall, who wrote that the drafters and ratifiers of the
Constitution “must be understood to have employed words in their natural sense,
and to have intended what they have said.”²⁸ Yet, despite its relatively uncontro-
versial premises, the moral reading has an important consequence for the status of
practices that existed when the Constitution was written. The moral reading’s first
proposition says that if the framers invoked moral principles, then the
Constitution’s meaning is determined by the meaning of those principles, not by
the framers’ views about how they would apply to contemporary practices. The
moral reading’s second proposition says that the framers did in fact invoke moral
principles. It follows that, in the words of Mark Greenberg and Harry Litman,
“original meaning, properly understood, must [allow for] the possibility that a
traditional practice is unconstitutional.”²⁹
The moral reading thus rules out a popular form of originalist argument which
supposes that “[w]hen a practice not expressly prohibited by the Bill of Rights
bears the endorsement of a long tradition of open, widespread, and unchallenged
use that dates back to the beginning of the Republic, we have no proper basis for
[holding it unconstitutional].”³⁰ Originalists must argue by reference to the prin-
ciples the framers invoked in the Constitution; they are not entitled to assume
that the framers’ practices were consistent with those principles.³¹ If the moral
reading were to succeed in disciplining originalist argument in this way, it would
greatly improve constitutional argument.
The moral reading does not, however, say anything about which principles the
framers invoked. One might (correctly) suppose that history is relevant to answer-
ing that important question. We have therefore not yet identified the features of
Dworkin’s interpretive method that render it so thoroughly moral. To do so, we
must proceed beyond the moral reading and investigate how Dworkin determines
which principles the American framers invoked.

IV. Ambiguous Text, Good Framers, and Moral Principles

Dworkin provides a general prescription about how to determine which moral

principles the Constitution invokes. He says that we must construct “different
elaborations” of constitutional phrases, “each of which we can recognize as a prin-
ciple of political morality that might have won [the framers’] respect,” and then we
must ask “which of these it makes most sense to attribute to [the framers], given
²⁸ Gibbons v. Ogden, 22 U.S. (9 Wheat.) 1, 188 (1824).
²⁹ Greenberg and Litman, supra note 17, at 571.
³⁰ Rutan v. Republican Party of Ill., 497 U.S. 62, 95–96 (Scalia, J., dissenting). For more extensive
discussion of this view, see Greenberg and Litman, supra note 17, at 574–582.
³¹ Id. at 597–617.
 Should Constitutional Judges be Philosophers? 13

everything else we know.”³² What else we know may include facts about constitu-
tional history; thus, as we have already noticed, Dworkin says that the principles we
attribute to the framers of the Fourteenth Amendment must fit the mindset of a
Congress trying to “capture the achievements and lessons of a terrible war.”³³
Yet, as we have already seen, the principles that Dworkin ultimately identifies are
exceedingly abstract. What leads Dworkin to those principles? He claims that the
constitutional text itself is responsible for his conclusions. According to Dworkin,
the Constitution’s “key clauses are drafted in the most abstract possible terms of
political morality.”³⁴ He supports this claim by pointing out that the Fourteenth
Amendment “commands ‘equal’ protection of the laws, and also commands that
neither life nor liberty nor property be taken without ‘due’ process of law.”³⁵ That is
true, of course. We can easily add other provisions to Dworkin’s list.³⁶
Dworkin’s argument, however, glosses over the complexity of constitutional
language. None of the Constitution’s expansive phrases sound unambiguously in
moral principle. On the contrary, all of them mix justice with positive law in
mysterious proportion. The Constitution demands not “equality” but “the equal
protection of the laws;” it protects not against “deprivations of liberty” but against
“deprivations of liberty without due process of law;” and it guarantees not “the
privileges and immunities of free citizens” but the “privileges [and] immunities”
attendant upon a particular legal status, that of “citizens of the United States.” The
Constitution’s sonorous phrases are hard to interpret because they couple diverse
terms without explaining the relationship among them. What does it mean to
guarantee the “equal protection of the laws”? Does it mean that law must live up
to the standards of equality? Or does it mean that equality must be toned down to
accommodate the practical, political functions of the law?
Dworkin’s construction of the Due Process Clauses is especially vulnerable.
Other interpreters have thought that those Clauses by their terms impose only
procedural restrictions upon the government. John Hart Ely, for example, said
that “there is simply no avoiding the fact that the word that follows ‘due’ is
‘process.’ ” In Ely’s view, “substantive due process” is an oxymoron akin to “green
pastel redness.”³⁷ Ely’s “process-means-process” argument would be stronger if the
Clauses said “without fair process” or “without legal process” or simply “without
due process,” rather than without “due process of law,” a combination that beefs
up “process” in a way that common sense cannot easily decode. Does it matter that
the Constitution adds “of law” to “due process”? Perhaps not to positivist legal
theorists, who believe that the meaning of legal rules depends entirely upon the
processes through which they were enacted. For these theorists, the reference to law
merely compounds the reference to process. But Americans have never understood

³² Dworkin, Freedom’s Law, supra note 1, at 9 (Author’s emphasis). ³³ Id.

³⁴ Id. at 72. ³⁵ Id.
³⁶ E.g., the Fourteenth Amendment also prohibits the states from abridging “the privileges or
immunities of citizens of the United States.”
³⁷ J. H. Ely, Democracy and Distrust: A Theory of Judicial Review 18 (1980).
14 Christopher L. Eisgruber

the concept of law in purely positivist ways. The American legal tradition has
consistently recognized a connection between law and justice. That connection is
written into the text of the Constitution, which declares itself to be “supreme
Law” and which lists among its purposes “to establish justice.”³⁸
So Dworkin might well be right about the meaning of the Due Process Clauses.
Nevertheless, the Constitution never quite articulates explicitly the grand moral
principles that Dworkin attributes to it. Moreover, his interpretation renders
much of the text redundant. On his view, for example, the Fourteenth
Amendment contains two consecutive clauses—the Due Process Clause and the
Equal Protection Clause—each of which protects all the rights appropriate to a
free government. The Fifth Amendment’s Due Process Clause does likewise, so
that the other eight amendments in the Bill of Rights appear superfluous. Of
course, it is entirely possible that the framers repeated themselves—perhaps they
took several stabs at expressing a difficult idea. Still, the apparent repetition fits
awkwardly with Dworkin’s highly textual form of argument. Unfortunately for
him, the constitutional text says both less and more than his philosophical
approach to the Constitution requires: less, because it never states a pure, compre-
hensive moral principle clearly; more, because it uses many duplicative phrases
where, according to Dworkin, one would suffice.
I do not believe that the constitutional text can, without more, bear the weight
that Dworkin puts upon it. There is, however, a (nontextual) reason why we
might interpret the Equal Protection Clause and the Due Process Clauses in the
expansive way that Dworkin recommends. The reason invokes a kind of interpre-
tive charity: all other things being equal, it will always make sense to attribute the
most noble principles and purposes to the framers.³⁹ We might rephrase our inter-
pretive task in this way: “The framers who drafted the Due Process Clauses and
the Equal Protection Clauses were not content to itemize specific instances of gov-
ernment misbehavior; instead, they drafted general principles to govern the
nation in the future. Which principles did they choose?” Why would we want to
say the framers did less than the best they could have done? Once we agree that the
framers might have put aside their prejudices to endorse justice and equality in
general, it seems insulting to assume they did anything less.
Given an ambiguous and potentially sweeping moral principle, we will face a
strong tug to read the best into that principle. Historical circumstances are
unlikely to provide much of a barrier. Suppose, for example, that we discover that
the framers of the Fourteenth Amendment all shared some ugly prejudice that
would have kept them from respecting equality in full measure.⁴⁰ Perhaps this
prejudice led them to enact unjust statutes in the course of their political careers.
Could we say that, in light of this prejudice, it “makes sense to attribute” to the
³⁸ For further discussion, see C. L. Eisgruber, Justice and the Text: Rethinking the Constitutional
Connection Between Principle and Prudence, 43 Duke L. J. 1, 48–53 (1993).
³⁹ On the inherent optimism of interpretation, see Dworkin, Freedom’s Law, supra note 1, at 38.
⁴⁰ It does not matter that the prejudice be very ugly. The argument would go through equally well
if we substituted “second-best view” for “ugly prejudice.”
 Should Constitutional Judges be Philosophers? 15

framers some dilute or truncated principle of equality consistent with their preju-
dices? No, for we will then be confronted with the fact that the framers could have
made their prejudices explicit, but chose not to. It is possible that the principle
they enacted embodied their prejudices, but it is also possible that the framers
enacted a broader principle which—while they might have hoped and believed it
to be consistent with their prejudices—transcended their prejudices. All other
things being equal, why not select the more flattering characterization?
For practical purposes, the only way to avoid associating an exceptionally broad
moral principle with the sweeping, enigmatic phrases of the Fourteenth
Amendment is to identify some reason which would lead the framers, at their very
best, to refrain from writing full, robust principles of equality and liberty into the
Constitution. Here lies the true engine that drives Dworkin’s philosophical treat-
ment of the Constitution. It is a powerful motor indeed. The logic that got us to
this stage in the argument is simple. The Constitution includes ambiguous phrases
that might refer to sweeping principles of political morality. The Constitution is
the standard against which Americans judge their government. Those two proposi-
tions, virtually uncontestable, create a strong presumption in favor of using moral
and political philosophy to interpret the Constitution’s abstract phrases: standards
for judging government ought to be good ones rather than bad ones. Dworkin has
seen these points clearly and exploited them powerfully.
The moral ambition of Dworkin’s constitutional method is thus not merely the
product of his intellectual taste. Nor, however, does it derive directly from the sur-
prisingly modest entailments of the “moral reading” of the Constitution. Instead,
Dworkin’s philosophical approach to adjudication results from a particular view of
constitutionalism: Dworkin tacitly assumes that the best framers would write
abstract, sweeping moral principles into their constitutions. This assumption has
considerable appeal. The best constitution-maker might well be the one who drafts
the most clarion and expansive statements of political morality. On the other hand,
American constitutional processes permit, and have in fact sponsored, the writing of
different kinds of provisions, some of which are detailed and others of which
(including the Equal Protection and Due Process Clauses) are murky. Is that because
Americans have been second-rate constitution-makers? Or might they reasonably
have aspired to do something besides stating principles of political morality?

V. Integrity and the Constitution

Dworkin’s own political theory suggests one reason why good constitution-
makers might aim at entrenching principles different from those recommended
by pure moral theory. Dworkin insists upon the existence of a value that he calls
“integrity.”⁴¹ Integrity demands that a nation’s political decisions manifest prin-
cipled consistency over time. According to Dworkin, integrity ranks with equality
⁴¹ Dworkin, Law’s Empire 166 (1986).
16 Christopher L. Eisgruber

and fairness as a political value. He tells us that integrity constrains what real, his-
torical polities can do. Utopian theorists may describe what political arrange-
ments best satisfy the demands of equality and fairness, but leaders of an ongoing
political enterprise must also concern themselves with the relation between their
present policies and earlier ones.⁴²
If there exists a value such as integrity, then surely it should matter to constitution-
makers, who aspire to design real governments, not utopian theories. Dworkin dis-
cusses integrity mainly in connection with the obligation of judges to respect past
precedents, and so one might suppose that constitution-makers need take it into
account only when fashioning judicial institutions. Yet, Dworkin’s theory does not
explain why integrity should be uniquely applicable to adjudication.⁴³ We might
conclude that it applies to government action more generally. If so, integrity could
give good constitution-makers a reason to choose principles that incorporated the
requirements of integrity along with those of equality and fairness. For example,
they might entrench a commitment to “traditional American principles of equality
and fairness” rather than simply to “equality and fairness.” And that, in turn, would
give us reason to interpret ambiguous constitutional language (such as “no state shall
deprive any person of life, liberty or property without due process of law”) to include
a reference to American tradition—because, on the assumptions just stated, that
would be the most charitable interpretation of what the framers said.
This line of argument discloses an irony in Dworkin’s defense of the moral
reading. Dworkin launched his extraordinary work in legal philosophy three
decades ago by observing that judges scrutinize precedent more intently in the
hardest cases, cases that cannot be resolved on the basis of precedent alone and so
force judges to draw visibly upon political and moral values. Dworkin refused to
dismiss the judicial fascination with precedent as mere delusion or smokescreen;
he devoted his jurisprudential career to identifying a conceptually rigorous middle
ground between deferring to past decisions and making fresh judgments of politi-
cal morality.⁴⁴ That is what led him to discover the value of integrity.⁴⁵ Now
Dworkin protests when constitutional scholars insist upon doing in the domain
of constitutional interpretation what he himself did when analyzing stare decisis.
That is, of course, merely ironic, not contradictory. Adjudication and constitu-
tional interpretation are different practices, and a solution appropriate to one
might fail with respect to the other. Nevertheless, Dworkin’s position ought to
make him mildly uncomfortable, for he argues not merely that some third posi-
tion between originalism and the moral reading is wrong but that none exists.
⁴² Dworkin, Law’s Empire, supra note 41, at 164–165.
⁴³ I owe this observation to Lewis Kornhauser. Dworkin does in fact recognize a principle that he
calls “integrity in legislation,” but it is the different principle of “integrity in adjudication” that
“explains how and why the past must be allowed some special power of its own in court.” Dworkin,
Law’s Empire, supra note 41, at 167.
⁴⁴ See, e.g., Dworkin, Taking Rights Seriously, supra note 3, at 81, 87, 112. See also R. M.
Dworkin, A Matter of Principle 147 (1985); Dworkin, Law’s Empire, supra note 41, at 15, 20,
130–131, 228.
⁴⁵ Dworkin, Law’s Empire, supra note 41, at 166.
 Should Constitutional Judges be Philosophers? 17

One has to wonder why there is a middle ground when reading precedent but not
when reading the Constitution.
In fact, I think the two practices too similar to justify the different conclusions
Dworkin reaches about them. If adjudicators have principled reasons to temper
moral demands (and constitutional principles) by reference to past practice, than
so too might constitution-makers and constitutional interpreters. That does not
count as a reason for rejecting the moral reading; the argument I have offered about
integrity accepts the two propositions discussed earlier (namely, that the
Constitution means what the framers intended to say, and that they intended to
state moral principles). The argument, in other words, is consistent with Dworkin’s
claim that there is no middle ground between the moral reading and strict versions
of originalism. The argument does show, however, that there exists a conceptually
coherent middle ground between strict originalism and Dworkin’s own application
of the moral reading—that is, between fidelity to highly specific historical practices
on the one hand and to the most abstract possible moral principles on the other.

VI. The Importance of Constitutional Purposes

The argument of the preceding section depended upon a particular conception of

integrity. Many people might reject that conception, either because they do not
recognize integrity as an independent value, or because they do not believe it
should be extended from adjudication to constitution-making. In this section I
describe other, more widely shared premises that, if accepted, might cause good
constitution-makers to refrain from constitutionalizing abstract moral principles.
The argument begins from the observation that the analysis of constitution-
making is itself an interpretive inquiry. In other words, our account of constitution-
making must both fit and justify it.⁴⁶ When we ask what good constitution-makers
would do, we must remember that they are making a constitution (rather than
designing the ideal political system, constitutional or not), and we must describe
their role in a way consistent with that ambition. We must accordingly ask what
purposes are served by constitutions like the American one. Many popular theo-
ries about the United States Constitution pay special attention to one fact about
it: the Constitution permits super-majorities to entrench specific judgments
against later change. Two hundred years ago, for example, many people appar-
ently thought that no foreign-born person should hold the office of President;
that no Senator should be younger than thirty; and that a jury trial should be
available in all federal torts cases in which the plaintiff sought more than twenty
dollars. Americans must abide by those judgments even if they now appear silly.
Americans may, of course, amend the Constitution, but Article V of the Constitution

⁴⁶ On “fit” and “justification” as dimensions of interpretation, see Dworkin, A Matter of

Principle, supra note 44, at 160; Dworkin, Law’s Empire, supra note 41, at 52–53, 239.
18 Christopher L. Eisgruber

makes amendment extremely difficult. The Constitution’s obduracy (that is, its
barriers against amendment) and its specificity are striking, and any good theory
about the Constitution’s role in the American political system must explain them.
People have used these features of the Constitution to argue that the point of
the Constitution is to strip present-day Americans of the ability to govern them-
selves and compel them to honor instead the specific judgments of earlier gener-
ations. Put more floridly, these theories maintain that the Constitution’s purpose
is to empower “the dead hand of the past.” For example, Supreme Court Justice
Antonin Scalia says that the Constitution’s “whole purpose is to prevent change—
to embed certain rights in such a manner that future generations cannot take
them away.”⁴⁷ In Scalia’s view, the decision to establish a written, obdurate consti-
tution reflects a lack of confidence in future generations. “A society that adopts a
bill of rights is skeptical that ‘evolving standards of decency’ always ‘mark
progress,’ and that societies always ‘mature,’ as opposed to rot.”⁴⁸
Scalia’s theory supplies a reason why later generations should refrain from using
their own, best moral judgment to interpret the Constitution’s abstract phrases;
indeed, on Scalia’s account, the whole point of constitutionalism is to avert the
exercise of judgment by later generations, whose judgment (it is feared) might be
rotten. Scalia’s theory belongs to a larger family of constitutional theories all of
which assert that the purpose of written constitutions is, for one reason or
another, to preclude Americans from engaging in open-ended moral argument.
For example, Walter Berns maintains that the purpose of the American
Constitution was to “negate and minimize” moral controversies which might dis-
rupt the peace that is an essential pre-condition of liberty and democracy.⁴⁹ He
argues that the exercise of “lawyer[s’] . . . private [moral] judgment” in constitu-
tional adjudication is inconsistent with that goal.⁵⁰
Dworkin, by contrast, does not offer any theory to justify the extraordinary
rigidity of the American Constitution. As a result, his expansive view of constitu-
tional rights appears to be the product of historical and textual accidents. The
Framers of the Constitution entrenched specific and now obsolete judgments
about (among other things) contractual liberty, guns, and juries. Nevertheless, the
Constitution came to include certain clauses which (according to Dworkin) pro-
tect every right we should demand from a government. America’s hard-to-amend
Constitution seems calculated to favor stability and conserve past practices, but it
turns out to contain sweeping moral principles which demand potentially radical
reforms. What stunning good fortune!

⁴⁷ Scalia, A Matter of Interpretation, supra note 20, at 40–41.

⁴⁸ Id. Many other lawyers and scholars, with viewpoints quite different from Scalia’s, also build
their constitutional theories around the fact that Article V permits super-majorities to entrench spe-
cific judgments against later revision. See, e.g., B. Ackerman, We the People: Foundations 14
(1991).
⁴⁹ W. Berns, Taking Rights Frivolously, in Liberalism Reconsidered 51, 59 (D. MacLean &
C. Mills, eds., 1983). ⁵⁰ Id. at 64.
 Should Constitutional Judges be Philosophers? 19

Dworkin’s theory of constitutional interpretation is thus incomplete. To answer

arguments like those made by Scalia, Berns, and others, Dworkin must offer an
alternative account of American constitutional institutions and super-majoritarian
amendment procedures.

VII. The Pragmatic Virtues of Super-majoritarian

Amendment Rules
To rebut “dead hand” theories of constitutionalism, we must explain why reason-
able constitution-makers might create super-majoritarian amendment procedures
if they aim to establish government on the basis of evolving standards of justice.
That project turns out to be easier than one might first suppose. The key idea is
this: we can regard super-majoritarian amendment procedures as mechanisms to
discipline and structure, rather than abrogate, the exercise of independent judg-
ment about political justice by subsequent generations.
I will briefly sketch one such account of super-majoritarian amendment pro-
cedures, an account that I elaborate at greater length in my book, Constitutional
Self-Government.⁵¹ There, I contend that super-majoritarian amendment pro-
cedures can serve at least three pro-democratic purposes. First, they prevent major-
ities from consolidating power at the expense of minorities. Various constitutional
institutions may protect the rights and interests of minorities. Federalism and
localism are examples: they permit national minorities to exercise power as local
majorities. Judicial review is another institution that might protect the rights of
minorities and individuals. If amendment procedures were thoroughly majoritarian,
national majorities might use them to consolidate their power. Super-majoritarian
amendment procedures make it more difficult for majorities to do so. This justifi-
cation for super-majoritarian procedures does not depend on the idea that there is
anything especially good about past practices or that the political judgment of
future generations is “rotten.” Instead, the idea is that super-majoritarian amend-
ment procedures safeguard political institutions that discipline current-day judg-
ment to take into account the interests of minorities as well as majorities.
Secondly, super-majoritarian amendment procedures help to establish stable
political institutions. Stable institutions are indispensable to effective govern-
ment. If people argue too much about how to count the votes or when to hold an
election, they will not be able to select and implement policies regarding national
security, social welfare, or civil rights. Under some circumstances, especially when
polities are young, flexible amendment procedures may destabilize institutions.
Political parties and interest groups may find it tempting to open and reopen
questions about institutional design either because of a public-regarding impulse

⁵¹ C. L. Eisgruber, Constitutional Self-Government 12–20 (2001).

20 Christopher L. Eisgruber

to perfect their institutions, or because they want to increase their own power.
Of course, the opposite problem also exists: if amendment procedures are too
demanding, people may find it impossible to make needed reforms. Nevertheless,
it is at least possible that super-majoritarian amendment procedures may help
people to establish and maintain effective political institutions. Once again, this
rationale for super-majoritarian procedures does not depend upon the idea that it
is desirable to preserve past practices or avoid the exercise of present-day judg-
ments about justice: on the contrary, it rests upon a recognition that people must
have stable political institutions in order to make and implement such judgments.
Thirdly, super-majoritarian amendment procedures create incentives that may
improve the quality of institutional reform. As Lawrence G. Sager has pointed out,
the American Constitution’s demanding amendment procedures encourage consti-
tution-makers and constitution-amenders to think about the longterm consequences
of reform.⁵² Because the Constitution is hard to change, constitution-makers are
reminded that their children and grandchildren must live with the consequences of
their decisions. If the Constitution were easier to amend, framers and ratifiers might
instead focus on short-term objectives, assuming that future majorities were free to
undo their work if it proved uncongenial to later circumstances. That would be a mis-
take: institutional reforms have lasting consequences even when they are not consti-
tutionally entrenched. For example, there is no constitutional provision that protects
the Department of Energy, but that does not mean that the Department is easily
abolished once it is created. On Sager’s account, the point of super-majoritarian
amendment procedures is not to preserve past practices, but rather to discipline
reforms so they do not favor short-term, specific practices and judgments at the
expense of evolving views about justice and the common good.
In Constitutional Self-Government, I argue that all three of these functions,
including the first one, should be understood as pro-democratic. More specific-
ally, I argue that democracy differs from pure majoritarianism, so that democracy
requires solicitude for the rights and interests of minorities as well as majorities.⁵³
For present purposes, we need not concern ourselves with these claims about
democracy. We can reject the “dead hand” theory, and accept Dworkin’s approach
to constitutional language, so long as we can regard the Constitution’s super-
majoritarian amendment procedures as serving purposes consistent with the goal
of creating a government that respects very abstract standards of justice. It is then
a separate question whether we should conceive of that government as demo-
cratic, or whether we should instead regard it as another form of government—
say, “constitutional democracy”—that limits democracy in order to promote
liberty or some other aspect of constitutional justice.

⁵² L. G. Sager, The Incorrigible Constitution, 65 N.Y.U. L. Rev. 893, 951–953 (1990).

⁵³ Eisgruber, Constitutional Self-Government, supra note 51, at 18–20, 52–56. See also
C. L. Eisgruber, Constitutional Self-Government and Judicial Review: A Reply to Five Critics, 37 U. S. F. L.
Rev. 115, 120–128 (2002); C. L. Eisgruber, Dimensions of Democracy, 71 Ford. L. Rev. 1723,
1725–1738 (2003).
 Should Constitutional Judges be Philosophers? 21

Nor need we conclude that Article V is perfect, or optimal, in order to conclude

that its purpose is consistent with the goal of honoring abstract standards of
justice. It is entirely possible that Article V is too demanding, so that its net effect is
to “empower the dead hand of the past.”⁵⁴ But that conclusion, if we endorsed it,
would not give us any reason to accept Scalia’s view, which is that a constitution’s
“whole purpose is to prevent change.” We have identified reasons why good
constitution-makers might make amendment difficult even if the goal of constitu-
tion-making is to create a government that honors evolving judgments about
abstract standards of justice. Those reasons remain valid even if America’s framers
miscalculated and made the American Constitution too hard to amend. It would
be perverse to assume that, if the framers designed suboptimal amendment pro-
cedures, we should interpret their more abstract constitutional provisions in a way
that magnifies, rather than ameliorates, their error.
If Dworkin endorsed a theory of super-majoritarian amendment procedures akin
to the one that I have proposed, he could easily employ that theory to buttress his
expansive readings of the Constitution’s liberty-bearing clauses. Perhaps Dworkin has
in mind another, quite different theory of constitutionalism. But in any event
Dworkin owes us some theory of constitutionalism.⁵⁵ For the moment, Dworkin’s
presentation of his theory relies too heavily upon claims about hermeneutics—claims
about the interpretation of language and law in general (as opposed to claims about
American constitutional procedures and institutions in particular). As a result, there is
a conceptual gap between his “moral reading of the Constitution” and his broad read-
ing of the Due Process and Equal Protection Clauses. The “moral reading” entails only
that those clauses state abstract moral principles; the moral reading tells us nothing
about which principles they state. In particular, it does not tell us whether those
Clauses state that “the government must treat everyone as of equal status and with
equal concern.”⁵⁶ Dworkin cannot remedy this deficiency by appealing to the text of
the Clauses; they speak with Delphic obscurity. Dworkin’s own theory suggests that
we should interpret those terms in a way that would make them consistent with the
provisions that a good drafter of the American Constitution would have enacted. But
to know what provisions a good drafter would enact, we must know what it means to
do a good job drafting an American-style constitution (that is, a constitution which is
written and hard to amend). In other words, we must know what purposes are
served by having an American-style constitution. And to know that, we must have
Dworkin’s theory of constitutional institutions, a theory he has not yet given us.

⁵⁴ For discussion of this possibility, see Eisgruber, Constitutional Self-Government, supra

note 51, at 20–25.
⁵⁵ As Jed Rubenfeld has said, Dworkin must pay more attention to “what kind of text the
Constitution is and who its interpreters are.” J. Rubenfeld, On Fidelity in Constitutional Law, 65
Ford. L. Rev. 1469, 1477 (1997). Unfortunately, Rubenfeld understands this inquiry as pertaining
to the Constitution’s legitimacy and authority, rather than its purposes or function. The demands of
political legitimacy are too weak to be helpful; many varieties of constitutionalism (and, for that
matter, non-constitutional government) will qualify as legitimate.
⁵⁶ Dworkin, Freedom’s Law, supra note 41, at 10.
22 Christopher L. Eisgruber

VIII. Conclusion
In my view, Ronald Dworkin has analyzed the structure of judicial reasoning in
American constitutional law more perspicuously than has any other commentator.
My aim here has been not to contest his position, but rather to clarify some of its pre-
suppositions. My interpretation of Dworkin’s “moral reading” has two important
consequences, both of which I hope that Dworkin might accept. First, although
Dworkin long ago called for a “fusion of constitutional law and moral philosophy,”
the moral reading does not require that result. The moral reading requires judges who
decide constitutional cases to focus on moral principle, but they might do so without
engaging in philosophical analysis. What matters most is not whether judges view the
Constitution through the lens of history or philosophy or something else, but
whether they aim at moral principle. There is a great difference between consulting
history at the expense of our judgments about justice (as Scalia purports to do)⁵⁷ and
consulting history to discipline and improve our judgments about justice. That dif-
ference, I believe, swamps the significance of methodological differences among
judges who agree that the Constitution demands judgments about moral principle,
but who disagree about whether those judgments should take the form of philosoph-
ical analysis or interpretations of American history. My argument thus permits a
negative answer to the question that forms the title of this essay: it is at least possible
that judges have no cause to study moral philosophy, and should continue, as lawyers
have traditionally done, to steep themselves in the lessons of political history instead.
Secondly, if judges fail to approach the Constitution in the proper spirit, they are
likely to do so because they subscribe to a mistaken view about the Constitution’s
purpose. Many people believe, as Scalia apparently does, that the point of constitu-
tionalism is to empower the “dead hand of the past” and so to deny the ability of later
generations to govern themselves on the basis of their own best judgments about
justice and the common good. It is easy to see why people reach that mistaken con-
clusion, and once they are in its grip, they believe themselves duty-bound to vindicate
historical opinion even at the expense of justice. To correct this mistake, one needs a
persuasive account of how constitutionalism facilitates government on the basis of
abstract standards of justice. One must explain, in particular, why the Constitution
quite reasonably prescribes specific, obdurate rules with respect to some issues, but
leaves other issues open to vigorous and heated debate. Dworkin’s argument for the
“moral reading of the Constitution,” with its almost exclusive emphasis upon what
the Constitution says, does not address the purpose of constitutionalism.

⁵⁷ Larry Kramer has described Scalia as embracing a kind of “judicial asceticism” under which “it
literally becomes necessary [for Scalia] to reach some results that [he] knows are unjust.” L. Kramer,
Judicial Asceticism, 12 Cardozo L. Rev. 1789, 1795 (1991).
 2
The Place of History and Philosophy
in the Moral Reading of the
American Constitution
James E. Fleming*

I. Introduction
Ronald Dworkin has long recognized that the fundamental questions of “What is
the Constitution?” and “How should it be interpreted?” are the central questions
of fidelity in constitutional interpretation.¹ From his first book, Taking Rights
Seriously,² to his book, Freedom’s Law,³ Dworkin has argued that commitment to
interpretive fidelity requires that we recognize that the Constitution embodies
abstract moral principles rather than laying down particular historical concep-
tions and that interpreting and applying those principles require fresh judgments
of political theory about how they are best understood. He now calls this interpre-
tive strategy the “moral reading” of the Constitution. Yet, narrow originalists such
as Robert H. Bork and Justice Antonin Scalia have asserted a monopoly on con-
cern for fidelity in constitutional interpretation, claiming that fidelity requires fol-
lowing the rules laid down by, or giving effect to the relatively specific original
understanding of, the framers and ratifiers of the Constitution.⁴ They have

* I prepared this essay for the conference, “Exploring Law’s Empire: The Jurisprudence of Ronald
Dworkin,” held at Princeton University on September 18, 2004. The essay is largely drawn from my
article, Fidelity to Our Imperfect Constitution, 65 Fordham L. Rev. 1335 (1997), which I prepared for
a symposium, Fidelity in Constitutional Theory, 65 Fordham L. Rev. 1247–1818 (1997), for which
Ronald Dworkin gave the keynote address.
¹ These questions of “What?” and “How?,” along with the question of “Who is to interpret?,” are
the basic interrogatives of constitutional interpretation. See W. F. Murphy et al., American
Constitutional Interpretation 17–20 (3d ed. 2003).
² R. Dworkin, Taking Rights Seriously 131–149 (1977) [hereinafter Dworkin, Taking
Rights Seriously].
³ R. Dworkin, Freedom’s Law: The Moral Reading of the American Constitution 1–38,
72–83 (1996) [hereinafter Dworkin, Freedom’s Law]; see also R. Dworkin, The Arduous Virtue of
Fidelity: Originalism, Scalia, Tribe, and Nerve, 65 Fordham L. Rev. 1249 (1997), reprinted in
R. Dworkin, Justice in Robes 117 (2006).
⁴ See R. H. Bork, The Tempting of America: The Political Seduction of the Law (1990);
A. Scalia, A Matter of Interpretation (1997); A. Scalia, Originalism: The Lesser Evil, 57 U. Cin.
L. Rev. 849 (1989).

The Place of History and Philosophy in the Moral Reading of the American Constitution. James E.
Fleming.
© Oxford University Press 2006. Published 2006 by Oxford University Press.
24 James E. Fleming

charged that constitutional theorists who reject these claims are “revisionists” who
disregard fidelity, thereby subverting the Constitution. Dworkin has vigorously
and cogently punctured the narrow originalists’ pretensions to a monopoly on
fidelity, arguing that commitment to fidelity entails that we pursue integrity with
the moral reading of the Constitution and that they, the narrow originalists, are
the real “revisionists.”⁵
I shall analyze two strategies for responding to the narrow originalists’ claim to
a monopoly on fidelity. Dworkin takes the first: Turn the tables on the narrow
originalists. He argues that commitment to fidelity entails the very approach that
they are at pains to insist it forbids, and prohibits the very approach that they
imperiously maintain it mandates. The second is taken by Bruce Ackerman and
Lawrence Lessig, to say nothing of Lessig’s sometime co-author, Cass R. Sunstein:
Beat the narrow originalists at their own game.⁶ Ackerman, Lessig, and Sunstein
advance fidelity as synthesis and fidelity as translation as “broad” or “soft” forms of
originalism that are superior, as conceptions of originalism, to narrow originalism.
What is “broad” or “soft” about their forms of originalism is that these theorists
conceive original understanding at a considerably higher level of abstraction than
do the narrow originalists.⁷ At the same time, they argue that the quest for fidelity
requires that we reject Dworkin’s moral reading.⁸ Indeed, Lessig and Sunstein

⁵ Dworkin, Freedom’s Law, supra note 3, at 74–76; R. Dworkin, Life’s Dominion: An

Argument About Abortion, Euthanasia, and Individual Freedom 125–129 (1993) [here-
inafter Dworkin, Life’s Dominion]. I take the term “integrity” from Dworkin’s conception of “law
as integrity.” See R. Dworkin, Law’s Empire 176–275 (1986) [hereinafter Dworkin, Law’s
Empire]. For an insightful analysis of Dworkin’s general conception of legal reasoning in relation to
fidelity, see G. C. Keating, Fidelity to Pre-Existing Law and the Legitimacy of Legal Decision, 69 Notre
Dame L. Rev. 1 (1993).
⁶ See B. Ackerman, We the People: Foundations (1991) [hereinafter Ackerman, We the
People]; B. Ackerman, A Generation of Betrayal?, 65 Fordham L. Rev. 1519 (1997); L. Lessig,
Fidelity and Constraint, 65 Fordham L. Rev. 1365 (1997) [hereinafter Lessig, Constraint]; L. Lessig,
Understanding Changed Readings: Fidelity and Theory, 47 Stan. L. Rev. 395 (1995); L. Lessig,
Fidelity in Translation, 71 Tex. L. Rev. 1165 (1993) [hereinafter Lessig, Fidelity]; L. Lessig & C. R.
Sunstein, The President and the Administration, 94 Colum. L. Rev. 1 (1994); C. R. Sunstein, Legal
Reasoning and Political Conflict (1996) [hereinafter Sunstein, Legal Reasoning]. Other
works illustrating the emergence of a form of broad originalism include M. J. Perry, The
Constitution in the Courts: Law or Politics? (1994); M. S. Flaherty, History “Lite” in Modern
American Constitutionalism, 95 Colum. L. Rev. 523 (1995); W. M. Treanor, The Original
Understanding of the Takings Clause and the Political Process, 95 Colum. L. Rev. 782 (1995).
⁷ See Sunstein, Legal Reasoning, supra note 6, at 171–182; B. Ackerman, Liberating
Abstraction, 59 U. Chi. L. Rev. 317 (1992).
⁸ For an example of Lessig’s rejection of Dworkin’s moral reading, see Lessig, Fidelity, supra note 6,
at 1259–1261. For an example of Sunstein’s rejection of Dworkin’s moral reading in favor of an alter-
native moral reading, see C. R. Sunstein, Earl Warren Is Dead, New Republic, May 13, 1996, at 35
(reviewing Dworkin, Freedom’s Law, supra note 3). For Ackerman’s rejection of Dworkin’s “rights
foundationalism” in favor of his own conception of “dualist democracy,” see Ackerman, We the
People, supra note 6, at 6–16. For examples of interpretations of Ackerman’s work as an attempt to
develop a broad form of originalism, see Flaherty, supra note 6, at 579–590; J. E. Fleming, We the
Exceptional American People, 11 Const. Commentary 355, 369–370 (1994); F. Michelman, Law’s
Republic, 97 Yale L.J. 1493, 1521–1523 (1988); S. Sherry, The Ghost of Liberalism Past, 105 Harv. L.
Rev. 918, 933–934 (1992) (reviewing Ackerman, We the People, supra note 6).
 History and Philosophy in Moral Reading of the Constitution 25

make the Borkish suggestion that Dworkin’s project is not one of fidelity, but one
of improvement.⁹ Thus, the broad originalists attempt to develop an intermediate
theory between narrow originalism and the moral reading.
Dworkin argues that the search for an intermediate theory is pointless and that
the moral reading is the only coherent strategy for interpreting the Constitution.¹⁰
I shall explore the reasons for constitutional theorists’ resistance to the moral read-
ing, and for their persistence in searching for an intermediate theory in the form
of a broad originalism. Dworkin offers one reason: They are in the grip of an
unfounded assumption, the “majoritarian premise,” which leads them to reject
the moral reading on democratic grounds. In Part II, I critique his analysis and,
more generally, assess his constitutional conception of democracy and his moral
reading as a substantive theory of the Constitution. Then, in Part III, I put for-
ward a second reason, which centers on the idea of fidelity: They are in the hold of
another problematic assumption, the “originalist premise,” which causes them to
reject the moral reading on “fidelist” grounds. There I assess Dworkin’s moral
reading as a theory of constitutional interpretation. I contend that the broad ori-
ginalists, like the narrow originalists, fundamentally misconceive fidelity. The com-
mitment to fidelity to the Constitution entails, as Dworkin argues, that we should
interpret it so as to make it the best it can be.¹¹ But broad originalists such as
Lessig mistake this commitment to fidelity as proof that Dworkin is an “infidel.”¹²
Ironically, in the name of interpretive fidelity, the broad originalists, like the nar-
row originalists, would enshrine an imperfect Constitution that does not deserve
our fidelity. Only under the moral reading do we have much hope of interpreting
our imperfect Constitution in a manner that might deserve our fidelity.¹³ Finally,
in Part IV, I suggest that the moral reading is a big tent, and urge liberal and pro-
gressive theorists who have resisted the moral reading in favor of questing for a
broad originalism to reconceive their work as coming within it: in particular, as
being in service of the moral reading by providing a firmer grounding for the
moral reading in fit with historical materials than Dworkin has offered.

⁹ Lessig & Sunstein, supra note 6, at 11 n.35, 85 n.336.

¹⁰ Dworkin, Freedom’s Law, supra note 3, at 14, 18.
¹¹ Dworkin, Law’s Empire, supra note 5, at 176–275; R. Dworkin, A Matter of Principle
146–166 (1985) [hereinafter Dworkin, A Matter of Principle].
¹² Lessig, Fidelity, supra note 6, at 1260.
¹³ Elsewhere, I have characterized the constitutional theory that I develop, constitutional construc-
tivism, as a “Constitution-perfecting theory,” as distinguished from a “process-perfecting theory.” See
J. E. Fleming, Securing Constitutional Democracy: The Case of Autonomy 4–5, 226–227
(2006). I mean “perfecting” in the sense of interpreting the Constitution with integrity so as to render
it a coherent whole, not in Monaghan’s caricatured sense of “Our perfect Constitution” as a perfect
liberal utopia. See H. P. Monaghan, Our Perfect Constitution, 56 N.Y.U. L. Rev. 353, 356 (1981).
Dworkin addresses the “perfect Constitution” objection, which is that his interpretations of the
Constitution always seem to have “happy endings” or “liberal endings.” Dworkin, Freedom’s Law,
supra note 3, at 36. He concedes that the Constitution is not perfect, for it does not protect “all the
important principles of political liberalism.” Id. Nonetheless, he argues that “[i]t is in the nature of
legal interpretation—not just but particularly constitutional interpretation—to aim at happy end-
ings.” Id. at 38. In that sense, Dworkin’s moral reading is also a Constitution-perfecting theory.
26 James E. Fleming

II. The Moral Reading and the Majoritarian Premise: or, The
Moral Reading as a Substantive Theory of the Constitution

In Freedom’s Law, Dworkin argues that the moral reading of the Constitution is
more faithful than the originalist strategy is to the text of the Constitution and the
conception of democracy it presupposes. He contends that “the only substantial
objection to the moral reading, which takes the text seriously, is that it offends
democracy.”¹⁴ Moreover, he argues that constitutional lawyers and scholars who
make this objection are in the grip of an unfounded assumption, the “majoritarian
premise.”¹⁵ This is the assumption that the fundamental value or point of democ-
racy is commitment to the goal of majority will. This premise undergirds a majori-
tarian conception of democracy that is not true to our scheme of government and
that indeed obscures the true character and importance of our system.¹⁶ As an
alternative, Dworkin offers a constitutional conception of democracy which con-
ceives the fundamental point or value of democracy to be concern for the equal sta-
tus of citizens.¹⁷ He then considers and rejects three arguments for the majoritarian
premise, which are rooted in liberty, equality, and community.¹⁸ I believe that
Dworkin’s arguments for the moral reading and against democratic objections
rooted in the majoritarian premise are sound. But I shall criticize his formulation of
a constitutional conception of democracy—or constitutional democracy—and his
own moral reading as a substantive theory of the Constitution.
First, Dworkin is right to lay bare and criticize the majoritarian premise and
the majoritarian conception of democracy that stems from it. For too long, that
premise and conception have hobbled constitutional theory by providing a mis-
guided and misleading account of our constitutional scheme. They have driven
constitutional theorists to regard as deviant or anomalous certain integral fea-
tures of that scheme. Most famously, that premise and conception underlie
Alexander M. Bickel’s anxious claim that judicial review is a “deviant institution”
that poses a “counter-majoritarian difficulty” in our democracy.¹⁹ Dworkin in
effect turns Bickel on his head,²⁰ for Dworkin’s formulation of the “majoritarian
premise” as an unfounded assumption is the inverted mirror image of Bickel’s
formulation of the “counter-majoritarian difficulty” as the root problem. On
Dworkin’s view, the fact that many constitutional theorists are obsessed with the
“counter-majoritarian difficulty” presents a serious problem, because it obscures
from them the true character of our system and prevents them from embracing
the moral reading.

¹⁴ Dworkin, Freedom’s Law, supra note 3, at 15. ¹⁵ Id. at 16. ¹⁶ Id. at 15–17.
¹⁷ Id. at 17–18. ¹⁸ Id. at 21–31.
¹⁹ A. M. Bickel, The Least Dangerous Branch: The Supreme Court at the Bar of Politics
16, 18 (2d ed. 1986).
²⁰ Similarly, Sunstein has suggested that Dworkin has stood Judge Learned Hand on his head. See
Sunstein, supra note 8, at 36.
 History and Philosophy in Moral Reading of the Constitution 27

But Dworkin would be wrong to suggest—and I do not believe that he does

so—that all democratic objections to the moral reading, in particular those
advanced by the broad originalists, are rooted in the majoritarian premise and the
majoritarian conception of democracy. For example, Sunstein makes democratic
objections to Dworkin’s moral reading from the standpoint of his own non-
majoritarian conception of democracy—deliberative democracy—and of his less
abstract, more pragmatic conception of legal reasoning.²¹ Moreover, Sunstein
advances these objections through developing an alternative moral reading of the
Constitution, rather than rejecting completely the idea of a moral reading. This
form of criticism is presumably the type that Dworkin would welcome, for it
engages the idea of a moral reading rather than wholly rejecting it.²²
Secondly, Dworkin is correct in arguing that a constitutional conception of
democracy—or a conception of constitutional democracy—better fits and justi-
fies our constitutional text and practice than does a majoritarian conception of
democracy. He is persuasive in contending that protection of, and respect for,
rights that are the conditions for moral membership in our political community
are themselves preconditions for the legitimacy of the outcomes of majoritarian
political processes.²³ Here Dworkin appears to have taken a page out of John Hart
Ely’s book, Democracy and Distrust,²⁴ in arguing for conceiving our rights as pre-
conditions for the legitimacy or trustworthiness of democracy. But unlike Ely,
Dworkin would include, among the conditions of democracy, certain “substan-
tive” rights such as moral independence, in addition to “procedural” rights like the
right to vote.²⁵
Dworkin is mostly right about what the conditions of moral membership in
our political community are. But the architecture of his constitutional theory is
problematic. I fear that Dworkin’s characterization of all of these substantive and
procedural rights as “democratic conditions” may lead to unnecessary trouble and
resistance. Many readers may resist his argument that substantive rights like moral
independence are “democratic conditions.” Even if they grant that both substan-
tive and procedural rights must be protected for the outcomes of the majoritarian
political processes to be legitimate or trustworthy, they may suspect that he is
pulling a fast one, or making it too easy, or being too clever by packing all of
the rights that constrain majoritarian political processes into the “democratic con-
ditions.”²⁶ To observe, as Dworkin might, that such readers’ objections seem to
²¹ See C. R. Sunstein, The Partial Constitution (1993) [hereinafter Sunstein, Partial
Constitution]; Sunstein, Legal Reasoning, supra note 6.
²² Dworkin, Freedom’s Law, supra note 3, at 38.
²³ Id. at 24; see Dworkin, Life’s Dominion, supra note 5, at 123.
²⁴ J. H. Ely, Democracy and Distrust (1980).
²⁵ Dworkin, Freedom’s Law, supra note 3, at 24–26, 349 n.5. For Dworkin’s earlier critique of Ely,
see R. Dworkin, The Forum of Principle, 56 N.Y.U. L. Rev. 469 (1981), reprinted in Dworkin,
A Matter of Principle, supra note 11, at 33 [hereinafter Dworkin, The Forum of Principle].
²⁶ L. G. Sager has made a similar critique of the architecture of Dworkin’s theory, although his
primary focus was on the theories of Ely, Ackerman, and Frank Michelman. See L. G. Sager,
The Incorrigible Constitution, 65 N.Y.U. L. Rev. 893, 942–948 (1990) (criticizing Dworkin’s
28 James E. Fleming

presuppose the unfounded majoritarian premise may be true, but unhelpful if the
aim is to persuade them to abandon it.
I believe that there is a more straightforward and plausible theoretical structure
through which to present conceptions of constitutional democracy like
Dworkin’s. Elsewhere, I criticize the architecture of constitutional theories such as
those of Ely and Sunstein, which attempt to frame or recast all of our basic liber-
ties, both substantive and procedural, as preconditions for representative or delib-
erative democracy.²⁷ I argue instead for a constitutional constructivism,²⁸ a
conception of constitutional democracy with two fundamental themes: first,
securing the basic liberties that are preconditions for deliberative democracy, to
enable citizens to apply their capacity for a conception of justice to deliberating
about the justice of basic institutions and social policies, and secondly, securing the
basic liberties that are preconditions for deliberative autonomy, to enable citizens
to apply their capacity for a conception of the good to deliberating about and
deciding how to live their own lives. Together, these themes for securing constitu-
tional democracy afford everyone the common and guaranteed status of free and
equal citizenship in our morally pluralistic constitutional democracy.²⁹ (This con-
ception has affinities to Dworkin’s view that the fundamental point or value of our
scheme of government is concern for the equal status of citizens.)³⁰ I offer my
account, constitutional constructivism, as the guiding framework that best fits
and justifies our constitutional text and underlying constitutional order.³¹
Moreover, I contend elsewhere that there are good reasons for conceiving our
basic liberties in terms of securing the preconditions for deliberative democracy
and deliberative autonomy instead of framing them as, or reducing them into,
preconditions for democracy.³² The first reason is prophylactic: Articulating a
constitutional constructivism with these two themes protects us against taking

“constitutive account”). In his recent book, Sager analyzes and criticizes Dworkin’s theory as being,
like Ely’s and Michelman’s theories, a “democratarian account.” L. G. Sager, Justice in
Plainclothes: A Theory of American Constitutional Practice 132–137 (2004) [hereinafter
Sager, Justice in Plainclothes].
²⁷ Fleming, supra note 13, at 4–5, 29–34, 43–51.
²⁸ I mean constitutional constructivism in two senses. First, I intend a general methodological
sense of constructivism, illustrated by Dworkin’s conception of constitutional interpretation as con-
structing schemes of principles that best fit and justify our constitutional document and underlying
constitutional order as a whole. Dworkin originally put forth this conception by analogy to Rawls’s
conception of justification in political philosophy as a quest for reflective equilibrium. Dworkin,
Taking Rights Seriously, supra note 2, at 159–168. Secondly, I intend a specific substantive sense
of constructivism, exemplified by John Rawls’s conception of the equal basic liberties in a
constitutional democracy such as our own as being grounded on a conception of citizens as free and
equal persons, together with a conception of society as a fair system of social cooperation. J. Rawls,
Political Liberalism (1993).
²⁹ Fleming, supra note 13, at 3–6, 61–74. I develop this theory by analogy to Rawls’s political
constructivism. See Rawls, supra note 28.
³⁰ See Dworkin, Freedom’s Law, supra note 3, at 17.
³¹ For examples of Dworkin’s formulations of the two dimensions of best interpretation, fit and
justification, see Dworkin, Law’s Empire, supra note 5, at 239; Dworkin, A Matter of Principle,
supra note 11, at 143–145; Dworkin, Taking Rights Seriously, supra note 2, at 107.
³² Fleming, supra note 13, at 78–79.
 History and Philosophy in Moral Reading of the Constitution 29

flights from substance to process by recasting substantive liberties as procedural

liberties or neglecting them. The second, related reason is architectonic:
Presenting our basic liberties in these terms illustrates that the two fundamental
themes of deliberative democracy and deliberative autonomy are co-original and
of equal weight. The third, more general reason is heuristic: Articulating our basic
liberties through these two themes keeps in view that our constitutional scheme is
a dualist constitutional democracy, not a monist or majoritarian representative
democracy. A final reason is elegance: the importance of being elegant (though
not too reductive) in constructing a constitutional theory. I originally advanced
these reasons for adopting the architecture of a constitutional constructivism with
the foregoing two themes as part of a critique of the architecture of process-
perfecting theories such as Ely’s and Sunstein’s, which recast our basic liberties,
substantive and procedural, as preconditions for representative democracy or
deliberative democracy, but they also apply with some force to the architecture of
Dworkin’s conception of such basic liberties as preconditions for democracy. That
is, the architecture of a constitutional theory with these two themes, which
together secure the preconditions for constitutional democracy, has these advan-
tages over the architecture of Dworkin’s theory.
Thirdly, and most importantly—to make explicit what has been implicit in my
critique of the structure of Dworkin’s constitutional theory—Dworkin never has
developed a moral reading as a general substantive liberal theory of our
Constitution and underlying constitutional democracy. To be sure, he has written
powerfully and cogently about the major constitutional issues of the day, and has
done so from a coherent and consistent viewpoint. Indeed, no one has made
greater contributions to constitutional theory than Dworkin has. But Dworkin
has not worked up a comprehensive yet elegant account of our basic liberties and
constitutional essentials as a substantive theory to beat Ely’s and Sunstein’s
process-perfecting theories.
That has been my project over the past decade or so. I have sought to develop a
Constitution-perfecting theory as an alternative to the process-perfecting theories
advanced by Ely and Sunstein.³³ According to the latter theories, the
Constitution’s core commitment is democracy, and judicial review is justified
principally when the processes of democracy, and thus the political decisions
resulting from them, are undeserving of trust. Process-perfecting theories are
vulnerable to the criticism that they reject certain substantive liberties (such as
privacy, autonomy, liberty of conscience, and freedom of association) as anom-
alous in our scheme, except insofar as such liberties can be recast as procedural
preconditions for democracy. Yet process-perfecting theories persist, notwith-
standing such criticisms, because no one has done for “substance” what Ely has
done for “process.” That is, no one has developed an alternative substantive
Constitution-perfecting theory—a theory that would reinforce not only the pro-
cedural liberties (those related to deliberative democracy) but also the substantive

³³ See id. at 4–5.

30 James E. Fleming

liberties (those related to deliberative autonomy) embodied in our Constitution

and presupposed by our constitutional democracy—with the elegance and power
of Ely’s process-perfecting theory.
That is what my book, Securing Constitutional Democracy, aspires to do. I develop
a Constitution-perfecting theory that secures both the substantive liberties associ-
ated with deliberative autonomy and the procedural liberties associated with delib-
erative democracy as fundamental, without deriving the former from the latter or,
worse, failing to account for substantive liberties altogether. Unlike process theo-
ries, it provides a firm grounding for rights of privacy and autonomy, along with
liberty of conscience and freedom of association, as necessary to secure individual
freedom and to promote a diverse and vigorous civil society. My theory also shows
how basic liberties associated with personal autonomy, along with those related to
democratic participation, fit together into a coherent scheme of basic liberties and
constitutional essentials that are integral to our constitutional democracy.
Finally, Dworkin is right to conceive courts as a “forum of principle,”³⁴ while
recognizing that legislatures and executives are also “guardians of principle.”³⁵
Some liberals and progressives, emphasizing Dworkin’s conception of courts as
“the forum of principle,” have criticized his theory for being too court-centered
and for ignoring “the Constitution outside the courts.”³⁶ That criticism, although
understandable, is plainly overstated. Dworkin has always made clear that legisla-
tures, executives, and citizens also have responsibilities to interpret the
Constitution.³⁷ Sanford Levinson recognized this early on, and appropriately
interpreted Dworkin as a constitutional “protestant” instead of a court-centered
“catholic” on the question, “Who is to interpret the Constitution?”³⁸
Dworkin makes a nod in the direction of endorsing Lawrence G. Sager’s well-
known view that certain constitutional principles required by political justice are
judicially underenforced, yet nonetheless may impose affirmative obligations out-
side the courts on legislatures, executives, and citizens generally to realize them
more fully.³⁹ Sager’s view is an important component of a full moral reading or
justice-seeking account of the Constitution. For it helps make sense of the evident
“thinness” or “moral shortfall” of constitutional law, while still offering a moral
³⁴ Dworkin, The Forum of Principle, supra note 25.
³⁵ Dworkin, Freedom’s Law, supra note 3, at 31.
³⁶ See Sunstein, Legal Reasoning, supra note 6, at 59–60; Sunstein, Partial Constitution,
supra note 21, at 9, 145–146, 374 n.35.
³⁷ See Dworkin, Freedom’s Law, supra note 3, at 31.
³⁸ S. Levinson, “The Constitution” in American Civil Religion, 1979 Sup. Ct. Rev. 123, 141 (inter-
preting Dworkin as a constitutional “protestant” on the question “Who is to interpret the
Constitution?”); see S. Levinson, Constitutional Faith 42–44 (1988). Dworkin has also referred
to his approach on this question as a “protestant” approach. See Dworkin, Law’s Empire, supra note
5, at 190, 413.
³⁹ Dworkin, Freedom’s Law, supra note 3, at 33–34. For Sager’s view, see Sager, Justice in
Plainclothes, supra note 26; L. G. Sager, Justice in Plain Clothes: Reflections on the Thinness of
Constitutional Law, 88 Nw. U. L. Rev. 410 (1993); L. G. Sager, Fair Measure: The Legal Status of
Underenforced Constitutional Norms, 91 Harv. L. Rev. 1212 (1978).
 History and Philosophy in Moral Reading of the Constitution 31

reading or justice-seeking account.⁴⁰ I would urge Dworkin to consider adopting

such a view. (Of course, many questions would remain concerning what is and
what is not judicially enforceable.) I believe that he could do so without under-
mining his arguments against the majoritarian premise.

III. The Moral Reading and the Originalist Premise: or, The
Moral Reading as a Theory of Constitutional Interpretation
Next, I shall consider another reason why the broad originalists have resisted the
moral reading, which centers on the idea of fidelity: They are in the grip of what I
shall call the “originalist premise.” This is the assumption that originalism, rightly
conceived, is the best, or indeed the only, conception of fidelity in constitutional
interpretation. On this view, fidelity by definition, or at least as practiced in our
constitutional culture, must be concerned with following the original meaning of
the text, the original understanding of the framers and ratifiers, or the like. The
originalist premise leads to objections to the moral reading on the ground that it is
“nonoriginalist,” “revisionist,” or not “fidelist.”
The originalist premise is expressed in its most extreme form by Bork, who
asserts that originalism is the only possible approach to constitutional interpreta-
tion that is faithful to the historic Constitution and consonant with the constitu-
tional design. He rejects all other approaches, most especially those like
Dworkin’s, as “revisionist.”⁴¹ In recent years, the originalist premise has also been
manifested in the emerging strain of broad originalism in liberal and progressive
constitutional theory. For example, Lessig evidently takes the view that original-
ism, by definition, is the only method of fidelity. Most strikingly, he has made the
Borkish assertion that Dworkin is an “infidel,” and he and Sunstein have sug-
gested that Dworkin does not even have a method of fidelity.⁴² I believe that the
originalist premise, as much as the majoritarian premise, drives the broad original-
ists’ resistance to Dworkin’s moral reading.
In unpacking what I have loosely called the originalist premise, I shall examine
several reasons why some liberal and progressive constitutional theorists have
resisted Dworkin’s moral reading in favor of searching for an intermediate theory in
⁴⁰ For a justice-seeking account or moral reading of the Constitution that is thicker, or counte-
nances less moral shortfall through judicial underenforcement than does Sager’s view, see S. A.
Barber, Welfare and the Constitution (2003); S. A. Barber, The Constitution of Judicial
Power (1993) [hereinafter Barber, Power]; S. A. Barber, On What the Constitution Means
(1984) [hereinafter Barber, Constitution]; S. A. Barber, Justice-Seeking Constitutionalism and Its
Critics, paper presented at the New York University School of Law Colloquium on Constitutional
Theory (Apr. 20, 1995) (unpublished manuscript on file with the author).
⁴¹ Bork, supra note 4, at 187–240. This is the obligatory footnote where I must acknowledge that
Raoul Berger is more extreme than Bork (or, for that matter, Scalia). See R. Berger, Government by
Judiciary: The Transformation of the Fourteenth Amendment (1977).
⁴² Lessig, Fidelity, supra note 6, at 1260; Lessig & Sunstein, supra note 6, at 11 n.35, 85 n.336.
32 James E. Fleming

the form of a broad originalism. More generally, I discuss the reasons for the
emergence of this strain of broad originalism. I contend that none of these reasons
is a good reason for the broad originalists not to endorse the moral reading, prop-
erly conceived. My general stance is to support broad originalism to the extent that
its proponents undertake it in service of the moral reading, but to criticize it to the
extent that they believe it is sustainable as an alternative to the moral reading.

A. The Turns to History and to Text, History, and Structure

First, the broad originalists seek to reclaim history, and indeed the aspiration to
fidelity, from the narrow originalists. They believe that liberals and progressives
ignored or neglected history for so long that they practically ceded it to conservatives.⁴³
The broad originalists undertook the “turn to history” to show that their constitu-
tional theories, aspirations, and ideals are firmly rooted in our constitutional history
and practice, and indeed provide a better account of our constitutional text and
tradition than do those of the conservative narrow originalists.
The liberal and progressive project of reclaiming history and fidelity from
the narrow originalists is understandable and laudable. But it is understandable
and laudable if undertaken in service of the moral reading, not as an alternative
to it. This project would explain a turn to history, but not necessarily a turn to
originalism. They are not the same thing.⁴⁴ And it would explain a turn to
history in order to pursue an historically grounded moral reading. But it would
not necessarily explain a turn to history that turns away from the moral read-
ing. The turn to history should not become an escape into history.⁴⁵ Why not
conceive the turn to history as doing “fit” work in support of a liberal or pro-
gressive moral reading rather than as a broad form of originalism that rejects
the moral reading?
Secondly, more generally, these liberals and progressives aim to ground their
arguments in the text, history, and structure of the Constitution, and they believe
that a broad originalism is more promising along these lines than is the moral read-
ing.⁴⁶ Some recite this trilogy of sources of constitutional meaning as if it were a
litany. Like the turn to history, the turn to text, history, and structure is an under-
standable and worthy project. Liberals and progressives should firmly ground their
⁴³ See L. Kalman, The Strange Career of Legal Liberalism 132–163 (1996).
⁴⁴ The major criticism I have of Kalman’s fine book is that she seems to treat the turn to history
and the turn to originalism as if they were the same thing.
⁴⁵ See C. Woodard, Escape into History, N.Y. Times (Sept. 15, 1996), §7 (Book Review), at 33
(reviewing Kalman, supra note 43).
⁴⁶ Among the enthusiasts of text, history, and structure are Sunstein, Akhil Amar, and Jeffrey
Rosen. See Sunstein, Partial Constitution, supra note 21, at 119–122; A. R. Amar,
Intratextualism, 112 Harv. L. Rev. 747 (1999); A. R. Amar & V. D. Amar, Is the Presidential
Succession Law Constitutional?, 48 Stan. L. Rev. 113 (1995); J. Rosen, A Womb with a View, New
Republic (June 14, 1993), at 35 (reviewing Dworkin, Life’s Dominion, supra note 5); “Life’s
Dominion”: An Exchange, New Republic (Sept. 6, 1993), at 43 (exchange between Dworkin and
Rosen concerning Rosen’s book review, supra).
 History and Philosophy in Moral Reading of the Constitution 33

arguments in text, history, and structure, not to mention practice, tradition, and
culture. But this turn is not necessarily a turn to originalism and against the moral
reading. Indeed, recourse to structure in constitutional interpretation typically
involves drawing inferences from political theory, not merely recovering, translat-
ing, or extrapolating from the original meaning of the text.⁴⁷ The turn to text,
history, and structure becomes a turn against the moral reading only if its pro-
ponents claim to be elaborating text, history, and structure without making
recourse to political theory. Such a claim would be problematic and implausible.
Why, then, do the liberal and progressive enthusiasts of text, history, and structure
cast their arguments as broad originalist arguments rather than as arguments in
support of better grounding the moral reading?
Thirdly, I suggest that the answer to the question—Why have the turns to history
and to text, history, and structure become turns to broad originalism and against
the moral reading?—is to be found in considerations of litigation strategy or judg-
ments about the types of arguments that are appropriate in our constitutional cul-
ture. The thought seems to be that our constitutional culture is largely originalist
(or positivist), and therefore that arguments in constitutional law, to be successful,
simply must be framed in an originalist mold. A view of this sort seems to animate
the work of broad originalists such as Ackerman, Lessig, and Akhil Amar. I have
heard a strong version of this view articulated roughly as follows: The only way that
liberals and progressives have any hope of persuading Justice Scalia to accept their
interpretations of the Constitution is to make originalist arguments.
To this view I have four responses. (1) The attempt to persuade Scalia that
fidelity to the Constitution leads to any liberal or progressive conclusions is a fool’s
errand. There can be no serious doubt that Scalia’s mind is ideologically impervi-
ous to liberal or progressive constitutional arguments.⁴⁸ Worse yet, this attempt
disfigures and debases constitutional theory by causing theorists to recast their
arguments in a narrow originalist mold dictated by Scalia.
(2) It is telling that the greatest liberal constitutional theorist-litigator of our
time, Laurence H. Tribe, has not adapted his constitutional theory to such an
originalist litigation strategy. To be sure, he has eschewed grand theory, as if to say,
“no theorists here, just us common lawyers.” But his conception of constitutional
interpretation in his academic writing is much closer to Dworkin’s theory than to
the broad originalist views of Ackerman, Lessig, and Amar.⁴⁹
⁴⁷ For examples of accounts of inferences from structure that recognize this, see C. L. Black, Jr.,
Structure and Relationship in Constitutional Law (1969); W. F. Harris, II, The
Interpretable Constitution 144–158 (1993); Fleming, supra note 13, at 90–91 (furthering the
“unfinished business of Charles Black”).
⁴⁸ Notwithstanding possible appearances to the contrary, Texas v. Johnson, 491 U.S. 397 (1989),
and R.A.V. v. City of St. Paul, 505 U.S. 377 (1992), cases in which Scalia supported stringent judicial
protection of freedom of speech, are not counterexamples. For instructive analyses of Scalia’s First
Amendment jurisprudence, as manifested in such decisions, see M. Tushnet, A Court Divided:
The Rehnquist Court and the Future of Constitutional Law 130–155 (2005).
⁴⁹ See L. H. Tribe & M. C. Dorf, On Reading the Constitution 17, 81–87 (1991); L. H. Tribe,
Taking Text and Structure Seriously: Reflections on Free-Form Method in Constitutional Interpretation,
34 James E. Fleming

(3) Our constitutional culture is not as originalist as the broad originalists seem
to assume. It certainly requires constitutional lawyers and scholars to pay homage
to history and to fit with historical materials, but that is not to say that it is origi-
nalist.⁵⁰ Originalism is an ism, a conservative ideology that emerged in reaction
against the Warren Court. Before Richard Nixon and Robert Bork launched
their attacks on the Warren Court, originalism as we know it did not exist.⁵¹
Constitutional interpretation in light of original understanding did exist, but
original understanding was regarded as merely one source of constitutional mean-
ing among several, not a general theory of constitutional interpretation, much less
the exclusive legitimate theory. Indeed, history was regarded as secondary to, and
merely as extrinsic evidence of, the meaning of text and structure.⁵² Scholars
wrote about the “uses of history” in constitutional interpretation rather than con-
tending that enforcing original understanding was the only defensible conception
of fidelity.⁵³ Moreover, original understanding, especially at a relatively specific
level, was understood to be largely indeterminate and inconclusive. As Justice
Jackson famously put it in concurrence in Youngstown Sheet & Tube Co. v.
Sawyer:⁵⁴
Just what our forefathers did envision, or would have envisioned had they foreseen modern
conditions, must be divined from materials almost as enigmatic as the dreams Joseph was
called upon to interpret for Pharaoh. A century and a half of partisan debate and scholarly
speculation yields no net result but only supplies more or less apt quotations from
respected sources on each side of any question. They largely cancel each other. And court
decisions are indecisive because of the judicial practice of dealing with the largest questions
in the most narrow way.⁵⁵
Regrettably, many constitutional lawyers and scholars in recent years seem to
have lost sight of this great wisdom. It is important to note that Laura Kalman, in her
fine intellectual history of recent constitutional theory, has practically suggested that

108 Harv. L. Rev. 1223 (1995) (criticizing the (broad originalist) theories of Ackerman and Amar);
L. H. Tribe, The Puzzling Persistence of Process-Based Constitutional Theories, 89 Yale L.J. 1063,
1072–1077 (1980), reprinted in L. H. Tribe, Constitutional Choices 9 (1985) (retitled The
Pointless Flight from Substance) (criticizing Ely’s theory for taking a “pointless flight from substance,”
just as Dworkin critiqued Ely’s theory for doing so, see Dworkin, The Forum of Principle, supra note
25). For a critique of the broad originalist theories of Ackerman and Lessig from a theoretical per-
spective similar to Tribe’s, see M. C. Dorf, Integrating Normative and Descriptive Constitutional
Theory: The Case of Original Meaning, 85 Geo. L.J. 1765 (1997).
⁵⁰ The Senate’s rejection of the Bork nomination was at least in part a rejection of Bork’s narrow
originalism. See Dworkin, Freedom’s Law, supra note 3, at 276–286, 287–305.
⁵¹ W. W. Crosskey may be an exception, but he was roundly criticized as exceptional. See, e.g.,
H. M. Hart, Jr., Professor Crosskey and Judicial Review, 67 Harv. L. Rev. 1456 (1954) (reviewing
W. W. Crosskey, Politics and the Constitution in the History of the United States (1953)).
⁵² See J. tenBroek, Admissibility and Use by the United States Supreme Court of Extrinsic Aids in
Constitutional Construction, 26 Cal. L. Rev. 287 (1938).
⁵³ See C. A. Miller, The Supreme Court and the Uses of History (1969); J. G. Wofford, The
Blinding Light: The Uses of History in Constitutional Interpretation, 31 U. Chi. L. Rev. 502 (1964).
⁵⁴ 343 U.S. 579 (1952). ⁵⁵ Id. at 634–635 (Jackson, J., concurring).
 History and Philosophy in Moral Reading of the Constitution 35

the best professional historians know better than to be originalists, but that some
constitutional lawyers and scholars who have taken the turn to history do not.⁵⁶
(4) Finally, we should put the following question to the broad originalists: If
our constitutional culture is so originalist, why do so many originalists complain
that so many constitutional law cases and so many features of our constitutional
practice cannot be justified on the basis of originalism?⁵⁷ The answer is that our
constitutional culture is not as originalist as the broad originalists have supposed.
Or that its commitment to originalism is more honored in the breach than in the
observance. Or that Dworkin is right in arguing that “[s]o far as American lawyers
and judges follow any coherent strategy of interpreting the Constitution at all,
they already use the moral reading,” but that there is a confused “mismatch”
between the role of the moral reading, which is embedded in our constitutional
practice, and its reputation, which is that it is illegitimate.⁵⁸

B. The Celebration of “Fit” to the Exclusion of “Justification”

Another reason why some liberal and progressive constitutional theorists resist the
moral reading and attempt to develop a broad originalism is that they believe that
Dworkin’s theory does not take history and “fit” seriously enough, or that it suffers
from a “problem of fit.”⁵⁹ Their objection has two aspects. In the first place, they
claim, Dworkin does not do the concrete groundwork necessary to show that his
interpretations of the Constitution adequately fit the historical materials includ-
ing original understanding and precedents. In the final analysis, they claim, he
will too readily reject as mistakes any historical materials that do not fit his polit-
ical theory. For both reasons, they are dubious about whether Dworkin’s theory, as
Dworkin himself practices it, actually constrains constitutional interpretation to
be faithful to anything other than his own liberal political theory.
In response, I would distinguish between Dworkin’s theory of fidelity as
integrity with the moral reading and Dworkin’s own application of it, and urge:
“Do as Dworkin says, not as he does.” That is, I would argue that Dworkin’s the-
ory of fidelity as integrity is the best conception of fidelity, but would concede that
Dworkin himself may not always satisfactorily do the fit work that his own theory

⁵⁶ Kalman, supra note 43, at 167–190; see J. N. Rakove, Original Meanings 3–22 (1996);
J. Appleby, Constitutional Conventions, N.Y. Times (July 21, 1996), §7 (Book Review), at 20 (reviewing
Rakove, supra). But see Rakove, supra at 7 (criticizing Jackson for overstating the point in the passage
from Youngstown quoted in text).
⁵⁷ See Bork, supra note 4; H. P. Monaghan, Stare Decisis and Constitutional Adjudication, 88
Colum. L. Rev. 723 (1988). For a highly instructive analysis of the gap between originalist theory
and our constitutional practice, see Dorf, supra note 49.
⁵⁸ Dworkin, Freedom’s Law, supra note 3, at 2, 4.
⁵⁹ For a broad originalist claim that Dworkin does not take fit seriously enough, see Flaherty,
supra note 6. For a positivist claim that Dworkin’s theory suffers from a “problem of fit,” see A. J. Sebok,
The Insatiable Constitution, 70 S. Cal. L. Rev. 417 (1997). For a narrow originalist critique along
these lines, see M. W. McConnell, The Importance of Humility in Judicial Review: A Comment on
Ronald Dworkin’s “Moral Reading” of the Constitution, 65 Fordham L. Rev. 1269 (1997).
36 James E. Fleming

calls for, or that he may do it too abstractly to satisfy these critics that he takes fit as
seriously as he should. Dworkin’s splendid essays in constitutional theory in the
New York Review of Books may aggravate such concerns. He writes these essays in a
style designed to reach and persuade a larger audience of citizens, not in a techni-
cal style to demonstrate to constitutional lawyers and scholars that he has done his
historical homework. Ironically, to the extent that Dworkin has indeed become,
in T.M. Scanlon’s estimation, “our leading public philosopher,”⁶⁰ he may have
diminished the appeal of his theory and his work to some constitutional lawyers
and scholars. For in their view, his “public philosophy” may not provide a good
model for the kind of scholarship that shows the proper regard for the aspiration
to fidelity, and that gives fit as well as justification its due.
Furthermore, some broad originalists evidently resist Dworkin’s moral reading
because they believe, as Bruce Ackerman once put it, that “fit is everything.”⁶¹ To
state the matter in terms of Dworkin’s well-known argument that the best inter-
pretation has two dimensions—fit and justification—they seem to believe that
fidelity is purely a matter of fit with historical materials, rather than also a matter
of justification in political theory.⁶² Fit and history do have a role in the quest for
fidelity to the Constitution, but a limited one. We should acknowledge the place
of history in constitutional interpretation—as a constraint that comes into play in
the dimension of fit—but should keep it in its place. Broad originalists tend to
exaggerate the place of history and to give it a greater role than it deserves and than
it is capable of playing.
History is, can only be, and should only be a starting point in constitutional
interpretation. It has a threshold role, which is often not dispositive. In the dimen-
sion of fit, history helps (or should help) screen out “off-the-wall” interpretations
or purely utopian interpretations, but often does not lead conclusively to any
interpretation, let alone the best interpretation. History usually provides a
foothold for competing interpretations or competing theories. It alone cannot
resolve the clash among these competing interpretations or competing theories.
Deciding which theory provides the best interpretation is not an historical matter
of reading more cases, tracts, or speeches or more scrupulously doing good profes-
sional history. To resolve the clash among competing interpretations or competing
theories, we must move beyond the threshold dimension of fit to the dimension of
justification. History rarely has anything useful, much less dispositive, to say at
that point.⁶³ In deciding which interpretation among competing acceptably
fitting interpretations is most faithful to the Constitution, we must ask further
questions: Which interpretation provides the best justification, which makes our
⁶⁰ T. M. Scanlon, Partisan for Life, N.Y. Rev. Books (July 15, 1993), at 45, 45 (reviewing
Dworkin, Life’s Dominion, supra note 5).
⁶¹ See Bruce Ackerman, Remarks at the New York University School of Law Colloquium on
Constitutional Theory, Nov. 16, 1993 (colloquy between Ackerman and Dworkin).
⁶² For Dworkin’s formulations of the two dimensions of best interpretation, fit and justification,
see sources cited supra in note 31.
⁶³ Indeed, as stated above, the best professional historians know better than to be originalists;
unfortunately, some constitutional lawyers and scholars do not. See supra text accompanying note 56.
 History and Philosophy in Moral Reading of the Constitution 37

constitutional scheme the best it can be, which does it more credit, or which
answers better to our best aspirations as a people?⁶⁴ These questions are not those
of an “infidel,” Lessig notwithstanding.⁶⁵ They are required by the quest for
fidelity in the sense of honoring our aspirational principles, not merely following
our historical practices or the original meaning of the text.⁶⁶ And the commit-
ment to fidelity is an aspiration to the best interpretation of the Constitution, not
merely to best fit with the historical materials or original meaning (or best transla-
tion of them). The view that fidelity is merely a matter of fit—or that “fit is every-
thing”—mistakenly assumes that the Constitution is defined, and exhausted, by
the historical materials.
More generally, some broad originalists may resist the moral reading because
they believe that fidelity requires following historical materials and eschewing
political theory. But broad originalists understand constitutional interpretation in
terms of “liberating abstraction,” or conceive original understanding at a relatively
high level of abstraction.⁶⁷ When they elaborate abstract original understanding,
they will find that they are not able to do so purely as a matter of historical
research, translation, or extrapolation. Instead, they will have to do so as a matter
of—and through recourse to—bounded political theory.

IV. Reconceiving the Moral Reading as a Big Tent

The upshot of my analysis of the reasons why the broad originalists have resisted
the moral reading in favor of trying to develop an intermediate theory is that we
should conceive the moral reading as a big tent that can encompass broad original-
ist conceptions such as those of Ackerman, Sunstein, and perhaps even that of
Lessig. Broad originalists have employed the argumentative strategy of using Bork
and Scalia, on the one hand, and Dworkin, on the other, as rhetorical foils or
extremes against which to set up their arguments.⁶⁸ This strategy leads to the unfor-
tunate results of caricaturing Dworkin’s arguments and, worse yet, obscuring simi-
larities and common ground between the moral reading and broad originalism.

⁶⁴ See Dworkin, Freedom’s Law, supra note 3, at 8–11; Dworkin, Law’s Empire, supra note 5,
at 176–275. ⁶⁵ Lessig, Fidelity, supra note 6, at 1260.
⁶⁶ For development of the idea that the Constitution embodies aspirational principles rather than
merely codifying historical practices, see Fleming, supra note 13, at 112–116, 226–227. For similar ideas,
see Barber, Power, supra note 40, at 60–61; Barber, Constitution, supra note 40, at 84–85; F. I.
Michelman, Super Liberal: Romance, Community, and Tradition in William J. Brennan, Jr.’s Constitutional
Thought, 77 Va. L. Rev. 1261, 1312–1320 (1991); Michelman, supra note 8, at 1496, 1514.
⁶⁷ See Sunstein, Legal Reasoning, supra note 6, at 171–182; Ackerman, Liberating Abstraction,
supra note 7.
⁶⁸ Compare Ackerman, We the People, supra note 6, at 10–16 (criticizing Dworkin) with
B. Ackerman, Robert Bork’s Grand Inquisition, 99 Yale L.J. 1419 (1990) (reviewing and criticizing
Bork, supra note 4); compare Sunstein, Legal Reasoning, supra note 6, at 48–53 (criticizing
Dworkin) with Sunstein, Partial Constitution, supra note 21, at 96–110 (criticizing Bork); see
also Lessig, Fidelity, supra note 6, at 1260 (“From the perspective of the two-step fidelitist, both the
originalist [such as Scalia] and the Dworkinian are infidels”).
38 James E. Fleming

Again, I would urge the broad originalists to reconceive their projects as being in
support of the moral reading, not as offering alternatives to it. They can help by
providing firmer grounding than Dworkin has offered for the moral reading in fit
with historical materials. (I do not mean to suggest that their own moral readings
are the same as Dworkin’s particular moral reading.) I shall close by giving three
reasons for embracing the moral reading, conceived as a big tent.
The first reason is hortatory: The moral reading exhorts judges, elected officials,
and citizens to reflect upon and deliberate about our deepest principles and high-
est aspirations as a people.⁶⁹ It does not command them to follow the authority of
the past. In a word, it rejects the authoritarianism of originalism, narrow or broad,
as inappropriate and unjustifiable in a constitutional democracy. As Christopher
L. Eisgruber points out, it is ironic if not absurd that originalists would impose the
“dead hand” of the past upon us in the name of popular sovereignty.⁷⁰ The moral
reading exhorts us to conceive fidelity in terms of honoring our aspirational prin-
ciples rather than merely following our historical practices and concrete original
understanding, which no doubt have fallen short of those principles. On this view,
fidelity is not subservient fealty.
The second, related reason is critical: The moral reading encourages, indeed
requires, a reflective, critical attitude toward our history and practices rather than
enshrining them. It recognizes that our principles may fit and justify most of our
practices or precedents but that they will criticize some of them for failing to live
up to our constitutional commitments to principles such as liberty and equality.⁷¹
Put another way, the moral reading does not confuse or conflate our principles
and traditions with our history, our aspirational principles with our historical
practices.⁷² Again, it recognizes that fidelity to the Constitution requires honoring
our aspirational principles, not following our historical practices and concrete
original understanding. That is, fidelity to the Constitution requires that we disre-
gard or criticize certain aspects of our history and practices in order to be faithful
to the principles embodied in the Constitution.
The final reason is justificatory: The moral reading, because it understands that
the quest for fidelity in interpreting our imperfect Constitution exhorts us to
interpret it so as to make it the best it can be, offers hope that the Constitution
may deserve our fidelity, or at least may be able to earn it. Ironically, despite their
pretensions to a monopoly on concern for fidelity, the originalists would enshrine
an imperfect Constitution that does not deserve our fidelity.⁷³
⁶⁹ I do not mean to imply that the moral reading necessarily requires completely theorized agree-
ments. But see Sunstein, Legal Reasoning, supra note 6, at 48–53 (criticizing Dworkin’s grand,
abstract theorizing and calling instead for “incompletely theorized agreements”).
⁷⁰ C. L. Eisgruber, The Living Hand of the Past: History and Constitutional Justice, 65 Fordham L.
Rev. 1611, 1613–1617 (1997). See also C. L. Eisgruber, Constitutional Self-Government (2001).
⁷¹ See Fleming, supra note 13, at 6, 98. ⁷² See id. at 227.
⁷³ Originalism, as an ism, has no firm footing in our constitutional culture, and it has no place there.
It is a species of authoritarianism that is antithetical to a free and equal citizenry. A regime of purportedly
dispositive original meanings is, at best, beside the point of constitutional interpretation and, at worst,
an authoritarian regime that is unfit to rule a free and equal people. For a similar view, see S. Freeman,
Original Meaning, Democratic Interpretation, and the Constitution, 21 Phil. & Pub. Aff. 3 (1992).
 History and Philosophy in Moral Reading of the Constitution 39

The moral reading frames questions of constitutional interpretation as matters

of principle, to be decided by reflection upon, and deliberation about, basic prin-
ciples and constitutional essentials, not mainly as matters of history that have
largely been decided (at least abstractly) for us by our forebears who are long dead
and gone. It underwrites a constitutional discourse that makes recourse to ques-
tions of principle themselves rather than primarily to other people’s views on other
subjects in other contexts. And the moral reading makes for a better constitutional
citizenry, not to mention better interpretations of the Constitution. It does not
reduce us to poring over other people’s opinions concerning these questions, nor
does it require us to put our arguments in the mouths of people long dead and
gone or to dress up our arguments in their antiquated garb. In other words, it
underwrites a deliberative citizenry, not an authoritarian one.
Thanks to Ronald Dworkin’s monumental contributions to constitutional theory,
we can see this clearly.
 3
How Constitutional Theory Found its Soul:
The Contributions of Ronald Dworkin
Rebecca L. Brown*

Like rain falling to a parched earth, Ronald Dworkin’s early work lighted upon a
field of constitutional thought desiccated by embarrassment over Brown v.
Board of Education.¹ From a distance of a half century, it is difficult to appreciate
the profound chagrin that had arisen from what now seems a simple judicial
declaration of equality. Yet the decision had hurled the world of constitutional
theory into decades of existential angst, leading it, temporarily, to lose a grasp
on its soul.
Taking Rights Seriously came on the scene in the 1970s and quickly began to
enrich the impoverished debate. Subsequent writings, culminating with Freedom’s
Law and its moral reading of the Constitution, increasingly refined Dworkin’s
ideas about law to apply more specifically to the interpretation of constitutional
text.² Ironically, while the academy had nearly imploded over the issue of how to
understand the Constitution’s promise of equality, it was Dworkin’s exposition of
a deep and foundational notion of equality that helped direct constitutional the-
ory toward recapturing its animating spirit. That rehabilitative effect, I suggest, is
at the heart of Dworkin’s contribution to constitutional theory.

* The author gratefully acknowledges the helpful contributions of Lisa Bressman, John
Goldberg, Bob Rasmussen, Christopher Yoo, and the participants in the Georgetown University
Law Center Constitutional Law Colloquium and University of Colorado School of Law faculty
workshop.
¹ 347 U.S. 483 (1954).
² In this essay I have taken the liberty of seeking to present intellectual, rather than chronological,
sequences in Dworkin’s work. This undoubtedly does some violence to the chronology of his corpus as
it actually unfolded. The Dworkin works on which I am primarily relying in this essay are Taking
Rights Seriously (1977); A Matter of Principle (1985); Law’s Empire (1986); Life’s Dominion:
An Argument about Abortion, Euthanasia, and Individual Freedom (1994); Freedom’s Law:
The Moral Reading of the American Constitution (1996) [hereinafter Freedom’s Law];
Sovereign Virtue: The Theory and Practice of Equality (2000) [hereinafter Sovereign
Virtue].

How Constitutional Theory Found its Soul: The Contributions of Ronald Dworkin. Rebecca L.
Brown.
© Oxford University Press 2006. Published 2006 by Oxford University Press.
42 Rebecca L. Brown

I. How Constitutional Theory Lost its Soul

Dworkin’s work, with its claim that there are “right answers” to hard cases,³ had
the misfortune, perhaps, to bear some resemblance—superficial, really—to an old
constitutional order that had come to be rejected with the New Deal. That old
order had permitted judges to defend their decisions from criticism by claiming a
passive role in interpretation that relieved them of responsibility for the jurispru-
dential choices they were making. In Lochner v. New York,⁴ for example, many
perceived that the Supreme Court had made special efforts “to insist the rule of
law ‘forced’ them to reject social welfare legislation.”⁵ This unattractive conflu-
ence of lack of candor and anti-progressive social ends fostered eventual revolt
against the notions of both an activist Supreme Court and a determinate
Constitution (both of which later had associations with Dworkin, although not
altogether accurately). Unsurprisingly, this revolt stoked a smoldering skepticism
begun by the realists early in the twentieth century, ultimately conscripting, in
some form, much of mainstream academia by the century’s midpoint.
The early realist-progressives were not averse to the existence or importance of
individual rights or the advancement of social ideals such as justice and liberty.⁶
Quite the contrary, the writers of this period evoked a genuine sense of societal
aspiration, including the betterment of the lot of the least well off.⁷ But the many
disappointments with what they saw as incorrect decisions from an intransigent
Supreme Court had stripped them of any trust in the judiciary as the guardian
of these rights.⁸ Indeed, opposition to judicial review in the pre-Brown period
was often explained by the rather pragmatic argument that judicial inquiry suf-
fered from a backward-looking perspective and excessively limited scope, making
it ill-suited to contribute to the evolution of the public good.⁹
To the extent there was constitutional theory at this time, its animating spirit
committed theorists to resisting any reading of the Constitution as an impediment
to badly needed social change. The Constitution was a backdrop for political insti-
tutions, enabling them to do what they deemed necessary to improve society, rather

³ See B. Bix, Jurisprudence: Theory and Context 94–97 (1996) (discussing the nature of
Dworkin’s “right answer thesis”). ⁴ 198 U.S. 45 (1905).
⁵ L. Kalman, The Strange Career of Legal Liberalism 18 (1996). The famous quote from
Justice Roberts in United States v. Butler captures the objectionably parsimonious description of the
Court’s task: “to lay the article of the Constitution which is invoked beside the statute which is chal-
lenged and to decide whether the latter squares with the former.” 297 U.S. 1, 62 (1936).
⁶ See, e.g., H. S. Commager, Majority Rule and Minority Rights 62 (1943) (arguing that
“majority will does not imperil minority rights” because “the people can be trusted” to protect minor-
ity interests).
⁷ Kalman, supra note 5, at 17–18 (describing the devotion of the realists to the regulatory state,
seeking to advance the public interest through administrative programs).
⁸ E. S. Corwin, Constitutional Revolution 89–90 (1941) (arguing that rights must generally
depend for their protection on legislatures).
⁹ R. Pound, Contemporary Juristic Theory 83 (1940).
 How Constitutional Theory Found its Soul 43

than constraining them from seeking societal change. Thus, constitutional theorists
were willing to accept the administrative state without serious constitutional objec-
tion because of its potential to expand the opportunity for good government.¹⁰
The Supreme Court’s approach to liberty claims during this period is quite
consistent with the view that the Constitution is best interpreted as imposing the
fewest obstacles on the political branches in the achievement of needed reforms.
The constitutional standard for assessing liberty claims held that individual liber-
ties would prevail unless restrictions were necessary in the common good. The
Lochner case, conceiving of the police power and the common good narrowly,
struck down regulatory measures in 1905. By 1937, however, the Court was ready
to understand the police power more broadly, with the common good recon-
ceived to include regulatory interference with private economic relations where
necessary to respond to the changing social and economic order.¹¹ Still purporting
to adhere to the balance of ordered liberty, the Court’s assessments of the relative
weights in the pans of the scale had altered considerably for the avowed purpose of
facilitating progressive legislative programs.
The ideals of democracy and rights shared a delicate coexistence as long as legis-
latures could plausibly be considered the most trustworthy repositories of the rights
of the people. Justice Frankfurter’s ode to populism spoke a powerful message. He
wrote that “[t]o fight out the wise use of legislative authority in the forum of public
opinion and before legislative assemblies rather than to transfer such a contest to
the judicial arena, serves to vindicate the self-confidence of a free people.”¹² At a
time when democracy was struggling for world opinion against, first dictatorship
and then communism, it is no wonder the prevailing commitment was a belief that
giving the people their voice would be a better source of freedom. This vesting of
public policy decisions in a holistic entity thought of as “the people,” through their
elected representatives, was hailed as the hallmark of democracy. The people would
enact and live under law, understood to be a system of rules, authoritative by virtue
of their source in the political branches of government.
This belief system proved fragile, unable to withstand the frontal assault created
by the forced acknowledgement of racial segregation.¹³ Such segregation, of
course, was nothing new, but (much as did slavery for the founders) it seemed to
have existed, for most of the century, by the side of the Constitution and its
accompanying promises of liberty and equality, without forcing anyone to notice
that perhaps rhetoric and reality did not square. The New Deal, with its emphasis
on addressing severe social and economic problems that united the country, took
¹⁰ See J. M. Landis, The Administrative Process 2 (1938) (“Without too much political theory
but with a keen sense of the practicalities of the situation, agencies were created whose functions
embraced the three aspects of government”).
¹¹ R. L. Brown, Activism is not a Four-Letter Word, 73 U. Colo. L. Rev. 1257 (2002) (arguing that
the mistake of Lochner was not in activist protection of liberty, but in artificially constrained notion of
“common good”). ¹² Minersville Sch. Dist. v. Gobitis, 310 U.S. 586, 600 (1940).
¹³ See R. G. McCloskey, The American Supreme Court 193 (1960) (racism of Adolf Hitler con-
tributed to a “feeling of dissatisfaction and guilt over America’s own patterns of race discrimination”).
44 Rebecca L. Brown

the spotlight away from those even uglier problems that divided it. Brown v.
Board of Education, to put it mildly, raised such a problem.
The blow to constitutional theory as a discipline was not a single coup de grace,
but rather a series of inner conflicts experienced and expressed in different, but
related, ways by several leading scholars of the time. Many have described this
period of intellectual history with depth and perspicacity;¹⁴ I wish only to high-
light some of the significant themes in order to lay the broad-based groundwork
for the symbolic role that I suggest Ronald Dworkin later came to play.
Learned Hand’s 1958 Holmes Lecture is widely viewed as the focal point of
debate. In that speech, he called into question the legitimacy of judicial review
itself,¹⁵ and challenged Brown as legislation from the bench.¹⁶ In a somewhat
more limited way, Herbert Wechsler also criticized Brown as unprincipled in its
unjustified preference of one group’s freedom of association over another’s.¹⁷ After
those two very public attacks on the legitimacy of the Supreme Court’s decision,
as one commentator describes it, “the dam of academic criticism of the Court
burst wide-open.”¹⁸ While many academics spoke in support of the outcome of
the Brown decision, they seemed apologetic for a possible appearance of lack of
principle, uncomfortable with any assertion of a necessary correlation between
segregation and inequality. Academics increasingly leveled criticism at other
activist decisions of the Court, consistently questioning the Court’s commitment
to principle and institutional integrity.¹⁹ Charles Black, departing from this trend,
courageously stepped forward and challenged the charade that allowed people to
doubt that segregation caused inequality: “How long must we keep a straight
face?”²⁰ he demanded, in reference to the claims that segregation had no connota-
tion of oppression or inferiority. He took on the unusual role of reassuring his col-
leagues about the legitimacy of both the outcome as a social policy and the role of
the Court in mandating it. “We as lawyers can without fake or apology present to
the lay community, and to ourselves, a rationale of the segregation decisions that
rises to the height of a great argument.”²¹
The words “fake or apology” seem particularly apt. They describe an academy
deeply unsure of itself, prone to taking different positions publicly and privately.
Most constitutional scholars recognized the moral wrong of segregation at some
level, but were highly uncertain of the legitimacy of enlisting the Constitution to
¹⁴ E.g., L. M. Seidman, Brown and Miranda, 80 Cal. L. Rev. 673 (1992); M. Tushnet, ed., The
Warren Court in Historical and Political Perspective (1993); B. Friedman, The Birth of an
Academic Obsession: The History of the Countermajoritarian Difficulty, Part Five, 112 Yale L.J. 153
(2002); Kalman, supra note 5.
¹⁵ Learned Hand, The Bill of Rights 10, 28–29 (1958) (suggesting that the inference of judi-
cial review from the Constitution was “not a lawless act” as long as it was limited to instances neces-
sary to prevent the collapse of constitutional government). ¹⁶ Id. at 54–55.
¹⁷ H. Wechsler, Toward Neutral Principles of Constitutional Law, 73 Harv. L. Rev. 1, 31–35 (1959).
¹⁸ Friedman, supra note 14, at 198.
¹⁹ See id. at 199–200 (discussing news reports regarding the increase in criticism of the Court
from recognized authorities on jurisprudence).
²⁰ C. Black, Jr., The Lawfulness of the Segregation Decisions, 60 Yale L.J. 421, 425 (1960).
²¹ Id. at 429.
 How Constitutional Theory Found its Soul 45

put an end to the practice. Their understanding of constitutional law embraced

only the thin guarantee of formally nonpreferential treatment, a facade of equal-
ity. It was not capacious enough to embrace Brown’s jarring declaration that “sepa-
rate” could be “inherently unequal” and thus inconsistent with the Constitution’s
textual promise of equal protection of the laws. In their personal moral codes,
these scholars could recognize a role for principle and could apply it to conclude
that racial segregation was a bad thing, causing injustice. But their commitment
to law as a system of rules, with no place for morality or principle, led them to
argue that the power of judges recedes when the law runs out, leaving them no
authority to resolve a constitutional issue by resort to principle.
One fascinating historical account of this period illustrates the schizophrenic
condition of the academy with its frequent use, in describing the views of the vari-
ous theorists, of the qualifiers, “publicly” and “privately.” Publicly the theorists
would take one position, while privately, in correspondence, for example, they
would hedge or qualify.²² Henry Hart, for one, a principal proponent of the Legal
Process school and its advocacy of judicial restraint and fidelity to legal rules, nev-
ertheless privately raged at Hand’s critique of Brown, which he felt would rob the
polity of “any principles of social order which are independent of the appetites and
wills of the contending groups.”²³ Wechsler, too, despite having devoted his
Holmes Lecture to the absence of principle underlying the Brown decision,
acknowledged deep personal conflict about his attack.²⁴ Another account
describes deep ambivalence, too, in Alexander Bickel and Philip Kurland, who
“advanced withering attacks on the Court, all the while approving” many of its
rights-protecting decisions.²⁵ The overall image portrays a Janus-faced academy,
plagued by self-doubt and inner conflict.²⁶
The culmination of this challenge to constitutional theory came with the early
work of Alexander Bickel.²⁷ Perhaps the reason that he came to be described later
as “the most influential scholar of his generation in the field of constitutional
law,”²⁸ is that he had the acumen to pinpoint the precise locus of the epidemic
malaise and to ease it with a palliative, if ultimately unsuccessful,²⁹ therapy. What
²² Kalman, supra note 5, at 33–37. See also Friedman, supra note 14, at 198 (suggesting Wechsler
“stumbled over the decision, unable to justify it, but all the while claiming agreement”).
²³ Kalman, supra note 5, at 35 and n.39 (quoting Hand’s papers).
²⁴ Wechsler, supra note 17, at 43.
²⁵ Friedman, supra note 14, at 251; see also M. Tushnet and T. Lynch, The Project of the Harvard
Forewords: A Social and Intellectual Inquiry, 11 Const. Comm. 463, 482 (1994) (describing Kurland’s
“self-doubt,” Archibald Cox’s slightly later “inner struggle,” and general difficulty among legal
process scholars to reconcile Brown with their school of thought).
²⁶ See also C. Black, Jr., The People and the Court 191 (1960) (“The greatest threat [to judi-
cial review] is a quieter one, working within the legal profession, in the pages of the law reviews, in
those self-doubts which the judges, like all other honest men in power, must recurrently feel”).
²⁷ See A. Bickel, The Least Dangerous Branch: The Supreme Court at the Bar of Politics
(1962).
²⁸ A. Kronman, The Lost Lawyer: Failing Ideals of the Legal Profession 24 (1993) (cited
in Kalman, supra note 5, at 37 n.42).
²⁹ See Friedman, supra note 14, at 219–221 & nn.286–299 (describing critiques of Bickel’s model of
judicial review, and Bickel’s own subsequent disavowal of much of the “counter-majoritarian difficulty”).
46 Rebecca L. Brown

his vision enabled him to see was that the underlying conflict about Brown was
really an identity crisis about democracy itself. In The Least Dangerous Branch, he
chose to press a highly simple resolution.³⁰ Democracy was easy to understand as
long as the important decisions were being made by the Congress or state legisla-
tures.³¹ Indeed, Bickel declared confidently that “the policy-making power of rep-
resentative institutions, born of the electoral process, is the distinguishing
characteristic of the system.”³² His emphasis on elections, along with his choice of
the word “distinguishing,” suggests that perhaps he was tacitly seeking to articu-
late democracy’s distinction from Communism, a salient competitor during this
Cold War period. On the matter of electoral accountability, clearly America’s sys-
tem of government had a strong claim to superiority, while on a metric of individ-
ual rights or entitlements, perhaps—unmentioned in Bickel’s definition of
democracy—the competitive advantage might be less clear.
Whatever Bickel’s motivation, his simplification of the concept of democracy,
reducing it to its most elemental populist foundations, captured a powerful and
widely felt sentiment. But if this was democracy, then the concept of individual
rights was in trouble. If democracy means the enactment into law of policies that a
majority of the people favor, and that preference is, by definition, reflected in the
decisions of representative institutions, then it follows that any effort to thwart or
overturn legislative decisions is anti-democratic. Indeed, that was precisely
Bickel’s conclusion about judicial review: he chose the words “undemocratic”³³
and “deviant,”³⁴ in addition to the perhaps less pejorative, and more famous,
“counter-majoritarian difficulty.”³⁵ Courts would be the last-resort guardians of
the most “clear cut” constitutional values, but would restrain themselves from
involvement in controversial political matters best left to the politically account-
able branches. “The more fundamental the issue,” Bickel wrote, “the nearer it is to
principle, the more important it is that it be decided in the first instance by the
legislature.”³⁶
The Least Dangerous Branch served to facilitate a seductive forgetfulness about
what a constitutional democracy is. It would not be understood as the “inarticulate
and complex”³⁷ idea encompassing a collection of interacting mechanisms work-
ing in different directions to ensure a combination of procedurally sound represen-
tative political government, along with “insight and wisdom and justice.”³⁸
³⁰ This book was not by any means the end of Bickel’s thinking on the subject, and in later works
he revised and, to some extent, complicated his views about democracy. See A. Bickel, The Supreme
Court and the Idea of Progress 175–181 (1970); A. Bickel, The Morality of Consent
(1975). My focus here is on the impact of the 1962 book.
³¹ Bickel paid little heed to the growing insights about the inexact correspondence between legis-
lative decisions and majority will, propounded by such theorists as Robert Dahl and David Truman.
See Kalman, supra note 5, at 38–39 (describing Bickel’s apparent rejection of pluralism) See also
Bickel, The Supreme Court and the Idea of Progress, supra note 30, at 84–86 (later recognizing
some basic interest group role, but still emphasizing the preeminence of the vote as the sole concern
of the Court). ³² Bickel, supra note 27, at 19.
³³ Id. at 17. ³⁴ Id. at 18. ³⁵ Id. at 16. ³⁶ Id. 161.
³⁷ Black, supra note 26, at 179. ³⁸ Id. at 182.
 How Constitutional Theory Found its Soul 47

Instead, the new understanding of democracy was much easier to comprehend: it

was majority rule.
But this was a loss for constitutional theory, as majority rule is flat and self-
contained; it looks neither backward for wisdom nor forward for aspiration. The
argument could no longer be (as it had been in the New Deal period) that the
popularly elected branches of government are most trusted to produce the best
social ends, and that, accordingly, they should be given as free a rein as possible in
the formulation of public policy. De jure segregation and its accompanying unjust
treatment of individuals, as well as the persecutions of the McCarthy era, had
foreclosed trust of that sort. If that were the argument that constitutional theory
propounded, it would have had to concede that there were no rights to begin
with—a thin disguise for the fundamentally skeptical point that there are, in fact,
no rights against the state.³⁹ Most were not willing to go that far. All that was left
for constitutional theory, then, was the argument that majority rule was an end in
itself, that whatever most of the people wanted they should have, in the name of
democracy. This exclusive focus on the process of government decision-making
suggested that courts should be careful, apologetic, humble, and concerned about
their own legitimacy.⁴⁰ Bickel’s grudging account gave no guidance as to how the
constitutional issues of the day should be decided, as long as they were resolved by
the right political body, which was, nearly always, the legislature or executive,
responding to majoritarian preference. This is a theory of constitutional law, but it
is a cynical theory, without a sense of overarching purpose of advancement toward
any substantive societal ideal. In short, his theory captured the sense of ambiva-
lence and self-doubt of constitutional theory generally. It took the last step toward
sapping the discipline of its soul.
Literature teaches us that people who lose their souls tend to vacillate in schizo-
phrenic torment between what they sense to be their true spirits and the tempta-
tions that draw them away. They are denied the luxury of inner peace and
self-knowledge. Dr. Faustus, who traded his soul to the devil in exchange for fabu-
lous magical power, is portrayed as frequently subjected to the opposing entice-
ments of the good angel, urging him to repent and serve God, and the evil angel,
exhorting him to follow his lust for power.⁴¹ Dorian Gray, too, having pledged his
soul in exchange for ageless beauty, is racked by a duality that is reconciled only in
his death.⁴² These figures sought the simplicity of an unmitigated good—for
them, power or youth—and sought, to their own doom, to reject the possibility of
balance among complex and competing human faculties.
And so it was, in less poetic form, for constitutional theory. Even if with the
best of intentions, the discipline succumbed to the enticement of simplicity over

³⁹ Taking Rights Seriously, supra note 2, at 146.

⁴⁰ See C. R. Sunstein, One Case at a Time: Judicial Minimalism on the Supreme Court
(1999) (carrying forward this theme in later work).
⁴¹ C. Marlowe, Doctor Faustus (Roma Gill ed. 1990).
⁴² O. Wilde, The Picture of Dorian Gray (Wordsworth ed. 1992).
48 Rebecca L. Brown

complexity, concrete answers over abstract ideals, losing track of what the
American Constitution had been since its inception. The academy’s schizophrenic
reaction to Brown—recognizing the necessity of its outcome, while still resisting
the institutional muscle that permitted the judiciary to reach it legitimately—is
symptomatic of this loss of core mission. Scholarship argued back and forth about
how best to hold the Court back, how to restrain it, from overstepping the narrow
bounds permitted it by a system conceived in majoritarian supremacy.⁴³ Judicial
review, and, with it, the enforcement of individual rights, was an embarrassment to
the Constitution rather than its highest calling. What served instead, as a final
cause, was elusive. The candor and commitment necessary to face up to a banish-
ment of rights from the democracy were lacking, yet the preservation of a shell with
no defining contours left a vacuum of direction and purpose. Thus, it plausibly
could be said in these times that “the institution of judicial review has somehow
faded out of the picture—or, what is far worse, has been converted into a mere rit-
ual of acquiescence, or in plainer terms a solemn if not particularly pious fraud.”⁴⁴
It was not appropriate to talk too much about rights, or even justice.⁴⁵ Anyone
who presumed to do so ran the risk of being accused of offending democracy or favor-
ing “judicial tyranny.”⁴⁶ Political rhetoric, following the lead of the academy, spoke of
aspirations of “strict construction” and “judicial restraint,”⁴⁷ both of which were code
words for reading rights in the Constitution as parsimoniously as possible.⁴⁸
Anything else, it was claimed, was “making” law rather than “interpreting” it.⁴⁹
With Ronald Dworkin came the restorative balm of candor and the frank com-
mitment to constitutional ideals. He was not literally alone, of course, in his aggres-
sive program of taking rights seriously. But his work stands out for its relentless
challenge to the false dichotomies that had become commonplace in constitutional
thinking: between making and interpreting law, between rank rights skepticism

⁴³ See E. Chemerinsky, The Supreme Court, 1988 Term—Foreword: The Vanishing Constitution, 103
Harv. L. Rev. 43, 73 (1989) (documenting the widespread adoption of the “majoritarian paradigm”,
“the idea that judicial review . . . is in tension with American democracy”).
⁴⁴ Black, supra note 26, at 191.
⁴⁵ See M. Horwitz, The Warren Court and the Pursuit of Justice, 50 WM. & Mary L. Rev. 5, 11
(1993) (describing how Chief Justice Warren was ridiculed for asking attorneys whether their posi-
tion was just: “Sophisticated legal scholars did not speak that way”).
⁴⁶ See, e.g., R. H. Bork, Neutral Principles and some First Amendment Problems, 47 Ind. L. J. 2–3
(1971) (arguing that judges necessarily abet “tyranny” if they exceed their proper sphere).
⁴⁷ See Taking Rights Seriously, supra note 2, at 31–37.
⁴⁸ See J. F. Simon, In His Own Image: The Supreme Court in Richard Nixon’s America 8–9
(1973) (by “strict constructionists,” President Nixon “meant that he would appoint judges who took
less active and reform-minded views”).
⁴⁹ See Changes Nixon May Make in Federal Courts, U.S. News & World Rep., Dec. 2, 1968, at 42
(quoting President Nixon as stating, “In my view, the duty of a Justice of the Supreme Court is to
interpret the law, not to make law, and the men I appoint will share that view”). The alleged
dichotomy has persisted. See President’s Remarks Announcing the Nomination of Clarence Thomas
to Be an Associate Justice of the Supreme Court of the United States and a News Conference in
Kennebunkport, Maine, 27 Weekly Comp. Pres. Doc. 868 871 (July 1, 1991) (then-President
George H. W. Bush’s suggestion that the nominee would “faithfully interpret the Constitution and
avoid the tendency to legislate from the bench”).
 How Constitutional Theory Found its Soul 49

and deference to political institutions, between self-government and judicial

activism. “Why has a sophisticated and learned profession,” he mused, “posed a
complex issue in this simple and misleading way?”⁵⁰ In restoring complexity to
constitutional democracy, his work both reinvigorated prior commitments and
brought new focus to the value so critical to the latter part of the twentieth century,
at the heart of the disagreement about judicial power, and foundational to
American democracy—the value of equality.

II. The Moral Reading of the Constitution

Ronald Dworkin set a place for principle at the table of constitutional theory. He
argued that the moral skeptics of the 1930s and 1940s were impoverished in their
rejection of morality as part of law. The positivist emphasis on law as command,
with authority commensurate to its source in the people, erroneously limited law to
a formal set of rules that had no use for morality in its enforcement and application.
To Dworkin, this was an unacceptable mask to impose on law of any kind, but his
attacks on positivism and utilitarianism had particular resonance for constitutional
law. Because Dworkin saw all law as a system of rights recognition and protection,
his vision readily lent itself to employment in the development of constitutional law
and theory. His understanding of law as comprising not just legal rules, but also
principles, provided an apt foundation from which to construct an interpretative
edifice for the Constitution itself, for Dworkin a rich source of such principle.
Dworkin’s method calls upon the judge faced with an adjudicative issue to dis-
cover principles by looking backward in time to the practices and traditions in the
nation’s history.⁵¹ Evolving from common law to constitutional law, the “moral
reading” of the Constitution asks judges to find what Dworkin calls the “best”
conception of constitutional moral principles that would fit the historical record,
which includes constitutional text, evidence of contemporaneous understanding,
precedent, and societal experience over time. These principles, in turn, inform the
resolution of current issues of the day, without necessarily determining outcomes.
For the first time, the idea of balancing individual liberty against the commu-
nity’s needs came under critical scrutiny. Dworkin argued for recognition of moral
rights against the government—those that it would be wrong for a government to
transgress even for the best of reasons rooted in the common good.⁵² These moral
claims that all individuals hold against their government are determined by a
political judgment regarding the limits of government power in the free society
that our Constitution establishes and maintains. This is an appropriate task for a

⁵⁰ Taking Rights Seriously, supra note 2, at 148.

⁵¹ See Law’s Empire, supra note 2, at 228–229; see also Bix, supra note 3, at 91 (explaining
Dworkin’s interpretative method).
⁵² See Taking Rights Seriously, supra note 2, at 139.
50 Rebecca L. Brown

judge who is not second-guessing the underlying choices regarding policy or

personal moral commitment that underlie the law, but rather is testing that enact-
ment against a constitutional principle such as justice or equality.⁵³
In seeking the principles to guide this exercise of political judgment, Dworkin
borrows a basic technique from his arch-opponents, the originalists. Like them,
he asks the judge to consider the textual provision at issue in its historical setting,
as well as whatever meaning can be derived. What he prescribes next, however,
expands the interpretative endeavor beyond what the originalist would do. His
investigation of the original intent is important in the search for the guiding prin-
ciples that were intended to guide society through time. How those principles will
be applied, however, is not governed by the original understanding, but rather by
the moral judgment of the present-day judge. That approach facilitates consis-
tency and fidelity to constitutional ideals such as liberty and equality, without
straitjacketing society with outgrown conceptions of what those concepts actually
require in specific cases. This effort to reach across temporal and ideological
canyons is the hermeneutical device that Dworkin offers in the quest for coher-
ence and integrity in constitutional interpretation. He anticipates moral progress.
This search for principle in our collective past supplies one defense against the
common charge that Dworkin gives judges too much power to inject personal
preference into constitutional meaning.⁵⁴ While Dworkin does not deny the
inevitable element of personal judgment in the Herculean job of identifying and
⁵³ I have struggled with the terminology before, in commenting upon Christopher L. Eisgruber’s
Constitutional Self-Government (2001), which shares some common ground with Dworkin.
See R. L. Brown, A Government For the People, 37 U.S.F. L. Rev. 5 (2002). One thing the two share is
the use of the word “moral” to describe the constitutional questions that courts are charged with
resolving. I continue to believe that the debate about principled judicial review would be edified if we
made a clearer distinction between personal-moral questions, such as whether it is morally acceptable
to terminate a pregnancy, and political-moral questions, such as whether it is morally permissible for
a government to dictate matters involving deep personal conscience and bodily integrity. Although
both types of issue may be presented in a single case, they are distinguishable, and importantly so, in
my view. The use of the word “moral” to describe both has led some to argue with Dworkin, I believe, at
cross-purposes. Richard Posner, for example, argued, in connection with his attack on Dworkin and what
he terms moral theory, that the Court in Roe v. Wade “ducked the moral question.” See R. A. Posner,
Problematics of Moral and Legal Theory, 111 Harv. L. Rev. 1637, 1703 (1998). I take it that he meant
that the Court avoided the question whether a fetus is a person—what I would call the personal-moral
question, or first-order moral question. But the Court in Roe did indeed answer the political-moral
question, or second-order moral question (what I believe Dworkin would view as the principal moral
question presented in the case), by resolving whether the government may regulate this particular
aspect of private decision-making. If we could substitute something like the words “political judg-
ment” for “moral judgment,” I believe some of the rabid objection to Dworkin’s work (as substituting
judicial morality for personal, and the like) could be calmed, or at least cabined to areas of genuine
disagreement. This does not go to the substance of Dworkin’s theory, but is rather an effort to stave
off one type of mischaracterization thereof.
⁵⁴ A related charge accuses Dworkin of bringing back the ghosts of natural law into constitutional
interpretation. Dworkin has vigorously defended his distinction between natural law and moral prin-
ciple, the latter being accessible through history and reason, tethered to societal values and traditions,
while the former is not. See S. Guest, Ronald Dworkin 84 (1991) (explaining that Dworkin does
not claim that “there is a ‘natural’ answer ‘out there’, which supplies an ‘objectivity’ to moral and legal
argument”).
 How Constitutional Theory Found its Soul 51

applying principle, he does deserve credit for seeking mightily to frame the inquiry
so as to avoid the appeal to the personal. He charges the judge with asking ques-
tions about the Constitution and the constitutional structure that we have, and
grounding the answer to that inquiry in reason brought to bear on shared experi-
ence and value, rather than personal experience and value.⁵⁵ His steadfast insistence
on the existence of accessible moral truth, while subjecting him to ridicule,⁵⁶
enriched the project of ascertaining appropriate limits on government power.
While the source of principle is in the past and present, its trajectory points
toward the future by relying on a judgment of how best to reconcile past practice
with the aspirations of the polity. What could be more inspiring to a discipline strug-
gling to find itself than a plea for “moral progress, and though history may show how
difficult it is to decide where moral progress lies, and how difficult to persuade others
once one has decided, it cannot follow from this that those who govern us have no
responsibility to face that decision or to attempt that persuasion.”⁵⁷
Dworkin’s method does not abandon law; but rather redefines what counts as
law. Unlike Hart, he does not ask the judge to step beyond the law and decide hard
cases on the basis of something other than law, like policy. Instead, he endows law
with a key component, principle, that prevents it from ever running out. The
moral reading of the Constitution, in particular, claims an authority deriving
from the Framers of the Constitution themselves, boasting not only consistency
with their original intent, but also fidelity to the text that they adopted as funda-
mental law. This claim attaches to every significant stage of constitutional draft-
ing: the original document, the Bill of Rights, and the Civil War Amendments.
The language of these rights-protecting provisions of the Constitution, phrased as
they are in “exceedingly abstract moral language”⁵⁸ such as the “right” of free
speech, “due” process, and “equal” protection, incorporates a set of abstract moral
principles as limits on governmental power.⁵⁹ He can thus accommodate prin-
ciple in the interpretative process without compromising the authority many
claim is vital to judicial legitimacy. If these constitutional principles seem, too
coincidentally, to correspond to basic tenets of liberal political thought, it is the
result, not of Dworkin’s preferences, but of the Framers’. They constructed our
nation’s basic commitments, after all, “in the bright morning of liberal thought.”⁶⁰
Central to the moral reading is the decision to conceive this set of constitu-
tional principles at its most general and aggregate level so as to establish an over-
arching guide for specific interpretations of the document. Dworkin has read the
principles set out in the Constitution, taken together, to commit the United States
to the proposition that “the government must treat all those subject to its domin-
ion as having equal moral and political status; it must attempt, in good faith, to

⁵⁵ See Freedom’s Law, supra note 2, at 10–11.

⁵⁶ See R. A. Posner, The Problematics of Moral and Legal Theory at 3 (1999) (calling the
idea of an accessible, objective moral order “spurious”).
⁵⁷ Taking Rights Seriously, supra note 2, at 147. ⁵⁸ Freedom’s Law, supra note 2, at 7.
⁵⁹ Id. ⁶⁰ Id. at 38.
52 Rebecca L. Brown

treat them all with equal concern; and it must respect whatever individual free-
doms are indispensable to those ends . . . .”⁶¹ This foundational insight has come
to be known as Dworkin’s notion of “equal concern and respect,” which he has
developed throughout his work, and has been, to some degree, the trademark of
his approach to constitutional interpretation. Although he initially attributed the
principle to Rawls,⁶² it is Dworkin who gave that principle life as a foundation for
constitutional theory.
It is the principle of equal concern and respect, enshrined into positive law in
the Constitution, that provides the starting place for analysis of all individual
rights. It is a principle much deeper and more substantive than a mere formal
equality. It contemplates real limits on government’s use of its powers in its treat-
ment of people: “Government must treat those whom it governs with concern,
that is, as human beings who are capable of suffering and frustration, and with
respect, that is, as human beings who are capable of forming and acting on intelli-
gent conceptions of how their lives should be lived.”⁶³ It thus embraces more than
mere equality of treatment, including also the rights we generally think of as
autonomy and liberty rights. These concepts of individual liberty are implicit in
the equality-based proposition that government can never constrain one person’s
pursuit of the good life on the rationale that another’s conception of a life well led
is superior.⁶⁴ This proposition helps to suggest both the contours of what kinds of
particular liberties must be respected and what kinds of reasons governments
must (or may not) offer in support of any state effort to curtail them. Both this
identification of protected liberty rights and the inquiry into state reasons are
evident in the Supreme Court jurisprudence under the Due Process Clause.
Sounding in liberty, these inquiries are latent in that thick conception of equality—
“the most radical conception of equality there is.”⁶⁵
Dworkin’s reliance on the text of the Bill of Rights and the Fourteenth
Amendment to supply the pedigree for the principle of equal concern and respect,
although apparently genuine, should not be overstated.⁶⁶ It is quite evident that

⁶¹ Id. at 8.
⁶² Taking Rights Seriously, supra note 2, at 181. Indeed, Dworkin devoted considerable atten-
tion to arguing that Rawls can be explained only by attributing to him a deep recognition of equal
concern and respect as a natural entitlement with which all persons are endowed, rather than as a
product of contract. Id. at 178–183. Whether or not this is the correct reading of Rawls, Dworkin
made clear that it is the correct reading of Dworkin. See id. at 272–273 (explaining the concept of
equal concern and respect as a “postulate[] of political morality”). ⁶³ Id. at 272.
⁶⁴ Dworkin has articulated the relationship between liberty and equality in different ways. At
times, he has seemed to suggest that equality may be viewed as prior to liberty because a just govern-
ment could not recognize a right to liberty that would conflict with a proper understanding of equality.
See R. Dworkin, What is Equality? Part 3: The Place of Liberty, 73 Iowa L. Rev. 1, 10 (1987). Yet he has
always resisted the idea that either liberty or equality is instrumental to the other, arguing that the two
are reconcilable ideals that merge into a fuller account of treating all persons with equal concern. See
Sovereign Virtue, supra note 2, at 123. ⁶⁵ Taking Rights Seriously, supra note 2, at 182.
⁶⁶ See E. B. Foley, Interpretation and Philosophy: Dworkin’s Constitution, 14 Const. Comment.
151, 154 (1997) (reviewing Freedom’s Law) (questioning the candor of Dworkin’s reliance on text).
 How Constitutional Theory Found its Soul 53

Dworkin’s commitment to the authority of the principle as intrinsic to our

Constitution derives from other sources, in addition to the collective decision by
politicians of prior centuries to concur in the use of a few fortuitous adjectives.
Although the abstract quality of the selected words does indeed support the intu-
ition, the principle of equal concern and respect is probably more accurately
understood, at least in part, as an inference from the fact and existence of the con-
stitutional democracy itself.
Like Bickel, Dworkin had the political acumen to recognize that the battle over
judicial review involved the very definition of democracy, and he addressed that
issue head on. He rejected all efforts to frame the question as how far democracy
can properly be compromised for the sake of rights—an apologetic Constitution.
Dworkin’s critics seemed to suggest that democracy should be understood as
falling at one end of a continuum, with justice in opposition at the other end, and
rights just trade-offs in between. Under this view, the difference between Dworkin
and Bickel would be simply where they would draw the line of acceptable trading
off. But Dworkin conceived the question quite differently.⁶⁷
For Dworkin, democracy is a means to attain justice, not its polar opposite. In
contrast to Bickel, who identified democracy with majoritarianism, Dworkin saw
democracy as instrumental to the achievement of equal status for all citizens.⁶⁸
Under this view of democracy, the legitimacy of government institutions is mea-
sured, not by the degree to which they respond to popular will, but by the degree
to which they provide the conditions necessary to achieve the moral objective.
Accordingly, the criticism of any institution as democratically “deviant” (Bickel’s
word for the unelected judiciary), would apply only if government failed to
respect and promote the goal of equal status for all.
This conception of the American democracy has two tremendous assets. First,
it does a much better job than does Bickel of explaining the Constitution that we
have. Those who seek to equate our democracy with majoritarianism have a diffi-
cult time explaining the various structures established in the Constitution for
tempering, buffering, and constraining the power of majority rule.⁶⁹ Under
Dworkin’s view, however, the Constitution’s structuring of the three branches
with different means and kinds of political accountability and diverse kinds of
institutional checks is entirely consistent with the end of equal concern and
respect. It portrays a robust and holistic organism whose parts work in interdepend-
ent ways toward a common objective. Secondly, Dworkin eliminates the need
for the “moral regret” that is inevitable if one sets up a dichotomy in which legiti-
mate government is poised in potential opposition to the project of doing what is
right (essentially, the Brown dilemma). This view entails a moral cost to be borne
whenever a court must overturn a legislative act that undermines the rights of
⁶⁷ Freedom’s Law, supra note 2, at 15. ⁶⁸ Id. at 17.
⁶⁹ See R. L. Brown, Accountability, Liberty, and the Constitution, 98 Colum. L. Rev. 531, 552–558
(1998) (arguing that if our Constitution was intended to establish a majoritarian government, it is a
failure).
54 Rebecca L. Brown

citizens.⁷⁰ In contrast, Dworkin sees the invalidation of unjust laws as the fulfill-
ment of the very end of legitimate government. When injustices are overturned or
avoided, government has achieved its highest aspiration, which is a cause for
moral celebration, not moral regret. Accordingly, his view makes possible an ener-
getic, rather than an apologetic, Constitution.
At the same time that Dworkin cut the constricting bonds of Bickelian restraint
and allowed the circulation to flow into the judiciary by charging it with an active
role in the protection of individual rights, he also exercised care to preserve his
interpretations of the Constitution as interpretations, not rewritings, of that
document. He has consistently refuted the critics’ attribution to him of a claim to
untether the judiciary from all constraint. The caricature of Dworkin’s view has
him arguing that only the judge’s opinion about objective moral reality is perti-
nent to his or her decision on a matter of constitutional interpretation. Dworkin,
however, has steadfastly persisted in his commitment to the “integrity” of the law,
by which he means that judges “must not deploy moral principles, no matter how
much they are personally committed to such principles, that cannot be defended
as consistent with the general history of past Supreme Court decisions and the
general structure of American political practice.”⁷¹ The task, as Dworkin has con-
sistently defined it, is the daunting job of gleaning both principles and their
appropriate application from the Constitution and society that we, at the present
time, have, recognizing that these are the product of influences bearing on the
present from many forces in our past.
This is a job that requires honest acknowledgment of not only the principles that
provide the ideals of a free society, but also the instances in our past that may not
represent our highest attainment of those ideals. This opportunity for judges to be
realistic and use what they know makes possible the open use of judgment and the
critique thereof, actually facilitating public debate about our societal ideals. It is a
dynamic approach to constitutional theory that restores to the Constitution a sense
of purpose and grants to the society that it enables the opportunity to evolve.
Where the prior constitutional theory had struggled with “fake and apology,”⁷²
Dworkin’s was candid. Where prior theory was schizophrenic, Dworkin’s was
single-minded. Where prior theory had sunk to “despair,”⁷³ Dworkin’s was opti-
mistic. Where prior theory accorded no affirmative role to the Constitution in con-
tributing to the aspirations of society, Dworkin’s expressly sought out the “best” of

⁷⁰ Some go so far as to suggest that invalidation of a law on constitutional grounds violates the
rights of those who supported it. See, e.g., R. H. Bork, The Tempting of America: The Political
Seduction of the Law 147 (1990) (when a judge recognizes new constitutional rights, “he violates
not only the limits to his own authority but, and for that reason, also violates the rights of the legisla-
ture and the people.”).
⁷¹ Freedom’s Law, supra note 2, at 319. The fuller concept of law as integrity is explored in Law’s
Empire, supra note 2. ⁷² See supra note 21, and accompanying text.
⁷³ See Black, supra note 20, at 428 (suggesting that “legal acumen has only one proper task—that
of developing ways to make it permissible for the Court to use what it knows; any other counsel is of
despair”).
 How Constitutional Theory Found its Soul 55

our traditions as part of the project of constitutional interpretation. And he recon-

ciled the two faces of academic thought about Brown.
While most of academia acknowledged that Brown had to be right, they appeared
to treat it as some sort of special case that, although inconsistent with their overall
theory of judicial review, could be justified by some special dispensation. Even Robert
Bork, one of the most committed originalists and rights skeptics of all, later conceded
that Brown was right even though it did not square with either originalism or rights
skepticism.⁷⁴ Bickel seemed to think that Brown was right, even though it did not
square with any application of his prescription of passive virtues as guides to judicial
behavior. Wechsler thought Brown was right even though he could not find a “neutral
principle” to justify it. The list could go on. But how strange it is that none of these
scholars appeared to take the admitted rightness of Brown, coupled with the inability
of their own theories to justify it, as evidence that their theories should be reexamined.
Dworkin’s theory not only tolerates Brown, but requires it.⁷⁵ The moral princi-
ple emanating from the Constitution and its history commits the government to
treatment of all its citizens as equals, with equal political status. It was quite obvious
by 1954 that official school segregation was not consistent with such equal status,
even if one chose to believe that in 1866 it had been.⁷⁶ Thus the Supreme Court
had no choice, if acting on the moral reading, but to hold the practice unconstitu-
tional. No beating of the breast was necessary, nor would the decision be a cause for
ambivalence and angst. Moral intuition and law had found reconciliation.
This accession to moral decision-making fit well in what was, after all, a his-
torical period in which explicitly moral issues held an unusually salient place in
public debate. Martin Luther King, Jr. was making his mark on history by level-
ing an explicitly moral challenge to segregation in public facilities.⁷⁷ Other civil
rights leaders called attention, in various ways, to the injustice of race discrimin-
ation, often framing the arguments as a moral imperative for civil rights reform.
The Cold War engaged Americans in explicitly moral comparisons between
democracy and socialism. The controversy over the war in Vietnam, too, was
expressed often in overtly moral terms, on the issue of the justice, both of the
war itself and of the government’s treatment of its citizens in connection with
the war. The feminist movement, also framing much of its message in moral
terms, added more voices to the cacophony of public moral argument that
characterized the times.⁷⁸
Considering the overall tenor of this political climate, it seems all the stranger
that, until Ronald Dworkin came along, most prominent constitutional theorists
⁷⁴ See Bork, supra note 70, at 81–84.
⁷⁵ See Law’s Empire, supra note 2, at 387–389 (discussing how Hercules would resolve Brown).
⁷⁶ Freedom’s Law, supra note 2, at 13.
⁷⁷ See M. Luther King, Jr., Letter from the Birmingham Jail, in Why We Can’t Wait (1964)
(asserting that “segregation is . . . morally wrong” and exhorting victims of injustice to exercise their
“moral responsibility to disobey unjust laws”).
⁷⁸ See G. Steinem, Wisdom Quotes, ⬍www.wisdomrevjone.com⬎ (“Law and justice are not
always the same”).
56 Rebecca L. Brown

of the time had found no place for moral argument in their thinking about society’s
constitutive commitments.⁷⁹ Dworkin’s conception of democracy, by contrast, was
deeply responsive to the moral demands of society at that time, and since.

III. Contributions
For some critics, Dworkin’s insistence upon the existence of right answers stirs
echoes in their minds of the formalists of the early twentieth century.⁸⁰ But the
ears of those who hear that echo deceive them. Dworkin does not claim that the
Constitution determinately answers all questions. To the contrary, he calls con-
stantly for human judgment to be interposed between the issue and the
Constitution for identification of the relevant principle. He explicitly recognizes
that the process of identifying principle “will leave many possibilities open.”⁸¹
“[T]houghtful judges must then decide on their own which conception does most
credit to the nation.”⁸² Right answers do exist, but no one necessarily has a way to
know what they are.⁸³ Their existence provides the aspiration for judges acting in
good faith.
While this may seem superficially detached, it is in fact grounded in American
identity and values. The people who constitute this democracy are given a voice in
constitutional discourse in at least two important ways. First, judges are charged
with finding “our” principles when they seek to apply a constitutional provision to
a case. Remaining faithful to constitutional practice by supplying a filtering func-
tion, the theory finds constitutional principle, not by direct appeal to people’s
preferences, but by interposing the judgment of a judge who is charged with tak-
ing the longer view, to locate principle in whatever evidence is available.
A second way Dworkin draws popular political views into the dynamic reading
of constitutional text is through a vigorous confirmation process. By offering a
robust and complex job description for those appointed to the bench, he provides
much opportunity for discussion of substantive issues and qualifications related to
⁷⁹ This is not to deny the significant contributions, of course, of fellow travelers in the journey
toward more robust recognition of individual rights under the Constitution. Any list of important
theorists who sought to advance the cause of rights during this time must include, among others,
T. Grey, Do We Have an Unwritten Constitution?, 27 Stan. L. Rev. 703 (1975); J. Skelly Wright,
Professor Bickel, the Scholarly Tradition, and the Supreme Court, 85 Harv. L. Rev. 769 (1971); L. Tribe,
Structural Due Process, 10 Harv. Civ. Rts-Civ. Lib. L. Rev. 269 (1975); F. Michelman, The Supreme
Court, 1968 Term—Foreword: On Protecting the Poor Through the Fourteenth Amendment, 83 Harv.
L. Rev. 7 (1969); C. Black, Structure and Relationship in Constitutional Law (1965);
Wellington, Common Law Rules and Constitutional Double Standards: Some Notes on Adjudication,
83 Yale L.J. 221 (1973).
⁸⁰ E.g., M. McConnell, The Importance of Humility in Judicial Review: A Comment on Ronald
Dworkin’s “Moral Reading” of the Constitution, 65 Fordham L. Rev. 1269, 1278 (1997).
⁸¹ Freedom’s Law, supra note 2, at 9. ⁸² Id. at 11.
⁸³ See Law’s Empire, supra note 2, at ix (explaining that “whether we have reason to think an
answer is right is different from the question whether it can be demonstrated to be right”).
 How Constitutional Theory Found its Soul 57

that job description. If a judge is expected to use judgment rather than eschew it,
his or her competence for that position should be tested by inquiries into all the
facets of human judgment that could be expected to influence the judge’s identifi-
cation and application of principle.
The strongest evidence that Dworkin’s is a theory about our Constitution and
our society, rather than some “Utopia” of his own fabrication,⁸⁴ is that he has
made a difference in the law of the land. It seems clear, at the least, that his ener-
getic defense of the role of rights in the constitutional scheme has changed the
terms of the theoretical and doctrinal debates to focus more on which rights to
protect, and how, rather than whether to protect them.
Chronologically, the first major sign that Dworkin was on to something
important in constitutional theory came in a somewhat hostile setting, John Hart
Ely’s extremely influential Democracy and Distrust. This is a book conspicuously
and explicitly committed to rejecting the attribution of substantive values to the
Constitution. Thus, one might not expect to find there a reliance on a core idea of
the one theorist perhaps most closely associated with a substantive reading of the
Constitution.⁸⁵ Yet in an unadorned footnote, the simple citation speaks volumes:
“34. R. Dworkin, Taking Rights Seriously 180 (1977).”⁸⁶ As the following discus-
sion shows, this indispensable citation, no mere flourish, supports the pivot upon
which turns Ely’s argument for a representation-reinforcing understanding of judi-
cial review. In order to show the (unacknowledged) significance of the role of
Dworkin’s idea to Ely’s work, it will be necessary to sketch out a short description of
the intellectual moves that contribute to the thesis of Democracy and Distrust.
Ely’s argument began with the observation that the concept of representation,
as understood by those who founded the American nation, was richer and more
complex than simply a structure of rulers and ruled. Indeed, as he recognized, it
embraced a unique amalgam of the two interests.⁸⁷ The idea was that the legisla-
tors “would live under the regime of the laws they passed and not exempt them-
selves from their operation: this obligation to include themselves among the ruled
would ensure a community of interest and guard against oppressive legislation.”⁸⁸

⁸⁴ Among those who have leveled this argument is J. C. Harrison, Utopia’s Law, Politics’
Constitution (Reviewing Freedom’s Law), 19 Harv. J.L. & Pub. Pol’y 917 (1995–1996).
⁸⁵ See J. Hart Ely, Democracy and Distrust 56–57 (1980) (discussing and rejecting the three-
part proposition that “moral philosophy is what constitutional law is properly about, that there exists
a correct way of doing such philosophy, and that judges are better than others at identifying and
engaging in it.”). ⁸⁶ Id., at 82 n.34.
⁸⁷ Ely’s book is notoriously short on historical support, but on this point it turns out that there is
good evidence of the claim that the American notion of representation was new and unique in ways
critical to the argument Ely was constructing. See R. L. Brown, Liberty, the New Equality, 77 NYU
L. Rev. 1491, 1512–1520 (2002) (exploring the historical foundations and implications of the repre-
sentation argument).
⁸⁸ Ely, supra note 85, at 78 & n.16; see Brown, supra note 87, at 1520–1528 (explaining the com-
munity of interests idea and arguing for its extension to support judicial review of modern-day liberty
claims).
58 Rebecca L. Brown

The principal enforcement mechanism for this obligation of governors to

governed was the possibility of being turned out of office on election day.
What was missing from this “insurance policy,” quite effective at protecting the
interests of most of the people through political accountability, was any protection
for those whose interests might differ from those of most of the people. The oppor-
tunity for electoral correction of unreasonable treatment of those minorities is not
readily apparent in this representative structure.⁸⁹ Ely’s explanation for this omis-
sion was that, for our forebears, the people were an essentially homogeneous group
whose interests did not vary significantly.⁹⁰ The importance of this claimed homo-
geneity, of course, is that it would, if true, largely take care of the minority problem
because “legislation in the interest of most would necessarily be legislation in the
interest of all”.⁹¹ Whatever incidental problems of this sort that might arise would
be taken care of either by the Bill of Rights or by the institutional structures of the
federal government.
Ely suggested that, as time went on in the nation’s early years and vital differences
within the population increased, these structural strategies became more and more
clearly insufficient to protect against the abuse of minorities by majorities. This is
the point at which Ely propounded the central claim of his book: that the traditional
American theory of representation, requiring representatives to honor the interests
of a majority of their constituents and be bound by the laws that bound them, had to
be and was extended so as to ensure that the representatives would, somehow, keep
faith with the interests of various minorities as well. The obvious question was how
this faith would be manifested in a system in which the majority has the power to
make policy and pass laws. It was clear that there would have to be losers in the
democratic process, and that the Constitution could not possibly be offended every
time some law was passed that an outvoted group disliked. Yet if Ely was going to
argue that courts may legitimately (which meant, for him, without evaluating the
merits of government policies or their justifications) police the representative
process for “malfunctions” of that process, then it was critical that he identify what
type of breach, of what obligation on representatives, would indeed constitute such
a malfunction. If the only obligation on representatives was faithfully to carry out
the wishes of the majority of their constituents, assuming all citizens had the oppor-
tunity to vote, then there would be no ground for judges to intervene when the
majority chose to burden groups of smaller numbers.
This presented a cavernous gap for Ely. He had argued, to this point, that it is
not the courts’ job to subject legislation to substantive screening. In this regard he
carried forward the thinking of many of the post-Brown advocates of judicial
restraint. Yet, unlike many of those earlier scholars, he was determined to justify

⁸⁹ Ely, supra note 85, at 78.

⁹⁰ Id. at 79. The accuracy of this statement is subject to some dispute, and Ely acknowledged that
homogeneity of the population was more a republican ideal than a reality, even in the 18th century.
⁹¹ Id.
 How Constitutional Theory Found its Soul 59

both Brown and much of the other judicial activism practiced by the Warren
Court. What the theory needed was a way to understand the process of “represen-
tation” as encompassing an obligation to engage in some basic substantive consid-
eration of the interests of those who may be on the losing side of policy decisions.
Without that more capacious definition of representation, it would seem that a
court limited to representation reinforcement would strike down no law unless
someone were excluded from the process altogether. The biggest challenge to con-
struction of this new understanding of representation was to anchor it in the
Constitution by showing where such an obligation on legislators to minority con-
stituents might reside.
Yet Ely took care of the problem in one bizarre sentence making no direct refer-
ence to the Constitution. The sentence addressed this very issue, the need to
maintain a community of interests between the representative and his or her
minority constituents in the quest to avoid majority tyranny. “Naturally,” wrote
Ely, “that cannot mean that groups that constitute minorities of the population
can never be treated less favorably than the rest, but it does preclude a refusal to
represent them, the denial to minorities of what Professor Dworkin has called
‘equal concern and respect in the design and administration of the political insti-
tutions that bind them.’”⁹²
And so it was that representation-reinforcement theory’s chasm, the need for a
new understanding of representation that could be policed by self-restrained
courts, was bridged by Dworkin’s idea of equal concern and respect. The critical
new definition of “representation”—which, in turn, would supply the boundaries
of legitimate judicial review—was built upon this core insight.
Once across this bridge, Ely was then free to proceed with a theory of represen-
tation reinforcement. If the notion of a democratic government imposes on repre-
sentatives an obligation to accord equal concern and respect to all their
constituents, then any legislation passed out of either disregard for, or intentional
malice toward, any group fails the test for “procedural” integrity and justifies judi-
cial intervention.
In pressing the importance of Dworkin to Ely, my purpose is not to join those
who have criticized Ely for claiming the impossible when he purported to liberate
the Constitution from substantive commitments.⁹³ I seek only to suggest that
Dworkin’s insight as to equal concern and respect is virtually indispensable to the
construction of any kind of theory that seeks to reconcile democracy and constitu-
tionalism, as Ely’s did. Even a constitutional theory that struggles mightily to

⁹² Ely, supra note 85, at 82 (citing R. Dworkin, Taking Rights Seriously 180 (1977)) (empha-
sis in original). An omitted internal footnote mentioned that, for persons to be “represented,” “their
interests are not to be left out of account or valued negatively in the lawmaking process.” Id. at 82 n.33.
⁹³ See L. H. Tribe, The Puzzling Persistence of Process-Based Constitutional Theories, 89 Yale L.J.
1063, 1077 (1980); M. Tushnet, Darkness on the Edge of Town: The Contributions of John Hart Ely to
Constitutional Theory, 89 Yale L.J. 1037 (1980).
60 Rebecca L. Brown

eschew Dworkin finds him pivotal to its effort to escape the leveling crush of pure
majoritarianism—a formidable tribute to a core idea.
But Dworkin’s conquest of Ely, and with him, much of the Legal Process school
of thought, is not the end of the story. As the agony over Brown slowly faded into
the distant past, the would-be heirs to the Legal Process school have undergone
fertilization, infancy and development into distinct, but identifiable, offspring. In
1996, Dworkin wrote Freedom’s Law, dedicated to attacking the “majoritarian
premise,” just as Taking Rights Seriously had styled itself, nearly twenty years
earlier, as an attack on positivism and utilitarianism. In many respects, little had
changed other than the names of the antagonists. The majoritarian premise,
which had led to the call for “strict construction” during the Nixon years, has
supported development or cultivation of such repackaged schools of thought as
judicial minimalism, originalism, pragmatism and the new institutionalism of
administrative law.
Just as Bickel called for judges to exercise their “passive virtues” to keep them-
selves out of societal contests, judicial minimalism calls for courts to leave as much
as possible undecided, leaving room for moral issues to be resolved through the
political processes.⁹⁴ This approach, I have suggested, is not respectful of the equal
status of all citizens, but rather entrenches enacted law without regard to the
strength or validity of the interests of those burdened by the codification of major-
ity preferences.⁹⁵ Despite protestations to the contrary, its failure to require judi-
cially evaluated reasons for divisive legislation renders it strongly majoritarian in
its effect. It is the heir to the “Passive Virtues” name.
The majoritarian premise has infected much of structural constitutional theory
as well.⁹⁶ Administrative law scholarship seeks to justify decisions based on who
makes them, and how accountable the decision-maker is to popular control, rather
than judging them on how well they are protecting the rights of individuals and
achieving good government. Calls for the President to have more control of admin-
istrative agencies, based on an accountability argument, have become a familiar
part of the constitutional separation-of-powers literature.⁹⁷ The administrative
issues involving deference to agency interpretations of statutes reflect the same
trend, both in the Chevron case itself and in the many scholarly paeans to it.⁹⁸

⁹⁴ Sunstein, supra note 40, at 3–6. ⁹⁵ See Brown, supra note 87, at 1538.
⁹⁶ See L. Schultz Bressman, Beyond Accountability: Arbitrariness and Legitimacy in the
Administrative State, 78 NYU L. Rev. 461 (2003) (describing the new views of the administrative
state as profoundly majoritarian).
⁹⁷ See, e.g., L. Lessig & C. R. Sunstein, The President and the Administration, 94 Colum. L. Rev. 1,
94 (1994) (the President must have control over agencies to preserve accountability, one of two “cen-
tral values of the framers’ original executive”); S. G. Calabresi & S. B. Prakash, The President’s Power to
Execute the Laws, 104 Yale L.J. 541 (1994) (arguing for a unitary theory of the executive branch,
based largely on accountability arguments); E. Kagan, Presidential Administration, 114 Harv. L. Rev.
2245 (2001) (arguing that increased presidential involvement in agency decisions can increase the
legitimacy of agencies by increasing majoritarian control of administrative lawmaking).
⁹⁸ See Chevron v. NRDC, 467 U.S. 837 (1984); A. Scalia, Judicial Deference to Administrative
Interpretations of Law, Duke L.J. 511, 518 (1989); L. H. Silberman, Chevron—The Intersection of
Law and Policy, 58 Geo. Wash. L. Rev. 821 (1990).
 How Constitutional Theory Found its Soul 61

Another reincarnation of vanquished opponents is the form of pragmatism that

is wholly consequentialist in its aspirations, which elicits the same responses from
Dworkin that the utilitarian approaches elicited in decades past.⁹⁹ Ultimately, this
approach revives the proposition that individual liberty can and should be sacri-
ficed to strong or powerful claims of the common good, a proposition that
Dworkin continues to refute. And, of course, the originalist/textualist school is
still alive and well, with Justice Scalia at the forefront of those urging those tech-
niques as means to constrain judges from judging.¹⁰⁰ To all these returning
opponents, Dworkin has refined and targeted his responses, but has not wavered
from his basic insights and commitments. He has forced contender after
contender to take him on, but still plausibly can claim that they have repeatedly
failed to justify any alternative to the moral reading.¹⁰¹
Thus Dworkin is not the St. George who valiantly slays the dragon. He is more the
Great Wall that steadfastly keeps the threatening hordes from storming the city, age
after age. His theories are as sound and important in responding to the modern skep-
tics as they were nearly four decades ago. His ideas have been developed and refined
over the years, and his practice of discussing real cases in publicly accessible lay publi-
cations has remained testament to his abiding commitment to the Constitution as a
means to facilitate the self-government that is our constitutional birthright.
In the Supreme Court, the influence of Dworkin’s work can be detected in sev-
eral places. One is the retreat from the representation-reinforcing approach to judi-
cial review of equal protection claims. That theory had been evident in the Court’s
equal protection jurisprudence, both before and after Ely’s book.¹⁰² The develop-
ment of suspect-class doctrine, the tiers of scrutiny, and the concept of motivation
as the touchstone of equal protection all reflect the sense that the goal of judicial
review of equality claims is to ensure that legislative classifications have not arisen
out of malice. But, facing different kinds of claims, the Court and several of its
members individually have retreated from that more process-based approach to
equal protection and displayed a greater willingness to engage in some version of
a moral reading of the Constitution. Whether one can lay credit—or blame—for
this development at Dworkin’s feet is hard to say, but clearly his work contributed
to an intellectual atmosphere in which this elevation of principle over mere repre-
sentational regularity could be put forward with a straight face.

⁹⁹ See R. Dworkin, Darwin’s New Bulldog, 111 Harv. L. Rev. 1718, 1719 (1998) (in challenge to
the claims of Richard Posner, lamenting uncertainty about “when the antitheory episode in our intel-
lectual history will have run its course.”). There are other forms of pragmatism that may be less
squarely in conflict with Dworkin, such as certain forms of practical reasoning which seek to identify
moral norms for use in the resolution of statutory issues. See W. N. Eskridge & P. P. Frickey, Statutory
Interpretation as Practical Reasoning, 42 Stan. L. Rev. 321 (1990).
¹⁰⁰ See A. Scalia, Originalism, The Lesser Evil, 57 U. Cinn. L. Rev. 849 (1989); A. Scalia, The Rule
of Law is the Law of Rules, 56 U. Chi. L. Rev. 1175 (1989).
¹⁰¹ See Freedom’s Law, supra note 2, at 14.
¹⁰² See, e.g., U S v. Carolene Prods., 304 U.S. 144, 152–153 n.4 (1938); City of Cleburne v.
Cleburne Living Center, 473 U.S. 432 (1985) (applying Ely’s factors to deny heightened scrutiny).
62 Rebecca L. Brown

The case I have in mind most conspicuously is Adarand v. Pena,¹⁰³ in which the
Supreme Court held that any government use of race in decision-making must be
subjected to strict scrutiny, regardless of motivation. Significant there was the heed-
lessness of the Supreme Court of the circumstance that the individuals claiming to
be burdened by the minority set-aside provision at issue possessed none of the classic
“suspect” features that would ordinarily lead a Court to fear that their interests
might not get adequate representation in the political process, features that had once
been a pre-condition to a heightened degree of judicial intervention. That would
have been the traditional analysis of such a case under a Bickelian or Elysian legisla-
tive-supremacy, representation-reinforcing view. Yet the Court took an entirely dif-
ferent approach. Instead it identified a moral principle, color-blindness, sought
(unsuccessfully, in Dworkin’s view) to ground it in constitutional text and prece-
dent, and determined that any compromise of that principle was deserving of the
most intrusive level of judicial scrutiny. This case is difficult to explain except on a
moral reading of the Constitution, even if the justices in the case may be criticized
for selecting a principle that is not well defended by reference to our nation’s history
and practices.
Fingerprints of the moral reading can be detected in the joint opinion in
Planned Parenthood v. Casey as well.¹⁰⁴ That opinion reflects a significant depart-
ure both from the earlier case that it purported to affirm (Roe v. Wade,¹⁰⁵ 1973)
and from other fundamental rights precedent. It acknowledged for the first time
“a realm of personal liberty which the government may not enter,” grounded in
American law by some authority other than a strictly textual application of the Bill
of Rights, an authority bearing many indicia of a Dworkinian principle.¹⁰⁶
Dworkin’s thick portrayal of equality continues to exert an important influence
today. The use of a principle of equal concern and respect, as a foundation for
implementing democracy, is as sound and robust today as it was thirty years ago.
Perhaps it is even more important than ever, as the society becomes increasingly
heterogeneous (the reason Ely was forced to make use of the principle), the legisla-
tive and electoral processes increasingly inaccessible, and the challenge of securing
good and fair laws through representative democracy increasingly daunting.
Perhaps most valuable to current constitutional law is the use of equality com-
mitments to give life to the protection of fundamental liberties. A recent Term of
the United States Supreme Court brought a stunning example of the importance
of Dworkin’s substratum of equal concern and respect to one of the divisive issues

¹⁰³ 515 U.S. 200 (1995).

¹⁰⁴ 505 U.S. 833 (1992). ¹⁰⁵ 410 U.S. 113 (1973).
¹⁰⁶ Compare Casey, 505 U.S. at 846 (“neither the bill of Rights nor the practices of States at the
time of the adoption of the Fourteenth Amendment marks the outer limits of the substantive sphere
of liberty which the Fourteenth Amendment protects”) with Griswold v. Connecticut, 381 U.S. 479
(1965) (seeking to ground right to marital privacy in penumbras emanating from textually enumer-
ated rights).
 How Constitutional Theory Found its Soul 63

of the twenty-first century.¹⁰⁷ In Lawrence v. Texas,¹⁰⁸ in which the Court invali-

dated the Texas law prohibiting same-sex sodomy, the Court relied on the protec-
tion of liberty under the Due Process Clause, and expressly declined, indeed, to
resolve the case under the Equal Protection Clause. But Dworkin’s principle of
equal concern and respect was in evidence nonetheless.
The Lawrence Court made a significant departure from the rhetoric and doc-
trine of prior cases raising liberty claims. In times past, it had been common for
the Court to make sharp distinctions between claims based in liberty and those
sounding in equality. Indeed, it was Justice Robert Jackson who, in 1949, warned
that the court should be wary of striking down state laws under the Due Process
Clause, but should be less reluctant to ensure equality.¹⁰⁹ More recently, Justice
Scalia expressed a similar instinct in the Cruzan case, when he rejected the use of
the Due Process Clause as a source of protection for liberty, arguing that it is the
Equal Protection Clause— “which requires the democratic majority to accept for
themselves and their loved ones what they impose on you and me”—that is “our
salvation.”¹¹⁰ For those uncomfortable with judicial review, equality was a more
palatable value than liberty, because it was ostensibly not substantive.
It was this mode of thinking that was responsible for the development of differ-
ent doctrinal frameworks for constitutional claims brought to protect liberty and
those seeking equal protection of the laws. Some cases confounded the distinc-
tion: is a law banning interracial marriage an affront to equality or to liberty? In
those cases, the Court muddled through, adhering as best it could, imperfectly at
times, to its distinct lines of analysis.¹¹¹ These lines had their roots in different
clauses of the Constitution and continued to reflect different intuitions about
judicial review and democratic process.¹¹²
But Dworkin supplies a different way to think about liberty and equality
claims. Starting from a notion of equal concern derived from the concept of
democracy itself, he leads us to an obligation of respect for whatever individual
freedoms are indispensable to the achievement of equal status. Thus, some

¹⁰⁷ A second case in the same Term, Grutter v. Bollinger, 539 U.S. 306 (2003), fulfilled a predic-
tion made by Dworkin in his discussion of whether racial diversity in a university setting would give
rise to a compelling state interest sufficient to justify race-conscious admissions policies. See
Sovereign Virtue, supra note 2, at 420–426 (accurately suggesting that the Bakke principle remains
good law, and recommending an emphasis, borne out in the case, on diversity as the predominant
state interest supporting such programs). I do not take the Court’s opinion in Grutter, however, to be
an application of his moral reading, particularly since it applies strict scrutiny as a result of Adarand v.
Pena, a precedent that Dworkin would not endorse (even if it does employ a moral reasoning of its
own). See id. at 417. ¹⁰⁸ 539 U.S. 558 (2003).
¹⁰⁹ Railway Express Agency, Inc. v. New York, 336 U.S. 106, 112 (1949) (Jackson, J., concurring).
¹¹⁰ Cruzan v. Director, Mo. Dep’t of Health, 497 U.S. 261, 300 (1990) (Scalia, J., concurring).
¹¹¹ See Loving v. Virginia, 388 U.S. 1 (1967) (deciding liberty claim with reference to the misce-
genation law’s affront to equality).
¹¹² See Brown, supra note 87, at 1491–1500 (discussing inter-relations between liberty and equal-
ity, the traditional judicial preference for equality, and the need for a more satisfactory merging of the
two ideals).
64 Rebecca L. Brown

substantive freedoms are derivative of one’s entitlement to be treated by one’s

government with equal respect.
The Supreme Court’s opinion in Lawrence appeared to approach its “liberty”
question with the principle of equal respect in mind. The Court found that the
particular non-public, non-commercial, consensual, private “common” sexual
practices engaged in by the two adult petitioners were “entitled to respect.”
Eschewing the more formalistic inquiry that had been the hallmark of the prior
liberty cases, which asked whether a particular claimed liberty is “fundamental”
under societal traditions and judicial precedent,¹¹³ this Court engaged in the
more substantive inquiry whether the state had “demean[ed] their existence or
control[led] their destiny.” This novel inquiry—and even the language in which it
is expressed, it seems to me—can be traced directly to Dworkin and the principle
of equal concern and respect.
A second way in which the Lawrence opinion both departed from prior case law
and appeared to embrace Dworkin was in its consideration of the state’s proffered
justifications for its law. A reading of prior liberty cases would have led one to predict
that the asserted interests of the state in the case would provide a significant focus of
discussion in the opinion. Accordingly, much speculation had been voiced on
whether the state of Texas would be able, if required, to assert a “compelling interest”
in support of its law. Even taking into account a prior holding that the moral beliefs
of a community could supply a rational basis for a law restricting non-fundamental
liberties,¹¹⁴ it seemed unlikely that such a rationale could be considered compelling
if the Court were to find the right to be, in fact, fundamental.
But the Court never got to that issue, any more than it had gotten to the “fun-
damental rights” determination itself. Having found that the law was demeaning
and that it failed to accord the necessary respect to its citizens, the Court went
on, simply, to hold that the state had proffered “no legitimate state interest which
can justify its intrusion into the personal and private life of the individual.”¹¹⁵
The state had argued in this case that it was entitled to support its law solely by
asserting “the State’s long-standing moral disapproval of homosexual con-
duct.”¹¹⁶ The Court, therefore, had an opportunity to discuss the sufficiency of
moral beliefs to meet doctrinal standards; or, in effect, to balance the strength of
the liberty of the petitioners against the community’s contrary interest, as prece-
dent had suggested it should do. But, departing from decades of jurisprudence
taking this more utilitarian approach, the Court declined to engage in the

¹¹³ See, e.g., Washington v. Glucksberg, 521 U.S. 707 (1997) (holding right of terminally ill to
assisted suicide not fundamental because not deeply rooted in tradition); Bowers v. Hardwick, 478
U.S. 186 (1986) (holding right to engage in homosexual conduct not fundamental because not
deeply rooted in tradition); Michael H. v. Gerald D., 491 U.S. 505 (1989) (plurality finding right of
biological father to relationship with child conceived in adultery not fundamental because not deeply
rooted in tradition).
¹¹⁴ This was the then-novel position taken in Bowers v. Hardwick, 478 U.S. 186 (1986).
¹¹⁵ 539 U.S. at 526.
¹¹⁶ Lawrence v. Texas, No. 02–102, Respondent’s Brief, Feb. 17, 2003, at 41.
 How Constitutional Theory Found its Soul 65

balancing. In the Court’s written opinion, one could almost detect the whisper of
Dworkin’s 1977 warning that government “must not constrain liberty on the
ground that one citizen’s conception of the good life . . . is nobler or superior to
another’s.” This postulate, he had emphasized, “is a conception of equality, not of
liberty . . . .”¹¹⁷ It appears that a quarter-century later, these insights have found
expression in the law of the land.
This aspect of Dworkin’s work shows great promise for addressing one of the
most difficult contemporary constitutional issues arising under the Due Process
Clause, the problem of evaluating the relative strengths of, on the one hand, a
claim to liberty, and, on the other, a state assertion of the need to restrict it. The
Court has been unable to ground this inquiry, on either side of the balance, in any
teleological commitments that would help frame the determination of a statute’s
validity. Wanting to avoid the charge that it is legislating from the bench, it has
struggled to identify some metric against which a claimed liberty right could be
evaluated, but has been unable to settle on one because it has articulated no over-
arching theory of what it is protecting or why. This noncommittal protection of
liberty rights has resulted in an often formalistic fragmentation of liberty into two
categories, some being “fundamental” and others “ordinary.”¹¹⁸ The conse-
quences of attaching one of these labels are huge, yet the normative impulses
underlying the dichotomy are elusive. This leads the Court more and more to
relegate claims to the default category of “ordinary” liberties, leaving them to the
mercies of legislative grace.
When the Court has attempted qualitative evaluation of liberties, its analysis has
been unsatisfying. For example, while acknowledging that some of the liberties it
has recognized as fundamental in the past have had something to do with auton-
omy, the Court, in the assisted suicide case, explicitly rejected autonomy as a defin-
ing feature of a fundamental right.¹¹⁹ The Court has buffeted around the idea of
societal tradition as a way to identify protected interests, but that has caused more
problems than it has solved. This approach to identifying fundamental rights has
persistently sparked the still unresolved controversy about the level of generality at
which to search for societal traditions, as well as the thorny question of which
history to select in cases of conflicting evidence.¹²⁰ Moreover, the tradition method
has never adequately responded to the more normative question why the
backward-looking reliance on past societal practice should provide the limits of

¹¹⁷ Taking Rights Seriously, supra note 2, at 273.

¹¹⁸ See R. L. Brown, The Fragmented Liberty Clause, 41 Wm. & Mary L. Rev. 65 (1999) (arguing
that this fragmentation inappropriately impedes the protection of liberty).
¹¹⁹ See Glucksberg, 521 U.S. at 727–728; (“That many of the rights and liberties protected by the
Due Process Clause sound in personal autonomy does not warrant the sweeping conclusion that any
and all important, intimate, and personal decisions are so protected.”).
¹²⁰ See G. R. Stone, L. M. Seidman, C. R. Sunstein, and M. V. Tushnet, Constitutional
Law 890–891 (4th ed. 2001) (describing dispute between Justice Scalia and Justice Brennan on the
appropriate level of abstraction at which to search for traditions).
66 Rebecca L. Brown

protected liberty going forward in an evolving society.¹²¹ Thus, the Court has not
succeeded in ordering different liberties into an intelligible hierarchy.
On the state’s side of the ledger, the doctrinal picture has been equally dim. The
cases have supplied no framework within which to measure the strength or validity
of state interests offered in support of regulatory restrictions on liberty. Moral argu-
ments, such as that made by Texas in support of its sodomy statute, are especially
difficult to quantify as “rational” or “compelling,” in the absence of some broader
theory addressing the appropriate use of state power in a free society. This lack of
guidance for evaluating the limits of regulatory authority has enabled those on the
side of regulating states to launch a liberty claim of their own. The claim is that the
citizens who support a statutory restriction on liberty have a “right” to enact their
beliefs into law without interference from the Constitution. According to this view,
it is a violation of their liberty to strike down laws they have passed to constrain the
objectionable behavior of others.¹²² The “freedom of association” claim, raised in
opposition to the Brown decision,¹²³ is one version of this argument. The Court,
having constructed no theory of how our Constitution protects liberty, has been
without the means to repudiate the devastating implications of this argument.
Dworkin’s work gives courts the tools with which to begin addressing both
types of deficit in the case law. On the liberty side, the principle of equal concern
and respect suggests some contours for a method of discerning which types of
liberty must be specially guarded in a constitutional democracy. Dworkin has
constructed arguments along these lines in addressing several of the major liberty
issues to arise in recent years; the analysis that these arguments share is the norma-
tive vision of liberty protection as a way to secure a certain minimal degree of dig-
nity and equal moral status for all citizens, a value traceable to both liberty and
equality ideals.¹²⁴ This anchoring of liberty to political theory provides guidance
in the understanding of why some claims might be more compelling than others.
Dworkin solves the problems of level of generality and conflicting traditions, so
irksome to court precedent, with his elements of fit and integrity in the interpreta-
tion of the Constitution.
Dworkin’s approach also provides a start for developing a principled way of
evaluating asserted state interests. His response, in particular, to those who make
the claim described above—that collective self-determination is compromised
by enforcement of individual rights—suggests contours for a method of evaluat-
ing state justifications for laws generally. He does not reject the community self-
determination argument on its face. After all, its claim, that a majority of a

¹²¹ See R. L. Brown, Tradition and Insight, 103 Yale L.J. 177, 200–205 (1993) (discussing several
theoretical objections to use of tradition as defining point for individual rights).
¹²² Freedom’s Law, supra note 2, at 21. Bork, for one, has explicitly made this claim. See supra
note 70.
¹²³ See Wechsler, supra note 17, at 35 (characterizing the “heart of the issue” in Brown as the choice
by a state between denying association to those who wish it and imposing it on those who would
avoid it). ¹²⁴ See Freedom’s Law, supra note 2, at 111.
 How Constitutional Theory Found its Soul 67

community has the right to determine the nature of the entire community through
law, is an essentially moral argument and, as such, must be taken seriously. In
order for the claim to succeed, however, the underlying interest in collective self-
government must itself meet moral standards. Thus, the community seeking to
determine its own values through legal restrictions on liberty must meet the mini-
mal obligation to respect the needs and prospects of all members of the affected
community. Accordingly, just as a community would obviously not have a moral
entitlement to employ its right to self-government so as to enslave some members
of the community for the benefit of the others (even if 99 per cent voted to enslave
the remaining 1per cent), it would also be precluded, more generally, from dictat-
ing what its citizens think about matters of political, moral or ethical judgment,
no matter how many members might share one view.
This recognition of a basic moral obligation on government provides a way of
beginning to distinguish and evaluate asserted government interests. A require-
ment that all laws accord equal concern and respect to all members of a community
can begin the process of distinguishing laws that legitimately seek to protect the
public good from those that simply impose contested visions of the good life on an
unwilling minority. Although admittedly difficult, this inquiry is far more substan-
tive and meaningful than the apparently ad hoc jurisprudence that led the Supreme
Court, without elaboration, to pronounce one day that a community’s moral
preferences are sufficient to support restraints on individual liberty (in Bowers v.
Hardwick), and then another day that they are not sufficient (in Lawrence v. Texas).
Dworkin’s work makes possible a meaningful, normative inquiry, and thus holds
the key to a coherent liberty jurisprudence for the twenty-first century.

IV. Conclusion

In the end, it is a question of optimism or pessimism. The schools of constitu-

tional thought to which Dworkin has sworn eternal opposition are characterized
by skepticism, cynicism, distrust, and pessimism. Dworkin’s works, by contrast,
have fought valiantly for the proposition that the American people can do better,
that our form of government was conceived under the sun of enlightenment, that
it committed us to the ideals expressed in the Preamble of the Constitution.¹²⁵
The hope of true equality, so far elusive to our society, is within our grasp, he
assures us: “we must hold to the courage of the conviction . . . that we can all be
equal citizens of a moral republic.” ¹²⁶ With his enduring faith, Ronald Dworkin
helped to reclaim for constitutional theory its itinerant soul.

¹²⁵ See E. J. McCaffery, Ronald Dworkin, Inside-Out (reviewing Freedom’s Law), 85 Cal. L.
Rev. 1043, 1058 (1997) (describing Dworkin’s resolute nonskepticism and belief in moral and social
improvement). ¹²⁶ Freedom’s Law, supra note 2, at 38.
 4
Coherence, Hypothetical Cases, and Precedent
S. L. Hurley*

Coherence accounts of practical reasoning in general, of which legal reasoning is

a particular case, postulate the existence of a theory, sought in deliberation,
which best displays as coherent the relationships among specific reasons for
action that conflict in application to the case to be decided. These relationships
are discovered in part by considering how the same conflicting reasons apply in
other cases, the resolution of which is settled. Ronald Dworkin’s account of legal
reasoning in Law’s Empire is an example of a coherence account. Coherence
accounts provide elements of various familiar conceptions of practical reasoning,
such as the Rawlsian conception of reflective equilibrium as the outcome of a
process of mutual adjustment between theory and intuition, and the decision-
theoretic conception of rationality in terms of adjustment between data about
preferences, criteria of choice, and principles of consistency.¹ I have elsewhere
discussed the general rationale and motivation for coherence accounts of prac-
tical reasoning, illustrated the process of seeking coherence through deliberation
with both legal and ethical examples, abstractly characterized the search for
coherence, and considered various problems to which the abstract characteriza-
tion gives rise.² In this paper I shall:
First, briefly set out my abstract characterization of deliberation as a search for
coherence;

* For helpful suggestions and comments on earlier drafts that saved me from various errors and
unclarities, I am grateful to Ronald Dworkin, Mark Greenberg, and Joseph Raz: needless to say,
responsibility for all remaining errors is entirely my own. This paper was originally, published in
10 (2) OXFORD J. LEGAL STUD. 221–251 (1990) and is reproduced here with permission.
¹ See J. RAWLS, A THEORY OF JUSTICE (1971); and R. L. KEENEY AND H. RAIFFA, DECISIONS WITH
MULTIPLE OBJECTIVES: PREFERENCES AND VALUE TRADEOFFS (1976).
² See my NATURAL REASONS (1989), chapters 4, 10, 11, 12, etc. Many background issues about
the nature and objectivity of coherentist practical reasoning are raised by the argument of this article
which I cannot here address at length in NATURAL REASONS. They include: the need for substantive
constraints on the description of the problem in terms of conflicting reasons, the relationships of
these constraints to the structure and process of coherentist deliberation and the explanatory charac-
ter of the resulting theories, the nature and limitations of the authority of theories about conflicting
reasons, and many others.

Coherence, Hypothetical Cases, and Precedent. S. L. Hurley.

© Oxford University Press 2006. Published 2006 by Oxford University Press.
70 S. L. Hurley

Secondly, consider various further examples of legal and ethical reasoning and
indicate briefly the way in which they provide instances of my general coherentist
characterization of deliberation;
Thirdly, point out the generally similar roles within coherentist accounts of
practical reasoning of settled actual cases and settled hypothetical cases as data to
be accounted for the sought-after theory or principles;
Fourthly, present an objection to coherence accounts of legal reasoning in
particular discussed by Kenneth Kress, who argues that they may give rise to
retroactive application of legal principles when new cases are decided between
the occurrence of the events litigated and litigation of them (the problem of
intervening cases); and
Finally, consider various possible responses to Kress’s argument. I shall briefly
evaluate and put aside the possible responses of giving intervening cases prospec-
tive effect only, and of accepting intervening case retroactivity as not so bad. I shall
then go on to diagnose the source of the problem by evaluating the relationships
among concept of coherence, hypothetical cases, and the doctrine of precedent.
I will show that the weaker coherentist requirement of treating like cases alike
within practical reasoning in general is not sufficient to give rise to the problem of
intervening cases, but that the stronger requirement imposed by the doctrine of
precedent within legal reasoning in particular is necessary for the problem to arise.
The distinction between the weaker and stronger requirements turns on the
asymmetrical treatment of actual and hypothetical cases by the doctrine of
precedent, by contrast with their symmetrical treatment in coherentist practical
reasoning in general. However, I shall then argue that, when this asymmetry is
properly understood, neither is the doctrine of precedent by itself sufficient to give
rise to the problem of intervening cases, but only does so when the intervening
legal decision is not ex ante correct, and that the role of hypothetical cases in
coherentist practical reasoning in fact limits the problem of intervening case
retroactivity. I shall conclude, following suggestions made by Dworkin, that
intervening case retroactivity does not present a problem for coherence accounts
such as his in particular.

I. An Abstract Characterization of Deliberation,

and Coherence Functions
Our subject matter is deliberation about what to do when the reasons that apply
to the alternatives conflict. The alternatives may be possible individual actions
(e.g. keeping one’s promise and risking someone’s life, or breaking it and avoid-
ing risk to life), or possible judicial actions (e.g. holding that a statute violates
the Equal Protection Clause, or that it does not). The reasons that apply to the
alternatives may be of various kinds, depending on the context of deliberation.
 Coherence, Hypothetical Cases, and Precedent 71

An individual may deliberate about his own conflicting self-interested reasons,

when no one else is significantly affected by the alternatives, or about a conflict
involving the interests of others; the applicable reasons may be ethical, aesthetic,
etc. A judge deliberates about conflicting legal reasons: legal doctrines, prece-
dents, rights, principles, policies, as expressed within various legal practices and
institutions, by cases, statutes, a constitution, legal scholarship, etc.
Coherence accounts claim that to say a certain alternative ought to be done is
to say that it is favoured by the theory, whichever it may be, that gives the best
account of the relationships among the various specific reasons that apply to the
alternatives in question. Which theory is in fact the best theory, which best
displays the applicable reasons as coherently related, is not specified by a coher-
ence account per se. Rather, it is left to deliberation to discover what the best
theory is. A coherence account merely claims that what ought to be done is
whatever alternative the best theory, whatever that is, favours. Thus the delib-
erator’s task is to discover the theory that best displays coherence. Deliberation
involves a process of constructing hypotheses about the content of what I call a
“coherence function’,” which represents the theory sought and which takes us
from alternatives ranked by specific reasons to all-things-considered rankings of
alternatives. Of course the coherence function must meet certain conditions,
and the all-things-considered ranking must meet certain conditions. For
example, the latter ranking should not be intransitive. However, I shall not be
considering the full range of these conditions in this article, but only those
embodied in the requirements of treating like cases alike and of the doctrine of
precedent.³
The process of deliberation can be analysed into five stages. At the first stage,
we specify the problem. We try to arrive at a characterization of the alternatives at
issue, and to determine what various specific reasons apply and how they rank
the alternatives. We here exercise our abilities to perceive the world in terms of
ethical, or legal, or other reason-giving concepts. (This is the issue spotting stage
familiar to law students; however, there is much more to legal deliberation than
issue spotting.) Assuming the reasons that apply conflict, that is, rank the
alternatives at issue differently, we proceed to the second stage, and examine more
carefully the various specific reasons that apply. Perhaps when we consider the
purpose of one reason, it will turn out to have a rather different import than
we originally thought in this particular case. At any rate, at this stage we develop
and firm up our local conceptions of the various specific reasons that apply,
without yet trying to arrive at a global conception of their relations to one
another.
At the third stage we begin gathering data, by looking for other issues to which
the conflicting reasons examined at stage two apply. In particular we are looking

³ For discussion of such conditions, see my NATURAL REASONS, supra note 2, chapter 12.
72 S. L. Hurley

for settled cases. They may have actually been decided, or may be posed as hypo-
thetical issues, the resolution of which can be taken as evident. By a “settled”
case, I mean a case which, if actual, is such that its resolution is clear to the
relevant decision-maker or decision-makers, and which, if hypothetical, is such
that its resolution would be clear to the relevant decision-maker or decision-
makers were the case to be considered. (That a resolution of a particular case is or
would be clear does not mean that it cannot be mistaken; settledness in particular
cases is a matter of what is or would be believed to be correct, not necessarily of
what is correct. However, it is of the nature of a coherentist account of what
should be done, in terms of coherence with settled cases in general, that not all
settled cases can be mistaken. I will return to this point below.) Thus, at least as a
conceptual matter, not all actually decided cases are settled, and some settled
cases are hypothetical, not actual (see Section VIII below for more on this use of
“settled”). We may give settled actual cases more weight than settled hypothetical
cases, or we may give them equal weight. The doctrine of precedent in law gives
settled actual cases more weight than settled hypothetical cases, though usually
only when the settled cases are those of the same court or higher courts of the
same jurisdiction; actual cases of lower courts or courts of other jurisdictions may
be treated on a par with settled hypothetical cases. However, as we shall see, this
difference in weight may only be significant under certain circumstances.
The fourth stage is the heart of the deliberative process. At this stage we engage
in all-out theorizing, looking for hypotheses which account for the resolutions of
issues we found at stage three. That is, we are trying to formulate hypotheses
about the relationships between the conflicting reasons under various different
circumstances present in the stage three cases, which account for those resolu-
tions. To this end we examine the stage three cases for distinctive circumstances or
dimensions which seem to enhance or diminish the weight of one of the conflict-
ing reasons in relation to the other. When we have formulated such an hypothesis,
we try to test it, by going back to stage three and looking for further settled cases in
which the same reasons apply and in which the circumstances identified by the
hypothesis are present. We thus go back and forth between stages three and four,
looking for settled actual and hypothetical cases that help us to refine our
hypotheses about the relationships between the conflicting reasons in various
circumstances.
Finally, at the fifth stage, we work out the consequences of the best hypothesis we
have arrived at for the original case at issue. That is, we apply that hypothesis about
the relationships between the applicable reasons to the circumstances present in the
case at issue. This hypothesis is a partial specification of a coherence function, which
takes us from the rankings of alternatives involving various circumstances or dimen-
sions by the conflicting reasons to an all-things-considered ranking.
This characterization of deliberation can be pictured in terms of a deliberative
matrix. The data gathered at stage three can be represented as follows, where in each
row, alternatives are ranked above or below one another by the applicable reasons.
 Coherence, Hypothetical Cases, and Precedent 73
Table 4.1.: The deliberative matrix

Resolution, all-things-
Reason X Reason Y considered

case at issue: alt a alt b ?

alt b alt a
settled actual cases: alt d alt c alt c
alt c alt d alt d
alt e alt f alt e
alt f alt e alt f
settled hypothetical alt g alt h alt h
cases: alt h alt g alt g
alt j alt i alt j
alt i alt j alt i
etc. etc. etc.

I present the matrix with only two conflicting reasons, X and Y, merely for
convenience; there is no restriction on the number of reasons that may be repre-
sented. (Indeed, multi-dimensional conflicts give rise to interesting theoretical
questions.)⁴ During stage four the alternatives are analysed and more fully char-
acterized in terms of various circumstances and dimensions of the cases which
may help to explain their resolution. These circumstances may be represented
by adding propositions, p, q, etc, to the alternatives in each row. The content of
such a proposition may be quantitative or non-quantitative. A hypothesis
would then take the form: “Reason X tends to outweigh Reason Y when it is the
case that p, while Reason Y tends to outweigh Reason X when it is the case that
q; when it is the case that both p and q, but not r, Reason X has more weight, but
when r is present as well, Reason Y has more weight,” and so on.⁵
⁴ Which I pursue in NATIONAL REASONS, supra note 2, chapters 12 and 13.
⁵ This schematization, applied to legal deliberation, may be compared to that employed by the
programme HYPO, developed by Edwina Rissland and her student Kevin Ashley, and discussed in a
Harvard Law School seminar on artificial intelligence and legal reasoning, conducted by Rissland
Autumn Term 1987. See K. D. Ashley, Modelling Legal Argument: Reasoning with Cases and
Hypotheticals, PhD dissertation, 1987, University of Massachusetts, Dept of Computer and
Information Science. While there are many differences in detail between the two approaches, I do
not believe there is any incompatibility in principle between them. The role of HYPO’s “dimen-
sions” is similar to the role of my propositions p, q, etc in the analysis of actual and hypothetical set-
tled cases at stage four. Perhaps I try to say a bit more than Rissland and Ashley do about the role of
hypotheticals in reaching legal conclusions, in that the answers to hypothetical questions feed back
into the resolution of the case at issue via the coherence function, but again I do not believe that
what I say is incompatible with their approach. Perhaps the most striking difference is that over
whether to use favourableness to conflicting legal doctrines or favourableness to plaintiff as opposed
to defendant, as the basic means of organizing the data. Often, within a narrowly limited area of the
law, such as trade secrets law, plaintiffs will typically represent one legal doctrine, and defendants
another, so that the two approaches are in principle quite similar. However, the doctrine-oriented
rather than party-oriented method of organization may have advantages when one comes to gener-
alize beyond a narrow limited area of the law, so that plaintiff and defendant no longer typically
74 S. L. Hurley

Two points of clarification may be helpful. First, I do not in this paper aim to
give a full account of what is distinctive about legal reasoning in particular, but
rather to consider what follows about legal reasoning from the fact that it has the
general features of practical reasoning, understood along coherentist lines. My
examples in the next section are intended to illustrate the application of the
coherentist account of practical reasoning in general to legal problems, rather
than to illuminate what is distinctive about legal reasoning in particular. I will go
on to consider the distinctively legal doctrine of precedent in the context set by a
view of legal reasoning as having the general features of coherentist practical
reasoning. This is important to understand for purposes of my discussion below
of the problem of intervening cases, since my eventual response to Kress in
Sections VII and VIII turns on the features that legal reasoning shares with prac-
tical reason in general, on a coherentist view of it, with respect to the role of
hypothetical cases. Such general features of coherentist practical reasoning limit
the effects of the distinctively legal doctrine of precedent in giving rise to inter-
vening case retroactivity, and thus illuminate the source of the problem. But I
would certainly not claim that the generally coherentist character of legal reason-
ing as a kind of practical reasoning plus the features of precedent I consider
provide a full account of what is distinctive about legal reasoning in particular;
this is not my purpose. Secondly, I would expect the account and illustrations of
practical reasoning to be controversial to the extent they represent a kind of
deliberative rationality with respect to conflicting ends, values or reasons, the
possibility of which has often been denied, for example, in favour of a view of
practical reasoning as exclusively instrumental, or in favour of more radically
sceptical or nihilistic views about practical rationality (in the legal context, con-
sider certain views associated with the critical legal studies movement, or par-
odies thereof ). However, I cannot join these issues here.

II. Some Examples of Legal and Ethical Deliberation Analysed

I have elsewhere considered in detail and at length a legal illustration of this gen-
eral account of practical reasoning, the stages of deliberation, and the deliberative
matrix. That discussion involved deliberation about the relationships between the
conflicting legal doctrines of estoppel and of consideration in cases now usually

represent particular legal doctrines. By organizing the cases according to legal doctrines directly
rather than by plaintiffs’ and defendants’ positions, one may hope to keep theoretical score as one
moves from one area of law to another in which the same doctrines apply, and to bring insights
about the relationships between legal doctrines from one area to the next.
Of course, different ways of perceiving what legal doctrines apply will yield different analyses, but
that is the way the law is, and an analysis which reflects this relativity of conclusion to starting point
may be illuminating. Moreover, we can in principle start with as many different legal doctrines as we
think may be relevant; again, there is no need to restrict the number of reasons weighed against one
another by the coherence function to two.
 Coherence, Hypothetical Cases, and Precedent 75

covered by the doctrine of promissory estoppel.⁶ I will not here give another
lengthy and detailed illustration, but rather will give several sketchy illustrations.
I hope in the former discussion to have persuaded readers that my account can be
made to work in detail. In this discussion I rather aim to persuade readers that the
account applies readily and intuitively in a wide range of cases. Accordingly I shall
not work out the details of my applications here, but shall merely briefly indicate
how the framework of my account would fit the examples.
Let us begin with the case of California v. Carney.⁷ In Carney, Fourth
Amendment issues were raised by a warrantless police search of a motor home,
parked in a downtown San Diego parking lot. The police had reason to believe
that Carney, the owner of the motor home, was exchanging marijuana for sex
acts. They observed a young boy enter the motor home and leave again an hour
and a quarter later. On questioning by the police, the boy said that such an
exchange had just occurred. The police then knocked on the door, and when
Carney came out they entered the motor home without a warrant and found
marijuana. A further search of the vehicle at the police station revealed more
marijuana, in cupboards and in the refrigerator. The parking lot where the war-
rantless search occurred was a short distance from a courthouse where a warrant
could easily have been obtained.
The case and its oral argument, rich in hypothetical cases, have been analysed
by Edwina Rissland to illustrate the idea of a “dimension,” which is used by the
case-based reasoning programme HYPO to generate hypothetical cases.⁸ My
treatment of the case essentially adapts her analysis to my framework, and illus-
trates, I believe, the compatibility of my framework and the notions of a deliber-
ate matrix and a coherence function, with Rissland’s dimension-based analysis.
As she points out, Carney is a case in which there is a conflict between a citizen’s
expectations of privacy, protected by the Fourth Amendment’s prohibition of
unreasonable searches and seizures, and the responsibilities and desires of the
police to investigate and control drug use and other prohibited activities. Thus,
the first row of our matrix for the case at issue involves at least two kinds of legal
reasons: I shall refer to them as reasons of Privacy, and reasons of the Police
Power. The alternatives in the case at issue are to allow or to disallow the warrant-
less search of the motor home in Carney under the Fourth Amendment. Reasons
of Privacy would favour disallowing the search, while reasons of the Police Power
would favour allowing it. We want to discover how reasons of Privacy and
reasons of the Police Power are related to one another with respect to warrantless
searches. Settled actual cases at the next several rows of the matrix tell us that
reasons of Privacy are augmented relative to those of the Police Power in circum-
stances in which the warrantless search at issue is of someone’s home; if we sum

⁶ See NATURAL REASONS, supra note 2, chapter 11.

⁷ California v. Carney, 471 U.S. 386 (1985). ⁸ See supra note 3.
76 S. L. Hurley

up these circumstances in proposition p, we can hypothesize that Privacy is

augmented relative to the Police Power when it is the case that p. By contrast,
settled cases tell us, the Police Power is augmented relative to Privacy when the
warrantless search at issue is of a vehicle; if we sum up the latter circumstances in
proposition q, we can hypothesize that Privacy is diminished relative to the
Police Power when it is the case that q. We have thus begun to specify a coherence
function which weighs Privacy against the Police Power in various circumstances.
While these hypotheses about the relations between Privacy and the Police
Power are well supported by settled cases, they really only set the stage for the
Carney problem, since the motor home in Carney is both a home and a vehicle.
We need a finer analysis of relevant circumstances in order to resolve the conflict
in this case. Thus we look for other settled cases, actual or hypothetical, in which
these two reasons come into conflict, in circumstances in which the home-vehicle
distinction is difficult to draw. On the basis of such cases, we try to arrive at a
hypothesis that will help police to draw the home-vehicle distinction in a reliable
and straightforward way in cases in which aspects of both home and vehicle are
present; that is, we would like to find a “bright-line” distinction, which will not
require police officers to make excessively subtle or difficult determinations. We
will try various hypotheses about the relevant circumstances, or dimensions of
cases, and will test them against various actual and hypothetical cases.
The fact that the thing to be searched is a vehicle seems to augment the Police
Power reason relative to those of Privacy in part because it is inherently mobile,
such that requiring a warrant would leave it time to “get away” and would greatly
frustrate legitimate exercises of the Police Power, and also because vehicles are not
usually used as homes, and thus not the objects of normal expectations of privacy.
These hypotheses are developed and tested in oral argument by reference to hypo-
thetical cases in which vehicles are also homes, but some of which are more inher-
ently mobile, or more home-like than others, owing to varying circumstances and
dimensions. The presence of wheels is obviously dispensable to inherent mobility
in the case of most boats, with respect to which presumably reasons of the Police
Power predominate. Nor is the presence of wheels decisive in the case of motor
homes with wheels but which are not self-propelling and which are permanently
connected to utilities supplies, with respect to which presumably reasons of
Privacy predominate. Location in a temporary parking lot as opposed to a perma-
nent motor home park strengthens the case with respect to inherent mobility and
hence with respect to the weight of Police Power reasons. But a counter-example
to that hypothesis would seem to be the case of someone very poor who cannot
afford a permanent spot for his vehicle but lives in it behind curtains, on the move
from day to day. The presence of the normal accoutrements of a home, such as a
refrigerator, in a vehicle, may strengthen reasons of Privacy; but again we would
seem to have a counter-example when the refrigerator is used solely to store
marijuana, as seemed to be the case in Carney, and not for home-like purposes.
 Coherence, Hypothetical Cases, and Precedent 77

From among these various circumstances and dimensions of hypothetical

cases, the best hypothesis will focus on objective factors that are straightforward
for the police to ascertain before deciding whether to search without a warrant
or not (as the contents of a refrigerator are not), so as to provide police with
“bright-line” guidance. This is because the investigation needed to apply an
excessively subtle distinction will make for a prior, unprincipled defeat of rea-
sons of Privacy, while a strong presumption in favour of home-likeness to
accommodate all unusual uses of vehicles as homes would in effect eliminate the
vehicle exception to the warrant requirement and would thus excessively com-
promise the Police Power. Thus the need for bright-line guidance in effect pro-
vides a way for the theory to get around the apparent counter-example to
inherent mobility hypothesis provided by the poor person who lives in his car.
The case does not really provide a counter-example, after all, since the police
cannot tell that such unusual vehicles are homes without privacy-infringing
investigation; and it is already settled by the vehicle exception that the alterna-
tive of disallowing searches of any vehicles which might possibly be homes has
been ruled out. The majority opinion in Carney indicates in effect that the best
hypothesis makes the relation between reasons of Privacy and those of the Police
Power in cases involving vehicles that may be homes turn on objective indica-
tions of mobility and use for transportation, rather than on the more difficult-
to-ascertain home-like uses of the vehicle. If the presence of the former
characteristics is expressed by proposition r, the majority’s hypothesis is that
when it is the case that r, as it was in Carney, reasons of the Police Power out-
weigh reasons of Privacy, and thus support the warrantless search at issue. The
majority implicitly regard this hypothesis as supported by the hypothetical cases
considered, given the need for bright-line guidance. By contrast, the dissenting
opinions seem to prefer a hypothesis which gives more of a role to the home-like
character of the place in strengthening reasons of Privacy.
Equal Protection cases raised under the US Constitution are a fertile source of
illustrations of the coherentist framework, as the jurisprudence of varying
degrees of relationship to various state interests served by challenged classifica-
tions lends itself immediately to representation within a deliberative matrix. In
discussing several such cases I shall not stop to assign labels, p, q, etc., to the
propositions employed in hypotheses about the relations among the conflicting
reasons in various circumstances. I shall merely identify the general categories of
reasons in play, and try to draw out the hypotheses about the relations between
the reasons in play that are implicitly tested by appeal to various settled cases. It
should be evident, however, how labels could be assigned and hypotheses located
within a deliberative matrix. In Equal Protection cases one category of legal
reasons is provided by the Equal Protection Clause itself: reasons of
Nondiscrimination, which oppose legal classifications which fail to treat
similarly situated citizens in similar ways. Further categories of reasons are
78 S. L. Hurley

provided by whatever state interests are supposed to be served by the challenged

classification.
In Parham v. Hughes,⁹ for example, the father of an illegitimate child challenged
a law which prevented him from recovering for the wrongful death of his child,
where an unmarried father who had filed papers to legitimize the child, or the
mother of the illegitimate child, would have been allowed to recover. If he could
be regarded as similarly situated with other parents of illegitimate children who
were allowed to recover for wrongful death, then reasons of Nondiscrimination
would favour striking the statute down. However, the state considered that it had
an interest in promoting the integrity of the legitimate family unit, and that rea-
sons of Legitimacy favoured restricting recovery for wrongful death to unmarried
fathers who had filed to legitimize their children and unmarried mothers. The
implication seems to be that such a restriction would provide incentives to unmar-
ried fathers to file to legitimize their children. Moreover, the state also claimed to
have an interest in avoiding difficult problems of proof of paternity, and that rea-
sons of Provability favoured the restriction of wrongful death suits for the death of
children born illegitimate to fathers who had filed to legitimize and mothers,
whose identity would rarely be in doubt. Thus we seem to have a conflict between
reasons of Nondiscrimination, on the one hand, and reasons of Legitimacy and of
Provability, on the other.
The court held that this state scheme did not violate Equal Protection or Due
Process by discriminating against unmarried fathers relative to unmarried mothers,
because fathers and mothers of illegitimate children are not similarly situated since,
under Georgia law, only fathers can by voluntary unilateral action make an illegiti-
mate child legitimate. That is, the plurality opinion seems implicitly to support the
hypothesis that when it is the case that men and women are not similarly situated
owing to a difference in their legally imposed abilities and status, that difference may
provide a basis for a statutory classification which does not constitute invidious dis-
crimination, such that, when such a classification is related in the right way to the
right kinds of state interests, the latter outweigh reasons of Nondiscrimination.
The dissent appeals to a hypothetical case to refute the plurality’s hypothesis.
Justice White writes:
There is a startling circularity in this argument. The issue before the Court is whether
Georgia may require unmarried fathers, but not unmarried mothers, to have pursued the
statutory legitimization procedure in order to bring suit for the wrongful death of their
children. Seemingly, it is irrelevant that as a matter of state law mothers may not legiti-
mate their children, for they are not required to do so in order to maintain a wrongful-
death action. That only fathers may resort to the legitimization process cannot dissolve
the sex-discrimination in requiring them to. Under the plurality’s bootstrap rationale, a
State could require that women, but not men, pass a course in order to receive a taxi
license, simply by limiting admission to the course to women.

⁹ 441 U.S. 347 (1979).

 Coherence, Hypothetical Cases, and Precedent 79

In the taxi licence hypothetical, men and women would not be similarly situated,
owing to their differing legally imposed abilities, i.e. their differing abilities to
enrol in the course. But here it is clearly absurd to suppose that this prior discrim-
ination, unscrutinized, could provide a justification for further discrimination in
the form of a statutory classification requiring only women to pass the course in
order to get a licence. Thus the hypothetical case provides a counter-example to
the plurality’s hypothesis. The implication of the dissenting opinion is that the
embedded distinction must itself be scrutinized, and not merely taken as given. In
Parham, the embedded distinction was one with respect to the ability to file papers
to legitimize: only men were able to do so. It is hard to see how a rule making it
impossible for women to file papers to legitimize children serves either the state
interest in promoting the legitimate family unit or in avoiding proof of paternity
issues. Permitting women to file as well as men would have no adverse effects
whatsoever on either of these state interests.
The dissent goes on to appeal to another counter-example, this time a type of
actual rather than hypothetical case, to the hypothesis that reasons of Legitimacy
outweigh those of Nondiscrimination if allowing members of one class but not
another to sue for wrongful death serves the state interest in promoting the
integrity of the legitimate family unit. The dissent points out that unmarried
mothers and fathers who file to legitimize but remain unmarried defy the integrity
of the legitimate family unit, just as do fathers who fail to file, but the former are
allowed to sue for wrongful death while the latter are not.
In Craig v. Boren, the question was whether a statute prohibiting the sale of
weakly alcoholic beer to males under 21 and females under 18 constituted denial
to males aged 18 to 21 of equal protection of the law. On behalf of the statute it
was urged that statistical evidence about the relative tendencies of males and
females aged 18 to 21 to drink and drive supported the gender line. Thus we have
a conflict between reasons of Nondiscrimination and the state interest in
Preventing Drunk Driving. The Court rejects the hypothesis that reasons of
Nondiscrimination can be overcome by the interest in Preventing Drunk Driving
when the latter is served by “statistically measured but loose-fitting generalities
concerning the drinking tendencies of aggregate groups.” In doing so it appeals to
hypothetical variations on the Craig v. Boren statute involving statistically
supported ethnic or racial lines instead of gender lines aimed at alcohol regulation.
The assumption made is that such ethnic or racial lines would not be acceptable,
despite statistical support, and that reasons of Nondiscrimination would predom-
inate in such cases: “In fact, social science studies that have uncovered quantifiable
differences in drinking tendencies dividing along both racial and ethnic lines
strongly suggest the need for application of the Equal Protection Clause in
preventing discriminatory treatment that almost certainly would be perceived as
invidious.” The Court illustrates its comments with reference to statistical
evidence to the effect that Jews, Italian Catholics, and black teenagers tend not to
be problem drinkers, in contrast to whites and North American Indians.
80 S. L. Hurley

Finally, consider the uses of hypothetical and actual cases in ethical argument,
as illustrated by the debate between Stephen Pepper and David Luban over
whether lawyers in an adversary system should help their clients to do legal but
unethical acts. Pepper argues that under many circumstances the answer is “yes.”
He argues roughly as follows. We hold the value of individual autonomy to be
more important than getting people to do the ethically right act in a wide range
of cases; for example, we allow that people should have the legal power to disin-
herit a child for marrying outside the faith, even though we may agree that it
would be wrong to do so. Individual autonomy in our complex society requires,
in many cases, legal assistance; only with the help of lawyers can people, in many
circumstances, be “first class citizens.” Therefore lawyers should give their clients
the legal help required for individual autonomy even when it permits them to do
unethical acts.
Luban replies by arguing that, while it is true that, since exercising autonomy is
good, helping people exercise autonomy is good, this is only half the story. “The
other half is that since doing bad things is bad, helping people do bad things is
bad. The two factors must be weighed against each other, and this Pepper does not
do.” That is, Pepper’s general hypothesis that when helping someone to do a legal
but unethical act is favoured by reasons of autonomy, reasons of autonomy
prevail, is too crude. Luban appeals to an analogous hypothetical case to defeat the
unqualified hypothesis:
Compare this case: The automobile, by making it easier to get around, increases human
autonomy; hence, other things equal, it is morally good to repair the car of someone who is
unable by himself to get it to run. But such considerations can hardly be invoked to defend
the morality of fixing the getaway car of an armed robber, assuming that you know in
advance what the purpose of the car is. The moral wrong of assisting the robber outweighs
the abstract moral goodness of augmenting the robber’s autonomy.¹⁰
Not only may reasons of autonomy sometimes be outweighed by the wrongness of
the act in question, but even if the balance of reasons favours allowing the agent to
do the unethical act himself, it does not necessarily follow that it will also favour
helping him to do it. The balance between autonomy and conflicting reasons may
be struck differently in different circumstances, i.e. with respect to omitting to
prevent as opposed to positively aiding.
Another argument Pepper makes is that allowing lawyers to weigh the wrong-
ness of acts and thus act as screens to filter out certain legally permissible acts is to
submit individuals to “rule by an oligarchy of lawyers.”¹¹ The implication of the
term “oligarchy” is that such weighing and screening by lawyers would constitute
an elitist centralization of ethical decision-making highly threatening to the

¹⁰ D. Luban, The Lysistratian Prerogative: A Response to Stephen Pepper, AMERICAN BAR

FOUNDATION RESEARCH JOURNAL 639 (1986).
¹¹ S. Pepper, The Lawyer’s Amoral Ethical Role: A Defense, A Problem, and Some Possibilities,
AMERICAN BAR FOUNDATION RESEARCH JOURNAL 617 (1986).
 Coherence, Hypothetical Cases, and Precedent 81

value of individual autonomy, and that, when this would be the result, reasons of
autonomy should prevail. Luban offers a counter-example to this implied
hypothesis:
. . . there is no oligarchy of lawyers, actual or potential, to worry about. An oligarchy is a
group of people ruling in concert, whereas lawyers who refuse to execute projects to which
they object on moral grounds will do so as individuals, without deliberating collectively
with other lawyers. The worry about a hidden Central Committee of lawyers evaporates
when we realize that the committee will never hold a meeting, and that its members don’t
even know they are on it. An analogy will clarify this. No doubt throughout history people
have often been dissuaded from undertaking immoral projects by the anger, threats, and
uncooperativeness of their spouses. It would scarcely make sense, however, to worry that
this amounts to subjecting autonomous action “to rule by an oligarchy of spouses.” There
is no oligarchy of spouses.
Luban seems to accept for the sake of argument that screening by a true Central
Committee might well be intolerable. But if weighing and screening by spouses does
not constitute an elitist centralization of ethical decision-making highly threatening
to the value of individual autonomy, it is not clear how this can be regarded as the
result when we substitute “lawyers” for “spouses.” Indeed, Luban suggests that infor-
mal social pressures are an essential complement to legal rules in regulating harmful
behaviour. (Note that I am not here endorsing Luban’s conclusions, but only using
his arguments to illustrate certain characteristic features of deliberation.)

III. Hypothetical Cases as Thought Experiments

The role of the settled cases appealed to in the above examples of deliberation is
analogous in some respects to the role of data in scientific theorizing. In both
areas, that is, one looks both for relevant data, or clear cases, and for generaliza-
tions that account for what seems to be clearly the case (though of course such
apparent clarity is not infallible in either area), and uses the latter generalizations
to make determinations about further cases. In neither area is the best account of
the data deductively entailed by the data. Nevertheless, in both areas the data in
some sense determines the best theory (or theories, in the case of a tie or moderate
degree of underdetermination of theory by data). The best theory (or theories—
I will hereafter omit the qualification, but it continues to apply) is some function
of the data, in the sense that if the best theory were other than what it is, the data
would have to be different in some way. This is just another way of
saying that situations that are relevantly similar in respect of data must be treated
consistently in theoretical respects, or, more briefly, that like cases should be
treated alike. With respect to legal deliberation, this general consistency require-
ment is that cases relevantly similar, in respect of applicable legal doctrines and
distinguishing circumstances, should be similarly resolved.
82 S. L. Hurley

Of course, the analogy should not be strained; some of the general roles of
ethical and legal deliberation and theorizing are very different from those of
scientific theorizing. Scientific theories are used to predict what will happen on the
basis of causal theories that account for what has happened in well-designed experi-
ments, and sometimes also for what it is thought would have happened under sig-
nificant counterfactual circumstances. The basis for scientific hypotheses are thus
experimental data and sometimes intuitions gathered in “thought experiments,”
such as Einstein’s famous thought experiments about flashlights emitting beams of
light in trains travelling at close to the speed of light. Scientific hypotheses generate
predictions which are then tested against the results of further experiments.
By contrast, ethical and legal deliberation can hardly be described as having the
role of predicting what will be done on the basis of causal theories. Rather, it has a
normative role: to give guidance in extending consistently to the case at issue a
series of settled ethical or legal judgments about what should be done when the
applicable ethical or legal reasons conflict. Its normative hypotheses thus aim to
account for clear resolutions of past cases in which the relevant reasons stood in
conflict, and also for clear resolutions of significant hypothetical cases, designed to
test the relationships between the conflicting reasons. Deliberative hypotheses are
used to generate not mere predictions of decisions and actions, but decisions and
actions themselves; hypotheses cannot be tested against the very decisions and
actions they generate. Rather, they are tested against cases, both actual and hypo-
thetical, in which the right answer about how a conflict of reasons should be
resolved is settled.¹²
Despite the differences between scientific, causal theorizing, and deliberative,
normative theorizing, it is important to recognize the way in which both kinds of
theorizing are responsible to the data to be explained. The requirement that the
sought-after hypothesis explains the data can be seen as the source of the deliber-
ative requirement of consistency, the general requirement that like cases be
treated alike. Some philosophers have regarded it as puzzling how one could hold
that both:
(1) the right answer about which alternative should be done is not entailed by
the nonevaluative facts about the alternatives (nonreductionism: what should be
done does not reduce to nonevaluative facts about the alternatives)
and:
(2) one must treat alike cases that are alike in respect of nonevaluative facts, so
that if there is a difference between two cases in respect of what should be done,

¹² I do not suggest that scientific, ethical or legal deliberation are to be understood instrumentally
rather than realistically; it is common ground between instrumentalism and realism that theoretical
propositions should be able to be used in practical roles, e.g. to predict, explain or guide. I am con-
cerned here to avoid overstating the analogy between the practical roles of scientific and deliberative
theorizing rather than to take a position on the further issues that divide instrumental and realistic
views of theories.
 Coherence, Hypothetical Cases, and Precedent 83

there must be some difference between them with respect to the nonevaluative
facts about the alternatives as well (supervenience: what should be done supervenes
on nonevaluative facts about the alternatives).
While there is no logical incompatibility between (1) and (2), the puzzlement
about how they can both hold can be expressed by asking: If the right answer is
not entailed by the nonevaluative facts, then what is the source of the require-
ment that cases alike in nonevaluative respects be treated alike?¹³ However,
recognition of the way in which right answers to questions about what should
be done reflects hypotheses that are required to explain the data about settled
cases provides a response to this puzzlement. The best theory about the data is
not entailed by—cannot be deduced from—the data. Nevertheless, it is essen-
tial to the notion of a theory responsible to data that the best theory, whatever
it may be, treats like cases alike. The source of this requirement is the essential
explanatory aspirations of theories: a theory the content of which varies inde-
pendently of the data it purports to account for to that extent does no explana-
tory work. This remains true even though the best theory is not entailed by
the data.¹⁴
Thus, the requirement that like cases be treated alike has its source in the
theoretical nature of judgments about what should be done when reasons conflict,
which is highlighted by coherence accounts. Moreover, the data to which theories
about what should be done are responsible are settled cases, both actual and
hypothetical. If in general the role of settled cases in deliberation is somewhat
analogous to that of experimental data in scientific theorizing, then the role of
hypothetical cases in particular may be regarded as somewhat analogous to that of
thought experiments.¹⁵ It is important for our understanding of the requirement
that like cases be treated alike that we include settled hypothetical cases, cases the
answer to which would be clear were they to arise, as well as settled actual cases
among the data to be explained. For reasons that will emerge in what follows, this
is particularly important in the case of legal deliberation, where the general
requirement that like cases be treated alike must be distinguished from the further
specific requirement imposed by the doctrine of precedent. My examples in the
previous section make clear that, since legal deliberation and argument often turn
not merely on settled actual cases but also on settled hypothetical cases, effective
legal reasoning often requires one to discover or construct revealing hypothetical
cases;¹⁶ a coherence account provides a framework for understanding the function

¹³ See, e.g., S. BLACKBURN, Spreading the Word (1984), chapter 6.

¹⁴ For further discussion of supervenience, irreducibility and explanation, see NATURAL REASONS,
supra note 2, chapter 14.
¹⁵ The effective postulation of hypothetical cases thus has something in common with effective
experimental design.
¹⁶ I do not suggest the existence of a positive professional obligation imposed on court or counsel
to discover or construct hypothetical cases, but of a normative requirement that arises from charac-
teristic features of legal reasoning as a species of practical reasoning. It may be that my account of the
84 S. L. Hurley

of hypothetical cases in legal reasoning. Hypothetical cases are not merely posed by
lawyers and judges, but answers to questions about how they should be resolved
are often taken for granted in a way which the argument of the case at issue
depends on. The answers to hypothetical cases may be implicitly assumed rather
than explicitly stated, but nevertheless they are often depended on in the reason-
ing of opinions, in a way that may be explicitly represented within a deliberative
matrix. This is to say that legal hypotheses are responsible to data about settled
hypothetical cases as well as settled actual cases; both are among the cases with
respect to which the requirement that like cases be treated alike must be under-
stood. Nevertheless, as we shall see, the doctrine of precedent imposes a further
constraint on legal deliberation in giving actual settled cases more weight than that
of hypothetical settled cases for purposes of determining what counts as treating
like cases alike.

IV. Kress on Coherence Accounts and Retroactivity: The

Problem of Intervening Cases
In an interesting article entitled “Legal Reasoning and Coherence Theories:
Dworkin’s Rights Thesis, Retroactivity, and the Linear Order of Decision,”
Kenneth Kress argues that the role within coherence theories such as Dworkin’s of
coherence with past decisions and deference to precedent makes for retroactive
applications of the law. In Kress’s view, retroactivity may occur if settled law
changes between the occurrence of the events being litigated and the adjudication
of them, since at adjudication the most coherent account of settled cases will be
responsible to actual cases decided after the occurrence of the events litigated.
What Kress calls the “ripple effect”:
. . . depends upon legal rights being a function of settled law and upon the temporal gap
between events being litigated and their eventual adjudication. Judicial decisions often
change the settled law. Often, if not always, the settled law will be changed between the
occurrence of events being litigated and their eventual adjudication. In consequence, a
litigant’s rights will sometimes also be changed. If changes in the settled law change the
dispositive legal right, a litigant who would have prevailed given the legal rights existing at
the time of the occurrence will lose because he no longer has that right at the time of adju-
dication. The opposite is true of the opposing litigant.
This is retroactive application of the law.¹⁷

latter is particularly influenced by characteristic features of the legal system in the United States,
such as the prominent role of the posing of “hypos” in Supreme Court oral argument, in legal
education, etc.
¹⁷ K. J. Kress, Legal Reasoning and Coherence Theories: Dworkin’s Rights Thesis, Retroactivity, and
the Linear Order of Decisions, 72 CAL L.R. 369, 380 (1984).
 Coherence, Hypothetical Cases, and Precedent 85

I shall refer to the problem Kress identifies as the problem of intervening cases.
Kress regards this as a particularly serious problem for Dworkin’s rights-oriented
version of a coherence account of legal reasoning, since Dworkin works his
account up in the course of criticizing Hart’s account for the scope it gives to
judicial discretion and for the retroactive applications of the law which judicial
discretion involves.
However, Kress regards the problem as generalizable: it applies to any coher-
ence account which is conservative in the sense that it adheres to the dominant
conception of precedent. According to the latter, legal truths depend in part on
prior legal decisions. The general form of the problem is that the best theory about
settled law may change between the time the events occurred and the time they are
litigated, owing to intervening legal decisions. “The mere historical fact of a prior
decision influences the decisions in later cases, and thus the law, because it
enlarges the settled law with which later decisions must cohere.”¹⁸ This is in effect
just to say that the prevalent conception of precedent gives more weight to settled
actual cases than to settled hypothetical cases for purposes of determining what
counts as treating like cases alike.

V. A First Response: Prospective Application

What is the correct response to or diagnosis of the problem of intervening cases?

Let us first briefly consider and dismiss, with Kress, the possibility that present
legal practice should be changed so as to base legal decisions on law settled at the
time the events adjudicated occurred rather than at the time of adjudication.
Kress suggests that to give decisions only prospective effect in this way is particu-
larly appropriate when decisions are legislative in character; as he puts it: “The
doctrine of prospective overruling is the legacy of legal realism, the doctrine that
maintains that judges legislate.” However, Kress considers and rejects this pos-
sible response to his problem, on theoretical grounds and on pragmatic grounds
(such as the difficulties of determining which events are relevant for purposes of
dating the law to be applied). I find his pragmatic arguments against this line of
response fairly persuasive. ¹⁹

¹⁸ Id. at 400.
¹⁹ Id. at 386–387: “It is unlikely that the revised . . . theory could be developed in detail, or utilized
by judges if it could, for several reasons. First, the possibility of temporally extended events and trans-
actions raises difficult problems in determining the date of the law to be applied. These problems are
compounded if several related but separable transactions are being litigated and there are multiple
issues. The prospect of applying the law at different times to different but related aspects of a compli-
cated transaction raises unappealing complexity. Further and perhaps insuperable complexity arises if
it is possible to analyze the overall transaction in multiple ways into different temporal components.
Unless we can be sure that one temporal analysis will be superior to all others, we will need rules to
choose among the various analyses. More important, it is unclear what precedential effect should be
given to a decision that applies law from many time periods, law that by definition differs from the law
that would be applied to an event that occurred at that time of adjudication if litigated at that time.”
86 S. L. Hurley

However, if Kress’s own objections to giving intervening cases prospective effect

only are persuasive, his problem is not just a problem for coherentists; it cannot be
avoided by “frank” assimilation of adjudication to legislation. Moreover, even if
we assume for the sake of argument that the correct response to the problem of
intervening cases is to give them prospective effect only, I am not persuaded that
this response would be inconsistent with a coherence account. To the extent that
retroactivity is the concomitant of a conception of precedent that gives more
weight to actual than hypothetical cases, a coherentist could argue that it is the
fact that actual decisions may change the law, through the operation of precedent,
that makes prospective application appropriate. But this law-changing effect of
precedent should not be confused with “legislative” discretion to make law for the
case at hand; these are two logically distinct issues. That is, the occurrence of a
judicial decision may change what the law requires from that point on, through
the operation of precedent, even though the judge had no antecedent discretion as
to how that case should be decided.²⁰ Even if only mistaken intervening decisions
can change the law, as Dworkin urges (see Section VIII) below, nevertheless a
mistaken decision that changes the law through the operation of precedent is
logically distinct from quasi-legislative discretion to make new law.

See especially Kress’s note 77, commenting on the line of criminal procedure cases involving
prospective-only application. Kress distinguishes the “more common form” of prospective-only
application, which nevertheless applies intervening decisions to pending cases, from the more
extreme form that would be necessary to avoid ripple effect retroactivity, which would deny the “new”
rule to pending cases as well. For the “more common form,” see Linkletter v. Walker, 381 U.S. 618
(1965), which did apply the intervening decision to pending cases.
It may be objected to that Kress has not addressed the line of criminal procedure cases subsequent
to Linkletter which explicitly discuss the application of intervening decisions to pending cases. See
Stovall v. Denno, 388 U.S. 293 (1967); Desist v. United States, 394 U.S. 244 (1969), especially Justice
Harlan’s dissent; Shea v. Lousianna, 470 U.S. 51 (1985), especially Justice White’s dissent. The Court
divides over the problem of intervening cases, and moreover, seems to change its own position.
Stovall and Desist come down in favour of nonretroactivity by refusing to apply the intervening deci-
sions to pending cases in which the relevant events of police conduct occurred prior to the interven-
ing decisions. However, Justice Harlan’s dissent in Desist, favouring precedent over nonretroactivity,
becomes the Court’s position in Shea, where Justice White dissents, arguing on nonretroactivity
grounds. A nice self-referential problem of intervening cases about the problem of intervening cases,
the logic of which I will not even attempt to untangle!
In reply, it should be kept in mind that the degree to which the Court is willing to consider
and countenance intervening case retroactivity in the criminal procedure context, where
police reliance on earlier cases is in question, may well not generalize. Thus, an acceptable degree
of retroactivity with respect to police reliance, which cuts in favour of criminal defendants,
may not be acceptable in general, that is, where the intervening case might cut against
criminal defendants and undermine their reliance rather than police reliance, or where it might
change the positions of civil litigants. It would be interesting to develop an integrated view of
retroactivity doctrine within and without the criminal procedure context, and to try to isolate the
special effects of the criminal procedure context on retroactivity doctrine, but I cannot do
so here.
²⁰ Kress does allow for this type of view. Kress, supra note 17, at 386; he does, at 382, distinguish
judicial discretion from judicial creation of law.
 Coherence, Hypothetical Cases, and Precedent 87

So, if the correct response to the problem of intervening cases is prospective-

only application, a coherence account is not thereby defeated, since there is no
inconsistency between a coherence account and this response. On the other hand,
Kress’s pragmatic reasons against this response to the problem are equally valid
given legislative accounts of judicial reasoning. So if this response is ruled out, it is
not only coherence accounts that are left with a problem. Let us thus suppose, for
the sake of argument, that the prospective-only response to the problem of
intervening cases is to be ruled out (but for a qualification relating to changes in
settled cases with extra-judicial sources, see the end of Section VII below).

VI. A Second Response: It’s Not So Bad

Consider a second possible response envisaged by Kress to the problem of inter-

vening cases, namely, that of accepting the retroactivity in question as, for one
reason or another, not all that bad. And at any rate, a coherentist might add, the
problem of intervening cases is not a problem for coherence accounts in particu-
lar, since they will still involve less retroactivity than accounts of adjudication as
ex ante discretionary do.²¹
Just how objectionable would intervening case retroactivity be? Perhaps coher-
entists should not regard it as a matter for particular concern. Kress offers
Dworkin something like this possible position in a footnote (though in fact this is
not Dworkin’s response to the problem):
Dworkin believes that the unfairness of retroactive application of law is not fully explained
by the rule of law requirements to give prior notice, satisfy justified expectations, and the
like. . . . Dworkin must take this position because the controversial nature of right answers
even when they exist means, as Dworkin admits, that right answers often cannot be
demonstrated and therefore often are not known in advance . . . Dworkin’s objection to
retroactivity appears to focus on the importance of giving a principled justification for
enforcing judicial decisions. Dworkin’s concern with retroactivity derives from the belief
that creating and applying new rights at the time of adjudication cannot be given a prin-
cipled justification . . . Arguably, therefore, ripple effect retroactivity with its consequential
disruption of expectations and failure to provide notice is not the kind of retroactivity that
would concern Dworkin.²²
Though Kress’s article was published before Law’s Empire, this suggestion
seems reasonably consistent with Dworkin’s views in Law’s Empire about respect
for precedent and coherence as expressing the integrity and personification of the

²¹ This tu quoque on behalf of coherence accounts seems to be borne out by the Court’s oscillation
between precedent and nonretroactivity in Stovall, Desist and Shea, supra note 19, where the issues do
not arise as a consequence of any particular theory of adjudication.
²² Kress, supra note 17, at 384.
88 S. L. Hurley

community. On the above suggestion, people would simply be on notice that

their claims were subject to requirements of coherence with possible intervening
precedents. Perhaps one could justify the retroactivity to people on the grounds
of their membership in a community with the virtues of integrity, which unfor-
tunately entail such retroactivity. People could of course try, perhaps through
their lawyers, to anticipate intervening precedents as best they could, by identify-
ing any relevant settled hypothetical cases. But the possibility would remain that
an unsettled hypothetical case might become a decided actual case and thus
create an unanticipated intervening precedent.
We might fill out this possible line of response further. I claimed there is no
inconsistency between admitting that the mere fact that a judge has decided a case
may change the law and claiming that in deciding it he has no discretion. Perhaps
the mere fact that the judge has decided the case changes the law, not through an
exercise of discretion, but as a result of his efforts to discover the right answer. The
retroactivity problem may arise because in law, as is not the case in science, the
mere fact of theorizing by a judge itself counts as a further piece of data for later
theorizing. But perhaps so long as the judge can give a principled justification for
his decision, even if it was not reliably predictable, the change to the law repre-
sented by his decision is not arbitrary. Recognizing such changes, it might be
argued, is itself supported by the considerations of coherence and community
integrity that support respect for precedent. People know that the legal system
gives weight, in this timeless sense, to intervening precedents, even if they do not
know that they are ahead of time. Is this really any more objectionable than not
knowing the right answer itself ahead of time?
Of course, we accept that it may be hard to determine what the best account of
settled cases would say about the case at issue, before it is decided. What the best
theory really is, and how it bears on the case at issue, may be quite controver-
sial.²³ But these are epistemological problems, not problems of true retroactivity;
changes in our knowledge of the law should be distinguished from changes in the
law. It simply is often extremely difficult to know what the law requires; perhaps
only the best lawyers know for sure, and we cannot afford them. But nevertheless
we are held to the law, and we must do the best we can to understand what it is. It
is unfortunate, and gives rise to unfairness, that the law is so complex and
difficult, and that access to it so expensive and restricted; but again these are not
problems of true retroactivity. It might be objected to that if from a litigant’s
practical point of view the effects of the “mere” epistemological problem and
those of “true” retroactivity are much the same, and we can live with the former,
then the latter cannot be so bad. But the stubborn intuition remains that true
retroactivity is somehow more profoundly unfair than are the effects of the

²³ This paper is not intended to address the problem of whether systematically wicked law is really
law; on this see, e.g., R. DWORKIN, LAW’S EMPIRE 102–108 (1986).
 Coherence, Hypothetical Cases, and Precedent 89

epistemological problem. Let us therefore, for the sake of argument, also put the
“not-so-bad” response aside.

VII. The Source of the Problem: Coherentism as

Opposed to Precedent
Consider what features of a system of decision-making are necessary or sufficient to
give rise to the problem of intervening cases. As we have seen, coherence accounts of
practical reasoning give settled cases an important role as the data, so to speak, to
which the theories or principles sought in deliberation are responsible. Moreover,
coherence accounts impose a general requirement with respect to settled law of
treating like cases alike (which I will sometimes refer to as “the weaker require-
ment”). That is, legal truths are a function of, among other things, settled law;²⁴ reso-
lution of the actual case at hand must be consistent with the resolution of
comparable settled cases. While, concerning any particular settled case, it is possible
that the best theory of settled cases may show that it is mistaken, if theoretical coher-
ence overall with settled cases is the standard, then it is not possible for all settled
cases to be mistaken; hence coherentist consistency involves a certain element of
conservatism. However, while these elements are present in coherence accounts of
legal reasoning in particular, they are not sufficient to give rise to the problem of
intervening cases or any distinctive issues about retroactivity. The latter only arise in
virtue of a further requirement (that I will sometimes refer to as “the stronger
requirement,” since it is not entailed by the weaker requirement), imposed by the
doctrine of precedent in the legal context in particular, namely, that actual settled
cases are to be given more weight, in virtue of having actually been decided by a
court, than hypothetical settled cases, or, in Kress’s words, that “prior decisions are to
be accorded weight which may influence the outcomes in later cases merely by virtue
of the fact that the decisions have occurred.”²⁵ Since the stronger requirement
imposed by the doctrine of precedent is not a necessary feature of all coherence theo-
ries, Kress is wrong to the extent that he suggests that the prospect of intervening
case retroactivity arises for all coherence theories.²⁶

²⁴ Kress, supra note 17, at 380. ²⁵ Id. at 400.

²⁶ Id. at 398 (“ . . . since nothing in the argument has relied on the details of any coherence theory,
it should be clear that ripple effects will occur in all coherence theories . . . ”). See also id. at 371 (“The
ripple effect will occur in any coherence theory with a principle of conservation. In adjudication,
precedent provides the conservative element”). However, compare his concessions to the distinction
between coherence theories in general and the legal doctrine of precedent in particular at the very end
of the article. Id. at 401–402. It is possible that, despite the way Kress introduces his problem by ref-
erence to coherence theories in general, not just in law, id. at 369–370, the remarks quoted above
should be interpreted as restricted to legal contexts, which involve the doctrine of precedent in
particular. But my primary purpose is to address Kress’s interesting problem on the merits, and the
substantive issues I consider arise however Kress’s remarks are interpreted.
90 S. L. Hurley

Decision theory provides one illustration of coherentist reasoning minus the

stronger requirement of precedent; within decision theory, coherence with settled
preferences provides a standard for the determination of problematic preferences,
but settled preferences about actual alternatives that have issued in actual
decisions are not given greater weight in principle than settled preferences about
hypothetical cases. Consideration of decision theory, in which no problem of
intervening cases arises, will thus help to bring into focus the way in which the
problem depends on the stronger requirement of precedent.
Suppose a decision theorist is trying to help a decision-maker arrive at decisions
about certain difficult issues where his criteria of choice or goals conflict. The
decision theorist aims to discover the decision-maker’s indifference curves with
respect to those conflicting criteria or goals in order to help him to make the deci-
sion among the actual alternatives that places him on his highest indifference
curve. To do this the decision theorist must depend on data provided by the deci-
sion-maker about what his preferences clearly are or would be in various cases
which are easier for him to decide about than is the case at issue. These cases need
not all involve feasible alternatives; some may be purely hypothetical. The exercise
is one of extrapolating consistently from the set of settled preferences to a determin-
ation of the case at issue. Perhaps some of the settled preferences of the decision-
maker are outliers and cannot be regarded as consistent with the rest within any
theoretically acceptable representation of the decision-maker’s preference space;
such settled preferences may be disregarded as “mistakes.” But it is of the nature of
this exercise that there cannot be too many such mistakes. Settled preferences are
here the data for which the decision theorist’s representation of the decision-
maker’s preference space must account. Theories may tell us that some of the data
they aim to account for is suspect or corrupt, and must be disregarded, but ulti-
mately theories are supported by data. The basis for regarding some settled prefer-
ences as mistaken is their failure to cohere with the rest; thus there cannot be a
basis for regarding too many of them as mistaken (though, of course, just what
counts as “too many” is difficult to specify).
The conservatism of this interpretative exercise is also attributed by a coherence
account to legal reasoning: not too many settled cases can be regarded as mistakes,
or the basis for regarding them as mistaken is itself undermined. However, these
elements of conservatism do not entail that settled actual cases should be given
more weight than settled hypothetical cases, and hence they are not sufficient to
give rise to the problem of intervening cases. The mere fact that a decision-maker
actually makes an intervening decision between the time at which certain events
occur and the time at which he must face a further problematic decision about
those events does not alter the set of settled cases with which his second problem-
atic decision must cohere. (Of course, a change of mind does alter the set of settled
cases; but this is a different matter.) A settled case may be equally settled, whether
actual or hypothetical. More generally, the set of settled cases does not coincide
 Coherence, Hypothetical Cases, and Precedent 91

with the set of actual cases. Not all actual cases are settled (some are problematic),
and not all settled cases are actual (some are hypothetical).²⁷
If the stronger requirement of adherence to precedent were not imposed on
legal reasoning in addition to the weaker elements of conservatism imposed by a
coherence account and shared with decision theory, no problem of intervening
cases would arise in law either. To see this, consider a revisionary conception of the
law as equally responsible in principle to both settled actual and settled hypothet-
ical cases, so that the change of a settled case’s status from hypothetical to actual
between occurrence of the events litigated and their adjudication cannot via
precedent change the law about the later case. Such a revisionary conception of
the law is compatible with a generally coherentist approach to practical reasoning,
but it omits the distinctively legal doctrine of precedent. Now, whether a case is
settled or not does not depend on whether it has actually been decided; a settled
case is equally settled whether its current status is actual or hypothetical, and
retroactivity does not arise in the way Kress envisages.
Kress considers the possibility that the law does not change purely in conse-
quence of legal decisions, but points out that this would be to purchase unchan-
ging right answers at the cost of dispensing with the doctrine of precedent. He
writes: “Judicial decisions change the settled law” and: “To deny that law and
legal rights change at all in response to new decisions is to deny that legal rights
are, at least in part, determined by settled law.”²⁸ While I think he is correct to
regard the denial of any change of law in response to new actual decisions as
giving up the doctrine of precedent, nevertheless, it is logically possible to hold
that legal rights are, at least in part, determined by settled law, without holding
that a change in a case’s status from hypothetical to actual is per se a change from
unsettled to settled. It is just this logical space that is occupied by the revisionary
conception of law we are now considering, which in effect does not incorporate
the doctrine of precedent as we know it. Of course, new legal decisions may
reflect extra-judicial changes in the settled law. But absent extra-judicial changes,
under the revisionary conception new legal decisions would not change settled
law, though they might change our awareness of it.
Kress further points out that, within Dworkin’s theory at least, “While at each
point in time, there are right answers to all or nearly all legal issues, the right
answer that is given may not be the same as that which would be given at another
point in time, in consequence of changes in institutional history.” However,
giving hypothetical and actual settled cases equal weight would not mean that the
law cannot develop and change, but merely that such developments and changes,

²⁷ Cf. Kress on the analogy between legal and scientific theorizing, id. at 392n, and on the charac-
ter of settled law, id. at 278. My use of the term “settled” here does not coincide with his, but I believe
that something like my use is essential to convey what is distinctive about coherentist views of prac-
tical reasoning, and of legal reasoning in so far as it is a species of practical reasoning.
²⁸ Id. at 380, 393.
92 S. L. Hurley

in both actual and hypothetical cases, would reflect extra-judicial developments

and changes in institutional history, social practices, etc.²⁹ Cases that were unset-
tled might become settled as the law develops, in response to extra-judicial devel-
opments. But under the revisionary conception a case would not change from
unsettled to settled merely by becoming the subject of a judicial decision and
changing status from hypothetical to actual.
I emphasize that I have invoked this revisionary conception of law not in order
to endorse it, of course, but merely to illustrate my claim that the problem of
intervening cases does not arise from the coherentist character of legal reasoning
per se, but rather from the further distinctive asymmetrical treatment of actual
and hypothetical cases by the doctrine of precedent in particular. The point of
the distinction between coherentist practical reasoning absent the doctrine of
precedent and legal reasoning with the doctrine of precedent is not to suggest we
should respond to Kress’s problem by doing away with precedent, but rather to
contribute to diagnosing the source of the problem. This distinction is important
because it is the general character of consistency and conservatism with respect to
hypothetical as well as actual cases in coherentist practical reasoning, which legal
reasoning shares and which the doctrine of precedent does not eliminate but
rather supplements, that supports the view I go on to assert in Section VIII below,
namely, that the special role of precedent in legal reasoning only makes a differ-
ence when the decided cases appealed to as precedent were not ex ante correct.
It may be objected that intervening case retroactivity would occur even under
the revisionary conception of law, so long as the law is subject to change such that
coherence with intervening, newly settled cases may be invoked to justify legal
decisions about events that occurred before the change. Perhaps the objector will
admit at this point that it would not matter whether the change in settled cases
occurs within the category of actual or hypothetical cases; it need not be the result
of a change in status from hypothetical to actual. But so long as the possibility of
change in the set of settled cases is admitted, even though the initiation of change
is distributed across actual and hypothetical cases, it may be claimed that the
problem of intervening cases would remain.
Advocates of coherence accounts are not committed to denying that the law may
change. However, they may wish, as Dworkin does, to deny that judges should initi-
ate such changes, as opposed to trying to discover what the best theory of settled law
requires. (There is nothing analogous to this division of labour between the judi-
ciary and other, law-making, branches of government, in the ethical and decision-
theoretic versions of the pursuit of coherence. In these latter cases, the deliberator
plays the legislative as well as the judicial role; his decisions may reflect changes of
mind that originate changes in the set of settled cases. At the same time, there is no
problem of retroactivity, although if someone changes his mind about what he
should have done after doing it, he may well regret having done it.) On the view that

²⁹ See and compare id. at 392–394.

 Coherence, Hypothetical Cases, and Precedent 93

the judicial role is to discover the law, not to make it, changes in judicial perception
of the content of settled cases, whether actual or hypothetical, ought to reflect extra-
judicial legal developments, such as changes in social practices and customs with
constitutive bearing on the law,³⁰ or in legislative or administrative background.
Such extra-judicial developments put members of society on fair notice, and
changes in legal theory which reflect such intervening changes should in principle be
applied prospectively only.³¹ However, legislative and administrative changes
usually are applied prospectively only. Moreover, changes in social practices usually
occur gradually enough that worry about such a change intervening between the
time of the events litigated and their adjudication may often seem contrived and
artificial. Perhaps there are occasionally abrupt, revolutionary changes in social
practices, but they are surely unusual. Perhaps in these unusual cases changes in
settled hypothetical cases which reflect the change in social practice should in theory
be given effect prospectively only (although there may be practical difficulties in
identifying such cases and the time of the change of social practice involved). Such
exceptions would have a rationale in the extra-judicial source of the change, and
could be admitted without threatening the normal judicial practice to the contrary.
Thus, extra-judicial changes in law do not threaten my claim that the problem of
intervening cases does not arise from the coherentist character of legal reasoning per
se, but rather from the distinctive asymmetrical treatment of actual and hypothetical
cases by the doctrine of precedent in particular.
In summary, the purely coherentist revisionary conception of law as lacking the
stronger requirement, imposed by the doctrine of precedent, would support the
following argument. Either cases intervening between the occurrence of events and

³⁰ See R. DWORKIN, TAKING RIGHTS SERIOUSLY 40–42 (1977).

³¹ An example would be given by a slight variation of the sequence of events in California Federal
Savings and Loan Association v. Guerra, 107 SCt 683, 479 U.S. 272 (1987). In Cal Fed, a pregnant
worker’s employer brought suit seeking a declaration that a California statute requiring employers to
provide leave and reinstatement to employees disabled by pregnancy was pre-empted by the federal
statute, Title VII, which prohibited sex discrimination. The court upheld the California Statute,
although in the 1976 case of General Electric Co v. Gilbert, 429 U.S. 125 (1976), it had interpreted
Title VII in a different manner, reasoning that discrimination against pregnant persons was not sex
discrimination because many non-pregnant persons are female. In 1978 Congress expressed its disap-
proval of the Gilbert reasoning by amending Title VII to make clear that it intended sex discrimina-
tion to include discrimination against pregnant persons. The disputed events in the Cal Fed case
occurred in 1982: the employee took a pregnancy leave, and when she informed her employer she was
ready to return to work she was told her job had been filled and there were no available positions.
She filed a complaint under the California statute and her employer brought a suit seeking a declar-
ation that the statute was pre-empted by Title VII’s ban on sex discrimination. The case was decided
in 1987.
Now suppose that between the Gilbert decision in 1976 and the 1978 amendment to Title VII
events like those disputed in Cal Fed occurred, but only came to trial after the Cal Fed decision. Then
Cal Fed would have the status of an intervening case. However, it would have been following an extra-
judicial initiation of an addition to the settled law, namely, the amendment of Title VII in 1978.
Under these circumstances it would seem that the legislative change should be given prospective
effect only.
94 S. L. Hurley

their adjudication reflect intervening extra-judicial changes in settled law or they

do not. If they do, the prospective-only application of the new legal hypotheses
supported by the newly settled law is appropriate for the same reason that prospec-
tive-only application is appropriate for legislative decision-making, namely to
avoid the unfairness of retroactivity. In such circumstances, we have judicial reflec-
tions of extra-judicial legal developments, not judicial law-making.³² Cases that
articulate such developments and incorporate them within a coherent legal theory
may change what we believe the law to be, but they do not in themselves change the
law, any more than an illuminating new scientific theory changes the truth about
what the best theory is merely by changing our beliefs about what it is. On the
other hand, if the intervening cases do not reflect extra-judicial changes, then either
they are settled or they are not. The mere fact that they have been actually decided
does not resolve this question. If an intervening case is not settled, the fact that it
has been decided between the occurrence of events and their adjudication does not
change the settled law. The fact that the case has been explicitly considered and
decided per se does not make it a case that is clear to the relevant decision-makers.
Moreover, it may have been a settled case before it was decided, if its resolution
would have been clear had it been considered. That is, either the intervening case is
not settled, despite being actual, or is settled, and was before it was decided. Either
way, there is no retroactivity. Thus the coherentist character of the revisionary con-
ception of law as merely requiring that like cases be treated alike and that hence not
too many settled cases be regarded as mistaken is not sufficient to give rise to the
problem of intervening cases; in addition, the doctrine of precedent is necessary.³³

VIII. The Source of the Problem: Precedent as

Opposed to Mistake or Underdetermination
There thus seems to be a—perhaps surprising—tension between the doctrine of
precedent, on the one hand, and the requirements of fairness which make retroac-
tivity objectionable, on the other. But perhaps the doctrine of precedent is not
sufficient to give rise to the problem of intervening cases either, though necessary.
It seems clear that the revisionary coherentist conception of law I sketched in the
last section, under which no distinction of weight at all is drawn between settled
actual and settled hypothetical cases, involves a substantial departure from the
doctrine of precedent as we know it. Nevertheless, the doctrine of precedent as we
know it is compatible with the significant influence of settled hypothetical cases; it
does not require that status as actual and status as settled coincide. Because of this,
the argument that applied to the revisionary conception of law to show that
intervening case retroactivity would not arise may be available, at least under
certain circumstances, in a legal system that does incorporate the doctrine of
³² As in the variation on Cal Fed, supra note 31. ³³ Cf. Kress, supra note 17, at 401–402.
 Coherence, Hypothetical Cases, and Precedent 95

precedent. I will briefly consider some cases to illustrate the point that status as
actual and status as settled do not in fact coincide. I will then go on to try to
sharpen the diagnosis of the intervening case retroactivity. I shall argue, largely
following suggestions made by Dworkin, that only actual intervening decisions
that are ex ante mistaken, or at least not uniquely correct (in cases of underdeter-
mination), but which the doctrine of precedent nevertheless presumptively
constitutes as judicially-initiated changes in the law, give rise to retroactivity.
Not all actually decided cases are settled; indeed, some actual cases, even some
that are settled, may be mistaken,³⁴ in virtue of their failure to cohere with other
settled, including actual, cases. Dependence on some actual cases is avoided, even
if they have not been explicitly overruled as mistaken, because they are highly
controversial, and/or considered uncertain, dubious, or poor authority. Moreover,
the doctrine of precedent may not treat actual cases of courts of other jurisdictions
as settled. An example of an intervening actual case which was nevertheless not
settled is that of Brown v. Porcher.³⁵ In Brown, the US Court of Appeals held that
the construal of a South Carolina statute to disqualify any claimant who voluntar-
ily left her most recent employment because of pregnancy violated a federal
statute providing that no person shall be denied unemployment compensation
under state law solely on the basis of pregnancy. Brown was decided in 1981; the
US Supreme Court denied certiorari in 1983. In 1980, however, events had
already occurred which gave rise to the closely related case on the same issue of
Wimberly v. Labor and Industrial Relations Commission, which went up on appeal
through the Missouri court system to the Missouri Supreme Court (in 1985), and
eventually reached the US Supreme Court (in 1987).³⁶ The Missouri Court of
Appeals had followed Brown, despite reservations concerning the soundness of its
reasoning; but the Missouri Supreme Court reversed, declaring that it had never
subscribed to the notion that Missouri state courts were bound to follow the deci-
sions of lower federal courts in construing federal statutes.³⁷ The Missouri
Supreme Court declined to follow Brown. “We do not mean to suggest that a
lower federal court’s construction of a federal statute is wholly irrelevant. The
courts of this state should ‘look respectfully to such opinions for such aid and
guidance as may be found therein’ . . . In some circumstances it may be appropri-
ate for a state court to defer to long established and widely accepted federal court
interpretations of federal statutes.” But it evidently felt that Brown was not a
settled case in this sense. The US Supreme Court eventually resolved Wimberly in
favour of the Missouri view and against Brown. But the point to be made here is
not that Brown was mistaken; perhaps it was correct. It is rather that Brown, the
intervening case, did not change the settled law because it was not a settled case,

³⁴ As Kress admits, id. at 378.

³⁵ 660 F.2d 100 (1981), cert denied, 459 U.S. 1150 (1983). Perhaps various abortion cases provide
other current examples. ³⁶ 107 SCt 821 (1987), 479 U.S. 511 (1987).
³⁷ Justice Donnelly, concurring in the result, denied that US Supreme Court decisions interpret-
ing the US Constitution are binding on the states.
96 S. L. Hurley

from the perspective of the Missouri Supreme Court. Thus, whether or not Brown
was correct, the best theory about settled law, which should determine the
outcome in the similar case which comes to trial after Brown, was not altered by
the intervening case. If the result in Brown were to have been upheld as the correct
result, there would nevertheless have been no retroactive application of the
decision in the intervening case, since, as the Missouri Supreme Court implied, it
was not a settled case. Continuing uncertainty about the right answer is just that,
not retroactivity.
Furthermore, there are many settled hypothetical cases which are more deeply
entrenched and would be more difficult to justify regarding as mistaken than
many actual cases: we are lucky that such cases have remained hypothetical rather
than actual. They have not become actual, in some cases, because the flagrant
violations of rights they would involve luckily have not occurred in our commu-
nities. But they are nonetheless clearly settled for being hypothetical. A good
example is found in Justice Marshall’s eloquent dissent in the recent case of
United States v. Salerno and Cafaro. The majority had upheld a statute permitting
the denial of bail altogether in certain cases against Due Process and Excessive
Bail Clause challenges, on the grounds in part that the statute is a regulatory
rather than a punitive measure. Justice Marshall writes:
The ease with which the conclusion is reached suggests the worthlessness of the achieve-
ment. The major premise is that “[u]nless Congress expressly intended to impose
punitive restrictions, the punitive/regulatory distinction turns on ‘whether an alternative
purpose to which [the restriction] may rationally be connected is assignable for it, and
whether it appears excessive in relation to the alternative purpose assigned [to it].’” The
majority finds that “Congress did not formulate the pretrial detention provisions as pun-
ishment for dangerous individuals”, but instead was pursuing the “legitimate regulatory
goal” of “preventing danger to the community” . . . Concluding that pretrial detention is
not an excessive solution to the problem of preventing danger to the community, the
majority thus finds that no substantive element of the guarantee of due process invali-
dates the statute.
Justice Marshall produces a settled hypothetical case as a counter-example to the
majority’s hypothesis about the punitive/regulatory distinction. He goes on:
This argument does not demonstrate the conclusion it purports to justify. Let us apply
the majority’s reasoning to a similar, hypothetical case. After investigation, Congress
determines (not unrealistically) that a large proportion of violent crime is perpetrated by
persons who are unemployed. It also determines, equally realistically, that much violent
crime is committed at night. From amongst the panoply of “potential solutions”,
Congress chooses a statute which permits, after judicial proceedings, the imposition of a
dusk-to-dawn curfew on anyone who is unemployed. Since this is not a measure enacted
for the purpose of punishing the unemployed, and since the majority finds that prevent-
ing danger to the community is a legitimate regulatory goal, the curfew statute would,
 Coherence, Hypothetical Cases, and Precedent 97
according to the majority’s analysis, be a mere “regulatory” detention statute, entirely
compatible with the substantive components of the Due Process Clause.³⁸
He regards the absurdity of this conclusion as a settled aspect of the law, and
nonetheless settled for the fact that the case envisaged is merely hypothetical. Here
is a settled hypothetical case that is more entrenched, and less likely to be regarded
as mistaken by the best legal theory, than many actual cases.
We have considered two cases in which the dissenting opinions appeal persua-
sively to hypothetical cases to defeat hypotheses on which the majority opinion
seems to turn, Parham and Salerno. Suppose for the sake of argument that these
decisions were mistaken up to the time of the decision, or ex ante mistaken.
Nevertheless the doctrine of precedent tells us we cannot assume them to be
ex post mistaken; we cannot infer an ex post mistake from an ex ante mistake
because the fact of the actual decision may change the law. It is possible that an
ex ante mistake is so serious that the enormity of the mistake outweighs its prece-
dential force, and it remains a mistake even ex post; but the mere fact of the deci-
sion has loaded the balance against this possibility. A mistake must be more
serious, must be more deeply incoherent with the settled law, including hypothet-
ical as well as actual cases, to be an ex post mistake than to be an ex ante mistake;
the difference is a matter of degree, a matter of how much must be uprooted to
accommodate the law to the mistake.
We see then that the doctrine of precedent permits determinately ex ante mis-
taken decisions to change the law;³⁹ such decisions, as intervening cases, may give
rise to retroactivity. Of course, the doctrine of precedent is not addressed to mis-
takes in particular; it does not perversely dignify ex ante mistaken actual decisions
as opposed to ex ante correct actual decisions by presumptively constituting only
the former as law henceforward. But precedential force is not needed to constitute
an ex ante correct decision as correct ex post; in this respect it is redundant (though
of course it may feature non-redundantly in the reasoning from previous prece-
dents that makes the decision ex ante correct). No special method of transition
between ex ante standing and ex post standing is needed in the case of correct deci-
sions; precedent has no law-changing work to do in these cases. Since resolutions
that are ex ante correct all things considered do not conflict with one another to
begin with, increasing the weight of a correct settled case as it changes from hypo-
thetical to actual makes no difference. Of course, courts may not know whether
the actual cases they defer to as having precedential force were ex ante mistaken or
not. Nevertheless, we may issue a challenge: how, on coherentist assumptions,

³⁸ U S v. Salerno, 107 SCt 2095, 2107–2108 (1987), 481 U.S. 739, 758–759 (1987).
³⁹ Cf. Kress’s claim, supra note 17, at 394, that “wholesale and final determination of the truth or
falsity of all possible legal propositions leaves no room for the operation of the dominant notion of
precedent.”
98 S. L. Hurley

could the precedential force of a decision, other than one reflecting extra-judicial
legal developments, change the law itself, as opposed to our beliefs about it, if that
decision were ex ante correct?
This in effect is Dworkin’s response to the problem of intervening cases. He
argues as follows.⁴⁰ He does consider that integrity requires us to give weight to
the fact that a case was actually decided so, even if it was decided wrongly. Either
an intervening case is wrongly or rightly decided. Suppose it is wrongly decided.
Then it will indeed change the law retroactively, and for the worse, since it was
wrongly decided. That is one reason it is so important for judges to reach the right
answer! We can accept that there is objectionable retroactivity in this case, but that
is not a consequence of the coherentist account of adjudication, or of the doctrine
of precedent, but rather of the fact that the wrong answer was reached. Suppose,
on the other hand, that the intervening case was rightly decided; then we are
presented with a challenge. How could a correct decision in an intervening case
make for retroactive change of the law?⁴¹ How could the affirmation of a correct
hypothesis itself change hypotheses in subsequent cases from right to wrong or
vice versa? It might make the correct answers to later cases more evident or
predictable, but this is not what Kress needs for his claim; retroactivity involves a
change in the law, not merely a change in our beliefs about the law.
For the moment put aside the possibility, which I will return to, that the correct
decision (not merely our beliefs about it) was ex ante underdetermined, rather
⁴⁰ In discussion.
⁴¹ Kress gives an example, in which the force of precedent is conceived in a mechanical way to
involve counting the number of steps required for privity of contract, which he claims is a paradigm
case of coherentist reasoning. See id. at 382–383. However, it can be dismissed as not an example of
Dworkinian, or more generally, coherentist reasoning at all since it does not involve consideration of
the reasons or principles that support particular results.
Some implicit support is founded in Supreme Court cases for Dworkin’s view that a correct deci-
sion in an intervening case does not change the law. In the criminal procedure context, see in particu-
lar the dissents by Justice Harland and Justice Fortas in Desist, supra note 19, Justice Harlan writes:
“ . . . If a ‘new’ constitutional doctrine is truly right, we should not reverse lower courts which have
accepted it; nor should we affirm those which have rejected the very arguments we have embraced.
Anything else would belie the truism that it is the task of this Court, like that of any other, to do jus-
tice to each litigant on the merits of his own case. It is only if each of our decisions can be justified in
terms of this fundamental premise that they may properly be considered the legitimate products of a
court of law, rather than the commands of a super-legislature.” See also H. Schwartz, Retroactivity,
Reliability and Due Process, 33 U. CHI. L. REV. 719, 748–749 (1966). In Hanover Shoe v. United Shoe
Machinery Corp, 392 U.S. 481 (1968), an antitrust case, the US Supreme Court discussed the inter-
vening case issue explicitly writing: “The theory of the Court of Appeals seems to have been that
when a party has significantly relied upon a clear and established doctrine, and the retrospective
application of a newly declared doctrine would upset that justifiable reliance to his substantial injury,
considerations of justice and fairness require that the new rule apply prospectively only.” The
Supreme Court, however, did not find before it a situation in which there was a clearly declared judi-
cial doctrine upon which a party had relied and which was overruled in favour of a new rule. The
intervening cases in question did not indicate that the issues they decided were novel, or that they
involved a departure from an earlier line of case or the need for innovative principles. “Whatever
development in antitrust law was brought about was based to a great extent on existing authorities
and was an extension of doctrines which had been growing and developing over the years.” Id. at
496–499.
 Coherence, Hypothetical Cases, and Precedent 99

than mistaken. To claim, following Dworkin, that precedent only makes a

difference given an ex ante mistake does not amount to saying that precedent does
not make any difference to the way people should reason about the law. One can
hold that decided cases do and should make a difference to the way people should
reason, via the doctrine of precedent, though they are ex ante mistaken— even if
one also holds that they make a difference only when they are ex ante mistaken,
since they must be either correct or mistaken, and if they are correct their actually
being decided makes no difference. The argument for the latter qualification
derives from the role of hypothetical cases in coherentist practical reasoning in
general; precedent is an additional special feature of legal reasoning, but it does
not eliminate the general role of hypothetical cases in legal reasoning as a special
case of practical reasoning. Precedent does and should make a difference to correct
reasoning about what the law is, and this is why, in part, the problem of retroactiv-
ity does arise. Reasoning from an ex ante mistaken precedent is not itself mistaken
simply because of the ex ante mistake.
The argument as to under what circumstances an intervening precedent can
change the law may be elaborated as follows. Either the intervening decided case
was ex ante correct or ex ante mistaken; moreover, either the corresponding hypo-
thetical case was settled in the same way as the actual case was decided, or it was
not so settled. Then there are four cases to consider. Suppose first that the decision
was ex ante (but not ex post) mistaken in fact, but nevertheless settled, i.e. it was, or
would have been, wrongly believed of the corresponding hypothetical case that
the (in fact) mistaken decision would be the correct decision for such a case. False
beliefs about the ex ante law have been replaced by true beliefs about the ex post
law, made true because the law has changed, in virtue of an ex ante mistaken prece-
dent that is good ex post law. The retroactivity problem arises with respect to what
the law is (though not with respect to beliefs about the law), but that is due to the
ex ante mistaken decision. Secondly, suppose that the decision was ex ante (but not
ex post) mistaken, though not settled as a hypothetical case. In this case the law and
beliefs about it may change as a result of the decision, but again retroactivity is
associated with an ex ante mistaken decision. Thirdly, suppose the decision was ex
ante correct and settled. In this case there is change neither in the law nor beliefs
about it, so no retroactivity problem can arise.
The fourth case is the interesting case: here the decision was ex ante correct,
but the corresponding hypothetical case was not settled in accord with the cor-
rect decision. The truth about the law has not changed as a result of the correct
decision, though what is believed about it may have changed (again, the change
from hypothetical to actual does not necessarily entail a change from unsettled to
settled, since a decided case may remain unsettled). If beliefs about the law do
not change as a result of the correct decision, neither the law nor beliefs about it
have changed, so it is hard to see how the correct decision in a subsequent case
could change. If beliefs about the law do change, it may be in one of (at least) two
ways. On the one hand, they may have been undetermined before, and may now
100 S. L. Hurley

be settled in accord with the correct decision. Could this change in belief, absent
a change in the law, be sufficient to change what the law is in a subsequent case?
Whatever considerations they were in virtue of which the intervening decision
was correct still obtain, and, to the extent they ever applied to the subsequent
case, apply equally after the intervening case is decided. How could the mere
addition of true beliefs about these unchanged considerations change the correct
decision in the subsequent case, as opposed to making it more evident? On the
other hand, beliefs about the law may have been mistaken before the intervening
case, and they may have changed to accord with the correct decision in the case.
How could this change from false to true beliefs about the intervening case
change the law, as opposed to beliefs about it, in the subsequent case? It does not
help to point out that according to a coherence account not too many settled
cases can be mistaken, because we are now supposing the prior beliefs were false;
if they were true, we are back to case three above, and there is no retroactivity.
Whatever considerations they were in virtue of which the prior beliefs about the
hypothetical version of the intervening case were false would apply also to the
subsequent case, to the extent relevant. If we continue to assume that the law in
the subsequent case was either determinately one thing or another at each point,
and no change of law with respect to the intervening case was brought about by
its correct decision, the challenge on coherentist assumptions still stands: how
could a decision in the subsequent case be changed from correct to mistaken or
vice versa merely as a result of a change in beliefs about the law in the intervening
case from false to true, given no change in the law itself with respect to that case?
Note that it is incorrect to object here that the determinacy assumption leaves no
room for the operation of precedent, since we have seen that it does leave room
for it in cases of mistaken decisions. What is at issue is whether precedent can
make a difference given correct intervening decisions.
At this point we should consider the consequences of suspending the assump-
tion of determinacy. If we allow that prior to a decision the right answer (not
merely beliefs about it) may be underdetermined, then precedent may make a
difference not just as a result of ex ante mistake, but also as a result of resolving
ex ante underdetermination. Dworkin, of course, would resist the possibility of
underdetermination in arguing for his right answer thesis, so it does not, in the
absence of an independent argument against the right answer determinacy thesis,
help Kress make a case against Dworkin in particular. But some coherentists
might accept the possibility of underdetermination, so it is relevant to the
general discussion with respect to coherentism.
Thus, from a theoretical point of view at least, intervening case retroactivity
can be laid at the feet, not of coherentist legal reasoning, or even of the doctrine
of precedent, but rather of mistaken decisions, and perhaps also of underdeter-
mination of the law, to the extent a given version of coherentism admits this
possibility. That legal mistakes are unfair to people is hardly news; the theoretical
novelty and interest of the problem of intervening cases thus looks to diminish to
 Coherence, Hypothetical Cases, and Precedent 101

the extent mistakes are its source. Moreover, the use of the problem as a means of
criticizing Dworkin’s version of coherentism in particular is limited by his
rejection of the possibility of underdetermination. The general shape of my
response to Kress with respect to the problem of intervening cases has been to
draw out the consequences of a coherentist view of practical reasoning in general
for legal reasoning and the doctrine of precedent in particular, which conse-
quences limit intervening case retroactivity. The argument has proceeded by
presenting a challenge to show how, given the role of hypothetical cases within
coherentist views of practical reasoning, a correct, as opposed to ex ante mistaken or
indeterminate, intervening case could give rise to retroactivity. Of course, we may
reject coherentist views and their consequences with respect to the doctrine of
precedent altogether, but then we depart from Kress’s subject matter. Note that
my aim has not been to dispute the interest of the retroactivity problem which
Kress has highlighted, but rather his conception of coherentism in general and
Dworkin’s version thereof in particular as the proper targets of the argument.
In pursuing my course of argument, however, I have at several points had to put
considerable weight on the distinction between changes in the law and changes in
our beliefs about the law in particular cases. Coherentism admits this theoretical
distinction on a case-by-case basis, even though it ties the right answers and beliefs
about them together globally in the way indicated in Section VII above, in that not
too many settled cases can be mistaken. However, perhaps from a more pragmatic
point of view, one more concerned with matters of notice and predictability,
which may be closer to that of litigants and potential litigants, this distinction has
been strained even so. Perhaps it is from this more pragmatic point of view that a
distinctive problem about intervening cases arises.
 5
Integrity and Stare Decisis
Scott Hershovitz*

Many think that stare decisis binds even the highest court in a jurisdiction to
follow precedents that were decided incorrectly. Indeed, the view is commonly
held by legal scholars¹ and judges² alike. But if that is what stare decisis really
requires, it is puzzling. What could justify a principle that requires courts to make
the same mistakes over and over again? Surely a better principle (one that most of
us endorse) is that people should own up to their mistakes and seek not to repeat
them. Could legal reasoning really be so different from everyday reasoning that
principle requires courts to make mistakes repeatedly rather than correct them?
We need to clarify the question before we can answer it. There are two varieties
of stare decisis—horizontal and vertical. Vertical stare decisis requires that lower
courts follow the decisions of higher courts. Horizontal stare decisis requires that a
court follow its own precedents. Vertical stare decisis is less mysterious than hori-
zontal. The deference lower courts show to higher courts facilitates coordination
among judges, and it has the potential to improve judicial decision making to the
extent that higher court judges have greater expertise than lower court judges.
Neither coordination nor expertise, however, can explain the practice of a court
considering itself bound by its own precedents. This essay explores the mystery of
horizontal stare decisis. For ease of exposition, I shall use “stare decisis” to refer to
horizontal stare decisis.
The view that stare decisis condemns courts to repeat their mistakes neither fits
nor justifies our legal practice. Fit is problematic because, with some regularity,

* Thanks to Jules Coleman, William Fetcher, Lewis Kornhauser, Scott Sharpiro, Seana Shiffrin,
and Nicos Stavropouls for helpful comments and conversations.
¹ See, e.g., L. Kornhauser, An Economic Perspective on Stare Decisis, 65 Chi.-Kent L.Rev. 63, 65
(1989) (describing stare decisis as “a practice that, paradoxically, demands that a court adhere to a
prior decision it considers wrong”); L. Alexander, Constrained by Precedent, 63 S. Cal. L. Rev. 1, 4
(1989) (“I shall focus on those situations . . . in which a subsequent court believes that, though a pre-
vious case was decided incorrectly, it must, nevertheless . . . decide the case confronting it in a man-
ner that it otherwise believes is incorrect”).
² “[T]here is nothing to do except stand by the errors of our brethren of the week before, whether
we relish them or not.” B. Cardozo, The Nature of the Judicial Process 150 (1921).

Integrity and Stare Decisis. Scott Hershovitz.

© Oxford University Press 2006. Published 2006 by Oxford University Press.
104 Scott Hershovitz

courts overrule precedents that they regard as mistaken. Justification is problematic

because, as we shall see, the values most often cited in defense of stare decisis—
efficiency and fairness—cannot underwrite a general practice of following incor-
rectly decided precedents.
The key to understanding the practice of stare decisis, I shall argue, lies else-
where. Specifically, it lies in the virtue Ronald Dworkin calls integrity. Integrity is
a value that is realized by patterns of behavior across time. The unique demand
that integrity makes upon both individuals and courts is that they recognize that
what they have done in the past affects what they ought to do now. Stare decisis, I
aim to show, promotes integrity in judicial decision making.
I also aim to show that stare decisis is a broader practice than we traditionally
conceive it. On an integrity-based view, stare decisis is the practice of engaging with
history, not just by following precedents, but also by distinguishing them and,
when appropriate, overruling them. Overruling a precedent, and sometimes even
distinguishing one, are often thought of as acts that run counter to the demands of
stare decisis. But if we think of stare decisis as a practice in which courts strive to
exhibit integrity in decision making, then we can see that distinguishing and over-
ruling precedents are ways that a court engages with its own history. As we shall see,
a court with no concern for the integrity of its own decision making would not
need to distinguish or overrule its precedents. It could simply ignore them.

Many investigations of stare decisis ask some version of the question, “What justi-
fies adherence to a decision known to be wrong?”³ The canonical expression of the
principle—stare decisis et not quieta movere—does not, on its face, require follow-
ing decisions known to be wrong. The Latin means “to stand by things decided,
and not to disturb settled points.”⁴ Of course, it is reasonable to assume that some
things that have been decided were decided incorrectly. So stare decisis will, on
occasion, require courts to conform to incorrectly decided precedents.
This observation, however, is insufficient to motivate the idea that to justify stare
decisis we need to explain why courts should adhere to precedents they regard as
wrongly decided. After all, by the same implication, the canonical formulation also
demands that courts follow precedents they regard as correct. Indeed, it demands
that courts follow settled points of law without reference to the correctness of the
decisions that settled them. So it is somewhat of a puzzle why those investigating
stare decisis tend to ask what justifies following incorrectly decided precedents
rather than, say, what justifies adhering to a decision irrespective of its merit.
³ See e.g., J. Macey, The Internal and External Costs and Benefits of Stare Decisis, 65 CHI.-KENT
L. REV. 93, 94 (1989) (“Professor Kornhauser approaches the issue of stare decisis by pondering the
question of what justifies adherence to a prior legal decision known to be wrong”) (discussing
Kornhauser, supra note 1); Alexander, supra note 1, at 4.
⁴ BLACK’S LAW DICTIONARY (8th ed., 2004).
 Integrity and Stare Decisis 105

The answer, I think, lies in a particular view many have of stare decisis. Richard
Fallon expresses the view as follows:
Because a court that believes a prior decision to have been correct can always reaffirm the cor-
rectness of its ruling without reliance on its precedential status, the force of the doctrine of
stare decisis lies in its capacity to perpetuate what was once judicial error or to forestall
inquiry into the possibility of legal error.⁵
Fallon actually expresses two views of stare decisis in this passage. The first is
that the force of stare decisis “lies in its capacity to perpetuate what was once judi-
cial error.” The second is that the force of stare decisis lies in its capacity to “fore-
stall inquiry into the possibility of legal error.” The first view is the one that is of
immediate interest.
The view that the force of stare decisis “lies in its capacity to perpetuate what was
once judicial error” is a common one. In response to a litigant who argued that a
precedent should be overturned because it was incorrect, Judge Posner recently
opined that stare decisis would be “out the window” if the incorrectness of a deci-
sion was a sufficient ground for overruling it.⁶ According to Posner, “no doctrine of
deference to precedent is needed to induce a court to follow the precedents that it
agrees with; a court has no incentive to overrule them even if it is completely free to
do so.”⁷ On this view, stare decisis has no force—does not make a difference—
insofar as it requires courts to follow correctly decided precedents. It is only when
stare decisis requires courts to follow incorrectly decided precedents that it makes a
difference to judicial reasoning. If this is right, to justify stare decisis, we would
need to justify a practice of following incorrectly decided precedents.
The structure of this argument will be familiar to students of jurisprudence.
Some argue that legitimate authorities are incapable of making a difference in
what their subjects ought to do. Either an authority directs a subject to do some-
thing she ought to do anyway, in which case the directive makes no difference in
the subject’s normative situation, or the authority directs a subject to do some-
thing she ought not to do. On the assumption that authorities that direct people
to do things they ought not to do are illegitimate, it appears that legitimate
authorities can make no difference in what their subjects ought to do. This claim
is known as the no difference thesis.⁸ Proponents of the no difference thesis are apt
to hold the view that the normative force of an authority lies in its capacity to bind
people to do things they ought not to do.
⁵ R. H. Fallon, Jr., Legitimacy and the Constitution, 118 HARV. L. REV. 1787, 1823 n.3 (2005)
(citing Michael Stokes Paulsen, Abrogating Stare Decisis by Statute: May Congress Remove the
Precedential Effect of Roe and Casey?, 109 YALE L.J. 1535, 1538 n.8 (2000) (“The essence of the
doctrine [of stare decisis] . . . is adherence to earlier decisions, in subsequent cases . . . even though
the court in the subsequent case would be prepared to say, based on other interpretive criteria, that
the precedent decision’s interpretation of law is wrong”)).
⁶ Tate v. Showboat Casino Marina Partnership, 431 F.3d 580, 582 (7th Cir 2005).
⁷ Id. at 582–583.
⁸ See J. RAZ, THE MORALITY OF FREEDOM 48 (1986) (“The no difference thesis asserts that the
exercise of authority should make no difference to what its subjects ought to do, for it ought to direct them
to do what they ought to do in any event”).
106 Scott Hershovitz

The most important problem with the no difference thesis is that sometimes
what a subject ought to do is underdetermined by the reasons she has. For example,
drivers have a reason to drive on the same side of the road as everyone else, but no
particular reason to drive on the left or right. An authority can make a difference by
giving drivers a reason to drive on the right, say, without requiring drivers to do
anything they ought not to do.⁹
The argument that stare decisis only makes a difference when it requires courts
to follow incorrectly decided precedents suffers from a similar defect. Stare decisis
can make a difference by requiring courts to follow precedents that were neither
correct nor incorrect when decided. Some dispute that there are such cases, but
they are a logical possibility if not a legal one. If a particular precedent represents
merely one among a number of permissible decisions that an earlier court could
have selected, writing, as it were, on a blank slate, then a principle requiring a later
court to follow that precedent makes a difference without demanding that the
court follow a precedent that was decided incorrectly.
Another problem with the view that the force of stare decisis lies in its capacity
to perpetuate judicial error is that it construes too narrowly the impact stare deci-
sis can have on the outcome of cases. Fallon points out that a court following a
precedent it regards as correctly decided can reach the same result without follow-
ing the precedent. At most, this supports the view that stare decisis makes no dif-
ference in what a court ought to decide (though we have just seen that even this is
not true). But even when a case is governed by a correctly decided precedent, stare
decisis may well make a difference to what the court considering it actually
decides. Stare decisis, after all, makes the later court more likely to read the earlier,
correctly decided precedent, and doing so may save that court from making errors
of its own.
The upshot is that the force of stare decisis does not lie in “its capacity to per-
petuate what was once judicial error.” Stare decisis can make a difference, even
when it requires adherence to correctly decided precedent. Thus, there is no rea-
son to think that to justify the practice of stare decisis, we need to justify the prac-
tice of following incorrectly decided precedents. We may need to justify following
incorrectly decided precedents as part of justifying the broader practice of simply
following precedent. However, if we start out by asking what justifies following
incorrectly decided precedents, we misconstrue the import of stare decisis and cut
ourselves off from the resources necessary to understand it.

II

A better question to ask is “What justifies adherence to a precedent irrespective of

its merit?” Adherence to a precedent irrespective of merits would, to use Fallon’s
words, “forestall inquiry into the possibility of legal error.” This second question is
⁹ Id. at 30.
 Integrity and Stare Decisis 107

better than our first, but it is also misleading in that we do not actually have a
practice of adhering to precedent without regard to merit. Courts in the United
States and (more recently) in the United Kingdom do, from time to time, over-
turn even long-standing precedent. Below, I shall argue that overturning prece-
dents is part and parcel of the practice of stare decisis, notwithstanding the fact
that when a precedent is overturned we often speak as if the demands of stare deci-
sis have been ignored or trumped. However, since much of the best work on stare
decisis sets out to explain why courts ought to follow precedent irrespective of
merit, we shall continue our inquiry into what might justify such a practice.
To justify adhering to judicial decisions irrespective of their merit, one needs to
pull off a challenging trick. One has to show both that judicial decisions are
deserving of deference as a class and that individual decisions deserve deference
even when a subsequent court believes that the earlier, precedential court made a
mistake. This is a difficult task, but it is not a task that we are unfamiliar with from
other contexts. To justify obeying an authority, one must pull off a similar trick.
Authorities provide merit-independent reasons for action.¹⁰ A mother can order
her daughter to play inside or outside. Her order provides a reason for her child to
act quite apart from its merit, quite apart from whether it would be better for the
child to play inside or out. This general truth is not without limits, of course.
A mother who orders her child to ingest arsenic greatly exceeds the scope of her
legitimate authority. But within limits, the merit of the order is irrelevant to the
question of whether it gives the child a reason to act. Every parent who has uttered
the words, “Because I said so” implicitly endorses this view of parental authority.
To explain when and why authoritative orders are deserving of deference,
Joseph Raz offers the normal justification thesis.¹¹ Roughly speaking, the normal
justification thesis holds that an authority is legitimate for a person if she will do a
better job of conforming to the reasons that apply to her by following the author-
ity’s orders than she would by following her own lights. Authorities can satisfy the
normal justification thesis through, among other things, special expertise, or the
ability to solve coordination problems. To illustrate: a doctor is an authority on
medical matters in virtue of her special knowledge; the state is an authority on
traffic matters in virtue of its ability to solve coordination problems.
When the normal justification thesis is satisfied, one has a reason to give an
authority’s orders deference as a class, as well as a reason not to deny deference
when one believes there is reason to doubt the merit of a particular order. If one
refused to follow the order of a legitimate authority whenever one believed that it
was in error, the advantage one would get from following the orders as a class
would be lost. Thus the normal justification thesis pulls off the trick when it
comes to authoritative orders—it justifies deference to authoritative orders as a
class, and deference to particular orders even when one has reason to believe they
are mistaken.
¹⁰ See J. Gardner, Legal Positivism: 51⁄2 Myths, 46 AM. J. JURIS. 199, 208–209 (2001).
¹¹ Raz, supra note 8, at 53. For a short introduction to Raz’s account of authority, see S.
Hershovitz, Legitimacy, Democracy, and Razian Authority, 9 LEGAL THEORY 201 (2003).
108 Scott Hershovitz

If we were interested in justifying vertical stare decisis, we could invoke the nor-
mal justification thesis. Higher courts are authorities for lower courts. The author-
ity is based on an ability to coordinate action, and perhaps (one might hope) on
expertise as well. Each lower court judge has reason to decide cases as most other
judges would decide them, and a practice of treating higher court cases as authori-
tative facilitates this aim. To the extent that higher court judges have superior
expertise to lower court judges (or even simply more time and resources to bring
to bear on a case), lower court deference to the ruling of higher courts will improve
lower court decision making.
Raz’s trick will not work for horizontal stare decisis, however. There is no reason
for the highest court in the land to believe that it will, on the whole, decide cases bet-
ter by conforming to its own previous decisions than it would by following its own
current lights. The coordination justification that works so well for lower courts has
no bearing.¹² Expertise cannot do the work either because later versions of a court
generally have more information available to them than earlier versions, and (again,
one hopes) members of roughly equal skill and knowledge.¹³ Thus, if the trick can
be pulled off for horizontal stare decisis, it has to be done some other way.

III

One common approach to justifying stare decisis is to argue that the practice is
efficient. Justice Cardozo took this tack, suggesting that “the labor of judges
would be increased almost to the breaking point if every past decision could be
reopened in every case.”¹⁴ Cardozo’s observation, if true, would be a premise in a
narrow argument for the efficiency of stare decisis, focusing on the conservation of
judicial resources.
One suspects that Cardozo is correct. If you always start with a blank canvas,
you spend a lot of time painting the background. But it is not entirely obvious that
Cardozo is right. Stare decisis may indeed conserve resources, but it also consumes
¹² When a court is composed of many members, and not all members of the court hear every case,
coordination concerns may well support horizontal stare decisis. Thus, circuit courts of appeals in the
United States, which typically hear cases in three-judge panels selected at random from a larger set of
judges, may use horizontal stare decisis as a means of coordinating action among all the judges on the
court. But such an explanation sheds no light on why the United States Supreme Court, for example,
would consider itself bound to follow its own precedent, since every Justice typically hears every case.
For a discussion of the impact different institutional structures have on the justification of stare deci-
sis, see Kornhauser, supra note 1.
¹³ F. H. Easterbrook, Stability and Reliability in Judicial Decisions, 73 CORNELL L. REV. 422, 423
(1988) (“In principle, modern judges have all the information available to their forbears, plus any dis-
coveries in the interim, and the benefit of hindsight. Judges often decide cases on the basis of predic-
tions about the effects of the legal rule. We can examine these effects—both for other strands of
doctrine and for the world at large—and improve on the treatment of the earlier case”).
¹⁴ CARDOZO, supra note 2, at 149. See also, J. P. Stevens, The Life Span of a Judge-Made Rule, 58
N.Y.U. L. REV. 1, 2 (“Perhaps the doctrine is of special interest to judges because it provides special
benefits for judges. It obviously makes their work easier”).
 Integrity and Stare Decisis 109

them. In a system with precedent, judges (and lawyers) spend time and money
researching prior cases. The revenues of Lexis and Westlaw are a cost of stare deci-
sis, as are the salaries of judicial clerks and many law firm associates.
Precedent allows judges to rely on previous cases, but it also demands that
judges distinguish cases, an occasionally laborious task. An armchair investigation
suggests that opinions do not grow shorter as the volume of precedent grows
larger. If anything, the trend is towards longer opinions with more citation and
quotation, and this is in part a consequence of stare decisis. Of course, many con-
troversies are settled without judicial involvement because stare decisis makes the
outcome of a case virtually certain. Additionally, stare decisis allows judges to dis-
pose of many cases (and many issues within cases) summarily. So even if published
opinions grow longer as a consequence of stare decisis, it may still be that judicial
resources are conserved by the practice.
Jonathan Macey has gone so far as to argue that stare decisis allows judges to
maximize their own leisure time by allowing them to free-ride on the efforts of
earlier jurists.¹⁵ Whatever its effect on judicial leisure time, it seems likely that
stare decisis conserves judicial resources, or at least allows those resources to be
deployed more efficiently. But the question is more complex than it appears at first
glance, and it would require serious empirical study to resolve. Importantly, even
if it were established that stare decisis conserves judicial resources, that fact would
not go very far towards justifying the practice. Judicial resources are just one kind
of resource among many, and if efficiency is to justify stare decisis, we must know
that the gains in conserving judicial resources are not offset by losses elsewhere.
That is, stare decisis must be efficient for society, not simply efficient for judges.
In an intriguing study of the efficiency of stare decisis, Lewis Kornhauser care-
fully delineates the potential sources of judicial error that could lead us to regard
previous decisions as mistaken (changes in values, changes in the world, improve-
ments in information, and incompetence).¹⁶ Kornhauser provides a nuanced
analysis of the value of stare decisis in the face of different sorts of error and in dif-
ferent institutional contexts. Rather than recap Kornahauser’s analysis, I want
instead to make some general remarks about the conclusions that Kornhauser
reaches and the prospects for an economic justification for the practice of follow-
ing precedents irrespective of their merit.
Kornhauser says that “the desirability of stare decisis will depend on the particu-
lar ‘facts’ of the situation the law seeks to govern.”¹⁷ He illustrates this by consider-
ing the desirability of stare decisis applied to the liability rules governing accidents
between drivers and pedestrians. Suppose that the value of walking is fixed. If the
value of driving is low, then strict liability will be preferred to negligence.
Conversely, if the value of driving is high, then negligence will be preferred.
Suppose further that in a regime with stare decisis, courts stick to the rules they
announce regardless of the actual value of driving, whereas without stare decisis,

¹⁵ Macey, supra note 3, at 111. ¹⁶ Kornhauser, supra note 1. ¹⁷ Id. at 86.

110 Scott Hershovitz

in each case, courts announce the legal rule that is optimal given the value of the
injurer’s activity.
To know whether stare decisis is desirable in these circumstances, one must
know certain facts about the relative valuations of the walking and driving.
Kornhauser explains as follows:
Suppose the valuations of the activities are such that both injurer and victim should adopt
moderate (or low) levels of the activity. Under stare decisis, the actor who escapes liability
will always adopt a higher level of activity. Under a practice of no stare decisis, however,
each actor will be uncertain whether she will bear the cost of an accident. This uncertainty
will induce her to adopt an activity level intermediate to the one she would adopt if she
escaped liability for certain. For certain relative values of activities, then, the uncertainty
over the legal rule induces the actors to adopt activity levels closer to the social optimum.¹⁸
Kornhauser’s illustration shows that to know whether stare decisis is an efficient
practice for courts to adopt with respect to the rules governing liability for acci-
dents between pedestrians and drivers, one must know something about walking
and driving.
This is hardly surprising. One would expect that the efficiency of a rule that
entrenches other rules against change would depend in part on the efficiency of
the rules entrenched and the nature of the behavior the rules govern. This has
important ramifications for the possibility of an economic justification of stare
decisis. The principal of stare decisis applies generally. With limited exceptions,
the application of stare decisis does not depend on the substantive area of law
involved. Stare decisis applies in contract as well as tort, in family law as well as
corporate law. It applies to the rules that govern liability between pedestrians and
drivers, as well as the rules that govern liability between farmers and ranchers.
Stare decisis is often said to carry less weight in the realm of constitutional law
because it takes a legislative supermajority to undo a Supreme Court decision.
Notwithstanding that, we do not observe the nuanced application of stare decisis
to specific classes of rules that we would expect to find were the efficiency of the
practice the driving force behind it. As Macey puts it:
[I]n Kornhauser’s model, the adjustment costs facing the relevant parties determine
whether social welfare is being maximized by a legal regime of stare decisis. But courts have
exhibited little, if any, ability or inclination to delve into the adjustment costs facing the
parties before them . . . .¹⁹
Efficiency seems a poor explanation of stare decisis as we know it in part because
the practice is insensitive to the facts that determine whether or not it is efficient
in any given context.
We should be skeptical of attempts to argue that a practice of following prece-
dent regardless of merit is across-the-board efficient because whether or not such a
practice is efficient depends crucially on the rules that the practice entrenches and

¹⁸ Id. ¹⁹ Macey, supra note 3, at 95.

 Integrity and Stare Decisis 111

on the particulars of the behavior those rules govern. However, one aspect of
efficiency that is sometimes appealed in order to justify stare decisis deserves
special attention: the value of certainty in the law. If judges were unconstrained by
precedent, some say, the law might shift unpredictably and people would be
unable to plan their affairs. This would have undesirable consequences (stunted
markets, stunted psyches, etc.), and to the extent that stare decisis ameliorates
these consequences, it may be instrumentally valuable.
No doubt, for particular areas of law, there is much to be said in favor of cer-
tainty. When it comes to the interpretation of insurance contracts, or the rules
governing secured transactions, it may well be more important to have a consis-
tently applied rule than to have a good rule. When certainty is sufficiently valu-
able to outweigh any loss from the entrenchment of a suboptimal rule, the need
for certainty will support a practice of adhering to precedent irrespective of merit.
We should be careful not to push certainty explanations too far. As Kornhauser
explains: “ ‘Certainty’ justifications for stare decisis often include ‘reliance’ or
‘planning’ arguments, but these arguments are only as strong as the value of the
planned conduct.”²⁰ To take an extreme but illustrative example, note that Plessy
v. Ferguson²¹ allowed Southerners to plan their affairs in the certainty that the fed-
eral government would not interfere with state-created racial caste systems. The
value of the activities planned in reliance on Plessy was hardly sufficient to warrant
continued adherence to the precedent. Indeed, it seems inappropriate to weigh
the “benefit” of segregation-based plans against the cost of segregation-caused
harms at all. Plessy is, as noted, an extreme example, and it represents a paradigm
of a case that ought to have been overruled. But the point is this: if one appeals to
certainty to justify following precedents irrespective of merit, then one must be
prepared to defend the value of the conduct planned in reliance on the rules
entrenched. It is easy to defend the value of conduct planned in reliance on stable
interpretations of insurance contracts, and impossible to defend the value of con-
duct planned in reliance on Plessy. In the middle lies a vast grey area in which the
value of conduct planned in reliance on previously announced legal rules may or
may not be sufficient to warrant a practice of deferring to those rules irrespective
of their merit. Certainty arguments, like efficiency arguments more generally,
may well support stare decisis for particular types of legal rules. But neither cer-
tainty nor efficiency can underwrite an across-the-board practice of deferring to
precedent without regard to merit.

IV

Fairness is another value commonly thought to underwrite stare decisis. But

arguments in favor of the fairness of stare decisis face an objection similar to

²⁰ Kornhauser, supra note 1, at 78. ²¹ 163 U.S. 537 (1896).

112 Scott Hershovitz

the one that plagues arguments in favor of its efficiency. Fairness may well
demand that courts treat the litigants in like cases alike, and stare decisis
encourages that. But fairness also makes demands regarding the outcome of
particular cases, and stare decisis can entrench unfair results. A practice that
accords deference to the rules announced in Plessy or Korematsu v. United
States,²² for example, entrenches particularly pernicious forms of unfairness
against legal change. Whether stare decisis is on the whole fair depends on the
relative magnitude of the gains in fairness from treating litigants in different
cases alike and the losses from entrenching particular unfair results. It is far
from obvious that, on the whole, stare decisis promotes fairness, and the ques-
tion is, in part, an empirical one.
Fairness does not hold out any more hope of justifying across-the-board stare
decisis than efficiency does, even if we accept the idea that litigants in different
cases ought to be treated alike. And we should have doubts about that claim, at
least as it is relevant to stare decisis. Fairness may demand that we apply the same
rules to litigants in neighboring courtrooms, but it does not clearly demand that
we apply the same rules to litigants separated by decades. Comparative claims of
fairness—claims of the form A deserves X because B received X—are only persua-
sive to the extent that the A and B are similarly situated. Litigants can frequently
claim the benefit of stare decisis even though they are far removed in both time
and place from the source of the precedent.
There are, of course, non-comparative claims of fairness that might support
stare decisis. Non-comparative claims of fairness are claims about how one
deserves to be treated irrespective of how others are treated. Many of the norms
we associate with due process are norms of non-comparative fairness. For
example, it is unfair to detain a person indefinitely without the opportunity for
a hearing before an impartial fact finder, and it would not make it fair to do this
as a matter of routine practice. It is also unfair to punish a person for violating
a criminal statute passed after the alleged violation, and it would not improve
the situation if people were commonly punished for ex post violations. More
relevant to stare decisis, perhaps as a matter of non-comparative fairness,
people ought to be given notice of the rules that will be applied to them, an aim
stare decisis could help facilitate. Or perhaps as a matter of non-comparative
fairness, courts shouldn’t disturb settled expectations. These sorts of claims
may well justify stare decisis in delimited areas of the law. But this is subject to
the point made above. To determine whether stare decisis is fair, gains in fair-
ness from providing notice or protecting settled expectations must be weighed
against the losses from entrenching unfair rules. Because of this, fairness, like
efficiency, is incapable of providing justification for an across-the-board
practice of following precedent irrespective of merit.
²² 323 U.S. 214 (1944).
 Integrity and Stare Decisis 113

We started our inquiry with the question, “What justifies adherence to a decision
known to be wrong?” We saw that this question is misleading because it does not
accurately capture what stare decisis requires. So we asked a different question:
“What justifies adherence to a precedent irrespective of its merit?” This question
better captures the canonical formulation of stare decisis, but it still leaves us with
an intractable problem. We can explain why lower courts ought to defer to the
decisions of higher courts simply by invoking Raz’s normal justification thesis.
But we cannot find an analogue that explains why the highest court in a land
should accord its own previous decisions similar deference. Neither efficiency nor
fairness does the trick.
Fortunately, to justify stare decisis, we need not justify adhering to decisions
irrespective of their merits. After all, courts do not do that. With some regularity,
courts overrule precedents and limit their scope by distinguishing them. We can
think of overruling and distinguishing as ways of breaching stare decisis. But we
can also see overruling and distinguishing as part of the practice of stare decisis.
On this broader view, the central demand of stare decisis is that courts engage with
the past and act with integrity.
We owe our understanding of the special connection between law and integrity
to the work of Ronald Dworkin. Dworkin was the first to recognize that integrity
is central to understanding our legal institutions. This was no trivial observation,
because (as we shall see below) integrity may not be a value for all institutions, per-
haps not even all law-like ones.
Before we can explore the connection between stare decisis and integrity, we
must get a fix on what integrity is. Providing a full account is too large an endeavor
to take on here, but some preliminary efforts will help us to get an alternative
account of stare decisis off the ground. According to Dworkin:
We want our neighbors to behave, in their day-to-day dealings with us, in the way we think
right. But we know that people disagree to some extent about the right principles of behav-
ior, so we distinguish that requirement from the different (and weaker) requirement that
they act in important matters with integrity, that is, according to convictions that inform
and shape their lives as a whole, rather than capriciously and whimsically.²³
Acting with integrity does not require that one act correctly. Rather, it requires
that one always act in accord with genuine convictions about what the right way
to act is. Integrity may seem like a second-best sort of value, one to be pursued
only when one cannot be confident of acting correctly. Indeed, Dworkin seems to
invite this understanding of integrity by suggesting that we demand it of others

²³ R. DWORKIN, LAW’S EMPIRE 166 (1998).

114 Scott Hershovitz

because we know that we disagree about what is right. Integrity is not a second-
best value, however, and our primary reason for demanding integrity in ourselves
and others is not our inability to agree about what is right. Rather, we demand
integrity because, whatever doubt we have about particular moral views, we are
confident that the demands of morality are coherent.²⁴ We are also confident that
morality does not demand that we act capriciously or whimsically in matters of
importance. Thus, if we are striving to act morally, we will act with integrity.
Let me put the point another way. Someone who acts with integrity may never-
theless do something she ought not to do from time to time. But someone who
acts without integrity, someone who acts incoherently or capriciously in matters
of importance, simply cannot be acting morally except by happenstance. A lack of
integrity signifies a lack of a commitment to act morally.
An individual displays integrity when her actions taken as a whole reflect a
commitment to a coherent and defensible moral view. The moral view must be
coherent because the demands of morality are coherent. However, commitment
to an evil moral view is no virtue simply because the view is coherent. Thus,
integrity also requires that one act in accord with a defensible moral view.
Otherwise, acting with integrity would not be a way of striving to act morally.
Of course, a defensible moral view is not necessarily a true one. This is why
Dworkin is right when he says that the demand that we act with integrity is
weaker than the demand that we act morally. However, acting with integrity is
part of striving to act morally, and that is the source of its value. We respect others
for their integrity even when we disagree with their actions because we recognize
their genuine commitment to acting morally.
Now we get to the important part for understanding stare decisis. Acting with
integrity requires recognizing that what one has done in the past is relevant to
what one ought to do now. Integrity requires a commitment to a moral view, and
one can only display a commitment to a moral view by a pattern of behavior across
time. Constantly shifting moral views are a sign of caprice, not integrity.
Importantly, refusing to change one’s moral views in the face of persuasive rea-
son to do so is also inconsistent with integrity. Remember that integrity is valuable
because it is an aspect of striving to act morally. If you are genuinely striving to act
morally, you will change your beliefs and behavior in response to persuasive argu-
ment or new evidence. A rigid refusal to change one’s moral convictions in the face
of new information is not a sign of integrity; it is a sign of obtuseness.
There are at least two situations in which integrity requires one to repudiate
one’s past. The first is when one’s moral convictions undergo genuine change.
Integrity requires acting in accord with one’s new moral convictions. If a person’s
moral convictions are constantly shifting, we will not say that she acts with integrity
even if she always acts in accord with her genuine moral convictions. This is

²⁴ I am using “morality” in its widest sense, in which the demands of morality are coextensive with
the demands of reason.
 Integrity and Stare Decisis 115

because we doubt that she has any commitment to her moral views or to acting
morally. But in the normal case, revisions in one’s beliefs and behaviors are not
only consistent with integrity, they are required by it.
One should also repudiate one’s past when one’s past behavior is inconsistent
with the moral commitments one has made. This is an all too common occurrence
for most of us. Integrity does not require that one repeat one’s mistakes; rather, it
requires correcting them to bring one’s behavior in accord with one’s moral com-
mitments.
These remarks are, of course, only exploratory. A full consideration of the nature
of integrity demands more space and attention than available here. However, we
have made enough progress in understanding what integrity demands of individu-
als that we can turn our attention to what integrity requires of courts.

VI

Courts are moral actors, and a court can display integrity in much the same way
that an individual can. A court displays integrity when its decisions reflect a com-
mitment to a coherent and defensible view of the rights and duties people have
under the law. Such a commitment can only be displayed by a pattern of decisions
across time. If a court’s rulings change capriciously, if it fails to pay heed to its own
pronouncements, we will doubt that it has any genuine commitment to the views
it expresses. On the other hand, if the court takes seriously what it has said in the
past and it displays consistency and coherence in action, we will believe that the
court acts on the basis of genuine convictions about the content of the law.
Why should courts act with integrity? For insight, let us turn once again to
Dworkin. He writes:
Integrity becomes a political ideal when . . . we insist that the state act on a single,
coherent set of principles even when its citizens are divided about what the right princi-
ples of justice and fairness really are. We assume, in both the individual and political
cases, that we can recognize other people’s acts as expressing a conception of justice or
decency even when we do not endorse that conception ourselves. This ability is an
important part of our more general ability to treat others with respect, and it is therefore
a prerequisite of civilization.²⁵
We want the state (and, derivatively, its courts) to act in accord with a single,
coherent set of principles for the same reason we want individuals to do so. We are
confident that morality provides a coherent vision of what we owe to one another,
and that that vision does not demand that the state or its agents act capriciously in
matters of importance. If the state acts incoherently or capriciously in matters of
importance, it cannot be acting morally except by happenstance. We may disagree

²⁵ DWORKIN, supra note 23, at 166.

116 Scott Hershovitz

about what morality requires of the state, but we want the state to strive to act
morally. Acting with integrity is a sign that it does so.
Dworkin breaks down the demands of integrity into two principles: the principle
of integrity in legislation and the principle of integrity in adjudication. The latter
principle, he says, explains “why judges must conceive of the body of law they
administer as a whole rather than as a set of discrete decisions they are free to make or
amend one by one, with nothing but a strategic interest in the rest.”²⁶ Stare decisis is
a means by which we promote this sort of integrity in judicial decision making.
A court that considers itself bound by the principle of stare decisis recognizes
that what it has done in the past affects what it ought to do now. Of course,
integrity in judicial decision making is no more a matter of slavishly repeating past
decisions, right or wrong, than it is for individuals. Just as individuals must some-
times repudiate their past, courts must do so as well. They must do so whenever
their convictions about the content of the law undergo genuine change, and they
must do so when they discover that their past decisions conflict with their genuine
commitments. Of course, if a court constantly shifts its views, we will doubt that
it has any genuine commitment to them, and we will not regard it as acting with
integrity. But, in the normal case, a court can overturn a precedent it regards as
mistaken without doing any violence to its integrity, and indeed, integrity may
demand that it do so.
There is good reason to think of stare decisis as a broader practice than simply
following precedent. If a court seeking to act with integrity has previously
announced a rule of law, it has three options: it can follow it, it can overrule it, or it
can distinguish the case. Overruling and distinguishing are as much ways of
engaging with the past as following is. They are ways of saying, “we recognize that
our prior decision is relevant in deciding what we ought to do now, but for these
reasons we are not following it here.” A court that did not consider itself bound by
stare decisis would not need to overrule or distinguish cases because it would not
recognize what it had done in the past as relevant to what it ought to do now.
An example of such an institution may help to make the point clear. Up until a
few years ago, allegations of student conduct violations at the University of
Georgia were adjudicated by an organization called the Student Judiciary.²⁷ The
Student Judiciary heard cases running the gamut from trivial infractions like
excessive noise in dorms to serious offenses such as DUI and sexual assault.
Members of the Student Judiciary sat on panels as judges in what were essentially
mini-trials to determine whether a student violated a rule and, if so, to impose an
appropriate sanction. Sanctions ranged from reprimand to expulsion.

²⁶ DWORKIN, supra note 23, at 167.

²⁷ In the mid-1990s I was a member of the Student Judiciary at the University of Georgia, and
I served for a year as its Chief Justice. The Student Judiciary has recently been replaced by the
University Judiciary, a group similar in function to the Student Judiciary, but somewhat different in
form.
 Integrity and Stare Decisis 117

The Student Judiciary had no system of precedent. A panel would not consider
the decision of a previous panel, even if a previous decision is precisely on point.
Thus, no case law grew up around the conduct rules or their implementation.
Each panel treated each case as if no others had preceded it.
The Student Judiciary shunned precedent for a variety of reasons. One reason is
that it was seeking to avoid some of the accoutrements of real legal systems. It did
not want to require student defendants to research previous decisions, nor did it
want to expend resources cataloguing them. For the Student Judiciary, efficiency
counseled against stare decisis.
The most important reason for shunning precedent, however, was the way the
Student Judiciary conceived of its mission. It understood itself to share the educa-
tional aims of the broader university. The organization believed that its responsi-
bility was to provide each student who appeared before a panel with the best
educational experience it could. A system of precedent might have gotten in the
way of tailoring each student’s hearing and sanction to their individual educa-
tional needs. Without precedent, maximum flexibility was maintained. The sys-
tem’s rejection of stare decisis was so complete that decisions from prior cases were
not even considered relevant, let alone dispositive.
Integrity was simply not a value the adjudicative practices of the Student
Judiciary recognized. This is interesting, because the Student Judiciary was as
court-like as an institution can be without being part of an actual legal system.
The Student Judiciary administered a system of conduct rules of general applica-
tion through adjudicatory bodies with judges and lawyer-like advocates. The
Student Judiciary was even an organ of the state.²⁸
I suspect that the Student Judiciary was, in fact, concerned with the integrity of
its decision making. Stare decisis is not the only way to promote integrity. The
Student Judiciary demanded that potential members participate in lengthy train-
ing, which promoted consistency and coherence in the decisions of its panels, and
members received continuing education as well. Nevertheless, in the individual
case, the Student Judiciary was designed to be unresponsive to claims that a prior
decision bound it to a course of action. Thus, unlike a court that adheres to stare
decisis, the Student Judiciary never had the need to distinguish or overrule one of
its prior decisions.
In contrast to the Student Judiciary, a court that adheres to stare decisis is differ-
ent not simply because it has a commitment to following its prior decisions, but
because it has committed itself to the idea that what it has done in the past is relevant
to what it ought to do now. Such a court answers to its precedents, by following
them, distinguishing them, and, on occasion, overruling them. Following, distin-
guishing, and overruling are all part of the pursuit of integrity in adjudication.

²⁸ Red & Black Pub. Co., Inc. v. Board of Regents, 427 S.E. 2d 257 (Ga. 1993) (holding that the
Student Judiciary is subject to Georgia’s Open Records Act, which applies to state agencies).
118 Scott Hershovitz

VII
This essay started with a question: could legal reasoning really be so different from
everyday reasoning that principle requires courts to make mistakes repeatedly
rather than correct them? Notwithstanding the traditional view of stare decisis,
the answer is no. Stare decisis does not require a court to blindly follow incorrectly
decided precedents. Nor does it require a court to stand by a precedent irrespective
of its merit. What stare decisis does require is that courts engage with the past and
act with integrity. They do this when they display a commitment to a coherent,
defensible view of the content of the law.
Now that we have an expanded view of the practice of stare decisis, it is reason-
able to wonder whether efficiency and fairness might not play a role in justifying it
after all. Neither efficiency nor fairness seemed promising as a justification for a
practice of following past decisions irrespective of their merit. But the broader
practice of engaging with the past by following, overruling, and distinguishing
precedent may well be both fair and efficient. Whether it is either, of course,
depends on how good a job courts do of it. The efficiency and fairness of stare
decisis conceived broadly still depends, in part, on the efficiency and fairness of
the decisions in particular cases.
The real place for efficiency and fairness in an explanation of stare decisis is not
so much in justifying the practice, but in giving it its contour. This essay has not
addressed the conditions under which a court ought to overturn one of its prece-
dents. In any given case, the question of whether a court should follow a precedent
depends crucially on matters of fairness and efficiency, and on a multitude of
other values as well. The fact that courts bother to engage with precedent at all,
however, is best explained by judicial aspirations to act with integrity.
 6
The Many Faces of Political Integrity
Dale Smith*

Crucial to a proper understanding of Ronald Dworkin’s approach to law and

adjudication is an understanding of what he means by “integrity.” The concept
of integrity is introduced in Law’s Empire¹ via a discussion of “checkerboard
solutions.” However, it is far from clear what checkerboard solutions are, let
alone what is wrong with them. This chapter engages in a search for an account
of what is wrong with checkerboard solutions that is both a plausible interpreta-
tion of Dworkin’s discussion and is defensible in its own right. In Part I, I present
Dworkin’s claims about checkerboard solutions. Part II then highlights an ambi-
guity in his discussion, before considering a number of different interpretations
of his position, each of which attempts to identify a single flaw inherent in
checkerboard solutions. I argue that none of these attempts succeed, before
canvassing the possibility that there is more than one thing wrong with such
solutions. While I suggest that this offers the most promising interpretation of
Dworkin’s writings on this topic, I go on (in Part III) to question whether this
account of what is wrong with checkerboard solutions can provide the support
for Dworkin’s broader claims about integrity that he believes it can. I argue that
Dworkin has failed to vindicate his claim that integrity should characteristically
trump justice in cases where they conflict, and that his discussion of checker-
board solutions provides only limited support for his claim that integrity offers a
good fit with our political beliefs and practices.
This chapter focuses primarily on Dworkin’s discussion in Law’s Empire, since it
is there that he discusses checkerboard solutions in the greatest detail. However,
since this chapter is ultimately concerned with the value of integrity, it should be
acknowledged that Dworkin has developed his views about the analysis of polit-
ical concepts like integrity since Law’s Empire. In particular, he now argues that we
can understand a political concept like integrity only in light of its relationship to

* I am grateful to Samantha Besson, Ronald Dworkin, Patrick Emerton, Jeff Goldsworthy, Scott
Hershovitz, Lewis Kornhauser and Nicos Stavropoulos for their helpful comments on earlier drafts of
this chapter.
¹ R. Dworkin, Law’s Empire (1986). All references in parentheses are to this work.

The Many Faces of Political Integrity. Dale Smith.

© Oxford University Press 2006. Published 2006 by Oxford University Press.
120 Dale Smith

other political concepts such as justice and legality.² On this view, we can fully
understand the concept of integrity only by constructing a holistic (and interpre-
tive) account of all our political values, at least part of which involves tracing the
interrelations between integrity and other political values. As Dworkin himself
points out, this is an enormous undertaking, which I cannot possibly perform in
this chapter. However, I hope that this chapter could be seen to represent a modest
and limited contribution towards that overarching goal, advancing to some extent
our understanding of the value of integrity and its relationship to other political
values such as justice and fairness.

I. What is a Checkerboard Solution?

Dworkin claims that there are two commonly accepted political values—justice
and fairness.³ However, he offers a somewhat idiosyncratic account of these two
values:
Fairness in politics is a matter of finding political procedures—methods of electing officials
and making their decisions responsive to the electorate—that distribute political power in
the right way. That is now generally understood, in the United States and Britain at least,
to mean procedures and practices that give all citizens more or less equal influence in the
decisions that govern them. Justice, on the contrary, is concerned with the decisions that
the standing political institutions, whether or not they have been chosen fairly, ought to
make. If we accept justice as a political virtue we want our legislators and other officials to
distribute material resources and protect civil liberties so as to secure a morally defensible
outcome. (164–165)
Dworkin claims that there is also a third political ideal—the ideal of integrity.
This ideal “requires government to speak with one voice, to act in a principled
and coherent manner toward all its citizens, to extend to everyone the same
substantive standards of justice or fairness it uses for some” (165). For example,
if the government relies on majoritarian principles to justify its decisions about
who may vote, it must respect the same principles when drawing electoral
boundaries. Similarly, if the government appeals to the principle that people
have a right to receive compensation from those who injure them carelessly as a
reason why manufacturers are liable for defective cars, it must give effect to
that principle in deciding whether accountants are liable for their mistakes as
well (165).

² R. Dworkin, Hart’s Postscript and the Character of Political Philosophy, 24 Oxford J. of Legal
Stud. 1, 17 (2004).
³ He mentions a further political ideal—procedural due process—but it plays no role in his argu-
ment, and so I shall leave it to one side.
 The Many Faces of Political Integrity 121

Dworkin labels this ideal “political integrity” to highlight the connection with
the notion of personal integrity, which involves acting “according to convictions
that inform and shape [one’s life] as a whole, rather than capriciously or whimsi-
cally” (166). Integrity becomes a political ideal when we insist that the state act on
a single, coherent set of principles even though its citizens disagree about what the
right principles of justice and fairness really are (166).⁴
To justify his claim that integrity is a distinct political ideal, Dworkin seeks to
show that it both fits with, and provides a sound justification of, our political
practices. It is in seeking to show that this ideal fits our political practices that he
introduces the notion of a checkerboard solution. He claims that we can explain
our opposition to checkerboard solutions only on the basis that we accept the
ideal of integrity.
So what is a checkerboard solution? This question will occupy us for much of
this chapter, but let us start with Dworkin’s initial characterization of the
phenomenon. He states that he is not using the word “checkerboard” to describe
statutes that make distinctions that claim to be justified on the basis of a policy
best served by the discriminations in question. Instead, he uses the word “to
describe statutes that display incoherence in principle and that can be justified, if
at all, only on grounds of a fair allocation of political power between different
moral parties” (435, n. 6).
Dworkin uses the connection between fairness and checkerboard solutions
alluded to in the last quotation to set up a puzzle that he believes only integrity can
solve. He points out that, since we all value fairness but disagree about moral
issues, it seems to follow that legislation on moral issues should be a matter “of trades
and compromises so that each body of opinion is represented, to a degree that
matches its numbers, in the final result” (178). For example, if people disagree
about whether justice requires compensation for product defects that the manu-
facturer could not reasonably have prevented, why should not the legislature
impose strict liability on car manufacturers but not on manufacturers of washing
machines?⁵ Or, if we disagree about the permissibility of abortion, why should not
Parliament make abortion criminal for women who are born in even-numbered
years but not for those born in odd-numbered years? (178).
Dworkin claims that this sort of solution to contested moral issues seems
mandated by fairness, since it secures the kind of proportional influence of
citizens over the political process that fairness (on Dworkin’s definition) appears
to recommend (435, n. 3). Allowing each group in society to choose some part of
that society’s abortion law, in proportion to their numbers, “is fairer (in our

⁴ Several commentators have suggested that there are important disanalogies between personal
integrity and what Dworkin calls “political integrity”: e.g. J. Raz, Ethics in the Public Domain:
Essays in the Morality of Law and Politics 307 (1995). However, I shall not pursue this issue here.
⁵ Though I shall question whether this is really an example of a checkerboard solution: see infra
text for note 18.
122 Dale Smith

sense) than the winner-take-all scheme our instincts prefer, which denies many
people any influence at all over an issue they think desperately important”
(179).⁶
However, Dworkin also claims that most of us would be dismayed by checker-
board statutes that treat relevantly similar actions (for example, abortions by
women born in even-numbered years and those by women born in odd-num-
bered years) differently on arbitrary grounds (179). This is not to say that we
always oppose political compromise, or even that we always oppose arbitrary com-
promises. Dworkin allows that we tolerate arbitrary distinctions with regard to
matters of policy (such as parking regulations), but claims that we do not tolerate
them when matters of principle are at stake. In other words, if we are concerned
only with promoting social welfare, then arbitrary distinctions are permissible,
but they are unacceptable where moral issues (or someone’s rights) are involved.⁷
When matters of principle are at stake, we accept that each point of view should
be allowed a voice in the deliberative process, but the collective decision must
aim to settle on some coherent principle that is then applied in all appropriate
circumstances (179).
Compromise on matters of principle is not always impermissible. Sometimes,
two independent principles that we hold come into conflict, in the sense that we
cannot satisfy both on this particular occasion. In such cases, we might give
weight to each of those principles in a certain relation (for example, by develop-
ing a tax scheme that seeks to respect both property rights and equality of
opportunity, to differing degrees). However, whatever relative weighting we give
those principles must flow throughout the scheme in question, and to other deci-
sions that involve the same two principles (435–436, n. 7). Dworkin claims that
this kind of conflict differs from the contradiction contained in checkerboard
statutes. In the latter case, one principle of justice is not outweighed or qualified
by another in a way that expresses a ranking of the two; instead, only a single
principle is involved, which is affirmed for one group and denied for another. It is
this which Dworkin claims we denounce: “If there must be compromise because
people are divided about justice, then the compromise . . . must be compromise
about which scheme of justice to adopt rather than a compromised scheme of
justice” (179).

⁶ See also (435, n. 3), where Dworkin explicitly states that fairness supports checkerboard solu-
tions. Dworkin does not specify precisely what sort of influence is required by the value of fairness,
but presumably he has in mind the ability to have one’s views reflected in political outcomes. On
this view, one has political influence to the extent that one’s favoured position is reflected in the posi-
tion adopted by the political process, rather than to the extent that one participates in the political
process. Whether this is an appropriate account of fairness falls outside the scope of this chapter: see
infra, note 17.
⁷ This is an over-simplification of Dworkin’s views. While he claims that integrity does not require
any simple form of consistency between policies (in the way that it does with regard to principles), he
claims that it does have some role to play even here (221–222).
 The Many Faces of Political Integrity 123

We cannot explain our opposition to checkerboard solutions by reference to

the value of fairness, since we have seen that checkerboard solutions promote that
value (by distributing political power more evenly than the alternatives). Nor,
according to Dworkin, is our antipathy to internal compromise⁸ based on the
ideal of justice. He acknowledges that it is tempting to argue that checkerboard
solutions are unjust because they treat different people differently for no good
reason (whereas justice requires us to treat like cases alike). However, he claims
that—while the defect in checkerboard solutions must lie in their distinctive
feature (that they treat people differently when no principle can justify this
distinction)—we cannot explain why this is objectionable if we focus solely on
considerations of justice (180–181).
Why not? Dworkin claims that checkerboard solutions can prevent instances
of injustice that would otherwise occur (181). For example, if we oppose abor-
tion, the abortion statute described above will have the merit of at least ensuring
that some foetuses (those belonging to women born on even-numbered years)
are protected. Moreover, if we are a minority in society, our choice may be
between supporting a checkerboard solution and allowing abortion in all cases. If
so, justice seems to favour the checkerboard solution, since this would minimize
instances of injustice.
Justice does not always favour checkerboard solutions. If we have majority
support within society, then we may be able to achieve the optimal outcome,
which will clearly not be the checkerboard solution. Nevertheless, unless we are
always part of the majority, checkerboard solutions will sometimes be the best
option available to us from the standpoint of justice. Yet, according to Dworkin,
internal compromises are always objectionable, and we cannot explain why this is
so if we focus solely on considerations of justice.⁹
A critic of Dworkin might argue that we can explain why checkerboard solu-
tions are always objectionable by reference to the ideal of justice. A checkerboard
solution (by definition) cannot be a fully just solution. At most, it is the least worst
alternative, because it minimizes injustices that we could not otherwise prevent.
Therefore, we always have reasons of justice to treat a checkerboard solution as
objectionable, even if we also think that it represents the most just (or least unjust)
outcome we can hope for under present circumstances.
Dworkin’s response is that, not only do we reject checkerboard solutions, but
many of us would prefer either of the alternative solutions to a checkerboard solu-
tion. For example, many opponents of abortion would (according to Dworkin)
prefer that abortion always be available than that we adopt a checkerboard solu-
tion regarding abortion (182). This attitude cannot be explained by reference to

⁸ Dworkin uses the phrases “checkerboard solution” and “internal compromise” interchangeably.
I shall follow this practice.
⁹ As we shall see, Dworkin concedes that sometimes checkerboard solutions may be acceptable,
but only because our distaste for them is outweighed by other considerations.
124 Dale Smith

the ideal of justice, since (from the viewpoint of an opponent of abortion)

making abortion available in all cases will produce more instances of injustice
than a checkerboard solution.
Dworkin recognizes that not everybody would condemn every checkerboard
solution. For example, some opponents of abortion might prefer a checkerboard
solution to a law that legalized abortion in all instances. However, he claims that,
if they would rank a checkerboard solution below all other alternatives when
considering other political issues, this shows that they recognize a reason not to
adopt checkerboard solutions, albeit a reason that is outweighed when the
substantive issue is particularly important (182–183). Alternatively, their opposi-
tion to checkerboard solutions can be seen from the fact that they would prefer a
statute prohibiting abortion except for women who became pregnant as the
result of being raped to a statute prohibiting abortion except for women born on
odd-numbered years, even if both statutes allowed exactly the same number of
abortions.¹⁰ This preference cannot be explained by reference to the ideal of
justice, since (we are assuming) there will be exactly the same number of injust-
ices under either statute.¹¹ Instead, this preference reflects the fact that the for-
mer statute gives effect to principles that opponents of abortion can recognize as
principles of justice (even though they reject those principles), whereas the latter
statute simply affirms for some people a principle that it denies to others (183).
To summarize, Dworkin claims that we recognize at least a pro tanto reason to
reject checkerboard solutions. However, he claims, fairness often supports
checkerboard solutions (given that people disagree about political issues) and
justice sometimes supports them (when the alternative would be even more
instances of injustice). Therefore, we must look elsewhere to explain our rejection
of those solutions. This suggests that there must be a third political ideal to which
we attach importance, since the first two ideals do not explain our attitude
towards internal compromises. Dworkin claims that this third ideal is integrity.
A state that adopts a checkerboard solution lacks integrity, because it must
endorse principles to justify part of what it has done that it must reject to justify
the rest (183–184). It is this fact that explains our opposition to checkerboard
solutions, despite the considerations of fairness that favour such solutions.
Checkerboard solutions are not the only way in which integrity can be flouted.
It is also violated whenever a community adopts different laws, each of which is

¹⁰ I shall refer to these as the rape-abortion law and the checkerboard-abortion law, respectively.
Possibly to make it more plausible that the rape-abortion law would permit the same number of
abortions as a checkerboard solution, Dworkin refers to a checkerboard solution whereby only
women born in one specified decade each century could have an abortion (183). However, for sim-
plicity, I shall discuss only one checkerboard solution regarding abortion—namely, the one referred
to in the text.
¹¹ Can we explain the preference on the basis that abortions in non-rape cases are worse (from the
viewpoint of justice) than abortions in rape cases? No, because Dworkin stipulates that, on the view
under consideration, it makes no difference whether the pregnancy is a result of rape (183).
 The Many Faces of Political Integrity 125

internally coherent but which cannot be defended together as expressing a coher-

ent ranking of different principles of justice and fairness. Dworkin recognizes that
our legal system “constantly” violates integrity in this way, since we cannot bring
all the rules our judges enforce together under a single coherent scheme of princi-
ple, but he claims that we regard this as a flaw in our legal system (184). Of more
importance for present purposes, however, is the fact that this illustrates that the
adoption of checkerboard solutions is not the only way to infringe integrity,
though Dworkin claims that it is a particularly flagrant and easily avoidable way of
doing so (184, 217).
Finally, Dworkin acknowledges that integrity is not always more important
than fairness and justice, and so law-makers should not always avoid breaches of
integrity. Indeed, despite checkerboard solutions being especially egregious viola-
tions of integrity, Dworkin suggests that there may be occasions on which even
a checkerboard solution may be justified. For example, a legislature that favours
no-fault compensation in defective product cases might find it impossible in prac-
tice to introduce such a scheme except for car manufacturers. The introduction of
such a scheme only for car manufacturers would be condemned by integrity, but
favoured by justice, and this might be a case where justice is more important than
integrity.¹² Nevertheless, Dworkin insists that while law-makers can occasionally
favour justice over integrity, they should not “characteristically” do so (218).¹³

II. What is Wrong with Checkerboard Solutions?

It is noticeable that Dworkin describes what is wrong with checkerboard solutions

in a number of different ways, not all of which are clearly synonymous. Here is a
(by no means complete) sample:
● checkerboard solutions “display incoherence in principle” (435, n. 6);
● checkerboard solutions “can be justified, if at all, only on grounds of a fair
allocation of political power between different moral parties” (435, n. 6);
● checkerboard solutions “treat similar [actions] differently on arbitrary
grounds”, in situations where matters of principle are at stake (179);
● checkerboard solutions concern a single principle “which is affirmed for one
group and denied for another” (435–436, n. 7);
●
checkerboard solutions “treat people differently when no principle can justify
the distinction” (180); and
●
checkerboard solutions do not give effect to anything that one can recognize
as a principle of justice (even if one rejects that principle) (183).

¹² The example is Dworkin’s (217–218).

¹³ Or else their claim to legitimacy will be undermined: see infra text between notes 45 and 46.
126 Dale Smith

Are these merely different ways of making the same point? Certainly, there are
significant similarities between several of the items on this list. For example,
treating similar actions differently on arbitrary grounds and treating people differ-
ently when no principle can justify the distinction may simply be two ways of
describing the same thing. However, even here, much depends on what we mean
by “arbitrary”. If “arbitrary” means unpredictable, or without legitimate author-
ity,¹⁴ then it is not so clear that the two descriptions are equivalent. There may be
a principle that justifies differential treatment for two people who perform similar
actions, even if the decision to treat them differently is unpredictable (for example,
because we were unaware of that principle before the decision) or lacks legitimate
authority (for example, because the decision-maker does not have authority to
deal with the relevant issues).¹⁵
Similarly, it is not immediately obvious that one cannot affirm a single prin-
ciple for one group and deny it for another without lapsing into “incoherence in
principle.” For example, many countries affirm the principle of freedom of associ-
ation for most of their citizens, but not for offenders who have been convicted and
sentenced to jail. Do such countries lack coherence in principle? Perhaps there is
more than one principle at stake in this example (and so this is not an example of
affirming a single principle for one group and denying it for another). However, at
least at first sight, more than one principle is at stake in Dworkin’s examples of
checkerboard solutions, as well.¹⁶
It is not clear what we should conclude from this. One possibility is that
Dworkin equivocates as to exactly what is wrong with checkerboard solutions,
switching haphazardly between different explanations. However, another possi-
bility is that—despite appearances—these various characterizations of checker-
board solutions in fact amount to the same thing. Or perhaps there is more than
one thing wrong with checkerboard solutions, reflecting the fact that there is more
than one principle of integrity that they violate (just as there is more than one
principle of justice or fairness).
What is clear, however, is that further consideration of Dworkin’s account of
checkerboard solutions is called for. I shall seek to provide an interpretation of
Dworkin’s position that both fits what he says about checkerboard solutions and is
sufficiently plausible in its own right to warrant consideration. More precisely, my
¹⁴ Both of these are possible meanings of “arbitrary” suggested by J. Waldron, Law and
Disagreement 167–168 (1999).
¹⁵ Dworkin may not be using the word “arbitrary” in either of these senses. If so, however, we need
to determine in what sense he is using the word (see infra Section A).
¹⁶ In fact, it is clear that checkerboard solutions need not concern only a single principle. Imagine
a variation of the checkerboard abortion law, under which abortions are permissible for women born
in the first four months of the year provided the abortion is performed in the first two trimesters of
the pregnancy, are never permissible for women born in the next four months of the year and are per-
missible for women born in the last four months of the year only if their pregnancy is the result of
rape. On any reasonable way of demarcating principles, there is more than one principle at stake in
this law, and yet the law is clearly a checkerboard solution, since women are treated differently
depending on the month in which they were born.
 The Many Faces of Political Integrity 127

aim in this Part is to find the best account of checkerboard solutions available to
Dworkin, before (in Part III) considering what implications that account has for
Dworkin’s broader theory. To this end, I shall begin by considering whether there
is a single feature of checkerboard solutions that can explain our attitude towards
such solutions, before considering whether some combination of such features
might underlie our opposition to internal compromise. However, in doing so, I shall
consider only the more promising interpretations of Dworkin’s views, rather than
attempting a comprehensive survey of suggestions as to what is wrong with
checkerboard solutions.¹⁷
Since I am searching for the best interpretation of Dworkin’s discussion of
checkerboard solutions, I shall assume that there really is something wrong with
such solutions. However, even making this assumption, not all of Dworkin’s
examples are equally convincing. Take his example of a statute that introduces
no-fault liability for car manufacturers but not for washing machine manufac-
turers. Might there not be a principled distinction between the two types of
manufacturer? For example, defective cars might cause personal injury more
often than defective washing machines, and we might consider it more import-
ant to compensate for personal injury than for other losses resulting from defective
products. Rather than concluding that this represents an exception to the
general rule that checkerboard solutions are unacceptable, I suggest that—if
there is a principled distinction—this shows that Dworkin’s example is not
really a checkerboard solution after all.¹⁸ Thus, I shall focus on another of
Dworkin’s examples, which clearly is a checkerboard solution—namely, a law
that prohibits abortions for women born in even-numbered years but not for
those born in odd-numbered years. Given the uncertainty about how to charac-
terize checkerboard solutions, I shall seek to derive guidance from this paradigm
example of an internal compromise.

A. Arbitrary Distinctions and Unreasonable Compromises

The first possibility I shall consider is that checkerboard solutions are defective
because they draw arbitrary distinctions between relevantly similar people, actions

¹⁷ Several commentators have argued that, contrary to Dworkin, we can explain why checker-
board solutions are undesirable by reference to the values of fairness or justice (e.g. S. Wasserstrom,
The Empire’s New Clothes, 75 Geo. L.J. 199, 249, n. 168 (1986); D. Reaume, Is Integrity a Virtue?
Dworkin’s Theory of Legal Obligation, 39 U. Toronto L.J. 380, 391–393 (1989)). However, I am
searching for the best interpretation of Dworkin’s views about checkerboard solutions, and so shall
leave to one side the possibility that Dworkin is wrong in claiming that the distinctive flaw of
checkerboard solutions is not that they are unfair and/or unjust.
¹⁸ A similar approach is adopted by G. C. Christie, Dworkin’s “Empire”, Duke L.J. 157, 185
(1987). Dworkin could argue that the proposed legislation is a checkerboard solution because—even
if a principled distinction could be drawn between car and washing machine manufacturers—this
was not the motivation for the distinction drawn in the legislation. Instead, that distinction was moti-
vated by considerations of political feasibility. However, the problem with checkerboard solutions
seems to lie in the nature of the distinction drawn, not in the motivation for drawing that distinction.
128 Dale Smith

or situations. As we have seen, there is textual support for interpreting Dworkin in

this way: he claims that checkerboard solutions treat similar actions differently on
arbitrary grounds, in situations where matters of principle are at stake (179).
Moreover, prima facie, this claim appears plausible. There does appear to be some-
thing arbitrary about allowing abortions for women born in even-numbered years
but not for those born in odd-numbered years.
However, there are two problems with this suggestion. First, the word
“arbitrary” often hides more than it illuminates. What does it mean to say that a
checkerboard solution is arbitrary? Different people use the word “arbitrary” to
mean different things,¹⁹ and so being told that checkerboard solutions are
arbitrary is not particularly helpful unless we are also told in what sense they are
arbitrary. The answer to this latter question may tell us what is wrong with
checkerboard solutions, but merely to claim that they are arbitrary will not further
our investigation very far. In other words, our investigation may start with the
proposition that checkerboard solutions are arbitrary, but it cannot stop there—it
must go on to consider in what sense they are arbitrary.
A similar objection can be levelled against the suggestion that checkerboard
solutions involve unreasonable compromises. Certainly, the checkerboard-
abortion law appears unreasonable, but the purpose of our inquiry is to determine
in what way it is unreasonable. It is not sufficient simply to note that this law
appears unreasonable.
Secondly, on at least one way of giving more determinate content to the word
“arbitrary,” it is not clear that checkerboard solutions are arbitrary. Dworkin
concedes that there are reasons, based on considerations of fairness, for adopting
checkerboard solutions. Checkerboard solutions seem to produce a more even
distribution of political influence than the alternatives, and might be championed
on this basis. Presumably, this is not enough to vindicate such solutions, but—if
by “arbitrary” we mean irrational or unreasoned—it may be enough to show that
they are not arbitrary. There are indeed reasons (of fairness) that support checker-
board solutions, even if those reasons are outweighed by other considerations.²⁰
It might be responded that there is still a sense in which checkerboard solu-
tions are unsupported by reason, and so are arbitrary. Considerations of fairness
might give us reason to compromise so that 50 per cent of women are eligible for
an abortion and 50 per cent are not, but this still does not support any particular
outcome. We could choose to allow abortions for those women who are born in
odd-numbered years, but we could alternatively allow abortions for those
women who are born in the first six months of each year. We have no reason to
choose one of these outcomes over the other, and so our choice is arbitrary.²¹

¹⁹ WALDRON, supra note 14, at 167–168 suggests three different meanings.

²⁰ Cf. A. Marmor, The Rule of Law and Its Limits, 23 Law and Philosophy 1, 29–30 (2004),
though Marmor may not accept Dworkin’s conception of fairness.
²¹ This objection was suggested by Scott Hershovitz.
 The Many Faces of Political Integrity 129

In assessing this objection, it is important to realize that we do have a reason to

adopt the former option (namely, a reason of fairness). Our problem is that we
also have a reason to adopt the latter option, and no reason to prefer one of these
two options to the other. This sort of situation is not uncommon. For example, if
we are designing traffic laws for the first time, we have a reason to require everyone
to drive on the right-hand side of the road (since this would reduce collisions), but
we also have a reason to require everyone to drive on the left-hand side of the road
(since this would also reduce collisions), and no reason to prefer the first option to
the second (or vice versa). In such situations, our decision may be arbitrary, but it
is not necessarily objectionable (as the traffic example shows). Therefore, it cannot
be this feature of checkerboard solutions alone that renders them objectionable.

B. Irrelevant Considerations
The most obvious explanation of what is wrong with the checkerboard-abortion
law is that its application depends on an irrelevant consideration, because it
makes the permissibility of abortion depend on whether one was born on an
even-numbered year, a consideration that is morally irrelevant to the issue at
hand. We can render this notion of “moral irrelevance” somewhat more precise
by saying that a morally irrelevant consideration is one that does not provide
even a pro tanto moral reason. Thus, to say that whether one was born on an
even-numbered year is morally irrelevant to whether one should be permitted to
have an abortion is to say that it does not provide even a pro tanto moral reason
for permitting (or not permitting) one to have an abortion.²²
It is interesting that Dworkin does not canvass this possible explanation in
Law’s Empire. Indeed, his discussion appears incompatible with it. Take his
suggested alternative to the checkerboard-abortion law—namely, allowing
abortions only for rape victims. Dworkin states that this alternative would be
seen to be superior to the checkerboard statute even if one believes “that it makes
no difference whether the pregnancy is the result of rape” (183). To believe that it
makes no difference whether the pregnancy is the result of rape is to believe that
whether one has been raped is irrelevant to whether one is permitted to have an
abortion (that is, it does not provide even a pro tanto reason for permitting an
abortion). Yet Dworkin expects someone who holds that view still to prefer the
rape-abortion statute to the checkerboard-abortion statute (and he regards this
preference as reflecting an opposition to checkerboard solutions that he shares).

²² This may be an unduly narrow conception of moral relevance. For example, even if one does
not believe that foetuses have souls, one may believe that whether they have souls is relevant to the
abortion debate, since—if they did have souls—this would provide a pro tanto reason to oppose
abortion. This suggests that a consideration may also be morally relevant if it would provide a pro
tanto reason were the non-moral facts different (e.g. were foetuses to have souls). However, nothing in
this Section (Section B) turns on this point.
130 Dale Smith

Therefore, he cannot believe that the distinction between checkerboard and

non-checkerboard statutes hinges on whether the statute is based on morally
irrelevant considerations. He believes that we should regard the rape-abortion
statute as preferable to the checkerboard-abortion statute even if we believe that
both are based on morally irrelevant considerations.
Admittedly, there is another possible interpretation of Dworkin’s discussion of
the two abortion statutes. He may be supposing that the anti-abortionist believes
that it makes no difference whether the pregnancy is the result of rape in the sense
that abortion will still be impermissible, not in the sense that it does not provide
even a pro tanto reason for permitting an abortion. In other words, he may be
supposing that the fact that the pregnancy is the result of rape makes a difference
to the pro tanto reasons that apply, but makes no difference to the overall conclu-
sion that one should reach. If so, his discussion is consistent with the claim that
the distinctive feature of checkerboard solutions is that they are based on
irrelevant considerations, since the rape-abortion law is based on a relevant
consideration (whether the pregnancy is the result of rape affects the pro tanto
reasons that apply, even if it does not alter our all-things-considered conclusion).
However, if Dworkin does believe that the distinctive feature of checkerboard
solutions is that they are based on irrelevant considerations, one might have
thought that he would qualify the hypothetical anti-abortionist’s view that the fact
of rape makes no difference, to make it clear that she does not believe that the
rape-abortion law is based on irrelevant considerations. One might also have
expected him to state that this is what is wrong with checkerboard solutions. It is
noticeable that, in the list I provided earlier of six ways in which Dworkin
describes what is wrong with checkerboard solutions, none refers to “irrelevant
considerations.” And Dworkin does not conclude that the rape-abortion law will
be treated as preferable by the anti-abortionist because it reflects a relevant consid-
eration, but rather because it “gives effect to two recognizable principles of justice,
ordered in a certain way, even though [the anti-abortionist rejects] one of the
principles” (183, footnote omitted).²³
For these reasons, the “irrelevant considerations” approach to what is wrong
with checkerboard solutions does not offer a plausible interpretation of Dworkin’s
discussion. Moreover, leaving aside whether it reflects Dworkin’s views, there are
significant problems with this approach. Imagine that Anna believes that whether
a pregnancy is the result of rape really is morally irrelevant, in the sense that it does
not provide even a pro tanto reason for permitting abortion. She believes that a
foetus is a human being, and the fact that one has been raped cannot provide even
a pro tanto reason for killing a human being. (It might provide a partial excuse in
certain circumstances, but cannot begin to justify such an action.)²⁴ Now imagine
²³ This explanation of what is wrong with checkerboard solutions (along with how it differs from
the current explanation) is considered in Section D.
²⁴ It could be argued that a consideration is morally relevant if it is relevant to either the justifiabil-
ity or excusability of a certain action (in which case, whether the pregnancy is the result of rape might
 The Many Faces of Political Integrity 131

that Anna must choose between the rape-abortion law and the checkerboard-
abortion law. She believes that both are based on irrelevant considerations, and so
cannot distinguish between them on this basis. However, is it not likely that Anna
will believe that the rape-abortion law is preferable to the checkerboard-abortion
law (again, assuming that both laws allow the same number of abortions)? The
former is likely to make more sense to her because she can see how someone could
come to support it (even though she believes that they are wrong to do so),
whereas she could not see how someone could come to support the latter.²⁵ This
suggests that what is distinctively objectionable about the checkerboard-abortion
law (our paradigm case of a checkerboard solution) is not that it is based on irrele-
vant considerations.
A number of objections could be offered to this argument, of which I shall
consider only one.²⁶ This objection starts by pointing out that, according to the
“irrelevant considerations” approach, something is a checkerboard solution if its
application depends on considerations that are in fact irrelevant. Which consid-
erations people believe to be irrelevant is beside the point. Therefore, it does not
matter whether Anna believes that the fact that a pregnancy is the result of rape
is morally irrelevant. What matters is whether that fact really is morally irrele-
vant. If it is, the rape-abortion law is a checkerboard solution; if it is not, the
rape-abortion law is not a checkerboard solution.
There is some truth to this objection. Under the “irrelevant considerations”
approach, what matters is whether a consideration is in fact irrelevant, not
whether it is believed to be irrelevant. However, we need to recall the structure of
Dworkin’s argument. He wants to show that integrity is a genuine political value
by showing that (inter alia) it fits our political practices. He tries to do this by
arguing that it is the fact that we value integrity that explains why we oppose
checkerboard solutions. This means that, for Dworkin’s purposes, what we
believe is pertinent. If Anna’s beliefs and practices show that she does not oppose
checkerboard solutions on the basis that their application depends on irrelevant
considerations, then this suggests that the “irrelevant considerations” approach
cannot give Dworkin what he wants. Either the “irrelevant considerations”
approach is mistaken (in which case, we must look elsewhere to discover what is
distinctively wrong with checkerboard solutions) or else Dworkin’s argu-
mentative strategy fails (because the correct explanation of what is wrong with
checkerboard solutions does not fit with the political beliefs and practices of

be a relevant consideration). However, it is doubtful whether Dworkin can accept this argument.
Of his three political values, fairness and integrity do not seem to have anything to do with excuses,
and he defines justice to be concerned with the right outcome, whereas excuses are relevant to blame-
worthiness (not to which outcome is correct). Moreover, the “irrelevant considerations” approach
would still face the further objection presented towards the end of this Section (Section B).
²⁵ I consider this as a possible explanation of what is wrong with checkerboard solutions in
Section D. ²⁶ This objection was suggested by Scott Hershovitz.
132 Dale Smith

people like Anna, since they distinguish between checkerboard solutions and
solutions based on irrelevant factors).²⁷
I am inclined to regard the first of these two possibilities as correct. If I believed
that whether a pregnancy was the result of rape is irrelevant to the permissibility of
abortion, I would still distinguish between the rape-abortion and checkerboard-
abortion laws on the basis suggested above—namely, that I can see how someone
could come to support the rape-abortion law, but I cannot see how someone could
come to support the checkerboard-abortion law.²⁸ This suggests that it is not just
Anna’s intuitions that run contrary to the “irrelevant considerations” approach.
Of course, both Anna’s and my intuitions may be mistaken, but some further
argument is required to show that this is in fact the case.
There is another counter-intuitive feature of the “irrelevant considerations”
approach. If we focus on whether a consideration is in fact irrelevant (as opposed to
whether it is believed to be irrelevant), then whether the rape-abortion law is a
checkerboard solution depends on the answer to a fairly complex moral question
(namely, whether the fact that a pregnancy is the result of rape provides a pro tanto
reason for permitting an abortion). Similarly, whether a law permitting abortions
in the first two trimesters only is a checkerboard solution depends on (inter alia)
whether personhood starts from the moment of conception and whether this
means that the age of the foetus is morally irrelevant. However, no such difficult
moral question need be resolved to determine that the checkerboard-abortion law
is a checkerboard solution—it is obvious that the permissibility of abortion should
not hinge on the year in which one is born. In other words, when we look at the
only clear case of a checkerboard solution that we have yet found, we can determine
that it is a checkerboard solution without having to evaluate the correctness of any
contested moral position. This suggests that the “irrelevant considerations”
approach fails to capture what is distinctive about checkerboard solutions, since it
does require us to evaluate the correctness of contested moral positions in order to
determine whether something is a checkerboard solution.
An obvious response is that the checkerboard-abortion law is a paradigm case of
a checkerboard solution precisely because its application clearly depends on irrele-
vant considerations. Whether other laws amount to checkerboard solutions will
be less clear, because it will be less clear whether the considerations on which their
application depends are irrelevant. However, to downplay the fact that we do not
have to evaluate the correctness of any contested moral position to determine that
whether a woman is born in an even-numbered year is morally irrelevant is to
overlook one of the most distinctive features of the checkerboard-abortion statute

²⁷ Remember that Dworkin’s discussion of the rape-abortion and checkerboard-abortion laws is

meant to show that the correct explanation of what is wrong with checkerboard solutions does fit with
the political beliefs and practices of people like Anna.
²⁸ Though this forms only part of the basis on which checkerboard solutions are distinguished
from other solutions to political problems: see infra Section F.
 The Many Faces of Political Integrity 133

(see Section D). To provide an account of what is distinctively wrong with such a
statute, we need to look beyond the question of whether its application depends
on an irrelevant consideration.

C. Coherence in Principle
A third possibility is that checkerboard solutions are undesirable because they lack
“coherence in principle”. Dworkin considers coherence in principle to be a
desideratum both of single rules and of sets of rules. However, the problem with
checkerboard solutions cannot be that they are inconsistent in principle with
other legal rules, since they are supposed to be objectionable in their own right,
without any need to consider other aspects of the legal system. Indeed, Dworkin
explicitly treats incoherence in principle between legal rules as a way of flouting
integrity that is separate from that inherent in checkerboard solutions (184).
Thus, I shall focus upon coherence in principle within a single rule.²⁹
Dworkin claims that legal decision-making should be a matter of principle, not
of political accommodation.³⁰ In other words, legal decisions must be defensible by
reference to a coherent set of principles. However, we have seen that (on Dworkin’s
account) checkerboard solutions can be defended by reference to (presumably
coherent) principles of fairness. Thus, if the defect in checkerboard solutions is that
they are not defensible on the basis of a coherent set of principles, it must be princi-
ples of justice that we should be considering, not principles of fairness.
With this point in mind, there is reason to ascribe to Dworkin the view that
checkerboard solutions are defective because they are not defensible by reference
to a coherent set of principles. Certainly, Dworkin believes that integrity involves
“a commitment to consistency in principle valued for its own sake” (167).
Moreover, he claims that a state that adopts a checkerboard solution lacks integrity
“because it must endorse principles to justify part of what it has done that it must
reject to justify the rest” (184). For example, the checkerboard-abortion law
endorses the principle that women should be allowed abortions at will when
explaining why women born in odd-numbered years are allowed abortions, and
then rejects that principle when determining whether women born in even-num-
bered years are allowed abortions.³¹ This suggests that Dworkin believes that
checkerboard solutions lack coherence in principle, because they both uphold and
reject certain principles of justice.

²⁹ Though coherence in principle between rules is discussed briefly infra in Part III, Section B.
³⁰ R. Dworkin, Life’s Dominion: An Argument about Abortion, Euthanasia, and
Individual Freedom 146 (1993). Dworkin is talking here about judicial decisions, but he demands
the same of legislators (at least when moral issues are at stake).
³¹ This may be what Dworkin has in mind when he says that checkerboard solutions are based
on a compromised theory of justice, as opposed to a compromise regarding which theory of justice to
use (179).
134 Dale Smith

Let us examine this explanation of what is wrong with checkerboard solutions

by considering another compromise regarding abortion, a compromise that
Dworkin presumably would not regard as a checkerboard solution. This compro-
mise involves allowing any woman to have an abortion during the first two
trimesters of her pregnancy, but allowing no woman to have an abortion during
the third trimester. Since this solution is (I am assuming, for the sake of the
example) a compromise on a matter of justice, it must represent a middle ground
between two or more principles of justice (rather than being the preferred solution
in its own right). For simplicity, let us imagine that these principles are that
abortion should always be permissible and that abortion should never be permissi-
ble. Yet does this not mean that this solution is objectionable on precisely the same
grounds as the checkerboard-abortion law, since it endorses the principle that
abortion should always be permissible when explaining why women in the first
two trimesters of their pregnancy are allowed abortions, and then rejects that
principle when determining whether women in their third trimester are allowed
abortions? Does not this solution also lack coherence in principle because it both
upholds and rejects certain principles of justice?
Surely not, one might respond. The difference is that, while this solution upholds
conflicting principles of justice, it does so in a certain relation determined by a fur-
ther, coherent principle of justice (namely, that—since the level of development of
the foetus is relevant when determining whether an abortion is permissible—
abortion should be permitted in the first two trimesters, but not in the third
trimester). In other words, while our two starting principles (that abortion is
always permissible and that it is never permissible) support different parts of the
rule, there is a further principle of justice that supports the whole rule.³² The same
is not true of checkerboard solutions. For example, there is no further principle of
justice that supports the whole of the rule adopted in the checkerboard-abortion
law, since there is no principle of justice that supports prohibiting abortion only
for women born in even-numbered years.
However, is there really no principle of justice that could support the whole of
the checkerboard-abortion law? On what grounds can we dismiss someone who
asserts that there is a principle of justice that supports this law? It is not enough
that any such principle would be incorrect. The set of checkerboard solutions does
not include all compromises that uphold an incorrect principle of justice (or else
the rape-abortion law may also be a checkerboard solution, if it turns out that
whether a pregnant woman has been raped should not be determinative of
whether she is allowed an abortion). Nor is it enough that the application of the
checkerboard-abortion law would depend upon irrelevant considerations (see
supra Section B). A more promising suggestion is that the (purported) principle

³² For the sake of the example, I am assuming that we do not regard this further principle of justice
as correct, but rather regard it as an acceptable compromise between the principles of justice that dif-
ferent people do regard as correct.
 The Many Faces of Political Integrity 135

that abortions should be prohibited only for women born in even-numbered years
is not recognizable as a principle of justice (whereas the principle that abortions
should be permitted for rape victims is recognizable as a principle of justice, even if
it turns out to be incorrect). However, this suggestion shifts the focus from
whether a solution both upholds and rejects a certain principle of justice to
whether it reflects anything recognizable as a principle of justice, and so is consid-
ered in Section D.
There is a further problem with the claim that checkerboard solutions are
defective because they lack coherence in principle. What motivates the require-
ment that legal decisions be defensible on the basis of a coherent set of princi-
ples of justice? It might seem obvious that no law should both uphold and reject
a principle of justice in the way that the checkerboard-abortion law does. Yet
(on Dworkin’s account) that law can be defended by reference to a coherent set
of principles of fairness. Is this not enough to ensure coherence in principle over-
all? Why do we also need coherence in terms of principles of justice? Have we
not seen that fairness may require us to abandon this more narrow type of
coherence, in order to ensure that everyone has an appropriate level of input
into political decision-making?³³ I shall delay consideration of this issue until
Part III, Section C.

D. Recognizable Principles of Justice

We have seen that a notable feature of the checkerboard-abortion law is that its
application depends on a morally irrelevant consideration (namely, whether the
pregnant woman was born in an even-numbered year). We have also seen that this
alone is insufficient to distinguish checkerboard solutions from compromises that
Dworkin considers more acceptable, such as a law restricting abortions to women
who become pregnant as a result of being raped. However, even someone who
considers the rape issue to be irrelevant to the permissibility of abortion can
acknowledge that other people treat it as a morally relevant consideration. The
same is not true of the checkerboard-abortion law—no one considers whether the
pregnant woman was born in an even-numbered year to be morally relevant in the
context of the abortion debate.
This suggests another way of characterizing checkerboard solutions. If a
compromise upholds something that we can recognize as a moral principle
(whether we accept it ourselves or not), then it is not a checkerboard solution; it is
only a compromise (on a matter of principle) that does not uphold anything we
can recognize as a moral principle that we condemn as an internal compromise.

³³ J. Waldron, The Circumstances of Integrity, 3 Legal Theory 1, 4 (1997) points out that some-
one might think it is more expressive of real integrity for a community to acknowledge that it is torn
between competing views of justice than to strive to cover up this tension by preserving coherence in
principle within the law.
136 Dale Smith

(In this context, “moral principle” must mean a principle of justice, since
Dworkin acknowledges that checkerboard solutions can be supported by recog-
nizable principles of fairness.)
This suggestion can be illustrated by reference to the rape-abortion law. I
suggested in Section B that some people may regard the fact that a pregnancy is
the result of rape as morally irrelevant, in the sense that it does not provide even a
pro tanto reason for having an abortion. Nevertheless, such a person could regard
the rape-abortion law as upholding a principle that is recognizably moral in
nature. In other words, she could view her opponents as putting forward a
genuinely moral position (albeit an incorrect one). Moreover, she could distin-
guish on this basis between the rape-abortion law and the checkerboard-abortion
law, since the latter cannot be seen to uphold a principle that is recognizably moral
in nature. Whether or not she accepts that principle herself, she can recognize that
the rape-abortion law upholds a moral principle, whereas the same is not true of
an abortion law based on the year in which the pregnant woman is born. The
latter turns on a factor that she cannot consider to be a moral consideration even
in the extended sense in which she can recognize the fact that a pregnancy is the
result of rape to be a moral consideration whilst regarding it as not providing even
a pro tanto reason for having an abortion. This suggests that—whether or not we
agree that it is irrelevant that a pregnancy is the result of rape—we can, at least in
principle, distinguish checkerboard solutions from principles that may be incor-
rect or based on morally irrelevant considerations but that are still recognizably
moral in nature.
This reading of Dworkin is not without textual support. For example, he states
that some compromises do not violate integrity “because they reflect principles of
justice we recognize though we do not ourselves endorse them” (436, n.8). He
gives the example of a reduction in the list of capital crimes on the basis of views
about moral culpability (or some other standard generally respected in the crimi-
nal law), as opposed to allowing criminals convicted of a capital offence to escape
death by drawing straws.³⁴ Similarly, when introducing the ideal of integrity,
Dworkin states that we assume “that we can recognize other people’s acts as
expressing a conception of justice or fairness . . . even when we do not endorse
that conception ourselves” (166). We would need to make that assumption in this
context only if integrity forbids only those actions that do not express a
(correct or mistaken) conception of justice or fairness.
Of course, this notion of a standard that is not recognizable as a principle of jus-
tice (or of a position that cannot be regarded as a genuinely moral position) is
vague. One could try to give it more analytical rigour by suggesting that the
question is whether, as a conceptual matter, a consideration can be regarded as

³⁴ Viewed in isolation, this example could be regarded as supporting either the current approach
or the “irrelevant considerations” approach. However, we saw in Section B that the latter represents
an implausible interpretation of Dworkin’s discussion of checkerboard solutions.
 The Many Faces of Political Integrity 137

moral in nature.³⁵ The principle underlying the rape-abortion law falls within the
concept of morality, since it holds that the permissibility of abortion depends on a
consideration (that a pregnancy was the result of rape) that can be treated as
morally relevant by someone who is not conceptually mistaken (though she may
be morally mistaken). The same is not true of conclusions drawn from the fact
that a pregnant woman was born in an even-numbered year.³⁶ However, this can
advance matters only to the extent that we have a clear understanding of what
constraints are imposed by the concept of morality (or, in other words, of what
can count as a moral reason, consideration or principle). This is itself a controver-
sial issue.³⁷
Be that as it may, there are other problems with this interpretation of Dworkin’s
position. Why must political compromises be based on recognizable principles of
justice? We have seen that checkerboard solutions may be based on recognizable
principles of fairness (giving everyone an input into the content of the law that is
proportionate to the support they can muster within society), and why is this not
enough to make a compromise recognizably moral in nature? Why should we
privilege justice over fairness in the way that (this interpretation of ) Dworkin
requires?
One might seek to be conciliatory at this point, and suggest that whether a
political compromise reflects recognizable principles of justice is one relevant
consideration, though whether it reflects recognizable principles of fairness is
also relevant. However, Dworkin must show more than that the former is a
relevant consideration, since he claims that integrity (on this interpretation, the
requirement that laws reflect recognizable principles of justice) should “charac-
teristically” prevail over other values. He therefore needs to show that our pursuit

³⁵ On this view, for something to be a “recognizable principle of justice”, it is not enough that one
believes that it is a principle of justice. Instead, what counts is whether it in fact satisfies the conceptual
requirements for being a principle of justice.
³⁶ The difference between the “irrelevant considerations” approach and the present approach
could then be viewed as follows. The “irrelevant considerations” approach requires us to identify the
correct conception of justice, since it asks whether a factor is morally relevant. Determining whether a
factor is morally relevant requires us to determine what pro tanto moral reasons exist, which depends
upon which conception of justice is correct. The present approach, on the other hand, requires us to
consider only the concept of justice, since it asks whether a principle satisfies the conceptual con-
straints on something being a moral principle (or, more specifically, a principle of justice). It does not
matter whether that principle is valid under the correct conception of justice, provided it falls within
the concept of justice. See R. Dworkin, Taking Rights Seriously 134–135 (1977) regarding the
distinction Dworkin draws between concepts and conceptions.
³⁷ E.g., R. Hare, Sorting out Ethics (1997) claims that a strong universalizability condition is
part of the very concept of morality, whereas J. Raz, Value, Respect, and Attachment 58–59
(2001) suggests that any such condition is the result of substantive moral reasoning. This means
that—if the present approach is correct—a statute permitting abortion only for white women is a
checkerboard solution on the former view, but not on the latter. However, presumably both views
would regard the checkerboard-abortion law, along with a law that allowed abortions only for women
who are more than 160 cm tall, as checkerboard solutions, since substantive moral reasoning is not
needed to recognize that these factors are morally irrelevant (in the way that it is arguably needed to
recognize that racial characteristics are morally irrelevant).
138 Dale Smith

of fairness should, in most cases, be limited by a requirement that political deci-

sions reflect recognizable principles of justice.
Dworkin might respond that he addresses these issues when discussing the
justification of integrity. In discussing checkerboard solutions, he is concerned
only with whether integrity fits our political practices, and so he cannot be
expected to have justified that ideal at this stage. I shall consider his attempt to
justify the ideal of integrity in Part III. However, even allowing that Dworkin’s
discussion of checkerboard solutions is meant to show only that the ideal of
integrity fits our political practices, are those practices really best explained on the
basis that we believe that compromises must reflect recognizable principles of
justice? Is it really this that underlies our attitude towards checkerboard
solutions? If not, on this interpretation the ideal of integrity cannot satisfy even
the dimension of fit.
In fact, the present suggestion (that checkerboard solutions are unacceptable
because they do not reflect recognizable principles of justice) does not fit well with
the view (ascribed to us by Dworkin) that there is something distinctively wrong
with checkerboard solutions. Even if the present suggestion captures some of what
is wrong with checkerboard solutions, it does not expose all that is wrong with
such solutions and cannot explain why we consider them to be distinctively flawed.
To see this, imagine that a legislature passed a law requiring everybody to worship
rocks (by spending a certain amount of time each week praying in front of a rock,
etc). No recognizable principle of justice (I assume) could justify requiring people
to worship rocks. We can imagine principles that purport to justify such a require-
ment (for example “rocks are sacred and deserving of supplication”),³⁸ but such
principles would not be principles of justice, because they would not be concerned
with achieving a morally defensible outcome (in any recognizable sense of
“morally defensible”).³⁹ Yet this law does not seem to be a checkerboard solution.
It lacks one of the distinctive features of the checkerboard-abortion law, which
permits some people to do X but prohibits others from doing so. The rock-
worshipping statute applies to everyone, without exception, rather than treating
different people differently in a way that is unjust. If this is right, the fact that a law
is not based on a recognizable principle of justice cannot be a sufficient condition
for it to be a checkerboard solution, since the rock-worshipping law satisfies this
requirement and yet is not an internal compromise. (Indeed, some sort of differ-
ential treatment seems required by the very use of the word “checkerboard”, with
its connotation of dividing an area into different regions.)
Dworkin might respond that the rock-worshipping law is not a checkerboard
solution because it is not supported by considerations of fairness (it is not a

³⁸ This example was suggested by Scott Hershovitz.

³⁹ Dworkin claims that justice is concerned with achieving morally defensible outcomes. If one
instead views justice as concerned with the righting of wrongs and/or the distribution of resources,
it becomes even clearer that the principle that rocks are sacred and deserving of supplication cannot
be regarded as a principle of justice.
 The Many Faces of Political Integrity 139

compromise between different views designed to give each side a proportionate

share of their desired outcome). However, it is not clear that a law must be
supported by considerations of fairness in order to be a checkerboard solution.
The checkerboard-abortion law would presumably remain a checkerboard
statute even if it was unsupported by any consideration of fairness (for example,
because everyone agreed that abortion should always be permitted), but was
instead introduced on a whim. Yet, even if this is a requirement for a law to be a
checkerboard solution, we can imagine it being satisfied with regard to the rock-
worshipping law. Imagine that society is divided about the proper relationship
between Church and State. Some people desire a complete separation between
the two institutions, while others want Christianity entrenched as the state
religion, others desire the same of Islam, etc. The rock-worshipping law might be
a (bizarre) solution to this disagreement—it gives anti-separatists some of what
they want (by entrenching a religion), but also gives pro-separatists some of what
they want (by refusing to entrench any religion actually endorsed within the
community). Still, the rock-worshipping law differs in an important respect
from the checkerboard-abortion law, in that it treats everybody according to the
same principle, without exception.⁴⁰

E. Treating People Equally

We have just seen that another noteworthy feature of checkerboard solutions is
that they treat people differently in a way that cannot be justified by reference to a
correct principle of justice. That is, they confer a right or benefit on one group and
deny it to another group without being able to justify this differential treatment.
For example, the checkerboard-abortion law denies an abortion to some women
but permits it for others, and does not do so in a way that can be justified by
reference to a correct principle of justice. Similarly, when Dworkin states that his
“model of principle” (which is based on integrity) “commands that no one be left
out . . . that no one may be sacrificed, like wounded left on the battlefield, to the
crusade for justice overall” (213), the suggestion is that people should not be
treated unequally even when this results in more justice overall.⁴¹ Where the
choice is between a checkerboard solution and a completely unjust solution,
adopting the checkerboard solution may ensure that at least some people get
treated according to just principles, but will treat people unequally (by granting
benefits to some people but not to others, where this differential treatment cannot

⁴⁰ It also differs, in the same way, from Dworkin’s other example of a checkerboard statute men-
tioned in Part I. Legislation that imposes no-fault liability upon car, but not washing machine, manu-
facturers treats victims of defective products differently depending on the sort of product that injured
them.
⁴¹ Admittedly, there is some controversy about how this metaphor should be interpreted. For a
somewhat different interpretation, see Waldron, supra note 33, 19–21.
140 Dale Smith

be justified by any correct principle of justice) and so should be rejected for that
reason.⁴²
Further support for this interpretation can be gleaned from Dworkin’s state-
ment, already noted, that checkerboard solutions involve a single principle, which
is affirmed for one group and denied for another, and it is this which we denounce
(436, n.7). He later states explicitly that any defect in checkerboard solutions
“must lie in their distinctive feature, that they treat people differently when no
principle can justify the distinction” (180).
In addition to fitting well with much of what Dworkin says about checker-
board solutions, this interpretation helps explain why the checkerboard-abortion
statute is an internal compromise (because it treats people differently without
any basis in justice) but the rock-worshipping law discussed in Section D is not
(because it treats everyone the same). It may also help to explain why checker-
board solutions are undesirable despite being supported by considerations of
fairness. It could be argued that fairness alone cannot justify treating people
differently where this is unsupported by any correct principle of justice. On this
view, whatever the arguments in favour of giving people equal input into their
community’s law-making, this cannot justify the difference in treatment
inherent in checkerboard solutions. This line of thought is pursued in Part III.
However, the fact that checkerboard solutions uphold rights or benefits for
some people but not others is insufficient to distinguish them from many other
legal decisions. For example, the rape-abortion law permits victims of rape to have
an abortion but precludes other women from having one, and so treats these two
groups differently. This by itself is insufficient to render the rape-abortion law
problematic, let alone to turn it into a checkerboard solution. Even more clear-cut
would be a constitutional provision that confers a right to freedom of association
on everyone except for criminals who are imprisoned pursuant to constitutionally
valid laws. Such a provision clearly upholds a right for some people that it denies
to others, yet is not a checkerboard solution.
Perhaps the distinctive feature of checkerboard solutions resides in the fact that
the differential treatment they require is not justified by correct principles of
justice. However, this borders on saying that a checkerboard solution is a wrong
solution (at least as seen from the viewpoint of justice), since any law that
unjustly differentiated between people would be a checkerboard solution. This
would fail adequately to distinguish checkerboard solutions from any other
decision that produces an unjust outcome (including the rape-abortion law, if
whether a pregnant woman has been raped should not be determinative of whether
abortion is permissible). There is supposed to be something especially wrong
about checkerboard solutions, which distinguishes them from decisions that are

⁴² This has led some people to suggest that “integrity” is simply Dworkin’s name for a type of
equality—namely, the equality manifested by applying the same legal principles to X that one has
applied to Y: e.g. L. Alexander & K. Kress, Against Legal Principles, 82 Iowa L. Rev. 739, 755 (1997).
 The Many Faces of Political Integrity 141

wrong merely because they draw distinctions that are based on incorrect princi-
ples of justice.
One alternative⁴³ is to ask whether the differential treatment required by a law
could be justified by reference to anything recognizable as a principle of justice,
rather than asking whether it can be justified by reference to a correct principle of
justice. This would combine the present interpretation (with its focus on differen-
tial treatment) with the interpretation considered in Section D (which concen-
trated on whether the decision could be justified by reference to a recognizable
principle of justice). As it suggests that there is more than one thing wrong with
checkerboard solutions, I shall consider this possibility separately in the next
section.

F. A Multitude of Sins?
So far, I have proceeded on the assumption that there is only one thing wrong
with checkerboard solutions, and have sought to isolate that single flaw. Having
failed to do so, I must now consider whether the reason that we reject internal
compromises is because they possess several flaws. I shall pursue only what I con-
sider to be the most promising line of thought in this regard—that a checker-
board solution both treats people differently in a way that is unjust and cannot be
justified by reference to anything recognizable as a principle of justice.
The first advantage of this suggestion is that it recognizes that the fact that a law
unjustifiably upholds a right or benefit for some people but not others is a neces-
sary, but not sufficient, condition for something to be a checkerboard solution. As
a result, this interpretation provides a basis for distinguishing between the
checkerboard-abortion and rape-abortion laws. Both treat people differently, but
the latter does so on the basis of a recognizable (though perhaps incorrect) princi-
ple of justice whereas the former does not. Secondly, this suggestion enables us to
differentiate checkerboard solutions from laws (such as the rock-worshipping
statute discussed in Section D) that are not based on recognizable principles of
justice but do not treat people differently, since we have now built into our defini-
tion of checkerboard solutions a requirement of differential treatment.
Again, there is some textual support for this interpretation. Dworkin contrasts
the rape-abortion statute with the checkerboard-abortion statute in the following
terms:
You [i.e. an opponent of abortion] see the first of these statutes as a solution that gives
effect to two recognizable principles of justice, ordered in a certain way, even though you
reject one of the principles. You cannot treat the second that way; it simply affirms for
some people a principle it denies to others. (183, footnote omitted)
⁴³ Another alternative sometimes hinted at in Law’s Empire is that we should ask whether the
differential treatment can be justified by any principle of justice that is consistent with the principles
underlying other legal decisions. However, we saw in Section C that checkerboard solutions are
meant to be problematic in their own right, not only when considered alongside other legal rules.
142 Dale Smith

We see here both limbs of the current suggestion. While Dworkin does not
explicitly state that both are necessary, the quotation is meant to represent the
view of someone who shares Dworkin’s intuitions regarding checkerboard
solutions.
I believe this to be the most promising interpretation of Dworkin’s discussion
of checkerboard solutions. Not only does it bring together (in a coherent fashion)
many of Dworkin’s comments about checkerboard solutions, but it also suggests
that there is something distinctively wrong with internal compromises (something
that previous interpretations of Dworkin’s position were unable to do). Moreover,
it can explain our intuitive dislike of checkerboard solutions, since treating people
differently in a way that cannot be defended by anything recognizable as a princi-
ple of justice appears intuitively objectionable.

III. Justifying Integrity

After considering several interpretations of Dworkin’s views about checkerboard

solutions, I concluded that the best interpretation is that Dworkin claims that
there are two things wrong with internal compromise—it treats different people
differently in a way that is unjust, and it cannot be supported by reference to
anything recognizable as a principle of justice. Henceforth, I shall refer to this as
the “favoured interpretation.” While it is, I think, the best interpretation of
Dworkin’s position, it is not without its difficulties, and I shall consider two
of these difficulties in Section C. First, however, I want to explore whether, if
Dworkin has provided an acceptable account of checkerboard solutions, this
account can support the various claims that he bases upon it. Thus, in Sections A
and B, I shall assume that the favoured interpretation provides a satisfactory
account of checkerboard solutions, and shall consider what implications that
account has for the rest of Dworkin’s theory of law.

A. Should Integrity Characteristically Trump Justice?

In assuming that the favoured interpretation provides a satisfactory account of
checkerboard solutions, I am assuming that it explains why internal compromise
is undesirable. In other words, it explains why there is always at least a pro tanto
reason to reject a checkerboard solution. However, Dworkin does not claim
merely that there is always a pro tanto reason to avoid internal compromises. He
also claims that integrity (on the favoured interpretation, the requirement that
political decisions concerning matters of principle must not treat people differ-
ently in a way that cannot be supported by any recognizable principle of justice)
should characteristically trump justice in cases where these two values conflict. In
particular, in cases where justice would be promoted by adopting a checkerboard
 The Many Faces of Political Integrity 143

solution, the pro tanto reason to reject such solutions should characteristically
prevail when reaching an all-things-considered judgment about what to do. In
this section, I consider whether Dworkin succeeds in establishing the truth of this
further claim.
Dworkin’s discussion of checkerboard solutions is meant to show only that the
ideal of integrity fits with our political beliefs and practices (because it explains
why we oppose internal compromises). His argument that integrity is indeed
valuable (and should characteristically trump justice in cases of conflict) comes
later in Law’s Empire. Let us now consider that argument.
Dworkin claims that integrity has a number of benefits, but he concentrates
primarily on its implications for political obligation. His argument that
integrity is necessary for political obligations to arise is lengthy and subtle, and
I can provide only a thumbnail sketch here.⁴⁴ Dworkin claims that political
obligations are a type of “associative obligation.” In other words, they are not
voluntarily incurred, but rather arise simply through one’s membership of a
political community. However, the political community must satisfy certain
criteria if political obligations are to arise. In particular, the community must
display pervasive and equal concern for all its members. Dworkin claims that
this requirement can be satisfied only if our political obligations are not
exhausted by explicit rules. If our obligations were exhausted by explicit rules,
then our political system could not display pervasive concern, because those
rules would inevitably contain gaps where we would be left to our own devices.
(In particular, when deciding what new rules to introduce, we would be free to
act in our own self-interest, without concern for other people.) However, non-
explicit obligations can be inferred from the explicit rules only if those rules are
coherent.⁴⁵ Moreover, the coherence between the society’s rules must extend to
the principles underlying those rules, since the non-explicit obligations will be
drawn largely from those principles. Therefore, coherence in principle (which
Dworkin often equates with integrity) is a necessary condition of political
obligation.
In Section C, I shall question the extent to which this explanation of why
integrity is valuable provides support for the favoured interpretation. However,
for now it is important to note that there is at least one way in which they seem to
fit together nicely. The favoured interpretation claims that there are two principles
underlying our opposition to checkerboard solutions—the principle that we
should not treat people differently in a way that is unjust, and the principle that
we should not treat people in a way that cannot be supported by any recognizable
principle of justice. Dworkin’s account of political obligations appears to support

⁴⁴ I draw particularly on Alexander & Kress, supra note 42, at 778 in offering this reconstruction
of an argument that occupies Dworkin for nearly 30 pages in Law’s Empire (195–224). I focus only
on those aspects of the argument that are relevant to my present concerns.
⁴⁵ More than mere consistency is required. E.g., one could not infer (a pervasive set of ) non-
explicit obligations from a set of three legal rules that bore no relationship to each other.
144 Dale Smith

the former principle. A society that does not respect this principle does not display
equal concern for all its members and so does not merit everyone’s allegiance.
Even if there are principles of fairness that support the differential treatment,
those who lose out under the checkerboard solution are being sacrificed to give
people an equal share of political influence, and it might be difficult to explain to
them why they should give their allegiance to such a society.
This, in turn, suggests that Dworkin is correct in claiming that integrity should
characteristically trump justice when these two values conflict. For political obli-
gations to arise, it is not necessary that political decisions always be perfectly just,
but it is necessary that such decisions display equal concern for all members of the
community, and this would appear to rule out checkerboard solutions (which
necessarily treat people differently in a way that is unsupported by any recogniz-
able principle of justice, and so fail to show equal concern for all members of the
community). Moreover, if we allowed justice to trump integrity on a regular basis,
we would no longer be able to infer non-explicit principles from the explicit rules
of our community (since those rules would lack the necessary coherence in princi-
ple). This would mean that not only would the law have ceased to show equal
concern, but it would also have ceased to show pervasive concern.
There is, however, a well-known objection to this line of reasoning. Dworkin’s
critics often argue that, where a just outcome cannot be obtained for everyone, it
is better to obtain that outcome for as many people as possible, even if this reduces
the coherence of the law and does not provide equal treatment for those who are
left out. For example, Raz suggests that “we accept the nearest approximation to
morally sound solutions that we can obtain, even though by doing so we may
reduce the coherence of the law.”⁴⁶ If this is right, we should be more willing to
accept checkerboard solutions (and other breaches of integrity) than Dworkin
suggests, where this is necessary to reduce injustice. We might still recognize a pro
tanto reason to avoid checkerboard solutions, but we would not regard that reason
as characteristically trumping considerations of justice.
However, this objection does not really engage with Dworkin’s claim that we
must not allow justice characteristically to trump integrity, or else our community
would cease to express equal and pervasive concern for all its members, and our
case for the existence of political obligations would collapse. To determine which
side in this debate is correct, it appears that we must weigh up the value of doing
justice to as many people as possible against the value of equal and pervasive con-
cern. However, rather than engaging in the difficult task of weighing up those val-
ues, I shall explore a different line of thought. I shall argue that the value of equal
and pervasive concern does not unequivocally support Dworkin’s position, and

⁴⁶ Raz, supra note 4, at 312–313. The context makes it clear that Raz endorses the view he is
ascribing to us. Similarly, Wasserstrom, supra note 17, at 267 asserts that, if he were an egalitarian in a
libertarian society, he would support legislation that made the community’s law more egalitarian,
even if this meant that the law was less coherent overall.
 The Many Faces of Political Integrity 145

that he has therefore failed to establish that integrity should characteristically

trump justice when these two values conflict.
While checkerboard solutions treat people differently in a way that we have a pro
tanto reason to avoid, so may a refusal to adopt checkerboard solutions. Refraining
from adopting checkerboard solutions may disadvantage those people whose moral
convictions are not dominant within the community, because their moral position
may not be reflected in the law at all.⁴⁷ Dworkin seems committed to regarding this
as a form of disadvantage, since he claims that the value of fairness supports
checkerboard solutions because they ensure that one’s moral position is reflected in
the law to the extent that one has support within the electorate.⁴⁸ This suggests
that, whatever one does, one may end up treating people differently in a way that is
undesirable. If one upholds a checkerboard solution, one will affirm a right or
benefit for some people that one denies (without justification) for others. However,
if one refuses to adopt checkerboard solutions, one may give weight to some
people’s political views but not others’ (for example, if law-makers always side with
the majority within society, in order to preserve coherence in principle).⁴⁹ If we
cannot avoid undesirable forms of differential treatment no matter what we do,
should we not embrace the form of differential treatment that minimizes instances
of injustice (which, we have seen, would involve adopting checkerboard solutions
in circumstances in which Dworkin would oppose their adoption)?⁵⁰
(Admittedly, checkerboard solutions not only treat people differently, but do so
by reference to something that is not recognizable as a principle of justice.
However, if this is the basis on which Dworkin believes that integrity should
characteristically prevail over justice, he must explain why a political compromise
is acceptable only if it reflects recognizable principles of justice. I consider this
issue in Section C.)
I am not claiming that integrity is unimportant. Rather, my point is that
integrity safeguards only one aspect of equal treatment, and so—while always a
relevant consideration—it may not be as important as Dworkin claims. We have
seen that checkerboard solutions can promote equal treatment as well as under-
mining it. I shall now argue that the sort of community that Dworkin believes is
necessary to give rise to political obligations has an equally equivocal relationship
to the requirement of equal treatment.

⁴⁷ Wasserstrom, supra note 17, at 267–268.

⁴⁸ If one has a different conception of fairness from Dworkin’s, the failure of the law to reflect my
considered views may not be unfair (especially if I was given an opportunity to participate in the
process by which the law was chosen and/or if my views are mistaken). However, the point made in
the text poses a problem for Dworkin, given his conception of fairness.
⁴⁹ The majority on one issue may be constituted differently from the majority on another issue,
but there may be some people who are rarely (if ever) in the majority.
⁵⁰ Alternatively, one could embrace whichever form of differential treatment minimizes unfair-
ness, but this would involve accepting checkerboard solutions whenever they produced the fairest dis-
tribution of political influence.
146 Dale Smith

To see this, consider a criticism of Dworkin offered by Alexander & Kress.⁵¹

They claim that a reduction in integrity may actually increase the extent to which
the law displays equal concern for all members of the community. They ask us to
imagine a society whose employment laws encourage nepotism. A statute is then
passed banning nepotism in private industry (but not in the public sector, because
there is insufficient political support for this change).⁵² This is a checkerboard
solution, because nepotism is still encouraged in the public sector, but is no longer
permitted in the private sector, and (we are assuming) there is no recognizable
principle of justice that could support the differential treatment involved in
permitting nepotism in the public sector but not the private sector. This statute
has reduced the integrity of the law, but Alexander & Kress suggest that it has
increased the extent to which the law displays equal concern for members of the
community, by reducing nepotism (which, by its very nature, is inconsistent with
equal concern).
They anticipate the objection that this checkerboard solution has actually
made things worse, because there is no longer a coherent set of principles that can
be inferred from the existing rules, depriving us of any basis on which to decide
novel cases (for example, whether nepotism is permitted by trustees with a discre-
tionary power to distribute trust funds). However, they respond, those who
approve of the new statute can at least say that it is based on a new and improved
conception of the concern that we owe to each other, a conception with implica-
tions for novel cases.⁵³ This is true, but that conception will not be pervasive—it
will be recognized with regard to private industry employment, but not with
regard to the public sector. This weakens the case for political obligation, because
society is divided into two camps (those who are treated according to correct
principles of justice and those who are not) in a way that cannot be justified by
reference to any recognizable principle of justice.
This suggests a more complicated picture than either Dworkin or Alexander &
Kress acknowledge. In one way, the checkerboard solution reduces the extent to
which people are treated differently, by eliminating nepotism in private industry
(and so ensuring that all applicants for a job in private industry are treated
equally). In another way, the checkerboard solution exacerbates the preferential
treatment that exists within society, by treating job applicants in private industry
differently from job applicants in the public sector (without there being any rec-
ognizable principle of justice that could justify this differential treatment). No
doubt, upholding integrity will sometimes result in an overall improvement in
terms of reducing (unjustifiable) differential treatment, and so improve the case

⁵¹ Alexander & Kress, supra note 42, at 779–780. I have changed their example slightly, to avoid a
complication they themselves note.
⁵² If it strikes one as implausible that such a law could ever be passed, imagine that there were a
number of high-profile corporate collapses blamed on the practice of nepotism in private industry,
whereas powerful lobby groups supported nepotism in the public sector.
⁵³ Alexander & Kress, supra note 42, at 780.
 The Many Faces of Political Integrity 147

for political obligation. However, Dworkin must show that this is characteristically
the case. Presumably, this will depend on a number of contingent facts about the
particular legal system being considered, and Dworkin does not even begin to
show that it is true of the particular legal systems he has in mind.⁵⁴
If I am right, Dworkin has failed to show that the value of integrity deserves
the place of special prominence that he reserves for it. In particular, he has not
given us sufficient reason to believe that integrity should characteristically prevail
over justice when these two values conflict. This has several significant implica-
tions for his broader legal and political theory, of which I shall briefly mention
two. First, it threatens to undermine his directions to both legislators and judges.
Dworkin claims that legislators should, when making new law, strive to keep the
law coherent in principle. He also claims that judges should, when identifying
and enforcing the law, seek to view the law as being coherent in principle (167).
However, these directions presuppose the pre-eminence of integrity. Coherence
in principle should be the primary goal of legislators and judges only if integrity
should characteristically trump justice. The fact that Dworkin has failed to estab-
lish that integrity should characteristically prevail means that the correctness of
his direction to legislators and judges is called into question.
Secondly, Dworkin’s failure to show that integrity should characteristically
prevail over justice means that his attempt to establish the existence of political
obligations is undermined. Dworkin argues that integrity must be valuable
because it is necessary for political obligations to arise. However, we can reverse
this line of reasoning and argue that—if integrity is not as valuable as Dworkin
claims—he has failed to demonstrate the existence of political obligations.
Unless integrity should characteristically prevail over justice, we lack sufficient
reason to achieve the level of coherence among (and within) explicit legal rules
that is required for us to be able to infer non-explicit legal principles to fill in gaps
in those rules. Yet, on Dworkin’s argument, such coherence is required in order
for political obligations to arise. Therefore, Dworkin’s failure to show that
integrity should characteristically prevail over justice means that he has failed to
show that we should achieve the level of coherence among (and within) explicit
legal rules that he claims is necessary to give rise to political obligations.

B. How Well Does Integrity Fit Our Political Practices?

Apart from raising concerns about whether Dworkin is right in claiming that
integrity should characteristically prevail over other values, my discussion of
what is wrong with checkerboard solutions also reveals another problem for

⁵⁴ Earlier in this section, we saw another way in which checkerboard solutions promote equal
concern—they give everybody some input into the content of political decisions. This, too, would
have to be weighed against the way in which checkerboard solutions undermine equal concern.
148 Dale Smith

Dworkin’s broader theory. That discussion has highlighted that there is not one,
but several, principles of integrity. There are at least two factors underlying our
attitude towards checkerboard solutions—a concern about differential treat-
ment, and a concern that legal decisions should reflect recognizable principles of
justice. If Dworkin is correct in claiming that it is integrity that explains our
attitude towards internal compromises, then both of these concerns must reflect
aspects of integrity. This in itself poses no problem for Dworkin. There are
numerous (correct) principles of justice, and there is no reason why there cannot
be several (correct) principles of integrity as well. However, in addition to the
two principles that underlie our attitude towards checkerboard solutions, there is
at least one more principle of integrity that has little to do with checkerboard
solutions. This is the principle that requires coherence between the legal princi-
ples that underlie and justify different legal rules, so that they form a single,
comprehensive vision of justice. We saw that Dworkin treats this principle as
part of the ideal of integrity, and that he recognizes that violation of this principle
poses a threat to integrity that is distinct from the threat posed by checkerboard
solutions.⁵⁵ Indeed, Kress suggests that it is this principle that Dworkin most
often refers to when discussing the ideal of integrity.⁵⁶ The problem is that, in
seeking to show that the ideal of integrity fits with our political practices,
Dworkin relies almost exclusively upon his discussion of checkerboard solutions,
and this principle of integrity is not relevant to that discussion. Therefore, even if
(via his discussion of checkerboard solutions) he has established that the first two
principles of integrity fit with our political practices, he has not shown that this
third principle of integrity does so.
Dworkin does suggest that, while our legal system often breaches this prin-
ciple, we regard this as a defect (184). However, more needs to be said to show
that a central principle of integrity satisfies one of only two conditions that
Dworkin imposes upon its acceptability (namely, the dimension of fit),⁵⁷ espe-
cially as he concedes that the evidence on this point is equivocal (since our legal
system often breaches this principle). This means that Dworkin’s defence of
integrity is incomplete. He may have shown that some principles of integrity fit
our political practices, but he has not shown that all do. Nor can he assume that,
because some principles of integrity satisfy that requirement, the rest must, too.
It is unlikely that every correct principle of justice fits our political practices, and
there is no reason to assume at the outset that integrity differs in this respect.

⁵⁵ Part II, Section C.

⁵⁶ K. Kress, Coherence and Formalism, 16 Harv. J.L. & Pub. Pol’y 639, 653 (1993). It is also this
principle that underlies Dworkin’s discussion of the examples at (165).
⁵⁷ Even if we regard fit as one aspect of justification (which seems to be Dworkin’s preferred posi-
tion: M. Greenberg, How Facts Make Law, 10 Legal Theory 157, 196–197, n.47 (2004)), the failure
to show that the third principle of integrity fits our political practices represents a failure to show that it
satisfies an important aspect of justification. Dworkin needs either to rectify this failure or else to show
that this principle of integrity satisfies other aspects of justification which make up for any lack of fit.
 The Many Faces of Political Integrity 149

This would not be such a problem for Dworkin if this particular principle of
integrity did not play such an important role in his jurisprudential theory. Once
we expand our focus beyond checkerboard solutions, Dworkin’s legal theory is at
least as much concerned with coherence in principle between laws as it is with
coherence in principle within laws. The fact that Dworkin fails adequately to
explain why we should accept this third principle of integrity should therefore be
of substantial concern to him.
Dworkin might seek to avoid this problem by arguing that there is in fact only
one principle of integrity (albeit one that can be violated in different ways—for
example, via checkerboard solutions or via a lack of coherence between laws). He
does claim that “the only basis we might have for opposing checkerboard
compromises . . . is the idea of integrity, that the community must respect
principles necessary to justify one part of the law in other parts as well” (210).
The requirement that the community respect principles necessary to justify one
part of the law in other parts as well provides the basis of Dworkin’s discussion of
coherence in principle between laws, and he appears to be suggesting here that it
also provides the basis for opposing checkerboard solutions. However, this
suggestion is puzzling. When seeking to achieve integrity between laws, we are
seeking to ensure that different laws cohere with each other (and that the prin-
ciples underlying those laws also cohere with each other). When seeking to erad-
icate checkerboard solutions, on the other hand, we are seeking to eliminate
individual laws that treat different people differently in a way that cannot be sup-
ported by any recognizable principle of justice. That there are two distinct sets of
principles in play here (one requiring coherence between legal rules and princi-
ples, the other requiring that an individual law not discriminate between people
in a way that cannot be supported by any recognizable principle of justice) seems
obvious. It is further supported by the argument in Part II, Section C that we
cannot explain what is wrong with checkerboard solutions on the basis that they
both uphold and reject certain principles of justice (something that may be said
of a body of laws that contains inconsistent legal rules or principles).

C. How Plausible is the Favoured Interpretation?

At the start of this Part, I suggested that—despite being the best interpretation of
Dworkin’s discussion on offer—the favoured interpretation is not unproblematic.
In this section, I shall briefly discuss two difficulties with that interpretation. First,
the problem of distinguishing between something that is recognizable as a princi-
ple of justice and something that is not remains. We saw in Part II, Section D that
it is not clear what counts as a “recognizable principle of justice.” I offered a
suggestion as to how this phrase could be rendered more precise (in terms of
which considerations or reasons satisfy the conceptual constraints on being a
moral consideration or reason), but I also pointed out that what these conceptual
150 Dale Smith

constraints are is disputed. To give more content to the notion of a “recognizable

principle of justice,” Dworkin would need to take sides in these disputes.
However, in the rest of this section I wish to focus on the second problem with
the favoured interpretation. Just as Dworkin’s argument that integrity fits with
our political practices does not apply to the principle of integrity requiring coher-
ence between legal rules and principles (see Section B), his argument that integrity
is valuable does not vindicate the principles of integrity that explain what is
distinctively wrong with checkerboard solutions. The argument that integrity is
valuable because it is necessary to give rise to political obligations may show that
coherence between legal rules and principles is important (since this enables us to
infer further legal principles, and hence to display pervasive concern for members
of the community). However, it does not show that it is important to avoid differ-
ential treatment that is unsupported by any recognizable principle of justice.
Admittedly, one of the two principles of integrity violated by checkerboard
solutions seems to fit well with Dworkin’s argument for political obligations. As
mentioned in Section A, treating people differently without adequate justification
appears to violate the requirement that the law display equal concern for all
members of the community. (Though it could be argued that one may display
equal concern even when unjustly treating people differently, if one is sincerely
trying to do the right thing by everyone.) However, it is not clear how Dworkin’s
discussion of political obligations is meant to show that there is something
distinctively wrong with differential treatment that is not supported by recogniz-
able principles of justice. In other words, it is not clear how Dworkin’s attempt to
show that integrity is valuable supports the principle of integrity requiring that
laws reflect recognizable principles of justice.
Moreover, it is not obvious how one could go about supporting that principle.
It is obvious why we should want our political decisions to be supported by correct
principles of justice, but checkerboard solutions are not alone in failing to satisfy
this requirement. The distinctive flaw of internal compromises is that the differen-
tial treatment that they require is not supported by recognizable principles of
justice, but why is it so important that our political decisions be supported by
such principles?
Dworkin might respond that our attitude towards checkerboard solutions
shows that we accept that it is important that our political decisions be supported
by recognizable principles of justice. If he is right in claiming that we oppose
checkerboard solutions, and that the distinctive feature of such solutions is that
they treat people differently in a way that cannot be supported by recognizable
principles of justice, does this not show that we oppose political decisions that lack
such support? The answer to this question might well be yes, but this does not
show that we are right to oppose political decisions of this type (or that we are right
to regard them as distinctively flawed)? To be confident that we are right to do so,
we need an argument that explains why political decisions must be supported by
recognizable principles of justice.
 The Many Faces of Political Integrity 151

It may be that we cannot display equal concern for all members of our commu-
nity if we adopt principles that are not recognizable as expressing equal concern. In
a society marked by moral disagreement, we cannot necessarily expect society to
adopt our conception of equal concern, but we can at least expect it to adopt a
recognizable conception. So long as the community adopts a recognizable concep-
tion of equal concern, we can feel that we are part of a genuine community, even if
we would prefer that it adopt a different conception.⁵⁸ However, it is not the case
that only recognizable principles of justice reflect a recognizable conception of
equal concern. In particular, checkerboard solutions are (at least in some cases)
based on principles of fairness that are recognizable as expressing equal concern
(since they are designed to produce an equal distribution of political influence). It
is unclear why this is not enough—why, that is, we need principles of justice that
we can recognize as expressing equal concern. Provided each side champions
recognizable principles of justice,⁵⁹ why should it matter if a recognizable princi-
ple of fairness requires us to compromise those principles of justice in a way that
does not reflect any further, recognizable principle of justice?
It could be argued that there is some sort of category error in resolving a matter
of principle (dealing with rights or other moral issues) by reference to something
that is not recognizable as a principle of justice. Dworkin defines justice as being
concerned with achieving morally defensible outcomes, and so it is principles of
justice that must be used to resolve matters of principle. Using something that is
not recognizable as a principle of justice reveals a misunderstanding as to the type
of solution that is required for such a problem and/or a conceptual misunder-
standing of what counts as a principle of justice.
However, this explanation is of little assistance to Dworkin. The requirement
that laws should reflect recognizable principles of justice is meant to be a princi-
ple of integrity, where integrity is meant to be a political value distinct from
justice and fairness. Avoiding category errors or conceptual misunderstandings,
on the other hand, would not seem to amount to a political value at all. (The
other principle of integrity relevant to checkerboard solutions, on the other
hand, does represent a political value, since it prohibits unjust differential treat-
ment.) Moreover, even if the importance of avoiding category errors or concep-
tual misunderstandings does reflect a distinct political value of integrity, it is not
clear that such a value should characteristically trump competing considerations
of justice or fairness. Obviously, it is desirable to avoid category errors and
conceptual misunderstandings, but it is not obvious that this is characteristically
more important than avoiding or minimizing instances of injustice or unfairness.
Therefore, this explanation does not seem well suited to supporting Dworkin’s

⁵⁸ A similar argument is suggested by J. Waldron, The Rule of Law in Contemporary Liberal Theory,
2 Ratio Juris 79, 83–84 (1989).
⁵⁹ This proviso is necessary so that we can know that they are putting forward views about justice
(since principles of fairness are meant to govern disagreements about justice).
152 Dale Smith

claim that checkerboard solutions are characteristically undesirable all things

considered.
To summarize, while the favoured interpretation provides the best account
available to Dworkin of what is wrong with checkerboard solutions, it leaves at
least two questions unanswered. First, it is not clear what counts as a “recognizable
principle of justice.” Secondly, it is somewhat mysterious why we should put so
much weight on the requirement that differential treatment be supported by a
recognizable principle of justice. Until these questions can be answered, there will
continue to be nagging doubts that—while there may well be reasons to oppose
checkerboard solutions—we cannot elucidate those reasons in the way that
Dworkin seeks to, by reference to the value of integrity.⁶⁰

IV. Conclusion

In his discussion of checkerboard solutions, Dworkin identifies (what at least

appears to be) a distinctive phenomenon, a phenomenon of which we intuitively
disapprove. This chapter has sought to provide an account of what is wrong with
checkerboard solutions that is both persuasive in its own right and is consistent
with Dworkin’s discussion. On the most promising such account, there are two
things wrong with checkerboard solutions—they treat people differently in a way
that is unjust, and they cannot be supported by reference to any recognizable prin-
ciple of justice. However, we have also seen that—even on this interpretation—
problems remain for anyone seeking to defend Dworkin’s broader claims about
integrity, the law and political obligations. Dworkin’s discussion of checkerboard
solutions may establish that there is a distinctive political virtue (integrity) that
provides a pro tanto reason to oppose internal compromise, but it does not estab-
lish that this pro tanto reason should characteristically prevail over countervailing
reasons provided by the values of justice or fairness. In other words, Dworkin may
have succeeded in showing that checkerboard solutions are always undesirable,
but he has not ruled out the possibility that often they may be the best we can
hope for. He therefore has not established that integrity merits the pre-eminent
position that he ascribes to it. This, in turn, calls into question his directions to
legislators and judges and his account of political obligations. Secondly, his
discussion of checkerboard solutions may show that some principles of integrity
fit with our political practices and beliefs, but it does not show that all principles
of integrity do so. In particular, his discussion of checkerboard solutions does not
show that the requirement of coherence in principle between laws satisfies the
dimension of fit.

⁶⁰ See supra note 17, which points out that a number of commentators have tried to explain our
opposition to checkerboard solutions by reference to the values of justice or fairness, not integrity.
 The Many Faces of Political Integrity 153

As a result, Dworkin is faced with a dilemma. The favoured interpretation of

his discussion of checkerboard solutions does not provide the support he needs for
some of his further claims about legal and political theory. Therefore, he must
either abandon those further claims or find a new way of arguing for them (pos-
sibly by suggesting an alternative to the various interpretations of his discussion of
checkerboard solutions suggested in this chapter). He cannot both embrace the
favoured interpretation of his discussion of checkerboard solutions and treat that
discussion as supporting his further claims about law and politics.
On the other hand, someone who does not share Dworkin’s further commit-
ments concerning legal and political theory can accept the favoured interpretation
without worrying about the above concerns. On this view, checkerboard solutions
are always flawed, but in a range of cases they may be the best we can hope for. It is
only someone who shares Dworkin’s further commitments who must go further
and argue that checkerboard solutions are, characteristically, undesirable all things
considered (and that there are further principles of integrity that also fit with our
political practices). It is only these claims that are unsupported by the favoured
interpretation.
Nevertheless, there are two independent concerns about whether, even on the
favoured interpretation, Dworkin has provided a satisfactory account of what is
wrong with checkerboard solutions. The favoured interpretation makes use of the
notion of a “recognizable principle of justice,” and it is not entirely clear how to
give content to this idea. Moreover, it is not clear how treating people differently
in a way that is unjust is made even worse when the differential treatment cannot
be supported by a recognizable principle of justice (though it can be supported by
a recognizable principle of fairness).
 7
Did Dworkin Ever Answer the Crits?
Jeremy Waldron

I.
I am not now, nor ever was, a member of the Critical Legal Studies (CLS)
movement. But I want to consider the adequacy of the answer Ronald Dworkin
gave almost twenty years ago to an objection (to his conception of law as
integrity), which he attributed to CLS though he himself referred to it in more
general terms as “internal skepticism.”¹ I want to consider the adequacy of his
response, not out of any desire to revive the intellectual fortunes of CLS and cer-
tainly not as a complaint that Professor Dworkin failed to do justice to CLS criti-
cisms during the brief lifetime of the movement. CLS was never particularly
interested in Dworkin, nor he in them.² But they had a common interest in what
I would like to call “the background elements” of a legal system—the principles
and policies that lie in back of the rules and texts that positivists emphasize.
Ronald Dworkin and Duncan Kennedy, for example, are both theorists of the
legal background. But what they say about it is very different. Dworkin thinks
recourse to the background affords the resources necessary for legal decision in
cases where the foreground law is disputed or indeterminate. Kennedy thinks the
background is so riven with contradiction as to be capable of offering spurious
support for everything and determinate support for nothing in legal reasoning.³
So I am interested in what one might think of as Dworkin’s descriptive
optimism—his view that the background elements (of a legal system like that of
the United States) are capable of bearing the weight of determinate argument
that he wants to assign to it in his theory of law.

¹ R. Dworkin, Law’s Empire 266–275 (1986).

² See A. Altman, Legal Realism, Critical Legal Studies and Dworkin, 15 Philosophy and Public
Affairs 205, 214–215 (1986) (“For the most part, proponents of CLS and Dworkinians have
ignored one another’s positions”).
³ See D. Kennedy, Form and Substance in Private Law Adjudication, 89 Harv. L. Rev. 1685
(1976).

Did Dworkin Ever Answer the Crits? Jeremy Waldron.

© Oxford University Press 2006. Published 2006 by Oxford University Press.
156 Jeremy Waldron

I will argue that, in order to answer the CLS critique, a defender of law as
integrity has to place considerable emphasis on what Professor Dworkin regards as
the constructive side of his argument. Instead of saying that the legal background is
coherent, Dworkin has to say that it is capable of being made coherent at the
hands of a sufficiently resourceful interpreter. Unfortunately the response cannot
rest there. Dworkin’s constructivism might licence an ingenious and versatile
manipulation of existing legal materials: we take the parts of the (often contradic-
tory background) that we like or approve, and we use them to add some sort of
doctrinal credibility to the positions we are aiming at. But it is not clear that this
fits the justification that Dworkin furnished for his method in Law’s Empire, for
that justification requires a party to argue, not merely that there are legal materials
which he can make use of in support of his brief, but that he is using background
materials in a way that keeps faith with the network of mutual commitments that
makes us the community we are.⁴ Unless the Dworkinian advocate can make such
an argument, he has nothing with which he can resist the urgings of a pragmatist,
to the effect that we should not bother with the background material at all, once
the foreground has been shown to be indeterminate, but we should make our case
directly for whatever is likely to promote the social good. To rebut that pragmatist
position—which Dworkin acknowledges has to be rebutted to make room for his
conception of law as integrity⁵—it is not enough to show that we can make some-
thing coherent and attractive out of the legal background. The Dworkinian advo-
cate has to show that we must attempt to make something coherent and attractive
out of the legal background. And what counts as a good faith attempt has to be
responsive to the ideas lying behind that “must.” My worry is that the construc-
tivism that Dworkin appeals to, in order to evade the CLS critique, succeeds (if it
does) only at the cost of cutting loose of the idea that the community is in some
sense already committed to a coherent and principled position which the
Dworkinian lawyer has a responsibility to unearth.
So: my aim in this paper is to explore the extent to which Professor Dworkin is
put to a hard choice between the agile and discerning constructivism he needs to
respond to CLS, on the one hand, and the integrity thesis about the commitments
of the community which he invokes to justify the claim that making coherent
sense of the existing legal materials, foreground and background, is something we
are morally required to do. I think Dworkin really is confronted with a dilemma
here, and I am not sure that a way through it can be negotiated. If he hangs on to
the integrity position, he makes it harder to respond to the skepticism of CLS. But
if he weakens the integrity requirement or loosens its connection to what is to
count as an appropriate mode of legal argument, then he leaves himself defenseless
against the pragmatist position that—clever though it is—the constructivism of
Dworkinian legal argument serves no useful purpose.

⁴ Dworkin, Law’s Empire, supra note 1, at 195–216.

⁵ Id. at 162–164.
 Did Dworkin Ever Answer the Crits? 157

As I said in my opening paragraph, I am not exploring Dworkin’s response to

internal skepticism in order to revive the moribund fortunes of CLS. The Crits
can look after themselves.⁶ I am exploring their critique and his response because
they highlight a connection between two different aspects of Dworkin’s jurispru-
dence, usually dealt with and criticized—if they are criticized—separately. On the
one hand, there is the question of whether his constructive or interpretive method
actually delivers the goods for judicial reasoning. On the other hand, there is a
philosophical question about the ultimate justification for his insistence that we
delve into the legal background rather than give up and resort to an overtly prag-
matic approach when foreground law runs out or reveals itself to be indeterminate.
The theory of integrity is supposed to give us our answer to the second question:
we must persist with the background in order to keep faith with the commitments
of our community. But what if the method for delivering the goods in hard
cases—creative constructive theory-building—turns out not to be a way of keep-
ing faith with our community? What if it makes a mockery of the idea of com-
munal commitments, which law as integrity bows to? Then the two parts of
Dworkin’s jurisprudence—the clever constructivism and the portentous theory of
integrity—sail past each other, and it will be hard to hang on to a sense of the
whole. We need to take this prospect very seriously; and if there is an adequate
Dworkinian response to the CLS challenge we must ensure that it does not fall at
this hurdle.

II.

The key to Dworkin’s jurisprudence is and has always been a view about what law
involves besides a heap of enacted rules (constitutions, codes, and statutes).⁷
Dworkin argues that as well as the enacted rules, which we see in the foreground
of legal analysis, a working legal system also comprises background policies and
principles, either those associated with particular enactments—as when we say
the policy of the Sherman Act is to foster competition—or those that pervade the
body of the law as a whole—as when we talk about “the well established policy of
English law of imposing a more extensive liability on intentional wrongdoers than
on careless defendants,”⁸ or as when we invoke the maxim that “[n]o one shall be
permitted to profit by his own fraud.”⁹ Dworkin believes that it is appropriate to

⁶ I must say, though, that my first-year law students notice at once the connection between
Kennedy’s work in CLS and Dworkin’s jurisprudence. And they do ask me what, if anything, has
been written to refute the implicit CLS critique. (They are too young to know that analytic legal
philosophers are supposed to disdain and ignore CLS ideas.)
⁷ The Model of Rules I, in R. Dworkin, Taking Rights Seriously (revised ed., 1977).
⁸ Three Rivers District Council and Others v. Governor and Company of the Bank of England (No 3)
[2003] 2 A.C. 1 (HL), at 162.
⁹ Box v. Lanier, 112 Tenn. 393, 409, 79 S.W. 1042, 1045 (1903).
158 Jeremy Waldron

invoke these principles and policies in legal argument and indeed that it is inap-
propriate not to. Not only are they important legal resources to be put to use in
hard cases where the foreground law is indeterminate, but they are also capable of
standing against foreground law (as the principle last-mentioned above stood
against the New York state legislation relating to wills in Riggs v. Palmer).¹⁰ Indeed
a case may be made that background principles and policies have to be assumed in
order to explain why the foreground law has the force it does.¹¹ Certainly any
jurisprudence that neglects these background elements is inadequate.
There is a lot in Professor Dworkin’s account of the legal background that I
do not want to dispute. In various places, he distinguishes between policies and
principles, as different types of background element: he associates “policy” with
social goals and “principle” with norms or reasons that command a particular
distribution of benefits, freedoms, and responsibilities.¹² I have no quarrel with
that, nor with the arguments that Dworkin rests on that distinction. I accept
Dworkin’s view that background principles (in the narrower sense) are capable
of sustaining claims of right.¹³ I also see little to quarrel with in what he says
about the logical character of the contribution that principles and policies make
to legal argument: he talks of the dimension of “weight,” which distinguishes
their contribution from the contribution made by rules.¹⁴ Finally, I have no
quarrel with what Professor Dworkin says about the kind of presence that prin-
ciples and policies have in a legal system. He argues—convincingly in my
view—that their status as law “lies not in a particular decision of some legisla-
ture or court, but in a sense of appropriateness developed in the profession and
the public over time.”¹⁵ (I think, too, that he is right to point up the difficulty
this poses for H.L.A. Hart’s concept of the rule of recognition.) None of these
matters is in dispute between Dworkin and the Crits. In fact, although CLS is
certainly less optimistic than Professor Dworkin is about the contribution that
these background elements can make to determinate legal reasoning, their criti-
cisms nevertheless presuppose that the legal background is at least as important
for jurisprudence as Dworkin claims it is. If anything they are less sympathetic

¹⁰ Riggs v. Palmer, 22 N.E. 188 (N.Y. 1889).

¹¹ See Dworkin, The Model of Rules I, supra note 7, at 37.
¹² See id. at 22. See also Dworkin, Hard Cases in Taking Rights Seriously, supra note 7, at 82 ff.
and Dworkin, Law’s Empire, supra note 1, at 221–224.
Dworkin notes that “principle” can be used in an inclusive sense to include both principles and
policies in these narrower meanings (The Model of Rules I, op cit., at 22), and it is worth noting that
there is also a similarly inclusive sense of “policy,” at least in judicial discourse, as when it is said, for
example, that it is “the general policy of the law that it is preferable that a successful defendant should
suffer the injustice of irrecoverable costs than that a claimant with a genuine claim should be pre-
vented from pursuing it” (Hamilton v. Al Fayed (No. 2) [2003] Q.B. 1175 (CA) 1178) or that “[t]he
policy of the law is to prevent A being judge in his own cause of the value of his life over B’s life or his
loved one C’s life, and then being executioner as well” (In Re A (Children) (Conjoined Twins: Surgical
Separation) [2001] Fam 147 CA, at 200).
¹³ Dworkin, Taking Rights Seriously, supra note 7, at 87.
¹⁴ Dworkin, The Model of Rules I, supra note 7, at 24 ff. ¹⁵ Id. at 39–45.
 Did Dworkin Ever Answer the Crits? 159

than Dworkin is to the positivist conception that he is attacking,¹⁶ for they

believe that a focus on purely foreground elements may radically underestimate
various forms of pathology endemic to (say) the modern American legal system.
In his early work, Dworkin imagined that the background elements of a legal
system might present themselves as discrete norms, which could be identified and
individuated by judges and advocates and spoken about almost in the way that
enacted rules are spoken about. Thus we might talk about the principle that no
man is to profit from his own fraud, and distinguish it from other principles
such as the principle that statutes are not to be given retroactive effect (if a non-
retroactive interpretation is possible) or the principle that intentional wrongdoing
is worse than equally harmful negligent wrongdoing. Even though background
principles are not enacted as texts and even though they have no canonical formu-
lation, still we do sometimes give a principle a verbal formulation and make it the
focus of legal argumentation, asking how it should be interpreted, and drawing
out its implications for particular cases. Judges and advocates do this all the time.
They make assertions about what the general policy of the law is or what princi-
ples pervade it; then they go on to use these assertions as premises for arguing
about the disposition of the cases in front of them. This gives the impression of a
sort of two-step process involved in the use of background elements in legal argu-
ment, analogous to the two-step process that is typically used in arguing about
textual rules. In step (1), we identify the norm in question—by the recognition
and citation of a validly enacted text (in the case of a rule) or by the more diffuse
argument used to discern the presence of a policy or principle in the law. Then, in
step (2), we argue for a particular disposition of the case in front of us, on the basis
of the bearing which we think the rule we have cited or the principle or policy
we have identified has on the facts of the case. This is a perfectly lucid model and,
as I have said, it is characteristic of much of what lawyers and judges say about
principles and policies.
Its disadvantage is the way it assimilates principles and policies to more discrete
forms of legal provision, such as statutory rules, as though the identification, indi-
viduation, and citation of the former were in the end not much different from the
identification, individuation, and citation of the latter.¹⁷ In fact, there are huge dif-
ferences. It’s not just that rules characteristically have a canonical formulation
which principles and policies lack; it is also that rules have (as it were) a fixed level of
generality, whereas the tendency of principles and policies is always to expand the
sphere of their application, something which has to do with their operating more
like reasons than like enacted provisions. If there is any limit on the generality of a
policy or principle, it is a matter of the esthetics or lucidity of its application to the

¹⁶ See, e.g. P. Gabel, Book Review of Taking Rights Seriously, 91 Harv. L. Rev. 302, 303–311
(1977).
¹⁷ This drew some of Dworkin’s early critics to the possibility that discrete principles may be “rec-
ognized,” within a positivist framework: see J. Raz, Legal Principles and the Limits of Law, 81 Yale L.J.
823, 843 ff. (1972). See Dworkin’s response in Taking Rights Seriously, supra note 7, at 68–71.
160 Jeremy Waldron

case with which we are primarily concerned. We state the principle so that its appli-
cation to the case at hand is patent, rather than a matter of further interpretation.
This suggests that the distinction between steps (1) and (2) is somewhat mislead-
ing, and that we might do better to think more directly about the relation between
the legal background and the instant case.
And that is the model suggested in some parts of Law’s Empire. The idea there is
that in any given case, a judgment for the plaintiff, say, is to be defended by show-
ing that it coheres better with existing legal materials than a judgment for the
defendant would. “Coheres better with” is understood as what Professor Dworkin
calls a post-interpretive claim. Counsel for the plaintiff offers a justificatory theory
which he says makes sense of all or most of the existing legal materials—prior
decisions, established doctrine, legislation, and so on—and he shows that that
theory would also justify a decision in favor of his client.¹⁸ The defendant offers a
contrary case, and counsel for the plaintiff will argue that the theory offered by the
defense is either less attractive in itself than the one that he has presented or covers
less of the existing materials than his does, or both. Of course, it is unlikely that
either of the theories on offer will stand in a justificatory relation to all the existing
legal materials. Some materials may be simply irrelevant—too far from the issues
in question in terms of the concerns they might be thought to embody, even
under the most abstract characterization. And some may have to be treated as out-
liers, relevant but incapable of being reconciled with the theory being put for-
ward. Also, it is unlikely that either theory—plaintiff ’s or defendant’s—will
amount to (what anyone regards as) a perfect theory of justice. Even allowing for
outliers, both theories will have to accommodate decisions (precedents, enact-
ments) that their proponents might not support if they were setting out their
favored account of justice on a blank slate. Still, plaintiff will argue that the theory
the defendant invokes to make the existing law look coherent is further from an
ideal theory of justice than the theory he (the plaintiff ) invokes, and defendant
will return the compliment. So they will argue back and forth on this dimension
of justice as well as on the dimension of how much of the legal record their rival
theories cover.
When we put the matter in this way, the distinction between background and
foreground elements in a legal system—between principles and rules—is seen as a
sort of façon de parler. What the existing law really amounts to is nothing more than
a mass of decisions—by constitution framers, by legislators, and by thousands of
earlier judges. Relying on background elements in current legal argument is not
really a matter of introducing a different set of provisions, viz. legal principles as
opposed to legal rules. It is rather a particular way of working with that existing mass
of decisions, though it may also involve self-conscious reference to the recorded
¹⁸ I shall leave out of account issues about the compartmentalization of law—see Dworkin,
Law’s Empire, supra note 1, at 250–254—since these do not really affect the criticisms I shall be con-
sidering. Everything I say can be qualified with reference to “local priority” or not, as the reader
wishes.
 Did Dworkin Ever Answer the Crits? 161

efforts of previous decision-makers (earlier judges, for example), to work with the
mass of previous decisions that confronted them in just the same sort of way.¹⁹
So there are these two ways of characterizing Dworkin’s approach: (i) his earlier
characterization in terms of norms called principles, i.e. norms that were comprised
in the law just like rules—only in the background of the law rather than in the fore-
ground; and (ii) his later approach which talked of rival theories put forward by
those who were working with the existing law to justify a current decision. I don’t
really think they are different, or that any difference between the two really matters.
A principle (or a policy) is nothing much more than a theory (or an element of a
theory) about what makes sense of the existing law, expressed directly in normative
terms.²⁰ We can say, in the voice of the theorist (or the current advocate) that the
best interpretation of the existing law—the theory that covers most of it and makes
it look good—is that it does not permit people to profit from their own fraud. Or
we can “point to” or “cite” a principle—“No one is to profit from his own fraud”—
more or less conscious of the fact that this in effect is something we read into the
law when we commit ourselves to the theory just mentioned. The latter character-
ization will seem particularly appropriate, when the mass of materials we are con-
fronting includes holdings by other judges which interpret the materials that
confronted them in just the way that we interpret the materials that confront us:
those holdings will enable us to point to something like “authority” for the princi-
ples we are relying on.²¹ The alternative characterization will be more useful, how-
ever, in cases where we have no such authority, but are striking out on a new
interpretive path, because we have a theory which we think better explains the
existing law (and generates a result for our client) than previous efforts at interpre-
tation do. Dworkin seems happy to move back and forth between these models in
Law’s Empire. Sometimes he says—in line with characterization (ii)—that a judge
“must choose between eligible interpretations [theories] by asking which shows the
community’s structure of institutions and decisions—its public standards as a
whole—in a better light.”²² And sometimes he says—in line with characterization
(i)—that his method “asks judges to assume . . . that law is structured by a coherent
set of principles.”²³ As I say, I don’t think anything much turns on this distinc-
tion;²⁴ I mention it only in order to make it clear that I think the two amount more
or less to the same thing, and that the critical points I am going to develop can be
applied to Professor Dworkin’s jurisprudence in both its phases.

¹⁹ This last point turns out to be quite important; see infra Section VI, text accompanying notes
56–58.
²⁰ Thus Dworkin answers the question of what principles the law contains by saying that it con-
tains those principles which belong to the soundest theory of the settled (foreground) law. See
Dworkin, Taking Rights Seriously, supra note 7, at 340 and Altman, supra, note 2, at 213.
²¹ . . . just as I pointed to “authority” for the various principles I cited, supra in notes 8, 9, and 12.
²² Dworkin, Law’s Empire, supra note 1, at 256. ²³ Id. at 243.
²⁴ These two ways of approaching the matter are also noted by A. Hunt, Law’s Empire or Legal
Imperialism, in Reading Dworkin Critically 9, 36–37 (A. Hunt ed., 1992). Hunt thinks the oscil-
lation between them matters much more than I do.
162 Jeremy Waldron

III.

Much of the CLS critique of Dworkin’s work is superficial:²⁵ for example, CLS
scholars charge that Dworkin’s jurisprudence is “elitist,” that he is insufficiently
concerned with “the deprived and disadvantaged in society,” that he is in denial
about the connection between law and politics, that his approach to justice is
too abstract, or that he “exhibits a profound antipathy for common consciousness
among the citizens.”²⁶ Dworkin himself, however, has identified a line of
CLS argument that poses a considerable threat to his enterprise.
The threat presents itself as follows. The law we confront in a modern democ-
racy is not the work of a single author. We do not live in a one party state, nor do we
even live in a unitary state. Law-makers of various persuasions coexist at various
levels or in the various tributaries of our legal system, and contribute to the growth
of the law from their diverse moral and political perspectives. Moreover, we live in a
society whose law is the repository of its checkered political history. When one
party succeeds another in electoral competition, it does not attempt to wipe the
legal slate clean and begin all over again; instead the latest generation of statesmen
work incrementally to make such modest changes in the law as their priorities dic-
tate and their political circumstances permit, leaving the remainder—the work of
the predecessors—to coexist comfortably or uncomfortably with whatever changes
they have managed to make.²⁷ The law in fact is a patchwork of provisions, the
work of a multitude of authors living and dead, with diverse and conflicting politi-
cal agendas, and diverse and conflicting commitments of principle.
In the midst of this patchwork—in the midst of this mess, we might say—
Dworkin’s lawyer is supposed to be able to find sufficient coherence to assert cred-
ibly that the law is pervaded by principles which favor his client’s case and to argue
credibly against any similar but opposing claim made on behalf of his opponent.
But why should we assume that this is possible? As Dworkin acknowledges, noth-
ing in the way the law was produced guarantees that the lawyer or the judge will
succeed in finding a coherent interpretation of it.²⁸ Or if he can find a coherent
interpretation of the patchwork, nothing in the way that the law was produced
guarantees that his success precludes similar success for his opponent: nothing
guarantees that the law has a shape amenable to a unique coherent justification of
principle.²⁹ Without that guarantee, or at any rate without a reasonable prospect

²⁵ See also Altman, supra note 2, at 215.

²⁶ See Hunt, supra note 24, at 39ff. and A. Hutchinson, Indiana Dworkin and Law’s Empire, 96
Yale L.J. 637, 600 (1987). See also Hutchinson, The Last Emperor, in Reading Dworkin
Critically, supra note 24, at 66 (common consciousness) and Gabel, supra note 16, at 314–315 (on
abstraction).
²⁷ See J. Waldron, The Circumstances of Integrity, 3 Legal Theory, 1 (1997), reprinted in
J. Waldron, Law and Disagreement (1999).
²⁸ Dworkin, Law’s Empire, supra note 1, at 273. ²⁹ Id. at 272.
 Did Dworkin Ever Answer the Crits? 163

of a unique coherence result, the Dworkinian style of legal argument is just a

recipe for the reproduction of indeterminacy.
The message of Critical Legal Studies is sometimes taken to be that law is sys-
tematically biased towards the interests of the ruling class. If that were so, then
there might possibly be some hope for Dworkin’s jurisprudence: the advocate or
judge hoping to dredge up some principled coherence from the law might expect
to find in it a more or less coherent version of ruling class ideology. But that has
always been a superficial reading of CLS. The most thoughtful CLS scholars have
emphasized the contradictory and conflict-ridden aspect of our law. Law, says
Roberto Unger, “is the product of real collective conflict, carried on over a long
time, among many different wills and imaginations, interests and visions.”³⁰ With
this sort of provenance, any given body of legal doctrine is bound to be messy, rich
in compromises, exceptions and contradictions.
Warring solutions to similar problems will coexist. Their boundaries of application will
remain uncertain. Interests and ideals favored in some domains will be discounted in
others for no better reason than the sequence in which certain decisive conflicts
took place and the relative influence enjoyed by contending parties of opinion at the
time. Intellectual fashions will join with preponderant interests to produce results that
neither interests nor fashions alone would have allowed us to predict. Defeated or
rejected solutions will remain, incongruously, in the corners of the law as vestiges of past
approaches . . .³¹
One would have to be naively optimistic—to the point of some sort of Hegelian
faith in the cunning of reason³²—to believe that these processes yield a single
determinate structure of principled reasons.
Indeed the criticism may be drawn even more tightly than that. In any piece of
high profile litigation, the parties will represent or say they are representing not
only their own interests but competing visions of what law has to offer in society.
But given our checkered political history, it is incredible that there will not already
be layers in the archeological midden of law to support each of the visions pro-
pounded by the rival parties. Plaintiff may be able to find liberal strains in the law,
corresponding to the times and the places where liberal principles have predomin-
ated in the law-making processes; and defendant will be able to find conservative
strains in the law corresponding to the times the law-making processes have been

³⁰ R. Unger, What can Legal Analysis Become? 65 (1996).

³¹ Id. at 66. By the way, awareness of this incoherence is by no means confined to CLS. See, e.g.
J. Raz, The Relevance of Coherence, in Ethics in the Public Domain: Essays in the Morality of Law
and Politics 277, 298–300 (1994) (“[L]aw is the result of the rough-and-tumble of politics, which
does not exclude the judiciary from its ambit, and reflects the vagaries of pragmatic compromises and
changing fortunes of political forces and the like. . . . [L]aw, therefore, does not form a coherent body of
principle of doctrine. . . . [T]here is no reason to expect the law to be coherent. By and large, one would
expect it to be coherent in bits—in areas relatively unaffected by continuous political struggles—and
incoherent in others.”)
³² See J. Waldron, Dirty Little Secret (Review Essay on Roberto Mangabeira Unger, What Should
Legal Analysis Become?) 98 Columbia L. Rev. 510 (1998).
164 Jeremy Waldron

wrested away from their liberal opponents. All that is there, waiting in the law for
the rival parties to come along and raid in their latest forays as litigators.
Duncan Kennedy offers the tightest version of this view.³³ He believes the con-
flicting tendencies in the law can be resolved into two opposed visions of the indi-
vidual and society—an individualist vision and an altruist vision—and that these
two visions are represented haphazardly in the law that confronts us as the high-
water marks of the ascendancy in law-making of first one tendency and then of the
other tendency.³⁴ Hence, what is or appears to be at stake for a given individual
engaged in litigation will usually present itself as one or other of these tendencies,
and Professor Kennedy’s position is that the body of law that any litigation draws
upon, and that the parties look to for support for their rival positions, is also riven
with the contradictory appearance of both positions. From the advocates’ point of
view, the result is an uncomfortable awareness that there is enough in the law for
his opponent to make an argument of principle every bit as convincing (or as
unconvincing) as his own. And from the judge’s point of view, there is the sense
that his decision might go either way: the judge may be predisposed to approach a
case in the individualist mode, but may find himself suddenly, as Kennedy puts it,
“shifting modes,”³⁵ attracted suddenly by the lines of available background argu-
ment that might sustain legal argument for the side that represents the altruistic
tendency. Either way, whether the judge shifts modes or not, or whatever the
direction in which the shift takes place, he will find no difficulty in constructing
an argument of exactly the sort, drawing on exactly the sort of elements, that
Dworkin recommends. For Kennedy is also a student of the legal background: the
only difference is that his more realistic scrutiny of it has not convinced him, as
Dworkin is convinced, that it yields to only one compelling argument of principle
in difficult cases.³⁶
Dworkin, as I said, acknowledges the importance of this objection. He calls it a
form of internal skepticism³⁷—to be distinguished from external or philosophical

³³ See Altman, supra note 2, for the view that Kennedy’s critique is the most dangerous for
Dworkin’s enterprise. As a matter of fact, Professor Dworkin barely refers to Kennedy’s article: there is
one reference, in a slightly different context, in Law’s Empire, supra note 1, at 438 n.26. (Dworkin’s
other references to Kennedy in Law’s Empire are to a bibliography of CLS that Kennedy
co-authored.) Still, Dworkin does acknowledge the substance of Kennedy’s position: he says that the
story of our legal system told by CLS is a story “of two deeply antagonistic ideologies at war within the
law, one drawn, perhaps, from communitarian impulses of altruism and mutual concern and the
other from the contradictory ideas of egoism, self-sufficiency, and judgmental moralism.” Law’s
Empire, supra note 1, at 272.
³⁴ Professor Kennedy also believes that these visions are present within each of us, and are not best
understood as associated, respectively, with particular parties, classes or factions in society. See
Kennedy, Form and Substance, supra note 3, at 1774–1776; see also D. Kennedy, The Structure of
Blackstone’s Commentaries, 205 Buffalo L. Rev., 28, 211–213 (1979).
³⁵ Kennedy, Form and Substance, supra note 3, at 1776.
³⁶ See also D. Kennedy, A Critique of Adjudication (fin de siècle) 34–37 (1977).
³⁷ Dworkin, Law’s Empire, supra note 1, at 79, 266–275. For an objection to this equation,
see P. Schlag, Missing Pieces: A Cognitive Approach to Law, 67 Texas L. Rev., 1195, 1198–1200
(1989).
 Did Dworkin Ever Answer the Crits? 165

skepticism about the very ideas of value and coherence, which he thinks is a dis-
traction in jurisprudence³⁸—and insists that it has to be taken seriously.³⁹ It is not
out of the question, he acknowledges, that the existing law could be so riddled
with inconsistency as to be unamenable to the sort of analysis he proposes. After
all, we know the foreground law is inconsistent, if not in its direct normative
implications,⁴⁰ then in the spirit that informs it—in the purposes, principles and
policies that must be assumed to make sense of it. The Dworkinian judge “knows
that legislative supremacy gives force to some statutes that are inconsistent in prin-
ciple with others . . . But he assumes that these contradictions are not so pervasive
and intractable . . . that his task is impossible.” The question now is: can that
assumption be defended?

IV.

Dworkin’s first pass at the objection is to deny that the extent of contradiction is as
great as the CLS scholars make it out to be. “The literature of critical legal studies
announces rather than defends these claims [about pervasive contradictions], as if
they were self-evident.”⁴¹ Pierre Schlag thinks this is just summary dismissal of
CLS criticism,⁴² but he is wrong and wrong, I suspect, for two reasons. He is
wrong, first, because Dworkin does not rest here on a flat denial. Indeed, he goes
on to mobilize other arguments, which I will consider in a moment. And Schlag is
wrong, secondly, because there may be more to the factual disagreement about the
extent of contradiction than meets the eye.
I shall return to the second point in Section VIII. But for the moment, I would
like to explore the first point, the one that does not involve Dworkin simply dis-
puting the extent of contradictions. After all, Dworkin is hardly in a position to
deny the existence of something like contradiction or incoherence in the law. On
the contrary, the background presupposition of his embrace of integrity is that the
settled law of any community has had multiple authors—framers, legislators,
judges, etc.—with quite radically differing visions of justice. (If there were no con-
flict between the various views about justice expressed in the enacted law, there
would be nothing for the Dworkinian ideal of integrity to do.)⁴³ So any attempt
to rebut internal skepticism simply by saying there are no contradictions, or by
saying that whatever contradictions there are, are low-level and unimportant,

³⁸ Dworkin, Law’s Empire, supra note 1 at 79–86.

³⁹ Moreover, it is “global” internal skepticism, not just piecemeal. See id. at 79. Calling internal
skepticism “global,” however, doesn’t make it the same as “external skepticism”; compare the mis-
apprehension about this in R. J. Lipkin, Beyond Skepticism, Foundationalism, and the New Fuzziness:
The Role of Wide Reflective Equilibrium in Legal Theory, 75 Cornell L. Rev. 811, 844 (1990).
⁴⁰ Foreground inconsistencies are managed by principles such as lex posterior.
⁴¹ Dworkin, Law’s Empire, supra note 1, at 273. ⁴² Schlag, supra note 37, at 1198.
⁴³ I argue this at length in Waldron, The Circumstances of Integrity, supra note 27.
166 Jeremy Waldron

would undermine the point of his preoccupation with integrity. Of course,

opinions about the extent and pervasiveness of such contradictions as there are
in the law might vary. And in the end, that’s what it may come down to. But my
initial suspicion is that Dworkin has not wanted to rely simply on the sort of limp
response that H.L.A. Hart gave to the critique of positivism put forward by the
Legal Realists—“Well, it’s not really as bad as all that.”⁴⁴ That would make his
jurisprudence contingent and precarious, and its applicability to real-world legal
systems perhaps unacceptably conditional.

V.

So what else does he say? Dworkin’s second response is to argue that this sort of
CLS skepticism neglects an important philosophical distinction between compet-
ing principles (such as autonomy and mutual concern) which may figure in a sin-
gle view about justice, and contradictory principles (such as equality and
inequality) which cannot possibly be combined in one coherent conception.⁴⁵
Referring to a tort law example he uses in Chapter 7 of Law’s Empire, Dworkin
says there is no contradiction in recognizing both a principle of collective solidar-
ity and a principle of individual fault-based responsibility: “on the contrary, any
moral vision would be defective if it wholly disowned either impulse. . . . No gen-
eral interpretation that denied either one would be plausible; integrity could not
be served if either were wholly disavowed.”⁴⁶ The challenge, he says, is to articu-
late a principled balance between them in the interpretive theory that we use to
justify our particular legal conclusions.
Now, what Dworkin says here is right, as far as it goes. Most moral theories, includ-
ing theories of justice, do have elements that stand in tension with one another; and it
is the mark of moral maturity not to regard this as a logical defect, but rather to see it
as an inevitable feature of our coming to terms with the plurality of values and princi-
ples.⁴⁷ But does pointing this out meet the objection? I fear it does not.
CLS scholars like Kennedy deny that the individualist and altruist elements
they discern in the legal background are merely prima facie principles or the sort of

⁴⁴ See H. L. A. Hart, Positivism and the Separation of Law and Morals, 71 Harv. L. Rev. 593,
614–615 (1957). Hart’s response to the Realist critique was “to emphasize that the hard core of set-
tled meaning is law in some centrally important sense and that even if there are borderlines, there
must first be lines. If this were not so the notion of rules controlling courts’ decisions would be sense-
less as some of the ‘Realists’—in their most extreme moods, and, I think, on bad grounds—claimed.”
See also H. L. A. Hart, The Concept of Law 136–141 (2d ed. 1994); H. L. A. Hart, American
Jurisprudence through English Eyes: The Nightmare and the Noble Dream, in his collection, Essays in
Jurisprudence and Philosophy 123, 126–132 (1983).
⁴⁵ Dworkin, Law’s Empire, supra note 1, at 269–275. ⁴⁶ Id. at 269–270.
⁴⁷ In this paper, I will not explore the tension between this view of moral pluralism and the
more holistic/hedgehog approach that Dworkin defends in other contexts. See, e.g. R. Dworkin,
Sovereign Virtue (2000).
 Did Dworkin Ever Answer the Crits? 167

in-tension ingredients that one would find combined or balanced in a well-

worked-out moral theory.⁴⁸ What they claim to find in the background of our law
are not rival tendencies but incompatible visions or incompatible theories of what we
owe to others and to the collective life of our society.⁴⁹ Each of these theories or
visions already balances egotistical against other-regarding influence. So, for
example, on Duncan Kennedy’s account, the individualist element in private law
is not just a vector, exerting an egoistic influence on the final social solution. It
already represents a determination of how self-regarding and other-regarding ele-
ments might be balanced. Kennedy observes that “individualism is sharply dis-
tinct from pure egotism, or the view that it is impossible and undesirable to set
any limits at all to the pursuit of self-interest,”⁵⁰ and it is distinct precisely in this
point that it already represents an attempt—the individualist attempt—to bal-
ance selfish and other-regarding tendencies. Something similar, Kennedy says, is
true of what he calls altruism:
[T]he altruist is unwilling to carry his premise of solidarity to the extreme of making every-
one responsible for the welfare of everyone else. The altruist believes in the necessity and
desirability of a sphere of autonomy or liberty or freedom or privacy within which one is
free to ignore both the plights of others and the consequences of one’s own acts for their
welfare.⁵¹
Once again, the altruism one finds in the legal background is not just a solidaristic
or communitarian tendency; it is already a view about how such a tendency
should be balanced against the egotistical tendencies that compete with it. Each of
the rival views about justice in society that Kennedy claims to discern in the back-
ground is already a view about how competing principles such as autonomy and
mutual concern should be ranked and weighed within a single conception. In that
sense, each does contradict the other’s weighting and ranking.
Dworkin ought to have no difficulty seeing this. After all, the presupposition of
his concern with integrity is that the authorship of enacted law reflects a diverse
⁴⁸ And similarly, CLS scholars argue that each of us is psychologically torn, not just between rival
tendencies of self-regarding and other-regarding impulses—which would after all be a very, very weak
claim and obviously true—but between rival visions of the relation between self and society. See
Kennedy, The Structure of Blackstone’s Commentaries, supra note 34.
⁴⁹ I should emphasize that it is not being asserted that these are fully worked-out and systematic
theories. There was a period, Kennedy says, when individualism did present itself as a systematic body
of theory: he is referring to the heyday of laissez-faire (Kennedy, Form and Substance, supra note 3, at
1729–1731). But with the disintegration of classical laissez-faire, both individualism and altruism
present themselves as fairly unsystematic bodies of theory. But still there is an important distinc-
tion—which Dworkin’s response elides—between an unsystematic theory (which unsystematically
balances various rival tendencies) and a prima facie principle or tendency (which would be the thing
to be balanced—systematically or unsystematically—with other tendencies).
Admittedly, Kennedy’s formulations are not always consistent on this. The formulation most use-
ful to Dworkin would be the following: “[M]odern individualism presents itself not as a deductive
system, but as a pole, or tendency or vector or bias, in the debate with altruism over the legitimacy of
the system of rules that emerged in the late nineteenth century.” Id. at 1732. Still, in context,
Kennedy’s reference is to haphazard and intuitive balancing exercises that contradict one another,
rather than to confrontation among the tendencies as between which balancing is required.
⁵⁰ Id. at 1714–1715. ⁵¹ Id. at 1718.
168 Jeremy Waldron

array of views about justice, not merely a diverse array of impulses (some selfish,
some communitarian) which would later have to be organized into a single theory
of justice by the Dworkinian interpreter.⁵² It is not a case of one statute’s having
been passed by an autonomy faction and another statute’s having been enacted by
the party of mutual concern, and of its never having occurred to anyone except the
judge (now faced with the two statutes together) that a suitably complex position
could accommodate both principles.
I have not said anything so far about Kennedy’s even more skeptical claim, that
(whether we regard individualism and altruism as competing balances or as com-
peting principles to be balanced) no metaprinciple of balance can be defended as
objectively correct. Kennedy says:
[I]t is futile to imagine that moral and practical conflict will yield to analysis in terms of
higher level concepts. The meaning of contradiction at the level of abstraction is that there
is no metasystem that would, if only we could find it, key us into one mode or the other as
circumstances “required.” . . . [W]e cannot “balance” individualist and altruist values . . .
except in the tautological sense that we can, as a matter of fact, decide if we have to. The
imagery of balancing presupposes exactly the kind of more abstract unit of measurement
that the sense of contradiction excludes.⁵³
This claim is supposed to cut off an approach that the Dworkinian lawyer might
take in the face of what we have said so far. Instead of saying that he is trying to
balance conflicting tendencies whose presence in the law does not reflect earlier
attempts to balance them, the Dworkinian lawyer could say that he is purporting
to replace the existing—and admittedly contradictory—balancings embedded in
the law by individualists and altruists, with a true balance, one that would be nei-
ther individualist nor altruist, but just correct. Kennedy wants to pre-empt this by
announcing that the idea of a true balance is simply out of the question.
To the extent he considers it, I think Dworkin regards this as an expression
of external skepticism about right answers in regard to balancing, and as such
he rejects it out of hand.⁵⁴ True, it may not be possible for the Dworkinian
lawyer to demonstrate that the new balancing—with which he proposes to
supersede the contradictory balancings embedded in the law by individualists
and altruists—is objectively correct. But non-demonstrability is not the same as
falsity or futility.
I think this last part of Kennedy’s critique is properly dismissed in this
way. But whether the dismissal ultimately helps Dworkin is a question we must
postpone until after we have examined the merits—in jurisprudence—of the
constructivist approach that we are now imagining the Dworkinian lawyer
undertaking.

⁵² Dworkin, Law’s Empire, supra note 1, at 178; see also Waldron, Law and Disagreement,
supra note 27, at 193–194. ⁵³ Kennedy, Form and Substance, supra note 3, at 1775.
⁵⁴ Cf. Dworkin, Law’s Empire, supra, note 1, at 267.
 Did Dworkin Ever Answer the Crits? 169

VI.

According to Professor Dworkin, the CLS or internal skepticism critique labors

under another misapprehension. The fact that various elements in our law have
conflicting inspirations does not mean that we cannot construct a theory that
resolves the contradictions into some sort of attractive coherence. After all,
Dworkin’s judge or lawyer is not simply reporting the provenance of the legal
materials he is dealing with: “He tries to impose order over doctrine, not to dis-
cover order in the forces that created it.”⁵⁵ The interpretive task is not to find out
the purpose or intention with which past legal decisions were actually made; it is
to make something good and coherent of the past decisions whether that good or
coherent understanding corresponds to what the decision-makers in the past had
in mind or not.
There is a quick version of this Dworkinian response and a longer version. The
quick version might go as follows. Dworkin might want to use this point about
constructivism to contest Professor Kennedy’s assumption that there are elements
in the existing legal record which just are individualist and that they are contra-
dicted by elements in the existing legal record which just are altruist, as though all
the elements come to us ready labeled in that way. In fact, Dworkin might want to
say, what we face is an array of decisions: say, the enactment of statute S1, a decision
for the plaintiff in case C1, a decision for the defendant in case C2, the enactment of
statute S2, and so on. Calling S1 and C1 altruistic and S2 and C2 individualistic, and
asserting therefore that the array of decisions is riddled with contradictions is itself
already an interpretation of the materials. Dworkin might protest therefore that
Kennedy is rigging the game by helping himself to these characterizations, as
though they were part of the preinterpretive specification of the materials.
This is a fair point, as far as it goes, and I shall return to it at the very end of
the paper. Some of what Kennedy says about altruism and individualism is that
they are characteristics of legal argument, not of legal materials: they are “two
opposed attitudes that manifest themselves in debates about the content of pri-
vate law rules,” and “they are helpful in the general task of understanding why
judges and legislators have chosen to establish or enact particular private law
doctrines.”⁵⁶ It looks, then, as though a Dworkinian lawyer might resist these
characterizations—leaving it open to himself to characterize the various mater-
ials in other ways—and thus resist the premise that he is confronted with inher-
ently contradictory materials. This strategy makes particular sense with regard
to legislative materials. We should be wary of any attempt to treat particular legis-
lators’ purposes as though they were canonical and on a par with the text of the
statute they enacted. Dworkin has been a fervent opponent of this mode of

⁵⁵ Id. at 273. ⁵⁶ Kennedy, Form and Substance, supra note 3, at 1713.

170 Jeremy Waldron

statutory interpretation,⁵⁷ and he must be allowed the advantage of that here.

So Dworkin’s interpreter need not accept that S1, say, is an altruistic statute; he
may want to fit it into his theory in a different way, which is at odds with what
its original sponsors thought.
With case law, though, the situation is more complicated and less helpful to
Dworkin. I don’t think he would (or should) want to associate his constructivist
approach to integrity with the claim that all we have to interpret is the actual holding
of a case—e.g., appellant loses, plaintiff has to pay $40,000 damages—and that we
can ignore the opinion produced by the court explaining that outcome, as though it
were the case-law analogue of legislative history. Unless one wants to adopt the posi-
tion of an extreme legal realist—like Jerome Frank⁵⁸—we will usually regard judicial
opinions accompanying decisions as part of the record we are required to interpret
and not as extraneous to it. Of course, those opinions are themselves partly interpre-
tive of the materials that their authors confronted. The judicial opinion accompany-
ing the decision in case C3 will be in part an interpretation of the decisions (and the
opinions) in cases C2 and C1; so the judge in C4 reading the opinion in C3 will be
reading the opinion of someone who was trying to do just what he is doing (so far
as C1 and C2 are concerned). What the judge is interpreting is thus in part a set of
interpretations. The activity of interpretation is recursive: one interprets the past
interpretations of others for they too are part of the community’s record. So one
interprets not just outcomes but modes of decision, patterns of argument, emergent
doctrines, and so on. And for this, it is not inappropriate to say, as Kennedy and
others sometimes say, that they are individualist or that they are altruist. One just has
to read them, and it may be very difficult—a Herculean task, in fact—to read the
altruism out of the decisions that one wants to reconcile with one’s own consistent
individualist interpretation or vice versa.
So there is no refuge for the constructivist in this quick gambit. The contradic-
tions among the materials he confronts may be non-negotiable. If this is the case,
then Dworkin’s constructive interpreter has to adopt a different and less glamorous
strategy. He has to pick and choose from among the existing contradictory decisions
the ones that his interpretation proposes to fit, and discard the rest. Confronting a
given set of legal materials, Dworkin’s judge or lawyer will try to construct a coherent
theory—or state an appropriate principle or set of principles—that implicates as
many of them as possible. Since there are contradictions (as opposed to mere ten-
sions) in the materials, no doubt a significant number of them will not be covered by
the theory or the principles he proposes. But this doesn’t mean he cannot come up
with a theory or a set of principles that fits a significant portion of them. (Indeed, if
the materials are pervaded by contradiction, as Kennedy asserts, then one would
expect Dworkin’s interpreter should be able to come up with a theory that fits
roughly half.)

⁵⁷ See Dworkin, Law’s Empire, supra note 1, at Ch. 9.

⁵⁸ J. Frank, Law and the Modern Mind (1970) at 111.
 Did Dworkin Ever Answer the Crits? 171

Now, a mischievous interpreter might take a different strategy: he might come

up with something purporting to be an interpretive theory, which was as inconsis-
tent as the patchwork of materials it purported to fit. But that would just show
bad faith—the bad faith that many opponents of CLS accuse its practitioners of
harboring: they want to make the law look bad. “Nothing is easier or more point-
less,” says Dworkin, “than demonstrating that a flawed and contradictory account
fits as well as a smoother and more attractive one.” But to sustain his internal skep-
ticism, the critic “must show that the flawed and contradictory account is the only
one available.”⁵⁹ The internal skeptic is not entitled to say that the interpretive
enterprise is doomed to failure unless he has tried as hard as he can: “[he] must
claim to have looked for a less skeptical interpretation and failed.”⁶⁰
Actually this last point won’t do, so far as the burden of proof is concerned.
Internal skepticism is not refuted by a showing that a non-contradictory
interpretation—a non-contradictory theory or a consistent set of principles—is
available. For of course one can select a non-contradictory subset from any mass of
inconsistent propositions: “p” on this sort of constructivism, would be a consistent
interpretation of {p, ~p}, and similarly individualism (or, alternatively, altruism)
would be a consistent interpretation—on this constructivist approach—of the
law as Professor Kennedy portrays it. After all, it’s a matter of basic logic that one
can make a case for anything on the basis of contradictory premises. So the onus
cannot be on the CLS skeptic to show that no case can be made. What must be
shown, in order to refute the internal skeptic, is either that only one such coherent
interpretation can be constructed or that a credible case can be made that one
of the available coherent interpretations of the contradictory materials is better
than any of the others.⁶¹ Too often, I think, Dworkin’s formulations neglect the
forensic adversarial context of legal argument: it is not enough to show that a lawyer
can come up with a legal argument; what he comes up with must be capable of
refuting and displacing the legal argument that his opponent is likely to come up
with as well.⁶²
Of course, all this needs to be leavened with some acknowledgment of whatever
is legitimate in Dworkin’s argument about right answers and demonstrability:
there can, he insists, be right answers to legal questions, even though there is no
acknowledged method of demonstrating that an answer is right.⁶³ I will assume

⁵⁹ Dworkin, Law’s Empire, supra note 1, at 274. ⁶⁰ Id.

⁶¹ This is seen clearly by M. Strassberg, Taking Ethics Seriously: Beyond Positivist Jurisprudence in
Legal Ethics, 80 Iowa L. Rev. 901, 929 (1995) (“[I]ndividual judges will disagree about whether and
how strongly a particular principle is embedded in our law. These interpretive disagreements expose
Dworkin to the criticism that judicial decisions relying on such principles are, in fact, a matter of
idiosyncratic judicial discretion. Indeed, a cynical view of the law . . . would suggest that institutional
support of some kind can be dredged up to support any principle favored by a judge”).
⁶² I sometimes wonder whether it would have been better for Dworkin to concentrate on an ideal
advocate rather than an ideal judge (Hercules), for by the time Hercules comes to write his opinion—
sitting as Dworkin imagines him on a single-person bench—there is just the legal argument he is
composing; there is no longer a competitor’s argument to be refuted.
⁶³ Dworkin, Taking Rights Seriously, supra note 7, at 216.
172 Jeremy Waldron

that this is roughly correct,⁶⁴ and that the case for internal skepticism cannot be
merely that Dworkin has offered no algorithm for correctness in legal argument.
Still, one is inclined to say: something must count as tending to show—or
there must be something one can do which counts as attempting to show—that
the argument one is making is better than any opposing argument one can
imagine being made by an opponent in the same case, drawing constructively
upon the same array of materials. I mean something must count as that, apart from
one’s consciousness that one’s own argument can be made or that it just is the
argument one is making. This is because one is also conscious—confronted as one
is, with this mass of contradictory materials—that one’s opponent’s argument can
be made as well.
I get the impression that Professor Dworkin is sometimes prepared to say that
this is a misplaced demand, analogous to what he regards as the unhelpful demand
that for a person who believes that moral propositions are objectively true or false,
something must count as trying to show that a given moral proposition is objec-
tively correct, over and above the mere assertion and re-assertion of that proposi-
tion.⁶⁵ At other times, however, he characterizes the phenomenology differently,
in a way that suggests that someone wedded to an interpretive argument may sen-
sibly regard himself as having reasons for thinking the argument he is wedded to is
correct. The characterization I have in mind is set out, in the first instance, in
regard to interpretive claims about literature:
Someone just converted to a new reading of Paradise Lost, trembling with the excitement
of discovery, thinks his new reading is right, that it is better than the one he has abandoned,
that those yet uninitiated have missed something genuine and important, that they do not
see the poem for what it really is. He thinks he has been driven by the truth, not that he has
chosen one interpretation to wear for the day because he fancies it like a necktie. He thinks
he has genuine, good reasons for accepting his new interpretation and that others, who
cling to the older view he now thinks wrong, have genuine, good reasons to change their
minds.⁶⁶
Now, I have no doubt that sometimes legal argument may have this character:
trembling with excitement, a jurist comes up with a new interpretation of existing
doctrine, a sense of something that others have missed, and of their having good
reasons to abandon their old tired reading in favor of one’s own. Sometimes it may
feel like that (though there would still be philosophical questions about what to
infer from such phenomenology). But is this how we expect it to be for most cases?
Is this phenomenology—the trembling excitement of discovery—likely to be
characteristic of legal work in the context of materials that are as contradictory as the

⁶⁴ But note that, anyway, the right answer motif is muted in Law’s Empire. I have discussed this
further in J. Waldron, The Rule of Law as a Theater of Debate, in Dworkin and his Critics ( J. Burley
ed., 2004)
⁶⁵ See Dworkin, Law’s Empire, supra note 1, at 81; R. Dworkin, Objectivity and Truth: You’d
Better Believe It, 25 Philosophy and Public Affairs 87 (1996).
⁶⁶ Dworkin, Law’s Empire, supra note 1, at 77.
 Did Dworkin Ever Answer the Crits? 173

CLS scholars say our existing American law is? For remember: Dworkin is offering
the constructive aspect of his argument as a way of finessing the CLS claim about
contradictions. Never mind how conflicted the law is, he says, never mind how
riddled it is with contradictions; one can always come up with some sort of con-
structive argument for one’s position. So the question is—if the legal materials are
as conflicted (and are known to be as conflicted) as, say, Duncan Kennedy thinks
they are⁶⁷—would one expect the process of constructive argument that Dworkin
promotes to have the sort of phenomenological flavor of “discovery” and “good
genuine reasons” that characterizes our imagined student’s new interpretation of
Paradise Lost? Given the knowledge that anyone faced with a set of contradictions
has about the opportunities that inconsistent premises offer for a multitude of
arguments in opposite directions, I think this is most unlikely.
We are contemplating the possibility that faced with an array of legal materials
as contradiction-ridden as Kennedy and other CLS scholars take them to be, a
constructive argument interpreting the law in favor of one side or the other in liti-
gation will be able to satisfy some very modest threshold of “fit.” Crudely, it will
make sense of roughly half of the materials it considers, and its proponent will be
conscious that any remotely competent work done on behalf of the other party
will also satisfy that threshold of fit, since it makes sense of the other half of the
(conflicting) legal materials under consideration. In these circumstances, should
we expect either advocate to make his argument with a burning consciousness that
he is right about the law and his opponent wrong? Should we expect the judge,
who entertains both arguments and understands the condition of the possibility
of each of them, to be conscious of a basis for discriminating between them?
Dworkin remains remarkably upbeat, as he faces the prospect of this even
rivalry between two equally viable theories purporting to make “sense” of the same
mess of contradictory legal decisions. What he says, at least about the judge’s situ-
ation, is this:
Hard cases arise, for any judge, when his threshold test [of fit] does not discriminate
between two or more interpretations of some statute or line of cases. Then he must choose
between eligible interpretations by asking which shows the community’s structure of insti-
tutions and decisions—its public standards as a whole—in a better light from the stand-
point of political morality. His own moral and political convictions are now directly
engaged.⁶⁸
The idea is that a judge, faced with two arguments that “fit” (in this quixotic
sense) an equal quantity of the contradictory legal materials, will resolve the issue
between them by considering which is superior on the merits, that is, on grounds
of justice or other substantial values. And that will be the tie-breaker.⁶⁹

⁶⁷ Of course, they may not be. But then Dworkin is thrown back on the position we discussed at
the end of Section IV: saying (Hart-style) to the skeptics: it is not really as bad as all that.
⁶⁸ Dworkin, Law’s Empire, supra note 1, at 255–256.
⁶⁹ See also Dworkin, Taking Rights Seriously, supra note 7, at 340–341.
174 Jeremy Waldron

To understand what this position amounts to, we need to recall that on the CLS
account the legal materials are not just contradictory in some technical sense.
They are contradictory as between the main views about justice (and other sub-
stantial values) already held by diverse factions in society, and also as between the
main values that are likely to be at stake in any piece of litigation to which it is sup-
posed that law as integrity should be applied. We need to remember, too, that this
is the hypothesis that Dworkin is supposed to be responding to, and in connec-
tion with which the passage just quoted above is supposed finally to refute the
internal skeptic’s critique.
Let me put it very schematically: suppose Kennedy is right and there are two
recognizable moral positions, Ind (for individualism) and Alt (for altruism) in the
law—contradicting one another, but spread throughout the settled law in an array
of decisions in numbered cases {Ind1, Alt2, Ind3, Alt4, Ind5, Alt6, . . . }. And now
once again, the old adversaries face off against one another in yet another case—
case 7, say. So the two sides set about constructing their arguments, following the
constructive method that Dworkin recommends to practitioners of law as
integrity. The individualist party cites a subset of the established cases {Ind1, Ind3,
Ind5} and he puts cases 2, 4, and 6 aside. He is conscious, of course, that his oppo-
nent is likely to cite an equally sized subset {Alt2, Alt4, Alt6}, putting cases 1, 3, and
5 aside. So what does the individualist party propose to clinch his argument? He
says that his argument with its support in half the case law is better than his oppo-
nent’s, because it is committed to Ind and because Ind is, as he believes, a better
moral theory than Alt. The reason that clinches the matter on his account—the
reason that is supposed to be capable of clinching the matter on Dworkin’s
account—is simply that in this conflict between two moral views that have come
to pervade the law, he has aligned himself with the better one!
I very much doubt whether Dworkin will be embarrassed by this (though I will
show shortly that he ought to be). The point of the critique implicit in my
schematic example tends to be obscured by the fact that when Dworkin says that
the judge’s “own moral and political values” are engaged in the choice among rival
theories that fit the existing legal materials equally well, we are not reminded that
those very values are likely to be implicated on one side of the contradictions that
pervade the law (on the CLS account). He sometimes writes as though the contra-
dictory legal materials—leading to the impasse of fit—are one thing, and the val-
ues the judge brings to his task are another. I don’t mean that Dworkin’s argument
rests on this misapprehension. He would probably bite the bullet and say that the
judge has no choice but to rely on his own views about justice, even when his own
views about justice are exactly the views that are represented (and contradicted) in
the diverse legal materials facing him. But the possibility is not one that he brings
to our attention. More important, he doesn’t draw to our attention the fact that
the attorneys for the parties⁷⁰ are already committed to one side or the other in

⁷⁰ See also supra note 62.

 Did Dworkin Ever Answer the Crits? 175

this debate; they are litigating the issue precisely because they are partisans of Ind
and Alt respectively. So the values they appeal to when they try to clinch their
argument are, in some sense, question-begging arguments. They assume the very
point at issue and they then claim that this assumption breaks the tie between
their own argument and that of their opponents!

VII.

An even more serious difficulty with Professor Dworkin’s constructivist response

to the internal skeptic is the question of its compatibility with the underlying
argument for integrity in Law’s Empire.
Briefly: Dworkin’s judge (and hence Dworkinian attorneys) are supposed to be
preoccupied with the interpretation of the existing legal materials because, like all
of us, they have an obligation rooted in political morality to keep faith with the
commitments of their community. This, according to Dworkin, is why it is worth
persisting in trying to make sense of the existing legal record, tainted as it is with
contradiction (if the skeptics are right). Absent that sense of obligation, the sens-
ible thing to do in legal practice would be to abandon wrestling with the messy
array of existing law and simply settle (and argue) cases pragmatically on the basis
of what will be best for a just and prosperous future. But if the skeptic is right, if
the legal materials are as contradictory as he says they are, then there is no coherent
record to keep faith with, and therefore no ground for the integrity-based case
against pragmatism to stand on. There may be clever moves—constructivist
moves—that a Dworkinian lawyer can make which look impressive and which, at
a stretch, have the feel of interpreting even the messy record that the internal skep-
tic postulates. But those moves do not and cannot engage the sort of integrity val-
ues that the interpretive exercise is ultimately answerable to. In other words, they
have no point or at least they don’t have the sort of point that integrity is supposed
to give them. Quite the contrary: these constructivist moves in the face of contra-
dictory materials make a mockery of the communitarian values underpinning
integrity. That, in the end, is the seriousness of the challenge posed by internal
skepticism: the challenge is not merely that Dworkin is giving Hercules an impos-
sible task; it is that the task given to Hercules makes no sense, if the legal world is as
the internal skeptic says it is.
There are several phases in this argument. The first is Professor Dworkin’s own
acknowledgment that in the end his own jurisprudence stands or falls with the
refutation of pragmatism. He phrases the point in terms of legal rights—that is,
rights which are generated by legal principles (or rights which are the upshot of
successful interpretive theories).
The pragmatist thinks judges should always do the best they can for the future, in the
circumstances, unchecked by any need to respect or secure consistency in principle with
what other officials have done or will do. . . . [Pragmatism] rejects what other conceptions
176 Jeremy Waldron
of law accept: that people can have distinctly legal rights as trumps over what would other-
wise be the best future properly understood. . . . I do not say this in any triumphant way.
The fact that a true pragmatist rejects the idea of legal rights is not a decisive argument
against that conception. For it is not self-evident that the idea of legal rights is attractive.
Or even sane.⁷¹
Indeed, as Dworkin goes on to say, it is quite easy to make the idea of legal rights
seem foolish. If a judge is convinced that the doctrine in a line of cases has proved
itself unjust, why should he accept the argument that a party has a legal right that
the injustice be perpetuated? Maybe if the party can show that he would suffer
more on account of his reliance on the injustice than his opponent would gain
were it corrected, or maybe if it can be shown that society is better off with this
doctrine as a basis for social predictability—maybe then there are reasons for
accepting his argument. But those arguments are themselves pragmatic and they
don’t affect the underlying point that it seems foolish to worry away at difficult
precedents, tangled doctrine, or obscure legislation to find out what the law really
says, when one suspects it will be unjust and when it is clear anyway that no one
is really relying on any particular interpretation and that interpretive argument
contributes nothing to predictability. Dworkin is adamant in Law’s Empire that
conventionalism, as a theory of law, falls at this hurdle—conventionalism cannot
withstand the pragmatist challenge⁷²—and he cannot consistently confront law
as integrity with a lower bar than this.
So a jurisprudence of integrity must be capable of explaining why we ought to
refuse the pragmatist invitation to decide cases simply in a way that is best for the
future. A jurisprudence of integrity must explain why it is not foolish to worry
away at difficult precedents, tangled doctrine, or obscure legislation in a quest for
a solution (to the case in front of us) which is consistent with what our commu-
nity has already committed itself to. And the account of legal argument that it gives
must connect with that explanation.
Now, Dworkin’s answer to the pragmatist challenge has to do with the condi-
tions for political legitimacy. Decisions by judges and other legal officials will be
enforced, and there is a question about the legitimacy of their enforcement and
the obligation that citizens have to defer to them—particularly those on the los-
ing side in the decisions. Professor Dworkin pursues a long and interesting argu-
ment suggesting that legitimacy depends on the extent to which the obligation to
accept legal outcomes can be presented as a form of associative or communal
obligation.⁷³ Associative obligation, he says, can be sustained only among the
members of a community “who share a general and diffuse sense of members’
special rights and responsibilities from or toward one another.”⁷⁴ Delving into
the legal record for principles or attempting to produce an interpretive theory

⁷¹ Dworkin, Law’s Empire, supra note 1, 160–162. ( I have reversed the order of the first two sen-
tences in this excerpt. That does not affect the sense.) ⁷² Id. Ch. 4.
⁷³ Id. at 186–216. ⁷⁴ Id. at 199.
 Did Dworkin Ever Answer the Crits? 177

which gives an attractive and coherent account of what has been done among us
already in the name of our association—both of these are ways of trying to show
what that “diffuse sense of members’ special rights and responsibilities” is. And
connecting those principles or that theory with the decision one is arguing for in a
particular case is a way of showing that parties in the case have an associative obli-
gation to abide by that decision, because they are already in a sense committed to it.
Dworkin argues that such a strategy has a huge legitimacy-advantage over the
pragmatist approach. In aiming to make society better for the future, the pragma-
tist naturally commands the support of those who accept his vision of social jus-
tice and the general good. But it is hard to see why anyone with a different or
opposed vision of social justice or the general good should accept the decisions of
a pragmatist official. Integrity, however, makes a different sort of claim—one
rooted ultimately in reciprocity. Even when we disagree with one another about
justice or the general good, we can see that some among us have been benefitted
by decisions applying principles of a certain sort, and we can appreciate the fair-
ness of allowing others to be benefitted by those principles, even when we disagree
with them, because we think that their consistent and open-ended application
across a diverse array of cases helps to establish us as a community of principle.
So: this is the case that Dworkin needs in order to make sense of modes of legal
argumentation that grapple—in this non-pragmatic way—with the obscure
implications of the set of decisions our community has already committed itself
to. Such arguments present our association with one another as a community of
principle and they use that as a basis for saying that a decision, now, that flows
from the principles we have unearthed has legitimacy even for those who would be
inclined in the abstract to oppose it. To claim this legitimacy advantage, we have
to be willing to view the existing record of our community in a certain light—“to
see and enforce it as coherent.”⁷⁵ A judge who accepts integrity must be prepared
to treat the parties who come before him in this spirit: “They are entitled, in prin-
ciple, to have their acts and affairs judged in accordance with the best view of what
the legal standards of the community required or permitted at the time they acted,
and integrity demands that these standards be seen as coherent, as the state speak-
ing with a single voice.”76 And it is in this spirit that Professor Dworkin under-
takes the enterprise which we have been discussing in Sections II through VI.
Let us return now to the internal skepticism of CLS and to the constructivism
that Dworkin uses to respond to it. The internal skeptic denies that it is possible
to view the record of a community like ours as one in which a certain set of coher-
ent principles have consistently determined legal outcomes. The record of our
community, he says, does not disclose any consistent “sense of members’ special
rights and responsibilities from or toward one another,” diffuse or otherwise. The
internal skeptic denies therefore that any resolution of an instant case can made

⁷⁵ Id. at 167. ⁷⁶ Id. at 218.

178 Jeremy Waldron

legitimate by showing that it flows from such a set of coherent principles. The record
of our community, he says, is one of incoherence, in which the principles necessary
to sustain some decisions are contradicted by the principles necessary to sustain
others. There is nothing, says the internal skeptic, corresponding to Dworkin’s
description of a consistent set of principles presupposed by all or most of the
existing legal materials, from which we could infer legitimate answers to the legal
questions that currently confront us. Professor Dworkin may say that “associative
obligations can be sustained among people who share a general and diffuse sense
of members’ special rights and responsibilities from or toward one another, a sense
of what sort or level of sacrifice one may be expected to make for another.”⁷⁷ But
Duncan Kennedy’s point is precisely that we are conflicted about what we owe one
another; we don’t share a single conception; we share contradictory ones, if we
share anything at all. Again, Dworkin may say that “[l]aw as integrity asks judges
to assume, so far as this is possible, that the law is structured by a coherent set of
principles . . . and it asks them to enforce these in the fresh cases that come before
them, so that each person’s situation is fair and just according to the same stan-
dards.”⁷⁸ But the skeptical case is that this is not possible, and that this instruction
simply cannot be carried out, because the existential assumption fails: law is not
structured by a coherent set of principles.
For myself, I find it difficult to see how any form of versatile constructivism can
displace or mitigate this verdict. A Dworkinian judge or lawyer might make an
attempt to establish a record that is consistent as far as it goes, in the sense of identi-
fying principles which offer a coherent explanation of some of the extant decisions.
But the internal skeptic thinks that any such attempt will be easy to discredit.
Whatever decisions it explains, it will leave an equal number of decisions unex-
plained (treated as outliers). So any such interpretation (in Dworkin’s own words)
will “show the record of the community in an irredeemably bad light, because
proposing that interpretation suggests that the community has characteristically
dishonored its own principles.”⁷⁹
Notice that this is not just a matter of there being rival theories available, and no
demonstrable way of settling the issue between them. We can accept Professor
Dworkin’s claim that “people are entitled to a coherent and principled extension
of past political decisions even when judges profoundly disagree about what this
means”⁸⁰ and his view that “consistency in principle [can] be important for its
own sake, . . . [even] when it is uncertain and controversial what consistency really
requires?”⁸¹ But those maneuvers do not save the position here. For what we face,
if the internal skeptic is correct, is not a dispute between two viable interpretations
which unfortunately cannot be settled in a demonstrable way; what we face are
two rival interpretations, each of which is discredited by the fact that it seems to
show a very considerable part of the community’s record in a very bad light.

⁷⁷ Id. at 199. ⁷⁸ Id. at 243. ⁷⁹ Id. at 257. ⁸⁰ Id. at 134. ⁸¹ Id. at 163.
 Did Dworkin Ever Answer the Crits? 179

The situation is not saved by a formulation which Dworkin sometimes uses, in

which law-as-integrity is understood in terms of an ethic of trying to portray the
law as coherent, rather than in terms of a duty to uncover the coherence that is
actually there. I mean formulations like the following:
We want our officials to treat us as tied together in an association of principle, and we want
this for reasons that do not depend on any identity of conviction among these officials,
either about fit or about the more substantive principles an interpretation engages. Our
reasons endure when judges disagree, at least in detail, about the best interpretation of the
community’s political order, because each judge still confirms and reinforces the principled
character of our association by striving in spite of the disagreement, to reach his own opin-
ion instead of turning to the usually simpler task of fresh legislation.⁸²
These formulations are quite common in Law’s Empire and elsewhere. Dworkin
says that integrity is upheld “when people in good faith try to treat one another
in a way appropriate to common membership in a community . . . and to see
each other as making this attempt, even when they disagree about exactly what
integrity requires in particular circumstances.”⁸³ He says that “[w]e gain even
through the attempt”⁸⁴ and that “[l]aw’s empire is defined by attitude, not terri-
tory or power or process”⁸⁵ and “[l]aw as integrity consists in an approach, in
questions rather than answers.”⁸⁶ But not everything that one does to advance
one’s own position through legal argument can count as a good faith attempt of
this kind. Nor are the conditions for a good faith attempt only subjective—such as
that the lawyer is doing the best for integrity that he can, or whatever. The attempt
must be something which it makes sense to embark on, and the critical or skepti-
cal case is that, given what participants must know about the pervasively contra-
dictory character of the existing legal materials, picking and choosing from among
them those that suit one’s purposes can hardly be described as a way of trying to
display one’s allegiance to their underlying coherence.

VIII.

Since the end of Section IV, I have been working on the assumption that the legal
background in a society like ours is about as riddled with contrary principles as the
CLS skeptics say it is. I proceeded on this assumption in order to see whether
Professor Dworkin had anything convincing to say in response to their skepticism
apart from a flat denial of its factual premise. I think we have concluded that he
has not. The distinction between tensions and contradictions does not do the
trick and neither does the move to constructivism nor the insistence that non-
demonstrability is compatible with objectivity.

⁸² Id. at 264 (emphasis added). ⁸³ Id. at 190 (emphasis added).

⁸⁴ Dworkin, Taking Rights Seriously, supra note 7, at 338.
⁸⁵ Dworkin, Law’s Empire, supra note 1, at 413. ⁸⁶ Id. at 239.
180 Jeremy Waldron

That leaves the flat denial. At the end of Section IV, I was inclined to doubt that
Dworkin wanted to rest his jurisprudence on this. It sounded flabby and uncon-
vincing, and appeared to make the whole enterprise contingent and precarious.
But maybe that was premature, and if we turn around the argument that we have
just been developing (in Section VII), we may be able to see why. I said that con-
structive interpretation by a Dworkinian advocate cannot count as a bona fide
attempt to keep faith with the commitments of his community if it is known to
the advocate that the record of the community is one of pervasive and systematic
contradiction. If the advocate knows that roughly half the decisions embody the
moral proposition that members of the community have very limited obligations
to one another and roughly half of the decisions embody the moral proposition
that members of the community have very extensive obligations to one another,
he cannot with a straight face pull out just the materials in the latter half and argue
on that basis that his opponent has a particular obligation to his client because this
is what we—the community—have always been committed to. But suppose his
knowledge of the extent and pervasiveness of the contradictions is not so clear.
After all, the illustration that we used in Section VI—an array of decisions in
numbered cases {Ind1, Alt2, Ind3, Alt4, Ind5, Alt6, . . . }—was a ludicrously over-
simplified version of the legal record that most advocates are likely to be con-
fronted with. Ludicrously over-simplified, and ludicrously over-clarified: the tens
of thousands of extant legal decisions that we find in the law reports and the
statute books present a much more equivocal and tangled record than this.
Confronted then with real-world legal materials, the Dworkinian advocate will
have in the back of his mind the possibility that he will not be able to make a
determinate case—or, as I emphasized in Section VI, the possibility that his
opponent may be able to make a case as determinate as the one that he makes—but
he may think it worth trying to see whether he can do any better, for integrity, than
that. If it seems to him that the prospect of success here is not out of the question,
then there is no reason to say—as I said about the overly simple case—that his
attempt just makes a mockery of integrity.
So the factual disagreement between Dworkin and the skeptics may come
down to this: Are the contradictions so clear and so pervasive that it is evidently
not worth trying to see whether a particular set of principles (or a particular inter-
pretation) fits the existing law? “Not worth trying” is the key predicate in this
formulation. It does not mean and it should not be taken to mean that there is no
legal argument that we can come up with. If we are prepared to pick and choose, we
can always come up with something, even if or (as I said earlier on the basis of ele-
mentary propositional logic) especially if we know the materials are inconsistent.
“Not worth trying” must refer to the conditions that would make the attempt
important or worthwhile; that is, it must refer to the general idea of integrity.
It must not be out of the question that our argument or our principles fit a very
significant portion of the legal materials, and it must not be out of the question
that this is not the case for our opponent’s argument. If the constructive strategy
that organizes our attempt rests on our having to say, even implicitly, that “the
 Did Dworkin Ever Answer the Crits? 181

community has characteristically dishonored its own principles,”⁸⁷ then it is not

an attempt worth making.
I think, in the end, this is where Dworkin should take his stand against the
Crits. He should say (and he does say): it is not clear up front that attempts to
argue in the mode of law-as-integrity are doomed to failure. If it were clear, we
should have no reason to resist the siren charms of pragmatism: forget the existing
law; ask instead what’s best for the future; and take one’s chances on the legitimacy
issue. But sometimes legal argument looks promising, and when it does we are
obliged to make the attempt (and the theory of integrity explains why). The
cruder versions of CLS jurisprudence were easy to discredit: it simply was not so
clear that law is systematically biased in favor of one social class as some CLS
scholars said it was. Law is more complicated than that. The more sophisticated
versions of CLS that I have been considering in this paper are certainly less easy to
dismiss, and they deserve as Dworkin says to be taken seriously.⁸⁸ But perhaps the
least convincing thing about them is the simplicity of their thesis about contradic-
tion: there are these two positions (individualism and altruism) and the existing
legal materials divide clearly and evenly between them. So long as we stick with
that simple picture, the futility of law-as-integrity is evident. But complicate and
muddy it somewhat, with more terms to the contradictions and arrays of less eas-
ily legible materials, then the appearance of futility recedes. And at that stage, as
Dworkin says,⁸⁹ there is nothing to do but try, for nothing else will reveal whether
an attempt can succeed while keeping faith with the motivations behind integrity.
The conclusions I have reached are perhaps not quite what Dworkin would want,
for two reasons. First, I have insisted that what one is trying to make is a determi-
nate legal argument, and that means one that excludes or promises to exclude the
prospect of a similar argument by one’s opponent. To repeat: nothing is easier than
for a constructive jurisprudence to come up with something that looks like a legal
argument. The test is whether there is a reasonable prospect that one can show
that nothing but this argument will keep faith with the commitments of our com-
munity. Secondly, I do think that Dworkin is simply wrong to think that this bur-
den can be discharged by using one’s own moral and political convictions to break
ties.⁹⁰ If the only thing to differentiate plaintiff ’s argument from defendant’s argu-
ment, as a matter of law, in plaintiff ’s eyes is that plaintiff is an altruist and—as the
plaintiff thinks—altruism is right (as a theory of justice), then plaintiff is arguing
as a pragmatist. He is not arguing in good faith in the mode of law-as-integrity
since, ex hypothesi, he concedes that the rest of his argument (the legal part) fails to
discharge the burden that integrity imposes and what he appeals to as a tie-breaker
has no connection with integrity, or nothing that would differentiate the stance of
a defender of law-as-integrity from a defender of pragmatism. The moral is that
the use of the tie-breaker, too, must satisfy the conditions of integrity. It can’t just
be wheeled out on an ad hoc basis to rescue an argument that would fail at the bar
of integrity without it.

⁸⁷ Id. at 257. ⁸⁸ Id. at 273–275. ⁸⁹ Id. at 274. ⁹⁰ Id. at 256.

 8
Associative Obligations and the
Obligation to Obey the Law
Stephen Perry

I. Political Obligation
One of the strands woven into the complex fabric of Law’s Empire (hereinafter LE)
is an argument that there exists, under certain conditions, a general moral obligation
to obey the law. Whether or not there can ever be such an obligation is an age-old
problem in political philosophy, and Dworkin’s argument is offered, in part, as a
contribution to that particular philosophical tradition. But it is more than that,
because the argument also constitutes an integral part of Dworkin’s general theory
of law. That is why I say it is one strand of a complex fabric; for Dworkin, political
and legal philosophy are inextricably connected. For many of the philosophers
who have addressed the question of political obligation, as I shall call it, there is no
such inevitable connection. Notice, to begin, that no respectable theory of politi-
cal obligation ever claimed that a person is obligated no matter what to obey the
laws of a legal system to which he or she is subject. Every minimally plausible the-
ory sets out certain conditions under which such an obligation is said to arise, and
Dworkin’s is no exception. Many such theories have, however, regarded these con-
ditions as ones that do not figure in any essential way in the concept of law itself.
They assume that law constitutes a system of norms the existence and content of
which can be established by, say, looking to certain kinds of social facts, and then
asking whether or not a given legal system meets a set of independently specifiable
conditions; if these conditions are met, then a general obligation to obey is said to
exist. Arguments based on consent are often (although not necessarily) of that
form; they look to an independently specifiable condition that asks whether or
not everyone who is subject to a legal system has validly consented to obey its laws,
whatever they are. The argument from fair play is also of this form; it looks to an
independently specifiable set of conditions that asks, first, whether the content of
the norms of the system show it to be a mutually beneficial scheme of cooperation,

Associative Obligations and the Obligation to Obey the Law. Stephen Perry.
© Oxford University Press 2006. Published 2006 by Oxford University Press.
184 Stephen Perry

and, secondly, whether or not those subject to the system have “accepted” its benefits.¹
There are, as Dworkin points out, well-known problems with both arguments if
they are regarded as the basis of a general obligation to obey the law, since it is never
the case that everyone subject to any given legal system has validly consented
to obey it or has accepted, in the appropriate sense, certain benefits it happens to
provide (LE at 192–195).² My present point, however, is that these arguments
treat the philosophical problem of political obligation as for the most part inde-
pendent of the philosophical problem about the nature of law.
Dworkin’s theory, as I noted, is not like this. Before I say anything about the
specifics of his view, however, let me first say a little more about the question of
political obligation itself.³ A general obligation to obey the law exists only if every
person who is subject to a given legal system has a moral obligation to obey each
and every law of the system because it is a law of the system. In saying that one has
an obligation to obey the law because it is the law I do not mean that one’s reason
for action in doing as the law requires must be that the law requires it; for the most
part, the law is indifferent to why one complies with the law so long as one does so.
To say that one has an obligation to obey the law because it is the law means,
rather, that at least one sufficient ground or basis of the obligation is the fact that
the law exists. The law need not be the only basis of the obligation. We clearly have
independent moral obligations not to assault and murder people, for example.
However, if one has a moral obligation to obey the law then the law must be a basis
of obligation in addition to one’s independent moral obligations. Often, of
course, the law modifies (or at least purports to modify) independent moral obliga-
tions, or to make them more precise where they are indeterminate, and there
would not be much point to law if it was not capable of at least sometimes doing
this. But it will not suffice to establish a general moral obligation to obey the law
to show that the law has had some effect on one’s moral obligations or, more gen-
erally, on one’s moral situation. It is, as I said, necessary to show that one has an
obligation to obey each and every law because it is the law. This challenge must be
met not just where the law reproduces independent moral obligations, but also
where the law makes moral mistakes, for example mistakes about what justice
requires. Theories of political obligation almost always place limits on the extent
to which the law can make moral mistakes and still give rise to a general obligation
to obey the law, but it is nonetheless no easy matter to show that there is ever an
obligation to obey an unjust law. And since no legal system is ever perfectly just, a
theory of political obligation would not have shown very much if it failed to show

¹ See H. L. A. Hart, Are There any Natural Rights?, 64 Phil. Rev. 185 (1955); J. Rawls, A Theory
of Justice 301–308 (2d ed., 1999).
² See A. J. Simmons, Moral Principles and Political Obligations (1979), especially
chs. 3–5.
³ In this paragraph I draw on material from my article Law and Obligation, 50 American
Journal of Jurisprudence 263 (2005). For very helpful (albeit skeptical) discussions of political
obligation, see Simmons, supra note 2; L. Green, The Authority of the State (1990), particularly
ch. 8; and J. Raz, Ethics in the Public Domain (1994), ch. 14.
 Associative Obligations and the Obligation to Obey the Law 185

that the general obligation to obey is not capable of encompassing at least some
unjust laws. Dworkin’s theory acknowledges this point. A general obligation to
obey the law can, of course, be prima facie and defeasible by other moral consid-
erations; it is possible that an obligation to obey an unjust law can exist, but be
overridden by the independent obligations of justice itself. Dworkin’s theory
acknowledges this point as well.
Here it will be helpful to bring in a distinction that Dworkin draws between the
grounds and the force of law (LE at 108–113). Dworkin does not employ the fol-
lowing formulation, but we can say that the grounds of law are the truth condi-
tions of propositions of the form “It is the law (of some specified jurisdiction) that
A,” where A is a normative proposition that states that some person or group of
persons has some normative status, for example that they have an obligation,
right, permission, liability, or power. The paradigmatic case is a proposition speci-
fying an obligation, and I will continue to concentrate on that case. (The properly
generalized version of the problem of political obligation asks whether the law sys-
tematically affects persons’ normative status in all the various ways that it claims to
do; for present purposes, we can ignore this complication.) For the sake of ease of
expression I will from now on omit the reference to a specific jurisdiction,
although it should always be understood as present. Suppose the proposition “It is
the law that everyone has an obligation to do X” is true.⁴ The force of law is its
normative force, which means its moral force.⁵ The concern when we are speaking
about force, in other words, is with the truth or falsity of normative propositions
such as “Everyone has an obligation to do X” when they figure in true proposi-
tions such as “It is the law that everyone has an obligation to do X.” To say that a
legal system has general normative force is simply to say that it gives rise to a gen-
eral obligation to obey. Force, as a general concept in legal philosophy, is con-
cerned with the connection between the truth of propositions of the form “It is
the law that A” and the truth of “A.” For the reasons given in the preceding para-
graph, it will not suffice to establish that there is a general obligation to obey the
law to show that for all propositions of the form “It is the law that ——,” both “It
is the law that A” and “A” are true. One must show, in addition, that at least one
sufficient basis for the truth of “A” is the truth of “It is the law that A.” Stated more
precisely, one must show that the various facts, whether social, moral, or both,
that serve to establish the truth of “It is the law that A” also serve to establish the
truth of “A.” A theory of political obligation is a theory about the conditions that

⁴ “Everyone” can be understood as referring to all persons who are subject to the particular legal
system; it does not matter for present purposes what, exactly, that means.
⁵ It is of course true that not all normativity is moral, but Dworkin assumes that legal normativity,
if it exists at all, is a type of moral normativity. Thus legal obligations (if they exist) are moral obliga-
tions, legal permissions (if they exist) are moral permissions, and so on. In my view this assumption is
correct, but I cannot discuss the point here. I will limit myself to the observation that it is not an
assumption that is in conflict with positivism, although not all positivists accept it. Kelsen and Raz
do, but Hart does not.
186 Stephen Perry

must obtain in order for this connection to hold in a systematic way for all the
laws of a legal system.
This brings me back to the observation that Dworkin’s theory of political obliga-
tion does not treat the question of whether there is ever a general obligation to
obey the law as independent from the philosophical problem about the nature of
law. Dworkin says that a theory of law is a theory about the grounds of law, which
means that it is a theory about the truth conditions of propositions of the form “It
is the law that A.” He is not always entirely clear about whether such a theory
applies only to the law of particular jurisdictions or whether it is meant to be more
general, but I think it is ultimately clear that it is meant to be more general: a the-
ory about the grounds of law is (part of ) a theory about the nature of law under-
stood as a general kind of social phenomenon. To say that a theory of law is about
the grounds of law is very vague, but I hope the formulation will suffice for present
purposes. It is meant to cover, for example, such theoretical views as Dworkin’s
own theory of law as integrity, Hart’s theory of the rule of recognition, and various
versions of the social sources thesis. I will say something about the specifics of law
as integrity in a moment. The point to be noticed for present purposes, however, is
that the theory holds that the grounds of law depend, in part, on considerations
having to do with the (potential) force of law. Dworkin does not put the point in
the following way, but I believe the essence of his view is this: law as integrity,
which is a substantive theory about the grounds of law, presupposes the truth of a
particular theory of political obligation, and it does so in such a way that the truth
of propositions of the form “It is the law that A” depends, in part and to a limited
extent, on whether or not “A” is capable of being made true by that theory of
political obligation.⁶ This is very abstract, and I will begin to fill in the details of
the view later in this section and in section III. But the abstract formulation suf-
fices to show us that, on Dworkin’s view, the theory of political obligation does
not take the content of law as given and then ask whether or not there is a general
obligation to obey; it is, rather, an integral part of the theory of the grounds of law,
and therefore itself has an effect on the content of law. This is, as I said, a claim
that Dworkin makes about law as integrity. But earlier in Law’s Empire he makes
an even more ambitious, methodological claim, which holds, in effect, that any
general theory about the grounds of law must presuppose some theory of political
⁶ Notice that the truth of propositions of the form “It is the law that A” is said to depend in part
not on the independent truth of the normative proposition “A,” in which case we would be dealing
with a certain radical kind of natural law theory, but rather on whether or not “A” is capable of being
made true by the theory of political obligation. In other words, the truth of “It is the law that A”
partly depends on whether or not “A” is capable of being made true by the truth of “It is the law that
A.” As we shall see, Dworkin’s own theory of law as integrity makes this determination holistically, by
looking to how “A” fits into the content of the law taken as a whole. There is no vicious or unavoid-
able circularity here. It should also be emphasized that the truth of propositions of the form “It is the
law that A” is said to depend only partly on the theory of political obligation. The most important
determinant of the truth of such propositions is, of course, social practice, such as facts about what
legislatures and courts have done. This is the crucial dimension of “fit” in Dworkin’s general theory of
interpretation.
 Associative Obligations and the Obligation to Obey the Law 187

obligation in much the same way as his own theory of law as integrity does.⁷
Dworkin summarizes the point by saying that “[a] general theory of law . . . pro-
poses a solution to a complex set of simultaneous equations” having to do with
both the grounds and the force of law (LE at 110). Different theories of law pre-
suppose different theories of political obligation (or else deny that there can be
such a theory). To determine which theory of law is correct we must, therefore, at
some point address the general question of political obligation, and this necessar-
ily takes us deep into issues of substantive political morality. This is, I believe, the
nerve of Dworkin’s general theory of interpretation, understood as a methodology
for doing philosophy of law.
I cannot discuss the general theory of interpretation here, except to make the
following brief observations. Although the truth conditions of propositions like
“It is the law that A” are affected by moral considerations having to do with the
capacity of the law to obligate us, they also depend crucially and unavoidably on
social facts, and in particular on facts about what legislatures and courts have
done. Given that social practice is, necessarily, the starting point of interpretation,
I do not think that Dworkin should be understood as saying that propositions of
the form “It is the law that A” can never be true unless some theory of political
obligation is true (although he is clearly more optimistic on this latter score than
most other contemporary theorists). What I believe he is saying, rather, is that the
practice of law with which we are familiar presupposes that there is a true theory of
political obligation. The practice is built on the assumption that certain kinds of
social facts serve as a systematic basis or ground for obligations. But this presupposi-
tion might be mistaken. It is possible that, as a general moral matter, the truth of a
proposition of the form “It is the law that A” can never be a sufficient basis for the
truth of “A”—i.e., the truth of such a proposition can never be a sufficient basis
for the existence of an obligation—in which case law as a general kind of social
practice would not make sense; complete skepticism about it would be justified. It
is also possible that, whether or not a general theory of political obligation is true,
a particular legal system is so unjust or evil that, within that system, the truth of a
proposition of the form “It is the law that A” is never, or almost never, a sufficient
ground for the truth of “A;” skepticism about the particular legal system would
then be justified. In both of these cases, Dworkin need not and should not deny
that propositions of the form “It is the law that A” could still be true just by virtue
of facts about social practice. In making this acknowledgment, however, we would
also have to acknowledge that, in the first case, the general practice was by its own
lights a radically defective one, and, in the second case, that we were dealing with a
degenerate instance of a legal system. (Cf. LE at 101–108.)
There is one more preliminary matter that needs to be mentioned before we
consider the specifics of Dworkin’s substantive views about political obligation.

⁷ “In much the same way” is deliberately vague. It is beyond the scope of this paper to explore
Dworkin’s general methodological claims about legal philosophy in any detail.
188 Stephen Perry

Dworkin holds, in keeping with a long and respectable tradition in political

philosophy, that the question of whether or not there is a general obligation to
obey the law is unavoidably connected to the question of when, if ever, the state is
morally legitimate: “A state is legitimate if its constitutional structure and prac-
tices are such that its citizens have a general obligation to obey political decisions
that purport to impose duties on them” (LE at 191). Given that the state, by means
of law, “purports to impose duties,” Dworkin is surely right that it is appropriate,
and indeed inevitable, to judge its moral legitimacy by reference to the standard of
whether or not it succeeds in creating those duties.⁸ This is particularly true given
that the state claims the authority not just to impose duties but to enforce them by
coercive means. As Dworkin notes, there may well be circumstances in which a
state is justified in the use of coercion even though there is no duty on the part of
those coerced to acquiesce, but these are exceptional (LE at 191). He is also appro-
priately cautious in describing the relationship between obligation and the justifi-
cation of coercion: obligation is “close to” a necessary condition for coercion, but
it is not sufficient. Although he is sometimes misunderstood in this regard,
Dworkin claims no more than the following: “[N]o general policy of upholding
the law with steel could be justified if the law were not, in general, a source of
genuine obligations” (LE at 191).
This brings me, finally, to Dworkin’s particular substantive theory of law as
integrity. As the above remarks make clear, Dworkin uses this term to refer both to
a theory of law, meaning a theory about the truth conditions of propositions of
the form “It is the law that A,” and to a related theory of political obligation.
Context should generally make clear to which I am referring, although for the
most part I shall be focusing on the theory of political obligation. Dworkin’s basic
approach to the problem of political obligation is interesting and original. He
argues that the general obligation to obey the law is an associative obligation,
meaning an obligation that arises within certain kinds of limited human associ-
ations. The paradigms are friendship and family. Dworkin describes a number of
conditions that he maintains are characteristic of associative obligations in gen-
eral, and then argues that law, when it meets versions of those conditions that are
appropriate to its character as a form of human association, likewise gives rise to
obligations. I will discuss the general conditions of associative obligation in some
detail in the following section, and accordingly will not set them out here. So far as
political obligation is concerned, Dworkin’s central claim is that it arises when a
legal system exhibits the distinct political ideal of integrity. Integrity, which is to
be distinguished from the other political ideals of fairness, justice, and procedural
due process, “requires government to speak with one voice, to act in a principled
and coherent manner towards all its citizens, to extend to everyone the substantive
standards of justice or fairness it uses for some” (LE at 165). The content of the
law depends, according to Dworkin, on interpretation, and the law should be

⁸ This is also Raz’s view.

 Associative Obligations and the Obligation to Obey the Law 189

interpreted as a whole so as to exhibit integrity, to the extent that this is possible.

(This is the upshot of the abstract idea that the truth conditions of propositions of
the form “It is the law that A” depend, in part, on whether or not “A” is capable of
being made true by the truth of “It is the law that A.”) If the content of the law
does exhibit integrity, then what would otherwise just be a “bare” community
becomes what Dworkin calls a “genuine” political community, and those who
belong to it have a general moral obligation to obey the law. This obligation is an
associative obligation similar to the obligations that hold between friends and
among members of a family.
In the following sections I shall discuss this argument for political obligation in
some detail. Here let me just sketch my main conclusions. In Section II I examine
Dworkin’s general analysis of associative obligations. I conclude that, while there
are some difficulties with the analysis, Dworkin has pointed to a genuine and dis-
tinct type of obligation that arises within certain kinds of limited relationships
among persons. I further suggest that the justification for these obligations, and
the reason for their distinctiveness, is that these relationships are ones of intrinsic
value. If that is so, then Dworkin’s strategy for establishing a general political
obligation requires us to show that a genuine political community—one whose
law exhibits integrity—has intrinsic value for its members. Properly understood,
this is, I believe, precisely Dworkin’s argument. In Section III I discuss this idea in
more detail and try to show that, despite the obvious differences between personal
relationships like friendship and family, on the one hand, and a political commun-
ity of the kind exemplified by modern states on the other, this line of argument is a
promising one. It may well be the only kind of argument for a general political
obligation to obey the law that has any chance of success.

II. Associative Obligations

Associative obligations are, very roughly, obligations that arise in connection with
certain limited human associations, like friendships and families. Such obligations
are, as Dworkin notes, special, which means that they are owed only to certain
persons and not universally. Are all special obligations also associative? This is not
a question that can be definitively answered before we have in hand an analysis of
associative obligations, but as an intuitive matter there does not seem to be any
good reason to think that they are. Suppose there is a general moral duty of easy
rescue, meaning a duty to assist another person who is in serious danger if one can
do so at no great cost to oneself. If you fall into a raging river and I am in a position
to throw you a life ring, then it is quite plausible to think that the duty I have in
those circumstances is a special one; I owe a duty to you that I do not owe to other
people. No doubt there is some formal sense in which the two of us are in associ-
ation with one another, but this does not look like what people typically have in
mind in speaking of associative obligations. Or think of the obligations that arise
190 Stephen Perry

when we exercise powers of promise or consent. If I promise to read a manuscript

of yours, then I come under a special obligation to you to do as I have promised.
Again, there is clearly a sense in which we are in association with one another, and
yet promissory obligations are often contrasted with, rather than treated as a
species of, associative obligations.
What is special, then, about associative obligations? Dworkin, who sometimes
uses the term “fraternal” instead of “associative,” offers the following general
analysis of the concept:
[T]he members of a group must by and large hold certain attitudes about the responsibil-
ities they owe one another if these obligations are to count as genuine fraternal obligations.
First, they must regard the group’s responsibilities as special, holding distinctly within the
group, rather than as general duties its members owe equally to persons outside it. Second,
they must accept that these responsibilities are personal: that they run directly from each
member to each other member, not just to the group as a whole in some collective
sense. . . . Third, members must see these responsibilities as flowing from a more general
responsibility each has of concern for the well-being of others in the group; they must treat
discrete obligations that arise only under special circumstances, like the obligation to help
a friend who is in great financial need, as derivative from and expressing a more general
responsibility that is active throughout the association in different ways. . . . Fourth, mem-
bers must suppose that the group’s practices show not only concern but equal concern for
all members. Fraternal obligations are in that sense conceptually egalitarian. They may be
structured, even hierarchical, in the way a family is, but the structure and hierarchy must
reflect the group’s assumption that its roles and rules are equally in the interests of all, that
no one’s life is more important than anyone else’s. Armies may be fraternal obligations if
that condition is met. (LE at 199–201)

Dworkin summarizes the four conditions in the following way: “The responsibil-
ities a true community deploys are special and individualized and display a perv-
asive mutual concern that fits a plausible conception of equal concern” (201). He
adds that “[t]he concern [these conditions] require is an interpretive property of
the group’s practices of asserting and acknowledging responsibilities—these must
be practices that people with the right level of concern would adopt—not a psy-
chological property of some fixed number of the actual members” (LE at 201).
There is much about this analysis that rings true. All four conditions appear to
hold for friendship, for example, which is surely a paradigm of a type of human
relationship that gives rise to associative obligations. I have some doubts, however,
about whether the same can be said of every associative obligation, or at least every
example of an associative obligation that Dworkin offers us. Clearly, as already
noted, the first condition is a necessary one; associative obligations are special, in
the sense that they are owed only to certain persons. The third condition, that
associative obligations express a general concern for the other person or persons
with whom one is associated, is probably a necessary one as well. But are all asso-
ciative obligations personal, in the sense defined by Dworkin’s second condition?
Certainly the obligations that exist in the paradigm cases of friendship and family
 Associative Obligations and the Obligation to Obey the Law 191

are personal in this sense. Recall, however, that Dworkin wishes to show that
political obligation is a type of associative obligation. If a general political obliga-
tion ever exists, it is far from clear that it is personal. In form, at least, it is not. A
general obligation to obey the law is usually thought to be owed, if to anyone, to
the community or state itself rather than to one’s fellow citizens considered one by
one. If the obligation is owed to the community it must of course be true that the
community is, or is capable of being treated as, a moral agent in some appropriate
sense, but this is far from being an obstacle for Dworkin; he insists that a proper
understanding of political and legal practice in fact requires moral personification
of the community. When it comes time to argue that the model of principle,
which is Dworkin’s preferred model of political community, meets the four stated
conditions, he writes with respect to the second condition that the model makes
the responsibilities of the relevant community “fully personal” because “it com-
mands that no one be left out, that we are all in politics for better or worse, that no
one may be sacrificed, like wounded left on the battlefield, to the crusade for
justice overall” (LE at 213). But this is not the sense of the second condition as
Dworkin originally defined it. He is not here arguing that the responsibilities in
question “run directly from each member to each other member.” The idea that
the model of principle “commands that no one be left out” sounds, rather, like a
restatement of the fourth condition, which holds that the concern expressed by
the responsibilities of an associative relationship must be an equal concern.
Turning directly to that fourth condition, what does it mean to say that the con-
cern expressed by the responsibilities of an associative relationship must be an equal
concern? Consider friendship. In this context the condition presumably means
that the obligations that two friends reciprocally owe to one another must express a
more or less equal level of concern. This is what Dworkin seems to have in mind
when he says that friends ought to have “roughly the same concern” for one another
even if they do not share exactly the same understanding of the responsibilities of
friendship (LE at 199). In the family context, however, the equality condition seems
to mean something different. Dworkin discusses the example of a culture that gives
parents the power to choose spouses for their daughters but not for their sons. The
culture accepts “in good faith” the equality of the sexes but thinks, mistakenly, that
equality of concern requires a paternalism for daughters that it does not require for
sons. Dworkin argues that, despite this moral mistake, it is possible that daughters
in this culture owe their parents a genuine obligation of obedience, although it is
also possible that the obligation might ultimately be overridden by other moral
considerations (LE at 204–205). I am not at present concerned with the question
of whether or not Dworkin is correct in making this claim. I mean only to point
out that the notion of “equal concern” must in this context have a quite different
sense from the one it has in the context of friendship. Here, it seems to mean some-
thing like the following: the moral powers that one holds over others within an
associative relationship may only be exercised in a way that manifests equal concern
for all those over whom the powers are held. Dworkin claims, controversially, that
192 Stephen Perry

children can come under an obligation to obey their parents even if parental power
is not so exercised, so long as it is exercised in accordance with a conception of equal
concern that is plausible and sincerely held. Whether or not this claim is true,
“equal concern” does not here refer, as it does in the case of friendship, to an equal
level of reciprocal concern that the members of an associative relationship owe to
one another. To make this clear, imagine the case of a single parent raising a single
child. The parent has an obligation to care for the child, and associated with this
obligation is a power to command obedience. The child has an obligation to do
as the parent requires. These two obligations—the obligation on the part of the
parent to care for the child, and the obligation on the part of the child to obey the
parent—are not reciprocal obligations like those that hold between two friends.
Both obligations are most plausibly understood as being grounded in the wellbeing
of the child alone. It is of course true that the child may eventually come under an
obligation to care for the parent in the latter’s old age, but this is not an obligation
that is reciprocal to the parent’s obligation to care for the child when the child is
young. This is not because the two obligations cannot exist at the same time. It is
because neither obligation is conditioned on the existence—past, present or
future—of the other. The parent of a child with a fatal illness has no less a duty to
care for the child—and arguably has a greater one—than the parent of a child with
a normal life expectancy, and a person close to adulthood who is adopted or
becomes a step-child may still come under an obligation to care for his or her new
parent in old age.
The main point I was concerned to establish in the last paragraph is that there
does not seem to be a single conception of equal concern that informs even the
paradigmatic associative relationships of family and friendship. The equality of con-
cern that must hold between friends, which is a form of reciprocity, need not hold
between parent and child. (Relations between siblings is a more complicated case,
which I do not discuss here.) Nor does there seem to be, in the case of friendship, an
analogue to the obligation to exercise parental powers over a number of different
children so as to manifest equal concern for each. (We do not, for example, owe the
same level of concern to all our different friends, and that is true even within a
group of mutual friends; the appropriate level of concern varies with the closeness
of the particular friendship.) It does not help to say that there is a more abstract
conception of equality underlying the specific conceptions that inform friendship
and family, because at a sufficiently abstract level all moral principles, and not just
those that underpin associative obligations, are, in Dworkin’s phrase, “conceptually
egalitarian.” None of this calls into question the obligatoriness of true associative
obligations, but it suggests, particularly when considered with the point made earl-
ier that some associative obligations might not be personal in character, that the
attempt to find a comprehensive set of attributes that hold for all associative rela-
tionships is unlikely to succeed. The problem is compounded when one considers
the many kinds of social practice that Dworkin regards as being, at least potentially,
associative in character: in addition to family, friendship, and political community,
 Associative Obligations and the Obligation to Obey the Law 193

he mentions academic colleagueship, union membership, and membership in an

army, and he clearly thinks there are others as well. Once we have failed to find
common ground even among the paradigm cases, how can we be sure that these
apparently quite different kinds of human association are, from a moral point of
view, properly grouped together? Relatedly, how can we be sure that other features
that seem to be essential to the paradigm cases of friendship and family—for exam-
ple, true emotional bonds of love, affection, or respect—are not essential features of
associative obligations in general, thereby ruling out such associations as political
community and (I would imagine) academic colleagueship?
Dworkin’s response to the above concerns might well be that the four condi-
tions are the basis of an “abstract interpretation of the yet more general practice of
associative obligation itself ” (LE at 197), and that it is this more general practice
that reveals the underlying unity of friendship, family, union membership, and so
on. Even if the proffered four conditions fall short as an interpretation of the more
general practice—they are, after all, only advanced by Dworkin as a first run at the
problem—the very existence of the general practice gives us reason to think, the
response would continue, that some satisfactory interpretation can be found. It is,
however, far from clear that any such general practice of associative obligation
exists, at least in the sense of a practice that self-consciously and reflectively mani-
fests the interpretive attitude. There could only be such a practice if people gener-
ally regarded the relevant patterns of behavior as a distinct practice, thought the
practice had value, and took the requirements of the practice to be sensitive to
what its value was taken to be (LE at 47).⁹ Even if friendship, family, political
community, and so on all exhibit the interpretive attitude within themselves,
there does not appear to be a comprehensive practice of associative practices, so to
speak, of which that is also true.
There is another kind of objection to Dworkin’s account of associative obliga-
tions that should be mentioned here. If it has merit it applies even if, contrary to
what I have suggested, the four conditions succeed in capturing common features
of the paradigm cases of associative relationships. An initial version of the objec-
tion runs as follows. Insofar as the analysis based on the four conditions simply
identifies certain attitudes held by the members of a group about their responsibil-
ities to one another, it is not clear why those attitudes should give rise to, or be
taken to be markers of, actual responsibilities within the group, as opposed to
mere beliefs about responsibilities. Denise Réaume has argued, for example, that
Dworkin’s analysis is ultimately no more than a variation on Hart’s practice theory
of rules, and as such it falls prey to Dworkin’s own critique of that theory.¹⁰

⁹ I take it that the existence of the interpretive attitude is a psychological or sociological matter. If
it were not, there would a danger that the term “interpretive” would simply be a label applied to prac-
tices that might, for whatever reason, be capable of being a source of obligation. The general account
of interpretation would lose its distinctiveness.
¹⁰ D. Réaume, Is Integrity a Virtue? Dworkin’s Theory of Legal Obligation 39 U. of Tor. L.J. 380,
402 (1989).
194 Stephen Perry

According to Dworkin’s critique, the fact that a group treats certain of its practices
as obligatory cannot, in and of itself, make them obligatory; something more in
the way of normative argument is required.¹¹ As it stands Réaume’s objection
skips too quickly over a crucial point, which is that the four conditions are
expressly stated by Dworkin to be interpretive properties of a group’s practices
rather than “psychological propert[ies] of some fixed number of actual members”
(LE at 201).¹² The third condition, for example, requires that these practices be
ones that “people with the right level of concern would adopt” rather than ones in
which some minimum proportion of members actually manifest concern for their
fellows (LE at 201).¹³ The four conditions together are meant to be an interpreta-
tion of what Dworkin takes to be a general practice of associative obligation, and
interpretations always have a dimension of justification or value. Dworkin’s
analysis thus cannot be regarded as a mere variant of Hart’s practice theory. That
having been said, however, we still run up against the question mentioned in the
previous paragraph of whether or not there is, in fact, a general interpretive prac-
tice of associative obligation. We also run up against the related problem of what
the value associated with such a practice might be. Although Dworkin is not
entirely clear about this point, he seems simply to invoke a single general value of
community or fraternity. But just as it is not clear that there is, in fact, a general
interpretive practice of associative obligation, it is similarly not clear that there is a
single value that can be regarded as accompanying such diverse forms of human
association as friendship, political community, union membership, and so on. If I
am right that the four conditions fail to capture common features of even the
paradigmatic instances of associative obligation, then we have at least some reason
to doubt the existence of both a general practice and a single value that underlies
every instance of such a practice.
In the preceding discussion I have tried to raise some doubts about Dworkin’s
analysis of associative relationships, but I have not called into question his basic
strategy for showing that a general moral obligation to obey the law can exist. He
begins, it will be recalled, with the premise that certain limited forms of human

¹¹ R. Dworkin, Taking Rights Seriously (1978), ch. 3.

¹² This does not call into question the point that the existence of the interpretive attitude, and
hence the character of a practice as an interpretive practice, is itself a psychological or sociological
matter. See further supra note 9.
¹³ It should, however, be borne in mind that the paradigmatic associative obligations of family
and friendship cannot exist unless each friend or family member actually feels certain emotions for,
and holds certain attitudes toward, the other person or persons in the relationship. I discuss this point
further below. Holding an attitude towards another person is different from holding an attitude
about the nature of the duties that you owe to that person, and the four conditions for the most part
express attitudes of the latter rather than the former kind. But in the case of the third condition the
two types of attitude converge; the mutual concern required by that condition must, in the case of
friendship and family, exist in a psychological sense. We must therefore take seriously the possibility
mentioned earlier that all associative obligations similarly require actual concern among the relevant
persons. This would obviously pose problems for treating political obligation as a type of associative
obligation.
 Associative Obligations and the Obligation to Obey the Law 195

association are generally and plausibly regarded as giving rise to obligations, and
then argues that political communities are associations of that same kind and
therefore they, too, are capable of giving rise to obligations. I believe this strategy is
a promising one, but if we are to give effect to it we must characterize associative
obligations in a somewhat different way from Dworkin’s analysis based on the
four conditions. (As I will suggest later, the appropriate characterization turns out
to be an important, if perhaps implicit, strand in Dworkin’s own discussion of
associative obligations.) My approach will be to focus, in the first instance, on the
paradigm associative relationship of friendship. Why do friends have obligations
towards one another? It seems almost misplaced to ask this question, and it is
tempting to answer it by saying that friends have obligations because that’s part of
what friendship is. So far as it goes this answer is completely correct, but it must be
understood as part of a deeper truth. As Aristotle and many others since have
observed, friendship has intrinsic value. It may also have instrumental value, of
course, but if a relationship between two persons was not intrinsically valuable—
and if its intrinsic value did not dominate its instrumental value for each of
them—then it would not be friendship. It is partly constitutive of friendship
that friends hold certain characteristic attitudes towards one another, the most
important of which are affection and respect. But the obligations of friendship—
primarily duties of loyalty and mutual assistance—are themselves partly constitutive
of the relationship.¹⁴ The attitudes, obviously, evolve over time, and the potential
friends must each make choices that permit the friendship to develop. At a certain
point, however, they just become friends, and the duties exist. This is not a matter
of exercising a power of promise or consent. As Dworkin says, it would be perverse
to describe friendship as a matter of assuming obligations; the shared history by
virtue of which friends come to owe obligations is, on the contrary, “a history of
events that attract obligations” (LE at 197). This seems exactly right. But the reason
this happens is that the relationship of friendship, constituted in part by certain
mutual duties, is intrinsically valuable. It might be objected that this is moral
bootstrapping. Doesn’t it amount to saying that the duties exist because it is
good that they exist? Yes, it does. But this kind of bootstrapping is pervasive
and unavoidable in morality. Duties are justified either instrumentally or non-
instrumentally, but in both cases the existence of the duty turns on its relationship
to value. Instrumental duties exist because they promote or contribute to the
creation of independently valuable states of affairs, or are likely to do so. Non-
instrumental duties exist because they are valuable in themselves, or partly con-
stitutive of a relationship that is valuable in itself. In each case, it is part of the
argument for the existence of the duty that it is good that it exist.
Dworkin is of course aware of the relationship between duty and value. One of
the two dimensions of interpretation is success in justifying a normative practice—
the other is success in fitting the practice—and justifications cannot help but

¹⁴ Cf. Raz, supra note 3, at 337.

196 Stephen Perry

invoke value. There is an interesting question about the relationship between asso-
ciative obligations and interpretive practices in general. Dworkin clearly expresses
the view in Law’s Empire that all associative relationships are either interpretive
practices themselves or are formed within broader interpretive practices, but does
he also think that all interpretive practices are (potential) associative relationships?
It is not entirely clear. Whatever Dworkin’s view on this question might have been
at the time that he wrote Law’s Empire, I wish to suggest that associative relation-
ships are a narrower category than interpretive practices generally.¹⁵ Taking friend-
ship as a paradigm, the suggestion is that we can identify a distinctive type of
obligation that arises because it is partly constitutive of some form of limited
human relationship that possesses intrinsic value. The further suggestion is that
Dworkin’s strategy for establishing the possibility of political obligation requires
one to argue that genuine political community—the term is Dworkin’s (LE at
211)—is an intrinsically valuable form of human association. Properly under-
stood, I believe that this is exactly Dworkin’s argument. Before defending that con-
clusion, however, let me say a bit more about associative obligations in general.
Notice, to begin, that this account of associative obligations does not presup-
pose a general interpretive practice of associative obligation, although it is consis-
tent with the existence of such a practice. The account is, in the first instance,
simply a first-order moral claim about the way in which certain kinds of human
relationships may give rise to obligations. Whether or not there is a general inter-
pretive practice of associative obligation is not, therefore, a question that needs to
be answered definitively for present purposes. I earlier argued that the four condi-
tions that Dworkin advances as characteristic of such a general practice are not
found together in every associative relationship. But it is important to emphasize
that those conditions are nonetheless likely to figure prominently in most such
relationships, although no doubt in different ways and configurations. A further,
related point to notice about the account I have offered is that just as it does not
require that there be a general interpretive practice of associative obligation, nei-
ther does it require that there be a single intrinsic value that all such obligations
share. The claim is simply that we can identify certain kinds of obligations that are
in an important sense similar to one another, for the reason that they are partly
constitutive of relationships that are intrinsically valuable. The account is consist-
ent with the idea that a single value underlies all associative obligations, but it is

¹⁵ It is also at least conceivable that associative relationships need not necessarily be, or be formed
within the context of, interpretive practices in Dworkin’s strong sense. For example, is it only possible
for two persons to be friends within a broader interpretive practice of friendship? It is almost imposs-
ible to imagine a human world in which most persons did not have ties of respect and affection with
others, but the interpretive attitude requires a certain self-consciousness about the general practice of
friendship, and, in particular, about what is taken to be its point or value, that might nonetheless be
absent from a particular culture. It is at least arguable that so long as individuals hold the appropriate
attitudes towards one another an intrinsically valuable relationship, together with its associated
duties, comes into being, regardless of whether these or any other persons hold views about the point
or value of friendship. I will not, however, pursue this line of thought further here.
 Associative Obligations and the Obligation to Obey the Law 197

probably more plausible to think that we are dealing with a plurality of values; the
intrinsic value of friendship would thus be different from the intrinsic value of
family, and different again from the intrinsic value, if any, of political community.
As has already been noted, it follows from the account of associative obligation
I have offered—that these are obligations that are partly constitutive of some
intrinsically valuable form of human relationship—that not all interpretive prac-
tices are, either in fact or potentially, associative relationships. There is nothing in
Dworkin’s general account of interpretation that rules out the possibility that the
value or point of some interpretive practices is instrumental, just as nothing rules
out the possibility that the value of some is non-instrumental. If the general char-
acterization is right, moreover, then Dworkin casts the net of associative obliga-
tion too widely: it seems unlikely, for example, that either union membership or
academic colleagueship is an intrinsically valuable form of human relationship.
This general account of associative obligation also makes clear that not all special
obligations are associative. Earlier I pointed to the example of a general duty of easy
rescue that is capable of generating special duties owed to particular individuals.
Although I shall not argue the point here, it seems plausible to think that both the
general and the special duties are justifiable solely on instrumental, consequentialist
grounds. The special duty is thus not an associative one. What about special obli-
gations that arise from the exercise of powers of promise and consent? This is an
interesting case. The power to promise is a power deliberately to place oneself
under new obligations to others, and the power to consent is, roughly, a power
deliberately or knowingly to change one’s normative situation in a broader sense.
(In addition to consenting to assume an obligation, one can, for example, consent
to waive one’s rights.) It would be very implausible to think that every relationship
that arises from the exercise of such powers is one of intrinsic value—consider
commercial contracts, for example—although their exercise can clearly contribute
to, or deepen, relationships that have such value independently; the best example is
probably the vows of marriage. If the relationships are not intrinsically valuable then
neither are the special obligations that constitute them; these obligations are not,
or at least are not necessarily, associative obligations. Even if this is true, however,
it is quite possible that the powers themselves are intrinsically valuable. Joseph
Raz, for example, has argued very persuasively that an important justification for
such powers, although not the only one, lies in the intrinsic value of being able to
create special bonds with others and, more generally, being able to fashion the
shape of one’s own moral world.¹⁶ (Notice that this is another example of the kind
of bootstrapping that I said earlier was unavoidable in morality; these powers exist
because it is good that they do so.) If this is correct then there is an important
affinity between assumed obligations and associative obligations, even though the
former are not necessarily instances of the latter.
¹⁶ J. Raz, Promises and Obligations, in Law, Morality, and Society: Essays in Honour of
H. L. A. Hart 210, 226–228 (P. M. S. Hacker and J. Raz., eds, 1977); J. Raz, The Morality of
Freedom 86–87 (1986).
198 Stephen Perry

I will offer, as a final general observation about associative obligations, a pair of

related points that were implicit in the discussion of the preceding paragraph.
There is no reason why associative obligations cannot coexist with instrumental
obligations within a single kind of relationship, and, indeed, no reason why the
same obligation cannot be justified on both instrumental and non-instrumental
grounds. The second point seems to hold true of families, for example, and pos-
sibly the first does as well. If there is love between a parent and a child then the
relationship between them is intrinsically valuable, and both the parent’s obliga-
tion to care for the child and the child’s obligation to obey the parent are partly
constitutive of that relationship. Both obligations undoubtedly have an important
instrumental justification as well, however, since both are necessary to ensure the
wellbeing and proper development of the child. The parent–child relationship is
one that by its nature is intrinsically valuable but that has, in addition, an import-
ant instrumental dimension. It also seems quite possible that certain relationships
that do not necessarily have intrinsic value, such as those created by contract, may
come to have an element of such value, and thus to give rise to associative obligations
that would not otherwise exist, because of the development of trust and respect
between the parties. Dworkin gives the interesting example of commercial part-
nerships (LE at 200). While I do not think that every such partnership necessarily
gives rise to associative obligations, it is quite plausible to think that at least some
do. It does not follow, of course, that associative obligations that exist within the
context of an essentially commercial relationship either are or should be enforce-
able as a matter of law. Within the legal systems we are generally familiar with the
fact that they are not so enforceable, and probably should not be. But that does
not mean that they do not exist.

III. Political Obligation and Integrity

It is time to return to the issue of political obligation. As was noted earlier,

Dworkin argues that a general obligation to obey the law arises when a legal sys-
tem possesses the special political virtue of integrity. Integrity, we are told,
“requires government to speak with one voice, to act in a principled and coherent
manner towards all its citizens, to extend to everyone the substantive standards of
justice or fairness it uses for some” (LE at 165). As this characterization makes
clear, integrity is a complex ideal which contains a number of different strands.
The core idea, however, is easily enough discerned: the state, acting through its
various institutions and officials, must always strive to ensure that both its laws
and the actions it takes to enforce those laws “express a single and comprehensive
vision of justice” (LE at 134). This idea is, in my view, properly treated by
Dworkin as central because it is connected in two distinct but related ways to the
concept of equality. First, if the state gives effect to a single conception of justice,
then its various officials will be acting, to put the point deliberately vaguely, in an
 Associative Obligations and the Obligation to Obey the Law 199

appropriately similar fashion: litigants can be assured, for example, that their case
will be handled in accordance with the same general principles whether they end
up in courtroom A or in courtroom B. In that sense government will, in
Dworkin’s phrase, “be speaking with one voice.” But equality in this sense is con-
sistent with great substantive inequality: members of one race, for example, might
consistently be treated very differently from members of another race, even
though any given individual would be treated exactly the same way in courtroom
A as he or she would have been treated in courtroom B. Integrity is connected to
substantive equality because it requires not just that the state act in accordance
with a single set of logically consistent standards, but that it act in accordance with
a single conception of justice. In Dworkin’s words, the government must “extend
to everyone the substantive standards of justice or fairness it uses for some.” This is
a tricky matter, because integrity is distinct from justice and hence does not
require that citizens be treated as justice in fact requires. Dworkin is assuming, I
believe, that while true justice necessarily entails equality of concern, equality of
concern does not necessarily entail true justice. This is a strong assumption, but it
has a great deal of plausibility. Dworkin at times speaks somewhat loosely of
integrity as requiring a plausible or coherent conception of equal concern (LE at
213). I think the better way to put the point, however, is to say that integrity
requires actual equal concern, which will take the form of a plausible or coherent
(but perhaps false) conception of justice. Integrity, in other words, is not just a
matter of sincerity or plausibility or coherence, but has genuine moral content.
That content is provided by the concept of equal concern, which occupies a moral
middle ground between simple consistency on the one hand, and true justice, on
the other. At any rate this will be my working assumption for present purposes.
When a legal system has integrity, Dworkin suggests, a “bare” political
community is transformed into a “genuine” political community, and its mem-
bers come under a general obligation to obey the law. Integrity, I have suggested,
is a matter of true equal concern. Equal concern is, in turn, the bridge between
Dworkin’s general analysis of associative obligations and his particular argument
for political obligation. While I have argued in the preceding section that
Dworkin’s general analysis, based on the four conditions, cannot be expected to
apply to associative obligations across the board, it is nonetheless very plausible
to think that the idea of equal concern would play a central role in justifying an
associative political obligation. The concern in question would be expressed by
the state, through its laws and governing actions, towards all its citizens. The
state is not, of course, a real person, from which two points immediately follow.
First, equal concern cannot in this context be a psychological matter; in the case
of political obligation, it must be an interpretive property (LE at 201). Secondly,
as Dworkin explicitly recognizes, the account requires that the state be morally
personified (LE at 167–175). (Recall that it is because political obligation runs
between the state and its citizens, rather than between citizens individually,
that political obligation cannot be personal in the sense defined by Dworkin.)
200 Stephen Perry

Why, then, might integrity make a moral difference of the kind to which
Dworkin points? The reason, I wish to suggest, is that when integrity is present,
the relationship between each individual citizen and the state—the state being
understood as an abstract moral personification of the general political
community—becomes, according to Dworkin’s argument, one that is intrinsi-
cally valuable. He writes that “a political society that accepts integrity as a politi-
cal virtue thereby becomes a special form of community, special in a way that
promotes its moral authority to assume and deploy a monopoly of coercive force”
(LE at 188). Dworkin goes on to describe other justifications for integrity that
are purely instrumental, such as the fact that it provides protection against
partiality, deceit, and corruption. The distinction between the two types of
justification suggests that he regards genuine political community as special
precisely because it has non-instrumental value. But whether or not I am correct
in attributing such a specific view to him, I believe that this idea is the basis of
the most compelling philosophical understanding of his argument for political
obligation.
In order to defend that claim, however, this interpretation of the argument
must be spelled out in greater detail. Why might it be thought that political com-
munity, even when it meets certain conditions pertaining to equal concern, could
be intrinsically valuable? It may well be the case that simply belonging to or iden-
tifying with certain kinds of communities is an intrinsic good for individuals. But
the good of belonging or identification seems unlikely to be enough to support an
obligation to obey. There are, after all, many kinds of communities other than
political ones—for example, ethnic, national, or religious communities—that
seem just as likely to provide such a good, and yet it is far from clear that this fact
can serve as the basis of an obligation to obey the relevant community’s norms.¹⁷
(This is not to deny that there might be other reasons to obey.) The claim must
therefore be, I think, that political community is a unique good, and one which is

¹⁷ Joseph Raz has argued that the law can have intrinsic value because it is one kind of standard of
conduct that a community can have, and “our perception of ourselves, of who we are, depends among
other things on our ability to identify with communities we live in, on our ability to belong to these
communities in the full sense of the word.” J. Raz, On the Nature of Law, 82 Archiv für Rechts
und Sozialphilosophie 1, 10 (1996). (Notice that he does not claim that political community is
unique in this regard.) But Raz also thinks that there is a voluntary aspect to identification and
belonging: “Identification with a community depends on our ability and willingness to accept the
standards which these communities endorse as our own. This ability . . . depends [among other
things] on one’s moral judgement giving basic approval to those standards.” Id. at 12. Elsewhere he
has similarly argued that respect for law, which is an attitude expressing identification with one’s
political community, can ground an obligation to obey that is, like the obligations of friendship,
“semi-voluntary”: “Such an attitude, if directed to a community which deserves it, is intrinsically
valuable. It is not, however, obligatory. One does not have a moral duty to feel a sense of belonging in
a community.” Raz, supra note 3, at 338. The attitude of respect for law is, in Raz’s view, not consent-
based but nonetheless semi-voluntary because, like friendship, it grows and develops over time. There
are obvious affinities between this view and Dworkin’s understanding of political obligation. But
Dworkin does not think that there is a voluntary or semi-voluntary aspect to political obligation, and
that is why he must make the stronger assumption about the value of political community that is
identified in the text.
 Associative Obligations and the Obligation to Obey the Law 201

necessary for individuals to live complete lives or to be completely fulfilled as

human or rational beings; there is intrinsic value for such beings in governing
themselves as a community by means of norms that they create for themselves.
Perhaps one might try to fill out this claim by drawing on the enlightenment ideal
of self-legislation, which Dworkin mentions in passing in Law’s Empire but does
not elaborate upon (LE at 189). I am not sure whether or not this claim is true,
although it has a certain intuitive plausibility. I will not, however, explore the issue
in detail here. The main point I want to make for present purposes is that it seems
that something like the claim I have identified must hold if political community is
ever to have intrinsic value of a kind that does not depend on consent or otherwise
have some voluntary aspect. Of course, the argument cannot be that political
community fulfills or completes human lives no matter what form it takes, if only
for the reason that governments can be thoroughly iniquitous or evil. It is only
genuine political community that can play this role, and that is where integrity
comes in. The basic idea, which again has much intuitive appeal, would appear to
be something like this. A political community, morally personified through the
state, only has intrinsic value for its citizens when it can be understood as treating
each citizen, considered individually, with equal dignity, concern, and respect. It
can only have intrinsic value, in other words, when it meets the demands of
integrity. A legal system without integrity might still have instrumental value—
even a deeply inegalitarian political order might, for example, be morally prefer-
able to social chaos—but it can only make an intrinsically valuable contribution
to human flourishing when integrity is present.
Even if it is true that a genuine political community has intrinsic value for its
citizens, however, we must still ask why it should be the case that citizens have a
general obligation to obey their community’s law.¹⁸ Even if the state treats every-
one with equal concern, and for that reason makes a certain kind of contribution
to human flourishing, why should citizens have an obligation to obey each and
every law, including in particular any laws that are unjust? The response to this
questions is, I believe, unnervingly straightforward. Political community is in part
constituted by norms that have been created by, or otherwise have their source in, a
certain structure of authority within the relevant society. Norms can be said to

¹⁸ Leslie Green raises a related question when he argues that “integrity [in individuals] creates no
duty to obey those who display the virtue,”and this does not change in the case of a personified com-
munity. L. Green, Associative Obligations and the State, in Law and the Community: The End of
Individualism? 93, 106 (A. Hutchinson and L. Green, eds., 1989). Green’s argument is much more
plausible if one assumes, as he at least implicitly does, that integrity cannot confer intrinsic value on a
political community. The analogy with integrity in an individual, which Dworkin himself makes, is
in fact quite misleading: the integrity of an individual cannot be a source of value for others in the way
that the integrity of a political community can, at least conceivably, make that community intrins-
ically valuable for its citizens. If I am right that the essence of integrity in the political context is a con-
cept of equal concern, then there is no analogue with the virtue of integrity in individuals; integrity in
the former sense is a creature of political morality alone. However, even if it is true that political
integrity can be a source of value in the way that I have suggested, it does not automatically follow
that citizens have a general moral obligation to obey the law.
202 Stephen Perry

exist in various ways; there is, for example, a clear sense in which a norm exists
simply because the members of a legislature enacted a law. But the argument for
political obligation that I am attempting to explicate holds that the relationship
between a political community and its citizens is, under certain conditions, intrin-
sically valuable. It seems to me to be implausible in the extreme to think that this
relationship could be morally valuable in that way unless the norms that are par-
tially constitutive of the relationship existed not just in a social or conventional
sense, but in the stronger sense of having the normative force that they purport to
have. They must, in other words, be valid norms, which in this context means
morally valid norms,¹⁹ and the obligations they purport to impose must be real
obligations. This is moral bootstrapping, but bootstrapping is no more problem-
atic here than it is in the case of friendship. Legal norms (and their attendant
obligations) are partially constitutive of the relationship between the state and
the citizen in exactly the same way that the obligations of friendship are partially
constitutive of friendship. It is in this claim, I believe, that the true power of
Dworkin’s argument’s for political obligation resides. Given the non-instrumental
nature of the argument, there is no basis for picking and choosing among laws
in the way that attempted instrumental justifications of a general obligation to
obey almost invariably permit.²⁰ It seems to me that not even non-instrumental
considerations of a different kind, such as considerations of justice, can negate the
obligation, in the sense of justifying the conclusion that it does not exist. (They can
defeat or override the obligation, as I will discuss further below, but that is quite
a different matter.) We are assuming, it should be recalled, that the state treats
all of its citizens with equal concern—since otherwise the law would not have
integrity—and we are also assuming that equal concern is a concept that has genu-
ine moral content. If the argument up to this point is correct, then equal concern is
the only moral property that the law must have if political community is to have
intrinsic value. Political community could not have such value unless the norms
that constitute it are morally valid and hence binding. No other moral considera-
tions are relevant to the existence of the obligation, although they may defeat or
override it.
The point that other moral considerations may defeat or override the general
obligation to obey the law is an absolutely crucial one, since otherwise the conclu-
sion that such an obligation exists will be unacceptable on its face. The point is not,
in fact, a controversial one; almost every contemporary theorist who discusses this
issue assumes that if there is a general obligation to obey the law, which most deny, it
is only prima facie in character. It is perhaps less obvious that the obligation, if it

¹⁹ See supra note 3.

²⁰ Cf. Raz supra note 3; Green supra note 3. Both Raz and Green view the value of law almost
entirely in instrumental terms, although Raz thinks that law can also have a non-instrumental dimen-
sion: see supra note 17. But both cheerfully concede that, precisely because the value of law is
(mainly) instrumental, there can be no general obligation to obey. As Raz puts the point, “much of
the good that the law can do does not presuppose any obligation to obey.” Raz, supra note 3, at 328.
 Associative Obligations and the Obligation to Obey the Law 203

exists, need not be particularly strong. It is completely consistent with the argument
of the preceding paragraph that the obligation is routinely overridden by considera-
tions of justice, or even by instrumental considerations. A related point is that a gen-
eral obligation to obey the law need not, by itself, be strong enough to justify the use
of coercion by the state. As was noted in section I, Dworkin is careful to note that
while obligation is close to a necessary condition for coercion, it is not by itself
sufficient (LE at 191). It should be borne in mind in this regard that not all laws are
enforceable by coercive means, or even enforceable in any legal sense at all; the laws
that govern the relations among different branches of government are only the most
obvious example. It is, again, completely consistent with the argument of the
preceding paragraph that even if there is a general obligation to obey, the coercive
enforcement of any given law can only be justified by reference to instrumental con-
siderations, such as deterrence, or different non-instrumental considerations, such
as retribution. Finally, the claim that there is a general non-instrumental obligation
to obey the law is completely consistent with the possibility that the obligation to
obey many individual laws can also be justified on instrumental grounds, such as the
need for coordination. As was remarked in section II, there is no reason why an
obligation cannot be justified both instrumentally and non-instrumentally.
There is one aspect of Dworkin’s discussion of integrity that I have not
mentioned up to this point, which is captured by the following passage:
Members of a society of principle accept that their political rights and duties are not
exhausted by the particular decisions their political institutions have reached, but depend,
more generally, on the scheme of principles those decisions presuppose and endorse. So
each member accepts that others have rights and that he has duties flowing from that
scheme, even though these have never been formally identified or declared. (LE at 211)
Elsewhere Dworkin goes beyond this claim, arguing that integrity may in fact
require the rejection of certain past political decisions as mistakes; despite
appearances, those decisions never contributed to the content of the law. The
general point that the content of the law may extend beyond past political deci-
sions, or to some extent depart from them altogether, flows from the role that
interpretation plays in Dworkin’s substantive theory of the grounds of law;²¹ it is
a consequence of the claim that there is a tradeoff between fit and justification,
which are the two dimensions of interpretation, rather than a lexical ordering. It
is because interpretation figures in this way in Dworkin’s substantive theory that
considerations of force can have an effect on the grounds of law. As was noted in
section I, within Dworkin’s substantive theory the truth of “It is the law that A”
depends, in part, on whether or not the normative proposition “A” is capable of
being made true by the truth of “It is the law that A.” The theory makes this
determination holistically, by looking to how “A” fits into the content of the law
taken as a whole.²² The upshot of all this is the following. The answer to the
²¹ Interpretation plays a separate but related role in Dworkin’s arguments about methodology in
legal philosophy. See the brief discussion of this issue in section I above. ²² See supra note 6.
204 Stephen Perry

question of whether or not there is a general obligation to obey the law of a par-
ticular legal system depends on whether or not the content of the law, considered
as a whole, has integrity, while the question of what is the content of the law
depends, in part, on whether or not law understood as having this content rather
than that would have integrity (and hence would give rise to a general obligation
to obey). This is complicated but not circular, or at least not viciously so. The
only comment that I will make here on the interpretive aspect of Dworkin’s the-
ory of the grounds of law is that, while it fits naturally with the claim that the
existence of political obligation depends on whether or not the law has integrity,
it is logically independent of that claim. One could make the claim that political
obligation depends on integrity even if one thought that the content of law was
determined solely by social facts. One could make the claim, in other words, as a
matter of pure political philosophy, and hence even if one held views about the
relationship between political and legal philosophy that were quite different
from Dworkin’s own.
Dworkin’s argument that there is a general obligation to obey the law is, at least
in the version I have presented here, obviously controversial. It depends on a num-
ber of strong assumptions. One is the claim that the concept of equal concern has
independent moral content; another is the claim that political community can
have intrinsic value; and yet another is the claim that political community is a
unique good which is necessary for human beings to live complete or completely
fulfilled lives. Any of these assumptions might be wrong. It is also possible that
even if the argument is correct, it does not establish very much of interest because
the conditions it requires will rarely, if ever, be met: it may be, for example, that
the moral property of equal concern cannot be comprehensively attributed to any
existing legal system any more than true justice can. The argument seems to me to
be an important one even so, if only because there is no plausible route other than
a non-instrumentalist one for establishing the existence of a general obligation to
obey the law. Enthusiasts of a predominantly instrumental approach to the value
of law have, for the most part, embraced the conclusion that there can be no
purely instrumental argument for the existence of such an obligation,²³ and in
this I think they are right. They believe that a general obligation to obey the law is
a fiction, and in any event could not have any moral significance; whatever the law
achieves that is of moral value, they argue, it does so whether there is such an
obligation or not. The instrumentalists may be correct. But perhaps it is not
entirely irrelevant that so many judges, legislators and ordinary citizens believe

²³ See supra note 20. John Finnis offers an argument for a general obligation to obey the law that is,
I believe, best regarded as combining both instrumental and non-instrumental aspects. See, e.g.
Finnis, The Authority of Law in the Predicament of Contemporary Social Theory, 1 Notre Dame J. of
L. Ethics, & Pub. Pol. 115 (1984). Finnis’s theory of political obligation is quite different from
Dworkin’s, in ways that it is not possible to elaborate here. I discuss the details of Finnis’s view at some
length in Perry, Law and Obligation, supra note 3.
 Associative Obligations and the Obligation to Obey the Law 205

that they really do have an obligation to obey the law. If the instrumentalists were
right, wouldn’t the news have filtered down by now? Perhaps the tenacity of the
belief in an obligation to obey the law should give them, and the rest of us, some
pause. One of the many merits of Dworkin’s argument is that it gives us a way of
thinking about the matter that helps to illuminate why that belief is as tenacious
as it is.
 9
Law’s Aims in Law’s Empire
John Gardner*

I
Does law have a purpose or point? Surely it does. The trickier questions are these:
Does law have a unifying purpose or point? Does law have a distinctive purpose or
point? Many think that, inasmuch as law has a unifying purpose—such as “the
guidance of conduct”—it is not a distinctive purpose. It is a purpose shared by
many things that are not law. And inasmuch as law has more distinctive pur-
poses—such as “being the final public arbiter of disputes” or “monopolizing the
use of force”—they are not unifying. Each such purpose is the purpose of some
law but not of all law. H.L.A. Hart’s book The Concept of Law is perhaps the best-
known defence of this conjunction of views.¹ Although he accepts that the law has
purposes, Hart advances a non-purposive (and more broadly non-functional)²
account of what legal norms have in common that distinguishes them from other
norms.
Ronald Dworkin belongs to a long tradition of writers who hold, by contrast,
that law has some purpose that is both unifying and distinctive. His book Law’s
Empire is an unusual contribution to this tradition in that it stands for the view
that law must have a unifying-and-distinctive purpose, but it does not commit

* Some of this paper reworks fragments of an earlier and more ambitious (but unpublished) paper
co-authored with Elisa Holmes. Thanks to Elisa for, inter alia, helping to disentangle the two differ-
ent projects that got tangled up in the earlier paper. Thanks to Jules Coleman, Scott Hershovitz and
Maris Köpcke-Tinturé for demonstrating the need to do so. And thanks to all these people for their
many other helpful suggestions, not all of which, alas, have found their way into this final text.
¹ H. L. A. Hart, The Concept of Law (1961). On some of law’s distinctive but non-unifying
purposes, see id. at 38–41. On some of law’s unifying but non-distinctive purposes, see id. at
189–195. See also Hart’s postscript to the second edition of The Concept of Law (2d ed., 1994), at
248–249. (Subsequent references are to the first edition unless otherwise indicated.)
² Not only is law not distinguished by its purposes; it is also not distinguished by its achievements.
For Hart, as for Kelsen, neither intended nor actual effects set legal systems apart from other norm-
ative systems. As Green summarizes the Hart-Kelsen view, law is “a modal kind and not a functional
kind.” L. Green, The Concept of Law Revisited, 94 Mich. L. Rev. 1687, 1711 (1996).

Law’s Aims in Law’s Empire. John Gardner.

© Oxford University Press 2006. Published 2006 by Oxford University Press.
208 John Gardner

itself to a final view on what this purpose is.³ In Chapter 3 of Law’s Empire,
Dworkin provisionally attributes a purpose to law in order to “organize[ ] further
argument about law’s character.”⁴ Even if he has this purpose wrong, he argues,
some unifying-and-distinctive purpose for law must be relied upon if arguments
about the nature of law are to get off the ground. We need “a statement of the
central concept of [our] institution that will allow [us] to see our arguments . . . as
arguments over rival conceptions of that concept.”⁵ To furnish such a “conceptual
statement”⁶ in the case of law, he says, we must find “the most abstract and funda-
mental point of legal practice.”⁷ For law is an “interpretive enterprise”⁸ and this
means that those who are interested in finding out what (else) is true about law
have to begin by taking an “interpretive attitude” to their subject. This in turn
means starting from the assumption that law “has some point” that sets it apart
and brings it together as the particular interpretive enterprise that it is.⁹
Here Dworkin’s argument proceeds transcendentally. Unlike many before him,
he does not try to show that law must have a unifying-and-distinctive purpose by
showing what unifying-and-distinctive purpose law has. Rather, he tries to show
that law must have a unifying-and-distinctive purpose by showing that we cannot
make sense of law without assuming one. The question of which unifying and dis-
tinctive purpose law has can be settled later. Meanwhile we can make do with a
provisional proposal that is pencilled in for the sake of argument.
So what purpose does Dworkin provisionally attribute to law for the sake of
argument? The following purpose, he says, is “sufficiently abstract and uncontro-
versial” to do the job:
Our discussions of law by and large assume, I suggest, that the most abstract and funda-
mental point of legal practice is to guide and constrain the power of government in the fol-
lowing way. Law insists that force [ ⫽ coercion] not be used or withheld . . . except as
licensed or required by individual rights and responsibilities flowing from past political
decisions about when collective force is justified.¹⁰
Dworkin’s Suggested Purpose, or DSP as I will call it for short, is complex and
intriguing. But is it plausible? I think not. At any rate, it is far from uncontrover-
sial. As many have pointed out, a legal system might still exist, and if it did would
still have a purpose or point, in a society of angels. Since the ex hypothesi perfect
population of such a society will be guided by the ex hypothesi perfect laws and
policies of their ex hypothesi perfect government, coercion by that government will
not be needed, and will not be used, to get them to fall into line with law. Nor,
therefore, will law be called upon to regulate government coercion. Yet there will

³ R. M. Dworkin, Law’s Empire (1986) [hereinafter “LE”]. ⁴ LE, 93. ⁵ LE, 92.
⁶ LE, 92.
⁷ LE, 93. For present purposes I assume that nothing turns, for Dworkin, on the distinction
between “law” and “legal practice.” ⁸ LE, 90.
⁹ LE, 47.
¹⁰ LE, 93. The square-bracketed insertion reflects what Dworkin says a few lines later, and is added
here to reduce the volume of quotation.
 Law’s Aims in Law’s Empire 209

still be co-ordination problems to which angelic law may provide the best solu-
tions (what side of the road to drive angelic vehicles on, what frequencies to allo-
cate to angelic cellphones, etc.). So angelic law may still exist and, if it does, it will
still have a purpose—even though there is no coercion by angelic governments for
it to “guide and constrain.”¹¹
There are several things one can learn from this exotic thought-experiment.
The most important lesson is this: That the licensing of government coercion in
ordinary human legal systems, even when successful, is a consequence of a deeper
failure. As Hart captures the point, the resort to legal coercion is a “pis aller,” a
“secondary provision[ ] for a breakdown in case the primary intended peremptory
reasons are not accepted as such.”¹² Whatever other purposes law may have, it
clearly has the purpose of providing law-subjects, including the government, with
normative guidance; that is to say, of subjecting their conduct to the governance
of norms. If only law were fully to succeed in this purpose, if only all law-subjects
were to use legal norms as they are supposed to be used, by being guided by them
qua norms, there would be no need, and no case, for the government to coerce
people into conformity with those same norms. So there would be no need for law
to regulate government coercion. If that much is true, then DSP, even if it is a dis-
tinctively legal purpose, is not a unifying one. It is a purpose of law only on those
occasions when law has failed to achieve its more unifying (albeit less distinctive)
purpose of providing normative guidance for use by its subjects.

II

Although I have just made my own sympathies clear, my interest here is not in
developing this critique of DSP.¹³ I am interested, rather, in an aspect of DSP that
the critique, as it stands, leaves unchallenged. Possibly law’s purpose, even in a
society of angels, is not merely to provide its subjects (including the government)
with normative guidance. Possibly law’s purpose, more specifically, is to provide
its subjects with justified normative guidance. This view has been defended at
length by John Finnis in Natural Law and Natural Rights. To Finnis, the “society of
angels” thought-experiment¹⁴ tends to suggest, not that law lacks a unifying-and-
distinctive purpose, but that law has a unifying-and-distinctive purpose quite
¹¹ Aristotle, Nicomachean Ethics 1179b30–1180b28; Hart, The Concept of Law, supra
note 1, at 195; J. Raz, Practical Reason and Norms 159 (1975); J. Finnis, Natural Law and
Natural Rights 266–267 (1980).
¹² Hart, Commands and Authoritative Legal Reasons in his Essays on Bentham 254 (1982).
¹³ One way to attempt a rescue of DSP might be to argue, with Kelsen, that there are no legal
norms except those that regulate the coercive conduct of state officials. In Kelsen’s words: “Law is the
primary norm which stipulates the sanction.” See General Theory of Law and State 63 (1949).
This view of Kelsen’s has not, however, survived Hart’s critique of it in The Concept of Law, supra
note 1, at 35–41.
¹⁴ Or “world of saints”, as Finnis prefers: Natural Law and Natural Rights, supra note 11,
at 269.
210 John Gardner

different from DSP. The purpose is that of providing co-ordination of conduct for
the common good.¹⁵ This purpose is in two respects more distinctive than that of
merely providing normative guidance to law-subjects. In the first place, according
to Finnis, law aims to provide normative guidance to law-subjects that works in a
distinctive way, namely by co-ordinating their conduct. In the second place, law
aims to provide normative guidance that lives up to a certain standard: it aims to
serve the common good. This second specification is the one that we are concerned
with here. It entails an important feature that Finnis’s suggested purpose shares
with Dworkin’s. For Finnis and Dworkin alike, law aims to guide its subjects prop-
erly, soundly, upstandingly, well. Law aims to be justified in the guidance that it
gives to those whom it aims to guide, such that what they do, when they are guided
by the law and make no mistakes about the law, will itself be justified.
One should be careful not to trivialize this thesis. In all norm-governed pur-
suits, questions of justification arise that are, so to speak, internal to the pursuit.
They are simply questions of whether the norms are being adhered to. In a game
of Monopoly, for example, I land on “Chance” and draw the “Go Directly to
Jail” card. The “banker” refuses to pay me the “salary” I ususally get for passing
“Go.” Is this refusal justified? The answer depends on the rules of Monopoly.
Barring exceptional circumstances (e.g. we are playing for the last space in the
lifeboat, or the “banker” is a cheat), the act of denying me the “salary” has no
significance outside the game, and there is no relevant extra-ludic standard of
assessment. In law, we are sometimes temporarily interested in justification con-
ceived in this purely internal way. On some occasions and for some purposes, we
merely want to know whether what we are doing is justified by the legal system’s
own rules, however quirky and technical. We want to know, for example, what
deadline the law sets for serving a notice to quit following a breach of lease,
because we want to know what the legal consequences will be of our postponing
service of the notice until after rent negotiations are complete. This is a superfi-
cially game-like question. As a lawyer, one may sometimes be tempted to think
of it on the model of Monopoly. But law, as Finnis rightly emphasizes, is not a
game.¹⁶ Everything done by law affects someone, or potentially affects some-
one, in a morally significant way. Thus a question of legal justification always
invites, and never eclipses, a further question of justification. The law sets, let us
suppose, a 14-day deadline for serving a notice to quit following breach of a
lease. But in one’s relations with one’s tenants, should one use this legal deadline
rather than some other? This is not a question about what is justified under the
legal norm. Ex hypothesi, a deadline of 14 days from breach of lease is justified
under the legal norm. Rather, it is a further question about whether one is
morally justified in relying on the legal norm as a justification for one’s action.
Law, on the view I am associating with Dworkin and Finnis, aims to live up to

¹⁵ Finnis, Natural Law and Natural Rights, supra note 11, e.g. at 334–336.
¹⁶ Id. at 305.
 Law’s Aims in Law’s Empire 211

this standard: to be something that its users are morally justified in using. In
Finnis’s terms, law can be “fully understood only by understanding [its] point,
that is to say [its] objective,”¹⁷ which is to make a distinctive kind of contribu-
tion to “practical reasonableness” (this expression being substituted only
because “the term ‘moral’ is of somewhat uncertain connotation”).¹⁸
No doubt Dworkin would resist this (Finnis-influenced)¹⁹ conceptualization
of the contrast between law and game-playing. Dworkin treats the moral ques-
tions inevitably raised by law as bearing (mainly) on the correct interpretation of
legal norms, not (or not mainly) on the propriety of their use once correctly
interpreted.²⁰ This is the second component of the “interpretive attitude” that
law calls for as an “interpretive enterprise.” Law not only “has a point”; law’s
content—what the law of any given jurisdiction says on any given question—is
also “sensitive to its point.”²¹ Thus, for Dworkin, what one is morally justified in
doing in one’s relations with one’s tenants affects (mainly) whether the law
should be understood as setting a 14-day deadline in the first place, not (or not
mainly) whether one is justified in using the legal deadline once the law is so
understood. It follows that in the legal domain, according to Dworkin, there is
normally no such thing as justification “internal to the pursuit” where this
implies leaving the question of moral justification open in determining what the
applicable legal norms are.
I will return to this issue later. For the time being, we need only note that
Dworkin’s famous view about the determination of legal content (the second
component of his “interpretive attitude”) does not drive any wedge between him
and Finnis on the question of law’s purpose or point (the first component). Finnis
and Dworkin alike suggest that law has a justificatory purpose. Moreover, for both
of them it is a morally justificatory purpose. Theirs is not the easy-to-accept thesis
that law is a normative pursuit and hence cannot but provide standards of justi-
fication of some kind. Theirs is the more substantial and contentious thesis that
law aims to provide standards of moral justification, and hence to be morally justi-
fied in the norms it provides. This thesis leaves open whether the aim in question
is to be served by interpreting laws morally, or by not following immoral laws, or
perhaps sometimes by interpreting laws morally and sometimes by not following
immoral laws. This disagreement can be bracketed for now. Our interest for the
time being is only in the Finnis-Dworkin “plateau”²² of agreement, which seems
to extend to the following thesis, if no further:
(␣) Law aims to be morally justified.
In formulating (␣) I have chosen “aims” rather than “has the purpose” or “has the
aim,” not only to save words, but also to avoid giving the impression that,
for either Finnis or Dworkin, (␣) fully captures law’s unifying-and-distinctive

¹⁷ Id. at 3. ¹⁸ Id. at 15. ¹⁹ Id. e.g. at 354–362. ²⁰ LE, 47–48.

²¹ LE, 47. ²² LE, 93.
212 John Gardner

purpose. For both of them (␣) captures no more than one (unifying but not
distinctive) aspect of law’s unifying-and-distinctive purpose.²³

III
Is Dworkin committed to (␣)? (␣) is entailed by DSP, which is Dworkin’s own “con-
ceptual statement” about law. But recall that DSP is only mooted provisionally by
Dworkin, as a working assumption to help us “organize[ ] further argument about
law’s character.” Does (␣) represent an equally provisional aspect of Dworkin’s
thought? Does he align himself with (␣) only for the sake of argument? Or would
(␣) also be entailed by every other suggested unifying-and-distinctive purpose for
law that Dworkin would be prepared to entertain as an alternative to DSP?
There is much in Law’s Empire to suggest that the answer is yes. Chapters four
to seven of the book are devoted to exploring the question: Which moral ideal is
the proper moral ideal for law (through its practitioners and officials) to aim at?
Dworkin famously answers: “integrity.”²⁴ His defence of this answer is conducted
on the footing that, while the choice of integrity as the proper moral ideal to be
aimed at is not itself conceptually determined—“law as integrity” is but one pos-
sible “conception” of law—nevertheless, it is part of the concept of law that law
aims at some moral ideal. Since it is (in turn) part of the concept of a moral ideal
that whatever aims at a moral ideal aims to be, at the very least, morally justified, it
seems that, for Dworkin, any acceptable conception of law is one that paints law
as aiming to be, at the very least, morally justified. So, in these subsequent chap-
ters of Law’s Empire, (␣) seems to be endorsed by Dworkin and not just mooted
for the sake of argument.
Yet there are also conspicuous themes in Dworkin’s work that seem to militate
against his endorsing (␣). In the following three sections I will consider three of
these themes.

IV

In recent work, Dworkin has expressed scepticism about the personification of

law. His scepticism is expressed in the context of a critique of Joseph Raz’s work on
the nature of law. What can Raz mean, asks Dworkin, when he says (as he often
does) that “law claims legitimate authority”? Dworkin answers:

²³ This opens up the possibility that thesis (␣) is also endorsed by Hart in The Concept of Law,
supra note 1, at 186–188. But Hart keeps his options open on this point, and in the postscript to the
second edition of The Concept of Law, supra note 1, at 248–249, he explicitly disclaims any attach-
ment to (␣). Law is not a game, but the difference, for Hart, does not lie in law’s aim. It lies, inter alia,
in law’s claim. See Section IV below on aims and claims.
²⁴ LE, 244. Hercules, the ideal judge “follows law as integrity.”
 Law’s Aims in Law’s Empire 213
This type of personification is often used in philosophy as a shorthand way of stating the
meaning or content of a class of propositions. A philosopher might say, for example, that
morality claims to impose categorical requirements, or that physics claims to reveal the
deep structure of the physical universe. He means that no proposition is a true proposition
of morality unless it accurately reports categorical (rather than only hypothetical) require-
ments or that no proposition is a true proposition of physics unless it correctly reports
physical structure. If we read Raz’s personification in this familiar way, we take him to
mean that no proposition of law is true unless it successfully reports an exercise of legit-
imate authority. But that would imply not that morality cannot be a test for law, as Raz
claims, but that it must be a test for law because, as he recognizes, no exercise of authority is
legitimate “if the moral or normative conditions for one’s directives being authoritative are
absent.”
It is difficult to find a sensible alternative reading of Raz’s personification. He sometimes
suggests that when he says “law” claims legitimate authority he means that legal officials
claim that authority; legal officials do this when they insist that they have a “right” to
impose obligations on citizens and that these citizens “owe them allegiance” and “ought to
obey the law.” It is one thing to suppose that legal officials often make such claims; it is
quite another to suppose that unless they make such claims, there is necessarily no law. In
fact many officials do not. Oliver Wendell Holmes, for example, thought the very idea of
moral obligation a confusion. He did not suppose that legal enactments replace the ordin-
ary reasons people have for acting with some overriding obligation-imposing directive, but
rather that these enactments add new reasons to ordinary ones by making the cost of acting
in certain ways more expensive. Whether a community has law does not depend on how
many of its legal officials share Holmes’ views. So we cannot make sense of Raz’s crucial
personification by supposing it to refer to the actual beliefs or attitudes of officials.²⁵
If sound, these criticisms tell against an important thesis that Raz implicitly
defends as a rival to (␣). This rival thesis is entailed by the thesis that law claims
legitimate authority. It says:
(␤) Law claims to be morally justified.
Dworkin’s criticisms of Raz on authority tell against (␤) because (␤) can be true
only if law is the kind of thing that can make claims, and in the quoted passage
Dworkin denies that law is that kind of thing. But should he deny it?
One may readily agree with Dworkin, as Raz does, that it takes a human being
to make a claim. Raz makes tolerably clear that, for these purposes, the relevant
human beings are legal officials.²⁶ Legal officials, he argues, are those who make
the claim mentioned in (␤). Yet at the same time Raz’s attribution of the claim to
law itself is not elliptical. It is not shorthand for “legal officials claim that law is
morally justified.” That is because legal officials, according to Raz, make the claim
mentioned in (␤) on law’s behalf. And that in turn is because it is part of the con-
cept of a legal official that, when someone acts as a legal official, she acts on law’s
behalf. So if (␤) is sound, then someone who does not claim moral justification for

²⁵ Dworkin, Thirty Years On, 115 Harv. L. Rev. 1655, 1666–1667 (2002).
²⁶ J. Raz, Ethics in the Public Domain 215–216 (paperback ed., 1995).
214 John Gardner

what she does is not, in doing it, acting as a legal official, for (in failing to make the
claim that law necessarily makes) she is no longer acting on behalf of the law. That
is the most important pay-off of (␤), and it holds only if the claim in (␤) is under-
stood (non-elliptically) to be law’s own claim, and not merely the claim of some
human beings.²⁷ It takes a human being to make a claim but it does not follow
that human beings are the only things that make claims. Law makes claims
through human beings acting on its behalf.
These remarks about (␤) matter for present purposes because they apply,
mutatis mutandis, to (␣) as well. True, law cannot make any claims except through
human beings acting on its behalf, i.e. legal officials. But by the same token law
cannot have any aims except through human beings acting on its behalf. For (␣) to
be non-elliptically true, there must equally be human beings who, on law’s behalf,
aim at the law’s being morally justified, and who are acting on law’s behalf (i.e.
count as legal officials) only if they aim at the law’s being morally justified.
Inasmuch as (␤) is offered by Raz as a rival to (␣), there must obviously be some
objections to (␤) that are not objections to (␣). But Dworkin’s way of rejecting (␤),
in the passage just quoted, would also commit him to rejecting (␣). Consider what it
means to make a claim to moral justification. Minimally, it means acting with the
aim that one be taken to be morally justified. Legal officials—those acting on behalf
of the law—make a claim to moral justification only if they aim that the law should
be taken to be morally justified by those law-subjects to whom it is addressed. This in
turn means that they aim either that law be morally justified or that it be mistaken for
something that is morally justified. It is the second possibility—the possibility of a
pretence or masquerade on the part of law and legal officials—that distinguishes (␤)
from (␣). Some things that would not count as law according to (␣)—because they
do not aim to be morally justified—still count as law according to (␤)—because they
masquerade as having that aim. Raz is careful to point out that, as compared with law
that lives up to the standard enunciated in (␣), law that merely masquerades as doing
so is a less central case of law.²⁸ Just as the central case or paradigm of anything that
has an aim is (ceteris paribus) the case in which it succeeds at that aim, so the central
case or paradigm of anything that makes a claim is (ceteris paribus) the case in which
it makes that claim sincerely. So the central case of law, according to Raz’s criterion
(␤), is the case of law that is morally justified. Relative to morally justified law, law
that merely aims to be morally justified but does not succeed in that aim is a less cen-
tral case of law. And relative to law that aims to be morally justified (whether or not it
succeeds in that aim), law that merely pretends to have that aim but does not really
have it is in turn a less central case of law. It is a highly deviant case of law. Yet the
claim that is present in the deviant case is also present in the central case. That is one
of the features that, for Raz, brings them all together as cases of law. Although there
may be law that does not have the aim mentioned in (␣), there is no possible law that
does not make the claim mentioned in (␤).

²⁷ Cf. P. Soper, Law’s Normative Claims in The Autonomy of Law (R. P. George ed., 1996).
²⁸ Ethics in the Public Domain, supra note 26, at 270.
 Law’s Aims in Law’s Empire 215

If we understand Raz’s thesis (␤) in this way, then Dworkin’s proposed

counterexample—Justice Holmes’ rule-sceptical attitude to law—serves as a
counterexample to (␤) only if it also serves as a counterexample to (␣). If truth be
told, it does not really serve as a counterexample to either (␤) or (␣). Admittedly,
in his extrajudicial writings, Holmes peddled the mistaken view that legal rules
cannot impose moral obligations on law-subjects but can only give them pruden-
tial incentives.²⁹ But were Holmes’ arguments and pronouncements from the
bench consistent with this view? Didn’t he sometimes use legal rules to help justify
his own legal rulings, thereby establishing that legal rules must be more than mere
incentives? And in the process didn’t he sometimes treat himself as bound to use
legal rules in justifying his legal rulings, thereby confirming that at least some legal
rules are rules of obligation? And, even when not, didn’t he at least insist that his
own legal rulings were morally obligatory, never mind why he made them? Or at
any rate didn’t he insist that his own legal rulings were morally justified even if not
morally obligatory? Or at the very least didn’t he insist on the moral justification
of the very ruling he was just in the process of making? It seems to me that, as a
legal official, Holmes could not but insist on the moral justification of the very
ruling he was just in the process of making. If he spoke from the bench in a way
that suggested that he did not insist on this, then it seems to me that he was not
speaking on behalf of the law—i.e. in his official capacity—when he did so. For
better or worse, he was replacing his official position with his personal position.
But even if you doubt all this, ask yourself: What, in Holmes’ judicial decisions,
could possibly count as a denial of the claim to moral justification that would not
equally count as a denial of the aim of moral justification? How could anything in
Holmes’ judicial decisions possibly serve as a counterexample to (␤) without
equally serving as a counterexample to (␣)? For the only relevant difference
between (␤) and (␣) is that (␤) allows for the extra possibility of officials who,
speaking on behalf of the law, pretend to be acting exactly as they would be acting
if only (␣) were true.
In invoking Holmes, therefore, Dworkin seems to be distancing himself not
only from Raz’s thesis that law makes a moral claim, but also from the thesis, appar-
ently so central to Law’s Empire, that law has a moral aim: that law’s distinctive-
and-unifying purpose is a morally justificatory purpose.³⁰

Perhaps more importantly, Law’s Empire itself gives us reason to doubt whether
Dworkin accepts, or could accept, (␣). Consider Dworkin’s statement of the first
assumption made by those who take the “interpretive attitude” to law. They

²⁹ O. W. Holmes, The Path of the Law, 10 Harvard Law Review 457, 461–462 (1897).
³⁰ Contrast LE, 172–175, where Dworkin seems to suggest that the law has a moral voice in which
officials, acting as officials, cannot but speak.
216 John Gardner

assume, he says, “that the practice [of law] does not simply exist but has value, that
it serves some interest or purpose or enforces some principle—in short that it has
some point—that can be stated independently of just describing the rules that
make up the practice.”³¹ There is some equivocation in this remark. To adopt the
interpretive attitude, says Dworkin, we need to assume that law serves a purpose.
Does this mean that law has a purpose? Or does it mean that law achieves that pur-
pose? The two claims are not, as they stand, incompatible. Many things have a
purpose that they also achieve. But the two claims are incompatible as soon as they
are understood as “conceptual statements” about law. That is because purposive
agency is agency that leaves open the logical possibility of failure. If it is impossible
to classify the actions of a certain agent into the categories “failure” and “success”
(because failure is conceptually ruled out) then that agent is not a purposive agent.
It does not have any purposes. And if a certain action cannot be classified as a fail-
ure or success (because failure is conceptually ruled out) then that action is not an
action with a purpose. And if a certain action cannot be classified as a failure or
success in its possession of a certain property (because failure to possess that prop-
erty is conceptually ruled out) then having that property cannot figure among the
purposes of that action. Accordingly, if law is such that whatever it does is morally
justified—if its being morally justified is part of what it means for it to count as
law—then it cannot at the same time aim at (have among its purposes that of )
being morally justified.
In short, thesis (␣) cannot possibly be true if the following rival thesis about law
is true:
(␹) Law is morally justified.
If Dworkin accepts (␹) then he cannot accept (␣).³² Moreover, he cannot accept
any “conceptual statement” for law of which (␣) is an implication. So he must also
abandon DSP, the unifying-and-distinctive purpose that he provisionally attaches
to law in Chapter 3 of Law’s Empire. Since by (␹) there is no logical space for law to
fail in the moral justification of state coercion—since law’s own moral justification
in doing whatever it does is conceptually secure—morally justifying state coercion
cannot be an aim or purpose of law. It can of course be an aim or purpose of some
people writing about law, such as Dworkin himself. Such law-favourers may aim
to morally justify (all) law in order to support the case for (␹). My point is only
that law itself (acting through its officials) cannot aim to be morally justified if (␹)
is true, whereas law itself (acting through its officials) cannot but aim to be morally
justified if (␣) is true.
Does Dworkin accept (␹), and hence eschew (␣)? Some other passages in Law’s
Empire maintain the same equivocation between (␹) and (␣) that afflicts

³¹ LE, 47.
³² Contrast J. Dickson, Evaluation and Legal Theory 106 (2001). Dickson allows Dworkin to
endorse both (␣) and (␹) together, and seems to see them as natural bedfellows.
 Law’s Aims in Law’s Empire 217

Dworkin’s statement of the “interpretive attitude” above. Consider the following

remarks on Nazi law in Chapter 3:
We need not deny that the Nazi system was an example of law, no matter which interpreta-
tion we favour of our own law, because there is an available sense in which it plainly was
law. But we have no difficulty in understanding someone who does say that Nazi law was
not really law, or was law in a degenerate sense, or was less than fully law. For he is not then
using “law” in that sense; he is not making that sort of preinterpretive judgment but a skep-
tical interpretive judgment that Nazi law lacked features crucial to flourishing legal systems
whose rules and procedures do justify coercion.³³
The words “do justify coercion” at the end of this passage might tempt one to sup-
pose that Dworkin means to commit himself to (␹) rather than (␣). By his “con-
ceptual statement” for law (DSP) he meant that whatever counts as law does
morally justify coercion, not that it has the purpose of doing so. And yet the pre-
ceding sentence decisively rules out this reading of the passage as a whole.
Dworkin sees morally justified law only as the central case or paradigm of law.
This is incompatible with his endorsing (␹). It is compatible with his endorsing
(␣) or (␤). It is also conducive to his endorsing either (␣) or (␤). For as we already
mentioned, both of these proposed theses share the implication (explicitly
endorsed by both Raz and Finnis) that morally justified law is the central case of
law, and hence that Nazi law is degenerate qua law. That Dworkin thinks there is
an available sense in which Nazi law “plainly” was law suggests that, as between
(␣) and (␤), he actually ought to favour Raz’s (␤). For there is ample evidence to
suggest a mere pretence of moral rectitude on the part of many Nazi officials, such
that Nazi law made moral claims that were not matched by genuine moral aims.
Be that as it may, the passage as a whole certainly militates against the view that
Dworkin accepts (␹). Morally justified law only represents law’s central case. So
“morally justified law” is not a tautology and “morally unjustified law” is not an
oxymoron. Both kinds of law are conceptual possibilities. So perhaps there is, after
all, nothing here to cast doubt on Dworkin’s allegiance to (␣). What seem to be
statements endorsing (␹), and hence denying (␣), are on closer inspection state-
ments denying (␹).

VI

Yet the problem of reconciling (␣) with other themes of Law’s Empire is not yet
over. Consider now Dworkin’s statement of the second assumption made by those
who take the “interpretive attitude” to law. They assume, he says, “that the
requirements of [law]—the behavior it calls for or judgments it warrants—are not

³³ LE, 103–104.
218 John Gardner

necessarily or exclusively what they have always been taken to be but are instead
sensitive to its point, so that the strict rules must be understood or applied or
extended or modified or qualified or limited by that point.”³⁴ Here we find traces
of the same equivocation between (␣) and (␹) that ran through Dworkin’s state-
ment of the first assumption, quoted above. If (␹) is true then the law’s “require-
ments,” the “strict rules,” are already morally justified and officials need not alter
them to make them so. If (␣) is true, on the other hand, then there may be
“requirements” or “strict rules” of law that are not already morally justified, in
which case legal officials, acting on behalf of the law, have the (␣)-given aim of
transforming them into morally justified requirements or rules. The words
“extended or modified or qualified or limited” plainly support the second reading
of the passage, and hence tend to confirm Dworkin’s allegiance to (␣) over (␹).
Legal officials have the (␣)-given aim of taking morally unjustified legal norms
(“strict rules”) and changing them (by extension, modification, qualification or
limitation) into morally justified legal norms. So there can be morally unjustified
legal norms that call for legal officials to change them with the (␣)-given aim of
making them morally justified. (␹) is false because (␣) is true.
The words “understood or applied,” on the other hand, can be read consistently
with either (␣) or (␹). Understanding and applying legal norms might be oper-
ations that legal officials perform upon legal norms with the aim of improving
them, i.e. with the aim of transfoming them into morally justified norms on the
occasions when they are not morally justified as they stand. On this reading of
“understood or applied,” (␹) remains false, for there can be morally unjustified as
well as morally justified legal norms. Yet one may also read the words “understood
or applied” consistently with (␹), and hence inconsistently with (␣). No legal
norm is morally unjustified, and therefore no legal official can possibly have the
aim of changing a legal norm from a morally unjustified one to a morally justified
one. The challenge for legal officials under the heading of “understanding and
applying” the “strict rules” is only to bring out what is already in these rules, i.e. to
understand each legal norm as the morally justified norm that it already is, and to
apply it accordingly.
Dworkin’s account of how judges should go about understanding and applying
the law—namely, by “constructively interpreting” it—is usually read in the latter
way, as an account according to which the law already means what it should mean
(i.e. what it would mean if it were morally justified) and hence only needs to have
its meaning brought out, not altered, by judges. This is what I referred to above as
“Dworkin’s famous view about the determination of legal content.” But on closer
inspection Dworkin’s official characterization of “constructive interpretation” in
Chapter 2 of Law’s Empire leans very strongly in the opposite direction:
Roughly, constructive interpretation is a matter of imposing purpose on an object . . . in
order to make of it the best possible example of the form or genre to which it is taken to

³⁴ LE, 47.
 Law’s Aims in Law’s Empire 219
belong. It does not follow, even from that rough account, that an interpreter can make of
[the object] anything he would have wanted it to be, that a citizen of courtesy who is
enthralled by equality, for example, can in good faith claim that courtesy actually requires
the sharing of wealth. For the history and shape of a[n] . . . object constrains the available
interpretations of it . . . . Creative intepretation, on the constructive view, is a matter of
interaction between purpose and object.³⁵
This passage has an odd start, when set against other passages in the same chapter.
Dworkin suggests that a purpose needs to be imposed on an object by a construct-
ive interpreter. If this were true it would compete with the thought that law has a
purpose. Perhaps Dworkin only means that the interpreter has to work out what
the purpose in question is? Or perhaps Dworkin means that, given that law has
among its purposes that of being morally justified, the interpreter still needs to do
the work of adjudicating between various different moral ideals for law (“concep-
tions”) to decide which particular laws would be morally justified ones and why?
Either way, the word “impose” seems ill-suited to capture how the interpreter is
supposed to relate to his or her object.
That constructive interpreters “impose” purpose on the object before them
seems, then, to be an infelicity in Dworkin’s formulation.³⁶ That constructive
interpreters thereby “make [something] of ” the object before them seems, how-
ever, to be an accurate statement of what Dworkin has in mind. There is always
an object of interpretation and the aim of the constructive interpreter is to
improve it, transform it into a better object of the same kind. In legal contexts,
thanks to the truth of (␣), that means a morally better object, a legal norm that
comes closer to conforming to the proper moral ideal for law, whatever that
ideal may be. In the process, there must of course be some preservation of some
aspects of the norm one started with. If there is nothing at all left of that norm
then what one did to it cannot count as interpreting it. But there must also be
some improvement. If one did not improve the norm one started with, one’s
interpretation of it was not constructive.
That seems to be the sense in which, according to Dworkin, constructive inter-
pretation is a matter of “interaction between purpose and object.” There is some
continuity in the object but there is also some improvement, some gravitation
towards the ideal that properly pertains to an object of that type. The key
Dworkinian contrast seems to be that between constructive interpreters, who aim
to give the same object a new and better meaning, and what Raz has called “con-
serving” interpreters, who aim to retrieve some meaning that the object already
has or has had.³⁷ Both differ from non-interpreting improvers—Dworkin calls
them “pragmatists”—who are simply inclined to replace the object outright with a
new and better one.

³⁵ LE, 52.
³⁶ See also J. Finnis, On Reason and Authority in Law’s Empire, 6 Law and Philosophy 357,
359–360 (1987).
³⁷ J. Raz, Interpretation without Retrieval in Law and Interpretation (A. Marmor ed., 1996).
220 John Gardner

If this reading is correct, then “understanding” and “applying” the law, when
tackled constructively, are also ways of “extending or modifying or qualifying or
limiting” the law. The law is not left as it was by the interpreting judge. This con-
clusion is crucial to the success of Law’s Empire if it is to be read as a defence and
explication of (␣). Throughout the book, Dworkin uses the situation of judges to
illustrate the force of (␣). But if judges are to have the aim, on behalf of the law,
that law be morally justified, there must be possible morally unjustified legal
norms for them to have and pursue this aim in respect of—morally unjustified
legal norms that they can render morally justified by their improving interpret-
ative interventions.³⁸
The most striking implication of this, if true, is that the Dworkin of Law’s
Empire has no significant axe to grind with the legal positivist tradition. In order
to continue grinding this axe, the Dworkin of Law’s Empire turns legal positivists
into advocates for “conserving” interpretations of legal norms. He thereby creates
an issue between himself and legal positivism, namely the issue of constructive
versus conserving interpretation. But none of the major recent writers commonly
thought of as legal positivists (Kelsen, Hart, Raz, Coleman) shows any predilec-
tion for conserving interpretations of legal norms.³⁹ The thing they all agree on,
what unites them as legal positivists, is that the law is made by legal officials, such
that if one wants to know what the law says on a given subject in a given jurisdic-
tion one needs to investigate what those officials did or said, not what they ought
to have done or said. The law is made up exclusively of norms that have been
announced, practised, invoked, enforced, or otherwise engaged with by human
beings acting on law’s behalf. We can represent this thesis as:
(␦) In any legal system, the law is made up of norms which are part of the law
only because some legal official engaged with them, and such an agent is
a legal official only because, by engaging with norms in certain ways, he
or she can make them part of the law.⁴⁰

³⁸ Cf. Nicos Stavropoulos’s claim that Dworkin “need not accept that any legal norms or rules are
individuated non-interpretively”: Interpretivist Theories of Law, Stanford Encyclopaedia of
Philosophy, ⬍http://plato.stanford.edu/entries/law-interpretivist/⬎. What does Stavropoulos mean?
Perhaps all he means is that each new constructive interpreter comes to the norm as already shaped
by some previous constructive interpretation of it. This is false (someone has to be the first judge
to read the first constitution, the first statute, etc). But even if it were true it would be trivial. Relative
to each act of interpretation there would still surely be a pre-interpretive norm, viz. one that was out-
put by a previous interpreter. Dworkin must accept that these pre-interpretive norms exist, or else an
act of constructive interpretation has norms neither as inputs nor as outputs, and so is in no sense
normative.
³⁹ Furthermore, Hart and Raz explicitly renounce any such predilection: see Hart, The
Concept of Law, supra note 1, at 200–202; Raz, Why Interpret?, 9 Ratio Juris 349, 360–363
(1996).
⁴⁰ For a perhaps more familiar formulation of (␦), see J. Gardner, Legal Positivism: 51⁄2; Myths, 46
American Journal of Jurisprudence 199 (2001).
 Law’s Aims in Law’s Empire 221

In Chapter 1 of Law’s Empire Dworkin promises that he will be arguing against

this “plain fact” view of law, as he calls it, the view that “what the law is in no way
depends on what it should be.”⁴¹ This suggests that he will be arguing against
thesis (␦). But the subsequent chapters of the book do not fulfil the promise.
Instead they argue, consistently with (␦), that judges should improve the law by
constructively interpreting it. It is true that, when the law is improved by con-
structive interpretation, the re-interpreted norms are part of the law because they
are (held by the judge to be) morally justified, or as close to morally justified as
they can be made. If the parenthetical words “held by the judge to be” are sup-
pressed in this statement, it sounds like an attack on (␦). It seems to make “what
the law is” depend on “what it should be.” But the parenthetical words cannot be
suppressed. For according to Dworkin’s own explicit characterization of construc-
tive interpretation, it is not a constructively interpreted norm’s being morally
justified that turns it into law. Rather, it is the “plain fact” of an official’s doing
something to a norm (viz. interpreting it constructively) that turns it into law.
Remember that constructive interpretation is an “interaction between purpose
and object.” There must be a human being—a legal official—who has the purpose
in question. There must also be an object—a legal norm—to which the purpose is
applied. The law is made by the interaction of the two, by the official’s engaging
with the norm as its interpreter, with the aim of yielding up a morally improved
legal norm. If that is the picture he has in mind, then Dworkin’s war with the legal
positivist tradition is over. He has no quarrel with (␦). He would have a quarrel
with (␦) only if the purpose of law were self-fulfilling, and needed no agent, no
constructive interpreter, to carry it out on law’s behalf. In that case, as I already
explained above, it would be wrong to think of it as law’s purpose, for there would
be no logical possibility of failure on the part of the law.
Thesis (␣), to put it another way, is perfectly compatible with thesis (␦).
Indeed, thesis (␣) presupposes thesis (␦).⁴² Dworkin himself shows why in his
more recent critique of Raz, quoted at length above.⁴³ For law to do things there

⁴¹ LE, 7. Strictly speaking, the words “in no way” make this view slightly more restrictive than (␦).
Some who endorse (␦), including Hart, understand it to be compatible with there being moral tests of
legal validity if (but only if ) those moral tests have been engaged with (announced, used, etc.) by legal
officials: see Hart, The Concept of Law, supra note 1, at 70–71. Elsewhere on the same page
Dworkin places a different restriction on the “plain fact” view. He formulates it as the view that “the
law is only a matter of what legal institutions have decided in the past . . . So questions of law can
always be answered by looking in the books where the records are kept.” Such a view may have been
held by Austin and maybe even Bentham but it is rejected by Kelsen, Hart, Raz and Coleman, all of
whom recognize the using of a norm, as well as the announcing of it, as a possible way of engaging
with it so as to make it into law. I am assuming that Dworkin means the “plain fact” view to be one
that could be taken by Kelsen, Hart, Raz and Coleman—in other words, that, in spite of his ambiguities,
he means it to correspond to (␦).
⁴² A point that is accepted and emphasized by Finnis. See Natural Law and Natural Rights,
supra note 11, at 232; also The Truth in Legal Positivism in The Autonomy of Law, supra note 27.
⁴³ See supra text accompanying note 25.
222 John Gardner

must be human beings who do those things on behalf of law. (␣) emphasizes that
legal norms are made (and modified) by people, for by (␣) legal norms are made
(and modified) with a certain aim, viz. that they be morally justified. If there are
any legal norms that are not made (or modified) by human beings then they can-
not have that or any other aim, so they would count as counterexamples to (␣). So
if Law’s Empire is a defence of (␣), it should also be read as a defence, albeit a back-
handed defence, of thesis (␦), which is the thesis that unites members of the legal
positivist tradition.
Of course the converse does not hold. One may be a legal positivist—a (␦)-
endorser—who does not endorse (␣). Raz is an example. As we saw, he endorses
(␤) over (␣). The difference, recall, is that (␤) allows for legal officials who only
pretend to aim that the law be morally justified: on law’s behalf they make the
claim to be morally justified but do not really have the aim. (␤) makes conceptual
space for such pretence, so that Nazi law is law, and Nazi officials are legal offi-
cials, even if many of them are only pretending to aim at moral justification. Of
course, while it makes logical space for it, (␤) does not make any moral space for
such pretence. (␤) is a thesis about law, not a thesis about the moral duties of
legal officials. Morally, as Raz says, judges should aim that the law they make or
modify be morally justified.⁴⁴ That they should have this aim does not entail
that, whenever they speak on behalf of the law, they do have this aim. It does not
entail that (␣) rather than (␤) is true, any more than it entails the opposite. It is
an independent problem. Nevertheless it is a problem that one confronts only if
one endorses (␣) or (␤) rather than (␹). For if (␹) is true then there is no logical
space for judges to act with the aim that the law they make or modify be morally
justified. For all the law they engage with is morally justified already, and their
only job is to bring this fact out.

VII

Of the three rival theses about law that we encountered, namely:

(␣) Law aims to be morally justified;
(␤) Law claims to be morally justified;
(␹) Law is morally justified;

⁴⁴ Sometimes, on Raz’s view, this moral imperative should carry judges beyond interpreting
the law to make it morally better. Sometimes they should effect moral improvements to the law in
ways that are more radical than mere interpretation would allow. They should sometimes decline to
follow the law. See, among many discussions spread across Raz’s work, The Authority of Law
(1979), ch. 10; Ethics in the Public Domain, supra note 26, ch. 14, and Incorporation by Law, 10
Legal Theory 1 (2004).
 Law’s Aims in Law’s Empire 223

the first two are legal-positivist theses.⁴⁵ They presuppose:

(␦) In any legal system, the law is made up of norms which are part of the law
only because some legal official engaged with them, and such an agent is a
legal official only because, by engaging with norms in certain ways, he or she
can make them part of the law.
Only (␹) is hard to reconcile with (␦). If law is made by people, then law is vulner-
able to moral error, for people are vulnerable to moral error. So if law is incapable
of being morally unjustified, as (␹) tells us it is, how can it be made by people?
There may be some who think that the value of having norms made by people is
such as to justify them morally, no matter how (otherwise) morally abhorrent are
the norms that are made. Surely some law, even Nazi law, is at least preferable to
chaos? This is a deeply unattractive moral position.⁴⁶ At any event, it is not
Dworkin’s position. Dworkin’s position is either that the law is morally justified
and hence not made by people (i.e. (␹) is true and therefore (␦) is false) or that law
aims to be morally justified and hence is made by people who pursue that aim on
behalf of the law (i.e. that (␣) is true and therefore (␦) is true). Law’s Empire leaves
us puzzled by suggesting that (␣) is true but (␦) is false. This is an impossible con-
junction of views. Much of the book suggests that we should hold Dworkin to (␣)
and accordingly discount his expressed (and widely-advertised) opposition to (␦).
That being so, we should be pleased to welcome Dworkin back into the best tradi-
tion of thinking about law, which is the legal positivist tradition.

⁴⁵ Robert Alexy has devoted much energy to showing that (␤), which he calls “law’s claim to moral
correctness” is incompatible with legal positivism. But this is only because he holds legal positivism to
the thesis that there is no necessary connection between law and morality. Thesis ( ␤) is clearly incon-
sistent with this, since it states a necessary connection between law and morality.
⁴⁶ A more common position is that the positivity of legal norms—the fact that they are made by
people—automatically lends some redeeming value to legal norms even when they are (otherwise)
morally abhorrent. I also find this position unattractive. In my view there can be immoral laws that
have absolutely no morally redeeming features.
 10
How Facts Make Law
Mark Greenberg*

I. Introduction
Nearly all philosophers of law agree that non-normative, non-evaluative, contin-
gent facts—descriptive facts, for short—are among the determinants of the content
of the law. In particular, ordinary empirical facts about the behavior and mental
states of people such as legislators, judges, other government officials, and voters
play a part in determining that content. It is highly controversial, however, whether
the relevant descriptive facts, which we can call law-determining practices, or law
practices (or simply practices) for short,¹ are the only determinants of legal content,
or whether legal content also depends on normative or evaluative facts—value
facts,² for short. In fact, a central—perhaps the central—debate in the philosophy
of law is a debate over whether value facts are among the determinants of the
content of the law (though the debate is not usually characterized in this way).
A central claim of legal positivism is that the content of the law depends only on
social facts, understood as a proper subset of descriptive facts. As Joseph Raz says,
“H. L. A. Hart is heir and torch-bearer of a great tradition in the philosophy of law
which . . . regards the existence and content of the law as a matter of social fact

* This chapter was originally published as How Facts Make Law, 10 Legal Theory 157 (2004). It is
reprinted here with permission from Cambridge University Press. For helpful comments on ancient
and recent predecessors of this chapter, I am very grateful to Larry Alexander, Andrea Ashworth, Ruth
Chang, Jules Coleman, Martin Davies, Ronald Dworkin, Gil Harman, Scott Hershovitz, Kinch
Hoekstra, Harry Litman, Tim Macht, Tom Nagel, Ram Neta, Jim Pryor, Stephen Perry, Joseph Raz,
Gideon Rosen, Scott Shapiro, Seana Shiffrin, Ori Simchen, Martin Stone, Enrique Villanueva, and
two anonymous referees for Legal Theory. Special thanks to Susan Hurley and Nicos Stavropoulos for
many valuable discussions. I would also like to thank audiences at the University of Pennsylvania,
New York University, UCLA, Yale University, the 2002 Annual Analytic Legal Philosophy
Conference, and the 2003 International Congress in Mexico City, where versions of this material
were presented. Finally, I owe a great debt to the work of Ronald Dworkin.
¹ For the moment, I will be vague about the nature of law practices. For more precision, see
Section II.B below. ² For some explanation of what I mean by “value facts,” see infra note 22 .

How Facts Make Law. Mark Greenberg.

© Oxford University Press 2006. Published 2006 by Oxford University Press.
226 Mark Greenberg

whose connection with moral or any other values is contingent and precarious.”³
In contemporary philosophy of law, there are two distinct ways of developing this
tradition: hard and soft positivism. Hard positivism denies that value facts may
play any role in determining legal content.⁴ Soft positivism allows that the rele-
vant social facts may make value facts relevant in a secondary way. For example,
the fact that a legislature uses a moral term—“equality,” say—in a statute may
have the effect of incorporating moral facts—about equality, in this case—into
the law.⁵ On this soft positivist view, however, it is still the social facts that make
the value facts relevant, and the social facts need not incorporate value facts into
the law. Hence according to both hard and soft positivism, it is possible for social
facts alone to determine what the law is, and even when they make value facts
relevant, social facts do the fundamental work in making the law what it is—work
that is explanatorily prior to the role of value facts. To put things metaphorically,
hard positivism and soft positivism hold that there could still be law if God
destroyed all value facts.
Ronald Dworkin is the foremost contemporary advocate of an antipositivist posi-
tion. According to Dworkin, a legal proposition is true in a given legal system if it is
entailed by the set of principles that best justify the practices of the legal system.⁶
Since the notion of justification on which Dworkin relies is a normative notion, a
consequence of Dworkin’s view is that the content of the law depends on value facts.
Understanding and resolving the debate between positivists and antipositivists
requires understanding the nature of the relevant determination relation—the
relation between determinants of legal content and legal content. The debate, as
noted, concerns whether law practices are the sole determinants of legal content.
It is difficult to see how one can systematically address the question of whether
A facts are the sole determinants of B facts without understanding what kind of
determination is at stake. But the positivist/antipositivist debate has so far been
conducted with almost no attention to this crucial issue.
A preliminary point is that the determination relation with which we are con-
cerned is primarily a metaphysical, or constitutive, one, and only secondarily an
epistemic one: the law-determining practices make the content of the law what it is.
To put it another way, facts about the content of the law (“legal-content facts”)
obtain in virtue of the law-determining practices. It is only because of this underlying
metaphysical relation that we ascertain what the law is by consulting those practices.
A second preliminary point, which should be uncontroversial, is that no legal-
content facts are metaphysically basic or ultimate facts about the universe, facts for
which there is nothing to say about what makes them the case. Legal-content facts,
like facts about the meaning of words or facts about international exchange rates

³ J. Raz, Ethics in the Public Domain 210 (1994). Raz also puts the point epistemically: the
content of the law “can be identified by reference to social facts alone, without resort to any evaluative
argument.” Id. at 211.
⁴ See, e.g., Raz, supra note 3, at ch. 10; J. Raz, The Authority of Law ch. 3 (1979).
⁵ See, e.g., J. Coleman, Negative and Positive Positivism, 11 J. Legal Stud. 139 (1982); H. L. A. Hart,
Postscript, in The Concept of Law (2d ed., 1997). ⁶ See R. Dworkin, Law’s Empire (1986).
 How Facts Make Law 227

(for example that, at a particular time, a U.K. pound is worth 1.45 U.S. dollars),
hold in virtue of more basic facts. The important implication for present purposes
is that the full story of how the determinants of legal content make the law what it
is cannot take any legal content as given. It will not be adequate, for example, to
hold that law practices plus some very basic legal-content facts (for example, legal
propositions concerning the relevance of law practices to the content of the law)
together make the law what it is, for such an account fails to explain what it is in
virtue of which the very basic legal-content facts obtain.
Descriptive facts about what people said and did (and thought) in the past are
among the more basic facts that determine the content of the law. I claim that the
content of the law depends not just on descriptive facts but on value facts as well.
Given the plausible assumption that fundamental⁷ value facts are necessary rather
than contingent, there is, however, a difficulty about expressing my claim in terms
of counterfactual theses or theses about metaphysical determination. Even if the
value facts are relevant to the content of the law, it is still true that the content of
the law could not be different from what it is without the descriptive facts being
different (since it is impossible for the value facts, being necessary, to be different
from what they are). Necessary truths cannot be a non-redundant element of a
supervenience base. Hence both positivists and antipositivists can agree that
descriptive facts alone metaphysically determine the content of the law.⁸
In order to express the sense in which the content of the law is claimed to
depend on value facts, we therefore need to employ a notion different from and
richer than metaphysical determination. We can say that the full metaphysical
explanation of the content of the law (of why certain legal propositions are true)
must appeal to value facts. I earlier put the point metaphorically by saying that if
God destroyed the value facts, the law would have no content. The epistemic
corollary is that working out what the law is will require reasoning about value.
As we will see, a full account of what it is in virtue of which legal-content facts
obtain has to do more than describe the more basic facts that are the metaphysical
determinants of legal content. The relevant determination relation is not bare

⁷ The point of the qualification “fundamental” is to distinguish basic or pure value facts—that,
say, harm is a relevant moral consideration—from applied or mixed value facts—that returning the
gun to John tomorrow would be wrong. The fundamental value facts are plausibly metaphysically
necessary, while the applied value facts obviously depend on contingent descriptive facts as well as on
fundamental value facts. This qualification does not affect the point in the text, since the contingent
facts are encompassed in the supervenience base of descriptive facts. That is, if the fundamental value
facts supervene on the descriptive facts, the applied value facts will do so as well.
⁸ The term “metaphysical determination” is typically used in a way that implies nothing about the
order of explanation or about relative ontological basicness. In this sense, that the A facts metaphys-
ically determine the B facts does not imply that the B facts obtain in virtue of the obtaining of the A
facts. Positivists and antipositivists can agree not only that descriptive facts alone metaphysically
determine the content of the law but also that the obtaining of the relevant descriptive facts is part of
the explanation of the obtaining of legal-content facts. In this paper, we will be concerned only with
cases in which the putative determinants are more basic than and part of the explanation of the deter-
mined facts. For convenience, I will therefore say that the A facts metaphysically determine the B facts
only when the B facts obtain at least in part in virtue of the obtaining of the A facts.
228 Mark Greenberg

metaphysical determination. (As we have just seen, if that were the relevant relation,
there would be no debate between the positivists and the antipositivists. Positivists
would win the debate trivially, since the descriptive facts alone fix the content of the
law.) I argue for a particular understanding of the metaphysical relation (between
the determinants and the legal content that they determine), which I call rational
determination. Rational determination, in contrast to bare metaphysical determin-
ation, is necessarily reason-based (in a sense that I elaborate in Section II.B).
A quick way to grasp the basic idea is to consider the case of esthetic facts.
Descriptive facts metaphysically determine esthetic facts. A painting is elegant in
virtue of facts about the distribution of color over the surface (and the like). But
arguably there need not be reasons that explain why the relevant descriptive facts
make the painting elegant. We may be able to discover which descriptive facts make
paintings elegant (and even the underlying psychological mechanisms), but even if
we do, those facts need not provide substantive aesthetic reasons why the painting
is elegant (as opposed to causal explanations of our reactions). On this view, it may
just be a brute fact that a certain configuration of paint on a surface constitutes or
realizes a painting with certain esthetic properties. (As noted below, facts about
humor provide an even clearer example.) In contrast, if it is not in principle intelli-
gible why the determinants of legal content—the relevant descriptive facts—make
the law have certain content, then it does not have that content.
Rational determination is an interesting and unusual metaphysical relation
because it involves the notion of a reason, which may well be best understood as
an epistemic notion. If so, we have an epistemic notion playing a role in a meta-
physical relation. (Donald Davidson’s view of the relation between the determi-
nants of mental content and mental content is plausibly another example of this
general phenomenon.)⁹ For this reason, I believe that the rational-determination
relation is of independent philosophical interest.
My main goal in this chapter, however, is to show that, given the nature of the
relevant kind of determination, law practices—understood as descriptive facts
about what people have said and done—cannot themselves determine the content
of the law. Value facts are needed to determine the legal relevance of different
aspects of law practices. I therefore defend an antipositivist position, one that is
roughly in the neighborhood of Dworkin’s, on the basis of very general philosoph-
ical considerations unlike those on which Dworkin himself relies.¹⁰
We have two domains of facts: a higher-level legal domain and a lower-level
descriptive domain. It is, I claim, a general truth that a domain of descriptive facts can
rationally determine facts in a dependent, higher-level domain only in combination

⁹ See infra notes 17 and 18.

¹⁰ Dworkin’s theory of law depends on a view about the nature of “creative interpretation.” In par-
ticular, he argues that to interpret a work of art or a social practice is to try to display it as the best that
it can be of its kind. See Dworkin, supra note 6, at 49–65. Dworkin’s central argument for the posi-
tion that legal interpretation is an instance of this general kind of interpretation is that this position is
the best explanation of “theoretical disagreement” in law. Id. at 45–96; see also R. Dworkin, Law as
Interpretation, in The Politics of Interpretation (W. J. T. Mitchell ed., 1983).
 How Facts Make Law 229

with truths about which aspects of the descriptive, lower-level facts are relevant to the
higher-level domain and what their relevance is. Without the standards provided by
such truths, it is indeterminate which candidate facts in the higher-level domain
are most supported by the lower-level facts. There is a further question about the
source or nature of the needed truths (about the relevance of the descriptive facts to
the higher-level domain). In the legal case, these truths are, I will suggest, truths about
value.
The basic argument is general enough to apply to any realm in which a body of
descriptive facts is supposed to make it the case by rational determination that facts
in a certain domain obtain. For example, if the relation between social practices,
understood purely descriptively, and social rules is rational determination, the
argument implies that social practices cannot themselves determine the content of
social rules. (At that point, we reach the further question of the source of the
truths needed in the case of social rules; the answer may differ from that in the
legal case.) Hence the argument is of interest well beyond the philosophy of law.
In this chapter, I will largely confine the discussion to the legal case.
In Section II, I clarify the premises of the argument and explain that they
should not be controversial. In Section III, I examine why there is a problem of
how legal content is determined. The content of the law is not simply the mean-
ings of the words (and the contents of the mental states) that are uttered in the
course of law practices. Something must determine which elements of law prac-
tices are relevant and how they combine to determine the content of the law. Next,
in Section IV, I argue that law practices themselves cannot determine how they
contribute to the content of the law. In Section V, I consider and respond to three
related objections. Finally, in Section VI, I examine what the argument has estab-
lished about the relation between law and value.¹¹

II. The Premises

In this section, I set out the two premises of the argument and make a number of
clarifications. The second premise will require a great deal more discussion than the
first. I take both premises to be relatively uncontroversial in many contemporary legal
systems, including those of, for example, the United States and the United Kingdom.

A. Premise 1: Determinate Legal Content

The first premise of the argument is the following:
(D) In the legal system under consideration, there is a substantial body of
determinate legal content.
¹¹ There are interesting connections between this chapter and G. A. Cohen’s recent Facts and
Principles, 31 Phil. & Pub. Aff. 211 (2003). Cohen’s paper came to my attention too late for me to
explore the connections here, however.
230 Mark Greenberg

My use of the term “determinate” (like my use of “determine”) is metaphysical,

not epistemic. That is, for the law to be determinate on a given issue is not for us
to be able to ascertain what the law requires on that issue (or still less for there to
be a consensus), but for there to be a fact of the matter as to what the law requires
with respect to the issue. Thus, when I say that there is a substantial body of deter-
minate legal content, I mean roughly that there are many true legal propositions
(in the particular legal system). What do I mean by “legal propositions”?¹² A legal
proposition is a legal standard or requirement. An example might be the proposi-
tion that any person who, by means of deceit, intentionally deprives another
person of property worth more than a thousand dollars shall be imprisoned for
not more than six months. For a legal proposition to be true in a particular legal
system is for it to be a true statement of the law of that legal system.¹³ D is consist-
ent with the law’s being indeterminate to some extent, and it is deliberately vague
about how much determinacy there is. I think it is obvious that D is true in the
legal systems of many contemporary nations.

B. Premise 2: The Role of Law-determining Practices

The second premise is:
(L) The law-determining practices in part determine the content of the law.
The basic idea behind L is that the law depends on the law practices. L thus rules
out, for example, the extreme natural-law position that the law is simply whatever
morality requires. I take it, however, that very few contemporary legal theorists
would defend this position or any other position that makes law practices irrele-
vant to the content of the law.
By the term “law practices” (or, more fully, “law-determining practices”) I mean to
include at least constitutions, statutes, executive orders, judicial and administrative
decisions, and regulations. Although it is unidiomatic, I will refer to a particular
constitution, statute, judicial decision, and so on as a law practice. So a practice, in
my usage, need not be a habitual or ongoing pattern of action. I need to clarify
what I mean by saying that a practice can be, for example, a statute. Lawyers often
talk as if a statute (or other law practice) is simply a text. It is of course permissible
to use the word “statute” (or “constitution,” “judicial decision,” etc.) to refer to the
corresponding text, and I will occasionally write in this way. But if law practices are
to be determinants of the content of the law, the relevant practice must be, for
example, the fact that a majority of the members of the legislature voted in a certain
way with respect to a text (or alternatively the event of their having done so), not
merely the text itself. So as I will generally use the term, “statute” (“constitution,”
etc.) is shorthand for a collection of facts (or events),¹⁴ not a text.
¹² The term is Dworkin’s. See Dworkin, supra note 6, at 4.
¹³ I will usually omit the qualification about a particular legal system.
¹⁴ I will hereinafter ignore the possibility of taking law practices to be composed of events rather
than facts.
 How Facts Make Law 231

In general, then, law practices consist of ordinary empirical facts about what
people thought, said, and did in various circumstances.¹⁵ For example, law prac-
tices potentially include the facts that, in a particular historical context, a legisla-
tive committee issued a certain report, various speeches were made in a legislative
debate, a bill that would have repealed a statute failed to pass, a concurring judge
issued a certain opinion, and an executive official announced a particular view of a
statute.¹⁶ Once I have clarified the claim that law practices partially determine the
content of the law, I will be able to say something more precise about what counts
as a law practice.
When L says that law practices determine (in part) the content of the law, what
sense of “determine” is involved? As noted above, a preliminary point is that L’s
claim is constitutive or metaphysical, not epistemic. That is, it is not a claim that
we use law practices to ascertain what the content of the law is, but that such prac-
tices make it the case that the content of the law is what it is.
I maintain that the relevant kind of determination is not bare metaphysical
determination but what we can call rational determination. The A facts rationally
determine the B facts just in case the A facts metaphysically determine the B facts
and the obtaining of the A facts makes intelligible or rationally explains the
B facts’ obtaining. Thus, L is the conjunction of two doctrines, a metaphysical-
determination doctrine and a rational-relation doctrine. Let me elaborate.
I will make the (uncontroversial, I hope) assumption that there are facts that
(1) are ontologically more basic than facts about legal content and (2) metaphysically
determine that the content of the law is what it is. The metaphysical-determination
doctrine is that these more basic facts that determine the content of the law non-
redundantly include law practices.
Metaphysical determination can be brute. If the A facts are more basic facts that
metaphysically determine the B facts, there is a sense in which the A facts explain
the B facts, for the A facts are more basic facts, the obtaining of which entails that
the B facts obtain. But there need be no explanation of why the obtaining of par-
ticular A facts has the consequence that it does for the B facts. To dramatize the
point, even a perfectly rational being may not be able to see why it is that particu-
lar A facts make particular B facts obtain.
The metaphysical-determination doctrine is not enough to capture our
ordinary understanding (which L attempts to articulate) of the nature of the

¹⁵ Hypothetical decisions arguably play a significant role in determining the content of the law,
but for purposes of this paper they will largely be ignored. Susan Hurley characterizes hypothetical
decisions as hypothetical cases that have a settled resolution. See S. L. Hurley, Coherence, Hypothetical
Cases, and Precedent, 10 Oxford J. Legal Stud. 221 (1990). Another possibility is to include any
hypothetical case that has a determinate right answer, even if there is disagreement on its resolution.
There would be disagreement about which hypothetical cases had determinate right answers and
therefore about which were determinants of legal content.
¹⁶ Nothing turns on how we individuate practices, at least in the first instance. E.g. a legislative
committee’s issuance of a report could be considered part of the circumstances in which a majority of
the legislature voted for a statute or could be considered a separate practice. Once the roles of differ-
ent elements of law practices are determined, there may be a basis for individuation.
232 Mark Greenberg

determination relation between the law practices and the content of the law. We
also need the rational-relation doctrine, which holds that the relation between the
determinants of legal content and legal content is reason-based. In the relevant
sense, a reason is a consideration that makes the relevant explanandum intelli-
gible.¹⁷ Here is one way to put the point. There are indefinitely many possible
mappings, from complete sets of law practices to legal content (to complete sets of
legal propositions). As far as the metaphysical-determination doctrine goes, it
could simply be arbitrary which mapping is the legally correct one. In other
words, the connection between a difference in the practices and a consequent dif-
ference in the content of the law could be brute. For example, it is consistent with
the truth of the metaphysical-determination doctrine that, say, the deletion of one
seemingly unimportant word in one subclause of one minor administrative regu-
lation would result in the elimination of all legal content in the United States—in
there being no true legal propositions in the U.S. legal system (though there is no
explanation of why it would do so). By contrast, according to the rational-relation
doctrine, the correct mapping must be such that there are reasons why law practices
have the consequences they do for the content of the law.
To put it metaphorically, the relation between the law practices and the content
of the law must be transparent.¹⁸ (For the relation to be opaque would be for it to
be the case that any change in law practices could have, so far as we could tell, any
effect on the content of the law. The effects on the content of the law could be
unfathomable and unpredictable, even if fully determinate.)
It bears emphasis that what must be rationally intelligible is not the content of
the law but the relation between, on the one hand, determinants of legal content
and, on the other, legal content. L holds not that the content of the law must be
rational or reasonable but that it must be intelligible that the determinants of legal
content make the content of the law what it is. For example, there must be a rea-
son that deleting a particular word from a statutory text would have the impact on
the law that it would in fact have.¹⁹

¹⁷ I will not attempt to spell out the relevant notion of a reason more fully here. One possibility is
that the best way to do so is in terms of idealized human reasoning ability. For example, the idea
might be that practices yield a legal proposition if and only if an ideal reasoner would see that they do.
The notion of a reason would hence be an epistemic notion. In that case, L would imply that the
metaphysics of law involves an epistemic notion; that is, what the law is would depend in part on
what an ideal human reasoner would find intelligible.
¹⁸ A useful comparison can be made to certain well-known positions in the philosophy of mind.
Donald Davidson’s radical interpretation approach to mental and linguistic content presupposes that
behavior determines the contents of mental states and the meaning of linguistic expressions in a way
that must be intelligible or transparent. D. Davidson, Radical Interpretation, in Inquiries into
Truth and Interpretation (1984); and D. Davidson, Belief and the Basis of Meaning, in Inquiries
into Truth and Interpretation (1984). Similarly, Saul Kripke’s “Kripkenstein” discussion presup-
poses that we must be able to “read off ” the contents of mental states from the determinants of con-
tent. S. Kripke, Wittgenstein on Rules and Private Language 24, 29 (1982). See infra note 25.
¹⁹ At this point in the text, I have deleted a paragraph that appeared in the original publication of
this chapter. I have also made a few changes in the next few paragraphs.
 How Facts Make Law 233

Another important clarification is that it is no part of rational determination that

the reasons in question must be value facts. The reasons that the determining facts
must provide are considerations that explain in rational terms why particular facts of
the target domain, as opposed to others, obtain. In general, non-normative (and
non-evaluative) facts can constitute reasons of the relevant kind. Here is an example
from a different domain. Consider the facts that (in a particular economy) the total
amount of demand deposits is 2 million units and the total amount of currency in
the hands of the public is 1 million units. With the fact that the money supply M1
consists in demand deposits plus currency in the hands of the public, these facts
make rationally intelligible that the total amount of M1 (in the economy in ques-
tion) is 3 million units.
As far as the rational-relation doctrine is concerned, it is an open question
whether there are non-normative, non-evaluative facts that could constitute rea-
sons for legal facts—and indeed whether there are value facts that could do so. (My
ultimate view, of course, is not that value facts could themselves provide such rea-
sons, but that both descriptive and value facts are needed.) The point is important
because otherwise positivists could not accept the rational-relation doctrine. The
strategy of my argument is to use the rational-relation doctrine, which, I claim
below, most legal theorists implicitly accept, to argue for the controversial conclu-
sion that value facts must be determinants of legal content.
In principle, conceptual truths (that are not value facts) about law could, with
law practices, make rationally intelligible the content of the law. For example, it
might be claimed that it follows from the concept of law that a validly enacted
statute makes true those propositions that are the ordinary meanings of the sen-
tences of the statute. On this view, the fact that a statutory text says that any per-
son who drives at more than sixty-five miles an hour commits an offense, together
with certain conceptual truths about law, makes it intelligible that the law requires
that one not drive at more than sixty-five miles an hour.
The general point, again, is that it is a matter for argument, not something pre-
supposed by L, what kinds of facts (if any) must supplement law practices in order
for the determining facts to provide reasons that explain why particular legal
propositions are true. In particular, L does not presuppose that value facts are
needed.²⁰
Why have I made the qualification that law practices partially determine the
content of the law? Law practices must determine the content of the law. But,
my argument continues, there are many possible ways in which practices could
determine the content of the law. (Put another way, there are many functions
that map complete sets of law practices to legal content.) Something other than

²⁰ At a later stage of analysis, we might find that there are restrictions on what kind of reasons the
determinants of legal content must provide. For example, it might turn out that legal systems have
functions and that in order for a legal system to perform its functions properly, the determinants of
legal content must provide reasons for action. See infra the last paragraph of Section VI.A. L does not
presuppose any such restrictions, however.
234 Mark Greenberg

law practices—X, for short—must help to determine how practices contribute

to the content of the law (that is, to determine which mapping is the legally cor-
rect one). So a full account of the metaphysics of legal content involves X as well
as law practices.
This conclusion can be expressed in two equivalent ways. We could say that
practices are the only determinants of legal content but that an account of legal
content must do more than specify the determinants. This formulation is particu-
larly natural if X consists of necessary truths.²¹ (A related advantage is that this
way of talking highlights that practices are what typically vary, producing changes
in the content of the law.) The second formulation would say that X and law prac-
tices are together the determinants of the content of the law. Because it is con-
venient to express this paper’s thesis by saying that X plays a role in determining
legal content (and because I want to leave open the possibility that X may vary),
this formulation seems preferable, and I will adopt it as my official formulation.
Accordingly, I will say that law practices are only some of the determinants of the
content of the law. (For brevity, however, I will sometimes omit the qualification
“partially” and write simply that law practices determine the content of the law.)

C. Law Practices as Descriptive Facts

Let me now return to the question of what counts as a law practice. I have said that
law practices consist of ordinary empirical facts about what people have thought,
said, and done—including, paradigmatically, facts about what members of consti-
tutional assemblies, legislatures, courts, and administrative agencies have said and
done. I want to be clear about the exclusion of two kinds of facts. First, law practices
do not include legal-content facts. Secondly, law practices do not include facts
about value, for example, facts about what morality requires or permits.²² The
law practices thus consist of non-legal-content descriptive facts. (For convenience,
I will generally write simply “descriptive facts” rather than “non-legal-content
descriptive facts”; this shorthand does not reflect a presupposition that legal-
content facts are value facts.) Let me explain the reasons for the two exclusions.
As I said, I am assuming that the content of the law is not a metaphysically
basic aspect of the world but is constituted by more basic facts. The reason for
the first exclusion—of legal-content facts—is that law practices are supposed to

²¹ See supra text accompanying notes 7 and 8.

²² By “facts,” I simply mean true propositions. Hence facts about value, or value facts, are true
normative or evaluative propositions, such as true propositions about what is right or wrong, good or
bad, beautiful or ugly. The fact that people value something or believe something is valuable is not a
value fact but a descriptive fact about people’s attitudes. For example, the fact, if it is one, that accept-
ing bribes is wrong is a value fact; the fact that people value honesty is a descriptive fact. This paper
does not attempt to address a skeptic who maintains that there are no true propositions about value.
One could use an argument of the same form as mine to argue that there must be value facts—for
without them there would not be determinate legal requirements. But a skeptic about value facts
would no doubt take such an argument to be a case of the legal tail wagging the value dog.
 How Facts Make Law 235

be the determinants of legal content, not part of the legal content that is to be
determined.
Suppose an objector maintained that the law practices that determine legal
content are themselves laden with legal content. It is certainly natural to use the
term “law practices” in this way. After all, the fact that the legislature passed a bill
is legal-content laden: it presupposes legal-content facts about what counts as a
legislature and a bill. Since legal-content facts are not basic, however, there must
be non-legal-content facts that constitute the legal-content-laden practices. At
this point, we will have to appeal to descriptive facts about what people thought,
said, and did—the facts that I am calling “law practices.” For example, the fact
that a legislature did such and such must hold in virtue of complex descriptive
facts about people’s behavior and perhaps also value facts. (If, in order to account
for legal-content-laden practices, we have to appeal not merely to descriptive facts
but also to value facts, so much the worse for the positivist thesis that the content
of the law depends only on descriptive facts.) The convenience of talking as if law
practices consisted in legal-content-laden facts about the behavior of legislatures,
courts, and so on should not obscure the fact that there must be more basic facts in
virtue of which the legal-content facts obtain. To build legal-content facts into law
practices would beg the question at the heart of this paper—the question of the
necessary conditions for law practices to determine the content of the law. (For
ease of exposition, I will continue to use legal-content-laden characterizations of
the law practices, but the law practices should, strictly speaking, be understood to
be the underlying descriptive facts in virtue of which the relevant legal-content
facts obtain.)
It is uncontroversial that certain kinds of facts are among the supervenience base
for legal content: roughly speaking, facts about what constitutional assemblies,
legislatures, courts, and administrative agencies did in the past. Of course, as just
noted, such characterizations are legal-content laden and are therefore shorthand
for non-legal-content characterizations of the law practices. (I do not mean, of
course, that it is uncontroversial exactly which facts of these kinds are relevant; I’ll
return to this point shortly.) There are at least two kinds of controversy, however,
about the determinants of legal content.
First, it is controversial whether value facts are among the determinants of
content. The reason for the second exclusion—the exclusion of value facts—is
that this paper tries to argue from the uncontroversial claim that law practices are
determinants of the content of the law to the conclusion that value facts must play
a role in determining the content of the law. If law practices were taken to be
value-laden, it would no longer be uncontroversial that they are determinants of
legal content. (On the other hand, even those theorists who think that value facts
are needed to determine the content of the law can accept that descriptive facts
also play a role.) Moreover, unless we separate the descriptive facts from the value
facts, we cannot evaluate whether the descriptive facts can themselves determine
the content of the law. In sum, by understanding law practices to exclude value
236 Mark Greenberg

facts, I ensure that L is uncontroversial and I prepare the way for my argument
that descriptive facts alone cannot determine the content of the law.
The second kind of controversy about the determinants of legal content is con-
troversy over precisely which descriptive facts are determinants. I have mentioned
some paradigmatic determinants of legal content. But there are other kinds of
descriptive facts—for example, facts about customs, about people’s moral beliefs,
about political history, and about law practices in other countries—that are
arguably among the determinants of legal content. Also, somewhat differently, it
is controversial which facts about judicial, legislative, or executive behavior are rel-
evant. There can be debate, for example, about the relevance of legislative history,
intentions of legislators and of drafters of statutes, legislative findings, judicial
obiter dicta, and executive interpretations of statutes. I propose to deal with this
second kind of controversy by leaving our understanding of law practices open
and non-restrictive.
There are several reasons for this approach. First, my argument is that prac-
tices, understood as composed of descriptive facts, cannot themselves determine
the content of the law. If I begin with a restrictive understanding of practices, my
argument will be open to the reply that I failed to include some of the relevant
facts. For this reason, I want to be liberal about which descriptive facts are part of
law practices. Secondly, my argument will not depend on exactly which descrip-
tive facts make up law practices. Rather, I will make a general argument that
descriptive facts—in particular, facts about what people have done and said and
thought—cannot by themselves determine the content of the law. Therefore it
will not matter precisely which such facts are included in law practices. Thirdly,
my view is ultimately that the question of which facts are part of law practices—
like the question of how different aspects of law practices contribute to the con-
tent of the law—is dependent on value facts. (Indeed, I will often treat the two
questions together as different aspects of the general question of the way in
which law practices determine the content of the law.) As we will see, that we
cannot in an uncontroversial way specify which are law practices and which are
not is one consideration in support of my argument for the necessary role of
value. All we need to begin with is some rough idea of law practices, which can
be overinclusive.
In sum, let law practices include, in addition to constitutions, statutes, and
judicial and administrative decisions, any other non-legal-content descriptive
facts that turn out to play a role in determining the content of the law.²³ Which
facts these are and what role they play are controversial, so we can begin with a
rough and inclusive understanding of law practices. One aspect of figuring out
how law practices contribute to the content of the law will be figuring out which

²³ This proviso does not make L the tautological claim that the determinants of legal content
determine legal content. L says that constitutions, statutes, judicial decisions, and so on are (non-
redundantly) among the determinants of content.
 How Facts Make Law 237

facts make a contribution and which do not. But there is no reason to expect a
clean line between law practices and other facts.²⁴
The exclusion of value facts should not be taken to suggest that law practices are
to be understood in solely physical or behavioral terms. To the contrary, as I explain
in the next section, I take for granted the mental and linguistic contents involved
in law practices. In other words, law practices include the facts about what the
actors believe, intend, and so on and about what their words mean.

D. Why L Should Be Uncontroversial

The metaphysical-determination doctrine should be relatively uncontroversial,
certainly for those who accept that there are determinate legal requirements.
Positivists, Dworkinians, and contemporary natural-law theorists as well as prac-
ticing lawyers and judges accept that constitutions, statutes, and judicial and
administrative decisions are (non-redundant) determinants of the content of the
law. That law practices may also include other descriptive facts to the extent that
those facts are determinants of the content of the law obviously cannot make the
metaphysical-determination doctrine controversial.
More generally, we began with the premise that there are determinate legal
requirements. What makes them legal requirements is that they are determined, at
least in part, by law practices. Contrast the requirements of morality (or, to take a
different kind of example, of a particular club). If law practices did not determine
legal content, there could still be moral requirements and officials’ whims, but
there would be no legal requirements. In order to think differently, one would
have to hold a strange view of the metaphysics of law according to which the con-
tent of the law is what it is independently of all the facts of what people said and
did that make up law practices, and law practices are at best evidence of that con-
tent. So I think it should be uncontroversial that law practices are among the
determinants of the content of the law.
As to the rational-relation doctrine, it is fundamental to our ordinary under-
standing of the law and taken for granted by most legal theory, though seldom
articulated. The basic idea is that the content of the law is in principle accessible to
a rational creature who is aware of the relevant law practices. It is not possible that
the truth of a legal proposition could simply be opaque, in the sense that there
would be no possibility of seeing its truth to be an intelligible consequence of the
law practices. In other words, that the law practices support these legal proposi-
tions over all others is always a matter of reasons—where reasons are considera-
tions in principle intelligible to rational creatures. (A corollary is that to the extent

²⁴ One natural understanding of “law practices” is more restrictive than the way I use the term.
According to this understanding, law practices are limited to (facts about) what legal institutions and
officials do in their official capacities. If we used the term “law practices” in this natural way, we would
need, in addition to the category of law practices, a category of other descriptive facts that play a role
in determining the content of the law.
238 Mark Greenberg

that the law practices do not provide reasons supporting certain legal propositions
over others, the law is indeterminate.)
I will not attempt to defend the rational-relation doctrine fully here but will
mention a few considerations. Suppose the A facts metaphysically determine
the B facts, but the relation between the A facts and the B facts is opaque. In that
case, how could we know about the B facts? One possibility is that we have
access to the B facts independently of our knowledge of the A facts. An example
might be the relation between the microphysical facts about someone’s brain
and the facts about that person’s conscious experience. Suppose that the micro-
physical facts metaphysically determine the facts about the person’s conscious
experience but that the relation is opaque. The opaqueness of the relation does
not affect the person’s ability to know the facts about his conscious experience,
because we do not in general learn about our conscious experience by working it
out from the microphysical facts. (Moreover, since we have independent know-
ledge of conscious experience, we might be able to discover correlations between
microphysical facts and conscious experience even if those correlations were not
intelligible even in principle.) To take a different kind of example, the micro-
physical facts may metaphysically determine the facts about the weather, and
the relation may be opaque, but again, we do not learn about the weather by
working it out from the microphysical facts.
A second possibility is that we do work out the B facts from the A facts but
that we have a nonrational, perhaps hardwired, capacity to do so. For example,
it is plausible that the facts about what was said and done (on a particular occa-
sion, say) determine whether what was said and done was funny (and to what
degree and in what way). And we do work out whether an incident was funny
from the facts about what was done and said. It is plausible, however, that the
relation between what was said and done and its funniness is not necessarily
transparent to all rational creatures; our ability to know what is funny may
depend on species-specific tendencies; that is, there may not be reasons that
make the humor facts intelligible; it may just be a brute fact that humans find
certain things funny.²⁵
Law seems different from both of these kinds of cases. First, our only access to
the content of the law is through law practices. It is not as if we can find out what
the law is directly or through some other route. And the whole enterprise of law-
making is premised on the assumption that the behavior of legislators, judges, and

²⁵ Compare the issue of how facts about our use of words determine their meaning. Natural
languages are a biological creation. Although many philosophers have thought differently (see
supra note 18), we cannot take for granted that the correct mapping from the use of words to their
meaning will be based on reasons. How, it may be objected, would we then be able to work out
from their use of words what others mean? The answer may simply be that we have a species-
specific, hardwired mechanism that rules out many incorrect mappings that are not ruled out by
reasons. In that case, an intelligent creature without that mechanism would not be able to work out
what words mean.
 How Facts Make Law 239

other law-makers will have understandable and predictable consequences for the
content of the law.
Secondly, we are able to work out what the law is and predict the effect on the
law of changes in law practices through reasons, not through some non-rational
human tendency to have correct law reactions to law practices.
When lawyers, judges, and law professors work out what the law is, they give
reasons for their conclusions. Indeed, if we find that we cannot articulate reasons
that justify a provisional judgment about what the law is in light of law practices,
we reject the judgment. By contrast, it is notoriously difficult to explain why
something is or is not funny, and we do not generally hold our judgments about
humor responsible to our ability to articulate reasons for them. A related point is
that we believe that we could teach any intelligent creature that is sensitive to reasons
how to work out what the law is.
It might be objected that although the epistemology of law is reason-based, the
metaphysics might not be. It is difficult to see how such an objection could be
developed. For present purposes, I will simply point out that when legal practi-
tioners give reasons for their conclusions about what the law is, they believe that
they are not merely citing evidence that is contingently connected to the content
of the law; rather, they believe that they are giving the reasons that make the law
what it is. The point is not that lawyers believe themselves to be infallible. Rather,
they believe that when they get things right, the reasons they discover are not
merely reasons for believing that the content of the law is a particular way, but the
reasons that make the content of the law what it is. Although they would never put
it this way, lawyers take for granted that the epistemology of law tracks its meta-
physics. And the epistemology of law is plainly reason-based.
Legal theorists generally take for granted some version of the claim that the
relation between law practices and the content of the law is reason-based. An
example is H.L.A. Hart’s argument that the vagueness and open texture of legal
language have the consequence that the law is indeterminate.²⁶ If bare metaphysical
determination were all that was at issue—if it were not the case that the relation
between practice and content were necessarily intelligible—the vagueness of
language would in no way support the claim that law was indeterminate. Similarly,
when legal realists or Critical Legal Studies theorists argue that the existence of
conflicting pronouncements or doctrines in law practices results in underdetermin-
ation of the law, their arguments would be beside the point if what was at stake
were not rational determination.²⁷
In general, the large body of legal theory that has explored the question of
whether law practices are capable of rendering the law determinate (and if so, how

²⁶ H. L. A. Hart, The Concept of Law ch. 7 (2d ed., 1997).

²⁷ See, e.g., A. Altman, Legal Realism, Critical Legal Studies, and Dworkin, 15 Phil. & Pub. Aff.
205 (1986); M. Kelman, Interpretive Construction in the Substantive Criminal Law, 33 Stan. L. Rev.
591 (1981).
240 Mark Greenberg

determinate) presupposes that law practices determine the content of the law in a
reason-based way. If the relation between law practices and the content of the law
could be opaque, any set of law practices would be capable, as far as we would be
able to judge, of determining any set of legal propositions. (As long as there are as
many possible sets of law practices as there are possible sets of legal propositions,
there is no barrier to the content of the law’s being fixed by the practices, and we
would have no warrant to rely on our assessment of other putative prerequisites
for practices to determine the content of the law.) In sum, the doctrine that law
practices rationally determine the content of the law captures a basic conviction
about the law that is shared by law-makers, lawyers, and legal theorists and is
supported by the epistemology of law.
Why does it matter to my argument that the relation between law practices and
the content of the law is reason-based? This paper explores the necessary conditions
for law practices’ making the content of the law what it is. The central argument is
that descriptive facts cannot determine their own rational significance—what reasons
they provide. The argument therefore depends on the claim that the descriptive
facts determine the content of the law in a reason-based way. It turns out that value
facts are needed to make it intelligible that law practices support certain legal
propositions over others.²⁸

E. The Scope of the Argument

Premises D and L tell us something about the scope of my argument. The argu-
ment is sound only for legal systems in which D and L are true. So my conclu-
sions are limited to legal systems in which there are legal requirements that are
determined in part by law practices. If there is a legal system in which there are
no determinate legal requirements, my argument would not apply to it.
Similarly, if there is a legal system in which law practices, understood as (facts
about) various people’s sayings and doings, do not play a role in determining the
content of the law, my argument would not apply to it. For example, perhaps
there could be a legal system in which the content of the law is determined
exclusively by the content of morality or exclusively by divine will. In this paper,
I do not address questions of the necessary conditions for something’s counting
as a legal system. It might be argued that a substantial body of legal require-
ments that are determined by practices of various officials or institutions is a
necessary condition for the existence of a legal system, but I do not intend to
pursue such an argument.

²⁸ Suppose that the relation between law practices and the content of the law were necessarily
intelligible only in a way that depends on some human-specific tendency. As long as practices must
provide considerations that are intelligible (even if only to humans), a version of my argument should
still go through.
 How Facts Make Law 241

III. Is There a Distinctively Legal Problem of Content?

We begin with our two premises: that the law has determinate content, and that
law practices in part determine that content. Our question is: What conditions
must be satisfied in order for law practices to determine legal propositions?
As I said above, since we are interested in problems of the determination of con-
tent only to the extent that they are peculiarly legal, we can take for granted the
content of sentences and propositional attitudes.²⁹ So the question is: How can a
collection of facts about what various people did and said (including the facts
about what they intended, believed, preferred, and hoped, and about what their
words meant) determine which legal propositions are true?
At this point, however, it must be asked whether there is a peculiarly legal prob-
lem of content. Once we take for granted the relevant mental and linguistic con-
tent, it may seem that no problem of legal content remains. Legal content is simply
the content of the appropriate mental states and texts. In this section, I consider
this possibility and argue that it is not at all plausible. The ordinary mental and
linguistic content of utterances and mental states of participants in law practices—
nonlegal content, for short—does not automatically endow the law with legal
content. Something must determine which aspects of law practices are relevant and
how together they contribute to the content of the law.
In the next section, I consider the possibility that, given the content of the
relevant utterances and attitudes, law practices themselves determine how they
contribute to the content of the law and thus can unilaterally determine the
content of the law. But before we turn to whether law practices can solve the
problem of legal content, we need to see what the problem is—why the nonlegal
content of law practices does not provide the content of the law. That is the
topic of this section.
In legal discourse, both ordinary and academic, constitutional or statutory pro-
visions and judicial decisions are often conflated with rules or legal propositions.
For example, lawyers will sometimes talk interchangeably of a statutory provision
and a statutory rule, or of a judicial decision and the rule of that case. In non-
philosophical contexts there is generally no harm in this kind of talk. Since
our question, however, is how law practices determine the content of the law, it is
crucial not to confuse law practices with legal propositions. For example, if one
assumed that a statute was the rule or proposition expressed by the words of the
statute, one might think that there was no problem of how law practices could

²⁹ There is no practical problem with taking these matters for granted and proceeding without
a solution to basic problems concerning how linguistic and mental content are possible. These
problems do not concern difficulties we encounter in practice in attributing linguistic and mental
content; the difficulty is in saying what it is in virtue of which a linguistic expression or mental
state has its content.
242 Mark Greenberg

determine legal content; or one might think that the only problem was how to
combine or amalgamate a large number of rules or propositions.
Although it would beg the question to take legal propositions for granted, we
do have the propositions that are the content of the utterances and mental states of
participants in law practices. What is wrong with the idea that those propositions
constitute legal content, so that law practices, once they are understood to include
facts about mental and linguistic content, automatically have legal content?
I will begin with the least serious problems—those concerning the attribution
of non-legal content. Although we are normally able to attribute attitudes to peo-
ple based on what they say and do and to attribute standard meanings to a large
number of sentences of a language we speak, there are difficulties in attributing
non-legal content to aspects of a putative law practice. Here are a few examples.
First, when I say that we can take for granted mental and linguistic content,
I mean that we need not ignore the mental and linguistic content that is available.
We should not, however, assume that all of the contents of the mental states of all
of the people involved in law practices are available. That would obviously be
false. In general, what is available in the standard reports of law practices is not
sufficient to attribute much in the way of attitudes to the people who actually per-
formed the actions and made the utterances; the fact that a particular legislator
voted for a bill or a certain judge signed an opinion is not in general sufficient to
attribute beliefs, intentions, hopes, and so on, to her. Moreover, the law restricts
what evidence of the intentions and beliefs of legislators and judges is acceptable to
determine the content of the law. Even when the intentions of a legislator or judge
are relevant to the content of the law, it is not the case that, say, her private letters
or diary may be a source of that intention. Something must determine which evi-
dence of legally relevant attitudes is legally acceptable.
Secondly, though many sentences of natural languages have standard meanings,
it is notorious that this is not true of some of the sentences uttered by those
engaged in making law practices. The point here is not that in legal contexts lin-
guistic expressions often have specialized meanings that are not straightforwardly
connected to their ordinary meanings. Rather, some of the contorted sentences in
the law books have no standard meaning in a natural language.
Thirdly, even when sentences taken alone have standard meanings, collections
of those sentences may fail to do so. In other words, the property of having a stand-
ard meaning (on a notion of standard meaning appropriate for present purposes)
is not closed under conjunction (for example, because context may introduce
ambiguity into an otherwise unambiguous sentence).
Setting aside these problems with ascertaining non-legal content, we can turn
to the more important question of the bearing of non-legal content on legal con-
tent. One problem is that the non-legal content of some elements of law practices
has, or arguably has, little or nothing to do with the legal content determined by
those practices. Consider sentences in statutory preambles, sentences in presiden-
tial speeches at bill-signing ceremonies, and sentences in judicial opinions that are
 How Facts Make Law 243

not necessary to the resolution of the issue before the court. Another example is
the actual but unexpressed hopes of the members of the legislature as to how the
courts would interpret a statute. Countless sentences are written and spoken at
different stages of law-practice making by people with myriad attitudes.³⁰
Something must determine which sentences’ and attitudes’ contents are relevant.
Another problem is that the contribution of a particular law practice to the
content of the law may not be the meaning of any text or the content of any per-
son’s mental state. The actual attitudes of appellate judges may be irrelevant;
instead the relevant question may be what a hypothetical reasonable person would
have intended by the words uttered by the judges or what would be the best, or the
narrowest, explanation of the result reached. Another possibility is that aspects of
law practices that contribute to non-legal content in one way contribute to legal
content in an entirely different way; facts about what was said and done may have
peculiarly legal significance. An obvious example is that common words such as
“malice” and “fault” are often used in legal discourse in a technical sense. To take a
more subtle instance, when a panel of several judges is badly split, it can be a com-
plex and tricky matter to ascertain the relevance to legal content of the meanings
of the words of the different judicial opinions.
Similarly, facts about the circumstances in which sayings and doings occurred
that have little to do with the non-legal content of the people’s attitudes and words
may significantly affect the content of the law. For example, in a judicial decision,
the fact that an issue is not in controversy arguably prevents the court’s statements
on that issue from making any contribution to the content of the law.
Even when the content of sentences and mental states is relevant to the content
of the law, there can be no mechanical derivation of the content of the law. For
example, how are conflicting contents to be combined? In general, there remains
the problem of how the non-legal contents associated with different law practices
interact with each other (and with other relevant aspects of law practices) to deter-
mine the content of the law.
We have surveyed a number of reasons why non-legal content—the meanings of
sentences and contents of mental states—does not simply constitute legal content.
But this way of thinking about the problem will have an artificial quality for those
familiar with legal reasoning. The idea that the non-legal content of law practices
constitutes their legal content presupposes roughly the following picture. Associated
with each law practice is a text (and perhaps some mental states). Once we have the
meanings of the texts and the contents of the mental states, each law practice will be
associated with a proposition or set of propositions. Ascertaining the law on a

³⁰ In the case of a judicial decision, for example, the possibly relevant sentences include sentences
uttered by the parties to the controversy, by lawyers, and by judges to lawyers and other judges. They
include sentences written by judges in orders and judicial opinions. Judicial opinions alone include a
large number and variety of sentences: they state facts, give reasons, summarize, make general claims
about the content of the law, state holdings; moreover, there are concurring and dissenting as well as
majority opinions.
244 Mark Greenberg

particular issue is just a matter of looking up the propositions that are applicable to
the issue. Even if this picture were accurate, we have discussed a number of reasons
why non-legal content would not automatically yield legal content. But the problem
is worse than these reasons would indicate. As I will now suggest, the whole picture
is wrongheaded. Law practices do not determine the content of the law by con-
tributing propositions which then get amalgamated.
Here is the real problem of legal content. There are many different law practices
with many different aspects or elements. There is an initial question of which facts
are parts of law practices and which are not. Are preambles of bills, legislative
findings, legislative committee reports, dissenting opinions, unpublished judicial
decisions, customs, the Federalist Papers, and so on to be included in law practices?
In my view, this question is really just part of a second question: Which aspects
of, for example, judicial or legislative practices are relevant to the content of the
law? Just to suggest the dimensions of the problem, here are some candidates for
the relevant elements or aspects of practices. With respect to a judicial decision: the
facts of the case, the judgment rendered, the words used by the court in the
majority opinion, the reasons given for the outcome, the judges’ beliefs, the judges’
identities, the level and jurisdiction of the court; with respect to a legislative action:
the words of the statute, the legislature’s actual intention (if there is such a thing),
the purposes that the words of the statute could reasonably be intended to imple-
ment, statements by the person who drafted the statute, speeches made during the
legislative debate preceding passage, the circumstances in which the legislature
acted, subsequent decisions not to repeal the statute.
Fourthly, once we know which elements of practices are relevant, the problem
of determining the content of the law is not simply a problem of adding or amal-
gamating the various relevant aspects of practices. One obvious point is that
some elements of practices are far more important than others, and elements of
practices matter in different ways. But more fundamentally, as anyone familiar
with legal reasoning knows, the content of the law is not determined by any kind
of summing procedure, however complicated. For example, judicial decisions,
constitutional provisions, and legislative history can affect what contribution
a statute makes. It is not that those practices contribute propositions that are
conjoined to a proposition contributed by the statute. The statute’s correct inter-
pretation may be determined by a potential conflict with a constitutional provision
or by the outcome of cases in which courts have interpreted the same or related
statutes.
To take a different kind of example, constitutional provisions, statutes, and
judicial decisions can have an impact on the contribution of judicial and adminis-
trative decisions to the content of the law by affecting our understanding of the
proper role of courts and administrative agencies. Or, differently, statutes can have
an impact on what judicial decisions mean by making clear what the legislature
cares about, thus affecting which differences between cases matter and conse-
quently whether past precedents control the present issue. A final example is that
 How Facts Make Law 245

the principle that a series of cases stands for is not the conjunction of the proposi-
tions announced in each case.
It is safe to conclude that the law does not automatically acquire content
when actions, utterances, and sentences involved in law practices are attributed
content. It is a mistake even to think that the issue is how to convert non-legal
content into legal content. We need to reject the simplistic picture in which
each law practice contributes to the content of the law a discrete proposition (or
set of propositions), which is the result of converting the non-legal content of
sentences and mental states into legal content. The bearing of non-legal content
on the content of the law is not mechanical. Once we root out any idea of a
mechanical conversion of non-legal content to legal content, it is clear that
something must determine which aspects of law practices are relevant to the
content of the law and what role those relevant aspects play in contributing to
the content of the law.

IV. Can Law Practices Themselves Determine How they

Contribute to the Content of the Law?
In this section, I consider the possibility that law practices can themselves deter-
mine how they contribute to the content of the law. I will argue that without stan-
dards independent of practices, practices cannot themselves adjudicate between
ways in which practices could contribute to the content of the law.
For convenience, let me introduce a term for a candidate way in which practices
could contribute to the content of the law. I will call such a way a model (short for
a model of the role of law-determining practices in contributing to the content of the
law).³¹ The rational-relation doctrine tells us that there are systematic, intelligible
connections between practices and the content of the law. It thus guarantees that
there are rules that, given any pattern of law practices, yield a total set of legal
propositions. A model is such a rule or set of rules.
A model is the counterpart at the metaphysical level of a method of interpreta-
tion at the epistemic level. (A model’s being correct in a given legal system is what
makes the corresponding theory of interpretation true.) Although the term is not
ideal, I use “model” rather than “method of interpretation” to signal that my con-
cern is constitutive or metaphysical, not epistemic; that is, the issue is how prac-
tices make it the case that the law’s content is what it is, not how we can ascertain
the law’s content from law practices. Because it is more idiomatic, however, I will
sometimes write in epistemic terms when discussing models.
(By way of analogy, it may be helpful to compare, on the one hand, the relation
between practices and the content of the law with, on the other, the relation
between words and the meaning of a sentence or group of sentences. The meaning
³¹ My thanks to Nicos Stavropoulos for suggesting this term.
246 Mark Greenberg

of a sentence depends in a systematic, intelligible way on the arrangement of con-

stituent words; analogously, the content of the law—in a given legal system at a
given time—depends on the pattern of law practices. A specification of the mean-
ings of individual words and of the compositional rules of the language is a speci-
fication of the rules by which the words determine the meaning of the sentence.
Analogously, a specification of a model is a specification of the rules by which law
practices determine the content of the law. In this sense, a model is the analogue of
the meanings of individual words and the compositional rules for the language.)
I will use the term “model” sometimes for a partial model—a rule for the
relevance of some aspect of law practices, for example, of legislative findings or of
dissenting judicial opinions, to the content of the law—and sometimes for a
complete model—all of the rules by which law practices determine the content of
the law. The context should make clear whether partial or complete models are in
question. The legally correct (or, for short, correct) model in a particular legal
system at a particular time is the way in which practices in that legal system at that
time actually contribute to the content of the law (not merely the way in which
they are thought to do so). Which model is correct varies from legal system to legal
system and from time to time within a legal system, since, as we will see, which
model is correct depends in part on law practices.
Models come at different levels of generality. More specific ones include the
metaphysical counterparts of theories of constitutional, statutory, and common-
law interpretation. Models can also be understood to include very general putative
ways in which law practices determine what the law requires. Thus Hart’s rule-
of-recognition-based theory of law and Dworkin’s “law as integrity” theory are
accounts of very general models. Very general models give rise to more localized
models of the contributions made by specific elements of practices.
Candidate models are candidate ways in which practices contribute to the con-
tent of the law. Since the issue of how practices contribute to the content of the law
has several components, models have several closely related roles: they determine
what counts as a law practice; which aspects of law practices are relevant to the
content of the law; and how different relevant aspects combine to determine the
content of the law, including how conflicts between relevant aspects are resolved.
The question of what determines how practices contribute to the content of the
law can therefore be reformulated as the question of what determines which mod-
els are correct. What settles, for example, the question whether the original-intent
theory of constitutional interpretation is true?
We can now turn to the main topic of this section: whether law practices can
themselves determine which model is correct. Certainly the content of the law, as
determined by law practices, concerns, in addition to more familiar subjects of
legal regulation, what models are correct. That is, the content of the law includes
rules for the bearing of law practices on the content of the law. For example, it is
part of the law of the United States that the Constitution is the supreme law, that
bills that have a bare majority of both houses of Congress do not contribute to the
 How Facts Make Law 247

content of the law unless the president signs them, and that precedents of higher
courts are binding on lower courts in the same jurisdiction.
The content of the law cannot itself determine which model is correct, however,
for the content of the law depends on which model is correct. If, for example,
statutes contributed to the law only the plain meaning of their words, the content of
the law would be different from what it would be if the legislators’ intentions made a
difference. Obviously, which legal propositions are true depends on which model is
correct. But as we have just seen, which model is correct depends in part on the legal
propositions. The content of the law and the correct model are thus interdependent.
This interdependence threatens to bring indeterminacy. Consider the law prac-
tices of a particular legal system at a particular time and ask what the content of
the law is. Suppose that if candidate model A were legally correct, a certain set of
legal propositions would be true, according to which model A would be correct.
And if candidate model B were correct, a different set of legal propositions would
be true, according to which model B would be correct. And so on. Without some
other standard, each mutually supporting pair of models and set of legal proposi-
tions is no more favored than any other pair.³²
Can law practices determine which model is correct? The prima facie problem
is that we cannot appeal to practices to determine which model is correct because
which model a set of practices supports itself depends on which model is correct.
But let us consider the matter in more depth. If practices are to determine which
model is correct, there are two possibilities.
First, a privileged foundational practice (or set of foundational practices) could
determine the role of other practices. This possibility encounters the problem of
how practices themselves can determine which practices are foundational. For
example, the fact that a judicial opinion states that only the rationale necessary to
the decision of a case is contributed to the content of the law cannot determine
that that is a correct account of the contribution of judicial decisions to the con-
tent of the law. Something must determine that the judicial opinion in question is
relevant and trumps other conflicting practices. A putatively foundational prac-
tice cannot non-question-beggingly provide the reason that it is foundational.

³² This note registers a rather technical qualification and can be skipped without losing the
main thread of the argument. A candidate model, given the law practices, may yield a set of legal
propositions that lends support to a different, inconsistent model. To the extent that this is the
case, we can say that the model is not in equilibrium (relative to the law practices). Models that are
in equilibrium (or are closer to it) are plausibly favored, other things being equal, over those that
are not (or are further from it). There is no reason to expect, however, that there will typically be only
one model that is closer to equilibrium than any other model. In fact, indefinitely many models are
guaranteed to be in perfect equilibrium (yet yield different sets of legal propositions). For example,
any model that includes a rule that practices (and thus the true legal propositions) have no bearing
on which model is correct is necessarily in perfect equilibrium. Without some independent stand-
ard for what models are eligible, there is no way to rule out such models. Hence the varying degree
to which different candidate models are in equilibrium does not ensure a unique correct model and
determinate legal content. See also the discussion of a coherentist solution in the text below.
248 Mark Greenberg

Moreover, it is unwarranted to assume that the significance of a putatively founda-

tional practice is simply its nonlegal content. Its significance depends on which
model is correct—the very issue the practice is supposed to resolve. In sum, a
foundationalist solution is hopeless because it requires some independent factor
that determines which practices are foundational (and what their contribution is).
Secondly, if no practices can be assumed to have a privileged status, the remain-
ing possibility is that all law practices together can somehow determine their own
role. Such a coherentist solution might at first seem to have more going for it than
the foundationalist one. The idea would be, roughly speaking, that the (total) law
practices support the model that, when applied to the practices, yields the result
that the practices support that very model. If no model is perfectly supported in
this way, the one that comes closest is the correct one.
The problem with this suggestion, crudely put, is that without substantive
standards that determine the relevance of different aspects of law practices, the
(total) law practices will support too many models. For any legal proposition,
there will always be a model supported by the practices that yields that proposi-
tion. Or to put it another way, the formal requirement that a model be supported
by or cohere with law practices is empty without substantive standards that deter-
mine what counts as a relevant difference. Suppose a body of judicial decisions
seems to support the proposition that a court is to give deference to an adminis-
trative agency’s interpretation of a statute. It is consistent with those decisions for
an agency’s interpretation of a statute not to deserve deference when there is a
reason for the different treatment. Such a reason could be, for example, that the
agency in the earlier cases, but not in the present case, had special responsibility
for administration of the relevant statutory scheme. But since the facts of every
case are different, if a model can count any difference as relevant, there will always
be a model that is consistent with all past practices yet denies deference to agency
interpretations of statutes.
As I have argued more fully elsewhere, such considerations show that practices
cannot determine legal content without standards independent of the practices
that determine which differences are relevant and irrelevant.³³ Hence law prac-
tices alone cannot yield determinate legal requirements. The point is a specific
application of a familiar, more general point that Susan Hurley has developed.³⁴
Formal requirements such as consistency are meaningful only in the light of sub-
stantive standards that limit which factors can provide reasons.
It would be missing the point to suggest that law practices themselves can deter-
mine the appropriate standards. Without such standards, a requirement of adher-
ence to practices is empty. In epistemic terms, we cannot derive the standards

³³ See M. Greenberg & H. Litman, The Meaning of Original Meaning, 86 Geo L.J. 569, 614–617
(1998).
³⁴ See S. L. Hurley, Natural Reasons 26, 84–88 (1989). Hurley credits Ramsey’s and
Davidson’s uses of arguments with similar import. See, e.g. D. Davidson, The Structure and Content of
Truth, 87 J. Phil. 279, 317–320 (1990).
 How Facts Make Law 249

from the practices because the standards are a prerequisite for interpreting the
practices.
It may be helpful to notice that the problem has a structure similar to that of
two famous philosophical puzzles: Nelson Goodman’s problem about green and
grue, and Saul Kripke’s problem about plus and quus.³⁵ In order for there to be
legal requirements, it must be possible for someone to make a mistake in attribut-
ing a legal requirement (if just any attribution of a legal requirement is correct, the
law requires that P and that not P and so does not require anything). One makes a
mistake when one attributes a legal requirement that is not the one the law prac-
tices yield when interpreted in accordance with the correct model. For any can-
didate legal requirement, however, there is always a non-standard or “bent” model
that yields that requirement. It is therefore open to an interpreter charged with a
mistake to claim that in attributing the legal requirement in question, she has not
made a mistake in applying one model but is applying a different model.
The proponent of the coherence solution will respond that law practices them-
selves support certain models. For example, in appealing to practices to decide
cases, courts have developed well-established ways of understanding the relevance
of those practices to legal content. The problem is that there will always be bent
models according to which the judicial decisions (and other practices) support the
bent models rather than the purportedly well-established ones. This kind of point
shows that there must be factors, not themselves derived from the practices, that
favor some models over others.
Here is an example.³⁶ Suppose that on February 1, 2005, a judge in a state court
in the United States must decide whether a woman has a federal constitutional
right not to be prevented from obtaining an abortion. Imagine that the judge
holds that the woman does not have such a right. It seems that the judge has mis-
read Roe v. Wade,³⁷ the seminal decision of the United States Supreme Court. The
judge claims, however, that according to the correct model of how judicial decisions
contribute to legal content, when constitutional rights of individuals are at stake
and strong considerations of justice support the claims of both sides, such deci-
sions should be understood as establishing a form of “checkerboard” solution.
According to such a solution, whether a person has the right in question depends
on whether the person is born on an odd- or even-numbered day.³⁸ Since Jane Roe
was born on an odd-numbered day (let us assume), Roe v. Wade’s contribution to
content is that only women born on odd-numbered days have a constitutional
right to an abortion.

³⁵ See N. Goodman, Fact, Fiction, and Forecast 72–81 (3d ed., 1973); S. Kripke,
Wittgenstein on Rules and Private Language 7–32 (1982). These puzzles involve concepts that
seem bizarre and gerrymandered. One challenge is to determine what it is that rules such concepts
out (at least in particular contexts), for if they are not ruled out, unacceptable results follow.
³⁶ The example borrows from Dworkin’s discussion of a “checkerboard” solution to the abortion
controversy. See Dworkin, supra note 6, at 178–186. Dworkin cannot be held responsible, however,
for my example. ³⁷ Roe v. Wade, 410 U.S. 113 (1973).
³⁸ See Dworkin, supra note 6, at 178–179.
250 Mark Greenberg

Before discussing the example, it must be emphasized that the point is not that
the judge’s position should be taken seriously; on the contrary, the example
depends on the fact that the judge’s position is plainly a non-starter. Since it is evi-
dent that the position cannot be taken seriously, there must be factors that rule out
models like the one in the example. The example makes the point that these fac-
tors must be independent of practices. Since the unacceptable positions that we
want to exclude purport to determine what practices mean, the factors that
exclude these positions cannot be based on practices. Moreover, there is no way to
rule out such positions on a purely logical level, since, as will become evident, it is
easy to construct self-supporting, logically consistent systems of such positions.
The claim is, then, that our unwillingness to take the judge’s position seriously
suggests that we must be depending on tacit assumptions independent of law
practices in determining which models are acceptable. Let us look at the example
to see why practices themselves cannot exclude the judge’s model.
The first objection to the judge’s position may be that the Supreme Court in
Roe v. Wade said nothing about the abortion right’s depending on birth dates. The
judge replies that according to the correct model, the reasons that judges give in
their opinions make no or little contribution to the content of the law. A second
objection may move to a different level: the practices of the legal system do not
support the judge’s model. Judicial decisions, for example, do not interpret the
contributions made by other decisions in such a checkerboard fashion, nor do
they ignore the reasons judges give. The judge, however, claims that according to
his model, judicial decisions have all along been using a bent model, according to
which the reasons judges give are significant until February 1, 2005, but not after-
wards. Similarly, the model specifies no checkerboard contributions to content
until that date, then requires them afterwards. All of the judicial decisions so far
are logically consistent with the hypothesis that they are using the bent model.
Obviously a third-level objection—that the practices do not support models that
give dates this sort of significance—can be met with the same sort of response.
In another version of the example, the judge might claim that according to
the correct model, in all cases involving the right to abortion, a Supreme Court
decision’s relevance to content ends, without further action by the Court, as soon
as a majority of the current Supreme Court believes that the decision was wrongly
decided. Since the judge believes that that is now the situation with regard to Roe
v. Wade, he claims that Roe v. Wade no longer has any bearing on the content of the
law. If it is objected that the judge’s position is not an accurate account of how
judicial decisions interpret past judicial decisions, the judge will claim that judi-
cial decisions have been following his model all along. Since (let us suppose) it
has never been the case before that a majority of the Supreme Court has disagreed
with a past Supreme Court decision on the right to abortion, the evidence of past
decisions supports the judge’s model, which treats only abortion rights cases
idiosyncratically, as strongly as a more conventional one.
 How Facts Make Law 251

The point should be obvious by now: these sorts of unacceptable models are
unacceptable because there are standards independent of practices that determine
that some sorts of factors are irrelevant to the contributions made by practices to
legal content. The practices themselves cannot be the source of the standards for
which models are permissible.
In this section, I have argued that practices themselves cannot determine how
practices contribute to the content of the law. Although I will not discuss the point
here, it is worth noting that my argument is not limited to the law. For example, the
argument shows that without standards independent of the practices, no set of
practices can rationally determine rules. What rules a set of practices rationally
determines will depend on what aspects of the practices are relevant and how those
aspects are relevant. And the practices cannot themselves resolve those issues.
Similarly, my argument does not depend on the complexities of contemporary
legal systems. My point therefore holds even for extremely simple cases. Even if
there were only one law-maker who uttered only simple sentences, and even if it
were taken for granted that the law-maker’s practices were legally relevant, the pre-
cise relevance of those practices would still depend on factors independent of the
practices. For example, there would still be an issue of whether the relevant aspect
of the practices was the meaning of the words uttered, as opposed to, say, the law-
maker’s intentions or the narrowest rationale necessary to justify the outcome of
the law-maker’s decisions.

V. Objections

I want now to consider three closely related objections. First, it may be objected to
that in practice there is often no difficulty in knowing which aspects of a practice
are relevant or which facts provide reasons. Bent models are not serious candi-
dates. Secondly, it may be objected to that practitioners’ beliefs (or other attitudes)
about value questions, not value facts, solve the problem of determining how
practices contribute to the content of the law. Thirdly, it may be said that in limit-
ing law practices to descriptive facts, I have relied on too thin a conception of law
practices. Properly understood, law practices can themselves determine the con-
tent of the law.
I replied to a version of the first objection in discussing the example of the abor-
tion-rights decision, but I will make the point in more general terms here. As I
have emphasized, the question of the necessary conditions for law practices to
determine the content of the law is a metaphysical, not an epistemic, question.
The problems that I have raised concerning how law practices determine the con-
tent of the law are not practical problems that legal interpreters encounter in try-
ing to discover what the law requires. Hence it is no objection to my argument
that legal interpreters do not encounter such problems.
252 Mark Greenberg

I have argued that there is a gap between law practices and the content of the
law that can be bridged only by substantive factors independent of practices. If
legal practitioners have no difficulty in crossing this gap—for example, in elimin-
ating bent models from consideration—that must be because they take the neces-
sary factors for granted. With respect to the example of the abortion-rights
decision, I argued that practices themselves cannot rule out the judge’s bent
models. Therefore our unwillingness to take the judge’s position seriously is evid-
ence that we are relying on tacit assumptions about what models are acceptable.
The lack of difficulty in practice suggests not that substantive constraints are not
needed but that they are assumed.
This point leads naturally to the second objection, which holds that it is the
assumptions or beliefs of participants in the practice that solve the problem of
how practices determine the content of the law. For example, it might be that a
consensus or shared understanding among judges or legal officials determines the
relevance of practices to the content of the law. Beliefs about value, not value facts,
do the necessary work.
As an epistemic matter, of course, we rely on our beliefs about value to ascertain
what the law is. But that is exactly what we would expect if the content of the law
depended on value facts. After all, in working out the truth in any domain, we
must depend on our beliefs. That we do so in a given domain in no way suggests
that the truth in that domain depends on our beliefs. Notice, moreover, that if the
content of the law depended on beliefs about value, then in order to work out
what the law was, we would have to rely on our beliefs about our beliefs about
value. For example, we might ask not whether democratic values favor intention-
alist theories of statutory interpretation, but whether there is a consensus among
judges that democratic values do so.
The most important point is that facts about what participants believe (under-
stand, intend, and so on) could not do the necessary work because such facts are
just more descriptive facts in the same position as the rest of the law practices. As
with the facts about the behavior of law-makers, we can ask whether facts about
participants’ beliefs are relevant to the content of the law, and if so, in what way.
Since the content of the law is rationally determined, the answers to these ques-
tions must be provided by reasons. As I have argued, the law practices, including
facts about participants’ beliefs, cannot determine their own relevance.
More generally, the same kind of argument explains why the questions of value
on which the content of the law depends must be resolved by substantive stand-
ards rather than by value-neutral procedures. In general, there are procedural ways
to resolve value questions—flipping a coin and voting are examples. Such pro-
cedures are in the same position as other law practices, however. There have to be
reasons that determine that a given procedure is the relevant one and what the
significance of the procedure is to the content of the law.
The third objection claims that the additional substantive factors are part of law
practices themselves. I have already addressed the suggestion that the law prac-
tices, conceived as facts about behavior and mental states, determine their own
 How Facts Make Law 253

relevance. The present objection is that my conception is too narrow. It somehow

fails to do justice to law practices to take them to consist of ordinary empirical
facts about what people have done, said, and thought. If the objection is to be
more than hand-waving, the objector needs to say what practices consist of
beyond such facts and how the enriching factor solves the problem. For example,
it would of course be no objection to my argument to claim that the descriptive
facts need to be enriched with value facts.
Another unpromising possibility, addressed in Section II.C above, is for the
objector to maintain that law practices are legal-content laden. According to this
version of the objection, facts about what counts as a legislature, who has author-
ity to make law, what counts as validly enacted, what impact a statute has on the
content of the law—in general, legal-content facts concerning the relevance of law
practices to legal content—are somehow part of the law practices. As argued,
however, unless legal content is to be metaphysically basic, there must be an
account of what determines legal content that does not presuppose it. It simply
begs the question to take law practices to include legal-content facts.
The objector challenges my conception of the law practices on the ground
that it is too restrictive. Here is one line of thought in support of my concep-
tion. We normally assume that law practices can be looked up in the law books.
But all that can be found in the law books, other than legal-content facts, are
facts about what various people—legislators, judges, administrative officials,
and so on—did and said and thought. If there is something else to law practices,
how do we know about it? To put the point another way, if I tell you all the facts
about what the relevant people said and did, believed and intended, you can
work out what the law is without knowing any more about the law practices. So
if there is an aspect of law practices other than these facts, it does not seem to
play a role in determining the content of the law. (It is true that you may have to
be skilled at legal reasoning to work out the content of the law, and that skill
may include an understanding of the significance of the practices to legal con-
tent. But I have already addressed the suggestion that it is participants’ under-
standings, rather than the substantive factors that are the subject of those
understandings, that do the necessary work.)

VI. The Need for Substantive

Factors Independent of Law Practices
I have argued that law practices cannot themselves determine the content of the
law because they cannot unilaterally determine their own contribution to the con-
tent of the law. There must be factors, independent of practices, that favor some
models over others. In this section, I sketch where this argument leaves us. In par-
ticular, I explain the sense in which the argument requires facts about value, and
the nature of the claimed connection between law and value.
254 Mark Greenberg

A. Value Facts?
In order for practices to yield determinate legal requirements, it has to be the case
that there are truths about which models are better than others independently of
how much the models are supported by law practices. Since practices must ration-
ally determine the content of the law, truths about which models are better than
others cannot simply be brute; there have to be reasons that favor some models
over others.
We have seen that law practices cannot determine their own contribution to the
content of the law. By contrast, value facts are well suited to determining the rele-
vance of law practices, for value facts include facts about the relevance of descriptive
facts. For example, that democracy supports an intentionalist model of statutes is, if
true, a value fact. What about the relevance of the value facts themselves? At least in
the case of the all-things-considered truth about the relevant values, its relevance
is intelligible without further reasons. If the all-things-considered truth about the
relevant considerations supports a certain model of the law practices, there can be no
serious question of whether that truth is itself relevant, or in what way. The signifi-
cance for the law of the fact that a certain model is all-things-considered better than
others is simply the fact that that model is better than others.
It might be suggested that an appeal to conceptual truth offers a way to avoid
the conclusion that the content of the law depends on value facts. The idea would
be that the concept of law (or some other legal concept), rather than substantive
value facts, determines that some models are better than others. As noted above,
conceptual truth is the kind of consideration that could provide reasons of the
necessary sort. The question is whether conceptual truth does so in the case of law.
My response begins with two points about what notion of conceptual truth this
kind of suggestion can rely on. According to what we can call a superficialist notion,
conceptual truths are truths about the use of concept-words, truths that are tacitly
known by all competent users of those words or are settled by community con-
sensus about the use of the words. Given such a notion of conceptual truth, we
should reject the idea that there are conceptual truths that can do the necessary
work. Ronald Dworkin famously argued that disputes about the grounds of law are
substantive debates, not trivial quarrels over the use of words.³⁹ Positivists have
generally responded by denying that they hold the kind of view Dworkin was
attacking. Thus, both sides agree that questions about which models are better than
others are not merely verbal questions that can be settled by appeal to consensus
criteria for the use of words. And both sides are correct on this point.
When, for example, Justices of the Supreme Court debate whether legislative
history is relevant to the content of the law, the dispute cannot be settled by appeal
to agreed-on criteria for the use of words. A lawyer or judge who challenges

³⁹ Dworkin, supra note 6, at 31–46.

 How Facts Make Law 255

well-established models is not ipso facto mistaken. For example, a lawyer could
advance a novel theory according to which New Jersey statutes make no contribu-
tion to the content of the law (on the ground, say, that there is a constitutional
flaw in New Jersey’s legislative process). The claim would not be straightforwardly
wrong merely because it goes against the consensus model, though it is likely mis-
taken on substantive grounds.
Secondly, we have seen that the practices of participants in the legal system can-
not be the source of the standards that support some models over others. It follows
that if conceptual truth is to be the source of the standards, conceptual truth must
not be determined by the practices of participants in the legal system; it must
depend on factors independent of our law practices.
The consequence of these two points is that if conceptual truth is to provide
the needed standards, it would have to be conceptual truth of a kind that is not
determined by consensus about the use of words and is not determined by our
law practices. I am sympathetic to such a notion of conceptual truth. Given
such a notion, however, it is not clear that an appeal to conceptual truth is a way
of avoiding the need for substantive value facts. Instead, the conceptual truths
in question may include or depend on value facts, for example, facts about fair-
ness or democracy. At this point, the burden surely rests on a proponent of the
conceptual-truth suggestion to offer a position that avoids the two problems
that I have just described without collapsing into a dependence on substantive
value facts.
A different kind of appeal to conceptual truth is possible. It could be argued not
that there are conceptual truths about which models are better than others, but
that conceptual truth determines that such issues are determined by a specific
internal legal value. This appeal to conceptual truth does not attempt to avoid the
need for value facts; it attempts to explain those value facts as internal to the law.
I will turn now to the nature of legal value facts. It is worth noting, however, that
an appeal to conceptual truth as the source of internal value facts will encounter the
same challenge as the appeal to conceptual truth to avoid the need for value facts.
Such an appeal requires an account of conceptual truth according to which truths
about the concept of law are independent of our law practices yet also independ-
ent of genuine value facts.
I have argued that the content of the law depends on substantive value facts.
What is the nature of those value facts? The most straightforward possibility is
that, other things being equal,⁴⁰ models are better to the extent that they are
favored by the all-things-considered truth about the applicable considerations—
the Truth, for short. In other words, the legally correct standard or value is simply
the truth about value. On this view, there is no special legal standard or value. For
example, the bearing of legislative history on the content of the law depends on

⁴⁰ “Other things being equal” because practices also play a role in determining which models are
better than others. See infra Section VI.B.
256 Mark Greenberg

considerations of democracy, fairness, welfare, stability—on every consideration

that is in fact relevant to the issue.
A second possibility is that, in the special context of the law, the all-things-
considered truth about the relevant considerations is that the standard for models
is not the general, all-things-considered truth about the relevant considerations
but some different standard. For example, it might be that, taking into account all
relevant considerations, the Truth is that the legally correct resolution of value
questions is the one that maximizes community wealth. According to this second
possibility, special legal value facts are genuine value facts; they are the consequence
of the application of genuine value facts—Truth—to the specific context of law.⁴¹
On this view, the fact that, say, wealth maximization is the virtue of models, is a
genuine value fact. A version of this possibility would allow the special legal value
facts to vary from legal system to legal system.
On the first and second possibilities, the content of the law depends on genuine
value facts in a way that is inconsistent with both hard and soft positivism. A posi-
tivist might try to argue that even if my argument so far is sound, there is a third
possibility. According to this possibility, there are substantive standards that
within the law do the work of value facts in resolving value questions but are not
genuine value facts. We might describe this possibility by saying that legal value
facts are internal to the law.
The hypothetical positivist’s suggestion that legal value facts are internal to the
law would have to mean more than that they have no application outside of law.
There could be legal value facts that were genuine value facts applicable only in
the legal context. In that case, the second possibility would be actual, and the con-
tent of the law would depend at base on genuine value facts. The third possibility
is supposed to avoid the conclusion that the content of the law depends on genu-
ine value facts. Perhaps the idea would be that legal value facts matter only to
those who are trying to participate in the legal system (and only to that extent).
(As with the second possibility, a version of the third possibility would allow that
the internal legal value can vary from legal system to legal system.)
I do not mean to suggest that the idea of internal legal values is unproblematic
or even fully coherent. I therefore do not need to explain exactly what it would
mean for there to be internal values. Nor do I need to explain what, other than the
Truth, could make it the case that there is a special legal value. I mention the idea
only because it seems to have some currency in philosophy-of-law circles. My
point is simply that I do not claim in this paper to have ruled out the view that the
content of the law depends on internal value facts rather than genuine ones.
I will briefly comment on the problems facing this view. We have already ruled
out the possibility that law practices determine their own relevance to legal

⁴¹ The position Dworkin calls “conventionalism” could be advanced as a version of possibility

two, though that is not exactly the way in which he presents it. See Dworkin, supra note 6, at
114–150.
 How Facts Make Law 257

content. Therefore something other than law practices would have to determine
the internal value standard—to make it the case that this standard was the relevant
one for the law (or for the particular legal system). It is difficult to see what that
could be other than the relevant considerations—the Truth. If we appeal to the
Truth, however, we have returned to the first or second possibility.
Any account of internal value facts thus faces a challenge of steering between
the law practices on the one hand and the Truth on the other. I have already
described the way in which an attempt to ground internal legal facts in concep-
tual truth faces this challenge. But the challenge confronts any account of
internal value facts. For example, suppose a theorist appeals to the function of
law or legal systems to ground internal value facts. On the one hand, as we saw
with conceptual truth, if the law’s function is going to provide the value facts
necessary for practices to determine the content of the law, that function must
be determined by something independent of the law practices. On the other
hand, if the law’s function is determined by the all-things-considered truth
about the relevant factors, an appeal to function is not a way of avoiding an
appeal to genuine value facts. Until we have an account of internal value facts
that meets the challenge, it is difficult to evaluate the potential of an appeal to
internal value facts.
An internal-value view faces a more substantive challenge as well. Internal
value facts would have to have appropriate consequences for the nature of law.
In a normal or properly functioning legal system, the content of the law pro-
vides reasons for action of certain kinds for certain agents. Whether the content
of the law can provide such reasons may depend on the nature and source of the
legal value facts. For example, it is plausible that for a legal system to be func-
tioning properly, the content of the law must provide genuine reasons for action
for judges. An internal-value theorist must explain how legal content deter-
mined exclusively by law practices and internal value facts can provide genuine,
as opposed to merely internal, reasons for action. More generally, we can investi-
gate the nature of legal value facts by asking what role such facts must play in a
theory of law.

B. The Role of Value Facts

Let us now turn to the role of value facts in determining the content of the law.
Since I do not want to beg the question against the possibility of a special legal
value (whether internal or not), I use “X” for that property in virtue of which
models are better than others. X might be, for example, (the promotion of ) wealth
maximization, the maintenance of the status quo, security, fairness, or morality.
(If there is no special legal value, X is the Truth, in the technical sense explained
above.) Note that the fact that a particular model is favored by X may be a descrip-
tive fact (for example, if X is wealth maximization). In that case, the relevant value
fact is that X is what the goodness of models consists in.
258 Mark Greenberg

I will make two clarifications about the role of X and then consider the implica-
tions for the relation between law and value. The first point is that X only helps to
determine which models are correct. X’s favoring model A over model B is neither
necessary nor sufficient for A to win out over B. As we saw in Section IV, practices
play a role in determining which model is better. Hence the model that is best all
things considered may not be the same as the model that is ranked highest by X
alone. (For simplicity I sometimes omit this qualification.)
In Section IV we discussed the interdependence between models and legal con-
tent. We saw that if we hold law practices constant, different candidate models
yield different sets of legal propositions. Without X, each mutually supporting
pair of model and set of legal propositions is as favored as any other such pair, and
indeterminacy threatens. X’s independence makes it possible for the interdepend-
ence of model and legal content not to lead to global indeterminacy.
In particular, what bearing practices have on the legally correct model depends
on which model is most X-justified in advance of any particular practices. For X
constrains the candidate models of practices and thus makes it possible for prac-
tices to determine anything. Practices themselves have something to say about the
second-order question of how practices contribute to the content of the law. But X
helps to determine what practices have to say on that question. Roughly speaking,
the legally correct model is the one that is most X-justified after taking into account
practices in the way that it is most X-justified to take them into account.⁴² In
other words, the legally correct model is the one that is most X-justified, all things
considered.
The second point can be brought out with an objection. Suppose it is
objected that X need determine only what considerations are relevant to the
content of the law but need not go further and determine how conflicts between
relevant considerations are to be resolved. According to this suggestion, X
would eliminate some candidate models as unacceptable but would have noth-
ing to say between models that give weight only to relevant aspects of law prac-
tices. The objector grants my argument that without an independent standard
of relevance, practices could not determine which models were correct. The
objector points out, however, that once we have an independent standard of
relevance, practices themselves might be able to determine which models are
correct.

⁴² In many legal systems, the practices, when taken into account in the way that is most X-justified
in advance of the practices, will support a model that is not the most X-justified in advance of the
practices. And when taken into account in accordance with that model, the practices may support yet
a different model. The question therefore arises of how important it is for a model to be supported by
the practices (taken into account in accordance with that model). (In the terminology of supra note
32, the more that a model is supported by the practices, the more the model is in equilibrium.) Since
X is the virtue of models, X is what determines how important it is for a model to be supported by the
practices. This is why it is fair to say, as I do in the text, that the legally correct model is the one that is
most X-justified after taking into account the practices in the way that it is most X-justified to take
them into account.
 How Facts Make Law 259

Here is a brief sketch of a reply to the objector. In order for there to be determin-
ate legal requirements, X must do more than determine what considerations are
relevant; X must favor some resolutions of conflicts between relevant considera-
tions over others. Otherwise, given the diversity of relevant considerations and the
complexity of factual variation, law practices will not yield much in the way of
determinate legal requirements. Inconsistent propositions of law (and inconsist-
ent models) will typically have some support from relevant aspects of law prac-
tices. Therefore, in order for there to be determinate legal requirements, X must
not only help to determine what considerations are relevant but must also help to
determine the relative importance of elements of law practices and how such
elements interact.
In fact there is a deeper problem with the objection. It assumes that there are
discrete issues of what considerations are relevant to the content of the law and
how the relevant considerations combine to determine the content of the law. It
may be convenient to separate the two kinds of issues for expository purposes, but
we should not be misled into thinking that they are resolved separately. It is not
the case that there is an initial, all-or-nothing determination of whether a type of
consideration is relevant and then an independent, further determination of the
relative importance of the relevant considerations. Rather, the reason that a con-
sideration is relevant determines how and under what circumstances it is relevant
and how much force it has relative to other considerations.
For example, legislative history’s relevance to the content of the law derives, let
us suppose, from its connection to the intentions of the democratically elected rep-
resentatives of the people. Thus, in order to determine how important legislative
history is relative to other factors, we need to ask exactly how it is related to the rele-
vant intentions and what the importance of those intentions is. The point is that
the contribution to content of some aspect of a law practice and how it interacts
with other relevant aspects depend on why the aspect is relevant. If this suggestion—
that relevance and relative importance are not independent questions—is right, then
in helping to determine the relevance of various considerations, X will necessarily
(be helping to) resolve conflicts between relevant considerations.
I have argued that there is a certain kind of connection between law and value. I
would like to conclude by saying something about the implications of this con-
nection. Just for the purpose of exploring these implications, I will assume that X
is morality. The point of this assumption is to make clear that even if morality were
the relevant value, the consequences for the relation between law and morality
would not be straightforward. As I will show, it would not follow that the content
of the law would necessarily be morally good or even that the moral goodness of a
candidate legal proposition would count in favor of the proposition’s being true.
First, although (by assumption) morality provides legally relevant reasons
independent of the content of the law, the legally correct model is not simply
whatever model is morally best (or most justified). “Morally best” here means
most supported or justified by moral considerations in advance of consideration of
260 Mark Greenberg

the practices of the legal system. The legally correct model need not be the morally
best one in this sense because, as we have seen, practices also have an impact on
which model is legally correct.
Secondly, morally good models do not guarantee morally good legal proposi-
tions. Even if the legally correct model was a highly morally justified one, the
content of the law might be very morally bad. A democratically elected and
unquestionably legitimate legislature could publicly and clearly promulgate
extremely unjust statutes, such as a statute ostensibly excluding a racial minority
from social welfare benefits. The judicial decisions may rely on highly morally
justified models, ones that, among other things, give great weight to such morally
relevant features of legislative actions as the clearly expressed intentions of the
elected legislators. The most justified model, all things considered, will be a
morally good one yet will yield morally bad legal content. In fact, in such a legal
system less justified models could yield morally better legal content than more
justified models. (In such cases, a judge might sometimes be morally obligated to
circumvent the law by relying on the less justified model.)⁴³
Although morally justified models do not guarantee morally good legal prop-
ositions, it might be suggested that part of what makes a model morally justified is
that it tends to yield morally good legal requirements.⁴⁴ For example, assume that,
other things being equal, a legal requirement is morally better the more it treats
people fairly. Some models will in general have a greater tendency to yield legal
requirements that treat people fairly. According to the suggestion under consider-
ation, that a model has such a tendency would be one factor supporting that
model.
Suppose that the suggestion were correct. According to one line of thought, it
follows that the content of the law would simply be whatever it would be morally
good for it to be (or more generally, whatever it would be most X-justified for it to
be). In that case the practices would be irrelevant. This line of thought might
therefore be taken to provide a reductio of my argument for the role of value in
determining the way in which practices contribute to the content of the law.
The line of thought is not sound, however. First, even if the tendency of a
model to yield morally good legal propositions counts in favor of that model, a
variety of other moral considerations favor models that make the content of the

⁴³ The relation between a judge’s moral obligations and morally justified models raises interesting
issues, but space does not permit discussion.
⁴⁴ At the extreme, for example, a model could hold that in some circumstances the goodness of a
candidate legal proposition tips the balance in favor of that legal proposition and against competing
candidates. (A different way to describe such a position would be to say that value not only can help
to determine which model is best, thus indirectly favoring some candidate legal propositions over
others, but also can favor candidate legal propositions directly. I will not use this terminology.) As I
say in the text, such a model may be less supported both by morality and by practices than models
that give less weight to content-oriented considerations. I suggest below (see the last four paragraphs
of this section), that the role that such a model assigns to value facts is outside the role that this paper’s
arguments support.
 How Facts Make Law 261

law sensitive to relevant aspects of law practices. A model may be morally better,
for example, to the extent that it respects the will of the democratically elected rep-
resentatives of the people, protects expectations, enables planning, provides notice
of the law, treats relevantly similar practices similarly, minimizes the opportunity
for officials to base their decisions on controversial beliefs, and so on.
Roughly, we have a distinction between content-oriented considerations and
practice-oriented considerations. The relative weight accorded by morality to
these two kinds of considerations is a question for moral theory that I will not take
up here. On any plausible account, however, morality will give substantial weight
to practice-oriented considerations. So the morally best model (considered in
advance of law practices) will make the law sensitive to relevant aspects of law
practices.
Secondly, as we have seen, the legally correct model also depends on the law
practices. Apart from the weight that morality gives to practice-oriented consider-
ations, the practices themselves may support models that make the law sensitive to
practices. (Contemporary positivists, my primary target in this paper, are likely to
be sympathetic to the view that practices support models that make the law sensi-
tive to practices.) For example, although I will not defend the claim here, in the
U.S. and U.K. legal systems, practices themselves strongly support models that
make the law sensitive to law practices. Practices are thus a second reason that the
role of value need not have the consequence that the all-things-considered best
model will be one that tends to yield morally good legal propositions. (Also, even
a model that has a tendency to produce morally good legal propositions may not
do so, given the law practices of a particular legal system.)
Thirdly and finally, if we reflect on the argument for value’s role in determining
the content of the law, we see that it supports only a limited role for value, one that
does not involve supplanting law practices or making them irrelevant. Our start-
ing point was that law practices must determine the content of the law and that
they must do so by providing reasons that favor some legal propositions over
others. The crucial step in the argument was that law practices cannot provide
such reasons without value facts that determine the relevance of different aspects
of law practices to the content of the law. The argument thus supports the involve-
ment of value facts in determining the content of the law only for a limited role:
determining the relevance of law practices to the content of the law.
We can apply this point to the specific question of to what extent a legal prop-
osition’s goodness can help to make it true: the goodness (in terms of morality or
of value X) of a candidate legal proposition is relevant to the proposition’s truth
only to the extent that its goodness contributes to making it intelligible that an
aspect of a particular law practice has one bearing rather than another on the content
of the law. I will call this the relevance limitation.
I want to emphasize that the point is only that the argument of this chapter
supports no more than such a limited role for value facts; the argument does not
show that the role of value facts must be so limited. Whether there is some other
262 Mark Greenberg

or more expansive role for value in determining the content of the law is left open.
This chapter’s argument for the conclusion that value facts play a role in determin-
ing legal content is that value facts are needed in order to determine the relevance
of law practices to the legal content. The argument therefore supports only that
role for value facts. There might, of course, be a different argument that shows,
say, that morality or some other value supplants the law practices (though of
course almost no contemporary legal theorist, least of all one of my positivist tar-
gets, thinks that there is such an argument).
Let us consider more specifically the implications of the relevance limitation.
The limitation does not imply that the goodness of a legal proposition can never
be relevant to its truth.⁴⁵ The goodness of a legal proposition will be relevant to
the extent that it has a bearing on the intelligibility of law practices’ supporting
that legal proposition over others.
A Dworkinian theory of law provides a helpful example.⁴⁶ Consider a model
according to which law practices contribute to the content of the law precisely that
set of legal propositions that best justifies those law practices. Whether this model
respects the relevance limitation will depend on the notion of justification involved
in the Dworkinian model. Consider a simplistic understanding of justification that
has the following implication: the set of propositions that best justify the law prac-
tices is that set that results from taking the morally best set of propositions and
carving out specific exceptions for the law practices of the legal system—exceptions
tailored in such a way as to have no forward-looking consequences. On this under-
standing of justification, the model would not respect the relevance limitation,
because value facts would not determine the significance of the practices; instead,
the practices would simply be denied any significance by a kind of gerrymandering.
On a more sophisticated notion of justification, to the extent that a legal pro-
position is bent or gerrymandered, it will be less good at justifying law practices.
(In the extreme case just considered, where a particular law practice is simply
treated as an exception without further application, that practice is not justified at
all by the propositions to which it is an exception.) I think it is plausible, though I
will not argue the point here, that, given a proper understanding of justification,
the Dworkinian model I have described respects the relevance limitation. (Below
I will consider a different model, often attributed to Dworkin, that arguably does
not respect that limitation.)

⁴⁵ It is easy to see that the goodness of a legal proposition could have evidentiary relevance to the
content of the law. Suppose that the intention of legislators matters to the content of the law. If there
is reason to believe that the legislators would have intended what is morally better (at least other
things being equal), the moral goodness of candidate legal propositions will have a bearing on their
truth because it will have a bearing on what the legislators intended. The discussion in the text con-
cerns the question of whether the goodness of candidate propositions can have constitutive rather
than evidentiary relevance.
⁴⁶ I say “a Dworkinian theory” rather than “Dworkin’s theory” to avoid questions of Dworkin
exegesis. I believe that the position I describe is the best understanding of Dworkin’s position. See also
infra note 47.
 How Facts Make Law 263

The relevance limitation implies that the goodness of a legal proposition is

never sufficient to make it true. That value facts are needed to determine the con-
tribution of law practices to the content of the law does not provide a basis for
making law practices irrelevant. To put it another way, that a candidate proposi-
tion is a good one does not make it intelligible that the law practices, regardless of
what they happen to be, support that proposition. It might be tempting to regard
a model on which the goodness of a legal proposition can, at least in some circum-
stances, be sufficient to make it true as the degenerate or limiting case of a model
that determines the relevance of law practices to the content of the law. The model
determines that in the relevant circumstances, practices have no relevance. But
though this description may be formally tidy, the argument that value facts are
needed to enable law practices to determine the content of the law provides no
support for a model on which value facts can make practices irrelevant. In other
words, though we can describe a putative “model” according to which practices
provide a reason favoring any particular set of legal propositions (the morally best
ones, for example), it does not follow that practices could provide such a reason.
What reasons practices provide is a substantive, not a formal, question.
We can apply this point to an intermediate case. Consider a model that
includes rules for the contribution of law practices to the content of the law but
also includes a rule of the following sort:
(R) If more than one legal proposition is supported by the (total) law prac-
tices (given the other rules of the model) to some threshold level, the legal
proposition that is morally best (of those that reach the threshold) is true.⁴⁷
I suggest that R is not supported by this paper’s argument for the role of value. In
general, that legal proposition A has morally better content than legal proposition
B does not ipso facto make it intelligible that law practices support A over B.
Adding the hypothesis that law practices provide strong support for both A and
B—support above some threshold level—does not change this conclusion. A

⁴⁷ Dworkin sometimes seems to suggest such a rule. See, e.g., Dworkin, supra note 6, at 284–285,
387–388; R. Dworkin, Taking Rights Seriously 340, 342 (1977). And his commentators typi-
cally interpret him in this way. See, e.g., L. Alexander & E. Sherwin, The Rule of Rules ch. 8
(2001); J. Finnis, On Reason and Authority in Law’s Empire, 6 Law & Phil. 357, 372–374 (1987);
Raz, supra note 3, at 222–223. I think that this is not the best understanding of Dworkin’s view (and
Dworkin has confirmed as much in conversation). On the best understanding, fit is merely one aspect
of justification, there is no threshold level of fit, and how much fit matters relative to other aspects of
justification is a substantive question of political morality. (The idea of a threshold of fit that inter-
pretations must meet to be eligible and beyond which substantive moral considerations become rele-
vant should be taken as merely a heuristic or expository device.) See R. Dworkin, A Matter of
Principle 150–151 (1985); R. Dworkin, “Natural” Law Revisited, 34 U. Fla. L. Rev. 165, 170–173
(1982); R. Dworkin, supra note 6, at 231, 246–247. A different point is that Dworkin sometimes
seems to suggest that there is an aspect of the question of the extent to which interpretations fit law
practices that is purely formal or at least not normative. See, e.g., Dworkin, Taking Rights
Seriously, at 107 (suggesting that how much an interpretation fits is not an issue of political philoso-
phy); see also Dworkin, Taking Rights Seriously, at 67–68 (perhaps suggesting that there are
aspects of institutional support that do not depend on issues of normative political philosophy).
264 Mark Greenberg

moral reason for favoring proposition A over proposition B is not itself a reason
provided by law practices, since it is independent of law practices. If this argument
is right, my argument for the role of value facts does not support a role like that
captured by R—one in which there is room for value facts to favor one legal
proposition over another independently of law practices. (Again, however, the
point is only that this paper’s argument does not support such a role for value
facts, not that such a role is necessarily illegitimate.)
In sum, even if value X were morality, it would not follow that the most morally
justified model would be legally correct, and even a morally justified model would
not guarantee morally good legal requirements. It is no part of the role of value
argued for in this chapter that the goodness of a proposition ipso facto counts in
favor of the proposition’s truth. The role of value is in determining the relevance of
law practices to the content of the law.

VII. Conclusion

I have argued that law practices, understood in a way that excludes value facts,
cannot themselves determine the content of the law. Different models of the con-
tribution of practices to the content of the law would make it the case that differ-
ent legal propositions were true, and a body of law practices cannot unilaterally
determine which model is correct. In order for there to be determinate legal
requirements, the content of the law must also depend on facts about value.
What is the role of such value facts? I have suggested that they support some
models over others—that is, they help to determine which features of law prac-
tices matter and how they matter. It is not that the goodness of a candidate legal
proposition counts in favor of its truth. Rather, the role of value is in helping to
determine how practices contribute to the content of the law. This chapter does
not attempt conclusively to rule out the view that the needed legal value facts are
internal to law. I have argued, however, that the proponent of such a view must
overcome significant obstacles to explain how internal legal value facts could be
independent of both law practices and genuine value facts. The chapter also
suggests a way forward: We can ask what the nature and source of legal value facts
must be in order for law to have its central features, for example, for a legal system
to be able to provide certain kinds of reasons for action.
 11
Hartian Positivism and Normative Facts:
How Facts Make Law II
Mark Greenberg

I. Introduction
In How Facts Make Law and other recent papers,¹ I argue that a full constitutive
account of the content of the law—of legal facts—must appeal to normative facts.
The project of HFML is to defend this position without assuming that legal facts
are themselves normative facts. The argument’s engine is a requirement that a con-
stitutive account of legal facts must meet. According to this rational-relation
requirement,² it is not enough for a constitutive account of legal facts to specify
non-legal facts that modally determine the legal facts. The constitutive determinants
of legal facts must provide reasons for the obtaining of the legal facts (in a sense
of “reason” that I develop). In HFML,³ I argue that non-normative, contingent
facts—descriptive facts, for short—do not provide such reasons without normative
facts.⁴
In the present paper, I focus on the rational-relation requirement. I deploy it in
three related projects. First, I respond to a family of objections that challenge me
to explain why normative facts and descriptive facts together are better placed to
provide reasons for legal facts than descriptive facts alone.⁵ A unifying theme of

¹ M. Greenberg, How Facts Make Law, supra Chapter 10 [hereinafter HFML]. See also
M. Greenberg, Reasons Without Values?, in 2 Social, Political, and Legal Philosophy (E.
Villanueva ed. in press 2006); Mark Greenberg, On Practices and the Law, 12 Legal Theory (in press
2006).
² In HFML, id., I use the term rational-relation doctrine. For elaboration, see infra text accompany-
ing notes 12–19. ³ HFML, supra note 1.
⁴ This paper was written substantially later than HFML, and my terminology has shifted slightly.
I use the term “normative facts” here in the way that I used “value facts” in HFML. Thus, I use the
term to include what are sometimes called “evaluative facts” such as facts about what is good or bad.
For further explanation of the notion of a normative fact, see HFML, supra note 1, note 22. I explain
another minor terminological shift in infra note 8 and accompanying text.
⁵ I am particularly grateful to Gideon Rosen for his written comments on HFML for the
Exploring Law’s Empire conference held at Princeton University in September 2004. In this paper, I

Hartian Positivism and Normative Facts: How Facts Make Law II. Mark Greenberg.
© Oxford University Press 2006. Published 2006 by Oxford University Press.
266 Mark Greenberg

the objections is that explanations have to stop somewhere; descriptive facts, it is

suggested, are no worse a stopping place than normative facts. For example, one
objection maintains that if we need a reason why descriptive facts have a particular
bearing on the legal facts, we need a reason why normative facts do so. Another
claims that any account of law will have to rely on some kind of brute fact about
law—in particular, one that can serve a bridge principle linking non-legal facts
and legal facts. If a Hartian account of legal facts requires a bridge principle link-
ing officials’ dispositions and attitudes to legal facts, a normative account of legal
facts will require a bridge principle linking normative facts to legal facts.
Rather than considering such objections in the abstract, I consider an inter-
locutor who uses the objections to defend a Hartian account of law.⁶ I choose a
Hartian account because it is the most influential version of legal positivism. The
second main project of the paper, accordingly, is to use the rational-relation
requirement to show why a Hartian account of law fails.
Thirdly, I spell out a consequence of the rational-relation requirement: if an
account of what, at the most basic level,⁷ determines legal facts is true in any pos-
sible legal system, it is true in all possible legal systems. For example, if a Hartian
account of legal facts is true in any possible legal system, it is true in all possible
legal systems. I use this all-or-nothing result in my critique of a Hartian account,
but the result is of interest in its own right. For example, a familiar strategy for
legal positivists is to argue that because Ronald Dworkin’s arguments against legal
positivism rely on features of the U.S. and U.K. legal systems, those arguments
cannot establish conclusions about all possible legal systems; in particular, they
cannot rule out the possibility of a Hartian legal system. My all-or-nothing result
makes this positivist strategy unavailable.

draw especially on his clear and powerful formulation of the objection. See G. Rosen, Comments on
Mark Greenberg, How Facts Make Law (unpublished manuscript, on file with author). Others who
have raised versions of the objection include Scott Hershovitz, Tom Nagel, Ram Neta, Ori Simchen,
and Enrique Villanueva. I would like to thank Scott Hershovitz, Herb Morris, Ram Neta, Keemin
Ngiam, Seana Shiffrin, David Sosa, and Nicos Stavropoulos for comments on a draft. I am indebted
to Scott Shapiro for many valuable discussions. I’m especially grateful to Scott Hershovitz for encour-
aging me to write this paper and for creating this volume.
⁶ I say “a Hartian account” rather than “Hart’s account” because I try to address the most powerful
and plausible version of a position in the neighborhood of Hart’s, rather than to be faithful to the
details of Hart’s own view. I will largely ignore questions of exegesis of Hart.
⁷ The point of the qualification “at the most basic level” is to exclude determinants of the legal
facts the relevance of which depends on the contingent facts of the legal system—on, to use a term
defined in the text two paragraphs below, the law practices of the legal system. As I explain in HFML,
law practices have an impact on the contribution of law practices to the content of the law. But in
order to satisfy the rational-relation requirement, there have to be additional determinants of the legal
facts, independent of the law practices. See HFML, Sections IV–V. These determinants are the ones
“at the most basic level.” For example, inclusive legal positivists think that normative facts can be con-
stitutive determinants of legal facts because of the dispositions and attitudes of legal officials.
According to inclusive legal positivism, therefore, the role of normative facts is not at the most basic
level. The ultimate issue in HFML and the present paper is whether normative facts must be among
the most basic constitutive determinants of the content of the law. See HFML, supra note 1. I will usu-
ally omit the qualification “at the most basic level.”
 Hartian Positivism and Normative Facts: How Facts Make Law II 267

It is important that the all-or-nothing result does not depend on the ultimate
conclusion that a constitutive account of legal facts must appeal to normative facts,
but follows immediately from the rational-relation requirement. (After all, my
ultimate conclusion obviously implies that a Hartian legal system is not possible.)
The rational-relation requirement is a relatively weak premise, which, I claim,
most legal theorists implicitly accept. It therefore may be surprising that it rules
out an ecumenical position according to which some possible legal systems are
Hartian and some are not.
A bit of terminology will be helpful. A legal fact⁸ is a true legal proposition—a
fact about the content of the law. That contracts for the sale of land must be in
writing is a legal fact in many legal systems. Law-determining practices, or law prac-
tices for short, are ordinary empirical facts (paradigmatically, facts about the say-
ings and doings of members of constitutional assemblies, legislatures, courts, and
administrative agencies) that are determinants of the content of the law. I call the
relevant facts “law-determining practices” rather than “legal decisions” because
the term “decisions” tends to suggest judicial decisions in particular.⁹
A model of the contribution of law practices to the content of the law—or, for short,
a model—is a (putative) way in which law practices contribute to the content of
the law. A model is thus the metaphysical counterpart of a method of interpreta-
tion. For example, a Hartian rule of recognition (understood as constitutive of
legal facts rather than as a way of identifying legal facts) is a candidate model.
The correct model (in a given legal system at a given time) is the way in which law
practices actually contribute to the content of the law (in the legal system at that
time), not merely the way in which they are taken to do so. Note that correctness is
therefore legal correctness, not, for example, moral rightness.
In the next section (Section II), I lay out the structure of the argument of HFML.
In Section III, I give a brief account of why a Hartian account of legal facts fails to
satisfy the rational-relation requirement. In Section IV, I show that a Hartian
account requires it to be true in all possible legal systems that acceptance of a rule of
recognition by officials makes that rule the correct model for the legal system. In
Section V, I use this result to undermine an appeal to Hartian intuitions. Our reflect-
ive understanding of law does not support the Hartian position that acceptance of a

⁸ In HFML, I use the uglier term “legal-content fact.” HFML, supra note 1, at 226.
⁹ For ease of exposition, I use legal-content-laden terms, such as “legislature” and “court” in char-
acterizing law practices. Strictly speaking, however, the law practices should be understood to be the
underlying descriptive facts in virtue of which the relevant legal-content facts obtain. Since legal facts
are not basic, there must be non-legal facts that constitute the legal-content-laden practices. These
more basic facts will include descriptive facts—the facts that I am calling “law practices.” For example,
the fact that a legislature enacted a statute must hold in virtue of complex descriptive facts about
people’s attitudes and behavior and perhaps also normative facts. (If, in order to account for legal-
content-laden practices, we have to appeal not merely to descriptive facts but also to normative facts,
so much the worse for the positivist thesis that the content of the law depends only on descriptive
facts.) The convenience of talking as if law practices consisted in legal-content-laden facts about the
behavior of legislatures, courts, and so on should not obscure the fact that there must be more basic
facts in virtue of which the legal facts obtain. For elaboration of the notion of a law practice, see
HFML, supra note 1, Section II.C.
268 Mark Greenberg

rule of recognition by officials makes that rule the correct model for the legal system.
Section VI shows that an account that appeals to normative facts does not run into
the problems faced by the Hartian account. I conclude in Section VII.

II. The Argument of How Facts Make Law

This section sketches the structure of the argument of HFML. It will be helpful
to present the argument in a slightly different form from that in which it appears
in HFML.
The position for which I argue in HFML can be described as follows: in any
legal system that has a certain three features, a full account of what determines¹⁰
the content of the law will make reference to normative facts.
The three features are captured by the following premises:
(1) Every legal fact is determined in part by law practices.
(2) There are many legal facts.¹¹
(3) Every legal fact is rationally determined by non-legal facts.
I think that these three premises are true in many contemporary legal systems,
including those of the United States and the United Kingdom. Proposition (3) is the
crucial one. The notion of rational determination is explained in HFML.¹² In brief,
a full constitutive account of the legal facts must do more than specify constitutive
determinants that modally determine the legal facts; the constitutive determinants
must constitute reasons why the legal facts obtain. Reasons, in the relevant sense, are
considerations that make the explanandum intelligible in rational terms, as opposed
to, say, emotional or aesthetic ones.¹³ In other words, the relation between the con-
stitutive determinants and the legal facts must be rationally intelligible.¹⁴ This is the
rational-relation requirement mentioned in the Introduction.
¹⁰ I use “determination” (“determines,” “determinants,” etc.) for constitutive, rather than modal,
determination. (When I mean modal determination, I will be explicit.) Thus, from the fact that the Y
facts supervene on the X facts (or the X facts fix the Y facts), it does not follow that the X facts are the
only constitutive determinants of the Y facts (or even that they are constitutive determinants of the Y
facts at all). For more on constitutive and modal determination, see M. Greenberg, A New Map of
Theories of Mental Content: Constitutive Accounts and Normative Theories, 15 Philosophical Issues
299 (2005).
¹¹ The point of this premise is to ensure that the conclusion (6), see infra 271, is not merely
vacuously true.
¹² HFML, supra note 1, 231–234. For elaboration, see also Greenberg, On Practices and the Law,
supra note 1. ¹³ See HFML, supra note 1, at 227–228, 231–234, 237–240.
¹⁴ See HFML, supra note 1, at 231–232. We could dramatize this point by saying that a full con-
stitutive account of legal facts must do more than specify the mapping or function from non-legal
facts to legal facts. It has to specify considerations that make it intelligible in rational terms why a
particular function is the operative one. But this way of characterizing the rational-relation require-
ment should not be understood to suggest that there are two fundamentally different kinds of
constitutive determinants—first-order determinants that are the arguments of the function and
second-order determinants that explain the function from first-order determinants to legal facts. See
infra text accompanying notes 21–22 and note 57 and accompanying text.
 Hartian Positivism and Normative Facts: How Facts Make Law II 269

Because of the rational-relation requirement, rational determination contrasts

sharply with constitutive determination in general. For in general, it is an open
possibility that the best we can do to explain why certain facts of a target domain
obtain is to specify the mapping or function from determining facts to target facts.
The determining facts need not provide reasons for the target facts. In HFML, I
use esthetic facts as an example.¹⁵ Arguably, the facts about the arrangement of
paint on a canvas need not provide reasons for the esthetic facts that they deter-
mine. A small difference in the arrangement of paint might make a clumsy scene
elegant, without providing a reason for the difference. In contrast, it cannot be a
brute fact that, say, a particular change in the wording of a statute would have a
particular impact on the legal facts.
Before completing the summary of the argument’s structure, I want to make
three clarificatory points about the rational-relation requirement. First, the rational-
relation requirement is only a necessary, not a sufficient, condition on the deter-
minants of the legal facts. It is therefore possible that a fact (or sets of facts) could
satisfy the rational-relation requirement, but not be a constitutive determinant of
the legal facts.
Secondly, I am sometimes asked whether putative or perceived (normative)
facts that turn out not to be facts at all could satisfy the rational-relation require-
ment.¹⁶ We must, of course, distinguish between beliefs (or facts about beliefs)
and the propositions that are believed. The present issue is not whether (false)
beliefs about normative matters (or the fact that people have those beliefs) could
satisfy the rational-relation requirement. That someone has a certain belief, what-
ever its content, and whether it is true or false, is an ordinary empirical fact, not a
putative normative one. In HFML, I consider and reject the idea that beliefs about
normative matters can substitute for normative facts in a constitutive account of
legal facts.¹⁷
A putative fact that turns out not to be a fact is merely a false proposition.
Consider, for example, the proposition that morality requires that commands
issued by members of a particular family have a particular impact on the law.
Could this false normative proposition provide a reason of the requisite sort? I am
inclined to give a negative answer, but it is not necessary to resolve the question.
Only facts can be constitutive determinants of facts. For it is the obtaining of a fact
or facts that makes it the case that another fact obtains. In the case of a false prop-
osition, there is nothing to do the metaphysical work. Hence, the important point
is that even if a false normative proposition could satisfy the rational-relation
requirement, it could not be a determinant of a legal fact.
Before turning to the third clarificatory point, I want to note a consequence
of the second one. Positivists cannot retreat to a position that concedes that

¹⁵ See HFML, supra note 1, at 227–228. See also Greenberg, On Practices and the Law, supra
note 1. ¹⁶ Thanks to Seana Shiffrin for pushing me to address this point.
¹⁷ See HFML, supra note 1, at 252.
270 Mark Greenberg

what are taken to be normative facts are needed to satisfy the rational-relation
requirement. The imagined positivist strategy would be to maintain that merely
putative normative facts can provide rational intelligibility, which is consistent
with the positivist view that normative facts need not be among the deter-
minants of legal facts. As noted, however, only real facts can be constitutive
determinants of anything. Hence, once it is conceded that putative normative
facts are needed to satisfy the rational-relation requirement, it follows that (real,
not merely putative) normative facts must be among the determinants of the
content of the law.
Thirdly, the rational-relation requirement is not a requirement that the legal
facts be shown to be good or valuable.¹⁸ I express the requirement by saying that
the determining facts must provide reasons for the legal facts, but as noted above,
a reason in the relevant sense is a consideration that accounts for an explanandum
in rational terms, not a justification. In general, as I explain in HFML, non-
normative considerations can constitute such reasons.
The point is important because otherwise positivists could not accept the
rational-relation requirement. The strategy of the argument is to use the rational-
relation requirement, which I claim most legal theorists implicitly accept, to argue
for the controversial conclusion that normative facts must be determinants of
legal facts. One can accept the rational-relation requirement, while taking it to be
an open question whether non-normative facts could themselves provide reasons for
the obtaining of legal facts—and indeed whether normative and non-normative
facts together could do so.
In HFML, I sketch some reasons for believing that (3) is true,¹⁹ at least in the
U.S. and U.K. legal systems and perhaps in all legal systems,²⁰ though I do not
attempt anything like a full defense of that position. I also suggest that most legal
theorists implicitly take for granted that (3) is true in the legal systems with which
they are concerned.
In fact, I am not especially concerned with the question of whether (3) is true in
all possible legal systems. Throughout this paper, I will simply assume that (3) is
true in the legal systems with which we are concerned, and I will omit any quali-
fication to that effect. Readers who believe that (3) is true only in some legal systems
can understand my arguments as applicable only to those legal systems.
In HFML,²¹ I argue that:
(4) The law practices cannot themselves rationally determine any legal facts.
I will not repeat that argument here. The basic problem with law practices is that
there are many possible ways in which they could bear on the legal facts, and they
cannot determine their own relevance. For this reason, I sometimes express the
rational-relation requirement by saying that there have to be reasons that determine

¹⁸ See HFML, supra note 1, at 232–233. ¹⁹ See HFML, supra note 1, at 237–240.
²⁰ To be more precise, all those in which the legal facts are determined by law practices.
²¹ See HFML, supra note 1, Section IV.
 Hartian Positivism and Normative Facts: How Facts Make Law II 271

the contribution of law practices to the legal facts. This way of putting things
should not mislead us into thinking that the rational-relation requirement is a
requirement not only that there be reasons for the legal facts, but that there be rea-
sons that explain why those reasons are reasons. That line of thought could suggest
that an explanatory regress lurks.²² But saying that there have to be reasons for the
contribution of law practices to the legal facts is just an intuitive way of summariz-
ing why law practices by themselves do not provide reasons for legal facts. What
the rational-relation requirement demands is not higher-order reasons, but deter-
mining facts that together provide reasons for the legal facts.
From (3) and (4), it follows that:
(5) Something other than the law practices is among the determinants of the
legal facts.
The objections that I want to consider in this paper accept (3), (4), and therefore
(5). But they deny that I can reach the conclusion:
(6) The legal facts are in part determined by normative facts.
So for purposes of this paper, we can assume (1)–(5).

III. Hartian Dispositions

Hart provides the most influential account of how non-normative facts determine
legal facts.²³ In this section, I explore how that account fares with respect to the
rational-relation requirement.
The existence of a rule of recognition does the main work in Hart’s account.
Although Hart would not put the point this way, on Hart’s account, a model of
how law practices contribute to the content of the law is legally correct in virtue of
the existence of a rule of recognition specifying that model. Hart’s account of what
makes it the case that a rule of recognition exists is an application of his “practice
theory” of rules.
According to the practice theory, a “social rule” is constituted by certain atti-
tudes and dispositions.²⁴ We can call these dispositions and attitudes Hartian
dispositions, and, following Hart’s terminology, we can say that people who have
such dispositions for a particular rule accept that rule.²⁵ The notion of Hartian

²² I discuss below the idea that if normative facts are needed to explain the relevance of law prac-
tices, an infinite regress is generated. See infra Section VI, especially note 57.
²³ H. L. A. Hart, The Concept of Law (2d ed. 1994).
²⁴ We can avoid the issue of what proportion of the people in a community need to have the
appropriate attitudes and dispositions because my criticisms of the Hartian approach will not depend
on that issue. When discussing examples, I therefore assume that all or the great majority of the rele-
vant people—in particular, the officials of a legal system—have the attitudes and dispositions.
²⁵ Because it avoids grammatical awkwardness, I will often use the “Hartian dispositions”
terminology rather than talking of “acceptance” of a rule.
272 Mark Greenberg

dispositions is relative to some rule, so we can talk of Hartian dispositions for a

given rule, or acceptance of a given rule. Given a rule R, for people to have
Hartian dispositions for R is for them to regularly act in accordance with R, to
regard R as the standard by which to guide their future conduct, to be disposed
to criticize or apply other kinds of social pressure to others who fail to follow R
(or threaten to do so), and to regard such criticism as justified.²⁶ (For con-
venience, I will sometimes use the term “social rule,” but I prefer “Hartian dispo-
sitions” because using the term “social rule” may lead us to slip into thinking that
we have established the existence of something more explanatorily substantial
than the specified set of dispositions and attitudes.)
Now, the canonical form of a Hartian social rule is that one is to ⌽, where
⌽ing is taking certain action under certain circumstances. For example, Hart
mentions the social rule that one is to take off one’s hat in church.²⁷ By contrast,
a typical formulation of a rule of recognition specifies what counts as law. Hart
gives the example of a rule of recognition that specifies that “what the Queen in
Parliament enacts is law.”²⁸ If we took such formulations of rules of recognition
seriously, rules of recognition would not fit into Hart’s practice theory of rules.
For that theory has nothing to say about rules that do not specify a course of
action.
Hart simply proceeds on the assumption that the Hartian dispositions for a rule
of recognition are what they would be if the rule were specified in terms of what
standards an official is to apply in dealing with matters that come before her in her
official capacity, rather than in terms of what is law.²⁹ That assumption is necessary
for the practice theory of rules to yield Hart’s account of law. (One way to put the
point is to say that the rule of recognition described as “what the Queen in
Parliament enacts is law” is more properly stated as “what the Queen in Parliament
enacts is to be applied in deciding matters that come before an official.” As long as
we are aware of the point, however, there will be no harm in following Hart in
informally formulating rules of recognition as specifications of what standards
are law.)
For example, consider R1, according to which the plain meaning of whatever
Rex pronounces is law (and if what Rex pronounces lacks a plain meaning, it has no
effect on the law). On Hart’s account, for R1 to exist is for the officials to decide

²⁶ See Hart, supra note 23, at 55–61, 255. Let us ignore any difficulties about the indeterminacy
of the rule for which officials have Hartian dispositions. If there are any such difficulties, they are
problems for the Hartian, not for me. For present purposes, I propose to give the Hartian the
strongest possible case by simply assuming that people’s Hartian dispositions uniquely determine, in
the way that Hart suggested, a particular rule. ²⁷ Id. at 55–57, 109.
²⁸ Id. at 107.
²⁹ Id. at 100–110. In his account of what constitutes a rule of recognition in a contemporary legal
system, Hart attributes a limited role to citizens who are not officials. Id. at 60–61. As it does not
affect the substance of the argument, I will omit reference to citizens in what follows. Thanks to Scott
Shapiro for help in formulating the assumption described in the text.
 Hartian Positivism and Normative Facts: How Facts Make Law II 273

cases according to the plain meaning of whatever Rex pronounces, to criticize other
officials if they fail to decide cases in that way, and so on. In other words, for R1 to
exist is, roughly speaking, for the officials to treat R1 as the correct model.
Suppose that the officials of a legal system accept R1. And suppose that Rex has
made only one pronouncement on the subject of foie gras, a pronouncement the
plain meaning of which is that the production of foie gras is prohibited. A Hartian
will now claim that the officials’ acceptance of R1 plus the fact of Rex’s pro-
nouncement make it a legal fact that the production of foie gras is prohibited.
It will be important in what follows that, on Hart’s account, what makes some-
thing a legal rule (other than the rule of recognition) is not that people have
Hartian dispositions for that rule. In our example, the law prohibits the produc-
tion of foie gras even if people lack Hartian dispositions for the rule that one is not
to produce foie gras. On Hart’s account, what makes a standard law is not that it is
accepted, but that it is identified by an accepted rule.³⁰
Given (3) (the rational determination premise), the non-legal facts have to make
rationally intelligible the obtaining of the legal fact that the law prohibits the
production of foie gras. It may be claimed that the officials’ Hartian dispositions
and Rex’s pronouncement together satisfy this requirement. Is this claim correct?
The facts about the officials’ Hartian dispositions are part of the law practices—
they are just more of the ordinary empirical facts about the attitudes and behavior
of various people that determine the content of the law. But by (4), law practices
cannot themselves rationally determine the legal facts. As I argue in HFML,³¹ the
problem is that ordinary empirical facts cannot determine their own relevance to
the legal facts.
Without repeating the arguments for (4), I will say very briefly why Hartian
dispositions in particular seem inadequate to satisfy the rational-relation require-
ment. Officials have Hartian dispositions for a rule that requires them to apply
certain standards. The Hartian needs this fact to explain why the standards are
law. The problem is that it is not clear why we should think that Hartian disposi-
tions for a rule have this significance. As a general matter—outside the legal
arena—a case has not been made that Hartian dispositions have explanatory
potency. For example, Hartian dispositions for a rule do not in general make the
rule binding on anyone or provide any reason for acting in accordance with the
rule.³² There are practices or organizations in which there are Hartian dispositions
for rules requiring people to haze new recruits, to sell children into sexual slavery,
and to use violence to extract “protection money” from shops. It does not follow
that these rules obtain or are binding on anyone in any non-trivial sense.

³⁰ See Hart, supra note 23, at 100–110, 256. Hart also suggest that, in addition to the rule of
recognition, “certain important legal rules” are social rules. Id. at 256.
³¹ HFML, supra note 1, at 178–185.
³² For fuller discussion of essentially this point, see Greenberg, On Practices and the Law, supra
note 1.
274 Mark Greenberg

The Hartian will likely respond that what is at issue is not whether the rule of
recognition is morally or all-things-considered binding or even whether there are
any (non-legal) reasons to act in accordance with it. Rather, the issue is simply
whether the rule is legally authoritative. According to this line of argument,
Hartian dispositions for a rule of recognition make it legally authoritative not
because of some general truth about Hartian dispositions, but because of some
truth specific to the legal case.
Consider, for example, Joseph Raz’s account of “why it is true that parliament-
ary legislation is binding on the courts:”
The answer is that this is so because of the practice of the courts which follow a rule to that
effect and because the rules practised by the courts of a legal system are rules of that system
according to the doctrine of identity.³³
Raz sees that the practice of the courts—their Hartian dispositions—by itself is
not an explanation. The “doctrine of identity”—a special truth about how things
are with law—is needed to complete the account. In the next two sections, we
will turn to the possibility that the Hartian could appeal to a truth specific to the
legal case.
It might be thought that there is an alternative to an appeal to a truth specific to
the legal case. A Hartian could argue that there is a realm that is broader than and
encompasses the legal domain in which acceptance of a rule of recognition has the
impact that the Hartian claims it has in the legal domain. One suggestion would
be that the relevant realm includes anything that is socially constructed. But this
suggestion is a non-starter. What makes something a celebration, a book, a
kitchen, or a fashion statement is not the acceptance of a rule of recognition.
A somewhat more promising suggestion would be that we should consider
rules of practices (including organizations, games, and so on). Notice, first, that
the Hartian cannot claim that ordinary legal rules are an instance of Hartian social
rules. As noted above, by the Hartian’s own account, the existence of a social rule S
is constituted by Hartian dispositions for S, not by officials having Hartian dis-
positions for a rule of recognition requiring officials to apply S. The Hartian needs
not a domain of Hartian social rules, but a domain in which acceptance of a rule
of recognition by officials charged with applying primary standards makes the
standards it specifies rules of the practice (regardless of whether there are Hartian
dispositions for those standards).³⁴ Call a domain that has this feature Hartian.
Many familiar social practices, such as etiquette, are obviously not Hartian since
they have no equivalent to legal officials, let alone to the acceptance by officials of
a rule of recognition.

³³ J. Raz, The Authority of Law 68 (1979).

³⁴ Hart maintains that in order for a legal system to exist, it is also necessary that the primary rules
specified in the rule of recognition must generally be obeyed. But there need be no Hartian disposi-
tions for those rules. See Hart, supra note 23, at 116–117, 256.
 Hartian Positivism and Normative Facts: How Facts Make Law II 275

There has been no serious attempt to pursue this approach to defending the
Hartian account of law. Rather than pursuing it here, I will simply register a few
comments about the task facing the Hartian who would take this route. First,
there is no obvious reason why an attempt to show that there is a domain of prac-
tices that is Hartian would fare any better than an attempt to show directly that
law is Hartian. (Indeed, Hart himself does not seem to recognize the possibility of
practices outside the legal domain in which acceptance of a rule of recognition
makes the standards it specifies rules of the practice.) Secondly, even if it were
shown that there is a Hartian domain, it would require further work to establish
both that law is a member of that domain, and that a legal norm is an instance of a
rule of a practice.
Finally, it would not be sufficient for the positivist to show that there are some
practices that are Hartian. As a preliminary matter, note that it is consistent with
my argument that there be some such instances. In the legal domain, constitutive
determination is rational determination. That is the source of the rational-relation
requirement. But it may be that in the case of some practices, the rules are not
rationally determined.³⁵ In such cases, there will be no requirement that the prac-
tices make the rules rationally intelligible. The correct model can be determined
arbitrarily, so it can be determined by Hartian dispositions, by normative facts, or
in any other way.
The Hartian seeks a non-legal truth that will help to explain the claimed role of
Hartian dispositions. If some practices are Hartian and some are not, that may
well be because they are members of a domain in which the determination of cor-
rectness can be arbitrary. In that case, however, the practices that happen to be
Hartian are no evidence of an explanatory truth that could help the Hartian.
(This is not to say that the Hartian needs an absolutely general truth. It might be
enough if there were an explanatorily significant class of practices—one that
forms a social kind—that is Hartian.)
In sum, the approach of identifying a Hartian domain, developing an
account of it, and showing that that account can explain the legal case seems
more daunting (and less developed) than the more direct approach of appealing
to a specifically legal truth. In this paper, at any rate, I focus on the latter
approach.
I close this section with a preliminary indication of why an account that
appeals to normative facts in addition to law practices is not vulnerable to the
problems that we have identified with respect to the Hartian account. We
should begin by saying something about the kind of normative facts that are at
issue. The relevant normative facts are facts about the bearing of law practices
on our legal obligations (or on other aspects of the legal facts). An example
might be that fairness supports a statute’s contributing its plain meaning to the

³⁵ See HFML, supra note 1, at 229.

276 Mark Greenberg

content of the law. Or that democratic reasons cut against a judicial decision’s
being able to create a standard that goes beyond what is necessary for resolution
of the dispute before the court.
We noted that Hartian dispositions are just more law practices, and thus sub-
ject to the arguments of HFML regarding why law practices cannot themselves
provide reasons for the legal facts. Unlike Hartian dispositions, normative facts
are not part of the law practices.³⁶ More substantively, the normative facts in ques-
tion seem to be just the sort of fact that, with law practices, could explain in rational
terms the obtaining of particular legal facts. Suppose fairness and democracy favor
plain meaning over all other models of the bearing of statutes on the content of
the law. On the face of it, that normative fact is the kind of thing that could sup-
plement facts about statutory language to yield an explanation of legal facts. I will
have more to say about this point in Section V below.
At this stage, we have reason to think that normative facts and law practices
together are not in the same position as law practices alone with respect to the
rational-relation requirement. It is therefore not open to an objector simply to
insist that if law practices cannot satisfy the explanatory demand, neither can norm-
ative facts and law practices together. In the next two sections, after developing the
Hartian’s position, we will take a closer look at strategies a Hartian might pursue to
argue that the Hartian account can satisfy the rational-relation requirement.

IV. Hartian Bridge Principles

We mentioned in the previous section the possibility that there are fundamental
truths about law that enable the Hartian to meet the rational-relation require-
ment. For the Hartian positivist, the natural candidate for such a truth is some-
thing along the following lines:
(7) For any rule R (that specifies that standards with certain features are
law),³⁷ officials’ Hartian dispositions for R make it the case that a legal sys-
tem’s law practices contribute to the content of the law in accordance with R
and only in accordance with R (and if officials do not have Hartian disposi-
tions for any such rule, then there are no legal facts).
Proposition (7) is in effect a bridge principle, one that takes us from ordinary
empirical facts—law practices—to legal facts. (7), along with facts about Hartian
dispositions and other law practices of a legal system, is supposed to make ration-
ally intelligible the legal facts of the system.
Before we examine whether (7) (or some similar bridge principle) can play the
role it is introduced to play, we need to become clearer about (7)’s scope. Is it a
claim about all possible legal systems or merely about some legal systems?
³⁶ See HFML, supra note 1, at 235–236.
³⁷ On the proper formulation of rules of recognition, see supra text accompanying notes 27–29.
 Hartian Positivism and Normative Facts: How Facts Make Law II 277

Let us use the term Hartian legal system for a legal system in which Hartian
dispositions for a rule of recognition make that rule the correct model for the legal
system (and in which that is the only way in which the correct model can be deter-
mined). It is tempting for the Hartian to rest on the claim that a Hartian legal sys-
tem is at least possible.³⁸ I want to raise the stakes by arguing that a bridge
principle, such as (7), has to be true in all possible legal systems in order for it to be
true of any legal system. If I am correct, it is not an option to retreat to the mere
possibility of a Hartian legal system—it’s all or nothing.

A. Hartianism: All or Nothing

Suppose that there are some possible legal systems in which (7) is false. The ques-
tion at issue in this subsection is whether there can be any legal system in which
(7) is true. Suppose, for purposes of reductio, that there is such a system, which we
can call H (for Hartian).
In H, the correct model is the rule of recognition for which the officials have
Hartian dispositions. Call this model M. According to the rational-relation
requirement, the determinants of the legal facts have to provide reasons for the
legal facts. So the determinants have to provide reasons that explain why M is the
correct model in H. There has to be more to those reasons than the facts about the
Hartian dispositions of the legal officials. By our initial assumption, there are legal
systems in which Hartian dispositions for a rule do not make that rule the correct
model. In other words, there are legal systems in which, even if the officials
accepted M, it would not be the correct model. Therefore, if the correctness of M
in H is not to be arbitrary, there has to be a reason why the Hartian dispositions
are operative in H, but not in those other legal systems—i.e. why H is Hartian.
If it is not to be a brute fact about H that it is Hartian, H must have some
property X in virtue of which it is Hartian.³⁹ It follows that (7) is not true in

³⁸ Rosen says: “One need not go all the way with Hart to think that it lies in the very nature of
legal systems that one way for a legal system to be in place is for Hart’s conditions to be satisfied.”
Rosen, supra note 5, at 6. This remark need not be understood to suggest that some legal systems
could be Hartian and others not. The remark seems to recognize that it would have to be a general
truth about law—“in the very nature of legal systems”—that Hartian dispositions for a rule of recog-
nition can make a model correct. Perhaps the possibility Rosen has in mind is that Hartian disposi-
tions for a rule are sufficient to make that rule the correct model in any possible legal system, but if
officials do not have Hartian dispositions for any rule of recognition, there are other ways for a model
to be correct. This possibility is consistent with there being a (disjunctive) principle about how legal
facts are determined that is true in all possible legal systems.
³⁹ Could it simply be a brute fact about a particular legal system (as opposed to about the law or all
possible legal systems) that it is Hartian, i.e. that there are no other facts about the legal system in
virtue of which the system’s being Hartian can be explained? The rational-relation requirement rules
out the possibility of a brute fact about a particular legal system. Accepting the possibility of such a
fact is tantamount to accepting that it can be arbitrary which model is correct. Notice that if it could
be a brute fact that a particular legal system was Hartian, two legal systems could be identical in every
respect, except that one was Hartian and one was not. Two such legal systems would be empirically
indistinguishable; they would not differ even in the participants’ dispositions, beliefs, and utterances.
278 Mark Greenberg

H. Hartian dispositions for a rule R are not sufficient to make it the case that R is
the correct model in H. Part of the explanation is that H has property X. The
supposition that (7) was true in H, but not in all possible legal systems has led to a
contradiction. Hence (7) is true in all possible legal systems or in none.
The foregoing argument does not exclude the possibility that in some legal sys-
tems, acceptance of a rule makes it the correct model, while in other legal systems
it does not.⁴⁰ It is just that there will have to be some property, X, in virtue of
which a given legal system falls in the former category. The full constitutive
account of the legal facts in legal systems in the former category will not be the
Hartian account because it will have to make reference to X. (And if the account is
to be a positivist one, X will have to be a non-normative property.)
Moreover, the revised account—the one that makes reference to X—will be
true in all possible legal systems. For if it were not, then there could be a legal sys-
tem with property X, but in which Hartian dispositions for a given rule fail to
make that rule the correct model for the legal system. But in that case, X is not the
property that makes Hartian dispositions operative, after all.
The point generalizes to any account of what, independent of the law practices,
determines the correct model of the contribution of law practices to the content of
the law. A full constitutive account of the legal facts in a particular legal system A
must specify some principle P that is independent of the law practices of A and
that, with those law practices, explains the legal facts.⁴¹ Suppose there is a legal
system B in which P is not true. In B, the legal facts can be different from those of
A in a way that is not explained by differences in the law practices. For example,
identical law practices would yield different legal facts in A and B. We have a viola-
tion of the rational-relation requirement.
Different models can of course be correct in different legal systems. If a prin-
ciple Q (about the relevance of law practices to the content of the law) is true in A,
but not in B, then that principle cannot be the whole story about what, inde-
pendent of law practices, determines the correct model. In order to satisfy the
rational-relation requirement, a constitutive account of the legal facts of A would
have to explain why Q is true in A. In other words, we will need a conditional
principle whose antecedent includes the condition that the legal system has the
critical property that makes Q applicable. The rational-relation requirement
ensures that there will be such a property. But the revised principle will be true
even in legal systems that do not possess that property. In sum, principles about
the relevance of law practices to the content of the law that are not themselves
derived from law practices must be true in all possible legal systems.
Hence the participants in two systems would not be distinguished even by their beliefs about whether
their own system was Hartian. It would thus be utterly mysterious what made one system Hartian
and not another.
⁴⁰ Another possibility is the one described in supra note 38 that Hartian dispositions are always
sufficient to determine legal facts, but that in the absence of such dispositions, there are other alter-
natives. The argument in the text applies, with appropriate modifications, to that possibility.
⁴¹ See HFML, supra note 1, Section IV.
 Hartian Positivism and Normative Facts: How Facts Make Law II 279

Returning to the Hartian account, we can conclude that, in order to satisfy the
rational-relation requirement, the Hartian needs a bridge principle whose scope is
all possible legal systems. Since the Hartian holds that (7) is a complete and
correct account of legal facts—that there is no property X—we can hereafter take
(7) as our candidate bridge principle.⁴² The Hartian needs (7) to be true in all pos-
sible legal systems. To save words, I will hereinafter understand “(7)” to make the
Hartian claim about all possible legal systems.⁴³

B. Does the Bridge Principle Enable the Hartian to

Satisfy the Rational-relation Requirement?
If (7) could be taken for granted, the Hartian account would satisfy the rational-
relation requirement. But taking (7) for granted would blatantly beg the question
in favor of Hartian positivists. A main reason for introducing the rational-relation
requirement into the debate was precisely in order to make progress in assessing
(7) and related propositions. (Indeed, if the Hartian account does not satisfy the
rational-relation requirement, we can conclude that (7) cannot be true.)
According to (3), all legal facts must be rationally determined by non-legal facts.
Now the Hartian might point out that (7) is not a legal fact in our technical
sense—that is, it is not part of the content of the law.⁴⁴ Instead it is some kind of
basic fact about law. Hence, the Hartian might argue that (7) is a non-legal fact,
which therefore can be counted among the determinants of the legal facts, and
need not itself be rationally intelligible in terms of non-legal determinants. The
issue, of course, is not whether (7) escapes the precise formulation that (3) gives to
the rational-relation requirement, but whether the reasons for believing in the
requirement apply with respect to (7).
The idea that motivates (3) is that all legal facts must be susceptible of being
made intelligible in terms of facts that are not part of the legal domain.⁴⁵
According to this idea, law is fundamentally not only a human creation, but one
that is constructed in such a way that the existence of particular legal facts must
always be fundamentally intelligible to rational creatures who know all the facts
except those that are specially legal. But if there are brute truths⁴⁶ about how the
law practices contribute to the content of the law, the requirement of rational

⁴² A positivist who thinks that some property X needs to be added to the specification of the cor-
rect bridge principle can substitute the appropriate bridge principle for (7) throughout. My argu-
ments will generally not be affected by the details of the bridge principle.
⁴³ In other words, the revised version of (7) is as follows: In all possible legal systems, for any rule R
(that specifies that standards with certain features are law), officials’ Hartian dispositions for R make
it the case that a legal system’s law practices contribute to the content of the law in accordance with R
and only in accordance with R (and if officials do not have Hartian dispositions for any such rule,
then there are no legal facts). ⁴⁴ See supra text accompanying note 8.
⁴⁵ See HFML, supra note 1, 226–227, 231–232, 237–240 See also Greenberg, Reasons Without
Values?, supra note 1; Greenberg, On Practices and the Law, supra note 1.
⁴⁶ See, e.g., HFML, supra note 1, at 227–228.
280 Mark Greenberg

intelligibility will be respected in name only. To put the point another way, to
allow a bridge principle (from non-legal facts to legal facts) to count as a
“non-legal fact” would trivialize the requirement that the relation between the
non-legal facts and the legal facts be rationally intelligible. Hence, in (3), the
phrase “non-legal facts” should be understood to exclude facts about how non-
legal facts contribute to the content of the law, even if those facts are not legal facts
in our technical sense.
We therefore find ourselves back to the question of whether (7) is rationally
intelligible in light of the non-legal facts, which is really just the original question,
addressed quickly back in Section III, of whether Hartian dispositions for a rule of
recognition R make it intelligible that the legal facts are determined in accordance
with R. Now that we have clarified what the Hartian needs—in particular that
what has to be rationally intelligible is that the Hartian bridge principle be true in
all possible legal systems—I want to address that question at greater length.
The Hartian is in a difficult position in trying to give reasons why Hartian dis-
positions for a given rule make it the case that law practices contribute to the con-
tent of the law in accordance with that rule. In Section III, we raised and set aside
the possibility of reasons that are not specifically legal. The present question con-
cerns reasons specific to law.
First, the Hartian cannot appeal to features of particular legal systems to explain
the relevance of law practices to the content of the law. The Hartian needs to explain
a truth about all possible legal systems. Secondly, empirical induction is not a
promising option. As we will see below, whether Hartian dispositions for a rule
make it the correct model in a particular legal system is not a straightforward empir-
ical question. Moreover, we will see that the evidence of our own legal system, if any-
thing, cuts against the Hartian. In the next section, I address the possibility that the
Hartian could explain (7) by appealing to our reflective understanding of law.

V. Can the Hartian Explain the Bridge Principle by Appealing

to Our Reflective Understanding of Law?
In the absence of non-legal reasons for (7), the Hartian will insist that understanding
why (7) is true—i.e. grasping the explanatory relevance of Hartian dispositions—
is part of understanding the nature of law, as revealed in our intuitions or convic-
tions about what legal systems are possible. No further explanation is needed or
could be given.⁴⁷
The Hartian may add a second point—that any account of legal facts will have
to appeal to some such brute fact about law, in particular about the way in which

⁴⁷ Rosen puts the point this way: “The positivist claim is that someone who fails to see the
explanatory force of the sociological facts in this sort of case simply fails to understand what it is for a
law to be the law of a given community.” Rosen, supra note 5, at 6.
 Hartian Positivism and Normative Facts: How Facts Make Law II 281

non-legal determinants of legal facts relate to legal facts. Hence, it cannot be an

objection to the Hartian account that it must appeal to a brute fact about law.
As to this second point, I address it in Section VI below. I show that it is not
true that an account of legal facts that appeals to normative facts as well as to law
practices has to appeal to a brute fact about law. Normative facts make such a
specifically legal bridge principle unnecessary.
As to the first point, I want to examine at some length the claim that under-
standing law requires understanding the explanatory force of Hartian disposi-
tions. To defend this claim, the Hartian appeals to our intuitions about the nature
of law. In addition to relying on an alleged intuition that (7) is true in all possible
legal systems, the Hartian may argue that a Hartian legal system seems to be possi-
ble. In this way, the Hartian may try to turn to her advantage my all-or-nothing
result—that, at the most basic level, principles concerning how law practices deter-
mine legal facts have to be true in all possible legal systems. That all-or-nothing
result implies that if a Hartian legal system is possible, all possible legal systems are
Hartian.
Given this dialectical situation, I think it is fair to offer a thought experiment
that tries to elicit the intuition that there could be a legal system in which (7) is
false. I do not put much store in our ability to divine what legal systems are pos-
sible through the kind of thought experiment I offer. For one thing, the all-or-
nothing result should make us cautious about the reliability of our judgments
about what legal systems are possible. That result puts our intuitions that different
kinds of legal systems are possible in sharp competition with each other.
Hence I do not advance my thought experiment as an affirmative argument
that a non-Hartian legal system is possible (and therefore that all legal systems
must be non-Hartian). Rather, the point is to counter the Hartian’s attempt to rely
on intuitions about what legal systems are possible.
Consider rule of recognition R2:
R2: The plain meaning of whatever the tallest person in the country
pronounces is law (and if what the tallest person pronounces lacks a plain
meaning, it has no effect on the law).
Imagine a legal system in which at time T1 all legal officials have Hartian disposi-
tions with respect to rule R2. That is, every legal official is disposed to apply the
plain meaning of whatever the tallest person in the country pronounces, is dis-
posed to criticize others who fail to do so, takes such criticism to be justified, and
so on. Suppose that years go by, and at time T2 a local legal theorist, Themis, pro-
poses that the practices of the officials are, and always have been, mistaken. She
points out that the wise king Rex I happened to be both very wise and very tall.
And she argues that the practice of treating the tallest person’s pronouncements as
law is best explained as the result of a confusion about whether Rex I’s wisdom or
height was the relevant criterion. It seems at least possible that Themis is correct
that the officials’ Hartian dispositions for R2 are mistaken, that it is not legally
correct to treat whatever the tallest person pronounces as law.
282 Mark Greenberg

Whether Themis is correct does not depend on whether the other officials come
to agree with her. But, to dramatize the story, we can suppose that they do. At time
T3, they come to agree not only about the current situation, but that their Hartian
dispositions for R2 were mistaken at T1, and have always been mistaken. Isn’t it at
least possible that they are correct? That is, couldn’t there be a legal system in which
the officials’ Hartian dispositions for a rule could turn out to be incorrect?
It might be objected that I have chosen an implausible or silly rule of recogni-
tion. But this objection misses the mark. In order for (7) to play the explanatory
role it is supposed to play, the merits of the rule of recognition cannot matter. If it
is a brute truth about legal systems that Hartian dispositions for a rule of recogni-
tion make it the correct model, then the merits of the rule are irrelevant. If, on the
contrary, Hartian dispositions for a given rule are effective in making that rule the
correct model only when the rule is sufficiently wise or sensible, then a constitu-
tive account of the legal facts that appealed only to (7) and the law practices would
be incomplete. A full account would have to make reference to facts about the
wisdom or sensibleness of the rule of recognition. Hence, if (7) is to play the
explanatory role it is supposed to play—making normative facts unnecessary—
the merits of the rule of recognition must be irrelevant. Therefore, in considering
whether there could be a legal system in which (7) is not true, it is fair game to
consider situations involving silly or bad rules of recognition. Indeed, we need to
consider such situations if we are to separate the work that is being done by accep-
tance of a rule from the work that is being done by the merits of the rule.
Let it not be objected that if the officials come to agree with Themis, they must
already have had a disposition inconsistent with Hartian dispositions for R2. This
objection is based on a mistake. From the fact that a person can be convinced that
doing X is the wrong thing to do, it does not follow that she was not previously
disposed to do X. A person’s dispositions can change. And, by hypothesis, that is
what has happened here. It is part of the description of the original situation that
the officials have Hartian dispositions for R2. The fact that they later decide that
their dispositions to follow R2 were mistaken, and consequently come to have
new dispositions, is consistent with that description.⁴⁸
We should not get too caught up in the details of an example. The issue is
simply whether there could be a legal system in which officials’ acceptance of a
rule is not the final word on how the content of the law derives from statutes,
judicial decisions, and other law practices. In such a legal system, judges and
other officials could unambiguously accept a particular rule of recognition, but

⁴⁸ We can also alter the facts to eliminate any worry of the sort addressed in the last paragraph.
We can suppose that none of the legal officials who are convinced by Themis at T3 were legal offi-
cials at T1—all of those who were legal officials at T1 have died or retired. Moreover, those who were
legal officials at T1 would not have been convinced by Themis that their acceptance of R2 was mis-
taken. In other words, the legal officials at T1 accepted R2, and were not disposed to reconsider that
acceptance.
 Hartian Positivism and Normative Facts: How Facts Make Law II 283

could nonetheless be mistaken. Remember that I am not using the example to

argue that such a legal system is possible, but merely to make the limited point
that our intuitions do not rule it out.
Setting thought experiments aside, it is worth noting that support for (7) is not
to be found in the evidence of our own legal system. One bit of evidence is that it
seems to be coherent for a lawyer to challenge any attitudes and dispositions of
officials about the correct model, even ones that are common to all officials and
have never been questioned.
Another piece of evidence concerns the way in which our discussions of the
relevance of law practices to legal facts actually proceed. When a lawyer, judge,
or theorist raises questions about what the correct model is, the ensuing discus-
sions typically make reference to value facts. Advocates of particular positions
appeal to normative facts—they give reasons why law practices should have one
impact rather than another. For example, they claim that their positions are
more consonant with democratic values, better protect the rights of minorities,
will lead to better states of affairs, are fairer, and so on. Dworkin has made this
point very powerfully, and I will not rehearse the evidence here.⁴⁹ The relevance
of this evidence is as follows. If our legal system were Hartian, then barring a
widespread misunderstanding, one would expect that debates over the correct
model would be resolved exclusively by appeal to facts about the actual attitudes
and dispositions of officials. How law practices should affect the legal facts
would be irrelevant.
A Hartian cannot respond to this evidence by maintaining that officials in our
legal system do not accept a rule of recognition. As (2) states, there are many legal
facts in our legal system. According to (7), however, without an accepted rule of
recognition, there would be no legal facts.
A Hartian might maintain that the explanation of the relevance of normative
facts is that officials in our system accept something like the following rule of
recognition: statutes, judicial decisions, and other law practices contribute to the
content of the law according to the model that is most supported by the relevant
values.⁵⁰ In other words, the Hartian could argue that our legal system is not one

⁴⁹ See R. Dworkin, Taking Rights Seriously (1977); R. Dworkin, Law’s Empire (1986).
⁵⁰ Alternatively, the Hartian might argue that the Hartian dispositions of the officials in our sys-
tem are for the following rule of recognition: statutes, judicial decisions, and other law practices con-
tribute to the content of the law in the way that the officials believe is most supported by the relevant
values. But notice that the hypothesis in the text is simpler and better accords with what the officials
themselves think and do. The officials do not think that the rule is to take law practices to contribute
to the law in the way that the officials believe is most supported by the relevant values. They think that
the rule is to take the law practices to contribute to the law in the way that is most supported by the
relevant values. For example, officials who know that they are in the minority with respect to some
dispute about the correct model do not automatically concede that they are mistaken, as they would
if they believed that the rule of recognition was the one described in this footnote. For discussion of a
related point, see M. Greenberg & H. Litman, The Meaning of Original Meaning, 86 Geo. L.J. 569,
608–609 (1998).
284 Mark Greenberg

in which values are ultimately doing the work—rather, at the most basic level,
dispositions to treat normative facts as relevant is what makes them relevant.⁵¹
It might be thought that this move by the Hartian leaves us with a stalemate.
Against this thought, however, I want to point out an asymmetry between the
positions. The positivist position is in tension with what the officials themselves
believe. In our legal system, when officials appeal to normative facts, they believe
that it is those facts, not other officials’ dispositions to be guided by those facts,
that ultimately matter.⁵² An official who appeals to fairness or democracy does not
think that fairness or democracy matters because all or most officials think it
does.⁵³ In contrast, the anti-positivist position is in harmony with what the offi-
cials believe. This asymmetry supports the view that our legal system is non-
Hartian because, other things being equal, a view that does not have the
consequence that the officials are systematically wrong about such a fundamental
matter is more plausible than a view that does. Again, the point of this brief dis-
cussion is not to argue that our legal system is not Hartian, but merely to refute
the claim that (7) is part of our reflective understanding of law, including of our
own legal system.
In sum, we have little reason to accept the Hartian’s claim that our understand-
ing of law includes an understanding that Hartian dispositions yield legal facts in
accordance with (7). If that is right, the Hartian needs to provide reasons why (7)
is true for all legal systems, and to do so without appeal to normative facts. We
remarked on the dimensions of this task in Section III. We can leave it as a chal-
lenge to the Hartian.

VI. Is an Appeal to Normative Facts Subject

to Parallel Objections?
I have argued that the Hartian account of the contribution of law practices to legal
facts does not fare well with respect to the rational-relation requirement. It
remains to address the claim that an account that appeals to normative facts in
addition to law practices encounters parallel difficulties.
It might be objected, for example, that in order for an account that appeals to
normative facts to be complete, a bridge principle, parallel to the Hartian one, will

⁵¹ As readers will have noted, we find ourselves in the vicinity of a familiar debate. See, e.g.
J. Coleman, Negative and Positive Positivism, in Ronald Dworkin and Contemporary
Jurisprudence 28 (M. Cohen ed., 1984); R. Dworkin, Reply to Coleman, in Ronald Dworkin and
Contemporary Jurisprudence 252 (M. Cohen ed., 1984). The rational-relation requirement
provides a new perspective on the debate. From this perspective, the issue is whether the practices of
officials or the normative facts to which they appeal are the reasons that, at the most basic level,
explain the legally correct model.
⁵² I think this claim is very plausible, but it is an empirical one, and my basis for it is admittedly
unsystematic. I will simply assume that it is true in what follows. ⁵³ See supra note 50.
 Hartian Positivism and Normative Facts: How Facts Make Law II 285

be required. The idea would be something like the following:

(8) That a legal system’s law practices should contribute to the law in accord-
ance with a model M makes it the case that they do contribute to the law in
accordance with M.
(The bridge principle could be refined by specifying how the term “should” is to
be understood. For example, only certain values might be relevant.) Such a bridge
principle, the objection continues, is no better off than the Hartian bridge prin-
ciple (7). The normative theorist must either explain a bridge principle like (8) or
rely on a brute fact about law.⁵⁴
We have already suggested the core of a response to this objection, however. In
Section III above, in criticizing the Hartian account, we observed that facts about
Hartian dispositions were simply more law practices. We noticed that these facts
would not provide reasons of the needed sort without a truth about the special
relevance of Hartian dispositions in the legal domain. The Hartian bridge prin-
ciple is just a way of formulating the Hartian’s claim about the relevance of
Hartian dispositions.
We also suggested in Section III that an account of legal facts that appealed to
normative facts does not encounter parallel problems. On the face of it, facts
about how law practices should contribute to the content of the law make ration-
ally intelligible how the law practices do contribute to the content of the law. We’ll
examine this claim further in this section.
To put the point another way, the normative facts in question are, or serve the
function of, a bridge principle. For they are precisely facts about the relevance of
law practices to legal obligations, rights, powers, and so on. Here are a few examples,
similar to those mentioned above.
(9) Fairness requires giving some precedential weight even to incorrectly
decided previous court decisions.
(10) Democratic values cut against legislative history’s having any impact on
the content of the law.
(11) All things considered, the relevant values support model M over all
other models of the bearing of law practices on the content of the law.
It might be objected that to think that these normative facts could serve as bridge
principles, explaining the impact of law practices on the legal facts, is to confuse
how things should be with how they are. There is an explanatory gap, the objec-
tion claims, between the fact that the law practices should have a certain impact on
the law and their having that impact on the law.
According to one way of understanding this objection, it makes a logical
point. The normative facts, with the law practices, do not logically entail the
legal facts. Rational determination does not require logical entailment, however.

⁵⁴ Rosen makes this objection. Rosen, supra note 5, at 6.

286 Mark Greenberg

The determining facts must both modally entail the legal facts and make the
obtaining of the legal facts intelligible in rational terms. Logical entailment may
be sufficient for rational intelligibility, but it is not necessary. (The inadequacy
of facts about Hartian dispositions was not that they do not logically entail legal
facts.)
The objection that there is a gap between the normative facts and the correct
model should instead be understood as challenging whether normative facts are
the right sort of material to combine with law practices to provide reasons for legal
facts. For example, it might be claimed that the fact that fairness militates in favor
of judicial decisions’ contributing to the law in a particular way does not provide a
reason that they do contribute to the law in that way.
Because what is at stake in the overall argument is precisely whether the content
of the law depends on normative facts, it would be question begging for the
objector to assume that there is an explanatory gap between normative facts and
the correct model on the ground that the content of the law is independent of
normative facts. (Equally, it would be question begging for me to assume that the
content of the law depends on normative facts.) The relevant question is not, at
this stage, the ultimate one of whether the content of the law is or is not independ-
ent of normative facts. (If we had the answer to that question, we would not be
engaged in the present discussion.) Rather, it is whether normative facts about the
bearing of law practices on legal facts are even the right sort of material to provide,
along with law practices, reasons for legal facts. To put it crudely, is the fact that it
would be fair or democratic or just for law practices to affect the content of the law
in a certain way the right kind of fact to make it rationally intelligible that law
practices do affect the content of the law in that way?
If we were trying to explain the occurrence of physical events, we might ques-
tion whether normative facts were the right sort of material. For such explananda,
causal explanations are needed, and normative facts will not feature ineliminably
in such explanations. Although it was long thought otherwise, in the physical
domain, how it would be good for objects to behave is no explanation of how they
do behave.
This platitude about causal explanation tells us little about rational intelligi-
bility. In fact, the platitude is consistent with the claim that facts about how it
would be good for objects to behave make it rationally intelligible that they
behave in that way. (In fact, it is tempting to speculate that part of the explana-
tion of why people long took for granted teleological explanations of occurrences
in the physical world is that people expected that the physical world would be
susceptible to being made intelligible in rational terms.) For rational intelligibil-
ity is neither necessary nor sufficient for causal explanation. On the one hand,
the best we can do to explain the occurrence of physical events may be to cite laws
or correlations that cannot be made intelligible in rational terms. And, on the
other, a putative causal explanation that makes a phenomenon intelligible may
be false.
 Hartian Positivism and Normative Facts: How Facts Make Law II 287

In contrast, the kind of explanation at issue here is not explanation of the

occurrence of events, but constitutive explanation—explanation of what makes it
the case that a fact of some target domain obtains. It is not a confusion to think
that what significance base facts should have for target facts could make intelli-
gible the significance that they actually have.
Donald Davidson’s radical interpretation theory of mind provides a useful
analogy.⁵⁵ Davidson holds that it counts in favor of an overall attribution of
beliefs and desires to a person that it makes the person believe and desire what he
or she should believe and desire. The basis for this position is not an empirical
hypothesis that humans are likely to believe and desire what they should, but a
constitutive thesis. Roughly, the thesis is that the constitutive determinants of
one’s propositional attitudes must make it intelligible that the person has the
beliefs and desires they have.
Davidson’s account is, of course, controversial, but what is typically thought to
be problematic is (among other things) the constitutive role Davidson gives to
rational intelligibility, not the claim that normative facts are the right sort of thing
to provide such intelligibility.⁵⁶
An objector might concede that normative facts can, with law practices, pro-
vide reasons for the legal facts, but insist that we still need an explanation of the
normative facts themselves. If the Hartian bridge principle needs an explanation,
it might be suggested, so do normative facts.⁵⁷

⁵⁵ For citations to Davidson and very brief discussion, see HFML, supra note 1, at 232 note 18,
238 note 25.
⁵⁶ I do not mean to endorse Davidson’s account. See HFML, supra note 1, at 238 note 25. Also, it
might be objected that Davidson’s view in the mental case is not parallel to my view of the determina-
tion relation in the legal case. According to this objection, what Davidson holds must be intelligible is
the content of the subject’s mental states, not the relation between determining facts and content facts.
On a better understanding of Davidson’s view—and on the understanding that is useful for our pur-
poses—the best interpretation of a person makes the person intelligible in light of his or her circum-
stances and behavior. For example, the attribution of a false belief on a particular issue makes the
person more intelligible rather than less if the person’s only evidence on the issue is misleading.
⁵⁷ Some commentators who have made objections in the general neighborhood of the one
described in the text have compared their objections to the point of Lewis Carroll’s famous dialogue
between Achilles and the Tortoise. See Lewis Carroll, What the Tortoise Said to Achilles, 4 Mind 278
(1895), reprinted in 104 Mind 691 (1995). On a straightforward interpretation of this comparison,
I would stand accused of mistakenly treating normative facts as premises rather than inference rules.
In other words, the objector’s claim would be that normative facts are analogous to inference rules
that allow one to move from law practices to legal facts. But this claim, whatever its merits, would be
no objection to my position that normative facts must figure in a full constitutive account of legal
facts.
The appeal to Lewis Carroll’s dialogue is perhaps more of a loose analogy than a direct application.
The idea seems to be that if I make an explanatory demand of the sort that makes it necessary to
appeal to normative facts, I open up an infinite regress. But this idea is mistaken.
In the text, I argue that normative facts and law practices together, unlike law practices alone, pro-
vide the requisite reasons for the law facts. The type of objection that I want to consider here grants
this claim, but maintains that the appeal to normative facts generates a further, higher-order explana-
tory demand. (And the satisfier of that further demand will generate a still higher-order explanatory
demand, and so on.)
288 Mark Greenberg

The answer to this objection is that the rational-relation requirement applies

only to legal facts. Specific normative facts about the relevance of law practices to
legal facts can be explained as the consequence of applying general normative
truths to the circumstances of legal systems.⁵⁸ For example, we might explain why
fairness requires giving some precedential weight even to incorrectly decided pre-
vious court decisions by appealing to more general truths about fairness as well as
to non-normative facts about the impact of court decisions on people’s lives.
Whether or not we can explain general normative truths—ones not specifically
concerned with legal systems—is not relevant for present purposes. Again, the
rational-relation requirement is that legal facts be rationally intelligible in light of
non-legal facts. There is no such requirement with respect to non-legal facts.
Hence, an objector who insists that explanations have to stop somewhere may
well be correct. But the important point for present purposes is that explanations
do not have to stop with legal facts, or facts about how legal facts are determined.
And given the rational-relation requirement, they cannot do so.
Finally, an objector might appeal to a thought experiment, parallel to my
thought experiment about the legal system with the height-based rule of recogni-
tion, to argue that we have little reason to believe that normative facts have a bear-
ing on the correct model in all possible legal systems. Just as I suggested that we can
conceive of a non-Hartian legal system, the objector suggests that we can conceive
of a legal system in which normative facts are not relevant to the correct model.
This objection gets the dialectical situation wrong. I appealed to the seeming
possibility of a non-Hartian legal system only to answer the Hartian’s appeal to
intuition in support of a brute truth about law. The normative account of legal
content does not rely on a brute truth about law, and thus does not need to appeal

There seem to be two ways to develop the infinite-regress objection. According to the first, if a con-
stitutive account of legal facts appeals to normative facts, it will then have to explain the obtaining of
the normative facts. This is the objection that I address in the text.
According to the second version of the objection, if a constitutive account of legal facts appeals to
normative facts, it will then have to explain the relevance of those normative facts to the legal facts. As
noted above, I sometimes express the intuitive inadequacy of law practices as reasons for legal facts by
writing that we need facts that explain the relevance of law practices to legal facts. See supra note 14
and text accompanying note 22. Similarly, I sometimes write that we need reasons for the mapping
from law practices to the content of the law. In using these formulations, I may have misled readers
into thinking that the rational-relation requirement is a requirement not only of reasons for the legal
facts, but also of reasons for those reasons. The requirement is only that the constitutive determinants
of the legal facts together provide reasons for the legal facts.
A constitutive account appeals to normative facts not to satisfy a second-order explanatory require-
ment, but simply to meet the first-order explanatory requirement that law practices do not meet
themselves. It is tempting to express what is missing from an account that appeals only to law prac-
tices by saying that we need facts that explain the relevance of the law practices to the content of
the law. But, once again, this is simply a way of expressing the requirement that something must
supplement the law practices if the constitutive determinants are to provide reasons for the legal facts.
If the law practices and the normative facts together provide the requisite reasons for the legal facts,
the rational-relation requirement is satisfied.
⁵⁸ For a sketch of this picture of normative facts, see Greenberg, On Practices and the Law, supra
note 1, section IV.
 Hartian Positivism and Normative Facts: How Facts Make Law II 289

to intuitions about what law is. It has been independently defended on the ground
that normative facts are the best candidate for what is needed in addition to law
practices to satisfy the rational-relation requirement. As pointed out above, we
should be skeptical of the reliability of thought experiments about what sorts of
legal systems are possible, especially in light of my all-or-nothing result. I have
used the rational-relation requirement to argue against a Hartian account. In the
face of this argument, thought experiments get no traction.

VII. Conclusion

In this paper, I have argued that normative facts and law practices together are bet-
ter placed to satisfy the rational-relation requirement than law practices alone,
including facts about officials’ Hartian dispositions. Normative facts about the
relevance of law practices to legal facts provide reasons why law practices have a
particular impact on the law, and facts about Hartian dispositions do not. That
law practices should have a particular impact on the content of the law makes
rationally intelligible that law practices do have that impact. For instance, that it is
fair for judicial decisions to have a certain precedential force is a reason why those
decisions in fact have that force. The mere fact that officials are disposed to give
decisions a certain precedential force does not by itself constitute such a reason.
Adding a purported truth about law, or bridge principle, to the effect that
certain attitudes and dispositions of officials have a certain bearing on which
model is legally correct does not help. If the officials’ attitudes and dispositions
do not provide reasons for the legal facts, the bridge principle is itself just a brute
fact about law.
By contrast, normative facts avoid the need to appeal to brute facts about law.
With law practices, they provide reasons for the legal facts. That it is fair or demo-
cratic for statutes or judicial decisions to have a particular impact on the law, com-
bined with facts about the particular statutes and judicial decisions of a legal
system, can explain the legal facts. We can sum up in an intuitive way by saying
that normative facts explain the relevance of the law practices to the legal facts. But
there is no further requirement to explain why the normative facts explain the rele-
vance of the law practices to the content of the law. Strictly speaking, the rational-
relation requirement demands only that the constitutive determinants provide
reasons for the legal facts.
Along the way, I argued that the rational-relation requirement yields the relatively
immediate result that, at the most basic level, legal systems cannot vary with respect
to what determines the relevance of law practices to the content of the law. The most
basic principles about the relevance of law practices to legal facts—the ones that do
not depend on law practices—must be true in all possible legal systems. One can
accept this result even if one rejects other parts of my argument for the conclusion
that normative facts must figure in a constitutive account of legal facts.
290 Mark Greenberg

The result has a variety of implications. It implies, for example, that our
convictions about what kinds of legal systems are possible cannot all be correct.
It therefore should make us dubious about the reliability of such convictions.
I mentioned that the result undermines the familiar claim that because Dworkin’s
arguments depend on properties of the U.S. and U.K. legal systems, those argu-
ments cannot show that Hartian legal systems are not possible. Of course, the
result is a double-edged sword. For example, an argument for the proposition that
one legal system is Hartian supports the conclusion that all possible legal systems
are Hartian.
 Response
Ronald Dworkin

There is not space for a detailed comment on all of the essays in this collection so I
concentrate on replying to arguments that challenge the views l have defended
and say much less about essays that elaborate and deepen those views. This
response is therefore very unbalanced, devoting much space to some authors and
little to others, but I hope it is not necessary to say that the imbalance does not
reflect any opinion about the relative importance of the essays. The essays with
which I mainly agree and therefore comment on little stand on their own, and are
each an important contribution to legal theory. I am grateful for all the essays, and
grateful particularly to Scott Hershovitz for his exemplary choice of writers
and his imaginative editing. Justice Breyer’s introduction to this collection is also
an important contribution to its themes: I am very grateful for his remarks and
welcome them as a brief account of his own constitutional philosophy.¹

Eisgruber

Professor Eisgruber has made an important contribution to constitutional theory

through his clear, illuminating and persuasive writings about the structure of a
general theory of constitutional adjudication. I agree with almost everything he
says here. I agree with his two-part account of the moral reading thesis, and also
with his suggestion that the moral reading, so understood, should not seem either
innovative or controversial.² I also agree that the moral reading, as we both under-
stand it, does not itself dictate how we should interpret particular constitutional
phrases or provisions.³ I further agree that, as he emphasizes, we must defend any
particular interpretation through a more general understanding of the point of
removing certain issues from the play of majoritarian politics. Constitutional
interpretation is not, as he stresses it is not, a matter of retrieving the psychological
state of various historical officials and citizens as they wrote, debated and ratified
forms of words. It is rather a matter of making sense of what they did by assigning
general purposes that both fit their situation and justify their adopting the
language they did.
¹ For a fuller statement of the latter, see his book, Active Liberty: Interpreting our
Democratic Constitution (2005). ² See R. Dworkin, Freedom’s Law 3 (1996).
³ I emphasized this in Freedom’s Law, at 7ff.

Response. Ronald Dworkin.

© Oxford University Press 2006. Published 2006 by Oxford University Press.
292 Ronald Dworkin

Eisgruber believes, however, that my own concern with the virtue of integrity in
legislation as well as adjudication counts against my very broad reading of certain
provisions of our own Constitution: in particular the equal protection and due
process clauses. He suggests that I would have given more effect to the value of
integrity by reading these clauses to enjoin fidelity not to equality and fairness in
the abstract but rather to “traditional American principles of equality and fair-
ness.” These are different instructions, he believes, because the latter requires
respect for the actual practices of American history and so would give history a
much more important role than I seem to allow it. I disagree: I think that the latter
instruction would not encourage but subvert integrity. Indeed it would provide
only what I have called bare consistency which I have argued is very different from
integrity.⁴
Integrity requires coherence in moral principle and coherence is especially
important in constitution-making when the moral principles in play are funda-
mental. Integrity commands that the nation speak with one voice on such import-
ant matters of principle, and it therefore demands that when its historical
practices—of racial segregation, for instance—cannot be seen to be consistent
with principles elsewhere recognized, those practices must be abandoned. Respect
for settled practices has often been urged as a ground for constitutional decisions:
that was the basis of the Supreme Court’s decision upholding Georgia’s ban on
homosexual sodomy in Bowers v. Hardiwick.⁵ But the Court has since declared for
integrity over settled practice: it has overruled its Bowers decision.⁶ Constitution-
makers can provide integrity in their constitutional protections only by mandat-
ing a moral test that allows the community to identify the principles to which it
takes itself to be committed even when it recognizes that it has not kept faith with
those principles in the past.
In any case, Eisgruber says, my account of constitutional interpretation
remains incomplete because I have not given a basis for the moral reading in a
more general theory that explains why a nation should embed certain constraints
in a constitutional document that can only be amended by a super-majority with
super dedication. I meant to offer such a theory in Chapter 1 of Freedom’s Law,
however. I said that that device creates and protects the right kind of democracy. I
distinguished between a majoritarian and what I there called a “constitutional”
conception of democracy;⁷ I argued for the latter and offered that better concep-
tion as part of a justification for the moral reading. So I agree with at least a sub-
stantial part of the justification for constitutional protection that Eisgruber
himself has advanced, which he summarizes in his present essay in these words:
“More specifically, I argue that democracy differs from pure majoritarianism, so

⁴ See R. Dworkin, Law’s Empire 219–224 (1988). ⁵ 487 U.S. 186 (1986).
⁶ Lawrence v. Texas, 539 U.S. 558 (2003).
⁷ I have called the latter a “partnership” conception of democracy in both earlier and later writing.
See R. Dworkin, Sovereign Virtue: The Theory and Practice of Equality ch. 10 (2000).
 Response 293

that democracy requires solicitude for the rights and interests of minorities as well
as majorities.”⁸ I hope I am right in supposing that we agree not only on the need
for a constitutional justification of the kind he describes but also on at least the
general form that such a justification should take.

Fleming

Professor Fleming’s study of constitutional interpretation is at once subtle, theor-

etical and practical. He suggests what he believes to be a better architecture for
constitutional theory than he finds in my account of a constitutional or partner-
ship democracy. He believes that my argument makes too much depend on the
claim that embedded constitutional rights protect democracy properly conceived:
he prefers an account that supposes two distinct goals: protecting both delibera-
tive democracy and deliberative autonomy. His formulation might well have the
polemical advantages he claims for it.
I think it important, however, to guard against what I believe to be a mistake in
political philosophy that Fleming’s formulation might encourage, which is to sup-
pose that there are two values at stake in constitutional adjudication—democracy
and autonomy—that might conflict, so that the decision whether to protect sub-
stantive “privacy” rights, for instance, could turn on which of these two values is
taken to be more important. If I am right, that supposed conflict is illusory. The
concept of democracy is an interpretive concept: we need to construct a concep-
tion of democracy that sets out a recognizable political value. I believe that no
conception of democracy satisfies that test unless it respects what Fleming might
regard as a distinct autonomy principle: that no one may submit to the judgment
of others in matters that he must, out of self-respect, decide for himself. Of course
people disagree about which matters those are. But any conception of democracy
that is restricted to process alone and ignores this autonomy dimension fails even
to state a pro tanto or prima facie political value. Process without substance only
pretends to be a political value. I agree with Fleming that some of my critics do not
make the mistake of confusing democracy with mob rule. But they may make the
equally important mistake of confusing democracy with fair elections.
Fleming is right that a great many constitutional scholars, including several who
would regard themselves as liberals, have taken up the idea that constitutional inter-
pretation is a matter of retrieving the intentions of statesmen who acted early in our
constitutional history. He calls these scholars “broad originalists.” The trend he
describes may be largely a matter of oscillating intellectual fashion. But it is
nevertheless important to consider (as not all these scholars do) why and in what way
their historical studies are relevant to contemporary adjudication. The explanation

⁸ Eisgruber, 20.
294 Ronald Dworkin

must confront a crucial distinction between two kinds of intentions that history may
uncover: the semantic intentions of the “framers” who wrote and enacted constitu-
tional provisions—what they intended to say in using the words they used—and their
expectation intentions—what they hoped or expected would be the impact those
words would have on constitutional law in concrete cases.
Chief Justice Roberts insisted on the importance of that distinction in his
Senate confirmation hearings,⁹ and Professor Eisgruber emphasizes it in his con-
tribution to this volume. I have myself insisted on the decisive importance of
semantic originalism: it seems crucial to discover what the framers meant to say
because that determines what propositions the Constitution expresses. History is
of course at the center of that investigation as Eisgruber says. But expectation orig-
inalism is a different matter. If the best interpretive reconstruction of the framers’
semantic intention concludes that they laid down abstract moral principles rather
than their own understandings of what those principles require in concrete cases,
then expectation originalism, whether narrow or broad, is beside the point. I agree
that my arguments for that conclusion about semantic intention are rough and
might be improved by more careful attention to historical sources, though I also
bear in mind Justice Jackson’s skepticism on that score. Finally, Fleming’s advice to
“Do as Dworkin says not as Dworkin does” sounds to me very good advice.

Brown

Professor Brown is much too generous in assigning me credit for the enormous
improvement in constitutional theory and law that followed the Supreme Court’s
decision in the Brown case. But she has nevertheless written an elegant and dra-
matic history of that development and in that way contributed to the intellectual
as well as legal history of the period. I hope her own faith can be sustained even if,
as many fear, the Supreme Court soon leads lawyers through a period of regres-
sion. There must be another renaissance of constitutional law around another
corner and her spirit will be necessary to its success.

Hurley

Susan Hurley’s characteristically imaginative and brilliant essay makes a very

important point: that hypothetical cases play an important role in interpretive legal
reasoning. The fact that “it goes without saying” that a particular hypothetical

⁹ See my article, Judge Roberts on Trial, 52:16 New York Review of Books 14 (October 20,
2005).
 Response 295

case would be settled one way or another sharply limits the eligibility of
interpretive claims that fit actual settled cases but do not fit what “goes without say-
ing” in that hypothetical case. She uses this insight to deepen the problem posed for
integrity by Kenneth Kress.
She also responds effectively to that problem. Kress pointed out that independ-
ent judicial decisions that are made between the time of events that give rise to a
legal dispute and the judicial resolution of that dispute may affect what resolution
is then appropriate. He fears that this phenomenon might produce retrospective
application of new law: the intervening decision may change the law so that the
dispute is adjudicated under a different legal structure from the structure in which
the parties acted. Hurley points out that this can happen, as a result purely of an
intervening decision, only when that intervening decision is mistaken, and she
adds that it is hardly surprising that a mistaken decision results in retrospective
application of new law. That is what happens in the mistaken decision itself and
Kress has only pointed out that the mischief is not confined to that decision but
infects even a few later decisions that are, in themselves, correct. It is worth pointing
out, however, that this later mischief will be avoided if the intervening decision is
very mistaken: sufficiently mistaken that it can be labeled a mistake in the sense
I described in Taking Rights Seriously.¹⁰ Mistakes of that character have no gravita-
tional force in future decisions.

Hershovitz

I am grateful to Professor Hershovitz for a sterling account of the connections

between the practice of precedent and the political virtue of integrity. I agree with
him that though considerations of fairness, efficiency and certainty may con-
tribute to the justification of stare decisis, we cannot understand the character and
pervasiveness of that practice without seeing, indeed emphasizing, its connections
with integrity. He ends his argument by suggesting that though fairness and effi-
ciency do not figure in the argument for integrity, integrity may nevertheless pro-
mote those other political virtues in the long run. That is a useful suggestion to
explore. In my response to Dale Smith’s essay in this volume I say that I regret a
suggestion I made in one passage in Law’s Empire that fairness conflicts with
integrity: I should have said only that we cannot appeal to the ideals of political
equality to justify our demand for integrity rather than checkerboard compromise
in matters of principle. Hershovitz raises the question, however, whether, even so,
integrity contributes to political equality on the right conception in the long run

¹⁰ R. Dworkin, Taking Rights Seriously ch. 4 (1976).

296 Ronald Dworkin

because political power is more equal when a majority accepts that what it
provides for itself it must provide for all.

Smith

Professor Smith has constructed a very careful and challenging review of my ideas
about integrity and his review prompts me to correct an important misstatement
in my book, Law’s Empire. I agree substantially with his conclusions about how to
understand my complaint about checkerboard statutes, though I would formulate
the principles in play somewhat differently than he does. He identifies two dis-
tinct principles of integrity: one condemns checkerboard statutes and the other
requires coherence among the different departments of a community’s law. He
thinks it “obvious” that these are different principles, so that my defense of one has
no consequences for the other.¹¹ I think, on the contrary, that only a single princi-
ple is needed. I discuss checkerboard statutes (which are very rare) only to illus-
trate what I thought to be an obvious way of infringing a general principle of
coherence. If law must be coherent over great stretches of doctrine, it must surely
be coherent within a single statute; if law need not be coherent across doctrine,
then why need it be coherent even within a single statute?
Smith also thinks that the best statement of my view about checkerboard
statutes requires a distinction between two distinct charges: first, that such statutes
unjustly discriminate and, second, that the discrimination cannot be justified by
appealing to any “recognizable” principle of justice. Once again I would prefer a
single ideal: integrity requires that the community’s law be justifiable through a
coherent scheme of principle that provides an eligible interpretation of that law.
The idea of justification does indeed presuppose that only some explanations can
justify. But an adequate justification of a community’s political record also presup-
poses not just a “recognizable” principle but one that would be recognized as a
moral principle there, so that it could make sense to attribute that principle to its
law-makers. There might be communities organized around astrological princi-
ples whose citizens believe that the oddness or evenness of the year of one’s birth
carries the moral significance that for many of us is carried by rape or danger to a
mother’s health. But the fact that that would be “recognizable” as moral in some
communities hardly makes it eligible as a justification for us.
Smith believes that I have failed to show why integrity, understood as he
describes, should characteristically “trump” other political virtues in legislation
and adjudication. He notes, first, a frequent objection to my claim that it should:
that a community should not adopt an ideal that entails more discrete instances of
injustice. I have frequently replied to that objection. There is no reason to think
that requiring judges to seek integrity rather than justice in individual cases will in

¹¹ Smith, 149.
 Response 297

fact produce more discrete unjust decisions in the long run.¹² Smith mainly
focuses on another political virtue, however: fairness. He asks why a checkerboard
statute, like a statute forbidding abortion only for women born in odd-numbered
years, could not be defended as improving political fairness even though it dis-
criminates among women in a way that cannot be justified by any moral principle.
Why should fairness be subordinated to integrity in the way I seem to suppose? I
argue that a political community is not legitimate unless it shows equal concern
for all its members, and that integrity in the community’s laws is essential to that
equal concern.¹³ But, Smith asks, why is fairness, which is the equal distribution
of political power, not also a requirement of equal concern? If it is, why is integrity
a more important requirement of equal concern than fairness?
These are important questions, and they call for the correction I mentioned.
But I will first say how I have answered them in books and articles other than Law’s
Empire. We must try to understand the various central political virtues, like jus-
tice, equality, liberty, political fairness, democracy and integrity, holistically so
that our account of each can be seen to be drawn from and reinforced by the
others.¹⁴ We can describe the concept of procedural fairness in politics in a very
abstract way as I did in Law’s Empire: fairness, I said, “is a matter of finding politi-
cal procedures—methods of electing officials and making their decisions respon-
sive to the electorate—that distribute political power in the right way.”¹⁵ But we
must then establish what “the right way” is. I argue that what political fairness
requires is not either equality of impact or equality of influence but rather a struc-
ture in which no one is denied vote or voice for impermissible reasons.¹⁶ So, once
that strict requirement is satisfied, political power should be distributed with an
eye to other values. In matters of policy, like decisions about where roads or air-
ports should be built, for instance, it is desirable that benefits and burdens should
be spread across the whole community in proportion to numbers and needs. So a
majoritarian decision procedure, in which compromises among interests are facili-
tated, is appropriate provided that that test of fairness is also met. In matters of
principle no such requirement holds, and fairness offers no objection to other
decision procedures including properly constructed judicial forums.¹⁷
¹² See my reply to Joseph Raz’s version of the objection in Dworkin and his Critics 383
(J. Burley ed., 2004).
¹³ I should say that Smith states my argument in what I take to be an odd way. He says that
integrity is needed so that law can be pervasive, that is, so that law can be extrapolated from recog-
nized explicit doctrine to cover cases unforeseen in that explicit doctrine. I think this puts the point
backwards: we need a legal culture in which law is extrapolated in that way in order to secure integrity
and equal concern.
¹⁴ See my article, Hart’s Postscript and the Character of Political Philosophy, 24 Oxford Journal of
Legal Studies 1 (2004), reprinted in R. Dworkin, Justice in Robes ch 6. (forthcoming 2006).
¹⁵ Law’s Empire, supra note 4, at 164.
¹⁶ See Sovereign Virtue, supra note 7, at chs 4 and 10. See also Freedom’s Law, supra note 2,
at Introduction.
¹⁷ I introduced the distinction between principle and policy in my earlier book, Taking Rights
Seriously, and explored the implications of the distinction for political fairness in several chapters of
that book.
298 Ronald Dworkin

Smith undertook to study only the arguments of Law’s Empire and I cite other
work, published before and after that book, not to criticize his essay but simply to
help in responding to the difficulties he raises. The overall structure and central
arguments of Law’s Empire are drawn from the more general arguments elaborated
in that other work. I emphasized in that book that integrity has no application to
matters of policy, for instance.¹⁸ But some of my remarks in Law’s Empire were
wrong and misleading. I said:
What is the special defect we find in checkerboard solutions? It cannot be a failure in fair-
ness (in our sense of a fair distribution of political power) because checkerboard laws are by
hypothesis fairer than either of the two alternatives. Allowing each of the two groups to
choose some part of the law of abortion, in proportion to their numbers, is fairer (in our
sense) than the winner-take-all scheme our instincts prefer, which denies many people any
influence at all over an issue they think desperately important.¹⁹
I made these remarks to show that integrity, which condemns checkerboard
statutes, is a political virtue distinct from political fairness. I could and should
have said only that fairness provides no argument against checkerboard statutes.
Since nothing in the way in which the checkerboard abortion law I imagined is
produced is necessarily unfair, we cannot explain our rejection of such laws by
appealing to political fairness. It was unnecessary for me to add the further, mis-
taken claim that fairness properly understood is advanced by such a statute or that
integrity and fairness so understood conflict. My argument for recognizing
integrity as a distinct virtue is in no way weakened by withdrawing that claim. I
am grateful to Smith for helping me to see and correct this mistake. But once the
mistake is corrected, and the argument of Law’s Empire made fully coherent with
my more general arguments in that book and elsewhere, his challenges to integrity
are all met.

Waldron

Jeremy Waldron has long been among the most astute legal philosophers. I am
grateful for his illuminating discussion here of what some writers have claimed to
be an important difference between my “early” and “later” suggestions about the
structure of law and legal reasoning. They believe that Hart and other positivists
answered my early arguments but have not so far responded effectively to my later
ones.²⁰ I agree with Waldron that the supposed difference between these two
stages of my work is largely illusory.
In this essay he resurrects a question I thought dead years ago. There is a radically
skeptical tradition in law: it surfaces cyclically. In the 1970s a group of legal

¹⁸ Law’s Empire, supra note 4, at 221. ¹⁹ Law’s Empire, supra note 4, at 179.
²⁰ I discuss this claim in ch 8 of my book Justice in Robes, supra note 14.
 Response 299

theorists whom Waldron calls “Crits” argued, much in the fashion of the legal “real-
ists” of decades earlier, that American and other mature legal systems are fissured
with “contradictions” so that any attempt to discover unifying principles cannot
succeed. I believe that judges should seek integrity in law and Waldron asks
whether I have ever answered the Crits’ claim that this is a pointless search. He cites,
as a powerful example of the Crits’ argument, Duncan Kennedy’s thesis that Anglo-
American law exhibits a deep schizophrenia. There are contradictory altruistic and
individualistic strains in the law, Kennedy says, just as there are in particular peo-
ple’s individual moral commitments, and these cannot be reconciled by any higher
order principle that orders or integrates them. This thesis poses a great danger,
Waldron believes, to my claim that a search for integrity in a political community’s
law is required to carry forward that community’s commitment to equal concern.
He describes this danger as a dilemma. My thesis that judges can construct success-
ful interpretations of their community’s legal practices is plausible only if I suppose
that a justifying scheme of principle—say an altruistic scheme—can count as a suc-
cessful interpretation of a community’s law even though it fits at best only half of
the legal material it claims to interpret. But if I do weaken my account of interpre-
tation in that way, then I can no longer claim that interpretation is a method of
identifying that community commitment of principle.
It is crucial, in assessing this challenge, accurately to describe the kind of con-
tradiction in legal materials that Waldron supposes the Crits to have demon-
strated in, say, American law. Kennedy argues that in different stages of American
legal history one or the other of his two supposedly contradictory strains, individ-
ualism or altruism, was dominant in the American legal culture.²¹ As we shall see,
it is unclear what that claim means. But in any case no contradiction between
what the law permits or requires in different historical stages of a community’s cul-
ture can pose any difficulty for a contemporary judge seeking integrity within
contemporary law. A judge may find integrity across the law that the community
now enforces even though that law is strikingly different from what was enforced a
century or more earlier. So Waldron must suppose that the Crits have established
more than an historical claim: that they have shown contradiction in the
American community’s contemporary legal practice. But Waldron offers no
examples at all of the conflicts he supposes endemic in that practice—perhaps he
thinks conflict so apparent that no examples are needed—so we must construct
our own example.
A judge must decide whether a defendant who broke a contract may be ordered
to perform as agreed or whether he may be held liable only for the economic dam-
ages the plaintiff actually suffered in consequence of the breach. The judge distin-
guishes two justifying schemes. Scheme A, which he calls an “individualist” scheme,
emphasizes the rationality of permitting “efficient” breaches of contract, that is,

²¹ See D. Kennedy, Form and Substance in Private Law Adjudication, 89 Harv. L. Rev. 1685,
1725ff. (1976) [hereinafter, Kennedy, Form and Substance].
300 Ronald Dworkin

breaches that improve the economic position of the breaching party without
damaging the economic position of the other party. Scheme B, which he calls an
“altruistic” scheme, emphasizes instead the moral importance of people keeping
their promises. Now distinguish between two kinds of interpretive situation the
judge might confront. He might find, first, that both schemes “fit” all the perti-
nent precedent cases in his jurisdiction in the following sense: a judge guided by
either Scheme A or Scheme B would have declared exactly the verdicts, given the
facts in each case, that the judges actually declared in all the precedent cases.
(Perhaps the plaintiff in each of these cases had originally demanded only eco-
nomic damages so that a judge guided by Scheme B would have concluded that
the plaintiff had waived his right that the contract actually be enforced.) So the
judge in the fresh case, in which the plaintiff does demand performance, cannot
say that either the “individualistic” or the “altruistic” scheme fits the legal material
better than the other. He must find some other way to adjudicate between the two
schemes. I shall call this a case of “rival” eligible interpretations.
Or the judge might find, second, that neither Scheme A nor Scheme B fits any
more than one half of the pertinent precedent cases in his jurisdiction. The plain-
tiff had demanded specific performance in all the cases: in half that demand was
denied, in half it was sustained, and there were no pertinent differences in the facts
of any of these cases that made it distinguishable from the judge’s new case. Since a
scheme of principle cannot be said to represent a commitment of the community
if it was rejected in as many cases as it was accepted, neither scheme provides
integrity and the judge must find some other basis for his decision. I shall call this
a case of “no” eligible interpretation.
Waldron does not suggest, as I read his essay, that integrity would be a false goal
if American law were shot through with rival-eligible-interpretation situations.
He is explicit that it is no-eligible-interpretation situations that he had in mind:
the schematic structure of conflict he offers as paradigm is explicitly such a case. I
agree that if no-eligible-interpretation situations were systemic in American law
the pursuit of integrity there would be silly. But as I said, Waldron offers no exam-
ples to support his supposition that this situation is pervasive in American law. He
apparently relies on Kennedy to support the supposition, but his reliance is very
much misplaced because, so far as I can see, Kennedy does not even claim, let
alone try to illustrate, an endemic no-eligible-interpretation situation in
American law. He does not claim even a single case of such a situation.
Kennedy says that “individualism” and “altruism” are rival ideologies that can-
not be combined in a rational overall scheme that combines elements of both. He
agrees that most people have elements of each in their moral personalities, but he
suggests that this produces not any kind of amalgamation of the two but only a
moral schizophrenia within us all. These are very implausible claims and
Kennedy’s account of the two values does not sustain them. He believes that
“liberals” are dominated by “individualism” and “communitarians” by altruism.
Contemporary political philosophy shows his mistake. Utilitarianism is the most
 Response 301

altruistic of influential political theories: it supposes that people have a basic duty
to count the welfare of others as equally important to their own welfare in identi-
fying moral and political principles. But it has provided the impetus for political
ideologies, including some forms of economic rationality, that Kennedy counts as
“individualistic” instead. More explicitly liberal political theories insist on indi-
vidual rights to liberty that communitarians reject but also insist on egalitarian
goals that communitarians also reject.²² These various accounts of justice claim to
do what Kennedy declares impossible: they offer a principled basis for recognizing
both individual rights and shared obligations of economic social justice not as
some Procrustean compromise but as flowing from more basic ideals of liberty
and equality. Nothing in Kennedy’s writings offers to show that the more sophisti-
cated integration of those values fails.Waldron is right that Critical Legal Studies
is now, as he puts it, “moribund.” That is in good part because its claimed founda-
tions in philosophy proved to be very fragile indeed.
We may put that broad criticism aside now, however, and assume that
Kennedy’s account of individualism and altruism as two wholly incompatible
mind-sets is persuasive. We must then notice that he claims, explicitly and repeat-
edly, that each of the two ideologies could justify almost all of the law in the sense
I indicated earlier: law-makers moved even wholly by one or the other might have
produced the very law that we have. Here are representative qualifications of his
claim that the ideologies he names clash:
When we set out to analyze an action, and especially a judicial opinion, it is only rarely pos-
sible to make a direct inference from the rhetoric employed to the real motives or ideals
that animate the judge. And it is even harder to characterize outcomes than it is personal-
ities or opinions. It will almost always be possible to argue that, if we look hard at its actual
effects on significant aspects of the real world, a particular decision will further both altru-
ist and individualist values, or neither. I will therefore avoid talking about “altruistic out-
comes” as much as possible.²³
Although individualism and altruism can be reduced neither to facts nor to logic,
although they cannot be used with any degree of consistency to characterize personalities or
opinions or the outcomes of lawsuits, they may nevertheless be helpful to this enterprise.²⁴
What enterprise? That is very far from clear, but here is Kennedy’s direct answer.
“The ultimate goal is to break down the sense that legal argument is autonomous
from moral, economic, and political discourse in general.”²⁵ I don’t know who
ever had that “sense”; I doubt Waldron means to suggest that I do.
So Waldron is wrong to appeal to Kennedy to support his own claim that con-
temporary American law is shot through with no-eligible-interpretation conflicts.
Kennedy argues at most that it is shot through with rival-eligible-interpretation
conflicts: that each body of law can be seen as representing the dominance of one

²² See, e.g., J. Rawls, A Theory of Justice (1971), and my book, Sovereign Virtue, supra
note 7. ²³ Kennedy, Form and Substance, at 1723.
²⁴ Kennedy, Form and Substance, at 1724. ²⁵ Id.
302 Ronald Dworkin

or the other of his two ideologies. I would reject that latter claim—it is plainly
flawed if, as I believe, Kennedy has failed to state any ideology that could justify
any decision at all. But I need not pursue that claim because, as I said, Waldron
does not suppose that rival-eligible-conflict would produce the dilemma he poses
for integrity.
How might he show that no-eligible-interpretation is pervasive? He accepts that
any statement that the law is shot through with contradiction must be the product
of—not something antecedent to—interpretation. What interpretive strategy could
yield that formidable result? In my illustration I supposed that the no-eligible-
interpretation requires showing that Schemes A and B each fit no more that half
the interpretive data of past judicial decisions in the sense I described. A law-
maker guided by either would impose the verdict of only those past decisions,
given the facts the precedents presented, and would reject the verdict of the other
half. That seems very unlikely in any jurisdiction with even a weak system of
precedent: at least sufficiently unlikely to call for some examples to show that it is
nevertheless the case. Any appellate decision holding that specific performance
may not be ordered in a certain type of contract case would overrule and remove
from the interpretive data set previous cases of that type ordering that remedy.
It may be, however, that Waldron contemplates a less dramatic kind of no-eligible-
interpretation conflict. He may have in mind a situation in which only Scheme A
fits in that way a settled practice of denying specific performance in contract cases
but only Scheme B fits a settled practice of allowing punitive damages in breach of
contract cases. That is also very unlikely, however, because contradiction between
doctrine in areas of law so closely interconnected would almost certainly also be
eroded by the practices of precedent and academic criticism and restatement.
Perhaps Waldron has in mind conflict between areas of law much more widely
separated: that contract law is inconsistent with Scheme A but tort law is inconsis-
tent with any denial of Scheme A, for instance. But once again that would need
careful illustration. It would also require making good a very ambitious claim:
that no theory of why tort and contract should differ can justify that doctrinal
distinction between the two domains.
Of course I cannot demonstrate that no conflicts of these kinds can be found. I
must await actual supposed examples. However, Waldron adopts an entirely differ-
ent strategy than I just described: he offers to show that no-eligible-interpretation
situations are more common than I suppose not by widening the areas of doctrine
that an interpretation is asked to justify but by widening the data set that an inter-
pretation must confront within each area. He says that a scheme of principle is
not a successful interpretation of a single past judicial decision, even if a law-
maker guided by that scheme would have ordered exactly the verdict it imposed
on the facts it presented, unless the scheme also fits the reasoning of the judge’s
opinion that accompanied the decision. Even if all the precedent cases refused to
award specific performance as a contract remedy, Scheme A would fit only those
decisions accompanied by opinions in which a judge declared principles that the
 Response 303

interpreter regarded as “individualistic” principles of economic rationality; it

would not fit those other decisions in which the judge cited concern for the situa-
tion of people forced to keep their word. This further argument rests on two
assumptions: first, that our legal record would be full of no-eligible-interpretation
doctrine if an interpretation were required to fit what the judge said as well as what
he did and, second, that legal interpretation should include that requirement.
The first assumption is not compelling; certainly not in the absence of any
examples offered to confirm it. True, the legal culture changes over time: what was
once regarded as the principled basis of some established rule of law gives way to a
different and more contemporary basis even though the rule itself is retained.
Many ancient common law rules are now thought justified by considerations very
different from the reasons why judges originally declared those rules. But Waldron’s
interpretive strategy would presumably require interpretation to fit only expres-
sions of opinion in the most contemporary decisions enforcing an old rule. The situ-
ation is different for old statutes because these do not require any contemporary
restatement for their force. A contemporary Congress might disown the principles
that inspired Social Security in the New Deal. But Waldron, sensibly, does not
insist that his strategy be applied to statutes.
His second assumption is more interesting, however, and we should consider it
independently. Waldron says that I would not (and should not) reject the need for
an interpretation of judicial decisions to fit opinions as well as verdicts.²⁶ The
“would not” is wrong: I have several times explicitly rejected that requirement,²⁷
and I have often taken Cardozo’s opinion in MacPherson v. Buick Motor Company,
which rejected it, as a paradigm of legal interpretation.²⁸ What about the “should
not”? Waldron says that if I rejected the requirement that interpretations must fit
opinions I would be embracing the legal “realism” of such writers as Jerome Frank.
Frank wrote in the era of obsession with Freudian explanations of absolutely
everything, and he championed a Freudian explanation of judicial reasoning. But
surely one can believe that legal interpretation is not hostage to judicial opinions
without also believing that such opinions are only rationalization of Oedipal lusts.
As I tried to explain in Law’s Empire, we must draw our interpretive strategies
in law from our sense of the point of integrity. If we suppose, as I do, that this point
is to keep faith with a requirement of equal concern then our strategies must
emphasize what we as a community do to or for people and give only a secondary
and derivative place to the reasons different actors in the legal process offer for what
we have decided to do. If Waldron disagrees, then that is the true focus of our pres-
ent disagreement, and he should offer a reason of political morality why judicial
opinions should have a more important role in interpretation than legislators’
opinions have. I conclude not that I have answered the Crits but that, as I suspected

²⁶ Waldron, 170.
²⁷ See, e.g., Law’s Empire, supra note 4, at 247–248, 284–285; Taking Rights Seriously, supra
note 10, at chapter 4. ²⁸ 217 N.Y. 382 (1916).
304 Ronald Dworkin

long ago, there is nothing in what they once said for me to answer. We cannot
be sure, before we look, that constructive interpretation can produce integrity in
any particular area of the law. But we have no reason to think, in advance, that
it cannot.
I should comment on one further aspect of Waldron’s argument. He imagines
that if I were persuaded that no-eligible-interpretation situations are pervasive in
law, I would try to rescue the idea of integrity from that debacle by proposing that
judges should choose between rival eligible interpretations by adopting the inter-
pretation that best matches their own moral convictions. He says that that strategy
would not provide integrity; rather, it would concede that there was no integrity
to find. I agree with his conclusion that no-eligible-interpretation cases imperil
integrity, as I said. But one objection he makes to the strategy is odd and would
also apply to my claim that judges should prefer the morally better interpretation
in rival-eligible-interpretation cases.
He says that judges who choose what they regard as the morally best result are
only taking sides between parties who themselves disagree about exactly which
result is morally best. I hope it is clear, in spite of some of what Waldron says, that
in recommending that judges choose on moral grounds I am not assuming that
their choice is right because it matches their own convictions. Of course it does.
But a choice is right only when—and then only because—those convictions are
themselves right. It is nevertheless true, of course, that in choosing they are taking
sides about the issue in controversy. How can they not do that? When the law is
unclear the judge must in the end take sides about the very issue that divides the
litigants: they disagree about what the law requires. Presumably that fact furnishes
no objection to the institution of adjudication. In my view what the law is
depends, in some cases, on what morality requires. So in taking sides on the for-
mer issue, which he must do, a judge necessarily takes sides on the latter as well.

Perry

Professor Perry’s discussion of associative obligations and political obligation clari-

fies important issues and brings significant problems to light. He supposes that if
associative obligations are genuine they must arise from practices that have what
he calls intrinsic value and he notes that if that is correct then some of my claims
about the range of associative obligations—in particular, my claims that unions
and college faculties can be the source of such obligations—must be mistaken.
I agree that both friendship and close relations with one’s family contribute to the
value of a life and that this is also true, for at least some people, of political activity.
But I am not persuaded that only practices that can be said to be intrinsically
rather than instrumentally valuable can be the source of associative obligations.
I believe that such obligations can arise when and because people are joined in cer-
tain kinds of relationships—typically relations of collaboration or partnership—
even when their lives would lack nothing if they were not.
 Response 305

I agree, of course, that normative conditions must be met in order that associa-
tive obligations be genuine. People do not have obligations of family, friendship or
political responsibility whenever or just because they or their friends, family or fel-
low citizens think they do: the conditions of special and equal concern that I
described and that Perry reviews must be met. But these may be met without sup-
posing that the practice they structure itself has intrinsic value. Perry suggests that
my statement of these conditions needs revision because they are not all met even
in all the circumstances in which he and I agree that associative obligations arise. I
argue, for instance, that the mutual concern that sponsors genuine associative
obligations must be both personal and equal; he doubts that these conditions are
met in the case of political obligation.
“A general obligation to obey the law,” Perry says, “is usually thought to be owed,
if to anyone, to the community or state itself rather than to one’s fellow citizens
considered one by one.”²⁹ My own view is different, however. I believe that the
duty is owed not to the collective entity but to its members as individuals. Of
course the duty is enforced by a collective entity: it is that state that prosecutes (or,
indeed, waives) the duty. But the state acts, in this instance as well as all others, as
the agent of the citizens acting through it. The legislation that establishes particular
political obligations in a democracy is the creation of the citizens in the same way
and it is a condition of ideal democratic legitimacy, in my view, that what these citi-
zens do collectively can be justified as showing equal concern by each for each.
Perry also doubts that the concern that grounds obligations of family is always
equal concern. Consider, he suggests, the obligation a parent owes to a new-born
child. Perhaps that child will in due course acquire an obligation to care for his
parent, but the parent’s immediate obligation is nevertheless independent of any
prediction that the child will accept that responsibility. But the needed equality of
concern is only a background assumption of the practice, not a condition whose
realization can be demonstrated in every moment of the relationship. If a respon-
sible adult makes plain his rejection of any concern for his parents, or for the other
members of the family, their associative obligations to him are at least attenuated
and, if the rejection is sufficiently decisive, ended.

Gardner

I am thoroughly bewildered by John Gardner’s contribution to this volume. He

argues throughout that I equivocate between accepting and rejecting a variety of
propositions to which he attaches Greek letters, but since I cannot understand
what any of those propositions means I suspect that I do not so much equivocate
about but ignore them. Gardner’s main claim is amusingly mischievous: he claims
that a careful reading of my book, Law’s Empire, shows that I am a legal positivist

²⁹ Perry, 191.
306 Ronald Dworkin

after all. But I do not understand his argument for that arresting claim any more
than I understand anything else.
My trouble begins—and perhaps ends—with Gardner’s breathtakingly
promiscuous personifications of law in all his Greek-letter propositions. He
spends much of his essay considering whether I embrace the proposition he labels
(␣): “Law aims to be morally justified”. He notices that I have objected to the
argumentative use some legal philosophers make of personification.³⁰ For exam-
ple, I do not think that Joseph Raz’s statement that the law necessarily claims
moral authority can be unpacked without making the resulting statement either
plainly false or incompatible with Raz’s positivist claims.³¹ Gardner labels Raz’s
statement (␤) and defends it, so we should pause over his defense before returning
to the mysteries of (␣). Gardner says that Raz means only that there is no law in a
community unless its legal officials claim moral authority for their decisions. I
anticipated this reading of Raz, and said that it cannot seriously be thought that
whether or not law exists in a particular community depends on the meta-ethical
or moral opinions of its officials. Suppose, I asked, the bulk of a community’s offi-
cials hold the moral views of Oliver Wendell Holmes? Gardner does not accept
this counterexample. He concedes that Holmes’ “extrajudicial” writings express
the view I describe, and he does not cite any judicial writings in which Holmes
states any contradictory view. But he asks a series of questions about Holmes’ atti-
tudes, expecting the answer “yes.” In fact, the best answer to all these questions,
taken to be questions about the historical Holmes, is very probably “no.” But only
one question counts: since Holmes was deeply skeptical of the very idea of moral
authority, he would have denied that his judicial opinions have moral authority.
In any case, to rescue Raz, Gardner would need to claim not only that I am
wrong about Holmes but that there would not be law in America if most judges
came to hold the meta-ethical opinions I ascribed to Holmes, and the implausibil-
ity of that suggestion remains. A group of South African judges met in a clandes-
tine conference with officials of the then illegal African National Congress in
Britain in the last years of apartheid: these judges made plain that they did not
think the law they administered daily had any moral authority. Would Raz or
Gardner think that these judges were not legal officials, or that if a great many
other South African judges shared their opinion there was no law there?
Gardner’s discussion of Raz’s claim is mysterious for a further reason, moreover.
He says that although Raz means that legal officials claim moral authority his per-
sonification is not to be understood as an elliptical statement exhausted by that
fact. Though he agrees with me that “it takes a human being to make a moral
claim” he adds that “it does not follow that human beings are the only things that

³⁰ I am guilty, I know, of personifying law and much else in the titles of books and articles. I doubt
this non-argumentative and harmless personification has played any part in the spate of argumenta-
tive personification now popular, but I apologize if it has. I believe that the personifications that figure
in my own arguments can all easily be dissolved.
³¹ See supra note 14, Justice in Robes, chapter 7.
 Response 307

make claims. Law makes claims through human beings acting on its behalf.” This
statement ruins any proposed redemption of Raz’s troublesome personification
because it simply invokes the personification all over again. Gardner uses “law”
presumably to refer either to a set of social practices or to a set of norms of some
kind. How can either have a “behalf ”? Is Hegel waiting in the wings? Or is there
some further dissolution of that further personification that allows Gardner’s
claim to make sense? Shall we say that it means only that legal officials claim that
legal requirements (rather than those of some other institution) have moral
authority? Then Gardner’s further elaboration of Raz’s claim is simply redundant.
So I remain perplexed what anyone could think (␤) means who supposes that it
is true. I have exactly the same problems with Gardner’s (␣). What can it mean to
say that law aims to be morally justified? I have argued, in Law’s Empire and else-
where, that the concept of law is an interpretive concept, and that lawyers must
therefore identify what the law requires or permits on some matter through a
process of interpretation that seeks the best available moral justification of the
broad legal practices of their community.³² Is (␣) a baroque way of saying the
same thing? If so, then of course I accept it. But I doubt that is what Gardner takes
(␣) to mean, because he says that (␣) is entailed by my suggestion that interpreters
might sensibly take the practice they are interpreting to have, as its goal, the
appropriate regulation of the use of coercive force by the state. I emphasize,
throughout Law’s Empire and elsewhere, that interpreters who begin in that way
may sensibly, and certainly without contradiction, adopt very different, more
detailed theories of legal interpretation from my own, including a theory very like
legal positivism.³³ Should we understand (␣) instead in something of the same
way Gardner wants us to understand (␤), leaving out the Hegel? Then we would
have to say that legal officials all reason about what the law is in the way I recom-
mend, which is undoubtedly false, or that they are not really legal officials unless
they reason in the way I recommend, which is tempting but I fear overreaching. I
cannot assign any sense to (␣) that fits what Gardner says about its implications
and that is in the least plausible, so I am not surprised that he makes such heavy
weather of deciding whether I believe that proposition or not.
What shall we make of Gardner’s further proposition (␥): “Law is morally justi-
fied”? Does that statement claim that every true proposition of law is morally jus-
tified? If so, it is puzzling why Gardner would spend any space pondering whether
I accept (␥). Of course I don’t. The mystery deepens when he explains that (␥) and
(␣) contradict one another. His explanation assumes that (␥) means that law is not
just always justified but is necessarily justified, so that the contradiction is like the
contradiction some (but not all) theologians find in the proposition that God
aims to be moral and necessarily is moral. How can one aim to be what one neces-
sarily is? In fact there is no contradiction here if, as Gardner suggests, (␣) includes
³² This is only a very rough summary of my account of legal reasoning. For a recent fuller state-
ment, see the Introduction to my recent book, Justice in Robes, supra note 14.
³³ Id. at chapter 8.
308 Ronald Dworkin

the proposition that no one is a legal official unless he aims at morally justified law.
Then the conjunction would mean that those who are legal officials, because they
aim to produce justified law, necessarily do produce justified law, which is not a
contradiction though it is of course very silly. So either (␣) or (␥) must mean
something different from what first appears if they are to be, as Gardner insists
they are, incompatible. There may be some sense of (␥) such that it is a genuine
interpretive question whether I accept it. But I have no idea what that is. In any
case, however, my remarks about Nazis, which Gardner quotes, have nothing to
do with (␥).³⁴ I was explaining how we should understand someone who says that
the Nazis had no law: we should understand him as denying that their legal
arrangements were such as to justify coercion. The undoubted fact that some
people sometimes use the sociological concept of law to carry that moral freight
doesn’t entail that true propositions of law are always or inevitably morally justi-
fied. I believe that reading the entire section of Law’s Empire from which Gardner
takes his quotation makes that plain.³⁵
Now we come to the supposedly most consequential of the Greek-letter
propositions:
(␦) In any legal system, the law is made up of norms which are part of the law
only because some legal official engaged with them, and such an agent is a
legal official only because, by engaging with norms in certain ways, he or she
can make them part of the law.
I assume that (␦) is meant to state a version of legal positivism.³⁶ If (␦) is right, I
assume, then the proposition that a plaintiff who cannot prove which manufac-
turer made the drugs that injured her is entitled to market-share damages from
them all is false unless and until some judge so rules, because only then has an offi-
cial “engaged” with that proposition in an appropriate way.³⁷ So when a judge
does declare that proposition, and does award market-share damages, his decision
cannot be seen as applying the law but must be treated as making up new law
through his “engagement” and applying it retroactively. That is a familiar view
among lawyers who call themselves positivists: it is no doubt Gardner’s own view.
I have argued against it steadily over three decades, repeating my rejection in a
great variety of recent publications some of which Gardner cites. Gardner’s claim
that I actually embrace that view is therefore indeed “striking.”³⁸

³⁴ Gardner, 217.
³⁵ See also the Introduction to Justice in Robes supra note 14. I there distinguish the sociological
concept of law which someone uses when he denies that the Nazis had law from the doctrinal concept
which Law’s Empire mainly explores.
³⁶ In any case, the formulation plainly needs tidying in various ways. We need to know more
about what “engaging” is and we need to take account of the fact that many people—district attor-
neys, for instance—are legal officials even though no amount of “engaging” by them makes law.
Presumably Gardner has or will remedy these obvious defects and I shall ignore them.
³⁷ See my description of “Sorenson’s Case” in Hart’s Postscript and the Character of Political
Philosophy, supra note 14. ³⁸ Gardner, 220.
 Response 309

Gardner produces this paradox by ignoring everything I have written about

interpretation and law since 1986 and instead pursuing a flatteringly close textual
analysis of a few lines in Law’s Empire. However, he misreads these lines in a way
that I and other writers had to correct years ago. He fixes on this passage from that
book: “Roughly, constructive interpretation is a matter of imposing purpose on an
object . . . Creative interpretation, on the constructive view, is a matter of interac-
tion between purpose and object.”³⁹ That is a general description meant to apply to
all examples of what I then took to be constructive interpretation: literary interpre-
tation, for example, as well as the interpretation of a variety of social practices.
Gardner apparently assumes that when this description is applied to legal interpre-
tation it means that an interpreter begins with something he takes to be a correct
statement of the law as it stands—that is, the “object” in question—and then trans-
forms that statement creatively so as to move the law closer to what it ought to be.
He “engages” with the existing law that is, in order to produce new and better law.
But everything in Law’s Empire contradicts that reading. The “object” of legal inter-
pretation is a set of practices that include tentatively taking certain propositions of
law to be true. Imposing purpose on that set of practices means finding a justifica-
tion for them that is then used to identify which propositions of law, including
those with whom no one has previously “engaged,” are in fact true.
Gardner’s mistake can be readily identified by reading the discussion of “Stages
of Interpretation” that figures in the same chapter as the passage to which he
devotes such attention.⁴⁰ There I contrast the “tentative” assumptions about what
the law is that are collected at what I call the “pre-interpretive” stage with judg-
ments about what the law “really” is at the “post-interpretive” stage, judgments
that may require concluding that some of the tentative assumptions about the law
at the earlier stage were “mistakes,” that is, not accurate statements of the law as it
actually is. This is a process through which, if I am right, lawyers come to recog-
nize the truth about law, not a process in which they create new law to replace
what they acknowledge the law already to be. No doubt arguments can be made in
defense of legal positivism, though I agree with Scott Shapiro, Jules Coleman and
other positivists that these arguments have yet to be made.⁴¹
I have so far said nothing about the discussion with which Gardner begins his
essay: he says I think that law has a unifying aim which is the aim of regulating a
political state’s use of coercion. He believes he refutes that proposition by imaging
law in a community of angels who obey that law unhesitatingly so that coercion is
not only never needed but never even contemplated. Surely, he insists, regulating
coercion cannot be the aim of law in such an angelic community. He apparently
thinks I believe it is a defining test of law that “it” has the aim of regulating

³⁹ Gardner, 219, quoting Law’s Empire, supra note 4, at 52.

⁴⁰ Law’s Empire, supra note 4, at 65ff.
⁴¹ See Justice in Robes, supra note 14, at chapter 8.
310 Ronald Dworkin

coercion so that no set of practices that cannot plausibly be said to serve that aim
could count as “law.” But I was talking about strategies of legal interpretation and
therefore about interpreters. I said that it would be helpful if a view about the
“point” of law were sufficiently widely accepted in a community so that, as I put it,
interpretive arguments could take place “on the plateau it furnishes” and proceed
by contrasting different conceptions of how that point is best understood and
achieved.⁴² I did not say that amenability to that interpretive strategy is a concep-
tual necessity so that anyone who rejected it would make a conceptual mistake.
On the contrary I conceded that there are “theories that challenge rather than
elaborate the connection [my suggestion] assumes between law and the justifica-
tion of coercion.”
I certainly did not say, as Gardner apparently believes I did, that this interpre-
tive strategy would be appropriate everywhere we find anything we might be
tempted to call law: it would not be appropriate if angels had to interpret God’s
law of angelic duties. Gardner’s misreading is another example of the serious
confusion that has been engendered in legal philosophy by a failure to distinguish
between the sociological and doctrinal concepts of law. We have great leeway in
deciding whether angelic law (or international law or Nazi law) should be called
“law.” The sociological concept of law in play in those questions is an indetermin-
ate criterial concept: we can give either answer so long as we make plain what
point we mean to make. We must not confuse that concept with the doctrinal
concept in play when we ask which interpretive strategies are appropriate to decid-
ing what the law of some community really is.⁴³ My suggestions about the justifi-
cation of coercion are claims about the right strategy in communities of the kind
lawyers are normally called upon to interpret.

Greenberg

Professor Greenberg’s essay is admirable in bringing the concepts and argumenta-

tive style of metaphysics to bear on traditional issues of jurisprudence. I applaud
him for reminding us that legal philosophy is part of general philosophy, not a
special insulated compartment of thought with its own distinct questions and
methods. I agree, as he notes, with his conclusion that “value facts are among the
determinants of the content of the law.”⁴⁴ But it is worth distinguishing two
forms of that claim: (1) value facts figure in the truth conditions of propositions of
law; (2) value facts figure in the justification for any theory about the truth condi-
tions of propositions of law. I emphasize the importance of that distinction for
jurisprudence in my recent book, Justice in Robes.

⁴² Law’s Empire, supra note 4, at 93.

⁴³ See Justice in Robes, supra note 14, Introduction and chapter 8. ⁴⁴ Greenberg, 225.
 Response 311

In How Facts Make Law, Greenberg argues only for proposition (2), and we
should therefore notice that the most influential contemporary versions of legal
positivism accept (2) and argue only against (1). They argue, for example, that
democracy is better served, or that a political community is more efficient, or that
authority is better respected, or that citizens are more likely to be critical of their
government, if judges accept that moral or other normative claims are not relevant
in deciding what the law requires. In that way these “political” positivists appeal to
“value facts” at what we might call the jurisprudential level of analysis to show the
irrelevance of such facts at the doctrinal level. It is true that some positivists argue
against (1) on what they claim to be purely conceptual grounds that are consistent
with rejecting (2) as well. I agree with Greenberg that these arguments are mys-
terious and in any case fail. But my arguments in Law’s Empire are meant to sup-
port (2) as well as (1).
I should add that I agree with Greenberg’s interpretation of my own views
about the interaction of the dimensions of fit and justification in legal interpreta-
tion.⁴⁵ As I have tried to explain elsewhere, these dimensions are best understood
as each drawn from a political value. I said in Law’s Empire and elsewhere that the
crude “threshold” account of the interaction of the two dimensions was meant
only to illustrate the difference between them.⁴⁶ However, I am uncertain that
Greenberg has located any metaphysical difficulty in that crude account. If we
supposed that two interpretations of the legal record fit that record equally well,
and decided between them on moral grounds, we would not be treating the legal
record as irrelevant to the content of the law. If the record had been different, the
better interpretation might not have been available at all.

⁴⁵ See Greenberg, supra Chapter 10, notes 46 and 47, and pages 262–63.
⁴⁶ See Dworkin and his Critics, supra note 12.
 Index
abortion autonomy 29–30, 80–1, 292
checkerboard solutions 121–4, 128–37,
139–41, 249–50, 297 beliefs 252, 269
coherence 133–5 Bill of Rights 52–3, 62
irrelevant considerations 129–33, 135 Brown v. Board of Education 41, 43–6, 48
justice 135–7 criticism of 44–5
law practices and the content of law 249–50 democracy 46
rape victims only, for 129–33, 135–6, equality 63
140–1 judicial review 53
actual settled cases moral reading of US Constitution, Dworkin’s
coherence 70–2, 85, 89–96, 98, 100–1 55, 58–9, 294
hypothetical cases, asymmetry between 70, public and private views of 45
91–2, 94 racial segregation 43–5, 55
jurisdictions, cases from other 95–6
mistakes 90, 95, 100–1 certainty, value of 111
state courts, precedents from 95–6 checkerboard solutions, Ronald Dworkin and
stronger requirement 89–90, 93–4, 98 abortion 121–4, 128–37, 139–41, 249–50, 297
weight of 85, 89–90, 93–4, 98 arbitrary distinctions 122, 125–9
altruism 164, 166–70, 174, 181, 299–302 coherence 120–2, 125–6, 133–5, 296
arbitrariness 122, 125–9 compromise 122–3, 127–9, 134, 136–8, 140–2
associative obligations consistency 133
academic colleagues 193, 304 criticism of 125–42
attitudes held by members of group 193 definition 121
commercial partnerships 198, 304 differential treatment 125–6, 139–41,
conditions 188, 190–5, 305 144–8, 296–8
consent 190, 197 equal treatment 139–41
Dworkin, Ronald 176, 178, 190–7, 304–5 fairness 120–5, 128–9, 133, 135–40, 145,
equal concern 191–2, 305 297–8
fairness 188–9 injustice, minimising 123
family relationships 190–4, 198, 304–5 integrity 119–53, 296–7
friendship 190–7, 304–5 interpretation 126–8, 140–2
gender equality 191–2 irrelevant considerations 129–33, 136
instrumental obligations 198 justice 120–5, 133–9, 140–1
integrity 188–9, 198–205 differential treatment 140–1, 296
interpretation 194–7 recognizable principles of 135–9
intrinsic value 195–7, 304–5 justification 125–6, 138, 140–2
Law’s Empire 196 law practices and the content of law 249–50
legal outcomes as 176, 178 Law’s Empire 119–42, 297
meaning 189–90 legislation 121–5
moral obligations 194–5, 197 morality 121–2, 129–33, 135–8
normative conditions 305 political compromise 137
obedience to law 183–9, 191, 194–5, 305 political integrity 119–53
political obligations 183–9, 191, 193–5, preferences 123–4
198–205, 305 principle, matters of 122, 133, 135
practices 194 compromise 122–3
promises 190 conflict between 122
special obligations 189–97, 305 ranking 122
value 195–8, 304–5 recognizable 135–9
assumptions or beliefs 252 rape, abortion only for victims of 129–33,
attitudes and dispositions 271–6, 280–3, 287, 135–7, 140–1
289 religion 138–9
314 Index
checkerboard solutions, precedent 86
Ronald Dworkin and (cont.) punitive/regulatory distinction 96–7
social welfare 122 settled 71–2, 81, 83–4, 90–1, 94–7, 294–5
unreasonable compromises 128 thought experiments, as 81–4
coercion indifference curves 90
justification 188, 217, 308 integrity 88, 98, 133, 143–8, 150, 152,
laws aims in Law’s Empire 208–9, 216–17, 162–5, 171, 175, 178–9, 292
309–10 intervening cases, problem of 84–101
political obligation to obey the law 188, 203 mistake 86, 95, 99–101
coherence, legal reasoning and 69–84 precedent 85–101
abortion 133–5 prospective application 85–7, 93
actual settled cases 71–2 retroactivity and 69–70, 84–9, 92–5,
hypothetical cases, asymmetry between 100–1
70, 91–2, 94 jurisdictions, cases from other 95–6
jurisdictions, cases from other 95–6 justice 133–5
mistakes 90, 95, 100–1 Kress, Kenneth 70, 74, 84–92, 98, 100–1,
state courts, precedents from 95–6 295
stronger requirement 89–90, 93–4, 98 law practices and the content of law 248–9
weight of 85, 89–90, 93–4, 98 law-making 162
alternatives legislation 86, 93, 147
non-reductionism 82 like cases alike, treating 82–4
ranking of 71 mistake 86, 90, 95, 97–101, 295
reasons that apply to the 70–3 model 248–9
background 156 moral reading of US Constitution, Dworkin’s
checkerboard solutions 120–2, 125–6, 292
133–5, 296 morality 114–16, 292
coherence function 71 non-evaluative acts 82–3
compromise 134 normative role of legal reasoning 82
conservatism 89, 90–2 political integrity 143–8, 150, 292
consistency 81, 88, 89, 133 practical reasoning 69–70, 89
data gathering 71–2 precedent 69–101, 114–16, 118
decision theory 90 actual settled cases, weight given to 85
deliberation discretion 88
abstract characterization of 70–4 fairness 94
deliberative matrix 72–3, 75, 77 hypothetical cases 86
ethical, examples of 74–82 intervening cases 85–101
examples of 74–81 jurisdictions, cases from other 95–6
legal, examples of 74–81 legislation 86
reasons 73 mistake 97–101, 295
stages of 71–2 prospective effect 85–7, 93–4, 97–8
determinacy, assumption of 100 retroactivity 94–101, 295
discretion 88 settled cases 83–94
Dworkin, Ronald 91–2, 133 state courts, decisions of 95–6
intervening cases, problem of 85, 95, stronger requirement 89–90, 93–4, 98
98–101 preferences 90
Law’s Empire 69, 87–8 principles 133, 135, 150, 158
political integrity 143–8, 150, 292 prospective application 85–7, 93–4, 97–8
ethical deliberation 74–82 punitive/regulatory distinction 96–7
extra-judicial developments 92–4 rationality 69
fairness 94, 133, 135 reasons 73
federal statutes, states’ interpretation of 95–6 reflective equilibrium 69
generalizations 81 retroactivity 69–70, 74, 84–9, 92–101, 295
hypothetical cases 70, 294–5 revisionary concept of law 92–4
actual settled cases 70, 91–2, 94 ripple effect 84–5
answers to 84 scientific theories 81–2, 94
generalizations 81 settled cases 83–94
mistake 97 actual 70–2, 85, 89–96, 98, 100–1
 Index 315
hypothetical 71–2, 81, 83–4, 90–1, 94–7, Brown v Board of Education 41, 43–6, 48
294–5 democracy 43, 46–8
like cases alike, treating 82–4 Dworkin, Ronald 41–2, 48–56
precedent 83–94 elections 46
weight of actual 85 equality in US Constitution, concept of
settled preferences 90 41–5
social practices, changes in 93 individual rights 46, 48
stare decisis 114–16, 118 legislatures 42, 43
state courts, precedents from 95–6 liberty claims 42–4
thought experiments 81–4 Justices of the Supreme Court
underdetermination 100–1 activism 42
communal obligations see associative liberty claims 41–4
obligations passivity 42
communitarianism 175, 300–2 restraint 48
compromise judicial review 44, 48
checkerboard solutions 122–3, 127–9, legislature, policy-making and 46
134–7, 140–2 majority rule 47, 48
coherence 134 New Deal 43–4
political 137, 142–3, 145, 152 police powers 43
principle, matters of 122 populism 43, 46–7
unreasonable 128 racial segregation 41, 43–5, 47
conservatism 34, 89, 90–2 social change 42–3
consistency constitutions see also constitutional judges,
checkerboard solutions 133 constitutional theory, United States
coherence 81, 88, 89, 133 Constitution
conservatism 89 constitution constructivism 28–9
integrity 15–16 constitutional purposes, importance of
moral reading of US Constitution, Dworkin’s 17–19
50 constitution-perfecting theory 29–30
stare decisis 115 democracy 26–8, 30
United States Constitution 15–16, 50 making 17–19
constitutional judges 1–4 moral reading of US Constitution, Dworkin’s
activism 42, 54, 59 6, 7
control of 2 super-majoritarianism 17–18
democracy 1–3 constructivism 156–7, 168–70, 173, 175,
Dworkin, Ronald 1–4 177–81
guardian or rights, mistrust of judges as 42 contents of the law see also law-determining
guiding principles, identification of 56 practices and contents of the law,
international human rights crimes 3 relationship between
interpretation 4, 42 context 11
job description 56–7 coordination of conduct for common good
jurisdiction 3 207–12, 215
liberty claims 41–4 counter-majoritarian difficulty 26
minimalism 60 Critical Legal Studies movement, Dworkin’s
moral decisions by judges 1 theory of integrity and 155–81
moral reading of US Constitution, Dworkin’s altruism 164, 166–70, 174, 181, 299–302
50–1, 54, 56–60 associative or communal obligations, legal
objectives, definition of 3 outcomes as 176, 178
passivity 42 background 155–61, 179
personal preferences 50–1 coherence 156, 162–5, 171, 175, 178–9
reason, use of 2–3 communitarianism 175, 300–2
restraint 48, 58 competing principles 166
standards, enforcement of 1–2 conflicting entities in law 163–4, 299
Supreme Court 41–4, 48 constructivism 156–7, 168–70, 173, 175,
constitutional theory 41–67 177–81
Bickel, Alexander The Least Dangerous contradictions 165–6, 171–4, 179–80, 299
Branch 45–7, 53–4 demonstrability 171–2, 180
316 Index
Critical Legal Studies movement, Dworkin’s deliberative matrix 72–3, 75, 77
theory of integrity and (cont.) democracy 27, 28–9
descriptive optimism 155 discrimination 78–9
discrete norms 159 drugs, searches for 75–6
egotistical tendencies 167 Equal Protection Clause 77–9
elitism 162–3 estoppel 74–5
equality and liberty, integration of 301 ethical deliberation
fairness 177 autonomy 80–1
fit 173, 300 examples of 74–81
foreground elements 160–1 lawyers enabling clients to do acts which
fraud, no one should benefit from own 161 are legal but not ethical 80–1
identification of norms 159 examples of 74–81
individual and society 163–4, 166–9 home-vehicle distinction, police powers of
individualism 164, 166–71, 174, 181, search and 75–7
299–303 hypothetical cases 76–80
interpretation 157, 166, 169–72, 175–8, illegitimate children, unmarried fathers and
299–304 78–9
judicial opinions 303 lawyers enabling clients to do acts which are
justice 160, 162, 166–8, 175–7, 181 legal but not ethical 80–1
justification 160 legal, examples of 74–81
law-making 162–3 non-discrimination 78–9
Law’s Empire 156, 160–1, 166, 175, 179 privacy 75–7
legislature 165, 169–70 promissory estoppel 74–5
legislative supremacy 165 reasons 73, 75–8
legitimacy 176–7, 181 search, police powers of 75–7
liberalism 300–2 stages of 71–2
moral theory 166–7, 172, 180, 299–300, democracy
304 Brown v Board of Education 45
no-eligible interpretation conflicts 301–4 conditions for 27–9
norms 161 constitutional democracy 20
phenomenology 172–3 constitutional judges 1–3
policies, principles distinguished from deliberative democracy 27, 28–30
158–60 elections 57–8, 293
positivism 159, 166 equal concern 53, 59, 62
post-interpretive claims 160 equality 26
pragmatism 156, 175–7, 181 judicial review 53
precedent 300, 302–3 justice 53–4
principles 158–61, 166–8, 177 majority rule 1, 2, 26–30, 53–4, 57–9, 61,
skeptics 155–7, 164–72, 177–80, 298–9 292–3
social goals 158 minorities, protection of 58
theories of legal background 155–7 pre-conditions for 28–9
utilitarianism 300–1 procedural liberties 30
culture 33–5 representation 57–9, 61
standards 4
dead hand of the past theory 19–22, 38 substantive liberties 30
decision theory 90 super-majoritarian amendment rules 20, 292
deference, stare decisis and descriptive facts
class, decisions deserving deference as a 107 Hartian positivism 265–6
fairness 112–13 law practices and the content of law 227–9,
individual decisions deserving 107 234–7
normal justification theory 107–8 normative facts 265–6
purpose of 103 rational-relation doctrine 265–6
deliberation relevance of 254
abstract characterization of 70–4 value facts 235–7, 240, 254
alcohol, prohibition on sale to males under determinacy
21 of 79 coherence 100
autonomy 80–1 legal propositions 230, 247
 Index 317
rational determinism 228–9 Brown v Board of Education 55
social facts 229 democracy 53, 59, 62
distinctive purpose of law 207–12, 215 Due Process Clause 52–3, 63
distinguishing cases 104, 109, 113, 116–18 Equal Protection Clause 7
doctrinal and sociological concept of law, Fourteenth Amendment 52–3
distinguishing between 310 gender equality 191–2
Due Process Clause individual rights 52
Dworkin, Ronald 13–15 integrity 199
equal concern 52–3, 63 liberty 63–6
interpretation 7, 13–15, 292 meaning 7
liberty 63–5 moral reading of US Constitution, Dworkin’s
moral reasoning 13–15, 21, 52–3, 63–5, 292 52–3, 55, 59, 61–6
positivism 13–14 parental powers 191–2
same sex sodomy 63–6, 292 political integrity 151
Dworkin, Ronald see also checkerboard solu- political obligation to obey the law 188,
tions, Ronald Dworkin and, Critical 199, 201–3
Legal Studies movement, Dworkin’s recognizable conception of 151
theory of integrity and, Law’s Empire, Equal Protection Clause
moral reading of US Constitution, deliberation 77–9
Dworkin’s, political integrity, Ronald Dworkin, Ronald 14–15
Dworkin and equal status and concern, meaning of 7
associative obligations 190–7, 304–5 historical reasoning 7
coherence 85, 95, 98–101, 133 interpretation 6–8, 14–15, 292
constitutional judges 1–4 moral reasoning 6–7, 14–15, 21, 38, 50–2,
constitutional theory 41–2, 48–56 63–4
Due Process Clause 13–15 non-discrimination 78–9
Equal Protection Clause 14–15 precedent 7–8
fidelity 23–5, 35–6 prejudice 14–15
Freedom’s Law 41 racially discriminatory laws, prevention of 6
historical reasoning 8 reasons 77–8
integrity 15–17, 104, 113–16 same sex sodomy 38, 50–2, 63–4
intervening cases 69, 85, 87–8, 95, 98–101 equality see also equal concern, Equal
majoritarian premise, US Constitution and Protection Clause
6, 19–21, 25–31, 292–3 checkerboard solutions 125–6, 139–41,
moral reading of US Constitution 26, 41, 144–8, 296–8
49–56 democracy 26
New York Review of Books, essays on constitu- differential treatment 125–6, 139–41,
tional theory in 36 144–8, 296–8
political obligation to obey the law 183–7, integrity, promoting reduction in 146–7
198–205 judicial review 63
precedent 6–8, 104, 113–16 justice 140–1, 144–8, 150–2, 296
super-majoritarian amendment rules 6, 19–21 justification 150–2
Taking Rights Seriously 41 liberty and, integration of 301
United States Constitution 15–17, 25–31, majoritarian premise 26, 28–30
41, 49–56, 292–3 moral reading of US Constitution, Dworkin’s
26, 38, 50–2, 63–4
efficiency, stare decisis and 108–12, 118 nepotism 146–7
elections 46, 57–8, 293 political integrity 144–8, 295–8
elitism 162–3 political obligation to obey the law 199
empiricism 11, 231 racial segregation 41, 43–5, 47
entrenchment United States Constitution 26, 28–30, 41–5
integrity 16 estoppel 74–5
stare decisis 110–12 ethics see morality
super-majoritarian amendment rules 20
United States Constitution 6, 16, 18, 20 facts make law 225–64, 310–11 see also
equal concern law-determining practices and con-
associative obligations 191–2, 305 tents of the law, relationship between
318 Index
facts make law (cont.) bent 249–52
abortion 249–52 better 254
aesthetic facts 228 coherence 248–9
antipositivitism 227–9 conceptual truth 254–5
beliefs 252 consensus 255
content of law, determinations of 225–37 legal propositions 248
descriptive facts 227–9, 234–7 legally correct 258–60
legal problem of content 241–5 morally best 259–60
metaphysical determination 226–8 relevant considerations 258–9, 263–4
non-legal content 241–5, 248 sentences, meaning of 245–6
rational determinism 228–9 specification of 246
reasons 233, 237–9 standards for 256
supervenience base 235 morality 259–60, 263–4
value facts 227–8, 235–6 non-legal content 241–5, 248
Critical Legal Studies movement 239 positivism 225–8, 256
descriptive facts 225–37 rational determinism 228–9
aesthetic facts 228 realism 239
content of law 227–9, 234–7 reasons 233, 237–40
executive, legislature or judicial behaviour, relevance limitation 262–3
facts about 236 relevant considerations 258–63
law practices 234–7, 252 sentences
reasons 240 legal content, effect of non-legal content
types of 236 on 242–3
value facts 235–7, 240 meaning of 242, 245–6
determinate legal requirements 237 non-legal content 242–3
Dworkin, Ronald 226–9, 226, 254, 262, words and 245–6
310–11 social facts 225–6
empirical facts 225, 231 standards 248–52
executive 236 statutes, constitutions, judicial and adminis-
foundational practices 247–8 trative decisions 230–1, 235, 237–47
Goodman’s problem about green and grue substantive factors 252–64
249 supervenience base 235
Hart, HLA 225–6, 239 text 243–4
higher-level facts 229 truth
indeterminate nature of law 239 conceptual 254–5
judicial behaviour, facts about 236 relevance of 254, 256
justification 226, 262 superficialist notions 254
Kripke’s problem about plus and quus 249 words, use of 254–5
law books 253 utterances and mental states 241–5
law-determining practices 236–45 value facts 225–8, 233, 235–7, 254–64,
descriptive facts 234–7, 252 310–11
justification 262 conceptual truth 254–5
law books 253 content of law 227–8, 235–6
legal requirements 237, 240 descriptive facts 235–7, 240, 254
reasons 237–40 exclusion 235–7
relevance 229 legal 256
substantive factors 252–64 models 254–5, 258
text 243–4 morality 259–62
types of 236–7 nature of 255–6
value facts, exclusion of 235–7 positivism 256
legal propositions 230, 246–8 relevance of 254
legal requirements 237, 240 role of 257–64
legislative history, relevance of 259 substantive 255–6
legislature 236, 253 truths 254–7
lower-level facts 229 values 252–4, 256–7
metaphysical determination 226–8 fairness
model 245–51 associative obligations 188–9
 Index 319
checkerboard solutions 120–5, 128–9, 133, acceptance of rules 278, 281–3
135–40, 144–8, 297–8 all legal systems 276–82, 288–90
coherence 133, 135 all or nothing 277–82, 288–90
deference 112–13 brute fact about law, appealing to 280–1
democracy 1, 2 content of law, legal facts relevant to
Hartian positivism and normative facts 288 278–80, 289
integrity 144–5, 151–3, 177, 292, 297–8 Hartian dispositions 280–1, 283–4, 289
justice 136–9 Hartian legal systems 277–82, 288
legislature 2 intuitions about nature of law 281–3, 288–9
majority rule 1, 2 law practices
non-comparative claims of 112 content of law, relevance to 278–80, 286
political integrity 144–8, 296–8 legal facts 276, 283–6
political obligation to obey the law 188, 198 legal facts
precedent 94, 111–12, 118 brute fact about law, appealing to 280–1
United States Constitution 292 law practices 276, 283–6
family relationships model of 277–9
associative obligations 190–4, 198, 304–5 non-legal facts, determined by 279–80
intrinsic value of 198 normative facts, appeals to 281–6, 288
federal statutes, state’s interpretation of 95–6 reasons 277
federalism 19 logical entailment 286
fidelity mistake 281–2, 288
Dworkin, Ronald 23–5, 35–6, 50, 282 model, determination of correct 277–9
improvement 25 281–3, 286, 288
integrity 25–6, 292 non-legal facts 279–80
moral reading of US Constitution 23–5, normative facts 281–6, 288–9
31–2, 33–4, 38, 50, 292 positivists and anti-positivists 284
originalism 23–5, 31, 33–4, 38 reasons 277
synthesis, as 24 recognition, rule of 277, 280, 282–4
United States Constitution 292 reflective understanding of law 280–4
fit relational-rational requirement 276–80,
Dworkin, Ronald 35–6 289
integrity 173, 300 Hartian dispositions 271–6
interpretation 36–7, 38 all-things-considered binding 274
moral reading of US Constitution 35–7, 38 attitudes and dispositions, constituted by
political obligation to obey the law 203–4 271–6, 280–3, 287, 289
Fourteenth Amendment 13–15, 52–3 bridge principles 280–1, 283–4, 289
Freedom’s Law 26, 60 identity, doctrine of 274
friendship law practices
associative obligations 190–7, 304–5 legal facts 275–6, 285
intrinsic value of 195–7, 304 normative facts 275–6
reasons 276
gender equality 191–2 legal facts
generalizations 81 acceptance of 273
government power, limits on 40–50, 52 law practices 275–6, 285
guiding principles of the United States reasons for 276
Constitution 50–6 legislation binding in courts 274
authority of 53 non-legal facts 273
Brown v Board of Education 55 normative facts 275–6
identification of 56 practice theory of rules 271
judges 56 precedent 289
personal preferences 50–1 Raz, Joseph 274
sources of 50–1, 54 reasons 276, 289
recognition, rule of 271–6
Hart, HLA see Hartian bridge principles, relational-rational requirement 271, 273
Hartian dispositions, Hartian posi- social rules 271–6
tivism and normative facts standards 272–5
Hartian bridge principles 276–85 truth, appeal to 274
320 Index
Hartian positivism and normative facts precedent 69–101
265–90 punitive/regulatory distinction 96–7
all or nothing result 266–7, 270 settled 71–2, 81, 83–4, 90–1, 94–7, 294–5
beliefs and facts 269 thought experiments, as 81–4
descriptive facts 265–6
fairness 288 identity, doctrine of 274
false propositions 269 individual rights 46, 48, 52
law-determining practices 267–8, 270–1, individualism 164, 166–71, 174, 181, 299–303
284–5 instrumental obligations 198
legal facts 265–7 integrity see also Critical Legal Studies
good or valuable 270 movement, Dworkin’s theory of
law-determining practices 268, 270–1, integrity and, political integrity,
284–5 Ronald Dworkin and
meaning of 267 adjudication, in 116–17
rational-relation requirement 265–6, associative obligations 188–9, 198–205
268–71, 288 checkerboard solutions 124–6, 136–8,
reasons 268, 270, 287 296–7
mind, radical interpretation theory of 287 coherence 88, 98
model of contribution of law practice to consistency 15–16
content of law 267 Dworkin, Ronald 15–17
parallel objections 284–9 entrenchment 16
physical events 286 equal concern 199
putative normative facts 270 fairness 292
rational intelligibility 286–7 fidelity 25–6, 292
rational-relation requirement 265–6, interpretation 16–17, 292
268–71, 286–9 irrelevant considerations 131–2
beliefs 269 justification 138
descriptive facts 265–6 laws aims in Law’s Empire 212
legal facts 265–6, 268–71, 288 legislation 116
non-legal facts 265 moral reading of US Constitution, Dworkin’s
normative facts 269–70,289 24, 35–6, 54, 292
putative normative facts 270 morally, striving to act 114–15
reasons 266, 268, 270, 287 political obligation to obey the law 186–9,
recognition, rule of 267–8 198–205
historical reasoning precedent 16–17, 104, 113–17, 292, 295
Dworkin, Ronald 5–9, 23–39, 49–51, United States Constitution 15–17, 24,
292–4 25–6, 54, 292
Equal Protection Clause 7 interpretation
internal role of 8 abstraction 37
moral reasoning 5–9, 23–39, 49–51, associative obligations 194–7
292–4 checkerboard solutions 126–8, 140–2
United States Constitution 6–8, 22 constitutional judges 42
history see also historical reasoning constructive 218–21, 308–9
fit 35–7, 38 creative 219, 309
moral reading of US Constitution, Dworkin’s Due Process Clause 7, 13–15, 292
32–8 Equal Protection Clause 6–8, 14–15, 292
originalism 32–5 favoured 142, 149–53
reclaiming 32–5 federal statutes, state’s interpretation of 95–6
stare decisis 104 fit 36–7, 38
human rights crimes 3 integrity 16–17, 119, 142–3, 149–53, 157,
hypothetical cases 166, 169–72, 175–8, 292, 299–304
actual settled cases 70, 91–2, 94 judges 4
answers to 84 justification 36–7, 38
coherence 70–2, 81–4, 86, 90–7, 294–5 laws aims in Law’s Empire 208, 211, 215–20,
deliberation 76–80 307–10
generalizations 81 majoritarian premise, US Constitution
mistake 97 and 30
 Index 321
mistakes 309 recognizable principles of 135–9, 148–53
moral reading of US Constitution, Dworkin’s single conception of 198–9
4–15, 21–39, 51–5, 291–3 justice, checkerboard solutions and 120–5,
no-eligible interpretation conflicts 301–4 135–9
norms 219 abortion 135–7
political obligation to obey the law 187–9, coherence 133–5
199–200, 203–4 compromise 135–7
United States Constitution 4–6, 16–17, 30, differential treatment 140–1, 296
292 Dworkin, Ronald 136–8
intervening cases, problem of fairness 136–9
coherence 84–101 integrity 136–8
mistake 86, 95, 99–101 irrelevant considerations 136
precedent 85–101 justification 138
prospective application 85–7, 93 minimising injustice 123
retroactivity and 69–70, 84–9, 92–5, 100–1 moral principles, recognizable 135–8
intuition about the nature of law 281–3, principles of, recognizable 135–9
288–9 rape, abortion only for victims of 136–7
irrelevant considerations religion 138–9
abortion 129–33, 135–6 justification
checkerboard solutions 129–33, 136 checkerboard solutions 125–6, 140–2
Dworkin, Ronald 129–31 coercion 217, 308
integrity 131–2 constructive interpretation 220
justice 136 deference 107–8
moral irrelevance 129–33 differential treatment 150–2
rape, abortion only for victims of 129–33, integrity 142–52, 160
135–6 interpretation 36–7, 38
laws aims in Law’s Empire 210–13, 306–8
judges see also constitutional judges moral 211–23, 306–8
integrity 303 moral reading of US Constitution, Dworkin’s
judicial process 2–3 36–7
opinions 303 normal justification theory 107–8
resources of 108–9 norms 218, 221
stare decisis 108–9 positivism 223, 308
judicial review stare decisis 104, 106–8, 113, 295
Brown v Board of Education 55
constitutional theory 44, 48 law-determining practices and contents of the
democracy 53 law, relationship between 226, 229–52
equality 63 abortion 249–52
individual rights 48 assumptions or beliefs 252
limits of 44 checkerboard solution 249–50
majoritarian premise, US Constitution and coherence 248–9
19, 29 conditions 241–5, 251
moral reading of US Constitution, Dworkin’s content-oriented considerations 261
53, 55, 63 descriptive facts, as 227–9, 234–7
super-majoritarian amendment rules 19 empirical facts 231
jurisdictions, cases from other 95–6 epistemic determinants 226, 228, 230, 240,
justice see also justice, checkerboard solutions 248–9, 251
and foundational practices 247–8
democracy 53–4 Hartian bridge principles 276, 278–80,
differential treatment 150–2 283–6, 289
integrity 120, 135–9, 142–53, 160, 162, Hartian dispositions 275–6, 285
166–8, 175–7, 181, 296–8 legal facts 226–8, 234–5, 268, 270–1,
justice-seeking 30–1 275–6, 284–5
majoritarian premise, US Constitution and mappings 232
30–1 meaning 230, 267
political obligation to obey the law 198–9, metaphysical-determination doctrine
202 226–8, 230–2, 234, 237–9, 251–3
322 Index
law-determining practices and contents of the South Africa 306
law, relationship between (cont.) unifying purpose of law 207–12, 215
models 245–51, 258 Law’s Empire see also law’s aims in Law’s
non-legal content 241–5 Empire
normative facts 275–6 associative obligations 196
partial determination of content of law 233 checkerboard solutions 119–42, 297
practice-oriented considerations 261 coherence 69, 87–8
rational determination 228–9, 231–3 integrity 119–20, 143, 156, 160–1, 166,
rational-relation doctrine 231–2, 237, 245 175, 179, 295, 297–8
reasons 233, 237–40, 276 irrelevant considerations 129
role of 230–4 political obligation to obey the law 183,
sentences 186–7, 201
meaning of 242–3, 245–6 Suggested Purpose 208–10, 212, 216
words and 245–6 legal facts
standards 248–52 acceptance 273
statutes, constitutions, judicial and adminis- brute fact about law, appealing to 280–1
trative decisions 230, 235, 237–47 good or valuable 270
transparency 232 Hartian bridge principles 277–86, 288
utterances and mental states 241–5 Hartian dispositions 273, 275–6, 285
values, beliefs about 252 Hartian positivism 265–71, 284–5, 287–8
value facts 227, 233–4, 252, 254–64 law practices 268, 270–1, 276, 283–6
law’s aims in Law’s Empire 207–23, 306–10 model of 277–9
agency, failure and success of 216 non-legal facts, determined by 279–80
claims, making 213–14 normative facts, appeals to 281–6, 288
coercion 208–9, 216–17, 308–10 rational-relation requirement 265–6,
constructive interpretation 220 268–71, 288
co-ordination of conduct for common good reasons 268, 270, 276, 277, 287
210 legal positivism see positivism
distinctive purpose of law 207–12, 215 Legal Process school 60
Dworkin’s Suggested Purpose 208–10, 212, legal propositions
216 determinacy 230, 247
guidance 210 facts make law 230, 246–8
Holmes, Oliver Wendell 215, 306 meaning 230
human beings making claims 214, 222, model 248
306–7 truth 230, 246–7
integrity 212 legal reasoning see also coherence, legal
interpretation 208, 211, 215–21, 307–10 reasoning and
justification 210–13, 217, 306–8 legislation
law officials, claims made by 213–14, binding on courts 274
307–8 checkerboard solutions 121–5
legal content, determination of 211 coherence 93, 147
mistakes 309–10 federal statutes, state’s interpretation of 95–6
moral ideal of law 212 history, relevance of 259
moral justification 211–23, 306–8 integrity 116
nature of law 212–13 law practices and the content of law 230,
Nazi law 217, 308 235, 237–47
norms 207, 209–10, 218–23, 307 political integrity 147
personification of law 212, 306–7 precedent 86
plain fact view of law 221 prospective changes 93
positivism 220, 222–3, 305–6, 308–9 Raz, Joseph 274
pragmatists 219 stare decisis 116
purpose, serving or having a 216 legislatures
Raz, Joseph 212–15, 217, 221–2, 306–7 constitutional theory 42, 43
sociological and doctrinal concept of law, democracy 43
distinguishing between 310 facts make law 253
society of angels thought-experiment 208–9, fairness 2
309–10 guardians of principle 30
 Index 323
integrity 165, 169–70 mistake
policy-making 46 actual settled cases 90, 95, 100–1
political obligation to obey coherence 86, 90, 97–101, 295
the law 187, 202 Hartian bridge principles 281–2, 288
supremacy 165 hypothetical cases 97
legitimacy 176–7, 181, 188 interpretation 309
liberty intervening cases, problem of 86, 95, 99–101
claims 42–4, 63–7 laws aims in Law’s Empire 309–10
demeaned existence of claimant, whether precedent 97–101, 295
state had 64 preferences 90
equal concern 63–6 stare decisis 103–7, 109, 115, 118
equality and, integration of 301 moral reading of US Constitution, Dworkin’s
fundamental liberties 64–5 5–12, 22, 49–56
majoritarian premise, US Constitution and abstraction 37, 51, 53
26, 30 administrative law 60
moral reading of US Constitution, Dworkin’s application of concepts 9–10
63–7 aspirational principles 38
police powers 43 Bill of Rights 52–3, 62
procedural liberties 30 Brown v Board of Education 55, 58–9, 294
racial segregation 43–4 coherence 292
states, liberty of 66 consequentialist pragmatism 61
substantive liberties 30 consistency 50
Supreme Court 42–3 constitutional law and 6, 7
tradition 64–5 context 11
critical attitude, encouraging a 38
majoritarian premise, US Constitution and culture 33–5
25, 26–31 dead hand of the past theory 38
administrative law 60 democracy 26, 53–4, 57–9, 61–2, 293
autonomy 29, 293 drafting 51
community 26 Due Process Clause 13–15, 21, 52–3, 63–6,
constitution-perfecting theory 29–30 292
constitutional constructivism 28–9 elections 57–8, 293
constitutional democracy 26–8, 30 empiricism 11
constitutional theory 47, 48 equal concern and respect, notion of 52–3,
counter-majoritarian difficulty 26 55, 59, 61–6
deliberative autonomy 29 Equal Protection Clause 6–7, 14–15, 21,
deliberative democracy 27, 28–30 63–6
democracy 26–30, 47–8, 53, 292–3 equality 26, 38, 50–2, 63–4
Dworkin, Ronald 25–31, 53, 58–60, 66–7, fidelity 23–5, 31–4, 38, 50, 292
291–3 fit 35–7, 38
equality 26, 28–30 Fourteenth Amendment 13–15, 52–3
fairness 1, 2 Freedom’s Law 26, 41, 60
forums of principle, courts as 30 government power, limits on 49–50, 52
guardians of principle, legislatures as 30 guiding principles 50–3
interpretation of Constitution 30 authority of 53
judicial review 29 Brown v Board of Education 55
justice-seeking 30–1 identification of 56
liberty 26, 30 judges 56
minorities, protection of 58–9 personal preferences 50–1
personal autonomy 30 sources of 50–1, 54
process-perfecting theory 29–30 Hart Ely, John. Democracy and Distrust
super-majoritarian amendment rules 19–21, 57–61
292–3 historical reasoning 5–9, 23–39, 49–51,
metaphysical-determination doctrine 226–8, 292–4
230–2, 234, 237–9, 251–3 history
mind, radical interpretation theory of 287 fit and 35–7, 38
minorities, protection of 19, 58–9 reclaiming 32–5
324 Index
moral reading of US Constitution, text 32–4, 49–50, 62
Dworkin’s (cont.) tradition 33, 64–5
improvement 25 utilitarianism 49, 60–1
individual rights 52 morality see also moral reading of US
integrity 24–6, 35–6, 54, 292 Constitution, Dworkin’s
intermediate theory 8–9, 25, 31–2, 37–8 associative obligations 194–5, 197
interpretation 4–15, 21–39, 51–5, 291–3 checkerboard solutions 121–2, 135–8
judges coercion 217
activism of 54, 59 coherence 74–82, 114–16
guiding principles, identification of 56 commitment to 114
job description 56–7 constitutional judges 1
minimalism 60 constructive interpretation 220
personal preferences 50–1 ethical deliberation 74–82
restraint 58 facts make law 259–60, 263–4
judicial review 53, 55, 63 ideal of law, moral 212
justice 53–4 integrity 114–15, 145, 166–7, 172, 180,
justification 36–7, 38 299–300, 304
Legal Process school 60 irrelevant considerations 129–33
liberty claims 63–7 justice 135–8
demeaned existence of claimant, whether justification 211–23, 306–8
state had 64 laws aims in Law’s Empire 212
fundamental liberties 64–5 model, morally best 259–60
states, liberty of 66 norms 218, 221
tradition 64–5 obedience to law 194–5
meaning of 9–10 political obligation to obey the law 184,
majoritarian premise 25, 26–31, 53, 58–60, 187–8, 200–3
66–7, 291–3 positivism 223, 308
minorities, protection of 58–9 recognizable moral principles 135–8
ordinary meaning of words 10–12 recognition, rule of 274
originalism 6–9, 12, 16, 23–6, 31–8 stare decisis 114–16
broad 24–5, 31–8, 293–4 striving to act morally 114–15
conservative ideology, as 34 value facts 259–62
culture 33–5
dead hand of the past theory 38 Nazi Law 212–13
fidelity 23–5, 31, 33–4, 38 New Deal 43–4
history, reclaiming 32–5 norms see also Hartian positivism and
narrow 24–5, 32 normative facts
revisionism 24, 31 associative obligations 305
semantic 294 authority, normative force of 105–7
text 32–4, 49–50 background 159, 161
philosophy 23–39 discrete 159
positivism 49 Hartian dispositions 275–6
practice 33 identification of 159
pragmatism 8, 61 integrity 161
precedent 7–8 interpretation 219
racial segregation 55 laws aims in Law’s Empire 207, 209–10,
representation 57–9, 61, 62 218–23, 307
revisionism 24, 31 law practices 275–6
same sex sodomy 63–6, 292 legal facts 281–6, 288
state determination 66–7 legal reasoning 82
states, liberty of 66 moral justification 218, 221
structure 32–4 political obligation to obey the law 185,
substantive theory of the US Constitution 201–2
26–31 principles 161
super-majoritarian amendment rules 292–3 putative normative facts 270
Supreme Court, Dworkin’s influence on 61 rational-relation doctrine 269–70, 289
Taking Rights Seriously 60 stare decisis 105–7
 Index 325
obedience to law see also political obligation to legislation, coherence of 147
obey the law more than one principle of integrity 149
associative obligations 183–9, 191, 194–5, moral solutions 145
305 nepotism in employment 146–7
moral obligation of 194–5 principles 148, 150
obligations see also associative obligations reduction in integrity 146
instrumental 198 political obligation to obey the law
moral 194–5 associative obligations 183–9, 191, 193–5,
political 183–9, 191, 193–5, 198–205, 305 198–205, 305
special 189–97, 305 coercion 188, 203
ordinary meaning of words 10–12 conditions 183, 185–6, 188
originalism cooperation 183–4
broad 24–5, 31–8, 293–4 Dworkin, Robert 183–7, 198–205
conservative ideology, as 34 equal concern 199, 201–3
culture 33–5 equality 199
dead hand of the past theory 38 fairness 188, 198
Dworkin, Ronald 6–9, 12, 16, 23–6, 31–8 fit 203–4
fidelity 23–5, 31, 33–4, 38 force of law 185
history, reclaiming 32–5 general obligations to obey law 184–5, 188,
moral reading of US Constitution 6–9, 12, 202–4
16, 23–6, 31–8, 49–50, 293–4 grounds of law 185
narrow 24–5, 32 instrumental considerations 203–5
revisionism 24, 31 integrity 186–9, 198–205
semantic 294 interpretation 187–9, 199–200, 203–4
text 32–4, 49–50 intrinsic value 201–2
overruling cases 104–5, 107, 111, 113, 116–18 justice 198, 202
single conception of 198–9
parental powers 191–2 legislature 187, 202
parliaments see legislatures legitimacy 188
personification of law 212, 306–7 morality 184, 187–8, 200–3
phenomenology 172–3 norms 185, 201–2
philosophers, constitutional judges as 5–22 scepticism 187
plain fact view of law 221 self-legislation 201
police powers 43 state, value of individual’s
political integrity, Ronald Dworkin and relationship with 200
119–53 unjust laws, obedience to 185
benefits of integrity 143 populism 43, 46–7
checkerboard solutions 119–53, 296–8 positivism see also Hartian positivism and
coherence 143–8, 150, 292 normative facts
compromise 142–3, 145, 152 anti-positivism 284
differential treatment 144–8, 150–2, Due Process Clause 13–14
296–8 facts make law 225–8
equal concern, recognizable conception hard 226, 256
of 151 Hartian bridge principles 284
equal treatment 144–7, 295–7 integrity 159, 166
explicit rules 143–4 laws aims in Law’s Empire 220, 222–3,
fairness 144–5, 151–3, 292, 297–8 305–6, 308–9
favoured interpretation 142, 149–53 moral justification 223, 308
fidelity 292 moral reading of US Constitution,
interpretation 119, 142–3, 149–53, 292 Dworkin’s 49
justice 120 social facts 225–6
integrity trumping 142–8, 296–8 soft 226, 256
principles of, recognizable 135–9, value facts 226, 256
148–53 practical theory of rules 271
justification for integrity 142–52 pragmatism
Law’s Empire 119–20, 143, 295, 297–8 consequentialism 8, 61
legality 120 integrity 156, 175–7, 181
326 Index
pragmatism (cont.) rational-relation doctrine
laws aims in Law’s Empire 219 beliefs 269
moral reading of US Constitution, descriptive facts 265–6
Dworkin’s 8, 61 Hartian bridge principles 276–80, 289
policy making 8 Hartian dispositions 271, 273
super-majoritarian amendment rules 19–21 Hartian positivism and normative facts
precedent see also stare decisis 265–6, 268–71, 286–9
coherence 205 law practices and the content of law 231–2,
discretion 88 237, 245
Dworkin, Ronald 6–8 legal facts 265–6, 268–71, 288
Equal Protection Clause 7–8 non-legal facts 265
fairness 94 normative facts 269–70, 289
Hartian dispositions 289 putative normative facts 270
hypothetical cases 69–101 reasons
integrity 16–17, 300, 302–3 constitutional judges 2–3
intervening cases, problem of 85–101 deliberation 73, 75–6
jurisdictions, cases from other 95–6 Equal Protection Clause 77–8
legal reasoning 69–101 facts make law 237–40
legislation 86 Hartian bridge principles 277
mistake 97–101, 295 Hartian dispositions 274, 276, 289
moral reading of US Constitution, Hartian positivism and normative facts 266,
Dworkin’s 7–8 268, 270, 287
prospective effect 85–7, 93–4, 97–8 law practices and the content of law 233,
retroactivity 94–101, 295 237–40, 276
settled cases 83–94 legal facts 268, 270, 287
state courts 95–6 judicial process 2–3
United States Constitution 16–17 legal facts 276
preferences stare decisis 106–7
checkerboard solutions 123–4 recognition, rule of
coherence 90 all-things-considered binding 274
judges, personal preferences of 50–1 Hartian bridge principles 277, 280, 282–4
mistake 90 Hartian dispositions 271–6
moral reading of US Constitution, Hartian positivism and normative facts 267–8
Dworkin’s 50–1 legal authority of 274
settled 90 morally binding 274
principles practice theory or rules 272
checkerboard solutions 122–3, 135–9 standards 272–5
coherence 133, 135 reflective equilibrium 69
compromise 122–3 reflective understanding of law 280–4
conflict between 122 reliance, certainty and 111
integrity 158–60 religion 138–9
justice 135–9, 148–53 retroactivity
moral, recognizable 135–9 coherence 70, 74, 84–9, 92–101, 295
norms 161 intervening cases, problem of 69–70, 84–9,
policy distinguished from 158–60 92–5, 100–1
ranking 122 precedent 94–101, 295
privacy 75–7 revisionary concept of law 92–4
process-perfecting theory 29–30 revisionism 24, 31
promissory estoppel 74–5 rule of recognition see recognition, rule of
punitive/regulatory distinction 96–7
scientific theory 91–2, 94
racial segregation 41, 43–5, 47, 55 search, police powers of 75–7
racially discriminatory laws, segregation 41, 43–5, 47, 55
prevention of 6 sentences
rape, abortion only for victims of facts make law 242–3, 245–6
checkerboard solutions 129–33, 135–7, law practices and the content of law 242–3,
140–1 245–6
justice 136–7 meaning 242–3, 245–6
 Index 327
model 245–6 reliance, certainty and 111
non-legal content 242–3 research, costs of 109
words 245–6 Student Judiciary of the University of
social change 42–3 Georgia 116–17
social goals 158 vertical 103, 108
social practices, changes in 93 statutes see legislation
social rules 271–6 super-majoritarian amendment rules 6, 17–21
social welfare 122 constitutional democracy 20
sociological and doctrinal concept of law, dead hand theories of constitutionalism 19,
distinguishing between 310 20–1, 22
South Africa 306 democracy 20, 292
special obligations 189–97, 305 Dworkin, Ronald 6, 19–21
standards entrenchment 20
constitutional judges 1–2 federalism 19
democracy 4 judicial review 19
facts make law 248–52 localism 19
Hartian dispositions 272–5 minorities, protection of 19
law practices and the content of law 248–52 moral reading 292–3
model 256 pragmatism 19–21
recognition, rule of 272–5 purpose of 19–20
stare decisis 103–18 reform, incentives for improving quality of
adjudication, integrity in 116–17 20
certainty, value of 111 short-termism 20
coherence 114, 115–16, 118 stable political institutions, establishment of
consistency 115 19–20
deference US Constitution
class, decisions deserving deference as a 107 Supreme Court
fairness 112–13 Dworkin’s influence on 61
individual decisions deserving 107 liberty claims, approach to 42–3
normal justification theory 107–8 moral reading of US Constitution 61
purpose of 103 police powers 43
distinguishing 104, 109, 113, 116–18 populism 43
Dworkin, integrity and 104, 113–16 racial segregation 43–4
efficiency 108–12, 118
entrenchment 110–12 Taking Rights Seriously 60 text
fairness 111–12, 118 Bill of Rights 62
non-comparative claims of 112 moral reading of US Constitution, Dworkin’s
higher court judges, expertise of 108 32–4, 49–50, 62
history 104 originalism 32–4, 49–50
horizontal 103–18 tradition 33, 64–5
illegitimacy 105 truth
integrity 104, 113–17, 292, 295 conceptual 254–5
judicial resources 108–9 facts make law 254–6
justification 104, 106–8, 113, 295 Hartian dispositions 274
legislation, integrity in 116 legal propositions 230, 246–7
liability rules 109–10 model 254–5
merits of decisions 104–7, 112–13, 118 relevance of 254, 256
mistakes 103–7, 109, 115, 118 superficialist notions 254
morality value facts 254–7
coherence 114–16 words, use of 254–5
commitment to 114
morally, striving to act 114–15 underdetermination 100–1
no difference theory 105–6 unifying purpose of law 207–12, 215
normal justification theory 107–8, 113 United States Constitution see also Due
normative force of authority 105–7 Process Clause, Equal Protection
overruling 104–5, 107, 111, 113, 116–18 Clause, majoritarian premise, US
planning, certainty and 111 Constitution and, moral reading
reasons 106–7 of US Constitution, Dworkin’s
328 Index
United States Constitution (cont.) certainty 111
amendments 6, 17–18 family relationships 198
change, prevention of 6, 18 friendships 195–7, 304
consistency 15–16 political obligation to obey the law 201–2
democracy, standards of 4 positivism 226
Dworkin, Ronald 15–17 state, value of individual’s relationship with
entrenchment 6, 16, 18 200
equality 41–5 substantive factors 253
fairness 292 value facts 225–8, 233, 254–64, 310–11
fidelity 282 see also value, values
Fourteenth Amendment 13–15, 52–3 conceptual truth 254–5
guiding principles of the United States content of law 227–8, 235–6
Constitution 50–6 descriptive facts 235–7, 240, 254
historical reasoning 5–8, 22 exclusion 235–7
integrity 15–17 legal 256
interpretation 4, 5–6, 16–17, 292 models 254–5, 258
judges’ approach to interpretation 4 morality 259–62
precedent 16–17 nature of 255–6
purposes of 17–19, 22 positivism 256
rigidity of, justification for 18 relevance of 254
social change 42–3 role of 257–64
substantive theory of 26–31 substantive 255–6
super-majoritarian amendment rules 19–21, truths 254–7
292–3 values see also value, value facts
unjust laws, obedience to 185 beliefs 252
utilitarianism 49, 60–1, 300–1 facts make law 252–3, 256–7
utterances and mental states 241–5 law practices and the content of law 252
substantive factors 253
value see also value facts, values
associative obligations 195–8, 304–5 words 10–12, 254–5