The Legacy of John Rawls

THOM BROOKS, FABIAN FREYENHAGEN

Continuum

THE LEGACY OF JOHN RAWLS

Continuum Studies in American Philosophy:

Dorothy G. Rogers, America's First Women PhilosophersThorn Brooks and Fabian Freyenhagen (eds), The Legacy of John RawlsJames A. Marcum, Thomas Kuhn's RevolutionJoshua Rust, John Searle and the Construction of Social RealityEve Gaudet, Quine on MeaningDouglas McDermid, The Varieties of PragmatismTimothy Mosteller, Relativism in Contemporary American Philosophy

THE LEGACY OF JOHN RAWLS

THOM BROOKS AND FABIAN FREYENHAGEN

continuum

Edited by

Continuum

The Tower Building, 11 York Road, London SE1 7NX

80 Maiden Lane, Suite 704, New York, NY 10038

www.continuumbooks.com

©Thorn Brooks, Fabian Freyenhagen and contributors, 2005

All rights reserved. No part of this publication may be reproduced or transmitted in

any form or by any means, electronic or mechanical, including photocopying, recording,

or any information storage or retrieval system, without prior permission in writing from

the publishers.

First published 2005

Paperback edition 2007

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.

ISBN-10: HB: 0-8264-7843-3

PB:0-8264-9987-2

ISBN-13: HB: 978-0-8264-7843-6

PB: 978-0-8264-9987-5

Library of Congress Cataloging-in-Publication Data

The legacy of John Rawls / edited by Thorn Brooks and Fabian Freyenhagen.

p. cm.

Includes bibliographical references and indexes.

ISBN 0-8264-7843-3

1. Rawls, John, 1921-. 2. Justice. 3. Liberalism. 4. Political ethics. 5. Political science-Philosophy.

I. Brooks, Thorn. II. Freyenhagen, Fabian.

JC251.L395 2005

320'.01-dc22

2005041911

Typeset by Aarontype Limited, Easton, Bristol

Printed and bound in Great Britain by Biddies Ltd, King's Lynn, Norfolk

CONTENTS

Preface

Motes on contributors viii

Abbreviations x

Introduction 1Thorn Brooks and Fabian Freyenhagen

1 The unity of Rawls's work 22LeifWenar

2 Self-realization and the priority of fair equality of opportunity 34RobertS. Taylor

3 Taking the distinction between persons seriously 50Anthony Simon Laden

4 Rawls and feminism: What should feminists make of 67liberal neutrality?Elizabeth Brake

5 Public reason and the moral foundation of liberalism 85Jon Mahoney

6 Dilemmas of public reason: pluralism, polarization, 107and instabilityRobert Talisse

1 Public reason and religion 124James Boettcher

8 John Rawls and the new Kantian moral theory 152

Ana Marta Gonzalez

vii

vi Contents

9 The Law of Peoples: the old and the new 177Chris Naticchia

10 The legacies of John Rawls 195Fred D'Agostino

Name index 213

Subject index 217

PREFACE

John Rawls was perhaps the most important moral and political philosopherof the last century. In November 2004, the Journal of Moral Philosophy pub-lished a special issue on "The Legacy of John Rawls" as a tribute to Rawls'sever-present influence on moral and political thought. Several of the papers inthis collection appeared in that issue. However, many others did not. All of thepapers included here were anonymously refereed and we would like to extendour sincere thanks to the members of the Journal of Moral Philosophy's editorialboard and our many referees for their advice—and their time—in decidingwhich submissions should be published.

In addition to our editorial board and referees, we would like to acknowl-edge our thanks to others as well. First, we must thank Philip de Bary for histireless and enthusiastic support of this project from the beginning. We aregrateful for the help of others at Continuum, such as Sarah Norman and IainBeswick, for their assistance in putting the original journal issue together aswell. Mention must also be made of Hywel Evans for his support of this pro-ject. We are also very appreciative of Tracey Brady and Laura Jarvis's helpwith preparation of the manuscript.

We must thank the Department of Philosophy at the University of Sheffieldfor their encouragement and backing on every matter of the Journal's activities,not least this book. In particular, we extend our very special thanks to LeifWenar. His advice on all aspects of this project has been simply invaluable.

Finally, to our partners, Meagan and Sabine, our most heartfelt apprecia-tion for their support over the course of this project. As always, we remain verymuch in their debt.

Thom Brooks

Fabian Freyenhagen

Sheffield, UK

November 2004

NOTES ON CONTRIBUTORS

James Boettcher is Assistant Professor of Philosophy at Saint Joseph'sUniversity in Philadelphia. He specializes in social and political philosophy,and has published several articles on the political liberalism of John Rawls.His current project involves an analysis of the role of religion in politicaldecision-making.

Elizabeth Brake is Assistant Professor of Philosophy at the University of Cal-gary, Canada. Her research interests include political philosophy, feministethics, Kant, and Hegel. Besides working on liberalism and feminism, she iscurrently writing on two topics: marriage and paternity.

Thom Brooks is Lecturer in Political Thought at the University of Newcastle,UK, and founding editor of the Journal of Moral Philosophy. He is editor of Rous-seau and Law. His work centres on British and German Idealism, democratictheory, and legal philosophy. He is currently writing a book on punishment.

Fred D'Agostino is Associate Professor of Philosophy and Head of the Schoolof Social Science at the University of New England in New South Wales, Aus-tralia. He is author of Free Public Reason (Oxford University Press, 1996) andThe Common Denominator (Ashgate, 2003), and has written numerous articlesand chapters on political theory and epistemology. He is currently workingon bringing materials in social psychology and organization theory to bearon issues in social epistemology.

Fabian Freyenhagen is currently completing his Ph.D. on Kant's and Ador-no's practical philosophy at the University of Sheffield, UK. He is the reviewseditor of the Journal of Moral Philosophy. He also works on Kantian practicalphilosophy more generally.

Ana Marta Gonzalez received her Ph.D. in Philosophy at the University ofNavarra, Spain, in 1997 writing on Aquinas's approach to the foundationsof morality. More recently, she became interested in Kant's moral philosophyand won a one-year Fulbright Fellowship at Harvard to do some research.She now teaches ethics at the University of Navarra.

Notes on contributors ix

Anthony Simon Laden is Associate Professor of Philosophy at the University

of Illinois at Chicago. He is the author of Reasonably Radical: Deliberative Lib-

eralism and the Politics of Identity (Cornell, 2001) as well as numerous articles onRawls, democratic legitimacy, and reasonable deliberation.

Jon Mahoney's primary research interests are in social and political philoso-

phy and ethics. He is currently Assistant Professor of Philosophy at Kansasstate University. Recent publications include "Objectivity, Interpretation,and Rights: A Critique of Dworkin" (Law and Philosophy 23 [2004]) and"Cosmopolitanism as a Moral Imperative" (Philosophy in the Contemporary

World 2.5 [2002]).

Chris Naticchia is Associate Professor of Philosophy at California state Uni-versity, San Bernardino, where he specializes in moral, political, and legal phi-

losophy. His recent work includes "Recognizing states and Governments"(Canadian Journal of Philosophy, forthcoming), which discusses the standardsthat states ought to use to recognize other states and governments as membersof the international community.

Robert Talisse is Assistant Professor of Philosophy at Vanderbilt Univer-sity. His research focuses on democratic theory, public deliberation, and poli-tical justification. He is the author of several articles in political philosophy,

and the hook Democracy After Liberalism (Routledge, 2005).

Robert S. Taylor is Assistant Professor of Political Science at the Universityof California, Davis. He specializes in contemporary Anglo-American politi-

cal philosophy and has recently published articles on Rawls's defense of thepriority of liberty (Philosophy & Public Affairs, 2003) and on self-ownership(Journal of Political Philosophy, 2004).

Leif Wenar received his doctorate in philosophy from Harvard in 1997.Among his scholarly articles on Rawls is "Political Liberalism: An Internal

Critique," which appeared in Ethics and The Philosopher's Annual. He is cur-rently Reader in Philosophy at the University of Sheffield, and for 2004—5holds a Fellowship in Justice and the World Economy from the CarnegieCouncil on Ethics and International Affairs.

ABBREVIATIONS

The following are abbreviations for the works of John Rawls as they arereferred to in the notes to chapters.

CP John Rawls, Collected Papers, ed. Samuel Freeman (Cambridge,MA: Harvard University Press, 1999).

JF John Rawls, Justice as Fairness: A Restatement, ed. Erin Kelly (Gam-bridge, MA: Harvard University Press, 2001).

LHMP John Rawls, Lectures on the History of Moral Philosophy, ed. BarbaraHerman (Cambridge, MA: Harvard University Press, 2000).

LP John Rawls, The Law of Peoples (Cambridge, MA: Harvard Univer-sity Press, 1999).

PL John Rawls, Political Liberalism (New York: Columbia UniversityPress, paperback edition, 1996).

PRR John Rawls, "The Idea of Public Reason Revisited" [1997] re-printed in CP, 573-616 and in LP, 129-80.

RH John Rawls, "Reply to Habermas" [1995] reprinted in PL, 372-434.

TJ John Rawls, A Theory of Justice (Cambridge, MA: Harvard Univer-sity Press, revised edition, 1999).

Thorn Brooks and Fabian Freyenhagen

John Rawls (1921—2002) is widely recognized has having rejuvenated andrevolutionized moral and political theory. As one of his fiercest critics,Robert Nozick, put it once:

Political philosophers now must either work within Rawls' theory orexplain why not. The considerations and distinctions we have developedare illuminated by, and help illuminate, Rawls' masterful presentation ofan alternative conception. Even those who remain unconvinced after wres-tling with Rawls' systematic vision will learn much from closely studying it.I do not speak only of the Millian sharpening of one's views in combating(what one takes to be) error. It is impossible to read Rawls' book withoutincorporating much, perhaps transmuted, into one's own deepened view.And it is impossible to finish his book [A Theory of Justice (eds)] without anew and inspiring vision of what a moral theory may attempt to do andunite; of how beautiful a whole theory can be.

That Rawls's theory is the focal point of moral and political philosophy, thatone has to either work within in it or explain why not, is something which is inall likelihood still true today, 30 years after the passage above was written. It isfor this reason that we take stock of the immense variety of Rawls's legacy inthis volume.

In this introduction, we want to do two things. First, we have a brief lookback at the work of Rawls, how it proved so influential, how it was criticized,and how it changed, partly as a result of these criticisms. The approach here ismore developmental than systematic, since the first chapter of this volumealready provides an interpretation of how Rawls's work can be seen as present-ing a coherent system. Second, we provide a glance ahead at the possible waysRawls's work will, at least in our opinion, continue to shape the discussion inmoral, political, and legal philosophy. Here our main aim is to make explicitthe breadth this continued influence is likely to take, with some of the chaptersin this volume pointing to more specific ways in which Rawls's legacy is set tolive on.

INTRODUCTION

The above-quoted passage taken from Nozick's Anarchy, State, and Utopia pro-vides a clue as to what made Rawls's theory as influential as it proved to be.Nozick speaks of Rawls's work providing a systematic theory, which illumi-nates distinctions and considerations people have drawn on since, forcingthem to incorporate much from it while presenting an "inspiring vision ofwhat a moral theory may attempt to do and unite." If one adds to this theability to make itself immune to traditional and new objections4 —an abilitywhich partly relies on integrating some of the elements of the objections—then one is, it seems to us, a good way along the road of understanding thealmost magnetic effect the "beautiful theory" of John Rawls had and con-tinues to have.

This is not to say that commentators agree that Rawls's work does present asystematic and consistent theory. Some of his critics and supporters alike chargehim with having moved against his original commitments in his adoption ofpolitical liberalism. In Chapter 1, Leif Wenar argues against these critics.He presents a systematic interpretation of Rawls's work according to which it isa unified whole structured around the ideas of justice and legitimacy. He alsohighlights the parallels in Rawls's theory ofjustice for domestic and global insti-tutions. Apart from providing an implicit defense against the charge of inco-herence, this chapter offers an excellent overview of the essential elementsof Rawls's theory. Consequently, Chapter 1 presents the ideal starting pointboth for reading this volume and for engagement with Rawls generally.

Justice as fairness: a comprehensive conception of justice

Looking at what Rawls's theory attempts to do, the first thing to note is whathe did not attempt to do. His conception of justice as fairness restricts itself tothe justice of social institutions, or what he calls "the basic structure ofsociety." So while one might call individuals or their actions and judgementsjust and unjust, Rawls did not set out to give an account of this. He concen-trated his efforts on the basic structure of society, because—as he famouslywrites—"its effects are so profound and present from the start." Initially,this restriction included a focus on the question of justice within states andnot between them, but later Rawls extended his account to the global basicstructure.

A quick glance at the index (or size, for that matter) of A Theory of Justicereveals that Rawls, while restricting himself to a theory of justice of the basicstructure of society, treats this subject extensively and in admirable depth.He sets out to find the principles of justice which should govern this structure,

2 Introduction

Taking stock

4

5

Introduction 3

provides a method for finding these principles, uses this method to comparedifferent alternatives, considers the question of civil obedience and our rightto civil disobedience, offers a theory of the good, engages in moral psychology,and enters into the relation between ideal theory and practical application.

All this was initially part of the project of defending a comprehensive doc-trine of liberalism which provided a contractarian theory of justice. Thetheory was contractarian in that the principles of justice were to be chosen ina way that represented what a general agreement would look like between freecitizens who are placed in reasonable conditions and who acknowledge havinga sense of justice. Rawls modeled these reasonable conditions in terms of theso-called original position. In the original position, people would be repre-sented by parties who are rational and mutually disinterested while behind aveil of ignorance (meaning that they only have access to the uncontroversialfindings of the sciences and common sense, but not to knowledge about thesocial position, sex, or conception of the good of those they are representing).The original position is supplemented and kept in check by the notion of a"reflective equilibrium." This notion is introduced to make sure that theprinciples selected in the original position match up with our considered jud-gements (and so can find actual agreement among those who are subject tothese principles). This two-fold methodology is used in order to examine dif-ferent candidates for the principles of justice for the basic structure. In parti-cular, Rawls argued that the following two principles of justice will come outon top in comparison with other candidates:

(1) Each person has the same indefeasible claim to a fully adequate schemeof equal basic liberties, which scheme is compatible with the samescheme of liberties for all;

(2) Social and economic inequalities are to satisfy two conditions:(a) they are to be attached to offices and positions open to all under

conditions of fair equality of opportunity; and(b) they are to be to the greatest benefit of the least advantaged mem-

bers of society (the difference principle).

The adoption of these principles would include giving the first principle lexicalpriority over the second and the first part of the second principle over thesecond part. A society that would be governed by these principles is calledby Rawls a "well-ordered society." In such a society, people accept andknow everyone else to accept the same (political) conception of justice. More-over, the basic structure of society is publicly known (or with good reasonsbelieved) to satisfy those principles. Finally, citizens have an effective senseof justice, i.e. they understand the principles of justice and (for the most part)act accordingly. Rawls's theory was not only a renewal and modification of

13

10

4 Introduction

classical liberalism, but also of Kantian morality, facing the then dominant

utilitarian theories as its main adversary. Even the two-fold procedure intro-

duced above finds its model to some extent in the way Kant proceeds in the

Groundwork of the Metaphysics of Morals. The influence of Kant, however, is

most visible in the conception of the original position itself, which was origin-

ally seen by Rawls as an interpretation of Kant's conception of morality.

It represents the idea of a rational agent choosing principles of justice indepen-

dently from the contingencies of nature and society and while being aware of

having principle-dependent desires (such as a sense of justice) that will need

realization as much as other desires. Equally, the idea of the primacy of liberty

(and the separateness of persons on which it relies) owes much to a Kantian

outlook, according to which no one can be used as a mere means to the welfare

of others. And, as it is also brought out in two chapters in this collection, Rawlstook over the idea of the primacy of the practical from Kant, too. Later,

some of the explicit Kantian influences had to be given up by Rawls in thewake of his shift to political liberalism. However, Rawls's modification of

Kant's conception of morality began already before this shift.

In Chapter 8, Ana Marta Gonzalez explores the way Rawls's theory mod-

ified Kantian morality, and influenced—especially through his Lectures on the

History of Moral Philosophy—a whole generation of moral theorists (such as Bar-

bara Herman, Christine Korsgaard, Thomas Hill, Jr.). Rawls, Gonzalez

points out, aimed to make Kant's moral philosophy more plausible, partly by

putting more emphasis on Kant's ethical writings other than the Groundwork

and partly by bringing Kant down to earth, linking his moral theory more to

current culture and re-interpreting the Kantian dualisms. Chapter 8 offers an

illuminating summary of what Kant did for Rawls and what Rawls did for the

Kantians. As mentioned, the main adversary for Rawls was utilitarianism,

with intuitionism also playing a role, albeit a less central one. One reason

for this might be that utilitarianism with its teleological structure of maximiz-

ing expected utility across people looks attractive for regulating the basic

structure of society, e.g. because it offers one matrix with which to weigh up

priorities. In fact, Rawls himself emphasized the suitability of utilitarianism

for justifying the institution of punishment in an early paper. Nonetheless,

Rawls rejected utilitarianism as providing the principles of justice, because—

as he famously put it—"Utilitarianism does not take seriously the distinction

between persons."

What it means to take this distinction seriously and how this disqualifies

utilitarianism—as well as providing responses to other objections to Rawls—

is explored in Chapter 3. Anthony Simon Laden argues here that the distinc-

tion between persons operates in Rawls's theory on four levels: the content

of the principles of justice, the form the argument takes, the conception of

(political) morality operative in the argument, and the conception of political

15

16

17

20

18

Introduction 5

philosophy all this implies. The reader should turn to Chapter 3 to see howRawls's theory is able to absorb and deal with a host of criticisms by its veryway of doing political philosophy.

And criticisms there have been plenty. Ever since Rawls's system came tofull prominence with the 1971 publication of A Theory of Justice there hasbeen an extensive and often fierce debate of his theory in all its differentaspects. After Theory Rawls was initially engaged in answering objections, inparticular in connection with distributive justice. The latter had become afocal point of the debate of his conception of justice, not least because ofNozick's detailed objections to it in his alternative account of how societyshould be governed, his 1974 Anarchy, State, and Utopia.

In this volume, Chapter 2 looks at a largely neglected aspect of Rawls's con-ception of distributive justice, namely, the priority of the principle of fairequality of opportunity over the difference principle. Robert Taylor recon-structs Rawls's argument for this priority, relying on the importance of theideas of self-realization and virtuosity. He also looks at the implications ofthis priority for public policy. This chapter represents a much-needed addi-tion to the literature on Rawls's two principles.

However, Rawls not only had to defend himself on questions of dis-tributive justice and the content of the principles of justice he proposed.He came into the firing line in respect to his general approach and assump-tions as well.

Feminists criticized Rawls's project for not paying sufficient attention togender inequality, or even blocking the focus on this issue, by endorsing liberalneutrality as well as by being committed to the public/private dichotomy.Some of these feminist critics, most notably Susan Moller Okin24, thoughtthat the Rawlsian framework could be adopted to make room for their con-cerns. And Rawls did indeed follow Okin's suggestion of viewing the familyas a social institution (and thereby regulated by the two principles of justice)and excluding the knowledge of one's sex from the items known in the originalposition. Other feminists have remained more critical of Rawls's project,partly because of more general worries about his conception of the personwhich they share with communitarians.

In Chapter 4, Elizabeth Brake continues Okin's project in suggesting thatfeminists have something to gain from working within and extending Rawls'sproject. She argues that the idea of fair equality of opportunity is not the onlyresource to be used to combat gender inequality. Rawls's endorsement of stateneutrality is, according to Brake, also beneficial for this cause and not—ascommentators like Catherine MacKinnon have maintained27 —an obstacle.The chapter by Brake is also of interest for a further reason. She offers Rawlsa defense of a qualified account of neutrality at the justificatory level, addres-sing a problem left unsolved in Political Liberalism.

21

23

24

26

7

25

6 Introduction

The debate between communitarianism and liberalism, which occupied theprofession for a significant amount of time, had also been partly sparked bythe publication of Theory.28 So-called communitarians objected to Rawls'saccount, among other reasons, for relying on a Kantian and liberal concep-tion of morality and persons. Such a conception is seen as objectionablebecause it involved a controversial conception of the good, which in itsformal and abstract character was too far removed from the ways of life andconceptions of the good which are integral to the identity of the people endors-ing them.

From Theory to Political Liberalism

In the early to mid-1980s, at the same time as the communitarian criticismsreached their peak, Rawls began to shift his own position. Commentators dis-agree as to whether this was in response to these criticisms or due to theattempt of making his own account more internally consistent. After a briefperiod of arguing for "Kantian constructivism," Rawls adopted what hecalled "political liberalism" (in contrast to the view of liberalism as a compre-hensive doctrine). He, thus, restricted the role of justice for the second time.Apart from focusing on the basic structure of society only (and not on indivi-dual actions and judgements), Rawls now suggested limiting the basis fromwhich the principles of justice could be drawn. He no longer saw himself asputting forward a comprehensive liberal doctrine with its Kantian conceptionof the person. Rather, Rawls now proposed to start with the conception of citi-zens which can be found in the public culture and traditions of modern demo-cratic societies. From this conception, according to which citizens are free andequal, one can move to a family of conceptions of justice (of which Rawls's twoprinciples is one example). One of these conceptions can then be used to laydown the constitutional essentials for such societies. Crucial in this shift was theacceptance of what Rawls calls the "fact of pluralism," i.e. the fact that inmodern societies people hold very different comprehensive conceptions of thegood which are—with a very few exceptions—all reasonable conceptions.This fact implied for Rawls that no one comprehensive conception of the goodcould govern the basic structure of society. Since there is reasonably disagree-ment about which particular conception of the good should be adopted, peoplecould object to a regime built on one conception alone. In order to sustain itselfsuch a regime would, thus, have to rely on oppressive use of state power—some-thing which goes against the very grain of liberalism. However, and this wasanother of Rawls's crucial (as well as controversial) claims, an agreement onhow to govern the basic structure is nonetheless possible. There is sufficient

8

33

Introduction 7

overlap between the different reasonable comprehensive conceptions of thegood (namely, the conception of citizens as free and equal) to form such anagreement. Thus, the agreement on constitutional essentials can arise out ofwhat Rawls calls an "overlapping consensus." For this to happen, the par-ties would have to rely on arguments which were independent of their particu-lar conception of the good (i.e. on public reasons). This does not exclude thatthey have conception-immanent reasons for accepting the agreement, too.In fact, Rawls hopes that they would, since this would add to the stability of awell-ordered society and thus make his vision more feasible.

There has been much debate about this shift to political liberalism, which isnot only limited to the already mentioned question about consistency withRawls's earlier work. One of the major issues has been whether it is desirableand possible to have a political conception of justice only. Jon Mahoney takesup this question of the moral foundation of political liberalism in Chapter 5.He argues that—contrary to what Rawls thinks—political liberalism cannotand should not be thought of as freestanding. Rather, it must claim to expressthe philosophical truth about practical reason and persons if it wants to avoidbeing either dogmatic or indistinguishable from a modus Vivendi.

Chapters 6 and 7 concern the question whether the idea of public reason istoo restrictive. Robert Talisse argues that public reason is unstable. Insistingon only reasonable comprehensive doctrines being admissible and limitingpolitical agreement to what can be argued for in public reasons would lead,according to Talisse, to the polarization of those excluded and thereby toinstability. James Boettcher, on the other hand, tries to show how Rawls'sidea of public reason is less restrictive than critics allege. Focusing particularlyon religion and public reason, his discussion aims to demonstrate that the con-straints in Rawls's idea of public reason are neither unfair, nor unfeasible, nordo they result in adverse social consequences. These two chapters offer—fromtwo opposed sides—an interesting perspective on the very topical issue of howwe can justify political decisions and what conceptions of the good can beexcluded in so doing.

The themes of pluralism and finding an agreement about the principlesgoverning the basic structure under the condition of reasonable pluralism arefurther explored in Chapter 10. Fred D'Agostino brings out how the back-ground of evaluative pluralism made Rawls search for a method of commen-suration of different proposals for constitutional essentials. He argues that theoriginal position is a form of commensuration as separation and that Rawls'smethod is ultimately pragmatist in nature. D'Agostino suggests that the back-ground of evaluative pluralism, the object of agreement on constitutionalessentials under this background condition and the pragmatist method ofachieving this represent the three important legacies of Rawls.

36

8 Introduction

Extending political liberalism outward: global justice

Apart from the discussions sparked by Rawls's move to political liberalism andthe issues which surround this, much of the recent debates have focused on theextension of Rawls's theory of justice to the international realm. In fact, theinterest in Rawls and global justice arose before Rawls himself consideredthis issue in detail. Some philosophers, most notably Charles Beitz, arguedthat principles of justice should not be conceived merely as intra-state—asRawls had conceived them at that time—but that principles of justice shouldbe conceived as being global. For Beitz and others, it was not obviouslytrue that individual states were the only units of sufficient cooperation andthat "the justice of the law of nations and of relations between states" wasbeyond the remit of the two principles of justice as Rawls had claimed it is.Beitz argued that significant cooperation existed at the global level as well.Hence, for him the only difference between Rawlsian justice at the domesticand global levels is that "[t]he principles of justice for international politicswould be the two principles for domestic society writ large." To choosethese principle for global justice we should, according to Beitz, adopt an"international original position."

The idea of using the original position to arrive at principles of justice for theinternational realm was something that Rawls himself had anticipated inTheory.41 Consequently, when he did finally turn to extending political liberal-ism to the global sphere, this was partly just due to him wanting to complete histheory. However, that Rawls felt the need to provide his oumaccountofthis exten-sion might have also resulted from his disagreement with the extensions offeredon his behalf by Beitz and Thomas Pogge. In contrast to these commentators,Rawls thinks that peoples (rather than individuals) should be represented in theinternational original position and that international redistribution should nottake the form of a global difference principle, but be limited to a duty of assis-tance. With the publication of Rawls's views on global justice, criticismturned from what Rawls had failed to develop to how he thought it should bedeveloped. In particular, the focus has been on the two issues just mentioned,namely, who should be represented in the international original position andwhat form international redistribution should take. Also, concerns havebeen raised about whether Rawls is too permissive in admitting non-liberal,but decent societies into the group of peoples being represented in the interna-tional original position. It is the latter issue which is taken up in Chapter 9,along with the attempt to fill in a surprising gap in the commentary on Rawls'stheory, namely a discussion of the two different versions of "Law of Peoples."Although being ultimately critical of both versions, Chris Naticchia identifiesimportant differences between them. This makes this chapter an importantresource for the engagement with Rawls on this highly salient issue.

38

40

s

43

41

Introduction 9

Looking ahead

So far in this introduction, we have sought to give an indication of the devel-opment of Rawls's thought and the debates surrounding it, while highlightinghow the topics covered in this volume fit into this story. It might be useful toalso look ahead and offer a suggestion as to what areas connected to Rawls'slegacy are, in our opinion, likely to remain live issues. Many of the most sig-nificant legacies John Rawls's work has left us are discussed in the essays com-prising this volume. However, while we cover the vast majority of relevantareas, no single volume could capture them all. Consequently, in the followingwe also highlight some of the additional areas where Rawls's work has madea contribution.

Completing the system

From within a Rawlsian perspective, there are two particular areas whichrequire further attention. Both our relations with nature and our duties tothose liberal citizens who cannot fully cooperate are issues where Rawls'ssystem is still in need of extension.

As for the former area, some work has already been done on combininga Rawlsian picture with environmental ethics and animal rights issues.We expect that this will continue and possibly intensify.

Equally, there has been a heightened interest in the second area, our dutiesto those who cannot fully cooperate. This can, for example, be seen from therelated, albeit wider issues in medical and bioethics which have been taken upby an increasing number of commentators who apply Rawls's framework orwork out its implications.

The ground in this area was broken by Norman Daniels. Since the late1970s, Daniels has been critically applying Rawls's theory of justice to consid-erations of distributive justice and health care. Daniels argues that healthcare should be regulated by principles of justice which ensure fair equality ofopportunity. He sets out not only to show how principles of justice are applic-able to bioethics, but also to demonstrate various policy outcomes we shouldadopt as well.

Nor is Daniels the only one engaged in these kinds of issues today. Oneincreasing area of concern is genetic engineering. While a relatively new area,some, such as David Resnik, have begun to apply Rawlsian principles to howwe should consider these questions. In this case, Resnik argues that geneticinequalities are acceptable in principle only if they are to everyone's advantageand do not interfere with the fair equality of opportunity open to all.

Genetic engineering is not the only specific area of bioethics where somehave sought to apply Rawlsian principles. Some employ Rawls's "original

47

48

49

50

10 Introduction

position" to answer questions about whether or not we should use placebos inclinical trials as their use may withhold treatment from patients. Othershave used Rawls's theory of justice more broadly to argue positions, such asthe fairness of patent systems and access to medicine.

Medicine, and the biological sciences related to it, continually presentsus with exciting new possibilities and ethical considerations. Rawls's viewson justice have been of increasing interest to scholars working on bio-ethics. We believe that this area will be a continuing area of great influence ofRawls's ideas.

Matters of debate

As we have seen already, Rawls's theory sparked a number of debates and con-troversies. Amongst these debates are some which have quieted down in recentyears and some which are still in full swing.

Debates which seem to have reached their height are, for example, thefamous liberal-communitarian debate and the topic of multiculturalism,which partly arose out of the former. Neither of them is directly discussed inthis volume. The reason for this is not that we believe either debate is conclu-sively decided one way or another. Our view is simply that debates in theseareas have quieted dramatically in the past few years, in contradistinction toother areas such as global justice or public reason.

Another example of a debate which though not settled has lost someof its salience is distributive justice. Here, our volume contains one essaywhich, as mentioned, takes up a neglected aspect of this debate, namely, thepriority of the principle of fair equality over the difference principle.Equally, while debates in feminism might have moved on slightly from discus-sion of Rawls's theory, Chapter 4 shows that this theory still offers interestingresources to feminism.

As for those issues which are still very much alive and will remain so for sometime, our volume covers the most relevant of them. Take the example of globaljustice. This is one area where Rawls's legacy, especially regarding the rela-tionship between his views and global justice, is still very much a matter ofdebate. A lot of new work has been published on this topic, most especiallythe latest collection of essays The Ethics of Assistance edited by Deen Chatter-jee. We believe that for the foreseeable future most interest in Rawls willrelate to considerations of global justice. This collection appropriately beginswith Leif Wenar's argument for why we should see consistency across Rawls'smajor works, from Theory through to The Law of Peoples. We end with ChrisNaticchia's examination of Rawls's different versions of The Law of Peoples.We do not want to suggest for a moment that global justice is the beginning

51

52

53

Introduction 11

and end of Rawls's legacy. However, this is the area that has generated themost interest in recent years, and this is reflected in this volume.

Similarly, the other issues related to Rawls's work which are still a matter ofintense debate are also covered by our volume. Thus, the question of the con-sistency of Rawls's work is addressed in Chapter 1, the debate about the rela-tion of Rawls's conception of justice with moral foundations is taken up inChapter 5, and Chapters 6 and 7 discuss Rawls's conception of public reason.

However, Rawls has been influential not only in terms of the content of moral

and political philosophy, but also in the way these disciplines proceed. Andperhaps his most lasting influence is in this respect as Chapters 3 and 10 high-light, among other things. Moreover, Rawls's influence on the way moral andpolitical philosophy is carried out extends further than just the way contem-

porary theorists approach current debates. He also changed our way of doinghistory of philosophy.

Rawls and the history of philosophy

It is this area where Rawls's legacy is being perhaps only slowly recognized.He held a sincere interest in the history of philosophy, especially in, but not

restricted to, the history of ethics. The study of intellectual history is a projectof great importance that can often further our understanding of more contem-

porary problems. Yet, for Rawls, our interest in important texts is not just lim-ited to what they can contribute to contemporary problems. In fact, studyingthese texts should be more than an attempt to show how each faces insurmoun-table problems in light of any philosophical advances we might have made:

We don't study them in the hope of finding some philosophical argu-ment, some analytic idea that will be directly useful for our present-dayphilosophical questions in the way they arise for us. No, we study Hume,Leibniz, and Kant because they express deep and distinctive philosophicaldoctrines.

This approach to the study of history of philosophy has been attractive tomany of his students, as is, for example, witnessed by a collection of essays writ-ten by his students from over the years. Given the popular appeal of much

else of his views, we imagine Rawls's approach here will become increasinglyattractive to philosophers with the recent publication of his Lectures on the His-

tory of Moral Philosophy.

It has been well known that Rawls held a particular interest in ImmanuelKant. Indeed, we have already mentioned Ana Marta Gonzalez's chapter inthis collection, entitled "John Rawls and the new Kantian moral theory,"

56

58

12 Introduction

which sets out to explain how Rawls's reading of Kant has been an importantinfluence on the work of Kantian scholars.

As the students' collection mentioned above makes clear, Rawls held aninterest in a number of philosophers other than Kant (the collection wasentitled Reclaiming the History of Ethics). One great merit of Rawls's recentlypublished Lectures is that it gives us at long last his mature views of Hume,Leibniz, and Kant. The Lectures also discuss Hegel. Many before had noticedimportant similarities between the views of Rawls and Hegel, such as StephenHoulgate and Sibyl Schwarzenbach. However, these scholars lacked muchdirect evidence: Rawls rarely mentions Hegel. Instead, scholars have had totry to draw strong inferences from Rawls's and Hegel's related ideas. Withthe publication of his Lectures, Rawls's particular interest in Hegel's conceptof "Ethical Life [Sittlichkeif]" is explicit and the importance of Hegel's workfor Rawls far more clear. Barbara Herman even claims that "the Hegel lec-tures sketch the bridge between Kantian moral thought and the liberalism ofRawls's own work."

We imagine that philosophers will continue to be interested in both Rawls'scontributions to the history of philosophy—and the contributions of impor-tant figures in the history of philosophy to Rawls's work.

Much of the attention has been on more obvious relationships betweenRawls and Western philosophers. Unfortunately, there have been far too fewattempts to relate Rawls's theory of justice as fairness with the Eastern philo-sophical tradition. We can only hope that future scholars will pursue thisroute too.

Reaching into the real world: Rawls and law

While Rawls's legacy will live on within academia, both in terms of theon-going debates that his work has sparked and the framework he has pro-vided, it is also of interest to see how his thought might exert influencebeyond the academic realm. Such influence was surely Rawls's hope, notmerely because every philosopher might wish for his or her theories to havean effect on the world, but also because of his concern with presenting what isa workable and feasible Utopia. True, Rawls's work contains a high level ofabstractions and theoretical complexity, but this he deemed necessary toarrive at what the nature of justice is under favorable conditions (whichmight not coincide with actual conditions) and was, hence, not apologeticabout it.63

One example where Rawls's work has reached into the actual working ofsocial institutions is law, with legal theorists and many judges showing espe-cially great interest in it. For example, the Fordham Law Review recently pub-lished a special symposium on Rawls and law, in which Ronald Dworkin says:

61

62

59

Introduction 13

Politicians around the world cite his ideas and American and other judgesappeal to his work, so we might talk about the impact that has already hadon the law in different countries. Or we might consider the impact that hemight have: We might ask what changes in American tax or tort law hisfamous difference principle would recommend.

Indeed, Rawls's work has the great benefit of being clearly applicable to law,in terms of supplying both useful principles of justice and a framework for con-sideringjustice.

Thus far, Rawls is very rarely mentioned in the decisions of high courts.Instead, he has been more popular with lower appellate courts. As Dworkinhints above, many judges have been interested in the idea of applying Rawl-sian ideas to help decide cases and issues of law. One example is the case Goetz v.Crossnan where the court argues:

Whatever ratio one selects, procedural protections must be fashioned togive some reasonable assurance that the pattern of erroneous releases anderroneous confinements will approximate the ratio that one finds tolerable.And in setting the ratio and formulating the appropriate procedural protec-tions to achieve it, we would do well to consider the matter from behindJohn Rawls's veil of ignorance: we should select a pattern we would findtolerable not knowing whether we would be one of those erroneously con-fined.68

There are many other examples we could also cite that take a similar position:Rawls's theory of justice is applicable to helping decide cases and issues of law.

Not all courts agree with this view, however. For example, in Martin v.Duggerthe court argues:

In resolving issues of this type, an economist is perhaps better situated thana federal judge. An economist could handle questions like this in the won-derful world of perfect competition. The court is not that lucky. Of course, ifthe court was a [sic] fortunate, it could situate itself behind a "veil of ignor-ance" ... Behind this veil, the court would "use the notion of pure proce-dural justice as a basis of theory" .. . The court would have no conceptionof the good, and would not understand its role within society ... The Rawls[sic] original position is an attractive situation. Unfortunately, this positiondoes not include an individual like Nollie Lee Martin, who stabbed PatriciaGreenfield only after his attempts to strangle her to death failed. Accord-ingly, the court will reach this issue fully cognizant of the fact that thiseighth amendment right arises only after a human like Martin has been con-victed of murder.

65

66

69

14 Introduction

Here the difference of opinion between Goetz and Martin is that the latterdecision expresses worries about the applicability of Rawls's ideas. Theargument in Martin for why the court thinks it is unable to decide the caseas if they sat behind a veil of ignorance seems to be as follows. The court isconstrained by the U.S. Constitution (in particular, Martin's protectionagainst cruel and unusual punishments guaranteed by the Eighth Amend-ment), and to decide as if they were in Rawls's original position would beas if they have a right to settle constitutional essentials for which this devicewas thought up. However, it is not and should not be the role of courtsto decide what the constitutional essentials are, but to judge in accordancewith them.

It is, however, arguable that the court misses the substance of Rawls's claim.One can grant that it is true as a point of law that the court in Martin is cer-tainly obliged to follow constitutional provisions and relevant judicial prece-dents. But this need not entail that Rawls's veil of ignorance has no currencyfor judges. In fact, it might be the other way around, i.e. because the originalposition is to be used—at least, on Rawls's picture—to establish the constitu-tional essentials, the courts might need to have recourse to it in order to estab-lish in what ways it has to uphold these essentials. For example, what are"cruel and unusual punishments" which are forbidden by the Eighth Amend-ment? Here, the use of the original position and veil of ignorance is far fromirrelevant, since it seems to provide a test for determining what kinds of pun-ishments are "cruel and unusual." Similarly, constitutional provisions andjudicial precedents are not set forever in stone: they may change over time.Consequently, it is important to have a procedure by which changes andamendments to laws and the Constitution could be tested to determinewhether they should be accepted or whether they are invalid given the consti-tutional tradition. Once more, it would seem that the original position withits veil of ignorance would be the relevant procedure, because it tells uswhether persons would agree to any constitutional protections, why theywould agree to them, and what would serve as these protections.

There are many further fruitful applications of Rawls's views on law. It istrue that his work has been far more popular with lawyers than judges: despitetheir being relatively few court decisions that mention him, his name appearsin over 4,200 documents on Westlaw.

Much of the attention here relates to the U.S. Supreme Court cases Vacco v.Quill—where the Court held that state bans against assisted suicide do not vio-late the Equal Protection Clause of the U.S. Constitution's FourteenthAmendment—and Washington v. Glucksberg—where the Court held that statebans against assisted suicide do not violate the Due Process Clause of the Four-teenth Amendment. Why the attention on Rawls's philosophy with respectto these two decisions relating to assisted suicide? This is due to the now

Introduction 15

famous "Philosopher's Brief" he wrote along with Ronald Dworkin, ThomasNagel, Robert Nozick, T.M. Scanlon, and Judith Jarvis Thompson.

One of the major issues raised by the Philosopher's Brief was the question ofto what extent should the Court take seriously the considered concerns of someof the most important moral and political philosophers in America. Somelegal theorists, such as Neomi Rao in an influential article published in theUniversity of Chicago Law Review, argued that it was right that the U.S. SupremeCourt decision took no notice of the Philosopher's Brief. Rao argues that theCourt only tends to use philosophers to "support" controversial decisionswhen the Court is attempting to justify certain political decisions beyond, orin contravention to, existing precedent. This view is challenged by others whoargue that there is nothing worrisome about using the arguments of philoso-phers to support legal decisions. Rao claims there is ample evidence to thinkthat judges have employed philosophers in order to justify decisions that con-stitutional provisions and relevant judicial precedents do not support, such asin the case Roe v. Wade. However, the use of philosophers—in this casePlato—only appears in a broad historical overview of Western civilization'sresponses to abortion and not as part of the distinct justification of the particu-lar decision. There is nothing illegitimate in itself in the use of philosophersby the courts in judicial decisions.

Moreover, one could argue that the U.S. Supreme Court was incorrect toignore the Philosopher's Brief, even if one disagreed about the relevance ofphilosophy to court decisions. Rawls et al. cite twenty cases, two statutoryprovisions, two journals, and one book on jurisprudence. They do not citephilosophers or their own works.

Other philosophers, such as Catherine MacKinnon, may well have beenmore successful than Rawls in helping frame and draft legislation in theUnited states and Canada. However, as this discussion has endeavored toshow, there is no good reason to think philosophers lack something of impor-tance to contribute to the practice and development of law in America andelsewhere. Furthermore, there is every reason to think that more work will bedone in the future on how Rawlsian principles of justice and important aspectsof it, such as the original position, the veil of ignorance, and the differenceprinciple, should help shape it.

Conclusion

We started this introduction with a quote from Robert Nozick according towhich post-Rawls moral and political philosophy must work within his frame-work or explain why it does not do so. We believe that this is still the case andwill continue to be the case for some time to come. This volume is meant to

73

76

16 Introduction

help those engaged in the field to understand better what Rawl's theory con-sists in and where the live issues in it lie. In this way it is hoping to pay tribute toa wonderful person, a great philosopher, and his beautiful theory.

Notes

1. R. Nozick, Anarchy, State, and Utopia (New York: Basic Books, 1974), 183.2. See Ch. 1 (LeifWenar, "The unity of Rawls's work").3. Nozick, Anarchy, State, and Utopia, 183.

4. In a 1998 interview Rawls described his aims in Theory as setting out "a certainclassical theory of justice—the theory of the social contract—so as to make itimmune to various traditional objections" (CP, 617; cf TJ, xviii/original edn,viii;CP, 614).

5. There are other factors which explain this, too. For example, C. Kukathas andP. Pettit point out that pre-Rawls twentieth-century political theory was charac-terized by a sharp divide between studies of what is desirable and what is feasible;Rawls is credited with putting them back together (see their Rawls—A Theory ofJustice and its Critics (Cambridge, UK: Polity Press, 1990), Ch. 1). Another impor-tant factor in explaining the influence of Rawls is not merely what he attempted todo, but how he attempted to do it. On this see Ch. 3, especially the fourth sectionand Ch. 10, the second section.

6. TJ, 7/original edn, 7. Rawls later developed his defense on the basic structure asprimary subject of justice. See e.g. PL, Lecture VII: "The Basic Structure as Sub-ject;" cf. JF, §§15f.

7. See his "The Law of the Peoples" (reprinted in CP, Ch. 24) and the revised LP.For a brief summary of Rawls's theory of global justice see Ch. 1; for a detaileddiscussion see Ch. 9 (Chris Naticchia, "The Law of the peoples: the old andthe new").

8. TJ, Ch. 3; PL, Lecture I, §4; JF, §6.9. TJ, 18f, 42-5/original edn, 20f, 48-51; JF, §10.

10. TJ, Ch. 3; JF, Part III. For a brief overview of the argument see Ch. 1.11. JF, 42f. We quote here the most considered version of the two principles. For a

comparison with the earlier versions see Rawls's "The Basic Liberties and TheirPriority" in PL, Lecture VIII and the summary thereof in JF, 42—5.

12. On the lexical priority of the first principle, see TJ, §§11, 39, 46, 82; PL, LectureV; JF, §30. On the arguments for this second priority see TJ, §§14, 46; JF, 43,163n44; and also Ch. 2 (Robert S. Taylor, "Self-realization and the priority offair equality of opportunity").

13. TJ, §69; PL, Lecture I, §6; JF, §3.14. TJ, xif, xviif./original edn, viif.15. Kant said he wanted first to work up "from common cognition to the develop-

ment of its supreme principle," and then down from "this principle and its sourcesback to the common cognition in which we find it used" (I. Kant; Practical Philo-sophy, translated by Mary J. Gregor (Cambridge, UK: Cambridge University

Introduction 17

Press, 1996), p. 47). The move up from common consciousness to the moral lawparallels the notion of reaching a reflective equilibrium; the move down is cap-tured well in Rawls by the original position and the four-stage sequence of adopt-ing and applying the principles of justice into legislation and practice (on thissequence see TJ, §31).

16. TJ, §40.17. See Chs 8 (Ana Marta Gonzalez, "John Rawls and the new Kantian moral

theory") and 10 (D'Agostino, "The legacies of John Rawls").18. On the latter see e.g. TJ, §7.19. See his 1955 paper "Two Concepts of Rules," reprinted in CP, Ch. 2.20. TJ, 24/original edn, 27.21. Rawls's conception of justice and his approach to moral and political philosophy

were foreshadowed in a series of earlier papers from the early 1950s onwards (seeCP, Chs 1-10).

22. See CP, Chs 11, 12.23. See note 1 for details.24. See e.g. her Justice, Gender, and the Family (New York: Basic Books, 1989).25. JF, §§4, 6, 50.26. See e.g. Seyla Benhabib, "The Generalized and the Concrete Other: The Kohl-

berg-Gilligan Controversy and Feminist Theory," in E. F. Kittay (ed.), Womenand Moral Theory (Totowa: Rowman and Littlefield, 1987), 154-77.

27. See e.g. Catherine MacKinnon, Towards a Feminist Theory of the State (Cambridge,MA: Harvard University Press, 1989).

28. For an excellent overview of this debate see Stephen Mulhall and Adam Swift,Liberals and Communitarians (Oxford, UK: Basil Blackwell, 1992).

29. For example, communitarians also rejected the idea of neutrality of the state,because it deprived the political sphere of the possibility of realizing communalgoods and identities that allegedly are its lifeblood. This objection might bebased on a misreading of Rawls to some extent (see e.g. Kukathas and Pettit,"Rawls," 11—16). After all, Rawls views—as early as Theory (TJ, §79)—politicaljustice as a good and the well-ordered society as "social union of unions" (see alsoJF,§60).

30. See e.g. Michael J. Sandel, Liberalism and the Limits of Justice (Cambridge, UK:Cambridge University Press, 1982).

31. Rawls explicitly denies any direct influence of communitarian criticisms onhis decision to abandon comprehensive for political liberalism (see PL (paperbackedn only), xix n6). However, the influence could have been more indirect.Perhaps, the external criticisms drove home the point to Rawls that modernsociety is pluralistic in a way that would make unstable a society organized onone comprehensive doctrine alone.

32. CP, Chs 16, 17. Kantian constructivism represents more a shift in emphasis thanin substance. In his 1980 Dewey Lectures, Rawls brings the Kantian conceptionof free and equal persons as reasonable and rational more to the forefront of histheory. He more openly connects this conception with the set-up of the originalposition, which is consequently seen as validating moral concepts without relying

18 Introduction

on claims about their reality (and in this sense "constructing" them). For a recentdiscussion of Rawls's constructivism (and a comparison with constructivism inKant) see O. O'Neill, "Constructivism in Rawls and Kant," in S. Freeman(ed.), The Cambridge Companion to Rawls (Cambridge, UK: Cambridge UniversityPress, 2003), 347-67.

33. PL, 36f.; JF, §11. That there is the possibility for their being more than one rea-sonable conception of the good has to do, according to Rawls, with the burdens ofjudgement, which e.g. include that the evidence might be conflicting and thatpeople might disagree over how to weigh it (PL, Lecture II, §2; JF, §11.4).

34. See PL, 37; JF, §11.3.35. PL, Lecture IV; JF, §§11, 58; CP, Ch. 20.36. Rawls even allowed in a later revision of his conception of public reason (PRR,

reprinted in CP, Ch. 26) that people can bring in these conception-immanentreasons in public debate, as long as they connect them at some stage withpublic reasons (see "Introduction to the Paperback Edition," in PL (paperbackedition only), xxxvii—1xii (1—1vii); JF, 96).

37. First, in 1993, in an article ("Law of Peoples," reprinted in CP, Ch. 24) and thenin an extended and revised version of this article published as a book (LP). On acomparison and critique of the versions see Ch. 10 below.

38. TJ, 7/original edn, 7f.39. Charles R. Beitz, "Justice and International Relations," Philosophy and Public

Affairs 4 (1975), 360-89 (363).40. Ibid., 366.41. He writes that "one may extend the interpretation of the original position and

think of the parties as representatives of different nations who must choosetogether the fundamental principles to adjudicate conflicting claims amongstates" (TJ, 331/original edn, 378).

42. For the latter, see Thomas Pogge, Realizing Rawls (Ithaca, NY: Cornell Univer-sity Press, 1989), Part 3.

43. Rawls contrasts and discusses both Pogge and Beitz in the book version of his essayon global justice (LP, 82f, 113-20).

44. See Simon Caney, "CosmopolitanJustice and Equalizing Opportunities," Meta-philosophy 32 (2001), 113-34; Andrew Juper, "Rawlsian Global Justice: Beyond'The Law of Peoples' to a Cosmopolitan Law of Persons," Political Theory 28(2000), 640-74; Thomas W. Pogge, "An Egalitarian Law of Peoples," Philosophyand Public Affairs 23 (1994), 195-224); and Pogge, "Rawls on International Jus-tice," Philosophical Quarterly 51 (2001), 246-53.

45. See Leif Wenar, "Contractualism and Global Economic Justice," Metaphilosophy32 (2001), 79-94.

46. See Ch. 10 on this.47. On Rawls and environmental ethics see e.g. Brent A. Singer, "An Extension of

Rawls" Theory of Justice to Environmental Ethics," Environmental Ethics 10(1988), 217—32; and the literature review by Daniel P. Thero, "Rawls and Envir-onmental Ethics: A Critical Examination of the Literature," Environmental Ethics17 (1995), 93-105. On the issue of animal rights and Rawls see e.g. Mark

Introduction 19

Bernstein, "Contractualism and Animals," Philosophical Studies 86 (1997), 49-72;Peter Carruthers, The Animals Issue: Moral Theory in Practice (Cambridge, UK:Cambridge University Press, 1992); Daniel A. Dombrowski, "Rawls and Ani-mals," International Journal of Applied Philosophy 12 (1998), 63-77; Mark Rowlands,"Contractarianism and Animal Rights," Journal of Applied Philosophy 14 (1997),235-47.

48. See Donald C. Ainslie, "Bioethics and the Problem of Pluralism," Social Philosophyand Policy 19 (2002), 1-28; Derek R. Bell, "Rawls and Research on CognitivelyImpaired Patients: A Reply to Maio," Theoretical Medicine and Bioethics 24(2003), 381-93; Giovanni Maio, "The Relevance of Rawls" Principle of justicefor Research on Cognitively Impaired Patients," Theoretical Medicine and Bioethics23 (2002), 45-53; Maio, "Research Ethics and the Principle of Justice as Fair-ness—A Restatement," Theoretical Medicine and Bioethics 24 (2003), 395-406;Jonathan D. Moreno, "Consensus, Contracts, and Committees," Journal of Med-icine and Philosophy (1991), 393-408; and Robert M. Veatch, "Egalitarian andMaximin Theories of Justice: Directed Donation of Organs for Transplant,"Journal of Medicine and Philosophy 23 (1998), 456-76.

49. See especially his Just Health Care (Cambridge, UK: Cambridge University Press,1985).

50. See David B. Resnik, "Genetic Engineering and Social Justice: A RawlsianApproach," Social Theory and Practice 23 (1997), 427-48. See also JonathanBrown, "Genetic Manipulation in Humans as a Matter of Rawlsian Justice,"Social Theory and Practice 27 (2001), 83-110 and Colin Farrelly, "Genes andSocial Justice: A Rawlsian Reply to Moore," Bioethics 16 (2002), 72-83.

51. See Stephen Senn, "Ethical Considerations Concerning Treatment Allocationin Drug Development Trials," Statistical Methods in Medical Research 11 (2002),403-11.

52. See David B. Resnik, "Fair Drug Prices and the Patent System," Health Care Ana-lysis 12 (2004), 91-115.

53. See Ch. 2.54. See Simon Caney, "Cosmopolitanism and the Law of Peoples," Journal of Political

Philosophy 10 (2002), 95-123; Eric Cavallero, "Popular Sovereignty and the Lawof Peoples," Legal Theory 9 (2003), 181-200; Pierre Laberge, "HumanitarianIntervention: Three Ethical Positions," Ethics and International Affairs 9 (1995),15—35; Chris Naticchia, "Human Rights, Liberalism, and Rawls's Law of Peo-ples," Social Theory and Practice 24 (1998), 345-74; David A. Reidy, "Rawls onInternational Justice—a Defense," Political Theory 32 (2004), 291-319; Kok-Chor Tan, "Liberal Toleration in Rawls's Law of Peoples," Ethics 108 (1998),276-95; Fernando R. Teson, "The Rawlsian Theory of International Law,"Ethics and International Affairs 9 (1995), 79-99; and Leif Wenar, "The Legitimacyof Peoples," in P. de Greiff (ed.), Global Justice and Transnational Politics (Cam-bridge, MA: MIT, 2002), 53-76.

55. TheEthics of Assistance: Morality and the Distant Needy (Cambridge, UK: CambridgeUniversity Press, 2004). See particularly the contributions by Charles Beitz,Thomas Pogge, Erin Kelly, Martha Nussbaum, and Peter Singer (Chs 2,8-10,13).

20 Introduction

56. LHMP, 329. Cf. Martha C. Nussbaum, "Conversing with the Tradition: John

Rawls and the History of Ethics," Ethics 109.2 (1999), 424-30 (425).

57. Andrews Reath, Barbara Herman, and Christine M. Korsgaard (eds), Reclaiming

the History of Ethics: Essays for John Rawls (Cambridge, UK: Cambridge UniversityPress, 1997).

58. A further volume of lectures, on the history of political philosophy, is forth-

coming.

59. See Peter Benson, "Rawls, Hegel, and Personhood: A Reply to Sibyl Schwarzen-

bach," Political Theory 22 (1994), 491-500; Stephen Houlgate, "Hegel, Rawls,and the Rational state," in Robert R. Williams (ed.), Beyond Liberalism and Com-munitarianism: Studies in Hegel's Philosophy of Right (Albany: SUNY, 2001), 249-73;

Sibyl A. Schwarzenbach, "Rawls, Hegel, and Communitarianism," Political

Theory (1991), 539—71; and Schwarzenbach, "A Rejoinder to Peter Benson,"

Political Theory 22 (1994), 501-7.

60. LHMP, 327-71.61. Barbara Herman, "Editor's Foreword," in LHMP, xi—xix (xv).

62. See Chung-ying Cheng, "Critical Reflections on Rawlsian Justice versus Confu-cian Justice," Journal of Chinese Philosophy 24.4 (1997), 417-26 and Ruiping Fan,

"Confucian and Rawlsian Views of Justice: A Comparison," Journal of Chinese

Philosophy 24 (1997), 427-56.63. See his "Introduction to the Paperback Edition," 1xii.

64. For example, see Frank I. Michelman, "Rawls on Constitutionalism and Con-

stitutional Law," in Samuel Freeman (ed.), The Cambridge Companion to Rawls

(Cambridge, UK: Cambridge University Press, 2003), 394-425; John E.Roemer, "Three Egalitarian Views and American Law," Law and Philosophy 20

(2001), 433-60.

65. Ronald Dworkin, "Rawls and the Law," Fordham Law Review 72 (2004), 1387-

405 (1387). On tort law, see Benjamin C. Zipursky, "Rawls in Tort Theory:

Themes and Counter-Themes," Fordham Law Review 72 (2004), 1923—40.66. However, see Robert G. Black and Others, Carrying on the Practice of Law under the Firm

Name of Black and Company v. The Law Society of Alberta [1989] 3 C.M.L.R. 288 Sup

Ct C (Canada), 313nn56, 58. Rawls is briefly mentioned in this case heard before

Canada's Supreme Court.

67. For example, see Uhl v. Thoroughbred Technology and Telecommunications, Inc., 2001

WL 987840, S.D. Ind., Aug. 28, 2001; Jensen v ARA Services, Inc., 736 S.W. 2d 374,

Mo., Sept. 15, 1987; and Parks v. Union Carbide Corp., 602 S.W. 2d 188, Mo. June10, 1980.

68. Goetzv.Crossnan, 967 F 2d, USLW 2010, 2nd Cir. (NY), June 10, 1992 at 39.

69. Martin v. Dugger, 686 F. Supp. 1523, 1569nl8, S.D. Fla., June 1, 1988.

70. On this issue Rawls suggests that the amendments to the Constitution should beaccepted only when they either (a) bring it "more in line with its original pro-

mise," or (b) "adapt basic institutions in order to remove weaknesses that come

to light in subsequent constitutional practices" (PL, 238f.).

71. See Vacco v. Quill, 521 U.S. 793 (1997) and Washington v Glucksberg, 521 U.S. 702

(1997).

72. See Paul J. Weithman, "Of Assisted Suicide and 'The Philosophers' Brief,"

Ethics 109 (1999), 548-78.73. See R. Dworkin, P.L. Zimroth, and A. Krash, "Brief of Ronald Dworkin,

Thomas Nagel, Robert Nozick, John Rawls, Thomas Scanlon, and Judith JarvisThompson as Amici Curiae in Support of Respondents," Glucksberg, 521 U.S. 702

(1997) (Nos 95-1858, 96-110), also published in Issues in Law and Medicine 15

(1999), 183-98.

74. Neomi Rao, "A Backdoor to Policy Making: The Use of Philosophers by the

Supreme Court," University of Chicago Law Remew f>5 (1998), 1371—401.75. See Thorn Brooks, "Does Philosophy Have a Place at the Supreme Court?" Rut-

gers Law Record 27 (2003), 1-35.76. Roev. Wade, 4:10 U.S. 113 (1973).

77. See Brooks, "Does Philosophy Have a Place at the Supreme Court?" 15-22.

78. It should be noted that it is not entirely clear that the Court actually avoids any

engagement with the Philosopher's Brief. (See ibid., 22—8.)

21Introduction

Of the many criticisms of Rawls's work, among the sharpest have been thosecharging inconsistency. Rawls's later work has been accused of contradicting

nearly everything that made his earlier work important.

Rawls's second book, Political Liberalism, attracted a great deal of censure on

this score. Many worried that Rawls's new-found concerns with stability and

consensus had resulted, in the words of one critic, in "a slighting of economic

justice and the plight of the worst-off, which was central in Theory of Justice."The difference principle, it was feared, had been "sacrificed," or at least

"drowned out." Bruce Ackerman complained that: "The egalitarian com-

mitment of A Theory of Justice does not survive the movement to Political Liber-

alism . .. Rawls is wrong, then, to suppose that his new commitment to politicalliberalism is compatible with his older commitments to the original position

and equality."

Other critics charged that the "radical change" of turning justice as fair-

ness into a political conception limited its relevance by making the theory

applicable only to societies that were already liberal. This "parochialism"

struck some as "disappointing"; other critics accused him of "morally criti-

cizable backsliding"; and at least one author claimed that Rawls "appears to

have jettisoned the project of justifying liberalism" altogether. Brian Barry,

reflecting on this "bad book by a famous author," remarked that "since there

is a widespread feeling that Political Liberalism does not succeed in fulfilling its

stated task, the conclusion is naturally drawn that the whole Rawlsian project

is fatally flawed." Perry Anderson's verdict was that "Rawls's new book is

thus not a development of his earlier work: it is an amputation of it. The

burden of Political Liberalism is an intellectual renunciation, rather than any

substantive addition."

Rawls's third book was then charged with double betrayal. The Law of Peo-

ples was accused of undermining both A Theory of Justice and Political Liberalism.

Some critics objected that Rawls's idea of a people was neither "clear enough"

nor "significant enough in the human world" to displace the focus on per-

sons in justice as fairness. Some were perplexed that the highly progressive

principles of A Theory of Justice were replaced in the international realm by

1

The unity of Rawls's work

Leif Wenar

1

4

11

12

"timid" principles from a "vanished Westphalian world." Others com-plained that Rawls's willingness to tolerate illiberal societies was not only "abetrayal of liberalism," but also "blatantly inconsistent" with his treat-ment of illiberal minorities in Political Liberalism.

In sum, Rawls's critics have charged "philosophical incoherence." What-ever the value ofRawls's particular arguments, the sense among many theo-rists is that these arguments do not fit together. Rawls's later additions havedragged down the original structure, and the best that Rawlsians can hopefor is that something could be built with the wreckage.

This chapter attempts to respond to these criticisms by laying out a systema-tic interpretation ofRawls's work as a whole. Rawls's work can indeed be seento present a unified theory, whose power comes from the mutual support of itsparts. Elsewhere I have drawn on this unifying interpretation to explain whyRawls went the way he did at particular points—for example, why he framedhis global theory in terms of peoples instead of persons, and why he rejected aninternational difference principle. In this chapter, I survey the entirety ofthe Rawlsian theoretical architecture, attempting to show how the majorstructural concepts in Rawls's works fit together. The challenge taken uphere is, essentially, to "use all of these Rawlsian words in one sentence." Theinterpretation will be successful if the reader agrees that this "sentence" laysout a consistent and interesting political philosophy, and one that remainstrue to the texts that Rawls wrote.

The reconstruction that follows is organized around the ideas of justice andlegitimacy. Justice is a familiar theme from Rawls's work, yet the interpreta-tion here assumes that the idea of legitimacy is at least as important to Rawls'sproject. An emphasis on legitimacy is essential, I believe, for understandingthe motivation behind many ofRawls's arguments, as well as how these argu-ments are intended to support one another. The importance of legitimacy toRawls's work has sometimes been suggested, but in my view it has not yet beensufficiently appreciated.

The reconstruction of Rawls's work is divided into four sections, corre-sponding with the four main topics that this work addresses. These fourtopics are: legitimacy within a liberal society, justice within a liberal society,legitimacy within a decent (but non-liberal) people, and legitimacy amongliberal and decent peoples. The reconstruction begins with what Rawls callsthe first task of liberal political theory.

Legitimacy within a liberal society

The first task of liberal political theory is to find principles to order a constitu-tional regime so as to be both legitimate and stable. The exercise of political

23The unity ofRawls's work

19

15

power in a liberal society is legitimate only when exercised in accordance witha constitution the essentials of which are acceptable to all citizens, regarded asreasonable and rational. Citizens are reasonable when they are ready to pro-pose and abide by fair terms of cooperation even at the expense of their owninterests, given that others are also willing to do so.

The task of finding legitimate principles for a liberal society is made difficultby the fact of reasonable pluralism. This is the fact that citizens of moderndemocratic societies will hold irreconcilable but reasonable comprehensivedoctrines—that is, irreconcilable but reasonable ideals of character and con-ceptions of what is valuable in human life. Because of reasonable pluralism,no comprehensive doctrine will be acceptable to all reasonable citizens.Therefore no comprehensive doctrine can be the basis for the legitimate exer-cise of political power in a liberal society.

Reasonable pluralism also makes the problem of stability acute. A stableliberal regime must be freely supported by a substantial majority of its politi-cally active citizens. Yet no liberal regime can maintain stability on the basisof the principles of a comprehensive doctrine without the oppressive use ofstate power.

No comprehensive doctrine can provide the content for the principles of alegitimate and stable liberal constitution. What other source could there be forthe content of these principles? There is only one source of fundamental ideasthat could serve as a focal point for all reasonable citizens of a liberal society:the public political culture of that society. The public political culture com-prises the political institutions of a society and the public traditions of theirinterpretation, as well as historic texts and documents that have become partof common knowledge. Reasonable citizens will understand that the publicpolitical culture of their society is the only source of ideas on which all can con-verge for the purpose of determining the basic terms of their cooperation.

The general solution to the problem of reasonable pluralism is thus to ordera liberal constitution according to the principles of a political conception ofjustice. A political conception is a moral conception of justice for the basicstructure whose principles are worked out from the fundamental ideas implicitin the public political culture of a liberal society. Since a political conceptionof justice stands free from all comprehensive doctrines, it is possible that theprinciples of a political conception will be acceptable to all reasonable citi-zens, and so possible that these principles can serve as the basis for legitimatecoercion. And it is possible for such principles to order society stably, sincethey can be the focus of an overlapping consensus.

In an overlapping consensus each reasonable citizen supports the politicalconception from within his or her own comprehensive view. Such a con-sensus is possible because reasonable citizens' comprehensive doctrines are

The Legacy of John Rawls24

22

24

27

32

likely to be (or to become) compatible with the ideas in the public politicalculture from which the principles of the political conception are derived.

Legitimacy thus requires that state power in a liberal society be exercised inaccordance with the values of a political conception of justice. Yet state poweris not the only political power that is exercised in a liberal society. Legitimacyimposes a moral duty of civility upon democratic citizens to appeal to theshared values of a political conception when they exercise political powerover each other, and especially when they debate and decide upon constitu-tional essentials and matters of basic justice. This is the duty of publicreason. Citizens are to appeal to and decide in accordance with shared politi-cal values when they vote in elections, when they campaign for political office,and when they explain their decisions as government officials.

What then are the ideas in the public political culture of a liberal societythat can be used to construct a political conception of justice? A fundamentalidea in the public political culture adequate for working up a political concep-tion of justice is the idea of society as a fair system of cooperation among freeand equal citizens. There are many ways of specifying these ideas of fairness,freedom, and equality. So there are many liberal political conceptions of jus-tice. Any of these conceptions of justice, if implemented, would satisfy theliberal principle of legitimacy: that the exercise of coercive political poweris fully legitimate only when this power is exercised in accordance with aconstitution the essentials of which all citizens may reasonably accept as freeand equal.

Each member of the family of reasonable liberal political conceptions of jus-tice will interpret differently the fundamental ideas of fairness, freedom, andequality. However, there is a limit to the latitude with which these ideas fromthe public political culture may be interpreted. All reasonable liberal politicalconceptions of justice will meet three criteria of liberal legitimacy. Theywill: (a) ascribe to all citizens the familiar liberal basic rights and liberties;(b) assign these rights and liberties special priority with respect to claims ofthe general good and perfectionist values; and (c) assure all citizens adequatemeans for taking advantage of these freedoms. These criteria require morespecifically that there be publicly funded elections, universal basic health care,and a not excessively unequal distribution of wealth and income.

When citizens (including officials) exercise political power in accordancewith the three criteria of liberal legitimacy, and within the bounds of publicreason, they satisfy the demands of legitimacy and thus the criterion of reci-procity among citizens. A legitimate society's stability is secured when itsbasic structure is effectively regulated by one of the family of reasonable politi-cal conceptions of justice (or a mix thereof), and when citizens who affirmsome member of this family are in an enduring majority.

25The unity ofRawls's work

34

36

39

41

43

Justice as fairness is a political conception of justice based on one specific inter-pretation of the ideas of fairness, freedom, and equality found in the publicpolitical culture of a liberal society. According to justice as fairness the free-dom of citizens has three aspects. Citizens are free in that they regard them-selves as having a capacity to form, revise, and pursue their conception of thegood; in that they believe themselves to be self-authenticating sources of validclaims on institutions; and in that they are viewed as capable of taking respon-sibility for their ends given the resources likely to be available to them. Citi-zens are equal in virtue of possessing to a minimal degree the basic moralpowers and the capacities that enable them to be fully cooperating membersof society. The specification of what citizens need—primary goods—isderived from this conception of the citizen. The conception of fairness usedin justice as fairness emphasizes that social and natural facts about citizensthat are arbitrary from a moral point of view should not be taken as basicwhen determining the distribution of primary goods.

The original position is a thought experiment meant to move from theseconceptions of fairness, freedom, and equality to determinate principles ofjustice. In the original position, rational representatives of reasonablecitizens choose principles of justice under conditions that are reasonable rela-tive to the conceptions of citizen and society outlined. For example, theidea that society should be a fair system of cooperation among equals ismodeled by the symmetric situation of the parties behind a veil of ignorance(that is, by their not knowing the particular circumstances of those theyrepresent, so that they are unable to favor those they represent in theirchoices). Since the conceptions of citizen and society are modeled in thesetup of the original position, the principles of justice that are selected withinthe original position should be the principles that are most congruent withthese conceptions.

Two principles would be selected in this original position: the first assuringequal basic rights and liberties, the second requiring fair equality of opportu-nity and that any inequalities of wealth and income be to the advantage ofthose worse off. These two principles are then to be further specified fromthe perspective of this original position into a fully determinate scheme of jus-tice through a four-stage sequence.

Since justice as fairness meets the three criteria of liberal legitimacy it is amember of the family of reasonable political doctrines. Justice as fairness istypical of such liberal conceptions, although it is also the most egalitarian ofthem. Some may believe that justice as fairness is the most reasonable con-ception of justice, while others may reasonably prefer other members of thefamily of reasonable political doctrines.

Justice within a liberal society

The Legacy of John Rawls26

48

49

50

52

54

57

46

The liberal principle of legitimacy specifies how coercive power may properlybe used within a liberal society. A more general principle of legitimacy isneeded to evaluate the use of coercive power in non-liberal societies. This gen-eral principle of legitimacy is: The exercise of coercive political power overpersons is legitimate only when it is exercised in accordance with a basic struc-ture that is acceptable to those persons, regarded as decent and rational.Persons are decent when they are ready to abide by the terms of a decentscheme of social cooperation even at the expense of their own interests, giventhat others are also willing to do so.

A society's basic structure will specify a decent system of social cooperationinsofar as it meets four conditions. First it must secure proper human rights,including rights to subsistence, security, personal property, and formal equal-ity before the law, as well as freedoms from slavery and some measure ofliberty of conscience. Second, its legal system must be such as to imposebona fide moral duties and obligations on all persons subject to it on mattersof law beyond those that concern human rights. A basic structure that satis-fies these first two criteria realizes a common-good idea of justice, in thatit takes into account what the society sees as the fundamental interests ofall persons. Third, its officials must hold and publicly demonstrate a sincereand not unreasonable belief that the law they administer is guided by acommon-good idea of justice. Fourth, it must give citizens a meaningfulrole in political discussions by providing opportunities for dissent, and byrequiring government officials to take this dissent seriously and give it a con-

, • , 65scientious reply.A society's basic structure must meet these four criteria in order to qualify

as a decent scheme of social cooperation. The institutions of a decent societymay be inegalitarian, and they may be based on a comprehensive doctrinethat is dominant in the local public political culture such as a religiousview. Liberals will not see such a society's laws as just, since these laws willnot be based on the ideas of fairness, freedom, and equality. Yet since decentsocieties are non-aggressive and their institutions are legitimate, liberal socie-ties have no justification for interfering in their affairs. To interfere in theaffairs of a decent society would be intolerant.

An outlaw state is not a decent society: it is either aggressive towardother peoples, or its officials violate the human rights of those within theirterritory, or both. Outlaw states need not be tolerated. Liberal societiesmay resist aggressive outlaw states in self-defense; and liberal societies mayintervene in the affairs of outlaw states to stop severe violations of humanrights, since officials who violate human rights have no legitimate authorityto exercise coercive political power in this way.

Legitimacy within a decent people

27The unity ofRauils's work

58

59

69

70

71

We require, finally, global principles to regulate relations among the mem-bers of the various liberal and decent peoples. Since these principles will becoercive, they must also pass a basic test of legitimacy. These principles must,that is, be acceptable to all persons regarded as rational and as either decent orreasonable. However, there is even more pluralism among individuals' com-prehensive doctrines globally than there is pluralism among individuals'comprehensive doctrines within a liberal society. So, as above, given thispluralism, no person's comprehensive doctrine can provide the content of theprinciples that will be used to coerce all. We must instead again look to apublic political culture as the focal source of fundamental ideas for the contentof a political conception. This time, we must look to the global public politicalculture to find the content for a political conception of a law of peoples.

The global public political culture contains few ideas about how personsliving in different societies should relate directly to one another. However,the global public political culture contains a wealth of ideas concerning howpeoples ought to relate to one another. The principles governing relationsamong the members of liberal and decent societies must therefore be principlesregulating conduct not among persons, but among peoples.

A people is a reasonable group of persons bound together by common sym-pathies and sharing a reasonably just or decent basic structure. Peoples arereasonable when they are ready to propose and abide by fair terms of coopera-tion even at the expense of their own interests, given that other peoples are alsowilling to do so. Liberal peoples and decent peoples are together known aswell-ordered peoples. The fundamental interests of well-ordered peoplesinclude protecting their citizens, their territory, their political independence,and their self-respect as peoples. Well-ordered peoples do not, however,have a fundamental interest in wealth above the level necessary to sustaintheir legitimate institutions; and for this reason there need be no (re-)distribu-tive principle for wealth among peoples above the level necessary to sustainthe legitimacy of each.

The fundamental idea in the global public political culture adequate forworking up a political conception of a law of peoples is the idea that peoplesought to relate fairly to each other as free and equal. Peoples are free in thatthey conceive of themselves as politically independent and self-determining:that is, as capable of making their own decisions concerning their territoriesand their citizens' well-being. Peoples are equal in that the fundamentalinterests of each are of the same importance as those of others. The idea offairness in the global public political culture emphasizes that the size, power,and prosperity of different peoples should not be taken as basic when deter-mining the terms of their cooperation.

Legitimacy among well-ordered peoples

The Legacy of John Rawls28

74

76

81

84

The global original position is a thought experiment meant to movefrom these conceptions of fairness, freedom, and equality to determinate prin-ciples of a law of peoples. This original position works by allowing rationalrepresentatives of reasonable peoples to choose principles for a law of peoplesunder conditions that are reasonable relative to the conception of peoples andtheir proper relations. Since the conceptions of peoples and their proper rela-tions are modeled in the setup of this original position, the principles of thelaw of peoples that are selected within it should be the principles that are mostcongruent with these conceptions.

The eight principles of the Law of Peoples would be selected in this orig-inal position. These principles state among other things that peoplesshould not instigate wars of conquest, should abide by their agreements,should honor human rights, and should assist those peoples that have fallenbelow the material conditions necessary to sustain legitimacy. The partiesin the second original position would also select standards of fairness fortrade, as well as guidelines to provide for cooperative organizations such asa world bank. The further interpretation of these principles and specifica-tion of these institutions is to be undertaken from the perspective of the globaloriginal position.

Finally, the basic requirement of legitimacy imposes a moral duty ofcivility on peoples to respect the bounds of global public reason. Membersof peoples are to appeal to and decide in accordance with shared global politi-cal values when they vote in elections, when they campaign for public office,and when they explain their foreign policy decisions as government officials.When peoples and their members exercise political power in accordance withthe eight principles of the Law of Peoples, and within the bounds of globalpublic reason, they satisfy the demands of legitimacy and thus the criterion ofreciprocity among peoples. The stability of the Society of Peoples is securedwhen relations among peoples are guided by the Law of Peoples, and whenwell-ordered peoples are capable of constraining any outlaw states that threa-ten the peace that obtains among them.

This concludes the interpretation ofRawls's work. Rawls's theories of justiceand legitimacy are not entirely finished, since they must still be extendedoutward to accommodate our relations to nature, and extended inward toexplain our duties toward our fellow liberal citizens who cannot cooperatefully with us. But on the topics treated here, Rawls's work is unified and com-plete. Rawls's system may still face objections that one or another argumentis unsound. Yet the charge of inconsistency is one criticism that Rawlsians

, 96can meet.

29The unity ofRawls's work

86

87

90

91

93

1. Stephen Holmes, "John Rawls and the Limits of Tolerance," The New Republic(11 October 1993), 39-47 (39).

2. Brian Barry, "John Rawls and the Search for Stability," Ethics 105 (1995), 874-915 (913).

3. Susan Moller Okin, "Review of Political Liberalism" American Political ScienceReviewQl (1993), 1010-11 (1010).

4. Bruce Ackerman, "Political Liberalisms," Journal of Philosophy 91 (1994), 364-86(374-5).

5. Bernard Williams, "A Fair state," London Review of Books (May 13, 1993), 7-8 (7) .6. Samantha Brennan and Robert Noggle, "Rawls's Neglected Childhood," in Vic-

toria Davion and Clark Wolf (eds), The Idea of a Political Liberalism (Boston:Rowmanand Littlefield, 2000), 46-71 (64).

7. Clark Wolf uses this phrase when reporting the criticisms of others, in "Funda-mental Rights, Moral Pluralism, and the Moral Commitments of Liberalism,"in Davion and Wolf, Idea of a Political Liberalism, 102-26 (124).

8. Bruce Brower, "The Limits of Public Reason," Journal of Philosophy 91 (1994),5-26(8).

9. Barry, "John Rawls and the Search for Stability," 915.10. Ibid., 915. Barry did not endorse as such the conclusion drawn in his remark,

although he did describe what he believed were great tensions between A Theoryof Justice and Political Liberalism.

11. Perry Anderson, "OnJohn Rawls," Dissent (Winter 1994), 139-44 (140).12. Thomas Pogge, "An Egalitarian Law of Peoples," Philosophy and Public Affairs

23 (1994), 195-224 (197).13. Stanley Hoffman, "Dreams of ajust World," New York Review of Books (2 Novem-

ber 1995), 52-6 (53).14. Allen Buchanan, "Rawls's Law of Peoples: Rules for a Vanished Westphalian

World," Ethics 110 (2000), 697-721.15. Buchanan ("Rawls's Law of Peoples," 697) uses this phrase in describing the

reactions of critics of Rawls's Law of Peoples.16. Kok-Chor Tan, "Liberal Toleration in Rawls's Law of Peoples," Ethics 108

(1998), 276-95 (283).17. Simon Caney, "Cosmopolitanism and the Law of Peoples," Journal of Political Phi-

losophy 10 (2002), 95-123 (106).18. See my "The Legitimacy of Peoples," in P. de Greiff and C. Cronin (eds), Global

Politics and Transnational Justice (Cambridge, MA: MIT, 2002), 53-76; and "Con-tractualism and Global Economic Justice," Metaphilosophy 32 (2001), 79-94; re-printed inT. Pogge (ed.), Global Justice (Oxford, UK: Blackwell, 2001), 76-90.An updated version of "The Legitimacy of Peoples" entitled "Why Rawls isNot a Cosmopolitan Egalitarian" is forthcoming in a collection of essays onRawls's Law of Peoples edited by Rex Martin and David Reidy.

19. Three excellent essays which emphasize the importance of legitimacy for under-standing Rawls's work are David Estlund, "The Survival of Egalitarian Justice in

Notes

The Legacy of John Rawls30

John Rawls's Political Liberalism" Journal of Political Philosophy 4 (1996), 68-78;

Allen Buchanan, "Justice, Legitimacy, and Human Rights," in Davion andWolf, Idea of a Political Liberalism, 73—89; and Burton Dreben, "On Rawls and

Political Liberalism," in Samuel Freeman (ed.), The Cambridge Companion to

Rawls (Cambridge, UK: Cambridge University Press, 2003), 316-46. On the

unity of Rawls's work, see also Daniel Weinstock's critical survey, "The Justifica-tion of Political Liberalism," Pacific Philosophical Quarterly 75 (1994), 165-85.

20. PL, xx, xxx, xli, 3-4, 47; JF, 1-2.

21. PL, xlvi, 136-40; JF, 40-1.22. PI,xliv, 48-54,81-6, 94; LP, 87-8; PAR reprinted in IP, 136; JF, 6-7.

23. PL, xviii, xxxviii, 3-4, 13, 36-7, 58-66, 175; RH, 374; JF, 3-6, 33-4.

24. PL, xli, 60-2, 146.

25. PL, xxxix—xl.

26. PL, 38; JF, 34.

27. PL, 37, 133-8; JF, 34, 84.28. PL, 45-6, 192; JF, 2-5, 34-5.

29. PL, 13-15; JF, 19-20.

30. PL, 11-15, 174-6; RH, 376; PAR, 43; JF, 26-7.

31. PL, xliv, 10, 40, 141-2; RH, 375.32. PI, 9, 15, 38-40, 132-49.

33. PL, xxi, 38, 134, 168-71; RH, 386-7; PRR, 172-3; JF, 32-3.

34. PL, xviii, xliii, xlvii-xlviii, 144-50, 158-68; JF, 33, 188-9, 192-8.

35. PL, xxii-xxiii, 1-lvii, 10, 137, 212-16, 227-30; IP, 55-6; PRR, passim; JF, 48,

90-2, 117-18.

36. PL, 215-16, 252-4; PRR, 133-6.37. PL, 11-20; JF, 5-8, 25-6; RH, 376; IP, 15.

38. PL, xlix, 167, 223-7; RH, 427-9; PRR, 141.

39. PI, xlix, 136-40, 216-17; RH, 393, 428; JF, 84, 141.

40. PL, xlviii, 6; IP, 14; PRR, 141.

41. PI, lvii-lix;IP, 49-51.

42. PL, xliv-xlvi, li, 16-18, 48-50; IP, 7, 14, 43n53, 114; PRR, 132-8, 140-1, 146-8,156-7, 168, 172-3.

43. PI,xlix-l, 14:0-72; RH, 391-2; JF, 9, 32, 84, 185-9, 199. A society stably ordered

by an overlapping consensus of reasonable comprehensive doctrines achieves a

reasonable and sufficient social unity.

44. PL, xxxvii, xliii, 4-5, 9, 167, 225-7; JF, xvii, 5-6, 39.45. TJ, 131-2, 475; PI, 19, 29-35; JF, 21-4.

46. TJ, 17, 441-9; PL, 19; JF, 20-1.

47. TJ, xiii; PL, 75-7, 178-90; JF, 57-61, 88, 188-9.

48. TJ, 13-14, 62-5, 82, 273-4; PI, 79; JF, 15-16, 55-7, 74-7, 124.

49. TJ, 15-19; PL, 22-8, 34-5, 45; JF, 14, 80-9.50. PL, xxii, 24-5, 77-81, 103-104; RH, 381; IP, 30-2; JF, 17-18,81-3.

51. TJ, 118-23; PI, 24-7, 79-80; JF, 15-18,87.

52. PL, xxii, 25-6, 72-3, 89-90, 103; JF, 16-18, 41-2.

53. TJ, 52-6, 266; PL, 5-6; JP, 42-50, 94-130.

31The unity of Rawls's work

54. TJ, 171-6; RH, 397-409; LP, 42n53; JF, 48.55. PL, xlviii—xlix.56. PL, 6-7; LP, 14.57. PL, xlviii-xlix; JF, 7-8, 39-41.58. This reading of The Law of Peoples, emphasizing legitimacy, is needed to make

sense of why Rawls is neither a cosmopolitan nor an international egalitarian.(See my "The Legitimacy of Peoples.")

59. LP, 64—8, 71—2, 86—8. Decent persons are responsible: they are capable of under-standing, recognizing, and acting in accordance with their duties and obligationsas specified by their society's conception of justice. (LP, 66, 71)

60. PL, 16, 109; LP, 65-8, 83, 93n6; JF, 6.61. IP, 65, 68, 78-81.62. LP, 65-7.63. Ibid.64. LP, 66-7.65. LP, 3n.2, 63, 72, 92.66. IP, 3, 64-7, 83.67. IP, 62-78.68. IP, 78, 83.69. IP, 59-62, 67-8, 83-4, 122.70. IP, 5, 48, 90.71. IP, 9, 80-1, 93n6, 94-5, 105-6.72. IP, 18-19, 40.73. IP, 54-5.74. IP, 15, 18, discussing the parallel third conditions of realistic Utopia.75. E.g., IP, 80n23, on Article 1 of the Universal Declaration of Human Rights.76. "The law of peoples proceeds from the international political world as we see it"

(IP, 83), with "familiar and largely traditional principles . . . from the history andusages of international law and practice." (IP, 57; cf. IP, 17, 36—7)

77. IP, 23-5, 59-68.78. IP, 25, 35.79. IP, 4, 62.80. IP, 28-30, 34-5, 47-8.81. See my "The Legitimacy of Peoples," 65—7.82. IP, 33-8, 111-12, 117-18.83. IP, 33-5, 69-70, 113-15, 121-2.84. IP, 32-3.85. IP, 30-5, 68-70, 115.86. IP, 58.87. IP, 39-43.88. IP, 37-8, 105-13.89. IP, 42-3.90. IP, 42.91. IP, 54-7.92. IP, 28, 35, 41, 56-7, 121-2.

The Legacy of John Rawls32

93. LP, 17-19, 27-30, 44-54, 64, 83-4, 122-6.94. TJ, 448-9; PL, 20-1, 245-6; JF, 176.95. PL, xlvi, 207-11, 241-7; LP, 86; PRR, 144-6.96. I am grateful to David Estlund, Stephen Macedo, Henry Richardson, Robert

Stern, and Paul Weithman for their criticisms and suggestions.

33The unity ofRawls's work

Introduction: fair equality of opportunity and its lexical priority

In his final statement of the two principles of justice, Rawls renders the secondprinciple as follows:

Social and economic inequalities are to be arranged so that they are both:

(a) to the greatest benefit of the least advantaged, consistent with thejust savings principle, and

(b) attached to offices and positions open to all under conditions of fairequality of opportunity.

I will hereafter refer to condition (b) as simply "fair equality of opportunity"or "FEO." FEO has two distinct components. First, FEO requires formalequality of opportunity or "careers open to talents," that is, it prohibits botharbitrary discrimination (on the grounds of race, gender, etc.) and monopolis-tic privilege (including barriers to entry in labor markets). Second, FEOdemands substantive equality of opportunity: all citizens must have a fairchance to achieve advantaged social positions, regardless of their social cir-cumstances (i.e. class status or family background). To achieve such fairness,the state must take action to prevent "excessive accumulations of propertyand wealth" and to maintain "equal opportunities of education for all."More specifically, the state must impose inheritance and gift taxes, limit theright of bequest, and subsidize education (whether directly through publicschools or indirectly through vouchers, tuition tax credits, loans, etc.).

Moreover, condition (b) of the second principle is lexically prior to condi-tion (a), that is, "fair [equality of] opportunity is prior to the difference prin-ciple" and consequently cannot be sacrificed for its sake. This priority rulemay seem unnecessary: under what possible circumstances would sacrificingFEO be to the "greatest benefit of the least advantaged," given especiallythat one of its goals is to "even out class barriers?" Rawls suggests that "itmay be possible to improve everyone's situation [including the least advan-taged] by assigning certain powers and benefits to positions despite the fact

2

Self-realization and the priority of fair equality of opportunity

Robert S. Taylor

1

3

4

that certain groups are excluded from them. Although access is restricted, per-haps these offices can still attract superior talent and encourage better perfor-mance." As an example, Rawls offers (though does not endorse) the claims ofBurke, Hegel, and others that "some sort of hierarchical social structureand a governing class with pervasive hereditary features are essential for the

public good."Though these claims are no doubt of historical interest, they may not seem

particularly compelling, leaving us still wondering whether the lexical prior-ity of FEO is really necessary. However, we do not need to rely on an example

as extreme as this one in order to recognize the possibility of conflict betweenFEO and the Difference Principle. For example, securing fair equality ofopportunity through educational subsidies could be exceedingly expensive,as a disadvantaged family and class background may make it difficult tobring a student up to a level of competence, much less excellence, at any

given task. Might it not be to the "greatest benefit of the least advantaged"to focus educational subsidies instead on those (often socially advantaged)students for whom such investment would offer the highest rate of returnand then tax them for the benefit of the poor? Rather than fighting a costly and

possibly futile battle against family and class privilege, one might insteadput such privilege to work for the least advantaged among us through re-distributive taxation. This thought seems to animate Rawls's own concept of"natural aristocracy," a kind of institutionalized noblesse oblige, and it is pre-cisely what the priority of FEO rules out as illegitimate. Thus, contraryperhaps to first impression, FEO's lexical priority has real bite: the least advan-

taged cannot trade off their fair opportunities to achieve office and position for the sake ofgreater monetary benefits.

Not surprisingly, this priority rule has been roundly criticized by manypeople, including Larry Alexander and Richard Arneson. Alexander seemsperplexed by the strength of the condition, believing that it makes a fetish of

our status as producers; moreover, he worries that FEO may become a "blackhole" for economic resources due to its lexical priority. Arneson writes that"enabling all individuals to have real opportunities for job satisfaction, educa-tional achievement, and responsibility fulfillment is not plausibly regarded asa justice goal that trumps all other justice values and should be pursued nomatter what the social cost."

Perhaps due to these criticisms, Rawls himself began to express doubts aboutthe lexical priority of FEO late in his life. Consider the following footnote(apparently written sometime in the early 1990s ) from justice as Fairness:

Some think that the lexical priority of fair equality of opportunity over thedifference principle is too strong, and that either a weaker priority or a

weaker form of the opportunity principle would be better, and indeed

35Self-realisation and the priority of fair equality of opportunity

8

12

more in accord with fundamental ideas of justice as fairness itself. At presentI do not know what is best here and simply register my uncertainty. How tospecify and weight the opportunity principle is a matter of great difficultyand some such alternative may well be better.

This latter-day ambivalence prompts the following question: what doesRawls's original defense of the lexical priority of FEO look like? We thus arriveat one of the most puzzling lacunae in all of his work. Apart from a single briefdiscussion (to which I will turn shortly), he fails to offer a?y< justification for thispriority rule. He defines the priority of FEO, illustrates it, etc., but never givesus an argument for it. This gap in his theory is made all the more surprising bythe almost obsessive care he takes in defending (with multiple arguments) theother major priority rule internal to justice as fairness, the priority of liberty.

Over the following pages, I attempt to defend Rawls against both his criticsand his own doubts by speculatively reconstructing his argument for the lexi-cal priority of FEO, building not only on the few clues he provides but also onother resources found in Theory, including especially the Aristotelian Principle(section 65) and the Humboldtian concept of social union (section 79). More-over, I show that this reconstruction can be defended against the criticism thatit commits Rawls to a substantive conception of the good, thereby jeop-ardizing the priority of right in his theory. As we shall see, this reconstitutedargument for the lexical priority of FEO strengthens the case for justice as fair-ness as well as having controversial implications for public policy.

Reconstructing Rawls's defense of the priority of FEO

Before starting my reconstruction, I should say a few words about method.Any attempt to reconstruct someone else's argument should hew as closely aspossible to their own words, methods, concepts, and (insofar as we can discernthem) intentions. As I reconstruct Rawls's defense of the priority of FEO, hisown words on the subject (which are few and vague, as we shall see) will pro-vide a rough guide. I will fill in the details using methods and concepts drawnfrom his own writings. Where interpolation or extrapolation is needed toadvance the argument, it will be carefully discussed and defended. The finalproduct of this effort should at the very least not be inconsistent with the spiritof Rawls's work; with luck, it will reflect his intentions and fit into the rest of histheory with a minimum of strain.

To begin, I will briefly examine Rawls's most powerful defense of the pri-ority of liberty, which will then serve as a model for a reconstructed defenseof the priority of FEO. In section 26 of Theory, Rawls says that "parties [inthe Original Position] regard themselves as having a highest-order interest in

The Legacy of John Rawls36

14

15

how all their other interests, including even their fundamental ones, are

shaped and regulated by social institutions." This highest-order interest inthe shaping of other interests (including the religious interest and the interestin integrity of the person) sits atop a "hierarchy of interests" and is lexicallyprior to all other interests, that is, it cannot be sacrificed to promote them. Asthe basic liberties are necessary conditions for the achievement of this highest-order interest, they cannot be sacrificed for the sake of other primary goods,such as office and position, income and wealth. Rawls's argumentative strat-egy is to justify a hierarchy of goods (basic liberties over other primary goods, asrequired by the priority of liberty) with a hierarchy of interests (a highest-order interest in shaping other interests over all other interests, including

fundamental ones). Of course, the success of such a strategy hinges on show-ing (inter alia) why this interest in the shaping of all other interests is ofsuch paramount importance, but its connection to our capacity for a concep-tion of the good, which is one element of our autonomy, makes such a showingpossible.19

So, one orthodox Rawlsian way to justify the lexical priority of FEO over

the Difference Principle (and therefore the lexical priority of fair opportunitiesto achieve office and position over income and wealth) is to justify the lexicalpriority of the interest that FEO supports over the consumption interest sup-ported by the Difference Principle. What kind of interest might this be?Rawls identifies it during his one very brief discussion of the priority of FEOand its defense:

I should note that the reasons for requiring open positions are not solely, oreven primarily, those of efficiency ... [The priority of FEO] expresses theconviction that if some places were not open on a basis fair to all, thosekept out would be right in feeling unjustly treated even though they bene-fited from the greater efforts of those who were allowed to hold them [as wasthe case with "natural aristocracy"]. They would be justified in their com-plaint not only because they were excluded from certain external rewards ofoffice but because they were debarred from experiencing the realization of self

which comes from a skillful and devoted exercise of social duties. Theywould be deprived of one of the main forms of human good.

Rather than concentrating on the interest in the "external rewards of office"(including salary and prestige), which after all bears a strong resemblance

to the consumption interest supported by the Difference Principle, I want tofocus on the interest in "the realization of self " that the holding of offices andpositions makes possible. If it can be shown that this interest is so important asto be lexically prior to the consumption interest, then the priority of FEO will

have been justified on orthodox Rawlsian grounds. Rawls's description of

37Self-realisation and the priority of fair equality of opportunity

20

16

18

self-realization as "one of the main forms of human good" suggests that suchan approach may be a promising one.

In order to demonstrate the importance of our interest in self-realization,however, we must first determine what self-realization consists of. Rawls saysthat realization of self comes from "a skillful and devoted exercise of socialduties." This skeletal explanation can readily be fleshed out by an examina-tion of the Aristotelian Principle (section 65), which motivates the achieve-ment of increasing virtuosity, and of the concept of social union (section 79),which provides the context for the development of such virtuosity. In thecourse of doing so we will see why and in what way Rawls believes that self-realization trumps consumption.

The Aristotelian Principle

Rawls defines the Aristotelian Principle in the following way: "other thingsequal, human beings enjoy the exercise of their realized capacities (theirinnate or trained abilities), and this enjoyment increases the more the capacityis realized, or the greater the complexity." The achievement of pleasure inincreasing virtuosity at any given activity is counterbalanced, however, by"the increasing strains of learning as the activity becomes more strenuousand difficult." This tradeoff between the pleasures and burdens of virtuosityat any given task determines how we will allocate resources (such as time andeffort) across tasks:

Every activity belongs to some chain. The reason for this is that humaningenuity can and normally will discover for each activity a continuingchain that elicits a growing inventory of skills and discriminations. Westop moving up a chain, however, when going higher will use up resourcesrequired for raising or for maintaining the level of a preferred chain.

Notice that in allocating resources the tradeoff is between virtuosity at differ-ent activities, not between virtuosity and consumption. In fact, Rawls onlyspeaks of our tendency to virtuosity being overridden when it comes intoconflict with justice itself or when various psychological proclivities (e.g. riskaversion and time preference) inhibit it. (I will return to both of thesepoints later.)

The importance of this tendency and its ramifications for institutionaldesign are spelled out in the following passage, which is rich with implicationsfor a defense of the priority of FEO:

The tendency postulated [i.e. the Aristotelian Principle] should be rela-tively strong and not easily counterbalanced. I believe that this is indeed

The Legacy of John Rawls38

24

25

the case, and that in the design of social institutions a large place has to bemade for it, for otherwise human beings will find their culture and form oflife dull and empty. Their vitality and zest will fail as their life becomes a, • . • 26tiresome routine.

The social duties attached to offices and positions provide valuable and (as weshall see) unique opportunities for the exercise and improvement of our abil-ities. FEO and its priority can be seen as creating and protecting institutionalspace for the use of our skills and guaranteeing resources (including educa-tional ones, in particular) to make their utilization effective. Consumptioncannot substitute for self-realization through the skillful discharge of socialduties for the very reasons alluded to in this passage: only increasing virtuositycan prevent life from becoming "dull and empty," whereas increasing con-sumption—though perhaps initially satisfying, especially where basic needshave yet to be met —has a tendency to become a "tiresome routine" itself,with titillation giving way to boredom and jadedness in an endless series ofaddictive cycles.

I should immediately note that Rawls never explicitly makes such a claimabout the lack of substitutability between self-realization and consumption.I am extrapolating here, but such an extrapolation is necessary to advance theargument: unless self-realization is of such a nature that consumption cannever substitute for it, we will be unable to defend the priority relationbetween the respective social primary goods (fair opportunities for office andposition versus income and wealth) that support them. Moreover, this extra-polation is consistent with many of Rawls's other statements about consump-tion. For example, during his defense of the Difference Principle, Rawls saysthat a "person choosing [according to a maximin rule] has a conception of thegood such that he cares very little, if anything, for what he might gain abovethe minimum stipend that he can, in fact, be sure of by following the maximinrule." This relative indifference to consumption beyond a "satisfactoryminimum" is consonant with (though it certainly does not imply) theabove claim about substitutability. I think these considerations militate infavor of provisionally accepting such a claim in order to see whether the recon-structed defense, considered as a whole, is compelling and broadly consistentwith Rawls's overall theory.

Before moving on to consider why the Aristotelian Principle should belinked to offices and positions in the basic structure of a just society, weshould consider another objection to the argument so far. The idea of mak-ing the Aristotelian Principle the foundation for a defense of the priority ofFEO might be criticized on the grounds that, as a mere factual premise, theprinciple has no moral force and cannot do the normative work that I wish itto do. That is, the Aristotelian Principle is, as Rawls repeatedly emphasizes, a

39Self-realization and the priority of fair equality of opportunity

28

"psychological law" or a "natural fact," a description of an evolved humantendency with obvious adaptive features. To argue from this innate disposi-tion toward virtuosity to a political principle that encourages and protects it isno more valid, so the criticism goes, than to argue from our innate disposi-tion toward violence to a political principle that endorses blood feuds andfactional warfare.

This criticism may lose its force, however, if its own premise is challenged:perhaps the Aristotelian Principle, despite Rawls's assurances, is somethingmore than a mere "psychological law." Few readers of Theory would questionthe moral role that autonomy plays in Rawls's theory; in sections 40 and 78,human autonomy is represented as the very ground of the moral law. Yet theform of its depiction, both there and elsewhere, is always factual in character:thus Rawls speaks of our "nature as free and equal rational beings" and of ourobservable capacities both for a conception of the good and for a sense of jus-tice, which are the constitutive elements of human autonomy. Thus, Rawlsoften clothes his normative premises in factual language, and so his use of suchlanguage when describing the Aristotelian Principle does not necessarily rob itof moral force. Given the importance of this premise in defending one of thekey elements of Rawls's justice as fairness, I believe we arejustified in ascribingto it more than merely factual significance.

Social union

Throughout his discussion of the Aristotelian Principle, Rawls constantlyhighlights the social context within which we develop our various skills. Theincreasing virtuosity of our fellow citizens, for instance, is a good for us, as theirimproved skills may help us to advance our own ends, may inspire us to similarforms of excellence, or may simply be a source of pleasure when they are pub-licly exercised.35 Such virtuosity can be developed and displayed in anumber of social settings, including even games and other forms of play.

This last observation raises the following important question: why wouldthe Aristotelian Principle bear any special relationship to offices and positionsin the basic structure, as required by the proposed defense of the priority ofFEO? In order to understand this connection, we must first examine the Hum-boldtian idea of social union, discussed by Rawls in section 79. As Rawls notesthere, individual men and women have neither the time nor the requisiteinborn potentials to achieve all the possible forms of human excellence. Theyare forced to specialize, choosing to develop some skills and allowing others tolie fallow. Fortunately, however, they can participate in and enjoy the com-plementary excellences of their fellow citizens through social cooperationin the pursuit of shared ends. Rawls's example of a symphony orchestra

The Legacy of John Rawls40

35

36

provides a nice illustration of these points: individual members of an orchestramay lack the time and/or ability to learn to play (or play well, at least) all or

even most instruments in an orchestra, but they can specialize by trainingthemselves on one or a few instruments and then cooperate with others in theorchestra to produce music together, thereby participating in the complemen-

tary excellences of their fellow musicians in the pursuit of a common goal.Now, as Rawls notes, such social unions can take many forms, many of

which are not properly thought of as part of the basic structure of society,which is the subject of justice. So, for example, friendships, chess clubs, artassociations, churches, and so on may be important examples of socialunions, but membership in them would generally not be regulated by FEO.What then distinguishes those social unions that are part of the basic struc-ture—governments, private and public corporations, universities, NGOs,and so forth—from social unions more generally? What makes them distinct

(inter alia) is that the offices and positions associated with them require amajor and usually dominant commitment of time and energy and act as theprimary sources of livelihood for those who hold them. The social duties asso-ciated with these offices and positions and the rich repertoire of skills necessaryto discharge them will consequently become a central focus of the lives of theofficeholders, especially their pursuit of virtuosity. Such centrality is thesource of the special connection between the Aristotelian Principle and the

offices and positions of the basic structure, and it explains why FEO is of suchoverwhelming importance: fair access to these positions is by far the mostimportant way (though certainly not the only way) to help citizens achievethe excellences of which they are capable.

A threshold condition for the application of the priority of FEO

Earlier I mentioned that the pursuit of virtuosity might legitimately be over-ridden if it conflicted with justice itself. For example, the priority of libertywould prevent the state from banning paeans to consumerism if its purpose indoing so was to keep citizens from being distracted from self-improving activ-ities. The first principle of justice is, in other words, prior to the second. ButFEO might be overridden, and its priority postponed, for reasons internal tothe second principle as well. For instance, Rawls notes that "the AristotelianPrinciple characterizes human beings as importantly moved not only by thepressure of bodily needs, but also by the desire to do things enjoyed simply fortheir own sakes, at least when the urgent and pressing wants are satisfied." Rawls isrecognizing here that the pursuit of virtuosity, at least for limited physicalbeings such as ourselves, has preconditions: we cannot effectively hone our

skills when we are racked by cold, thirst, hunger, or other such afflictions.

41Self-realization and the priority of fair equality of opportunity

38

Thus, the priority of FEO would have to be relaxed if such relaxation were

necessary to allow the accumulation of sufficient income and wealth to make

the pursuit of virtuosity itself feasible.

This last example raises a larger question: under what conditions does the

lexical priority of FEO come into effect? Rawls explicitly addresses this issue in

Political Liberalism, drawing a parallel between the first and second principles

ofjustice:

The notion of fair equality of opportunity, like that of a basic liberty, has a

central range of application which consists of various [non-basic] liberties

[such as free choice of occupation and freedom of movement] together

with certain conditions under which these liberties can be effectively exercised ... Just

as in the case of basic liberties, I assume that this range of application can be

preserved in ways consistent with the other requirements ofjustice, and in

particular with the basic liberties.

In the case of the basic liberties, these conditions include an unspecified level of

social, legal, and economic development (especially a modicum of material

comfort). Something similar is evidently intended for FEO: adequate

socio-political and material resources must be available before the priority of

FEO goes into effect, where "adequate" means whatever level is necessary for

the liberties associated with FEO to be "effectively exercised."

Given what Rawls has said on this subject, we can speculatively reconstruct

the nested set of thresholds for the application of the lexical priorities of liberty

and FEO. Begin with his general conception ofjustice, in which "all social

values—liberty and opportunity, income and wealth, and the social bases of

self-respect—are to be distributed equally unless an unequal distribution

of any, or all, of these values is to everyone's advantage." This general con-

ception presumably applies to all societies below a certain level of develop-

ment. Once the requisite level of social, legal, and economic development has

been reached, however, the first priority rule (the priority of liberty) comes

into play; given the priority of the first principle to the second, an increasing

social resource base must first be used to secure the priority of liberty. As the

resource base continues to grow, though, a point will eventually be reached

where the second priority rule (the priority of FEO) will come into effect; the

special conception ofjustice will then be fully implemented.

The priority of FEO versus the priority of right?

An important question now arises: does the commitment to self-realization

that I argue is implicit in the lexical priority of FEO simultaneously commit

The Legacy of John Rawls42

41

42

Rawls to a substantive conception of the good for its defense, thereby jeopar-dizing the priority of right in his theory? Arneson asks much the same questionand answers it as follows:

Within Rawls' theory, which eschews any social evaluation of people'sconceptions of the good, there does not seem to be a basis for affirming thatthe goods of job satisfaction and meaningful work trump the goods thatmoney and other resources distributed by the [Difference Principle] canobtain. From the different perspectives afforded by different and conflictingconceptions of the good, individuals will differ on this question . . . For some,work satisfaction and entrusted responsibility fulfillment may loom verylarge; for other individuals, quite other goods are crucial.

Interestingly, Rawls's own criticisms of moderate perfectionism (in section 50of Theory) for being inconsistent with the priority of right may militate againstthe proposed defense of the priority of FEO. He says there that the "criteria ofexcellence are imprecise as political principles, and their application to publicquestions is bound to be unsettled and idiosyncratic, however reasonably theymay be invoked and accepted within narrower traditions and communities ofthought." Granted, the perfectionism involved in Rawls's privileging of thepursuit of excellence through office and position is extremely weak and plura-listic compared to, say, Nietzsche's perfectionism. Nevertheless, it is initiallyunclear why the overriding importance ascribed to self-realization throughwork is any less "unsettled and idiosyncratic" than that ascribed by Nietzscheto creating and elevating Ubermenschen.

Might there be some way for Rawls to respond to these criticisms and toshow that the priority of FEO and the priority of right are in fact consistent?One possible response is that just as reasonableness (i.e. our capacity for a senseof justice) and rationality (i.e. our capacity for a conception of the good) arefacets of our autonomy, of our independence from natural and social contin-gency, so is self-realization. Since human autonomy is the very ground of themoral law in Rawls's doctrine of right, as we noted earlier, this response mayrescue him from the charge of inconsistency.

But how could self-realization possibly be construed as a facet of our auton-omy? To see how, first consider why rationality is such a facet. Rationalitymight at first seem heteronomous because unduly influenced by our needsand desires, which are themselves often the products of natural and social con-tingency. Rawlsian rationality is detached and critical, however, and requiresthat moral agents distance themselves somewhat from their immediate wants:

The aim of deliberation is to find that plan which best organizes our activ-ities and influences the formation of our subsequent wants so that our aims

43Self-realisation and the priority of fair equality of opportunity

45

and interests can be fruitfully combined into one scheme of conduct. Desiresthat tend to interfere with other ends, or which undermine the capacity forother activities, are weeded out; whereas those that are enjoyable in them-selves and support other aims as well are encouraged.

Far from blindly serving what Plato called the "manifold beast" of desire,rationality schedules, prioritizes, tempers, and prunes desires as well as orga-nizing them into a coherent plan of life. By doing so, it exemplifies our auton-omy and demonstrates that we are more than simply the resultant vectors ofgenetic, familial, and social forces.

Rationality's task of designing and implementing a plan of life requires theutilization of external resources, including especially the generic, liquid formof such resources—money. But there is a middle term, so to speak, between aplan of life and the external resources needed for its realization: internalresources, including skills, drive, and self-discipline. As we noted above, self-realization is solely concerned with cultivating such resources, just as ration-ality is focused on organizing and culling desires. Moreover, like rationality,self-realization may be impeded by refractory animal impulses. As Jon Elsterhas noted, akrasia, myopia, and extreme risk-aversion can act as barriers tothe development of internal resources: creating such resources in ourselves isinitially painful (hence akrasia and myopia as barriers) and not guaranteed tosucceed (hence risk-aversion as a barrier). So self-realization is in large parta struggle against these natural inertial tendencies, as Rawls himself inti-mates, and our success at it is as emblematic of our autonomy as the struggleof rationality against untoward desires.

I do not intend to suggest here, of course, that people develop their skills onlyas a way to advance their life plans. As Rawls emphasizes, "the AristotelianPrinciple characterizes human beings as importantly moved ... by the desireto do things for their own sakes." That is, the perfection of one's skills can benot only a means to, but also constitutive of, one's plan of life. Such dualityshould not present a problem: virtuosity, like health, is both good in itself andgood for what it makes possible.

Given the Kantian provenance of Rawls's theory, it is illuminating to notethat these three facets of autonomy—reasonableness, rationality, and self-rea-lization—are paralleled by the three varieties of maxim in Kant's practicalphilosophy: maxims of morality, maxims of prudence, and maxims of skill,respectively. As the connections between reasonableness and morality andbetween rationality and prudence are fairly clear, I will focus on the third ofthese connections, that between self-realization and skill. For Kant, themaxims of skill are technical imperatives: they determine not "whether theend [sought] is rational and good . . . but only what one must do in order toattain it ... Since in early youth it is not known what ends might occur to us

The Legacy of John Rawls44

48

in the course of life, parents seek above all to have their children learn a greatmany things and to provide for skill in the use of means to all sorts of discretionaryends." Once individuals have reached adulthood, of course, they continue todevelop these and other skills in the pursuit of their chosen ends, and, as wehave seen, this quest for excellence is the prime element of self-realization.But Kant also emphasizes that the development of one's skills is a self-regard-ing duty (if an imperfect one): "as a rational being he necessarily wills that allthe capacities in him be developed, since they serve him and are given to himfor all sorts of possible purposes." Thus, for Kant as for Rawls, self-realiza-tion is a moral imperative.

Kant goes on to note that these three varieties of maxim are "clearly dis-tinguished by dissimilarity in the necessitation of the will." Whereas themaxims of morality bind rational agents unconditionally, maxims of prudencehave force only in relation to a universal subjective end (i.e. happiness), andmaxims of skill constrain only insofar as agents actually will the ends to whichthese maxims specify the means. Hence, a hierarchy exists among the maxims:morality limits the pursuit of happiness, which in turn dictates the develop-ment of certain skills. A parallel hierarchical relation holds among Rawls'sthree priorities: the priority of right (grounded in our reasonableness) is para-mount; the priority of liberty (grounded in our rationality, as noted earlier)comes second but is the first priority "internal" to justice as fairness; finally,the priority of FEO (grounded in our capacity for and interest in self-realiza-tion) comes third. Thus, while all three of the facets of autonomy are emblem-atic of our independence from natural and social contingency, the degree ofindependence differs, and this dissimilarity motivates the hierarchical relationamong both them and the priorities that they ground.

Conclusion

I have argued in this chapter that the lexical priority of FEO in Rawls's justiceas fairness can be successfully defended against its critics, despite his owndoubts about it. Using the few textual clues Rawls provides, I speculativelyreconstructed his defense of this priority, showing that it is grounded on ourinterest in self-realization through work. This reconstructed defense made lib-eral use of concepts already present in Theory, including the Aristotelian Prin-ciple (section 65), which motivated the achievement of increasing virtuosity,and the Humboldtian concept of social union (section 79), which provided thecontext for the development of such virtuosity. I also showed that this com-mitment to self-realization, far from violating the priority of right in Rawls'stheory, stems directly from his underlying commitment to autonomy, which isthe very foundation of the moral law in his doctrine of right.

45Self-realization and the priority of fair equality of opportunity

56

Alternative defenses of FEO's lexical priority are no doubt possible. Forexample, one implication of this priority is that eliminating social inequalities(i.e. those arising from family and class privilege) is infinitely more importantthan counteracting natural inequalities (i.e. those arising from differences inability and ambition). Thus, one might be able to provide a basis for the lex-ical priority of FEO by arguing that social inequalities are infinitely worsethan natural inequalities. Why might this be so? Perhaps social but not naturalinequalities prevent us from being full and equal participants in the basicstructure of a well-ordered society or cause special injury to the self-respect ofthose denied fair opportunities, owing to the fact that social inequalities seemmore a product of conscious human action and even human design than nat-ural inequalities. Thus, the social dependency implicit in Rawls's idea of"natural aristocracy" might be deemed infinitely more degrading than thenatural dependency that is arguably implicit in the Difference Principle itself,which makes the income of the least advantaged dependent in large part on(properly motivated) able and ambitious people.

Without denying the promise of such alternatives, I do want to point outtwo advantages of the self-realization defense. First, it is clearly based onRawls's text, as I noted near the beginning of the second section of thischapter. When Rawls argues that those denied fair opportunities would be"debarred from experiencing the realization of self which comes from a skillfuland devoted exercise of social duties," he seems to be indicating his preferredway of defending FEO's lexical priority. We are not bound, of course, tofollow Rawls's lead, but given his own words and the way that the resultingself-realization defense fits neatly within his theory, a certain deference maynot be inappropriate. Second and more importantly, the self-realizationdefense is shown in the third section of this chapter to flow from the sameunderlying commitment to autonomy that ultimately grounds not only thepriority of right but also the priority of liberty. This defense thus serves as aconstituent element of a unified, autonomy-based defense of the three priori-ties injustice as fairness. Other approaches to defending the priority of FEOwould likely lack this coherentist justification.

Whichever approach to defending the priority of FEO that we ultimatelydecide to take, we must still ask: why is its defense so important? Given Rawls'sadmission that the argument for the Difference Principle is "unlikely ever tohave the force of the argument for the two prior principles," most of the powerand distinctiveness of justice as fairness would appear to derive from the twointernal priorities of liberty and FEO. Therefore, a persuasive defense ofFEO is a vital support for his theory, the success of which would otherwisedepend mostly, if not exclusively, on the defense(s) of the priority of liberty.But the implications of a compelling defense of FEO's lexical priority extendmuch further than this. Though the United States has utterly failed to provide

The Legacy of John Rawls46

59

either fair equality of opportunity or a satisfactory social minimum for its own

citizens, its relative emphasis on the former (especially in the form of subsidies

for higher education) may draw some support from the lexical priority of

FEO: the decision to commit resources to state colleges and universities, sub-

sidized student loans, and so on, rather than to broader financial support for

the poor, may be partially justified by the modest perfectionism of the self-rea-

lization defense.. Thus, far from being an obscure and poorly motivated com

panion to the priority of liberty, the priority of FEO is arguably its peer in

terms of both its importance to justice as fairness and the controversialness of

its policy implications.

Notes

1. TJ, 266.2. Ibid., 62; CP, 141; JF, 67n.3. TJ, 63.4. Ibid., 245; CP, 141; JF, 51, 161.5. TJ, 11, 266.6. Ibid., 63.7. Ibid., 73.8. Ibid., 264. Also see Bernard Williams, "The Idea of Equality," in Peter Laslett

and W.G. Runciman (eds), Philosophy, Politics, and Society (Oxford, UK: BasilBlackwell, 1962), esp. 125-9, and Michael Lessnoff, "John Rawls' Theory ofJustice," Political Studies 19.1 (1971), 63-80, esp. 75ff.

9. TJ, 57, 64-5.10. Larry Alexander, "Fair Equality of Opportunity: John Rawls" (Best) Forgotten

Principle," Philosophy Research Archives 11 (1985), 197-207; Richard Arneson,"Against Rawlsian Equality of Opportunity," Philosophical Studies 93.1 (January1999), 77-112. Also see Thomas Pogge, Realizing Rawls (Ithaca, NY: CornellUniversity Press, 1989), 161-96.

11. Alexander, "Fair Equality of Opportunity," 198, 202-3, 205-6. On FEO as a"black hole," also see Arneson, "Against Rawlsian Equality of Opportunity,"81-2, and Pogge, Realizing Rawls, 169.

12. Arneson, "Against Rawlsian Equality of Opportunity," 99.13. Email correspondence with Erin Kelly, editor ofJF (April 14, 2003).14. JF, 163n. There is some textual evidence (admittedly indirect) in TJ suggesting

that Rawls is responding to Alexander's criticisms. For example, a passage in theoriginal edition (87) that was sharply criticized by Alexander ("Fair Equality ofOpportunity,'' 199-200) —who felt that it implied that FEO would never conflictwith the Difference Principle—is missing in the revised edition (TJ , 76).

15. See sections 26, 33, and 82 in TJ, as well as "The Basic Liberties and Their Prior-ity" in PL.

16. TJ, 131.

47Self-realization and the priority affair equality of opportunity

60

61

17. Ibid. ,476.

18. Ibid., 131.

19. As I have recently argued in "Rawls's Defense of the Priority of Liberty:A Kantian Reconstruction," Philosophy and Public Affairs 31.3 (Summer 2003),

247-72.20. TJ, 73; emphasis added. This account of FEO's priority may explain Rawls's

position on the distribution of educational resources: "resources for education

are not to be allotted solely or necessarily mainly according to their return as

estimated in productive trained abilities, but also according to their worth inenriching the personal and social life of citizens, including here the least favored."

(TJ,92)21. Alexander speculates that the lexical priority of FEO might be defended using the

Aristotelian Principle but does not attempt such a defense himself. (Alexander,

"Fair Equality of Opportunity," 205—6)

22. TJ, 374.23. Ibid., 376.

24. Ibid., 378.

25. Ibid., 376-8.

26. Ibid., 377.

27. Ibid., 379.28. Ibid., 134.

29. Ibid., 135.

30. Rawls also says in PL that:

were the parties [in the Original Position] moved to protect only the material

and physical desires of those they represent, say their desires for money andwealth, for food and drink, we might think that the original position modeled

citizens" heteronomy rather than their rational autonomy. But at the basis

of the parties" reliance on primary goods is their recognition that these goods

are essential all-purpose means to realize the higher-order interests connected

with citizens' moral powers and their determinate conceptions of the good.(PL, 76)

As I will show later, self-realization can be understood as one of these "higher-

order interests."

31. The basic structure is the subject of justice and includes

the political constitution and the principal economic and social arrangements.

Thus the legal protection of freedom of thought and liberty of conscience, com-

petitive markets, private property in the means of production, and the monog-

amous family are examples of major social institutions. Taken together as one

scheme, the major institutions define men's rights and duties and influence

their life prospects, what they can expect to be and how well they can hope todo. (TJ, 6-7)

32. TJ, 375, 376.

33. Ibid., 222, 452, 455; emphasis added.

The Legacy of John Rawls48

34. Ibid., 442. Thus Rawls asserts that "these claims [about human moral capacities]depend solely on certain natural attributes the presence of which can be ascer-tained by natural reason pursuing common sense methods of inquiry." (TJ,442n)

35. TJ, 373, 375-6.36 Ibid., TJ, 374-5, 377.37. Ibid., 458-9.38. Ibid., 459n.39. Ibid., 378.40. Ibid., 379; emphasis added.41. Ibid., 228, 363-4; emphasis added.42. TJ, 54-5, 132, 474-6.43. Ibid., 54-5.44. Rawls argues that "the case for certain political liberties and the rights of fair

equality of opportunity is less compelling [than the case for 'liberty of conscienceand the rights defining integrity of the person'] . . . It may be necessary to forgopart of these freedoms when this is required to transform a less fortunate societyinto one in which all the basic liberties can be fully enjoyed." (TJ, 217)

45. Arneson, "Against Rawlsian Equality of Opportunity," 98—9.46. TJ, 290.47. Ibid., 286.48. Ibid., 360-1.49. Plato, Republic 589a.50. Jon Elster, "Self-Realization in Work and Politics: The Marxist Conception of

the Good Life," Social Philosophy and Policy 3.2 (Spring 1986), 97-126, here 107-8.51. TJ, 376-7.52. Ibid., 379.53. Immanuel Kant, Groundwork of the Metaphysics of Morals, trans. Mary Gregor

(Cambridge, UK: Cambridge University Press, 1997), 26-7. Also see H.J.Paton, The Categorical Imperative: A Study in Kant's Moral Philosophy (Philadelphia:University of Pennsylvania Press, 1947), 89-96.

54. Kant, Groundwork, 26; original emphasis.55. Ibid., 32-3.56. Ibid., 27.57. TJ, 63-5, 73-8.58. Ibid., 73.59. Ibid., xiv; cf. ibid., 220.60. The United States and South Korea are tied for the second-highest level of

post-secondary enrollment in the world (72% of the relevant age group), behindonly Finland (84%). (Figures from The Economist Pocket World in Figures (2004Edition), 74.)

61. I thank Debra Satz, Tamar Schapiro, Allen Wood, Rob Reich, Peter Euben,Peter Stone, Steven Kelts, Neal D'Amato, Simon May, and Lael Weis for theirhelpful comments and suggestions.

49Self-realisation and the priority of fair equality of opportunity

"Utilitarianism," John Rawls tells us a mere two dozen pages into A Theory ofJustice, "does not take seriously the distinction between persons." If we readthis sentence as expressing the heart of Rawls's criticism of utilitarianism as abasis for a conception of justice for a democratic society, we can draw twoconclusions. First, an adequate conception of justice for a democratic societyneeds to take the distinction between persons seriously, and second, Rawls'sown alternative: justice as fairness, does.

The first of these conclusions is, it seems to me, uncontroversial. Never-theless, it is worth noting that some degree of that lack of controversy comesfrom the fact that we are thinking about conceptions of justice here, and, inparticular, of what might be called political justice: the justice of basic socialand legal institutions, what Rawls calls the "basic structure." For while anynumber of desirable goals, from efficiency to environmental protection, mightnot require taking the distinction between persons seriously, the very conceptof justice surely does. Whether we define the concept of justice with Polem-archus in Plato's Republic as "giving to each what is owed to him" or withRawls as applying to institutions when "no arbitrary distinctions are madebetween persons in the assigning of basic rights and duties and when the rulesdetermine a proper balance between competing claims to the advantages ofsocial life," the very idea of justice makes reference to the distinction betweenpersons.

My interest here, however, lies with the second claim: that Rawls's justice asfairness does take the distinction between persons seriously. On one level, thisclaim is no more controversial than the first one, and to the extent that it iscontroversial, the controversy is rather well-trodden. Nevertheless, many ofRawls's readers, both sympathetic and critical, have not fully appreciatedeither what taking the distinction between persons seriously requires, or howdeeply Rawls's attempt to do so structures his philosophical work. Rawls, Iargue, offers us not only an alternative set of principles of political justice fora democratic society, but an alternative conception of the role of political phi-losophy in such a society. Taking the distinction between persons seriouslythus has implications not only for the principles of justice we endorse, but also

3

Taking the distinction between persons seriously

Anthony Simon Laden

for the relationship between political philosophy and social policy, and thushow we might, to use a Wittgensteinian turn of phrase, go on in political phi-losophy in a Rawlsian way.

I start by distinguishing four levels at which it makes a difference if we takethe distinction between persons seriously. With respect to each level, I makefour claims. First, Rawls takes the distinction between persons seriously andutilitarianism does not. Second, principles of justice for a democratic societyought to take the distinction between persons seriously. Third, whether or notone considers it important to take the distinction seriously at a given level isinfluenced by one's attitude at other levels, so that the most plausible andcoherent versions of both justice as fairness and utilitarianism will differ onall four levels. Finally, certain criticisms of justice as fairness for not takingthe distinction between persons seriously at one level fail to appreciate that italso takes this distinction seriously at higher levels. Having climbed into thephilosophical stratosphere, I conclude with some more mundane suggestionsabout the roles political philosophy should play in political life if we accept myclaims about the advantages of Rawls's approach on the way up.

We can label the levels at which taking the distinction between persons ser-iously matters as follows: (1) the content of principles, (2) the form of the argu-ment in favor of those principles, (3) the conception of (political) moralitymore generally, and (4) the conception of moral and political philosophy.While Rawls explicitly discusses the first two levels in A Theory of Justice, thethird and especially the fourth level remain relatively undiscussed, in partbecause Rawls himself mentions them only occasionally, and arguably onlyfully appreciated their significance toward the end of his career, with theseries of papers he published after Political Liberalism. Rehearsing the familiarterritory of the first two levels will, however, help to prepare the way for dis-cussion of the third and fourth levels.

Rawls begins his discussion of what he calls "classical utilitarianism" in Theoryby pointing out that it does not treat the distinction between persons seriouslyat the first level, that of content, and works from there to the second level, thatof the form of argument. Conceding that "it is impossible to deny the initialplausibility and attractiveness of" classical utilitarianism, Rawls goes onimmediately to point out its failure to pay attention to the distinction betweenpersons at the level of content: "The striking feature of the utilitarian view ofjustice is that it does not matter, except indirectly, how this sum of satisfactionsis distributed among individuals." At the level of content, that is, utili-tarianism does not take the distinction between persons seriously because it

Level 1: the content of the principles

Taking the distinction between persons seriously 51

5

does not pay attention to issues of distribution, except indirectly. Rawls's owntwo principles ofjustice, of course, pay a great deal of attention to distribution.In fact, it is not much of a stretch to say that distribution is all that they payattention to. A basic structure satisfies the two principles ofjustice to theextent that it guarantees or is highly likely to produce a particular distribu-tional pattern of basic liberties, opportunities, and goods. Such attention todistribution is a means of taking the distinction between persons seriouslybecause it requires us to be sensitive to who has what.

Note here two points about the difference between the content of utilitari-anism and that of Rawls's two principles. First, the content of an adequateconception ofjustice for a democratic society needs to treat the distinctionbetween persons seriously. Whatever else is true about a democracy, it seemsclear that one of its fundamental features is that it must take individual mem-bers of the demos seriously as distinct individuals. Presumably, what made util-itarianism a progressive philosophy of choice in the nineteenth and earlytwentieth centuries was the extent to which it does take each person seriouslyas an equally good repository of utility.

Second, and more interestingly, according to some of Rawls's critics, justiceas fairness fails to take the distinction between persons sufficiently seriouslyat this level. Robert Nozick argues that justice as fairness pays too muchattention to particular end-state patterns, losing sight in the process of theactual transactions between people that led to those end-states. In otherwords, Nozick claims that in order to take the distinction between personsreally seriously, a theory must pay more attention to what different per-sons do, and not treat them merely as points on a distributional curve.I return to this criticism below.

After pointing out that utilitarianism pays what he takes to be the wrong sortof attention to questions of distribution, Rawls goes on to trace this failure to aproblem with the "most natural way" of arriving at utilitarian principles: "toadopt for society as a whole the principle of rational choice for one man."According to Rawls's reconstruction here, the most natural argument for anaggregative principle is one that relies on an aggregative device, "conflatingall persons into one through the imaginative acts of the impartial sympatheticspectator."

Classical utilitarianism thus does not take the distinction between personsseriously at this second level because the argument on behalf of its principlesrelies on adopting a principle of choice for a single person and applying it

Level 2: the form of the argument

The Legacy of John Rawls52

to a whole society. In contrast, Rawls explains, justice as fairness, "being acontract view, assumes that the principles of social choice, and so the princi-ples of justice, are themselves the object of an original agreement." Theargument for the two principles made within justice as fairness, an argumentwhose central piece invokes the device of the original position, thus takes thedistinction between persons seriously because it involves showing that eachperson would agree to the principles of justice from his or her own distinctiveperspective.

Note that this criticism of utilitarianism is, indeed, different from the onemade at the first level. It is conceivable that the conflation of all persons intoone via the sympathetic imagination of an impartial spectator would lead us toendorse the two principles of justice or some other principle that places directemphasis on distribution, just as it is also conceivable that principles that paidno direct attention to distribution could be generated on the basis of a differ-ent kind of argument (as Rawls thinks is the case for what he calls "averageutilitarianism").

Nevertheless, there is an important connection between the attention onepays to the distinction between persons at each level. What makes the lack ofdirect attention to distribution shown by utilitarianism appear to be a prob-lem is in large part a concern to make an argument for principles of justicethat takes the distinction between persons seriously. If, on the other hand, weregard an aggregative device as an appropriate means to argue for principlesof justice, then we are unlikely to be overly concerned that the principles itselects have an aggregative content. Thus, what looks like a failure on thepart of utilitarianism at the first level from Rawls's perspective may seem tothe utilitarian who does not take the distinction between persons seriouslyat the second level as not a failure but merely the result of prior theoreticalcommitments.

The importance, within a democratic society, of taking the distinctionbetween persons seriously at this level becomes clear when we recall Rawls'sinsistence on what he calls publicity. Rawls insists that in developing prin-ciples of justice for a democratic society, we must develop a. public conceptionof justice. It is not enough to develop principles that can only achieve theirdesired result if applied behind the backs of democratic citizens, since themeans to do so would of necessity be undemocratic. Conceptions of justice fora democratic society cannot adopt the structure of what Bernard Williamscalls "Government House utilitarianism." But that means that both theprinciples and their justification must be publicly known (or knowable) byall citizens and that their application can proceed in light of such knowledge.An argument in favor of principles of justice for a democratic society, then,will only be successful if it can be addressed to each citizen individually, thatis, only if it takes a form that treats the distinction between persons seriously.

Taking the distinction between persons seriously 53

Rawls claims that a contract doctrine, suitably framed, does this to a greaterextent than the impartial spectator argument for utilitarianism.

The fact that the two principles of justice are principles that we could allagree to from an appropriate perspective provides grounds for a reply toNozick's criticism. Nozick treats justice as fairness as if it imagined a particulardistributional pattern descending from the heavens and imposing itself on ademocratic society. Once we understand that Rawls aims to take the differ-ence between persons seriously at this second level, however, that picture ofjustice as fairness has to be rejected. Rather, we are to imagine members of ademocratic society coming to agree that the justice of their society depends ontheir conducting their distinct affairs with one another within the confines ofcertain structures designed to bring about certain distributive results. Fromthe perspective of justice as fairness as a contract doctrine, its concern withdistribution ceases to look like a failure to take seriously the distinctionbetween persons. Even if the focus on distribution alone treats individuals asmere points on a distributional curve, the argument in favor of that particularpattern insists that each citizen be taken seriously as a distinct person whoseagreement must be secured for the principles to be just.

But while taking the distinction seriously at the second level provides Rawlsa response to Nozick, his means of doing so leads to a new sort of criticism.Utilitiarianism fails to mount an argument that takes the distinction betweenpersons seriously because it "conflates all persons into one through the imagi-native acts of the impartial sympathetic spectator." But surely, this secondline of criticism goes, justice as fairness also fails to take the distinction betweenpersons sufficiently seriously insofar as it appears to conflate all persons intoone through the imaginative act of a choice behind the veil of ignorance.As any number of critics have pointed out, and as Rawls himself seems toadmit, the invocation of the veil of ignorance, which makes the parties in theoriginal position indistinguishable from one another, has the effect of conflat-ing them into one person, of transforming an agreement that takes the dis-tinction between persons seriously into a choice that does not. This line ofcriticism is also mistaken, in part because it fails to take heed of the fact thatRawls aims to take the distinction between persons seriously at the third level:that of the conception of political morality itself.

The difference between utilitarianism and justice as fairness at this level is thedifference between what T.M. Scanlon calls "philosophical utilitarianism"and "contractualism" in his article "Gontractualism and Utilitarianism."It is brought out beautifully and forcefully by Christine Korsgaard:

The Legacy of John Rawls54

Level 3: the conception of (political) morality

15

To later generations, much of the moral philosophy of the twentieth centurywill look like a struggle to escape from Utilitarianism. We seem to succeed indisproving one Utilitarian doctrine, only to find ourselves caught in the gripof another. I believe that this is because a basic feature of the consequential-ist outlook still pervades and distorts our thinking: the view that the businessof morality is to bring something about... The subject matter of morality is notwhat we should bring about, but how we should relate to one another.If only Rawls has succeeded in escaping Utilitarianism, it is because onlyRawls has fully grasped this point. His primal scene, the original position,is one in which a group of people must make a decision together. Their taskis to find the reasons they can share.

On this reading of justice as fairness, its subject matter is how we relate to oneanother, and not a set of principles or institutions that we (whoever that turnsout to be) are to bring about. As Scanlon puts the point, for contractualism,the idea of general agreement is "in a more fundamental sense, what moralityis about."17

There are several points to be made about this contrast. The first is to ex-plain what it has to do with taking the distinction between persons seriously.Why does thinking that the business of morality is to bring something aboutfail to take the distinction between persons seriously, while thinking that thebusiness of morality is how we relate to one another succeed in doing so?A moral theory that tells me what I ought to bring about is likely to address meas if I am the sole agent in the universe. Let me explain what I mean by that.The moral theory tells me what I ought to do or bring about. No doubt it tellsme to do things to and for other people, but in telling me to do those things, itimplicitly treats those other people as more or less complicated objects. Theseobjects may be the recipients of my action, they may be the helpmeets of thataction, or they may get in my way, but they will not really be seen as my part-ners in action in any sort of strong way. On this reading of consequentialism, Ican be sensitive to the differences among those whom my action affects, butthere is a sense in which I will fail to treat them as persons in doing so.

Of course, consequentialist theories do not have to be as solipsistic as allthat. They can, if they are any good as moral theories, also take themselves toapply to all other people as agents, as persons, and so in some sense such atheory will not treat me as the only agent so much as a fully representativeone. But notice that from this perspective, a theory that takes any given agentas fully representative of all the rest is not really taking the distinction betweenpersons seriously. This helps explain the relative clumsiness that utilitariantheories have displayed in trying to deal with collective actions, whether inthe form of rule or coordination utilitarianism. In each case, the utilitarianimagines lots of people doing the same thing, all acting side-by-side, but not

Taking the distinction between persons seriously 55

really together, and tries to relate a single agent's action to that action multi-plied. Utilitarianism thus does not take the distinction between persons seri-ously at this third level because it does not treat the recipients of actions aspersons and does not treat the agents who perform them as distinct.

In contrast, if we take the subject of morality to be how we relate to oneanother, then we are led to take rather seriously the particular parties tothose relationships as both persons and distinct. To see this point, it helpsto further specify the relationships that are the subject of morality on thisview. They are intersubjective, which means that they are constituted by nor-mative relations between and among a certain kind of being of whom personsare the prime (and perhaps the only) example, rather than by causal relationsbetween and among beings capable of setting certain causal chains in motion,a class which includes persons but not in any special place. Thus, the rela-tionship I form with my fellow citizens as free and equal co-authors of thelaws that govern us is intersubjective because it is essentially normative andwhile it is conceivable that beings other than human beings could form thatsame relationship with me, it would require that they resemble human beingsin important ways. In contrast, the relationship that a doctor bears to me inso-far as it is at her discretion and under her control how much pain medication Iget need not be intersubjective, since it is constituted by her capacity to set inmotion certain causal chains that will affect me, and as such, the same rela-tionship could be set up between a robot and a dog.

The very possibility of intersubjective relationships requires that there bedistinct persons, and so any theory that focuses its attention on such relation-ships will need to take the distinction between persons seriously. Moreover, ifas in contractualist moral theories, what determines the moral quality of anintersubjective relationship is whether the parties to it can agree to its terms,then we are going to have to be particularly sensitive to the different and dis-tinct positions from which the various parties to that relationship evaluate it.

Taking the distinction between persons seriously at this level is also an im-portant criterion for an adequate conception of justice for a democratic soci-ety. Here, we should note two further fundamental facts, one about justice andthe other about democratic societies. Questions of justice arise only insofaras there exist distinct individuals with competing claims. This requires notonly that the individuals be distinct, but that they somehow be related, suchthat their claims can truly be said to compete. Interestingly, Rawls claims thatwhen seen from this perspective, utilitarianism is not so much guilty of con-flating all persons into one, as of being too individualistic:

This assimilation of justice to a higher order executive decision, certainlya striking conception, is central to classical utilitarianism; and it alsobrings out its profound individualism, in one sense of that ambiguous word.

The Legacy of John Rawls56

It regards persons as so many separate directions in which benefits and bur-dens may be assigned; and the value of the satisfaction or dissatisfaction ofdesire is not thought to depend in any way on the moral relations in whichindividuals stand, or of the kinds of claims which they are willing, in the pur-suit of their interests, to press on each other.

Principles of justice, then, cannot take the distinction between persons ser-iously by ignoring their relationships. Turning to democratic societies, notethat what, ultimately, makes them democratic is the nature of the inter-subjective relationship that citizens form with one another, a relationshipthat Rawls characterizes as shaped by freedom and equality. If we were deter-mining principles of justice for an authoritarian regime, we might ask our-selves what sorts of institutional and legal reforms the dictator should bringabout, and then judge the dictator's justice by reference to a description ofthose reforms. If, however, the fundamental feature of a democratic society isthe nature of the intersubjective relationship among its citizens, and justiceis the first virtue of the institutions that constitute that relationship, then thequestion of justice for a democratic society must ultimately be the question ofthe nature of that relationship. And so a political theory ofjustice for a demo-cratic society had better take the intersubjective relationship among citizensas its primary focus.

We can also note that the choice of whether or not to adopt a form of argu-ment for principles ofjustice that takes the distinction between persons ser-iously will be influenced by one's position at this third level. If the business ofmorality is to bring something about, then it makes sense to find a means tofigure out what that something is. Nothing in that project need push us to takethe distinction between persons seriously, or regard it as a failure if we do not.On the other hand, if the business of morality is the nature of our intersubjectiverelationships, then there is not only a reason to take the distinction betweenpersons seriously at this third level, but a reason to make taking the distinctionbetween persons seriously at the first two levels a criterion of success, becausethose relationships are in part constituted by the principles we adopt to regu-late our conduct and the arguments we can offer one another to justify them.

Once we see that justice as fairness focuses on our intersubjective relation-ships as citizens, we can see why it is a mistake to see the veil of ignorance as asign that Rawls fails to take the distinction between persons seriously. I'll men-tion three considerations. First, the criticism that justice as fairness conflatesall individuals into one via the veil of ignorance fails to take sufficient heed ofthe fact that the original position is a device of representation, and not anactual event. Thus, the parties who are arguably conflated via the veil ofignorance are not the persons the distinction between which an adequatetheory ofjustice must take seriously. The original position is a device that

Taking the distinction between persons seriously 57

helps us to clarify our judgements about which arguments in favor of variousprinciples ofjustice are ones we can share with our fellow citizens. The point ofthe argument for justice as fairness, as we have just seen, is to work out the basisof a political agreement among actual citizens. This requires that the form ofthe argument for justice as fairness extends beyond the veil of ignorance toinclude considerations about whether actual citizens can agree to the twoprinciples, and so the failure of the veil to take the distinction between personsseriously need not call into question the ability of the argument as a whole totake that distinction seriously.

Second, even within the perspective of the original position, the parties areto consider whether or not a society well-ordered by its preferred principleswill be stable. One of the features of the society ofjustice as fairness thatmakes it stable is that individual citizens, pursuing a variety of conceptions ofthe good, and occupying a variety of social positions, can endorse the argu-ments in favor of the two principles ofjustice, and see that a society that fulfillsthese two principles will provide to a sufficiently high degree the social bases oftheir self-respect. These factors contribute to individual citizens developing asense ofjustice, and thus to the stability of the society. The stability of a well-ordered society, we might surmise, is precisely the result of the agreement thatshows that the argument for the two principles ofjustice takes the distinctionbetween persons seriously at the second level. From the perspective of the con-trast drawn at the third level, where we learned that for justice as fairness, suchagreement is what justice is fundamentally about, we can thus see that thequestion of stability rather than the question of preferability from behind theveil of ignorance must form the core of Rawls's argument.

Third, as I said at the beginning of my remarks, I think we can understandthe development of Rawls's thinking about justice as fairness over the course ofhis career as stemming from an increasing recognition of what is necessary inorder fully to take the distinction between persons seriously. Thus, we shouldnot be surprised to see in his later writings, from Political Liberalism onward,a de-emphasis on the choice aspects of the original position argument, and agreater emphasis on the questions of actual agreement among actual citizens,of the sort that constitutes an overlapping consensus or which is needed tosatisfy the liberal principle of legitimacy.

Even though the fact that justice as fairness takes the distinction betweenpersons seriously at the third level shows that it does so to a sufficient degreeat the second level, we can understand a third line of criticism as suggestingthat justice as fairness fails to take the distinction between persons sufficientlyseriously at this third level. This worry comes out most clearly in criticisms ofRawls for being insufficiently democratic such as that of Jiirgen Habermas,or from those skeptical of the very idea of moral theory, such as BernardWilliams. What unifies these seemingly different lines of criticism is the

The Legacy of John Rawls58

claim that although Rawls focuses on the agreement of citizens as the centraldomain of political philosophy, these citizens are merely the imaginary con-structs of a political theorist. When it comes to real distinctions between realcitizens in the actual world, the criticism goes, the very fact that Rawls hasdeveloped a theory of justice shows that he has failed to take these distinctionsseriously. As Habermas puts the point:

The higher the veil of ignorance is lifted and the more Rawls's citizens takeon flesh and blood, the more deeply they find themselves subject to princi-ples and norms that have been anticipated in theory and have alreadybecome institutionalized beyond their control. In this way, the theorydeprives the citizens of too many of the insights that they would have toassimilate anew in each generation ... It is not possible for citizens to viewthis process as open and incomplete, as the shifting historical circumstancesnonetheless demand. They cannot reignite the radical democratic embers ofthe original position in the civic life of their society, for from their perspec-tive all of the essential discourses of legitimation have already taken placewithin the theory; and they find the results of the theory sedimented in the

24constitution.

The problem Habermas points to here is that in developing a theory of jus-tice, Rawls seems to have already done all the work that democratic citizensare supposed to do themselves in coming to agreement. In doing so, he fails totake the distinction between persons seriously, because he fails to leave openand incomplete the fundamental shape of the agreement that distinct citizenswill achieve. In the terminology of this chapter, Habermas here claims thatRawls fails to take the distinction between persons sufficiently seriously atthis level because, as a theorist of justice, he stands outside the ongoing rela-tionships of citizens to one another.

The problem to which Habermas points, then, is not so much with the con-tent of the principles of justice, or the arguments for them, or even their self-understanding as constituting the relations of free and equal citizens to oneanother, but rather with the way, as he understands it, Rawls conceives therole of political philosophy itself. Understanding Rawls's reply to Habermason this point, then, requires moving up to the final level at which it is impor-tant to take the distinction between persons seriously: the conception of politi-cal philosophy.

In order to clarify the contrast at the fourth level, we can start by thinkingabout how what I have elsewhere called a "theoretical approach" to political

Taking the distinction between persons seriously 59

Level 4: the conception of political philosophy

philosophy fails to take the distinction between persons seriously. Those whotake a theoretical approach to political philosophy hold that it is the reasoningof the theorist that ultimately determines the shape of just political principles,and so their justification depends entirely on the soundness of the theorist'sreasoning. The modus operandi of a theoretical approach is to develop a theoryand then apply it. Utilitarianism stands as the paradigm case of such anapproach. Utilitarians claim to have a correct account of value or the good,and then go on to draw inferences from that account about what sorts of beha-vior or institutional structure would make human lives go better or worse.This self-understanding leaves two tasks for the utilitarian political or moralphilosopher. First, she can work at refining or reforming the theory, provid-ing a more precise or better account of utility or its sources in human life, oroffering a novel way to understand how to aggregate it into a single maxi-mand. Second, she can apply the theory to real-life moral or political ques-tions. Such application need not be straightforward or routine. Think, forinstance, of Mill's argument that utilitarianism supports individual rightsin Chapter 5 of Utilitarianism, or Rawls's argument that utilitarians can sup-port the practices of promising and punishment in his "Two Concepts ofRules." Nevertheless, what is important for my purposes here is that neitherendeavor requires engaging with the actual people who are going to be subjectto the theory or its applications. The utilitarian theorist acts as if her job is towork out a correct theory and figure out how to apply it, rather than to par-ticipate in the process of coming to reasonable agreement with her fellowcitizens. She takes up what Rawls describes as an administrative view-point, rather than the perspective of a citizen. In doing so, she fails to treatthe distinction between persons seriously because she fails to take seriously thevarious and sundry grounds that different people might have for objecting,not merely to the correctness of her theory, but to her presumed intention toimpose it on them.

Habermas's remarks quoted above help us to see why such inattention to thedistinction between persons at this fourth level poses a problem for a concep-tion of justice for a democratic society. Within a democratic society, theauthority to determine principles of justice lies always in the hands of citizens,not in the hands of princes or bureaucrats or philosophers. Thus, doing politi-cal philosophy within and for a democratic society requires abandoning theperspective of the theorist favored by utilitarians and many other political phi-losophers and adopting the perspective of the citizen.

That is, however, what I think Rawls does, and to my mind this aspect of hisphilosophical project represents his most important break with utilitarianism.Rawls describes political liberalism as applying the principle of toleration tophilosophy itself, aiming as it does to reconcile different comprehensive doc-trines without deciding which among them might be true. Similarly, we might

60 The Legacy of John Rawls

describe Rawls's approach to philosophy as applying the principle of contrac-tualism to the activity of philosophy itself. Contractualism takes agreement tobe, in Scanlon's words, "what, at a fundamental level, morality is about."What I am suggesting here is that we understand Rawls as taking reasonableagreement to be what, at a fundamental level, political philosophy is about.

Moral and political philosophy, then, are forms of moral and politicalactivity, in both an ennobling and a deflationary sense. In an ennoblingsense because philosophers are not just sitting on the sidelines spouting ulti-mately irrelevant theory, but interacting with one another and other non-philosophers in ways that help to constitute the very relationships that arethe subject matter of morality and politics. In a deflationary sense becauseonce philosophers leave the sidelines, we have no special status or authority.As Rawls puts the point, "Injustice as fairness, there are no philosophicalexperts. Heaven forbid!" Note that this conception of philosophy's role isnot a late addition to Rawls's work, prompted by the criticisms of Habermasand others. It is a very long-standing characteristic of his philosophy, whichcan be summarized in the claim that justification is always justification to aparticular other.

Once again, the choice to take the distinction between persons seriously atthis level affects how we view the importance of doing so at the third level.If our job as political philosophers is to offer a correct theory, and we take thatto be a worthwhile (and not merely an academic) exercise, it is natural to con-clude that the value of such an enterprise is specifying what it is that moralitytells us to bring about, and thus that the business of morality is to bring some-thing about. Similarly, if we take the business of morality to be to bring some-thing about, it is natural to give the job of figuring out what that something isto moral and political philosophy conceived as an enterprise of theory con-struction. If, however, we are fundamentally interested in the intersubjectiverelationships we form with one another, we cannot hope to form the appropri-ate ones by ignoring or denying our place within those relationships, as thetheorist implicitly does. We are thus led to find a way to do political philoso-phy that does not involve taking a theoretical approach. Such an attempt is,however, fraught with its own dangers. Chief among these is that it may seemto lead to the conclusion that there is really nothing for political philosophy todo. That conclusion strikes me as too hasty, and I think we can see in Rawls amodel of what there is to do and how to do it.

Let me conclude, then, by suggesting three distinct but related roles for politi-cal philosophy (and philosophers) to play on this conception. The first is the

Taking the distinction between persons seriously 61

Climbing back down: going on in a Rawlsian way

construction of what Rawls calls arguments in public reason. Within this role Iinclude not only formulating arguments for particular policies and principles,but also evaluating the arguments made within the broader public sphere andthe narrower political sphere in terms of their reliance on public reasons.While this role for political philosophy suggests and will include such focusedinterventions as the evaluation of various arguments for and against therecognition of same-sex marriages, it can also include much larger questions,such as the choice between what Rawls calls "welfare-state capitalism" and"property-owning democracy." After Rawls says that injustice as fairnessthere are no philosophical experts, he goes on to point out that "citizensmust, after all, have some ideas of right and justice in their thought and somebasis for their reasoning. And students of philosophy take part in formulatingthese ideas but always as citizens among others." Note that such interven-tions into actual political debate should not be understood as the applicationof a theory. When political philosophers endeavor to make or evaluate argu-ments in public reason, we are addressing those arguments to our fellow citi-zens as equal participants in a shared endeavor. The success of our efforts isto be measured by the response from our fellow citizens, and not by a standardof correctness internal to the discipline of philosophy.

Second, in some cases, the problems with a democratic society will rundeeper than disputes about particular policies or the fine-tuning of principles.The deliberation of citizens can generate legitimate political authority only ifthat deliberation involves the real engagement of free and equal citizens.Thus, there are times when what appear to be perfectly good public reasonarguments should be rejected because part of what makes them attractive is abackground of systematic injustice that undermines the very possibility of rea-sonable deliberation. Faced with such injustices, citizens need to make rathermore complex public reason arguments, arguments that lay out the precondi-tions for political deliberation to be legitimacy conferring. Such arguments,turning as they do on conditions of possibility and more fully developedaccounts of legitimacy and justice, are ones to which political philosophy iswell-positioned to contribute. We can see an example of such an argument inRawls's discussion of campaign finance in Political Liberalism, and the bases forfurther such arguments in his remark in "The Idea of Public Reason Revis-ited" that public reason relies on a criterion of reciprocity that holds that

when those terms are proposed as the most reasonable terms of fair coopera-tion, those proposing them must also think it at least reasonable for others toaccept them, as free and equal citizens, and not as dominated or manipu-lated, or under the pressure of an inferior political or social position.

Third, citizens of particular democratic societies will, at various times,face problems for which the conceptual tools of philosophy will be needed.

The Legacy of John Rawls62

The problem to which Political Liberalism addresses itself is what Rawls, follow-ing Kant, describes as a crisis of reasonable faith. Very roughly, citizens in apluralistic society might come to doubt whether a constitutional democraticregime is possible given that citizens cannot be brought to agree on funda-mental matters without the use of oppressive force. Faced with such a crisis offaith, we may find ourselves unable to muster the commitments and efforts atcompromise and self-sacrifice necessary to make such a pluralistic democracywork. So this lack of faith is a political, not merely a philosophical, problem.Nevertheless, a large part of its solution lies within the conceptual domain ofphilosophy, insofar as our faith can be restored by a philosophical demonstra-tion of the conceptual coherence of a pluralistic democracy. Rawls describesthis role for philosophy as "philosophy as defense." Note that engaging in phi-losophy as defense may require the construction of rather elaborate and com-plex theoretical structures, such as those of justice as fairness and politicalliberalism. Nevertheless, the point of such theoretical construction is not thatof the philosopher as theorist, concerned to work out the correct principles ofjustice, as it were once and for all, but the philosopher as citizen, concerned toaddress his own and his fellow citizens' political predicaments.

All three of these roles for political philosophy are responsive to the par-ticular political conditions in which we live. If we understand political phi-losophy as I have argued Rawls came to understand it, then what there is forpolitical philosophers to do in this vein will depend to a very large degree onwhat we take to be the problems we face as citizens: where we find agreementin our societies fragile or non-existent, where it seems that the breakdown ofsuch agreement matters, and what sorts of crises of faith we face. If you areused to thinking of political philosophy in a more Platonic guise, as are mostpolitical philosophers and theorists, you may find this interpretation of ourwork rather disappointing, as if Rawls has condemned us to a series of small(and thus minor) tasks. Such a reaction would be a mistake, however. To seewhy, we only need to note that it was to accomplish just such narrow tasks thatRawls wrote arguably the two most important books on the subject of politicalphilosophy published in the last 50 years, books that no one would describe assmall. The size of a philosophical task is not wholly determined by its scope.Among John Rawls's many philosophical strengths, one of the greatest andleast appreciated was his capacity to see just how difficult apparently smalltasks can be. As I have argued, central among these is taking the distinctionbetween persons seriously.

Taking the distinction between persons seriously 63

Notes

1. TJ, 24/original edn, 27.2. Ibid., xvii—xviii/original edn, vii—viii.

3. Plato, TheRepublic (trans. G.M. Grube; rev. C.D.C. Reeve; Indianapolis: Hack-ett, 1992), 6 (Book I, 332al).

4. TJ, 5/original edn, 5.5. See, in particular, the "Introduction to the Paperback Edition" of PL; RH; and

PRR reprinted in CP.6. TJ, 23/original edn, 26.7. Robert Nozick, Anarchy, State, and Utopia (New York: Basic Books, 1974), 149-64,

183—231, esp. 228. Amartya Sen makes a somewhat different criticism of Rawls atthis level for his use of primary goods as an index of equality. Sen claims that thisindex fails to take seriously not merely the differences between people but the dif-ferences in those differences. (Amartya Sen, Inequality Reexamined (Cambridge,MA: Harvard University Press, 1995).)

8. TJ, 23-4/original edn, 26-7.9. Ibid., 24/original edn, 27.

10. Ibid., 25/original edn, 28.11. Bernard Williams, Ethics and the Limits of Philosophy (Cambridge, MA: Harvard

University Press, 1985), 108-10.12. A similar invocation of Rawls's concern to treat the distinction between persons

seriously at this second level can, I think, also provide a starting point for replyingto Sen's criticism. Such a reply would invoke the role of considerations of stabilityin the argument for the two principles, and point out that the argument for thoseprinciples needs to be acceptable to all citizens given their various differencesfrom one another, and so the index in which inequalities are measured withinthe two principles must be similarly acceptable. Of course, showing that primarygoods fits this criterion within the confines of justice as fairness better thanSen's capabilities would require further arguments that are beyond the scope ofthis chapter.

13. TJ, 25/original edn, 27.14. See, for instance, Jean Hampton, "Contracts and Choices: Does Rawls Have a

Social Contract Theory?" Journal of Philosophy 11 (1980), 315-38.15. T.M. Scanlon, "Contractualism and Utilitarianism," in A. Sen and B. Williams

(eds), Utilitarianism and Beyond (Cambridge, UK: Cambridge University Press,1982), 103-28.

16. Christine Korsgaard, "The Reasons We Can Share," in her Creating the Kingdom ofEnds (Cambridge, UK: Cambridge University Press, 1996), 275-310 (275); ori-ginal emphasis.

17. Scanlon, "Contractualism and Utilitarianism," 128.18. This description is intentionally vague so as not to require commitment to a par-

ticular view of what is essential to being a person, or what constitutes setting acausal chain in motion.

19. I am grateful to Tamar Shapiro for pressing me to be clearer about the specialnature of the relationships in question and their connection to taking the distinc-tion between persons seriously.

20. John Rawls, "Justice as Reciprocity," reprinted in CP, 190-224 (218).

The Legacy of John Rawls64

21. Even this line of reply concedes too much to the original objection, however.

There are good reasons to see even the parties in the original position as distinct,

if not different, and thus engaged in coming to an agreement rather than making

a choice. (See, for instance, my "Games, Fairness and Rawls's^4 Theory of Justice,"Philosophy and Public Affairs 20 (1991), 189-222.)

22. One might ask whether this shift in emphasis represents a change in doctrine or

merely a change in presentation. I tend to think it is the latter, inspired by what

Rawls took to be misreadings of the structure of the argument in Theory. Many

readers of Theory took it to be attempting to derive principles ofjustice from prin-ciples of rationality. Such an interpretation leads one to read the original position

as the central feature of the argument for the two principles and to read that part

of the argument as an argument within rational-choice theory. But those are mis-

takes. I say more about this in my "The House that Jack Built: Thirty Years of

Reading Rawls," Ethics 113 (2003), 367-90.23. Jiirgen Habermas, "Reconciliation through the Public Use of Reason: Remarks

on John Rawls's Political Liberalism," Journal of Philosophy 92 (1995) ,109-31; Wil-

liams, Ethics.24. Habermas, "Reconciliation," 128; original emphasis.

25. I use the term "theoretical approach" and argue that Rawls does not adopt suchan approach in my Reasonably Radical: Deliberative Liberalism and the Politics of Iden-

tity (Ithaca, NY: Cornell University Press, 2001), 15-17.

26. John Stuart Mill, Utilitarianism, ed. Roger Crisp (Oxford, UK: Oxford University

Press, 1998). John Rawls, "Two Concepts of Rules," reprinted in CP, 20-46.

27. For a vivid example of this lack of connection between the question of what makes

the theory correct and what would lead people to endorse it, see Mill's Utilitarian-ism, where he completely separates the question of the proof of utilitarianism

(its correctness) from those regarding its "sanctions" (what would lead people to

adopt it).

28. Rawls, "Justice as Reciprocity," 216-18.29. Astute readers will notice the addition of "reasonable" to the contractualist for-

mula. Although I take it that Scanlon, too, means to speak of reasonable agree-ment and not any agreement, and so this does not represent a substantive

difference with Scanlon's formula, I here make the reasonableness condition

explicit to mark a difference between the view I am attributing to Rawls and aform of conventionalism or relativism that held that any agreement is as good as

any other, and all are decisive to questions of political philosophy.

30. RH, 427.

31. See TJ, original edn, 580—1, but see also Rawls, "Outline of a Decision Procedure

for Ethics," reprinted in CP. Burton Dreben has perhaps most clearly recognized

the difference between Rawls and utilitarians (and to Dreben's mind, most otherphilosophers) at this level. (Burton Dreben, "On Rawls and Political Liberal-ism," in S. Freeman (ed.), The Cambridge Companion to Rawls (Cambridge, UK:

Cambridge University Press, 2003), 316—46) But Dreben reads this as a radicalnew direction undertaken in Political Liberalism and the work that followed it,

Taking the distinction between persons seriously 65

rather than, as I would argue, the working out of certain fundamental commit-ments that have been there all along. One reason Dreben misses this continuity isthat he fails to appreciate the motivations for Rawls's approach, motivations thathave as much if not more to do with his conception of democracy than with hisconception of philosophy. If we see that Rawls's conception of political philoso-phy is the consequence of his conception of democracy, then it is easier to see itspredecessors in the social contract tradition. Most notable here is Rousseau'sremark that he writes from the perspective of a citizen, and Kant's remark that"reason has no dictatorial authority; its verdict is always simply the agreement offree citizens." (Immanuel Kant, Critique of Pure Reason, trans. N.K. Smith (NewYork: St. Martin's Press [1781/87] 1929), 593 (A738/B766).)

32. Though these terms do not appear in Theory, Rawls claims in later work that jus-tice as fairness is consistent with property-owning democracy, but not with wel-fare-state capitalism. His most detailed discussion of the difference between thesetwo systems is in JF, 135—52.

33. PL, 427.34. PRR reprinted in CP, 578. I think much of Catharine MacKinnon's more theore-

tical work can be read as engaged in just such arguments, laying out the means bywhich the structural inequalities of a patriarchal society serve to silence the voicesof women in political deliberation. (See my "Radical Liberals, Reasonable Fem-inists: Reason, Power and Objectivity in the Work of MacKinnon and Rawls,"Journal of Political Philosophy 11.2 (2003), 133—52, and, more generally, my Reason-ably Radical, esp. 131-58.)

35. I presented an earlier version of this paper at a symposium on "Rawls, Utilitar-ianism and Social Policy," at the Central Division meetings of the AmericanPhilosophical Association in April 2004. I benefitted greatly from the discussionon that occasion, and in particular from questions from Alyssa Bernstein, RahulKumar, and Henry Richardson, which have led me to clarify several points in thischapter. I would also like to thank Samuel Fleischacker, David Owen, andTamar Schapiro for their comments on earlier drafts, and Thorn Brooks for theinvitation to publish this here.

The Legacy of John Rawls66

4

Rawls and feminism: What should feminists make of liberal neutrality?

Elizabeth Brake

Introduction

Among feminist philosophers, there has been substantial debate over theimplications of Rawls's liberalism for women's equality. Some feminists, suchas Susan Moller Okin, have accepted the fundamental structure of Rawls'sliberalism, but have criticized Rawls himself for ignoring gender and remain-ing ambiguous regarding the status of the family, the structure of which affectswomen's life-chances: "Rawls's theory neglect [s] .. . gender." As Okin writesin her reply to Political Liberalism, Rawls did not clarify in that work whetherhe endorsed substantive or merely formal equality for women. In her conclud-ing words, she asks, "what does Rawls mean to say about justice between thesexes?" However, Okin herself argues that Rawls's theory ofjustice, if appliedto the social structures that perpetuate women's inequality, has great poten-tial for changing those structures; she accepts Rawls's theory (once it is madesensitive to gender) as a suitable theory of feminist justice.

Other feminists, however, criticize liberalism as an ideology whose promiseof equal rights obscures the mechanisms of oppression. Notably, CatharineMacKinnon has argued that liberal freedoms serve male power and obscurethe extent of women's subordination. For example, freedom of speech hasbeen used to protect pornography, which MacKinnon argues is harmful towomen; this freedom, as MacKinnon sees it, protects the interests of men pre-cisely where those interests are at odds with women's, while it appears to haveno gender bias (since freedom of speech is every citizen's right, regardless ofgender). In MacKinnon's view, the supposedly objective, neutral state is infact "male." By this she means not only that law-makers and justices applysupposedly gender-neutral laws in sexist ways, but also that those supposedlygender-neutral laws are themselves subtly, but powerfully, discriminatory.The best example of this (historically) is the family: the area which liberalismhas protected as private has been precisely the site of women's oppression.Given the unequal balance of power between men and women, protectingthe family sphere from judicial scrutiny masked injustices. Again, in MacKin-non's view, it is not coincidental that protected freedoms (such as speech)coincide with the areas in which women's inequality is now, in her view,

maintained. One of MacKinnon's central points, drawn from Marxist theory,is that because many choices are products of oppressive conditioning, liberalfreedom of choice will perpetuate oppression.

MacKinnon's view is in contrast to Okin's liberal feminism, for while Okinalso recognizes that social conditioning and free individual choices perpetuatesexual inequality, she argues that while "Rawls's Theory ... does not discusssuch injustices of gender, [it] has great potential for doing so." Martha Nuss-baum too has defended the project of liberal feminism against MacKinnon'scritique. A crucial difficulty with MacKinnon's view is that her critique ofliberalism seems to depend on a denial of freedom. Okin and Nussbaum —like many feminists—are also concerned with how social pressures shapewomen's choices, but MacKinnon goes much further in suggesting thatwomen's choices cannot be truly free within a patriarchal society. In thischapter, I argue that Rawls's liberalism can indeed be an ally for feminism.But in contrast to Okin and Nussbaum, I focus primarily on the issue of liberalneutrality, a topic suggested by MacKinnon's work. While neutrality is oftentaken to be at odds with feminism—since, for example, it seems feminist edu-cation in schools would conflict with it—I argue that feminists should wel-come neutrality as a moral ideal in the process of justification, and thatneutrality itself will require substantive feminist reform.

From MacKinnon's perspective, liberal neutrality is a deceptive fiction,concealing the state's patriarchal bias. In contrast, I argue that the liberalaspiration to neutrality supports feminist goals. I discuss two kinds of neutral-ity: neutrality at the level of justifying liberalism itself, and state neutrality inpolitical decision-making. Both kinds are contentious within liberal theory.Rawls's argument for justice as fairness has been criticized for non-neutralityat the justificatory level. Rawls himself, in Political Liberalism, noted this prob-lem in A Theory of Justice. I will defend a qualified account of neutrality atthe justificatory level, arguing that the exclusion of certain creeds from thejustificatory process is justified. I then argue that the justification process Idescribe offers a justificatory stance supportive of the feminist rejection ofstate-sponsored gender hierarchy. Further, I argue that liberal neutrality atthe level of political decision-making will have surprising implications forgender equality. While—as Okin avows—Rawls himself may have remainedambiguous on how to address gender inequality, his theory implies that thestate must seek to create substantive, not merely formal, equality.

Liberal neutrality is the doctrine that the state should remain neutralbetween competing conceptions of the good, where an individual's conception

The Legacy of John Rawls68

Liberal neutrality

of the good is whatever plan of life she has, subject to certain rational

constraints. Such conceptions may include, for example, commitment to reli-

gious beliefs, or to feminism, or to the traditional gender-structured family.

Neutrality has been taken up widely by liberal theorists. For example,

Ronald Dworkin has denned liberalism as the view that the government must

"treat its citizens as equals" and argued that this requires that "political

decisions must be, so far as is possible, independent of any conception of

the good life, or of what gives value to life." Thus, the foundational liberal

tenet of moral equality directly implies liberal neutrality. Will Kymlicka

writes: "A central feature of contemporary liberal theory is its emphasis

on 'neutrality'—the view that the state should not reward or penalize par-

ticular conceptions of the good life but, rather, should provide a neutral

framework within which different and conflicting conceptions of the good can

be pursued."13

The doctrine of neutrality is often thought to be problematic, at many

levels. First, some liberals have rejected the aspiration to neutrality between

conceptions of the good, arguing that a commitment to moral equality need

not imply a commitment to seeing all ways of life as comparable. Equal respect

for individuals, on this view, does not require equal respect for whatever ends

and values they possess. An argument associated with this view is that liberal

neutrality involves an impoverished view of human life and human possibility

and is likely to produce citizens in thrall to such a view. I will not address this

criticism further here. Second, neutrality in political decision-making appar-

ently presents serious impediments to such decision-making. For example,

H. Tristram Engelhardt has argued that liberal neutrality conflicts with

nationalized healthcare. He points out that such a system must either provide,

or not provide, services such as abortion, assisted reproductive technologies,

and physician-assisted suicide. Either way, he argues, it will privilege some

comprehensive doctrines over others: for instance, a system providing abor-

tion will privilege doctrines that see it as a legitimate medical procedure over

those which see it as morally impermissible. I will return to this issue in the

final section of this chapter. A third problem is that liberal neutrality seems to

require that the justification for liberalism also be neutral, that is, that instead

of basing liberalism on a controversial conception of human nature or of the

good (as Mill arguably did), its defender must rest his or her case on principles

that can be accepted by all reasonable persons. This issue is the subject of an

important change that Rawls made in Political Liberalism to the view he

defended in A Theory of Justice.

The background to this change is that in Theory, Rawls had defined a poli-

tical system as stable when it motivates its citizens to act according to its prin-

ciples of justice: "One conception of justice is more stable than another if the

sense of justice that it tends to generate is stronger and more likely to override

What should feminists make of liberal neutrality? 69

disruptive inclinations and if the institutions it allows foster weaker impulsesand temptations to act unjustly." To demonstrate the stability of justice asfairness, Rawls tried to show that citizens would be motivated to act justlywithin a system regulated by the principles of justice. To this end, he arguedthat the "disposition to take up ... the standpoint of justice accords with theindividual's good," giving as a reason for this "the Kantian interpretation [ofthe theory of justice]: acting justly is something we want to do as free and equalrational beings." But this argument employs a premise about the nature ofhuman good. It gives a Kantian conception of the good as a comprehensivedoctrine, one in which acting justly accords with an agent's good. Rawlsdefines a "comprehensive doctrine" as a theory of value which applies to awide range of subjects, such as "what is of value in human life, .. . ideals ofpersonal character, as well as ideals of friendship and of familial and associa-tional relationships." Liberal neutrality, however, requires neutralitybetween comprehensive doctrines. Thus, since assuming a Kantian compre-hensive doctrine conflicts with neutrality between comprehensive doctrines,Rawls rejects this move in Political Liberalism.

Not only did the account of stability illicitly employ a Kantian conceptionof the good as a comprehensive doctrine, but moreover, Rawls writes in Politi-cal Liberalism, Theory treated justice as fairness itself as a comprehensive doc-trine. Rawls had there presented a conception of a "well-ordered society" inwhich "everyone accepts and knows that the others accept the same principlesof justice, and the basic social institutions satisfy and are known to satisfy theseprinciples` However, the idea of a society in which "everyone accepts ...the same principles of justice" conflicts with neutrality if justice as fairnessis understood as a comprehensive doctrine. Rawls attempts to correct theseproblems in Political Liberalism by adjusting the scope of justice as fairness;it is not comprehensive, but narrowly political. As such, it is the possiblesubject of an overlapping consensus between various reasonable comprehen-sive doctrines—"religious, philosophical and moral"—which can agree to theregulatory principles of justice. Justice as fairness, as political, applies onlyto the basic structures of society and can slot into the various comprehensivedoctrines found therein.

In making these changes, Rawls was in part motivated by the circumstancesof contemporary American society, especially the deep divisions over religion.In an interview, he explained the focus on religion in Political Liberalism asmotivated by his concern "about the survival, historically, of constitutionaldemocracy ... the problem is how do you see religion and comprehensive secu-lar doctrines as compatible with and supportive of the basic institutions of aconstitutional regime." Nevertheless, it is not clear either that the changesmade in the later work will increase the appeal of his theory for those already

The Legacy of John Rawls70

holding incompatible comprehensive doctrines or that Rawls has made aplausible case for the stability of justice in a society deeply divided overquestions of value. Some have suggested that the original account makes amore convincing case for the stability of the principles. Moreover, PoliticalLiberalism's restriction of the scope of justice continues to fail to be neutralbetween all comprehensive doctrines.

First, the argument for justice as fairness depends on a claim of moral equal-ity. Rawls claims that human beings are morally equal in virtue of possessingthe potential "for a conception of the good . . . [and] for a sense of justice."But the features of individuals which Rawls picks out as constitutive of moralequality are themselves not neutral, because they reflect a conception of whatis important about human beings. Justice as fairness is not neutral in onerespect in which it claims to be neutral because it privileges one conceptionof the good, that is, one in which the individual's plan of life is an object ofrational choice. Someone who believes that autonomy is not especially impor-tant, or that humans are not equal in more important respects, might resistthis claim. Thus, Rawls's derivation of the principles ofjustice appears illegiti-mately to ignore competing conceptions of the good. The response that neu-trality is not foundational but derived from the ideal of moral equality appearsto beg the question.

Distinguishing between political and comprehensive doctrines, as Rawlsdoes in Political Liberalism, does not meet this objection, because such adistinction is itself controversial. For by assuming that the political concep-tion of the individual and its associated model of moral equality are thebasis for defining political principles, Rawls has ignored comprehensiveconceptions which would model political principles on alternative concep-tions of the individual and his or her morally significant features. Forinstance, a comprehensive doctrine might simply deny that the politicalsphere is separable: examples of religions that would base law on religiousteachings spring readily to mind. Thus the religious believer who seeks unityof Church and state, or a legal system based on the Old Testament, mightrespond to Rawls that his restriction of the scope ofjustice does not make thetheory either neutral or acceptable to the believer. Or, a doctrine mightallow such a distinction, but carve it differently: some feminist analyses ofoppression are an example (consider the claim that "the personal is politi-cal"). Thus, MacKinnon might say in response to Political Liberalism that thedistinction between comprehensive doctrines and the political implicitlyexcludes feminist considerations (which take issue with the content of variouscomprehensive doctrines and their effects in society) from justice as fairness.In the next section, I turn to the question of the extent to which such a responseis justified.

What should feminists make of liberal neutrality? 71

In the next section I will argue that neutrality at the level of political decision-

making can serve feminist goals. Here I will argue that the restriction of

justice as fairness to the political is less fruitful; moreover, neutrality at the

level of theory justification must be qualified as "skeptical" or "agnostic,"

not pluralist. Such a procedure derives the ideal of moral equality and the

policy of political neutrality (in decision-making by the liberal state) from

the absence of any justification for unequal treatment. The procedure I will

sketch is morally neutral—since it presupposes no claims about the good —

but uses an epistemic argument to exclude doctrines that deny equality.

It retains neutrality at the justificatory level only in a qualified manner. While

some illiberal comprehensive doctrines must be tolerated, and their expres-

sion protected, in a liberal state, they can be ignored in the derivation of

the principles governing such a state. My account is closer to Theory, which

I see as more compatible with a feminist perspective in this respect, than to

Political Liberalism.

From a feminist perspective, it may be a mistake not to see justice as a com-

prehensive doctrine. Recall that a comprehensive doctrine applies to "what is

of value in human life, . . . ideals of personal character, as well as ideals of

friendship and of familial and associational relationships." According to

most feminist views, justice should regulate these spheres—especially that of

the family. Justice is not only a virtue of political institutions but also a

virtue within the apparently private spheres of family and civil society.

Some qualifications are important. The scope of justice should be distin-

guished from the question of the legitimate extent of state interference in indi-

vidual lives. The state might promote, but not enforce, the ideal of justice in all

spheres of life. Also, justice is not the only virtue. Principles of justice can be

supplemented with other comprehensive doctrines. Third, the content of prin-

ciples of justice may be different at micro and macro levels (for example, the

difference principle might not be the relevant principle of distribution withinthe family). But the underlying theory of justice, deriving from the ideal of

moral equality and equal respect, should be consistent from the macro to the

micro level.

From the liberal feminist viewpoint, in which justice should apply to the

family, the restriction of justice to the political makes the theory less attrac-

tive. From the Marxist feminist viewpoint, restricting justice to the political

begs the question of how the political ought to be defined (for example, if the

political is defined as any relationship characterized by a dynamic of power,

Rawls's distinction is incorrect). And within liberal theory, the restriction to

the political does not seem to offer a significant advantage. As I have indicated

above, it does not speak to those whose alienation from the theory Rawls

The Legacy of John Rawls

Feminism, justificatory neutrality, and William Clifford

72

sought to overcome, such as religious fundamentalists. However, allowingjustice to be comprehensive conflicts with neutrality at the level of politicaldecision-making, so I will defer this question.

To return to the issue of justification, I think a defender of Rawls can givegood epistemic reasons for rejecting certain creeds from consideration in thejustification procedure, creeds which, for instance, would define humans asunequal or maintain that the Church ought to be the highest arbiter in politi-cal affairs. Rawls restricts himself to being neutral among reasonable views,but he allows religious views to be counted as reasonable. I will suggest thata stronger account of rational constraints on belief shows why it is legitimatefor certain conceptions of the good to be disregarded at the level of politicaljustification. Of course, this may not appeal to the believer, but, as I havepointed out, Rawls's more conciliatory approach is unlikely to, either. Thisapproach will illuminate the appeal of Rawls's liberalism to feminism, forit begins with an ideal of moral equality incompatible with views of genderhierarchy. Indeed, most sexist and racist views (and other forms of illegiti-mate discrimination) are excluded from the justificatory process on the view Iwill describe.

In an 1877 paper, William Clifford argued that it was ethically wrong tohold religious beliefs, or at least, to hold any beliefs "without sufficient evi-dence." Clifford argued that there is a normative requirement to evaluateour beliefs: "it is wrong always, everywhere, and for any one, to believe any-thing upon insufficient evidence." Clifford gives an example of a ship-ownerwho sends his ship to sea sincerely believing in its sea-worthiness although helacks sufficient evidence for this belief. The ship-owner is guilty of the deaths ofthe passengers when the ship sinks. Even had it not sunk, Clifford adds, hewould still have been guilty, for he had no right to believe as he did. Cliffordaims to establish a duty to doubt on several grounds, both consequentialistand deontological.

Clifford's consequentialist arguments are interestingly resonant with dis-cussions of feminism, liberalism, and stability. Where feminists warn thatbelief in gender hierarchy has subtle effects on women's actual status, Cliffordsomewhat confusingly paraphrases Jesus: "He who truly believes that whichprompts him to an action has looked upon the action to lust after it, he hascommitted it already in his heart." Believing without warrant leads theincautious believer down the wide plain road of epistemological vice. Sincebelief is not "a private matter" but helps create society, vicious habits ofbelief infect posterity. The person who believes without warrant makes her-self credulous, and thereby does a "great wrong towards Man," for (like Millin On Liberty), Clifford claims that if society becomes a den of the easily con-vinced, it is poised to "sink back into savagery." These arguments take thecritical examination of beliefs as instrumentally valuable.

What should feminists make of liberal neutrality? 73

Clifford also suggests that "the faculty of belief" is valuable for its own sakeand that failing to reflect critically on our beliefs fails to respect it: "Belief, thatsacred faculty which prompts the decisions of our will, and knits into harmo-nious working all the compacted energies of our being, is ours not for ourselves,but for humanity." He castigates credulity, like lying, as a failure to reverethe truth:

Habitual want of care about what I believe leads to habitual want of care inothers about the truth of what is told to me . . . It may matter little to me,in my cloud-castle of sweet illusions and darling lies; but it matters much toMan that I have made my neighbors ready to deceive. The credulous manis father to the liar and the cheat.

A credulous citizenry is, indeed, perhaps as great a threat to the stability of aliberal state as a quarrelsome but critical one. However, I am not equipped toevaluate these causal claims. I want to investigate the following ideas, sug-gested by Clifford, as relevant to Rawls's justification procedure: credulity,belief on insufficient evidence, is irrational (a bad strategy), and beliefsformed without sufficient evidence are unreasonable and lacking in respectfor humanity.

First, credulity is a bad strategy, both for the individual and the group, inpolicy debate. When discussion is undertaken for some purpose—such aschoosing principles of justice—a rule of honesty is rationally justified sincethe knowledge that one's interlocutors may be lying will make progress muchmore difficult. But credulity will also undermine debate by making truth-telling unnecessary. (A rule against credulity might also be thought of as partof a Habermasian discourse ethic.) Someone might respond that lying is asophisticated and time-saving form of communication; honesty is not alwaysthe best policy. But even if lying might be rational, believing too easily wouldnever seem to be a good strategy in debate. In some cases, believing somethingwe ought not to believe (on purely epistemic grounds) could be the best strat-egy (for example, when the belief will motivate us to action). But while credu-lity might be rational when one chooses it to motivate oneself, the dangers ofindiscriminate credulity in debates of consequence clearly outweigh the possi-bilities of benefit. In most human interactions, doubt is rational. Though theremay be isolated exceptions, doubt as a habit is rational, for credulity letsreason linger, in paraphrase Clifford, in cloud-castles of illusion, where itskeenness may atrophy in the delights of darling lies. One might, therefore,justify the exclusion of the beliefs of the credulous from the original position byintroducing a rule of doubt.

At the more fundamental level of justification, Clifford's suggestion thatcredulity is an epistemological vice explains why beliefs without sufficient

The Legacy of John Rawls74

evidence are not reasonable, that is, why they fail to respect the process ofthe giving of reasons. Credulity eases the distinction between the true and thefalse. Clifford suggests that the truth demands reverence: belief is no lightmatter. To believe the truth of some claim is to make a judgement about thenature of the world. Further, belief invests the believer in its object. Our abil-ity to believe, or not to believe, is basic to our rational maneuverings. Themetaphysical magician, who waves the wand of his conviction without dis-crimination, granting this hypothesis truth and that not, mocks reason itself.Believing without sufficient evidence undermines the distinction betweentruth and falsity because it fails to be precise about that distinction; if I taketo be true what could just as easily, to my knowledge, be false, I have failed toacknowledge the proper boundary between the two and the gravity of myjudgment, both as it reflects on the world and as it implicates me. If this isright, credulity abuses humanity just as lying does in Kant's view, by failingto show respect for rational nature in oneself and others. Unwarranted belieffails to respect reason as an end in itself by using it as a toy to indulge prejudice,inclination, and fancy.

Rawls argues that political justification must be reasonable, in the sensethat such justification gives reasons which one could expect another to accept(whether or not they do in practice). According to Rawls's idea of politicallegitimacy, the "exercise of political power is fully proper only when it is exer-cised in accordance with a constitution the essentials of which all citizens asfree and equal may reasonably be expected to endorse in the light of principlesand ideals acceptable to their common human reason." Rawls invokes thisprinciple in justifying the idea of a public reason, that is, that considerationsintroduced into deliberations be limited to principles that citizens can rea-sonably be expected to endorse. Of course, reasonable people can disagreedue to burdens of judgement. But Rawls would count more beliefs as reason-able than would my proposed Cliffordian constraints. My suggestion extendsthe category of the unreasonable to include beliefs without sufficient evidence,which I take to include most forms of revealed religion, atheism, and beliefsabout the essential natures of particular races, men, or women. For what Clif-ford's arguments purport to do is raise the bar for the reasonable—for what wecan, or should, take as a reason.

Rawls's view that political justification must be reasonable involves norma-tive assumptions about respect for individuals and the justification for theexercise of power over them. Giving reasons—or at least acting in a way forwhich one could give reasons that others could reasonably be expected toaccept—is required (in Kantian terms) to treat others as ends in themselves,or "as beings who must .. . be able to contain in themselves [or share] the endof the very same action." Rawls's political constructivism locates politicallegitimacy precisely in the possibility of agreement, or sharing of ends. As I

What should feminists make of liberal neutrality? 75

have noted above, such a conception of political justification is not neutralbetween comprehensive doctrines, since some will reject reason as the appro-priate form of political justification. My suggestion has been that, given thisaccount of political justification, liberalism may exclude all unwarrantedbeliefs from its justificatory process. This is not to argue, however, that a lib-eral society should not be tolerant of different conceptions of the good, for suchtolerance is fundamental to liberalism. Political neutrality issues from the veilofignorance.

Thus, the justification for liberalism need be even less neutral, with regardto competing doctrines, than Rawls suggests. An "agnostic" neutrality canexclude more competing doctrines from consideration. Further, the motivat-ing ideal of liberalism can be generated on Gliffordian grounds, for the Rawl-sian ideal of moral equality can be defended epistemically by ruling outunreasonable doctrines. Claims of essential inequality cannot be defendedbecause individual exceptions can always be found, and it is difficult to seewhat evidence could be produced to justify claims about, for instance,women's nature. Predictions about the abilities of classes of people cannot bemade with accuracy, so no legitimate reason for unequal distribution of rightsbetween men and women can be given. Further, as John Stuart Mill argued inThe Subjection of Women, we cannot make inferences about women's "nature"from the characteristics women exhibit in an unequal society in which womenand men are subject to different expectations and upbringings. Under currentconditions, we can have no grounds for making general claims about women'sinnate propensities and abilities. Someone might suggest instead that thedistribution of rights or primary goods be proportionate to abilities. But howcould the relevance of abilities be proven? The thought is that an unequal dis-tribution of rights must be justified, and no sufficient evidence can be given forprivileging certain groups or abilities. In the absence of such evidence, thedefault must be equality.

Rawls's defense of reasonableness in political justification (and my ex-tended version of it) cohere with feminist values by excluding sexist beliefsand hierarchical political systems. The exclusion of such beliefs avoids genderbias in the theory; Rawls's liberalism is radically, foundationally, egalitarian.MacKinnon's critique of liberal neutrality as essentially male may have pur-chase instead as a critique of the misguided application of liberal principles.For example, as suggested before, the historical exclusion of marriage fromjustice reflected a gender bias in the construction, and the application, of thelaw. This construction failed to appreciate that serious injustices which calledfor legal recourse existed in the supposedly private realm, and, further, thatthe privacy of marriage (if privacy is understood as absence of interferenceby the state) was illusory since the borders of that privacy, and aspects of theinteraction within it, were regulated by law.

The Legacy of John Rawls76

Liberal feminism has drawn on the principle of equal opportunity to arguethat liberalism is committed to state action to reverse gender inequality.When this principle is applied to gender inequality and the gendered struc-tures of society, it requires the state to address these conditions through lawand redistributive measures such as state-supported childcare, equitablelaws of property division during marriage and on divorce, flexible workinghours, parental leave for both parents, and gender-free schooling. As abasic structure of society—by Rawls's admission—the family is subject to theprinciples of justice. Prima facie, some feminist reforms, such as education inmoral equality, seem incompatible with state neutrality. For example, manyfeminists hold that justice between the genders cannot be achieved so long aschildcare remains primarily women's responsibility. Shared parenting, flex-ible working hours, and state-supported creches may support this goal and becompatible with neutrality. But education that encourages boys and girlsequally to envision themselves as care-givers—or billboard advertisementsencouraging men to change diapers—might conflict with neutrality by teach-ing children a controversial conception of the good. However, again deferringthis conflict between neutrality and feminism, I want to focus on differentimplications that neutrality will have for feminist goals.

MacKinnon's complaints about liberal neutrality focus on how the abstrac-tion of liberal theory ignores the concrete context and effects of its application.Thus, a right to free speech in a (hypothetical) context where the media is con-trolled by sexist men will secure the protection of sexist speech, with ensuingbad consequences for women. However, if we accept MacKinnon's view thatstanding practices often have a deeply patriarchal element, then neutrality inpractice will require radical change. First, certain apparently private socialstructures arguably impede women's equality. But, second, these structuresare not purely private—for instance, it is mistaken to conceptualize civilsociety as a distinct sphere from the state and as free from state interference.Thus, state neutrality will require substantive change in state policies regulat-ing these apparently private spheres, when standing regulation is premised onor promotes a certain conception of the good. Raz and Kymlicka have distin-guished neutrality in deliberation—the state's avoidance of invoking concep-tions of the good in policy-making—and in consequences—the state'savoidance of promoting, as a result of its actions, a conception of the good.I will adduce both kinds in the following discussion.

Take the example of having children and a career. Women are formally freeto choose a career, but, to a much greater extent than men, face a conflictbetween pursuing a career and having children. External obstacles to pursu-ing both include the expense of good childcare and the lack of flexibility in

Feminism and political neutrality

What should feminists make of liberal neutrality? 77

56

working hours imposed by many careers. The structures for pursuit of socialprimary goods are fitted for someone without the responsibilities of parenting.Because women typically take on greater childcare responsibilities than domen, these structures unduly penalize women and tend to reinforce the gen-dered division of labor by forcing parenting women out of the workplace.Green argues that such situations call for state redress on the basis of equalopportunity, since changes are needed to ensure that all individuals canpursue their conceptions of the good. But it is also the case that such struc-tures are not neutral. The arrangement of working hours, for example, washistorically formed on the basis of a gendered division of labor. Further, theireffects unduly promote one conception of the good—that in which primarycare-givers for young children stay home, while their partners work outsidethe home to support the family.

It is crucial to my argument to note that the state directly maintains theworking environment predicated on this conception in a host of ways. Thestate sets minimum wage and overtime laws, places constraints on workinghours, and regulates employee rights through labor law; indirectly, it supportscompanies through government contracts; and as an employer, it enforces itsown policies for those employed by the state. It also has more indirect effectsthrough provision of services such as public transport (which affects access toemployment). Once we consider the state's involvement in all of these areas, itbecomes clear that the state has substantial effects on working conditions.Neutrality between conceptions of the good in this area will require creatingpolicy on grounds that respect the variety of lifestyles persons might choose,and seeking to ensure that the consequences of policy do not indirectly pro-mote one conception of the good.

Another example is marriage. As Okin argues, although the family has beenseen as private, it is a major determinant in the distribution of social goods, andthus family arrangements must meet the demands of justice. It is a life-shapinginstitution, the "gender structure [of which] is itself a major obstacle to equal-ity of opportunity," affecting the "opportunities of girls and women." Again,reform of the family seems required on grounds of equal opportunity. But neu-trality too requires substantial change to marriage and family law. Simply byrecognizing marriage, the state fails to be neutral—and even if marriage wereextended to same-sex partnerships, this would be the case. For by recognizingmarriage, the state picks out a certain type of relationship—monogamous,permanent, between two persons—as worthy of recognition. Again, both indeliberation and effects, the state privileges this type of relationship—corre-sponding to a certain conception of the good—over other arrangements.When we consider how many of the basic structures of society are shaped bystate regulation, and recognize—through feminist or other critique—how

The Legacy of John Rawls78

those structures assume and promote controversial conceptions of the good, itseems that neutrality will require substantial change.

Ronald Dworkin has argued that a liberal state is required by neutrality toact to secure a conception of the good under threat. Taking the example ofenvironmental conservation, he argues that a liberal state is not permitted tosupport conservation on the grounds that it is part of "a superior conceptionof what a truly worthwhile life is." But it may be permitted and even requiredto do so on the basis that non-intervention "is not neutral amongst competingideas of the good life, but in fact destructive of the very possibility of some ofthese." If wildlife and nature are destroyed, the conception of the goodwhich involves them will no longer be open to pursuit.

Similarly, gender-structured social practices destroy the possibilities thatwould be available to women if they were fully equal members of society.The gendered division of labor within the family, the devaluing of women'swork, and employment that penalizes those with small children close offalternatives that should be left open. Different courses of action are madeunavailable by the social conditions themselves. Just as environmentaldestruction destroys the possibility of outdoor pursuits, the hegemony ofgender-structured social practices destroys the possibility for women to liveas they could in a society in which women were not systematically disadvan-taged. The force of this point is not that women's lives could be better. Thepoint is that the world that is made impossible by social practices is one inwhich women could live without the impediments to success and the strainson their psychology, their social relations, and their attitudes to and expecta-tions of work, love, and parenting, which currently exist as a consequence ofoppression. The closure of this possibility is unjust, and not just because it vio-lates equal opportunity, but because it violates neutrality. Even when indivi-dual women are able to overcome barriers, they do not have access toconceptions of the good that would be available to them in a society wherewomen were fully equal. Also eliminated are the unknown possibilities of thegood, for women, of living as fully equal members of society.

In this chapter, I have tried to show that feminism has more to gain fromstate neutrality than has been widely recognized. Justificatory neutrality asagnostic, like the foundational ideal of moral equality, implies a rejection ofsexist and other oppressive doctrines. In practice, state neutrality implies sub-stantial change, once the state's involvement in the apparently private sphereis recognized, and the disproportionate influence of a sexist conception of thegood on those structures—and concomitant promotion of that ideal—is seen.Liberal neutrality need not be inconsistent with feminist goals; indeed, it maysupport them. In fact, if neutrality does require these far-reaching changes,the other conflicts between feminism and neutrality, or neutrality and justice

What should feminists make of liberal neutrality? 79

as a comprehensive doctrine, may become less pressing. Attitudinal change

or feminist education may be less important when primary care-givers

are no longer penalized in their pursuit of a career, the law of marriage

and divorce is reformed, and other substantial economic and legal changes

are effected.

1. Susan Moller Okin, Justice, Gender, and the Family (New York: Basic Books, 1989),

Ch. 5, quotation from 89. See also Veronique Munoz-Darde, "John Rawls,Justice in the Family, and Justice o^the Family," The Philosophical Quarterly 48.192

(1998), 335-52; she writes of Rawls that "the family is both treated as a distinctand fundamental institution, and never discussed in any detail." (See Munoz-

Darde, "John Rawls," 337) As these writers have pointed out, in Theory, Rawls

assumed that the family as it stands is just. (See TJ original edn, 467—8, and sec-

tions 70 and 71.)2. Susan Moller Okin, "Political Liberalism, Justice, and Gender," Ethics 105.1

(1994), 23-43 (43). Rawls replies to Okin in PRR reprinted in LP, 129-80, sec-tion 5, and in JF, IV.50. Andrew Smith has argued that these replies fail to

address the problems raised by liberal tolerance of religious groups which oppress

women; see "Closer But Still No Cigar: On the Inadequacy of Rawls's Reply to

Okin's 'Political Liberalism, Justice, and Gender'," Social Theory and Practice 30.1

(2004), 59-71.3. See Okin, "Political Liberalism, Justice, and Gender," 42—3. More recently, Okin

has addressed the tensions between liberalism and democracy which arise in the

context of conferring group rights on groups which oppress women. (See Susan

Moller Okin, with respondents, in Joshua Cohen, Matthew Howard, andMartha C. Nussbaum (eds), IsMulticulturalismBadfor Women? (Princeton: Prin-

ceton University Press, 1999).)4. See Catharine MacKinnon, Feminism Unmodified: Discourses on Life and Law (Cam-

bridge, MA: Harvard University Press, 1987) and Toward a Feminist Theory of the

State (Cambridge, MA: Harvard University Press, 1989), esp. 157-70. See also

Carole Pateman, The Sexual Contract (Cambridge, UK: Polity Press, 1988).

These feminist critiques of liberalism are influenced by Marxism. One can com-pare MacKinnon's claims to Marxist claims that liberal ideology serves capital-ists while obscuring workers' oppression. The views of MacKinnon and Pateman

should be distinguished from the feminist critique of liberalism motivated by an

ethics of care or relationality; for an example of that view, see Virginia Held,"Non-Contractual Society: A Feminist View," in Marsha Hanen and Kai Niel-sen (eds), Science, Morality, and Feminist Theory, Canadian Journal of Philosophy, Sup-

plementary Volume 13 (1987), 111-37.5. MacKinnon, Toward a Feminist Theory, 195—214.

6. Ibid., 161-2.

Notes

The Legacy of John Rawls80

What should feminists make of liberal neutrality? 81

7. Okin, "Political Liberalism" 42.

8. Martha Nussbaum, Sex and Social Justice (Oxford, UK: Oxford University Press,

2000), 77—80. Indeed, though MacKinnon critiques liberalism, Nussbaum writes

that we may see her as "a kind of Kantian liberal, inspired by a deep version of

personhood and autonomy," 79.

9. See John D. Walker, "Liberalism, Consent, and the Problem of Adaptive Prefer-

ences," Social Theory and Practice 21.3 (1995), 457-71, and my "A Liberal Response

to Catharine MacKinnon," Southwest Philosophical Studies 22 (2000), 17-23.

10. More fine-grained distinctions can be made; see pp. 883—4 in Will Kymlicka,

"Liberal Individualism and Liberal Neutrality," Ethics 99.4 (1989), 883-905,

and Joseph Raz, The Morality of Freedom (Oxford, UK: Oxford University Press,

1986). One might distinguish the two sorts of neutrality I have in mind by think-

ing of them as positioned before and after the original position. Justificatory neu-

trality is neutrality in the argument for the theory of justice; political neutrality is

the neutrality exercised by the state in accordance with the principles ofjustice.

11. TJ original edn, 92-3, and PI, 19.

12. Ronald Dworkin, "Liberalism," in Stuart Hampshire (ed.), Public and Private

M>™% (Cambridge, UK: Cambridge University Press, 1978), 113-43 (127-9).

13. Kymlicka, "Liberal Individualism," 883.

14. John Skorupski, "Liberal Elitism," in his Ethical Explorations (Oxford, UK:

Oxford University Press, 1999), 193-212.

15. H. Tristram Engelhardt, "Freedom and Moral Diversity: The Moral Failures of

Health Care in the Welfare state," Social Philosophy and Policy 14.2 (1997), 180-96.

16. See, for example, Colin Bird, "Mutual Respect and Neutral Justification," Ethics

107.1 (1996), 62—96: "it is not enough that the liberal state embrace an ethic ofneutrality. The justification for this liberal stance must itself display a certain

kind of neutrality or impartiality by avoiding arguments which rely on 'contro-

versial' claims about the nature of the good life." (See Bird, "Mutual Respect

and Neutral Justification," 62.)

17. TJ original edn, 454. See also PI, 142.

18. TJ original edn., 567, 572. One way to understand the principles ofjustice is as

the object of choice of a free rational will. Rawls calls his method "Kantian con-

structivism": his principles, like Kant's moral law, are constructed through

reason rather than intuited. (PL, 99-107)

19. PL, 13.

20. Ibid., xvi—xvii.

21. TJ original edn, 453—4. Rawls dismissed this conception of a well-ordered society

as "unrealistic" in PI, xvi.

22. TJ original edn, 454, and see PL, xvi.

23. See PL, xvi—xvii. He also revises the account of stability.

24. PL, 15.

25. PI, 11.

26. Interview with Bernard Prusak, "Politics, Religion and the Public Good:

An Interview with Philosopher John Rawls," Commonweal 125.16 (1998), 12-18,

reprinted in CP, 616-22.

82 The Legacy of John Rawls

27. See Michael Huemer, "Rawls's Problem of Stability," Social Theory and Practice22.3 (1996), 375—96. For further discussion, see also Samuel Schemer, "TheAppeal ofPolitical Liberalism," Ethics 105.1 (1994), 4-22.

28. See, for example, Okin's "Political Liberalism, Justice, and Gender." Also, SusanMendus has argued that Rawls's argument for the congruence of justice with theagent's good need not invoke a comprehensive conception of the good, and so theaccount in Theory need not be inconsistent—see "The Importance of Love inRawls's Theory ofJustice," British Journal of Political Science 29.1 (1999), 57-75.

29. TJ original edn, 561.30. To review some of the difficulties associated with establishing an account of moral

equality, see Bernard Williams, "The Idea of Equality," in his Problems of the Self(Cambridge, UK: Cambridge University Press, 1973), 230-49.

31. A related criticism, made by Adina Schwartz in "Moral Neutrality and PrimaryGoods," Ethics 83.4 (1973), 294-307, and by Thomas Nagel in "Rawls on Jus-tice," Philosophical Review 82.2 (1973), 220-34, is that Rawls's individualism andhis argument that contractors in the original position will seek to maximize theirprimary goods is incompatible with socialist views of the good, or indeed, withthose of members of religious orders who take vows of poverty. (See also Kym-licka, "Liberal Individualism," 886—93.)

32. Paul F. Campos raises such a criticism in "Secular Fundamentalism," ColumbiaLaw Review 94.6 (1994), 1814-27.

33. Of course, Rawls means only for justice as fairness to be compatible with all rea-sonable comprehensive doctrines (PL, 210), and he would surely view a doctrinewhich denied the separation of Church and state as unreasonable. It should beobvious why this response will not placate the believer in question.

34. This is also Okin's concern; but the concern deepens in proportion to the range ofsocial practices which one sees as constitutive of, and reinforcing, inequality.

35. Cf. Okin's "Political Liberalism, Justice, and Gender" on this point.36. PL, 13.37. See Jean Hampton, "Feminist Contractarianism," in Louise Antony and Char-

lotte Witt (eds), A Mind of One's Own (Oxford, UK: Westview Press, 1993),227-56.

38. I use "civil society" in Hegel's sense. On this point, see Okin, Justice, Gender, andtheFamily, and John Tomasi, "Individual Rights and Community Virtues," Ethics101.3 (1991), 521-36.

39. Makingjustice comprehensive suggests a much stronger liberalism. For example,it allows that the liberal view ofjustice will come into conflict with comprehensiveviews about the good, which include belief in gender hierarchy. This will producesocial instability; if liberalism is to be stable, it must inculcate the principles ofjustice and associated ideals in education. This is the conflict Okin discusses inMulhculturalism. In response to the idea that restricting justice to the politicalwill promote stability, I say that justice, whether political or comprehensive,will be unstable unless ideals of moral equality are taught.

40. Rawls distinguishes the rational and the reasonable. Rational applies to means-end reasoning, and reasonable to a willingness to enter discussions.

What should feminists make of liberal neutrality? 83

41. Of course, on Rawls's description, they should be excluded in the original positionbecause the contractors do not know their own race or sex. But my comments hererespond to the objector who asks why Rawls is entitled to set up the original posi-tion in a race- or gender-blind way.

42. Clifford's view has been challenged. William James responded to it in "The Willto Believe" (1896), reprinted in John J. McDermott (ed.), The Writings of WilliamJames (Chicago: University of Chicago Press, 1978), 717—35. Susan Haack hasmore recently argued that it is too demanding in her "The Ethics of Belief Recon-sidered," in Lewis Hahn (ed.), The Philosophy of Roderick M. Chisholm (Chicago:Open Court, 1997), 129-44. A major problem with Clifford's view is the diffi-culty of specifying what evidence counts as sufficient. I will bracket these seriousworries for two reasons. First, I am concerned with ruling out certain value judge-ments and religious beliefs from the justificatory process, not establishing thatbelief in them is ethically wrong. Second, I am more interested in comparing Clif-ford's arguments with Rawls's than with defending Clifford's conclusions.

43. William Kingdon Clifford, "The Ethics of Belief," in his Lectures and Essays (edsLeslie Stephen and Frederick Pollock; London: Macmillan, second edn, 1886),339—63 (346), originally published in Contemporary Review (1877).

44. Clifford, "Belief", 342.45. Ibid.46. Ibid., 345; John Stuart Mill, OnLiberty (London: Penguin Books, 1985 [1859]).47. Clifford, "Belief", 343.48. Ibid., 345-6.49. Ibid., 346.50. As a character in David Lean's 1962 film Lawrence of Arabia suggestively says,

"A man who tells lies, like me, merely hides the truth. But a man who tells half lieshas forgotten where he put it."

51. Kant's views on lying extend beyond the famous false promising example in theGroundwork; in the "Doctrine of Virtue," Book 1, Ch. 1, he writes that: "By a lie ahuman being throws away and, as it were, annihilates his dignity as a humanbeing." (Immanuel Kant, Practical Philosophy, ed. Mary Gregor (Cambridge,UK: Cambridge University Press, 1996), 552—3) Part of the argument he givesfor this is that lying contradicts the purpose of the faculty of communication.

52. PL, 137.53. Rawls in fact claims the theory of justice is reasonable, not true, so Clifford's com-

ments would have to be adjusted, mutatis mutandis, to emphasize the evaluation ofreasons as opposed to the evaluation of true belief.

54. From Kant's Groundwork of the Metaphysics of Morals, also to be found in Gregor'sPractical Philosophy, 80 (4:430).

55. John Stuart Mill, The Subjection of Women, ed. Susan Moller Okin (Indianapolis:Hackett, 1988 [1869]).

56. Okin, Justice, Gender, and the Family, 175—9. Okin also bases her arguments on thefamily's influence on moral development, in response to Rawls's account of moralpsychology in Part III of Theory. She argues that citizens will not develop a senseof justice so long as fathers and mothers possess unequal shares of power. Like

84 The Legacy of John Rawls

Mill, she notes that habitual inequality in personal life ill equips us to treat others,in any circumstances, as free and equal. (Also see Karen Green, "Rawls, Women,and the Priority of Liberty," injanna L. Thompson (ed.), Women and Philosophy,Australasian Journal of Philosophy, Supplementary Volume 64 (1986), 26—36.)Green uses the liberty principle, as well as the principle of equal opportunity, inan argument for liberal feminism; however, invoking the liberty principle in thiscontext seems to me problematic.

57. TJoriginal edn, 7.58. See Kymlicka, "Liberal Individualism", 883—4. Raz (Morality) argues that

Rawls endorsed neutrality in consequences, and Kymlicka that Rawls endorsedjustificatory neutrality. Since both are implicated in the same way in my argu-ment, I will consider both.

59. Green, "Rawls," 35.60. Okin, Justice, Gender, and the Family, 16.

61. Feminists and Marxists have argued that monogamous marriage indirectly pro-motes other conceptions of the good (such as the value of property ownership).(See Christine Overall, "Monogamy, Non-Monogamy, and Identity," Hypaha:A Journal of Feminist Philosophy 13.4 (1998), 1-17 and John McMurtry, "Mono-gamy: A Critique," TheMonist56 (1972), 587-99.)

62. Dworkin, "Liberalism," 141.

In Political Liberalism Rawls argues for the following conception of justice:

The aim ofjustice as fairness.. . is practical: it presents itself as a conceptionof justice that may be shared by citizens as a basis of reasoned, informed,and willing political agreement. It expresses their shared and public politi-cal reason. But to attain such a shared reason, the conception ofjusticeshould be, as far as possible, independent of the opposing and conflictingphilosophical and religious doctrines that citizens affirm. In formulatingsuch a conception, political liberalism applies the principle of tolerationto philosophy itself. The religious doctrines that in previous centuries werethe professed basis of society have gradually given way to principles of con-stitutional government that all citizens, whatever their religious view, canendorse. Comprehensive philosophical and moral doctrines likewise cannotbe endorsed by citizens generally, and they no longer can, if they ever could,serve as the professed basis of society.

Rawls believes a freestanding conception of political liberalism is the mostpromising liberal conception ofjustice because of its ability to accommodatethe fact of reasonable pluralism, and promote a stable, well-ordered society.Political liberalism also imposes moral constraints on the kinds of reasonsthat can be invoked to defend principles ofjustice. In other words, politicalliberalism is the best version of liberalism because it accomplishes the practi-cal and moral aims of liberalism better than those versions of liberalism —including the version defended in A Theory ofjustice—that defend liberalism asa comprehensive moral doctrine. The most striking difference betweenA Theory ofjustice and Political Liberalism is that the earlier work affirms a con-ception ofjustification that treats principles ofjustice as true whereas the laterwork affirms a conception ofjustification that treats principles ofjustice asacceptable to fair-minded citizens in a liberal democracy.

There are a number of objections to the idea of a freestanding political lib-eralism. Three are relevant to this chapter. One claims that Rawls's method-ology in Political Liberalism leaves unclear whether he is affirming Kantian or

5

Public reason and the moral foundation of liberalism

Jon Mahoney

1

3

86 The Legacy of John Rawls

Hobbesian liberalism. According to Kantian liberalism, justice requires thatpolitical principles be understood as principles that reasonable persons moti-vated by moral reasons can agree to. According to Hobbesian liberalism, jus-tice requires that political principles be understood as principles that rationalpeople motivated by a desire to advance their self-interest can agree to. Kan-tian liberalism requires a moral agreement. Hobbesian liberalism settles for amodus vivendi. Rawls insists that political liberalism affirms a Kantian yet free-standing moral conception of justice and thus that this conception requires an"overlapping consensus" that is affirmed for moral rather than strategic rea-sons. However, Rawls characterizes a freestanding yet moral doctrine as adoctrine that inherits moral content from the reasons citizens have for affirm-ing it. This implies that the moral content of political liberalism emerges froma joint agreement by citizens who may have any number of reasons for accept-ing a liberal conception of justice. If political liberalism is freestandingbecause its justification emerges from a consensus, there is no way to specifyin advance that only reasons offered by reasonable citizens qualify as legiti-mate reasons for affirming political liberalism. And thus political liberalismis indistinguishable from Hobbesian liberalism.

A second objection contends that the reasonable rejectability criterion formoral justification must be taken to be true and thus one of liberalism's coreprinciples cannot be regarded as freestanding. The reasonable rejectabilitycriterion specifies that no agent can be forced to accept terms for social co-operation that he or she may reasonably reject. Rawls claims that many in-compatible doctrines qualify as reasonable. Compatibility with freestandingprinciples of justice rather than truth is what qualifies a doctrine as reason-able. Therefore, political liberalism should refrain from judging the truth orfalsity of any such doctrine, including itself. This is among the reasons Rawlsclaims that political liberalism applies the principle of toleration to itself.However, political liberalism must at least affirm the truth of its own founda-tional principle of justification. The substitution of reasonableness for truthcannot apply to the principle of reasonable rejectability itself for this wouldrender liberalism utterly groundless.

Finally, a third objection claims that those principles admissible by liberal-ism can only be regarded as justified if their soundness can be establishedfrom a shared, impartial point of view of practical reason. Rawls believes thatpolitical liberalism can be affirmed for separate yet equally justified reasons.A reason for affirming political liberalism qualifies as justified if it is affirmedfrom the standpoint of a reasonable comprehensive doctrine. Since there are,according to Rawls, a number of conflicting yet reasonable doctrines thataffirm liberal principles, different doctrines can for different reasons affirmthe same principles of justice. For instance, Kantian, utilitarian, and religiousendorsements can all qualify as reasonable. This implies that the standpoint

Public reason and the moral foundation of liberalism 8 7

of practical reason from which agents affirm liberalism does not qualify as ashared point of view on the basis of which fundamental political principlesare endorsed. Rather, freestanding principles of justice become part of ashared vision for political life only after citizens affirm such principles. Butunless there is a point of view—and if not of practical reason, then what? —that can be construed as independent of all comprehensive doctrines, thenthe means for identifying a doctrine or reason as reasonable is quite mysteri-ous. To claim, as Rawls does, that a doctrine or citizen is reasonable becauseit or he/she affirms the right kind of principles of justice—and not just forany reason whatsoever but for moral reasons—seems to presuppose a sharedmoral point of view.

Each of these objections has been advanced by a number of liberal critics ofRawls. Those who offer such objections share a common argumentativestrategy, which is to show that Rawls's attempt to defend a Kantian concep-tion ofjustice without a Kantianjustificationis unsuccessful. Rawls believes hecan answer this charge. Anyone familiar with Rawls's work knows that hispainstaking attempt to present a systematic theory of liberalism is intendedto equip his theory with an immunity to criticism. In fact, Rawls either antici-pates or responds to all three objections claiming they can be rebutted, accom-modated, or ignored because irrelevant. The purpose of this chapter is todemonstrate that Rawls cannot resort to any of these strategies if the threecharges against political liberalism are presented in the right way.

The first section considers the relationship between reasonableness andpublic reason. The second section considers Rawls's conception of citizensand their capacities. My exposition of Rawls's position focuses on the ideas ofthe reasonable and public reason as these concepts are used in Political Liberal-ism and in Rawls's more recent essay "Public Reason Revisited." In my view,the best way to argue against Rawls is first to establish that his conceptionof citizens and their capacities is indispensable to the moral foundation ofpolitical liberalism. In the third section I present an argument based on thisstrategy that purports to show that liberalism cannot be understood as a free-standing doctrine.

Reasonableness and public reason

A reasonable citizen is someone willing to seek fair terms for social coopera-tion. Fair terms for social cooperation are those that can be jointly affirmedby all in a manner that affirms the equal standing of each. There are otherways that Rawls uses the concept of reasonableness, but for present purposesthe capacity for reasonableness, a feature of the capacity for public reason, isthe most important. Reasonableness in this sense is to be characterized as a

88 The Legacy of John Rawls

capacity of citizens who want to regulate their lives in common through termsall can accept, where acceptability requires that such terms be fair.

Rawls claims that the idea of public reason is contained in the idea of thereasonable. The idea of the reasonable contains the idea of public reason inthe following way: reasonableness is a capacity to be fair in one's treatment ofothers and fairness supplies the minimal moral content to public reason.Public reason provides the norms that citizens, public officials, and judgesshould follow when making claims about matters of justice. The idea ofpublic reason "is a view about the kind of reasons on which citizens are torest their political case in making their political justifications to one anotherwhen they support laws and policies that invoke the coercive powers ofgovernment."

The subject of public reason is justice broadly construed. A citizen com-plies with what Rawls terms the ideal of public reason, when, for example,she demands equal employment opportunities on the grounds of politicalequality. By contrast, a citizen violates the ideal of public reason when hedemands that others convert to his religion as a condition for equal politicalstanding. As Rawls puts it, "we say that ideally citizens are to think of them-selves as if they were legislators and ask themselves what statutes, supported bywhat reasons satisfying the criterion of reciprocity, they think it would be mostreasonable to enact."

Public reason imposes a normative constraint on the kinds of considera-tions citizens are entitled to base their political judgements on and the waysin which citizens are to conduct themselves in their relations with othercitizens. Should there be disputes about whether a reason falls within therange of acceptable public reasons, such disputes can be adjudicated by con-sidering whether the reason in question could be accepted by all without vio-lating liberalism's moral commitment to rights and equality.

One might claim that the idea of the reasonable is rooted in the idea of prac-tical reason understood as a capacity to present, solicit, and revise one's rea-sons for affirming principles. Rawls denies that liberalism must accept thisconnection between the reasonable and practical reason. Reasonableness isunderstood to be a normative concept that imputes a capacity to citizens;practical reason in the traditional sense used in comprehensive moral doc-trines is understood to be a normative concept that imputes a capacity to per-sons. Political liberalism tries to avoid philosophical disputes about whichreasonable people can disagree and thus tries to remain neutral on questionsabout practical reason. According to Rawls, the idea of public reason does notrefer to the capacities of natural persons, and thus is to be distinguished fromthe capacity for practical reason. This is required in order to claim that theidea of public reason is part of a freestanding rather than comprehensive doc-trine. In this way, the idea of reasonableness and the idea of public reason is

12

14

Public reason and the moral foundation of liberalism 89

said to be suitably nonsectarian; it is compatible with the sectarian views of allcitizens willing to affirm political liberalism. The attempt to derive the reason-able and public reason from a comprehensive doctrine is a mistake, in Rawls'sview, because accommodating the fact of reasonable pluralism requires mini-mizing controversies that are likely to erode support for liberal principles.In other words, liberalism understood as a comprehensive moral doctrine isto be rejected as the foundation for society's charter for political and legalinstitutions, because any comprehensive doctrine can be reasonably rejected.In Justice as Fairness: A Restatement, Rawls claims that were the state to affirm acomprehensive liberalism its exercise of political power would be oppressive:"This is as true of the liberalism of rightness as fairness, as it is of the Chris-tianity of Aquinas or Luther."

Citizens and their capacities

In Lecture II of Political Liberalism, "The Powers of Citizens and theirRepresentation," Rawls offers an account of the capacities citizens can beassumed to possess in a liberal democracy. He claims that his discussion of thecapacities that make possible reasonableness and public reason can be under-stood as applying exclusively to citizens and the freestanding liberalism theyendorse. Rawls's formulation of these capacities is supposed to be so generic asto be acceptable to any reasonable citizen regardless of his or her comprehen-sive doctrine. He does not assert that natural persons lack these capacities; heclaims only that political liberalism is neutral on this and other issues aboutwhich reasonable people disagree.

The capacities of citizens include what Rawls terms the two moral powers:the capacity for a sense of justice—the reasonable—and the capacity to for-mulate and revise conceptions of the good—the rational. Rawls further ela-borates the two moral powers by claiming political liberalism regards citizensas having: (a) object-dependent desires (i.e. causally determined preferences),(b) principle-dependent desires (i.e. desires motivated by principles whichmay be affirmed for either moral or non-moral reasons), and (c) conception-dependent desires (i.e. desires motivated by a reflectively endorsed conceptionof value). This characterization tracks a familiar distinction between moti-vating desires—desires that motivate independently of reasons—and motivateddesires—desires that are motivated by reasons; object-dependent desires aremotivating; principle- and conception-dependent desires are motivated.

The primary reason Rawls claims political liberalism must impute suchcapacities to citizens is that such capacities are central to political autonomy.This conception of political autonomy includes the capacity for principle- andconception-dependent desires and these capacities serve to explain part of the

17

19

90 The Legacy of John Rawls

moral foundation for public reason. Citizens who affirm the ideal of publicreason have a conception-dependent desire to affirm reasons that others withthe same conception-dependent desire also affirm. I interpret Rawls's accountof reasonableness and conception-dependent desires as one way to character-ize the reasonable rejectability criterion for moral justification. There is infact no difference between claiming that reasonable citizens share a concep-tion-dependent desire to find fair terms for social cooperation and claimingthat reasonable citizens accept as justified only those reasons that no reason-able citizen can reject. Of course, there could be a difference between thenorms of public reason and the reasonable rejectability criterion, yet Rawlscharacterizes the content of public reason in such a way that any reason thatsatisfies the norms of public reason will satisfy the reasonable rejectability cri-terion. Likewise, any reason that violates the norms of public reason will vio-late the reasonable rejectability criterion. In each case the content of thereasonable is specified in terms of a counterfactual claim about what citizenswill and will not affirm if they are reasonable. Of course, the reasonable has tobe understood as part of a freestanding political doctrine and thus neitherpublic reason nor reasonable rejectability are to be understood as a conceptionof moral justification for all moral norms. Moreover, since political liberalismdemands a conception of the citizen as free and equal, the account of citizensand their powers presented in Lecture II is a crucial part of the moral founda-tion of a freestanding political doctrine. The minimal moral content to politi-cal liberalism can, according to Rawls, be presented in a manner that anyreasonable citizen can affirm, regardless of his or her comprehensive doctrine.The moral content to political liberalism can, in other words, be affirmed asthe basis for an overlapping consensus endorsed by reasonable persons moti-vated to find fair terms for social cooperation.

If Rawls's position is sound, his Kantian conception of the citizen is fit toserve as part of the content of justice that is affirmed within an overlappingconsensus of reasonable persons jointly committed to finding fair terms forsocial cooperation. Since this conception can be affirmed within an overlap-ping consensus, it is neutral toward reasonable comprehensive doctrines,including those that deny a Kantian conception of moral justification. More-over, given the practical aim of political philosophy, it is more important toachieve agreement about principles of justice than it is to achieve agreementabout the reasons for them. Therefore, if a conception of justice can beaffirmed for separate but equally legitimate reasons, political liberalism canaccommodate the fact that different people will have different reasons foraffirming the Kantian conception of justice advanced by political liberalism.

There are good motives for adopting this strategy of conflict avoidance. It isa fact about the modern world that, as Montaigne put it, "the more we talkabout such things [e.g. human nature, moral truth, the good life, etc.], the

Public reason and the moral foundation of liberalism 91

more we disagree—even with ourselves." Thomas Paine expressed the sameidea when he claimed, "I do not believe that any two men, on what are calleddoctrinal points, think alike who think at all." Why not try to accommodatedisputes about human nature (and others like it) by showing how such dis-agreements can be contained within an overlapping consensus on basic politi-cal norms? This question lies at the center of Rawls's attempt to elucidate thecontent of public reason with principles that do not presuppose any controver-sial moral doctrine.

From one perspective, Rawls's strategy is appealing. Liberalism has alwaysbeen advanced as a conception of justice that affirms that reasonable peoplecan disagree about many but not all things. Reasonable people can, forinstance, disagree about whether Aristotle or Kant provides the best theoryof justice. But no reasonable person can, as Locke famously put it, compel aperson to affirm a belief by fire and sword. According to Rawls, his free-standing conception of political liberalism is the first to succeed in consist-ently fulfilling the practical aim of liberal theory—accommodating reason-able pluralism—while at the same time defending moral principles that allreasonable citizens can accept. However, I believe it is possible to present thecharge that political liberalism cannot be freestanding in a way that Rawlscannot answer.

The argument against political liberalism

The moral foundation of liberalism can be defended in one of three ways:

option 1: as a conception one accepts as a result of one's affirmation of poli-tical liberalism;

option 2: as a conception one must affirm as a presupposition for political lib-eralism;

option 3: as philosophical truth about practical reason and persons.

The criterion for evaluating each option is: does it provide a justification forthe moral foundation of liberalism? I use the expression "the moral foundationof liberalism" to refer to the conception of the citizen and his or her powers andthe reasonable rejectability criterion.

My claims are: if we endorse Option 1, then it is impossible to distinguish amoral consensus from a modus Vivendi. If we endorse Option 2, then the moralfoundation of liberalism must be affirmed dogmatically. Since Option 1 ren-ders political liberalism indistinguishable from Hobbesian liberalism, it isunacceptable to anyone who believes a just society is chartered by moral prin-ciples affirmed for moral reasons. Option 2 does demonstrate how political

25

92 The Legacy of John Rawls

liberalism can be understood as both freestanding—which makes it com-patible with the fact of pluralism—and moral—which provides political liber-alism with the reasonable rejectability criterion and other central moralconcepts. But Option 2 is unstable; it requires that the central moral conceptsrequired by liberalism be affirmed as both reasonable and part of a free-standing doctrine. I call this dogmatic liberalism because it affirms a moralfoundation for which no justification is provided. That leaves Option 3.In my view this is the preferred option for liberals who want to advanceliberalism as a political doctrine with a moral foundation.

Option 1

Rawls claims that not all reasons for affirming liberalism qualify as justified;political liberalism is not a doctrine of moral skepticism or complete neutralityabout the reasons people have for affirming a conception of justice. However,any endorsement of political liberalism that reflects moral reasons that expressa moral commitment to fair terms for social cooperation qualifies as a justifiedreason for affirming the Kantian yet merely political conception of justiceadvocated by political liberalism. This is one reason Rawls believes he canrebut the objection of those who claim that a Kantian conception must beendorsed by a Kantian justification. The idea of an overlapping consensusshows how conflicting doctrines can affirm for different reasons the sameKantian conception of justice. The moral foundation for political liberalismis part of an overlapping consensus among reasonable citizens, that is, citi-zens with a conception-dependent desire to find fair terms for social coopera-tion. Though such citizens also hold a variety of incompatible comprehensivedoctrines they can agree on fundamental principles of justice because theyagree that the charter for political society should be just. However, whatjustifies the claim that only moral reasons for affirming political liberalismare legitimate?

Option 1 is a strategy for justifying the moral foundation of liberalism thatlinks justification to affirmation by citizens who affirm the moral content of anoverlapping consensus for a number of different reasons. According to Option1, the moral content to political liberalism is neutral regarding the variousreasons citizens have for affirming this moral content. There is some textualsupport for the claim that this is a feature of political liberalism. At one pointin his discussion of the senses in which political liberalism is freestanding,Rawls states, "In what sense are citizens free? .. . The relevant meaning offree is to be drawn from the political culture of [a democratic] . . . society."Recall that the conception of citizens and their powers is presented by Rawlsas part of the justification for the idea of public reason and reasonable reject-ability. If this conception is drawn from the political culture of a democratic

27

28

29

Public reason and the moral foundation of liberalism 93

society, then the moral content to political liberalism is inherited. Of course,the democratic culture from which such a conception is drawn may embodythe central moral concepts of liberalism. But that is beside the point. The issuehere concerns the justification for the moral foundation of liberalism.

With Option 1 political liberalism requires what it cannot achieve. On theone hand, it requires that only reasons motivated by fairness qualify as legiti-mate reasons for affirming political liberalism. On the other hand, from theperspective of this conception of political liberalism, there is no way to dis-tinguish reasons for affirming political liberalism that are motivated by fair-ness and reasons motivated by self or strategic interests. In other words, thedistinction between the reasonable and the rational cannot be invoked to dis-tinguish legitimate from illegitimate reasons for affirming political liberalismif political liberalism presupposes that citizens already endorse the moralfoundation of political liberalism. In order to rebut the charge that an over-lapping consensus is compatible with a modus vivendi, Rawls must show thatpolitical liberalism is entitled to make moral demands on citizens concerningthe reasons they have for accepting a conception of justice independentlyof the reasons citizens actually endorse.

Characterizing Option 1 in this way is instructive in part because it explainsthe appeal behind the charge that Rawls's Kantian conception requires aKantian justification. Option 1 brings to light the tension between what wemight call the political and the moral aims of political liberalism. However,the reason Option 1 cannot be affirmed in a way that earns the right to distin-guish an overlapping consensus from a modus vivendi is not that one must affirma Kantian justification to earn such a right. Rather, one must affirm a moralconception, Kantian or otherwise, that has priority over any reasons citizenshave for affirming political liberalism. This moral conception has to be pre-sented as one that can be invoked prior to any reasons one might offer infavor of political liberalism.

Option 2

According to Option 2, political liberalism is neutral regarding those reasonsfor affirming political liberalism that qualify as reasonable (i.e. reasons moti-vated by fairness) yet rejects as illegitimate reasons that are not motivated byfairness. On this interpretation, political liberalism is addressed to reasonablecitizens in a way that imposes minimal moral requirements on the reasons theyhave for endorsing liberalism. In this way, political liberalism can serve as thefocus of an overlapping consensus that both accommodates pluralism andaffirms a moral conception of justice.

Option 2 is a better representation of political liberalism than Option 1.Consider, for example, Rawls's claim that there are two senses in which a

94 The Legacy of John Rawls

liberal doctrine may be political. A liberal doctrine is political in the rightway if it imposes the following requirement: only moral reasons for affirmingliberalism are justified reasons. By contrast, a liberal doctrine is political in thewrong way if it imposes the much weaker requirement: any reason for affirm-ing liberalism is a justified reason. Political liberalism is supposed to be politi-cal in the right way, because it requires that citizens affirm it for public ratherthan private, and moral rather than instrumental reasons. An overlappingconsensus that affirms political liberalism is therefore not compatible with aHobbesian modus vivendi. Rather, it is a moral agreement for moral reasons byreasonable citizens who want to regulate their lives in common with fair termsfor social cooperation.

Option 2 is thus an improvement over Option 1 because it imposes the rightkind of moral demand on citizens. It also qualifies as a freestanding conceptionbecause it presents the moral demands it imposes on citizens in a way thatallows any moral reason advanced by a reasonable citizen to qualify as a justi-fied reason for affirming liberalism. In "Justice as Fairness: Political not Meta-physical," Rawls claims

the aim of justice as fairness as a political conception is practical, and notmetaphysical or epistemological. That is, it presents itself not as a con-ception of justice that is true, but one that can serve as the basis of informedand willing political agreement between citizens viewed as free and equalpersons.

This passage illustrates that for Rawls a legitimate agreement aboutfundamental principles of justice is a consensus that affirms terms for socialcooperation compatible with the idea of "citizens viewed as free and equal.""Free" and "equal" represent moral constraints that must be satisfied in orderfor a consensus aboutjustice to qualify as legitimate. A moral conception of theperson is presented by political liberalism as a moral demand that citizensmust recognize in the reasons they have for affirming liberal principles. More-over, this moral demand requires reasonableness rather than truth regardingthe various reasons citizens may have for affirming a liberal conception of jus-tice. This feature of political liberalism illustrates how a freestanding doctrinecan make moral demands on the reasons citizens have for affirming liberalism.

The primary reason for preferring Option 2 over Option 1 is that it presentspolitical liberalism in a way that makes clear why a modus vivendi is incompati-ble with a liberal conception of justice. Option 2 is also compatible with theidea of a freestanding doctrine. As one commentator puts it, "the reasonable... is the political equivalent or analog of the true." Though the analogy ispartial, this is a helpful characterization. Reasonableness is a standard thatreasons must satisfy in order to qualify as legitimate. The reasonable thus

31

Public reason and the moral foundation of liberalism

plays a justificatory role in political liberalism. Like truth, reasonableness is atest for legitimacy. At the same time, the reasonable is far more inclusive thantruth. Many reasons that will not generate an agreement about truth willnevertheless be acceptable as reasons for affirming liberal principles of justice.This is an appealing feature of political liberalism because it shows how poli-tical liberalism can accommodate the range of reasons citizens may have foraffirming a liberal conception of justice.

The problem with Option 2 is that there is no way for this conception toprovide a justification for the moral foundation of liberalism; its moral contentis a presupposition that is used to issue an unjustified declaration of the con-cepts and principles that citizens must endorse as a precondition for endorsingan overlapping consensus. Dogmatism is an apt characterization for apoliticalliberalism that regards its moral foundation as a legitimate source of claims yetrefuses to provide a justification for this foundation.

Suppose we substitute "reasonable" for "true" and claim that political lib-eralism is committed to affirming the reasonableness of its own moral foun-dations. Reasonableness could then serve as a justificatory standard forevaluating political claims. We could then defend the thesis that some claimsare justified because they are reasonable and leave unanswered questionsabout the truth of such claims. There is a familiar strategy in moral philosophythat might be deployed here. For example, some philosophers believe they candefend cognitivism and thereby establish that moral claims are objective with-out having to commit to a traditional version of realism. This strategy is notavailable to a methodology that espouses a freestanding political liberalism.Political liberalism needs some account of how the concepts and principlesthat make possible the distinction between acceptable and unacceptable rea-sons for affirming liberalism are justified concepts and principles. However, inorder to remain freestanding, political liberalism must refrain from advancinga theory of justification that can serve to ground its own moral foundation.The substitution of "reasonable" for "true" does nothing to allay the chargethat political liberalism offers a declaration instead of a justification.

My claim here is that political liberalism will be dogmatic if it is presented asfreestanding and moral. My argument for this claim is based on four of thefeatures of political liberalism discussed in this chapter:

(1) Reasonableness is a willingness to seek fair terms for social cooperation.(2) The demand that one be reasonable is a demand that liberalism makes

on citizens independently of their comprehensive doctrines; reasonable-ness is a kind of categorical demand.

(3) Political liberalism applies the principle of toleration to itself and thismeans, according to Rawls, that political liberalism must refrain fromaffirming the truth of its own moral foundation.

34

95

96 The Legacy of John Rawls

(4) Political liberalism is not groundless; it has moral content and this moralcontent can justify liberal demands for reasonableness.

These central features of political liberalism rule out appeals to truth as away to justify the moral demands of liberalism. More importantly, the in-sularity required to formulate a doctrine as freestanding precludes affirminga conception of justification that one could appeal to in order to rebut thecharge that a declaration has been substituted for a justification; politicalliberalism offers a declaration of, rather than justification for, its moralfoundation.

The conclusion that follows from this argument against political liberalismis, in my view, decisive. Unless there is an independent justification fordemanding the priority of the reasonable over the rational, one that can beinvoked to assess the reasons citizens have for affirming a conception of justice,political liberalism will lack a justification for the moral demands it makeson citizens. Invoking an independent justification for why only moral reasonsfor affirming political liberalism presupposes a moral point of view whoseauthority overrides attempts to affirm political liberalism for non-moral rea-sons. This point of view cannot be justified by a freestanding doctrine. Anyjustification for the priority of the reasonable over the rational will requirea moral conception of the person and his or her capacities; and this moralconception has to be defended as part of a moral rather than freestandingdoctrine.

David Estlund has offered an interpretation of political liberalism that onecould use to argue against my position. Estlund claims that on the moststraightforward interpretation, political liberalism cannot be neutral towardits own foundation; it must affirm the truth of its principle of moral justifica-tion. Assuming the conception of truth advanced by liberalism does not spe-cify any controversial claims about the nature of morality, admitting the truthof its own foundation renders political liberalism both coherent and freestand-ing. This renders political liberalism coherent because in affirming the truth ofthose principles and concepts required to express its moral demands on citi-zens, liberalism is given a foundation. Yet political liberalism remains free-standing because it takes no stand on controversial theses about the nature oftruth in moral and political philosophy. Estlund claims this strategy can beused to rebut the charge that the very idea of a freestanding yet moral doctrineis incoherent.

According to Estlund, political liberalism can affirm what he calls a mini-mal conception of truth without being committed to any specific thesis aboutmoral truth. On this view, political liberalism affirms the truth of reasonablerejectability yet offers no metaethical theory ofjustification to substantiate themoral basis for this principle. There are two ways—each has undesirable

Public reason and the moral foundation of liberalism 9 7

consequences—that one might conjoin Estlund's proposal with the idea of afreestanding political doctrine. According to one, justification comes fromcitizens and their comprehensive doctrines (i.e. not political liberalism itself).According to the other, justification is alleged to come from within a free-standing liberal doctrine. I will briefly consider each strategy for affirmingthe truth of political liberalism. I argue that the first interpretation ofEstlund's proposal yields a conception of liberalism that is incompatible withthe idea of a freestanding doctrine. I argue that the second interpretation"declares" rather than "justifies" the moral demands that political liberalismimposes on citizens.

Suppose the truth affirmed by political liberalism is the truth of a free-standing doctrine. Suppose also that citizens who affirm the truth of politicalliberalism do so because, in their view, affirming political liberalism enablesthem to achieve cognitive harmony between their comprehensive doctrinesand a liberal conception of justice. Asserting the truth of political liberalismis therefore licensed by individual citizens deriving the truth of political liber-alism from their comprehensive doctrines. If the comprehensive doctrines onthe basis of which citizens affirm the truth of political liberalism demand thepriority of the reasonable over the rational, then an agreement about justicewill be a moral consensus.

This characterization of the truth of political liberalism raises an importantquestion. If the truth of political liberalism is derivative, does this entail thatpolitical liberalism is not freestanding? It depends. The answer is no if politicalliberalism (or some central feature of political liberalism) does not demand ofcitizens that they affirm its truth. This is the strategy that Rawls wouldprefer. The answer is yes if political liberalism (or some central feature ofpolitical liberalism) is affirmed as true. This is the strategy that would followfrom Estlund's position if it were used to support the derivative status of poli-tical liberalism's truth. Each characterization of political liberalism comeswith a serious cost. The first is merely a restatement of Option 1. The secondis incompatible with the idea of a freestanding doctrine. If the moral founda-tion of political liberalism is affirmed as true and if this truth is inherited fromone or more comprehensive doctrines, then it necessarily follows that inaffirming the truth of the moral foundation of political liberalism one alsodenies that political liberalism is freestanding. There is, however, a secondinterpretation of Estlund's proposal, one that more closely matches the con-ception of political liberalism he wants to defend.

Suppose the truth of political liberalism can be invoked without having toappeal to a comprehensive doctrine that supports political liberalism. On thisview, political liberalism is a freestanding doctrine that affirms the truth of itsmoral content. In his discussion of different conceptions of the foundation ofpolitical liberalism, Estlund offers the following commentary:

98 The Legacy of John Rawls

We might . . . distinguish three possible versions of political liberalism andits foundations: a wholly procedural version avoids appealing to any stan-dard of truth or correctness outside of acceptability to reasonable citizens.A version that appeals to the truth of the acceptance criterion regardless ofits acceptability to reasonable citizens would be a dogmatic substantive lib-eralism. These ought to be rejected in favor of an undogmatic substantivepolitical liberalism in which no doctrine is available in justification unlessit is acceptable to reasonable citizens, not even this doctrine itself (thismakes it undogmatic), because such an acceptability criterion is true or cor-rect independent of such acceptability (this makes it substantive).

In light of the issues discussed in this chapter, this is a very tantalizing pro-posal. It promises to show that one can affirm political liberalism in a waythat satisfies the practical aim of political philosophy as Rawls defines it,while at the same time affirming the truth of the reasonable rejectability cri-terion. Estlund's proposal could therefore be conjoined to the idea of a free-standing liberal doctrine in a way that clearly distinguishes politicalliberalism from Hobbesian liberalism. The priority of the reasonable over therational is a feature of the moral foundation of political liberalism and thisfoundation is affirmed as true. This is a very appealing version of Option 2.

Estlund's attempt to remain faithful to Rawls's claim that political philoso-phy should be practical, while rejecting Rawls's claim that political liberalismcan remain neutral regarding the truth of its own foundational principle ofjustification, is a well-motivated attempt to defend an improved version ofRawls's project. Unfortunately, the methodological strategy Estlund advo-cates is itself open to the charge of dogmatism. This is a result of his attemptto combine "true" with the idea of political liberalism. The claim that aconception of political liberalism is undogmatic because it affirms that "nodoctrine is available in justification unless it is acceptable to reasonable citi-zens" does not suffice to show that the moral foundation of liberalism isundogmatic. Rather, it shows only that if one affirms political liberalism,then one can undogmatically demand of citizens that they offer reasons thatothers cannot reasonably reject. But what is the justification for affirming poli-tical liberalism in the first place? If one's response to this question is, "becausethe moral foundation of political liberalism is true" then one had better havesome justification to back up this claim. According to Estlund, political liber-alism need only assert its foundational principle ofjustification to be true in theminimal sense. However, if reasonable rejectability is affirmed as true, thenit must be reasonable to ask "what entitles political liberalism to affirm thepriority of the reasonable over the rational?" Otherwise, "declaration"rather than "justification" is the most apt characterization of the status of thefundamental moral demand that political liberalism makes on citizens. Yet

38

Public reason and the moral foundation of liberalism 99

the very idea of asserting the truth of reasonable rejectability but only in theminimal sense precludes offering such a justification. Therefore, invoking"truth" to make demands under these restrictions (i.e. that one offers no jus-tification for the claim whose truth is asserted) is a form of dogmatism.Estlund's proposal turns out to be a version of dogmatic liberalism.

On Estlund's view we should loosen the reigns and modify the third featureof political liberalism listed above—that political liberalism must refrain fromaffirming the truth of its own moral foundation—and claim instead that poli-tical liberalism must admit the truth of its own foundational principle of justi-fication. I have tried to show that this is an unsuccessful attempt to salvagepolitical liberalism. Unless one is committed to providing a justification forthe moral foundation of liberalism the charge of dogmatism obtains even ifwe substitute "true" for "reasonable."

Option 3

Option 3 accepts the demand that liberals are required to offer a justificationfor the moral foundation of liberalism. I will not attempt to offer a completejustification here. Instead, I propose a strategy that treats the reasonablerejectability criterion and the capacities of citizens as part of a partially com-prehensive moral doctrine. In calling liberalism a partially comprehensivedoctrine I intend to convey the following: liberalism is a political doctrinethat does not attempt to defend a theory that encompasses all values, yet lib-eralism does regard its moral foundation as justified and true. On this view,liberalism encompasses values that define principles that are to be expressedwithin what Rawls terms the basic structure of society.

Moreover, since on my view liberalism does regard its moral foundation asjustified and thus true, liberalism is not freestanding. The most compellingreason to advocate this conception of liberalism is that it offers a principleddefense of a moral point of view that affirms the priority of the reasonableover the rational. The justification for this point of view rests on a conceptionof citizens, the reasonable rejectability criterion and practical reason. Thisstrategy for defending liberalism also has the advantage of being free of theproblems inherent in Options 1 and 2.

In restricting the class of legitimate reasons for affirming a liberal con-ception of justice to reasons motivated by reasonableness, Rawls invokes acounterfactual constraint: the reasonable rejectability criterion for moral jus-tification. Reasonable rejectability is a normative requirement to which allreasons for affirming political principles must conform. The most well-knownversion of the reasonable rejectability criterion for moral justification isdefended by Thomas Scanlon. Although he acknowledges the similaritiesbetween his own and Scanlon's position, Rawls is adamant in claiming that

41

42

100 The Legacy of John Rawls

political liberalism does not require a commitment to a moral point ofview or principle that extends beyond the scope of the practical and limitednormative aims of political philosophy. Whereas for Scanlon, reasonablerejectability is a contractarian principle of moral justification, for Rawls, rea-sonable rejectability is a feature of a freestanding political liberalism. It isclear, therefore, that Rawls does not intend liberalism's principle of justifica-tion to be interpreted in the way I am suggesting here. The main reason forthis is that reasonable rejectability understood as a criterion for moral justifi-cation would have to be part of a comprehensive moral doctrine because ittakes sides in debates about which reasonable people can disagree. Forexample, one might affirm a comprehensive moral doctrine that rejects con-tractualist models of justification altogether. Political liberalism, in Rawls'sview, should be presented in such a manner that people who hold such viewscan affirm a Kantian conception of justice without having to modify theircomprehensive moral doctrine. However, if neither Option 1 nor Option 2can offer a satisfactory account of this central feature of liberalism, then thebest strategy is to defend a moral position that can provide a justificationfor adopting reasonable rejectability as the criterion of moral justification.Provided this characterization of reasonable rejectability is interpreted aspart of a justification for the moral foundation of liberalism, Rawls's owncharacterization of the capacities of citizens as discussed in the first sectionof this chapter is a good starting point.

Rawls claims that liberalism imputes to citizens a capacity for principle-and conception-dependent desires. If we jettison the claim that liberalism is afreestanding doctrine, then this feature of liberalism is naturally construed as aclaim about practical reason. This commits liberals to a conception of practi-cal reason that affirms the priority of reason over desire. This priority is exhib-ited by the judgement sensitivity of at least some of the desires citizens haveconcerning their political interests and of the demands they are entitled tomake on other citizens regarding such interests. One straightforward implica-tion of the idea of a conception-dependent desire is that of a desire whosemotive comes from something other than a desire. This is exactly the sort ofclaim that liberals should defend as a straightforward, though controversial,claim about persons and their capacities. There is no reason to qualify thiscapacity for practical reason as a capacity of citizens rather than persons ifthe idea of this capacity is understood to be part of the moral foundationfor liberalism.

It is instructive in this context to consider Rawls's analysis of Kant's distinc-tion between empirical and pure practical reason. When an agent deliber-ates about action from the standpoint of instrumental reason he or she adoptsthe point of view of empirical practical reason. The principles of action

44

Public reason and the moral foundation of liberalism 101

considered by empirical practical reason correspond to what Kant termedhypothetical imperatives. By contrast, when an agent deliberates aboutaction from a moral point of view he or she adopts the point of view of purepractical reason. The principles of action considered by pure practical reasonare evaluated by what Kant called the categorical imperative (i.e. the moralpoint of view). This distinction between types of practical reasoning parallelsRawls's distinction between the rational and the reasonable. The most signifi-cant feature of the moral point of view is that it imposes the following restric-tion on practical deliberation: "the reasonable frames the rational and limits itabsolutely." In this way the reasonable frames the point of view from whichpersons can for legitimate reasons affirm liberal principles of justice. Legiti-mate reasons satisfy the reasonable rejectability criterion; illegitimate reasonsdo not. In connecting the categorical demand for reasonableness to a con-ception of citizens and their capacities, including the capacity for practicalreason, the liberal strategy for justifying its moral foundation is a versionof Option 3.

This characterization of practical reason is well-suited for a liberal-contrac-tarian conception of moral justification that avails itself of core Rawlsian con-cepts such as reasonableness and public reason. Charles Larmore provides aclear account of the requirement that one's reasons for affirming principles ofjustice satisfy public and shared criteria of justification:

Publicity really amounts to the demand that the reasons each person has toendorse the principles [of justice] be reasons the person sees others to haveto endorse them as well. It requires that the principles ofjustice be groundedin a shared point of view.

A similar claim is offered by Habermas when he claims:

reasonable citizens cannot be expected to develop an overlapping consensusso long as they are prevented from jointly adopting a moral point of viewindependent of, and prior to, the various perspectives they individuallyadopt from within each of their comprehensive doctrines. The notion ofreasonableness is either .. . too weak to characterize the [justification fora] conception of political justice, or it is defined in sufficiently strong terms,in which case what is ... reasonable is indistinguishable from what ismorally right.

Rawls of course claims that political liberalism can be understood to rest on ashared point of view so far as reasonable persons can affirm the same principlesofjustice and thus participate in an overlapping consensus. However, since

48

49

102 The Legacy of John Rawls

Options 1 and 2 yield unacceptable results, this shared point of view cannot be

understood as part of a freestanding conception of justice. Rather, the shared

point of view of practically reasoning agents must be understood as a precon-

dition for a moral agreement among reasonable persons. Furthermore, the

moral constraints imposed by this point of view should be affirmed nondogma-

tically; they should be justified rather than merely declared.

Combined, the reasonable rejectability criterion and the characterization

of citizens and their capacities should be presented as a way to defend the

special status of persons. For partially comprehensive liberals, reasonable

rejectability is a contractarian principle of justification that satisfies—better

than alternative principles—a commitment to "respect for persons," a long-

standing ideal of liberalism. Option 3 treats the conception of the person as

part of a defense of the reasonable rejectability criterion for moral justifi-

cation. And the defense of this criterion appeals to considerations that must

be argued for on philosophical, rather than freestanding political grounds.

Conclusion

Rawls's contribution to contemporary liberal political philosophy is both sub-

stantive and methodological. On the one hand, Rawls's conception of justice

defines the current framework of thought for egalitarian liberalism. On the

other hand, the position advanced in Political Liberalism presents a novel view

about how liberal principles should be defended in modern, pluralistic socie-

ties. In this chapter I have been concerned with this aspect of the Rawlsian

legacy. My objections to the idea of a freestanding political doctrine take noth-

ing away from Rawls's egalitarian liberalism. The substantive conception of

justice defended by Rawls is, in my view, the best version of egalitarian liberal-

ism. A Theory of Justice and Political Liberalism are as yet unmatched as a defense

of a substantive liberal conception of justice. It is important, however, that

liberals offer the best possible justification for their position.

If the argument presented in this chapter is sound, then the moral founda-

tion for liberalism cannot be defended as part of a freestanding political doc-

trine. This does not mean that liberalism should be defended as a complete or

full-blown comprehensive doctrine that aspires to define values for human life

as such. Liberalism is a political doctrine in the sense that it defines the values

that underlie public institutions and the ideals of citizenship. These political

values, however, require a moral justification that is not provided by a meth-

odology that treats the basic moral concepts of liberalism as freestanding. Lib-

eralism is one moral doctrine among others and must be accepted or rejected

in light of traditional criteria of moral justification.

50

51

Public reason and the moral foundation of liberalism 103

Notes

1. PI, 9-10.2. According to Rawls, a doctrine "is comprehensive when it includes conceptions of

what is of value in human life, and ideals of personal character, as well as idealsof friendship and of familial and associational relationships, and much else that isto inform our conduct, and in the limit to our life as a whole." (PL, 13) By con-trast, "Political liberalism . .. aims for a political conception of justice as a free-standing view. It offers no specific metaphysical or epistemological doctrinebeyond what is implied by the political conception itself." (PL, 10)

3. Jean Hampton, "Should Political Philosophy Be Done Without Metaphysics?"Ethics99July (1989), 791-814 (791-2).

4. Hampton, "Political Philosophy," 800.5. PL, 15.6. David Estlund, "The Insularity of the Reasonable: Why Political Liberalism

Should Admit the Truth," Ethics 108 (January 1998), 252-75 (253).7. Jiirgen Habermas, "Reasonable Versus True: Or the Morality of Worldviews,"

in P. de Grieff and C. Cronin (eds), The Inclusion of the Other (Cambridge, MA:MIT Press, 1998), 75-101 (88-9).

8. These objections have been advanced by, among others, Estlund, "The Insularityof the Reasonable;" Habermas, "Reasonable Versus True;" Hampton, "Politi-cal Philosophy;" Charles Larmore, The Morals of Modernity (New York: Cam-bridge University Press, 1996) and "The Moral Basis of Political Liberalism,"Journal of Philosophy XCVI.12 (1999), 599-625; Joseph Raz, "Facing EpistemicDiversity: The Case of Epistemic Abstinence," Philosophy and Public Affairs 19(1990), 3—46; Kok-Chor Tan, Toleration, Diversity, and Global Justice (UniversityPark: Pennsylvania state University Press, 2000); and Leif Wenar, "Political Lib-eralism: An Internal Critique," Ethics 106 October (1995), 32-62.

9. PRR in LP, 129-80.10. "Rawls refers to reasonable principles ofjustice, reasonable judgments . . . a rea-

sonable overlapping consensus .. . reasonable norms . . . the virtue of reasonable-ness . . . reasonable pluralism . .. reasonable agents or persons" among manyother senses of reasonableness." (Wenar, "Political Liberalism," 32)

11. PI, 62.12. PRR, 165.13. The conception of equality affirmed byjustice as fairness is: "social and economic

inequalities are to be arranged so that they are both (a) to the greatest expectedbenefit of the least advantaged and (b) attached to positions and offices open toall."(TJ, 72)

14. PRR, 135.15. Rawls states that public reason "has five different aspects: (1) the fundamental

political questions to which it applies; (2) the persons to whom it applies (govern-ment officials and candidates for public office); (3) its content as given by a familyof reasonable political conceptions ofjustice; (4) the application of these concep-tions in discussion of coercive norms to be enacted in the form of legitimate law for

104 The Legacy of John Rawls

a democratic people; and (5) citizens" checking that the principles derived fromtheir conceptions ofjustice satisfy the criterion of reciprocity." (PRR, 133)

16. PRR, 172.17. JF, 187-8.18. PI, 47-88.19. Ibid.,81.20. WzW.,81-6.21. For an instructive account see Jonathan Dancy, Moral Reasons (Oxford, UK:

Blackwell, 1993), 9.22. Rawls discusses the similarities between his conception of reasonableness and rea-

sonable rejectability in PL, 49, 85.23. Quoted by Larmore, Morals of Modernity, 151.24. Quoted in Ibid., 169.25. Rawls also claims that reasonable citizens accept the burdens ofjudgement. The

burdens ofjudgement include the very demanding evidentiary and normative cri-teria that can rightly be invoked in moral argument. Rawls claims there are six:(1) Evidence: evidence relevant to resolving disagreement is "conflicting and com-plex, and thus hard to assess and evaluate." (2) Relevance: evaluating the respec-tive relevance of various considerations often results in conflicting and complexjudgement. (3) Indeterminacy: all moral and political concepts "are vague andsubject to hard cases." (4) Personal experience: in modern society diversity of reli-gion, culture, conceptions of the good life, etc. will shape persons' moral sensibil-ities in different ways. (5) Normative considerations: conflicting comprehensivedoctrines and kinds of normative considerations—moral, legal, religious, and poli-tical, for instance—are an unending source of disagreement. (6) Restrictions onvalues that can be realized within any society. Following Isaiah Berlin, Rawlsclaims there are inevitable limits to the range of values that can be pursued andrealized within a political society. Rawls argues that political liberalism but notcomprehensive liberalism satisfies the burdens ofjudgement. (PL, 54—8)

26. John Locke, Letter Concerning Toleration (The Hague: M. Nijhoff, 1963).27. This phrase is used by David Estlund to characterize a version of liberalism that

"appeals to the truth of the acceptance criterion [i.e. reasonable rejectability]regardless of its acceptability to reasonable citizens." (Estlund, "The Insularityof the Reasonable," 256) I characterize dogmatic liberalism more broadly toinclude any version of liberalism that asserts a moral foundation without provid-ing a justification for this foundation.

28. For a version of this argument see Habermas's "Reasonable Versus True," 77—8.29. JF, 21.30. Ibid., 188-9.31. John Rawls, "Justice as Fairness: Political not Metaphysical," in CP, 388-414

(394).32. Daniel A. Dombrowski, Rawls and Religion: The Case for Political Liberalism

(Albany: state University of New York Press, 2001), 84.33. See for instance Christine Korsgaard's Sources of Normativity (New York: Cam-

bridge University Press, 1996).

Public reason and the moral foundation of liberalism 105

34. One might object at this point and claim that I am ignoring Rawls's Kantian con-structivism, which is the method of justification used to defend the original posi-

tion as an ideal bargaining situation and as a device for selecting the two

principles of justice. However, in Political Liberalism Kantian constructivism ispresented as part of a freestanding conception rather than as a justification for

the claim that liberalism should be freestanding.

35. Estlund, "The Insularity of the Reasonable."36. Estlund claims "we can distinguish between minimal and substantial senses of

truth. A statement P is true in the minimal sense if and only if P. 'All people areequal' is true in the minimal sense if and only if all people are equal." (Estlund,

"The Insularity of the Reasonable," 263)

37. Rawls claims that "if any of those reasonable comprehensive doctrines supports

only true moral judgements, the political conception itself is correct, or close

thereto, since it is endorsed by a true doctrine. Thus, the truth of any one doctrine

in the consensus guarantees that all the reasonable doctrines yield the right con-ception of political justice, even though they do not do so for the right reasons as

specified by the one true doctrine." (PL, 128) It is important to stress that thisclaim does not imply that one can affirm the truth of political liberalism as a

reason for why others must accept it as a conception of justice; at least not onRawls's interpretation of political liberalism.

38. Estlund, "The Insularity of the Reasonable," 256.39. Ibid.

40. Ibid., 270-1.

41. For us the primary subject of justice is the basic structure of society, or more

exactly, the way in which the major social institutions distribute fundamentalrights and duties and determine the division of advantages of social cooperation.By major institutions I understand the political constitution and the principal

economic and social arrangements." (TJ , 6)

42. Thomas Scanlon, What We Owe To Each Other (Cambridge, MA: Harvard Uni-

versity Press, 1998).

43. PL, 49-50.44. I do not intend to imply that on my view liberalism should be unconcerned with

the fact of pluralism. Rather, Option 3 affirms a conception of liberalism that is

committed to providing a principled moral justification for why the exercise ofpolitical power is to be constrained by a liberal conception of toleration. I have

not addressed the question of how liberalism, on my conception, should construeliberal toleration. It does follow from my argument that liberalism requires a

moral justification for its conception of toleration and that this justification

cannot be part of a freestanding doctrine.45. This expression is from Scanlon, What We Owe, 18—22.

46. LUMP, 230-2.47. Ibid., 230-1.

48. Charles Larmore, "Public Reason," in Samuel Freeman (ed.), The Cambridge

Companion to Rawls (New York: Cambridge University Press, 2003), 368-93 (371).

49. Habermas, "Reasonable Versus True," 77.

106 The Legacy of John Rawls

50. For a version of this argument, see Larmore, The Morals of Modernity and "TheMoral Basis of Political Liberalism."

51. Draft versions of this paper were presented at the 2003 Eastern Division Meetingof the American Philosophical Society in Washington, DC, Auburn University,Kansas state University, and the University of North Florida. I am grateful toaudiences at each location for helpful criticisms and suggestions. I also thanktwo referees for the Journal of Moral Philosophy for providing a number of helpfulsuggestions.

Dilemmas of public reason: pluralism, polarization, and instability

Rawls has not adequately resolved the dilemmas created by the "irreconcil-able values" that are the starting point of his political liberalism.

Rawls's freestanding liberalism

A little over a decade ago, John Rawls launched a searing criticism of liberalpolitical theory. The force of the criticism is captured by Rawls's claim that"the question the dominant tradition has tried to answer has no answer."By this, Rawls meant that the traditional aspiration of liberal theory to pro-vide a firm philosophical foundation for liberal politics was profoundly mis-guided. According to Rawls, this aspiration, common to Locke, Kant, Mill,and others, is misguided due to the fact of reasonable pluralism. Reasonablepluralism is the claim that moral epistemology is such that conflicts betweenfundamental philosophical, moral, and religious views—what Rawls calls"comprehensive doctrines"—are intractable even among fully rational per-sons. Since liberalism is committed to a conception of legitimacy according towhich the political order must be justifiable "to every last individual," andsince the fact of reasonable pluralism means that there is no comprehensivedoctrine that can win the consent of all reasonable citizens, a radical conclu-sion follows: The project of trying to propose a philosophical justification forliberal politics is inherently self-defeating.

Accordingly, in Political Liberalism and related work, Rawls proposedto apply "the principle of toleration to philosophy itself." This is done by jet-tisoning the philosophical aspiration of liberal theory; rather than proposingphilosophical premises from which liberal commitments are supposed tofollow, the political liberal begins by "looking to the public culture" of amodern liberal society and attempts to identify a "shared fund of implicitlyrecognized basic ideas and principles." In this way, Rawls proposes a liberal-ism that is "freestanding." The objective of a freestanding liberal theory,then, is not to provide a proof of liberalism, but to "formulate [characteristi-cally liberal] ideas and principles clearly enough to be combined into a politi-cal conception of justice congenial to our most firmly held convictions."

6

Robert B. Talisse

1

2

5

9

108 The Legacy of John Rawls

According to Rawls, the task of collecting and systematizing the principlesimplicit in a liberal society is the "most we can expect" from liberal philoso-phy, "nor do we need more."

Rawls's political liberalism has been criticized from all sides. Liberal theor-ists such as Jean Hampton, Ronald Dworkin, and Samuel Schemer haveargued that a freestanding liberalism is an impoverished liberalism, a liberal-ism that in the end cannot give anyone a reason to be a liberal. Other theor-ists, both liberal and otherwise, have argued that it is strictly impossible forliberal theory to "stay on the surface, philosophically speaking," becausethere are substantive and contestable philosophical claims latent within the"considered convictions" from which Rawls begins.

These debates are still very much alive in the literature, and it is difficult todiscern whether there will be any clear victor. Hence my argument in the cur-rent chapter deviates slightly from the typical critical strategies. Here I shalllaunch an internal objection to Rawls's political liberalism. The argumentwill focus on Rawls's conception of public reason and will show that politicalliberalism must fail according to its own criteria of success. More specifically,the argument will show that whereas Rawls invokes public reason as amechanism for maintaining political stability, recent work by Gass Sunsteinshows that there are sound empirical reasons for thinking that the strictures ofpublic reason will in fact generate increasing levels of instability.

Although the argument targets Rawls's position in particular, the generalline of objection can be brought against a variety of views of public discourse incurrency. The Rawlsian models of proper public discussion proposed byThomas Nagel, Bruce Ackerman, Stephen Holmes, and Charles Lar-more are vulnerable to the criticism I shall raise. Additionally, influentialversions of deliberative democracy, such as those offered by Joshua Cohen,and Amy Gutmann and Dennis Thompson, are subject to my criticism.It may also be the case that the discursivisms ofjiirgen Habermas and Karl-Otto Apel are similarly jeopardized. I of course cannot argue for all of theseclaims in a single chapter; the point is that a successful critique of Rawls willimplicate many other theorists.

Background: stability and consensus

First some background must be set in place. The main question driving Politi-cal Liberalism is that of how it could be possible for citizens to achieve consensuson a single conception of political justice under conditions of reasonable plur-alism. One obvious answer is what Rawls called a modus Vivendi agreement.Where liberalism is adopted as a modus vivendi, each citizen sees the politicalorder as an acceptable compromise between the unattainable best state of

11

18

19

24

25

Dilemmas of public reason: pluralism, polarization, and instability 109

affairs—namely, a politics based entirely on his own comprehensive doc-trine—and the avoidable worst state of affairs—namely, a political orderthat completely reflects a doctrine that opposes his own. Although the sugges-tion that justice prevails where no one gets what he most wants, but everyoneavoids what he most fears enjoys a distinguished lineage stretching throughHobbes back to Plato's Glaucon, it is rejected by Rawls. According to Rawls,a modus vivendi agreement can last only for as long as there is a relative balanceof power among the conflicting doctrines. Since a modus vivendi liberalism isreally nothing more than a truce, it can provide no motivation for an indivi-dual to uphold the agreement should the balance of power turn decisively inhis favor. A modus vivendi liberalism is "political in the wrong way," andhence inherently unstable.

By contrast, Rawls envisioned a liberalism that could win an overlapping con-sensus among citizens. Where a liberal order is endorsed in an overlapping con-sensus, each individual sees the liberal conception as the appropriateexpressions of his own comprehensive doctrine in the political sphere. Thus,although citizens do not share a common justificatory account of liberalism,each supports the political order "for its own sake" and "on its ownmerits"; consequently, individuals will continue to endorse the liberalorder regardless of the balance of power among contending comprehensivedoctrines. In this way, political stability is possible under conditions of plural-ism. Since only a freestanding liberalism can win an overlapping consensus,only a freestanding liberalism can be stable.

Public reason in freestanding liberalism

Rawls realized that politics is not exhausted once an overlapping consensus isachieved: laws must be made, campaigns waged, elections held, votes cast,and cases decided. In these endeavors, too, citizens confront the obstacles toagreement posed by the fact of reasonable pluralism. Hence Rawls proposeda mode of public political discourse in which citizens "conduct their funda-mental discussions within the framework of what each regards as a politicalconception of justice based on values that the others can reasonably beexpected to endorse." In public political discussion, then, citizens "shouldbe ready to explain the basis of their actions to one another in terms eachcould reasonably expect that others might endorse as consistent with theirfreedom and equality." This means that, as in political justification gener-ally, citizens "are not to appeal to comprehensive religious and philosophi-cal doctrines" in properly public discussion.

Of course, certain questions cannot be understood except in terms of deepphilosophical commitment. These will be questions about which we should

28

29

30

33

110 The Legacy of John Rawls

expect citizens to disagree sharply. In light of this, Rawls contends that "a lib-

eral view removes from the political agenda the most divisive issues, serious

contention about which must undermine the bases of social cooperation."

Democratic deliberation hence applies not to public policy questions gener-

ally, but only to what Rawls calls "constitutional essentials" and "questions

of basic justice." Hence Rawls identifies the supreme court as the "exem-

plar" of public reason.

Citizens holding comprehensive doctrines that compel them to seek a

politics based in "the whole truth" who consequently insist upon appeal-

ing to nonpublic reasons in public discourse fail to recognize the duty of

civility, and so fall short of the "ideal of democratic citizenship." More

importantly, citizens who reject the idea of public reason in this way are ipso

facto unreasonable, and consequently may be dealt with coercively. Rawls

writes: "[A] given society may also contain unreasonable, irrational, and

even mad, comprehensive doctrines. In their case the problem is to contain

them so that they do not undermine the unity and justice of society."

As this brief survey shows, public reason places restrictions on the agenda and

the vocabulary of democratic deliberation; political liberalism is in this way an

instantiation of what Stephen Holmes has called "the politics of omission."

For the present purposes, it is important to observe also that public reason

aims at an agreement that is nonepistemic. To explain: When operating

within public reason, citizens do not aim for outcomes based on the epistemi-

cally best reasons, but rather for the outcome that best concurs with the basic

judgements and intuitive liberal principles already assumed. That is, public

reason takes the basic commitments of a liberal society as "fixed points" and

requires—as a condition of reasonableness—that citizens' contributions to

public discourse recognize them. In other words, on the Rawlsian view, the

reasonablenss of a doctrine or belief is primarily a matter of content rather

than justification, and the reasonableness of a person is a matter of which views

he holds rather than of the extent to which he supports his views by reasons. In this

sense, public reason is an epistemically closed system; no antiliberal position,

no matter how tightly argued or well-supported, could be reasonable in a

Rawlsian liberal order. In fact, there could be no properly public discussion

about the merits of liberalism itself, and no public discussion about the appro-

priateness of public reason as a model of liberal political discourse.

The exclusion objection

There is a common line of criticism that attacks public reason for being

inherently exclusionary. In an ironic instance of overlapping consensus, the

34

35

36

37

38 39

40

42

44

45

48

Dilemmas of public reason: pluralism, polarization, and instability 111

"exclusion objection," as we shall call it, is found in the work of theoristswho otherwise agree on very little else. For example, radical democrats suchas Seyla Benhabib, Nancy Frazer, and Iris Young argue that sincepublic reason proceeds from a strict delineation of a properly "political" voca-bulary, it implicitly privileges the status quo. Public reason therefore comesto nothing more than an apologia for existing power structures that crowdout and silence the voices and concerns of the less powerful. Natural Lawtheorists Robert George and Christopher Wolfe agree. They claim: "Publicreason . . . almost always has the effect of making the liberal position thewinner in morally charged political controversies. It does this in effect byruling out of bounds substantive moral argument on behalf of nonliberal posi-tions." The civic republican Michael Sand el has lamented the "politicalcosts" of public reason, arguing that, "Public reason is too spare to containthe moral energies of a vital democratic life." And the worry is not restrictedto antiliberal theorists; liberals, too, have criticized public reason. The liberalWilliam Galston has expressed his concern: "It is difficult to imagine that anyliberal democracy can sustain conscientious support if it tells millions of itscitizens that they cannot rightly say what they believe as part of democraticpublic dialogue."

The exclusion objection has been met with both a clarification and a quali-fication. As for clarification, Charles Larmore explains:

Rightly conceived, [public reason] does not thwart the uninhibited politicaldiscussions which are the mark of vigorous democracy. We can argue withone another about political issues in the name of our different visions of thehuman good while also recognizing that, when the moment comes for a leg-ally binding decision, we must take our bearings from a common point

r • 56oi view.

The restrictions of public reason apply only to decision-making contexts, notpolitical discussion generally, hence the requirement that "the most divisiveissues" be removed from the public political agenda does not quell discussionamong citizens in nonpublic domains. In fact, like Larmore, Rawls affirmsthat lively debate about controversial issues, conducted by means of nonpub-lic reasons, is a vital activity within the "background culture" of a liberaldemocracy. Now for the qualification. In work following Political Liberalism,Rawls introduced "the proviso" as a revision of public reason's vocabularyrestrictions. According to the proviso, citizens in properly public discussionmay invoke reasons drawn from their comprehensive doctrines provided thatthey are prepared "in due course" to offer public reasons to supplement thenonpublic ones.

112 The Legacy of John Rawls

This rejoinder is not satisfying. Although the exclusion objection is oftenformulated as to suggest that it is aimed at public reason's agenda and voca-bulary restrictions, it should be understood instead to be aimed at the non-epistemic character of public reason. What public reason excludes is not theradical democratic, Thomist, and civic republican positions, but rather the rea-sons associated with those doctrines. More precisely, public reason cannotrecognize a Thomist's reasons as reasons. Accordingly, even an irrefutableproof of the Thomistic doctrine of ensoulment is insufficient to render reason-able a Catholic's public opposition to abortion. Presumably, this is due tothe fact of reasonable pluralism, which has it that a logically sound demonstra-tion of x is insufficient for a proof of the falsity of all views inconsistent with x.

But why should our Thomist, or anyone else for that matter, endorse such apluralism? Here, the political liberal must be careful. He cannot offer a philo-sophical argument for pluralism—to do so would be to violate the very idea of afreestanding liberalism. The question of why one should accept reasonablepluralism is a question to which political liberalism "does not speak." Thiswill strike the Thomist as dishonest and hypocritical; however, to object to thepolitical liberal's silence is to be unreasonable, and thus someone the liberalstate must endeavor to "contain."

In this way, public reason is epistemically exclusionary; regardless of howwidely it is construed, public reason cannot acknowledge the epistemic force ofthe arguments advanced in favor of nonliberal positions, and cannot givereasons why the force of those arguments should be disregarded. Rawls's pro-viso confirms this: that persons who advance nonpublic reasons in politicalcontexts are bound to supply public reasons "in due course" indicates thatnonpublic arguments ultimately can do no justificatory work no matter how epis-temically sound they may be. Although public reason may allow everyone avoice, it grants a hearing to only a few.

The political liberal will concede this point but question its critical force.He may argue as follows: Citizens who insist on presenting arguments thatpresume the truth of their own comprehensive doctrines are failing at properdemocratic citizenship because they implicitly reject the fact of reasonablepluralism, recognition of which is necessary for the stability of a liberalregime. Thus the insistent Thomist, civic republican, and radical democratare all destabilizing forces, and surely a liberal democracy, like any regime,should be expected to endeavor to secure its own stability. So whereas it maybe the case that public reason cannot duly recognize the epistemic merits of allviews, there is no reason why it should be required to do so.

This is a cogent reply, but it confronts a difficulty: There is good reason toexpect that the nonepistemic character of public reason will generate instabil-ity. The argument for this claim draws from some recent work in what wemight call "group epistemology," to which we now turn.

Dilemmas of public reason: pluralism, polarization, and instability 113

Group polarization

Gass Sunstein has called attention recently to the statistical regularity knownas group polarization. Group polarization means that "members of a deliberatinggroup predictably move toward a more extreme point in the direction indi-cated by the members" predeliberation tendencies." The term "extreme"here does not refer to points on a spectrum of opinion; it is rather denned intern-ally, that is, only by reference to persons" doxastic tendencies prior to discus-sion. Simply put, "like-minded people, after discussions with their peers, tendto end up thinking a more extreme version of what they thought before theystarted to talk." Citing Sunstein's own examples, consider that,

(1) A group of moderately profeminist women will become more stronglyprofeminist after discussion.

(2) After discussion, citizens of France become more critical of the Unitedstates and its intentions with regard to economic aid.

(3) After discussion, whites predisposed to show racial prejudice offer morenegative responses to the question of whether white racism is responsiblefor conditions faced by African-Americans in American cities.

(4) After discussion, whites predisposed not to show racial prejudice offermore positive responses to the same question.

Group polarization "has been found all over the world and in many diversetasks" and does not discriminate along educational, class, ethnic, gender, orpolitical lines; it has been shown to be operative in judicial panels, legisla-tures, political parties, religious organizations, and civic groups. Moreover,the polarization effect is greatly amplified in cases of "enclave deliberation,"which is "that form of deliberation that occurs within more or less insulatedgroups, in which like-minded people speak mostly to each other over extendedperiods of time."

Group polarization shows that deliberative bodies of like-minded personsare epistemically unstable. The concern here is not the instability as such —change of belief is not necessarily bad. Rather, the danger is that the shifts inbelief occur only in one direction and without regard for reasons. That is,when a group polarizes, the members come to adopt increasingly moreextreme versions of their former positions, and this movement is not occa-sioned by the introduction of better arguments. When groups polarize it isbecause of social dynamic features, not reasons.

According to Sunstein, the antidote to polarization is a vibrant "culture offree speech" that prizes or even rewards dissent. Such a culture employsblocks to epistemic insularity and takes positive steps to ensure that citizensare exposed to appropriately rich "argument pools." Sunstein offers several

114 The Legacy of John Rawls

interesting policy recommendations designed to counteract polarization;

however, we cannot discuss them here.

Public reason, group polarization, and instability

That public reason is likely to generate deliberative enclaves should be plain.

Citizens are very deeply committed to comprehensive doctrines that conflict

with political liberalism on several levels. Citizens of faith present a conspicu-

ous, though not the only, example. Many religious believers hold not only that

abortion is a grave moral evil, but also that their opposition to abortion must

not be relegated to the "background culture" of society. As Nicholas Wolter-

storff has argued, for many liberal democratic citizens, "their religion is not

... about something other than their social and political existence; it is alsoabout their social and political existence." Interestingly, the situation is not

altogether different for the radical democrat, who disagrees with the political

liberal not only on substantive issues of justice and equality, but also on the

question of the very nature of the political. That is, part of what is at stake in

our most central controversies is the "character of public life itself, as well asthe meaning and scope of accepted political values." Public reason cannot

countenance a public deliberative space in which these fundamental issues

can be reasonably engaged; it must "put the 'values' of public reason beyond

political contestation." That is, public reason must generate epistemically

excluded groups.

Consider now the predicament of those persons that political liberalism

epistemically excludes: Believing, correctly, that there is no point in raising

their arguments in public, they will likely form small groups devoted to the

advancement of their position; these groups will meet regularly to discuss the

group's views and devise strategies for disseminating their message. Condi-

tions will be ripe for polarization. As the groups polarize, individuals will not

only come to hold more extreme versions of their initial position, but will come

to see themselves as excluded, victimized, and oppressed; naturally, they will

also grow increasingly dismissive of opposing views, and will regard those that

affirm them as either evil or benighted. In this way, polarized groups are also

epistemically crippled; that is, they grow increasingly unable and unwilling

to engage in reasoned discussion with those with whom they disagree. Fanati-

cism will set in, the overlapping consensus will give way to a modus vivendi, and

hence precisely the kind of instability Rawls sought to avoid will result.

More importantly, a different kind of instability is likely to emerge, namely,

the kind associated with hatred and violence.

It may be objected that I have been merely speculating. However, if it

will be granted that certain regions of our public political discourse closely

Dilemmas of public reason: pluralism, polarization, and instability 115

approximate the model of public reason, it can be shown that my specula-tions about epistemic exclusion are not implausible.

Carol Swain has recently published an alarming study of what she calls the"new'' white nationalist movement in contemporary America. Swain's ana-lyses are based on interviews conducted with ten prominent white nationalists;the transcripts of these interviews are available in a book edited by Swain andRuss Nieli. An examination of the strikingly similar narratives offered by thewhite nationalists reveals the pattern described above. For example, both Wil-liam Pierce, the recently deceased founder of the neo-Nazi National Allianceand author of the infamous novel that inspired Timothy McVeigh to terror-ism, and Lisa Turner, the Women's Information Coordinator of the whitesupremacist World Church of the Creator, claim to have been motivated bywhat they perceived to be a systematic refusal on the part of mainstreamsociety to engage their ideas. It is important to note that the complaint is notthat people were not convinced of their positions, but rather that the publicspace of political reason giving was closed to them; hence they characterizemainstream white America not as mistaken about race, but as "brainwashed,""conditioned," and "propagandized." They thus were forced to "build"their own "infrastructure" for disseminating their ideas. They both pointto the Internet as the most effective recruitment tool. Perhaps not surpris-ingly, Sunstein has shown that, insofar as it enables individuals to pre-selectand filter the information to which they will be exposed, the Internet is apowerful source of polarization.

Swain explicitly draws the connection I have suggested between epistemicexclusion and group polarization:

Sunstein's analysis [of group polarization] seems to describe somethingclearly at work among many of the white nationalist leaders interviewed... I believe that one reason why many of the members and potential mem-bers of their organizations have such little exposure to alternative view-points is because of the overall feebleness and lack of honesty that currentlydominates discussion about controversial racial issues in America.

Believing that America is "increasingly at risk of a large-scale racial con-flict," Swain makes a recommendation similar to Sunstein's: "What is mostneeded now ... is for white nationalists to be heard and debated in main-stream forums where their data and ideas can be openly evaluated and sub-jected to critical assessment." Despite the maneuvers designed to loosen theagenda and vocabulary restrictions of public reason, its nonepistemic charactermeans that the kind of debate called for by Swain is not possible within politi-cal liberalism. Nor are public debates concerning a wide array of other contro-versial moral issues. Stating Swain's and Sunstein's point more generally,Sandel observes:

116 The Legacy of John Rawls

Where political discourse lacks moral resonance, the yearning for a publiclife oflarger meanings finds undesirable expressions. Groups like the "moralmajority" and the Christian right seek to clothe the naked public squarewith narrow, intolerant moralisms. Fundamentalists rush in where liberalsfear to tread.

To summarize: In the name of pluralism and the "absolute depth" of the"latent conflict" between citizen's comprehensive doctrines, political liber-alism devises for the sake of stability a politics lacking wholly in political con-testation. But as I hope to have shown, a politics without contestation is anepistemically unstable politics, and an epistemically unstable politics is notonly politically unstable in Rawls's sense, but also dangerously volatile.

A proposal sketched: deliberative agonism

A decade ago, John Rawls taught us to place the following question at theheart of our political theorizing: Is it possible for persons divided at fundamen-tal levels to nonetheless reach a stable consensus on a single conception of poli-tical justice, and, if so, how? Today Rawls's question confronts us not merelyas philosophers, but as citizens. I have argued that the answer provided byRawls is insufficient; however, I accept Rawls's arguments that a traditionalliberalism is nonviable and a modus Vivendi is unstable. There must be someadditional option for democratic politics. Although there is little space left tofashion a positive alternative in any detail, I shall sketch an approach that Icall "deliberative agonism."

Like political liberalism, deliberative agonism acknowledges the fact of rea-sonable pluralism and so rejects traditional modes of liberal theorizing. Thedeliberative agonist agrees further with the political liberal that "to justifythe exercise of collective political power is to proceed on the basis of a freepublic reasoning among equals," and so is committed to a model of democ-racy that is deliberative or reason-based rather than aggregative or prefer-ence-based. However, contra the political liberal, the agonist does not beginfrom a distinction between the "reasonable" and the "unreasonable," andso does not circumscribe a "special domain of the political" that is cleansedof controversy, and moreover does not stipulate a special mode of "public"reasoning. Instead, the deliberative agonist endorses an epistemic view of poli-tical discourse and a "decentered" conception of the political that counte-nances multiple and interrelated civic spheres as sites of fully public dialogue.Consequently, the deliberative agonist rejects the idea underlying politicalliberalism that a stable politics can proceed only from a consensus on basicprinciples. Instead, the deliberative agonist begins from the very justificatory

Dilemmas of public reason: pluralism, polarization, and instability 117

practices of reason-giving that make disagreement, objection, dissent, andcontestation possible.

Hence I follow Claude Lefort in envisioning a democratic politics "foundedupon the legitimacy of a debate as to what is legitimate and what is illegiti-mate." Such a politics "requires a real debate about possible alterna-tives," and so cannot begin from consensus, overlapping or otherwise, atthe level of "substantive beliefs" but rather must work at the epistemic levelof "procedures, processes, and practices for attaining and revising beliefs."The aim of such deliberation is not the implausible one of total consensus, butrather outcomes that are sufficiently based in processes of collective reasoningto motivate even those who disagree with them to continue to participate.In this way, the processes of political deliberation are ongoing; they aim not toovercome conflict, but rather to preserve the conditions under which "agonis-tic confrontation among adversaries" may continue without devolving into an"antagonistic struggle between enemies."

To be sure, I have not claimed that ongoing deliberative processes are guar-antors of stability, progress, or justice. Nor do I pretend that practices ofreason-giving are necessarily free of ideological distortions and structuralinequalities. However, it is a virtue of the epistemic conception of delibera-tion that it can be self-reflexive and thus self-critical. Moreover, though it isepistemic insofar as it aspires to be reason-sensitive, the deliberative agonismI have sketched does not entail any particular epistemology; it is committedonly to what Wilfrid Sellars characterized as "the logical space of giving rea-sons." The self-reflexive and epistemologically mimimalist conception ofdeliberation I have proposed, then, offers to us the chance of retrievingerrors, repairing injustice, exposing sites of illegitimate power, and challen-ging received conceptualizations of all of those terms.

To be sure, a deliberative agonism may in the end fare no better than poli-tical liberalism, for it too will have to confront group polarization and otherchallenges. Alas, to paraphrase Isaiah Berlin, there can be no politics withoutrisk. Thus, if the foregoing arguments are correct, our choice is not between arisky agonism and a secure political liberalism, but rather between a politicsprepared to recognize its risks and one whose hazards include those generatedby its own delusions.

Notes

1. James Bohman, Public Deliberation (Cambridge, MA: MIT Press, 1996), 73.2. PL, 135.3. Rawls's pluralism is epistemic because it is genera ted by what he calls "the burdens

ofjudgment." (PL, 54—58) It is important to distinguish Rawls's pluralism from

118 The Legacy of John Rawls

the recent neo-Berlinian varieties offered in John Gray, Two Faces of Liberalism(New York: The New Press, 2001) and William Galston, Liberal Pluralism (Cam-bridge, UK: Cambridge University Press, 2002), which are robustly metaphysi-cal and thus not of the sort Rawls could endorse. There are questions concerningthe cogency of Rawls's pluralism that cannot be taken up here, but see my "Rawlson Pluralism and Stability," Critical Review 15 (2003), 173-94.

4. Jeremy Waldron, "Theoretical Foundations of Liberalism," in his Liberal Rights(Cambridge, UK: Cambridge University Press, 1993), 37.

5. Note the "fact of oppression," according to which "a continuing shared under-standing on one comprehensive religious, philosophical, or moral doctrine canbe maintained only be the oppressive use of state power." (PL, 37) From this itfollows that all traditional liberalisms would require oppression, and are there-fore self-defeating.

6. PL, 10.

7. Ibid., 8.8. Ibid., 10.

9. Ibid., 8.

10. Ibid., 124.11. John Rawls, "Justice as Fairness: Political Not Metaphysical," reprinted in CP,

388-414.12. Jean Hampton, "Should Political Philosophy be Done Without Metaphysics?",

Ethics W (1989), 791-814.13. Ronald Dworkin, "Foundations of Liberal Equality," in Stephen Darwall (ed.),

Equal Freedom (Ann Arbor: University of Michigan Press, 1995), 190—306.14. Samuel Schefner, "The Appeal of Political Liberalism," in his Boundaries and Alle-

giances (New York: Oxford University Press), 131—48.15. Dworkin, "Foundations of Liberal Equality," 192. Cf. Brian Barry in Culture

and Equality (Cambridge, MA: Harvard University Press, 2001), 331n.27:"Rawls has by now abandoned most of the ideas that made A Theory of Justiceworthwhile."

16. Rawls, "Justice as Fairness," 395.17. PL, 8.18. See David Estlund, "The Insularity of the Reasonable: Why Political Liberalism

Must Admit the Truth," Ethics 108 (1998), 252-75; Chantal Mouffe, The Demo-cratic Paradox (New York: Verso, 2000), 24; and Seyla Benhabib, The Claims of Cul-ture (Princeton: Princeton University Press, 2002), 111. In Democracy's Discontent(Cambridge, MA: Harvard University Press, 1996), Michael Sand el produces afully consistent civic republican reading of the principles and intuitions underly-ing our institutions.

19. My argument shall engage primarily the account of public reason given inRawls's Political Liberalism. To be sure, most of Rawls's later career was devoted toclarifying the doctrine of public reason. (See John Rawls "Commonweal Inter-view," reprinted in CP, 616-22; PRR reprinted in CP, 573-615; and JF, 89ff.)I shall have occasion to discuss these later developments; however, I contend thatthe fundamental core of the view remained unchanged.

Dilemmas of public reason: pluralism, polarization, and instability 119

20. Thomas Nagel, "Moral Conflict and Political Legitimacy," Philosophy and PublicAffairsl6 (1987), 215-40.

21. Bruce Ackerman, "Why Dialogue?" Journal of Philosophy 86 (1989), 16-27.22. Stephen Holmes, "Gag Rules or the Politics of Omission," in his Passions and Con-

straint (Chicago: Chicago University Press, 1995), 202-35.23. Charles Larmore, The Morals of Modernity (Cambridge, UK: Cambridge Univer-

sity Press, 1996).24. Joshua Cohen, "Procedure and Substance in Deliberative Democracy," injames

Bohman and William Rehg (eds), Deliberative Democracy (Cambridge, MA: MITPress, 1997), 407-38.

25. Amy Gutmann and Dennis Thompson, Democracy and Disagreement (Cambridge,MA: Harvard University Press, 1996).

26. JiirgenHabermas, Between Facts and Norms (Cambridge, MA: MIT Press, 1996).27. Karl-Otto Apel, "The A Priori of the Communication Community and the Foun-

dations of Ethics," in his Towards a Transformation of Philosophy (London: Rout-ledge, 1980).

28. PL, 142.29. Ibid., 148.30. Ibid., 226.31. Ibid., 218.32. Ibid., 225.33. Ibid., 224. It is worth noting further that Rawls takes this restriction to apply to

citizens' voting as well; they must not decide how to cast their votes by appealto their comprehensive doctrines. (On this see PL, 215.)

34. Ibid., 157.35. Ibid., 214. Rawls continues, "This means that political values alone are to settle

such fundamental questions as who has the right to vote, or what religions are tobe tolerated, or who is to be assured fair equality of opportunity, or to hold prop-erty. These and similar questions are the special subject of public reason."

36. PL, 216. Thus, "To check whether we are following public reason we might ask:how would our argument strike us presented in the form of a supreme court opi-nion? Reasonable? Outrageous?" (PL, 254)

37. PL, 243.38. Ibid., 217.39. Ibid., 98. As I note below, Rawls later qualified this claim with "the proviso."

(PRR, 584)40. PL, 59.41. "It is unreasonable for us to use political power . . . to repress comprehensive

doctrines that are not unreasonable." (PL, 61) Hence it may be fully reason-able in some cases to use political power to repress unreasonable comprehensivedoctrines.

42. PL, xvi. Cf. "That there are doctrines that reject one or more democratic free-doms is itself a permanent fact of life, or seems so. This gives us the practical taskof containing them—like war and disease—so that they do not overturn politicaljustice." (PL, 64n.l9)

120 The Legacy of John Rawls

43. Benhabib, Claims of Culture, 108.

44. Holmes, "Gag Rules," 202ff.

45. Rawls explicitly claims that justification is a "practical" and not an "epistemolo-gical" problem. (PL, 44)

46. There is a fascinating literature examining whether public reason is capable of pro-

ducing determinate outcomes on questions of basic justice; see especially David

Reidy, "Rawls's Wide View of Public Reason: Not Wide Enough," Res Pubhca 6

(2000), 49-72.47. PL, 124.48. Cf. Bohman, Public Deliberation, 86. According to Rawls, the basic liberal princi-

ples "meet the urgent political requirement to fix, once and for all, the content of

certain political basic rights and liberties, and to assign them special priority.

Doing this takes those guarantees off the political agenda." (PL, 161)

49. Seyla Benhabib, "Toward a Deliberative Model of Democratic Legitimacy," inher (ed.) Democracy and Difference (Princeton: Princeton University Press, 1996),67-94.

50. Nancy Frazer, "Rethinking the Public Sphere," in Craig Calhoun (ed.), Haber-

mas and the Public Sphere (Cambridge, MA: MIT Press, 1992), 122-36.51. Iris Marion Young, Inclusion and Democracy (New York: Oxford University Press),

36ff. See also Young, "Activist Challenges to Deliberative Democracy," injamesFishkin and Peter Laslett (eds), Debating Deliberative Democracy (Oxford, UK:

Blackwell, 2003), 102-20.

52. See Lynn Sanders, "Against Deliberation," Political Theory 25 (1997), 347—76;and Brooke Ackerly, Political Theory and Feminist Social Criticism (Cambridge,

UK: Cambridge University Press, 2000), 52f.53. Robert George and Christopher Wolfe, "Introduction," in Robert George

and Christopher Wolfe (eds), Natural Law and Public Reason (Washington, DC:

Georgetown University Press, 2000), 2. Cf. Stanley Fish, "Mutual Respectas a Device of Exclusion," in Stephen Macedo (ed.), Deliberative Politics (New

York: Oxford University Press, 1999), 91. Consider the infamous footnote

in which Rawls claims that any comprehensive doctrine that would reject awoman's "duly qualified right" to abortion "in the first trimester" is "to that

extent unreasonable"; he further contends that "we would go against the ideal of

public reason if we voted from a comprehensive doctrine that denied this right."

(PL, 243n.32)

54. Michael Sandel, Liberalism and the Limits of Justice, second edn (Cambridge, UK:Cambridge University Press, 1998), 217.

55. William Galston, "Diversity, Toleration, and Deliberative Democracy," in

Macedo (ed.), Deliberative Politics, 43.

56. Charles Larmore, "Public Reason," in Samuel Freeman (ed.), The Cambridge

Companion to Rawls (Cambridge, UK: Cambridge University Press, 2003), 383.57. PL, 157.

58. Ibid., 220.59. Ibid.,\i;PRR, 591.

60. PAR, 591.

Dilemmas of public reason: pluralism, polarization, and instability 121

61. The adequacy of the proviso is disputed. Even Larmore, himself a political lib-

eral, rejects it; see his "Public Reason," 386.

62. See, for example, Joshua Cohen's "Democracy and Liberty," in Jon Elster (ed.),

Deliberative Democracy (Cambridge, UK: Cambridge University Press, 1998),185—231. In particular, see Cohen's discussion of the Papal encyclical Evangelium

Vitae, in which an anti-abortion argument is presented that claims to be

independent of any specifically religious claims. Cohen asserts, without argu-ment, that the Pope appeals to a "conception of reason" that is "itself sectarian."

(Cohen, "Democracy and Liberty," 196)63. Contrast the philosophically robust pluralisms developed in Gray's Two Faces of

Liberalism and Galston's Liberal Pluralism', both theorists criticize Rawls on this

point. Rawls is defended against Gray's criticism in my "Two-Faced Liberal-

ism," Critical Review 14 (2002), 441—58; and Galston is criticized in my "Can

Value Pluralists be Comprehensive Liberals?" Contemporary Political Theory 3(2004), 127-39. See also Galston, "Liberal Pluralism: A Reply to Talisse,"Contemporary Political Theory 3 (2004), 140-7.

64. PL, 128. The political liberal must take a vow of "epistemic abstinence" even

about the epistemic value of his own commitments. On this, see Joseph Raz,

"Facing Diversity: The Case of Epistemic Abstinence," in his Ethics in the Public

Domain (New York: Oxford University Press, 1994), 60-96 and Estlund, "TheInsularity of the Reasonable."

65. PL, xiv. See Stout, Democracy and Tradition (Princeton: Princeton University

Press), 76, who claims that:

One message being preached nowadays in many of the institutions where

future preachers are being trained is that liberal democracy is essentially hypo-critical when it purports to value free religious expression . . . Over the next sev-

eral decades this message will be preached in countless sermons throughout the

heartland of the nation.

Cf. Lucas Swaine, "A Liberalism of Conscience," Journal of Political Philosophy

11 (2003), 369-91.66. "The zeal to embody the whole truth in politics is incompatible with an idea of

public reason that belongs with democratic citizenship." (PRR, 574)

67. Cf. Robert Goodin, Reflective Democracy (New York: Oxford University Press,

2003), 178.

68. Cass Sunstein, "The Law of Group Polarization," in Fishkin and Laslett (eds),Debating Deliberative Democracy, 81.

69. Cass Sunstein, Why Societies Meed Dissent (Cambridge, MA: Harvard University

Press, 2003), 112.

70. Cass Sunstein, Designing Democracy (New York: Oxford University Press,

2001), 23.71. Sunstein, "The Law of Group Polarization," 82.72. Sunstein, Why Societies Need Dissent, 111.

73. Cass Sunstein, Republic.com (Princeton: Princeton University Press, 2001), 75—6.74. Sunstein, Why Societies Need Dissent, 112.

122 The Legacy of John Raw Is

75. Sunstein, "The Law of Group Polarization," 84.76. Nicholas Wolterstorff, "The Role of Religion in Decision and Discussion of Poli-

tical Issues," in R. Audi and N. Wolterstorff, Religion in the Public Square (Mary-land: Rowmanand Littlefield, 1997), 91.

77. Bohman, Public Deliberation, 86.78. Ibid.79. Russell Hardin, "The Crippled Epistemology of Extremism," in Albert Breton

(ed.), Political Extremism and Rationality (Cambridge, UK: Cambridge UniversityPress, 2002), 1-34.

80. Hence Hardin:

Winston Churchill reputedly quipped that fanatics are people who cannotchange their minds and will not change the subject. He got their epistemologyjust right in his first point. But perhaps he got them wrong in his second point.It is not so much that they will not change the subject. Rather, they cannotchange it, because they have no other subject. That is the nature of theircrippled epistemology, without which they would not be fanatics. (Hardin,"The Crippled Epistemology of Extremism," 21)

81. Sunstein, Why Societies Meed Dissent, 12.82. On this, see Sandel, Liberalism and the Limits of Justice, 216.83. Carol Swain, The New White Nationalism in America (Cambridge, UK: Cambridge

University Press, 2002).84. Carol Swain and Russ Nieli (eds), Contemporary Voices of White Nationalism in Amer-

ica (Cambridge, UK: Cambridge University Press, 2003).85. Swain and Nieli, Contemporary Voices of White Nationalism in America, 258, 264.86. Ibid., 261. Cf. Carol Mason, KillingforLife (Ithaca, NY: Cornell University Press,

2003).87. Swain and Nieli, Contemporary Voices, 250, 266.88. Sunstein, Republic.com.89. Carol Swain, The New White Nationalism in America, 10.90. Ibid., 423.91. Ibid.,35.92. Sandel, Liberalism and the Limits of Justice, 217.93. PL, xxvi.94. A more detailed presentation can be found in my Democracy After Liberalism (New

York: Routledge, 2005).95. Joshua Cohen, "Procedure and Substance in Deliberative Democracy," 412.96. Deliberative democracy is often contrasted with aggregative models. See Bernard

Manin, "On Legitimacy and Political Deliberation," Political Theory 15 (1987),338—68; Benhabib, "Toward a Deliberative Model"; John Dryzek, DeliberativeDemocracy and Beyond (New York: Oxford University Press, 2000), Ch. 1; JamesBohman, "The Coming of Age of Deliberative Democracy," Journal of Politi-cal Philosophy 6 (1998), 400—25; Samuel Freeman, "Deliberative Democracy:A Sympathetic Comment," Philosophy and Public Affairs 29 (2000), 371-418;Iris Marion Young, Inclusion and Democracy, Ch. 1; and Jack Knight and James

Dilemmas of public reason: pluralism, polarization, and instability 123

Johnson, "Aggregation and Deliberation: On the Possibility of DemocraticLegitimacy," Political Theory 22 (1994), 277-96.

97. PL, 139.98. Habermas, Between Facts and Norms, 298.99. Claude Lefort, Democracy and Political Theory (Minneapolis: University of Minne-

sota Press, 1988).100. Mouffe, The Paradox of Democracy, 113.101. Benhabib, "Toward a Deliberative Model," 73.102. Bohman, Public Deliberation, 33.103. Mouffe, The Democratic Paradox, 117.104. See Jiirgen Habermas, "On Systematically Distorted Communication," Inquiry

13 (1970), 205—18; and Young, "Activist Challenges." These issues are taken upin my "Deliberativist Responses to Activist Challenges," Philosophy and SocialCriticism, forthcoming.

105. The author would like to thank the Center for Ethics and Public Affairs of theMurphy Institute of Political Economy at Tulane University for its generoussupport during the completion of this work.

7

Public reason and religion

James Boettcher

The long-standing question of the relationship between religion and liberal-ism remains fiercely disputed. A popular book of the 1990s suggests that con-temporary liberal philosophy is one of the social forces in the United States,encouraging the trivialization of religion and a corresponding "culture of dis-belief." Stephen Garter writes that: "Today's political philosophers seepublic dialogue as essentially secular, bounded by requirements of rationalityand reason. It is not easy to fit religion into that universe, which is why somereligiously devout people find themselves at war with the dominant trends incontemporary philosophy." Of course, the dominant trends in contemporarypolitical philosophy have for decades been shaped directly by the work ofJohn Rawls. The publication of Political Liberalism coincided with—and cer-tainly also helped to sustain—a growing interest among both legal theoristsand philosophers in the question of the role of religion in the public sphere.Yet, while Rawls's own political turn over the course of the 1980s seems tohave been inspired mainly by his recognition of the importance of relig-ious and philosophical pluralism, Political Liberalism, with its central idealof public reasoning, has been criticized extensively for misunderstandingthe proper relationship between religion and liberal-democratic citizenship.In the so-called "war" between religion and contemporary philosophy, itwould seem that some of the principal battles are being fought over Rawls'swork and the controversy it has engendered.

It is not the aim of the present chapter to contribute to any battles, as I hopethat the metaphor of a cultural or intellectual "war" turns out to be mis-placed. One of the goals of the chapter is to show instead that Rawlsianpublic reasoning is consistent with much reasonable religious belief and prac-tice. But, as the controversy surrounding the infamous "abortion footnote" toPolitical Liberalism illustrates, Rawls is partially to blame for much of the mis-understanding of and widespread suspicion about the idea of public reason.The claims of that footnote seem to rely on a narrow and exclusionist interpre-tation of public reasoning which Rawls ultimately rejects. They also seem tobe based on the incorrect assumption that an idea of public reason can some-how generate the correct answers to contested political questions, circumvent-ing actual public political debate.

Public reason and religion 125

Nor is understanding the idea of public reason made easier by the fact thatthe idea's exposition in Lecture VI of Political Liberalism does not include themore convincing "wide view" of public reasoning advanced by Rawls in sub-sequent writings. Only in the book's second "Introduction," and with thewisdom of hindsight, does Rawls acknowledge that with its main philosophi-cal question political liberalism reaches out specifically to religious believers.But even Rawls's most direct engagement with religion in "The Idea of PublicReason Revisited" fails to respond adequately to many of the difficult ques-tions posed by critics. Does the idea of public reason demand too much of reli-gious believers? Does it compromise the integrity of their faith commitments?Are the requirements of public reasoning feasible for religious believers? Arethese requirements fair? Would they have adverse consequences for the publicpolitical deliberation of a liberal-democratic society marked by deep and per-vasive religious conviction?

These are just some of the questions that have been raised over the last twodecades in what has become a rather sizable body of literature on liberalism,religion, and the public sphere. In reviewing this literature along with the rele-vant critical discussions of political liberalism, I shall specifically investigatethe relationship between religion and Rawlsian public reasoning, guided bythe aim of reconciling religious believers to public reason's ideal. I begin bypresenting a wide interpretation of Rawls's wide view of public reason whichspeaks directly to the role of religious discourse and argument in politicaldecision-making (the first two sections). A suitable interpretation of the ideapublic reason must explain the ways in which this idea both encourages citi-zens and officials to present and discuss publicly their religious views andinstructs citizens and officials sometimes to restrain their appeal to their reli-gious views in accepting and advancing political justifications. I then turn tothree main varieties of criticism which are motivated by religious concernsand which generally target liberal interpretations of political justificationand citizenship (third section). Jointly these criticisms suggest that restraintson religious discourse and argument are infeasible, unfair, and politically dis-advantageous. However, with respect to Rawls's wide view of public reason,I maintain that these objections miss the mark. Finally, I examine brieflyadditional criticisms of Rawls's view, also motivated by religious concerns,which raise specific questions about the determinacy of public reasoning andthe grounds of public reason's status as a moral duty of citizenship (fourth sec-tion). These criticisms, I suggest, call for further clarification and elaborationof the conceptual structure of political liberalism. Although I do not addressall of the concerns about public reasoning that might be raised by religiouslyminded critics, I hope to show that the idea of public reason is less restrictivethan many critics have supposed, and that the forms of restraint that it doesrequire of citizens and officials are neither infeasible nor unfair.

126 The Legacy of John Rawls

The idea of public reason

Public reason, in the most general sense, refers to the publicly accessible rea-soning and deliberation of democratic citizens and officials concerning justiceand the common good of their society. As a special duty of citizenship, Rawls's"idea of public reason" has a more definite structure, concerning when, how,and to whom it applies. When fundamental political questions are at stake, allpolitical officials and, ideally, all citizens as well are morally obligated to turnto the values of a reasonable political conception of justice as the basis of theirpolitical decision-making. A reasonable political conception of justice appliesonly to the basic structure of society and aims at securing fair terms of coopera-tion for all persons as free and equal citizens. And it should be possible toarticulate and present a reasonable political conception independently of thecomprehensive religious or philosophical doctrines in which it may be morallyembedded. In addition to relying on a reasonable political conception, citi-zens and officials are prepared to justify their decisions to one another inpublic reason by satisfying what Rawls refers to as the liberal principle oflegitimacy based on the criterion of reciprocity. According to this principle,"[o]ur exercise of political power is proper only when we sincerely believethat the reasons we would offer for our political actions—were we to statethem as government officials—are sufficient, and we also reasonably thinkthat other citizens might reasonably accept those reasons."

As an ideal of citizenship, then, the idea of public reason requires that citizensand officials both search for adequate political justifications and sometimesrestrain their appeal to certain doctrinaljustifications for the exercise of coercivepower. These requirements apply both to the political choices of citizens andofficials, and to the public presentation of arguments justifying those choices. Ide-ally all citizens and officials should be at least ready to explain how their politi-cal choices are supported by suitable political justifications in public reason.But the idea of public reason is not intended to govern all questions or delibera-tive settings. It applies neither to political society's "background culture" ofchurches, clubs, associations, professional societies, and the like, nor to theforms of media which stand between this culture and the official "public politi-cal forum'' of legislatures, courts, offices, campaigns, and voting booths. Andwithin the public political forum, the requirements of public reason apply,strictly speaking, only to fundamental political questions, i.e. constitutionalessentials and matters of basic justice. In short, the idea of public reason governsboth the decisions and the justificatory discourse of citizens and officials in theirattempt to work out the fundamental terms of their political associationthrough voting, campaigning, and official political decision-making.

The content of public reason consists primarily of the values associatedwith a political conception of justice, along with guidelines of inquiry and

Public reason and religion 127

judgement. A complete analysis of public reason's content would be beyondthe scope of the present chapter, but a few remarks may be helpful before turn-ing to the relationship between public reason and religion. First, in contrast tothe ideal of political justification worked out by Robert Audi, Rawlsian publicreason is not equivalent to a standard of secular reason that would admitnon-religious comprehensive justifications where religious justifications areinadmissible. Rawls defines secular and religious reasoning as forms of"comprehensive," or "nonpublic," reasoning which should be distinguishedfrom the public reasoning associated with a political conception of justice.Unlike a political conception, a comprehensive doctrine, as an exercise ofboth theoretical and practical reason, applies to the whole of a person's life,and specifies virtues, obligations, and ideals that transcend the domain ofthe political. Second, the content of public reason is not restricted to Rawls'sown conception of justice, "justice as fairness." To be sure, each citizenis expected to endorse a coherent and complete political conception, and toavoid insincere, ad hoc appeals to contradictory claims or competing interpre-tations of what justice demands. But, from the general standpoint of politicalliberalism and the citizenry at large, the content of public reason is not consti-tuted by a single political conception of justice.

A related point is that the idea of public reason is not intended to bypass theprocess of political deliberation and judgement by supplying ready-madeanswers to difficult political questions. A political conception of justiceshould be capable of ordering values in such a way that pressing politicalissues may be addressed. But, insofar as citizens are expected to hold differentconceptions of justice, there is no authoritative canon of reasons and argu-ments that, once they are articulated by the theorist, would render additionalpolitical debate and discussion unnecessary. Likewise, public reason does notpresent &populist conception of political justification that would require a citi-zen to adopt and present only those arguments that would in fact be endorsedby all other citizens. Rather, the idea of public reason instructs each citizen toseek agreement by identifying first-person political justifications that satisfythe criterion of reciprocity and address others as reasonable. But, from thethird-person standpoint of citizens generally, the results of public reasoningare often inconclusive, and a citizen should expect to encounter disagreementwith other sincere and reasonable citizens.

Public reason and religion: the wide view

Even in this very brief exposition of the idea of public reason, I have alreadyreferred to "restraints" on public argument and to the idea that certain claimsand arguments may be classified as "inadmissible." This kind of language may

128 The Legacy of John Rawls

be misleading inasmuch as it suggests that citizens are obligated to keep mostor all comprehensive (i.e. "nonpublic") discourse and argument out of thepublic political forum. Public reason should not be interpreted, however, asan exclusionist principle according to which only public reasons are admissiblein the political forum. Rather, as I see it, the idea of public reason serves as arestraint on discourse and argument only in the following sense: Claims andarguments that are based on nonpublic reasons fail to serve as adequate politi-cal justifications, and, absent supporting public reasons, should not be adoptedor presented as justifications for decisions concerning constitutional essentialsand matters of basic justice. That such arguments are, absent supportingpublic reasons, insufficient as political justifications on fundamental politicalquestions, however, does not mean that they are insignificant, irrelevant, orpublicly inadmissible.

At no point does Rawls endorse a conception of public reasoning that wouldexclude all nonpublic discourse and argument from the process of politicaldecision-making. While Rawls seems to have been drawn initially to an exclu-sionist position, the first edition of Political Liberalism offers by contrast the so-called "inclusive view" of public reason. The inclusive view, influenced inpart by Lawrence Solum, would admit nonpublic reasons, even in the publicpolitical forum, under certain historical and socio-political conditions. Butthis view is subsequently abandoned by Rawls in favor of an even more "per-missive" position, namely, the "wide view" of public reason. The main fea-ture of the wide view is its proviso concerning the introduction of the nonpublicdiscourse and argument of comprehensive doctrines into political decision-making. According to the proviso:

[RJeasonable comprehensive doctrines, religious or nonreligious, may beintroduced in public political discussion at any time, provided that in duecourse proper political reasons—and not reasons given solely by compre-hensive doctrines—are presented that are sufficient to support whateverthe comprehensive doctrines introduced are said to support.

Examining a few questions about the proviso will enable us to understand theways in which a wide interpretation of Rawls's wide view of public reasonmight accommodate religious citizens.

First, the proviso suggests that comprehensive doctrines may inform politi-cal decision-making as long as supporting public reasons accompanythem. A related but logically distinct question is whether citizens mustalways advance public reasons any time they introduce their comprehensivedoctrines in the public political forum. All comprehensive discourse andargument that meets the proviso is publicly admissible. But must all compre-hensive discourse and argument in the public political forum meet theproviso? Should the proviso be interpreted as allowing comprehensive

Public reason and religion 129

discourse and argument in the public political forum only ?f supporting publicreasons are also presented?

I submit that it should not be interpreted in this fashion. The provisorequires the presentation of public reasons insofar as comprehensive reasonsare presented in support of some position or decision about a fundamentalissue. As I see it, the proviso's notion of "support" is best understood to meanthat all adequate political justifications must always include a justification withinthe domain of public reason. If a citizen adopts or presents religious or othernonpublic reasons as considerations that count toward the justification of herposition on a fundamental issue, she is obligated also to identify (and some-times present) a sufficient argument for that position in public reason; other-wise she should exercise restraint by not appealing to her comprehensivedoctrine for justificatory purposes. Of course, in much political argument,and certainly in cases of constitutional argument, the proviso will typicallyadmit nonpublic reasons only if supporting public reasons are provided. How-ever, as Robert Audi has observed, even our discourse in the public politicalforum is not necessarily limited to justificatory or evidential purposes. A citi-zen might introduce a religious doctrine simply in order to express her back-ground perspective or depth of conviction on a particular issue. As long as thisexpression is not intended as a reason or argument that counts toward the jus-tification of a position or decision on a fundamental political question, thestrict terms of the proviso need not apply.

A second set of questions, raised by Rawls, concern the specific details ofsatisfying the proviso. How should comprehensive doctrines be expressed inthe public political forum? When exactly should citizens satisfy the proviso?How quickly must they do so after the introduction of their comprehensivedoctrines? Are all persons equally obligated by the proviso? One possibilityfor honoring the proviso would involve some citizens or officials serving asproxies or spokespersons for others. Is this sort of representation allowed bythe proviso? Without definitively answering these questions, Rawls simplysuggests that the norms governing the proviso should enable us to understandthat it is satisfied in good faith: The details of how to satisfy the proviso "mustbe worked out in practice" and depend on "the nature of the public politicalculture" and the "good sense and understanding" of its citizens.

What does it mean to exercise good sense and understanding in this context?In attempting to satisfy the proviso, citizens are guided by their disposition tobe reasonable and their desire to be recognized by others as reasonable. A rea-sonable citizen seeks fair terms of political cooperation, treats others as freeand equal and as interested in exercising basic moral powers, and acknowl-edges what Rawls calls the burdens of judgement, or the idea that sincereand reliable reasoners are apt to disagree about how to resolve complex anddifficult issues. The aspiration to be recognized as reasonable motivates

130 The Legacy of John Rawls

citizens to reason and act in a manner that reassures others of their good faitheffort to offer public reasons for their political choices.

This attempt on the part of citizens to reassure one another of their reason-ableness leads to a final question. Does political liberalism actually encouragethe introduction of religious and other nonpublic reasoning in the public poli-tical forum? Rawls addresses this question only briefly, suggesting that theremay be "positive reasons" for introducing comprehensive doctrines into poli-tical debate. The common assumption among many critics of political liber-alism, based perhaps on some of Rawls's own remarks, is that the idea of publicreason at best seeks merely to tolerate religious discourse and argument. As I seeit, however, there is a strong case to be made within political liberalism forencouraging the public presentation of certain forms of religious discourse andargument.

This case turns on what we might call the transparency problem. The idea ofpublic reason asks citizens to aim at agreement about fundamental politicalquestions, but not to expect it. Specifically, the criterion of reciprocity requireseach citizen to search for political justifications that others might reasonablyaccept, while the Rawlsian burdens of judgement instruct citizens to expectthat others may very well interpret evidence or fundamental political valuesin different ways and arrive at competing answers to complex questions. Citi-zens are to advance political justifications that they believe to be sound andmost reasonable, and which they believe that others, who are also disposed toseek fair terms of cooperation among free and equal citizens, can accept as atleast reasonable, that is, as at least consistent with political liberalism's under-lying conceptions of the person, society, and judgement. The hope is thatcitizens can recognize and appreciate competing claims and arguments as atleast reasonable, despite their ongoing disagreement about which conceptionofjustice is valid or what precisely justice requires of basic social institutions.But just how are they to do that?

One of the ways in which a citizen can recognize other citizens, along withthe claims and arguments of other citizens, as reasonable is to learn moreabout the conceptions ofjustice and comprehensive doctrines that supportand motivate those claims and arguments. By asking how another citizen'sconception ofjustice leads to a different conclusion on a particular matter oflaw or policy, a citizen is able to achieve a kind of critical distance from herown political judgements and a better understanding of how others mightreasonably disagree with her. She is better able to appreciate how anothercitizen's conception ofjustice offers an interpretation of political liberal-ism's basic commitment to fair terms of cooperation among free and equalcitizens. The political virtues at stake in many cases of disagreement—civicengagement, toleration, self-criticism, and fair-mindedness—are also exer-cised when a citizen sincerely attempts to understand how another citizen's

21

Public reason and religion 131

comprehensive doctrine supports a reasonable political conception of justice.Good citizens should not only remain aware of religious diversity, but alsomake an effort to understand different religious traditions and imagine howvarious questions might be approached from within these traditions.

In "The Idea of Public Reason Revisited," Rawls has identified differentforms of nonpublic discourse that would enable citizens, even in the publicpolitical forum, to examine and discuss the relationship between their politicaljudgements, competing conceptions of justice and religious and philos-ophical comprehensive doctrines. In declaration, citizens might explain theircomprehensive views to one another in order to demonstrate how theseviews support reasonable political conceptions of justice. In conjecture, citizensbegin with claims that they believe others to accept, and attempt to show howthese claims support a particular conclusion on questions of law or policy.In witnessing, citizens introduce their comprehensive views in order to explainthe basis and depth of their dissent, including their religiously motivated dis-sent, from a legitimately enacted decision. Through declaration, conjecture,and witnessing, and perhaps through other forms of comprehensive discourseand argument as well, citizens attempt to learn from one another about thepolitical implications of their religious views. They also attempt to reassureothers, with whom they disagree on particular matters of law or policy, oftheir shared commitment to reasonable political conceptions of justice.

Thus, according to a wide interpretation of Rawls's wide view, religiouscitizens are not just permitted but even encouraged to present their compre-hensive views. This presentation might involve, though it is certainly not lim-ited to, non-justificatory religious conversations, religious justifications thatmeet the proviso, declarations of faith that express the depth of a citizen's poli-tical convictions or the urgency of some matter of basic justice, conjecturesabout how a political question might be understood within another citizen'sparticular religious tradition, and political dissent through witnessing. Theseforms of discourse and argument amount to more than mere tolerance oragreeing to disagree, but to less than a political contest over the correct doc-trinal solution to shared political problems. Citizens can engage one anotherin a spirit of "intellectual solidarity" and "ecumenical political dialogue,"while continuing to satisfy the proviso and recognize public reason's ideal.Or, as Rawls puts it:

Citizens' mutual knowledge of one another's religious and nonreligious doc-trines expressed in the wide view of public political culture recognizes thatthe roots of democratic citizens' allegiance to their political conceptions liesin their respective comprehensive doctrines, both religious and nonreli-gious. In this way citizens' allegiance to the democratic ideal of publicreason is strengthened for the right reasons.

132 The Legacy of John Raw Is

But a religious citizen's allegiance to public reason's ideal would also seem todepend on how successfully political liberalism can address the questions withwhich this chapter began. Critics have identified a number problems encoun-tered at the intersection of religious commitment and public reasoning. Someof these problems might be formulated specifically as difficulties with politicalliberalism, while others might be associated more generally with the notionthat citizens should sometimes exercise restraint in their public religiousdiscourse and argument. I shall begin with familiar objections to the doc-trine of restraint in an attempt to determine whether and to what extent theseobjections are applicable in the case of Rawls's wide view of public reason.

Public reason and religion: three main criticisms

Are the requirements of public reasonfeasible?

Public reason's ideal is realized by persons who have successfully adopted asocial role, namely, the role of democratic citizen. This role draws on thebasic moral powers of persons, as well as their willingness to recognize publicreason's structure and exercise certain intellectual capacities. Good citizensmust be able to evaluate evidence, engage in valid forms of reasoning, applyprinciples of justice, deliberate together, and critically assess their own beliefsand convictions. Through these capacities as well as their dialogue withothers, citizens are able to make responsible judgements about which of theirown claims and arguments are sound and based on adequate public reasons.This form of self-understanding is essential if they are to formulate politicaljustifications that are consistent with the criterion of reciprocity. Howeverthe practice among citizens of categorizing their own reasons and argumentsas "public" and "nonpublic" may also appear to be rather unrealistic.

First, at times it will be quite difficult for a citizen to determine which of hervarious beliefs should count as public reasons. A citizen must attempt to deter-mine the extent to which the force of a claim or argument derives from a poli-tical conception of justice that might be presented independently of hercomprehensive religious doctrine. She should also consider how her reasoningwould be received if she were a judge addressing a court or legislator address-ing her constituents. A citizen must also continually bear in mind the op-posing comprehensive views and the different political conceptions of othercitizens. She should attempt to imagine just how accessible, reasonable, andsignificant her claims would appear if she believed what others reasonablybelieve, both politically and comprehensively. Yet these forms of self-exami-nation, self-criticism, and imagination may turn out to be quite difficult to sus-tain. To set aside one's religious convictions in political deliberation requires,as Kent Greenawalt has observed, "an exceptional discipline," and so "[i]t is

Public reason and religion 133

doubtful whether one should recommend to ordinary people a self-restraintthat is so hard to perform."

This difficulty is compounded by the fact that in other contexts citizens maygrow accustomed to deliberating from within their religious doctrines aboutpolitical questions. The role of citizen is after all one social role among themany that are usually integrated into a complex personal identity. As morethan just citizens, persons in the background culture of civil society willreason in different ways about the issues that affect their lives and commu-nities. Thus on the basis of her religious doctrine a citizen may arrive at amoral judgement, or even a political one, about the issues underlying a consti-tutional essential or matter of basic justice. Yet the same citizen is then askedto set aside those judgements, or at least some of the grounds for them, forthe purpose of reasoning publicly with others who do not share her compre-hensive view. A citizen may conclude that she is being asked to separate herdeliberation from an essential part of her person, an experience that may beboth awkward and alienating. Thus some critics have argued that Rawlsunderestimates the difficulty of asking religious believers to ignore what theytake to be the highest moral authority. As Christopher Eberle observes,because the obligation to obey God is for many theists both overriding andtotalizing, it is not easily relegated to a nonpublic realm. Others criticshave argued that the fact that citizens in a political culture like that of theUnited states do regularly introduce religious arguments as justifications inpolitical decision-making is a reason that they should continue to do so. Cer-tainly it seems implausible to ask citizens to maintain an artificial barrierbetween their deepest convictions and their political commitments.

I believe that these concerns show public reason to be a demanding ideal,but not an impossible one. To engage in inquiry after suspending all or mostof our beliefs would be an arduous philosophical task. But this is not whatpublic reason requires. That we cannot suspend all of our beliefs at once doesnot mean that we cannot bracket some of them or examine certain beliefsin light of others. In determining which of our beliefs should serve as publicreasons we have to start somewhere, namely with the constitutive normsand values of a political conception of justice. Conversations with others,along with the forms of self-questioning, thought-experiment, and imagina-tion already discussed, enable us to sharpen and revise these determinationsin specific cases. They also help us to appreciate which considerations aremost relevant for the issue at hand. We then attempt to draw conclusionsbased on our public reasoning, comparing these conclusions to our priorintuitions and judgements or to the conclusions we reach from within ourcomprehensive doctrines.

But even if reasoning of this kind is possible, would it contribute to anexperience of a compartmentalized existence, alienating citizens from the

134 The Legacy of John Rawls

political process, or from themselves? Perhaps in some cases it would, althoughanswers to this question will vary in light of the historical and social conditionsof particular public cultures. In general, however, there are several reasonswhy public reason need not be understood as alienating. First, the experienceof having obligations associated with a recognized role or status is a familiarone for most people. In addition to our obligations as citizens, we have obliga-tions to others as friends, family members, neighbors, co-workers, or simply asfellow human beings. While not all of these relationships require us to adhereto a specific standard of reasoning, we do assign weight to certain kinds ofconsiderations when we ask ourselves questions such as "What would a goodfriend do in this situation?" Moreover the very nature of a relationship some-times plays an essential part in establishing what we owe to others. Andsometimes the sources of our obligations are manifold. Spousal obligations, forexample, may derive from a shared history, personal promises, a publiclyrecognized marriage, and a relationship to God. The moral life providesnumerous examples of obligations the content and source of which are deter-mined in part by roles and particular relationships.

It is also important to observe that the idea of public reason does not requirecitizens to maintain an "airtight barrier" between their beliefs or betweendomains of discourse. As I have argued, the restraints on reasoning that areimposed by this idea apply only to claims and arguments that contribute topolitical justifications concerning fundamental political questions in thepublic political forum. And even acknowledging these restraints, citizensmay still endorse alternative (nonpublic) arguments for their positions onthese questions. The wide view of public reason, as I have interpreted it, notonly allows but also encourages citizens to present religious claims and argu-ments publicly alongside their political justifications. Each citizen, moreover,is expected to remain aware of religious diversity and to acknowledge that thepolitical conceptions of many fellow citizens are morally embedded in thosecitizens' comprehensive religious and philosophical views.

Finally, the experience of a separation in judgement, unavoidable in somecases, need not always be an alienating experience. After all, in thinking abouta fundamental political question, a citizen may discover that public reasonconfirms the judgements of nonpublic reason. In this case, reasoning "sepa-rately" about an issue, reasoning in terms other than those of a comprehensivedoctrine, shows that doctrine, or at least some of its claims or conclusions, to besupported by independent grounds. A citizen's allegiance to her religious doc-trine may be strengthened by this kind of support. Of course there are boundto be occasions when public reason seems to contravene the conclusions of areligious view. Such cases should motivate a citizen to consider carefullywhether her religious reasoning, as well as her public reasoning, is sound.Often it is not obvious exactly what religious doctrines recommend politically.

Public reason and religion 135

In most cases, starting from the authority, teachings, or basic texts of a reli-gious doctrine, a number of inferential steps must be taken before specific deci-sions about matters of law and public policy are reached. Misinterpretationand mistakes in judgement are as much a possibility along this path as theyare in the domain of public reason.

Are the requirements of public reason fair?

It is not only in arguments about feasibility that one encounters the chargethat the idea of public reason compartmentalizes various dimensions of theperson. This charge often serves as a premise in a related argument againstpublic reason, namely, an argument about fairness. According to this argu-ment, even if it is possible for citizens sometimes to separate their judgementsso that on the appropriate occasions they rely on sufficient public reasons, it isunfair to require such separation from religious citizens whose public and non-public lives are otherwise guided and unified by religious ideals. Charges thatthe idea of public reason is unfair may be understood in terms of two basic con-cerns. A first concern is that politically the idea of public reason would be dis-proportionately burdensome for religious citizens. Second, critics sometimessuggest that the idea of public reason would interfere significantly with theintegrity of religious belief and practice. I shall address both of these concerns,beginning with the first.

Critics of public reason worry that it requires religious citizens to exercisemeasures of restraint and self-examination that are not required, or perhapsnot required in the same way, of nonreligious citizens. Nicholas Wolterstorffobserves that religious reasons referencing God or sacred religious texts aremore obviously comprehensive and so "nonpublic" than reasons derivedfrom a secular comprehensive doctrine. An argument based on the Torah,for example, is more likely to be interpreted by citizens at large as nonpublicthan one based on utilitarianism or nationalism. Thus the idea of publicreason would seem to be unfair because it politically disadvantages religiousbelievers vis-a-vis citizens who hold nonreligious comprehensive doctrines.A related argument is advanced by Patrick Neal, who suggests that personswho hold liberal comprehensive doctrines stand in a privileged positionwithin political liberalism. Comprehensive liberals can for the most partassume that their comprehensive beliefs are congruent with the conclusionsof political liberalism, while citizens of faith understandably would want tocheck these conclusions against their background convictions. Imaginingthat their positions were reversed, Neal suggests that a comprehensive liberalwould want to carry out the same analysis of her background convictions if shewere asked to accept principles of justice that were developed within a reli-gious tradition.

136 The Legacy of John Rawls

Although much depends on the public culture in question, let us supposethat religiously based reasoning is generally easier to recognize than someother forms of nonpublic reason. It is not clear how this fact translates into acase of unfairness. The idea of public reason does not single out religion as atarget of restriction or restraint. Rawlsian political liberalism does not essen-tially turn on the religious-secular distinction, nor does it demand specialforms of self-examination or self-criticism among only religious believers.Conscientious citizens understand the unreasonableness of simply invoking asecular comprehensive point of view when constitutional essentials or mattersof basic justice are at stake. They also reject as unreasonable such politicaladvocacy when it is motivated by hostility toward religion or by an effort topromote secularization, atheism, or a nonreligious doctrine such as compre-hensive liberalism. Moreover, political liberalism certainly permits citizensto affirm, as the content of their public reasoning, a reasonable political con-ception that may be associated with a religious view, provided that such a con-ception can also be presented as freestanding to others.

The idea of public reason is fair in the sense that it applies in the same way toall persons qua citizens. It attributes to all citizens the same justificatoryresponsibilities and the same requirements of restraint in reasoning. To besure, some religious believers, based on their theological orientation, aremore likely to accept the general idea that the moral-political judgementsassociated with their religious convictions also admit of reasoned justification.Thus the idea of public reason might be more politically disadvantageous forcertain doctrines in practice, insofar as it would demand adherents of thosedoctrines, in comparison to other citizens, to exercise more restraint on someissues. Leaving aside an analysis of which citizens and doctrines, religious ornonreligious, might encounter potential political disadvantages, it is still farfrom obvious that these disadvantages represent a form of unfairness. A firstpoint is that citizens who hope to deliberate together about common purposesmight find that appealing solely to sectarian religious justifications is politi-cally ineffective. Absent an attempt to satisfy the proviso or to translate reli-gious claims into a public-political idiom, such justifications will likely fail toconvince many fellow citizens. Second, norms of political cooperation oughtto protect citizens' interests in affirming and pursuing reasonable comprehen-sive doctrines, but these norms cannot be fashioned to secure the vitality orcontinuity of particular religious doctrines. As Rawls observes, drawing onIsaiah Berlin, "[a] just liberal society may have far more space than othersocial worlds but it can never be without loss." One way to substantiate thecharge of unfairness would be to show that public reason is so disproportio-nately restrictive that either only a small set of doctrines or only nonreligiousdoctrines could flourish under its ideal. But this outcome is, I think, ratherunlikely, especially considering the extensive religious freedoms guaranteed

Public reason and religion 137

by all reasonable political conceptions of justice that would comprise the con-tent of public reason.

It is also worth bearing in mind that in some public cultures, such as that ofthe United States, the great majority of the citizens and officials who wouldactually apply the idea of public reason also identify as religious and remaindeeply sympathetic to religiously inspired conceptions of justice. Yet thesesame citizens might still disagree about many religious and theological issues,as well as about how religious convictions and practices should inform politi-cal decision-making. In the United States, for example, it is simply not thecase that citizens who support a strong religious voice in politics agree alsoabout a particular political vision or agenda that would be disadvantaged bypublic reason. And an increasingly religiously diverse population wouldonly seem to heighten the possibilities for disagreement about religiouslyinformed political issues or about the appropriate role of religion in publiclife. In this context, the idea of public reason would respond as much to dis-agreements among religious believers and traditions as it would to disagree-ments between religious and nonreligious citizens.

What about the second charge of unfairness, i.e. that public reason wouldinterfere significantly with the integrity of religious beliefs and practices?A religiously grounded wholeness or integrity is an ideal, perhaps the highestideal, in the lives of many religious believers. A good life is one in whichreligious faith, for example, inspires and informs all of one's actions, com-mitments, and decisions, including one's political decisions about whetherto support a particular candidate, campaign, policy, law, or judicial deci-sion. For some citizens it follows from a religious ideal that their politicaldeliberation and decision-making should not be separated from their deep-est religious convictions. To the extent that public reason at times requiresthem not to base their judgements on those convictions, it represents aninfringement on the exercise of their religion. Hence Wolterstorff con-cludes that the idea of public reason is an unfair violation of the freedom ofreligious citizens:

[T]he liberal assumes that requiring religious people to debate and act poli-tically for reasons other than religious reasons is not in violation of their reli-gious convictions . . . He assumes, in other words, that though religiouspeople may not be in the habit of dividing their lives into a religious compo-nent and a non-religious component, and though some might not be happydoing so, nonetheless, their doing so would in no case be in violation of theirreligion. But he is wrong about this. It is when we bring into the picturepeople for whom it is a matter of religious conviction that they ought tostrive for a religiously integrated existence—then especially, though notonly then, does the unfairness of the liberal position come to light.

138 The Legacy of John Rawls

There are a number of issues that must be addressed by way of a response tothis criticism. An initial question is whether the idea of public reason would infact interfere with an ideal of religious integrity. This is undoubtedly a possi-bility. I have already argued that public reasoning could have the effect ofstrengthening a person's commitment to a religious doctrine, by providingsupport to that doctrine's claims and conclusions. Moreover, in the wideview, citizens may offer both public and religious reasons in the public politi-cal forum, and they are encouraged to engage in religious dialogue that helpsto illuminate the nature of reasonable disagreement. Some citizens may insist,however, that for them living a religious life involves drawing directly from areligious doctrine in their political deliberations, without having to considerthe requirements of public reason when fundamental questions are at stake.For these citizens, the idea of public reason, even as I interpret it, will interferewith the goal of integrity. Wolterstorff is right to recognize this interference asa genuine moral-political problem.

But in precisely what sense is the idea of public reason unfair to religiousbelievers who are also committed to the ideal of integrity? We might beginaddressing this question by setting aside the least convincing answers to it.A first point is obvious: The idea of public reason does not deny a religiouscitizen's political right to speak and act as she sees fit. The principles of a rea-sonable political conception of justice legally protect, though they do notencourage, a citizen's refusal to recognize the idea of public reason in herpolitical speech and action. Second, public reason is not unfair in the senseof placing a burden—in this case, an obstacle to integrity—uniquely onreligious citizens. The tension between public reason and integrity high-lighted by Wolterstorff could arise in the pursuit of a nonreligious compre-hensive doctrine as well as a religious one. A deep ecologist, for instance, mightbelieve that the requirements of public reason fail to place sufficient weight onthe value of nonhuman nature. Even a comprehensive liberal might findthat the requirements of public reason interfere with her attempt to organ-ize her life around values of creativity and experimentation that not allcitizens share.

In claiming that public reason infringes on the "free exercise of theirreligion," Wolterstorff ostensibly means that it prevents people from fullypracticing their religion, or living in accordance with its ideals and direc-tives. That is, an obligation to reason publicly appears to drive a wedgebetween a person's religious convictions and the political choices that wouldotherwise be attached to those convictions as religious ideals and obligations.In order to get a clear view of this problem, we should examine the type ofconvictions that might conflict directly with the idea of public reason.Assume that what we may call, from a political standpoint, first-order reli-gious convictions consist of religiously based teachings about God, sacred

Public reason and religion 139

texts, creation, the human person, salvation, moral choices, the aims oflife, and so on. If it counts among a person's second-order religious convic-tions that she is obligated always to base her political advocacy and choiceprimarily on these first-order religious convictions, and if the obligation toreason publicly is incompatible with the obligation always to base politi-cal advocacy and choice primarily on such first-order convictions, then theidea of public reason would seem to require the rejection of at least one ofa person's (second-order) religious convictions. Thus, one might conclude,public reason would infringe unfairly upon the freedom to live in accordancewith a religious doctrine.

Several considerations, however, militate against such a strong conclusion.First, we should remember that, like the problem of integrity, the issue ofinfringement is not restricted to religious believers. Any citizen, religious ornonreligious, could claim that it counts as a belief within her comprehensivedoctrine that fundamental political choices should be made primarily onthe basis of that doctrine even in the absence of supporting public reasons.Second, the unfairness of this supposed infringement depends upon howwidely one interprets the freedom of free exercise. Interpreted in absoluteterms, according to which no religious obligation could be overridden ormorally criticized, such freedom would be incompatible with the fact of rea-sonable pluralism. Religious or nonreligious majorities could use an absolutistunderstanding of integrity and free exercise to justify using state power inorder to promote a particular comprehensive doctrine, including, for exam-ple, atheism. Assuming that such a conception of religious freedom is unten-able, it follows that not all restrictions on religiously motivated actions andobligations amount to unfair infringements on free exercise.

The question is whether the particular restrictions associated with the Rawl-sian conception of public reasoning are consistent with free exercise, and Isubmit that they are. These restrictions are unlikely to interfere significantlywith most religious belief and practice. They focus less on the content of a reli-gious doctrine than on the political relationship characteristic of a pluralisticcitizenry that includes a diverse body of religious believers. The content of areligious doctrine is affected directly by the idea of public reason only insofaras that doctrine issues second-order claims about whether and how to draw onfirst-order religious convictions in setting the fundamental terms of politicalassociation. Even where this conflict exists, it affects only one aspect of a per-son's comprehensive doctrine by presenting that person with opposing obliga-tions. As I discuss below, this potential conflict of obligations may indeed turnout to present a problem for political liberalism's account of public reason'sstatus as a moral duty of citizenship. But it is not a problem that essentiallyturns on the unfairness of public reason or on the violation of the free exerciseof religion.

140 The Legacy of John Rawls

Would the practice of public reason have adverse social and political consequences?

If the arguments presented thus far are correct, the idea of public reason, as anobligation of liberal-democratic citizenship, is neither infeasible nor unfair toreligious believers. But religiously minded citizens might encounter a thirdvariety of criticism, based on consequentialist arguments about the implica-tions of realizing public reason's ideal. According to this variety of criticism,the actual practice of public reasoning would significantly impede or impover-ish public political deliberation. In the literature on the role of religion in thepublic square, different concerns of this sort have been linked to proposals forrestraints on religious discourse and argument in political decision-making.These restraints are said to privatize what should remain public. Specifically,they are said to weaken or silence critical religious voices that protest injustice,and to deprive citizens of the many social and political benefits that might comefrom robust religious-political dialogue. Critics often turn to Rawls's ownexamples of Abolitionism or the civil rights movement as examples of strugglesthat were sustained by religious conviction and made effective use of publicreligious argument. Surely we would not want to condemn the actors in thesemovements for running afoul of the guidelines of public reason.

These concerns lead to understandable objections—indeed perhaps deci-sive objections—to the restraints associated with an exclusionist interpreta-tion public reason. But they are not objections that are easily applied toRawls's wide view of public reason. Political liberalism does not reduce reli-gious discourse to private reflection, as if religious commitment were a merepreference or "hobby." Far from being a private affair, religious discourseconstitutes part of what Rawls calls the "social reason" of the background cul-ture of civil society where there are no proposed restrictions on the presenta-tion of religiously based arguments. Religiously based arguments, includingarguments supporting campaigns against injustice and oppression, representan important contribution to the background culture that shapes the politicalattitudes and commitments of citizens. And, under the terms of the proviso,these arguments are also admissible in the public political forum. Moreover,a wide view of public reason, properly interpreted, encourages citizens toengage one another and to aim at a richer understanding of how differentcomprehensive views can nourish reasonable political commitments and con-ceptions ofjustice.

Of course, in a public culture with a high degree of religious identificationand church involvement, citizens are more likely to have religiously basedpolitical convictions. One problem, discussed already, is that citizens withsuch convictions would be incapable of dividing their judgements into"public" and "nonpublic" categories. But we can now examine a related prob-lem that might be associated with the fact that many citizens do regularly

rely on religious convictions as the basis of political choice: Because citizens doin fact rely on religious convictions in this way, their convictions should betested, as Michael Perry puts it, in "the to-and-fro of public political argu-ment." In one of Perry's examples, a citizen might be opposed on biblicalgrounds to the legal recognition of same-sex unions. Irrespective of that citi-zen's attempt to exercise the virtues of citizenship by attempting to locate sup-porting public reasons for her position, a thoroughgoing examination of thebiblical grounds for her opposition to same-sex unions might reveal thosegrounds to be inadequate. Perhaps only the public testing of her biblicallybased commitments would encourage her to change her position on the issue.In the case of a bad argument that nevertheless informs the political choices ofcitizens, why should we subject unsound public reasoning to more scrutinythan unsound religious reasoning?

As I see it, however, the wide view of public reason provides ample room forthe critical testing of religious convictions. Challenging another citizen's reli-giously based argument for a position is not always equivalent to offering ajustification for the opposite position, a justification which would require sup-porting public reasons. But, even in cases where a citizen's challenge toanother citizen's religiously based argument also serves as a justification forthe first citizen's position on a fundamental political issue, that citizen needonly recognize and adhere to the terms of the proviso. A wide view of publicreason would, for example, enable a citizen to reject another citizen's bibli-cally based arguments for opposition to same-sex unions, and/or present bothreligiously based counter-arguments and public reasons favoring the legalrecognition of same-sex unions. Moreover, discussion of religious argumentsfor and against the recognition of same-sex unions is the kind of discussionthat is likely to occur for the most part outside of the public political forum.Much critical testing will be carried out in the background culture where therestrictions of public reason do not apply.

But this reply leads to an additional problem. It is possible that the require-ments of public reason, regardless of where they are supposed to apply, wouldhave a significant effect on religious discourse and argument in the back-ground culture. Paul Weithman argues that the norms which govern officialpolitical decision-making "tend, as a sociological matter, to be taken as normsfor other discourse as well." Churches and religious believers might end upengaging in "self-censorship" by limiting their conversations to the terms ofpublic reason, even when they are not required to do so. This practice of self-censorship could in turn give rise to many of the negative social and politicalimplications that seemed at first not likely to result from public reasoning.That is, such self-censorship might weaken or silence religious criticisms ofinjustice, prevent enriching encounters with diverse religious believers, andstand in the way of sufficient testing of religiously based political arguments.

141Public reason and religion

48

Would citizens be capable of sustaining, in their official and unofficial delib-erative practices, the distinction between the norms governing the back-ground culture and norms governing the public political forum? Would ablurring of this distinction lead to a more dialogically impoverished publicsquare? I shall not pursue these questions further, though it seems that theanswers to them ultimately depend upon empirical evidence about the practiceof public reason which is simply not available. However, these questions dopoint to potential dangers in the practice of public reasoning, and they suggestthat the citizens who would realize public reason's ideal must meet a relativelyhigh standard of political virtue. These citizens must have both the capacityand the willingness to make the distinctions in judgement required by the ideaof public reason. Weithman's concerns are connected to what he understandsto be a central problem with the "standard approach" to the question of reli-gion and public reasoning. The standard approach, exemplified by the Raw-Isian approach, begins by positing an ideal to which citizens must aspire,without sufficiently examining the conditions that make good citizenship pos-sible. An alternative approach, endorsed by Weithman among others, is tobegin instead with sociological and political facts about a particular publicculture, and, in light of the empirical evidence, to present a conception ofliberal-democratic citizenship that more effectively preserves the valuablepublic-political contributions made by churches and religious believers.

Public reason and religion: problems with political liberalism

The criticisms examined so far, while often leveled at Rawlsian public reason,might be addressed to different liberal conceptions of political justifica-tion and different proposals for restraint in religious discourse and argument.I have only attempted to show that these criticisms are, for the most part,either inapplicable or unwarranted in the case of Rawls's wide view of publicreason. But there are additional concerns about religion and public reasonthat might be addressed more directly to the Rawlsian view, raising ques-tions about the conceptual machinery of political liberalism. Before conclud-ing, I shall briefly examine two such concerns.

A first concern is that political liberalism does not contain an adequateaccount of why the idea of public reason is a moral duty of citizenship that pre-sents citizens with binding moral obligations that would override, in the caseof a conflict, competing religious ideals and obligations. According to politicalliberalism, a religious citizen who rejects the requirements of public reasonacts unreasonably. That is, such a citizen fails to respect others as free andequal, capable of exercising their basic moral powers and entitled to fairterms of cooperation. But, with a number of critics, we might ask: Why does a

The Legacy of John Rawls142

50

respect for persons require more than a good faith effort to seek a sound justi-fication and provide an honest explanation of the grounds for one's politicalchoices? Why do choices made solely on the basis of a religious doctrine repre-sent a failure to treat others as free and equal? Why should religious citizenswho accept the norm of respect for persons also accept the obligation some-times to exercise restraint? There seems to be an explanatory "gap" betweenpolitical liberalism's underlying conceptions of the person and society and itsmore specific requirements of public reason. To close this gap, we wouldneed an argument that demonstrates why the requirement sometimes to exer-cise restraint, as a moral obligation of citizenship, is entailed by political liber-alism's conception of the person and society. While there may be a convincingargument that satisfies this desideratum, readers are unlikely to find it expli-citly articulated in the pages of Political Liberalism. Much will depend on justwhat it means to treat others as free and equal, that is, as free to affirm reason-able conceptions of the good and equally entitled to reasons justifying thearrangement of shared political institutions.

A second concern is that, absent an appeal to religious or other comprehen-sive beliefs, public reason does not lead to determinate conclusions in hardcases. Numerous critics have presented the charge of indeterminacy againstRawls's idea of public reason. The publicly accessible values of a politicalconception ofjustice are said to be too limited and abstract to sustain plausiblearguments, especially when the disagreement among citizens reaches beyondthe domain of the political to include scientific or philosophical matters.Or, sometimes the same political values, interpreted or weighed differently,might be cited in support of opposite conclusions on a single issue. But we mustbe clear about just what is problematic in cases of indeterminacy in politicaljudgement. The problem is not that most arguments in public reason wouldbe radically indeterminate because they could be based only on abstract politi-cal values. On the contrary, public reason's content also consists of specificprinciples ofjustice, guidelines of inquiry, scientific judgements, and many ofthe epistemic resources of philosophy, common sense, and reasoned reflectionon personal and shared experiences. Nor is the problem that citizens mightregularly fail to locate political justifications that all other reasonable citizenswould in fact accept. Citizens in public reason aim at agreement but do notexpect it; indeed, disagreement between citizens is assumed to be the "normalcase." In the case of a "stand-off," Rawls argues, citizens should "vote for theordering of political values they sincerely think most reasonable."

A more likely problem is that for some fundamental political issues —what we might call special cases of indeterminacy—judgements that rely on see-mingly comprehensive philosophical or religious claims may turn out to beunavoidable. Some political questions involve disagreements over morethan how to interpret or weigh basic political values. With respect to the

143Public reason and religion

55

56

52

abortion issue, for example, the judgement of who counts as a person wouldseem to be both required for any plausible political justification and informeddirectly by various philosophical commitments. Citizens who attempt tofollow political liberalism's instructions for the case of a stand-off may con-clude that no ordering of political values is by itself sufficient to sustain a deter-minate political judgement unless the question of personhood is firstresolved. It is not clear why a sincere and reasonable citizen violates publicreason's ideal by relying on philosophical or religious convictions in such spe-cial cases of indeterminacy.

The problem of indeterminacy also raises questions about how to under-stand political liberalism. On the one hand, according to Rawls, a reasonablepolitical conception of justice should be complete, so that it provides answersto most fundamental political conceptions. Rawls argues that the values ofpublic reason "are not puppets manipulated from behind the scenes by com-prehensive doctrines." On the other hand, political liberalism should beready to admit the political justifications of reasonable religious citizens,some of whom endorse reasonable political conceptions that originate in,but could be presented independently of, their comprehensive doctrines.Indeed, many citizens, especially religious citizens, will not arrive at a protanto justified political conception by adopting the perspective of the originalposition. Rather they will formulate political justifications by translatingsome of the teachings of their comprehensive doctrine into the politicalvalues of public reason, seeking sound arguments which satisfy the criterionof reciprocity and which others might accept as at least reasonable. To besure, Rawls is right to insist that sincere and reasonable religious citizensshould not appeal to political conceptions on an ad hoc basis, manipulatingtheir public reasoning so as to decide issues on the basis of doctrinal grounds.Yet, in the effort to work out reasonable political conceptions from their reli-gious traditions, these citizens might end up adopting political conceptionsthat either present lacunae in reasoning or include seemingly "comprehen-sive" beliefs in special cases of indeterminacy. Tolerating this incompletenessmay turn out to be the price that political liberalism must pay for the impor-tant goal of making room for multiple reasonable approaches to the questionof what justice demands.

The problem of indeterminacy and the problem of the moral grounding ofpublic reason's requirements call for further discussion of the very structureof political liberalism. And there are sure to be additional questions aboutpublic reason and religion. But, one of the main goals of this chapter has been

Conclusion

The Legacy of John Rawls144

58

59

60

to offer an interpretation of Rawls's idea of public reason that shows it to befeasible, fair, and acceptable, all things considered, in a pluralistic societymarked by widespread religious conviction. The idea of public reason doesrequire restraint in the appeal to a religious doctrine, but only when that doc-trine is used, absent supporting public reasons, in the public political forum asa justification for the exercise of coercive power with respect to constitutionalessentials and matters of basic justice. This means that the restrictions imposedby the idea of public reason do not apply to much religious discourse and argu-ment. Indeed, in response to what I have called the transparency problem,political liberalism might even be understood to encourage certain forms ofreligious discourse and argument in the public political forum. Successfulpublic reasoning depends on citizens who can recognize opposing claims, con-ceptions of justice, and comprehensive philosophical and religious doctrines asreasonable. And this recognition of reasonableness depends on citizens' will-ingness to engage one another in attempting to understand the sources andnature of ongoing disagreements. The idea of public reason is thus consistentwith enriching and fruitful public religious dialogue among citizens who arecommitted to seeking political justifications that satisfy the liberal principleof legitimacy based on the criterion of reciprocity. Religious believers andother citizens who participate in such dialogue while at the same timeacknowledging the requirements of public reason are building the bonds oftrust and civic friendship.

Public reason and religion 145

Notes

1. Stephen L. Carter, The Culture of Disbelief (New York: Basic Books, 1993). For anoverview of contemporary debates about the relationship between religion andliberalism, see Paul Weithman's introductory essay, "Religion and the Liberal-ism of Reasoned Respect," in his (ed.) Religion and Contemporary Liberalism (NotreDame: University of Notre Dame Press, 1997), 1-37.

2. Carter, Culture of Disbelief, 42.3. For a treatment of this topic within the legal academy, see especially the numer-

ous contributions to the following symposia: "The Role of Religion in PublicDebate in a Liberal Society," San Diego Law Review 30 (1993), 643-916; "TheReligious Voice in the Public Square," Loyola of Los Angeles Law Review 29(1996), 1401-542; and "Religiously Based Morality: Its Proper Place in Ameri-can Law and Public Policy?" Wake Forest Law Review 36 (2001), 217-570.

4. An exception is the defense of political liberalism presented by Daniel A. Dom-browski in his Rawls and Religion (Albany: SUNY Press, 2001).

5. PL, 243, n.32. Rawls briefly discusses the "troubled question of abortion" in orderto show how comprehensive religious or philosophical doctrines support, or fail tosupport, a reasonable balance of political values. Rawls cites familiar politicalvalues, but offers no argument, in concluding that, by any reasonable measure,

the balance of these values would support a woman's right to abortion, at leastduring the first trimester of pregnancy. On the question of abortion, the equalityof women is said to be an "overriding" value, and any comprehensive doctrinethat would lead to a denial of abortion rights is "to that extent unreasonable."(PL, 243, n.32) Numerous commentators—indeed most of Rawls's critics andmany of his sympathizers—have turned to this footnote in order to support acritical challenge of one sort or another. (See, for example, John Finnis, "Abor-tion, Natural Law and Public Reason," in Robert P. George and ChristopherWolfe (eds), Natural Law and Public Reason (Washington, DC: Georgetown Uni-versity Press, 2000), 75-105.)

A retraction of sorts is offered in "The Idea of Public Reason Revisited," whereRawls suggests that the original footnote was intended only to "illustrate" theway in which comprehensive doctrines can come into conflict with publicreason, assuming that public reasons are available on only one side of a givenquestion. On the abortion question, however, Rawls admits that there are reason-able arguments that both satisfy the requirements of public reason and deny aright to abortion. (See PKR reprinted in CP, 605-6, n80 and n82.)

6. In the text's second "Introduction," Rawls formulates the main question of Poli-tical Liberalism as follows: "How is it possible for those affirming a religious doc-trine that is based on religious authority, for example, the Church or the Bible,also to hold a reasonable political conception that supports a just democraticregime?" (PL, xxxix.)

7. CP, 578.8. I thank Michael Perry who, after reading an earlier version of this chapter, sug-

gested that I make this distinction clear. (See also Perry, "Why Political Relianceon Religiously Grounded Morality is Not Illegitimate in a Liberal Democracy,"Wake Forest Law Review 36 (2001), 217-50 (228).) As Charles Larmore hasobserved, Rawls does not always clearly distinguish between political discus-sion and political decision-making. (See Charles Larmore, "Public Reason," inSamuel Freeman (ed.), The Cambridge Companion to Rawls (New York: CambridgeUniversity Press, 2003), 368-93, especially 382-3.)

9. PL, 217. Rawls does not specify the extent to which ordinary citizens, who aresupposed to be ready to explain their political choices in public reason, must actu-ally present their public reasons for these choices. In a representative democracy,it would seem to be especially important that legislators, judges, and other politi-cal officials actually present and publicly discuss their public reasoning. Indeed,Rawls suggests that the idea of public reason applies directly to political officials,while public reason's ideal applies also to citizens, who are to think of themselvesas iflhey were officials. (CP, 577) For a discussion of how requirements of publicreasoning might apply differently in the cases of citizens and various governmen-tal officials, see PI, 215-16 and CP, 575-8. See also Kent Greenawalt, Private Con-sciences and Public Reasons (New York: Oxford University Press, 1995).

These distinctions are important, but for the sake of convenience I shallrefer hereafter to both citizens and officials simply as "citizens," using the terms"idea" and "ideal" interchangeably with respect to public reasoning.

146 The Legacy of John Rawls

10. CP, 576.11. In a number of writings, Robert Audi has advanced principles of secular reason-

ing and motivation for the citizens of a liberal democracy. (See especially RobertAudi, Religious Commitment and Secular Reason (New York: Cambridge UniversityPress, 2000).)

12. CP, 583.13. Ibid., 581.14. PL, 147-54.15. In a number of essays, Solum develops a similar interpretation of public reason.

(See Lawrence Solum, "Constructing an Ideal of Public Reason," San Diego LawReview 30 (1993), 729-62; Solum, "Situating Political Liberalism," Chicago-KentLaw Review 69 (1994), 549-88; Solum, "Inclusive Public Reason," Pacific Philoso-phical Quarterly 75 (1994), 217-31; and Solum, "Novel Public Reasons," Loyola ofLos Angeles Law Review 29 (1996), 1459—85.) Solum claims to have introduced thedistinction between the exclusive view and the inclusive view in correspondencewith Rawls in 1990. (See Solum, "Situating Political Liberalism," 562, n.68.)

16. See PL, lii and PRR, 591—4. The wide view is Rawls's considered view, andin what follows I shall leave aside the details of Political Liberalism's original"inclusive view" of public reason. For criticism of the inclusive view, see PaulWeithman, "Taking Rites Seriously," Pacific Philosophical Quarterly 75 (1994),272—94. For a defense of the inclusive view, as preferable to the wide view, seeCharles Larmore, "Public Reason," 386—7.

17. CP, 591; cf, PL, li-lii.18. Audi, Religious Commitment and Secular Reason, 75—8.

19. CP, 592.20. Ibid., 591,593.21. For full discussion of this interpretation of reasonableness as a standard that

applies to both citizens and their claims and arguments in public reason, see my"What is Reasonableness?" Philosophy and Social Criticism 30.5-6 (2004), 597-621.

22. CP, 594.23. "Intellectual solidarity" is David Hollenbach's term. See David Hollenbach,

S.J., The Common Good and Christian Ethics (New York: Cambridge UniversityPress, 2002). For Hollenbach's understanding of the ways in which his approachis "generally compatible" with certain elements of the Rawlsian conception ofpublic reasoning, see especially The Common Good and Christian Ethics, 165—70.

"Ecumenical political dialogue" is Michael Perry's term from his, Love andPower: The Role of Religion and Morality in American Politics (New York: Oxford Uni-versity Press, 1991). It requires citizens to adopt attitudes offallibilism and plur-alism, and, in presenting politically relevant religious arguments, to honorstandards of public accessibility and public intelligibility. I should note thatPerry's view has changed over the years, and that my wide interpretation ofRawls's wide view of public reason has more in common with Perry's original dis-cussion of ecumenical political dialogue and ecumenical political tolerance, fromLove and Power, than with Perry's subsequent accounts of the role of religion inpolitical decision-making, discussed below.

Public reason and religion 147

24. CP, 592.25. Here I am indebted to David Rasmussen for many valuable discussions.26. Kent Greenawalt, Private Consciences and Public Reasons, 138. Greenawalt's own

account of public reason also recommends certain restraints on reasoning,though these restraints are intended to be more flexible and context dependentthan the ones proposed by Rawls. See also Greenawalt, "Religion and AmericanPolitical Judgments," Wake Forest Law Review 36 (2001), 557-70.

27. See, for example, Patrick Neal, "Political Liberalism, Public Reason, and the Citi-zen of Faith," in George and Wolfe (eds), Natural Law and Public Reason, 171—98.

28. ChristopherJ. Eberle, Religious Convictionin Liberal Politics (New York: CambridgeUniversity Press, 2002), 145.

29. Michael Perry, Under God? (New York: Cambridge University Press, 2003),39—41. See also Carter, The Culture of Disbelief, 55—6 and David Hollenbach,S.J., "Contexts of the Political Role of Religion," San Diego Law Review (1993),

877-902 (889).30. Robert Audi discusses this possibility in his Religious Commitment and Secular Reason,

116-41.31. Nicholas Wolterstorff, "The Role of Religion in Decision and Discussion of Poli-

tical Issues," in Robert Audi and Nicholas Wolterstorff (eds), Religion in the PublicSquare (New York: Rowman and Littlefield, 1997), 67-120 (105). See also Wol-terstorff, "Why We Should Reject What Liberalism Tells Us about Speaking andActing in Public for Religious Reasons," inWeithman (ed.), Religion and Contem-porary Liberalism, 162—81.

32. Patrick Neal, "Political Liberalism," 171-98.33. Here I adapt Michael Perry's defense of his own standard of "public accessibility"

in Love and Power, 107. For the charge that Perry's standard of public accessi-bility is disadvantageous for theologically conservative and traditionalist citizens,see David Smolin, "Regulating Religious and Cultural Conflict in PostmodernAmerica: A Response to Professor Perry," Iowa Law Review 76 (1991), 1067—1103 and Sanford Levinson, "Religious Language and the Public Square," Har-vard Law Review 105 (1992), 2061-79. See also Daniel O. Conkle, "Different Reli-gions, Different Politics: Evaluating the Role of Competing Religious Traditionsin American Politics and Law," Journal of Law and Religion 10 (1993—4), 1—32.Conkle advocates a stronger religious presence in the public square, but he alsochallenges the principle of religious equality, i.e. the principle that all religionsshould be treated equally.

34. PL, 197n.32. See also, Rawls's "Fairness to Goodness," reprinted in CP, 267-84.35. Some of the significant political and theological differences among persons in the

United States who generally support church involvement in politics are presentedin the July 2003 report by the Pew Forum on Religion and Public Life, Religion andPolitics: Contention and Consensus (Washington, DC: The Pew Research Center,2003). I rely also on the April 2001 report by the Pew Forum, American Views onReligion, Politics and Public Policy (Washington, DC: The Pew Research Center,2001). See also Robert Booth Fowler, Allen D. Hertzke, and Laura R. Olson,Religion and Politics in America, second edn (Boulder: Westview Press, 1999),

The Legacy of John Rawls148

especially 137—73, discussing the Christian Right and African American Chris-

tianity, both of which include many evangelical Protestants.

36. On the growing religious diversity of the United States, see Diana Eck, A NewReligious America (New York: HarperCollins, 2001).

37. Wolterstorff, "The Role of Religion," 105.

38. Ibid., 116. See also Wolterstorff, "Why We Should Reject What Liberalism

Tells Us," 176—7. A similar argument is advanced by Christopher Eberle, who

turns to the Rawlsian notion of the "strains of commitment" in order to suggest

that parties in the original position would reject restraints that potentiallythreaten their moral identity. (See Eberle, Religious Conviction in Liberal Politics,

140-8.)39. On this problem, see my "Ecologism, Environmental Protection and Liberal-

Democratic Decision-Making," Interdisciplinary Environmental Review 5 (2003),

1-16.

40. Wolterstorff, "Why We Should Reject What Liberalism Tells Us," 176.41. In an essay that aims at a reconciliation of the views of Rawls, Audi, and Wolter-

storff, James Sterba provides an argument that is relevant here:

It would not do to correct one unfairness by imposing a similar or even greater

unfairness. We need to determine, therefore, whether a minority, religious or

otherwise, that loses out to a religious majority might also be unfairly treated,and if it is unfairly treated, whether that unfairness needs to be addressed as

much, or even more so, than the unfairness to which Wolterstorff has drawnour attention.

Though I am less convinced than Sterba that Wolterstorff has drawn our atten-

tion to an element of unfairness in public reason, I accept this basic argument.(See James P. Sterba, "Rawls and Religion," in Victoria Davion and Clark

Wolf (eds), The Idea of Political Liberalism (New York: Rowman & Littlefield,

2000), 34-45 (40).)

42. This general problem is also discussed by Rawls in "Constitutional Liberty and

the Concept of Justice," reprinted in CP, 87.43. For the criticism that restraints on religious discourse and argument would

weaken religious critiques of injustice, see Michael Sandel, "Political Liberal-ism," Harvard Law Review 107 (1994), 1765-94; David Smolin, "Cracks in the

Mirrored Prison: An Evangelical Critique of Secularist Academic and JudicialMyths Regarding the Relationship of Religion and American Politics," Loyola of

Los Angeles Law Review 29 (1996), 1488-1512; and Timothy P. Jackson, "TheReturn of the Prodigal? Liberal Theory and Religious Pluralism," in Religion and

Contemporary Liberalism, 182—219.

For the related criticism that restraints on religious discourse and argument

would deprive citizens of the benefits of robust religious-political dialogue,see Jeremy Waldron, "Religious Contributions in Public Deliberation," SanDiego Law Review 30(1993), 817-48; Jeffrey Stout, Democracy and Tradition (Prin-

ceton: Princeton University Press, 2004), 75—7; Philip L. Quinn, "Political

Liberalisms and Their Exclusions of the Religious," in Weithman (ed.), Religion

Public reason and religion 149

and Contemporary Liberalism, 138—61; and Wolterstorff, "The Role of Religion,"especially 110-11.

44. For a reply to this criticism, see David A.J. Richards, "Public Reason and Aboli-tionist Dissent," Chicago-Kent Law Review 69 (1994), 787-842. See also Pi, lii.

45. See, for example, the exclusionist view of Richard Rorty in his "Religion As Con-versation-stopper," in Philosophy and Social Hope (New York: Penguin Books,1999), 168-74.

46. On this point, see also Dombrowski, Rawls and Religion, 116.47. Perry, Under God?, 39-40. See also Perry, Religion in Politics (New York: Oxford

University Press, 1997), 44—5. Daniel Conkle also stresses the importance of test-ing religious convictions in his "Different Religions, Different Politics," espe-cially 28-30.

48. In Under God?, Perry rejects the idea that, just in virtue of the obligations of citi-zenship, citizens are required sometimes to exercise restraint in their political reli-ance on religiously grounded moral beliefs. But he also argues that Christiansshould be wary of religious arguments against the legal recognition of same-sexunions. (See Perry, Under God?, 55—97, and Perry, "Christians, the Bible, andSame-Sex Unions: An Argument for Political Self-Restraint," Wake Forest LawReview 36 (2001), 449-86.)

49. Weithman, Religion and the Obligations of Citizenship (New York: Cambridge Uni-versity Press, 2002) 141.

50. Ibid., especially 36—92. Both Hollenbach, in The Common Good and Christian Ethics,and Greenawalt, in Private Consciences and Public Reasons, seem to be interested in anapproach to liberal-democratic citizenship that would examine more carefullyempirical evidence about the role of religion in the public culture. See also JohnA. Coleman, S.J., "Public Religion and Religion in Public," Wake Forest LawReview 36 (2001), 279-304.

51. See especially Christopher J. Eberle, Religious Conviction in Liberal Politics, 81—151and Eberle, "What Respect Requires—And What It Does Not," Wake Forest LawReview 36 (2001), 305-52. See also Stout, Democracy and Tradition, 72-3; Wolter-storff, "The Role of Religion," 105-9; Perry, Religion in Politics, 59; and Neal,"Political Liberalism, Public Reason and the Citizen of Faith," 196—7.

52. Perry, "Religious Arguments in Public Political Debate," Loyola of Los AngelesLaw Review 29 (1996), 1421-58 (1452-7). Paul Weithman reconstructs and criti-cally examines an argument for why citizens should recognize the duty to reasonpublicly. (See Weithman, "Citizenship and Public Reason," in George andWolfe (eds), Natural Law and Public Reason, 125-70.)

53. For a discussion of the indeterminacy problem, see Perry, Religion in Politics,57—61; Quinn, "Political Liberalisms and Their Exclusions of the Religious,"150-1; Sandel, "Political Liberalism," 1778; Robert P. George and ChristopherWolfe, "Introduction," in Natural Law and Public Reason, 1—9; and Peter de Mar-neffe, "Rawls's Idea of Public Reason," Pacific Philosophical Quarterly 75 (1994),232-50(235).

54. As Sharon Lloyd has observed, one need not rely on the "deep" claims of a com-prehensive doctrine in order to make use of "argument by counterexample,

The Legacy of John Rawls150

thought-experiment, simple observation and uncontroversial empirical theory."

(See Lloyd, "Relativizing Rawls," Chicago-Kent Law Review 69 (1994), 709-35

(721).)55. PRR, 605.56. Here I reject the notion that citizens would regularly or typically encounter the

problem of indeterminacy on fundamental political questions. (On this point, see

Perry, Religion in Politics and PL, liii.)

57. PRR, 605.

58. Ibid., 585.59. Ibid., 582-3.

60. As I have argued elsewhere, Economic Justice for All, the U.S. Catholic bishops'1986 pastoral letter on the U.S. economy can be understood as a form of public

reason. (See my " 'Political, Not Metaphysical': Reading the Bishops' Letter as a

Form of Public Reason," Proceedings of the American Catholic Philosophical Association

77 (2003), 205-19.) Unless it is combined with other elements of Catholic doc-trine or social teaching, however, the bishops' letter does not present a "com-

plete" political conception of justice in the Rawlsian sense. Tolerating such

incompleteness would enable us to welcome a citizens' reliance on the letter as asource of political justification.

Public reason and religion 151

In recent years increasing attention has been paid to the new approach toKant's ethics developed by some American scholars. At first glance, whatconnects these authors is their attempt to offer an account of Kant's moral phi-losophy much closer to ordinary moral reasoning and able to challenge theusual charges of formalism and rigorism that, at least since Hegel and Schiller,have accompanied every rendering of Kant's moral philosophy. Beyond thelogical differences among proponents, this approach presents itself as a correc-tion of the conventional reading of Kant prevalent in the Anglo-Americanworld for many years. According to the new Kantians, the conventional read-ing, largely focused on Kant's Groundwork, lends itself to the usual criticisms offormalism and rigorism, thereby crediting other related charges, such asKant's alleged inability to explain moral motivation or its inherent tendencyto jeopardize the ethical integrity of the human agent.

By contrast, the new approach tends to belittle these criticisms insofar as itclaims to offer a more comprehensive reading of Kant's texts, which, byexpanding the deliberative aspect of Kant's practical reason, makes histheory more fit to address ordinary ethical problems. Indeed, were we togather under a single heading most of the topics these authors emphasize, wecould certainly focus on their account of practical reason and deliberation,intended to highlight the first-person perspective proper to Kantian moralphilosophy. As Korsgaard puts it:

Moral philosophy is the extension and refinement of ordinary practicaldeliberation, the search for practical reasons. This makes Kant's enterprisevery different from that of philosophers who talk about morality and themoral agent from the outside, third-personally, as phenomena that are inneed of explanation. Kant's arguments are not about us; they are addressed

3to us.

It is from this perspective that we can account both for their review of thetraditional versions of the Categorical Imperative, and for their developmentof other aspects perhaps neglected in the past. Particularly noteworthy is the

Introduction

Ana Marta Gonzalez

John Rawls and the new Kantian moral theory

8

2

emphasis on the following topics: the particular role of the HypotheticalImperative in practical deliberation; the complementarity—rather thanopposition—of the Hypothetical Imperative to the Categorical Imperative;the crucial role of the Categorical Imperative in Moral Judgment; and,finally, the way in which the different formulations of the Categorical Impera-tive account for significant moral differences in the way we deliberate aboutthe morality of particular actions and even provide the ground for a theory ofvalue. Barbara Herman's remarks, in the Preface to The Practice of MoralJudgment offers a brief overview of the shift in perspective and its implications:

It has been one of the givens of Kant's ethics that everything there was to sayabout moral judgment belonged to the interpretation of the CategoricalImperative tests, and that all of the difficulties in this area were a speciesof problems about universalization: the difficulty of deriving content froma formal procedure and the problem of action description are the two mostfamous. I argue that it is much better to see the Categorical Imperative andits tests as an aspect of moral judgment—setting its terms, I would say —but needing to be placed in a framework that can explain moral perception,deliberation, and (internal) criticism. How the Categorical Imperativeworks cannot be understood apart from a reasoned view of the kind ofresults it is able to generate and of its place in the moral agent's complexfield of response and deliberation ... I argue for the unusual view that Kan-tian moral judgment depends on the availability of an articulated concep-tion of value—in particular, of the value of the fully embodied person.

Now, at first glance, and precisely because of its emphasis on the practicaldimension of the Categorical Imperative, the new approach seems to owe agreat deal to Onora O'Neill's book on the Categorical Imperative. Indeed,in a context still heavily marked by utilitarianism, O'Neill's work representedone of the first attempts to make sense of Kantian moral theory in terms verysimilar to those of the new Kantian moral theory. Yet there is something dis-tinctive in the new Kantian moral theory that the connection with O'Neill'swork does not enable us to grasp entirely. In the case of the new Kantian moraltheory, the focus on practical reason is intended both to highlight the centralrole of the concept of "rational nature" in Kant's moral philosophy, and tomake room for a more satisfactory account of the Kantian moral agent, anaccount that succeeds in showing the relevance of character for moral judge-ment. Once again, Herman's words help illustrate this aspect of the new Kan-tian moral theory:

A great deal of recent criticism of Kantian ethics has targeted its thin con-ception of the person, the inadequacy of its idea of character, its stultifyingrestriction on admissible moral motivation, and its mistaken views about

John Rawls and the new Kantian moral theory 153

4

6

the place of impartial moral requirement in a good human life. These criti-cisms live off the mistaken view of moral judgment as involving algorithmicemployment of tests, and its attendant picture of the moral agent as seekingto bring her will into conformity with principles of duty. With this view ofjudgment out of the way, it becomes possible to see that Kant's notionsof virtue and character are in no way peripheral to the understanding ofmoral judgment and action. We are able to consider the nature of a Kantianmoral agent—what motives, feelings, thoughts, and commitments guideher deliberations and actions. There is then room to develop an account ofmoral personality that places moral activity within the ongoing practicalcommitments of a good life.

Thus, while the analysis of the process of deliberation and moral judgementis at the core of the new Kantian moral theory, it is the reference to humannature, broadly understood, that better explains the distinctiveness of thenew approach. Certainly, the thesis of the centrality of the concept of"human nature" in ethics, controversial as it is when attributed to Kant,loses something of its implausibility when we take it to mean "rationalnature" and come to understand the deep revision of the concept of reasoneffected by Kant himself. Still, such a formulation is likely to surprise morethan one scholar, for at least verbally it clearly confronts the conventionalreading of Kant, who repeatedly rejects any kind of reference to the particula-rities of human nature within ethics.

Now, if we ask for the origin of this approach to Kant's ethics, we would findthat for many proponents a confessed common source is to be found in JohnRawls's Lectures on Ethics at Harvard beginning in the 1960s, lectures thatO'Neill herself had the opportunity to attend. In what follows, I have triedto explore the extent of Rawls's influence on the development of this "newKantian moral theory" by highlighting some aspects of Rawls's own interpre-tation of Kant, and demonstrating its connection to the characteristic toneand topics raised by the new Kantians.

Rawls's influence on the new Kantian moral theory

According to some of his former students, at the time when Rawls delivered hislectures on the history of ethics at Harvard, scholars rarely paid attentionsimultaneously to the philosophical arguments and to the history of philoso-phy, much less attempted to show the relevance of the history of ethics for con-temporary issues. Meta-ethics largely dominated the academic discussion, andit was unusual to address substantive ethical problems. Against this back-ground, Rawls's lectures "offered reconstructions of classical arguments thatreclaimed their power and their capacity to inform contemporary concerns."

The Legacy of John Rawls154

9

1 2

It was precisely this approach to the history of philosophy that would influ-ence the future work of many of Rawls's students. This influence has nowbecome particularly evident in the work of some prominent scholars who areadvancing this new interpretation of Kant's ethics. Thomas Hill, BarbaraHerman, Christine Korsgaard, and Andrews Reath all studied under Rawlsand share his basic commitment to discuss the history of ethics in a way rele-vant to the contemporary world. Besides being specifically devoted to Kant'smoral philosophy, they seem likely to echo Rawls's own approach to Kantin significant ways. As Korsgaard observes, Rawls's influence on contempor-ary Kantian scholarship has been both methodological and substantive,though we should not artificially separate the two aspects. While Rawls'smethodological influence on contemporary Kantian scholarship lies in his con-viction that textual analysis and philosophical reflection on a philosopher'swork should go together, Rawls's substantive influence relates radically to hisconception of Philosophy as a deeply practical project, an approach thatcould itself be considered the development of a possibility embedded inKant's own philosophy.

Philosophy as a deeply practical project

Indeed, according to the interpretation developed by Susan Neiman—also aformer student of Rawls—Kant's transformation of the concept of reasonwould amount to a new conception of philosophy not so much as a theoreticalundertaking—directed to the acquisition of knowledge—as a practical andmoral one, where the important thing becomes the possible contributionof philosophy to the fulfillment of the human vocation. Rawls's attitudetoward philosophy is, in this sense, undeniably Kantian, since Rawls's owntheory clearly aims at the realization of a moral ideal. From this perspective,it credits Kant's asserted "Primacy of Practical Reason."

Asserting the primacy of practical reason, indeed, means to orientate one'sthinking according to an ideal of reason, such that the leading questionbecomes this one: how must we think of ourselves and of the world in orderto fulfill our moral vocation—to shape a just society? Rawls's doctrinemeets this requirement. Yet, does this suffice to qualify a doctrine as Kantian?As Rawls himself points out: "Kant's view is marked by a number of dualisms,in particular, the dualisms between the necessary and the contingent, formand content, reason and desire, and noumena and phenomena. To abandon thesedualisms as he meant them is, for many, to abandon what is distinctive inhis theory." However, Rawls goes on to explain why he disagrees with thisinterpretation, arguing that Kant's "moral conception has a characteristicstructure that is more clearly discernible when these dualisms are not takenin the sense he gave them but reinterpreted and their moral force reformulated

John Rawls and the new Kantian moral theory 155

1 4

13

within the scope of an empirical theory. One of the aims of A Theory of Justicewas to indicate how this might be done."

While we can certainly speak of similarities between Kant's and Rawls'sconceptions of philosophy, Rawls's comments signal the existence of aclear difference between them. This difference points at the particular twistKant's philosophy is likely to receive when developed against the Americanphilosophical tradition. Actually, the text that has just been quoted suggeststhat Rawls considers the reinterpretation of Kant's dualisms an essentialcondition for making Kant's view effective as a moral ideal—perhaps inRawls's own theory. Now, it is Rawls himself who observes that such a reinter-pretation would bring his own theory closer to Dewey. Thus, in the first ofhis Dewey Lectures, after praising Dewey's attempt to adapt what is valu-able in Hegel's theory "to a form of naturalism congenial to our culture,"he added: "There are a number of affinities between justice as fairness andDewey's moral theory which are explained by the common aim of overcomingthe dualisms in Kant's doctrine."

Indeed, very much in a Deweyan spirit, overcoming the Kantian dualismsamounts to assuming, as already present in the public culture, many of therequirements of a conception of justice. As Rawls asserts, "On the Kantianview that I shall present, conditions for justifying a conception of justice holdonly when a basis is established for political reasoning and understandingwithin a public culture." That basis is specifically provided by the Americanpolitical tradition:

We are not trying to find a conception of justice suitable for all societiesregardless of their particular social or historical circumstances. We want tosettle a fundamental disagreement over the just form of basic institutionswithin a democratic society under modern conditions. We look to ourselvesand to our future, and reflect upon our disputes since, let's say, the Declara-tion of Independence. How far the conclusions we reach are of interest in awider context is a separate question.

Rawls's conception of justice therefore seeks to articulate the ideals alreadyimplicit in the common sense or the public culture of a specific democraticsociety. Not aiming at the universal as such, his conception is designedto satisfy the requirements of justice within a particular political tradition.For this very reason, Rawls's attempt amounts to a justification of socialinstitutions through a method certainly indebted to the Kant of the transcen-dental deductions. It could also, however, be considered equally close toHegel or Dewey, in that Rawls applies that justification procedure to socialinstitutions in a way that takes a certain political tradition for granted:

156 The Legacy of John Rawls

1 6

1 7

1 9

20

What justifies a conception ofjustice is not its being true to an order antece-

dent to and given to us, but its congruence with our deeper understanding of

ourselves and our aspirations, and our realization that, given our history

and the traditions embedded in our public life, it is the most reasonable doc-

trine for us. We can find no better basic charter for our social world.

In this way, the primacy of practical reason, an unequivocal Kantian fea-

ture of Rawls's political philosophy, receives a cultural specification. The

transcendental philosophy is to accomplish its justifying mission within a par-

ticular political culture. Accordingly it is not surprising that Rawls rejects the

charge of "formalism" usually directed against Kant since Hegel first made

his case. Indeed, reading Kant from within a certain political culture—that

is, reading him in a Hegelian key—overcomes that objection. All the content

needed by practical reason is already present in the public culture.

It seems to me that this feature of Rawls's interpretation of Kant is also pre-sent in the new Kantian moral theory. Whenever these authors reject the tra-

ditional objection of formalism, the underlying assumption is that there is

no need to read Kant in an abstract way; it is equally possible to read him from within a

particular culture. On this view, the agent does not need to deprive herself of

her moral experience when she is engaged in the process of moral delibera-

tion and judgement, that is, when she is to make use of the categorical impera-

tive. This point is particularly clear in Barbara Herman's account of moral

judgement:

An agent who came to the GI [Categorical Imperative] procedure with no

knowledge of the moral characteristics of actions would be very unlikely to

describe his action in a morally appropriate way. Kant's moral agents are

not morally nai've. In the examples Kant gives of the employment of the GI

procedure (G422—423), the agents know the features of their proposed

actions that raise moral questions before they use the GI to determine their

permissibility. It is because they already realize that the actions they want

to do are morally questionable that they test their permissibility.

Now, if we ask where this antecedent knowledge comes from, Herman's

answer ultimately manifests the cultural link that distinguishes this new Kant:

It is useful to think of the moral knowledge needed by Kantian agents

(prior to making moral judgments) as knowledge of a kind of moral rule.

Let us call them "rules of moral salience." Acquired as elements in moral

education, they structure an agent's perception of his situation so that

what he perceives is a world with moral features. They enable him to pick

John Rawls and the new Kantian moral theory 157

2 2

2 3

out those elements of his circumstances or of his proposed actions thatrequire moral attention.

Of course, while Rawls's position, as conveyed in the Dewey Lectures, couldbe read as implying an abandonment of the universal scope of a Theory of Jus-tice, to focus on some problems proper to modern liberal democracies—andthus to a particular culture—Herman's point is a more general one, playingat a more basic level: the categorical imperative works solely on the assump-tion of some rules of moral salience, no matter the culture. And yet, since theserules, according to Herman, are acquired "as elements in moral education,"each individual would find those rules embedded in his/her culture. In otherwords: although Herman's point, unlike Rawls's, is a universalist one, Her-man's non-formalistic way of reading Kant surely has been inspired byRawls's own sensibility for the cultural embodiment of reason, for, after all,the rules of moral salience may be defined differently in different cultures.

Herman's "rules of moral salience," however, are not the only aspect ofthe new Kantian Moral Theory in which the cultural linkage resulting fromthe Kant-Dewey "marriage" becomes evident. Another aspect can be foundin the frequent references to the plausibility of a theory as a decisive factorfor its acceptance. "Plausibility" is, indeed, the word most often employed bythese authors to describe their goal, both as Kantian scholars and moral phi-losophers. The new Kantians seek to offer a reconstruction of Kant's moralphilosophy that is, as Thomas Hill puts it, "as plausible as possible," inorder to put it in dialogue with contemporary ethics. Both aspects are sum-marized in Barbara Herman's description of her own approach to Kant'stexts: an attempt to provide an interpretation "which makes sense of the textand makes the texts make sense."

Now, as Thomas Hill points out, "Kant's ethics is most plausible when seenas a less comprehensive account of morality than he thought;" plausibility,then, demands "sympathetic reconstruction and extension of certain coreKantian ideas but also critically abandoning some of Kant's ideas on par-ticular issues that prove to be untenable and unwarranted by Kant's morebasic theory." Hill's remark seems a clear echo of Rawls's previously quotedjustification for reinterpreting and reformulating Kant's dualisms. A remark-able example of this tendency to solve or mitigate Kant's dualisms is Kors-gaard's interpretation of the phenomenon-noumenon distinction in terms of twodifferent standpoints, an interpretation that deprives the distinction of someof those "scandalous" ontological reverberations, which, in Herman's words,make it "unacceptable to us" :

This view is not, as so many have supposed, an ontological or metaphysi-cal theory according to which we exist simultaneously in two different

158 The Legacy of John Rawls

2 4

2 5

2 8

26

"worlds", one somehow more real than the other. As I understand it, it goeslike this: In one sense the world is given to us, it appears to us, and we arepassive in the face of it. We must therefore think of the world as generatingthe appearances, as giving them to us. The world insofar as it appears to us isphenomenal; the world insofar as it generates the appearances is noumenal.We can only know the world as phenomenal, that is, insofar as it is given tosense, but we can think of it as noumenal. So there are not "two worlds", butrather one world which must be conceived in two different ways. And all ofthese points apply above all to ourselves. When we view ourselves as phe-nomena, we regard everything about ourselves, including inner appear-ances such as thought and choices, as parts of the natural world, andtherefore as governed by its laws. But insofar as we are rational, we alsoregard ourselves as active beings, who are the authors of our thoughtsand choices.

While Korsgaard's interpretation certainly finds support in Kant's texts,her conciliatory approach does contrast with other well-known interpreta-tions. We just need to think of Jaspers, for whom the phenomenon-noumenondistinction meant a contradiction bound to arise precisely because the philo-sophical idea Kant wanted to express cannot be held in any logical expres-sion. For Jaspers it was this tension that made Kant's thinking "a matrixof seemingly inexhaustible possibilities." Indeed, one of these possibilities—but only one of them—is what we are considering right now. And as a parti-cular interpretation among others, it will inevitably prove controversialamong other Kant scholars. What I would like to highlight in regard tothe new Kantian moral theory, however, is what I take to be its distinctivehallmark: namely, its explicit reference to the ordinary and common experi-ence as hermeneutical keystone, able to develop a contemporary version ofKant's ethics that can be put in dialogue with other contemporary ethicaltheories.

While the allusion to common experience could be taken as an invitation tofigure out how the teleological concepts we use in ordinary life to articulateour moral experience can be justified within a Kantian framework, it canalso imply a more cultural claim. As one scholar recently pointed out, the allu-sion to common experience as the ultimate reference of philosophical reflec-tion is quite characteristic of twentieth-century philosophy. At the sametime, however, the meaning of "common or ordinary experience" is in manyways elusive, and always open to interpretation. What seems moral commonsense within a certain culture can look otherwise in a different one. There is notheoretically neutral conception of everyday life.

From this perspective, Rawls's decision to develop his own proposalagainst the background of the American political tradition proved a wise

John Rawls and the new Kantian moral theory 159

3 2

3 3

3 6

3 7

3 9

methodological restriction. Indeed, it is perhaps because of the implicit parti-cularization of his otherwise abstract discourse that Rawls's theory hasproven, in fact, to be of interest to people coming from different traditions.What the latter suggests is that Kant's transcendental deduction worksagainst the background of particular experiences, to the extent that we suc-ceed in highlighting their essential features. Of course, the identification ofthose essential features remains in need of further explanation, but Rawls canskip this elucidation since his declared aim is practical, not epistemological.Particularly, his Political Liberalism is expressly designed to leave such episte-mological and metaphysical questions aside.

We could assume that, insofar as the new Kantian moral theory follows thepractical path, it can also leave those questions aside. But does the reference to"plausibility" within the context of the new Kantian moral theory have afunction similar to Rawls's reference to the American political tradition? Inother words: is it backed by a similar transcendental deduction? What is itthat makes a theory plausible? Actually, the only criteria of plausibility thatan ethical theory should meet, according to Hill, is its compatibility withscientific knowledge. Scientific knowledge, indeed, appears as a major fea-ture of our contemporary world view that any plausible moral theory has tokeep in mind: "Moral theory is not science, of course, but any moral theorythat is worthy of contemporary support should, in my opinion, at least be com-patible with empirical explanations regarded as well established in the currentscientific community."

What counts as well established in the current scientific community is, ofcourse, as changing a matter as what counts as current community. Given theuncertainty of scientific facts and their dependence on conventional interpre-tations, at least in some aspects moral theory must be open to fluctuation alongwith those changeable assumptions. Now, insofar as these references to ourcultural and scientific practices represent a distinctive feature of the new Kan-tian moral theory, what this theory suggests is a revision of the role played bythe empirical in Kant's own moral theory. Such a revision is intended to makesense of the otherwise "too dry" pure moral theory, developed by Kant in theGroundwork and the Second Critique. By contrast, the so far somewhat neglectedtext of the Metaphysics of Morals is receiving increasing attention. The idea isnot so much to blur Kant's distinction between the pure and the empirical inethics, but to show how both aspects interact in practice while maintainingwhat we could call a "Kantian framework."

Kant's moral constructivism

The distinctive features of the "Kantian framework" as assumed by the newKantian moral theory can, once again, be traced back to Rawls's view of

160 The Legacy of John Rawls

4 1

Kantian moral constructivism, as opposed to "rational intuitionism." Rawlsregards "rational intuitionism" as the dominant moral theory "from Platoand Aristotle onward until it was challenged by Hobbes and Hume"; ques-tionably he uses the same title to refer to "the view exemplified in the Englishtradition by Clarke and Price, Sidgwick and Moore, and formulated in itsminimum essentials by W.D. Ross."

Rawls describes "rational intuitionism" as characterized by two features:"first, the basic moral concepts of the right and the good, and the moralworth of persons, are not analyzable in terms of nonmoral concepts (althoughpossibly analyzable in terms of one another); and, second, first principles ofmorals (whether one or many), when correctly stated, are self-evident propo-sitions about what kinds of considerations are good grounds for applying oneof the three basic moral concepts, that is, for asserting that something is(intrinsically) good, or that a certain action is the right thing to do, or that acertain trait of character has moral worth."

By contrast, Kant's moral theory requires "that there is no such order ofgiven objects determining the first principles of right and justice amongfree and equal moral persons." Otherwise the ethical principle would beheteronomous, while for Kant the ethical principle must be autonomous.At this point, however, two clarifications are needed.

In the first place, Rawls observes that "a Kantian doctrine of autonomyneed not deny that the procedures by which first principles are selected aresynthetic a priori," provided that such procedures be "suitably founded onpractical reason, or, more exactly, on notions which characterize persons asreasonable and rational." In other words, we must distinguish between thefirst principles and the procedures to select them: while the first are entirelya priori, the second can include some appeal to experience.

A further clarification regards the so-called "moral facts": to call a moraldoctrine "constructivist" does not mean to consider that "moral facts, muchless all facts, are constructed. Rather, a constructivist procedure providesprinciples and precepts that specify which facts about persons, institutions,and actions, and the world generally, are relevant in moral deliberation.Those norms specify which facts are to count as reasons" ; but the facts them-selves are already "available in our everyday experience or identified by the-oretical reason."

While we may once again recognize in those words the claim that moraltheory is not supposed to replace everyday moral experience but rather togive an account of it, we could nevertheless still ask how we come to qualify afact as a "moral fact," in the absence of any kind of "moral intuition." Bar-bara Herman's notion of "rules of moral salience," to which I referred above,could provide a provisional answer to this problem. But if we ask whence therules of moral salience come, we have but two options left: either admitting a

John Rawls and the new Kantian moral theory 161

4 3

4 4

4 8

kind of moral intuition, or else developing a more or less sophisticated natur-alistic account. In her article "Making Room for Character," BarbaraHerman explores this latter path, developing a complex account of theorigin of those rules that connects them with the natural history of desire.While the account as such may not easily be classified as Kantian, it certainlymanages to be very plausible.

Without entering in this problem himself, Rawls did argue that Kant'smoral doctrine can be described as constructivist because the content of themoral law, and the relevance of whatever moral facts we find in our ordinaryexperience, is to be determined as a result of applying the Categorical Impera-tive procedure. Rawls sees this procedure as working in four steps.

First of all, "we have the agent's maxim, which is, by assumption, rationalfrom the agent's point of view"; he specifies further that "the maxim is alsoassumed to be sincere" and its form is that of a particular hypothetical impera-tive: "I am to do X in circumstances C in order to bring about Y."

The second step would be the generalization of the maxim to be achieved:"Everyone is to do X in circumstances C in order to bring about Y."

The third step introduces the reference to a law of nature: "everyone alwaysdoes X in circumstances C in order to bring about Y (as if by a law of nature)."

And finally, Rawls says, we would be required to add the law of nature weourselves have imagined to the existing laws of nature, in order to see whatkind of world would arise.

At this point the test would consist in asking ourselves two things: (a) WouldI really be able to act on my maxim within the perturbed social world? and(b) Would I possibly be willing to act on my maxim in the perturbedsocial world?50

Now, in order to overcome some difficulties arising from the analysis ofKant's own examples, Rawls suggests introducing two ideas: the notion of"true human needs," along with that of "two limits of information." Takentogether, argues Rawls, these conditions require that we "see ourselves asproposing the public moral law for an ongoing social world enduring overtime." Now, while the introduction of the notion of "true human needs" isintended to avoid formalism, the requirement of limiting the information,which immediately suggests Rawls's own "veil of ignorance," is intended to"enable us to see what Kant means when he says that the moral law disclosesour freedom to us."

The disclosure of freedom takes place precisely to the extent that we main-tain the priority of pure practical reason over empirical practical reason.Thus, while empirical practical reason is at work when the agent formulateshis or her maxim, the imposition of restrictions on empirical practical reasonby introducing the universalization requirement discloses an essential fea-ture of the human agent: he is not merely a rational agent, able to formulate

The Legacy of John Rawls162

49

54

hypothetical imperatives, but a reasonable agent, able to subject himself to auniversal law. Accordingly, we can imagine a particular agent coming to theGI procedure with a particular set of inclinations, desires, and ends that hewould want to realize, deliberating about the best possible means to realizethose ends according to "The Hypothetical Imperative," and, as a result, gen-erating a particular maxim in the form of a hypothetical imperative. Now, thisis the maxim, which, once subjected to the universalization requirement,brings about the moral content, in terms of permissibility or impermissibilityof the particular action. Beyond this, what the whole procedure shows is thatthe human agent is endowed with two different powers—the rational and thereasonable—which reveal the underlying conception of the human person:namely, "the conception of free and equal persons as reasonable and rational,a conception that is mirrored in the procedure."

Rawls insists that this conception of the person is not constructed, but rather ismirrored in the GI procedure, which is not constructed either. Indeed, accord-ing to him, the GI procedure is merely laid out, since we must take it as implicitin everyday moral reasoning. This is something affirmed by Kant himself inseveral places, most noticeably in the Second Critique, when he speaks of the"Typic of Moral Judgment." Thus, both the GI procedure and the concep-tion of the person it mirrors—along with the conception of a society of suchpersons—are at the basis of Kant's moral constructivism.

For Rawls, then, Kant's moral constructivism follows straightforwardlyfrom his demand for autonomy, and thereby from a conception of the personwhose denning characteristics are freedom and equality. Both features becomeparticularly evident in the third formulation of the Categorical Impera-tive, the Kingdom of Ends, which most properly represents the moral idealelicited from Kant's moral theory. But while freedom and equality, as denn-ing characteristics of persons, are fully disclosed in Kant's explanation of theKingdom of Ends, the basic features of the person as reasonable and rational,capable of an effective sense of justice and of pursuing a conception of thegood, are already present in the formula of Universal Law. "Moral personal-ity' ' is the term Rawls employs to refer to these moral powers:' 'The first poweris the capacity for an effective sense of justice, that is, the capacity to under-stand, to apply and to act from (and not merely in accordance with) the prin-ciples ofjustice. The second moral power is the capacity to form, to revise, andrationally to pursue a conception of the good."

Unlike the capacity for justice, the idea of "pursuing a conception of thegood" could sound not entirely Kantian. Yet in his article "Themes inKant's Moral Philosophy," after showing how the Categorical Imperativeworks, Rawls suggests the several ways in which the concept of good finds aproper place in Kant's moral theory: (1) the conception of happiness as organ-ized by The (as opposed to a particular) Hypothetical Imperative; (2) the

163John Rawls and the new Kantian moral theory

5 5

5 7

5 9

The Legacy of John Rawls

60fulfillment of true human needs; (3) the good as fulfillment in everyday lifeof what Kant calls "permissible ends"; (4) good will, as the supreme (althoughnot complete) good of persons; (5) the good as the object of the morallaw, which is the realm of ends; (6) Kant's conception of the complete orhighest good.

Through the specification of the various ways in which the concept of goodis to be found in Kant's ethics, Rawls intended to clarify the way in which theReasonable—which issues the principles of justice by following the CategoricalImperative—supervenes over the Rational—thereby making it possible tospeak of a "moral good." Accordingly, although the agent comes to the GIprocedure with certain conceptions of the good, what counts properly as a"moral good" would be a result of the procedure, that is, a result of the restric-tions imposed by the Reasonable over the Rational.

Thus, by suggesting that these two uses of reason—the Rational and theReasonable—were implicit in Kant's moral theory, Rawls was highlighting,against the background of his own theory of justice, the practical implicationsof the Kantian concept of the (human) person. He holds this conception of theperson as playing a central role in Kant's moral philosophy. At the sametime, he also made clear that, "unless this conception (of the person) and thepowers of moral personality it includes—our humanity—are animated, as itwere, in human beings, the moral law would have no basis in the world."As the example he quotes suggests, by saying this he was trying to stress thefact that reason—or humanity for that matter—must be embodied in real,particular, human beings. Therefore it is not humanity in general, but human-ity realized in particular human beings, that makes morality something real. Now,since the source of particularization in Kant comes mainly from the empiricalside of nature, the latter demand brings us back to the importance of theempirical.

As has already been pointed out, the attempt to show how the empiricalworks within Kant's ethics is a distinctive mark of the new Kantian moraltheory. The attempt, of course, has to face the traditional objections of form-alism and rigorism, particularly challenging when we deal with the problem ofmoral motivation. Before taking up this topic, however, I would like to pointout that the main aspects of the Rawlsian interpretation of Kant that I havebrought up here are easily recognizable in the writings of the new Kantians.

Human nature and practical reason

In fact, both Rawls's focus on the person and moral personality, on the onehand, and his account of practical rationality developed in terms of the dis-tinction between the Reasonable and the Rational, on the other, are two

164

6 2

6 5

noticeable characteristics of the new Kantian moral theory. Thus, Thomas

Hill's introduction of the "Hypothetical Imperative" as a principle that

works in parallel, instead of colliding, with the "Categorical Imperative,"

represents an attempt to clarify Kantian practical deliberation, largely in

debt to Rawls's own distinction. The same could be said of Korsgaard's

characterization of both the hypothetical and categorical imperatives as con-

stitutive principles of actions.

Likewise, the noticeable shift of attention from the Formula of Universal

Law to the Formula of Humanity—especially evident in Korsgaard—could

be interpreted in light of Rawls's discovery of the centrality of the person in

Kant's moral philosophy, although it cannot be literally traced back to him.

The latter is clear, for even if Rawls insisted on pointing at the conception

of the person behind Kant's use of the categorical imperative, thereby suggest-

ing the idea of rational nature as the source of value, he did not develop a

particular argument for this point, based on the interpretation of Kant's For-

mula of Humanity. Incidentally, Rawls limited himself to pointing out that

the Formula of Humanity should not be interpreted as introducing new

requirements beyond those already made explicit in the formula of Universal

Law. At the same time, he suggested that, as long as "humanity" is under-

stood as "our pure practical reason together with our moral sensibility," the

Formula of Humanity could not be interpreted in isolation from the Meta-

physics of Morals.

While in saying this, Rawls may have encouraged contemporary attempts

to relate the Groundwork and the Metaphysics of Morals, it should be kept in mind

that at that point he was taking humanity as synonymous with "moral person-

ality." In doing so, however, he would depart from Kant, who had drawn a

sharp distinction between humanity and personality. According to that dis-

tinction, humanity would include pragmatic, but not moral reason. Following

Kant's distinction, however, as well as his definition of Humanity in terms of

"the capacity to propose an end to oneself," Korsgaard would manage to

relate the Formula of Humanity to a theory of value which is ultimately

grounded in rational nature. According to her, then, the Formula of Human-

ity would command us to respect "the capacity for the rational determination

of ends in general, not just the capacity for adopting morally obligatory

ends" (which would not be humanity, but personality):

Humanity, completed and perfected, becomes personality, so that in treat-

ing the first as an end in itself we will inevitably be led to realize the second.

Thus, in the Critique of Practical Reason, humanity in one's own person and

personality are spoken of as if they were the same thing (G2, 87). But the

distinctive feature of humanity, as such, is simply the capacity to take a

rational interest in something: to decide, under the influence of reason,

John Rawls and the new Kantian moral theory 165

6 7

6 8

7 1

that something is desirable, that it is worthy of pursuit or realization,that it is to be deemed important or valuable, not because it contributes tosurvival or instinctual satisfaction, but as an end—for its own sake.

Now, focusing on the distinctive human capacity of "taking a rational inter-est in something," and, more precisely, on the fact that such capacity is notonly good as an end, but also intrinsically valuable, Korsgaard advances anargument to show the implications of Kant's Formula of Humanity for thedevelopment of a theory of value. The argument is not without problems,though. It is certainly true that Kant himself sets the basis for a theory ofvalue in the Groundwork, when he introduces the distinction between dignityand price, distinguishing further between market price and fancy price, andemphasizing that "autonomy is the ground of the dignity of human natureand of every rational nature." However, the argument for an intrinsic valueof humanity somewhat disconnected from personality (and morality) couldsound surprising, given the intrinsic connection that Kant himself introducesbetween dignity and morality. Thus, he says: "morality, and humanity insofaras it is capable of morality, is that which alone has dignity." Along the samelines, at the beginning of the Groundwork, Kant had stated clearly that only agood will is to be regarded as intrinsically good.

Korsgaard's argument for the intrinsic value of humanity, however, beginsby taking into account the necessary connection between "reason'' and' 'good"in the following terms: "a rational action must be done with reference to an endthat is good." Now, in Kant's account, it is not any prior knowledge of the goodthat provides us with a reason for acting, but rather it is our having a sufficientreason for (doing) something that justifies a particular end as good: "a good endis one for which there is a sufficient reason." With that in mind, Korsgaardpoints to the passage where Kant argues for the Formula of Humanity by pro-viding a kind of regressive account of "reasons for acting" until he finds a"sufficient reason" only in the rational being as the only being who repre-sents his or her existence as an end in itself. Korsgaard posits that the reasonfor this claim lies in the very fact that in choosing any other object, a rationalbeing not only takes that thing to be valuable for him (subjectively), but takeshimself as necessarily valuable. In other words, Korsgaard maintains that:

rational choice has . . . a value-conferring status. When Kant says: "rationalnature exists as an end in itself. Man necessarily thinks of his own existencein this way; thus far it is a subjective principle of human actions" (G 429),I read him as claiming that in our private rational choices and in general inour actions we view ourselves as having a value-conferring status in virtue ofour rational nature. We act as if our own choice were the sufficient conditionof the goodness of its object: this attitude is built into (a subjective principleof) rational action.

The Legacy of John Rawls166

7 2

7 4

7 8

Accordingly, Korsgaard's thesis rests in two considerations: (1) one does notchoose something because she discovers some value in it. It is rather her choos-ing something what makes it valuable to her eyes. (2) This very operationinvolves taking oneself—one's rational nature—as the ultimate source ofvalue. Although it is far from certain that Kant actually developed this argu-ment, the argument as such is certainly Kantian, for the way it gives priority tothe rational over the good. In doing so, it is certainly constructivist. At thesame time, the focus on the agent involved in this kind of reasoning can betraced back to Rawls's own focus on the person and moral personality.

The issue of moral motivation

As I suggested above, an important aspect of Rawls's influence on the newKantians lies in the implicit invitation to review the role of the empiricalin Kant's moral theory. This naturally leads our attention to the way the newKantians respond to the objections of formalism and rigorism traditionallyaddressed to Kant's ethics. Here I shall solely focus on the more particularissue of moral motivation, which is somehow at the intersection of both prob-lems. Thus, while alleged rigorism consists in the demand "to act from themotive of duty alone," formalism would impose the impossible requirementof acting because of the universality of the law, regardless of its content. Yetboth things seem either undesirable (so Schiller's objection against actingfrom duty alone), or simply impossible—as Hegel pointed out in regard tothe categorical imperative, and as the moral philosophers coming from anempiricist tradition have argued specifically in regard to motivation.

Since the empiricist objection goes back to Hume's skepticism about practi-cal reason, traditional Kantians have usually rejected the charge as missingthe real point of Kant's position—namely, that pure practical reason doesexist. In other words, in order to be practical, or to move to action, reasondoes not need any antecedent sensible feeling or expectation. Reason canmove us to act by the mere representation of the law. Yet can a purely formallaw be practical at all? At this point, all depends on understanding that theuniversal character of reason requires that it has to be determined by some-thing equally universal. For Kant, this "something universal" is nothingother than the law. To the extent that we act according to the universal law,we are acting well. On the other hand, to the extent that we act according to ageneral—but not universal—principle, we are, however unnoticeably, lettingsomething empirical get in the way. We are not acting as autonomous agents;we are acting badly.

It seems to me that one cannot abandon this picture without abandoningKant. Yet the picture should be purified of some misunderstandings implied

John Rawls and the new Kantian moral theory 167

in the expression, "acting from the motive of duty." Perhaps a good way to

clarify those misunderstandings is to distinguish more clearly between the

question of "the determining ground of morality" and the question of motiva-

tion in general. For Kant, the determining ground of morality, as just said, is

the universality of the law, and this is also what Kant comprises under the

notion of "duty." But Kant also distinguishes between duties according to

their content, so that it must be possible to speak of motivation in a more quali-

fied sense. In other words: in performing a duty of beneficence, we must

certainly be determined by duty—and thereby, by some form of categorical

imperative—but, at the same time, there must be something that allows us,

as agents, to distinguish between the duty of beneficence and some other

duties. I believe this is the question both Herman and Korsgaard have in

mind when they place so much weight on Kant's distinction between "incen-

tives (Triebfedern}" and "motives (Bewegungsgrund}."

Kant introduces this distinction in the following passage of the Groundwork:

The subjective ground of desire is an incentive; the objective ground of voli-

tion is a motive; hence the distinction between subjective ends, which rest on

incentives, and objective ends, which depend on motives, which hold for

every rational being. Practical principles are formal if they abstract from

all subjective ends, whereas they are material if they have put these, and

consequently certain incentives, at their basis.

The former passage suggests that incentives are usually sensible, though it is

not always so (think of the moral incentive). Their role is to instigate in us the

possibility of an action. Now, any action is supposed to involve a maxim (sub-

jective principle for action), which must be checked against the Categorical

Imperative. Before doing so we are still at the level of empirical practical

reason (recall Rawls). Only when we have checked the maxim against the

Categorical Imperative, and ascertained its (possible) universalization, are

we in a condition to determine our reason for the morally good, because it is

only then that the representation of the universal law can determine the uni-

versal nature of our reason, and only then that the so-called moral incentive

properly arises.

At this point, it is important to notice that, while it is the universality of the

law—and thereby the Categorical Imperative—that determines reason,

the Categorical Imperative finds its deliberative field within the material pro-

vided by the empirical practical reason, in the maxim of the action suggested

by the incentive. Were it not for this maxim, and ultimately the original incen-

tive for action, we could not have even applied the Categorical Imperative.According to Herman it is in the maxim where we should look for the proper

motive for an action, because it is the maxim that shows what the particular

168 The Legacy of John Rawls

8 0

reasons for an action are. Her account of "motives" as "reasons for acting," isthus to be opposed to the empiricist conception of "motives as desires"—where "desires" are taken as "causes," rather than as "reasons." The sameconclusion follows from Korsgaard's definition of motive as "an incentive plusa principle." In both cases we get a picture that permits us to maintain alegitimate diversity within the motive of duty—that is, acting from themotive of duty does not involve the neutralization of all significant features ofmoral character, as expressed in the virtues. Thus, Herman writes:

The man of sympathetic temper responds to suffering and takes thatresponse to give him a reason to help. Only then does he act from themotive of sympathy. An action that is done from the motive of duty is per-formed because the agent finds it to be the right thing to do and takes itsrightness or requiredness as his reason for acting. He acts from the motiveof duty with a maxim that has moral content.

Does Herman's and Korsgaard's move involve the rejection of any Kantianprinciple? It all depends on how one understands the last words in Kant's textquoted above: "Practical principles are formal if they abstract from all subjec-tive ends, whereas they are material if they have put these, and consequentlycertain incentives, at their basis." Have they put a particular incentive at thebasis of the Categorical Imperative? Or have they rather taken the incentivemerely as the occasion to apply the Categorical Imperative?

While there can surely be controversial points in their interpretation ofKant, it seems to me that the basic account of moral motivation they developremains within a Kantian framework, at least as Rawls conceives of it. Thushe writes:

Let's ask how the Cl-procedure exhibits the moral law as sufficient of itselfto determine the will. Here we should be careful not to interpret this featuretoo strongly. I do not think Kant wants to say, and certainly he does notneed to say, that the moral law determines all the relevant aspects of whatwe are to do. Rather, the moral law specifies a scope within which permissi-ble ends must fall, and also limits the means that may be used in their pur-suit, and this goes part way to make the moral law sufficient of itself todetermine the will. Of course, particular desires determine which permissi-ble ends it is rational for us to pursue, and they also determine, within thelimits allowed, how it is rational for us to pursue them. This leeway I view ascompatible with Kant's intentions.

Now, the scope within permissible ends fall is the scope determined bythe universalization procedure: if the maxim passes the test of universaliza-tion, the type of action reflected in the maxim is permitted. Otherwise it is

John Rawls and the new Kantian moral theory 169

8 3

8 4

81

prohibited. Yet the maxim was originally formulated on the basis of a particu-lar incentive for action. In other words: incentives provide us with ends. We donot arrive at the Categorical Imperative without anything to pursue. And yet,according to Herman and Korsgaard, the incentive alone is not the motive foraction. The motive is rather the maxim along with the incentive.

By contrast, the determining ground of a. good action will be its possible uni-versalization—its ability to become a universal law. On the other hand, if onedetermines oneself to act apart or against this possible universalization, oneacts badly. How can this be possible, given that our universal reason can onlybe determined by a universal principle? According to Kant this happensbecause the human being "reverses the moral order of his incentives in incor-porating them into his maxims." In other words, instead of making one'sdesire of a certain good conditional to the fulfillment of the categoricalimperative, one would make the fulfillment of the moral law conditionalto the acquisition of that good. In Rawls's terminology, instead of subordinat-ing the rational to the reasonable, one would be subordinating the reasonableto the rational.

Conclusion

Rawls's approach to Kant may look controversial to those familiar with theconventional reading of Kant. This is not strange, since Rawls himself triedover the years to go beyond that conventional reading—largely based onKant's Groundwork—to discover the ethical relevance of many of his otherwritings, whereby he could make his case for a more plausible Kant.

If, as I have argued through this chapter, many central points of the newKantian moral theory, can be explained in the light of Rawls's interpretationof Kant, many of the objections raised against the new Kantians could beclarified along the same lines: as a revision of Kant's ethical work, intended tomake it look more plausible. To what extent the resulting Kant can still becalled "Kantian" in the old sense remains an open question.

Notes

1. See Robert B. Pippin, "On Allen Wood's Kant's Ethical Thought," Inquiry 43(2000); Robert B. Pippin, "A Mandatory Reading of Kant's Ethics?" Philosophi-

cal Quarterly 51.204 (2001), 386-93; and Donald H. Regan, "The Value of

Rational Nature," Ethics 112 (2002), 267-91.

2. A good account of the usual criticisms addressed to Kantian ethics can be foundin Thomas Hill, Dignity and Practical Reason in Kant's Moral Theory (Ithaca, NY:

Cornell University Press, 1992) (hereafter DPR); see also Barbara Herman, The

170 The Legacy of John Rawls

8 5

8 6

John Rawls and the new Kantian moral theory 171

Practice of Moral Judgment (Cambridge, MA: Harvard University Press, 1993)(hereafter PMJ).

3. C. Korsgaard, Creating the Kingdom of Ends (Cambridge, UK: Cambridge Univer-sity Press, 1996), xii (hereafter CKE).

4. See especially T. Hill, "The Hypothetical Imperative," in DPR, 1—17 andC. Korsgaard's "The Locke Lectures," unpublished.

5. See especially C. Korsgaard, "Kant's Formula of Humanity," in CKE, 106—32.6. PMJ, ix.

7. Thus, referring to the charges of formalism and rigorism, O'Neill wrote: "As Iworked on Kant's writings, I came to believe that neither of these charges can bemade to stick. The Categorical Imperative can guide action and does not lead torigorism." (Onora O'Neill, Acting on Principle: An Essay on Kantian Ethics (NewYork: Columbia University Press, 1975), vii)

8. For a criticism of this aspect, though not from a Kantian perspective, see Regan,"The Value of Rational Nature."

9. PMJ, x.10. SeeLHMP.

10. Thus, Thomas Hill refers to Rawls Lectures at Harvard in 1962 (DPR, 17); Bar-bara Herman refers to his Lectures in 1977 (PMJ, 50).

11. C. Korsgaard, B. Herman, and A. Reath (eds), Reclaiming the History of Ethics:Essays for John Rawls (Cambridge, UK: Cambridge University Press, 1997), 183.

12. Ibid., 1.13. See C. Korsgaard, "Rawls and Kant: On the Primacy of the Practical," in Pro-

ceedings of the Eighth International Kant Congress, Vol. I (Milwaukee: Marquette Uni-versity Press, 1995), 1165-73.

14. See S. Neiman, The Unity of Reason (New York: Oxford University Press, 1994).Neiman argues that Kant transformed the concept of reason, disconnecting itfrom knowledge, to get a pure regulative concept.

15. J. Rawls, "A Kantian Conception of Equality," reprinted in CP, 254-66 (264).16. CP, 264.17. J. Rawls, "Kantian Constructivism in Moral Theory," reprinted in CP, 303—58

(304).18. CP, 305.

19. Ibid., 305-6.20. On the concept of transcendental deduction, see D. Henrich, "Kant's Notion of a

Deduction and the Methodological Background of the First Critique," in E. For-ster (ed.), Kant's Transcendental Deductions: The Three Critiques and the "Opus Postu-

mum" (Stanford: Stanford University Press, 1989), 29—46.21. See R. Pippin, Idealism as Modernism: Hegelian Variations (Cambridge, UK: Cam-

bridge University Press, 1997).22. Rawls, "Kantian Constructivism," in CP, 306—7.23. PMJ, 75.24. Ibid., 77.

25. In suggesting that Rawls's sensibility for the cultural embodiment of reason mayhave some influence on Herman's reading of Kant, I do not deny the existence of a

172 The Legacy of John Raw Is

more basic difference between the late Rawls and Barbara Herman. Thus, in PL,John Rawls claims that the normative basis of our conception ofjustice is based onthe contingent fact that we have a liberal political culture. According to KylaEbbels Duggan, with whom I have contrasted this point, Herman would not liketo go so far. Herman would still claim—in a clear Kantian manner—that thereare principles which apply to all, regardless of any kind of cultural differences.I would like to thank the anonymous referee for asking me to clarify this point.

26. T. Hill, Human Welfare and Moral Worth: Kantian Perspectives (Oxford, UK: Clar-

endon, 2002), 279.27. Thus, Thomas Hill describes his project as follows: "My own project for some

time has been to see how far Kant's basic moral theory, properly understood andmodified as necessary, can be made plausible as at least a candidate for seriousconsideration in contemporary philosophical discussions." (HWMW, 310)

28. SeePAfJ, viii.29. HWMW, 309.30. Ibid.,310.31. PMJ, ix.32. CKE, xi. Generally, in Korsgaard's interpretation it is not entirely clear

whether she ascribes the distinction between phenomenon and noumenon to thedistinction between understanding and reason in general, or, rather, to thedistinction between theoretical and practical reason. While the text just quotedseems to support the first option, at other times Korsgaard seems inclined toascribe the noumenal perspective merely to practical reason. Although thelatter approach is in tune with Kant's asserted primacy of practical reason, itwould make it difficult to defend the unity of reason. (See Neiman, The Unity ofReason, 143)

33. See Kant's argument in Groundwork, Part III, 4:45 Iff34. Phenomenon and thing in itself are untenable notions from the standpoint of

objective knowledge, but in their failure they are indispensable. If they are takenas tangible entities, they lead to a distortion. Two worlds arise, one in the fore-ground, the other in the background. The two are related but each seems tohave a separate existence of its own. The background world becomes a realm ofphantasms, whose contents all stem from our world. But for Kant there is only oneworld. What is touched upon in transcending thought is not another world, butno world at all. And insofar as it exists, it exists in this nonworld. A theory of twoworlds is not Kantian, but only an inevitably contradictory mode of expression.(K.Jaspers, Kant (New York: Harcourt Brace, 1962), 39)

35. Jaspers, Kant, 19.36. "Charity directs us, when interpreting a theory, to prefer readings that make it

more plausible unless textual considerations to the contrary are compelling. Allthe more, this policy makes good sense if our aim is to develop a contemporaryversion of the theory in question." (HWMW, 266)

37. Korsgaard's references to teleology, for instance, invite us to follow this line.38. S. Rosen, The Elusiveness of the Ordinary: Studies in the Possibility of Philosophy (New

Haven: Yale University Press, 2002).

John Rawls and the new Kantian moral theory 173

39. See Rosen, The Elusineness of the Ordinary, 100. In this chapter, Rosen contrasts the

underlying assumptions of Aristotle and Kant regarding everyday life.

40. Reflecting the ordinary sense of our moral terms, I take it, is a prima facie, butby no means decisive, consideration for including a particular conception (e.g.

of conscience) in our moral theory. An entirely revisionist moral theory is un-

likely even to get a hearing, but there are many possible considerations fornot automatically adopting current (or even persistent) "common sense."

For example, it may presuppose what is contrary to (not just beyond) our best

scientific knowledge. (HWMW, 297-8)41. HWMW, 287.

42. I think that it is at least controversial to include Aristotle along with, say, Moore,because the former did have a concept of practical reason, which puts him, in cer-

tain aspects, closer to Kant than to the other intuitionists. Perhaps the division of

the history of ethics between intuitionists and constructivists is too rough to be

entirely fair. It assumes that reason must be either intuitive or merely regulative.Yet for many centuries both aspects were supposed to play a role in the single

faculty of reason. What the ancients understood by nous and dianoia, later called

intellectus and ratio, were two dimensions of the same intellectual power. As two

different dimensions of a single intellectual faculty, ancient and medieval reason

find no proper equivalent in the modern theory of knowledge; as a result, in ethi-cal matters either one favors an intuitionist approach, or a constructivist one. Butthis either/or was not as clear in pre-modern ethics.

43. Rawls, "Kantian Constructivism," 343.

44. Rawls, "Kantian Constructivism," 343—4.

45. Ibid., 345.46. Ibid., 346.

47. J. Rawls, "Themes in Kant's Moral Philosophy," in Forster (ed.), Kant's Trans-

cendental Deductions, 101.48. Ibid.

49. Much of the work of moral judgment takes place prior to any possible appli-cation of rules in the eliciting of the relevant moral facts from particular cir-

cumstances . . . The central difficulty for Kantian theory comes from the iden-

tification of the aspect of character that makes moral judgment possible with a

capacity that involves, or requires for its development, the nonrational facul-

ties. But if we are ever to have a Kantian ethics liberated from its noumenal

baggage, this is just the sort of fact that must be accommodated . . . I believethat the key to getting this right involves rethinking the basic relation between

desire and motive: the way desires are or can be the occasion for motives and

the way rational motives in turn affect the structure of natural history of desire.(B. Herman, "Making Room for Character", in Stephen Engstrom and Jenni-

fer Whiting (eds), Aristotle, Kant, and the Stoics: Rethinking Happiness and Duty

(Cambridge, UK: Cambridge University Press, 1996), 36, 37)

50. See Rawls, "Themes," 82-4.

51. The first limit is that we are to ignore the more particular features of persons,

including ourselves, as well as the specific content of their and our final ends and

174 The Legacy of John Rawls

desires (4:433). The second limit is that when we ask ourselves whether we canwill the perturbed social world associated with our maxim, we are to reason as ifwe do not know which place we may have in the world. (Rawls, "Themes," 86)

52. Rawls, "Themes," 86. Herman has criticized the requirement of publicity.She argues that it introduces "a new locale of moral opacity." And asks: "Whyshould publicity be determinative? What reason do we have for thinking thatin satisfying publicity under universalization a maxim has universal form?"(PMJ, 227) In her view, "the publicity requirement is an expression of a morebasic value claim."

53. "Of course for this idea to work, we require an account for those needs. And herecertain moral conceptions, rooted in our shared moral sensibility, may beinvolved." (Rawls, "Themes," 86)

54. Ibid.55. Ibid., 9956. According to Kant, "this is how even the most common understanding judges"

(CPrR 5: 70). He explains further:

for the law of nature always lies at the basis of its most ordinary judgments, eventhose of experience. Thus it has the law of nature always at hand, only that incases where causality from freedom is to be appraised it makes that law ofnature merely the type of a law of freedom, because without having at hand some-thing which it could make an example in a case of experience, it could not pro-vide use in application for the law of a pure practical reason. (CPrR 5: 70)

57. Rawls, "Themes," 99.58. By a kingdom I understand a systematic union of various rational beings

through common laws. Now since laws determine ends in terms of their univer-sal validity, if we abstract from the personal differences of rational beings aswell as from all the content of their private ends we shall be able to think of awhole of all ends in systematic connection (a whole both of rational beings asends in themselves and of the ends of his own that each may set himself), that is,a kingdom of ends, which is possible in accordance with the above principles.For, all rational beings stand under the law that each of them is to treat himselfand all others never merely as means but always at the same time as ends inthemselves. But from this there arises a systematic union of rational beingsthrough common objective laws, that is, a kingdom, which can be called a king-dom of ends (admittedly only an ideal) because what these laws have as theirpurpose is just the relation of these beings to one another as ends and means.(Kant, Groundwork, Part II, 4:433)

59. Rawls, "Kantian Constructivism," 313.60. As it was pointed out before, the second conception is designed expressly to meet a

need of reason: to have objective content. For this, however, Rawls had to intro-duce the publicity requirement. Barbara Herman has criticized this point. (SeePMJ, 227)

61. This points at "the social world that would come about (at least under reasonablyfavorable conditions) if everyone were to follow the totality of precepts that result

John Rawls and the new Kantian moral theory 175

from the correct application of the Cl-procedure." It is this social world whichdefines Rawls's idea of a "moral conception": "a moral conception is not torevolve around the good as an independent object, but around a conception ofthe right as constructed by our pure practical reason into which any permissiblegood must fit." (Rawls, "Themes," 93)

62. "I shall use the secular term 'realized realm of ends,' and I assume that his com-plete good can be approximated to in the natural world, at least under reasonablyfavorable conditions. In this sense it is a natural good, one that can be approached(although never fully realized within the order of nature)." (Rawls, "Themes,"90—4) Most likely, Andrews Reath's clarification of the double meaning —secular and religious—of Kant's notion of the "highest good," followed by anargument which shows that only the first one is really consistent with Kant's defi-nition of a practical end, could also be in debt to Rawls.

63. In characterizing human persons I have used the phrase "reasonable andrational." The intention here is to mark the fact that Kant uses verniinftig toexpress a full-bodied conception that covers the terms "reasonable" and"rational" asweoftenusethem.. . It is useful to use "reasonable" and "rational"as handy terms to mark the distinction that Kant makes between the twoforms of practical reason, pure and empirical. The first is expressed as animperative in the categorical imperative, the second in the hypotheticalimperative. (Rawls, "Themes," 81-113 (87, 88))

64. "By contrast, rational intuitionism requires but a sparse conception of the person,based on the idea of the person as knower." (Rawls, "Themes," 97)

65. Rawls, "Themes," 100.66. "Recall here Kant's thought that to commit suicide is to root out the existence of

morality from the world." (Metaphysics of Morals 6:422—3).67. See "The Hypothetical Imperative," in DPR, 17.68. See especially her Locke Lectures, II and III.69. "Our humanity is our pure practical reason together with our moral sensibility

(our capacity for moral feeling). These two powers constitute moral personality,and include the power to set ends; they make a good will and moral characterpossible." (Rawls, "Themes," 89)

70. Kant, Religion within the Boundaries of Mere Reason, 6:27—8.71. Korsgaard, "Kant's Formula of Humanity," in CKE, 111.72. Korsgaard, "Kant's Formula of Humanity," in CKE, 114.73. Korsgaard shows the relevance of making this distinction: happiness, for Kant, is

valuable as an end, but it is not intrinsically valuable. A good will is both valuableas an end and intrinsically valuable. And so it is with humanity—even (and herewould lie a possible difficulty) if it is not yet equivalent to personality.

74. Kant, Groundwork, Part II, 4:436.75. Ibid., 4:435.76. Korsgaard, "Kant's Formula of Humanity," in CKE, 120.77. The passage begins already at 4:428, with Kant showing the difference between

relative and necessary ends; then Kant goes on to affirm the principle: "I say thatthe human being and in general every rational being exists as an end in itself, not

176 The Legacy of John Rawls

merely as a means to be used by this or that will at its discretion; instead he must in

all his actions, whether directed to himself or also to other rational beings, always

be regarded at the same time as an end." Then, Kant offers a list of possible ends,distinguishing between those which have relative worth (objects of the inclina-

tions, the inclinations themselves as sources of needs, beings whose existence rests

not on our will but on nature, which he calls things), and finally rational beings or

persons, "because their nature already marks them out as an end in itself, that is,

as something that may not be used merely as a means, and hence so far limits all

choice (and is an object of respect)."78. Korsgaard, "Kant's Formula of Humanity," in CKE, 122—3.79. See G.W.F. Hegel, Elements of the Philosophy of Right, ed. A. Wood, trans. H.B.

Nisbet (Cambridge, UK: Cambridge University Press, 1991).

80. Kant, Groundwork, Part II, 4:428.

81. Kantian motives are neither desires nor causes. An agent's motives reflect his

reasons for acting. An agent may take the presence of a desire to give him areason for action as he may also find reasons in his passions, principles, or prac-

tical interests. All of these, in themselves, are "incentives (Tnebfederri) " not

motives, to action. It is the mark of a rational agent that incentives determine

the will only as they are taken up into an agent's maxim. Indeed, it is only when

an agent has a maxim that we can talk about his motive. (Herman, "On theValue of Acting from the Motive of Duty," inPMJ, 11-12)

82. See Korsgaard, The Locke Lectures.

83. PMJ, 12.

84. Rawls, "Themes," 109.

85. Kant, Religion, 6:36.86. I would like to thank Professor Christine M. Korsgaard for providing me with

some material for this chapter, as well as for a useful conversation on the topic of

this chapter. I am also indebted to Kyla Ebbels Duggan for her feedback on the

similarities between the late Rawls's and Barbara Herman's reading of Kant.Finally, I would like to thank Melissa Moschella and Thorn Brooks and Fabian

Freyenhagen of the Journal of Moral Philosophy for their help with revisions.

In a previous paper, I argued that John Rawls's defense of basic human rightsas a requirement of international justice fails, and that the source of this failureis his reluctance to impose liberal values on societies whose traditions and cul-ture are not liberal. If basic human rights were to be defended, I suggested,we must either apply a comprehensive liberalism to all societies (whetheror not their traditions and culture are liberal) or else restructure the globaloriginal position so that representatives of individuals, not peoples, select theprinciples of justice (a course Rawls rejects as biased toward liberal individu-alism) . Both options, I argued, would require abandoning the effort to find a"political" conception of international justice.

My previous paper was based on the argument Rawls presented in his arti-cle "The Law of Peoples." Since then, however, Rawls has considerablyexpanded and modified his argument in a book by the same title. Althoughthe changes now strengthen his defense of basic human rights (primarily byrequiring greater gender equality in societies whose traditions and culturereject it), I shall argue that they still do not go far enough or else produce ten-sions of their own. The remedy, once again, I shall suggest, is that we mustreject Rawls's reluctance to impose liberal values more widely if we are to jus-tify basic human rights.

If this argument is sound, the upshot would be this. Many (if not most) cri-tiques of Rawls's view seem to fall into one of two camps. On the one hand,there are those who challenge his abstention from liberal individualismand subsequent embrace of the independent moral significance of peoples.On the other hand, there are those who accept (sometimes arguendo) hisembrace of the independent moral significance of peoples, yet maintain that,nonetheless, more recognizably liberal principles can still be justified (particu-larly in the area of global economic distribution). This chapter, however,adds an arguably stronger critique, for it holds not that stronger liberal prin-ciples can be justified even if we accept the independent moral significance ofpeoples, but that weaker principles (ones guaranteeing basic human rights)would fail to be justified.

Before I turn to my defense of these claims, I want to say something aboutwhat is at stake in defending them. At stake is not only our understanding

9

The Law of Peoples: the old and the new

Chris Naticchia

3

6

178 The Legacy of John Rawls

of why Rawls—arguably the most prominent political philosopher of thetwentieth century—decided to alter his final, extended contribution to politi-cal philosophy, or even our estimation of his legacy in the area of internationaljustice (matters that will primarily interest Rawls scholars). More impor-tantly, what is at stake is our understanding of the limits to which wecan avoid appealing to liberal values in order to justify basic human rights.If our best attempts to justify basic human rights ultimately show that appealsto liberal values are unavoidable, then we should regard such appeals as nomore of an imposition than the expectation that all societies must protectbasic human rights. Even more significantly, if such appeals justify liberalfreedoms that go beyond basic human rights, then arguments in support ofbasic human rights would also justify international efforts to advance furtherliberal reforms within nonliberal societies.

To defend these claims, I will begin by rehearsing the main difficulties withthe argument Rawls presented in his earlier article. Then I will assess the mostsignificant changes that appear in the book version of that argument. Afterthis, I will discuss the deeper motivation for the changes and sketch someideas for better handling the concerns that underlie them.

The old Law of Peoples

In his first version of the law of peoples, Rawls argued as follows. A just law ofpeoples is the outcome of two separate international sessions of the originalposition. In the first, only delegates from liberal societies are represented.In the second, delegates from both liberal and nonliberal—what Rawls callshierarchical—societies are represented. As Rawls characterizes them, hierarch-ical societies are nonexpansionist states, usually organized around compre-hensive religious doctrines, that privilege a state religion over others, butprotect dissidents from persecution and respect basic human rights. Thehuman rights that count as basic, according to Rawls, are minimum rights tolife, liberty, personal property, and "a measure" of freedom of conscience andassociation. They do not include liberal-style freedom of speech, democraticpolitical rights, or equal liberty of conscience. In both cases, he argues, theoutcome is the same. From behind a veil of ignorance that excludes knowledgeof the size, population, military strength, natural resources, and economicdevelopment of their society, they select principles that guarantee the freedomand independence of their peoples, allow a right to wage war in self-defense,impose a duty of nonintervention, enable them to ratify and require them toobserve treaties, place limits on the conduct of war, and obligate them to pro-tect basic human rights.

9

The Law of Peoples: the old and the new 179

Let us look more closely at this argument. One thing to notice right awayabout it is this: in the international version of the original position, the veil ofignorance is thinner than it is in the domestic version. In the domestic version,the veil prevents the parties from knowing their particular conceptions of thegood. In the international version, by contrast, the veil allows the delegates toknow their particular conceptions of domestic justice—the analogue, in thiscase, of conceptions of the good. Yet Rawls says nothing to justify the disanal-ogy. This enables one to raise the following objection. If, as Rawls claims inPolitical Liberalism, "the fact that we affirm a particular . .. comprehensivedoctrine is not a reason for us to propose, or to expect others to accept, a con-ception of [domestic] justice that favors those of that persuasion," then simi-larly, the fact that we affirm a particular conception of domestic justice is not areason to propose or to expect others to accept a law of peoples that favors ourview. But if this consideration warrants excluding knowledge of conceptionsof the good in the domestic case, then it should also warrant excluding knowl-edge of conceptions of domestic justice in the international one.

Now how serious this objection is will clearly depend on whether excludingsuch knowledge changes the outcome of the procedure. If the delegates wouldchoose the same principles anyway, then Rawls can simply amend his proce-dure and get the same result. By contrast, if the principles chosen would bevery different, then the objection carries greater force. So we need to deter-mine whether the principles chosen would indeed be different, and if so, towhat extent.

To make matters manageable, consider the duty of nonintervention and theobligation to respect basic human rights—arguably the two most importantprinciples—with the understanding, shared by Rawls, that the latter takesprecedence if the two come into conflict. Thus, if basic human rights arebeing violated, intervention of some type (not necessarily forcible) is justi-fied. For Rawls, this standard represents how high to set the bar before inter-vention becomes morally permissible, and it is with justifying this claim thatwe will be primarily concerned.

Suppose, then, that the delegates no longer know their conceptions ofdomestic justice. Since the delegates are ignorant of the military strength andother resources of their societies, they would plausibly reason that it is farworse to be conquered by, at war with, or intervened in by expansionist socie-ties than it is to hold expansionist tendencies in check. Therefore, by applyingmaximin, they would adopt a duty of nonintervention. Then, charged withadvancing their society's conception of domestic justice, but not knowingwhether they represent liberal or hierarchical societies, they would plausiblyreason that it is worse to be forced to conform to principles incompatible withtheir society's conception of domestic justice—a liberal society forced toadvance some comprehensive religious doctrine, for example, or a religious

180 The Legacy of John Rawls

society forced to embrace liberal freedoms—than it is to tolerate the existenceof the other kind of society. So by applying maximin once again, they wouldagree to mutual toleration rather than risk forced conformity. And since bothliberal and hierarchical societies protect basic human rights, they wouldembrace a principle guaranteeing their protection—but no stronger principlerequiring additional liberal freedoms. Even if we use a thicker veil, therefore,we can still generate principles requiring nonintervention and respect for basichuman rights—principles that permit nonliberal societies to exist and hence(as Rawls puts it) "express liberalism's own principle of toleration of otherreasonable ways of ordering society."

However, this reply exposes a critical assumption that requires justification.The reply assumes that the delegates know that they must represent either lib-eral or hierarchical societies (even if they do not know which type they repre-sent)—both of which respect basic human rights. To see why, suppose theyknew that they could represent what Rawls calls outlaw societies—ones thatare either expansionist or do not respect all of the basic human rights (orboth). In that case, the delegates would still select a duty of noninterventionfor the reason given before (that it is far worse to be conquered by, at war with,or intervened in by expansionist societies than it is to hold expansionist ten-dencies in check). That decision would prohibit expansionist outlaw societies.jVonexpansionist outlaw societies would still be in the running, though. Then,reasoning as before—that it is worse to be forced to conform to principlesincompatible with their society's conception of domestic justice than it is totolerate the existence of the other kinds of societies—they would again agreeto mutual toleration rather than risk forced conformity. But this time, thescope of their toleration would have to be wider: it would have to includesome nonexpansionist outlaw societies, that is, ones that do not respect all ofthe basic human rights. So, in this case, the delegates would select a principleslightly weaker than one requiring the protection of all basic human rights—itwould require protecting some but not all of them—or to put it another way,they would have a slightly narrower interpretation of what the basic humanrights are.

If we use a thicker veil, then, we can generate a principle requiring respectfor the full array of basic human rights only if we exclude outlaw societies fromthe original position (and enable the delegates to know that they areexcluded). But what justifies their exclusion? For Rawls, each aspect of theoriginal position must be given supporting grounds. What supporting groundscan be given for this aspect? Rawls claims that outlaw societies are properlyexcluded because they fail to meet what he describes as necessary conditionsfor being a member in good standing in a reasonable law of peoples: they mustbe nonexpansionist, respect (the full array of) basic human rights, and beguided by a common good conception of justice.

The Law of Peoples: the old and the new 181

Now immediately we should see that if part of the price of admission, so tospeak, to the original position is being nonexpansionist and respecting basichuman rights, then it is hardly surprising that the delegates select principlesrequiring nonintervention and respect for basic human rights (which theyalready observe). But then their emergence from the original position wouldnot seem to give them any additional moral force. It would be like requiringthe parties in Rawls's domestic original position to know that they eachembraced his first principle of justice. In that case, if they selected that princi-ple, its emergence from the original position would hardly constitute a justifi-cation of it.

Fortunately, Rawls does not need to rely on either of these conditions, for tworeasons. First, as we have just seen, the delegates would still choose a principlerequiring nonintervention even if they knew that they might represent expan-sionist societies. Second, since Rawls considers respect for basic human rights tobe a consequence of the (third) condition that a society must be guided by acommon good conception of justice, he could simply substitute this conditionfor the condition requiring respect for basic human rights and still generate aprinciple requiring such respect. So we can amend Rawls's procedure by drop-ping the first two conditions and relying solely on the third: the price of admis-sion to the original position is observing a common good conception of justice.

But notice that this move merely pushes the issue back. What supportinggrounds can be given for making a common good conception of justice theprice of admission? A common good conception of justice, according toRawls, is one that "takes impartially into account what it sees not unreason-ably as the fundamental interests of all members of society." He then sug-gests that it is because common good conceptions of justice are "notunreasonable" that the original position should include delegates who couldpotentially represent them (but not other views). But why are such concep-tions not unreasonable? Were he to use the term "reasonable" as he does inPolitical Liberalism—as involving the exercise of theoretical and practicalreason and drawing upon a tradition of thought and doctrine—then manycommon good conceptions would qualify as reasonable, since many of themdraw upon doctrines involving theoretical and practical reason. But then sotoo would some conceptions that deny certain basic human rights, for theexact same reason. This probably explains why Rawls resorts to a differentuse of the term in the law of peoples, where by "reasonable" he means anyconception of domestic justice that admits "a measure of liberty of conscienceand freedom of thought, even if these freedoms are not in general equal forall members of society as they are in liberal regimes," and even if they permit"an established religion with certain privileges," provided that no religionsare persecuted. (I will say more about this issue later, under the heading"Deeper motivation.")

21

182 The Legacy of John Rawls

Our reconstruction of Rawls's argument, then, is this. If we use the thickerveil, we can generate a principle requiring respect for the full array of basichuman rights only if we exclude outlaw societies from the original position.What justifies the exclusion of outlaw societies is that they do not possess acommon good conception of justice. The ground for excluding societies thatdo not possess a common good conception of justice is that their conceptionsof justice are unreasonable. And the reason that they are unreasonable is thatthey do not recognize liberty of conscience and freedom of thought.

We now come to the crux of the matter. What can Rawls say to someonewho objects that this sets the price of admission to the original position toolow—that conceptions of domesticjustice that admit only "a measure" of reli-gious freedom, and allow it to be unequal (perhaps quite unequal) amongmembers of society are not reasonable? What resources, in other words, areavailable from within Rawls's theory to respond to such a challenge?

Notice that one cannot respond by claiming that to exclude any more socie-ties would be intolerant, since the issue is precisely where the limits of tolera-tion are to be drawn. Besides, some societies are already excluded. Why notmore? Nor can one respond by claiming that to exclude any more societieswould be inconsistent with liberalism's principle of toleration for other waysof organizing society, since it's not necessarily the case that only liberal oneswould remain. For example, suppose that a hierarchical society with an estab-lished religion and widely unequal religious freedoms nonetheless allows"a measure" of religious freedom (a very meager one) for members of religiousminorities: permitting them to pray in groups of no more than three. As longas those members did not invite sanctions with massive public disobedience,and as long as the government did not adopt intrusive measures to detectprivate violations, enforcing these policies would not rise to the level of perse-cution (and would not violate their other basic human rights as Rawls char-acterizes them). Yet if we consider a conception of domestic justice thatpermits hierarchical societies to be organized like this to be unreasonable thenwe have grounds for excluding at least some hierarchical societies from the ori-ginal position. Gall these societies extreme hierarchical societies. Still, hierarch-ical societies with less significant inequalities—call them moderate hierarchicalsocieties—would not be excluded. So other societies besides liberal ones wouldstill be included if we drew the limits of toleration tighter.

The point, however, is not merely that Rawls's theory seems to lack theresources to respond to the challenge that he sets the price of admission tothe original position too low. The challenge itself seems independently plausi-ble. Moreover, when we consider the historic connections between religionand gender discrimination, it gains further strength. In many religion-basedsocieties, for instance (not just poor ones), girls receive far less education thando boys. The illiteracy rates for women are often twice that of men. Girls are

The Law of Peoples: the old and the new 183

subject to early, arranged marriages that further limits their opportunities.Divorce is far more difficult for women to obtain than it is for their husbands.The cumulative impact of these inequalities profoundly influences the controlthat women have over the course of their lives and severely limits their abilityto exit the religious culture into which they are born. Given that hierarchicalsocieties may have a state religion, and given that these restrictions on girlsand women are often based on religion, it follows that such religion-basedgender inequalities are also permitted by hierarchical societies (even moder-ate ones). Yet if we consider a conception of domestic justice that permits hier-archical societies to be organized like this to be unreasonable, then we havegrounds for excluding more of them.

As these remarks suggest, we should reject as too lax the view that a concep-tion of domestic justice is not unreasonable if it allows only "a measure" ofliberty of conscience and freedom of thought. We should raise the bar. Andindeed, Rawls does precisely that in his new law of peoples. But as I shall alsoargue, Rawls himself does not raise it high enough. We should raise it higher.

The new Law of Peoples

Of the most significant changes to Rawls's argument, one is merely cosmetic:hierarchical societies are now called decent (or decent hierarchical) ones. Butothers are substantive. For example, whereas Rawls, in the article version,held that a society must be nonexpansionist, respect basic human rights, andbe guided by a common good conception of justice in order to be representedin the original position, he drops the condition requiring respect for basichuman rights in the book version. This change saves Rawls from the chargethat he makes respect for basic human rights the price of admission to an ori-ginal position from which he hopes to derive a principle mandating theirrespect—of presupposing exactly what the original position is supposed to jus-tify. But since the book version still insists that a society must be nonexpansion-ist—since it makes this the price of admission to an original position fromwhich Rawls hopes to derive a principle of non-intervention—he still seemsto presuppose at least part of what the original position is supposed to justify.Nonetheless, we can save Rawls from the charge of presupposing what the ori-ginal position is supposed to justify if we simply drop the condition that asociety must be nonexpansionist. Since the delegates are ignorant of the mili-tary strength and other resources of their societies, they would plausiblyreason that it is far worse to be conquered by, at war with, or intervened inby expansionist societies than it is to hold expansionist tendencies in check.So they would still choose a principle of nonintervention.

184 The Legacy of John Rawls

Even if we amend Rawls's view in this manner, though, the other require-ment would remain: the requirement that a society must be guided by acommon good conception of justice in order to be represented in the originalposition. Yet we may still justifiably ask whether common good conceptions ofjustice really are (as Rawls puts it) "not unreasonable." But now, in the bookversion of the law of peoples, Rawls strengthens what is required for a concep-tion of justice to be not unreasonable—presumably, to block precisely theobjections that we raised in the last section. For example, whereas in the arti-cle Rawls held that a conception must admit "a measure of liberty of conscienceand freedom of thought, even if these freedoms are not in general equal for allmembers of society as they are in liberal regimes" —a characterization com-patible with widely unequal religious freedoms—in the book he holds that aconception must admit "a sufficient measure of liberty of conscience and free-dom of religion and thought," even if they are unequal. Sufficient, we mightask, for what? Elsewhere in the book he phrases it even more strongly, requir-ing that a conception must admit "a sufficient measure of liberty of conscienceto ensure freedom of religion and thought." And when he describes a societythat has such a conception of justice—the fictitious Kazanistan—he expresslylimits religious inequalities to the holding of higher political and judicialoffices: "Islam is the favored religion, and only Muslims can hold the upperpositions of political authority and influence the government's main decisionsand policies.. . Yet other religions are tolerated and may be practiced withoutfear or loss of most civic rights, except the right to hold the higher political orjudicial offices." Indeed, Rawls insists that "they are not subjected to arbi-trary discrimination, or treated as inferior by Muslims in public or social rela-tions," which suggests that they are otherwise treated equally.

These remarks suggest two different requirements for a conception ofjusticeto be not unreasonable, one that limits religious inequalities to the higher poli-tical and judicial offices (if we take the Kazanistan discussion to define thatrequirement), and one that allows wider inequalities (if we take it merely toillustrate it)—though not as wide as those permitted in the article. SinceRawls introduces Kazanistan as an example of a society whose conception ofjustice is not unreasonable, I will interpret his remarks as permitting widerinequalities, although I will also consider the alternative interpretation atthe end of this section.

There is another respect in which Rawls strengthens what's required for aconception ofjustice to be not unreasonable. As we saw in the last section, it iscompatible with Rawls's earlier characterization of a hierarchical society forit to permit significant inequalities in rights between men and women. Butnow, in the book, Rawls limits the permissible range of gender inequalities,just as he does religious ones. There must be "equal justice for women," heclaims, for two reasons. First, the inequality and subjection of women are

31

The Law of Peoples: the old and the new 185

among the leading causes of population pressures that burden a society rela-tive to what its economy can sustain. Thus, removing the inequality and sub-jection of women and giving them equal justice is a good strategic reason forenabling a society to meet its members' subsistence needs, as required by thebasic human right to life. (Rawls cites the Indian state of Kerala, which gavewomen rights to receive education, to own property, and to hold wealth, asone of several places where this strategy significantly reduced birth rates. )Second, and more directly connected with the interests of women, the subjec-tion of women simply violates their basic human rights.

Now I do not think that either of these reasons adequately supports Rawls'sview that there must be equal justice for women. That giving women equaljustice is a good strategy for reducing birth rates, which is in turn a good strat-egy for enabling a society to meet its members' subsistence needs, is an empiri-cally contingent reason for guaranteeing basic rights—the sort Rawls elsewhererejects as too insecure. At best, it supports the position that equal justice forwomen is required in burdened societies that are unable to meet their mem-bers' subsistence needs, but not elsewhere. The second reason, that the subjec-tion of women violates their basic human rights, argues for upholding theirbasic human rights but no more. It does not argue for any right of nondiscrimi-nation—a right conspicuously missing from Rawls's list of basic humanrights—as the phrase "equal justice for women" seems to require. PerhapsRawls means to suggest, through his example of the Indian state of Kerala,that women must have some rights to education, wealth, and property,though not rights equal to men. (I am assuming that there can be some genderinequalities that intuitively seem to constitute discrimination, but do not riseto the level of subjection.) But this would be a nonstandard use of the phrase"equal justice for women." Rather than attribute this nonstandard usage toRawls, I will interpret his claim that there must be equal justice for women asrequiring a right to nondiscrimination on the basis of gender in spite of theapparent inability of his reasons to support that view.

On this interpretation of the new law of peoples, then, Rawls is claimingthat conceptions of justice are not unreasonable if and only if they allow a suf-ficient measure of religious freedom to all of their members, even if these free-doms are unequal, and they guarantee equal justice for women. This is the newprice of admission to the original position. I would like to raise three questionsabout this claim.

First, why does Rawls seem to permit religious discrimination while barringgender discrimination? By allowing unequal—though not too unequal—reli-gious freedoms, Rawls seems to permit some form of what we would intuitivelyconsider to be religious discrimination. In contrast, by insisting on equal jus-tice for women, Rawls seems to require a right to nondiscrimination on thebasis of gender. What justifies the difference?

186 The Legacy of John Rawls

The difference seems to involve an inconsistency in principle, the sort thatRonald Dworkin has called "checkerboard" law. This is law that makesarbitrary distinctions in matters of principle—for example, a law makingabortion illegal for women born in even years but not odd ones. By contrast,a law making abortion illegal except in cases of rape, incest, or when the life orhealth of the mother is at stake, is not. Whether we think that the exceptionsshould be broader, or whether we think there should be none at all, we at leastrecognize that the distinction drawn is a principled one. But in Rawls's case,if official discrimination is a wrong that should be prohibited as a matter ofjustice, then it seems arbitrary to exempt religious discrimination from thescope of that principle. If so, then any conception of justice that permits suchdiscrimination as a matter of law would seem to allow checkerboard law. Butintuitively, this seems wrong. Instead, we should treat religious and genderdiscrimination the same.

Now one might want to reply on Rawls's behalf as follows. For Rawls, poli-tical philosophy should be practical: it should produce an outcome that can besupported by adherents of diverse comprehensive worldviews, whether theyare views of the good, or of justice. Thus, Rawls may be perfectly justifiedin allowing religion—his paradigm case of a comprehensive worldview—tobe used as a basis for making political distinctions, but barring gender frombeing used.

But notice what this reply assumes: it assumes that in trying to satisfy thispractical aim, in trying to gain the support of diverse comprehensive world-views, it is nonetheless permissible to impinge upon these views to someextent. Gender discrimination, even if based on religion, is impermissible.(Presumably, it would be easier to achieve the practical aim if no impinge-ments at all were permitted.) So the practical aim is not overriding. But oncewe accept some compromise in the practical aim, other impingements seemequally motivated. For instance, we might claim that racial and ethnic discri-mination, like gender discrimination, should be impermissible, even if they arebased on religion. And now religious discrimination, itself based on religion,begins to look like an isolated case. Why does the practical aim allow only thiskind of discrimination, but not others? We seem to face the problem of check-erboard law all over again, even after taking Rawls's interest in the practicalaim of political philosophy into account.

Even if one remains unpersuaded by these considerations, however, there isa more fundamental problem with the reply: it ignores what Rawls refers to asthe two stages of his presentation. Rawls speaks of his domestic principles,justice as fairness, as being presented in two stages. The first stage involvesthe argument from the original position, where parties choose principlesfrom behind a veil of ignorance without investigating whether they can bethe focus of an overlapping consensus. The second stage involves testing the

The Law of Peoples: the old and the new 187

principles to see if they can be the focus of such a consensus. Presumably, thesame sequence is to be followed for his international principles. But at thispoint we are still trying to determine "the most philosophically favored inter-pretation" of the international original position—we are still working out thefirst stage. After that, we can consider the likelihood of the outcome becomingthe focus of an overlapping consensus. But such (practical) considerations arenot to influence the setup of the original position yet.

Second, suppose that we do treat religious and gender discrimination thesame, not by barring them both but by allowing both kinds of inequality(although everyone has "a sufficient measure" of religious freedom, andwomen have some rights to education, wealth, and property). This avoidsthe problem of allowing checkerboard law. But it faces the original problem.What can Rawls say to someone who objects that this still sets the price ofadmission to the original position too low—that conceptions of domestic jus-tice that permit discrimination on the basis of gender or religion but restrict itfrom being too excessive are not reasonable?

It is tempting here to appeal to Rawls's notion of a consultation hierarchy.According to Rawls, a common good conception of justice requires eitherdemocratic political rights, as in liberal societies, or a consultation hierarchy,in nonliberal societies. A consultation hierarchy includes "a family of represen-tative bodies whose role in the hierarchy is to take part in an established proce-dure of consultation and to look after what the people's common good idea ofjustice regards as the important interests of all members of the people." In aconsultation hierarchy, members of society are not regarded as "separate indi-viduals deserving equal representation. . . according to the maxim: one citizen,one vote," as they would in liberal societies. Rather, "persons belong firstto estates, corporations, and associations—that is, groups." Thus, "[e]achperson belongs to a group represented by a body in the consultation hierar-chy." And "as members of associations, corporations, and estates," persons"have the right at some point in the procedure of consultation ... to expresspolitical dissent, and the government has an obligation to take a group's dis-sent seriously and to give a conscientious reply." Hence, in response to thecharge that he sets the price of admission to the original position too low,Rawls might reply that any inequalities on the basis of gender or religion thatissue from a consultation hierarchy are reasonable since the interests of womenand religious minorities were considered in that process.

This would be an effective reply, perhaps, if we were to assume that womenand religious minorities accepted discrimination willingly and reflectively —as opposed to, say, simply acquiescing in it, or accepting it on the basis of adap-tive preferences. But unless Rawls is going to allow acquiescence or adaptivepreferences a role in determining what counts as reasonable inequalities—andgive up the attempt to find an Archimedean point for assessing them—his

39

188 The Legacy of John Rawls

reply must rest on their being accepted in some more robust sense. And then itis hard to see why women and religious minorities would not insist that discri-mination against them neglects their fundamental interests—interests whichthe consultation process is supposed to uphold.

Third, suppose that we treat religion and gender the same, not by "levelingdown" the status of women to that of religion, as we just did, but by "level-ing up" the status of religion to that of women. On this view, conceptions ofjustice are not unreasonable if and only if they guarantee equal justice forwomen and for minority religious groups by requiring a right to nondiscrimi-nation on the basis of gender or religion. But there is one exception: the higherpolitical and judicial offices may still be reserved for members of the favoredreligion. Now in this case, where inequalities in rights and freedoms are indeedvery limited, we may be inclined to agree that such conceptions are not unrea-sonable, and that therefore this is appropriate as the price of admission to theoriginal position. But it is important to notice that this brings them quite closeto the family of liberal conceptions of justice—the main differences being thatthe nonliberal conceptions require rights of political participation that arecollective rather than individual, and that they permit some religious restric-tions on the higher political and judicial offices. Since these nonliberal con-ceptions are almost liberal, we might ask whether there is some other way ofaccounting for the intuition that they are not unreasonable. I will say moreabout this in a moment. What I want to do now is look at what I think is thedeeper motivation for the changes we have discussed.

Deeper motivation

I think that ultimately one source of the problem—that is, the problem ofdeciding which conceptions of justice are not unreasonable with respect tothe limits they impose on a society's domestic practices—is this. When anagent in the original position represents groups of individuals rather than indi-viduals, the resulting principles will be silent on what goes on within eachgroup, since the principles the agent must agree to are supposed to govern therelations of the groups with one another, not their relations with their ownmembers. (Feminists were first to notice this in their critique of Rawls'sinitial claim that the parties in the domestic original position were heads offamilies, a claim which they argued made the internal affairs of familiesopaque to claims of justice. ) But in his international original position,Rawls refuses to have the delegates represent individuals; instead, they mustrepresent peoples. As a result, the delegates' choice of principles under uncer-tainty does not impose any new limits on a society's domestic practices that arnot already determined by the price of admission. What this means is that the

The Law of Peoples: the old and the new 189

price of admission to the international original position is what establishesthose limits—it (not the choice of principles under conditions of uncertainty)does the normative work. And so, when we determine what the price of admis-sion is, we are determining what those limits ought to be. If intuitions differ,though, about whether the limits determined by the price of admission are tooweak, then the theory cannot produce "a resolution which we can affirm onreflection," as Rawls claimed his domestic theory could do with differingintuitions over the proper distribution of wealth and income. We are leftwith a mere clash of intuitions.

Given this, and given the intuitive implausibility of allowing inequalities inrights and freedoms to be too wide, it is only natural that Rawls sought tostrengthen his view of what makes a conception of domestic justice not unrea-sonable (since what counts as being "not unreasonable" just is the price ofadmission). But as I have suggested, the most plausible way of interpretingthat view is as requiring such conceptions to be almost liberal. This is aboutas strong as Rawls can make his view and still be consistent with his claimthat liberalism requires the toleration of other ways of organizing society.So we might describe Rawls's deeper motivation for making this change asinvolving, on the one hand, the need to strengthen his view of what makes aconception of justice not unreasonable, and on the other hand, the need tohonor what he conceives of as liberalism's principle of toleration.

But there is another way of accommodating this result. We could claim thatonly liberal conceptions of domestic justice are not unreasonable, but for prac-tical rather than moral reasons concerning tolerance, we should exempt socie-ties that follow nonliberal, but almost liberal, conceptions from even themildest forms of intervention (like political and economic pressure) aimed atreforming them in a liberal direction. Instead, societies that follow liberal con-ceptions should simply lead by example.

This is not Rawls's view, of course, since for him the exemption is grounded inmoral reasons concerning tolerance. Indeed, it is one he officially rejects. Yetthere are times when he seems sympathetic to it. "All societies undergo gradualchanges," he claims, "and this is no less true of decent societies than of others.Liberal peoples should not suppose that decent societies are unable to reformthemselves in their own way. By recognizing these societies as bonaf.de membersof the Society of Peoples, liberal peoples encourage this change." Thisemphasizes a practical rationale for exempting decent societies from any out-side pressures to liberalize in favor of letting them reform themselves. Second,he claims that "in view of the possible inequality of religious freedom, if for noother reason, it is essential that a hierarchical society allow and provide assis-tance for the right of emigration." The suggestion here seems to be that mem-bers of a religious minority should have an effective right of exit so that, if theywish to, they may seek entry into a society that guarantees equal rights of

190 The Legacy of John Rawls

religious freedom, as in a liberal society. If so, then it almost seems as thoughRawls thinks they ought to have, as a fallback position, the freedom to enter aliberal society—which comes very close to suggesting that all persons every-where are entitled to liberal rights and freedoms somewhere.

The upshot is that we have two ways of accounting for the intuition thatsocieties that are almost liberal should be exempt from even milder forms ofpolitical and economic pressure aimed at reforming them in a liberal direc-tion. One way is grounded in moral reasons concerning tolerance; the otheris grounded in practical concerns about the efficacy of exerting such pressureon societies that are already almost liberal. Although I shall not argue thepoint here, the advantage, it seems to me, belongs with the latter view, sinceRawls's version of the former appears to rest on nothing more than an intui-tion: an intuition about where the limits of tolerance are to be drawn. By con-trast, the latter provides an explanation of those limits (a deflationary one). Buteven if we favor the former, reconstructed Rawlsian view, it is important tonotice this: we are still appealing to liberal values of equal justice and non-discrimination, and committing ourselves to advancing them in cultures accustomed to privileges for men and certain religions, even if we accept exceptionsfor the higher political and judicial offices. This requires liberal reforms that gobeyond securing basic human rights as Rawls characterizes them. Hence, notonly does this reconstructed argument support basic human rights. It also jus-tifies international efforts to promote equal justice for women and religions.

Wider implications

These findings imply that international efforts to encourage societies to liber-alize are in principle justifiable as long as those societies are not already almostliberal. Of course, the specific form such efforts take will need to be justified.From the fact that a society violates some requirement of justice, it does notfollow that we may respond in any way we like. The response itself must beproportionate, just, and rationally designed to achieve its end. We can there-fore imagine a range of acceptable responses depending on the severity of theviolations—from military intervention with appropriate authorization inthe worst cases, to political and economic pressure, to diplomatic and eco-nomic incentives.

Rawls's theory itself has little to say about which of these efforts would bejustified in what circumstances, since it is, by design, an ideal theory—a theoryof the state of affairs we should aim to achieve—rather than nonideal theory—a theory which indicates the steps we may take in transitioning to the ideal.Nonetheless, if the argument of this chapter is sound, the basic point, con-cerning the ideal that he defends, remains: even if we accept Rawls's

The Law of Peoples: the old and the new 191

procedure, we should reject his reluctance to impose liberal values morewidely, and to insist instead that equal justice for both women and religionsdeserves to be among our international moral imperatives.

Notes

1. C. Naticchia, "Human Rights, Liberalism, and Rawls's Law of Peoples," SocialTheory and Practice 24.3 (Fall 1998), 345-74.

2. By "we," I mean "you and I, the reader and author, or agents acting on ourbehalf, whether international or national."

3. Another alternative that we might apply is not a comprehensive liberalism, but areligious comprehensive doctrine, interpreted so as to support basic humanrights. This too would require abandoning the effort to find a political conceptionof international justice. My assumption throughout, though, which I do notattempt to defend here, is that the best defense of basic human rights is broadlyliberal. Obviously it is of the first importance should this assumption prove mis-taken.

4. John Rawls, "The Law of Peoples," in CP, 529-64.5. LP.6. See, for instance, Charles Beitz, "Rawls's Law of Peoples," Ethics 110.4 (July

2000), 669—96; Simon Caney, "Cosmopolitanism and the Law of Peoples," Jour-nal of Political Philosophy 10.1 (2002), 95-123; and Thomas Pogge, "Rawls onInternational Justice," Philosophical Quarterly 51.203 (April 2001), 246-53.

7. See, for instance, Allen Buchanan, "Rawls's Law of Peoples: Rules for a VanishedWestphalian World," Ethics 110.4 (July 2000), 697-721; and Pogge, "Rawls onInternational Justice."

8. In calling The Law of Peoples "his final, extended contribution to political philoso-phy," I am of course excluding two later works by Rawls, his LUMP and his JF,on the grounds that the first is in the history of moral philosophy and the second isa reworking of his previous ideas in TJ and PL.

9. Here I intend merely to assert that there are limits to which we, in attempting toprovide a philosophical justification for basic human rights, can avoid appealingto liberal values. This is compatible with states themselves professing officialagnosticism about the philosophical basis for such rights.

10. As I mention above in note 3, I do not attempt here to defend the claim thatbroadly liberal defenses are best, only that, if they are, we cannot bracket themas tightly as Rawls seeks.

11. A more fully developed version of the argument contained in this section appearsin my "Human Rights, Liberalism, and Rawls's Law of Peoples."

12. Nor do they include rights to nondiscrimination on the basis of religion or gender,which are also typically regarded as requirements of liberalism. The importanceof this fact will be seen in section II.

13. Rawls, "The Law of Peoples," 538-40.

192 The Legacy of John Rawls

14. See Thomas Pogge, "An Egalitarian Law of Peoples," Philosophy and Public Affairs23.3 (Summer 1994), 195-224 (206).

15. PL, 24.16. I take it that the principle enabling them to ratify and requiring them to observe

mutually advantageous treaties follows from the principle guaranteeing theirfreedom and independence as a people, which I take to be the correlative of theduty of nonintervention. I also take it that the principles allowing them the rightto wage war in self-defense and to place limits on the conduct of wars follow fromthe principle imposing a duty of nonintervention. This suggests that the duty ofnonintervention is the basis for the rest of these principles, which are derivative(with the exception of the one guaranteeing the freedom and independence oftheir people, which is correlative). The only remaining principle to account foris the obligation to protect basic human rights. Hence the claim that the duty ofnonintervention and the obligation to protect basic human rights are the twomost important principles.

17. As I explain later in the text, we may think of intervention to effect change asinvolving different types of policies that range over a continuum, from forcible(military) intervention on one end of the spectrum, to diplomatic and economicincentives at the other, with various forms (and degrees) of political and economicpressure in between.

18. Rawls, "The Law of Peoples," 530.19. It would be mistaken to suppose that such societies must be devoted to a senseless

reign of domestic terror. It could be just like a hierarchical society except that itscomprehensive religious doctrine denies certain human rights. For example, itmight deny a guarantee of minimum economic security (a subsistence right takenby Rawls to be part of the minimum right to life); instead of being a claim-right(as the term "guarantee" suggests), it might merely be a liberty-right. Alterna-tively, it might deny one aspect of the right to freedom by requiring forced laborfrom members of a certain ethnicity (but only to an extent consistent with socialstability). For more on this point, see my "Recognition and Legitimacy: A Replyto Buchanan," Philosophy andPublic Affairs'28.3 (Summer 1999), 249 and "HumanRights, Liberalism, and Rawls's Law of Peoples," 351. For Rawls's view of mini-mum economic security, see Rawls, "The Law of Peoples," 546.

20. Rawls, "The Law of Peoples," 545-7.21. Ibid. ,545.22. Ibid., 547.23. PL, 59. Elsewhere, Leif Wenar suggests that Rawls sometimes seems to define

"reasonableness," or "reasonable doctrine," differently: as what reasonable per-sons affirm. That of course shifts the burden to defining "reasonable person." AsWenar suggests, though, Rawls's characterization of this notion seems to presup-pose a partially comprehensive Kantianism, or at least a conception of indivi-duals' highest-order interests, which would not be supported by many doctrinesin an overlapping consensus. (See Leif Wenar, "Political Liberalism: An InternalCritique," Ethics 106.1 (October 1995), 32-62.) Since (in this context) concep-tions of domestic justice are the analogue of such doctrines, I take it that Rawls

The Law of Peoples: the old and the new 193

cannot consistently resort to this usage of the term either. This would thus provide

additional reason for his resorting to a different use entirely (as explained later inthe paragraph).

24. Rawls, "The Law of Peoples," 547.25. Susan Moller Okin, " 'Mistresses of Their Own Destiny': Group Rights, Gender,

and Realistic Rights of Exit," Ethics 112.2 (January 2002), 205-30 (216-22).26. LP, 64.

27. Rawls, "The Law of Peoples," 547 (emphasis added).

28. Ibid., 74 (emphasis added).29. Ibid., 64 (emphasis added).

30. Ibid., 75-6.

31. Ibid.,76.

32. Ibid., 9, 110.

33. Ibid., 110.

3 4 . Ibid., I I I .35. The worry is similar to Rawls's worry in A Theory of Justice that utilitarian-

ism provides an unacceptably contingent basis for securing the basic rights and

liberties.

36. Ronald Dworkin, Law's Empire (Cambridge, MA: Harvard University Press,1986), 178-84.

37. Ibid., 178.

38. PL, 140 n7. This observation was also made (in a separate context) in my

"Human Rights, Liberalism, and Rawls's Law of Peoples," 367.

39. "So while a political conception ofjustice addresses the fact of reasonable plural-

ism," claims Rawls, "it is not political in the wrong way: that is, its form andcontent are not affected by the existing balance of political power between com-

prehensive doctrines." (See PL, 142.) Although Rawls's remark addresses his

domestic theory ofjustice, it presumably applies to his international theory as well.

40. IP, 71.41. Ibid., 71.

42. Ibid., 72.43. Ibid., 71-2.

44. Ibid., 72.

45. "Human Rights, Liberalism, and Rawls's Law of Peoples," 357.

46. Since then, Rawls has indicated that the parties to his domestic original position

are individuals rather than family heads, which should accommodate this con-cern. For other views, see Susan Moller Okin, "Political Liberalism, Justice, and

Gender," Ethics 105.1 (October 1994), 23-43 (29-32, 35-7); and John Exdell,"Feminism, Fundamentalism and Liberal Legitimacy," Canadian Journal ojPhilo-

sophy 24.3 (September 1994), 441-63.

47. TJ, 19.

48. IP, 61.49. Ibid., 74.

50. \ say "almost" because having the freedom—or a liberty right—to enter a liberal

society does not strictly entail having an entitlement—or claim right—to such

194 The Legacy of John Rawls

entry. But since the liberty would be of no value without the claim, one mightplausibly interpret Rawls (or, alternatively, the best Rawlsian view) as beingcommitted to the latter.

51. Rawls claims that the

long-run aim is to bring all societies eventually to honor the Law of Peoples andto become full members in good standing of the society of well-ordered peoples.Human rights would thus be secured everywhere. How to bring all societies tothis goal is a question of foreign policy; it calls for political wisdom, and successdepends in part on luck. These are not matters to which political philosophyhas much to add .. . What to do on these questions is ... essentially a matter ofpolitical judgment and depends upon a political assessment of the likely conse-quences of various policies. (See LP, 93.)

To understand the continuing importance of John Rawls's work, we need tounderstand the background, the object, and the method of his fifty-year questas a political thinker.

The background to Rawls's investigation was a (carefully circumscribed)acknowledgment of a certain kind of evaluative pluralism. In particular,as early as A Theory of Justice, we find Rawls casting his activities in relation tothe fact, as he saw it, that there is no public conception of the good. Later, inPolitical Liberalism, Rawls wrote of a reasonable pluralism of comprehensivedoctrines. Although these formulations are not quite the same, Rawls in bothcases acknowledged the fact and the importance of a diversity in moral judge-ments that is likely, absent coercion, to persist even in the face of strenuousefforts to address it.

The object of Rawls's work was to develop a method of commensuration thatwould enable us, the free and equal citizens of a democratic society, to identifya common basis for our dealings, in search of mutual benefit, with one another.Because of diversity, indeed reasonable plurality, of beliefs and values, therewill be diversity, too, in the numerous, incompatible proposals for the basicstructure of society or, as Rawls put it later, for the constitutional fundamen-tals that define the basic terms of our association. How are we, collectively, torank these proposals and, in particular, to identify, as best for us, a single,determinate proposal that could serve as the commonly agreed and self-sus-taining basis for our association? To answer this question, to identify a way ofcommensurating these proposals was, I submit, the object of Rawls's work.His twofold critique, in A Theory of Justice, of utilitarianism (which com-mensurates in a Procrustean manner) and of intuitionism (which embracesincommensurability) was, perhaps, the surest indication of Rawls's preoccu-pation with this vital issue.

Finally, the method used by Rawls in his work was broadly, though tacitlyrather than explicitly, pragmatist in character, especially in the sense thatRawls was aiming, now explicitly, to develop principles of political association

10

The legacies of John Rawls

Fred D'Agostino

Rawls's threefold legacy

196 The Legacy of John Rawls

that generate their own support in the attitudes and actions of those whosebehavior they are intended to guide. Rawls was aiming, crudely, to developprinciples of justice which can function, socially, as tools for the arrangement,critique, and rearrangement of existing social institutions. And he understood,profoundly, that, in order to play such a role, these kinds of principles wouldhave to secure uptake; articulating such principles would be merely an aca-demic exercise unless they were embraced.

In the sections which follow, I will comment on each of these aspectsof Rawls's project, with the aim of displaying some of the main features ofhis enterprise that remain, in fact, to be fully developed and whose develop-ment will, I submit, constitute Rawls's (at least medium-term) legacies topolitical thought.

Rawls's pluralism

Any discussion of pluralism should begin, I think, with recognition of the factof diversity. People have different opinions, indeed form different deliberatedjudgements about what is good and even about what is right, and these differ-ences in attitudes are likely to survive, if not entirely unchanged, all sorts ofattempts to bring the parties involved to agreement. This is diversity and it isan all-but-inescapable feature of our daily lives. Indeed, many of the mostfamiliar institutions of our societies and cultures can be thought of as respon-sive or any way responsible to this fact of diversity. (Foremost among them, aswe will see in the next section, is a system that allocates individual rights ofjudgement and decision.)

In moderately realistic situations, there is disagreement even about whatthe facts are. Consider, for instance, the disputes, likely to be persistent,about how to interpret forensic evidence during a criminal trial. Even expertwitnesses can disagree, in good faith and on the basis of highly deliberatedengagements with evidentiary materials, and can persist in their disagreementeven after being confronted with it in adversarial and searchingly challengingcircumstances, for example of cross-examination. The situation is all the morestriking in the case of values and norms.

None but the most thorough-going relativist is likely to urge the institu-tional acknowledgment of diversity about the facts. But even relatively hardyopponents of relativism are forced, or so they often seem to feel, to recognizethat diversity on the evaluative front does indeed require institutional recog-nition. Whereas we might, even in the face of factual disagreement, hold firmto the idea that the truth is singular and indivisible, this kind of commitmentrequires more in the realm of values than most theorists and ordinary peopleare prepared to give in societies such as ours. Most of us, at least outside the

The legacies of John Raw Is 197

academy and the more fundamentalist religious and ethical circles, think thatdiversity of opinion or judgement about value matters is unlikely to be reducedby merely intellectual means; that its reduction will require, instead and asRawls himself thought, machineries of persuasion or coercion that cannotthemselves be morally approved.

Most of us, in other words, are pluralists: we acknowledge diversity in a not-entirely-grudging way. For pluralists, diversity is some kind of fact about ourlives that calls for institutional recognition, for example in acknowledging thatmany issues are matters of discretion: It is up to each individual (conscien-tiously deploying their various faculties in moderately favorable circum-stances) to decide for themselves what it is appropriate for them to believe orvalue in various situations where uncoerced consensus is not to be achieved.

There are, of course, various modes of commitment to broadly pluralistideas and there are, of course, various devices by means of which institutionalrecognition can be given to the kinds of diversity that might be more thanmerely superficial (e.g. in cloaking the mistaken beliefs or misguided valuesof some in our midst).

Isaiah Berlin and others of his ilk (including Stuart Hampshire and BernardWilliams) are strong pluralists in the sense of believing that diversity of eva-luations, especially, reflects deep and objective facts about the human situa-tion and does not merely arise from the (inevitable) imperfections of humanreasoning or of human sympathy. On this account, as John Gray compen-diously puts it:

[Values] are many, they often come into conflict with one another and areuncombinable in a single human being or a single society, and ... in manysuch conflicts there is no overarching standard whereby the competingclaims of such ultimate values are rationally arbitratable.

As Isaiah Berlin himself says:

[In the face of such diversity] it is better to face [the] intellectually uncom-fortable fact than to ignore it, or automatically attribute it to some defi-ciency on our part which could be eliminated by an increase in skill orknowledge; or, what is worse still, suppress one of the competing valuesaltogether by pretending it is identical with its rival—and so end by distort-ing both.

While Rawls did not explicitly reject (or tacitly accept) such a strong plural-ism, he did embrace some form of pluralism proper; he did not treat the man-ifest politically relevant diversity of our sorts of societies as a merely superficialphenomenon. In particular, Rawls believed, and argued vigorously, that

198 The Legacy of John Rawls

there is diversity, especially of kinds relevant to the organization of social lifebetween free and equal citizens interacting for their mutual benefit, thatcannot be eliminated except through coercion or manipulation. He referred,in particular, to "the practical impossibility of reaching reasonable and work-able political agreement. .. especially an agreement that might serve the poli-tical purpose, say, of achieving peace and concord." Of course, Berlin'sstrong pluralism is based, in the end, on an ontological commitment aboutthe nature of objective value. Since Rawls hoped "to avoid, so far as possible,disputed philosophical theses" (as strong pluralism certainly is), his own tac-tical pluralism was based, instead, on the so-called "burdens of judgment."These are those "hazards involved in the correct (and conscientious) exercisof our powers of reason and judgement in the ordinary course of political life,"which make it improbable that "conscientious persons with full powers ofreason, even after full discussion, will all arrive at the same conclusion."They include, inter alia, the fact that, in relevant cases, evidence bearing ondisputed propositions is conflicting and hard to interpret and that our con-cepts are vague, open-textured, and contested.

Perhaps most importantly, and returning to the issues raised by what, inA Theory of Justice, he had called intuitionism, Rawls recognized that muchhuman deliberation takes place in the presence of a plurality of choice-rele-vant standards of assessment. There is, and his claim to this effect should notsurprise us given his opposition to utilitarianism, no unitary metric, as GassSunstein (and others) calls it, in terms of which the options for choice canbe arrayed for comparison. There are, instead, a plurality of such metricsand, crucially, no canonical way of weighting these metrics to arrive ata choice. While any given individual might be able to weight, indeed have acharacteristic way of weighting, these metrics, of trading-off the values whichthey measure, there is no particular way of weighting or trading-off that isbinding on all (say, rational and well-informed) individuals. There is, rather,reasonable diversity in choosing in a multidimensional setting because, asRawls himself put it, people "may disagree about [the] weight" to be assignedto the various choice-relevant dimensions, and "so arrive at different [overall]judgments."

This is, I believe, one of the most strategic of all Rawls's many observationsabout the human condition. It grounds his opposition to utilitarianism; itexplains the place, in the unfolding argumentative narrative of A Theory of Jus-tice, of his discussion of intuitionism; and it grounds his approach to corn-mensuration in the ethico-political realm, and, in particular, his claims onbehalf of individual rights of conscience and the like (see the next section).Nevertheless, there is something unfinished about his discussion, about ourdiscussions, of this point. In particular, neither Rawls nor most of his com-mentators have treated even his weaker, tactical pluralism as anything other

The legacies of John Rawls 199

than an inconvenience. Certainly, it has not generally been treated as, rather,one of the most glorious opportunities for humankind. Oddly, among hisrough contemporaries, the philosopher and historian of science, ThomasKuhn, has, perhaps, seen this positive side of such intuitionism more clearlythan most.

If two different individuals use the same standards to evaluate options forchoice, but weight them differently and hence arrive at different decisionsabout which option to choose, then this can, Kuhn sees, have some importantadvantages (relative to a situation where both arrive at the same decision).In particular, this divergence in decisions can spread risk in a particular way,at least in circumstances where decision-makers are exploring a relativelyunknown terrain. Suppose that two scientists are trying to choose betweentwo different (and incompatible) elaborations of some paradigmatic frame-work they are both committed to. (This is a routine aspect of normal science,according to Kuhn.) Obviously, if both reach the same conclusion aboutwhich elaboration is better, they will concentrate their efforts, and this maybe advantageous. On the other hand, if the two reach different conclusions,then a larger portion of the terrain will be explored by them and this too maybe advantageous. After all, where both agree, they might still be wrong inthinking that the particular, agreed elaboration is the right one. Where theydisagree, there is a greater chance that the right elaboration, whatever thatmight be, will be explored by one of them. This, then, is an advantage of notresolving disagreement prematurely, but, rather, letting it become a resourcerather than a problem.

All this has obvious applicability to situations more centrally of concern topolitical (and ethical) theorists. Where individuals do not (spontaneously)agree (absent coercion) on how to weight choice-standards and, hence, onwhat choice to make, they may benefit, individually and collectively, fromthe fact that they make different choices. Each gets to see how other choices(than the one that was made) pan out and, hence, to learn something aboutthese options that probably could not have been learned in any other way.(As many theorists have emphasized and as certainly follows from the finitudeand fallibility which underpin the burdens of judgement, it is difficult to geteither the facts or the preferences straight ex ante in situations of even moderatecomplexity. )

Exploring the circumstances in which there are social and individual bene-fits from the social recognition of diversity is, I think, a substantial project ofcontinuing research and, assuredly, a legacy of Rawls's own work on the bur-dens ofjudgement. At the same time, and as is obvious, the fact of diversity andthe inappropriateness, in democratic societies, of dealing with it through coer-cive means, raises the crucial problem of understanding how social order ispossible and, in particular, how free and equal citizens committed, as Rawls

200 The Legacy of John Rawls

later put it, to a diversity of reasonable comprehensive doctrines can find amutually agreeable (and stable) basis for the organization, for mutual benefit,of their social lives.

Rawls and incommensurability

That Rawls was a commensurator is not a proposition that I have very fre-quently seen or heard asserted. The literature about incommensurability isvery nearly completely innocent of any reference to Rawls as an importantcontributor to the project of bringing, into a situation of disciplined compari-son, options which are, prima facie, incomparable. Nevertheless, Rawls didmake two crucial contributions to our understanding of the prospects for com-mensurating the (apparently) incommensurable. These were: (a) the normal-izing method of commensuration that is implicit in the use of original positionargumentation, and (b) the distributive model for commensuration whichemphasizes the importance, and provides social recognition for, the devolu-tion of decisional responsibility to individual citizens. I take these pointsin order.

Normalization and the original position

If an individual A ranks an option Oj higher than O2 because it is betterin relation to his understanding of the good (or given his weighting of thedecision-relevant standards of assessment) and B ranks O2 higher than Ojbecause it is better in relation to her different understanding of the good,then these options are, for any collective constituted by A and B, prima facieincomparable, or incommensurable, with one another. This is a situationthat Rawls had clearly in mind when he suggested, in A Theory of Justice, that,absent a public conception of the good, there could be, within a teleologicalframework, no public conception of the right and, in particular, no concep-tion that would rank-order options relating to constitutional fundamentals orthe basic structure of society, where a consensus of reasonable judgementsis clearly vital (as it might not be in other cases). Without consensus at thislevel, there would be, as Rawls clearly understood, no basis for securing thebenefits of cooperation for mutual advantage. We would not, in this situation,inhabit a well-ordered society. How are we to tackle this difficulty?

If it is diversity that creates a difficulty of commensuration, then perhapsthis difficulty can be overcome if we reduce diversity compatibly with givingit social recognition. It is this seemingly impossible feat that Rawls essayed inhis work, especially in A Theory of Justice.

The legacies of John Raw Is 201

How did Rawls propose to reduce diversity? By normalizing the social con-tractors—by making each of them the same as all the others. And oncethey are the same, of course, their judgements about constitutional fundamen-tals will also be the same and the problem of commensuration will be solved.Of course, not all normalizations are legitimate. Rawls himself recognizedthat we can reduce (at least expressed) diversity of judgement by coercionand many of his commentators drew attention—they thought damagingly—to the fact that diversity is lessened through ideological mechanisms, for exam-ple through attempts, crude or subtle, to control what people believe andvalue. So the project of normalization, if it is to be successful within the frame-work of a democratic culture, has to meet two criteria. First, there has to be atechnology of normalization, or reduction of diversity. Secondly, there has to bea rationale for the use of such a normalizing technology.

Rawls's technology of normalization is familiar. It involves the originalposition and, in particular, the veil of ignorance. We are asked to imaginethat social contractors, considering what constitutional fundamentals mightserve to promote cooperation for mutual benefit between free and equal citi-zens, do so subject to a restriction on their knowledge. No individual knows,for instance, "his place in society, his class position or social status . . . his for-tune in the distribution of natural assets and abilities, his intelligence andstrength, and the like." Accordingly, no individual can, in his own delibera-tions about constitutional fundamentals, rank proposals about them in accor-dance with how well he is likely to fare if these proposals are implemented.How is he to rank them, then? In accordance with how any person is likely tofare if they are implemented. And, according to Rawls, when individualsdeliberate in this way, how one person ranks proposals is the same as how anyother person ranks them—this is the element of normalization, and com-parability, on a collective level, is therefore achieved.

The reduction of diversity which is effected by the original position, and,specifically, by the veil of ignorance, enables us to render commensurable, ina collective sense, those proposals about the basic structure which, otherwise,could not be ranked in a way that could, in the face of antecedent diversity, beconsidered socially binding. Of course, commensurability is purchased atwhat might be considered a high price, namely, the normalization of thesocial contractors. How is this normalization to be justified? Rawls's answerwas, I think, threefold.

First of all, there is, if you like, a specifically ethical rationale for the use oforiginal position argumentation and for the deployment of the veil of ignor-ance. After all, what it prevents the use of is, precisely, information that itwould be ethically improper to use as a basis for the assessment of alterna-tive basic structures. It prevents the use of information that people would,characteristically, use in a self-interested way. As Rawls said: "One excludes

202 The Legacy of John Rawls

knowledge of those contingencies which set men at odds and allows them to beguided by their prejudices." Blocking the use of such information forces indi-viduals to think impartially, that is, ethically, about the terms of their associa-tion with one another. On the other hand, the information about themselvesthat is available to individuals deliberating about the basic structure does,according to Rawls, represent them as free and equal moral agents. Thereis, then, nothing arbitrary, according to this reasoning, about the reductionsof diversity effected by deployment of the veil of ignorance. Insofar as eachperson, in a quasi-Kantian way, thinks of herself as denned in terms of herstatus as a free and equal citizen of a well-ordered society, she has reason toaccept her normalization (as, specifically, morally improving).

Secondly, as Rawls notoriously claimed, and spent three decades trying todemonstrate, an appropriately devised specification of the initial situation (forthe choice of the basic structure) will yield principles regulating social life thatcan be affirmed, in reflective equilibrium, against our considered judgementsof justice (as suitably pruned and adjusted). These principles, in other words,generate their own support when they are made the basis for social life. Thisis Rawls's well-known preoccupation with what he came to call, in PoliticalLiberalism, the question of stability. I will consider this argument further inthe next section.

Finally, and this is what I was alluding to earlier, when I mentioned thedistributive model, what suitably normalized parties to the original positionagree to actually provides a great deal of social/institutional recognition oftheir antecedent diversity in normative judgements. For the basic structure,which Rawls suggested we would agree to, is one which prioritizes a system ofindividual rights, especially of opinion, conscience, and the like, and hencecreatively evades the demand for more specific forms of commensuration thatwould be incompatible with our status of free and equal citizens (given thereasonableness of much of the diversity of judgements that we experience).

This will bear some explication. (It is point (b) above.)

Liberty of conscience and commensuration via separation

I said, earlier, that Rawls, as commensurator, was concerned to provide forthe assessment of proposals about the basic structure while also giving duerecognition to (reasonable) doctrinal diversity. The original position answersthe first need. What serves the second? Simply the fact that suitably nor-malized social contractors agree to an allocation of rights which protects andnurtures diversity. According to Rawls, "it seems that equal liberty of con-science is the only principle that the persons in the original position canacknowledge." He continues: "They cannot take chances with their libertyby permitting the dominant religious or moral doctrine to persecute or to

The legacies of John Raw Is 203

suppress others if it wishes." What the contractors agree to, then, is preciselya device, a system of liberties, including, crucially, a liberty of conscience, thatprotects diversity of belief, commitment, and behavior. As he puts it in PoliticalLiberalism, "within the scope allowed by the basic liberties and the other provi-sions of a just constitutional regime, all citizens can pursue their way of life onfair terms."

All this is straightforward, but it points, I believe and as others havealso seen, to a new way of understanding what commensuration might actu-ally amount to. It points, in short, to a second way of commensurating thatneeds to be given more recognition by philosophical (and other) discussantsof this topic.

Sometimes, when we need to decide, collectively, how to arrange our lives,what is at stake is a specific concrete plan of action that we, the relevant par-ties, all need to play a role in executing. For instance, if friends are trying todecide how to spend time together, they might see this as involving comingto an agreement about, say, where to have dinner and what, specifically andconcretely, to do afterwards (go to the movies or perhaps the pub). Such col-lective decisions are, typically, domains of compromise and negotiation, butwhat emerges from these deliberations is an agreement on something specificthat will be a joint engagement by all involved parties. Gommensuration, inthis case, is achieved when we have a collectively agreed ordering of concreteand specific behavioral options—we prefer Pizza Hut plus movie to McDo-nald's plus pub.

Obviously, this is not what commensuration a la Rawls amounts to. It is notthat the social contractors agree, for instance, about what they, or the ordin-ary people whom they represent, will believe. Indeed, such an approach to theproblem of commensuration would be entirely out of keeping with Rawls'slarger project which is, precisely, to safeguard the diversity of reasonable com-prehensive doctrines. In what sense, then, can the results of original positiondeliberation be considered a form of commensuration?

There are, in fact, two primary modalities of commensuration, and Rawls'swork illustrates the second, which has, in my view, been too little commentedon in general.

The first and more familiar modality (which is not Rawls's) involves devel-oping a field of calculation in which the various options and the standards usedto assess them are brought together, and trade-offs among the standards areidentified to the degree required to produce an unequivocal ranking of theoptions. Cost-benefit analysis is a good example of this modality of com-mensuration. You might have two or more options for choice, O1; . . . On,and two or more standards for assessing the options Si, .. . Sm, and, oncewe determine how to establish a common currency among the standards,we can identify a determinate ranking of the options and hence render them

28

204 The Legacy of John Rawls

commensurable. Since, in typical cost-benefit analyses, many of the standardscan already be expressed in monetary terms (input costs and potential rev-enue), commensuration will usually involve pricing out those standardswhich are not already expressed that way, for example by determining stake-holders" willingness to pay.

There is, however, a kind of commensuration without trade-offs, as theoristssuch as John Gray, Gass Sunstein, and Friedrich Hayek have recognized.Gray puts the matter clearly:

The importance of several property for civil society is that it acts as anenabling device whereby rival and possibly incommensurable conceptionsof the good may be implemented and realized without recourse to any col-lective decision-procedure . .. One may even say of civil society that it is adevice for securing peace by reducing to a minimum the decisions on whichrecourse to collective choice—the political or public choice that is bindingon all—is unavoidable.

This is, if you like, commensuration as separation. Instead of bringingtogether the various standards and options and stakeholders in a field of calcu-lation, we keep them apart. By separating the spheres (of decision-making),we ensure that diversity is respected and facilitated: each person is the judgeof their own situation and makes the choice that best reflects their understand-ing of it.

And this is, of course, what Rawls's allocation of rights of conscience (andthe like) also ensures, just as several property does. Just as each person, in aregime of freedom of commerce, can dispose of and acquire the goods that hewants, so too in a regime of several mentalities can each person form andexpress the beliefs and values that she thinks appropriate.

The only remaining question is more obscure: What makes this a system ofcommensuration? The crucial point is this. Although individuals in a systemof freedom of conscience do not agree about what, substantively, it is appro-priate to believe, they do agree about a distributional issue, namely, abouthow beliefs and commitments should be distributed among individuals. Eachof us, from behind a veil of ignorance, agrees, in effect, that distribution ofattitudes is best in which each individual has been able to freely (and con-scientiously?) arrive at the attitudes by which she will organize her ownlife. This distribution of attitudes to individuals is ranked ahead of any otherdistribution by the original position argumentation. It is in this sense, thisdistributional sense, that Rawls's arguments lead to a commensuration ofoptions. (And the options are distributions, not concrete and specific beha-vioral alternatives.)

This, I think, is a legacy of Rawls's work that deserves exploitation.How can we rethink the project of social decision-making or social choice to

30

31

32

The legacies of John Rawls 205

recognize and give proper emphasis to such a distributional modality ofcommensuration and to the normalizing technology that gives rise to it? Thismodel is authentically democratic. It tries, however precariously, to balancethe need for consensus with the value, the positive, affirmative value of diver-sity. It deserves a more prominent place in our thinking about the polity and togive it such a place would be a suitable tribute to Rawls's attempt to articulateand illustrate it.

Rawls's pragmatism

Perhaps the most interesting, and neglected, aspect of Rawls's legacy is themodel for a form of pragmatism in ethical theorizing that he developed inA Theory of Justice. Although Rawls's pluralism and his ideas about originalposition reasoning have received considerable attention, discussion of Rawlsas a pragmatist is largely limited to non-English-speaking authors.

In fact, A Theory of Justice provides a detailed, generalizable model for prag-matically oriented theorizing in ethics and politics across a range of issues.This model is enunciated, specifically, in Rawls's analysis, in sections 22 and23, of the circumstances of justice and of the formal constraints on our under-standing of justice. The basic ideas are obvious enough, if not much discussed,and a short summary (and amplification) will therefore suffice.

First of all, we need to understand that Rawls was trying to determine, ineffect, which principles of sociability are fit to play a certain role in the orga-nization of our collective lives. His analysis of justice was therefore not a con-ceptual analysis, but, rather, an exercise in armchair social theory. Thequestion was not, or anyway was not exhausted by: What is the current under-standing of justice in our society? (This is what theorists who assimilateRawls's work to the coherence theory seem to think.) The question is, rather:What sort of understanding ofjustice, if it were propagated and if there wereuptake by most citizens, would function effectively in the circumstances whichmake such an understanding socially important? Rawls's analysis was, then,pragmatic, not conceptual. Rawls put this clearly enough:

The intuitive idea ofjustice as fairness is to think of the first principles ofjustice as themselves the object of an original agreement in a suitablydefined initial situation. These principles are those which rational personsconcerned to advance their interests would accept in this position of equal-ity to settle the basic terms of their association. It must be shown, then, thatthe two principles ofjustice are the solution for the problem of choice pre-sented by the original position. In order to do this, one must establish that,given the circumstances of the parties, and their knowledge, beliefs, and

33

206 The Legacy of John Rawls

interests, an agreement on these principles is the best way for each per son to securehis ends in view of the alternatives available.

It is, I repeat, a fact about how principles might function that justifies thechoice of these principles as principles of social coordination for our society.That they enable "each person to secure his ends," subject to certain circum-stances, conditions, and constraints, is their justification, not that they reflectsome antecedent understanding of what justice is, metaphysically or concep-tually. (This, by the way, shows why Rawls's approach is not subject, or atleast is not subject for the reasons which are usually adduced, to the chargethat it provides inadequate ethical leverage against such existing understand-ings ofjustice as may, of course, reflect ideological thinking. Rawls's approachis meant, specifically, to correct for mistaken understandings that mightnevertheless be widely diffused. And the basis for correction is, of course, apragmatic one: How well does this understanding facilitate the achievementof certain goals?)

In view of how little comment this aspect of Rawls's approach hasattracted—that is, its pragmatic orientation—it is difficult to exaggeratethe importance of these considerations, not only for Rawls's specific project,but, indeed, for ethico-political theorizing in general. From a pragmaticpoint of view, the question is always, What is good in the way of belief? Howcan our aims as individuals and collectively best be promoted by our systemof beliefs and practices? This methodological readjustment is, I think, a contri-bution by Rawls to our thinking in these areas that is truly revolutionaryin potential.

Let's see, in detail and with some amendments, how pragmatic analysisworks in relation to normative concepts and principles. Rawls explicitly iden-tified two sorts of considerations that are relevant to such analysis and impliesa third.

First of all, Rawls noted that, in order to determine what sort of principlesmight be fit to play a certain role, we must understand what circumstancesmake it necessary to develop and propagate such principles. And the reason-ing, largely implicit in Rawls, is obvious enough. Suppose, for instance, thatscarcity of supply relative to demand for social primary goods, in Rawls's ter-minology, is characteristic of our situation. This is part of what makes the pro-pagation of distributional principles and practices necessary: given scarcityand certain other factors, people will not collectively and automatically self-equilibrate to ensure that demand does not outstrip supply. But, of course, thisfact must also be taken account of in the development of precisely these prin-ciples, which, in particular, will not be fit to play the role ofjustice if they justassume away the problem of distribution by presupposing, for instance, thatindividuals will spontaneously adjust their demands to the supply available

34

The legacies of John Rawls 207

to fulfill them. (Bruce Ackerman's account, in Social Justice in the Liberal state, isparticularly clear on the importance of these circumstances. )

Secondly, Rawls noted that, in order to determine what sort of principlesmight be fit to play a certain role, we must understand what (formal) con-straints on such principles are reasonable to impose, at least tentatively, as anexpression of the function which we expect such principles to discharge.(Given the pragmatism of Rawls's approach, the epithet "formal" was,I think, unfortunate.) Again, the reasoning is obvious. If we expect principlesof justice to play a role in settling certain sorts of disputes that might arise inour society, then, obviously, they will have to exhibit certain sorts of features.One of Rawls's constraints is, of course, that "a conception of right mustimpose an ordering on conflicting claims," a requirement which, accordingto Rawls, whose pragmatism was plainly in evidence here, "springs directlyfrom the roles of its principles in adjusting competing demands." (If we arein dispute and appeal to the right as a basis for settling our dispute, but it, theright, fails to order our claims, then it contributes nothing to settling the dis-pute we'd tried to use it, as a tool, to settle. (This is, of course, Rawls's way ofrecognizing the importance of commensuration!)

Finally, I note, distinguishing what Rawls himself had run together, that, inorder to determine what sort of principles might be fit to play a certain role, wemust understand what capacities and attitudes human beings are likely tobring to the situations, in which these principles might be deployed, that willsupport their deployment in those situations. (This is the "possible" aspect ofRawls's analysis of the "normal conditions under which human cooperation isboth possible and necessary.") In this case, principles and practices cannot bepropagated, let alone play a role in adjusting people's relations with oneanother, if, for instance, there is some (relatively) insuperable barrier, cogni-tive, affective, or institutional, to their successful uptake. (Rawls's analysis offeasibility considerations in Part Three of A Theory of Justice is directed, in part,to an examination of these sorts of issues.) In my book Free Public Reason,I argued, for instance, that the "reasonableness" of individuals is, in this sense,a capacity, or perhaps an attitude, that needs to be widespread in a given com-munity if certain sorts of social relations are to be possible in that community.This, I will say, is a condition for justice.

It is circumstances, conditions, and constraints, in the senses identified, thatplay a crucial and largely unnoticed role in Rawls's pragmatic analysis of jus-tice. In short, we try to identify principles of justice such that:

because of the conditions for justice—for example, people's reasonable-ness;these principles can meet the demands specified by the formal constraintsonjustice;

37

208 The Legacy of John Rawls

in the circumstances of justice—for example, despite the relative scarcityof supply with respect to demand.

Notice, in particular, that an analysis conducted on these terms cannot be con-fused with conceptual analysis, even if that ideal is interpreted rather loosely.Although there may be certain conceptual elements involved in articulatingthe constraints on justice, even in this case functions are to the fore—What dowe want to use the principles (and practices) of justice to do? And, certainly,claims about the conditions for and circumstances of justice are, althoughusually highly abstract and general, matters of fact rather than matters ofmeaning. We are trying to design a tool for use by certain kinds of agents toaccomplish certain sorts of purposes in a certain kind of environment, and ourproblem is one of practical functional design, not of conceptual analysis ormetaphysical speculation about The Good or The Right.

This, I submit, is the proper way to conceptualize Rawls's method oftheorizing about justice. He was primarily concerned to design evaluativetools that "can perform their practical function well." A pragmatic approachis, on this account, the right one for principles and practices that superintendthe social relations of human beings. They are tools and should be theorized assuch. As Richard Rorty says: "Pragmatism treats every such [conceptualiza-tion] ... as an experiment, designed to see if we can get what we want at acertain historical moment by using a certain language." Or, as ElizabethAnderson puts it:

[VJalue judgments are justified by showing that they can perform theirpractical function well. This is done by showing that it is rational to usethem to guide our deliberations and attitudes. So instead of saying that it isrational to value something because it is good, pragmatism says that it isgood because it is rational for us to value it.

The suggestion that political theory is based on the functional design of socialtools for collective use is, then, Rawls's third great legacy to political theory, orso I maintain. Rawls, like Quine before him, stands revealed, ultimately andcontrary to widespread earlier understandings, as some form of unselfcon-scious, perhaps even diffident, pragmatist about the work of the politicalthinker. This point is still, of course, insufficiently appreciated and Rawls'spragmatism is, therefore, his least noticed legacy to political thinking.

Remembering Rawls

To say that, without A Theory of Justice, there might not be any (non-Marxist)political theory in Anglophone societies is, very nearly, a cliche. Certainly,

40

The legacies of John Rawls 209

what political theorizing we have exists in the light cast by that great work. It

became a touchstone to thinkers working both at the level of high theory and

at the level of concrete practice. Like that other great work of post-war Amer-

ican philosophy, Thomas Kuhn's The Structure of Scientific Revolutions, it has,

however, been as widely misunderstood as it has been discussed and praised.

Perhaps surprisingly, there is even something common to their two projects.

Both were concerned about the prospects for commensuration in the face of

plurality. One appropriate response to Rawls's legacy would be to learn to

read his work, for example through Kuhn, in a way that draws on work outside

the sometimes too closely drawn boundaries of political and ethical theory.

Rawls himself was expansive in his intellectual range of interests. Reading

Rawls through the work of cognitive psychologists, decision theorists, legal

theorists, epistemologists, and the like would be an appropriate tribute to his

own breadth of reading.

Notes

The phrase "John Rawls" yields more than 54,000 hits on the Google Internetsearch engine. The Philosopher's Index shows more than 1,700 hits on "Rawls," andthe Infotrac multidisciplinary database of articles, ASAP, shows more than 2,400articles indexed to the name "Rawls," across a wide range of fields. These datashow one way of measuring Rawls's influence and legacy.See my article "Pluralism and Liberalism," in Gerald Gaus and ChandranKukathas (eds), The Sage Handbook of Political Theory (London: Sage, 2005) forreferences. The most important are: W.B. Gallic, "Essentially Contested Con-cepts," Proceedings of the Aristotelian Society (1955—56), 167—98; Stuart Hampshire,Morality and Conflict (Oxford, UK: Basil Blackwell, 1983); Isaiah Berlin, Four

Essays on Liberty (London: Oxford University Press, 1969); and, more recently,J. Donald Moon, "Liberalism, Autonomy, and Moral Pluralism," Political Theory

31 (2003), 125-35.See my book Incommensurability and Commensuration (Aldershot: Ashgate, 2003).See my article "Rituals of Impartiality, "Social Theory and Practice?! (2001), 65—81.See my article "Expertise, Democracy, and Applied Ethics," Journal of AppliedPhilosophy 15 (1998), 49-56. For a subtle account articulating directly to Raw-Isian concerns, see Martha Nussbaum, "Moral Expertise? Constitutional Narra-tives and Philosophical Argument," Metaphilosophy 33 (2002), 502-20.See Andrew Fiala, "Toleration and Pragmatism," Journal of Speculative Philosophy

16 (2002), 103—16, and also my article "Relativism and Reflective Equilibrium,"TheMonistll (1988), 420-36.See my article "Ethical Pluralism and the Role of Opposition in Democratic Pol-itics," The Monist 73 (1990), 437-63, and, more recently, John Gray, "WherePluralists and Liberals Part Company," International Journal of Philosophical Studies6 (1998), 17-36.

1.

2.

3.4.5.

6.

7.

TM

41

210 The Legacy of John Rawls

8. John Gray, Post-Liberalism (New York: Routledge, 1993), 65.

9. Berlin, Four Essays on Liberty (London: Oxford University Press, 1969), 1.

10. PL, 63.

11. PL, 56,58.12. Cass Sunstein, "Incommensurability and Kinds of Valuation," in Ruth Chang

(ed.), Incommensurability, Incomparability, and Practical Reason (Cambridge, MA:

Harvard University Press, 1997), 238.

13. PL, 56.

14. See my article "Incommensurability and Commensuration," Studies in the Historyand Philosophy of Science 32 (2000), 429-47. See also Peter Achinstein, "Subjective

Views of Kuhn," Perspectives on Science 9 (2001), 423-32.

15. T.S. Kuhn, The Essential Tension (Chicago: University of Chicago Press, 1977),

331.

16. See Paul Hoyningen-Huene, Reconstructing Scientific Revolutions (Chicago: Univer-

sity of Chicago Press, 1993), esp. 151—2. See also my article "A 'Demographic'Approach to the Rationality of Science: The Wave Model," Methodology and

Science 26 (1992), 244-56.17. Note that this demonstration does not depend on treating diversity as an empirical

consequence of strong pluralism. That diversity in judgements or in weighting isvaluable is not a proposition whose demonstration depends on such ontological

matters. Tactical pluralism is enough to ground it. And this is why I said Rawls-

ian business in this area was unfinished.18. See Christopher Cherniak, Minimal Rationality (Cambridge, MA: MIT Press,

1986), and also my article "Transcendence and Conversation: Two Conceptions

of Objectivity," American Philosophical Quarterly 30 (1993), 87-108.19. The economist James March has made a great deal of this point. See, in particu-

lar, "The Technology of Foolishness," injames March and Johan Olsen (eds),

Ambiguity and Choice in Organizations (Bergen: Universitetsforlaget, 1976). March

emphasizes, in particular, that rational-choice models are mistaken in (largely

tacitly) assuming that choice is a function of preferences. Rather, he says, choices

are made, frequently, in order to understand what our preferences might be —rather than as an expression of them.

The canonical anthology edited by Ruth Chang, Incommensurability, Incomparabil-

ity, and Practical Reason contains, in its index, but three references to Rawls and

none of these amounts to a portrayal of him as someone preoccupied with ques-

tions of commensurability. Searching The Philosopher's Index on the Boolean pro-duct of "Rawls" and "incommensurability" also yields only three entries (one by

John Gray), but, again, none of them treats Rawls, as I propose to do, as "theGreat Commensurator." Charles Altieri is a notable exception: "Rawls seeks a

social philosophy capable of developing a model of judgment that can negotiate

competing and 'incommensurable' visions of the good without subordinatingthem to any encompassing metatheory." (Charles Altieri, "Judgment and Justice

under Postmodern Conditions", in Reed Dasenbrock (ed.), Redrawing the Lines

(Minneapolis: University of Minnesota Press, 1989), 76)

20

The legacies of John Rawls 211

21. To use the word "normalization" (and its cognates) in this context is, inevitably,to sound Foucauldian echoes. This may seem surprising or perhaps undignified ina discussion of Rawls's work. I do not agree with this response. Foucault, largelyunnoticed by those analytic philosophers who scoffed, and still scoff, at his work,has a powerful, also quasi-Kantian, critique of what Rawlsians might call illiberalsocial orders. See my paper "Two Conceptions of Autonomy," Economy and Society27 (1998), 28-49, and, more recently, Paul Healy, "A'Limit Attitude': Foucault,Autonomy, Critique," History of the Human Sciences 14 (2001), 49—68.

22. See my Incommensurability and Commensuration for an overview of the literature onincommensurability in value theory. This particular way of characterizingincommensurability in a social context brings out the relevance, to the under-standing of incommensurability, of the work of Kenneth Arrow on social choice.For helpful commentary, see, for instance, Dennis Mueller, Public Choice II (Cam-bridge, UK: Cambridge University Press, 1989), esp. Ch. 20.

23. TJ, 137.24. Ibid., 19.25. PL, 98.26. I develop an interpretation of this goal in my paper "Rituals of Impartiality."27. TJ, 207.28. PL, 155.29. Most notably John Gray. See his Post-Liberalism, 314: "The importance of several

property for civil society is that it acts as an enabling device whereby rival andpossibly incommensurable conceptions of the good may be implemented and rea-lized without recourse to any collective decision-procedure."

30. See, for instance, Peter Bogetoft and Peter Pruzan, Planning with Multiple Criteria(Amsterdam: North-Holland, 1991).

31. Gray, Post-Liberalism, 314.32. Cf. Sunstein, "Incommensurability and Kinds of Valuation," 234: "[T]he

traditional liberal effort to use law so as to create diverse social spheres—families,markets, politics, religious organization—makes space for different kinds ofvaluation."

33. "Rawls" AND "pluralism" gets 70 "hits" in The Philosopher's Index; "Rawls" and"original position" gets 113. A Google search on "Rawls' pragmatism" yieldstwo hits, one now inaccessible and one to my Stanford Encyclopedia article on"Original Position." The Philosopher's Index gives half-a-dozen hits on "Rawls"AND "pragmatism," but none is specifically on Rawls as a pragmatist.

34. TJ, 102-3, emphasis added.35. Bruce Ackerman, Social Justice in the Liberal state (New Haven: Yale University

Press, 1980).36. TJ, 133-4.37. Free Public Reason (Oxford, UK: Oxford University Press, 1996).38. See G.A. Cohen's "Facts and Principles," Philosophy and Public Affairs 31 (2003),

211—45. On Cohen's account, pragmatically derived principles of justice cannotthemselves be "ultimate" principles, which must on Cohen's analysis rest on

212 The Legacy of John Rawls

fact-insensitive principles, but only regulatory principles. Cohen (242) antici-

pates and counters my own most obvious reply:

"Fine. I accept your distinction between basic principles and principles of reg-ulation. But why should I care about basic principles? I care about what hap-

pens in the world, and the principles of regulation that we adopt in the light of

the facts determine that." "The response is unsustainable [Cohen replies]

because we necessarily have recourse to basic principles to justify the principles

of regulation that we adopt. . ."

Certainly we do, and, for the pragmatist, that principle is, simply, that weadopt for the purposes of living together for mutual advantage and as free and

equal citizens those principles that advance this project. Of course, it may be

difficult to ground or defend such a principle, but, anyway, it seems (a) pragmatist,

(b) ultimate, and (c) fact-insensitive, while (d) likely to yield something like the

two principles of justice in something like contemporary circumstances in whatRorty calls "the North Atlantic democracies."

39. Richard Rorty, "Inquiry as Recontextualization," in his Objectivity, Relativism,

and Truth (Cambridge, UK: Cambridge University Press, 1990), 98-9.40. Elizabeth Anderson, "Practical Reason and Incommensurable Goods," in

Chang (ed.), Incommensurability, Incomparability, and Practical Reason, 91—2.

41. Thomas Kuhn, The Structure of Scientific Revolutions (Chicago: University of Chi-cago Press, 1970).

NAME INDEX

Achinstein, Peter 21 OnAckerly, Brooke 120nAckermann, Bruce 22, 30n, 108, 119n,

211nAinslie, Donald C. 19nAlexander, Larry 35, 47n, 48nAltieri, Charles 21 OnAnderson, Elizabeth 212nAnderson, Perry 22, 30nApel, Karl-Otto 108, 119nAquinas, Thomas 89,112Aristotle 38-41, 44, 45, 91, 161, 173nArneson, Richard 35, 43, 47nArrow, Kenneth 21 InAudi, Robert 122n, 127, 129, 147n,

148n

Barry, Brian 22, 30n, 118nBeitz, Charles 8, 18n, 19n, 191nBell, Derek R. 19nBenhabib, Seyla 17n, 111, 118n, 120n,

122nBenson, Peter 20nBerlin, Isaiah 117, 118n, 136, 197, 198,

209n, 210nBernstein, Alyssa 66nBernstein, Mark 18nBird, Colin 81nBoettcher, James 7, 124-51Bogetoft, Peter 21 InBohman, James 117n, 120n, 122nBrake, Elizabeth 5, 67-84Brennan, Samantha 3 OnBrooks, Thorn 1—21, 66n, 176nBrower, Bruce 30nBrown, Jonathan 19nBuchanan, Allen 30n, 31 n, 191 n

Campos, Paul F. 82nCaney, Simon 18n, 19n, 30n, 191 nCarruthers, Peter 19n

Carter, Stephen L. 124, 145n, 148nCavallero, Eric 19nChang, Ruth 21 OnChatterjee, Deen 10Cheng, Chung-ying 20nCherniak, Christopher 21 OnClifford, William 73-76, 83nCohen, G. A. 211-12nCohen, Joshua 80n, 108, 119n, 121n,

122nColeman, S.J.John A. 150nConfucius 20nConkle, Daniel O. 148n, 150nCronin, Ciaran 30n

D'Agostino, Fred 7, 17n, 195-212D'Amato, Neal 49nDaniels, Norman 9Dancy, Jonathan 104nDasenbrock, Reed 210nDavion, Victoria 30n, 149nDe Greiff, Pablo 30nDe Marneffe, Peter 150nDeweyjohn 156, 158Dombrowski, Daniel A. 19n, 104n,

145n, 150nDreben, Burton 3In, 65—6nDryzekjohn 122nDuggan, Kyla Ebbels 172n, 176nDworkin, Ronald 12-13, 15, 20n, 2In ,

69, 79,81n, 84n, 108, 118n, 186, 193n

Eberle, Christopher J. 133, 148n, 149n,150n

Eck, Diana 149nElsterjon 44,49nEngelhardt, H. Tristram 69, 81nEstlund, David 30-ln, 33n, 96-9,

103n, 104n, 105n, 118n, 121nEuben, Peter 49nExdelljohn 193n

214 Name Index

Fan, Ruiping 20nFarrelly, Colin 19nFiala, Andrew 209nFinnis,John 146nFish, Stanley 120nFleischacker, Samuel 66nFoucault, Michel 21 InFowler, Robert Booth 148-9nFraser, Nancy l l l , 1 2 0 nFreeman, Samuel 20n, 120n, 122nFreyenhagen, Fabian 1—21, 176n

Gallic, W. B. 209nGalston, William 111, 118n, 120n,

121nGaus, Gerald 209nGeorge, Robert P. 111, 120n, 146n,

150nGlaucon 109Gonzalez, Ana Marta 4,11, 17n, 152-76Goodin, Robert 12InGray, John 118n, 12In, 197, 204, 209n,

210n,211nGreen, Karen 78, 84nGreenawalt, Kent 132-33, 146n, 148n,

150nGutmann, Amy 108, 119n

Haack, Susan 83nHabermas, Jiirgen 58, 59, 65n, 74, 101,

103n, 104n, 105n, 108, 119n, 123nHahn, Lewis 83 nHampshire, Stuart 8In, 197, 209nHamptonjean 64n, 82n, 103n, 108,

118nHardin, Russell 122nHayek, Friedrich 204Healy, Paul 21 InHegel, Georg Wilhelm Friedrich 12,

20n, 82n, 156, 157, 167, 176nHeinrich, D. 17InHerman, Barbara 4, 12, 20n, 153, 155,

157-8, 161-2, 168-9, 170-ln,171-2n, 173n, 174n, 176n

Hertzke, Allen D. 148-9nHill, Jr., Thomas 4, 155, 158, 160, 165,

170n, 171n, 172nHobbes, Thomas 86, 91, 94, 98, 109,

161Hoffman, Stanley 30n

Hollenbach, S.J., David 147n, 148n,150n

Holmes, Stephen 30n, 108,110,119n,120n

Houlgate, Stephen 12, 20nHoward, Matthew 80nHoyningen-Huene, Paul 21 OnHuemer, Michael 82nHume, David 11,12,161,167

Jackson, Timothy P. 149nJames, William 83nJaspers, Karl 159, 172nJohnson, James 122—23nJuper, Andrew 18n

Kant, Immanuel 4, 6, 11, 12, 16n, 17n,44-5, 49n, 63, 66n, 70, 75, 81n, 83n,85-7, 90-3, 100-1, 105n, 107,152-70, 170n, 171n, 172n, 173n,174n, 175n, 176n, 202, 21 In

Kelly, Erin 19n, 47nKelts, Steven 49nKnight, Jack 122-3nKorsgaard, Christine M. 4, 20n, 54,

64n, 104n, 152, 155, 158-9, 165-7,168-70, 171n, 172n, 176n

Krash, A. 2InKuhn, Thomas S. 199, 209, 2lOn, 21 InKukathas, Chandran 16n, 17n, 209nKumar, Rahul 66nKymlicka, Will 69, 77, 81n, 82n, 84n

Laberge, Pierre 19nLaden, Anthony Simon 4, 50-66Larmore, Charles 101, 103n, 104n,

105n, 106n, 108, 111, 119n, 120n,121n, 146n, 147n

Laslett, Peter 47nLean, David 83nLefort, Charles 117, 123nLeibniz, Gottfried Wilhelm 11,12Lessnoff, Michael 47nLevinson, Sanford 148nLloyd, Sharon 150nLockejohn 91, 104n, 107Luther, Martin 89

MacKinnon, Catherine 5, 15, 17n, 66n,67-8, 71, 76, 77, 80n, 81n

Name Index 215

McDermott, JohnJ. 83nMcMurtry, John 84nMcVeigh, Timothy 115Macedo, Stephen 33n, 120nMahoney, Jon 7,85-106Maio, Giovanni 19nManin, Bernard 122nMarch, James 21 OnMartin, Rex 30nMason, Carol 122nMay, Simon 49nMendus, Susan 82nMichelman, Frank I. 20nMill, John Stuart 1, 60, 65n, 69, 73, 76,

83n, 83-4n, 107Montaigne 90Moon, J.Donald 209nMoore, G. E. 161, 173nMoreno, Jonathan D. 19nMoschella, Melissa 176nMouffe, Chantal 118n, 123nMueller, Dennis 21 InMulhall, Stephen 17nMunoz-Darde, Veronique 80n

Nagel, Thomas 15, 82n, 108, 119nNaticchia, Chris 8, 10, 16n, 19n,

177-94Neal, Patrick 135, 148n, 150nNeiman, Susan 155, 171 n, 172nNieli, Russ 115, 122nNietzsche, Friedrich 43Noggle, Robert 30nNozick, Robert 1, 2, 5, 15, 16n, 52, 54,

64nNussbaum, Martha C. 19n, 20n, 68,

80n, 81n, 209n

O'Neill, Onora 18n, 153, 154, 17InOkin, Susan Moller 5, 30n, 67, 68, 78,

80n, 81n, 82n, 83-4n, 84n, 193nOlsenjohan 21 OnOlson, Laura R. 148-9nOverall, Christine 84nOwen, David 66n

Paine, Thomas 91Pateman, Carole 80nPaton, H.J. 49n

Perry, Michael 141, 146n, 147n, 148n,150n

Pettit, Philip 16n, 17nPierce, William 115Pippin, Robert B. 170n, 171 nPlato 15,50,64,109,161Pogge, Thomas W. 8, 18n, 19n, 20n,

47n, 191n, 192nPolemarchus 50Prusak, Bernard 81 nPruzan, Peter 21 In

Quine, W. V. O. 208Quinn, Philip L. 149n, 150n

Rao, Neomi 15, 2InRasmussen, David 148nRawls, John 1-16, 17n, 18n, 22-4, 29,

34-47, 50-63, 67-77, 85-95,97-102, 103n, 103-4n, 105n, 107-12, 114, 116, 117n, 118n, 124-33,136, 140, 142-5, 145-6n, 146n, 152,154-65, 167, 168, 170, 172n,177-91,191n, 193n, 195-209, 209n

Razjoseph 77, 81n, 84n, 103n, 121nReath, Andrews 20n, 155, 171n, 175nRegan, Donald H. 170n, 171 nReich, Rob 49nReidy, David 19n, 30n, 120nResnik, David 9, 19nRichards, David A. J. 150nRichardson, Henry 33n, 66nRoemer, John E. 2 OnRorty, Richard 150n, 208, 212nRosen, Stanley 172n, 173nRoss, W. D. 161Rousseau, Jean-Jacques 66nRowlands, Mark 19nRunciman, W.G. 47 n

Sandel, Michael 17n, 111, 115, 118n,120n, 122n, 149n, 150n

Sanders, Lynn 120nSatz, Debra 49nScanlon, T. M. 15, 54, 55, 64n, 65n,

99-100, 105nSchapiro, Tamar 49n, 64n, 66nScheffler, Samuel 82n, 108, 118nSchiller, Friedrich 167Schwartz, Adina 82n

216 Name Index

Schwarzenbach, Sibyl A. 12, 20nSellars, Wilfrid 117Sen, Amartya 64nSenn, Stephen 19nSidgwick, Henry 161Singer, Brent A. 18nSinger, Peter 19nSkorupski, John 81nSmith, Andrew 80nSmolin, David 148n, 149nSolum, Lawrence 128, 147nSterba,James 149nStern, Robert 33nStone, Peter 49nStoutjeffrey 121n, 149n, 150nSunstein, Cass 108, 113-14, 115, 121n,

122n, 204, 210n, 211nSwain, Carol 115, 122nSwaine, Lucas 12InSwift, Adam 17n

Talisse, Robert Basil 7, 107-23Tan, Kok-Chor 19n, 30n, 103nTaylor, Robert S. 5, 16n, 34-49Teson, Fernando 19nThero, Daniel P. 18nThomson, Judith Jarvis 15Thompson, Dennis 108, 119n

Thompson, Janna L. 84nTomasi, John 82nTurner, Lisa 115

Veatch, Robert M. 19n

Waldron, Jeremy 118n, 149nWalker, John D. 81nWeinstock, Daniel 31nWeis, Lael 49nWeithman, Paul 21n, 33n, 141-2, 145n,

147n, 149n, 150nWenar, Leif 2, 10, 16n, 18n, 19n, 22-33,

103n, 192-93nWilliams, Bernard 30n, 47n, 53, 58,

64n, 82n, 197Wolf, Clark 30n, 149nWolfe, Christopher 111, 120n, 146n,

150nWolterstorff, Nicholas 114, 122n, 135,

137-8, 148n, 149n, 149nWood, Allen W. 49n

Young, Iris Marion123n

111, 120n, 122n,

Zimroth, P. L. 2InZipursky, Benjamin C. 20n

SUBJECT INDEX

Aristotelian principle 36, 38-40, 41, 44,45

autonomy 37, 40, 43-5, 46, 89, 152,161, 166-67

basic liberties 25, 26, 37, 42, 52, 110,177, 203

"basic structure of society" 2, 16n, 24,25, 27, 28, 39, 40, 41, 46, 48n, 50, 52,195

bioethics 9ff, 19n, 69burdens of judgement 18n, 75, 130

Categorical Imperative 152—3, 162—5,167-70, 171n, 175n

communitarianism 5, 6, 10, 17ncomprehensive conception of the good

(also "comprehensive doctrines")6, 24ff, 27, 28, 60, 70ff, 72, 79, 80,82n, 85, 86ff, 92, 97, 102, 103n, 105n,109, 119n, 120n, 127-30, 131, 177,179, 182, 186, 191n, 203, 211n

constitutional essentials 6, 7, 14, 20n,24, 203

constructivism 6, 17fn, 161, 163Contractualism 3, 16n, 18n, 18fn, 30,

54f, 56, 61,64n, 65n

decent peoples (also "decent societies")8, 14, 27, 28, 29, 32n, 183, 189

difference principle 3,5,8,10,13,15,22,23,34,35,37,39,42,46,47n, 183,189

difference principle, global 8, 183, 189distinction between persons 4, 50—66distributive justice 5, 9, 10, 34-49, 76,

177, 200,206

eighth amendment 13ffequality 22, 25, 26, 27, 29, 57, 64n, 67,

69, 71, 76, 77, 79, 84n, 94, 116, 130,143, 185, 189, 190, 191n, 202

fair equality of opportunity 3,5,9, 10,16n, 26,34-49, 79, 186

feminism 5, 10, 67-84, 185, 188, 191nfreedom of speech 67 ,77global justice 8, lOff, 16n, 18n, 19n,

27-29, 30n, 177-94Goetzv. Crossnan 13ff, 20n

healthcare 9, 19n, 25human rights 19n, 27, 31n, 32n, 162,

177-81, 183, 190, 191n, 202Hypothetical Imperative 153,163,

165

intuitionism 4, 195, 198-99

law 12-15, 20n, 111, 186, 211nleast advantaged, the (also "worst off")

3, 22, 26, 34, 35, 46, 48nlegitimacy 2, 19n, 23ff, 30n, 58, 62, 72lexical priority 3, 16n, 34—49liberalism, political 2, 4, 6f, 8, 17n, 23ff,

60,63,67,85-106, 136

Martin s. Dugger 13ff, 20nmulticulturalism 10

neutrality 5, 17n, 67-84, 92

Original Position 3, 4, 5, 7, 8, 9f, 13, 14,15, 17n, 18n, 22, 26, 29, 36, 48n, 53,54, 55, 57ff, 59, 65n, 178-9, 181,188-9,201,203,205

original position, global 8, 18n, 29, 83n,177, 178-9, 181, 187, 188-9

outlaw societies 180overlapping consensus 7, 24, 3 In, 58,

86,92,93, 109, 186

perfectionism 43,47Philosopher's Brief 145ff, 21 n

218 Subject Index

pluralism 6, 7, 19n, 24, 28, 30n, 92, 112,139, 193n, 195, 196-200, 205, 210n

political conception ofjustice 6ff, 22,24-6,27,71,75,76, 126, 127, 131

principles ofjustice, two 3, 5, 8, 14, 17n,26, 34, 42, 52, 53, 54, 58, 64n, 65n,72,96, 182, 187, 196,207

priority of liberties 3, 4, 36, 41, 42, 45,47, 47n, 48n

priority of right 36, 42-5, 46, 161, 208procedural justice 13public reason 7, 10, 11, 18n, 25, 29, 30n,

62, 107-51, 195public political culture 3, 6, 24ff, 27, 28,

115, 195publicity 53

reasonable conditions 3, 26, 86, 89ff,164, 193n

reasonable doctrine (also "Conception")6ff, 17n, 25, 26, 3In, 70, 87ff, 89ff,93, 107, 126, 144, 164, 177, 192n,193n, 202

reasonable person 17n, 24, 26, 28, 29,43, 45, 89-91, 94, 98, 104n, 202

reciprocity 25, 29, 62

reflective equilibrium 3, 17n, 202religion 7, 69-71, 73ff, 75, 80n, 82n, 85,

86, 109, 112, 124-51, 182-83, 184,186, 188, 189-90, 191n, 197

Roe v. Wade 15

self-realisation 5, 16n, 34-49self-respect 28, 42, 46sense ofjustice 3, 4, 40, 43stability 7, 22, 23f, 25, 29, 30n, 31n, 58,

64n, 109

toleration 19n, 30n, 60, 72, 86, 107,180, 182, 190

truth 7

United States Supreme Court 14ff, 2 Inutilitarianism 4, 50-66, 195, 198

Vaccov. Quill 14, 20nveil of ignorance 3, 13, 14, 15, 26, 54,

59, 57ff, 178-80, 182,201

Washington v. Glucksberg 14, 20nwell-ordered society 3, 7, 17n, 58, 70,

81n