John Rawls 2007933334, 9780754627135, 9781315251431

Citation preview

John Rawls

International Library of Essays in the History of Social and Political Thought Series Editor: Tom Campbell Titles in the Series: Hannah Arendt Amy Allen

T.H. Green John Morrow

James Madison Terence Ball

Martin Heidegger Stephen Mulhall

Friedrich Hayek Norman Barry

Jean-Jacques Rousseau Timothy O 'Hagan

Charles-Louis de Secondât Montesquieu David Carrithers

Michel Foucault David Owen

Emile Durkheim Roger Cotterrell

John Rawls David A. Reidy

Vilfredo Pareto Joseph Femia

Immanuel Kant Arthur Ripstein

Jean Bodin Julian H. Franklin

Jeremy Bentham Frederick Rosen

David Hume Knud Haakonssen

Theodor Adorno James Schmidt

Edmund Burke Iain Hampsher-Monk

Thomas Hobbes Gabriella Slomp

Talcott Parsons John Holmwood

Friedrich Nietzsche Tracy Strong

Thomas Aquinas John Inglis

Isaiah Berlin Scott Veitch

Aristotle George Klosko Georg Hegel Dudley Knowles Thomas Paine Bruce Kuklick Max Weber Peter Lassman Mary Wollstonecraft Jane Moore

John Rawls

Edited by

David A. Reidy University of Tennessee, Knoxville, USA

Routledge

Taylor & Francis Group

LONDON AND NEW YORK

First published 2008 by Ashgate Publishing Published 2016 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN 711 Third Avenue, New York, NY 10017, USA Routledge is an imprint of the Taylor & Francis Group, an informa business

Copyright © David A. Reidy 2008. For copyright of individual articles please refer to the Acknowledgements. All rights reserved. No part of this book may be reprinted or or utilised in any form or by any electronic, mechanical, or other now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Notice: Product or corporate names may be and are used only for identification infringe.

or to

Wherever possible, these reprints are made from a copy of the original printing, but these can themselves be of very variable quality. Whilst the publisher has made every effort to ensure the quality of the reprint, some variability may inevitably remain. British Library Cataloguing in Publication Data John Rawls - (International library of essays in the history of social and political thought). 1. Rawls, John, 1921-2002 2. Liberalism 3. Justice 4. Political science - Philosophy I. Reidy, David A., 1962320.5T092 Library of Congress Control Number: 2007933334

ISBN 9780754627135 (hbk)

Contents Acknowledgements Series Preface Introduction PART I

THEORIZING JUSTICE

1 Liam B. Murphy (1999), 'Institutions and the Demands of Justice', Philosophy and Public Affairs, 27, pp. 251-91. 2 Joseph Raz (1982), 'The Claims of Reflective Equilibrium', Inquiry, 25, pp. 307-30. 3 Aaron James (2005), 'Constructing Justice for Existing Practice: Rawls and the Status Quo', Philosophy and Public Affairs, 33, pp. 281-316. 4 Jon Mandle (1997), 'Justice, Desert, and Ideal Theory', Social Theory and Practice, 23, pp. 399–425. 5 Sibyl A. Schwarzenbach (1991), 'Rawls, Hegel, and Communitarianism', Political Theory, 19, pp. 539-57. PART II

vii ix xi

3 45 69 105 133

WHAT JUSTICE DEMANDS

6 Norman Daniels (1990), 'Equality of What: Welfare, Resources, or Capabilities?', Philosophy and Phenomenological Research, 1, pp. 273-96. 7 Robert S. Taylor (2003), 'Rawls's Defense of the Priority of Liberty: A Kantian Reconstruction', Philosophy and Public Affairs, 31, pp. 246-71. 8 Thomas Pogge (2004), 'Equal Liberty for All?', Midwest Studies in Philosophy, 28, pp. 266-81. 9 Will Kymlicka (1989), 'Liberal Individualism and Liberal Neutrality', Ethics, 99, pp. 883-905. 10 Andrew D. Williams (1995), 'The Revisionist Difference Principle', Canadian Journal of Philosophy, 25, pp. 257-81. 11 Steven Wall (2003), 'Just Savings and the Difference Principle', Philosophical Studies, 116, pp. 79-102. 12 Samuel Schefifler (2003), 'What is Egalitarianism?', Philosophy and Public Affairs, 31, pp. 5-39. 13 Susan Moller Okin (2004), 'Justice and Gender: An Unfinished Debate', Fordham Law Review, 72, pp. 1537-67.

169 193 219 235 259 285 309 345

vi

John Rawls

PART III A POLITICAL LIBERALISM 14 Joshua Cohen (1994), 'A More Democratic Liberalism', Michigan Law Review, 92, pp. 1503–46. 15 Jeremy Waldron (1994), 'Disagreements about Justice', Pacific Philosophical Quarterly, 75, pp. 372-87. 16 Charles Larmore (1999), 'The Moral Basis of Political Liberalism', The Journal of Philosophy, 96, pp. 599-625. 17 James W. Boettcher (2004), 'What is Reasonableness?', Philosophy and Social Criticism, 30, pp. 597-621. 18 Philip L. Quinn (2001), 'Religious Citizens within the Limits of Public Reason', The Modern Schoolman, 78, pp. 105-24. 19 Micah Schwartzman (2004), 'The Completeness of Public Reason', Politics, Philosophy and Economics, 3, pp. 191-220.

379 423 439 467 493 513

PART IV A LIBERAL FOREIGN POLICY 20 Kok-Chor Tan (2001), 'Critical Notice: John Rawls, The Law of Peoples: With the "Idea of Public Reason Revisited'", Canadian Journal of Philosophy, 31, pp. 113-32. 21 Samuel Freeman (2006), 'The Law of Peoples, Social Cooperation, Human Rights and Distributive Justice', Social Philosophy and Policy, 23, pp. 29-68.

545

Name Index

605

565

Acknowledgements The editor and publishers wish to thank the following for permission to use copyright material. Blackwell Publishing for the essays: Liam B. Murphy (1999), 'Institutions and the Demands of Justice', Philosophy and Public Affairs, 27, pp. 251-91. Copyright © 1999 Princeton University Press; Aaron James (2005), 'Constructing Justice for Existing Practice: Rawls and the Status Quo', Philosophy and Public Affairs, 33, pp. 281-316. Copyright © 2005 Blackwell Publishing, Inc.; Norman Daniels (1990), 'Equality of What: Welfare, Resources, or Capabilities?', Philosophy and Phenomenological Research, 1, pp. 273-96; Robert S. Taylor (2003), 'Rawls's Defense of the Priority of Liberty: A Kantian Reconstruction', Philosophy and Public Affairs, 31, pp. 246-71. Copyright © 2003 Princeton University Press; Thomas Pogge (2004), 'Equal Liberty for All?', Midwest Studies in Philosophy, 28, pp. 266-81; Samuel Scheffler (2003), 'What is Egalitarianism?', Philosophy and Public Affairs, 31, pp. 5-39. Copyright © 2003 Princeton University Press; Jeremy Waldron (1994), 'Disagreements about Justice', Pacific Philosophical Quarterly, 75, pp. 372-87. Copyright © 1994 University of Southern California. Cambridge University Press for the essay: Samuel Freeman (2006), 'The Law of Peoples, Social Cooperation, Human Rights and Distributive Justice', Social Philosophy and Policy, 23, pp. 29-68. Copyright © 2006 Social Philosophy and Policy Foundation published by Cambridge University Press. Canadian Journal of Philosophy for the essays: Andrew D. Williams (1995), 'The Revisionist Difference Principle', Canadian Journal of Philosophy, 25, pp. 257-81; Kok-Chor Tan (2001), 'Critical Notice: John Rawls, The Law of Peoples: With the "Idea of Public Reason Revisited"'', Canadian Journal of Philosophy, 31, pp. 113-32. Copyright Clearance Center for the essays: SibylA. Schwarzenbach(1991), 'Rawls, Hegel, and Communitarianism', Political Theory, 19, pp. 539-57. Copyright © 1991 Sage Publications, Inc.; Will Kymlicka (1989), 'Liberal Individualism and Liberal Neutrality', Ethics, 99, pp. 883-905. Copyright © 1989 University of Chicago; Susan Moller Okin (2004), 'Justice and Gender: An Unfinished Debate', Fordham Law Review, 72, pp. 1537-67; Joshua Cohen (1994), 'A More Democratic Liberalism', Michigan Law Review, 92, pp. 1503^6. The Journal of Philosophy for the essay: Charles Larmore (1999), 'The Moral Basis of Political Liberalism', The Journal of Philosophy, 96, pp. 599-625. Copyright © 1999 The Journal of Philosophy, Inc. Sage Publications for the essays: James W. Boettcher (2004), 'What is Reasonableness?', Philosophy and Social Criticism, 30, pp. 597-621. Copyright © 2004 Sage Publications;

viii

John Rawls

Micah Schwartzman (2004), 'The Completeness of Public Reason', Politics, Philosophy and Economics, 3, pp. 191-220. Copyright © 2004 Sage Publications. Springer Science and Business Media for the essay: Steven Wall (2003), 'Just Savings and the Difference Principle', Philosophical Studies, 116, pp. 79-102. Copyright © 2003 Kluwer Academic Publishers. Taylor & Francis for the essay: Joseph Raz (1982), 'The Claims of Reflective Equilibrium', Inquiry, 25, pp. 307-30. (Taylor & Francis Ltd, http://www.informaworld.com) Social Theory and Practice for the essay: Jon Mandle (1997), 'Justice, Desert, and Ideal Theory', Social Theory and Practice, 23, pp. 399-425. Copyright © 1997 Social Theory and Practice. Every effort has been made to trace all the copyright holders, but if any have been inadvertently overlooked the publishers will be pleased to make the necessary arrangement at the first opportunity.

Series Preface The International Library of Essays in the History of Social and Political Thought brings together collections of important essays dealing with the work of major figures in the history of social and political thought. The aim is to make accessible the complete text with the original pagination of those essays that should be read by all scholars working in that field. In each case, the selection is made from the extensive available literature by an established expert who has a keen sense of the continuing relevance of the history of social and political thought for contemporary theory and practice. The selection is made on the basis of the quality and enduring significance of the essays in question. Every volume has an introduction that places the selection made in the context of the wider literature, the historical period, the contemporary state of scholarship and the editor's particular interests. TOM CAMPBELL Series Editor Centre for Applied Philosophy and Public Ethics (CAPPE) Charles Sturt University Canberra

This page intentionally left blank

Introduction i. John Rawls (1921-2002) stands alone in the latter half of the twentieth century as the preeminent political philosopher of the English-speaking world and as perhaps the only political philosopher of the twentieth century to have produced texts likely to remain canonical for the next century. In the tradition of political philosophy he stands with the likes of Hobbes, Locke, Rousseau, Kant, Hegel, Mill and Marx (though his work did not range so widely across philosophy generally as most of these greats), all of whom we continue to read today. It is, accordingly, entirely appropriate, notwithstanding the fact that Rawls passed away only a few years ago, to publish in a single volume in this series some of the best essays written about his work. I have tried to select essays that will serve to orient the serious student wishing to engage Rawls's work both sympathetically and critically, that is, with understanding. I have tried also to select essays that have significantly shaped recent scholarly debate over Rawls's work or that in my judgment will or should significantly shape such debate in the near future. Selecting the essays was no easy task. The secondary literature on Rawls's political philosophy is enormous. Indeed, a general search of The Philosopher's Index, a commonly used database for bibliographic research in philosophy, shows more than two thousand 'hits' for 'Rawls', roughly the same number as given by searches for 'Hobbes' or 'Rousseau' and considerably more than given by searches for 'Sidgwick', 'Collingwood' or 'Nozick'. Comparable results obtain for searches of Worldwide Political Science Abstracts and other bibliographic databases. This constitutes some evidence for the claim above regarding Rawls's canonical status. But it also suggests the difficulty of identifying some 20 essays that would serve the serious reader in need of a detailed and reliable map into Rawls's thought and the philosophical and political debates it has engendered Happily, I did have the advantage of a few constraints. This volume reproduces only journal essays; it includes no book chapters or excerpts from books. While this means some important works are excluded (for example, Robert Nozick's and Brian Barry's influential criticisms of Rawls, or criticisms pressed in essays written for the collections edited by Norman Daniels, Samuel Freeman or Rex Martin and myself), it does narrow the field from which to select essays for inclusion.1 But, still, there are thousands of journal essays devoted to the study of Rawls's political philosophy. Additional constraints narrowed the field yet further. Two multivolume collections (5 and 4 volumes, respectively) of journal essays on Rawls's work were published in 1999 and 2003 (respectively).21 have tried to avoid reproducing here essays already collected and reproduced 1 See Nozick (1974); Barry (1973); Daniels (ed.) (1975); Freeman (ed.) (2003); Martin and Reidy (eds) (2006). 2 See Richardson and Weithman (eds) (1999); Kukathas (ed.) (2003).

xii

John Rawls

in those volumes. I have not succeeded completely on that front, but I have succeeded to a large extent. In some cases steering clear of essays already collected in these volumes meant only avoiding essays that I would have excluded in any event on other grounds, most often because they are now mainly of historical interest, the debates they initiated either having come to an end or transcended the terms of their initiation (for example, Michael Sandel's well-known essay (1984)). In other cases it meant preferring more recent contributions to an ongoing debate - contributions unavailable to the editors of those other multivolume collections. A few final constraints bear mention. I have selected only essays published in English. I have selected no more than one essay per author and have tried to select essays from a diverse range of journals (17 different journals, including two law reviews). And I have tried to include as much of the best of recent work on Rawls as possible, with an eye towards constructing a volume that will remain useful for some time. I have organized the essays into four parts. The first, titled 'Theorizing Justice,' collects essays concerned with various presuppositions - methodological, epistemological, psychological etc. - to Rawls's political philosophy. The second, titled 'What Justice Demands', collects essays addressed to the substantive content of Rawls's conception of justice, which he names 'justice as fairness'. This content is most fully articulated in Rawls's first book, A Theory of Justice (1971) and remains largely unchanged throughout his later work. The third part, titled 'A Political Liberalism', gathers essays provoked by Rawls's second book, Political Liberalism (1993, 1996) and subsequent essay 'The Idea of Public Reason Revisited'(1997). In Political Liberalism Rawls reformulates the 'congruence argument' given in Part III of A Theory of Justice for the stability of 'justice as fairness'once institutionally embodied. He also develops the important and controversial idea(l) of public reason and liberal principle of political legitimacy. The fourth part, titled 'A Liberal Foreign Policy', includes two essays that examine Rawls's attempt in his third book, The Law of Peoples (1999), to set out a principled international political morality from the point of view of a liberal democratic people. The essays collected here range, then, over the entire core of Rawls's political philosophy. Still, they don't range over all aspects of his thought. They do not address, for example, Rawls's readings of or relationships to the great political philosophers of the past. With the recent publication of Rawls's Lectures on the History of Moral Philosophy (Herman, 2000) and Lectures on the History of Political Philosophy (Freeman, 2007), the secondary literature in this area of Rawls scholarship is likely to grow in coming years. Serious students of Rawls will want to keep an eye on this literature. A complete understanding of Rawls's work will include a good understanding and sound evaluation of Rawls's readings of and relationships to, for example, Rousseau, Hegel and Sidgwick, among others. The essays collected here also leave unexamined many issues that arise as soon as one tries to extend Rawls's political philosophy to specific issues of institutional design or public policy. While this is in some sense unfortunate, the focus of this volume is Rawls's political philosophy. Accordingly, essays the central claims of which are as dependent on contingent institutional facts, shifting demographic conditions, complex statistical analyses, or controversial causal claims as they are on anything one might take away from Rawls's political philosophy simply fall outside the focal scope of this volume. While this volume collects 21 of the best essays on Rawls's political philosophy, and while a great deal of what has been excluded can be excluded without regret, there are nevertheless numerous essays that could not be included that I regret not being able to include: David

John Rawls

xiii

Gauthier (1974) on Rawls's treatment of natural endowments, Norman Daniels (1979) on reflective equilibrium, Amartya Sen ( 1990) on primary goods, Iris Young ( 1989) on citizenship and social unity, G.A. Cohen (1997) on the incentives argument for the difference principle and Rawls's institutionalist approach to justice, Kok-Chor Tan's (2004) reply to Cohen, Thomas Pogge (1994) on Rawls's liberal foreign policy, William Galston (1989) on liberal state neutrality and toleration, Kent Greenawalt (1994) on public reason, Leif Wenar (2004) on the unity and continuity of Rawls's work over time, and so on (this list is by no means exhaustive). In a volume like this one, it is inevitable, though still regrettable, that essays fully deserving of inclusion will be excluded. I can only hope that my selections, taken all together, seem reasonable in light of the merits of the essays included, the purpose of this volume, and the various constraints I faced as editor mentioned above.

II. John (Jack) Bordley Rawls was born 21 February 1921, in Baltimore.3 His father was a selftaught and successful lawyer. His mother was an artist and politically active homemaker. Jack was the second of five sons. Two of his younger brothers died, however, while he was a child. Both died from diseases Jack unintentionally passed on to them, the first when Jack was just seven years old, the second when he was just eight. Later in life while on an academic leave at Stanford University Rawls became friendly with a psychotherapist who informally diagnosed him as suffering from a deep sense of guilt for the death of his two brothers. Rawls neither affirmed nor denied the informal diagnosis, instead allowing only that the deaths no doubt affected him profoundly. Rawls attended mostly private elementary and secondary schools and did very well. The high school he attended, a boarding school, was quite religious (High Church Episcopal). Though he apparently did not enjoy his time at the school, he did well. The reason he did not enjoy his time had less to do with the religious orientation of the school and more to do with the lack of academic challenge and intellectual and personal freedom. Rawls entered Princeton University in 1939. After trying several subjects as possible majors, he eventually settled into philosophy. Walter Stace (a British empiricist with utilitarian leanings - despite an early book on Hegel - who had had a religious experience as a young man and retained a lifelong interest in mysticism) and Norman Malcolm (who had studied under Wittgenstein) were among his teachers there. Slowly Rawls acquired a philosophical interest in religion and religious questions. He had always been conventionally religious and even contemplated eventually entering the seminary. While an undergraduate philosophy major at Princeton, his interest in religion deepened and took a distinctively philosophical-theological turn. His senior thesis, 'A Brief Inquiry into Sin and the Meaning of Faith: An Interpretation Based on the Concept of Community', is a deeply thoughtful 170-page examination of the idea of sin as the refusal of, and faith as an openness to, the interpersonal spiritual core, the living community of all persons, including God, at the heart of the universe. 3 Biographical information in this section is drawn from Chapter 1 of Thomas Pogge's 2007 biographical sketch; Samuel Freeman's discussion of Rawls's life in the Appendices to his 2007b work; John Rawls's 1991 interview; and materials held in Rawls's archived essays at Pusey Library, Harvard University.

xiv

John Rawls

Rawls elected to graduate early (summa cum laude in philosophy) from Princeton so as to enlist in the army to fight in the Second World War. He served from 1943 to 1946 and spent two years in the Pacific Theatre as an infantryman, earning a Bronze Star. Several experiences during the war - the Holocaust, the tragic and often senseless loss of good men, a chaplain urging soldiers to pray for God to guide their bullets, Hiroshima, and so on - caused Rawls to abandon the theistic idea of a personal God to whom one might pray and any lingering thoughts about entering the seminary. He did not abandon, however, his commitment to the idea of an interpersonal spiritual core, a community of persons at the heart of our experience in the universe. He described himself later in life as a non-theist. In 1946 Rawls entered the graduate program in philosophy at Princeton. He subsequently spent a year at Cornell, with Malcolm (who had since moved to Cornell) and Max Black, both of whom were working on Wittgenstein's philosophy. Upon returning to Princeton, Rawls met and then married his wife, then Margaret Fox. She, like Rawls's mother, was a highly intelligent portrait artist who would come to be politically active. She and Rawls shared a deep interest in and appreciation of art and art history. (Rawls himself painted - mainly in the abstract expressionist style - until the early 1960s; he cited Pollock and de Kooning as his artistic héros.) In 1950 Rawls completed his dissertation, 'A Study in the Grounds of Ethical Knowledge: Considered with Reference to Judgments on the Moral Worth of Character', with Stace, with whom he had also worked on his undergraduate thesis, as his director. From 1950 to 1952 Rawls taught as an instructor at Princeton. He began to study outside philosophy: economics, game theory, and the history of the United States, especially constitutional history. He became uncertain as to whether and how he might contribute to philosophy. At the urging of J.O. Urmson, the Oxford philosopher who had visited Princeton in 1951, Rawls applied for and won a Fulbright fellowship and spent 1952-53 at Oxford. This was perhaps the most important year in his intellectual life. He found himself reoriented in philosophy and set on the path that would eventually lead to his life's work. This could be significantly explained by his rich intellectual and friendly relations with H.L.A. Hart, Isaiah Berlin, Stuart Hampshire, Gilbert Ryle and others central to Oxford philosophy during a very productive period there. Several key ideas seem to have germinated during the year at Oxford, including the idea of modelling not just social cooperation but also our moral reasoning about the principles that ought to govern social cooperation as rule-governed games the rules of which must be acceptable to all players from a common and acceptable point of view prior to any particular play of the game. This idea, once connected to the idea already present in the dissertation that there is no foundationalist justificatory path to such constitutive principles, eventually took Rawls to his own reconstructed social contract doctrine. According to this doctrine, we, you and I justify to each other the constitutive principles of the just society we seek by showing first that we would agree to them if we were engaged in a 'game' of moral reasoning (the original position) the rules of which (the veil of ignorance, primary goods) each of us could accept from a common and acceptable point of view. To show that we could each accept from a common and acceptable point of view the rules constitutive of this 'game' of moral reasoning, we show each other that our use of them is proper. This we do by showing that the principles (of justice) agreed to in this 'game' of moral reasoning consistently and coherently order (and improve) our considered judgments about particular cases, that they belong to our moral sense or sense of justice in 'reflective equilibrium'. That, at least, is the aim or hope.

John Rawls

.XV

In 1953 Rawls returned to the United States and spent the next decade and a half systematically working out these ideas and their implications and presuppositions. He initially took a position as Assistant Professor of Philosophy at Cornell and was tenured there in 1956. In 1959 he accepted an invitation to spend a year as a visiting professor at Harvard. In 1960 he joined a fledgling philosophy department at the Massachusetts Institute of Technology (MIT). In 1962 he moved to Harvard as a permanent member of the philosophy department there. He retired in 1991, though he continued to teach for a few years thereafter. During his active professional life he influenced philosophy as much through his teaching and training students, both graduate and undergraduate, as through his publishing.

III. It is easy to (mis)read Rawls's life work as a somewhat idle academic exercise of theoretical reason - a kind of philosophical contemplation of the pure form of justice. But Rawls's life work is as much, perhaps more, a product of the exercise of his practical as his theoretical reason. The theoretical exercise of reason aims at knowing about or contemplating its objects. The practical exercise of reason aims at making or perfecting its objects. Practical philosophy concerns the practical exercise of reason. Rawls devoted his life work to the practical philosophical defense of constitutional liberal democracy against critics from both the right and the left. He did political philosophy with the aim of helping to create, not merely to know or contemplate the form of, a just and stable constitutional liberal democracy. Of course, to make or perfect anything in a reasoned way, one must have some sense of a plan or design at which to aim. And this plan or design ought in the first instance capture what it is we want under favorable but not impossible or wildly unrealistic conditions. Should we find ourselves confronted with less than favorable conditions, we can of course adjust our plan or design accordingly. But to do so in a reasoned way, and always with a sense of what it is we are giving up and a reasoned hope for what we might still achieve, we need some sense of what it is we would aim to make and perfect given more favorable conditions. By showing us that constitutional liberal democracy is what we should and would aim to make and perfect given favorable conditions, that it is something we may and should reasonably hope and thus work for, even as we soldier on under less than favorable conditions, Rawls joined and contributed to all those engaged in the practical task of securing and perfecting constitutional liberal democratic regimes. The task Rawls joined is not just a practical task, it is a historical task. Rawls does not argue for constitutional liberal democracy as that which all and any agents possessed of practical reason should and would aim to make and perfect. Rather, he argues for it as that which we, you and I should and would aim to make and perfect as historically situated agents already possessed not only of practical reason but also of a particular institutional inheritance and normatively rich shared self-understanding. That Rawls's contribution to the practical and historical struggle for constitutional liberal democracy is philosophical in no way diminishes its practical import. Without the kind of reasoned hope for the possibility of a just and stable constitutional liberal democracy to orient their deliberations and actions, contemporary citizens are, under real world conditions, all too susceptible to letting their political disagreements drive them apart, undermine mutual respect and social cooperation, and lead them to either apathetic resignation from or dogmatic

xvi

John Rawls

fanaticism within political life (historically two of the greatest political sources of evil in the world). One might naturally wonder what it is about real world conditions in aspiring liberal democracies that makes this so. But one need not wonder long. All constitutional liberal democracies today face (at least some of the) serious problems faced today by the United States: entrenched economic and social inequalities corrosive of both democracy and selfrespect; deep ethnic, religious, moral and cultural cleavages that threaten political trust and solidarity; a tendency toward individualistic and aggregativist majoritarian conceptions of democracy - vote your pocketbook! vote your religious preferences! - that put many public goods and liberal freedoms at risk and undermine democratic deliberation; a materialist, consumerist background culture that leaves citizens inclined toward libertarian conceptions of the state as a private association and without an adequate understanding or appreciation of civic friendship as a great collective good; and so on. Under real world conditions it is all too easy to think it foolish to hope, let alone work, for a just and stable constitutional liberal democracy, to devote oneself to building out of the status quo a more perfect union. IV.

Part I of this volume takes up several of the novel presuppositions - methodological, epistemological, psychological, and so on - that frame and structure Rawls's political philosophy. The first of these is that the proper first subject of a theory of justice is not interpersonal relations or individual transactions between persons but rather the basic social structure of a society, understood holistically as a complex and rule-governed network of institutions - the economy, government, a system of property, the family, civil society, and so on, taken all together as dynamically interactive. Of course, we may and do properly speak of justice in interpersonal relations and individual transactions. But justice in this context means something different from and must be theorized separately from justice as applied to the basic social structure. It follows that we cannot infer from the fact that an egalitarian principle of justice ought to govern the basic social structure that it ought also to govern interpersonal relations or individual transactions. The latter constitutes a separate inquiry. So while the demands of political justice and interpersonal morality must be consistent and perhaps minimally coherent, they need not be isomorphic or rooted in the same moral soil. To illustrate: Rawls famously holds that the rules giving rise to economic inequalities within the basic social structure of a constitutional liberal democracy must satisfy his egalitarian 'difference principle'. But he also holds that citizens not only must be politically and legally free to follow rules in their interpersonal relations that do not satisfy this principle, they also as a moral matter probably ought not try to follow rules that satisfy this principle in their interpersonal relations. Of course, citizens must internalize the difference principle and the egalitarian commitments it expresses as regulative of their basic social structure. And their morality of interpersonal relations must be at least consistent with their so doing. Otherwise, their society would likely prove unstable without unacceptable coercion. But citizens need not affirm and act from the difference principle in their daily personal life. Liam Murphy (Chapter 1) characterizes Rawls's view in this regard as 'dualist', since it presupposes that institutions and interpersonal conduct constitute two separate subjects of justice, each with their own distinctively appropriate principles irreducible to some common

John Rawls

xvii

master principle. He characterizes his own view as 'monist', since it posits a single subject of justice, ranging across both institutions and interpersonal conduct, to which the most fundamental principles of justice apply (even if second-order principles apply separately to institutions and interpersonal conduct). Drawing on and adding to earlier attacks on Rawlsian dualism by Robert Nozick and G.A. Cohen, Murphy presses the case for monism. He argues that in the end Rawls unjustifiably slides from the truth that securing justice is a collective responsibility best realized in large part through institutional design to the wrong-headed dualist view that the primary and first subject of justice is exclusively the design of the basic social structure. Along the way, he provides readers with an extremely valuable overview of a debate over Rawls's approach to political philosophy that is likely to remain important for some time. A second novel presupposition framing Rawls's political philosophy is that principles of justice can be neither deductively (or otherwise inferentially) justified from self-evident foundational premises nor justified themselves as self-evident and foundational. Instead, they can be justified only by showing that they coherently order our considered judgments at various (ideally at all) levels of abstraction and across diverse (again, ideally all) subjects, and thus genuinely belong to our cultivated moral sense or sense of justice. Once citizens can publicly defend before one another their basic social structure by reference to principles in 'reflective equilibrium' with the full range of their other considered judgments, no further public justification is possible or, as a practical matter, necessary. Rawls proposes that we approach this practical ideal of full public justification by first seeking out as individuals principles that coherently order our considered judgments in specific and fairly narrowly circumscribed domains (for example, tax policy or punishment), and then slowly extending the process to wider and wider domains (political judgments generally, then all moral judgments, and so forth). Eventually, we must reason publicly with others in search of common principles of justice capable of coherently ordering our considered judgments across as wide a domain of our common experience as possible. For in practical political philosophy justification is always justification to another. These steps toward what Rawls calls 'full reflective equilibrium', toward a full public justification of principles of justice, are best understood as analytically rather than temporally distinct, for we may and properly do engage in them all simultaneously and in an ongoing way (with lots of feedback loops, and so on, in our reasoning). Joseph Raz (Chapter 2) wrestles with Rawls's idea of reflective equilibrium, proposing various interpretations of and lodging various objections to it, interpretations and objections that have continued to inform debate over the meaning and philosophical soundness of reflective equilibrium as a method of inquiry and standard of justification in political philosophy, or normative moral theory more generally. On Raz's view, reflective equilibrium is at best a method for arriving at, or a standard for assessing, complete descriptions or explanations of our moral beliefs. It contributes little to nothing to the justification of our moral beliefs or to our normative reasoning when making moral decisions. Raz's essay makes it clear that we cannot determine the nature, import and plausibility of reflective equilibrium without getting clear also about the aims and purposes of the project of which it is a part. It invites readers to think hard about the latter to try to make sense of the former. Aaron James (Chapter 3) argues that Rawls, both in his early and later work and contrary to initial appearances, does not aim to deliver from the mountaintop of pure reason the blueprint

xviii

John Rawls

of a just society. Rather, he aims to exercise his practical reason and clarify his sense of justice, and to stimulate us to join with him in this undertaking, by constructing out of his and our institutional inheritance, public culture and historically shaped self-understandings, an ideal present but unrealized within the status quo of our existing political practice, a realistic ideal for which we might and should struggle, exercising our practical reason and guided by our sense of justice. On James's account, Rawls's philosophical method is neither purely descriptive nor purely prescriptive. It is rather both; it is interpretive. James raises important questions not only about the role of history and a sometimes alleged conservatism in Rawls's political philosophy but also about the unity of Rawls's thought over time. Many scholars would affirm James's characterization of Rawls's approach with respect to his later work but deny it with respect to his earlier work. On James's view, there is more unity and continuity to Rawls's thinking than often meets the eye at first glance. The extent to which Rawls's political philosophy constitutes a unified, continuous whole will be a key question for decades to come in Rawls scholarship. A striking feature of Rawls's political philosophy is his apparent severing of justice from desert. After all, the root requirement of justice has long been said to be that each should get what he or she deserves. Yet Rawls appears not to ground his conception of justice in desert. This feature of Rawls's political philosophy is explored by Jon Mandle (Chapter 4). Here Mandle rejects a common reading of Rawls. According to it, Rawls holds that, taken apart from and prior to any basic social structure, people do not deserve their natural talents or initial starting points in life and thus do not deserve any of the goods that might flow in social and economic life from them. A just basic social structure then cannot distribute goods to people according to their desert since people do not deserve the one thing they bring to the basic social structure, their natural endowments or talents. This common reading of Rawls has given rise to debate over Rawls's conception of persons, his alleged treatment of natural endowments and talents as a 'common asset', and his supposed fundamental commitment to a distributive principle that is fully 'endowment insensitive and choice sensitive'. Mandle argues that this common reading is a misreading of Rawls and his understanding of the relationship between justice and desert and that the debates it has birthed are not wellconceived. He argues that what Rawls holds is that a just basic social structure cannot be defined in terms of distributing goods to persons according to their moral desert. This is because we cannot know a person's moral desert without knowing whether she has satisfied her moral obligations. But to know whether a person has satisfied her moral obligations we must know what her obligations are. Obligations (in contrast to natural duties) arise only in the context of reasonably just institutions or practices. Thus, to determine whether she has satisfied her obligations, we must first determine the justice of the institutions or practices she participates in. But for this we need principles of justice to assess those institutions or practices. Thus, we cannot define a just basic social structure as one that distributes goods to persons based on their moral desert. For we cannot know a person's moral desert without knowing first whether her basic social structure is reasonably just. This point, central to Rawls's approach to justice, has nothing to do with how Rawls conceives of persons metaphysically or the nature of a person's claim to their own natural endowments or talents. Mandle's essay serves, inter alia, as a cautionary reminder: not every debate over Rawls's political philosophy has been well motivated by Rawls's arguments. A good deal of Rawls scholarship in the coming decades

John Rawls

xix

will involve clearing away the underbrush of poorly motivated debates that somehow gained enough inertia to occupy many journal pages over many years. Rawls's political thought is standardly characterized as Kantian. And indeed in many respects it is. Yet it is easy to overread the influence of Kant on Rawls's thought. Rawls draws from a wide range of historical influences, from Aristotle to American pragmatism. Sibyl Schwarzenbach (Chapter 5) brings out several important Hegelian themes within Rawls's political philosophy. Of course, unlike Hegel, Rawls offers, and by Political Liberalism explicitly argues for, a political philosophy more or less detached from any particular and comprehensive metaphysical system. Nevertheless, like Hegel, Rawls thinks political philosophy properly aims at reconciling us to the real or actual by using reason to overcome or transcend the deep cultural conflicts, the practical impasses, that mark our historical moment. And like Hegel, Rawls begins with a political conception of the person both drawn from post-Reformation modernity and (contrary to many critics) innocent of the dualisms that characterize Kant's thought. Finally, like Hegel, Rawls offers a rich conception of community, or, as Rawls prefers, social union, as both intrinsically good and the necessary context for the full development of individuality. Schwarzenbach makes a powerful case against the atomistic, individualistic, acquisitivist reading of Rawls that so many communitarian criticisms rest on, and she invites readers of Rawls to rethink common views about the 'tradition' of political philosophy to which he belongs. V.

The essays collected in Part II concern key claims by Rawls regarding the substantive demands of justice. These demands Rawls articulates most generally in two principles. These principles govern the rules constitutive of the basic social structure. The first Rawls calls the liberty principle. It holds that 'each person has the same indefeasible claim to a fully adequate scheme of equal basic liberties'[...] 'compatible with the same scheme of liberties for all'. The second, known as the difference principle, allows social and economic inequalities so long as they satisfy two conditions: 'first, they are attached to offices and positions open to all under conditions of fair equality of opportunity; and second they are [...] to the greatest benefit of the least-advantaged members of society'.4 Rawls assigns the first principle 'lexical priority' over the second, and the first half of the second principle (concerned with fair equality of opportunity) lexical priority over the second half (concerned to ensure that economic inequalities work to the maximal advantage of those advantaged least by them). To say a principle has lexical priority is to say that we ought to satisfy it before others and that once we satisfy it we ought not violate or encroach on it for the sake of satisfying other principles lexically ranked lower. These principles evidently express a certain (complex) egalitarian conviction about the demands of justice. Roughly, they hold that a just basic social structure will distribute certain key resources - the primary goods: liberties, opportunities, wealth and income - equally unless an unequal distribution would be acceptable from the point of view of those who would gain the least from it. Norman Daniels (Chapter 6) assesses various criticisms of Rawls's appeal to primary goods or basic resources as the appropriate metric for determining whether 4

For this statement of the two principles, see Rawls (2001) p. 42,

XX

John Rawls

the basic social structure satisfies our egalitarian convictions about justice. These criticisms, levelled by G.A. Cohen, Richard Arneson, Amartya Sen and others, share a common root idea. Because primary goods are merely resources or means to what we ultimately care about - whether welfare or self-realization or something else - and because persons will differ in their ability to convert primary goods into whatever it is we ultimately care about, an egalitarian distribution of primary goods is not the same thing as, and may fall short of, an egalitarian distribution of whatever it is we most care about and may thus fail to satisfy our egalitarian convictions. Daniels defends Rawls against these criticisms. Along the way he sheds important light on Rawls's egalitarian commitments, a topic explored more fully by Samuel Sheffler (Chapter 12). Sheffler addresses a (mis)reading of Rawls already suggested at several points above, namely that a just basic social structure is one that distributes goods to persons based solely on their choices and that immunizes persons from bad luck. As Elizabeth Anderson, Susan Hurley and others have argued, this view, dubbed 'luck egalitarianism', is arguably problematic on its own terms. But it is especially problematic as a reading of Rawls. Sheffler explores the various ways in which luck egalitarianism fails both to fit Rawls's texts and arguments and, regardless of pedigree, on its own terms. In the process he invites readers to rethink what we mean when we say that Rawls is, or that as liberals we are, committed to both freedom and equality as fundamental political values. A central feature of Rawls's conception of justice is the lexical priority he assigns to his liberty principle. Indeed, this feature constitutes one of the most important distinctions between Rawls's conception of justice and utilitarian, perfectionist or intuitionist alternatives (the main alternatives against which he argues in A Theory of Justice). Robert Taylor (Chapter 7) reviews several objections to the lexical priority Rawls assigns his liberty principle. He is led by these to reject two of Rawls's arguments for the lexical priority of liberty. But he finds more plausible a third argument, one rooted in Kantian conceptions of the person and autonomy. However, on Taylor's view, this argument presupposes the Kantian metaphysics allegedly built into A Theory of Justice, a metaphysics Rawls explicitly sets to the side when he recasts 'justice as fairness' as 'political, not metaphysical' in Political Liberalism. Taylor raises fundamental questions not only about whether Rawls can justify the lexical priority he assigns his liberty principle, a signal substantive demand of justice on Rawls's view, but also about whether the substantive demands of justice generally on Rawls's view can be sustained within the metaphysically thin context of his later reconstructed stability or congruence argument. Thomas Pogge (Chapter 8) focuses on a second aspect of the lexical priority Rawls assigns his liberty principle. Rawls allows liberties to be constrained only where so doing strengthens the overall system of liberty for all, and he allows them to be constrained unequally only when acceptable to those given lesser liberty. Focusing on the principles that ought to govern criminal law, a paradigmatic constraint on liberty, Pogge argues that Rawls's liberty principle ultimately yields a system of liberty a good deal closer to that associated with utilitarianism (it even allows for the punishment of innocents, on Pogge's view) and at odds with many of our considered judgments. In an age in which liberty and security seem increasingly in tension, Pogge's essay invites us to think hard about whether and how Rawls's conception of justice might profitably frame our deliberations over how best to structure a fully adequate system of liberty.

John Rawls

xxi

A familiar criticism of Rawls's view of the substantive demands of justice is that, by assigning priority to a fully adequate system of liberties, it embraces a liberal commitment to state neutrality with respect to the value or worth of various ways of life citizens might pursue, a neutrality that is neither possible nor attractive. Will Kymlicka (Chapter 9) assesses this line of criticism and in the process deepens our understanding of Rawls's commitment to liberal state neutrality. Kymlicka also draws our attention to an important but arguably unstable distinction in Rawls's thought between the public political culture within which principles of justice regarding the basic social structure are articulated and defended, and the background culture of civil society within which the merits of various ways of life are tested, challenged, displayed, and so on. Rawls's second principle of justice concerns both fair equality of opportunity and unequal distributions of wealth and income in the economy. Andrew Williams (Chapter 10) examines several interpretations of the latter half of this requirement of economic justice. He criticizes one important and attractive interpretation in particular, developed by Rex Martin, according to which a just economy is one that tends towards the most egalitarian of all possible Paretoefficient distributions of wealth and income. Along the way Williams raises fundamental questions about both the content of and justification for Rawls's difference principle. Readers should know that this is a matter that Rawls continued to struggle with; in Rawls 2001, he makes important revisions to his argument for the difference principle. Rawls's principles of justice, including the difference principle, are meant to govern a society's basic social structure in perpetuity. Thus, they necessarily presuppose some notion of just savings between generations. This Rawls recognized from the beginning. But over time Rawls revised both the content and justification of his principle of just savings. Steven Wall (Chapter 11) outlines the changes in Rawls's view of just savings and then argues that in the end Rawls is saddled with an insurmountable tension between the demands of just savings and of the difference principle. He proposes to resolve this tension not by limiting the former to intergenerational and the latter to intragenerational justice, as if these were separate domains of justice, but rather by rejecting the difference principle construed as an egalitarian principle aimed at minimizing inequalities (subject to Pareto efficiency considerations) in favor of a less egalitarian principle of distributive justice (to apply to contemporaries in perpetuity) that privileges benefits to the least well-off in absolute terms. In the final essay of Part II, Susan Okin, a leading and in many respects sympathetic feminist critic of Rawls's political philosophy, surveys (in Chapter 13) a range of feminist criticisms of Rawls's work and then engages Rawls's response to particular feminist criticisms regarding his treatment of gender and the family. Roughly, the central objection is that Rawls's principles of justice leave unaddressed and unchallenged, indeed in some ways tacitly presuppose, the social reproduction of gender hierarchy within the family and other segments of civil society. Since gender hierarchy is a form of injustice, Rawls's principles of justice fall short of the mark. There can be no doubt that Rawls was himself committed to securing justice for women. Whether his theory and principles of justice fulfil, or might without radical revision fulfil, that commitment remains one of the more pressing questions in political philosophy.

xxii

John Rawls

VI.

In Part III of A Theory of Justice Rawls argued that in a society governed by his two principles of justice citizens would eventually find themselves socialized by the general pattern of institutional life into a kind of comprehensive or partially comprehensive Kantian moral doctrine according to which doing justice would be part of their own good. Justice as fairness would therefore generate its own support, it would be stable in the right way. By 1980 Rawls found this argument unsatisfactory. The problem was that in a society governed by his two principles, it was inevitable that there would always be a plurality of reasonable comprehensive or partially comprehensive moral, religious and philosophical doctrines, many anti-Kantian or at least non-Kantian. The stability argument from Part III of A Theory of Justice had to be reworked. Reworking that argument and following out the implications of various changes introduced occupied Rawls for most of the next two decades, from 1980 to 2000. The upshot of his thought here is most completely expressed in Rawls, Political Liberalism (1993, 1996) and Rawls, 'The Idea of Public Reason Revisited' (1997). The essays constituting Part III of this volume address themselves to this aspect of Rawls's political philosophy. Joshua Cohen, one of the most subtle and penetrating of Rawls's interpreters, reviews (in Chapter 14) Political Liberalism and its most significant doctrines (the ideas of reasonable pluralism, of a freestanding political conception, of public reason, of an overlapping consensus, etc.) shortly after its publication. It constitutes an invaluable point of departure for entry into the debates over Rawls's later work. Jeremy Waldron (Chapter 15) extends Rawls's idea that certain burdens of judgment lead inevitably in a free society to reasonable disagreement over conceptions of the good and comprehensive or partially comprehensive moral, religious and philosophical doctrines to the idea that they lead inevitably also to reasonable disagreement over conceptions of justice themselves. But if citizens reasonably disagree over conceptions of justice, then it is hard to see how any exercise of philosophical reason, as opposed to a contingent political settlement, can stabilize a free society by underwriting an overlapping consensus over any particular conception of justice. In this essay, Waldron raises a question that continues to trouble readers of Rawls. Given what Rawls says about reasonable disagreement in a free society, why isn't it 'reasonable disagreement all the way down', to including conceptions of justice? And if it is, how can there be any substantive basis to social unity in a free society? Questions of the sort pressed by Waldron have caused many readers to wonder about the substantive moral basis of Rawls's political liberalism. What, if anything, still stands in the face of the reasonable disagreements so characteristic of modernity? Charles Larmore (Chapter 16) argues that it is a moral commitment to respect for persons in political life where the exercise of coercive force is always on the horizon. Using the context of a wellknown exchange between Rawls and Habermas in the Journal of Philosophy, Larmore argues that both Rawls and Habermas miss this common basis of their political philosophies. It is this commitment that compels us to search for some way around our many reasonable disagreements to principles we might all reasonably affirm as the political basis of our shared social life. Of course, being morally compelled to undertake the search does not by itself guarantee success. But it does suggest the possibility of success, for so long as we share this moral commitment it is not reasonable disagreement for us 'all the way down'. If this is right,

John Rawls

xxiii

though, then Rawls's political liberalism is a bit less 'freestanding' than Rawls sometimes characterizes it. It presupposes a deep moral commitment to respect for persons in political life, a contingent and historical commitment constituting the 'we' for whom reasonable disagreement is a challenge in political life. James Boettcher (Chapter 17) steps back from the discussions advanced by Cohen, Waldron and Larmore to inquire into Rawls's fundamental idea(s) of the reasonable. This complex idea has seemed to many to be little more than Rawls's way of referring in a veiled and ad hoc manner to the authority of his own intuitions. Reconstructing Rawls's account of the reasonable in its many faces or applications, Boettcher defends Rawls against this charge. By so doing, he offers one line of response to Waldron's criticisms and adds to the line of analysis developed by Larmore. Reasonable citizens will, on Rawls's view, endeavor to organize themselves in political life on terms that can be publicly justified to all reasonable citizens as free equals. This means, given the reality of reasonable disagreement over comprehensive doctrines, including religious doctrines, they will refrain from trying to organize themselves in political life on terms that can be justified only from within a comprehensive or partially comprehensive doctrine some reasonable citizens reasonably reject. This has proved a controversial claim that has drawn many critics, including Nicholas Wolterstorff, Chris Eberle, Jeffrey Stout and others. Philip Quinn (Chapter 18) defends Rawls's idea(l) of 'public reason' as 'the best attempt so far to work out in detail a liberal political theory according to which the costs liberal regimes should impose on religious citizens are, though real, close to minimal'. Micah Schwartzman (Chapter 19) takes up another line of objection to Rawls's idea(l) of public reason. The objection is that, once the resources of comprehensive or partially comprehensive doctrines are set to the side, citizens restricted to public reason will find themselves either unable to reason collectively to agreement, or, more radically, to reason individually to determinate judgments, over basic political questions. Schwartzman defends the Rawlsian idea(l) against both strands of this objection, concluding that while public reasons cannot by themselves settle for us collectively or individually some of our most intractable moral controversies, this constitutes no good reason to retreat to a politics of competing nonpublic reasons. There are ways to politically settle such controversies consistent with the idea(l) of public reason, and, given the values at stake, on Schwartzman's view, we ought to avail ourselves of these before turning to a politics of non-public reasons.

VII. Part IV of this volume collects just two essays addressed to Rawls's third book The Law of Peoples (1999). In this book Rawls undertakes to articulate and defend the principles that ought to govern a liberal democratic people as it interacts on the international stage with other decent peoples, whether liberal and democratic or not, as well as with outlaw regimes, burdened societies and benevolent absolutisms. Many readers of Rawls were surprised and disappointed by the views he sets forth in The Law of Peoples: his insistence on liberal toleration of and respect for non-liberal and non-democratic but decent, well-ordered peoples; his failure to affirm a global difference principle as a principle of global economic justice; his conception of basic human rights; and his framing the problem to which he addresses himself

xxiv

John Rawls

as one of international rather than global justice, thereby apparently assuming rather than arguing for an internationalist system of states. Rawls had first given an Oxford Amnesty Lecture, subsequently published as an essay in 1993, titled 'The Law of Peoples'. It attracted important critical attention from Thomas Pogge, Fernando Tesón, Charles Beitz, Allen Buchanan and others. But Rawls's final statement was not made until 1999 in the book of the same title, after Rawls had had time to digest criticisms. He largely stuck to his guns. The most significant revision was to incorporate explicitly into the law of peoples a duty of humanitarian assistance. Kok-Chor Tan's essay (Chapter 20) is among the best of the early and influential criticisms of Rawls's The Law of Peoples. Tan argues that Rawls's own liberal premises, properly understood, lead to a more robustly cosmopolitan conception of global justice, one in which non-liberal and non-democratic peoples do not enjoy respect as members in good standing in the international order and in which global economic inequalities are governed by a principle more demanding than Rawls's humanitarian duty of assistance. Samuel Freeman (Chapter 21) both defends Rawls's approach and conclusions in The Law of Peoples and criticizes the approach and conclusions of his critics, including Tan. Freeman argues that Rawls is correct to see a fundamental distinction between the domestic and international cases and to give priority to the former, as the basic or paradigmatic case of social cooperation, in his theorizing of justice. He defends Rawls's conception of human rights as both more demanding and better suited to the role of public reason within international politics than critics have allowed. And he explains Rawls's reasons for insisting on toleration of decent well-ordered peoples, even if non-liberal or non-democratic, and for rejecting a global difference principle in favor of a duty of assistance, understood as a duty of justice. Freeman's essay invites readers to think about why it is that well-ordered social cooperation is normatively significant (commands respect) both for its members and for non-participant outsiders, even when it is neither liberal nor democratic in nature. VIII. Taken all together the essays in this volume constitute, in my judgment, an invaluable starting point for serious study of Rawls's political philosophy. Readers eager to push beyond these essays can of course read into the literature and authors they cite. Of course, not everything is or could be covered. I have included little devoted to critical assessment of the reasoning of agents in the original position or other technical features of Rawls's original position argument for his principles of justice. This is in part because Rawls came to rely less and less on that argument, judging it to be just one way of organizing the fundamental idea(l)s out of which his conception of justice is constructed: the idea(l)s of social cooperation on fair terms, of citizens as the free and equal co-authors of those terms, and of their society as well-ordered. The basic question for Rawls, and for us, is what our commitment to these idea(l)s practically means for us.

John Rawls

XXV

References Barry, B. (1973), The Liberal Theory of Justice: A Critical Examination of the Principal Doctrines in 'A Theory of Justice ' by John Rawls, Oxford: Oxford University Press. Cohen, G.A. (1997), 'Where the Action Is: On the Site of Distributive Justice', Philosophy and Public Affairs, 26.1, pp. 3-30. Daniels, N. (ed.) (1975), Reading Rawls: Critical Studies on Rawls s 'A Theory of Justice ', Palo Alto, CA: Stanford University Press. Daniels, N. (1979), 'Wide Reflective Equilibriumand Theory Acceptance in Ethics', Journal of Philosophy, 76.5, pp. 256-82. Freeman, S. (ed.) (2003), The Cambridge Companion to Rawls, Cambridge: Cambridge University Press. Freeman, S. (ed.) (2007a), Lectures on the History of Political Philosophy, Cambridge, MA: Harvard University Press. Freeman, S. (ed.) (2007b), Justice and the Social Contract: Essays on Rawlsian Political Philosophy, Oxford: Oxford University Press. Galston, W (1989), 'Pluralism and Social Unity', Ethics, 99A, pp. 711-26. Gauthier, D. (1974), 'Justice and Natural Endowment: Toward a Critique of Rawls's Ideological Framework', Social Theory and Practice, 3, pp. 3-26. Greenawalt, K. (1994), 'On Public Reason', Chicago-Kent Law Review, 69.3, pp. 669-89. Herman, B. (ed.) (2000), John Rawls: Lectures on the History of Moral Philosophy, Cambridge, MA: Harvard University Press. Kukathas, C. (ed.) (2003), John Rawls: Critical Assessments, 4 vols., London: Routledge. Martin, R. and Reidy D. (eds) (2006), Rawls s Law of Peoples: A Realistic Utopia?, Oxford: Blackwell Publishing. Nozick, R. (1974), Anarchy, State and Utopia, New York: Basic Books. Pogge, T. (1994), 'An Egalitarian Law of Peoples', Philosophy and Public Affairs, 23.3, pp. 195-224. Pogge, T. (2007), John Rawls: His Life and Theory of Justice, Oxford: Oxford University Press. Rawls, J. (1971), A Theory of Justice, Cambridge, MA: Harvard University Press. (1991), 'John Rawls: For the Record - An Interview', in the Harvard Review of Philosophy, 1, pp. 38-47. Rawls, J. (1993), 'The Law of Peoples', in Stephen Shute and Susan Hurley (eds) On Human Rights, New York: Basic Books. Rawls, J. (1993, 1996), Political Liberalism (revised, expanded paperback), New York: Columbia University Press Rawls, J. (1997), 'The Idea of Public Reason Revisited', University of Chicago Law Review, 64, pp. 765-807. Rawls, J. (1999), The Law of Peoples, Cambridge, MA: Harvard University Press. Rawls, J. (2001), Justice as Fairness: A Restatement (E. Kelly éd.), Cambridge, MA: Harvard University Press. Richardson, H. and Weithman, P. (eds) (1999), The Philosophy of Rawls: A Collection of Essays, 5 vols., New York and London: Garland Publishing. Sandel, M. (1984), 'The Procedural Republic and the Unencumbered Self, Political Theory, 12.1, pp. 81-96. Sen A. (1990), 'Justice: Means versus Freedoms', Philosophy and Public Affairs, 19.2, pp. 111-21. Tan, K.-C. (2004), 'Justice and Personal Pursuits', Journal of Philosophy, CI, no. 7, pp. 331-62. Wenar, L. (2004), 'The Unity of Rawls's Work', Journal of Moral Philosophy, 1.3, pp. 265-75. Young, I. (1989), 'Polity and Group Difference: A Critique of the Ideal of Universal Citizenship', Ethics, 99.2, pp. 250-74.

This page intentionally left blank

Parti Theorizing Justice

This page intentionally left blank

[1] Institutions and the Demands of Justice LIAM B. MURPHY

I. INTRODUCTION In the first sentence of the first section of A Theory of Justice Rawls writes that "justice is the first virtue of social institutions." He soon elaborates: 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 fundamental rights and duties and determine the division of advantages from social cooperation. Somewhat later in the book, we find this crucial passage: The principles of justice for institutions must not be confused with the principles which apply to individuals and their actions in particular circumstances. These two kinds of principles apply to different subjects and must be discussed separately1 In this article I defend the contrary view: all fundamental normative -principles that apply to the design of institutions apply also to the conduct of people. I have benefitted enormously from the comments of many people, including members of audiences at Brown University; the University of Maryland, College Park; New York University; Tufts University; the University of Virginia; and a panel at the 1998 Annual Meeting of the American Political Science Association. I am particularly indebted to Richard Arneson, G. A. Cohen, Ronald Dworkin, Lewis Kornhauser, Thomas Nagel, Thomas Pogge, Joseph Raz, Carlos Rosenkrantz, and the Editors of Philosophy 6» Public Affairs. The financial support of the Filomen d'Agostino and Max E. Greenberg Research Fund of the New York University School of Law is gratefully acknowledged. i. Cambridge, Mass.: Harvard University Press, 1971. The three passages are found at pp. 3, 7, and 54-55-

4

John Rawls

252

Philosophy Public Affairs

Like all of his philosophical thought, Rawls's conception of justice as a normative realm distinct from that of ethics or morality has been extremely influential. It is true that the general idea that distributive justice must be promoted through the structural reform of society rather than individuals' do-gooding is not new with Rawls; it has long been taken for granted by pretty much everyone with egalitarian sympathies. Likewise, Rawls is not the first philosopher to discuss the importance of social institutions in forming people's interests, character, and sense of justice.2 But Rawls offers a novel philosophical interpretation of the role of institutions that has been very widely accepted. For Rawls, the significance of institutions is not just causal; in a fundamental way, institutions are what normative political theory is all about. It is only this idea of the foundational role of institutions in normative political theory that I question. Thus Rawls is surely right to highlight the formative significance of institutions.3 And it is obviously true that, as a practical matter, it is overwhelmingly preferable that justice be promoted through institutional reform rather than through the uncoordinated efforts of individuals—a point worth emphasizing in an era characterized by the state's abandonment of its responsibility to secure even minimal economic justice and by politicians' embrace of "volunteerism" as a supposed substitute.4 Though Rawls's conception of the difference between political justice and personal ethics now seems very much the mainstream view, it is a 2. See, e.g., A Theory of Justice, §69. 3. As Rawls says, "everyone recognizes that the institutional form of society affects its members and determines in large part the kind of persons they want to be as well as the kind of persons they are" ("The Basic Structure as Subject," in Political Liberalism [New York: Columbia University Press, 1993], p. 269). 4. I refer primarily to the United States in the wake of the Personal Responsibility and Work Opportunity Reconciliation Act of 1996, which was signed by President Bill Clinton on August 22,1996. When addressing a rally of volunteers at the Presidents' Summit for America's Future in Philadelphia on April 27,1997 (where he was joined by former Presidents Jimmy Carter and George Bush as well as General Colin Powell), President Clinton had this to say: "We're still losing too many kids to crime, to drugs, to not having decent income in their home and to not having a bright future. And we're here because we don't think we have to put up with it, and we believe that together we can change it. Isn't that right?" (Quoted in James Bennet, "At Volunteerism Rally, Leaders Paint Walls and a Picture of Need," New York Times, April 28,1997.). For a thorough explanation and evaluation of the effects of the Personal Responsibility and Work Opportunity Reconciliation Act, see Helen Hershkoff and Stephen Loffredo, The Rights of the Poor (Carbondale and Edwardsville: Southern Illinois University Press, 1997).

John Rawls

253

Institutions and the Demands of Justice

significant innovation in political philosophy.5 The clearest contrast comes from the utilitarians who, despite their intense interest in institutional design, have always used the same criterion to evaluate institutions and personal conduct.6 But I will not defend utilitarianism in particular. I will make the general argument that any plausible overall political/moral view must, at the fundamental level, evaluate the justice of institutions with normative principles that apply also to people's choices. We should not think of legal, political, and other social institutions as together constituting a separate normative realm, requiring separate normative first principles, but rather primarily as the means that people employ the better to achieve their collective political/moral goals.7 Despite the reference to "separate subjects" in the above quotation, 5. In "The Basic Structure as Subject" (first published in 1978), pp. 259-60, Rawls accepts Hugo Adam Bedau's criticism of his claim to the contrary in A Theory of Justice, pp. 10-11; see further Bedau, "Social Justice and Social Institutions," Midwest Studies in Philosophy^ (i978):i59-75. As Bedau says, "[b]ecause institutionalism may now seem so plausible and attractive, we might not realize how relatively novel it is even today" (p. 162). In 1995, Brian Barry writes: "If Rawls had achieved nothing else, he would be important for having taken seriously the idea that the subject of justice is what he calls 'the basic structure of society' Rawls's incorporation of this notion of a social structure into his theory represents the coming of age of liberal political philosophy" (Justice as Impartiality [Oxford: Clarendon Press, 1995], p. 214). An important example of Rawls's approach can be found in the increasing willingness of economic analysts of law to employ a utilitarian criterion (rather than some criterion of efficiency that does not require interpersonal comparisons of utility) for the evaluation of legal rules while ignoring the dramatic implications the utilitarian criterion would have if employed by people; the assumption must be that the criterion does not apply to personal choice. See, for example, Louis Kaplow, "A Fundamental Objection to Tax Equity Norms: A Call for Utilitarianism," National Tax Journal 48 (i995):497-5i46. As Rawls notes in "The Basic Structure as Subject," pp. 260-61. See also Henry Sidgwick, The Methods of Ethics, 7th ed. (London: Macmillan, 1907; repr., Indianapolis: Hackett, 1981), pp. 457-59, and (for a note on Bentham) pp. 87-88. As Rawls also notes, libertarianism, as defended by Robert Nozick in Anarchy, State, and Utopia (New York: Basic Books, 1974), also rejects the idea of special principles for the design of institutions, see Rawls, "The Basic Structure as Subject," pp. 262-65; see also the brief discussion of Nozick in the text below. Many other contemporary opponents of Rawls's idea could be cited; for one example, see Joseph Raz, The Morality of Freedom (Oxford: Clarendon Press, 1986), p. 4 and passim; and see Bedau, "Social Justice and Social Institutions," pp. 162-64, for a survey of some discussions of justice roughly contemporaneous with A Theory of Justice. 7. Since I am arguing that we should see normative political theory as^oritinuous with ethics or morality, terminology becomes a problem. Should we talk here about a person's "moral" goals or her "political" goals? Each is somewhat misleading, as are slogans such as "it's all politics" or "it's all morality."

5

6

John Rawls

254

Philosophy & Public Affairs

Rawls clearly does not believe that the subjects of justice and morality are entirely distinct. Both in A Theory of Justice and Political Liberalism the theory of justice is treated as part of moral theory, where the latter is understood broadly to include, at least, principles appropriate for the design of institutions and principles appropriate for the guidance of personal action.8 So the two subjects Rawls has in mind are both moral subjects in a broad sense. I am interested in the specific claim that the two practical problems of institutional design and personal conduct require, at the fundamental level, two different kinds of practical principle. I will use the label "dualism" for this claim and "monism" for its denial. Monism is of course compatible with the existence of specifically political principles of a nonfundamental kind, such as the principle that taxation should be levied according to taxpayers' "ability to pay." What monism rejects is any defense of such a principle by appeal to a fundamental one that does not also apply directly to people's conduct. It should therefore be clear that monism does not have the absurd implication that all morally defensible legal principles are ipso facto valid moral principles. In rejecting the distinction Rawls draws between politics and morality, I am not rejecting the distinction between law and morality. Thus it might be appropriate to enact legislative or constitutional provisions that are not plausible moral principles. Rawls's difference principle could possibly be an example.9 Monism is also fully compatible with such views as that the duties of state officials are very different from the duties of private citizens, or that the duties of citizens are very different from the duties of stateless people. These are questions of personal conduct; different moral views are sensitive in different ways to political roles and political context. Monism is even compatible with the idea that just institutions have some kind of intrinsic expressive value: just institutions could have expressive value that is not reducible to the value of their effects even though the principles that tell us that the institutions are just are not fundamentally principles about institutions. What monism rejects, then, is that there could be a plausible fundamental normative principle for the evaluation of legal and other institutions that does not apply in the realm of personal conduct. What mo8. See A Theory of Justice, e.g., p. 120; Political Liberalism, p. n. 9. For the difference principle, see A Theory of Justice, pp. 75 ff., and p. 264 below.

John Rawls

255

Institutions and the Demands of Justice

nism rejects is the kind of principle Rawls says his difference principle is.10 The example of the difference principle is appropriate, for my discussion will focus exclusively on distributive justice. This is because almost all the substantive arguments for and against dualism that I am aware of concern distributive justice.11 More narrowly still, a main hope behind all the arguments for dualism seems to be that it will help with a fundamental problem faced by, specifically, egalitarian theories of distributive justice: the problem of the unreasonable demands such theories of justice may potentially impose on some people. As the focus on distributive justice should suggest, I understand the case for dualism to be largely independent of the main themes of Political Liberalism. It is true that Rawls includes dualism in his definition of the "political conception of justice" introduced in that book.12 The language of dualism also figures in the account of a political conception of justice as one that stands free of, and is compatible with, a range of "comprehensive views."13 But this central issue for Political Liberalism of how a conception of justice can be justified and stable in a society made up of people with a range of different moral, philosophical, and religious views in fact seems to be neutral between dualism and monism.14 Rawls does not believe that an overlapping consensus of reasonable comprehensive doctrines will necessarily include all of the principles of justice found in his own preferred theory of justice, justice as fairness.15 Thus there is room for reasonable disagreement about principles for the design of institutions too—we cannot say that the overlapping consensus just comprises those principles from the various reasonable doctrines that apply to the basic structure of society. Moreover, there seems to be no reason why there could not be an overlapping consensus on a set of principles—a subset of all the moral/political 10. "The difference principle ... applies to the announced system of public law and statutes and not to particular transactions or distributions, nor to the decisions of individuals and associations, but rather to the institutional background against which these transactions and decisions take place" ("The Basic Structure as Subject," p. 283). 11. A reason for rejecting dualism that I do not discuss, which does not turn on issues of distributive justice, would be that it is incompatible with Raz's conception of political authority; see Raz, Morality of Freedom, chap. 4. 12. See p. 11. 13. See pp. 12-15. 14. See esp. Lecture IV. 15. See, e.g., ibid., pp. 223-27.

7

8

John Rawls

256

Philosophy &> Public Affairs

principles adhered to by various groups in the polity—that apply not only to the basic structure but also to people. Of course, an overlapping consensus that embraces only principles that apply just to institutions is also possible. But the possibility of a dualist overlapping consensus counts neither for nor against the plausibility of dualist egalitarian conceptions of justice. My aim in what follows is to evaluate dualism and monism on their own terms; I want to investigate what arguments are available for either view that do not just amount to arguments for some particular moral/ political theory. I do not suppose, of course, that the evaluation of dualism and monism can be in some sense metanormative or nonnormative. The aim is rather to adduce normative considerations in favor of either view whose force can extend beyond the confines of any one fully worked out theory of justice or morality. Dualism has been criticized before, perhaps most prominently by Robert Nozick, who posed the pertinent question: "In virtue of what features of the basic structure, features not possessed by microcases, do special moral principles apply that would be unacceptable elsewhere?"16 But the arguments for dualism that I will consider were all offered after Nozick wrote his critique. And though several points I will make along the way repeat points made by Nozick, the arguments for monism I will discuss—the recent incentives argument of G. A. Cohen and my own argument concerning nonideal theory—are not at all concerned with the issue of entitlements that frames his critique.17 What binds together all the arguments I will consider is an underlying concern to describe a plausible and robust egalitarian theory of distributive justice that nevertheless appears to make reasonable demands on people in just and unjust circumstances. That this Rawlsian project is worthy I take for granted; my aim is to show that dualism hinders rather than helps it. Nevertheless, as I will argue in my concluding section, there is an aspect of Rawls's institutional approach to justice that we should embrace even though we reject dualism. If we think of justice as being about institutional design we will naturally think of people's responsi16. Nozick, Anarchy, State, and Utopia, p. 205. Thomas Pogge, Realizing Rawls (Ithaca: Cornell University Press, 1989), pp. 25-28, rightly notes that Rawls's distinction does not concern size; but this point does not undermine the substance of Nozick's discussion. 17. "The only reason I have thought of for discounting microtests of the fundamental principles is that microsituations have particular entitlements built into them." Nozick, Anarchy, State, and Utopia, p. 206.

John Rawls

257

Institutions and the Demands of Justice

bility for justice as being in an important sense collective: we together, through our institutions, secure justice. Indeed it seems possible that the idea that justice is a collective responsibility is one of the motivations behind Rawls's dualism. I fully embrace that idea, and I see in it the basis of an answer to the central problem, mentioned above, of describing a robust egalitarian conception of justice that nevertheless makes reasonable demands on people. But we can accept that justice is fundamentally a matter of collective obligation without accepting that it is fundamentally a matter of institutional design. II. THE DIVISION OF LABOR The first argument in favor of dualism turns on what Rawls calls an institutional division of labor. Rawls offers this argument in "The Basic Structure as Subject," where he also notes that the reasons for (what I call) dualism were not fully explained in A Theory of Justice.18 Rawls had said in the book that the "basic structure is the primary subject of justice because its effects are so profound and present from the start."19 But of course the immense practical significance of the major social institutions in determining people's life prospects does not at all support the claim that special criteria are required for the evaluation of those institutions. The division-of-labor argument, by contrast, does promise to provide such support. In "The Basic Structure as Subject" Rawls says that the basic structure of society comprises background institutions that preserve justice by correcting the outcomes of ordinary market transactions among people. As defined by Rawls, institutions in the basic structure do not apply to people in their daily lives and are thus to be distinguished from legal rules that do, such as those governing contractual relations. "What we look for," Rawls says, is an institutional division of labor between the basic structure and the rules applying directly to individuals and associations and to be followed by them in particular transactions. If this division of labor can be established, individuals and associations are then left free to advance their ends more effectively within the framework of the basic 18. See "The Basic Structure as Subject," p. 25811. 19. P. 7-

_9

10

John Rawls

258

Philosophy & Public Affairs

structure, secure in the knowledge that elsewhere in the social system the necessary corrections to preserve background justice are being made.20 Note that Rawls here seems to offer a narrower characterization of the basic structure than he did in A Theory of Justice.21 In a passage from the book already quoted, Rawls defines the basic structure of society as "the way in which the major social institutions distribute fundamental rights and duties and determine the division of advantages from social cooperation." In the passage just quoted from "The Basic Structure as Subject," Rawls seems to narrow the account of the basic structure, and thus the subject of justice, to those institutions that preserve background justice without impinging directly on people's daily lives. Thus the institution of contract law, which does impinge upon people's daily lives insofar as it rules out force and fraud in contractual relations, does not count as part of the basic structure, whereas on the broader characterization offered in A Theory of Justice it would seem to. I will return to these two conceptions of the basic structure shortly. The ideal of a division of labor is developed further by Thomas Nagel in Equality and Partiality.22 Though Nagel writes of "a moral division of labor," the basic idea is the same. It is to take the business of securing justice off people's plates in their day-to-day lives. Assuming that any plausible theory of justice is egalitarian and will thus—in a capitalist society at least—include a redistributive component, we face the question of how best to achieve that redistribution. The idea that people must constantly think about how well-off other people are and adjust their behavior accordingly is morally unappealing. People lead freer and better lives, facts of normative significance, if they can devote most of their concerns to their own affairs without always monitoring levels of wellbeing or degrees of social inequality. A division of moral labor that leaves to background institutions the job of securing a just distribution in society and holds people responsible for only a limited set of negative duties would enable us to eat our cake and have it too. We would achieve our egalitarian aims without making ourselves miserable in the process. 20. Pp. 268-69.

21. As Pogge points out, Realizing Rawls, pp. 21-24. 22. New York: Oxford University Press, 1991, chaps. 6, 9. Nagel does not offer the division of labor argument in defense of dualism.

JohnRawls

259

Institutions and the Demands of Justice

Now it should be noted that Nagel believes that the idea of the division of moral labor brings with it serious problems. One theme of Equality and Partiality is that, once a division of moral labor is in place and people are for the most part concerned mainly with their own interests, it is hard to see how they can be expected to continue to support the background institutions that, for some of them at least, operate so strongly to their detriment. We want to liberate people from day-to-day concerns about how well everyone else's lives are going, but we rely on some such concern when we expect people to support just institutions.23 I will return to this problem briefly in the next section. I am just as attracted to the ideal of the division of labor as is Rawls or Nagel. But unlike Rawls, I do not think that this counts in favor of the view that justice is a separate subject requiring different principles.24 It is true that the ideal of the division of labor presents a further clear motivation for paying special attention to the design of institutions. For these institutions, we now see, have two virtues: not only do they secure justice more effectively than could people acting without institutions, they also minimize the costs people must sustain to secure justice. But none of this supports the idea that different normative principles apply to institutional design on the one hand and personal conduct on the other. Note first that Rawls's narrow conception of the basic structure, and thus the account of an institutional division of labor that builds on it, appear to depend on an inaccurate assumption about the way different kinds of legal institutions impact on people. We should not attempt to secure justice via those legal rules "applying directly to individuals and associations and to be followed by them in particular transactions"; the work of securing economic justice is left to the institutions of the basic structure which do not apply directly to individuals. But this distinction between background institutions and those legal rules that apply directly to people and associations is not tenable. The central institutions of the basic structure, on the narrow account, are those of taxation and transfer. We may say that the legal rules governing transfer payments do not "apply" to the recipients of those payments. Nothing similar can be 23. See chaps. 9 and 10. Thomas C. Grey succinctly stated the problem in his review of A Theory of Justice, see "The First Virtue," Stanford Law Review 25 (i973):323-2524. See "The Basic Structure as Subject," pp. 265-69.

11

12

John Rawls

260

Philosophy Public Affairs

said about the collection side of the system, as the legal rules that make up the institution of taxation obviously apply directly to individuals and associations. The idea is perhaps that whereas the rules of contract law apply to every agreement, and do so obtrusively, the rules of tax apply to individuals and associations infrequently and unobtrusively; but this too is inaccurate. Income taxation of individuals by withholding is comparatively unobtrusive, but it is just one aspect of the regime. As Anthony Kronman points out in an extensive discussion of this point, sales tax applies to every instance of certain kinds of transaction made by individuals.25 He could also have mentioned that income taxation can be very obtrusive indeed for those not subject to withholding; even though the tax does not apply to each transaction, many self-employed persons probably think about taxation more or less constantly.26 If we consider commercial associations, it is obvious enough that the taxation aspects of major corporate deals generate a significant proportion of the legal work (and expense) involved.27 Kronman goes on to argue that it may be no less intrusive on people and associations to promote distributive justice through the law of contracts. He also suggests that this might be the more efficient route to take.28 Neither possibility seems at all likely to me. But what is clear is that there is no general reason, no reason of principle, to think that some parts of the law will always be less obtrusive, less restrictive of individual liberty, than others. It is possible to design a taxation and transfer scheme of great obtrusiveness; wage earners could be required to write weekly checks to specified recipients, for example. This would be an absurdly inefficient way of securing distributive justice, but that is the point. Whether taxation and transfer, on the one hand, or contract law, on the other, can better achieve the aim of securing justice while leaving people as free as possible to pursue their own interests unhin25. See "Contract Law and Distributive Justice," Yale Law Journal 89 (i98o):so3-4. 26. Indeed, sales taxation is potentially much less obtrusive, and thus less costly to individuals, than is income taxation even for those subject to withholding. Where sales tax is incorporated into the final sale price, the consumer gives the existence of the tax no thought at all. Certain other forms of consumption tax are likewise invisible to the consumer. See Deborah L. Paul, "The Sources of Tax Complexity: How Much Simplicity Can Fundamental Tax Reform Achieve?" North Carolina Law Review 76 (i997):i5i-22i; Joel Slemrod and Nikki Sorum, "The Compliance Cost of the U.S. Individual Income Tax System," National Tax Journal 37 (1984) 1461-74. 27. See also Kronman, "Contract Law and Distributive Justice," p. 504. 28. Ibid., pp. 507-10.

John Rawls

261

Institutions and the Demands of Justice

dered by the machinery of justice is in large part an empirical question, subject to changing circumstances.29 So the narrow definition of the basic structure Rawls appears to work with in "The Basic Structure as Subject" does not correspond to any principled doctrinal or structural division in legal systems. Thus this account of the basic structure would make it impossible to know which parts of our legal system should be regulated by principles of justice, and which parts should not. Rawls's concern is to isolate out those institutional structures which do not apply directly to people, or do so less intrusively, as the proper site for the application of principles of justice. But it is obviously not feasible to approach taxation policy, for example, under the assumption that part of the law of taxation is in the basic structure and part of it is out. The division-of-labor argument for dualism, as I have so far construed it, therefore employs a conception of the basic structure that is in fact incompatible with dualism, for dualism must be taken to imply that we can easily identify the realm in which principles of justice apply. All this suggests that Rawls did not after all intend the account of the basic structure offered in "The Basic Structure as Subject" to differ significantly from that offered in A Theory of Justice. Indeed a passage in the article suggests that Rawls understands contract and property law and the background institutions of taxation and transfer as together belonging to the one structure that should be evaluated as a whole by principles of justice.30 Perhaps this is the best interpretation of what Rawls writes, since it is indeed hard to see how legal rules applying directly to people could be regarded as entirely outside the purview of justice, so that, for example, sales taxes, or principles of unconscionability in contracts, could in principle not be evaluated on grounds relating to distributive justice.31 However, reverting to the broader notion of the basic structure does not help the division-of-labor argument for dualism. If the proper site 29. See ibid., p. 507 30. See the only full paragraph on p. 268. The interpretation that Rawls did not see himself as narrowing the account of the basic structure is supported by the fact that at other points in "The Basic Structure as Subject," and in portions of Political Liberalism that were written later, Rawls defines the basic structure in the same broad terms used in A Theory of Justice, see Political Liberalism, e.g., pp. 35, 258. 31. On the latter in the U.S. context, see Uniform Commercial Code §2-302; Restatement (Second) of Contracts §208; E. Allen Farnsworth, Contracts, 2nd ed. (Boston: Little,

13

14

John Rawls

262

Philosophy &> Public Affairs

for the application of principles of justice is a basic structure that includes, at least,32 all of law, it might seem that we have found a structural divide—that between law and nonlaw—that corresponds to the desired division of labor. But in fact there is no reason to think that the application of principles of justice to legal institutions alone removes from people the burdens of securing justice. It would be possible, for example, to enact a positive law requiring citizens to aim at justice continually in their daily lives. A criminally enforced regime of "a thousand points of light" may be a thoroughly unappealing idea, practically and normatively, but it is perfectly possible. The ideal of the division of labor thus provides no support one way or the other for the view that principles of justice apply exclusively to the design of institutions. The ideal does, of course, offer guidance on how our institutions might best be designed, but that is a different matter. Monists can easily agree with that. To develop this last point, it may help to sketch a conception of justice that rejects dualism, but embraces the ideal of the division of labor. The view I will sketch is one that sees distributive justice as based on some principle of beneficence. That is, on this view distributive justice is about benefitting people, increasing their well-being. This is of course what utilitarians believe, and utilitarianism is one beneficence-based approach. But utilitarianism is distinctive for claiming that beneficence is all there is to political morality; utilitarianism is a one-principle theory. The view I will sketch explains distributive justice in terms of beneficence but leaves open what other principles of political morality there may be. For example, one can believe that distributive justice should be understood in terms of beneficence and also believe that people have rights of various kinds that constrain efforts to promote well-being. Furthermore, there is no reason why a beneficence-based conception of distributive justice must agree with utilitarianism that what is to be promoted is aggregate well-being, without regard to its distribution. A parBrown, 1990), pp. 323-39; Duncan Kennedy, "Distributive and Paternalist Motives in Contract and Tort Law," Maryland Law Review 41 (i982):s63-658; Michael J. Trebilcock, The Limits of Freedom of Contract (Cambridge, Mass.: Harvard University Press, 1993), pp. 97-101. 32. I discuss the question of which institutions count as part of the basic structure further in the next section. See also Pogge, Realizing Rawls, pp. 22-23.

John Raw I s

263

Institutions and the Demands of Justice

ticularly attractive principle of beneficence that is sensitive to distribution, and that qualifies, to my mind, as an egalitarian principle, is what Derek Parfit calls the principle of weighted beneficence.33 This principle holds that when promoting well-being, one should give (some) priority to promoting the well-being of those who are worse off. Benefits to worse-off people matter more, they have greater moral weight.34 If we think of distributive justice in terms of weighted beneficence we see no distinction between the principles that govern the design of institutions and those that govern personal conduct. Within the constraints imposed by whichever rights we recognize, institutions should be designed so that they maximize weighted aggregate well-being over time; likewise people, within the same constraints, should act so as to promote the same thing: aggregate weighted well-being over time. Approaching social justice via a principle of weighted beneficence, we embrace monism. However, when designing institutions, it will make eminent sense from within this perspective to remove as much as possible the burdens of promoting weighted well-being from people's day-today lives. It will make eminent sense to set up background structures of taxation and transfer such that people can for the most part pursue their own interests, "secure in the knowledge that elsewhere in the social system the necessary corrections to preserve background justice are being made." If the background institutions are doing their job properly, people will not have to think too much about promoting general well-being, 33. See Equality or Priority? (The Lindley Lecture, The University of Kansas, 1991). The phrase "weighted beneficence," which does not appear in this publication, comes from Parfit, "On Giving Priority to the Worse Off," unpublished ms., 1989. T. M. Scanlon describes a version of weighted beneficence in "Rawls' Theory of Justice," in Norman Daniels, éd., Reading Rawls (Stanford: Stanford University Press, 1989), pp. 197, 201. 34. The beneficence-based conception of justice should not be confused with the charity view criticized in Thomas Nagel, "Poverty and Food: Why Charity Is Not Enough," in Peter G. Brown and Henry Shue, eds., Food Policy (New York: The Free Press, 1977), pp. 54-62. A beneficence-based approach neither renders matters of justice optional, nor automatically treats existing legal entitlements as legitimate (see ibid., p. 56). Brian Barry criticizes a conception of international justice based on beneficence (or "humanity") in part for leaving status-quo legal entitlements in place; see "Humanity and Justice in Global Perspective," in J. Roland Pennock and John W. Chapman, eds., Ethics, Economics, and the Law (New York: New York University Press, 1982), pp. 243-50. Again, any approach to justice that does do this is obviously defective, but there is nothing in the idea that justice should be understood in terms of beneficence that leads to such a conclusion.

15

16

John Rawls

264

Philosophy &> Public Affairs

and this liberation is, from the point of view of beneficence, all to the good. III. COHEN, INCENTIVES, AND THE FAMILY The natural fit between the ideal of the division of labor and the account of justice as weighted beneficence confirms that the attractiveness of the division of labor does not in itself count for or against dualism. I turn now to an argument of G. A. Cohen to the effect that the combination of dualism and an institutional division of labor will almost always produce injustice. We turn then from an argument in favor of dualism to an argument against it. In his recent article, "Where the Action Is: On the Site of Distributive lustice," Cohen argues for the view that "principles of distributive justice ... apply to the choices that people make within the legally coercive structures to which, so everyone would agree, principles of justice (also) apply."35 That is, he presents an argument for monism. To explain Cohen's argument I need first to introduce a different argument he made some years earlier. In his Tanner Lectures, "Incentives, Inequality, and Community," Cohen presented an argument against the standard interpretation of the operation of Rawls's difference principle.36 The difference principle holds that social and economic inequality is justified only where it occurs in an overall scheme that improves the expectations of the least advantaged members of society.37 Cohen focuses on the question of why inequality might improve the prospects of the worst-off group in society and thus be justified by the difference principle. The standard assumption is that in a fully equal society people will lack incentives to work hard, take risks, etc. If, however, high earnings, and thus inequality, are tolerated, people, especially "talented" people,38 will have the right incentives: they will work hard, and this will redound to everyone's benefit. This kind of inequality is "necessary" to improve the prospects of 35. Philosophy & Public Affairs 26, no. i (winter 1997) :s. 36. Reprinted in Stephen Darwall, éd., Equal Freedom (Ann Arbor: University of Michigan Press, 1995), pp. 331-97- See also, "The Pareto Argument for Inequality," Social Philosophy and Policy 12 (winter 1995). "Where the Action Is," pp. 5-10, contains a helpful summary of the earlier argument. 37. See, e.g., A Theory of Justice, pp. 75 ff. 38. I endorse Cohen's remarks about the focus on "talented" people in standard discus-

John Rawls

265

Institutions and the Demands of Justice

even the people who will be worst off in the unequal society and is thus sanctioned by the difference principle. Cohen's objection here is that the inequality is only "necessary" to improve the prospects of the worst-off group because without incentives, those who could work hard will not. If people were equally productive with or without incentives, there would be no need to depart from equality. Thus the difference principle, as standardly interpreted, actually tolerates inequality which is not necessary to improve the prospects of the worse off. For there is another possibility: the talented could simply choose to work hard without the additional rewards. Cohen concludes that for the difference principle to be really satisfied there must be an egalitarian ethos in society, an ethos that will lead people to choose to work hard even in the absence of inequality-creating incentives. Cohen's incentives argument was intended to show that Rawls's own account of justice, properly interpreted, tolerated far less inequality than Rawls and others had supposed.39 To adapt Cohen's argument to the terms of my discussion, however, we can say that whereas Nagel raises doubts about the feasibility of a division of labor that leaves to background institutions the job of promoting justice, Cohen argues that a society characterized by such a division of labor is unlikely to be just. In other words, Cohen believes that it is not possible to create a society where (to repeat Rawls's words once more), individuals and associations are ... left free to advance their ends more effectively within the framework of the basic structure, secure in the knowledge that elsewhere in the social system the necessary corrections to preserve background justice are being made.40 For the basic structure, on its own, is not capable of making those corrections for all possible patterns of individual choice. If justice is to be secured, people have to make decisions contrary to their own interests—they have to choose actions that will promote justice. sions of this issue. As he says, all that needs to be true of the relevant people "is that they are so positioned that, happily for them, they do command a high salary and they can vary their productivity according to exactly how high it is" ("Where the Action Is," pp. 6-7). 39. See, e.g., "Incentives," p. 378. The argument contains a great deal of detail and attention to Rawls's texts that I will not consider here. For a summary of the argument that Rawls himself should say that a society that provides economic incentives is unjust, see "Where the Action Is," pp. 7-9. 40. "The Basic Structure as Subject," p. 269.

17

18

John Rawls 266

Philosophy &> Public Affairs

Cohen here brings out a further serious problem with the ideal of the division of labor. Or rather, he gives us reason to modify the content of the ideal. As I originally presented it, the ideal of the division of labor described a world where people can be relatively unconcerned about issues of justice in their daily lives—in two senses. First, their motivations and their deliberations about what to do need not focus on how best to secure justice. Second, they can typically act in whatever way will promote their self-interest. Cohen has in effect established that we should not be too hopeful about the degree to which we can attain the second of these two elements of the original ideal. A passage from "Incentives, Inequality, and Community" makes the point well: It is not true that, in the society I have in mind, a person would have to worry about unfortunate people every time he made an economic decision. Liberals would regard that as oppressive, and, whether or not they are right, one function of the egalitarian ethos is to make conscious focus on the worst off unnecessary. What rather happens is that people internalize, and—in the normal case—they unreflectively live by, principles which restrain the pursuit of self-interest and whose point is that the less fortunate gain when conduct is directed by them.41 I am not sure that the word "ethos" is ideal for Cohen's purposes, but the point is clear. He is not saying that justice requires that all people develop benevolent or equality-loving characters. He is saying that in a just society it is inevitable that many people will quite typically have to make choices contrary to their self-interest. In the ideal of the division of labor as I originally presented it, people have to choose contrary to interest in only a small number of cases—to support the just regime, and, within the just regime, to comply with a narrow range of negative duties. It is compatible with the revised ideal that people must choose contrary to self-interest more or less all the time. But the revised ideal is nevertheless certainly worth having, since it points to a just world where people's deliberative and emotional lives need not be swamped by concern with the aims of justice. Moreover, if we take seriously Nagel's concerns about the tension within the original ideal caused by the need for people to make the often severely contrary-to-interest 41. P. 384, note omitted.

John Raw I s

267

Institutions and the Demands of Justice

choices necessary to sustain the regime, it might in the end appear that the revision to the ideal forced on us by Cohen's incentives argument is not so dramatic after all.42 Though Cohen's incentives argument operates in the context of Rawls's own theory of distributive justice, the results seem to have quite general application. Thus, for example, I think that a beneficence-based view of justice can also hope to accommodate only the revised and weaker form of the ideal. However I will not here attempt to spell out how this story would go. Having set out Cohen's incentives argument (and having digressed a little on its impact on the ideal of the division of labor), we are now in a position to present his case against dualism. Cohen notes that Rawlsians would object to his incentives argument on the ground that it mistakenly assumes that principles of justice apply to the ordinary day-to-day choices of people when they in fact apply only to institutions in the basic structure.43 In response to this "basic structure objection" to his incentives argument Cohen presents an independent argument against the idea that the basic structure alone is the subject of justice. I will lay out that argument presently. But I find it striking that Cohen does not offer the simple reply to this objection that I believe he is entitled to, viz. "to the contrary, my argument provides a strong reason to reject the idea that principles of justice apply only to the basic structure." He is entitled to this reply, I believe, because dualism is not the natural or default view, departures from which require conclusive justification. At any rate, I can for my own case say that Cohen's incentives argument counts as a strong objection to dualism. If dualism makes it impossible even to discuss how people ought to change their behavior, within an institutional scheme, in order to better achieve the aims of justice, then that seems to be a very good reason to reject dualism. Cohen's demanding conclusions about what justice requires of people may not, in the end, be right, but it cannot be the case that they are simply irrelevant to the topic of justice. 42. Cohen's result may, however, undermine one very important aspect of the ideal of the division of labor, viz. that it promises the compatibility of distributive justice and a market economy—important on most egalitarian theories because of its informational value. On this issue see Nagel, Equality and Partiality, pp. 91-95; Joseph Carens, Equality, Moral Incentives, and the Market (Chicago: University of Chicago Press, 1981). 43. "Where the Action Is," p. 10.

19

20

John Rawls 268

Philosophy &> Public Affairs

I will now very briefly summarize the main argument Cohen himself presents against the idea of the basic structure as subject.44 Cohen's concern is with the question of which institutions are in the basic structure and which are not. He notes that Rawls sometimes lists the institution of the family as part of the basic structure. Susan Moller Okin has argued that this must be the case and that A Theory of Justice pays insufficient attention to the application of principles of justice to genderbased inequalities within families.45 Cohen does not dispute that practices within families should be judged according to principles of justice. But Okin is wrong, he says, to think that Rawlsian principles of justice can be used in this way. Rawlsian principles of justice do not apply to ordinary day-to-day choices, but only to the design of institutions; and although institutions that depend on the coercive powers of the state are not sustained by ordinary day-to-day choices of people, institutions not sustained by state coercive powers, such as the family, are indeed instead sustained by people's ordinary day-to-day choices. So if the family is subject to principles of justice, principles of justice apply to people's ordinary day-to-day choices. Thus, contrariwise, to maintain the view that principles of justice do not apply to individual choices, Rawls must accept that they do not apply to the institution of the family. But that, Cohen submits, would be contrary to Rawls's stated concern with the justice of those social structures that have a profound impact on a person's life prospects. Since it is that concern which motivates the focus on the basic structure to begin with, Rawls's best option is to abandon the idea that justice does not apply to the ordinary choices people make against the backdrop of the basic structure. In an endnote to his article, Cohen expresses doubts about his claim that legally coercive institutions, such as those of taxation and transfer, are not sustained by people's day-to-day choices.461 agree that the role individual choice plays in maintaining a social institution is by no means obvious, and that the contrast Cohen draws in his argument against Rawls may not be tenable. But Cohen ends by saying that he can concede this, since the upshot would be that all institutions are sustained by individual choice, and thus that Rawls's separation of two normative spheres would be even more clearly undermined. This seems to 44. Ibid., pp. 17-23. 45. Susan Moller Okin, Justice, Gender, and the Family (New York: Basic Books, 1989), chap. 5. 46. "Where the Action Is," pp. 28-30.

John Rawls

269

Institutions and the Demands of Justice

me to be wrong. If all institutions—those enforced by the coercive power of the state and those not—are fundamentally sustained by the choices people make, that does not mean that a Rawlsian could not distinguish between principles that apply to institutions and principles that apply to ordinary personal choice. For the choices that sustain institutions such as the family are not like the choices we make about how to expend our resources, or even the choices we make about how hard to work. What is striking about the choices that sustain the family is that they form a pattern of choice, more or less uniform across society. A person's decision about whether to work hard for less money is of course also in some ways affected by the choices others are making, but not in the same way as are the choices of husbands and wives about how to divide domestic labor. To say otherwise seems to deny the reality of the family as a special kind of social form; it would be to make any identifiable pattern of behavior in society, such as a majority tendency to self-seeking behavior, as much of an institution as the family is, and thus also as great a constraint on social and personal life as the family ÍS.47

Cohen is right to insist that those who believe that justice is about the design of institutions and not individual choice owe us a better account of what institutions are, precisely. But I do not believe that his own remarks on the nature of institutions yet refute dualism. I cannot offer a theory of the nature of social institutions and the role of individual choice in their maintenance, but it does seem that a plausible account of institutions that could be used to block Cohen's argument may be available. Nevertheless, even if Cohen's official reply to the basic- structure objection to his incentives argument is unsuccessful, we have seen that his incentives argument by itself provides a direct reason to abandon dualism. IV. INSTITUTIONS AND PEOPLE'S OBLIGATIONS So far we have encountered one powerful reason to reject dualism and no reason to accept it. I turn in the next two sections to sketch a further pair of arguments in dualism's favor. The arguments are Rawlsian in spirit, though not explicitly offered by him. While the division-of-labor 47. It may be that Cohen would simply accept this (to my mind) tendentious implication. See his reference to the "ethi that sustain gender inequality, and inegalitarian incentives," ibid., p. 23, and the discussion of blame in the final section of the article, especially p. 26.

21

John Rawls

22

270

Philosophy &> Public Affairs

argument for dualism is somewhat indirect, claiming in effect that the dualist approach brings with it a certain benefit, the arguments I will next consider both provide direct grounds for believing that the principles that govern institutional design do not apply to people's choices. In each case, the idea is that the existence of institutions gives rise to a special kind of normative problem, one that institutions are responsible for, but people are not. As background for these arguments, and for my own later argument for monism, I must first explain how, on the dualist view, the virtues of institutions relate to the responsibilities of people. Some such connection there must be. Justice may be a virtue of institutions, and we may thus be able to say that particular institutions are in themselves just or unjust, but people must be subject to some kind of requirement to support those institutions, for they cannot become virtuous and remain so all of their own accord. I am not here repeating the point that it is hard to imagine a set of institutions that manages to secure distributive justice without impinging upon people's daily lives to some degree. Rather, the focus now is on those special choices—as opposed to what I have been calling ordinary or day-to-day choices—that people need to make in order that the right kinds of institutions exist over time; such choices would include those of how to vote in elections.48 (It will be recalled that Nagel's skepticism about the feasibility of the division of labor turned on the need for these institution-supporting choices.) Of course Rawls does not ignore the need for a principle connecting people to institutional design and survival. He postulates that people have a natural duty to support just institutions.49 In addition, "we are to assist in the establishment of just arrangements when they do not exist, at least when this can be done with little cost to ourselves."50 This is 48. See also Nagel, Equality and Partiality, p. 89. As Barry Friedman pointed out to me, raising a further problem for dualism, the line between institution-supporting choices and day-to-day choices is not bright: Is it, for example, my institution-supporting duty to engage people in political discussion to help keep the sense of justice of my fellow citizens true and alive? 49. See A Theory of Justice, pp. 115, 333-37. As A. John Simmons has pointed out, it is hard to see how this natural duty to support just institutions is supposed to link people with any particular set of just institutions; see Moral Principles and Political Obligations (Princeton: Princeton University Press, 1979), chap. 6. See also Ronald Dworkin, Law's Empire (Cambridge, Mass.: Harvard University Press, 1986), pp. 190-216; Jeremy Waldron, "Special Ties and Natural Duties," Philosophy & Public Affairs 22, no. i (winter i993):3-3O. 50. A Theory of Justice, p. 334; see also p. 115.

John Rawls

271

Institutions and the Demands of Justice

precisely what one would expect. The dualist view is that principles of justice are used to evaluate institutions and do not apply to people's own choices. Principles of justice therefore describe the responsibility of institutions. But then the responsibility that people have in respect of justice must be to support and bring about just institutions. If people have any responsibility at all in respect of justice it must presumably be this one, since the principles of justice do not describe their responsibility (what else would their responsibility be?). And people must in any case have this responsibility for the dualist view to make sense: since institutions are not agents and don't actually have any responsibilities at all, it is only people who can ensure that institutions satisfy principles of justice. These points allow us to put the basic disagreement between monism and dualism in somewhat different terms than have been used so far. Whereas monism holds that people have direct responsibility for justice, dualism holds that as far as justice is concerned, the responsibility of people is mediated by institutions. If just institutions must aim at equality, monism holds that people must aim at equality too; dualism holds, by contrast, that people must aim at the existence of institutions that aim at equality. V INTERACTION If justice is in some fundamental way all about institutional design, it is natural that people's responsibility in respect of justice would concern institutions, and not directly what just institutions concern themselves with. But why, to return to our main question, would we think that justice was all about institutional design? One possibility, which emerges most clearly in the work of Thomas Pogge, turns on the idea that interaction of a certain regularized kind generates new kinds of obligations.51 In Pogge's view, the obligations people have to other people merely as 51. See Realizing Rawls, pp. i, 20-28; 273-80; "Cosmopolitanism and Sovereignty," Ethics 103 (1992)148-57; "A Global Resources Dividend," in David A. Crocker and Toby Linden, eds., Ethics of Consumption: The Good Life, Justice, and Global Stewardship (Lanham: Rowman &> Littlefield, 1998), pp. 501-7, 527-28; "The Bounds of Nationalism," in Jocelyne Couture, et al., eds., Rethinking Nationalism, Canadian Journal of Philosophy, Supplementary Vol. 22 (Calgary: University of Calgary Press, 1998), pp. 463-504. In what follows I draw a view out of Pogge's writings; I do not claim to be offering a full account of his actual position.

23

24

John Rawls

272

Philosophy & Public Affairs

such are fairly weak, but the obligations people have to others with whom they are interacting according to reasonably well-established ground rules are strong. These strong obligations direct people in the first instance to those ground rules of interaction themselves: people have an obligation not to impose unjust ground rules of interaction on others.52 Pogge's perspective on distributive justice is cosmopolitan and his argument for global rather than merely domestic justice draws heavily on the institutional approach. The world's rich "are causally deeply entangled in the misery of the poor and [we] cannot extricate ourselves from this involvement so long as their misery continues."53 This causal link, coupled with the fact that the current scheme of institutions—the current nature of our global causal entanglement—could be replaced with a feasible alternative under which the world's poor would do very much better, turns the current complacent behavior of most of the world's rich into rights-violating behavior; we are violating the rights of people not to have unjust institutions imposed on them.54 This is only the sketchiest summary of a complex argument, but we see in it the general form of the dualist position. Institutionally governed interaction—causal entanglement according to more or less regular ground rules of economic and political interaction—generates an obligation that, absent that interaction, does not exist. But that obligation is not to do what principles of justice tell us just institutions would do. Our responsibility is to make sure that the institutions do that, so that we are no longer violating other people's rights by imposing unjust institutions on them. My main objection to dualism (in Section VII) will be that all views of this kind are implausible precisely because they have us aim at the good of institutions rather than at the good institutions can best do. But apart from that general point, I find this version of dualism implausible on its 52. See, e.g., Realizing Rawls, p. 32. 53. "A Global Resources Dividend," p. 505. For extensive discussion of global interdependence as establishing the domain of global distributive justice, see Charles Beitz, Political Theory and International Relations (Princeton: Princeton University Press, 1979), pp. 143-53; Beitz later rejects the relevance of interdependence, see "Cosmopolitan Ideals and National Sentiment," Journal of Philosophy 80 (1983) 1595, citing the criticism of David A. J. Richards, "International Distributive Justice," in Pennock and Chapman, eds., Ethics, pp. 287-93; see also Pogge, Realizing Rawls, p. 241, n. 3. 54. See "A Global Resources Dividend," pp. 504-7.

John Rawls

273

Institutions and the Demands of Justice

own terms. Pogge's overriding purpose is to establish the moral importance of global inequality. To do that, he believes it important to distinguish injustice, the imposition of unjust institutions, from the mere failure to assist other people or alleviate existing inequality in the world. The aim is to shore up the importance of (genuine) injustice by contrasting it with a less significant moral concern and cautioning people not to confuse the two. Thus there is posited a strong moral discontinuity: the existence of institutionally governed interaction makes all the moral difference; it defines the group of people among whom justice is an issue. Now it is true that we are today all connected via global institutions of the kind Pogge mentions. And indeed, since any causal explanation of the current world distribution of resources or well-being would run in part through the prevailing institutional set-up, we can say that the connection is causal. But it remains unclear to me why the fact of causal connection should have any significance for the question of justice. If the world's poor could be shown all to be worse off because of the advent of global institutions, then perhaps the causal role of institutions would have some intuitive significance—the idea could be that while people need not do very much to help each other, they must not hurt each other, and this is what institutions have done. But the upshot of this would be that the rich have a duty to make sure that the institutions they control leave no one worse off than she would have been in the absence of global institutions. This clearly falls far short of Pogge's conception of global justice; nor is it the case that he wants the significance of institutions to depend on their having made people worse off (think about the domestic case).55 But since that is so, it is a mystery why the fact that institutions are of causal significance should entail that they have special significance for justice as well.56 55- But see "The Bounds of Nationalism," pp. 491-94, for some discussion which suggests that Pogge may after all be sympathetic to this line of thought; I interpret him to be here offering a second-best argument to an audience skeptical about more robust claims of global justice. The argument is nevertheless very weak on its own terms, because of the implausibility of the premise that global institutions have made the world's poorest worse-off than they would have been in the absence of institutional interaction with the world's rich. Pogge, ibid., p. 494, writes that the rich could not disprove this claim; this seems so, since any claim of this kind is wildly speculative, but why do the rich have the burden of proof? 56. It would not help to say that in not promoting feasible alternative schemes under which the poor do better, the rich cause the poor to be worse off. This would be to appeal

25

26

John Rawls

274

Philosophy &> Public Affairs

Pogge does not discuss extensively his reasons for thinking that institutions mark out the scope of justice, and it may be that he has other arguments in mind. But the full argument would have to be rather compelling, because the claimed moral discontinuity is so stark. We are to imagine that a remote community not yet involved in any external trade or otherwise affected by global institutions would make no claims of justice on us, but that as soon as we begin to trade with them, thus including them in the world economy, something like the difference principle would suddenly govern our relations—requiring the immediate establishment (or extension) of institutions of taxation and transfer.57 I find it incredible that the application of the ground rules of economic interaction could have this much moral significance.58 It should be noted that one could advance a monist version of Pogge's argument. It could be thought that the existence of institutional interaction simply defines the scope of our (strong) obligations in respect of human inequality or well-being; on this view people's responsibility to the kind of negative causation that some say is involved when I fail to rescue a drowning person; see, e.g., John Kleinig, "Good Samaritanism," Philosophy &> Public Affairs 5, no. 4 (summer 1976) 1382-407. But if that is plausible, the rich also cause the terrible plight of the world's poor by not doing other things, not connected to institutions, such as contributing to private development projects. Scanlon, "Rawls' Theory of Justice," p. 202, plausibly (though for the contrary view, see Barry, "Humanity and Justice," p. 233) treats Rawls's focus on institutions as equivalent to his stipulation that the problem of justice is to find fair terms of cooperation among people engaged in cooperation for mutual advantage. For this aspect of Rawls's view, see, e.g., A Theory of Justice, p. 4; Political Liberalism, pp. 15-18; for a succinct reconstruction— endorsed by Rawls, Political Liberalism, p. i?n.—see Allan Gibbard, "Constructing Justice," Philosophy &> Public Affairs 20, no. 3 (summer 1991) 1264-79. Scanlon's interpretation is congenial to Pogge's cosmopolitan version of justice as fairness, as he suggests that on Rawls's view "considerations of justice apply at least whenever there is systematic economic interaction; for whenever there is regularized commerce there is an institution in Rawls' sense" ("Rawls' Theory of Justice," p. 202); see also Pogge, Realizing Rawls, p. 241. But since neither Rawls nor Pogge defends dualism by reference to the conception of justice as fair terms of cooperation, I will not pursue this possible ground for the special significance of institutions for justice; see also n. 58. 57. Pogge makes clear that prior to the trading there would still be fairly weak duties of morality owed to such a community; but there will be no (strong) duties of justice; see, e.g., "Cosmopolitanism and Sovereignty," p. 51. 58. The discontinuity here objected to is made worse if the significance of institutions for justice is explained via Rawls's conception of justice as fair terms of cooperation (see n. 56). For relevant discussion see Gibbard, "Constructing Justice"; Nozick, Anarchy, State, and Utopia, pp. 185-86; Richards, "International Distributive Justice"; Allen Buchanan, "Justice as Reciprocity versus Subject-Centered Justice," Philosophy &> Public Affairs 19, 3 (summer i99o):227-52.

JohnRawls

275

Institutions and the Demands of Justice

would be to promote equality or general well-being among those with whom they are interacting in the relevant way and not necessarily, as in the dualist version, to promote the justice of the ground rules of that interaction. (Indeed some remarks of Pogge suggest that this might be his own view.)59 But the possibility of this monist interpretation does not show that my objection to the moral discontinuity Pogge's view entails is not really an objection to dualism. For dualist views all require some such account of the special significance institutions have for the question of justice, whereas monist views do not. Of course I have no general argument against the possibility that institutions have the kind of special significance sufficient to provide the initial motivation for a dualist view. My strategy in this section and the next is therefore simply to question the motivation behind two particular arguments for dualism. VI. DEMOCRATIC LEGITIMACY Pogge's argument was originally offered as an explicit elaboration and defense of Rawls's dualism. The next dualist view, which turns on the notion of democratic legitimacy, echoes the Rawlsian idea that a theory of justice applies in the first instance to a closed political community as well as the deep connections in Rawls's work between distributive justice and the political values of a constitutional democratic state.60 As with the interaction view, I will not attempt to do more than sketch the main features of the argument, so that we can form some rough idea of its potential strength. Furthermore, though my presentation of the argument from democratic legitimacy is inspired mainly by writings of Ronald Dworkin, I do not claim here to be offering an accurate account of Dworkin's own views.61 59- As, for example, when he writes of a "sharp and strong negative duty not to collaborate in the imposition of an unjust institutional scheme without working toward institutional reform or at least helping the victims of injustice" ("Three Problems with Contractarian-Consequentialist Ways of Assessing Social Institutions," Social Philosophy and Policy 12 (i995):242-43, my emphasis). 60. For these Rawlsian ideas, see, e.g., Political Liberalism, pp. 11-15. 61. The argument from democratic legitimacy was originally put to me, as an interpretation of Rawls's case for dualism, by Ronald Dworkin and Thomas Ñagel in a session of their Colloquium for Law, Philosophy, and Social Theory at NYU School of Law. I have also benefitted greatly from reading Dworkin, "The Roots of Justice," unpublished ms. 1997. For points similar to those offered in this section see G. A. Cohen, "If You're an Egalitarian, How Come You're So Rich?" in the published version of Cohen's 1996 Gifford Lectures (Cambridge, Mass.: Harvard University Press, forthcoming).

27

28

John Rawls

276

Philosophy & Public Affairs

The central claim of the argument from democratic legitimacy is that the strictest requirements of distributive justice—the requirement of "equal concern" in Dworkin's account—apply only within the boundaries of a particular state. Two features of states are especially significant in explaining this fact. First, they claim a monopoly on the legitimate use of force within their jurisdictions, and second, when they speak, they speak in the name of an entire political community.62 Since states have these features, their treatment of their citizens is subject to special moral/political constraints that do not apply to people or ordinary associations, or to states in their relations with other states. In particular, if states are to use force legitimately, and to speak legitimately in the name of the entire political community, they must satisfy the strictest requirements of distributive justice in their treatment of their citizens. (Weaker requirements of distributive justice may apply to interstate relations and between rich and poor individuals wherever located.) Thus, the duties that people have in respect of full-fledged distributive justice all flow from their roles as citizens, or, at least, as members (somehow defined) of the relevant political community. Absent membership in some such community, a person's duties in respect of distributive justice will be different in kind, and weaker. And the duties that members of political communities have in respect of full-fledged distributive justice are that they promote and support just state institutions. (As with the interaction view, one could perhaps offer a monist version of the democratic legitimacy view63—though it is unclear how people's direct efforts at, say, promoting equality could improve the legitimacy of the state.) I am more attracted to the cosmopolitanism of Pogge than the statebased conception of distributive justice on which this second argument depends.64 The increasingly global nature of what Pogge calls the ground 62. See further Dworkin, Law's Empire, chaps. 5 and 6. Cf. Rawls, "The Basic Structure as Subject," p. 277: "there are no social ends except those established by the principles of justice themselves " 63. Dworkin, in a brief passage addressing the duties of rich people in the nonideal case, contemplates (but does not endorse) the principle that we should "live with only the resources we think we would have in a fair society, doing the best we can, with what is left, to repair injustice through private charity" ("The Foundations of Liberal Equality" [1990] in Darwall, éd., Equal Freedom, p. 261). This would be to reject dualism. 64. Dworkin and Rawls are of course not the only philosophers who defend the view that distributive justice is in the first instance an internal matter for each state. For a

John Rawls

277

Institutions and the Demands of Justice

rules of economic interaction, not to mention the existence of powerful political institutions such as the International Monetary Fund, puts severe pressure on the isolationist version of dualism. The kind of genuine self-government that inspires the argument from democratic legitimacy is increasingly out of the reach of states, with the possible exception of a few that are extremely powerful economically or militarily. Of course there is nothing in the view that rules out the development of multinational or global states; but in their absence the smaller and less powerful states seem to be in danger of falling outside the domain of full-fledged distributive justice. If isolationism were plausible, however, the account of the special significance of institutions for justice that turns on the conditions of democratic legitimacy would seem to have more promise than the account that appeals to causal entanglement. Nevertheless, and even apart from the problems facing any isolationist conception of justice, I find this second account of the special significance of institutions for justice prima facie as implausible as the first, and for the same reason: the moral discontinuity it appeals to is simply too dramatic. We are asked to believe that people, as individuals, owe only limited duties to nonintimates, but that they nevertheless oughtvto accept that their government act according to strict egalitarian principles. It seems prima facie quite implausible that, properly understood, the egalitarian's commitment to general equality in her community flows entirely via her concerns with legitimate governance. Consider the question of why, on this view, any powerful and resource-rich person would ever think it a good idea to become a member of a political community. Of course, for most people there is no choice in the matter, but the thought experiment is, I think, revealing. Most of us think that (non-minimal) states are valuable because of what they do for the quality of people's lives or for other values, such as equality. Whether or not the current system of nation states does more harm than good, some geographical division of the labor of government seems likely to be part of the best total institutional scheme.65 But on the view being considered, equality and the well-being of strangers are no part of the reasons why we value well-known argument for this view see Michael Walzer, Spheres of Justice (New York: Basic Books, 1983), chap. 2. 65. See Pogge, "Cosmopolitanism and Sovereignty," for a very interesting suggestion of a territorial nesting of political authority.

29

30

John Rawls

278

Philosophy & Public Affairs

states. So long as there is some minimalist or modus vivendi alternative that would satisfy the requirements of prudence, why would we then accept the cost of forming genuine political communities?66 Putting the point the other way around, those of us of an egalitarian disposition must, on this view, reject utterly the appeal of the Utopian ideal of the withering away of the state. Once again we have a dualist view motivated by an implausibly strong claim about the significance of institutions for justice, in this case focusing on the institutions of a particular sovereign state. Of course the two special features of states are undoubtedly of normative significance. If wrongs committed by way of official state action are worse than the wrongs of private citizens, these features would figure in any explanation of that asymmetry. My point is the specific one that the two features of states are not sufficient to motivate the idea that distributive justice in its full-fledged form is to be understood essentially as a constraint on legitimate governance. Once again I have not done more than scratch the surface of the argument. One or other of the two motivations for dualism may well be defensible against my intuitive objections, and there may be other motivations for dualism I am not aware of. But it does seem fair to say that we have not found any overwhelmingly strong reasons in favor of departing from the more traditional monist view. We can now turn to what I think is the strongest, or at least clearest, reason to resist any such departure. VII. INSTITUTIONS AND NONIDEAL THEORY The premise of my argument is that an acceptable theory of justice must have acceptable implications for both ideal and nonideal theory. Rawls makes it very clear in A Theory of Justice that in his view the appropriate order of argument is to address ideal theory first: we ask first what principles of justice would be appropriate in a world where institutions are stable and supported by all people. Only after our ideal has been specified, he says, can we begin to think about nonideal situations.67 But even if Rawls is right about the correct order of argument, the nonideal theory 66. There is some similarity between this point and Nozick's criticism of Rawls's use of the notions of cooperation and reciprocity; see Nozick, Anarchy, State, and Utopia, pp. 193, 223. 67. See, e.g., A Theory of Justice, pp. 8-9, 245-46.

John Rawls

279

Institutions and the Demands of Justice

that is generated by the ideal theory must itself be plausible. I am not saying that a theory of justice is incomplete until every detail of nonideal theory is fully worked out, but rather that if we follow Rawls and devote most of our attention to ideal theory we should at least make sure that our ideal account has no clearly implausible implications for the associated nonideal theory.68 I find this initial premise hard to doubt. After all, none of us will ever live in a society where all people fully support and comply with the principles of justice, not even via an overlapping consensus.69 What we will always appeal to in practice are principles of nonideal theory. If our theory has implausible implications for the nonideal case, the theory may have some intellectual interest, but it would fail as a normative political theory. That is the rather simple background premise of my argument. And the remainder of my argument is rather simple too, amounting to a single intuitive claim. It seems to me that any political theory that accepts Rawls's bifurcation of the normative realm into one set of principles for institutions and another for people will yield an implausible account of what people should do in nonideal circumstances. Thus there is a general reason to reject dualism. To see this, recall that on the dualist view the duty of people in respect of justice is to support and promote just institutions, not to aim directly at the ends those institutions strive to bring about. Cohen's incentives argument shows, I believe, that this position is problematic even for ideal theory. The institutions may be the best they can be, but if people can do more to promote the aims of justice a view that refuses to extend the principles of justice to personal conduct is prima facie deficient. But the underlying issue emerges most clearly when we focus on nonideal theory. Recall Rawls's brief remarks on the natural duty of justice for the 68. See Annette Baier, "Theory and Reflective Practices," pp. 207-27 in Postures of the Mind (Minneapolis: University of Minnesota Press, 1985), p. 212: "Most moral theorists who outline and defend norms which are valid in conditions of full compliance ... give some recognition to the need for supplementary accounts of what moral norms hold good in more realistic conditions, and give some promissory notes about working back from the world they have presented in their theory to the actual world. Maybe such promises can be kept, but only when they have been kept will we know just how much guidance the full-compliance world provided in the journey away from it, back to the actual situation in which we act." 69. On the difference between Rawls's earlier notion of a well-ordered society and his later notion of an overlapping consensus, see Rawls, Political Liberalism, pp. xvi-xvii.

31

32

John Rawls

280

Philosophy & Public Affairs

nonideal case: we are to assist in the establishment of just arrangements when they do not exist, at least when this can be done with little cost to ourselves. This means that in an unjust society, a society of great inequality or great suffering by the worst-off, people are required to promote institutions that will alleviate the inequality and suffering. On the dualist view, our concern with inequality or the amount of suffering abroad is necessarily mediated through institutional structures. But if equality or well-being is the underlying concern that produces a theory of justice, why would people not be directly concerned about these things? If people have a duty to promote just institutions, why do they lack a duty to promote whatever it is that just institutions are /or? A bad answer will be that the most effective way to promote equality or wellbeing is to promote just institutions. This is a bad answer because when it is correct it is the answer that anyone should give. As I have said, if we accept monism we regard institutions as of overwhelming practical significance. In many nonideal situations the best thing an individual can do to alleviate inequality or suffering is to support appropriate institutional reform. But monism does not see the responsibilities of people as fundamentally mediated by the question of institutional design in the way that dualism does. Note that this point is similar to, but not the same as, the objections made to the two motivations for dualism. I objected to the interaction and democratic legitimacy views on the ground that the reasons offered for dualism did not seem strong enough tq justify the stark moral discontinuities these views involved. These objections, though important at the level of abstract theory, may not seem to have that much importance—especially not that much practical importance—once we are on the right side, as it were, of the discontinuities. The current objection is to the practical implications of granting a fundamental mediating role to institutions within the domain of justice, whatever that may be. The case to focus on is of course a nonideal situation where it is not true that the best way for people to alleviate inequality or promote wellbeing is to promote just institutions. For here monism tells people to do what they can to bring about an improvement directly. If injustice is about inequality, people should do what they can to reduce it. If they can have a greater impact on inequality by aiming directly at its reduction than they would have if they directed their energies to institutional

John Rawls

281

Institutions and the Demands of Justice

reform, this is what they should do. Dualism tells a different story: even if the individual could do more to reduce inequality, alleviate suffering, or whatever, by direct action, this is not what justice requires her to do. Justice requires her to promote just institutions even if she is sure that the aim of the just institutions she is promoting would be better served if she herself pursued that aim directly. How could this be right? The point is clearest in the international context. Here a dualist would believe, following Pogge, that justice requires an egalitarian set of institutions to replace the mostly informal and decidedly inegalitarian institutions that currently prevail.70 But it could not be right that an individual rich First Worlder is required to devote her resources to the Quixotic task of promoting just international institutions. Such a person could clearly do so much more to alleviate suffering or inequality by doing what she can on her own—by giving money to humanitarian aid agencies. With a stark example like this, dualism starts to seem fetishistic. Though less stark, I believe the point is equally valid for the domestic case. What is the best thing a just American billionaire can do with the fraction of his fortune that he doesn't really need?71 Donate it to the Democratic Party? Run for president or found a "third" party? Set up a political action committee to pressure members of Congress to pursue a more just welfare policy? Or might it be better for him to donate the money directly to inner-city hospitals or schools around the country (with adequate oversight, of course)? This last example illustrates a further important point.72 To say that billionaires should do what they can directly, rather than through the reform of institutions, is not to say that they should ignore currently available institutional structures. It may be best for them to use existing institutional structures as the most efficient way to promote well-being or equality. It bears repeating that the point I am making has nothing whatsoever to do with some idea that individualized charity is generally superior to institutional programs as a means for the achievement of 70. See Pogge, Realizing Rawls, Part III, "Globalizing the Rawlsian Conception of Justice"; "A Global Resources Dividend." 71. In late 1998 there were 189 American billionaires; see Forbes Magazine, October 12, 1998. 72. I am indebted in this paragraph to Lewis Kornhauser.

33

34

John Rawls

282

Philosophy & Public Affairs

justice.73 Though this may .sometimes be the case in the domestic context too, my argument does not in fact depend on its being a serious possibility. Rather, the fundamental point is that it is not always best, from the point of view of justice, to devote one's efforts to making institutions, considered in themselves, just. Dualism tells us that principles of justice govern the design of institutions. Monism tells us to use whatever means are available to promote justice in a society. Those best means might include the promotion of fully just institutions, the support and funding of existing unjust institutions, or some direct noninstitutional effort to promote well-being or equality. Furthermore, even when our efforts are best directed toward improving or creating institutions, it may sometimes be appropriate to aim at institutions that are not fully just; in a particular nonideal world, fully just institutions may be dangerously unstable.74 It might be thought at this point that dualism could be adapted to answer my concerns, at least for those cases where less than perfect institutions seem to be the best instruments for the achievement of the aims of justice. It might be thought that an appropriate nonideal theory for dualism would instruct us to design whatever institutions would, in the prevailing circumstances, best promote the underlying aims of the ideal theory that describes fully just institutions. In other words, dualism could embrace the idea of principles of justice for the design of nonideal institutions in a nonideal world. But such a position would actually be less intuitively plausible than the standard version of dualism. For once we accept that the principles that govern the design of ideal institutions essentially describe means to an end, the oddness of thinking that justice is concerned with some means to that end but not others becomes rather evident.75 It should now be clear that it is not an accident that Rawls is committed to both the primacy of the basic struc73. Indeed, like Cohen, and for similar reasons, I find it somewhat distasteful even to think about individual efforts to promote justice; see "If You're an Egalitarian, How Come You're So Rich?" 74. Cf. Cohen, "Incentives," p. 395: "[According to an ancient Marxist wisdom, justice is not the first virtue of institutions in conditions of scarcity. Under those conditions a just distribution may be impossible to achieve, since powerful people will block it. In that case striving for justice may make everyone worse off, and unjust laws and institutions should not be 'reformed or abolished.' " (Quoting Rawls, A Theory of Justice, p. 3). 75. A related point is made by Michael Phillips in "Reflections on the Transition From Ideal to Non-Ideal Theory," Nous, 19 (1985):563-64.

John Rawls

283

Institutions and the Demands of Justice

ture as the subject of justice and the primacy of ideal theory; the two positions are mutually supporting. It is worth emphasizing that a dualist cannot respond by simply agreeing that the responsibility a person has in respect of justice in the nonideal case is to do whatever she can to promote the aims expressed in the principles of justice. That is the straightforward abandonment of dualism, of Rawls's central idea that principles of justice do not apply directly to individuals. Of course, we saw that defenders of the interaction view and perhaps even the democratic legitimacy view could adopt monist versions of their theories of justice; but that is not what concerns us here. Perhaps dualist views could provide principles of some third kind to guide personal conduct in the nonideal case, but I have no idea what considerations would govern the formulation of such principles.76 The point I have made is simple and intuitive. Where people live within the domain of justice they must concern themselves with the substantive political/moral aim expressed in the principles of justice, be that the aim of equality, the aim of increasing weighted well-being, or some other aim. It is not credible that what fundamentally matters is that the relevant institutions promote equality or well-being, rather than that equality or well-being be promoted. Even if the existence of just institutions has some kind of expressive value, that could scarcely be more important than the actual achievement of greater equality or increased well-being. The point can be missed if we think only about ideal theory. It is tempting to shrug off Cohen's incentives argument on the ground that any society whose institutions are doing the best they can must after all be just enough.11 But if we look to nonideal theory and the nonideal case—the actual case of flagrant failure by most societies and by the world as a whole to secure justice—then that response is not available. Whatever motivation one might have for embracing the dual76. In his most recent article Pogge suggests that people in nonideal circumstances must either promote just institutions or, failing that, compensate the victims of the current unjust institutional regime; see "The Bounds of Nationalism," p. 503. This does not seem to be compatible with a dualist interpretation of his view, since we do not compensate people for an ongoing violation of their rights, we stop violating their rights—which in this case means that we stop imposing unjust institutions on them. 77. I am not suggesting that Cohen himself has ignored the importance of the nonideal case; see "If You're an Egalitarian, How Come You're So Rich?"

35

36

John Rawls

284

Philosophy 6» Public Affairs

ist view, it is hard to see how it could survive reflection on its implications for the theory we need in the actual world. VIII. PURE PROCEDURAL JUSTICE My discussion may be thought to reflect a fundamental misunderstanding of the nature of Rawlsian political theory, and thus of dualism. I have in mind the idea that while Rawlsian political theory expresses a conception of pure procedural justice, my arguments, which make frequent reference to the aims of just institutions, proceed from an incompatible consequentialist perspective. Pure procedural justice, in Rawls's definition, "obtains when there is no independent criterion for the right result: instead there is a correct or fair procedure such that the outcome is likewise correct or fair, whatever it is, provided that the procedure has been properly followed." As an example, Rawls mentions a fair bet.78 Pure procedural theories of justice thus evaluate a social system on the basis of the intrinsic character of its procedures or practices.79 A consequentialist theory of justice, by contrast, evaluates the justice of a society by looking to the effectiveness of its procedures in achieving a certain social goal: a certain degree of equality in society, a maximal promotion of weighted well-being, etc. On first sight it might seem that a pure procedural theory of justice is compatible with dualism only.80 But in fact nothing in the idea that the justice of a society's practices is determined, not by the degree to which those practices promote a certain result, but rather by their intrinsic character, forces us to confine our interest to a society's institutional practices. Nagel offers examples of what a pure procedural theory would be concerned with in the intrinsic character of a society's procedures: "what kinds of causes they permit to determine social outcomes, 78. A Theory of Justice, p. 86. 79. See Nagel, "Justice and Nature," Oxford Journal of Legal Studies 17 (1998) 1304. 80. It might seem that the second half of p. 282 in "The Basic Structure as Subject" shows that Rawls sees a connection between the idea of pure procedural justice and dualism. But I do not read Rawls as there offering the idea of pure procedural justice as further support for dualism. Rather, he is explaining the operation of the difference principle— understood both as incorporating an element of pure procedural justice and as applying to background institutions only (and thus delivering the benefits of the division of institutional labor). The claim that a "fair distribution can be arrived at only by the actual working of a fair social process over time" does, however, require further comment; see p. 287 below.

John Rawls

285

Institutions and the Demands of Justice

whether they discriminate between people on grounds that are unfair, whether they fail to treat people as they deserve, whether they penalize people for things that are not their fault, and so forth."81 It seems clear enough that all these things could be determined by social practices that fall short of institutional practices. As we have said, it is of course desirable to draw a normative distinction between (say) unfair discrimination by the state and unfair discrimination by an individual. But the normative difference between official and unofficial wrongful action is not what marks out the idea of pure procedural justice; that distinction can be made within the terms of an account of pure procedural justice that evaluates the intrinsic character of both institutional and noninstitutional social practices. Moreover, not all institutions represent the state in its official capacity. So the division between consequentialist and pure procedural political theories cuts across the division between monism and dualism. My discussion of what will "better achieve the aims of justice" is innocent too, since it need not refer to the achievement of a lower level of inequality or a higher level of well-being in society; the aims of justice are also better achieved— according to pure procedural political theory— when the intrinsic character of social practices conforms more nearly to some ideal. It might now be said that though there is a sense in which we can say that the aims of a pure procedural theory of justice are better achieved the more nearly the intrinsic nature of a society's procedures conform to some ideal, this does not implicate individuals at all, since they can do nothing to change a society's practices. But why would we think that people are incapable, on their own, of improving a society's procedures or practices? It is true that no one person can single-handedly make a society's practices perfect, but then neither can any one person eradicate suffering or inequality. To develop this point, we can consider what is for us the most important item on Nagel's list of characteristics of social practices—"what kinds of causes they permit to determine social outcomes." There are a range of possible criteria for the evaluation of the way social outcomes are generated in a society. Nagel himself explores (and rejects) the possibility that naturally caused inequalities are not unjust: if this were 81. "Justice and Nature," p. 304.

37

38

John Rawls

286

Philosophy Public Affairs

right, we would say that a society's procedures are just if purely natural differences between people determine social outcomes, but unjust if socially caused differences do.82 Similarly, Dworkin and others argue that it is unjust if social outcomes are determined by (some kinds of) brute luck, by events over which a person had no control.83 It is clear enough that an individual can do something about eradicating inequalities caused in either of these ways. Her efforts will have to be more sophisticated than if she were simply concerned with inequality of well-being however caused, but that raises no fundamental bar to action. If she is much richer than another person, for reasons that are social, or, alternatively, due to brute luck, a significant transfer from her would make it the case that in her society, social factors or brute luck are less productive of inequality than they otherwise would be. Even if she does not make these transfers on a regular basis, it is accurate to say that each particular transfer has the effect of making her society's procedures better. It might next be objected that this kind of incremental improvement in a society's practices is not what pure procedural accounts are about. Such accounts aim rather to describe what fully just social practices look like. But of course it cannot be the case that pure procedural theories fail to rank as better and worse social practices that fall short of the ideal; if this were so, such theories would be entirely useless to human beings doomed to apply only nonideal theory.84 So we must assume that the nearer a society's procedures conform to the ideal, the more just is the society. And thus, since social practices go beyond institutional practices, it must be accepted that in nonideal circumstances a person may best be able to promote justice by improving noninstitutional social practices. My argument against dualism applies to consequentialist and pure procedural theories of justice alike. 82. See "Justice and Nature." 83. See "What Is Equality? Part 2: Equality of Resources," Philosophy &> Public Affairs 10, no. 4 (fall 198:0:283-345. Dworkiris view is a good deal more complicated than is suggested by this distinction, and indeed not all inequalities caused by factors over which a person has no control are thought by Dworkin to be unjust; see also "What Is Equality? Part i: Equality of Welfare," Philosophy &> Public Affairs 10, no. 3 (summer i98i):i85-246; "What Is Equality? Part 3: The Place of Liberty," Iowa Law Review^ (1987)11-54; "Foundations of Liberal Equality." G. A. Cohen defends the idea that inequalities caused by factors over which a person has control are not unjust in "On the Currency of Egalitarian Justice," Ethics 99 (l989):906-4484. Dworkin himself makes this point in "What Is Equality? Part 3: The Place of Liberty," pp. 38-45-

John Rawls

287

Institutions and the Demands of Justice

But if all this is right so far, what now do I say about these words of Rawls: "A fair distribution can be arrived at only by the actual working of a fair social process over time in the course of which, in accordance with publicly announced rules, entitlements are earned and honored. These features define pure procedural justice."85 Does not this passage make it clear that pure procedural justice cannot be promoted by people in the way described? Dworkin makes a similar remark, in the context of a brief discussion of nonideal theory: "since a just distribution cannot be established counterfactually, but only through just institutions, we are unable to judge what share of our wealth is fair."86 These remarks by Rawls and Dworkin are somewhat misleading, I believe, or at least easily misinterpreted. Shortly after the just-quoted words, Rawls writes: "if it is asked in the abstract whether one distribution of a given stock of things to definite individuals with known desires and preferences is more just than another, then there is simply no answer to the question." This is so, but, in the context of Rawls's criteria of justice, not obviously relevant. It is true that compared to a criterion of justice that distributes goods to definite individuals with known desires and preferences the difference principle is a principle of pure procedural justice. But that does not mean that a fair distribution, in the relevant sense of a distribution of primary social goods in accordance with the difference principle, can be arrived at only by the actual working of a fair social process over time. The same point applies for the fragment of Dworkiris theory of distributive justice presented earlier. It is true that what counts as a fair distribution of resources at any one time cannot be determined just by looking at who has what. On a view that distinguishes between different causes for inequality in resources, a more subtle inquiry is required; among other things, we need to know whether one person has more than another because of choices they both made, or whether the disparity is due instead to brute luck. But this does not mean that we do not know what a fair distribution is without setting up a procedure and letting it run; a fair distribution is one where, among other things, there are no inequalities of resources that are due to factors beyond people's control. Now Dworkin's actual theory of distributive justice is very complicated, and it may be that informational problems would make it impossible for people to promote justice on 85. "The Basic Structure as Subject," p. 282; see also, A Theory of Justice, p. 88. 86. "Foundations of Liberal Equality," p. 261.

39

40

John Rawls

288

Philosophy Public Affairs

this account. But that would be a practical rather than a theoretical bar. There is thus no reason in principle why on (monistic variants of) either Rawls's or Dworkiris account of distributive justice people cannot take direct steps to make their society more just, using the most effective available means.87 IX. THE DEMANDS OF JUSTICE Rawls's division-of-labor argument explicitly links dualism to the attractiveness of a less intrusive, and in that sense less demanding, theory of justice; I endorsed the importance of the division of labor for the issue of demands but rejected the connection with dualism. In the arguments that I drew on to present the interaction and democratic legitimacy views, Pogge and Dworkin also give an important place to the issue of the demands of justice. In both cases, though in different ways, the institutional view is presented as more reasonable or plausible with respect to demands than any view that holds people directly responsible for 87. This discussion illustrates the fact that the notion of pure procedural justice makes sense only relative to some baseline distribution, a distribution that the pure procedural theory makes no judgment about. (For a related point, though addressed to Nozick's notion of an entitlement theory of justice, see Robert J. van der Veen and Philippe Van Parijs, "Entitlement Theories of Justice," Economics and Philosophy \ [1985] 170-74.) Thus the difference principle, which demands that institutions achieve a certain result, namely equality of primary social goods except where inequality is to the benefit of the worst off group, and which would thus at first sight seem to be a consequentialist criterion, nevertheless counts as a criterion of pure procedural justice relative to a "distribution of a given stock of things to definite individuals with known desires and preferences," since according to the difference principle people's desires and preferences are not relevant in the determination of distributive shares. The reason Rawls finds it relevant to point this out, I believe, is this. It is natural, when thinking about distributive justice, to look directly to the actual distribution of well-being or goods to particular people and ask whether that distribution is just. This distribution is what ultimately matters to the people involved. This is why it is appropriate to point out that, relative to this distribution in particular, justice as fairness embodies pure procedural justice: it pays no attention to the question of which people— with their particular desires and preferences—occupy which places in the distribution of social goods. (See Pogge, Realizing Rawls, p. 203: "[the difference principle] embodies an element of pure procedural justice insofar as it regulates how individuals get 'distributed' over the various index positions.") The upshot of this point would seem to be that the distinction between pure procedural and consequentialist theories of justice is not all that deep. Both kinds of theory give people a result to aim at; the only difference is that in the case of pure procedural theories the result in question is typically not a certain distribution of well-being or resources (which is what people really care about), but rather a distribution of well-being or

John Rawls

289

Institutions and the Demands of Justice

achieving the aims of justice.88 It is in the issue of demands, I suspect, that the most powerful underlying motivation for dualism can be found. The standard way of thinking about the problem of what are reasonable moral/political demands focuses on the cost or sacrifice imposed on complying agents.89 If we take this view of the problem, however, dualism is actually of no help. It is true that monistic nonideal theory, if it requires people to do as much as they can to promote equality or wellbeing, seems to be extremely demanding on many well-off people— especially in a cosmopolitan version. But a requirement that people do whatever they can to bring about the existence of just institutions would be just as demanding. Dualism changes the content of the demands of justice but does not evidently reduce their extent. One reason why dualism might be thought to help with the problem of demands is that we naturally consider the dualist position through the lens of ideal theory, and it is clear that the demands on each person are less under ideal institutions—which ensure that all people make their contributions— than in a nonideal situation of partial compliance. Even for the ideal case, however, the standard approach to the problem of demands might lead us to count, say, the difference principle as extremely demanding on those people who would do better in some alternative, less egalitarian social system.90 resources that can be traced to certain causes. This is why I do not follow Nagel in referring to pure procedural theories of justice as deontological ("Justice and Nature," p. 304). Deontological moral principles condemn or require actions of certain kinds, without reducing the goodness or badness of the action to the goodness of badness of its results. Deontological political principles must then condemn societies that allow, say, brute luck to determine social outcomes, even if doing so promotes, say, the level of aggregate wellbeing. The idea is not that it is a good result if brute luck does not determine social outcomes; rather, allowing brute luck to determine social outcomes is simply something that societies must not do. But since societies are not super-agents, but just collections of people, the only way to understand this "must not do" is that people should be more concerned about whether practices in their society allow brute luck to determine social outcomes than about other goals, such as the aggregate level of well-being. When pure procedural or "deontological" theories of distributive justice are incorporated into people's practical lives, they seem to collapse into consequentialist theories. 88. See Pogge, e.g., Realizing Rawls, p. 34, "A Global Resources Dividend," p. 502; see Dworkin, "The Roots of Justice." 89. See, e.g., Shelly Kagan, The Limits of Morality (Oxford: Oxford University Press, 1989), chap. i. 90. Nagel puts the issue of the sacrifice (measured against a baseline of feasible alternative systems) that can be reasonably be required of people in an ideally just society at the center of Equality and Partiality.

41

42

John Rawls

290

Philosophy & Public Affairs

In any case, it is clear that for nonideal theory, the theory we need, dualism neither helps nor hinders with the problem of demands as standardly understood. Rawls himself is clearly aware of this. His natural duty of justice for the nonideal case contains the caveat "at least where this can be done with little cost." Here he does not make the implausible assertion that the demands of justice on people in nonideal circumstances only apply when they bring little cost— such a view would doom us to injustice pending some miracle of general supererogation. Rather, in writing "at least," he leaves the problem of demands for another day. What I want now to suggest is that Rawls's commitment to dualism can be read in such a way as to suggest an entirely different approach to the problem of demands. We can start by noting that Rawls includes no cost caveat to the natural duty of justice for the ideal case. And he explicitly rejects the relevance of comparing people's levels of well-being under justice as fairness with the levels they would enjoy in some feasible alternative scheme.91 On the standard approach to the issue of demands, this seems odd, since we after all can very well make sense of such comparisons. If the natural duty of justice requires billionaires to vote for the difference principle, supposing they could vote instead for laissez-faire, is it not clear that the natural duty of justice is asking rather a lot of them? Why shouldn't there be a question of the reasonableness of the demands, somehow measured, made under ideal compliance?92 One reason may be that there actually is no appropriate baseline available for such assessments of demands—in either the ideal or the nonideal case.93 Another may be that the standard approach simply misdiagnoses our intuitive concerns about demands. Rawls's focus on the basic structure and his assertion of a limit to individual sacrifice in the nonideal case only lend support to a diagnosis of our concerns about demands that looks, not to the extent of the demands people face, but rather to the question of what the appropriate demands on a complying agent are in nonideal situations. When we think institutionally, we naturally think collectively. We think that justice is something for all of us as a group, 91. See "The Basic Structure as Subject," 278-79. 92. See Nozick, Anarchy, State, and Utopia, pp. 189-97; G. A. Cohen, Self-Ownership, Freedom, and Equality (Cambridge: Cambridge University Press, 1995), pp. 223-26; and n. 90. 93. See Liam B. Murphy, Moral Demands in Nonideal Theory (New York: Oxford University Press, forthcoming), chap. 3.

John Rawls

291

Institutions and the Demands of Justice

rather than for each of us individually. In the ideal case of full compliance we are all doing our share, and, we might think, there is no issue about whether the demands, on all of us, are reasonable. On this interpretation, it is no accident that Rawls adds no cost caveat to the natural duty of justice for the ideal case; the point would not be that the demands in the ideal case are moderate enough, but rather that no issue of demands arises since we are all doing our fair share. The real issue of demands emerges when only some of us do our part in carrying out a collective obligation. Here the utilitarian story, which requires each person to keep benefitting others up to the point where this does as much harm as good, despite the fact that each would need to do much less if all were doing their share, can be regarded as failing to take into account the collective nature of the demands of justice. One way of taking the collective nature of the demands of justice into account would be to regard the duty of each person in nonideal circumstances as limited to her share of the collective burden.94 Here I mention a view about reasonable demands in the nonideal case that I hold but which obviously cannot be attributed to Rawls. But we do have here another possible reason why we might be tempted to think that dualism helps with the problem of demands. Even if Rawls's commitment to the basic structure as the subject of justice does not reflect a sense that justice is a collective obligation, one that we must shoulder together, it naturally suggests such a view. If I am right, adopting such a view allows us to make progress with the problem of demands; and what it certainly does is explain why we might think that there can be no problem of demands under full compliance. In any case, however, dualism itself plays no role here. The idea that justice is a collective obligation is fully compatible with monism; nothing in the idea that my responsibility in respect of justice is my share of our collective responsibility in respect of justice necessarily implicates institutions. This is fortunate since, as I have argued, it is a mistake to think that justice is fundamentally all about institutional design. 94. See L. Jonathan Cohen, "Who Is Starving Whom?" Theoria 47 (i98i):6s-8i; Parfit, Reasons and Persons (Oxford: Clarendon Press, 1984), pp. 30-31; Murphy, "The Demands of Beneficence," Philosophy &> Public Affairs 22, no. 4 (fall i993):26y-92, and Moral Demands in Nonideal Theory, much work would be required before the view presented in the last work could take into account problems of institutional design.

43

This page intentionally left blank

[2] The Claims of Reflective Equilibrium Joseph Raz Balliol College, Oxford

As well as propounding a theory of justice, Rawls's A Theory of Justice advocated a method in moral philosophy, the method of reflective equilibrium. My aim is to raise doubts concerning the significance and value of this method. I will first briefly describe the process of reasoning dubbed by Rawls Reflective Equilibrium. I will then examine some of the claims made for it by Rawls, Nielsen, and others.

I. The Method of Argument by Reflective Equilibrium 'Let us assume that each person beyond a certain age and possessed of the requisite intellectual capacity develops a sense of justice under normal social circumstances' (p. 46).l The sense of justice is here conceived as a skill in judging things to be just and unjust and in supporting these judgments with reasons. Provisionally one can say that the object of moral theory is to provide 'a formulation of a set of principles which, when conjoined to our beliefs and knowledge of the circumstances, would lead us to make the judgments that our sense of justice disposes us to make' (p. 46). This provisional description of the process of establishing a reflective equilibrium has to be augmented and modified in four important respects. First, though the quotation above is cast in the plural, suggesting that the inquiry is into the beliefs which people of our culture generally tend to accept, this impression is misleading. The inquiry is into each person's moral beliefs and the reflective equilibrium is one achieved by a person between his disposition to make certain moral judgments and to adduce certain reasons to back them and the moral principles which would lead to the making of those very moral judgments. Rawls hopes that different people may converge round the same reflective equilibrium, but he remains uncommitted on this point and provides no reason for believing in such a convergence.2 Secondly, though the quotation sounds as if the equilibrium is between principles provided by moral philosophy and particular moral judgments made by a person, this is not really Rawls's intention. Moral philosophy provides principles which, given the person's beliefs, would lead him to

46

John Rawls

308 Joseph Raz make all the judgments he is disposed to make, be they general or particular.3 One possibility is that the principles chosen by moral philosophy will just be a subset of that person's explicitly avowed beliefs, from which all his other beliefs follow. Thirdly, the equilibrium is achieved not between the principles and all our moral judgments but between those principles and our considered moral judgments only. Those are 'those judgments in which our moral capacities are most likely to be displayed without distortion' (p. 47). Tor example, we can discard those judgments made with hesitation, or in which we have little confidence. Similarly, those given when we are upset or frightened, or when we stand to gain one way or the other can be left aside' (p. 47). 'Considered judgments are simply those rendered under conditions favorable to the exercise of the sense of justice' (p. 47). The person making the judgment is presumed, then, to have the ability, the opportunity and the desire to reach a correct decision' (p. 48). They are in fact similar to the conditions which 'single out considered judgments of any kind' (p. 48). Fourthly, the person whose views end in reflective equilibrium is presented 'with all possible descriptions to which one might plausibly conform one's judgments together with all relevant philosophical arguments for them' (p. 49). Confronted with these alternatives a person may abandon some of his pre-philosophical considered judgments in order to be able to embrace a set of principles which would not have been in equilibrium with them. Thus the process of achieving an equilibrium itself may change the person's existing moral views. Rawls's procedure is not a foundationalist but a holistic one. No moral judgment is immune from change on the route to a reflective equilibrium.4 Nor does Rawls stipulate an upper limit to the permissible change. However unlikely, it is possible that by the time the agent reaches reflective equilibrium he may have abandoned all his prephilosophical moral judgments. To summarize, the method of argument advocated by Rawls begins with a person who accepts many moral judgments of various degrees of generality. He is confronted by all possible sets of moral principles and the philosophical arguments for them. In the process of considering them he may or may not abandon some or all of his initial moral beliefs and he may or may not accept other moral beliefs. The process is at an end when all his considered moral judgments are adequately supported by a set of principles that he accepts in combination with his other beliefs. This is an idealized description of the method. In practice no one can be presented with all the possible sets of moral principles, nor is there any guarantee that the process of reflective equilibrium will ever end. A person may continually vacillate between two moral systems, finding the one more

John Rawls

Claims of Reflective Equilibrium

47

309

convincing at one time and the other at another, or being convinced that one moral theory provides the best solution to one problem but that another incompatible theory is better suited to solve another. For our purposes we can confine our attention to the credentials of the idealized method. I will not challenge it on the ground that the impossibility of following it in practice undermines its value. What does the method of reflective equilibrium described above actually establish? A person starts the argument disposed to assent to a class of moral judgments.5 He ends up disposed to assent to a class of moral judgments which may or may not be identical with the class he started from. If he has any reason at all to have greater confidence in his moral views at the finishing line than he had at the start, it must derive from the nature of the process which led him from start to finish. We are told that in between he has examined all the possible sets of moral principles and arguments for them. It is presumably the quality of this examination which provides the grounds to trust its outcome. But Rawls says nothing about it. The individual understands the principles and the arguments he is confronted with. He contemplates them and ends up with certain moral views, which may or may not have been affected by his contemplation. We are told nothing more. May we infer that none of the other features of the process affect the reliability or acceptability of its outcome? Suppose I reach reflective equilibrium but that you know that my deliberations were greatly influenced by a large number of clearly false factual assumptions, or were arrived at by a series of logical howlers such as affirming the consequent, or were based on accepting a demonstrably false epistemological theory. Are you bound to say that my current views have profited from my engaging in the process and reaching reflective equilibrium? To say yes seems too absurd to bear contemplating. Yet if we reject this possibility what rooni is left for the method of reflective equilibrium? Are you to be allowed to reject the outcome of my deliberations simply on the ground that the Kantian method of moral argument is the only valid one and that it leads to the affirmation of Kantian moral principles which I rejected because though I examined Kant's moral method I failed to see its merit? If so, we are back to the usual philosophical argument about the merits and demerits of various methods of argument and of various theories. The method of reflective equilibrium is then not a method in moral philosophy at all. It simply advocates that our judgment be informed (look around to make sure you didn't commit a mistake through oversight) and consistent. I wish to suggest a third interpretation of Rawls's intentions. Inasmuch as one accepts in the process of one's deliberations non-moral factual, logical, or philosophical views or procedures, then one's resulting moral

48

John Rawls

310 Joseph Raz theory is flawed if those are flawed. But one's moral views held as a result of the method of reflective equilibrium cannot be flawed on the ground that one was unduly influenced by any other factor or consideration.

II. Teasing Out the Implications of One's Views What credence can one give to the method of reflective equilibrium? I will explore the problem in several stages, starting by making modest claims for the method and proceeding to examine progressively bolder claims which have been made for it. Some of the modest claims are rejected because they are too modest. They do not justify regarding the method as the basis of all argument in moral theory. Some are rejected on other grounds, but all of them are rejected provisionally only. For each of the more ambitious claims embraces the more modest ones, and if accepted would vindicate them regardless of the initial reservations about them when presented on their own without the backing of the deeper and more profound claims concerning the force of reflective equilibrium. The first modest claim is that the principles adopted in reflective equilibrium may help o'ne clarify the implications of one's own views. 'We can note', says Rawls, 'whether, in cases where our present judgments are in doubt and given with hesitation, these principles offer a resolution which we can affirm on reflection' (p. 19). One possible interpretation of this remark is that the improvement brought about by achieving a reflective equilibrium is in having moral views on more issues, or in having fewer doubts about moral issues. But this interpretation is unlikely to be correct if only because there are several fatal objections to the view it encapsulates. Having fewer doubts and having views on more issues is an improvement if there is reason to trust the additional views. Otherwise it is arguable that it is better to be committed to moral views on fewer issues. So the method of reflective equilibrium is welcome on this ground only if there is some other reason to believe that it leads to the endorsement of reliable views. Secondly, this interpretation seems to assume that there are clear and correct moral views on all issues. But it is possible that many issues are morally indeterminate. If so, there will inevitably be moral doubts and hesitations. For it is of the essence of indeterminacy that one who knows all there is to know about morality will be in doubt whether some concrete cases are indeterminate or not. Finally, there is nothing in the method of reflective equilibrium to suggest that its use will lead to more doubts being resolved than created. The result of applying the method may be that a person who was sure of his judgment on all issues now sees the other side of every coin and becomes very doubtful of many of his concrete moral

John Rawls

Claims of Reflective Equilibrium

49

311

judgments (while being quite sure about his new set of principles which are the source of his doubts in concrete cases), I would therefore suggest that the quoted remark draws attention to a different virtue of the method of reflective equilibrium. It is likely to make people better aware of the implications of views which they already hold.6 Our firm beliefs may well help us decide cases about which we are hesitant, and engaging in the process of reflective equilibrium may well help us see how. The undoubted value of seeing clearly the implications of our own views on many occasions cannot, however, blind us to the fact that these implications cannot have more force than the views from which they derive. The method of reflective equilibrium cannot be the method of moral theory unless it can provide us with grounds for confidence in our views, and not only with an understanding of their implications regarding which we were in doubt.

III. Arguments from Agreement The second minimalist interpretation is clearly stated by Nielsen: '[I]t is important to see the use of reflective equilibrium . . . as a method for making progress in moral argument, moving from initial disagreement in some of our considered judgments to agreement and consensus.'71 think it fair to take Nielsen to be pointing to the fact that the use of the method produces consensus as a reason for trusting its product. It may of course be the case that the very existence of consensus on certain matters, or in certain circumstances, lends support to the agreed views. Daniels, for example, comments that if the method of reflective equilibrium does lead to consensus then it will dispel '[o]ne traditional worry, that moral judgments are not objective because there is insufficient agreement about them. . . .'8 This is no doubt true, though Daniels is quick to admit that he knows of no reason which shows that the method of reflective equilibrium would lead to consensus. All he can do is detail his hopes that it might. I have already mentioned that Rawls does not claim that his method tends to produce a consensus among those who use it, and it may well be argued that the absence of convergence does not dispose of Nielsen's point. His argument is that the method of reflective equilibrium can at least be trusted to the extent that it does lead to agreement. It can be trusted by people who are led by it to agree with each other. They are justified in regarding the principles they agree to as the principles which should govern their relations. Perhaps there is support for such a view in the following remarks of Rawls's:

50

John Rawls

312 Joseph Raz [Justification is argument addressed to those who disagree with us, or to ourselves when we are of two minds. It presumes a clash of views between persons or within one person, and seeks to convince others, or ourselves, of the reasonableness of the principles upon which our claims and judgments are founded. Being designed to reconcile by reason, justification proceeds from what all parties to the discussion hold in common. Ideally, to justify a conception of justice to someone is to give him a proof of its principles from premises that we both accept, these principles having in turn consequences that match our considered judgments. Thus mere proof is not justification. A proof simply displays logical relations between propositions. But proofs become justification once the starting points are mutually recognized, or the conclusions so comprehensive and compelling as to persuade us of the soundness of the conception expressed by their premises. It is perfectly proper, then, that the argument for the principles of justice should proceed from some consensus, (pp. 580-1).

Rawls is here addressing himself to an objection directed at his contractarian procedure. But he ties up his remarks with his method of reflective equilibrium, and their generality allows one to consider them in this context as well. They certainly seem to confuse 'justifying the conclusion to him' with cjustifying the conclusion'. One justifies a conclusion to a person by producing a valid argument which starts from premises agreed to by that person (either because he accepts them initially or because he comes to accept them as a result of contemplating the argument). But it does not follow that by justifying the conclusion to that person (who may be oneself) one has provided a justification of that conclusion. The person may be very gullible and happy to accept rather silly premises. All that has been established is that he believes that the conclusion has been justified. It does not follow that it has been justified. Rawls may shift his ground and argue that even if his comments are misleading when applied to justification generally, they are valid when confined to moral justification, for 'one of the aims of moral philosophy is to look for possible bases for agreement where none seem to exist' (p. 582). Even if one accepts this view, however, I doubt if it can be used to support the method of reflective equilibrium. Many would agree that it is valuable to secure agreement based on moral truths. So much may be implied by the fact that they are moral truths. But it is a long and hazardous way from here to the view which Rawls may hold, namely that the fact that consensus can be achieved concerning a certain set of moral principles is a reason for holding them to state moral truths. Secondly, the argument recommends moral agreement on moral issues. It does not recommend any special way of achieving it. It therefore does not recommend the method of reflective equilibrium which has yet to be shown to be at all efficient in securing agreement. And it does not recommend it above other ways of achieving agreement, except possibly for its efficiency, if it is at all efficient. I have dwelt on the arguments from consensus and agreement, for it is

John Rawls

Claims of Reflective Equilibrium

51

313

these that Nielsen has emphasized. My conclusion is that they bear no weight when standing on their own. It remains to be seen whether some deeper interpretation of the method of reflective equilibrium will reinstate them by providing the required theoretical underpinning.

IV. Reflective Equilibrium as a Heuristic Device The arguments from agreement canvassed above have the attraction that if successful they bypass knotty epistemological problems such as whether there are any moral truths and how to establish what they are. If we are morally bound by what we agree to, regardless of whether or not it is true, then why bother with moral truth if agreement is possible in ignorance of it? Of course all this would mean is that agreement on a view makes it 'true for those who agree to it' ? but that is enough at least to reduce the need to know whether anything is true regardless of agreement with it. Rawls has a second and very different way of understanding the force of the method of reflective equilibrium which appears to represent it as advancing moral philosophy while bypassing epistemological problems. But in fact it is a first step towards the solution of such problems. He anticipates it in A Theory of Justice where he says: '[I]f we can find an accurate account of our moral conceptions, then questions of meaning and justification may prove much easier to answer. Indeed some of them may no longer be real questions at all' (p. 51). This remark is developed in The Independence of Moral Theory' in a direction which as I will show below is inconsistent with the main thrust of the discussion of reflective equilibrium in A Theory of Justice. I suggest that for the time being we put aside the idea of constructing a correct theory of right and wrong, that is, a systematic account of what we regard as objective moral truths. Since the history of moral philosophy shows that the notion of moral truth is problematical, we can suspend consideration of it until we have a deeper understanding of moral conceptions. . . . So provisionally we may bracket the problem of moral truth and turn to moral theory: we investigate the substantive moral conceptions that people hold, or would hold, under suitably defined conditions. In order to do this, one tries to find a scheme of principles that match people's considered judgments and general convictions in reflective equilibrium.9

It is important to realize that the method of reflective equilibrium is here advocated not as a second best. It is not merely something to keep moral philosophers occupied and to stop them from impeding the progress of moral philosophy sby giving way to the impulse to answer questions one is not yet equipped to examine'.10 The method of reflective equilibrium is in fact a necessary step towards the solution of the problems of moral epistemology:

52

John Rawls

314 Joseph Raz [Jjust as the theory of meaning as we now know it depends on the development of logic from, let's say Frege to Gôdel, so the further advance of moral philosophy depends upon a deeper understanding of the structure of moral conceptions and their connections with human sensibility.11

It is far from clear how Rawls envisages that moral theory will contribute to moral epistemology. It is clear, however, that despite his reference to reflective equilibrium his independent moral theory has little or nothing to do with reflective equilibrium. According to the conception of The Independence of Moral Theory', One thinks of the moral theorist as an observer, so to speak, who seeks to set out the structure of other people's moral conceptions and attitudes. Because it seems likely that people hold different conceptions, and the structure of these conceptions is in any case hard to delineate, we can best proceed by studying the main conceptions found in the tradition of moral philosophy and in leading representative writers, including their discussions of particular moral and social issues.12

There is no trace here of Rawls's insistence that 'the views of the reader and the author are the only ones that count' (p. 50). Rawls and his readers are merely observers delineating the contours of the different moral systems current among people generally. Their own moral views have no special status. Nor has reflective equilibrium much relevance to this conception of moral theory. It may help the observer capture the contours of coherent moral systems. It may, in other words, be a device for ridding people's moral beliefs of inconsistencies and freeing moral systems of conflicts with known facts and physical, social, or philosophical theories. '[O]ne seeks the conception, or plurality of conceptions, that would survive the rational consideration of all feasible conceptions and all reasonable arguments for them.'13 But if reflective equilibrium has no further functions, then it is dispensable in favour of deriving coherent moral conceptions from the writings of moral philosophers where they are to be found more or less ready made, I have little doubt that moral theory so conceived as the phenomenology of moral opinions is a valuable heuristic device likely to improve our understanding of the moral experience. This role of reflective equilibrium, however, assigns it little or no value in validating any moral view. It may appear that surviving the test of reflective equilibrium is a necessary condition of moral truth. But this impression is hollow. All we are told is that incoherent moral systems are to be discounted. It is chimerical to think of reflective equilibrium as a test of coherence, because the process of transition from the initial pre-philosophical state to the finishing line in reflective equilibrium is yet again entirely unspecified. All the other interpretations of the method that we have examined or will examine claim that the very fact of acceptance of a moral system or of agreement,

John Rawls

Claims of Reflective Equilibrium

53

315

consensus, or convergence on a moral system in reflective equilibrium gives reason to believe in it. Not so the view advanced in The Independence of a Moral Theory'. There the very survival of a moral system gives it no credence. I have no more reason to trust the moral system I accept in reflective equilibrium than to accept the one you do. Neither does the fact that several people agree to one system in reflective equilibrium show that it is likely to be valid, nor does it show that it is valid for them, that they are bound to govern their relations by it. But if survival is in itself no advantage, then it is the quality of the reasoning which leads from the start to reflective equilibrium which determines the quality of the product. Having been given no instruction as to the nature of that reasoning, there is here in fact no role for reflective equilibrium as such. We are forced to fall back on all the coherence-establishing arguments familiar from elsewhere.

V. Describing our Moral Sense Rawls's own main claim for his method is this. It is a way (the way?) to establish a theory of one's moral sense, of one's moral capacity. 4[O]ne may think of moral philosophy at first (and I stress the provisional nature of this view) as the attempt to describe our moral capacity' (p. 46).14 Why should moral philosophy be interested in describing our moral sense? One answer which suggests itself is that our moral capacity is constitutive of morality and determines its content. This view is susceptible of various interpretations. Our moral capacity may turn out to be genetically determined and to be the same in all members of our species. It may be genetically determined but variable within the species. Or it may be culturally determined allowing for greater or smaller variations between cultures and within cultures. Because of these various possibilities the task of describing our moral views in reflective equilibrium is not to be equated with the task of moral philosophy as a whole. Furthermore, not only can it be accomplished without a complete semantic and epistemológica! account of moral discourse and of moral beliefs, but its accomplishment is an essential step in developing an appropriate semantic and epistemological account. For knowing the structure of moral systems which survive the test of reflective equilibrium, and knowing their number and their degree of similarity, may help determine whether or not morality as a whole or any part of it is a biological species-uniform phenomenon, etc. But holding moral philosophy to be the theory of our moral sense is no re-affirmation of the role of reflective equilibrium as a heuristic device, which was criticized in the previous section. For this time we start from

" 54

John Rawls

316 Joseph Raz the assumption that our moral capacity is constitutive of morality and determines its content. Hence, to find out the moral principles which bind me, all I have to do is study my moral sense. Moral epistemology and semantics must await an answer to questions like 'is my moral sense the same, in whole or in part, as that of all members of my culture or of the whole human race?' But long before we will be anywhere near to answering those questions I will know what is morally required of me, for I will have reached (or at least I will have significantly approximated) a state of reflective equilibrium. What is the force of this interpretation of morality and the role it assigns to reflective equilibrium? One of its strengths is to make clear an important difference between a psychological and a philosophical theory of our sense of justice. The former may take the form of correlations between, say, the level of sugar in one's blood and one's tendency to approve or to reject certain moral propositions. Or it may relate one's experiences in infancy to one's views about authority. A philosophical description of a moral sense will not include such principles. A metaphor which comes to mind is that while psychology may describe a person's moral disposition from the outside, philosophy is confined to an insider's view of the sense of justice. A philosophical description takes the form of a set of moral principles, and they are principles which the person whose sense of justice is investigated would accept as valid. Not only do they represent his moral judgments as a coherent system, but they are acknowledged by him to do so. This guarantees that the principles do not merely match his moral judgments. They present them in the way that he perceives them, interrelated in the way that he himself (at least after achieving equilibrium) interrelates them. To distinguish a theory of a moral sense which is subject to these constraints from others I shall call Rawls's philosophical theory a theory of the internal constitution of a moral sense. The view I attribute to Rawls is that morality is the internal constitution of the moral sense, and that the method of wide reflective equilibrium is the one by which that internal constitution can be best ascertained. This interpretation of Rawls is, subject to a few differences, R. M. Dworkin's Dworkin highlights 'Rawls's exciting if imprecise idea that [the deep moral principles reached in reflective equilibrium] are constitutive of our moral capacity'.15 He then provides a strikingly vivid interpretation of the role of the method of reflective equilibrium: It treats intuitions of justice not as clues to the existence of independent principles, but rather as stipulated features of a general theory to be constructed, as if a sculptor set himself to carve the animal that best fits a pile of bones he happened to find together. This 'constructive' model does not assume, as the natural model does, that principles of justice have some fixed, objective existence, so that descriptions of these

John Rawls

Claims of Reflective Equilibrium

55

317

principles must be true or false in some standard way. It does not assume that the animal it matches to the bones actually exists.16

I find this description very helpful except that it includes and lays much stress on one idea familiar from Dworkin's own work but not part of Rawls's method. Dworkin represents a moral system in reflective equilibrium as richer and more extensive than the pre-philosophical judgments which provide its starting-point. The sculptor creates an animal incorporating a few found bones. We know that this picture is misleading inasmuch as found bones may be discarded. In fact, strictly speaking, none of them needs to be included in the final sculpture. But the mistake that I am particularly concerned to correct here is that a moral system in reflective equilibrium needs to be richer and more comprehensive than the prephilosophical bones which served as the starting-point for its construction. There is no such requirement in Rawls's description of his method. It is true, however, that Rawls on occasion expresses himself in a way suggesting that he has hopes that the method of reflective equilibrium will enrich one's sense of justice by leading to views being held on more possible situations, and to more refined and complex views being maintained. At one point he says: A correct account of moral capacities will certainly involve principles and theoretical considerations which go much beyond the norms and standards cited in everyday life; it may eventually require fairly sophisticated mathematics as well. (p. 47)

This may be a reasonable prediction of likely developments in the psychology of decision and action., But since the correct account Rawls is referring to consists in setting out the internal constitution of the moral capacities, the sophisticated mathematics he is envisaging will have to be part of the principles which people will accept as representing, at the most abstract level, their moral views. Does moral advance depend on the mastery of sophisticated mathematics? Is such a development necessary or desirable? The desirability of a moral change in such a direction is itself a matter of moral controversy to be settled presumably by the method of reflective equilibrium itself. But is there no reason to think that such a development is necessary to represent correctly our moral capacities as they are at the moment? An argument to that effect could be that it is impossible to derive our many and immensely varied particular moral judgments from the usually avowed vague and prima facie moral principles, regardless of how much factual information is used in the process. This may be true, but the argument fails nevertheless. It presupposes that morality has a certain structure wherein particular judgments 'are derived' from general principles. It may well be that morality does not have this character. It is possible

56

John Rawls

318 Joseph Raz that our moral capacity is exercised not by 'deriving' the particular from the general, but by exercising a skill, i.e. using judgment within a framework of principles. According to this conception of morality our moral capacity can be partially explained through an analogy with expertise in French wines. Judgments of the relative merits of French wines are 'objective' and can be true or false. There are even some general principles which govern them. For example, a wine should have a balance of acid and tannin. A wine should not be too mature for its age. But the expert's judgments are not derivable from such principles and his factual value-free beliefs. They rely on a skill in judging wines, which is, at least up to a point, teachable. Though of course it is not taught by deriving particular judgments from brute facts and abstract principles. It belongs to this conception of morality that even those with a good moral judgment are not infrequently in doubt about the precise and ultimate verdict on a situation. Morality so conceived is likely to include a fair degree of indeterminacy. In more recent writings Rawls seems to move in such a direction. He stresses that his 'view accepts from the start that a moral conception can establish but a loose framework for deliberation and judgment which must rely very considerably on our powers of reflection and judgment'.17 Even though the method of reflective equilibrium may lead to a change of one's moral views, there is no general reason to hope or expect that it would radically change the degree of specificity of our moral principles or change their vague or prima facie nature.

VL Reflective Equilibrium and Our Moral Sense - A Critique I will argue that if morality is the internal constitution of our moral capacity then the method of reflective equilibrium is futile. Its decisive failure is its inability to guide our choice between moral views. My first objection merely claims that this view of morality cannot justify Rawls's discussion of ill-considered moral judgment. It is followed by two further objections which show that his method cannot offer guidance in choosing between conflicting views held concurrently or diachronically. (1) If the purpose of the method of reflective equilibrium is to help individuals discover the internal constitution of their moral capacity, then why should the process start by excluding a class of moral judgments called 'unconsidered judgments' from the examination? After all, are they not an expression of our moral capacities just like any other moral judgments we make? One cannot answer this question by saying that the judgments thus excluded are ones we have no confidence in and that we lack confidence

John Raw I s

Claims of Reflective Equilibrium

57

319

in them because they are ill considered. If we had no confidence in them they would not be our judgments and would not be included in the starting-point anyway. Some of them may be judgments we are unsure of, but that is no reason to exclude them at the start. All it means is that they are more likely to be jettisoned in the process of reaching an equilibrium. During that process any view, however well considered, may be abandoned. What can justify abandoning any of our views at the start? Any one of them may help illuminate the moral capacity which gave it birth. It is tempting to think that Rawls's view, that only considered moral judgments need be in equilibrium with the moral theory which is accepted in reflective equilibrium, is a pointer to a realist interpretation of his views. That is an interpretation which regards the study of our moral capacity as a way of collecting evidence about the nature of an independently existing moral reality. But it is equally possible to see it as a remnant of an earlier stage in the development of the doctrine of reflective equilibrium. It may be a left-over from the times when Rawls tended to a foundationalism which saw moral philosophy as providing a theory predicting all the moral judgments a person ever made. If the equilibrium takes existing moral judgments as fixed points to be matched by a moral theory, the task may well seem impossible unless at least a class of ill-considered judgments is excluded. According to this suggestion the mature doctrine of reflective equilibrium can dispense with the doctrine that only considered judgments count. Whatever the origins of Rawls's elimination of all but considered judgments, I tend to agree with the above suggestion that their elimination contributes nothing to the doctrine of reflective equilibrium. I would like to go one step further and argue that Rawls's elimination of ill-considered judgments is inconsistent with the interpretation of morality as the internal constitution of our moral capacity to be discovered through the method of reflective equilibrium. Consider the following argument in support of Rawls's view about illconsidered judgments. It proceeds through an analogy, often drawn for other purposes by Rawls himself, between moral philosophy and linguistic theory. Some assume that it is part of the task of linguistic theory to describe our linguistic competence to tell a grammatical sentence from an ungrammatical one. It may be thought that a theory of that competence would consist in linguistic rules, e.g. that sentences without a verb are ungrammatical, or that adjectives must have the same gender as the noun they qualify. However, it would be naïve to assume that such a theory of this linguistic competence could accurately predict our performance. We are likely to discover that people deviate from the predictions of the theory. The deviations will perhaps be explained as a result of haste, tiredness,

58

John Rawls

320 Joseph Raz anxiety, absent-mindedness, etc. That is, we will use additional and inore general psychological theories to explain why we sometimes fail to recognize that a sentence is grammatical or that it is ungrammatical. The same, it can be argued, is true of our moral capacity. The theory of our moral capacity consists in moral principles which can be used to predict our moral judgments. But the theory will not provide totally accurate predictions. To improve our predictive power the specifically moral theory has to be augmented by other more general psychological theories about behaviour under stress, about absent-mindedness and the like. Considered moral judgments are those which can be explained and predicted by the theory of our moral capacities by itself. Ill-considered moral judgments are those whose explanation makes use of the other psychological theories. Reasoning along these lines one could see why the ill-considered judgments will not be in reflective equilibrium with the moral theory which accounts for our moral capacity. But the reasoning here outlined is not Rawls's. It assumes that we discover which of our moral judgments are and which are not considered by working out a theory accounting for most of our moral judgments and then declaring those and only those which are • explained by it to be our considered judgments. Rawls, however, suggests that we identify our considered moral judgments prior to, and independently of, accepting any moral theory in reflective equilibrium. He assumes that this can be done by identifying those features which are 'favourable for deliberation and judgment in general' (p. 48). But much deliberation and judgment in general concerns independent existences, and it is far from clear that the conditions favourable for deliberation and judgments about 'the external world' are also conducive to deliberation and judgment expressive of an internal capacity not concerned with the descriptions of an external reality. We do not normally assume that music written in a rage is any the worse than music written in tranquil recollection. (2) This last comment leads naturally to the second and more fundamental objection to Rawls's exclusion of all but our considered moral judgments. Assume that we have identified our considered moral judgments. They are those which express our moral capacity by itself (i.e. they are predicted by the theory of our moral capacity by itself). Other moral judgments express our moral capacity as distorted by some other, psychological dispositions. Have we any reason to regard the ill-considered judgments less trustworthy? All that has been established so far is that their prediction and explanation are more complex since they require reliance on other psychological theories beside those which explain our specifically moral capacity. It is no answer to this question to say that they are not expressions of our moral sense. This is true by hypothesis, as we are concerned with moral judgments which cannot be explained as the expression of our moral

John Rawls

Claims of Reflective Equilibrium

59

321

capacity alone. The question I am raising is whether this is a reason to discount them, to regard them as false or unreasonable or untrustworthy.18 Can one answer this question by adverting to one important difference between the Rawlsian conception of morality as the internal constitution of our moral capacity and linguistic theory? As accounted for above, linguistic theory describes linguistic behaviour from the outside. Language users are unaware of its theory and even linguists do not judge the grammaticality of sentences by the use of the theory (except when they wish to test the theory itself). We know which sentences are grammatical independently of linguistic theory. The theory predicts our judgments. It does not explain how we come to make them; it does not describe how we reason about questions of grammar.19 Moral theory provides us not just with principles from which we can infer which moral judgments we will make. These very principles are also principles which we ourselves recognize as organizing our moral views in the best way we can think of. This, it may be claimed, explains our lack of confidence in 'ill-considered' moral judgments. They are at odds with our own conception of our trusted moral views. By hypothesis they are not merely impure products of our moral capacity but also untrusted products of our mind, for Rawls's very conception of the theory of our moral capacity, of the theory of its internal constitution, requires that it concerns all the moral judgments which we trust. The internal constitution of our moral capacity is just our way of representing the foundations of all moral judgments in v^hich we have confidence. This answer is, however, flawed. It assumes too much. All that we are entitled to accept is that our views on some occasions are at odds with our views on other occasions. We may also note that our views on most occasions are adequately represented by a certain moral theory, whereas on a small number of occasions our views differ from those represented by that theory. Suppose that we find ourselves in one of those small number of occasions. At that moment we accept moral judgments inconsistent with those we hold on most occasions. We know that later we are likely to change our mind, but do we have now a reason to distrust our present judgments? Do we now have any reason to discount them? I am not suggesting that we do not. But I am suggesting that the view of morality as the internal constitution of our moral disposition, as so far interpreted, lacks the resources to answer this question adequately. (3) At the beginning of this section I suggested that the Rawlsian conception of morality as the internal constitution of our moral sense cannot motivate his claim that we are able to identify our considered moral judgments in a way which is not essentially dependent on their identification as those judgments which we accept in reflective equilibrium. The second

60

John Rawls

322 Joseph Raz objection to his treatment of the subject is much more fundamental. Given that even after reaching reflective equilibrium we are likely from time to time to make moral judgments which are out of line with our considered moral views, does Rawls's theory under the present interpretation provide us with any reason to discount the ill-considered judgments? I suggested that it lacks the resources to do so. Essentially the same difficulty is raised by my third objection to the view that morality is the internal constitution of our moral capacities. The first two objections concern the standing of ill-considered moral views. That is, they concern the status of views which deviate from our normal views during the current period of our life, but which we firmly believe in during certain short periods. The third objection inquires into the status of our previous views, and of our future views, assuming that they may well differ from the views we currently hold in reflective equilibrium. Imagine a person, let us call him Jones, who when young held perfectionist romantic moral views regarding a person's life as worth while only inasmuch as it is effectively dedicated to the service of genius, be it the genius in himself or in others. Now in his middle age, Jones has abandoned those views in favour of utilitarianism. There may well be one psychological theory which will comprehensively explain his moral views during both stages. It may, for example, rely on generalizations such as: A person who believes in his own unlimited potential will tend to endorse perfectionist, romantic views; once a person recognizes his own limitations he will tend to endorse a morality emphasizing needs and the fragility of human achievement. Such generalizations have no place in Rawls's conception of the theory of our moral sense, since they describe it from the outside. The internal constitution of one's moral sense cannot encompass both of Jones's moral systems, because they are incompatible. Jones now rejects his youthful views, and the moral system he accepts in reflective equilibrium is utilitarian and has no trace of romantic perfectionism in it. The question is: Does the view of morality as the constitution of the moral sense discovered by reaching reflective equilibrium provide Jones with any reason to believe that his former views were less reasonable or worth while than his current ones? One possible answer is that it does if, but only if, Jones's current moral views are in reflective equilibrium. The fact that they are the product of the method of reflective equilibrium gives them credence. Does it? One cannot rely here on the soundness of the reasoning used in the process of reaching reflective equilibrium. Rawls makes no independent claims for the reliability of these ways of reasoning (except in so far as they concern non-moral issues such as factual assumptions or logical procedures). To the extent that they are valid their validity is to be inferred from the fact

John Rawls

Claims of Reflective Equilibrium

61

323

that they are the forms of reasoning which the method of reflective equilibrium leads to. They cannot, therefore, give us any reason to trust the method of reflective equilibrium, since that method is our only ground for having confidence in those forms of reasoning. Dworkin has suggested a different reason for trusting the method of reflective equilibrium. It rests, he suggests, on the 'assumption that men and women have a responsibility to fit the particular judgments on which they act into a coherent program of action'.20 But morality is unlikely to be a programme of action. It is more likely to be a mixture of restraining principles and an assortment of partial, approved goals from which each person may choose, as well as much else. In any case it is not clear why one should strive for coherence, at least if that term means, as Dworkin clearly takes it to mean, something more than logical consistency. If there is such a responsibility, it should be justified by the method of reflective equilibrium and not be used to justify that method, since if such a responsibility exists it is but one of our moral responsibilities. This is indeed the way Rawls himself treats the issue. He clearly admits the possibility that morality consists of a plurality of independent principles. It is even possible that it may consist of such a plurality of principles without ordering principles to resolve conflicts. The only way therefore to dispute this view is to set forth recognizably ethical criteria which account for the weights which, in our considered judgments, we think appropriate to give to the plurality of principles' (p. 39). Coherence is to be established through the method of reflective equilibrium, and no assumption of coherence is presupposed by that method itself. All we know of a person who follows the method of reflective equilibrium is that he has thought long and hard about moral issues and is (or rather was) aware of a large number of different moral systems. Does this awareness give us any reason to trust our views in reflective equilibrium? Not unless we are given a reason to think that moral views one holds while aware of their implications and of all their alternatives are more trustworthy than moral views held without this awareness. The reason we are looking for has to establish that the views held in awareness of their nature are more likely to be an accurate representation of our moral capacities. One such reason may be that one is less likely to change one's views if one is aware of their implications and of all the alternatives to them than if one is not. If that is the reason for preferring views held in reflective equilibrium then it is not a very powerful one. There might be other ways of establishing that people are unlikely to change their views. Indeed there are even other ways of fortifying them against changing their views. On the other hand, one might have good grounds for suspecting that, though some people hold their views in reflective equilibrium, they are very likely

62

John Rawls

324 Joseph Raz to .change them. It is clear that none of these critical reflections has much force if our moral dispositions are largely genetically determined. There is ample evidence, however, to suggest that our genes do not uniquely determine our moral views, and that even in one society there are many people who vary significantly in their moral outlook. If so, it is hard to understand how reflective equilibrium can be trusted as the guide to the true internal constitution of our moral sense. It may perhaps be argued that I have raised the wrong question. Jones need not ask what reason the fact that his current views are held in reflective equilibrium gives him to trust them, or to prefer them to his youthful views. He already trusts his present views and prefers them to his old ones. We conceded as much when we said that they are his current views. It is not the purpose of the method of reflective equilibrium to give him reason to believe in what he believes already. Its purpose is to explain to him, and to us, the structure of his views. But this cannot be so. This explanation may do for narrow reflective equilibrium, i.e. for the unrevised provisional description of the method given on page 307 above. At that stage we assumed that the method consisted in taking a person's pre-philosophical views as fixed points and constructing a body of principles from which they follow (with the aid of that person's other beliefs). This task may plausibly be explained as an attempt to represent the structure of his views. But if that is what Rawls is aiming at, why move to a wide reflective equilibrium in which people are required to survey a variety of moral theories and to choose among them, even though the choice may involve abandoning much of their earlier beliefs? Let me put the point in a slightly different form. Imagine that Jones, who holds utilitarian views in reflective equilibrium, meets a friend who urges him to allow himself to be brainwashed in a special process which will turn him into a puritan fundamentalist. Does Rawls's theory, according to the interpretation we are now considering, help Jones decide what to do? Suppose that Jones begins to be attracted to fundamentalist puritan ethics. Does Rawls's theory advise him how to react to this new attraction? (I am assuming that he has not yet changed any of his moral views. It is just that he sees more and more force in the puritan point of view. I further assume that his conversion will not involve non-moral howlers or falsehoods.) If the method of reflective equilibrium merely explains the structure of Jones's views, then it clearly has no advice to give him in his present soul-searching. It does not provide a way of arguing for one moral theory or another. It is merely a way of explaining the meaning and structure of our views to us. The interesting point is that, even if the fact that Jones

John Rawls

Claims of Reflective Equilibrium

63

325

holds his views in reflective equilibrium is a reason for him to have confidence in them, this interpretation of the method does not help Jones decide what to do, because according to it the reason for trusting views held in reflective equilibrium is that they represent our sense of justice. But if Jones will allow himself to be brainwashed so as to accept another moral theory, then that will be proof that his sense of justice, at least once the change happened, is not represented by his earlier views. And therefore the fact that his views are held in reflective equilibrium provides no reason against changing them, even if the change does not conform to the method of reflective equilibrium. The problem that all my objections ultimately point to is that all this interpretation of the method does is to explain to us our present moral sense. It provides no reason against moral change, nor does it provide a reason for preferring our considered judgments to others. In the absence of a deterministic theory limiting the possibility of change quite drastically, this interpretation of moral philosophy and of the role of reflective equilibrium fails in its most fundamental task. It fails to explain that morality is normative and binding. Morality provides reasons for certain actions and beliefs which do not directly and exclusively depend on the fact that we already believe that there are reasons for such actions and beliefs. A moral theory which, like Rawls's according to the interpretation we are considering, fails to explain this fails in its main task.

VII. Rawls and Blackburn: The Multi-tiered Moral Sense The argument of the previous section did not challenge the view of morality as the internal constitution of our moral capacity. It showed that if this is what morality is, as Rawls seems sometimes to suggest, then the method of reflective equilibrium does not contribute much to the study of morality and to moral deliberation generally. Simon Blackburn has advocated a view of morality which also regards it as the internal constitution of our moral dispositions.21 None of the objections raised in the previous section poses any difficulty for his view. It differs from the view of morality canvassed above in that Blackburn regards our moral dispositions as multi-tiered. We possess dispositions to approve of certain moral dispositions and to disapprove of others. These dispositions can explain how we can reject ill-considered judgments. They account for our belief that we are fallible and that we may come to have better (or worse) moral views in the future. The projective theory indeed denies that the standard of correctness derives from conformity to an antecedent reality. It does not follow that there is no other source

64

John Rawls

326 Joseph Raz for it. And there is: to moralize at ail involves commitment to some way of using an input of information to determine an output of reaction. And we are extremely sensitive (and surprisingly unanimous: the general character of the wise man is usually painted in the same way) to the features in such dispositions which we count as flaws and excellences. But we often do not know how much our own sensibilities betray such weaknesses, and how much our age and culture encourages dispositions which, could we see them, or could we improve upon them, we would be unable to endorse. Of course, these evaluations of dispositions are themselves 'subjective': they are ours. But there is no circularity in using our own evaluations to enable us to assess, refine, improve upon, our own evaluations, any more than there is in rebuilding Neurath's boat at sea. Nor is anything given an axiomatic status (although, as I have remarked, our beliefs about what makes a good sensibility are often very firm): relying upon other planks we can criticize each plank in turn.22

Blackburn's theory differs from Rawls's primarily in attempting to answer the fundamental semantic and epistemological questions which Rawls declines to explore. But it leads to a view of moral theory not unlike Rawls's. It is essentially based on our pre-philosophical moral views, while enabling us to question and reject any number of them. This is not the place to examine the adequacy of Blackburn's account. But it is worth pointing out that his view can be used both to explain and to downgrade the method of reflective equilibrium. If the method has any appeal, that appeal derives from and is explained by a second- (or third-) order disposition that some or all of us share. This is a disposition to approve of moral views held as a result of engaging in the method of reflective equilibrium. This disposition is itself 'subjective'. Some people who have moral capacities may not share it. (Though of course we may condemn them for this lack.) Rawls's own theory allows for the possibility that when different people use the method of reflective equilibrium they may end up holding different moral views. But if the method is to be based on the foundation provided by Blackburn,23 even the persuasiveness of the method itself depends on subjective preferences and is not a requirement of rationality. Furthermore, even people who possess a preference for views held in reflective equilibrium are likely to have other higher preferences as well. The method of reflective equilibrium need not be the sole source of the validity of various forms of moral reasoning. They may be grounded directly in some of our other higher-order preferences.

VIII. The Possibility of Moral Realism I argued that if morality is the internal constitution of our moral capacity, then the method of reflective equilibrium is not the fundamental method of moral philosophy. But perhaps we should not view our moral capacity

John Rawls

Claims of Reflective Equilibrium

65

327

as constitutive of morality but merely as evidence pointing to the existence of a moral reality beyond it. Could one regard the method of reflective equilibrium as providing evidence for the existence of moral facts? One cannot object to this interpretation of the method of reflective equilibrium on the ground that it makes the method into a secondary, parasitic way of establishing moral truths. We would regard the study of astronomers as a secondary and parasitic method of studying astronomy. But the same reasoning does not necessarily apply to morality. The system of morality approved of by the method of reflective equilibrium may well include not only moral principles of right action but also, among others, principles specifying modes of valid moral arguments. These principles of moral reasoning will then be used to solve moral problems. But this does not mean that the method of reflective equilibrium is secondary to reasoning by the use of the established methods of moral argumentation. On the contrary, it is the method of validating all other forms of moral argument. They are valid only if they form a part of a moral system which is in reflective equilibrium. R. M. Dworkin has argued that Rawls's method is inconsistent with moral realism. He begins by claiming that a realist interpretation would require holding our pre-philosophical moral views as fixed points of reference and would not allow for the many modifications of them contemplated by Rawls when he describes how we could abandon pre-philosophical moral views to conform to a set of principles which we find attractive. Dworkin then proceeds to explain that he does not deny that an individual may recognize that some of his moral views are misguided in the same way in which he can recognize that some of his sense perceptions are misguided: I am in a different position from the color-blind man who need only come to understand that others' perceptions differ from his. If I believe that my intuitions are a direct report from some moral reality, I cannot accept that one particular intuition is false until I come to feel or sense that it is false. The bare fact that others disagree, if they do, may be an occasion for consulting my intuitions again, but if my convictions remain the same, the fact that others may explain them in a different way cannot be a reason for my abandoning them, instead of retaining them in the faith that a reconciliation of these with my other convictions does in fact exist.24

While this point is sound it hardly affects a possible realist argument based on convergence. It is not a person's divergence from others which would make him lose confidence in some of his views. It is the fact that the best explanation of the fact that those others converge away from him is that they converge around the truth. Hence the method of reflective equilibrium is consistent with moral realism. The way the method of reflective equilibrium can be used as evidence of the existence of moral facts was best explained by Daniels: (1) In a given area of inquiry, the methods used are successful in the sense that they

66

John Rawls

328 Joseph Raz produce convergence and a growth of knowledge; (2) the only plausible account of the success of these methods is that they lead us to better and better approximations to truths of the kind relevant to the inquiry; (3) therefore, we should adopt a realist account of the relevant objects of inquiry.25

Daniels does not propose any argument along these lines. It is at least in part doubts about the possibility of any sound argument of this form which led Blackburn and others to a rejection of moral realism. I find the rejection of moral realism premature. Consider the question of whether or not one should be a realist concerning the existence of states, or of legal systems. States and legal systems are social institutions. They are the product of human activity. Though I will offer here no argument to that effect, it seems to me very likely that the best explanation of why we have many of the beliefs we do have about states and the law is that we have them because of the existence of states and of legal systems with appropriate properties. If that is so then we should be realists about states and the law. But if there is such an explanation it is likely to have much wider applications. It is likely to establish a realist stance regarding many social institutions. If morality is a social institution, if it is constituted by a network of social practices of a certain kind, then we may well have to endorse a realist view of morality. This at any rate seems to me to be the most promising route to moral realism. If this is so, however, then reflective equilibrium has little to recommend itself. If morality is a social institution, then a certain degree of convergence within a society is to be expected. But it will not be best explained by the fact that people tend to converge because of reflective equilibrium. The explanation will be that they have the views they have because of their awareness of socially constituted moral facts. Here we have something which may be empirically testable. Daniels's proof requires that the use of reflective equilibrium tends to increase convergence more than moral education and the internalization of moral practices in other ways. The view of morality as a social institution assumes that the opposite is the case.

IX. Conclusion Perhaps the attraction of the method of reflective equilibrium can be explained as follows. One of the dogmas of much twentieth-century moral philosophy is that discourse divides into moral and non-moral, and that no combination of non-moral premises can rationally justify preferring one moral view to its contradictory. Many are unhappy with the dogma and yet uncertain what precisely is wrong with it. Rawls's method neither

John Rawls

Claims of Reflective Equilibrium

67

329

endorses the dogma nor rejects it, It offers moral progress without taking up any position on the way in which discourse can or cannot be divided into moral and non-moral, or on the ways in which non-moral premises can support moral conclusions. His way of by-passing the great stumbling block of twentieth-century moral philosophy is to recommend a method based on coherence and survivability. The basic flaw in his reliance on survivability is that (unless underpinned by an as yet elusive argument for moral realism) it fails in the basic requirement of a method of moral argument, i.e. the ability to guide the agent's choice of moral views. As to coherence (inasmuch as it is more than consistency), its appeal is based on an unwarranted assumption. It is the assumption that a moral outlook is superior to its rivals (other things being equal) if it can be represented by fewer and simpler principles, and if it yields more determined moral views on more issues. But morality may not be like that. Moral knowledge may consist in the possession of a skill rather than in propositions derivable from abstract principles. And any acceptable morality may include a high degree of indeterminacy. More legitimate is the attraction the method seems to me to derive from two of its features. First, it implies the proposition that the less likely one is to abandon a moral belief just because one has acquired more information, the more trustworthy that belief is. Secondly, it combines conservatism with reform. It builds on our existing views and yet allows us to revise any number of them. But however appealing one may find these two features of the method, they are not enough to validate it. They are shared by many other conceptions of morality and of moral progress. For example, these two features are equally implied by Blackburn's theory (see Section VII above) which has no room for reflective equilibrium except in a subordinate role, and by the view that morality is a social institution (Section VIII above).26 NOTES 1 All page references in the text are to J. Rawls, A Theory of Justice, Harvard University Press, Cambridge, Mass. 1971. Since Rawls is happy to apply his method to morality in general I shall refer to a sense of justice or to a moral capacity, interchangeably. 2 Ibid., p. 50. 3 J. Rawls, 'Outline of a Decision Procedure for Ethics', Philosophical Review, Vol. 60 (1951), No. 2. Rawls allowed only particular pre-philosophical judgments to count. But he has later come to revise this view. See his The Independence of Moral Theory', Proceedings and Addresses of the American Philosophical Association, Vol. 48 (1974-75), p. 7. 4 The Independence of Moral Theory', op. cit., p. 8. 5 Strictly speaking I should add here and later his disposition to withhold his assent from some moral judgments.

68

John Rawls

330 Joseph Raz 6 I have here in mind any justification of a moral view on the ground that moral judgments which one is firmly committed to lend it support. 7 K. Nielsen, 'Grounding Rights and a Method of Reflective Equilibrium', Inquiry, this issue, p. 293. 8 Norman Daniels, 'Wide Reflective Equilibrium and Theory Acceptance in Ethics', Journal of Philosophy, Vol. 76 (1979), p. 274. 9 The Independence of Moral Theory', op. cit., p. 7. 10 Ibid., p. 10. 11 Ibid., p. 6. See also A Theory of Justice, op. cit., pp. 51-52. 12 The Independence of Moral Theory', op. cit., p. 7. 13 Ibid., p. 8. Is this all one may expect from philosophy generally? Is it the only form of non-coercive philosophy, one which does not offer proofs but explains how certain views are possible? Is it, that is, an example of R. Nozick's conception of proper philosophical arguments as explained in Philosophical Explanations, Clarendon, Oxford 1981, Ch. 1. 14 The provisional nature of this statement anticipates the broadening of a narrow reflective equilibrium describing the moral capacity as it is into a wide reflective equilibrium which describes a moral capacity which may have changed in the process of reaching the equilibrium. This point will be discussed later in this section. I use 'moral sense' and 'moral capacity' interchangeably since that is the way Rawls uses these expressions. They do not refer to the capacity to form moral judgments in general, but to the tendency to endorse certain specific moral judgments. 15 R. M. Dworkin, The Original Position', N. Daniels (Ed.), Reading Rawls, Blackwell, Oxford 1975, p. 26. 16 Ibid., p. 28. Dworkin's use of 'constructivism' is not to be confused with Rawls's in 'Kantian Constructivism in Moral Theory', Journal of Philosophy, Vol. 77 (1980), p. 515. 17 'Kantian Constructivism in Moral Theory', op. cit., p. 560. 18 The only,reason we discount aberrant linguistic judgments is that they fail to conform to our language, which is a social institution and not merely the expression of a mental capacity, not merely an acquired taste. 19 It is not my aim to condone or to condemn such linguistic theory. I only remark that this is all that theories like Chomsky's and others are entitled to claim for themselves. In any case this is the kind of theory with which the analogy with moral philosophy drawn above was made. 20 The Original Position', op. cit., p. 28. 21 See his 'Moral Realism', J. Casey (Ed.), Morality and Moral Reasoning, Methuen, London 1971, p. 101; Truth, Realism, and the Regulation of Theory', Midwest Studies in Philosophy, Vol. 5 (1980), p. 353; 'Rule Following and Moral Realism', S. H. Holtzman and C. M. Leich (Eds.), Wittgenstein: to Follow a Rule, Routledge & Kegan Paul, London 1981, p. 163. 22 'Rule Following and Moral Realism',.op. cit., pp. 175-6. 23 I should explain that Blackburn himself does not discuss the relation of his views to Rawls's and bears no responsibility for the following speculations. 24 The Original Position', op. cit., p. 34. 25 'Wide Reflective Equilibrium and Theory Acceptance in Ethics', op. cit., p. 278. 26 I am grateful to R. Gavison, J. Griffin, P. M. S. Hacker, and D. Heyd for helpful comments on a draft of this article.

[3] Constructing Justice for Existing Practice: Rawls and the Status Quo AARON JAMES

Even the most sympathetic reader of John Rawls's work can feel surprised and puzzled by one of his last published works, The Law of Peoples.1 One might have thought the powerful lesson of Rawls's career, and the very point of asking what would be accepted from his original position, is the essential importance of what is owed or justifiable to individual "free and equal" persons.2 Yet, in finally turning to justice outside the domestic context, Rawls poses the central question as one of what societies, or "peoples," owe to other societies. He could have offered reasons why, in the global context, we owe to individuals only what we owe to the societies of which they are members. But Rawls provides no such argumentation. There would be little difficulty in assuming a social order roughly like that which presently exists merely for the sake of "non-ideal theory." But Rawls is clear that ideal theory requires starting from "the international political world as we see it."3 One is tempted to conclude that Rawls misunderstood his own conception of I am particularly indebted to William Bristow, Marshall Cohen, and the Editors of Philosophy & Public Affairs for comments on this article. I also thank Samuel Freeman, Christine Korsgaard, Sharon Lloyd, Martin O'Neill, Thomas Ricketts, Gerasimos Santas, T. M. Scanlon, Cynthia Stark, Nicholas White, and the Southern California Law and Philosophy Group for comments or relevant discussion. 1. John Rawls, The Law of Peoples (Cambridge, Mass.: Harvard University Press, 1999). 2. As is argued by "cosmopolitans" such as Charles R. Beitz, Political Theory and International Relations, rev ed. (Princeton, N.J.: Princeton University Press, 1999); Beitz, "Cosmopolitan Ideals and National Sentiment," The Journal of Philosophy 80 (1983): 591-600; Beitz, "Rawls's Law of Peoples," Ethics no (2000): 669-96; Thomas Pogge, Realizing Rawls (Ithaca, N.Y.: Cornell University Press, 1989); Pogge, "An Egalitarian Law of Peoples," Ethics 23 (1994): 195-224; Brian Barry, "International Society from a Cosmopolitan Perspective," in International Society: Diverse Ethical Perspectives, ed. David R. Mapel and Terry Nardin (Princeton, N.J.: Princeton University Press, 1998). 3. Rawls, The Law of Peoples, p. 83.

70

John Rawls

282

Philosophy &> Public Affairs

how to reason about what justice requires. It is unlikely that Rawls would have committed such an egregious error, however, and there is a more charitable but no less striking conclusion: however well we thought we understood Rawls's domestic theory, it is we that have not fully appreciated what Rawls has been up to all along. I want to consider, in an exploratory spirit, the possibility that Rawls has all along been following a single, abstract "constructive" method, which begins from existing social practices. It is as follows: (1)

Identify an existing social practice, including its point, or the goods it is meant to realize. Assume circumstances favorable to its continuance. (2) Identify the practice's participants. Assume general compliance with its terms. (3) Design a suitable original position. That is: a. Represent each full participant in the practice as appropriately motivated, by an interest in the goods the practice is meant to create; b. Draw a veil of ignorance, behind which (i) all parties have the same information, and (ii) no one has knowledge of the facts that would undermine the fairness of an agreement on terms for distribution of the relevant goods; (4) Determine which terms of organization such parties would choose (among a list of candidate principles). Treat these terms as necessary conditions for the practice's being justly organized, that is, as principles of social justice. On this characterization of Rawls's method, original position reasoning has no authority as such; it must be grounded in independent judgments about what social practices exist and what kinds of agents participate in them. In other words, even if certain principles, and not the alternatives, would be chosen from a certain original position, it may well be that the alternative principles would be chosen from a different original position, which represents participants as "free and equal" in a different way4 Some further basis is needed to determine which original 4. That is, different goods might be in question, the parties may have greater or lesser information, and they may or may not be risk averse or self-interested. What cannot vary is only the "formal constraints of concept of right." Rawls, A Theory of Justice (Cambridge, Mass.: Harvard University Press, 1971), pp. 130-36, hereafter TJ.

John Rawls

283

Constructing Justice for Existing Practice: Rawls and the Status Quo

position is authoritative. For Rawls, that further basis is the extent to which the original position in question is "suitably" or appropriately "tailored given the agents modeled and the subject at hand."5 Or as Rawls also puts the point: In justice as fairness the principles of justice for the basic structure of society are not suitable as fully general principles: they do not apply to all subjects, not to churches and universities, or to the basic structures of all societies, or to the law of peoples.... It is the distinct structure of the social framework, and the purpose and role of its various parts and how they fit together, that explains why there are different principles for different kinds of subjects.6 Seeing Rawls as beginning from existing practices helps to explain how he could focus on persons in the domestic context and peoples in the global setting. He takes each context to require quite different judgments about what social practices exist and which agents participate in them. Although Rawls says very little in general terms about how these judgments are to be made, what he does say in very early papers, such as "Justice as Fairness," suffices to set the stage for his later work. Rawls characterizes "social institutions" or "practices" as "any form of activity specified by a system of rules which define offices, roles, moves, penalties, defenses, and so on, and which gives the activity its structure," offering as examples "games and rituals, trials and parliaments, markets and systems of property"7 Although he assumed that human individuals have a "certain logical priority," Rawls is clear that a practice could have as its participants "nations, provinces, business firms, churches, teams, and so on."8 In these terms, his focus on individuals in A Theory of Justice reflects his judgment that major domestic institutions assign offices and roles chiefly to individual persons.9 In the same way, his attention to 5. Rawls, The Law of Peoples, pp. 32-33, 86. 6. Rawls, "The Law of Peoples," in Collected Papers: John Rawls, ed. Samuel Freeman (Cambridge, Mass.: Harvard University Press, 1999), pp. 532-33. For parallel remarks, see Rawls, Political Liberalism, paperback ed. (New York: Columbia University Press, 1996), p. 262. 7. Rawls, "Justice as Fairness," in Collected Papers, ed. Samuel Freeman, p. 47, n. i. 8. Ibid., p. 48. 9. Although Rawls's notorious "heads of households" assumption suggests that individuals as such were not the loci of his concern in the domestic context either.

71

John Rawls

72

284

Philosophy & Public Affairs

whole societies in The Law of Peoples reflects his judgment that "international law and practice" constitutes the basic structure of global society, and that the participants in these practices are not individuals as such, but societies and their government representatives. Thus, a "global original position" that represents each person of the world would have been appropriate were the global scene more like domestic society.10 But an "international original position" is more appropriate to the "subject at hand," relations between societies.11 Although it is widely assumed by philosophers under Rawls's influence that existing practices set limits on what fundamental principles of justice could require, this assumption has only recently attracted critical attention.121 will not offer any conclusive assessment of this assumption. I do, however, want to further develop its underappreciated role in Rawls's thought, in a way that sheds light on the nature of Rawls's overall project. Admittedly, the suggestion that Rawls has always taken the subject at hand to be the practices that simply happen to exist squares more readily with The Law of Peoples than his earlier and better-known A Theory of Justice. It may therefore appear that Rawls simply changed his method. My initial task will be to show that Rawls has indeed reasoned from existing practices all along. This is clear in his early work, and it is compatible with his limited conception of ideal theory, as expressed within A Theory of Justice itself. I will also explain how, if this is correct, Rawls's method has "interpretivist" underpinnings; original position reasoning is partly grounded in what has come to be called 10. As argued in Beitz, Political Theory and International Relations. For skepticism whether the global and domestic contexts are analogous, see Brian Barry, "Humanity and Justice in Global Perspective," in Ethics, Economics, and the Law, ed. J. R. Pennock and John W. Chapman (New York: NYU Press, 1982). 11. This hardly implies that only relations between whole societies matter. Any of various practices associated with globalization can, for instance, still have principles of social justice (even if the Law of Peoples would forbid using force to promote their fulfillment). I outline a broadly Rawlsian account of fair trade in my "Distributive Justice without Sovereign Rule: The Case of Trade," Social Theory and Practice 31 (2005). Thus Rawls can admit what Sen calls "plural affiliations," and his view is not essentially a form of "national particularism," as suggested in Amartya Sen, "Justice across Borders," in Global Justice and Transnational Politics, ed. Pablo De Greiff and Ciaran Cronin (Cambridge, Mass.: MIT Press, 2002). 12. G. A. Cohen, "Facts and Principles," Philosophy & Public Affairs 31 (2003): 211-45 argues that principles grounded in any facts, not simply facts about existing practices, cannot be fundamental. I present a Rawlsian reply in an unpublished manuscript.

John Rawls

285

Constructing Justice for Existing Practice: Rawls and the Status Quo

"constructive interpretation" of existing practices. What this brings out is that Rawls is largely unconcerned with pure moral ideals; even his concern with the ideal theory requirements of social justice is not, and never has been, fundamentally divorced from philosophical anthropology. Although my aim is not overall assessment, the plausibility of seeing Rawls as reasoning from existing practices depends in part on whether we can cast such reasoning in a favorable light. If it is clearly indefensible, interpretive charity will require reading Rawls in some other way. The obvious difficulty is that original position reasoning as described above appears to be biased toward, or at least sensitive to, not simply the unalterable facts of human life, which Rawls takes to be neither just nor unjust, but also contingent features of the status quo. If the procedure described above is all there is to reasoning about social justice, and it directs us to focus on existing practices, it will follow that principles of justice cannot tell us what fundamental social organization to have in the first place. Moreover, if the procedure is all there is to reasoning about justice, and it leads us to focus on participants, we will be forced to ignore what should be the real possibility that a practice treats nonparticipants unjustly. Many have argued that the second kind of "status quo bias," toward those that happen to participate in a practice, is a serious flaw in Rawls's domestic theory13 As I explain in due course, insofar as there is a potentially significant concern of status quo bias, it is not this concern but the first, more fundamental concern with existing practices. Since Rawls does not take original position reasoning to be all there is to reasoning about social justice, he can account for the justice claims of nonparticipants by reasoning of another kind. Rawls does assume, however, that all reasoning about what social justice requires of us begins from existing practices, so the charge of bias toward existing practices needs to be taken head on. Although I will not put that charge fully to rest, I will suggest that there is little reason, short of overall assessment of Rawlsian 13. See, e.g., Brian Barry, Justice as Impartiality (Oxford: Clarendon, 1995); Allen Buchanan, "Justice as Reciprocity Versus Subject-Centered Justice," Philosophy & Public Affairs 19 (1990): 227-52; Allan Gibbard, "Constructing Justice," Philosophy & Public Affairs 20 (1991): 264-97; Robert E. Goodin, "What Is So Special About Our Fellow Countrymen?" Ethics 98 (1988): 663-86; Amartya Sen, "Open and Closed Impartiality," Journal of Philosophy XCIX (2002): 445-69.

73

74

John Rawls

286

Philosophy &> Public Affairs

justice, why beginning from existing practices manifests any obvious and objectionable form of bias. Rawls's reliance on "constructive interpretation" means he is not forced to describe our practices in ways that enshrine obvious forms of injustice; indeed, he can describe the practices that ground original position reasoning in moralized terms. Moreover, the idea that existing practices constrain what fundamental principles apply has an attractive rationale in the idea that principles of justice are moral constraints instead of ideal goals. PRACTICES IN IDEAL THEORY Rawls's focus on existing practices in his later work could be viewed as a departure from his considered view about justice as such, as expressed in A Theory of Justice. I begin by challenging this interpretation. Even in his early work Rawls explicitly assumes that reasoning about what justice requires of us must proceed from the social practices we actually have, and it seems unlikely that A Theory of Justice is simply an anomaly within his work as a whole. In "Two Concepts of Rules," Rawls urged the importance of distinguishing between justification of a practice and justification of a particular action falling under a practice.14 Rawls argued that utilitarianism is most plausible when it is seen as a standard of justification of the former kind, so it was natural to conclude that he would himself see the justification of a practice in much the same way as a utilitarian does, as a question of what practice to have in the first place given the likely consequences of all possible forms of organization. This turns out, however, not to be what Rawls had in mind. As "Justice as Fairness" makes plain, his own view would be that justification of a practice is quite unlike wholesale utilitarian justification of practices, and much more like justification of an action under a practice, which depends in part on the structure of practice in question. In describing what he called the conjectural situation, the immediate precursor to the original position, Rawls has us imagine, first, that the social practice in question is "already established, there is no question . . . as to how they will set up these practices for the first time," and second, that people choose terms of social cooperation in full knowledge of their actual social 14. Rawls, "Two Concepts of Rules," in Collected Papers.

John Rawls

287

Constructing Justice for Existing Practice: Rawls and the Status Quo

position.15 Rawls thought we could focus in this way on practices "already established," because even fully informed parties would be pressed toward impartiality out of rational self-interest; being uncertain about where in society they might wind up over time, they would choose principles that do not seriously disadvantage any particular social position. With impartiality secure, justification of a practice as socially just would otherwise be responsive to the kind of practice it is, and unconcerned with whether an independent goal of well-being or even equal distribution could be better promoted by disbandment or fundamental reorganization. Of course, Rawls might have changed his approach in A Theory of Justice] he could have assumed that justification is wholly insensitive to what practice is in question. Rawls's famous criticism of the "system of natural liberty" can indeed seem incompatible with focus on existing social organization. Rawls writes, "Intuitively, the most obvious injustice of the system of natural liberty is that it permits distributive shares to be improperly influenced by ... factors so arbitrary from a moral point of view" (TJ, p. 72). According to one common interpretation, here Rawls wittingly or unwittingly commits himself to the "luck-egalitarian" view that justice is realized only when the influence of arbitrary factors on distributive shares has been eliminated.16 The moral point of having a basic structure, then, is to compensate for undeserved inequality. If this is correct, Rawls would have broken sharply from his earlier approach, which allows "no question . . . h o w . . . [to] set up ... practices for the first time" (even, presumably, if an alternative would better promote the goal of eliminating undeserved misfortune). He would also have later rejected or ignored his earlier views in allowing that the domestic basic 15. Rawls, "Justice as Fairness/' p. 53. 16. Proponents of "luck egalitarianism" include G. A. Cohen, "On the Currency of Egalitarian Justice," Ethics 99 (1989): 906-44, and Richard Arneson, "Equality and Equality of Opportunity for Welfare," in Equality: Selected Readings, ed. Robert Westmoreland and Louis Pojman (New York: Oxford University Press, 1997). On the trend toward interpreting Rawls in these terms, see Samuel Scheffler, "What Is Egalitarianism?" Philosophy & Public Affairs 31 (2003): 5-39. Although the position is suggested by some of Rawls's well-known remarks in the first edition of A Theory of Justice, revisions made as early as 1975 for the German translation (the original appeared in 1971) make it plausible to hold that much of the suggestion is rhetorical overstatement. For discussion, see Thomas Nagel, "The Writings of John Rawls," in Concealment and Exposure (New York: Oxford University Press, 2002), pp. 82-83.

75

76

John Rawls

288

Philosophy &> Public Affairs

structure's fundamental requirements do not apply to global organization. For if justice is realized only when no one suffers from undeserved misfortune, what is required in the domestic and global contexts could only differ because different organizational means are needed to promote this more fundamental goal. This interpretation is not inevitable, and insofar as a different interpretation presents Rawls's appeal to arbitrariness in terms consistent with his earlier and later work, the presumption should be in its favor. For one thing, Rawls explicitly denied one version of the view that justice requires compensation for undeserved inequality, which he called the principle of redress.17 More important, his famous remark never claims that the influence of arbitrary factors is to be eliminated as much as possible, but only that distributive shares cannot be "improperly influenced" by arbitrary factors. This allows us to reconstruct the appeal to arbitrariness in terms of Rawls's earlier and later practice-sensitive approach. What Rawls sees in A Theory of Justice is that his earlier "conjectural situation" allows fully informed parties who start out well off to exploit knowledge of their advantageous position; they can safely bet that things will not get intolerably bad and hold out for terms that disadvantage those who start out in less well-off positions. Rawls therefore draws the veil of ignorance, in order to eliminate both the influence of the distributional status quo and other factors "arbitrary from a moral point of view." But this was not tantamount to introducing a new positive goal of eliminating any influence of arbitrary factors. The appeal to arbitrariness can instead have an entirely negative role, as a claim about what is arbitrary in light of the kind of practice the basic structure of society is. Let us say that a factor is arbitrary from a moral point of view just in case it cannot be part of a (good) justification of some activity that needs to be justified, because the factor in question is irrelevant to this purpose. That is, under the relevant conditions, the factor provides no reason at all, or no relevant reason, or no relevant reason of any weight, why the activity in question should be justified. Further morally 17. Rawls clearly distinguishes the Difference Principle from the principle that "undeserved inequalities [of birth and natural endowment] call for redress" or "are somehow to be compensated for" (27, pp. 100-01). Scheffler in "What Is Egalitarianism?" draws attention to this on p. 25.

John Rawls

289

Constructing Justice for Existing Practice: Rawls and the Status Quo

informed judgments are needed in order to determine what factors are arbitrary or irrelevant as reasons. But, for Rawls, any such judgments must be made against a background conception of what kind of activity or social practice is in question, a conception that itself generates initial presumptions about what factors are relevant. In competitive sports, the talented player who continually scores and leads the team to victory may well, in all fairness, receive greater shares of "the credit," pay, and esteem, simply because winning and excellence cannot be separated from its very purpose and point. What would be arbitrary, at least presumptively, is distributing the benefits of victory, or victories themselves, according to an irrelevant factor such as wealth (although only presumptively, because wealth might determine who wins without "improper influence/' if, in practice, it simply influences who becomes most talented). Similarly, for Rawls, it is now understood, at least in societies such as our own, that organization as a basic structure is, by its very nature, the coordination of activity in order to produce otherwise unavailable "primary social goods." Although it is in this case morally relevant whether or not one actively participates in the creation of these goods (by being a working or child-rearing, law-abiding good citizen), it is arbitrary or morally irrelevant, at least presumptively, how talented one is. A scheme that rewards the talented might still be justified, but only if the practice-grounded presumption of irrelevance can be removed by further moral judgments that directly or indirectly treat talent as a relevant and weighty consideration. Thus, Rawls allows that incentives for the better endowed meet the presumption against giving talent a determining role, so long as the reason in favor of this influence is that the incentives-driven scheme is ultimately of the greatest possible benefit of the worst-off person. Here talents are made relevant as a consideration for other reasons (namely, the incentives-based scheme advances the basic structure's aim of realizing primary social goods as well as each person's interest in being better off rather than worse off). It is a further question whether the system of natural liberty might be similarly justified, but, in Rawls's view, for any of several reasons the moral judgments needed to make talent a relevant factor are not available. For one thing, one will normally only count as "talented" because the basic structure of society is organized in a particular way, for example, because effort and ability "depends in large part on fortunate family circumstances, for

77

78

John Rawls 290

Philosophy &> Public Affairs

which [one] can claim no credit" (77, p. 104). As a variable endogenous to the social structure, talent, or even its exercise, could have no bearing favorably or unfavorably on the justification of the structure itself. Perhaps some beneficial traits are truly "natural," in the sense that they would create social benefit in virtually any basic structure. Even in such cases, however, it may not seem clear why, if one central function of the basic structure is the momentous task of assigning economic life prospects, such "natural" facts should be any more relevant that the number of hairs on one's head. Further, even if having natural gifts is some reason for one to benefit, this reason maybe so weak as to be effectively irrelevant if benefits are only possible because others do their part in the same scheme. The reason in question would not clearly override or even strongly counterbalance the claim anyone else has to share in socially created goods, in virtue of doing his or her part in the scheme that creates them. A defender of the system of natural liberty could of course reject this "intuitive" argument about what factors are arbitrary in the light of the basic structure's aim. What is important for our purposes is that, in Rawls's view, this disagreement is, in part, about what a basic structure is, about its understood nature and purposes, not about what people are owed independently of any practices. For Rawls, it is a fundamental mistake for libertarians to argue, as Robert Nozick did, that we have certain rights against the basic structure of society, simply because we would have those rights in a state of nature; any such rights must be defended anew, as the proper conception of justice in the light of what the existing basic structure is.18 If correct, libertarians have only two responses to Rawls's arguments. They can accept that there is a presumption against the relevance of natural endowment, in light of what the basic structure is, but argue that this presumption can be defeated. Or they can reject Rawls's conception of the basic structure's nature. It may be held, for instance, that the basic structure's essential aim is nothing other than to remedy the inconveniences of private rights protection. In order to defend this interpretation as a claim about the nature of existing basic structures, it would have to be quite generally understood 18. Robert Nozick, Anarchy, State, Utopia (New York: Basic Books, 1974). Nozick would of course deny that justice depends on existing practice.

John Rawls

291

Constructing Justice for Existing Practice: Rawls and the Status Quo

and accepted in the history, traditions, and contemporary opinions of societies we are familiar with. That may seem doubtful, or at least open to serious question. On the other hand, Rawls's own conception of the basic structure is also open to question on the same grounds. Can it really be grounded in ordinary behavior and opinion, or does it merely present a social ideal? I explain below how Rawls does indeed see his conception of the basic structure of society as an interpretive characterization. For present purposes, what is important is that Rawls's appeal to the arbitrariness of natural endowment is not a reason to view A Theory of Justice as an anomaly within Rawls's work. It can be reconstructed in practice-sensitive terms that are consistent with his early and later writings. We also find positive, albeit indirect evidence of continuity with his early and later work within A Theory of Justice itself, in the limited nature of his conception of ideal theory The very idea of "ideal theory" can seem to preclude assigning any fundamental relevance to existing practices. Is not the question precisely what practices to have in the first place? But this is not Rawls's question. For Rawls, ideal theory includes only two kinds of idealization, assumptions of "favorable conditions" and "strict compliance" (77, p. 245), both of which have determinate content only as idealizations of some particular kind of social practice. An assumption of "favorable conditions" does have some practiceindependent content. Circumstances such as relative isolation or heavenly abundance can render most any form of organization impossible or unnecessary. Conditions that are favorable in this general sense, however, may have quite different implications for practices of different kinds. Conditions may be "unfavorable" for a basic structure of domestic institutions unless "individuals coexist at the same time on a definite geographical territory" having "roughly similar . . . physical and mental powers" (77, pp. 126-27). Yet the same "circumstances of justice" may not be required to make international practices concerning the use of force "possible and necessary."19 The assumption of "strict compliance" also has general practice-independent content: according to Rawls, we are to 19. As Rawls suggests when he says of the circumstances of domestic justice: "if one supposes that the concept of justice applies whenever there is an allotment of something rationally regarded as advantageous or disadvantageous, then we are interested in only one instance of its application" (TJ, p. 8).

79

80

John Rawls 292

Philosophy &> Public Affairs

suppose that "principles of justice are chosen on the supposition that they are generally complied with. Any failures are discounted as exceptions" (77, p. 245). Yet here, again, what failures we are to discount depends on what practice is in question. In an established domestic basic structure, we are to ignore tax dodgers and the like. In an established practice for the use of international force, we are to ignore aggressive "outlaw states." To say ideal theory involves idealization oí some practices is not to say which practices we are to consider. It is not to say we should focus on existing practices. What, however, is the alternative? Suppose we simply imagine a practice, construct a suitable and authoritative original position, and reason to its principles of justice. Does this tell us what justice actually requires of us? For all we have said, the imagined practice, although logically possible or conceivable in theory, might be impracticable given the fixed conditions of human life. In that case, its realization could not constitute ideal justice in our world, and its principles would make no actual claim on us. Suppose, then, that we restrict ourselves to feasible practices that simply happen not to be established. Still, the principles applicable to such a practice, P, might be incompatible with the principles we might similarly construct for another feasible but not established practice, Q. If P and Q could not co-exist, or if they could co-exist only at the cost of injustice, then, again, neither class of principles would tell us what justice actually requires of us. Some further basis is needed to determine which of any principles that might apply to us do in fact specify what justice requires. Notice that this further basis is not provided by Rawls's "natural duty of justice," the duty "to support and to comply with just institutions that exist and apply to us" and to "further just arrangements not yet established," at least "without too much cost to ourselves" (TJ, p. 115). This duty clearly assumes an independent answer to the question of what institutions or arrangements are just. If the justification of any principles within ideal theory must be for a kind of social practice, then our question Which practices are we to consider? remains unanswered. It is sometimes assumed that the natural duty of justice requires us to realize whatever institutions are needed to do justice, whether by creating institutions where none exist, as in a state of nature, or by fundamentally reorganizing existing practices. Notice, however, that this does not follow from Rawls's formulation of the duty. The duty's first part, "to

John Rawls

293

Constructing Justice for Existing Practice: Rawls and the Status Quo

support and to comply with just institutions that exist and apply to us," is clearly meant to guide conduct with respect to existing practices. The same can be true for the duty's second part. To "further just arrangements not yet established" is, first and foremost, to further just arrangements not yet established within existing practices, that is to say, to support reform of the existing basic structure of institutions according to its applicable principles. In addition, the duty can require the creation of new practices when they are needed either (i) to sustain existing just practices (say, under changing conditions), or (ii) to reform existing practices according to their applicable principles (I offer examples of this from the global context below). The duty may still offer no guidance for agents wholly aside from any existing practices; it may not tell agents to leave the state of nature and create a basic structure, for instance. Indeed, if Rawls's argument for the duty is any indication of its intended content, it offers no such guidance: the argument Rawls gives explicitly assumes an established basic structure and its applicable principles.20 One could still argue, as a further claim, that natural duties would require the creation of a basic structure even when it does not exist. Rawls himself never clearly advances this thesis.21 The question remains, then, Which practices is ideal theory to consider? According to a conception of ideal theory stronger than Rawls's own, we are to consider the practices needed in order to optimize a certain ideal goal (such as the greatest happiness, equal distribution of

20. TJ, pp. no and 334-36. Rawls does note that "it would be possible to choose many of the natural duties before those for the basic structure without changing the principles in any substantial way" (TJ, p. no). Yet he is quite clear that the natural duty of justice is an exceptional case. Unlike many natural duties, and like "obligations" such as the principle of fairness, Rawls says "the duty to support just institutions" presupposes "principles for social forms." He assumes an established basic structure and its principles for all "principles for individuals," despite his admission that many could be independently defended, largely for the sake of simplicity. See TJ, p. no. 21. Even if we assume Rawls would accept a stronger, broadly Kantian duty to leave the state of nature and submit to the rule of law, what this entails might vary according to prevailing circumstances and practice. It may not involve submission to the modern state, but, as in many parts of present-day rural Africa (not to mention pre-modern ages), submission to local custom and dispute resolution by a tribal chief or local elders. One might argue that we need a still stronger, practice-insensitive version of the duty of natural justice in order to explain how existing practices might be in some sense fundamentally unjust. I consider whether Rawls has adequate critical resources below.

81

82

John Rawls

294

Philosophy &> Public Affairs

welfare or opportunities, or a ranked plurality of such values).22 In this case, principles chosen from an original position for a constitutional democracy would only actually apply if a constitutional democracy is optimal according to the favored values as compared to alternative schemes. Similarly for any principles that would be chosen from an original position for practices concerning the use offeree between societies; they would only actually apply if a system of societies is more optimal by the favored values than alternatives such as a single global constitutional democracy. In other words, Rawls's argument for his Two Principles as applied to a constitutional democracy, within a system of such societies, could be readily undercut; it would only need to be shown that such practices would not exist within a more optimal alternative global scheme.23 Now notice the crucial point: Rawlisan ideal theory includes no such optimality condition. According to Rawls, to insist on an equalitypromoting or other optimality condition beyond "the elimination of arbitrary distinctions and the establishment, within the structure of a practice, of a proper balance between competing claims" would fail to "distinguish that sense of equality which is an aspect of the concept of justice from that sense of equality which belongs to a more comprehensive social ideal."24 We will see below that Rawls does not take original position reasoning to exhaust reasoning about social justice. Here Rawls opts for a more restrictive choice. He simply excludes from the purview of social justice, in the sense that applies paradigmatically to "the structure of a practice," optimality according to any "comprehensive social ideal." Since there is no optimality condition on what practices we have in the first place, the implication seems to be this: no principles are 22. For similar views, see Arneson, "Equality and Equality of Opportunity for Welfare," Cohen, "On the Currency of Egalitarian Justice," Liam Murphy, "Institutions and the Demands of Justice," Philosophy & Public Affairs 27 (1998): 251-91, andAmartya Sen, "Rights and Agency," in Consequentialism and Its Critics, ed. Samuel Sheffler (Oxford: Oxford University Press, 1988), pp. 187-223. 23. Such as the overlapping domains of sovereignty described by Thomas Pogge, "Cosmopolitanism and Sovereignty," Ethics 103 (1992): 48-75. 24. Rawls, "Justice as Fairness," p. 48. Moreover, Rawls assigns no positive value to equality in itself (Rawls, The Law of Peoples, pp. 113-15), following T. M. Scanlon, "The Diversity of Objections to Inequality," Lindley Lecture: University of Kansas (1997). For discussion of similar considerations, see Charles Beitz, "Does Global Inequality Matter?" Metaphilosophy^\ (2001): 95-112.

John Rawls

295

Constructing Justice for Existing Practice: Rawls and the Status Quo

"more ideal" than the ideal theory principles that apply to the practices we actually have. The thought can be put in this way: Rawls claims that "[conceptions of justice must be justified by the conditions of our life as we know it or not at all" (77, p. 454). In his view, the conditions of life as we know it include not simply the unalterable conditions of human life, but forms of social organization that could well be fundamentally reorganized or abandoned, given enough time and favorable circumstances. They include, for instance, modern constitutional democracies in a world of distinct societies.25 What I have argued can be summarized as the claim that Rawls assumes what might be called the Existence Condition: any (fundamental, ideal theory) principle of social justice has as a condition of its application the existence of some social practice.26 The Existence Condition raises significant questions about how the practices that do exist are to be identified, and why Rawls interprets our social forms as he does. Before turning to this issue, however, it will be useful to first consider whether focus on existing practices might have a general rationale. One possible rationale is that a principle of justice functions not as an ideal goal, or optimality condition, but as a moral constraint: a requirement on the manner in which a form of activity is carried out. In understanding that a principle is a constraint, agents understand at least that (i) the principle applies to a form of activity of a specified kind, and that (ii), although optional from the point of view of the principle itself, the 25. In The Law of Peoples, Rawls takes an international order for granted without arguing or even overtly supposing that this is our only practically possible option. He does suggest, following Kant, that a world government would either be "a global despotism" or "rule over a fragile empire torn by frequent civil strife" (p. 36). He does not clearly take this to show that an order much like our own is inevitable, however. If he had, the "laws and tendencies" that make it inevitable would have to be identified independently of the order in question. The only fixed "laws and tendencies" Rawls mentions are those of the idealized existing social order itself (see p. n). Indeed, Rawls emphasizes the possibility of institutional reorganization. He says that, in contrast with laws and tendencies such as "the fact of reasonable pluralism," "the limits of the possible are not given by the actual, for we can to a greater or lesser extent change political and social institutions and much else" (p. 12). 26. As suggested above, this is not to say we cannot imagine any practice we like and ask what justice would require of it, if it did exist. The claim is simply that any principles we take to apply under hypothetical conditions will not necessarily tell us what fundamental requirements of justice actually require of us. We need to consider existing practices only insofar as this is our question.

83

John Rawls

296

Philosophy &> Public Affairs

activity in question, if undertaken, cannot be carried out in certain specified ways. You may not be required, morally, to promise to meet me at 2:00 P.M., but if you do promise, you can fail to show only if you provide fair warning that you will not be there (unless certain excusing conditions obtain). Similarly, as previously isolated societies, we might not be required by justice to trade goods or services (perhaps both societies are already very well off). But insofar as we do sustain a system of trade, for the sake of its many benefits, principles of fairness constrain its structure. In the domestic context, we may not be required by any principle of social justice to leave the state of nature and create the domestic rule of law (this being a matter of natural duty). But so long as cooperation is indeed established, there are limits on how its advantages and disadvantages can be distributed. A constraint interpretation is appealing because it enshrines the intuitive priority of considerations of social justice as a structural feature of what a principle of social justice is. Toleration of social injustice, except under special conditions, is then not simply failure to promote a worthy ideal, but a kind of wrongdoing we have overriding reason to avoid. By contrast, those who view justice as consisting, at least in part, in the realization of an ideal goal such as equal distribution (according to some measure) often concede that justice may have no special priority over other values.27 For when inequality is not great and everyone is very well off, it is implausible to insist that the promotion of ideal equality is of overriding significance, whatever the cost to other worthy values.28 If principles of justice are constraints on social practices, the Existence Condition can be explained as follows. Understood as a constraint, a principle of social justice would apply to the world only when a group 27. See, for example, Christopher Ake, "Justice as Equality," Philosophy & Public Affairs 5 (1975): 69-89, p. 73; Amartya Sen, Inequality Reexamined (Cambridge, Mass.: Harvard University Press, 1992), pp. 92-93; Larry Temkin, Inequality (NewYork: Oxford University Press, 1993), p. 292. 28. A similar difficulty arises for an appeal to "priority for the worse off," the idea that benefits matter more, morally, the worse off people are. (See Thomas Nagel, "Equality," in Mortal Questions [Cambridge: Cambridge University Press, 1979], and Derek Parfit, "Equality and Priority," Ratio 10 [1997]: 202-21.) When people are already very well off, it may be better if benefits go to less well-off persons, but this may seem to have little urgency as compared to other worthy goals. One can insist that it is nevertheless of overriding importance to shift benefits even to those quite well off, simply because failing to do so allows an injustice. This, however, can seem unmotivated.

John Rawls

297

Constructing Justice for Existing Practice: Rawls and the Status Quo

of agents is indeed organized into a particular kind of social practice. What the Existence Condition reflects is that such a principle can be respected in either of two ways: either by regulation of the structure in accord with the principle, or by disbandment or reorganization into a different social form. So long as the group is in fact organized into the relevant kind of practice, it cannot permissibly be organized in specified ways (for instance, in ways that assign unequal burdens among its participants). When the practice is not actually being undertaken, however, the relevant constraint does not apply. It does not then specify what justice actually requires. A constraint interpretation must allow a distinction between ideal and non-ideal theory. How could it do so? When conditions are not ideal, because they are not ideally favorable or because compliance is partial, how could principles of justice function as constraints instead of mere worthy ideals? Rawls's answer can have two parts. On the one hand, a previously established practice can face various threats to its existence or integrity. There may be impending unfavorable conditions (such as extreme scarcity, natural disaster, or outside attack), or compliance may threaten to unravel (because of mutual distrust or insufficient assurance). Under such extreme conditions, when the social preconditions for application of the relevant ideal principles of justice are in question, Rawls can suppose that such principles wait in suspension while, for quite different reasons of morality or prudence, steps are taken to preserve the practice's basic integrity.29 On the other hand, these are normally exceptional cases; it is perfectly commonplace for the basic form of a practice to be more or less secure. Conditions still may not be ideally favorable and compliance maybe partial, but requirements of both ideal and non-ideal theory apply. Ideal theory principles apply even if their fulfillment is not immediately possible. As constraints, the principles provide overriding reason to enact any adjustments required to make their fulfillment possible, at any cost to ideals other than justice. When all such possible step-wise routes involve further injustices, the force of the constraint is that injustice cannot be traded to advance any values other than justice. Just as for Rawls liberty can only be sacrificed for the 29. The threat to security is of course the dictator's favorite excuse (not to mention the wartime president who quickly curbs civil liberties). I take the fact that this permission can be abused to reflect its genuine presence.

85

86

John Rawls

298

Philosophy &> Public Affairs

sake of greater liberty, "an injustice is tolerable only when it is necessary to avoid an even greater injustice" (77, p. 4). Much as principles concerning the protection of liberty have lexical priority over principles concerning economic redistribution within domestic society, all principles of justice have lexical priority over the promotion of any further ideals.30 RAWLS AND CONSTRUCTIVE INTERPRETATION So there is reason to suppose that Rawls begins reasoning about justice from existing practices, that he accepts the Existence Condition. There is also reason to doubt this. The Existence Condition saddles any reasoning about what principles actually apply with a difficult interpretive problem: we have to identify and characterize an existing social practice, and for this we are required to interpret the attitudes and behavior of actual people. But is Rawls really engaged, even in part, in this interpretive enterprise? He can appear to be simply recommending certain practices as morally preferable, not seeking to describe practices that are already established. After all, his characterizations of domestic and global society are not only morally laden, but controversial: hardly descriptions of the most straightforward sort. However, as I will now argue, Rawls is indeed engaged in interpretation of practices that do already exist. The initial interpretive problem is this: among the many and various existing social practices that can plausibly be said to exist, which should we attend to? Rawls's answer is that we should attend first to the most "basic" of existing social structures.31 According to Rawls, our basic structures consist of the major institutions of domestic society and relations among societies. But why focus on these broadly political structures, as opposed, say, to gender or familial or religious ones? Rawls's answer is broadly moral: major domestic institutions are of particular moment because their effects are "so profound and present from the 30. This is what Rawls calls "The Priority of Justice over Efficiency and Welfare" (Ibid., p. 302). 31. Although Rawls need not treat every social practice that could properly be said to exist as though it is as legitimate as any other. There is no presumption, for instance, that even "the practice of slavery" in the antebellum South might have been just if only it were better organized; its moral status is in question until it is determined what our basic structures are and what principles apply to them.

John Rawls

299

Constructing Justice for Existing Practice: Rawls and the Status Quo

start," while relations between societies are basic because, along with domestic institutions, they have caused the "the great evils of human history."32 Rawls's appeal to moral considerations here can seem circular. Will not our sense of the profundity of certain effects and the greatness of certain evils rely on the very principles of justice Rawls seeks to justify? Not necessarily. Initially, Rawls can suppose the social structures to which he assigns methodological priority, as our most basic structures, are indeed widely agreed to exist as a matter of descriptive sociology. Thus he picks out the cases of his first concern, the societies of contemporary North America and Western Europe, under the description "modern constitutional democracies," focusing on their major political, legal, economic, and social institutions, organized into a single system. Similarly, he presents relations between societies, in the first instance, under the description "international law and practice," focusing initially on their most uncontroversial concern, the use of force. Rawls can then defend the methodological priority of such structures, characterized in the most general sociological terms, on the basis of two minimal moral claims. The first claim is that such structures are in presumptive need of justification, if for no other reason than that they tend to affect human interests to which most moral or political theories assign significance.33 The second claim is that the domestic and international political structures in question generally tend to affect these interests more significantly than social structures of other kinds.34 Together, these claims justify the priority of domestic and international political structures, but they do so leaving entirely open the question what principles of justice apply. 32. Rawls, TJ, p. 7; Rawls, The Law of Peoples, pp. 6-7. 33. Here "interests" can be broadly construed to include any of the range of claims or objections people might raise in practice against the structures in question. He can remain neutral between different theories on which of these claims give rise to a demand for justification. 34. Of course, gender, kinship, and religious structures also profoundly affect most human interests. I cannot consider the matter in detail, but the thought may be that reasonably just political structures will often substantially compensate for significant harms within these other structures, while reasonably just structures of these other kinds are less likely to substantially compensate for the significant harms of political structures. For Rawls's discussion of why political injustice has created the "great evils of human history," see The Law of Peoples, pp. 6-7.

87

John Rawls

00

00

300

Philosophy &> Public Affairs

These minimal moral claims narrow the interpretive problem posed by the Existence Condition, but they do not resolve it. In designing an original position, we need to know, at the very least, what goods the relevant practices are supposed to create and who the relevant participants are. The sociological characterizations of our basic structures just offered do not provide this information. For this reason, Rawls offers further, moralized characterizations: The aim of the basic structure of a modern constitutional democracy is to create primary social goods, and to do so as a cooperative scheme among persons for mutual or reciprocal advantage.35 The aim of international law and practice is to create goods of peace, national autonomy, and to uphold basic domestic justice, and to do so in a way that reflects mutual societal recognition.36 What relation do these morally laden characterizations of our basic structures bear to their initial sociological descriptions? Rawls's larger reflective equilibrium methodology allows him to work at both levels without clarifying their precise relation. Yet he does say, quite expressly, that the former, morally laden characterizations are "implicit" in the latter, more neutral descriptions, or at least in their associated "public cultures" (PL, p. 13-14). The "fundamental idea" of society as a cooperative venture for mutual or reciprocal advantage is, Rawls says, "implicit in the public political culture of a democratic society"; it is implicit, that is, in "the political institutions of a constitutional regime and the public traditions of their interpretation (including those of the judiciary), as well as historic texts and documents that are common knowledge." In characterizing our basic structures in moral terms, Rawls takes himself to be making explicit what is implicitly understood in our basic structures, as identified in uncontroversial, sociological terms.37 35. For Rawls's emphasis on reciprocal and not merely mutual advantage, see Political Liberalism (hereafter PL), p. 16. 36. Rawls suggests that mutual respect is an "essential part of the basic structure . . . of the Society of Peoples" on p. 122 of The Law of Peoples. 37. Rawls does say that the idea of a well-ordered society is an "idea of practical reason," whose content is gleaned by "reflect[ing] on how these ideas [ideas of practical reason] appear in our practical thought" (PL, p. 108). This may seem to suggest that the idea of a well-ordered society is not a creature of existing practices and their interpretations, but this is not necessarily the case. One possibility is that the idea of a well-ordered society appears in our practical thought because it is implicit in the public political culture of a democratic society. Its content will then be as much a matter of social interpretation as a priori reflection. Another possibility is that the abstract and unspecified idea of a wellordered society is itself a priori, while its content—that which grounds original position reasoning—is given by what is implicitly accepted in a public political culture.

John Rawls

301

Constructing Justice for Existing Practice: Rawls and the Status Quo

In other words, Rawls is engaged in what has come to be called constructive interpretation. As Ronald Dworkin describes it, constructive interpretation of a social practice proceeds in three stages.38 At Dworkiris first, "pre-interpretive" stage, we establish a common object of interpretation, by tentatively identifying a practice in uncontroversial terms. Thus Rawls initially identifies our basic structures in uncontroversial, sociological terms, as the major institutions of a "modern constitutional democracy" and as "international law and practice." At Dworkiris second, "interpretive" stage, we propose a characterization of the accepted "purpose or aim in the practice," a purpose or aim that bears some rational relation to the structure identified at the pre-interpretive stage. This is the role of Rawls's moralized descriptions of our basic structures, as cooperative schemes for the sake of the relevant, specified goods. Finally, at Dworkirfs third, "post-interpretive" or "reforming" stage, we specify requirements that must be fulfilled in order for the practice to achieve the goal or aim identified at the interpretive stage. For Rawls, it is at this stage that original position reasoning comes into play.39 In these terms, the crucial question whether Rawls's moral characterizations of our basic structures, offered at the interpretive stage, can be seen as interpretations of existing practices, identified at the preinterpretive stage. Given the Existence Condition, a proposed characterization of the nature and purposes of a practice cannot be a mere worthy moral ideal; it must also be in some sense generally accepted among participants. Rawls's sociological descriptions of our basic structures are sufficiently uncontroversial so as to be generally accepted; or at least they fulfill the central need of the pre-interpretive stage, to identify the object of interpretation, in relatively uncontroversial terms. Yet, as we have seen, it is only Rawls's further, moral characterizations of our basic structures, as cooperative schemes subject to requirements of reciprocas. Ronald Dworkin, Law's Empire (Cambridge, Mass.: Belknap Press, 1986), pp. 65-66. 39. In this way, Rawls denies Walzer's thesis claim that a suitable sociological interpretation of a "sphere" of human activity and the social meanings of its associated goods will determine their criterion of just distribution (Michael Walzer, Spheres of Justice [New York: Basic Books, 1983]). Because original position reasoning has a role, social meanings are not, by themselves, sufficient. The practice-dependence of original position reasoning also distinguishes Rawls from Dworkin, who writes, "a theory of justice is not required to provide a good fit with the political or social practices of any particular community, but only with the most abstract and elemental convictions of each interpreter" (Dworkin, Law's Empire, on p. 425, n. 20).

89

90

John Rawls

302

Philosophy &> Public Affairs

ity and mutual recognition, which ground original position reasoning. But these characterizations clearly will not be accepted by everyone. For instance, some may not see domestic institutions as having any distributive aim. In the international case, some may see the goal of international law as the promotion of distributive justice, and not merely, as Rawls would have it, the goals of keeping peace, respecting autonomy, and upholding only the most basic of human rights. Such disagreements call into question whether Rawls's favored interpretations are ultimately defensible. If nothing but bald moral assertion could justify Rawls's favored characterizations, we cannot charitably view him as fundamentally engaged social interpretation after all. I take it the basic problem here concerns the defensibility of Rawls's interpretive proposals, not whether it is coherent for Rawls to take himself to be describing our existing basic structures. There is a problem of coherence when a practice description is offered in controversial moral terms but the putative object of interpretation is not independently identified. If Rawls provided no independent identification of our basic structures, it would be unclear how he could possibly be describing an existing practice, instead of merely favoring a practice as morally preferable. I will return to the larger issue of defensibility, but it will be useful to first consider in greater detail why Rawls's use of moral vocabulary does not jeopardize the coherence of his claim to be explicating what is "implicit" in the public political cultures associated with the basic structures of domestic and global society Notice, first, that although any characterization of a practice must be in some sense "generally accepted," Rawls need not admit any controlling significance to such indicators of general opinion as the latest polls, the dominant political ideology, or even promulgated and largely accepted legislation. Firstly, all of these can be incompatible with other things generally accepted, which better reflect the character of a practice. Explicit constitutional provisions are the most obvious example, but the conflict can also be more informal: a conflict, for example, with principles, traditions, or interpretations that are generally "understood" (even if never codified), or deeper presuppositions of general understanding. What counts as either constitutional or understood is of course itself a matter of interpretation. Secondly, the standard of successful interpretation is not set by approximation to any privileged notion of "general acceptance." At the most abstract level, what matters is simply

John Rawls

303

Constructing Justice for Existing Practice: Rawls and the Status Quo

which interpretation is best according to various criteria: the best interpretation is, by comparison with the alternatives, more consistent, coherent, and simpler; and yet it provides a richer, more comprehensive, and more perspicuous account of the practice's various relevant and salient features, including such elements of its public culture as texts, traditions, accepted principles, implicit understandings, and guiding historical examples.40 Any given indicator of general opinion is merely one among many such factors that need to be accounted for. There is, then, no general reason why ascribing moral content to the basic structure of society should automatically reduce to bald moral assertion. For one thing, the factors any interpretation must explain may themselves have moral content. Constitutions may include requirements of equal treatment or due process, for instance; or progressive taxation and welfare institutions may have the expressly stated and officially promulgated aim of reducing undeserved hardship and inequality of opportunity. One can, moreover, move from such elements to a more general moral characterization of a larger basic structure without abandoning one's claim to be offering an interpretation. The proposed characterization must simply be an interpretive proposal, which is subject to canons of interpretation distinct from those that govern mere reasoning "on the merits" about what practices to have in the first place. Although one could also approve of our practices under their proposed moral descriptions, approval is not essential; in principle, an amoralist engaged in pure philosophical anthropology could reach the same moralized interpretive conclusions. Rawls makes both highly specific and highly general claims about the understood moral content of our existing basic structures. He claims quite specifically with H.L.A. Hart that certain rights are part of the understood nature of a viable legal system.41 Moreover, in noting that it is now understood within international law and practice that 40. I do not mean to defend this perhaps controversial conception of social interpretation, but merely to suggest that it is the most natural one for Rawls's purposes. I therefore ignore possible objections. 41. In PL, p. 109, n. 15, he mentions "H.L.A. Hart's minimum content of natural law" (H.L.A. Hart, The Concept of Law [Oxford: Oxford University Press, 1961], pp. 189-95). He also flirts with the strong requirements for a legal system defended by Philip Soper, A Theory of Law (Cambridge, Mass: Harvard University Press, 1984). His approach in The Law of Peoples is similar, but he qualifies his endorsement of Soper's approach on p. 66, n. 5.

91

John Rawls

92

304

Philosophy Public Affairs

governments no longer enjoy certain traditional powers of states (to wage war and to unrestricted internal autonomy), Rawls attributes the change to acceptance of quite general if largely inchoate standards of legitimacy.42 Rawls is more ambitious still in attributing to general understanding an overarching moral conception of the basic structure of modern constitutional democracies. Thus he claims that it has come to be understood in the modern era, at least within the public culture of constitutional democracies, that the goal of its basic institutions is not any perfectionistic ideal, or even an ideal goal of equal distribution, but the creation and distribution of various primary goods. It is now understood, Rawls claims, that such a basic structure is, by its very nature, a scheme that can be viewed by its members as a scheme of cooperation. Among other things, this is understood to mean that the scheme is subject to assessment under the concepts of fairness, reasonable acceptability, or reciprocity (and not, for example, in terms of independent concepts of entitlement or desert).43 If Rawls can coherently see himself as describing our existing basic structures in moral vocabulary, he can also coherently suppose that those who would disagree with his proposed characterizations do not fully understand what our basic structures are. Rawls says that the public political culture in a democratic society is "at least familiar and intelligible to the educated common sense of citizens generally" (PL, p. 14). By implication, it may not be familiar or intelligible to those who are uneducated or formally educated and out of touch with what citizens generally understand. This is not to say we should expect agreement among those of "educated common sense," who are familiar with and find intelligible the main elements of a public political culture. When deep presuppositions are at issue, and successful interpretation requires thoughtful and skilled judgment, divergence in interpretation is to be expected. None of this is to say that Rawls's moral characterizations are correct, or even that he has sufficient resources to account for the kinds of interpretive disagreements we find in actual constitutional democracies. So, 42. Rawls, The Law of Peoples, p. 27. 43- Rawls, PL, p. 16. Rawls also mentions as essential for cooperation that each person can see the scheme as advancing his or her own good, in terms of primary social goods, and that the scheme coordinates behavior by rules that are generally accepted, and known as accepted, as proper guides for conduct.

John Rawls

305

Constructing Justice for Existing Practice: Rawls and the Status Quo

for all we have said, a persuasive defense may still require bare moral assertion. Moreover, while Rawls does expressly present himself as explicating what is implicit in our public political cultures, and while he does occasionally appeal to defining historical examples and public texts or traditions, he does not explicitly carry out the dirty work of interpretive argument.44 In part, this can be explained by his usual assumption that the best defense is a good offense; he proceeds by first "working out" a positive characterization of our basic structures, leaving defense against alternative characterizations as a secondary task. Still, Rawls's conspicuous lack of interest in this secondary task may seem to suggest he could not have taken his stated interpretive ambition very seriously (perhaps realizing that, short of moral assertion, he lacks sufficient resources for proper interpretive defense). I suggest an alternative explanation: Rawls's resources for interpretive argument do have considerable depth, and his awareness of this fact explains why he neglects detailed interpretive defense in favor of other theoretical tasks (such as accounting for reasonable disagreement about what justice requires of society, when it is generally understood as a cooperative scheme). Insofar as Rawls is proposing interpretive characterizations of our basic structures, this precludes any direct appeal to moral considerations. At the same time, canons of interpretation are not always wholly independent from those governing evaluation. Even in the purest cases of third-person interpretation, when one is interpreting the practices and sustaining moral or normative attitudes of an alien group, interpretive charity will require provisional reliance on one's own assessments of merit.45 The relation between interpretation and assessment can be even closer in cases of first-person interpretation. When one is interpreting one's own present beliefs, for example, there may be no difference between what a judgment of what one believes and what one is to believe, on the merits.46 First-personal self-interpretation is hardly a 44. In Political Liberalism Rawls is of course very concerned with reasonable disagreement in conceptions of justice. His solution to the problem involves appeal to the generally accepted "fundamental idea" of society as a scheme of cooperation, which he spends comparatively little time defending. 45- As is required for interpretive charity generally. See, for instance, Donald Davidson, Inquiries into Truth and Interpretation (Oxford: Oxford University Press, 1984). 46. See Richard Moran, Authority and Estrangement (Princeton, N.J.: Princeton University Press, 2001) on "transparency," pp. 60-65.

93

94

John Rawls

306

Philosophy &> Public Affairs

sufficient basis for interpreting the attitudes of many different people. But it is not always irrelevant, either. Rawls is not an interpreter of an alien group, but an interpreter of structures and a public political culture with which he is intimately associated. As a presumptively competent member of a practice, he can take his own interpretation of the practice's nature as (defeasible) evidence of common opinion. Although any characterization informed by self-interpretation will be informed by evaluative judgment, it also will be subject to standards for interpretation of relevant features of the practice and of the attitudes of other practitioners. At the very least, for example, a characterization must rationalize the structure of any independently identified organization. Charity will weigh, at least presumptively, against attribution of any moral goal that the group systematically and grossly fails to realize (however certain one feels, based on self-interpretation, that the goal in question is the practice's essential point). Any characterization must of course square with and explain the various texts, traditions, implicit understandings, and defining historical examples plausibly said to constitute a public political culture. Rawls's central interpretive burden is to identify the relevant texts, traditions, and defining historical examples and to advance his interpretation as the best way of accounting for them, by comparison to alternative interpretations, according to all the relevant interpretive criteria. I will return to how this argument might go momentarily. Supposing for the moment that the argument could be persuasive, we can return to the question whether Rawls has resources to account for those who would disagree. What, if anything, might Rawls say about those who disagree if he is to adequately defend his own favored interpretation? To the extent dissenters must be disqualified, Rawls cautions that, when explaining disagreement arising from "lack of reasonableness, or rationality, or conscientiousness . . . we must be careful that the evidence for these failings is not simply the fact of disagreement itself" (PL, p. 121). One possible source of evidence that does not simply assume one's own interpretation is that certain texts or traditions must be accounted for by any plausible interpretation of a practice. Given that this is so, the fact that someone's interpretation wholly ignores them shows him to be "oversimplifying" as an interpreter. Similarly, one could argue, without simply asserting one's own interpretation, that a certain interpretation does not give some salient feature sufficient weight, and

John Rawls

307

Constructing Justice for Existing Practice: Rawls and the Status Quo

on this basis charge that someone who accepts it is being "unbalanced" or "one-sided." Beyond this, one can argue that dissenting interpreters have fallen into internal inconsistency, factual error, or reliance on unsupportable empirical assumptions. In the context of many controversial political issues (issues about the status of racial minorities or homosexuality, for instance), this can go a long way In addition, dissenters might be shown to lack general moral understanding. When constitutional provisions or informal traditions make explicit reference to moral concepts such as equality or liberty or both, full interpretive understanding of those and other understood moral commitments in need of interpretation will require moral understanding of the associated concepts. Those unaccustomed to the delicate task of reasoning with all of the moral concepts that might be in question, including reasoning about any of their systemic relations, may be prone to interpretive misunderstanding. The likelihood of misinterpretation, as a result of failed moral understanding, is only greater for those motivated by perceived threats to vested interests, political party loyalty, or mere unwillingness to entertain unfamiliar moral views. In short, Rawls has ample resources to defend his own proposed interpretations in light of existing interpretive disagreement. The danger in such a defense is that Rawls's own, first-personal sense of what a "full understanding" of accepted moral terms entails will outrun what is in fact generally understood. At the same time, there is little reason to think Rawls necessarily crosses the line into bald moral assertion. Indeed, he not only shows interpretive restraint, but is open to the charge of undue conservativism. For instance, he ascribes to international law and practice the goal of promoting a limited set of basic rights, and not justice more generally. He does so despite the fact that clear-cut elements of the public political culture, such as the Universal Declaration of Human Rights, might be cited as evidence that more ambitious moral goals have already been generally accepted. Whether Rawls overreaches ultimately depends on the merit of his moralized characterizations as proposed interpretations. It is not my aim to defend Rawls here, but, in order to indicate how his interpretations could be more plausible than others, it is worth briefly mentioning how the argument against one alternative might proceed. Consider a libertarian interpretation that sees the understood aim of domestic social institutions as limited to the protection of rights of non-interference,

95

96

John Rawls

308

Philosophy &> Public Affairs

including, say, rights against most coercive taxation. Rawls can argue that, by comparison, his own characterization of existing major institutions better accounts for the preponderance of key features that any interpretation would have to explain.47 Unlike Rawls, the libertarian interpretation will have difficulty explaining why, in most existing constitutional democracies, funding of public goods (such as parks and roads) by coercive taxation is widely understood, across the political spectrum, as an appropriate function of government. Even in the case of a quasi-public good such as education, which is understood to have some redistributive functions, disagreement tends to arise about how much coercively extracted tax support to provide, not whether it may be provided at all. Similarly, the appropriateness of some welfare institutions, which quite expressly serve the most basic claims of need, is relatively uncontroversial, even if this is, again, funded by taxation without consent. The libertarian can argue that these assumptions are not central features of existing institutions, and so that the proposed interpretation need not take full account of them. This, however, runs afoul of charity: if people did generally take the aim of institutions to be the protection of rights to non-interference, why do so many routinely assume just what this precludes, the appropriateness of taxation for public goods and minimal redistribution? Alternatively, the libertarian can argue that other key features are better explained in terms of rights of non-interference. But what are these features? If the interpretive advantage is supposed to be a plausible account of accepted values of life and liberty, for instance, Rawls's interpretation can also take these values into account: they are generally understood to be central determinants of any fair terms of cooperation. And if on the libertarian's strongest points Rawls does equally well, the points that lean in his favor (such as those mentioned above) will render his interpretation, on balance, the better one. STATUS Quo BIAS? I conclude, then, that Rawls grounds original position reasoning in constructive interpretation of our basic structures. I return now to the 47. I take this to be the standard governing interpretation generally, whether of society or of art or of literature.

John Rawls

309

Constructing Justice for Existing Practice: Rawls and the Status Quo

question whether Rawls's practice-sensitive method is biased toward the status quo. I will not address the well-known objection that Rawls's larger reflective equilibrium methodology is unduly conservative. I focus on concerns of bias that arise within that methodology. In order to achieve reflective equilibrium, it is crucial that we adjust our use of original position reasoning so that its output accords with our "considered judgments."48 The problem is that, if we limit ourselves to original position reasoning, we seem systematically prevented from doing this, being exclusively concerned with existing practices and those who happen to participate in them. Original position reasoning does lead us to systematically ignore nonparticipants in a given practice.49 This is despite the fact that our considered judgments presumably include the possibility that nonparticipants can be unjustly dominated, neglected, or—the favorite example of Rawls's critics—excluded from participation.50 When assessing a particular kind of practice as just or unjust, why dwell on its participants, instead of considering everyone the practice affects? The most plausible kind of answer is that participants have a special kind of social justice claim, which warrants special or at least separate attention. This has no immediate implication for the status of nonparticipants. They may have: (i) (ii) (iii) (iv)

no moral claim; some moral claim, but not a claim of social justice; claims of justice, but no special claim to equal (or otherwise fair) shares in comparison to participants; or claims of justice, and the same claims as participants to equal shares.

48. This is not a significant constraint if our considered judgments are strongly influenced by existing practice. I assume there are reasonably uncontroversial considered judgments for which this is not the case. 49- Rawls is quite explicitly concerned in the domestic context only with "citizens . . . normally active and fully cooperating members of society over a complete life" with "physical and psychological capacities within a certain normal range" (Rawls, "Social Unity and Primary Goods," Collected Papers, pp. 368-69). For parallel passages, see Rawls, PL, p. 20, and Rawls, "The Law of Peoples," p. 531. In the international context, ideal theory similarly considers only "well-ordered" societies. See Rawls, The Law of Peoples, pp. 64-68. 50. For example, Buchanan, "Justice as Reciprocity Versus Subject-Centered Justice," p. 237.

97

98

John Rawls

310

Philosophy &> Public Affairs

Position (iv) is precluded if participants have a special kind of justice claim. But this hardly implies (i), the shocking position that the severely handicapped, for example, would have no moral claim against a society to aid or even non-interference.51 Neither does position (ii) follow: the position that the domestic severely disabled, refugees, asylum seekers, or stateless people pose moral and "humanitarian" concerns, but not issues of social justice. For there is also the possibility of position (iii): although the principles derived from any particular original position only concern participants in some practice, nonparticipants also have claims of social justice against mistreatment, which are justified in some other, perhaps contractualist way. On T. M. Scanlon's contractualism,52 for instance, domination, negligence, or exclusion may violate principles for the regulation of a practice that no appropriately motivated and informed person could reasonably reject.53 Would Rawls deny this last, accommodating view? To answer this question, let us assume, as Rawls seems to, that the scope of social justice, in the sense that applies paradigmatically to the organization of a social practice, depends on how we characterize reasoning to principles for this concept's application. In that case, this concept will be exhausted by internal, participation-based claims only if the original position reasoning is all there is to reasoning about its application. Rawls does sometimes appear to hold that any principle of social justice will be derivable from some application or other of original position reasoning, in which case reasoning about social justice is entirely participationbased.54 Yet this is not Rawls's considered view. For instance, it is quite 51. As on some interpretations of David Gauthier, Morals by Agreement (Oxford: Clarendon Press, 1986), p. 268. 52. T. M. Scanlon, What We Owe to Each Other (Cambridge, Mass.: Harvard University Press, 1998). This possibility is in accord with Sen's claim (in Sen, "Open and Closed Impartiality") that original position reasoning must be supplemented. I take the natural alternative to be a contractualist framework such as Scanlon's, not the Smithian, "impartial spectator" approach that Sen recommends. 53. For example, the relevant principles might include: principles of Collective Noninterference, which proscribe various forms of force, coercion, or other control; principles of Collective Due Care, which require precautions for the exercise of dangerous forms of power; and principles of Collective Exclusion, which prohibit the turning away of refugees, boat people, those seeking political asylum, and so on. 54. See, e.g., Rawls's cryptic comments on "unity by appropriate sequence" (PL, pp. 261-62).

John Rawls

311

Constructing Justice for Existing Practice: Rawls and the Status Quo

clear from the role of human rights in The Law of Peoples that Rawls takes original position reasoning to be only one part of reasoning about what justice requires. Although he claims that both liberal and non-liberal societies would choose, from their respective original positions, the principle that "peoples are to honor human rights," this turns out to be a mere formality. Only "decent" societies are represented in the international original position, and they by definition respect human rights.55 Rawls does argue for his favored list of human rights, but only on the grounds that it has the appropriate functional role, as a standard of international legitimacy.56 The question, then, is why original position reasoning should ever be the appropriate way to reason about justice. In order to provide an answer, we need to specify a special participation-based claim. In his early, stage-setting work, Rawls suggests a claim against exploitation. "A practice will strike the parties as fair," Rawls writes, "if none feels that, by participating in it, they or any of the others are taken advantage of, or forced to give in to claims which they do not regard as legitimate."57 In other words, once people are no longer "forced to give in to claims which they do not regard as legitimate," it is a further question whether a practice treats its participants in an arbitrary or unfair manner. It will usually be a difficult and practice-sensitive question what counts as full "participation" or "compliance" with the terms of a practice, in the sense that secures one's status as a contributor. (Is mere non-interference enough? Can one unwittingly participate?) But if we assume this can often be worked out in particular cases, the general thought is that when many people act in specified, coordinated ways in order to realize goods that would not otherwise exist, those who have contributed, by having had a hand in those goods' creation, can lay claim to 55. Rawls, The Law of Peoples. The principles are enumerated on p. 37. For the role of the decency assumption, see pp. 6sff. 56. Specifically, a society's arrangements must conform to a "common good conception of justice." For Rawls, a conception that does not include guarantees of security and life would not be reasonably characterized as a common good conception, although a conception that does not honor rights to freedom of association and democratic participation might be so characterized, at least under certain circumstances. Human rights therefore include rights to security and life, but not rights to freedom of association and democratic participation. See Ibid., pp. 65-80. 57. Rawls, "Justice as Fairness," p. 59.

99

100

John Rawls

312

Philosophy &> Public Affairs

their enjoyment.58 If participants then receive little benefit, or lesser benefits than others who have done no greater part, simply, say, because the group or its influential members divert benefits away from them, they can reasonably claim, on the basis of having done their part, that they are being "taken advantage of" or exploited for the gain of others. By contrast, those who do not actually contribute, by having some role in the practice, are in no position to press this particular kind of claim. This kind of claim is in one way sensitive to the status quo, because its rationale is grounded in facts about who happens to actually contribute to a practice. But this sensitivity is limited, since there is no reason nonparticipants cannot complain of exploitation or injustice of some other kind. For example, those who are not given an opportunity to participate might complain of exclusion. But if they are indeed excluded, they cannot also make a participation-based claim against exploitation, for the simple reason that one cannot complain one's contribution is being exploited for the gain of others if one is prevented from making any contribution at all. So Rawls's method is not open to the charge of systematic bias against nonparticipants, so long as Rawls allows that original position reasoning is not all there is to reasoning about social justice. This will not eliminate the charge of bias toward existing practices, however. I have argued that Rawls's method begins from existing practices. Rawls's account of human rights suggests that justice is not participation-based, but it also suggests that justice is practice-sensitive. A list of human rights, for Rawls, is part of an answer to the question, When can members of a society reasonably insist on being free from outside military, political, or economic interference? It is not an answer to the question, Why should "international law and practice" as it is currently understood exist in the first place? Does Rawls's practice-sensitivity objectionably bias reasoning about justice toward the status quo? That is, does it prevent us from capturing our considered judgments in reflective equilibrium? There are different ways to approach this question. One way simply takes for granted that 58. To be sure, Rawls famously denies that participants can lay greater contributionbased claims to social goods as compared to other participants. The talented are not being "taken advantage of" by the less talented who may seem to offer less but equally do their part.

JohnRawls

313

Constructing Justice for Existing Practice: Rawls and the Status Quo

a Rawlsian account of justice is unsatisfactory, all things considered. If this is so, a plausible explanation why is that the Existence Condition imposes an unacceptable bias toward the status quo. But this charge of status quo bias has little obvious force independently of argument that Rawlsian justice is inadequate, all things considered. Since I am not taking a position on that issue, I will simply set this kind of charge to one side. It is quite another matter, however, whether the Existence Condition systematically prevents us from accounting for reasonably uncontroversial considered judgments. In my remaining remarks, I provide reasons why this is not clearly the case. The charge of status quo bias, I suggest, chiefly depends for its force on arguments that Rawlsian justice is inadequate, all things considered. The Existence Condition does place significant limits on the critical power of social justice assessment. If a principle applies to the world only insofar as an appropriate kind of social practice exists, that principle cannot itself be used to criticize either the existence or non-existence of the kind of practice that conditions its application. Suppose, for example, that for any particular family, there are principles that require a fair division of labor. In that case, whether family practices are 19505 suburban North American, traditional Korean, or Italian Mafia, they involve an interpretation of a basic form of kinship organization, a form of organization that conditions the application of the fair division of labor principle. Although any instance of that basic form might be unjustly organized (labor might be divided unfairly), the bare fact that people are organized into the relevant basic social form, into a kinship structure, cannot itself be unjust. The bare existence of the basic form cannot at once condition the application of the equal labor principle and violate that selfsame principle. For to say that an application condition for the fair labor principle is fulfilled, that is, to say that the relevant kinship structure exists, is only to say that the practice may or may not fulfill the fair labor principle, not that there is occasion for evaluating the existence of the basic kinship structure itself as just or unjust. The Existence Condition entails that the existence or non-existence of a basic kind or form of a practice (which conditions the application of some principle) can at most be criticized as unjust in relation to requirements that apply in the first instance to some further existing practice. The injustice must be relational or extrinsic. To illustrate this possibility, suppose that to even approximate Rawlsian Fair Equality of

101

102

John Rawls

314

Philosophy &> Public Affairs

Opportunity state institutions would have to eliminate the traditional family, because anything like it breeds uncorrectable inequality in early childhood education, undermining equal chances to compete for positions of power and reward. Suppose also that the family can be eliminated without violating liberties protected by principles of justice by a powerful system of rewards within a new and pervasive social ethos. In this case, justice would require that we either alter kinship structures or that we retain our families, dismantle the other institutions, and find a path to just tribalism. For Rawls, which practices must be abandoned then depends on whether the principles that apply to the state have priority over those that apply to the family, or vice versa (where priority is specified by "priority rules") (PL, p. 262). Because Rawls holds that the principles that apply to the basic structure have "regulative primacy with respect to the principles and standards appropriate for other cases" it seems that families as we know them would have to go (PL, pp. 257-58). Do these limitations prevent us from accounting for our reasonably uncontroversial considered judgments of injustice? In fact, the Existence Condition is quite permissive. It is compatible, for instance, with the possibility that, as practiced, each and every existing practice is socially unjust, because it fails principles applicable to the kind of practice it is. Moreover, the very existence of any nonbasic structure ("the practice of slavery," for instance) can itself be unjust if it is incompatible with a just basic structure. The same cannot be said for basic structures themselves; their very existence cannot be socially just or unjust. Yet, for several reasons, this is not clearly objectionable independently of whether or not Rawlsian justice is adequate in the final analysis. First, as we have seen, what basic structures do exist is open to morally informed, constructive interpretation. Second, when this is combined with further, substantive contractualist argument, Rawls's method may give rise to quite significant demands (including, for example, the Difference Principle). Third, it is still possible for justice to require revolution. To illustrate this possibility, consider that the Existence Condition does not preclude the following kind of argument for (albeit gradual) revolution in the global order. The Existence Condition provides no reason to deny that we ought in some cases to create new practices as a matter of justice, on the grounds that existing practices require organizational supplementation in order to be just. This is arguably the case for the present global order. For it is

JohnRawls

315

Constructing Justice for Existing Practice: Rawls and the Status Quo

plausible to claim, as a matter of interpretation, that the modern system of territorial states is, in part, an ongoing system of control over natural resources and their socially created benefits. It may also be plausible to argue that a scheme of property can permissibly deprive individuals of access to natural resources and goods they might use as means of subsistence only insofar as they are offered the benefits of such access in some other way (in the form, say, of socially guaranteed opportunities to earn or to buy food, water, and shelter). For Rawls, the content and grounds of any such argument will be affected by what is required to prevent the deterioration of assets, the central aim of a scheme of property.59 But this basic requirement itself is one with which many political theories will in one form or another agree (although perhaps for different, practice-insensitive reasons). Since by most accounts the existing state system fails this basic demand, we can reasonably conclude that the creation of effective international practices of aid and provision is not morally optional but a requirement of justice on the global system of resource control. Now, it may well be that our attempts at assistance will never be sufficient so long as the existing state system is in place. There may be no way to check the opposing pressures of abusive wayward regimes, the over-empowerment of multinational corporations, heightened vulnerability to unemployment and destitution wrought by unregulated markets, and the ensuing absolute deprivation and untoward psychological, economic, and political outcomes of material inequality. This is not necessarily a reason why practices of aid and provision are not required, after all. The Existence Condition allows for the possibility of piecemeal revolution when two practices cannot both be justly organized. In that case, meeting the demands of justice on the underlying system of resource control may require such extensive piecemeal modification of the present order that the result could hardly be recognized as the modern state system. Justice could well require global revolution. (To allow for such a possibility may well be part of Rawls's motivation for seeing the world not as a society of self-interested "states" but of distinct, potentially reasonable "peoples.")60

59. Rawls, The Law of Peoples y p. 8. 60. Ibid., pp. 23-30.

103

104

John Rawls

316

Philosophy &> Public Affairs

In conclusion, then, the Existence Condition affords considerable latitude in how our existing practices are identified and criticized. This shifts questions of status quo bias to questions of whether Rawlsian justice is plausible or adequate in the final analysis. It is fair to wonder why reasoning about fundamental justice should be sensitive to existing practices at all, an issue on which Rawls is largely silent.611 have already suggested, however, that Rawls's belief in the priority of justice over other values assumes principles of justice function as constraints and not ideal goals, and that the conditional applicability of constraints in turn explains the Existence Condition. This hardly immunizes Rawls from the charge of status quo bias (it does not show that Rawlsian justice is adequate or plausible in the final analysis). But it does show that focus on the practices actually being undertaken can reflect seriousness about fundamental justice. As moral constraints on our basic structures, principles of justice are by nature of normative moment; what they require is not simply another worthy ideal, but that certain things be done, even at great cost to competing goals. In the light of this, Rawls can go on the offensive: it is not Rawls, but those who treat justice as merely one value among others, who fail to register the seriousness of what justice demands.

61. In the way of explanation, Rawls offers only the methodological remark: "We must always start from where we now are, assuming that we have taken all reasonable precautions to review the grounds of our political conception and to guard against bias and error" (Ibid., p. 121). This approach is natural given Rawls's aim to "reconcile . . . us to our political and social condition," and not simply to describe what may be an unlikely utopia (Ibid., p. n).

[4] Justice, Desert, and Ideal Theory Jon Mandle Justice requires that a person get what he or she deserves. This claim is widely accepted, although the exact nature of the relationship between justice and desert becomes controversial as these concepts are specified in greater detail. It would seem, however, that one of the most well-known theories of justice, justice as fairness, rejects this claim. Something of a consensus has emerged in the literature concerning Rawls's view about the relationship between justice and desert, and most contemporary discussions of desert make at least a passing remark about his so-called "rejection of desert." I wish to dissent from this understanding and suggest an alternative account. I begin by presenting what I take to be the consensus view concerning Rawls's rejection of desert. I explain why I believe this common understanding of his position is mistaken by focusing on the core passages on which this interpretation is based. I then present what I take to be a more adequate and straightforward interpretation of Rawls's claims concerning desert. This will lead us to a discussion of ideal theory and why Rawls believes such an investigation has priority over non-ideal theory. I defend this priority, although, with Rawls, I certainly acknowledge that addressing the question of ideal theory should not be mistaken for a complete theory of justice. It is only a first step, but a necessary one. Rawls's rejection of desert as a basis for the just distribution of social resources is well known and frequently cited in the secondary literature. At least his conclusion seems to be straightforward. Two quotations will suffice to illustrate this. At one point Rawls says, 'There is a tendency for common sense to suppose that income and wealth, and the good things in life generally, should be distributed according to moral desert. Justice is happiness according to virtue Now justice as fairness rejects this conception. " Two pages later, he puts the point this way: 'The

106

John Rawls

400

Jon Mandle

essential point is that the concept of moral worth does not provide a first principle of distributive justice. While the first passage might appear to deny any logical relation between justice and moral desert, the second refines this view. Rawls is not claiming that there is no connection between justice and desert. Rather, he is claiming that we cannot define justice in terms of desert. So, he is rejecting the platitude with which I began— "justice requires that a person get what he or she deserves"—only if we take this to assert a certain logical priority of the concept of desert to the concept of justice. That is, he is rejecting the view that we can determine what a person deserves prior to and independent of considerations of justice, and then simply define justice as giving that person what she deserves. Note that in both passages I have quoted, Rawls explicitly refers to moral desert. In this sense, what we deserve is based on and is proportional to our moral virtue. I want to begin by emphasizing this point: Rawls reserves the term "desert" for evaluations based on moral worth. There are numerous senses in which we use the term "desert" in ordinary conversation. Clearly, the sense of desert that I just mentioned, in which the grounds of a desert claim are found in the moral virtue of a person, is only one of these. But Rawls restricts his use of the term "desert" to these cases. What about other entitlement claims that individuals can make, and that ought to be respected, on grounds other than moral virtue? We often speak, for example, of someone deserving a job or position of authority, or of a person deserving compensation for having their property destroyed or a contract violated. At one point, Rawls calls such notions "deserving in the ordinary sense, but generally he gives a different label to such claims. When such claims are legitimate, and ought to be respected, Rawls says they express "legitimate expectations." Rawls thinks of social institutions as specifying certain properties or actions on the basis of which individuals make entitlement claims. When the institutions are just, these claims express "legitimate expectations." That is, just institutions will specify (among other things) which properties qualify a person for a job or position, the scope of property rights, and what actions constitute negligence or violation of a contract. As Rawls expresses the point: "Having done various things

JohnRawls

Justice, Desert, and Ideal Theory

107

401

encouraged by the existing arrangements, [people] now have certain rights, and just distributive shares honor these claims. A just scheme, then, answers to what men are entitled to; it satisfies their legitimate expectations as founded upon social institutions. " These entitlements are the joint product of the properties and actions of the individual and (assuming the institutions are just) the institutional rules within which the individual is a participant. It is important to be clear that a claim to certain legitimate expectations is not to be confused with a claim concerning the "intrinsic worth" or the moral virtue of the individual involved. To illustrate this, Rawls has us consider a just economic order that relies, at least in part, on market mechanisms. One acquires a legitimate expectation, a claim on social resources that society and other individuals are morally required to respect, on account of the legal contracts one has signed and (ultimately, in a perfectly competitive market) on the basis of one's marginal productivity. But, as Rawls points out, "the extent of one's contribution (estimated by one's marginal productivity) depends upon supply and demand. Surely a person's moral worth does not vary according to how many offer similar skills, or happen to want what he can produce. No one supposes that when someone's abilities are less in demand or have deteriorated . . . his moral deservingness undergoes a similar shift. " So, if a person invents a new efficient widget-maker or works from nine to five to comply with a contract, we might intuitively say that such a person "deserves" a certain compensation, and it would be unjust for society or other individuals to deprive her of that compensation. Although Rawls would agree that such a person is entitled to compensation, the basis of this assessment is not the fact that she is morally virtuous (she may or may not be) but rather the compliance with a valid contract within a just institutional context. Therefore, he would say that this claim expresses a legitimate expectation rather than desert. One further preliminary point is necessary. It is crucial to remember that Rawls is interested in arriving at principles of justice with which to evaluate the basic structure of society. The basic structure consists in "a society's main political, social, and economic institutions, and how they fit together into one unified system of social cooperation." When he asserts, therefore, that

108

John Rawls

402

Jon Mandle

"the concept of moral worth does not provide a first principle of distributive justice," implicit is the addendum, "for the purpose of evaluating the basic structure. " In particular, he is not denying that we can evaluate particular rules, policies, and programs within a just basic structure by the degree to which they respond to appropriate pre-institutional criteria. For example, we may say intuitively that in a spelling-bee, the best speller deserves to win. We criticize the rules of the contest if they don't generally identify the best speller. The rules of the spelling-bee do not themselves define who the best speller is. Rather, we have a pre-institutional criterion to which the rules themselves ought to conform. (Of course, spelling ability is not pre-institutional in the sense of being independent of the basic structure of society, but it is independent of the rules of the spelling-bee.) If the rules are set up properly, Rawls would agree that the best speller has a legitimate expectation to win the prize. Spelling ability is not a moral virtue, however, so Rawls himself would not apply the term "desert." The point here is that Rawls denies none of this common sense, and this is clear when we attend to the emphasis on the basic structure. What Rawls's "rejection of desert" amounts to, therefore, is a rejection of the claim that basic social justice requires the distribution of happiness according to a prior conception of moral virtue. Why does Rawls reject this precept, which he admits is one of common sense? Versions of what I will call the "consensus understanding" of Rawls's argument (or important components of it) have been defended by, among others, Robert Nozick, Michael Sandel, Alan Zaitchik, George Sher, David Gauthier, John Kekes, G.A. Cohen, and William Galston—a widely diverse group of philosophers, to say the least. This reconstruction of Rawls's view begins by emphasizing two passages that supposedly represent Rawls's metaphysical beliefs about the nature of the self. The first is Rawls's claim that "the self is prior to the ends which are affirmed by it." The second is this: "We see then that the difference principle represents, in effect, an agreement to regard the distribution of natural talents as a common asset and to share in the benefits of this distribution whatever it turns out to be." The thought is this. The first passage (allegedly) shows that Rawls is committed to a metaphysical view in which the self is I O

JohnRawls

Justice, Desert, and Ideal Theory

109

403

"radically disembodied." According to Sandel, Rawls believes that "all endowments are contingent and in principle detachable from the self." It is this metaphysical picture of the self, Sandel believes, that underlies Rawls's defense of the difference principle. Sandel draws the connection between the metaphysical picture of the self and the "common asset" interpretation of the difference principle as follows: "[T]he difference principle acknowledges the arbitrariness of fortune by asserting that I am not really the owner but merely the guardian or repository of the talents and capacities that happen to reside in me, and as such have no special moral claim on the fruits of their exercise." Because these talents and capacities are contingent and not really constitutive properties of the self, we cannot be said to deserve them. William Galston explains the next crucial premise in the argument this way: ccRawls's argument, familiar since A Theory of Justice and reinforced in the Dewey Lectures, rests on the proposition that desert claims cannot be based on something undeserved." Since we cannot be said to deserve our talents and abilities (they are only contingently related to us), we cannot be said to deserve anything based on our possession of them. There is one final step. Sandel asks who does have a claim to our talents and capacities if not the individual? He answers that liberals assume that "'society' in some indeterminate sense... has a prior claim on whatever assets the individual does not." Since the individual does not have any special claim to ownership of his talents and capacities, they belong to society, which may treat our natural talents (and, indeed, whatever other properties it fancies) as common (or collective) property to be distributed in whatever way it chooses. (I should note in passing that speaking of "ownership" of talents is SandeFs language. Rawls never speaks of ownership in this context.) As Sandel observes, if this interpretation of Rawls is correct, there are even more profound consequences than those that Rawls (allegedly) draws. Not only can justice not be founded on a prior idea of desert, the idea of desert itself becomes virtually incoherent. As Sandel puts it: "no one can strictly speaking be said to deserve anything because no one can be said to have anything, at least not in the undistanced, constitutive sense of possession

110

JohnRawls

404

Jon Mandle

necessary to a desert base." Alan Zaitchik makes a similar observation when he claims that this interpretation of Rawls results in "a completely general argument which alleges that no desert theory could be true for the simple reason that no one ever deserves things.' On this view, even one's moral virtue is only contingently attached to one's self, and so not deserved, and so unsuited to underwrite a claim to desert. Indeed, Sandel observes this when he repeats his conclusion and then takes the argument even one step farther: 'The principles of justice do not mention ... desert because, strictly speaking, no one can be said to deserve anything. Similarly, the reason people's entitlements are not proportional to nor dependent upon their intrinsic worth is that, on Rawls's view, people have no intrinsic worth. This is an ingenious interpretation and if it were true, it would certainly explain why Rawls would think that the principles of justice do not mention desert. But the reason for this would not have anything to do specifically with the concept of justice. Rather, on this interpretation, Rawls would reject the connection between justice and desert because he denies the coherence of the concept of desert at all (at least for people). Many critics, rightly in my view, take this to be a kind of reductio ad absurdum. They oppose Rawls's "rejection of desert" (the rejection of the logical priority of desert to justice) because they think that his argument requires a rejection of the concept of desert altogether. Before presenting my own interpretation of Rawls's account, I will argue that this consensus view is mistaken. It will be worth our while to consider in detail the two passages that I said formed the core of the consensus interpretation, because they have been the focus of considerable attention (and I think misunderstanding) even among those not directly concerned with questions of desert. And it must certainly be admitted, once again, that the interpretation championed by Sandel and others has enjoyed remarkably widespread acceptance. But it is sustained only by reading these core passages out of context. The first thing to notice is that the "common asset" passage (which is supposed to be the conclusion) comes before the "self passage (which is supposed to be the premise). Indeed, it precedes the latter passage by over 450 pages. If Rawls were arguing from a metaphysical conception of the self

John Rawls

Justice, Desert, and Ideal Theory

111

405

to a conclusion about who is the "owner" of our 'talents and capacities," this would be a highly convoluted, or maybe disguised, presentation. Clearly this does not refute the consensus interpretation, but perhaps it will give us pause before agreeing with Sher that "[w]e know that Rawls wants to move from the premise that people do not deserve their character or abilities to the conclusion that people also do not deserve the advantages that those 'natural assets' make possible."2 I deny precisely this assumption, even though, as Galston has it, this is indeed the "familiar" argument. To defend my claim that this is not the argument that Rawls actually makes, we must attend more closely to the context of those two core passages. The final chapter of A Theory of Justice is called 'The Good of Justice." In that chapter, Rawls reflects on whether the account of justice that he has presented would be stable. Specifically, he is concerned with what he calls "congruence" between justice (or, more generally, the right) and the good. "It remains to be shown," he says at the beginning of the chapter, "that given the circumstances of a well-ordered society, a person's rational plan of life supports and affirms his sense of justice." As part of this chapter, he contrasts justice as fairness with "hedonistic utilitarianism" in an effort "to show how just institutions provide for the unity of the self and enable human beings to express their nature as free and equal moral persons. " As Rawls uses the term, "teleological" theories are those in which "the good is defined independently from the right, and then the right is defined as that which maximizes the good. Teleological theories differ from one another in their account of the good, that is, what is to be maximized by right conduct. But consider a teleological theory that recognizes different goods associated with different individuals. Utilitarianism is such a theory, as it recognizes that different individuals can be happy or not in various degrees. A teleological theory that recognizes different goods must explain how these goods are to be compared, combined, and maximized, in order to determine right conduct. Hedonism appears to answer this demand by presenting a dominant end (namely, pleasure) that allows the comparison, combination, and maximization of the goods of different

112

John Rawls

406

Jon Mandle

individuals. But although it may provide a dominant end that allows for the aggregation of these different goods, Rawls argues, "[i]t seems obvious that hedonism fails to define a reasonable dominant end.' On reflection, he suggests, we do not take the sensation of pleasure to be the sole end to which all of our actions (or rational or valuable actions) are ultimately directed. Our goals and values are much more diverse than that. Rawls takes the attraction of hedonism as a way of aggregating the good, together with its failure to define a reasonable end, to be revealing: Hedonism is, one might say, the symptomatic drift of ideological theories insofar as they try to formulate a clear and applicable method of moral reasoning. The weakness of hedonism reflects the impossibility of defining an appropriate definite end to be maximized. And this suggests that the structure of ideological doctrines is radically misconceived: from the start they relate the right and the good in the wrong way.

Rawls is claiming that we can recognize the failure of teleological theories by considering the following dilemma. We can identify an end that is measurable and specific enough for aggregation and maximization (such as pleasure) only by making it implausible as a dominant end. We could, alternatively, choose an inclusive end, by giving a formal definition of each person's good (we could call it "eudaimonia" if we like) but this will come at the cost of leaving it indeterminate what these goods consist in, and how they are to be compared and combined. Rawls continues the above passage as follows: We should not attempt to give form to our life by first looking to the good independently defined. It is not our aims that primarily reveal our nature but rather the principles that we would acknowledge to govern the background conditions under which these aims are to be formed and the manner in which they are to be pursued. For the self is prior to the ends that are affirmed by it; even a dominant end must be chosen from among numerous possibilities. There is no way to get beyond deliberative rationality. We should therefore reverse the relation between the right and the good proposed by teleological doctrines and view the right as prior.

When Rawls speaks of "our nature" in this context, it would seem that he means our rational and moral nature, what he sometimes calls (for example, on the subsequent page) our "moral personality." The claim is that it is not through our identification with some substantive dominant end that we identify ourselves as

JohnRawls

Justice, Desert, and Ideal Theory

113

407

free and equal. Rather, it is through our affirmation of fair institutions within which to pursue our goods that we express ourselves as rational and moral persons. Whether we accept this argument against teleological theories is not here the point. What is important for us is simply the fact that in this context the claim about the priority of the self to its ends has nothing whatever to do with the "common asset" interpretation of the difference principle. Indeed, it has nothing to do with the difference principle and Rawls's specific conception of justice at all. Rawls is concerned with the structure of moral theory and practical reasoning, how we view our deliberative capacities in relation to our ends, not whether we can be considered the "owners" of our talents and capacities. Furthermore, notice that Rawls doesn't mention or even allude to our "talents and capacities" in this passage. The only priority he is discussing concerns the self and its ends (its conception of the good). Rawls does not make any claim at all about the priority of the self with respect to our natural talents or abilities. Yet, this is what Sandel needs to sustain his interpretation. Sandel tacitly slides from Rawls's assertion of priority of the self over its ends to an assertion of the priority of the self over its natural talents and abilities. He sustains this by claiming that Rawls postulates a self prior to and independent of all its properties, including not only its ends, but also its natural abilities, and even its moral character. But Rawls never makes such claims. On the other hand, there are passages where Rawls asserts that we do not deserve our natural talents and abilities, and perhaps Sandel is relying on these. Don't these passages imply that the self is independent not only of our ends, but also of our natural talents and abilities, and perhaps even our moral character? Consider the following example of such an assertion: It seems to be one of the fixed points of our considered judgments that no one deserves his place in the distribution of native endowments, any more than one deserves one's initial starting place in society. The assertion that a man deserves the superior character that enables him to make the effort to cultivate his abilities is equally problematic; for his character depends in large part upon fortunate family and social circumstances for which he can claim no credit. The notion of desert seems not to apply to these cases.

114

JohnRawls

408

Jon Mandle

It is crucial to recognize, as we have seen, that by "desert" Rawls means "deserving on account of moral virtue." He takes it to be a truism that the initial social position into which we are born cannot be a reflection of our moral virtue, precisely because it is our initial starting place. In this sense, it cannot be something that we deserve (it can't be based on one's moral virtue), although it may or may not be something to which we are entitled (or "deserve in the ordinary sense"). Similarly, the native endowments and talents with which we find ourselves are not a reflection of our moral virtue. Once we recognize this, he argues, we will come to see, as well, that the "superior character that enables him to make the effort to cultivate his abilities" cannot be a reflection of our moral virtue, either. Note that there is no claim here that one is not responsible for one's moral character; he is onlv considering the capacity to develop one's talents and abilities. As Thomas Pogge puts it: Rawls is not saying that there is no moral reason to leave endowments where we find them—of course there is—but that differences in endowments do not, within the conception of social justice, signify differences in moral worth. This claim does not presuppose that native endowments are a matter of luck or fortune. Even if people had to have worked very hard to be bright or athletic, these traits would still be arbitrary from a moral point of view. Those who don't care to become bright or athletic through hard work are not therefore to be considered morally inferior as far as our shared public conception of justice is concerned.

So, I conclude that contrary to Sandel, Rawls nowhere commits himself to the priority of the self to its abilities or talents, nor does he anywhere claim that a person is not entitled to her talents and abilities. He does claim that a person doesn't deserve them, but this only means that the entitlement is not based on one's moral virtue. This interpretation might seem to be refuted by the second core passage on which the consensus interpretation is based, so let us turn to examine it. This passage occurs in the second chapter of A Theory of Justice, entitled 'The Principles of Justice." In this chapter Rawls's concern is to present and explain the principles that he believes would be chosen in the original position. As he tells us in the introduction to this chapter, "[n]ot until the next chapter do I take up the interpretation of the initial situation and begin the argument to show that the principles considered here

JohnRawls

Justice, Desert, and Ideal Theory

115

409

would indeed be acknowledged [T]he aim [of this chapter] is to explain the meaning and application of the principles.' And he emphasizes this point, later saying that this discussion is only an attempt to "prepare the way for the favored interpretation of the two principles so that these criteria, especially the second one, will not strike the reader as too eccentric or bizarre." Having presented an interpretation of the principles in the first several sections of this chapter, Rawls proceeds to further elaborate their content by contrasting them with what he calls "the principle of redress." The principle of redress holds that "in order to treat all persons equally, to provide genuine equality of opportunity, society must give more attention to those with fewer native assets and to those born into the less favorable social positions. The idea is to redress the bias of contingencies in the direction of equality.' Rawls stresses that the difference principle is not the same as the principle of redress. Yet, there are certain similarities. He contrasts the difference principle with the principle of redress: Thus although the difference principle is not the same as that of redress, it does achieve some of the intent of the latter principle. It transforms the aims of the basic structure so that the total scheme of institutions no longer emphasizes social efficiency and technocratic values. We see then that the difference principle represents, in effect, an agreement to regard the distribution of natural talents as a common asset and to share in the benefits of this distribution whatever it turns out to be. Those who have been favored by nature, whoever they are, may gain from their good fortune only on terms that improve the situation of those who have lost out.

There are several points about this passage that need to be emphasized. First, what the difference principle shares with the principle of redress is a rejection of "technocratic values" and subservience to the single goal of "social efficiency." The difference principle does not aim at the rectification of natural inequalities. Second, and most importantly for us, the "common asset interpretation" of the difference principle is just that: a possible interpretation of the principle, an alternative formulation that helps explain its content. It is not, and is not presented as, an argument for the difference principle. As Rawls stresses repeatedly, the argument for the difference principle occurs in the next chapter when he considers what would be chosen from the original position. Here he is only claiming that if society chooses to regulate its basic structure by

116

JohnRawls

410

Jon Mandle

his principles of justice, this choice would "represent, in effect, an agreement to regard the distribution of natural talents as a common asset..." (my emphasis). There is no thought here of arguing from the social ownership of the natural talents to the difference principle, as Sandel suggests. Furthermore, there is not even any thought that the difference principle requires that we regard the natural talents themselves as being distributed (or redistributed) by some central authority or society as a whole. The question that Rawls is addressing is not the distribution of those talents, but rather, under what conditions can those with natural talents gain further on the basis of their good fortune. That is, Rawls is not suggesting even metaphorically that the talents themselves are to be regarded as common assets. The difference principle concerns only the cooperative, institutional scheme within which individuals claim shares of social resources on the basis of their skills, luck, and effort. Jeremy Waldron introduces the right distinction when he observes that "owning a talent is one thing, and benefiting from it in a scheme of social cooperation is quite another. Rawls is only interested in the latter. Those who have the good fortune to be endowed with desirable natural abilities are surely entitled to those abilities, but they are not necessarily entitled to any particular institutional scheme within which to exercise those talents. Even more to the point, they certainly don't deserve any particular institutional scheme. As Rawls says, "[t]he natural distribution is neither just nor unjust; nor is it unjust that persons are born into society at some particular position. These are simply natural facts. What is just and unjust is the way that institutions deal with these facts.' But there is no question of the skills themselves being redistributed. Rawls is only concerned with the design of the institutional mechanisms through which society rewards those who have these skills, as well as those who make greater effort or who are simply lucky. As Waldron, once again, writes: In any conceivable society, which might be the subject of speculations about justice, talents will be nurtured and exercised in an institutional framework. That framework will make a difference to what happens, socially and economically, when somebody exercises his talents— There is no sense to the idea that talents can simply be exercised by those who own them apart from any social framework whatsoever. And there is no sense to the idea that there is a natural

John Rawls

Justice, Desert, and Ideal Theory

117

411

phenomenon called 'reaping the benefits of one's talents' that is understood apart from the social arrangements and institutions that define one's relationships to other people.

It is this institutional framework within which talents are nurtured, developed, encouraged, exercised and rewarded that is the subject of Rawls's investigation. And the principles that he provides to answer this question include the difference principle, the idea that a just institutional scheme will reward the exercise of one's talents and skills in ways designed to benefit the least advantaged. So, I conclude that Rawls does not make the argument that the consensus understanding attributes to him. Rawls never says that the self is prior to its capacities and talents. He never says that the grounds of desert claims must themselves be deserved. He never says that we are not responsible for our moral character. He never claims that society has some kind of prior claim over an individual's capacities and talents. He does claim on many occasions that our entitlement to our talents and capacities is not based on our moral virtue, but he takes that to be a triviality. The interesting question to which he addresses himself concerns the institutional context within which individuals exercise their talents and acquire legitimate claims to social resources. If, as I have argued, the consensus interpretation is incorrect, why does Rawls think that justice cannot be based on a logically prior concept of moral desert? I believe there are two reasons, one of which has only emerged clearly in Political Liberalism. Evaluation of the justice of the basic structure should be based only on principles that "all citizens as free and equal may reasonably be expected to endorse in the light of principles and ideals acceptable to their common human reason. This is Rawls's "liberal principle of legitimacy." But as Rawls now emphasizes, "the diversity of reasonable comprehensive religious, philosophical, and moral doctrines found in modern democratic societies is not a mere historical condition that may soon pass away; it is a permanent feature of the public culture of democracy. Therefore, in evaluating their constitution and other matters of basic social justice, citizens should respect the limits of public reason. In particular, they should only present arguments concerning basic social justice in terms that all reasonable persons

118

John Rawls

412

Jon Mandle

could accept. They should not present arguments that rely on their full comprehensive doctrine. The problem with designing the basic structure to reward moral desert is that an assessment of the moral virtue of an individual requires a (fully or partially) comprehensive moral doctrine. We cannot assume that there is (or will be) a single standard of moral virtue that all reasonable people will accept. Any attempt to design the basic structure to reward moral virtue would rely on a particular comprehensive moral doctrine. But this would violate the liberal principle of legitimacy because some citizens could reasonably reject that standard. Basing justice on moral desert violates the liberal principle of legitimacy and exceeds the bounds of public reason. It should be emphasized that the problem, here, is the reliance on a comprehensive moral doctrine to assess individual moral virtue. This would not be a problem if we were to rely exclusively unpolitical virtues, that is, only those virtues that all reasonable comprehensive doctrines would endorse. Justice as fairness itself relies on such a notion of the political virtues. But nobody, including Rawls, believes the political virtues exhaust the moral virtues, and therefore any attempt to base the design of the basic structure on full moral desert is unacceptable. Still, why couldn't we define justice as rewarding people according to the degree to which they manifest the political virtues? This would not violate the limits of public reason, since all reasonable comprehensive doctrines recognize these virtues, even while they disagree about others. To understand the problem with this suggestion, we must examine a feature of Rawls's argument that many commentators, both defenders and critics of justice as fairness, mention and then drop without considering its implications. Rawls believes justice as fairness must begin as an exercise in "ideal theory." Every political theory makes idealizations of one sort or another, but justice as fairness makes a rather startling assumption: it starts by considering the choice of principles for a well-ordered society. Rawls believes that we must begin our investigation of justice with ideal theory in this sense. Three idealizing conditions suffice for a society to be well ordered. First, citizens are assumed to have an effective sense of justice. Second, the principles of justice that they endorse are

John Rawls

Justice, Desert, and Ideal Theory

119

413

widely shared (unlike their comprehensive moral doctrines). Third, the basic structure of the society satisfies that widely shared conception. Rawls is clear that actual societies are rarely, if ever, well-ordered in this sense. He thinks that our society, for example, fails to greater or lesser degrees on all three criteria. Nonetheless, for reasons we will examine below, Rawls thinks we should consider this idealized case first. Notice that the conditions necessary for a society to be well ordered do not by themselves specify particular principles of justice. They only express the assumption that whatever citizens take the most appropriate principles of justice to be, they will be instantiated in both the behavior of the citizens generally and in the basic structure. The obvious question is the following: doesn't this idealizing move of assuming a well-ordered society vitiate the need for the virtue of justice in the first place? According to one understanding of justice, the answer is clearly yes. If we take justice to require rewarding people according to their moral virtue (basing justice on moral desert), it would seem that ideal theory would abstract away from the important basis of differentiation between individuals. This idealization would seem to bypass precisely the interesting issue that gives rise to the importance of justice in the first place, namely, that people are not equally virtuous and therefore are not equally deserving. Many critics of liberalism, such as John Kekes, press Rawls on exactly this point. Kekes accuses Rawls of failing to "face evil," that is, of a failure to confront the fact that people have an unfortunate tendency to cause unjustified harm to one another. The strength of this tendency varies among individuals and circumstances, as does the countertendency to produce good (or, I suppose, justified good). One's desert is dependent on the balance of these two tendencies, or what Kekes calls a person's "moral merit. 'The core belief of justice," according to Kekes, "[is] that people ought to get what they deserve. The essential idea," he continues, "is not merely that desert creates a claim, but also that the claim should be based on some specific morally good or evil state, characteristic, relationship, or action. ' John Hospers and W.D. Ross take a similar view of justice, and as we have seen Rawls concede, it seems to be a

120

John Rawls

414

Jon Mandle

precept of common sense. On this view, justice is a response to immorality (or, more precisely, differential levels of moral virtue). Rawls's understanding of justice, or, better, the circumstances that constitute the occasion for the virtue of justice, is different from that of Kekes. Following Hume, Rawls calls the conditions in which justice finds its occasion Public Affairs, who made many extremely valuable suggestions for improvement. I thank Sarah Goldberg for research assistance. i. Elizabeth Anderson, "What is the Point of Equality?" Ethics 109 (1999): 287-33?.

310

JohnRawls 6

Philosophy & Public Affairs

religion, sex, and ethnicity is unjust, and that distributive inequalities resulting from such discrimination are unjust as well. It also holds that people of equal talent from different social classes should have equal access to the social positions for which their talents qualify them, and that it is unjust if inequalities result from a society's failure to provide this kind of equal opportunity. Thus, the prevailing morality agrees with luck egalitarianism in rejecting certain kinds of inequalities deriving from unchosen features of people's circumstances. However, in rejecting all inequalities of advantage resulting from differing circumstances, luck egalitarianism goes far beyond the prevailing political morality, which, against a background of nondiscrimination and equal opportunity, is prepared to tolerate significant distributive inequalities deriving from differences of talent and ability. By contrast, luck egalitarianism denies that a person's natural talent, creativity, intelligence, innovative skill, or entrepreneurial ability can be the basis for legitimate inequalities. Consider next the acceptability of inequalities deriving from people's choices. If some people make more money than others because they choose to work longer hours, then the prevailing morality certainly agrees with luck egalitarianism that that is not in itself objectionable. Unlike luck egalitarianism, however, the prevailing political morality does not go so far as to say that any extra income deriving from people's choices should, in principle, be exempt from redistributive taxation. In determining one's tax burden, the prevailing morality makes no attempt to identify, let alone to shield from taxation, the portion of one's income that is traceable specifically to one's choices as opposed to one's natural abilities. The upshot is that although there are substantial areas of overlap between luck egalitarianism and the prevailing political morality, luck egalitarianism is in one way much more willing than the prevailing morality to engage in redistributive taxation, but also in one way much less willing. In this sense at least, luck egalitarianism is both more and less egalitarian than the prevailing political morality/ 2. I do not mean to imply that these two factors cancel each out so that there is no net difference in the redistributive implications of the two positions. Luck egalitarians differ among themselves about the how much redistribution is justified» in part because they disagree about the extent to which actual inequalities are the result of differences in people's circumstances as opposed to differences in their choices. The "prevailing political morality" also encompasses a range of positions on the extent of legitimate redistribution, for although, as I have said, adherents of the prevailing morality have no principled objec-

JohnRawls 7

What is Egalitarianism?

Unlike Rawls's theory, which became the focus of intense critical scrutiny as soon as A Theory of Justice appeared in print, luck egalitarianism has been relatively slow to attract critical attention, despite the impressive level of influence it has attained and the lively debates that have taken place among proponents of its different variants. This state of affairs has begun to change, however, and in this article I wish to make a modest contribution to the project of critical examination that Anderson and others have initiated.31 am indebted to Anderson's discussion in many ways, some of which I will acknowledge more specifically as I proceed. But my primary emphases will be different from hers, and I hope that my main points are sufficiently independent as to be of interest in their own right. Luck egalitarianism is often presented as an extension and generalization of some Rawlsian arguments whose substantive implications Rawls himself failed fully to appreciate. Thus, luck egalitarianism is often said by its advocates to be truer than Rawls's own conception of justice to some of his fundamental insights. In part because of its putative Rawlsian pedigree, perhaps, luck egalitarianisiris supporters have done less than one might expect to provide an independent defense of the position at the fundamental level. In this article, however, I argue that luck egalitarianism can draw little support from Rawls. In addition, I present reasons for doubting whether it is either a plausible position in its own right or a compelling interpretation of egalitarianism. My article is organized as follows. In Section I, I will summarize a familiar version of the recent history of political philosophy—a version that locates the origins of luck egalitarianism in Rawls's thought but tion to taxing income that derives from people's choices, they disagree among themselves about how much taxation is justified and why. In general, however, most luck egalitarians think that the economic regimes of contemporary liberal societies should be significantly more redistributive than those regimes have usually been in practice, and more redistributive than the prevailing political morality would allow. As I note in Section II, it is therefore striking that the rise of luck egalitarianism occurred during a period in which liberal societies were in fact becoming much less redistributive. 3. See, in addition to the paper of Anderson's cited above, Jonathan Wolff, "Fairness, Respect, and the Egalitarian Ethos/' Philosophy &> Public Affairs 27 (1998): 97-122; Seana Valentine Shiffrin, "Paternalism, Unconscionability Doctrine, and Accommodation," Philosophy &> Public Affairs 29 (2000): 205-50; Timothy Hinton, "Must Egalitarians Choose Between Fairness and Respect?" Philosophy &> Public Affairs 30 (2001): 72-87. Debra Satz has also pursued related themes in her as yet unpublished work.

311

312

John Rawls 8

Philosophy &> Public Affairs

insists that he himself fails to develop the view in a consistent or thoroughgoing way. Although I will make clear my disagreement with the interpretation of Rawls as an incipient luck egalitarian, I will postpone a complete account of the bases for my disagreement until Section III. In the intervening section (Section II), I will present some reasons for doubting the plausibility of the luck-egalitarian position. In the course of my discussion, I will argue that any form of distributive egalitarianism, if it is to be persuasive, must be rooted in a more general conception of equality as a moral value or normative ideal. With that argument as background, I will return in Section III to Rawls. I will attempt to show that his work, properly understood, provides no support for luck egalitarianism; his position is better understood as helping to illustrate how a plausible form of distributive egalitarianism can be anchored in a conception of equality as a social and political ideal. Finally, in Section IV, I will consider some attempts to demonstrate that luck-egalitarian principles can also be anchored in a broader conception of equality.

I As I have noted, it is often said that luck egalitarianism has its origins in Rawls's work, but that Rawls himself fails to develop the view in a consistent or thoroughgoing way. According to this interpretation, Dworkin subsequently builds on the basic Rawlsian insight to provide the first systematic formulation of a luck-egalitarian position, a formulation that is more faithful to Rawls's original insight than is Rawls's own conception of justice. Alternative versions of luck egalitarianism are then developed in response to Dworkin and Rawls. Will Kymlicka offers a clear description of this putative progression of ideas. Kymlicka says that "[o]ne of Rawls's central intuitions ... concerns the distinction between choices and circumstances,"4 and that Rawls is "motivated"5 by the desire to produce a theory that is "ambition sensitive" but "endowment insensitive"—a theory that makes people's fortunes depend on their choices but not on their natural endowments or other unchosen circumstances. However, Kymlicka adds, "Rawls seems not to have realized 4. Will Kymlicka, Contemporary Political Philosophy (Oxford: Clarendon Press, 1990), p. 70. 5. Ibid., p. 76.

John Rawls

9

What is Egalitarianism?

the full implications of his own argument/'6 for "while Rawls appeals to th[e] choices-circumstances distinction, his difference principle violates it/'7 By contrast, he continues, "Dworkin accepts the 'ambition sensitive' and 'endowment insensitive' goal that motivated Rawls's difference principle. But he thinks a different distributive scheme can do a better job living up to that ideal/'8 There are two primary aspects of Rawls's work that are usually cited as evidence for his (imperfectly developed) luck egalitarianism. The first is his informal moral argument for the superiority of his two principles of justice to the laissez-faire "system of natural liberty." That rival scheme allows for the operation of a free market economy constrained only by a background of equal liberty and formal equality of opportunity. Rawls observes that the system of natural liberty permits the distribution of resources to be "strongly influenced"9 by people's natural attributes and the social positions into which they are born. But these factors, he argues, are "arbitrary from a moral point of view/'10 Because they are so arbitrary, we should not allow the distribution of resources "to be improperly influenced by" them.11 Instead, we should strive "to mitigate the influence of social contingencies and natural fortune on distributive shares."12 Rawls goes on to say that this line of thought does not, "strictly speaking,"13 count as an argument for his conception of justice, since he is undertaking to develop a kind of social contract theory and, as he understands it, this means that the "official" arguments for his conception must be arguments that establish that it would be rational for the parties in the original position to choose it. Nevertheless, the "unofficial" moral argument that I have sketched is often cited as the initiating formulation of the luck-egalitarian idea that inequalities deriving from unchosen circumstances are unjust. The second aspect of Rawls's view that has been understood as luck egalitarian in its tendency is the appeal to responsibility that he makes in the course of defending primary social goods as the appropriate basis 6. Ibid., p. 72. 7. Ibid., p. 76. 8. Ibid. 9. John Rawls, A Theory of Justice (Cambridge: Harvard University Press, 1971)» p. 72. 10. Ibid. 11. Ibid. 12. Ibid., p. 73. 13. Ibid., p. 75.

313

314

John Rawls

10

Philosophy & Public Affairs

for interpersonal comparisons. Some critics argue that primary goods are not a reasonable basis for such comparisons because a theory that relies on them takes no account of variations in the costs of satisfying different people's preferences. Two people with identical bundles of primary goods may have very different levels of utility or welfare because one of the people has simple tastes that can be inexpensively satisfied while the other has preferences that are very costly to satisfy Since primary goods take no account of these variations, a theory that uses them as the basis for interpersonal comparisons will treat the two people as enjoying the same level of well-being and will make no provision for the greater difficulty that the second person has in satisfying his preferences. In response, Rawls argues that we have "a capacity to assume responsibility for our ends/'14 and that "citizens can regulate and revise their ends and preferences in the light of their expectations of primary goods/'15 In a just society, therefore, "those with less expensive tastes have presumably adjusted their likes and dislikes over the course of their lives to the income and wealth they could reasonably expect; and it is regarded as unfair that they now should have less in order to spare others from the consequences of their lack of foresight or self-discipline."16 This response has been cited as an early expression of the luck-egalitarian position that people must bear the cost of their own choices and that economic inequalities are therefore legitimate provided they arise from differences in the choices that people make.17 Despite what he sees as the luck-egalitarian strands in Rawls's thought, Kymlicka argues that Rawls's own theory of justice is incompatible with luck egalitarianism in at least two significant respects. First, the difference principle does not make any special provision for those who have special medical needs, even when those needs result from unchosen natural conditions, since the principle takes only primary social goods, such as money, into account when assessing individuals' well-being. Thus, the principle treats two individuals with identical bundles of primary social goods as being equally well-off, even if one of the individuals is in normal health while the other suffers from a congenital medical 14. Rawls, "Social Unity and Primary Goods/' in John Rawls: Collected Papers, ed. Samuel Freeman (Cambridge: Harvard University Press, 1999), pp. 359-97> at p. 369. 15. Ibid., p. 370. 16. Ibid., pp. 369-70. 17. See, for example, Kymlicka, p. 75.

John Rawls

11

What is Egalitarianism ?

condition that is very costly to treat. Second, the difference principle seeks to maximize the position of the worst-off social class, as measured by an index of primary goods, even if the reason that some members of that class have small shares of those goods is that they freely choose to work shorter hours because of a preference for greater leisure over greater income. In this way, the difference principle interferes with the processes by which differences in people's choices may generate economic inequalities, and it requires some people to bear the costs of the free choices made by others. The fact that these features of the difference principle are incompatible with a luck-egalitarian position has led some interpreters to conclude that Rawls's view is inconsistent and that he fails to appreciate fully the implications of his own luck-egalitarian arguments and premises. I think that this interpretation is misleading, and that it is a mistake to construe Rawls as appealing to a general distinction between circumstances and choices, which his difference principle then fails to respect. I do not see any evidence that Rawls relies on such a distinction or that he ever expresses the aim of producing a theory that is "endowment insensitive" but "ambition sensitive." In my view, the best explanation of the fact that Rawls's theory of justice does not respect the distinction between choices and circumstances is that Rawls is not attempting to respect it. He simply does not regard the distinction as having the kind of fundamental importance that it has for luck egalitarians. To be sure, Rawls does make the arguments against the system of natural liberty and in favor of a reliance on primary goods as the basis for interpersonal comparisons, which have been seen as congenial to a luck-egalitarian position. But Rawls deploys those arguments in the service of a view that is substantively quite different from luck egalitarianism. Nor does Rawls's view have the same theoretical motivations as the ones articulated by luck egalitarians. Richard Arneson, for example, when introducing his preferred version of a luck-egalitarian position,18 18. Richard J. Arneson, "Equality and Equal Opportunity for Welfare/' Philosophical Studies 5$ (1989) : 77-93. More recently, Arneson has abandoned the version of luck egalitarianism defended in this article, in favor of a view that he calls "responsibility-catering prioritarianism." See, for example, Richard J. Arneson, "Equality of Opportunity for Welfare Defended and Recanted/' Journal of Political Philosophy 7 (1999): 488-97; and "Luck Egalitarianism and Prioritarianism," Ethics no (2000): 339-49. In the second of these two articles Arneson describes responsibility-catering prioritarianism both as an "(outlier) member of the luck egalitarian family" (p. 340) and as "a close cousin of luck egalitarianism" (p. 341).

315

316

John Rawls

12

Philosophy & Public Affairs

says that it "is addressed to egalitarians"19 and that it seeks to answer the following question: "Insofar as we care for equality as a distributive ideal, what is it exactly that we prize?"20 His reply is that "the idea of equal opportunity for welfare is the best interpretation of the ideal of distributive equality."21 In posing this question and developing his reply, Arneson takes himself to be entering into a discussion initiated by Rawls and Dworkin. Their writings, he says, "have debated the merits of equality of welfare and equality of resources taken as interpretations of the egalitarian ideal."221 will postpone until Section IV the question of how best to characterize Dworkin's enterprise, but I believe it is a mistake to present Rawls as aiming to persuade an audience of fellow distributive egalitarians that equality of resources is a better interpretation of the egalitarian ideal than equality of welfare. As I shall try to show, Rawls's theory has different aims and ambitions, and relies on different moral premises. Thus, it is hardly surprising, and certainly no sign of inconsistency, that his theory should end up having a different content as well. Of course, someone might insist that whatever Rawls himself may think, a luck-egalitarian position does in fact follow from his arguments against natural liberty and in favor of primary goods, once the full implications of those arguments are properly understood. To the extent that he fails to recognize this consequence, his theory is so much the worse for that. I do not find this convincing. In Section III, I will outline briefly the kind of alternative view that Rawls's arguments seem to me intended to support. Although there are certainly areas of overlap between Rawls's position and luck egalitarianism, and although there are certainly strands of argument in his work that luck egalitarians are bound to regard as congenial, my own view is that, on the whole, Rawls's arguments support his own position rather than luck egalitarianism. In certain fundamental respects, moreover, Rawls's position seems to me more attractive than luck egalitarianism. For the moment, however, I want to set Rawls's view to one side and explain some of the reasons for my reservations about the luckegalitarian position. 19. 20. 21. 22.

Arneson, "Equality and Equal Opportunity for Welfare," p. 77Ibid. Ibid. Ibid.

JohnRawls

13

What is Egalitarianism ?

II

As I have already suggested, many discussions of luck egalitarianism are addressed primarily to other egalitarians. They simply assume that an ideal of distributive equality has an important role to play in our thinking about distributive justice, and they ask how that ideal is best to be understood. Arneson makes the point explicitly. His arguments, he writes, are "addressed to egalitarians, not their opponents."23 As with Arneson's own essay, most of the luck-egalitarian literature is devoted to considering two closely related questions. The first question is what exactly egalitarians wish to equalize.24 This question is often formulated as a question about the correct "metric" of equality. The candidates for equalization that have been debated include welfare, resources, opportunity for welfare, and access to advantage, among others. The second question is which forms of disadvantage should receive compensation in the name of equality. Here there has been discussion of physical handicaps, medical needs, limited talents, unfavorable social positions, unsuccessful gambles, expensive tastes, expensive religious commitments, undesirable aspects of temperament, and so on. Although the debates about how best to respond to these questions have been intense, it is by no means clear how much the different versions of luck egalitarianism differ in their practical, political implications. Arneson, speaking of the choice between some version of "equality of resources" and his own preferred principle of "equal opportunity for welfare," concedes that "the practical implications of these conflicting principles may be hard to discern, and may not diverge much in practise."25 But if the practical import of the intra-egalitarian debates is not always clear, there is no doubt that there are great differences, of considerable political significance, between luck egalitarians as a group and nonegalitarians. Yet luck egalitarians have spent relatively little time defending the moral core of the luck-egalitarian position against skep23. Ibid. 24. Indeed, in some luck-egalitarian writings, this is not only the first question but also the first sentence. John Roemer, for example, begins his essay "Equality of Talent" by asking: "If one is an egalitarian, what should one want to equalize?" And Roemer's essay "A Pragmatic Theory of Responsibility for the Egalitarian Planner" begins: "What should an egalitarian seek to equalize?" Both essays are included in Roemer's Egalitarian Perspectives (New York: Cambridge University Press, 1994), pp. 119-47anci 179-96, respectively. 25. Ibid., p. 87.

317

318

JohnRawls

14

Philosophy 6» Public Affairs

tics and proponents of other views. An analogy would be a debate among defenders of rival versions of utilitarianism in which the practical stakes were unclear and little attempt was made to persuade others that maximizing some version of utility was the right approach to take in the first place. Of course, there is certainly a place for internal discussions, among the proponents of any broad philosophical position, about how that position is best to be interpreted. Utilitarians do in fact engage in such discussions, as do proponents of every other philosophical position that is appealing enough to attract a wide and diverse group of supporters. Important as these internal discussions are, however, they do not obviate the need to motivate, explain, and defend the core of a philosophical position to those who are not yet persuaded. In the specific case of luck egalitarianism, moreover, this task is especially important in light of the stark contrast between the theory's political implications and the political context in which it was developed. The rise of luck egalitarianism as an important position within political philosophy, which took place during roughly the last two decades of the twentieth century, coincided with a marked increase in inequalities of income and wealth in the United States and other liberal societies.26 Those decades were characterized by a strong trend, intensified by the collapse of the Soviet Union, toward increased privatization and reliance on the market, and a steady erosion in political support for distributive egalitarianism of any kind. Thus there was a sharp disparity between the luck egalitarianism that was becoming increasingly influential in philosophical discussions of distributive justice, and the actual distributive practices of the societies in which those discussions took place—and this despite the fact that one of the aims of luck-egalitarian theorists is to demonstrate that egalitarians need not, in general, be hostile to markets, and must indeed rely upon them in important ways. Of course, the existence of a disparity between luck-egalitarian theory and actual political practice does not by itself tell against the luck-egalitarian position. lustice is a normative notion, and so no proposed principle of justice can be falsified simply by 26. For some of the relevant U.S. data, see U.S. Census Bureau, Income Inequality (1967-1998), at www.census.gov/hhes/income/incineq/p6o204/. For data on other liberal democracies, see Timothy M. Smeeding, "Changing Income Inequality in OECD Countries: Updated Results from the Luxembourg Income Study," Luxembourg Income Study Working Paper No. 252 (March 2000).

John Rawls

15

What is Egalitarianism?

pointing out that actual practice does not, in fact, to conform to it. Yet the starkness of the contrast between luck-egalitarian theory and contemporary political practice makes it especially important to address fundamental questions about the moral defensibility of the core luckegalitarian idea. What needs to be considered is whether luck egalitarians have latched on to an idea of equality that has real appeal as a social and political value. The importance of addressing this issue can be brought out in another way. A moment ago I suggested a parallel between debates among luck egalitarians and intra-utilitarian debates. The relation between luck egalitarianism and utilitarianism is an interesting and, I think, an instructive one, which merits closer scrutiny. For those who interpret Rawls as an early adherent of luck egalitarianism, his defense of primary goods as the basis for interpersonal comparisons has been taken to place him squarely on the side of those who advocate equality of resources, and in opposition to those who advocate some version of equality of welfare. Yet Rawls' defense of primary goods is of course part of an argument against utilitarianism, not equality of welfare. His target, in other words, is the position that we should maximize welfare, not that we should equalize it. Now for utilitarians, the notion of utility or welfare—whether it is interpreted hedonistically or is instead construed as consisting in the satisfaction of preferences—plays a double role. First, it supplies the content of the utilitarian theory of what is genuinely valuable or good, both for an individual and sans phrase. Second, it is treated as the appropriate measure to use in assessing personal well-being. These two roles are related, inasmuch as it is the utilitarian theory of the good that underwrites the reliance on utility as the measure of individual well-being. And both of these features of utilitarianism are distinct from its third crucial claim, which is that the right or just thing to do is to maximize overall utility. Rawls is sensitive both to the distinctions and to the relations among these three features of the utilitarian position. He rejects the utilitarian theory of the individual good, which he regards as indefensibly monistic, in favor of the pluralistic view that the good for an individual consists in the successful development and execution of a rational plein of life. And the switch from a monistic to a pluralistic account of the good helps in turn to motivate his rejection of utility in favor of primary goods as the measure of personal well-being, as well as his rejection of maxi-

319

320

John Rawls

16

Philosophy & Public Affairs

mizing conceptions of justice in favor of the idea that what justice requires is the establishment of a fair framework of cooperation within which people can pursue their diverse plans of life. When the position that one ought to maximize aggregate welfare is replaced by the position that one ought to equalize welfare levels, certain objections to the maximizing view are transformed. For example, utilitarianism is often said to be embarrassed by the possibility of "utility monsters"—that is, by the possibility of people who are unusually efficient at converting resources into utility. The objection is that if there were such people, utilitarianism would dictate that resources should be channeled disproportionately to them, without regard for the distributive implications of such a policy. When utilitarianism is replaced by a welfarist version of luck egalitarianism, the problem of utility monsters is replaced by the problem of expensive tastes. Here the objection is that if one aims to equalize welfare, then the distribution of resources in society is held hostage to those whose tastes and preferences are unusually costly to satisfy—that is, to people who are unusually inefficient at converting resources into utility. Welfarist egalitarianism is said to imply that resources should be channeled disproportionately to these people, without regard for the way in which doing so may depress the levels of well-being that others can achieve. In view of the relation between these two objections, it would be instructive to compare the respective conclusions that critics have drawn from them. It would also be useful to explore other connections between utilitarianism and luck egalitarianism. But although a number of luck-egalitarian writers have in fact compared the two positions, their comparisons usually begin from the premise that utilitarianism is most plausible when construed as a theory of equality, rather than as a teleological view that is concerned with maximizing the good.27 Since, as 27. Dworkin makes this claim in Sovereign Virtue (Cambridge: Harvard University Press, 2000), pp. 62-64. It is developed at length by Kymlicka (Contemporary Political Philosophy, chap. 2 and 3, and Liberalism, Community, and Culture [Oxford: Clarendon Press, 1989}, chap. 3), who absorbs it into (what might be called) his "egalicentric" narrative of the history of political philosophy. Kymlicka argues that Rawls's theory of justice is motivated by inadequacies in the utilitarian conception of equality, and that Dworkin's theory of justice is then motivated by inadequacies in Rawls's conception of equality. ("Dworkin's theory was a response to problems in Rawls's conception of equality, just as Rawls's theory was a response to problems in the utilitarian conception of equality" [Contemporary Political Philosophy, p. 85).) Eric Rakowski seems also to endorse the claim in his Equal Justice

John Rawls

17

What is Egalitarianism ?

these writers invariably go on to point out, utilitarianism is actually quite implausible when construed as a theory of equality, their discussions rarely amount to more than a rehearsal of the advantages of luck egalitarianism. But the interpretation of utilitarianism as a theory of equality is itself implausible in my view,28 and so a less narrowly structured investigation of the relations between utilitarian and luckegalitarian principles would be welcome. Discussions of luck egalitarianism have also done a less thorough job, in general, of addressing the most fundamental questions about the justification of that position, than utilitarians and their critics have done in addressing the analogous questions about the justification of utilitarianism. For example, luck-egalitarian debates between welfarists and resourcists do not always distinguish clearly between the question of whether welfarism is an adequate theory of the good, and the question of whether it is the individual good itself, or instead the means of achieving it, that egalitarians should be seeking to equalize. And the idea that justice requires the equal distribution of something is often simply taken for granted as the starting point of discussion. In my view, the luck-egalitarian position is open to doubt on a number of grounds. Perhaps the most obvious difficulty is that the degree of weight that the luck egalitarian places on the distinction between choices and circumstances seems, on its face, to be both philosophically dubious and morally implausible. Philosophically, the question is whether the distinction is deep enough to bear the kind of weight that luck egalitarians place on it. Some luck-egalitarian writings seem implicitly to suggest that whatever is assigned to the category of unchosen (Oxford: Clarendon Press, 1991), pp. 23-25. Amartya Sen, who does not himself put forward a luck-egalitarian theory but whose emphasis on the question "equality of what" has exerted a powerful influence on the development of luck egalitarianism, speaks of the " 'hidden* egalitarianism in utilitarian philosophy" (Inequality Reexamined [Cambridge: Harvard University Press, 1992], p. 13). By this he means that utilitarianism attaches equal importance to the utilities of all people, and that this "egalitarian foundation is ... quite central to the entire utilitarian exercise" (p. 14). Sen is surely correct to say that the fact that utilitarians attach equal importance to the interests of all people is a very significant feature of their view. However, it is a further step to the conclusion that utilitarianism is best thought of simply as a theory of equality Not every theory that favors equality in some respect is best interpreted as being motivated, ultimately, by a conception of what equality requires. 28. For reasons that are well explained in Samuel Freeman, "Utilitarianism, Deontology, and the Priority of Right," Philosophy & Public Affairs 23 (1994): 313-49.

321

322

John Rawls

18

Philosophy &> Public Affairs

circumstance is a contingent feature of the causal order, which is not under the individual's control and does not implicate his or her personhood, whereas voluntary choices are fully under the control of individuals and constitute pure expressions of their agency. But this contrast is, of course, untenable. In any sense of identity that actually matters to people, unchosen personal traits and the social circumstances into which one is born are importantly, albeit not exclusively, constitutive of one's distinctive identity.29 And, in any ordinary sense of "voluntary," people's voluntary choices are routinely influenced by unchosen features of their personalities, temperaments, and the social contexts in which they find themselves. In his defense of a version of luck egalitarianism,30 G. A. Cohen concedes that reliance on the distinction between choices and circumstances may leave luck egalitarians "up to our necks in the free will problem," but he says—perhaps with some irony?—that "that is just tough luck."31 Cohen points out that one of the achievements of luck egalitarianism is to demonstrate that egalitarians can incorporate "the most powerful idea in the arsenal of the anti-egalitarian right: the idea of choice and responsibility."32 Yet, in so doing, luck egalitarianism invites the objection that, like the political philosophies of the antiegalitarian right, it tacitly derives much of its appeal from an implausible understanding of the metaphysical status of the category of choice. This objection would be easier to dismiss if the luck-egalitarian account of the significance of choice were morally compelling. But in fact that account seems on its face to be morally implausible. It is morally implausible, for example, that justice requires individuals to be fully compensated for disadvantages that derive from unchosen features of their circumstances but not to be compensated at all for disadvantages that result from their voluntary choices. As Anderson has argued, it is morally implausible that choice should have that kind of significance or make that degree of difference. On the one hand, there are many unchosen personal attributes that may be disadvantageous but for which we do not, in fact, demand compensation from others. On the other hand, the fact that a person's urgent medical needs can be traced to his own 29. This is a theme of several of the essays included in my Boundaries and Allegiances (New York: Oxford University Press, 2001). 30. G. A. Cohen, "On the Currency of Egalitarian Justice," Ethics 99 (1989): 906-44. 31. Ibid., p. 934. 32. Ibid., p. 933.

John Rawls

19

What is Egalitarianism?

negligence or foolishness or high-risk behavior is not normally seen as making it legitimate to deny him the care he needs. Still less do people automatically forfeit any claim to assistance if it turns out that their urgent needs are the result of prudent or well-considered choices that simply turned out badly33 We are neither so systematically alienated from the unchosen aspects of our own identities nor so uniformly confident of and identified with our role as choosers as to regard the presence or absence of choice as having this kind of make-or-break significance. This helps explain why the appeal of luck egalitarianism may seem tacitly to depend on a form of metaphysical libertarianism, for libertarianism may appear to promise a basis in metaphysics for a dichotomy that would otherwise seem so stark as to be morally untenable.34 Some luck-egalitarian writers have sought to address these difficulties by qualifying the luck-egalitarian principle or limiting its application in various ways. And some writers have suggested that certain of the difficulties can be avoided by drawing the line between choices and circumstances differently. Dworkin, for example, in explaining the differences between his view and Cohen's, argues that the line should be rooted in "ordinary people's ethical experience."35 If the distinction is drawn in this way, he says, then various aspects of one's personality that are not in any 33. See Anderson, "What is the Point of Equality," pp. 296-300. 34. In "Responsibility, Reactive Attitudes, and Liberalism in Philosophy and Politics" (Philosophy &> Public Affairs 21 [1992]: 299-323, reprinted in Boundaries and Allegiances, pp. 12-31), I argued that contemporary liberalism has been exposed to conservative attack because of a perceived tendency among many liberals, both in politics and in philosophy, to rely on a reduced conception of individual agency and responsibility. The points made in the last two paragraphs might be taken to suggest that what luck egalitarianism has done, in effect, is to overcompensate for that perceived tendency. 35. Dworkin's reply to Cohen takes up sections II through IV of chap. 7 of Sovereign Virtue. The quoted phrase occurs on pp. 289-90. There is additional discussion of the ethical interpretation of the distinction between choices and circumstances on pp. 322-25. It is worth noting that in his original formulations of his equality of resources scheme, Dworkin did not characterize the relevant distinction as a distinction between choices and circumstances at all. He tended to speak instead of a distinction between the person and his circumstances. See, for example, Sovereign Virtue, pp. 81 and 140. He was criticized for this by Cohen (in Section IV of "On the Currency of Egalitarian Justice"), who argued that the relevant notion is choice rather than personality In more recent discussions, such as those cited at the beginning of this note, Dworkin does indeed characterize the distinction as a distinction between choices and circumstances (or choice and chance), but he preserves a link with his earlier characterization by asserting that people's "choices reflect their personalities" (Sovereign Virtue, p. 322). My argument in the text is that this does not yield a distinction between choices and circumstances that is helpful to Dworkin's position.

323

324

John Rawls

Philosophy & Public Affairs

20

straightforward sense chosen—including one's ambitions, tastes, preferences, convictions, and traits of character—will nevertheless fall on the choice side of the line. The reasons are twofold. First, all of these features of personality are relevant to the choices that one makes, either because they supply motives for one's actions or because they "affect [one's] pursuit of" one's ends.36 Second, people normally identify with these aspects of their personalities and see themselves as having to take "consequential responsibility" for them, in the sense that they expect to have to bear the costs of possessing them.37 One conclusion Dworkin draws from this is that people cannot normally demand compensation for expensive values or tastes. Even though such tastes are not actually chosen, they nevertheless fall on the choice side of the line by his standards, at least when people identify with them. Eric Rakowski, who takes a similar position, says that people's values and beliefs are "constitutive elements of themselves for which they must assume responsibility,"38 and that even though one does not choose one's preferences, one can nevertheless choose to strengthen or weaken them, so that they too are appropriately situated on the choice side of the line.39 However, this way of drawing the line between choices and circumstances leads to further difficulties. By the criteria that Dworkin and Rakowski suggest, people's talents and abilities, no less than their ambitions and preferences, may also deserve to be situated on the choice side of the line. After all, people's talents and abilities are often relevant to their choices, either because they shape people's motives or because they affect people's pursuit of their ends. In addition, people frequently view their talents, no less than their values and preferences, as importantly constitutive of their identities. And even if talents are themselves unchosen, people can nevertheless choose whether to develop them. Furthermore, many people also expect to take "consequential responsibility" for their talents, in the sense that they believe they are entitled to the differential rewards that such talents may enable them to secure. If this is correct, then the effect of broadening the category of choice to include not only the actions one actually chooses but also the various constitutive features of oneself that underlie one's choices, is to cast doubt 36. 37» 38. 39.

Dworkin, Sovereign Virtue, p. 322. Ibid., p. 290. Rakowski, Equal Justice, p. 63. Ibid., pp. 57-72.

John Rawls

21

What is Egalitarianism?

on the capacity of the distinction between choices and circumstances to support the substantive positions that luck egalitarians favor.40 There is a further point that a number of critics have made. Luck egalitarianism, as it is often presented, appears to treat equality as being an essentially distributive ideal whose fundamental aim is to eliminate the effects on distribution of "brute luck." Because of this concern to neutralize brute luck, the luck egalitarian arrives at allocative decisions on the basis of judgments that are strongly "inward looking." That is, an individual's claim to be compensated in the name of equality for some disadvantage always depends on a judgment about the source of the disadvantage in different aspects of the self. We cannot know whether an individual's disadvantage entitles her to egalitarian compensation without disentangling the respective contributions made by her will, on the one hand, and by unchosen features of her talents and personal circumstances, on the other hand, to the processes that put her at that disadvantage. For this reason, luck egalitarianism encourages her to look inward in deciding whether she has a legitimate claim on fellow citizens, and, as Anderson and Wolff have emphasized,41 it encourages those fellow citizens both to scrutinize the deepest aspects of her self and to arrive at heavily moralized judgments about the degree of responsibility she bears for her own misfortune. In all of the respects I have mentioned, the luck-egalitarian conception of equality diverges from a more familiar way of understanding that value. Equality, as it is more commonly understood, is not, in the first instance, a distributive ideal, and its aim is not to compensate for misfortune. It is, instead, a moral ideal governing the relations in which people stand to one another. Instead of focusing attention on the differing contingencies of each person's traits, abilities, and other circumstances, this 40. At one point (Sovereign Virtue, pp. 260-63), Dworkin himself insists on the importance, for ethical purposes, of identification with aspects of one's unchosen circumstances. These "parameters," which help define what counts as a good life for an individual, are distinguished from those circumstances that count instead as "limitations" on one's ability to lead a good life. However, as Anthony Appiah suggests in his review of Sovereign Virtue, it is very unclear how a recognition of the ethical role of parameters is to be reconciled with the assignment of fundamental significance to the distinction between choices and circumstances. (See K. Anthony Appiah, "Equality of What?" The New York Review of Books April 26 [2001}: 63-68.) 41. Anderson, "What is the Point of Equality?" (at, for example, p. 310); Wolff, "Fairness, Respect, and the Egalitarian Ethos" (especially pp. 113-18).

325

326

John Rawls 22

Philosophy &> Public Affairs

ideal abstracts from the undeniable differences among people. It claims that human relations must be conducted on the basis of an assumption that everyone's life is equally important, and that all members of a society have equal standing. As Anderson insists, in defending a version of this ideal, equality so understood is opposed not to luck but to oppression, to heritable hierarchies of social status, to ideas of caste, to class privilege and the rigid stratification of classes, and to the undemocratic distribution of power. In contrast to the inward-looking focus of luck egalitarianism, it emphasizes the irrelevance of individual differences for fundamental social and political purposes. As a moral ideal, it asserts that all people are of equal worth and that there are some claims that people are entitled to make on one another simply by virtue of their status as persons. As a social ideal, it holds that a human society must be conceived of as a cooperative arrangement among equals, each of whom enjoys the same social standing. As a political ideal, it highlights the claims that citizens are entitled to make on one another by virtue of their status as citizens, without any need for a moralized accounting of the details of their particular circumstances. Indeed, it insists on the very great importance of the right to be viewed simply as a citizen, and to have one's fundamental rights and privileges determined on that basis, without reference to one's talents, intelligence, wisdom, decisionmaking skill, temperament, social class, religious or ethnic affiliation, or ascribed identity. Of course, things are not quite so simple. The social and political ideal of equality, as I will call it, itself has distributive implications. Furthermore, even if that ideal emphasizes the importance, for certain purposes, of abstracting from the differing contingencies of individuals' situations, it must also concede the necessity of attending to such differences for other purposes. People may claim equal rights as citizens, but the interpretation and application of those rights will often depend on features of their individual circumstances. And special circumstances may at times give rise to special rights. Still, one will think about the distributional implications of equality very differently than luck egalitarianism tempts us to do if one insists that, in the end, the relevant question is about the bearing on distribution of a morally-based ideal of human social and political relations—and not about the optimal way of reflecting in our economy a metaphysical distinction between individuals' choices and their unchosen circumstances. Granted, the social and

JohnRawls

23

What is Egalitarianism?

political ideal of equality may itself be understood in different ways.42 But unless distributive egalitarianism is anchored in some version of that ideal, or in some other comparably general understanding of equality as a moral value or normative ideal, it will be arbitrary, pointless, fetishistic: no more compelling than a preference for any other distributive pattern. If one keeps firmly in mind the fact that questions about egalitarian distributive norms must be controlled by some broader understanding of equality, then the appeal of luck egalitarianism seems to me limited. As I have already suggested, many people accept what I have called the social and political ideal of equality. That ideal does not support the ambition of purging the influence of brute luck from human relations, which is just as well since one has only to describe that ambition for its folly to be evident. As Anderson argues, questions of distribution are important, for people who are committed to the social and political value of equality, not because a properly designed set of distributive institutions can help to minimize the influence of luck, but rather because certain kinds of distributive arrangements are incongruous with that social and political value. Clearly, for example, people whose basic needs have not been met—people who lack adequate food, clothing, shelter, education, or medical care—cannot participate in political life or civil society on a footing of equality with others, or can do so only with great difficulty. Even if basic needs have been met, a society cannot be considered a society of equals if the resources that individuals have available to pursue their most cherished ends is left entirely at the mercy of market forces. Moreover, significant distributive inequalities can all too easily generate inequalities of power and status that are incompatible with relations among equals. Thus, those who accept the social and political ideal of equality will have compelling reasons to avoid excessive variations in people's shares of income and wealth, and this will mean, among other things, that they have reason to oppose institutions that allow too much scope for differences in people's natural and social circumstances to translate into economic inequalities. But, on the one 42. One important interpretative issue, about which there is no general consensus, is what the ideal implies about the relations among people who are not citizens of the same country or members of the same political society. Debra Satz discusses this issue in her unpublished paper, "Inequality of What and Between Whom? When and Where Does Inequality Matter? The Case of Inequality Between Nations."

327

328

John Rawls

24

Philosophy &> Public Affairs

hand, this is very different from having the general ambition of eradicating the distributive effects of brute luck, and, on the other hand, it does not assume that inequalities arising from people's choices are always acceptable. From this perspective, the most important questions concern not the neutralization of luck but rather, for example, the nature of people's "basic needs/' the proper criteria for political institutions to use in distinguishing between genuine needs and what are merely very strong preferences, the appropriate measure for the social and political institutions of a complex modern society to employ when assessing the wellbeing of individual citizens, and, especially, the degree of material inequality that is compatible with a conception of society as a fair system of cooperation among equals. Depending on how these questions are answered, people who are committed to the social and political ideal of equality may end up supporting a system that tolerates either more or less inequality of income and wealth than luck egalitarianism does. In either case, however, the motivation for their position will be different from the motivation for a luck-egalitarian outlook. Since the position that adherents of the ideal favor will have been developed in response to questions that differ, at least in part, from the ones that luck egalitarians ask, the basis for their position will lie in considerations that differ from the ones that luck egalitarians cite.43 Ill

All of this helps to explain why it is misleading to treat Rawls's theory of justice as representing a kind of incipient luck egalitarianism. Notwithstanding his remarks about the moral arbitrariness of people's natural attributes and of the social positions into which they are born, his failure to claim that justice requires people to be compensated for all disadvantages resulting from unchosen features of their circumstances is no mere oversight. Despite occasional remarks to the effect that his con43. In "Equality and Justice/' in Ideals of Equalityy ed. Andrew Mason (Oxford: Blackwell, 1998), pp. 21-36, David Miller draws a distinction between distributive equality and social equality that is similar to the distinction 1 have drawn in this section, although Miller takes a different view of the relations between the two notions than I would. In the same volume, Richard Norman ("The Social Basis of Equality," pp. 37-51) invokes Miller's chapter in defending what he calls "socially-located egalitarianism."

John Rawls

25

What is Egalitarianism ?

ception of justice "nullifies the accidents of natural endowment and the contingencies of social circumstance as counters in [a] quest for political and economic advantage/'44 it is quite clear from his discussion as a whole that the underlying motivation for Rawls's theory of justice is not the general elimination of the influence of brute luck on distribution. Indeed, Rawls explicitly emphasizes the differences between his own theory and what he calls "the principle of redress." This is "the principle that undeserved inequalities call for redress; and since inequalities of birth and natural endowment are undeserved, these inequalities are somehow to be compensated for."45 Rawls says that the principle of redress is plausible "only as a prima facie principle, one that is to be weighed in the balance with others."46 His own theory, accordingly, "gives some weight to the considerations singled out by the principle of redress,"47 but it is clearly "not the same as"48 that principle. Rather than trying to devise a conception of justice that will minimize the effects of brute luck, Rawls aims to identify the most reasonable conception of justice to regulate the basic structure of a modern democratic society. For the purposes of this enterprise, a society is conceived of as a fair system of cooperation among free and equal people, each of whom is taken to have the capacity for a sense of justice and the capacity to develop and pursue a rational plan of life which is constitutive of his or her good. Different plans are rational for different people, and the human good is irreducibly heterogeneous. Rawls cites the "moral arbitrariness" of natural attributes and social contingencies, not because his ultimate aim is to extinguish the influence of all arbitrary factors but rather because he thinks the arbitrariness of the factors he cites serves to undermine both an important objection and an influential alternative to his view. The objection is that those people who are more talented or intelligent or hard-working than others deserve greater economic rewards than his theory would permit them to secure. And the alternative that many of those who press this objection favor is "the system of natural liberty," which, as we have seen, allows people to compete in a largely 44. Rawls, A Theory of Justice, p. 15. 45. Ibid, p. 100. 46. Ibid., p. loi. 47. Ibid., p.ioo. 48. Ibid., p. 101. Rakowski believes (mistakenly, in my view) that Rawls relies on the principle of redress as a premise of his argument. See Equal Justice, pp. 112-14.

329

330

John Rawls

26

Philosophy & Public Affairs

unregulated free market, constrained only by the need to respect one another's basic liberties and by a requirement of formal equality of opportunity. Rawls believes that the appeal of the system of natural liberty is morally spurious. The extent to which the system allows people's material prospects to be influenced by their natural assets and the social circumstances into which they are born is indefensible. It is indefensible because it is incongruous with people's status as equals and because the distribution of those contingencies does not itself have any moral basis. If we take seriously the idea that all citizens have equal standing in society, and the idea that each has an equally important interest in developing and pursuing a rational plan of life within a fair cooperative framework, then it is inappropriate to set up an institutional scheme that makes people's chances of carrying out their plans depend so heavily on natural and social contingencies that in themselves have no moral authority. Rawls's emphasis on the moral arbitrariness of people's natural attributes and social starting points is meant to undercut our tendency to treat those factors as morally authoritative, especially when doing so would compromise something morally fundamental. What is relevant for Rawls, in other words, is the conjunction of two points. The first point is that the distribution of natural and social contingencies lacks any moral basis. The second point is that a system that allows the economic distribution to track the distribution of those contingencies too closely will compromise the status of some citizens as equals, for it will undermine their ability to satisfy the equally legitimate interest that each citizen has in developing and pursuing a rational plan of life that is constitutive of his or her good. If Rawls is right, the conjunction of these two points gives us reason to reject the system of natural liberty, once we conceive of society as a fair system of cooperation among free and equal people. But the importance of these points neither derives from nor commits Rawls to the general ambition of neutralizing all of the distributive effects of bad brute luck. For Rawls, what is fundamental is the status of citizens as equals, and the moral arbitrariness of people's natural and social starting points is important because it helps to clarify the distributive implications of taking equal citizenship seriously. Rawls's defense of his reliance on primary goods as the basis for interpersonal comparisons of well-being also looks very different when it is

John Rawls

What is Egalitarianism?

27

not seen through the lens of luck-egalitarian concerns. In saying that citizens in a well-ordered society are expected to take responsibility for their ends, in the sense that they are expected to "regulate and revise their ends and preferences in the light of their expectations of primary goods/'49 Rawls is certainly not making a claim about the moral significance of a putative metaphysical distinction between voluntary choices and unchosen circumstances. Instead, he is making an observation about how his principles of justice serve, in effect, to allocate responsibility between society and the individual. He writes: This conception [of justice] includes what we may call a social division of responsibility: society, the citizens as a collective body, accepts the responsibility for maintaining the equal basic liberties and fair equality of opportunity, and for providing a fair share of the other primary goods for everyone within this framework, while citizens (as individuals) and associations accept the responsibility for revising and adjusting their ends and aspirations in view of the all-purpose means they can expect, given their present and foreseeable situation.50 As Rawls notes, this allocation of responsibility would make no sense if, in general, people were unable "to moderate the claims they make on social institutions"51 because they lacked the capacity to adjust their plans in light of the resources they could expect to have at their disposal under the terms of a fair distribution. By the same token, the allocation of responsibility would make no sense if society as a whole lacked the capacity to establish institutions capable of guaranteeing the basic liberties and fair equality of opportunity. In fact, as Rawls suggests, it does not seem unreasonable to suppose that people normally do have the capacity to adjust their plans in light of their fair expectations. What is more important for the purposes of our discussion, however, is that the suggestion that they have such a capacity is not a metaphysical thesis about the relation of causation and the will, nor, in any case, does it provide the motivation for the allocation of responsibility. People are asked to accept responsibility for their ends, in Rawls's sense, not because the metaphysics of the will makes it fitting that people should bear the costs of their choices, but rather because it is reasonable to expect people to 49. Rawls, Collected Papers, p. 370. 50. Ibid., p. 371. 51. Ibid.

331

332

John Rawls

28

Philosophy & Public Affairs

make do with their fair shares. And what makes shares fair, according to Rawls, is not that they compensate people for all unchosen disadvantages while leaving them to bear the costs (or reap the rewards) of their voluntary choices. Shares are fair when they are part of a distributive scheme that makes it possible for free and equal citizens to pursue their diverse conceptions of the good within a framework that embodies an ideal of reciprocity and mutual respect.52 In addition, Rawls emphasizes that his reliance on primary goods as the basis for interpersonal comparisons is limited to the special case of social justice. As he says, "Primary goods are not... to be used in making comparisons in all situations but only in questions which arise in regard to the basic structure. It is another matter entirely whether primary goods are an appropriate basis in other kinds of cases/'53 What is distinctive about the case of social justice is that interpersonal comparisons must be capable of providing grounds for adjudicating among conflicting claims in a way that all citizens can recognize as fair. It follows, Rawls believes, that we need a practical basis for making these comparisons and that this basis must lie in features of people's situations that are publicly accessible and can be appraised without violating people's liberties or subjecting them to unduly intrusive examination. As he says, "the idea is to find a practicable basis of interpersonal comparisons in terms of objective features of citizens1 social circumstances open to view/'54 This is in striking contrast to the inward-looking focus of luck egalitarianism. Whereas the aim of neutralizing the distributive effects of brute luck requires intrusive and conceptually problematic judgments about the inner sources of people's disadvantages, the aim of adjudicating fairly among the claims of free and equal citizens requires judgments that rest on a practicable and public basis. None of this is to suggest that Rawls's views are beyond criticism. It is perfectly possible to argue, for example, that the difference principle in52. For a related discussion of the passage from Rawls discussed in this paragraph, see T. M. Scanlon, "The Significance of Choice/' in The Tanner Lectures on Human Values, vol. 8, ed. Sterling McMurrin (Salt Lake City: University of Utah Press, 1988), pp. 151-216, at 197-201.1 have also discussed Rawls's doctrine of "responsibuity for ends" in section IV of "Responsibility, Reactive Attitudes, and Liberalism in Philosophy and Politics" (see Boundaries and Allegiances, pp. 29-30). 53. Rawls, Collected Papers, p. 364. 54. Rawls, "The Priority of Right and Ideas of the Good/' in Collected Papers, pp. 449-72 at pp. 454-55.

John Rawls

29

What is Egalitarianism ?

appropriately rewards those who are among the worst-off, as measured in terms of primary goods, only because they prefer greater leisure to greater income and so choose to work at less demanding jobs or not to work at all. It is also perfectly possible to argue that Rawls's reliance on primary social goods as the measure of well-being renders him incapable of taking adequate account, as an acceptable theory of justice surely must, of those who have special medical conditions that are unusually costly to treat. Indeed, Rawls himself concedes the force of both of these criticisms. In response to the first, he indicates that despite some reservations, he is prepared to contemplate an expansion of the list of primary goods to include leisure time.55 In response to the second, he acknowledges the importance of making provision for those with special medical needs but treats this as a "problem of extension"56 which is to be deferred until after "the first problem of justice'*57 has been addressed. This problem concerns the relations among "citizens who are normally active and fully cooperating members of society over a complete life."58 Accordingly, for the purpose of addressing this problem, Rawls makes the frankly idealized assumption that all citizens have, "at least to the essential minimum degree, the moral, intellectual, and physical capacities that enable them to be fully cooperating members of society over a complete life/'59 Elsewhere he describes himself as assuming "that all citizens have physical and psychological capacities within a certain normal range."60 He speculates that the problem of special medical needs can be dealt with "at the legislative stage when the prevalence and kinds of these misfortunes are known and the costs of treating them can be ascertained and balanced along with total government expenditure."61 These tentative lines of response may or may not prove adequate. Perhaps they are too ad hoc or require intuitionistic balancing of a kind that Rawls had hoped to avoid. But it is a mistake to regard the criticisms to which he is responding as the thin end of a wedge: that is, as leading in55. See Rawls, Collected Papers, pp. 252-53 and 45556. See Rawls, Political Liberalism (New York: Columbia University Press, 1993), pp. 21-22, and Collected Papers, p. 531. 57. See Rawls, Collected Papers, pp. 259 and 368. 58. Ibid., p. 368. 59. Rawls, Political Liberalism, p. 183. 60. Rawls, Collected Papers, p. 368. 61. Rawls, Political Liberalism, p. 184.

333

334

John Rawls

30

Philosophy &> Public Affairs

exorably to luck egalitarianism. The aim that informs Rawls's response to the problem of special medical needs, and that would also have to inform any better response, is the aim of "restor[ing] people by health care so that once again they are fully cooperating members of society."62 This, once again, is entirely independent from and does not commit Rawls to the aim of compensating people for all disadvantages resulting from unchosen circumstances. Instead, the aim of enabling people to be fully cooperating members of society provides an independent standard for judging which disadvantages should be compensated. By this standard, some disadvantages should be compensated even if they result from bad "option luck," whereas others should not be compensated even if they result from bad "brute luck." For example, the standard may require that people be provided with medical care even if their need for care results from choices that they made voluntarily but that turned out badly. On the other hand, it does not require that people be compensated for expensive tastes simply because those tastes result from unchosen features of their upbringing, for such compensation is not normally necessary to enable people to be cooperating members of society. As Rawls says, "We don't say that because the preferences arose from upbringing and not choice that society owes us compensation. Rather, it is a normal part of being human to cope with the preferences our upbringing leaves us with."63 In discussing the relation between Rawls's view and the luck-egalitarian position, my primary concern has not been interpretative. By emphasizing the differences between the two positions I have hoped instead to accomplish two things. The first is to make clear that the philosophical underpinnings of luck egalitarianism are not to be found in Rawls's work. Thus, if luck egalitarianism is to be supplied with a compelling motivation, that motivation will need to come from somewhere else; it 62. Ibid. 63. Rawls, Political Liberalism, p. 18511. Of course, luck egalitarians disagree among themselves as to whether people should receive compensation for expensive unchosen tastes, for they disagree about whether such tastes are properly subsumed under the heading of "choice" or of "circumstance." Dworkin, as we have seen, assimilates most preferences to the "choice" side of the luck-egalitarian divide, despite the fact that they are not themselves chosen, and he concludes that in general no compensation is called for. For Rawls, however, the question is not whether preferences are properly subsumed within the category of "choices." The question, instead, is whether compensation for expensive tastes is necessary to enable people to participate in the scheme of social cooperation.

John Rawls

31

What is Egalitarianism ?

cannot simply ride piggy-back on Rawls's remarks about the arbitrariness of the natural lottery or about the need for citizens to take responsibility for their ends. My second aim has been to illustrate how a plausible form of distributive egalitarianism can be anchored in a more general conception of equality as a social and political ideal. Rawls's theory shows how this can be done. For Rawls, people are conceived of as free and equal citizens, and the aim is to determine which principles of distributive justice are most appropriate for a modern democratic society whose members are so understood. In other words, the question is which principles of justice are most consistent, in modern conditions, with the freedom and equality of persons. Equality is understood as a social and political ideal that governs the relations in which people stand to one another. The core of the value of equality does not, according to this understanding, consist in the idea that there is something that must be distributed or allocated equally, and so the interpretation of the value does not consist primarily in seeking to ascertain what that something is. Instead, the core of the value is a normative conception of human relations, and the relevant question, when interpreting the value, is what social, political, and economic arrangements are compatible with that conception. IV

As I have suggested, proponents of luck egalitarianism often give the impression that they conceive of equality as an essentially distributive ideal. They begin from the premise that there is some currency that should be distributed equally and then proceed to investigate what that currency might be. I have argued that this is misguided. Any form of distributive egalitarianism, if it is to be at all plausible, must be anchored in some more general conception of equality as a moral value or normative ideal. This does not by itself disqualify luck egalitarianism as the best form of distributive egalitarianism since it is open to luck egalitarians to argue that it can be so anchored. Yet I have also argued that once we remember that egalitarian distributive norms must flow from a broader conception of equality, the appeal of luck egalitarianism is limited. It may be protested that in developing these arguments, I have not done justice to the luck-egalitarian position. Luck egalitarians often present their view as expressing the intuitive idea that it is unfair if some

335

336

John Rawls

32

Philosophy &> Public Affairs

people are worse-off than others owing to factors beyond their control.64 And, in response to my arguments, they may assert that this idea in turn is rooted in a conception of people as having equal moral worth, so that luck egalitarianism does flow from a broader conception of equality as a moral value. It does not treat equality as being an essentially distributive ideal. However, the idea that it is unfair if some people are worse-off than others owing to factors beyond their control captures only part of the luck-egalitarian position. Luck egalitarians also believe that it is not unfair if some people are worse-off (or better-off) than others as a result of their voluntary choices. Thus, the luck-egalitarian claim must be that the position that results from combining these two formulations is rooted in a conception of people as having equal worth. This claim requires argument. Most theories of justice purport to treat people as having equal moral worth or importance. If luck egalitarians believe that their position is the best expression of the equal worth of persons, then that claim requires some defense; it cannot simply be assumed. This is particularly true because it is by no means clear that most people actually have the intuitions to which luck egalitarians appeal. To be sure, there are some contexts in which it does seem unfair that people should suffer disadvantages as a result of circumstances beyond their control, and there are some contexts in which it does not seem unfair that people should suffer disadvantages as a result of their voluntary choices. But these observations alone do not suffice to establish the intuitive credentials of luck egalitarianism, for luck egalitarianism asserts something more sweeping. It claims that, in general, it is unfair if people are better or worse off as a result of factors beyond their control and that, in general, it is not unfair if people are better or worse off as a result of their voluntary choices. It is far from clear that, in its generalized form, this claim enjoys widespread intuitive support.65 The more common or intu64. For versions of this formulation, see Arneson, "Equality and Equal Opportunity for Welfare/' p. 85, and Larry Temkin, Inequality (New York: Oxford University Press, 1993), p. 200. 65. The generalized luck-egalitarian claim bears a superficial similarity to the equally generalized claim that it is inappropriate to blame people, or to judge them morally responsible, for things that are beyond their control. The latter claim, which is of course central to discussions of free will and responsibility, is indeed widely accepted. Despite the superficial similarity between the two claims, however, there are also important differences. The claim about blame and responsibility articulates a precondition for responsible agency

John Rawls

33

What is Egalitarianism ?

itive view, I believe, is that the fairness or unfairness of differences in advantage resulting from, on the one hand, factors beyond people's control and, on the other hand, people's voluntary choices, is highly dependent on the prevailing social context and institutional setting. For example, if I have a less successful career as a philosopher than you do because your superior philosophical gifts enable you to refute all my arguments, then, contrary to what the generalized claim might lead us to expect, most people would not regard that as unfair. Nor would most think it unfair if a naturally gifted professional athlete were offered a more lucrative contract than his less talented teammate. On the other hand, most people would consider it outrageous if an emergency room doctor left an injured patient untreated simply because the patient's injury resulted from a foolish but voluntary decision. And few would think it acceptable to deny legal counsel to an indigent defendant on the ground that her inability to pay for an attorney was the result of poor financial decision making on her part. If this is correct, then the intuitive credentials of luck egalitarianism cannot be taken for granted, and it is all the more important for luck egalitarians to explain how exactly their position is supposed to follow from a conception of persons as having equal moral worth. Moreover, I have argued that equality is most compelling when it is understood as a social and political ideal that includes but goes beyond the proposition that all people have equal moral worth. It is this ideal that we invoke when we say that our society should be organized as a society of equals. The case is analogous to other human relationships that we take to be governed by an ideal of equality. When we say, for example, that a friendship or a marriage should be a relationship of equals, we do not mean merely that the participants are of equal moral worth but also that their relationship should have a certain structure and character. that strikes most people, upon reflection, as compelling. The luck-egalitarian claim, by contrast, is a first-order normative judgment that treats equality as the default distributive baseline and claims that departures from that baseline are justified if and only if a suitable condition of control is satisfied. As I argue above, the intuitive plausibility of that normative judgment cannot be taken for granted. Moreover, such plausibility as it does have may depend on seeing it, wrongly in my view, as a natural extension of the claim about the preconditions of responsible agency. This is related to my earlier observation (at pp. 18-19 above) that luck egalitarianism may tacitly depend for its appeal on a perception that it can be validated metaphysically.

337

338

John Rawls

34

Philosophy &> Public Affairs

Similarly, I believe, our notion of a society of equals expresses a normative ideal of human relations. Luck egalitarians must either reject this ideal or argue that it too supports their distributive principles. It is not clear what grounds they have for rejecting it, but, as I have already indicated, I see no reason to think that it supports the generalized luckegalitarian position. Since, as I have argued, neither complete choicesensitivity nor complete endowment-insensitivity seems morally plausible in its own right, and since neither reflects our intuitive understanding of the implications of equality, we have no reason to think that either would be a feature of a society of equals. Ronald Dworkin is one writer who can be interpreted as attempting to anchor luck-egalitarian principles in an ideal of equality that goes beyond the proposition that all people are of equal worth, but his ideal of equality is not the same as the social and political ideal I have described. Dworkin states that a "full theory of equality" seeks to answer the question of what it is to treat people as equals. A theory of distributive equality addresses one aspect of this question, namely, which scheme of distribution may be said to treat people as equals. Thus Dworkin presents his equality of resources scheme neither as a set of first principles nor as rooted merely in the idea that people are of equal worth but rather as part of a spelling out of the idea of equal treatment. In effect, then, he can be interpreted as attempting to anchor luck-egalitarian principles in a more general ideal of equality—namely, the ideal of treating people as equals. Dworkin says that treating people as equals means treating them with equal concern. Indeed, he says that "[e]qual concern is the sovereign virtue of political community" and that "[n]o government is legitimate that does not show equal concern for the fate of all those citizens over whom it claims dominion and from whom it claims allegiance."66 Some writers have denied that the ideal of equal concern really does support luck-egalitarian principles. G. A. Cohen, for example, says that, contrary to what Dworkin supposes, "it seems quite unclear that a state which forthrightly refuses to pursue a norm of strict distributive equality ipso facto shows failure to treat its subjects with equal respect and concern."67 Anderson goes further, arguing that luck egalitarianism fails "the most 66. Dworkin, Sovereign Virtue, p. i. 67. G. A. Cohen, If You're an Egalitarian, How Come You're So Rich? (Cambridge: Harvard University Press, 2000), p. 165.

John Rawls

35

What is Egalitarianism?

fundamental test any egalitarian theory must meet: that its principles express equal respect and concern for all citizens."68 Whether or not Dworkin's ideal of equality supports his distributive principles, however, there is a prior question about the adequacy of that ideal itself. It is noteworthy, I think, that when, in his famous two-part article "What is Equality?,"69 Dworkin first broaches the question of how the ideal of treating people as equals might best be applied to issues of distribution, he does so with reference to the example of a wealthy man who is deciding how, in drawing up his will, to divide his wealth among his children, each of whom has different needs, ambitions, and tastes. He then moves on to consider a putatively analogous question that arises in "an ordinary political context," namely, the question of how "officials" should make decisions about the distribution of resources among citizens.70 Now the first thing to notice is that the model of testator and heirs is an asymmetrical model, with one party distributing benefits and the others receiving them. Thus, when this model is applied to questions of economic distribution within society, citizens are represented, again asymmetrically, as objects of treatment by some kind of centralized subject. In this case, there are important questions that arise about the identity of the subject. Is it, as Dworkin's language seems at times to suggest, a group of "officials"? Is it, as he suggests at other times, "the distributive scheme" itself? Or is it, as some of his formulations appear to imply, the society or community as a whole? And how is any of these to be understood as a locus of agency? Whatever the answers to these questions may be, what is really striking is that the model of testator and heir does not describe a relationship among equals at all. An autocratic parent might write his will in such a way as to treat all of his children as equals in Dworkin's sense, but that would not transform his family into an egalitarian social unit. Similarly, an autocratic government might impose an economic system 68. Anderson, "What is the Point of Equality?" p. 289. 69. Reprinted as chap, i and 2 of Sovereign Virtue. 70. Dworkin, Sovereign Virtue, p. 13. Dworkin is not the only luck-egalitarian writer who uses the model of testator and heirs to illustrate what I refer to in the text below as an "administrative conception" of equality. In his article "Equality and Responsibility" (Boston Review10 [April/May 1995]: 3-7 at p. 3)» John Roemer uses the same model to illustrate his own version of an egalitarian distributive theory. And Roemer says elsewhere that the aim of his theory is to provide "an algorithm" ("A Pragmatic Theory of Responsibility for the Egalitarian Planner/' p. 182) that could be used by "an egalitarian planner" (p. 180) to "implement" (p. 182) the "correct egalitarian ethic" (p. 180).

339

340

John Rawls

36

Philosophy &> Public Affairs

that treated individuals as equals in Dworkin's sense/1 but that would not transform the society into an egalitarian political community. In short, Dworkiris ideal of equality, as applied to questions of distribution, is not itself a model of social or political equality at all; it is perfectly compatible with social hierarchy, inasmuch as it involves one relatively powerful party choosing how to distribute resources among those with relatively less power.72 Of course, it might be protested that there is no reason why the same model could not also be applied to the distribution of power itself. But then who would be doing the distributing? In fact, however, Dworkin does not advocate the equal distribution of power. After distinguishing what he calls "the two modes of power,"73 impact and influence, he goes on to argue, "If a community is genuinely egalitarian in the abstract sense—if it accepts the imperative that a community collectively must treat its members individually with equal concern—then it cannot treat political impact or influence as themselves resources, to be divided according to some metric of equality the way land or raw materials or investments might be divided."74 Accordingly, Dworkin would allow a "limited place for equality of impact but none for equality of influence."75 The upshot is that Dworkin's ideal of equality 71. Indeed, Dworkin himself says that a "benevolent tyranny" might bring about "a more egalitarian distribution than a democracy could" (Sovereign Virtue, p. 187). 72. In a discussion of this point, Stephen Perry objected that the device of the auction, which plays a much more developed role in Dworkin's argument than does the model of testator and heirs, is not comparably hierarchical. In the auction, a number of shipwreck survivors on a desert island are given an equal number of clamshells, which they use to bid for available resources. Since no role is envisioned for a government, and since all of the participants in the auction have the same status (leaving aside the role of the auctioneer), the auction might be thought to provide a model of genuine social equality. But the primary role of the auction, as Dworkin describes it, is to tell us what kind of distribution of resources should properly be counted as an equal distribution. It is intended to undercut the idea that distributive egalitarians want each person at all times to have the same amount of wealth. The auction is not meant to illustrate a more general ideal of social equality, nor do the participants in the auction ask themselves what distributional practices are appropriate to a society of equals. Indeed, Dworkin gives no general characterization at all of the social relations among the participants. Instead, they are simply stipulated to accept the assumption that resources shall be "divided equally among them" (Sovereign Virtue, pp. 66-67), and the auction presupposes that aim. Dworkin's underlying idea of equality, which is not itself illustrated by the auction, is the (asymmetrical) idea that a government must show equal concern for all of its citizens; as he insists, equal concern is "the special and indispensable virtue of sovereigns" (Sovereign Virtue, p. 6). 73- Dworkin, Sovereign Virtue, p. 199. 74. Ibid., pp. 209-10. 75. Ibid., p. 200.

John Rawls

37

What is Egalitarianism ?

does not require or even permit an equal distribution of power; and the kind of equal distribution of resources that it does require is not incompatible with social hierarchy or even, as he himself says, with "benevolent tyranny."76 Dworkin has developed his ideal of equality in great detail over a period of many years, and he has applied it with great force to an impressive range of issues. The observations I have made do not detract from the many virtues of his view. But those observations do, I think, suggest that Dworkin's ideal represents what might be called an administrative conception of equality. A "genuinely egalitarian" community, he believes, is one that accepts the abstract principle that people should be treated as equals (or treated with equal concern) and whose "officials" then administer social and political institutions in accordance with the best conception of what treatment as equals requires. But an egalitarianism that begins from the question of how best to administer or operationalize an abstract principle of equal concern contrasts sharply with an egalitarianism that begins from the question of what relationships among equals are like and goes on from there to consider what kinds of social and political institutions are appropriate to a society of equals.77 If the 76. Ibid., p. 187. Dworkin would of course reject a "benevolent tyranny" for other reasons. His point (and mine) is merely that considerations of distributional equality do not in themselves furnish grounds for doing so. 77. My characterization of Dworkin's conception of equality as administrative is closely related to G. A. Cohen's description of it as "statocentric" (see If You're an Egalitarian, How Come You're So Rich?p. 165). Cohen writes: Within Dworkin's theory of equality, the locus of the norm of equality proper... is in the relationship between the state and those whom it claims the right to govern. Because it claims that right, the state must treat its citizens with equal respect and concern, on pain of being a tyranny, and it must therefore distribute resources equally to its members. But if the state fails to do so, then no analogous duty falls on individuals. It is not the individual's duty to treat everyone (relatives, friends, and strangers alike) with equal respect and concern, (p. 164) This is perhaps an appropriate place to record my sense that there is a tension within Cohen's own writings between his apparent endorsement, in "On the Currency of Egalitarian Justice," of choice as an inequality-legitimating factor and his criticism, in his Tanner Lectures ("Incentives, Inequality, and Community," in The Tanner Lectures on Human Values, vol. 13, ed. Grethe Peterson [Salt Lake City: University of Utah Press, 1992}, pp. 263-329) and in If You're an Egalitarian, How Come You're So Rich? of incentive-based inequalities as unjust precisely because of their choice-dependence. I am inclined to view this as symptomatic of a broader tension between the luck-egalitarian attitude toward choice and the attitudes associated with what Cohen calls an egalitarian "ethos." The emphasis on the importance of a choice-constraining egalitarian ethos is quite congenial to the social and political ideal of equality, but the luck-egalitarian motivation for such an emphasis is less clear.

341

342

John Rawls

38

Philosophy &> Public Affairs

administrative conception of equality supports luck-egalitarian distributive principles—and this is the claim that Anderson denies and about which Cohen expresses doubts—then Dworkin's version of luck egalitarianism is indeed anchored in a more general ideal of equality. But people who think of themselves as egalitarians need to consider how well the administrative conception captures the value of equality as they understand it. It will already be clear that, in my view, the social and political ideal of equality has greater resonance for most people. This does not mean that there is no room for distributive egalitarianism or even for Dworkin's administrative conception of equality. But it does mean that these distributive and administrative ideas will be most compelling if they can be shown to have strong roots in the ideal of a society of equals.78 The recent history of political discourse in liberal societies seems to me to confirm this. I have already noted that the rise of luck egalitarianism among philosophers took place during a period of giddy triumphalism about markets that largely displaced a concern with economic equality in many liberal societies. Yet it is a striking fact that during this very same period, the power of equality as asocial and political ideal continued to make itself felt. One of the great preoccupations of liberal societies in recent decades has been with the normative implications of diversity: with questions about how best to accommodate differences of race, gender, religion, culture, and ethnicity. And these questions have almost always been addressed through an (implicitly if not explicitly) egalitarian lens. The central issue, in effect, has been this: What must a modern liberal society do to constitute itself as a genuine society of equals given that its membership will inevitably be highly diverse in many socially and normatively salient respects? It is this profoundly egalitarian question that has animated the intense debates in liberal societies about pluralism and multiculturalism; about race, gender, and ethnicity; and about the politics of identity. One lesson we should learn from the very intensity of these debates, and from the remarkable social changes that have accompanied them, is the continuing vitality of the idea of a society of equals. And this in turn tells us something—or 78. Dworkin comes close to recognizing the importance of that ideal in his discussion of campaign finance reform in chap. 10 of Sovereign Virtue, in which he places great emphasis on the notion of "citizen equality." But that is not the notion of equality to which he appeals in defending his version of luck egalitarianism.

John Rawls

39

What is Egalitarianism ?

so it seems to me—about the appeal of distributive egalitarianism. What it tells us is that the most compelling versions of that idea, both politically and philosophically, will be those whose source in an ideal of genuine social equality can be vividly and convincingly demonstrated.79

79. Thus, I believe that Will Kymlicka gets things more or less backward when, in chap. 9 of Liberalism, Community, and Culture (Oxford: Clarendon Press, 1989), he seeks to ground a conception of equality for minority cultures in Dworkin's luck-egalitarian distributive principles.

343

This page intentionally left blank

[13] JUSTICE AND GENDER: AN UNFINISHED DEBATE Susan Moller O kin*

INTRODUCTION The revival of Anglo-American political theory and the revival of feminism in what came to be called its "second wave" both emerged early in the second half of the twentieth century. To situate both, consider the publication dates of some of the foundational works of each movement: Simone de Beauvoir's The Second Sex1 appeared in 1949 and was translated into English by 1952; Isaiah Berlin's Two Concepts of Liberty2 was published in 1958, as was John Rawls's Justice as Fairness',3 Betty Friedan's The Feminine Mystique4 appeared in 1963 and Kate Millett's Sexual Politics5 in 1970; John Rawls's complete A Theory of Justice6 came out in 1971. Yet despite the fact that these two revivals were contemporaneous, there was for a couple of decades very little dialogue between the participants in each of them. Feminist political theorists tended to turn their attention to the * Marta Sutton Weeks Professor on Ethics in Society, Stanford University; Matina S. Homer Distinguished Visiting Research Professor, The Radcliffe Institute for Advanced Study, Harvard University 2003-2004. Special thanks to David Miller, Jerry Cohen, and other members of the All Souls' Oxford Political Theory seminar (February 2002); to Cass Sunstein, Martha Nussbaum, David Strauss and other members of their seminar at the University of Chicago Law School (November 10, 2003), and to Linda McClain, Marion Smiley, Tracy Higgins and other participants in the Fordham University School of Law Conference "Rawls and the Law" (November 7-8,2003) for their helpful comments on several different versions of this paper. An earlier version of the paper was translated into French by Patrick Savidan, and appeared as Justice et identité générique: un débat en cours, in 2 Comprendre: Le Lien Familiel 67 (Françoise de Singly & Sylvie Mesure eds., 2001). [Editors' Note: Professor Okin died just as she was completing this Article. The Fordham Law Review recognizes her contribution to the field as a scholar, and is proud to publish her Article as part of this conference on "Rawls and the Law."] 1. Simone de Beauvoir, The Second Sex (H.M. Parshley ed. & trans., Knopf 1953) (1949). 2. Isaiah Berlin, Two Concepts of Liberty (1958). 3. John Rawls, Justice as Fairness (1958), reprinted in John Rawls, Collected Papers 47 (Samuel Freeman éd., 1999). 4. Betty Friedan, The Feminine Mystique (1963). 5. Kate Millett, Sexual Politics (1970). 6. John Rawls, A Theory of Justice (1971).

346

John Rawls

1538

FORDHAM LAW REVIEW

[Vol. 72

historical canon of political thought, rather than to contemporary works.7 Feminist philosophers embarked on classifying and analyzing the theories of the feminist movement itself.8 And feminist activists, at least in the United States and Great Britain, concerned themselves to a large extent with consciousness raising and with urgent public policy issues such as reforming rape law, drawing attention to domestic violence, and trying to get the Equal Rights Amendment passed. On the other hand, the leaders of the newly re-burgeoning field of political philosophy—John Rawls, Ronald Dworkin, Brian Barry, and Steven Lukes, for example—paid no attention in their writings to the feminist movement and its ideas.9 They appeared to assume, as had most of those in the long tradition that preceded them, that inequalities between the sexes were "natural" rather than political, and were therefore not an appropriate concern for philosophical inquiry. By the end of the century, however, things had changed significantly, as a full-scale debate emerged between John Rawls and various feminist critics of his already classic work, A Theory of Justice. Here, I shall give an account of this debate, and try to advance it a little further. I start by defining some of its key terms and pointing to some of the most relevant concepts of Rawls's theory of justice. Next, I briefly address some feminist critiques of Rawls that have emerged as parts of feminist critiques of liberalism as a whole, together with feminist responses to such critiques. Then I give an account of the debate about gender and justice between John Rawls and those who have critiqued his theory from a feminist perspective but have also valued, and used in their own work, some of its central concepts and arguments. Finally, I give a brief response to Rawls's recently published response to critics on the subject of gender, justice, and the family. I shall argue that, although Rawls eventually responded directly to those I term his "constructive" feminist critics, the distinction between comprehensive and political liberalism that he introduces in Political Liberalism severely diminishes the capacity of his theory of justice to answer feminist criticism. Only by allowing that his principles of justice apply directly to the internal life of families—which Rawls clearly resists—and by restricting "reasonable conceptions of the good" to those that are non-sexist, could one revise 7. See, e.g., The Sexism of Social and Political Theory: Women and Reproduction from Plato to Nietzsche (Lorenne M.G. Clark & Lynda Lange eds., 1979); Jean Bethke Elshtain, Public Man, Private Woman: Women in Social and Political Thought (1981); Susan Moller Okin, Women in Western Political Thought (1979). 8. See, e.g., Alison Jaggar, Feminist Politics and Human Nature (1983). 9. Michael Walzer was a partial exception. In Spheres of Justice (1974), he included a brief chapter on the family, which addressed issues of gender to some extent.

John Rawls

2004]

JUSTICE & GENDER

347

1539

the theory so that it both includes women and has an effective and consistent account of moral development. For clarification, let me first define "justice" and "gender," as I use the terms in what follows. By "justice" I mean social justice or, more precisely, distributive justice—that is to say, the ways in which goods and benefits, burdens and responsibilities, should be allocated within society. Theories of justice are centrally concerned with which initial or acquired personal characteristics or positions in society should affect the social distribution.10 They are about whether, how, and to what extent, beginnings should affect outcomes. And yet, although a person's sex is an initial personal characteristic that is frequently accorded great social significance and has a vast influence on many outcomes, contemporary theories of justice have by and large sidestepped the question of whether, or how, this should be so. "Gender" is a more complicated term. It once applied mainly to language—in English as well as in languages in which far more words are gendered. In its new, much larger meaning, the word refers to the deeply entrenched social institutionalization of sexual difference. This new meaning, developed by feminist scholars in many disciplines, reflects the fact that we now think that so much of what has traditionally been thought of as innate, sexual difference is socially produced or constructed. A whole range of factors—from the almost exclusively female nurture of infants and small children and the toys given to girls and boys to our use of language and our cultural and religious systems of belief—contribute to making us into the women or men, the gendered persons, we become. Moreover, the structure of our societies assumes gender, at the same time helping to perpetuate it. Institutions, from workplaces and schools to legislatures, assume that persons are independent beings who neither need daily care nor give it to others, while they largely ignore the needs of those who in fact provide most of this daily care. Such gendered institutions and social structures contribute to the inclusion and privileging of men and the disproportionate burdening of women and their exclusion from positions of power, wealth and authority. A wide range of academic disciplines, especially in the social sciences (economics being the most resistant to the need for reconstruction), shore up these arrangements by making gendered assumptions that bear little or no relation to the real world, such as that unpaid work has no value, and that families are headed by benevolent altruists. Theories based on such assumptions help to legitimate and reproduce the gendered social institutions and structures that keep women, as a sex, relatively powerless, even in societies in which we have formal legal equality with men. 10. Some have criticized this distributive paradigm of justice, but I cannot address that issue here. See, e.g., Iris Marion Young, Justice and the Politics of Difference (1990).

348

John Rawls

1540

FORDHAM LAW REVIEW

[Vol. 72

Feminist scholars in many disciplines have explored the many dimensions of the social and psychological construction of our gendered identities and our gendered societies, and have uncovered the historical and cultural variability of gender.11 The significance of late twentieth century work on sex, gender, and the relation between the two is vast. Feminist scholarship has undermined centuries of argument that started with the notion that not only the distinct differentiation of male and female, but also the division of labor by sex and the domination of women by men, is natural and inevitable. It has challenged theories that claim to be about individuals, but are actually about adult males (with unmentioned wives in the background), as well as theories based on the idea that families are benevolent despotisms. It thus opens up for discussion and debate many issues not previously thought of as political or economic, or even considered in discussions of justice. Not surprisingly, feminist theorists have raised many questions about both past and contemporary theories of social justice, including Rawls's prominent theory.

I. JUSTICE AS FAIRNESS AND THE BASIC STRUCTURE Rawls states early on that his theory of justice is about "the basic structure" of society. This basic structure, he says, is "the way in which the major social institutions distribute fundamental rights and duties and determine the division of advantages from social cooperation."12 He makes clear from the start that the political constitution and the principal social and economic arrangements are such "major institutions" because [tjaken together as one scheme, [they] define men's rights and duties and influence their life prospects, what they can expect to be and how well they can hope to do. The basic structure is the primary subject of justice because its effects are so profound and present from the start.13

Not surprisingly, given this definition, Rawls specifies that the family, or rather more precisely "the monogamous family" is one such major social institution.14 According to his initial, crucial, characterization of "the primary subject of justice," then, it seems that how families "distribute fundamental rights and duties and determine the division 11. Some important examples are Nancy Chodorow, The Reproduction of Mothering (1978); Dorothy Dinnerstein, The Mermaid and the Minotaur: Sexual Arrangements and Human Malaise (1976); Linda J. Nicholson, Gender and History: The Limits of Social Theory in the Age of the Family (1986); Joan Wallach Scott, Gender and the Politics of History (1988); Michelle Z. Rosaldo, The Use and Abuse of Anthropology, 5 Signs 389 (1980). 12. Rawls, A Theory of Justice, supra note 6, § 2, at 7. 13. Id. 14. See id.

John Rawls

2004]

JUSTICE & GENDER

349

1541

of advantages from social cooperation" should be part of it.15 Accordingly, a reader might expect that, among other issues concerning family justice, the traditional division of labor by sex and its many social, economic and political ramifications will constitute matters of concern in the ensuing theory. As is well known, Rawls claims that the principles of justice chosen in the original position will be fair in the sense that, in arriving at them, persons cannot be partial to themselves, to their personal characteristics or their own conceptions of the good. Since in the original position "[n]o one knows his situation in society nor his natural assets,... therefore no one is in a position to tailor principles to his advantage."16 The two principles that Rawls argues would be so chosen are the principle of equal basic liberty and the difference principle (to which is appended the condition of "fair equality of opportunity"). "[T]he first requires equality in the assignment of basic rights and duties, while the second holds that social and economic inequalities ... are just only if they result in compensating benefits for everyone, and in particular for the least advantaged members of society."17 The only justifiable deviations from socioeconomic equality, according to the difference principle, are those that—indirectly, by providing incentives that increase the size of the economic "pie" that is to be divided—work so as to benefit the least advantaged.18 And the principle of fair equality of opportunity requires that persons of equal abilities and motivations have the same chance of attaining any position in the just society, regardless of their circumstances of origin.19 Much of Part II of A Theory of Justice, entitled "Institutions," is devoted to explaining how the principles of justice would influence the design of the constitutional framework and the various institutions of a just society. In a great mélange of relatively "pure" theory and applied theory, Rawls discusses how rights-establishing constitutions, legislatures, legal systems, economic systems, systems of taxation, intergenerational considerations, constraints on majority rule, and the treatment of conscientious objectors might all operate so as to translate the principles of justice, which he has argued can be derived from the original position, into practice in a just or "well-ordered" society. Significantly, though, Rawls does not discuss how the principles would influence either the internal structures and workings of the family, or their relations with the wider society. Families are 15. Id. 16. Id. § 24, at 139. 17. Id. § 3, at 14-15; see also id. § 26, at 150-61 (presenting a fairly brief account of how the two principles are arrived at in the original position). 18. Id. § 12, at 73. 19. Id. § 14, at 83-90.

350

John Rawls

1542

FORD H AM LAW REVIEW

[Vol. 72

mentioned only twice in this middle section of the book, in connection with the constraints they place on equality of opportunity, and in connection with intergenerational justice, for example as institutions that might play any role in either translating principles of justice into practice or in obstructing such translation, or that might be structured or organized with such roles in mind. Apart from this, they go unmentioned. Subsequently, in Part III, Rawls gives a fairly detailed account of moral development. For he thinks that a just society, such as he argues for, is not easy either to achieve or to maintain. He warns that it will be difficult for people who, once the veil of ignorance is lifted, realize that they could benefit from a less egalitarian social distribution to sustain their commitment to the principles of justice they have chosen in the original position, when their impartiality was assured. Therefore, in order to try to increase the stability of the just society, he devotes considerable attention to the development of a sense of justice in the society's citizens. The family, otherwise largely unmentioned, features prominently as a participant in this process. But rather than discussing its suitability as a nurturing ground for justice, considering how justice might apply to the family, or exploring alternative forms of family, Rawls simply says of it: "given that family institutions are just."20 As we shall see, even those feminists who value major aspects of the theory have serious problems with its lack of explicit discussion of the family and its facile assumption that families are just. But before I discuss their critiques, I wish first to look briefly at the ideas of feminist critics who find Rawls's whole liberal enterprise alien to what they regard as good feminist ways of thinking.

II. FEMINISM AND LIBERALISM: FRIENDS OR FOES? Much feminist political theory has been critical of liberal political theory in general and of Rawls's theory in its entirety. Linda McClain and Martha Nussbaum have addressed such critiques in quite comprehensive and very illuminating papers.21 It is important, as we discuss these critiques of liberalism and responses to them, to distinguish between feminist critiques of liberal theories—including Rawls's theory—in their entirety, and feminist critiques of specific, 20. Id. § 75, at 490 (emphasis added). 21. Martha C Nussbaum, Sex & Social Justice 55-80 (1999); Linda C. McClain, "Atomistic Man" Revisited: Liberalism, Connection, and Feminist Jurisprudence, 65 S. Cal L. Rev. 1171 (1992). For an account and criticism of Iris Marion Young's and Seyla Benhabib's critiques of liberalism, from which I draw in what follows, see Susan Moller Okin, Reason and Feeling in Thinking About Justice, 99 Ethics 229 (1989). Neither McClain, Nussbaum, nor I address Bonnie Honig's postmodernist feminist critique of A Theory of Justice in Political Theory and the Displacement of Politics (1993), which focuses on a couple of pages of Rawls's 600-page book.

John Rawls

2004]

JUSTICE & GENDER

351

1543

though sometimes very important, aspects of liberal theories, such as the public/private dichotomy and the (explicit or implicit) assumption that the subjects of liberalism are the male heads of households. The latter critiques, I contend, can be addressed without abandoning the central tenets of liberalism and liberal theory; the former, which I shall discuss first, cannot. Liberal political theory, as well as liberal jurisprudence, has been found by many feminist theorists to be "male" or "masculine." As Linda McClain writes, responding to such critiques, but first summarizing their essence: A central theme of the critique has been that the law embodies a masculine perspective in emphasizing autonomy and the individual over interdependency and the community. Liberalism has been viewed as inextricably masculine in its model of separate, atomistic, competing individuals establishing a legal system to pursue their own interests and to protect them from others' interference with their rights to do so. Hence, it is said that liberal, masculine jurisprudence has exalted rights over responsibilities, separateness over connection, and the individual over the community.22

Many of these charges have been specifically aimed at Rawls's A Theory of Justice. Related criticisms often brought against liberal theories, again including Rawls's, are that they abstract overly from actual human beings and thus obscure, or render irrelevant, differences among them, and that they stress the importance of rationality, undermining the importance of the emotions both as definitive of human nature and as crucial for moral theorizing.23 Those who characterize liberalism in these various ways are often termed "relational feminists" or "cultural feminists." They often advocate alternative conceptions of the person and of social relations, more in line with those found in communitarian theories. When women's experience of the world is taken into account, they argue, connection, responsibility, community, caring and attention to differences—neglected or deprecated in liberalism—will take their proper place in moral and political theorizing. Relational feminism has its early roots in two influential theories of the early 1980s: Carol Gilligan's famed theory about women's "different voice," and communitarianism, especially that of Michael Sandel. Gilligan's feminist theory of moral development began as a critique of that of the Kohlberg school, which owes much to Rawls's work and prioritizes justice, rights, and principles. Gilligan, pointing out that Kohlberg's empirical work had dealt exclusively with male 22. McClain, supra note 21, at 1173-74 (citations omitted). 23. Examples of the former type of critique are Seyla Benhabib, The Generalized and the Concrete Other and Iris Marion Young, Impartiality and the Civic Public, both in Feminism as Critique (Seyla Benhabib & Drucilla Cornell eds., 1987). Examples of the latter are found in Jaggar, supra note 8, passim.

352

John Rawls

1544

FORDHAM LA W REVIEW

[Vol. 72

subjects, contrasts with it a morality built on care, contextuality, and concern for others, which she finds more characteristic of the female subjects she studies.24 SandePs critique of Rawls depends in part on an idealization of the traditional family—an institution Sandel claims is beyond the need for justice; but it is also based on a critique of the Rawlsian "self/' which Sandel interprets as disembodied and unembedded in any social or cultural reality.25 Two prominent feminist political philosophers, Iris Marion Young and Seyla Benhabib, developed critiques of Rawls's liberal theory of justice that resonate with themes central to Gilligan's and SandePs work. Young argues against the ideals of impartiality and universality in moral reasoning, claiming that they attempt to eliminate difference and "otherness," and create a false dichotomy between reason and feeling—overvaluing the former and denigrating or devaluing the latter. She thus considers Rawls's theory to be as rationalist, monological, and abstracted from particularity as Kant's.26 Benhabib argues, similarly, that in universalistic moral theories such as Rawls's and Kohlberg's, "ignoring the standpoint of the concrete other leads to epistemic incoherence."27 In Rawls's original position, she claims, "the other, as different from the self, disappears.... Differences are not denied: they become irrelevant."28 With only a "generalized other," Benhabib concludes, "what we are left with is an empty mask that is everyone and no one."29 Such relational feminist critiques, however, often caricature or misunderstand liberal thought (as does much of the communitarian critique). They take as representative of liberalism ideas that do play some part in the liberal tradition; but they tend to misinterpret these ideas or their roles in the theories. For example, they treat the characteristics of parties in the original position as if the original position were less a heuristic device than a theory of the self or of human nature. They interpret "autonomy," a concept central to most liberal theory, as if it meant atomism rather than self-determination. And they tend to draw a sharp distinction between an "ethic of justice" and an "ethic of care," finding the latter absent from liberal thought.30 I have argued that Rawls's liberal theory of justice can plausibly be 24. Carol Gilligan, In a Different Voice (1982). On Gilligan's theory, see Susan Moller Okin, Thinking Like a Woman, in Theoretical Perspectives on Sexual Difference 145-59 (Deborah Rhode éd., 1990). 25. Michael Sandel, Liberalism and the Limits of Justice (1982). 26. Iris Marion Young, Toward a Critical Theory of Justice, 1 Soc. Theory & Prac. 279 (1981); Young, supra note 23. 27. Benhabib, supra note 23, at 89. 28. Id. 29. Id. 30. McClain, supra note 21, at 1174-75, 1203 & n.157. Again, the terms are Gilligan's.

John Rawls

2004]

JUSTICE & GENDER

353

1545

read as combining both of these approaches. Thus read, it should confound such critics.31 The actors (or actor) at the crux of the theory—the parties in the "original position" who arrive at the principles of justice—are indeed said by Rawls to be "rational" and "mutually disinterested." Thus they appear, superficially, to fit the relational feminist version of atomistic "liberal man." The charge that difference and otherness are absent appears initially to be borne out by Rawls's suggestion that those in the original position are identical.32 However, those who stress only the characteristics of rationality, mutual disinterest and sameness fail to note that the actors Rawls so describes have another, crucially important, characteristic, which gives his "veil of ignorance" its name. They do not know any of their personal attributes or attitudes, or the social position they will find themselves in, in the society whose principles of justice they are choosing. Indeed, Rawls foresaw misperceptions of his theory that focus on the Rawlsian "self" as rational and disinterested, and warned that the theory is likely to be regarded as based on egoism if the mutual disinterest assumption about those in the original position is highlighted in isolation from the other specifications. Seen instead in their totality, he says, the combination of qualities ascribed to them "forces each person in the original position to take the good of others into account."33 As I have argued, those behind the Rawlsian veil of ignorance can be perceived as thinking only for themselves, only because they do not know which self they will turn out to be and, therefore, must consider the interests of all possible selves equally In the absence of knowledge about their own particular characteristics... they must think from the position of everybody, in the sense of each

in turn.34

For actual persons (rather than participants in a heuristic device), who of course know who they are, to do this requires not rational egoism but, rather, strong empathy and a readiness to listen to the very different points of view of others. Rather than exemplifying an overly abstracted and rationalistic "ethic of justice," Rawls's liberal theory of justice is built to a large extent on an ethic of care; it is at least as concerned with certain important emotions as with rationality, and with human differences as with similarities. As McClain 31. See Okin, supra note 21. 32. See Rawls, A Theory of Justice, supra note 6, § 24, at 139 ("[S]ince the differences among the parties are unknown to them, and everyone is equally rational and similarly situated, each is convinced by the same arguments. Therefore, we can view the choice in the original position from the standpoint of one person selected at random."). 33. Id. § 25, at 148. 34. Okin, supra note 21, at 244.

354

John Rawls

1546

FORDHAM LAW REVIEW

[Vol. 72

concludes, referring to Rawls's explicit rejection of indifference as a value for the just society, "Rawls's theory envisions not a disembodied individual, but a society where recognition of interdependency leads to mutual respect—and where mutual respect entails not the right to be let alone but an almost tender-hearted solicitude for one another."35 For real persons to have the strong sense of justice Rawls expects of the citizens in his well-ordered society requires that those citizens have well-developed capacities for empathy, care, and concern for others—certainly not the self-interest and instrumental rationality that are but a part of his prescription for the party or parties in the original position.36 However, the achievement of such qualities places a special burden on families, which Rawls's theory requires to be sources from which moral development originates. And, paradoxically, the families envisioned in both A Theory of Justice and Political Liberalism seem incapable of meeting this challenge. McClain, Martha Nussbaum, I and others who have defended Rawls's version of liberalism against anti-liberal feminist critiques have done so in part because we think that liberalism properly understood, with its radical refusal to accept hierarchy and its focus on the freedom and equality of individuals, is crucial to feminism. While we agree that many liberal theorists, past and present, have failed to fulfill their commitment to the individual, in part by relegating women to "natural" subordination within families, in part by simply assuming—but then paying no attention to—all the work women do to produce and reproduce the supposedly "independent" male self, we argue that consistent and fully developed liberalism, quite radically revised so as to include women, has great potential for feminism. Indeed, as Nussbaum deftly puts it, most of liberalism's failing is not that it is too individualist, but that, in its views of the family it is not individualist enough.37 Whatever the faults of past liberals, surely no contemporary liberal theory should be able to get away with ignoring more than half of human adults. Nussbaum relates how women in various parts of the world, where liberalism does not in general prevail, are appealing to liberal concepts such as autonomy, rights, and self-respect to make their case for women's equality, freedom, and even bodily integrity.38 And McClain points out that, even in cases in which the ethic of care might be expected to yield better 35. McClain, supra note 21, at 1209. 36. To be sure, Rawls's construction of the original position is designed so as to eliminate from the formulation of the principles of justice biases that might result from particular attachments to others, as well as from particular facts about the self. But surely impartiality of this kind is a reasonable requirement of any theory of justice. 37. Nussbaum, supra note 21, at 65. 38. Id. at 55-56.

John Rawls

2004]

JUSTICE & GENDER

355

1547

arguments for feminist-friendly positions, liberalism can both be more persuasive and yield better results.39 Many feminists, as early as the seventeenth century, have appealed to the arguments of liberal or egalitarian theorists who themselves ignored or excluded women in order to make claims on behalf of women.40 Similarly, in the late twentieth century, some feminist theorists have built feminist arguments on prominent liberal theories, once the latter have been subjected to the types of revision necessitated by feminist critique. This is especially true of Rawls's A Theory of Justice.

III. CONSTRUCTIVE FEMINIST CRITIQUES OF RAWLS'S A THEORY OF JUSTICE

In 1974, very soon after A Theory of Justice was published, philosopher Jane English pointed out an aspect of its theory-building unquestioned by other early commentators—the passing assumption that those in the original position are "heads of families" rather than all persons, or all adults. In a paper mainly concerned with the issue of justice between generations, she argued that "[b]y making the parties in the original position heads of families rather than individuals, Rawls makes the family opaque to claims of justice."41 English noted briefly that, with respect to some issues, such as the division of labor between the sexes, this could well be a problem: "For example, suppose that, due to efficiency, all families gain significantly if the natural childbearers are universally appointed as child rearers."42 Such a policy would be good from the point of view of Rawls's "heads of families," because it is good for families considered as single entities. However, it may be far from just or fair from the point of view of some individuals within those families, particularly if caring for children negatively affected their other personal, economic or political opportunities. For one thing, it would clearly violate Rawls's fairly demanding standard of fair equality of opportunity. Surprisingly, however, given that the traditional division of labor was one of the main issues being raised and challenged by the burgeoning feminist movement at the time, English's important critical insight about Rawls's instantly famous theory lay fallow for 39. McClain shows that liberal arguments are stronger than relational feminist or ethic of care arguments in the case of the duty to rescue a stranger, while the application of arguments grounded in the ethic of care to such issues as abortion are fraught with problems. See McClain, supra note 21, at 1228-62. 40. For example, Mary Astell built on Locke's liberalism, Mary Wollstonecraft built in part on inconsistencies in Rousseau's egalitarianism, and John Stuart Mill, in The Subjection of Women (1869), extended some liberal and egalitarian utilitarian arguments to women. 41. Jane English, Justice Between Generations, 31 Phil. Stud. 91,95 (1977). 42. Id.

356

John Rawls

1548

FORD H AM LAW REVIEW

[Vol. 72

nearly a decade, and a more developed feminist critique of Rawls was slower to emerge.43 In 1983, however, Australian political theorist Deborah Kearns zeroed in on one of the problems—one that has preoccupied feminist scholars of Rawls since. The theory, despite its initial inclusion of the family as part of the basic structure of society to which the principles of justice were to apply, went on to treat women, the family, and love as outside of the realm of justice. But this very exclusion, in its failure to remark upon a gendered division of labor in the family that involved injustice, threatened to undermine the development of a sense of justice in the children who were to be the just society's future citizens. Kearns concluded that it was because Rawls recognized that "the family is not necessarily a just institution [that]... he effectively removes [it] from the scope of the principles of justice."44 In 1987, focusing largely on Rawls's A Theory of Justice, I took up the argument that, although the social institution of gender seemed ripe for analysis in contemporary theories of justice, it was unaccountably absent from them. When raised, moreover, it caused significant problems.45 While the relation between the other initial or acquired characteristics of persons and how they fared in social distributions was at the core of theories of justice, somehow discussion of the social differentiation between the sexes and the allocation of burdens and benefits along the lines of gender was missing. Most significantly, as Kearns pointed out, the sexual division of labor in the household, which was taken for granted by most families and affects every other aspect of our lives, was scarcely touched upon in any of the leading contemporary theories of justice. I have argued that, in A Theory of Justice, Rawls not only almost completely ignores gender, but he almost completely ignores women. He omits sex from the list of personal characteristics that are veiled from those in the original position and, as English pointed out, (ominously) specifies that those who reason in the original position are the "heads of families."46 Though he indicated several years later that sex was to be regarded as such a contingent characteristic, he gave no indication of recognizing that this required substantial revision of major aspects of his theory.47 However, with this change, 43. Jane English herself died in a tragic accident within several years. 44. Deborah Kearns, A Theory of Justice—and Love; Rawls on the Family, 18 Pol: Australasian Pol. Stud. Ass'n J. 36 (1983). 45. See Susan Moller Okin, Justice and Gender, 16 Phil. & Pub. Aff. 42 (1987). I argue that Walzer's Spheres of Justice (1974) pays more attention to the division of labor than does Rawls's A Theory of Justice, but that the feminist aspects of Walzer's argument, taken seriously, undermine his book's defense of cultural relativism. 46. See, e.g., Rawls, A Theory of Justice, supra note 6, § 22, at 128. 47. In Fairness to Goodness, 84 Phil. Rev. 537 (1975), he writes: "That we have one conception of the good rather than another is not relevant from a moral standpoint. In acquiring it we are influenced by the same sort of contingencies that

John Rawls

2004]

JUSTICE & GENDER

357

1549

considerable revision of the theory seemed called for. First, the "heads of families" assumption would have to be retracted or, at least, explained and justified. Second, if those in the original position, and in the two subsequent stages of Rawls's four-stage "unveiling"—the constitutional convention and the legislative process—were not to know (among all the specifics about themselves) what sex they would be once the veil was completely raised, they would surely be deeply concerned about many aspects of social gendering and sex discrimination as well as matters affected by biological sex differences.48 Third, families would certainly have to be taken seriously as part of the basic structure of society. In A Theory of Justice, Rawls's arguments alternate references to "men" and "his" with more neutral references to "persons" and "parties." He shows no sign of recognition either that this terminology is problematic or that to assume that the individuals who reason about justice are the heads of families takes issues of justice within families right off the agenda. At the time he was writing, this kind of omission was only just beginning to be noticed in various academic disciplines. In the early 1970s, philosophers and political theorists still routinely used male pronouns and referred to people as "men." Economists assumed—as Rawls appeared to—that the family could be treated as if it were a single individual, and as if only one of its members (its male head) counted as "a rational economic actor;" and sociologists cleaved to the structural functionalist view that rigid sex roles were necessitated by efficiency in the fulfillment of the essential functions that the family performs for society.49 Even in this intellectual climate, however, especially given that Rawls himself listed the family as one of the basic social institutions to which the theory of justice was to apply, the "heads of families" assumption sounded immediate warning bells for any feminist reader. But when Rawls relates, in Part II of A Theory of Justice, how the two principles of justice are applied to the basic institutions of society, as I said lead us to rule out a knowledge of our sex and class." Id. 48. As for the fourth stage of the sequence, Rawls thought that, even ideally, judges and administrators applying the laws could know all the facts about themselves without prejudice to justice. Strongly doubting this, in Justice, Gender, and the Family, I evoked a cartoon in which three robed, male justices look down at their very pregnant bellies, as one says: "Perhaps we'd better re-consider that decision." Susan Moller Okin, Justice, Gender, and the Family 102 (1989). Recent work by Nancy Crowe bears out my and the cartoonist's concern. Crowe shows, for example, in work based on twenty years of evidence, that the probability that a Republican male judge on a three-judge federal court panel will vote in favor of a female plaintiff in a sex discrimination case is 28%, whereas the probability a Republican female will vote this way is 53%. Nancy Crowe, Women in Black: The Effect of Judges' Sex on Legal Decisions (Oct. 20,2003) (paper given at the Radcliffe Institute for Advanced Study), 49. See any philosophical journal of the late sixties, for this usage. The prevailing assumption in economics was argued in detail by Gary S. Becker, A Treatise on the Family (1981). Talcott Parsons & Robert F. Bales's Family, Socialization and Interaction Process (1955) exemplifies the prevailing view in sociology.

358

John Rawls

1550

FORD H AM LA W REVIEW

[Vol. 72

earlier, he discusses at length their application to all of the previously named institutions of the basic structure—except the family. The questions of whether and how the family, in its traditional or any other form, is a just social institution and how or whether it translates the principles of justice into social practice are never raised. In one significant passage, Rawls does address the issue of formal, legal sex discrimination. He states that inequalities in basic rights that are based on "fixed natural characteristics," such as "[distinctions based on sex ... race and culture" are acceptable only if "justified by the difference principle," and that such "inequalities are seldom, if ever, to the advantage of the less favored."50 However, this reasoning is applied only to basic rights, such as freedom of speech, religion, and suffrage. In addition, it does not address the very real problem that formal equality between members of different races or sexes often bears little relation to actual equality. In the case of sex equality, this especially is so in the context of a long history of oppression and exclusion of women—a history in which all major social institutions have been constructed on the assumption that someone other than their participants is taking care of the young, the old, the sick, and even the participants' own daily needs.51 The only place where A Theory of Justice says much about the family is in the section about moral development. Here, as I mentioned earlier, while acknowledging the significance of families for moral development, Rawls simply assumes that they are just. Given their prominent role in his theory of moral development, it is particularly striking that he would simply assume the justice of families, a presumption he affords no other part of the basic structure. Moreover, he writes very much as though he is thinking in terms of traditional, gendered family structure and roles.52 Thus I surmised, departing from Kearns's interpretation, that Rawls omitted discussion of what justice would require in families not because he recognized that families were not just, but rather because he perceived no injustice in traditional, gendered family arrangements. I agreed, however, with Kearns's conclusion that Rawls's theory of justice 50. Rawls, A Theory of Justice, supra note 6, § 16, at 99. It is interesting that the difference principle is introduced here in qualifying a basic right, whereas in general the equality of basic rights takes precedence over the difference principle. It is also interesting that, in effect, Rawls's suggestion is in line with the legislation that Congress felt obliged to pass when a federal appeals court bizarrely interpreted the Pregnancy Discrimination Act as meaning that no state legislation could favor pregnancy, by giving pregnant women any entitlements not shared by everyone. 51. One of the best explanations of this point is made by Catharine MacKinnon, Difference and Dominance, in Feminism Unmodified (1987). 52. See Rawls, A Theory of Justice, supra note 6, § 71, at 467-68. He says, for example, that the content of the ideals that one learns first within the family "is given by the various conceptions of a good wife and husband, a good friend and citizen, and so on." Id. at 468; see also Okin, supra note 48, at 96 (providing a greater elaboration of this point).

John Rawls

2004]

JUSTICE & GENDER

359

1551

contains serious problems, not only for any reader who was concerned with justice for women, but also for the coherence of his arguments about moral development.53 Families are obviously, as Rawls recognized at the beginning of the book, an important part of the basic structure of society, strongly influencing the life chances of their members—especially women, when the families are genderstructured. And though Rawls relied heavily, in his account of the development of a sense of justice that he considered essential to the stability of a just society, on families that he assumed were just, he provided no reasons at all for this assumption, nor any discussion of changes in social structures or policies that might better justify it. These problems with Rawls's theory of justice seemed in part due to the weight of the Western tradition of political thought, in which questions having to do with women and the division of labor between the sexes are so often ignored or, at best, attributed to "nature."54 Even when issues of sex and gender preoccupied the great philosophers, as they did Rousseau, they tended until the 1970s to be virtually ignored by subsequent interpreters and teachers of the tradition.55 Within the tradition (if that word is interpreted liberally), while rare feminists such as Mary Astell, Olympe de Gouges, Mary Wollstonecraft, John Stuart Mill, and George Bernard Shaw had challenged it, the norm was to assume, or to argue, that the division of labor between the sexes and the sexual hierarchy it was presumed to justify were natural, not political.56 Thus, only men belonged in the realm of culture and political life, where they—unburdened by day-today preoccupations and freed from the partial, emotional ties associated with intimacy—could debate serious concerns such as justice amongst themselves, capable due to their superior rationality of following just principles. As Carole Pateman and Fran Olsen both brilliantly pointed out, the dichotomy between the "public" and the 53. One of the oddest rejections I ever received from a journal was in response to my submission of Justice and Gender to Ethics, in 1986. The rejection letter was accompanied by several readers* reports, the essence of one of which was that, although the paper succeeded in showing that Rawls's theory of justice excluded women from its subject-matter, the author had failed to say anything significant about "the theory itself." I wondered if this reason for rejection would have been given in response to a critique of a theory of social justice for not including men. 54. Two of the clearest examples of such thinking are in Book I of Aristotle's Politics and Book V of Rousseau's Emile. 55. The works mentioned in notes 1-6 supra had challenged this, but too recently to have had much effect on political theorists writing in the 1970s. The main exceptions to this interpretation of the canon were a few Straussians who, on the one hand, seized upon the arguments defending the subordination of women and, on the other, contended that those philosophers who argued for the equality of the sexes either did not mean what they wrote (for example Plato), or were not "great" (for example John Stuart Mill). Among the first general contemporary political theories to reflect on and critique aspects of gender were Michael Walzer's Spheres of Justice (1983) and Philip Green's Retrieving Democracy: In Search of Civic Equality (1985). 56. See also Okin, supra note 48, at 25-40.

360

John Rawls

1552

FORDHAM LAW REVIEW

[Vol. 72

"private" spheres of life had long served to separate out and to shape the sphere in which men, perceived as autonomous, independent, and often self-interested individuals, had rights and made contracts, from the sphere in which women took care of the daily needs of the supposedly autonomous men and of children—the sphere in which bonds were assumed to be naturally hierarchical and motivations altruistic. Olsen's most brilliant point is that "[bjecause the state is deeply implicated in the formation and functioning of families," the notion that it can choose whether to intervene in the formation and functioning of families is nonsense; the only real question is how it intervenes.57 Thus contemporary feminists were arguing powerfully, in the footsteps of predecessors such as Wollstonecraft and John Stuart Mill, that neither families nor the divisions of labor within them can be conceived of as natural, in the sense of unaffected by coercive laws.58 While Rawls did not expressly spell out the public/private distinction in A Theory of Justice, his assumption that the parties who developed the principles of justice were "heads of households" relied heavily on it, as did his stated reason for the assumption—that heads of families, even if otherwise assumed to be motivated by rational selfinterest, could be expected, altruistically, to take into account the interests of other members of their own families. Moreover, since Rawls's just society requires its actual members (as opposed to the artificial construct that was the party in the original position) to be 57. Frances E. Olsen, The Myth of State Intervention in the Family, 18 U. Mich. J. L. Reform 835, 837 (1985). 58. Carole Pateman, Feminist Critiques of the Public/Private Distinction, in The Disorder of Women (1987); Olsen, supra note 57. Reading both these articles, but especially Olsen's, was for me a "scales falling from my eyes" experience. See Okin, supra note 48, at ch. 6. Regardless of such work, Gerald Cohen, at the very moment of acknowledging the influence of feminist theory, falls prey to the notion that families are free from legal coercion. In Where the Action Is: The Site of Distributive Justice, he considers the idea that Rawls's basic structure might consist only of the major social institutions that are "coercive (in the legal sense)." G.A. Cohen, Where the Action Is: The Site of Distributive Justice, 26 Phil. & Pub. Aff. 3,18-20 (1997). His "signal example" of a major institution that "can depend far less on law than on convention, usage, and expectation" is the family. Id. But of course the law frequently extends its coercive arm to enforce familial responsibilities, and the threat that it may do so influences a great deal of apparently "uncoerced" family behavior. Millions of non-custodial parents are required by law to pay child support, have their wages garnished, and are sometimes imprisoned for failure to pay; millions of married and (especially) divorced individuals are required by law to support each other financially in various ways; at least hundreds of thousands of parents are charged with child abuse or with child neglect. In October 2003, a single mother in Connecticut, who was employed 60 hours a week, was convicted of felonious neglect of her home and family after her 12-year old son—bullied mercilessly at school—killed himself. Marc San tora, After Son's Suicide, Mother Is Convicted of Unsafe Home, N.Y. Times, Oct. 7, 2003, at Bl. Cohen's apparent unawareness of the extent of use of coercive laws that lurk behind the "usages and expectations" of families is an astonishing example of both the power and the longevity of the notion that families are havens of privacy and voluntary action.

John Rawls

2004]

JUSTICE & GENDER

361

1553

neither self-interested nor wholly altruistic, but instead to have well developed "senses of justice," this reliance on the public/private dichotomy was a significant problem. It seemed to undermine rather drastically his theory of moral development, and therefore to reveal a rather serious internal problem for the theory. Whereas Kearns had asked: "[H]ow can a sense of justice develop in an unjust institution?"59 I questioned strongly whether Rawls's bald, unargued statement that the family was a just institution sufficed as a basis for his reliance on families at the heart of his theory of moral development.60 In Justice and Gender and Justice, Gender, and the Family, I attempted to extend the argument beyond critique towards developing a feminist reading of Rawls's theory of justice. I argue that, in spite of the problems noted, the feminist potential of Rawls's method of thinking about justice and his conclusions is considerable. Once the veil of ignorance is understood as hiding from its participants their sex as well as their other particular characteristics and circumstances and the unjustified "heads of households" assumption is relinquished, the original position is a powerful concept for challenging the gender structure. Once we dispense with the traditional liberal assumptions about public versus private, political versus non-political spheres of life, we can use Rawls's theory as a tool for feminist criticism with which to think about how to achieve justice between the sexes, both within the family and in society at large. As a result, the theory of justice would be improved. It would reflect the presence of women's, as well as men's, points of view in the original position. And it would no longer be weakened by its reliance, for the first stages of moral development, on a basic social institution that was either naively assumed to be just or to which the principles of justice were not considered to be applicable. In Justice, Gender, and the Family, I venture to spell out some of the arguments that might be made, and the conclusions that might follow, from using this revised version of Rawls's theory. I ask how a theory of justice that applied Rawls's two principles to families as well as to the other institutions of the basic structure of society would change gendered assumptions and practices. In particular, how might representatives who did not know whether they were to be men or women in the society they were planning or legislating for employ law, education and other public policy to change the division of labor in families so as to promote equality of fair opportunity and the equal worth of political liberty for women? How might laws and other policy instruments apply to gender relations in families and workplaces, especially, the requirement that permissible inequalities 59. Kearns, supra note 44, at 41. 60. Okin, supra note 48, at 97 & n.28.

362

John Rawls

1554

FORD HAM LA W REVIEW

[Vol. 72

benefit the least advantaged? In the late-twentieth century era of significant transition in and disagreement about relations between the sexes, it seemed to me that two types of public policies, resulting mostly from application of the principle of equality of fair opportunity and the difference principle, could alleviate the injustices of gender. The first would encourage men and women to share the public and the domestic, the paid and the unpaid roles and responsibilities of family life, equally, so that both might participate on an equal footing in their various roles—at work, in civil society, and in politics—in the non-domestic spheres of life. Such policies, I argued, would need to include subsidized early child care and after-school care for children, flexible working hours for parents and other caregivers, genderneutral parental and other family-related leave, and firmly enforced anti-discrimination law in all necessary areas. The second type of policies would protect those (perhaps mostly, but not exclusively, women) who choose to undertake the bulk of unpaid family work, from the vulnerabilities they now incur. Such policies would include equal division of the earner's paycheck between the earning and the non-earning spouse, and family law ensuring that, in the event of divorce, both post-divorce households would have the same standard of living.61 In his recently published response to feminist critics, Rawls agreed with most, if not all, of these specific suggestions. I will merely introduce this recently published response now, and summarize and discuss it in the last two sections of the paper. Meanwhile, I will turn to his previously published but post-A Theory of Justice work, especially Political Liberalism, in which Rawls's response to feminist criticism of A Theory of Justice was mostly not to respond. Moreover, his responses to other critics as well as certain other changes he made in his theory made his problems with feminism worse. To these I shall now turn.

IV. How POLITICAL LIBERALISM COMPOUNDED RAWLS'S PROBLEMS WITH FEMINISM For some years before Rawls's Political Liberalism was published in 1993, elaborating further his theory of justice and responding to many of its critics, a manuscript authored by him entitled Women and the Family had been in circulation among his students and their students. In it, Rawls took up some of the feminist critique discussed above. However, to the surprise of some readers, this response was not included in the new book, though revised versions of parts of it have recently appeared in print, included in The Idea of Public Reason

61. For a fuller account of my argument, see id. at 170-86.

John Rawls

2004]

JUSTICE & GENDER

363

1555

Revisited62 and in Justice as Fairness: A Restatement.6* Thus, in Political Liberalism, Rawls addresses feminist critics only very briefly and obliquely, in his Introduction. He mentions that a number of "major matters" had been omitted from A Theory of Justice, including "the justice of and in the family"—though he reminds us that he did "assume that in some form the family is just."64 He remarks that since, in contemporary society, "among our most basic problems are those of race, ethnicity, and gender," his chosen focus on toleration, in the new book, might seem dated.65 Then, noting that A Theory of Justice had been criticized for not dealing with problems of gender and the family, he briefly notes that he thinks his conceptions and principles can be addressed to them. He does not attempt to do so, though, beyond stating tellingly that, "the equality of the Declaration of Independence which Lincoln invoked to condemn slavery can be invoked to condemn the inequality and oppression of women."66 Not only did Political Liberalism not respond, beyond this intriguing and elusive passage, to feminist critics. Certain aspects of the newly elaborated theory—notably Rawls's determination to differentiate his political theory of justice from a more comprehensive theory—made his problems with feminism worse, despite the promising suggestion made at its outset. Not surprisingly, Political Liberalism was very soon critiqued by feminist theorists. The new framing of the theory, as a discussion about how a just society of citizens with diverse conceptions of the good can be stable over time, we argued, revealed even more clearly the internal problems caused for the theory by Rawls's neglect of justice within families. Central aspects of the revised theory rendered even more problematic than before Rawls's passing assumption that families are just institutions. For he now made it abundantly clear that in his just, pluralist society, "reasonable" conceptions of the good included religions that both preached and practiced highly sexist modes of life. "[Ejxcept for certain kinds of fundamentalism," he supposes, "all the main historical religions... may be seen as reasonable comprehensive doctrines."67 62. John Rawls, The Idea of Public Reason Revisited (1997), reprinted in Collected Papers 573,595 (Samuel Freeman éd., 1999). 63. John Rawls, Justice as Fairness: A Restatement 162-68 (Erin Kelly éd., 2001). 64. John Rawls, Political Liberalism xxxi (1993). 65. Id. at xxx. 66. Id. at xxxi. See my two alternative interpretations of this passage in Susan Moller Okin, Political Liberalism, Justice, and Gender, 105 Ethics 23,39-43 (1994). 67. Rawls, Political Liberalism, supra note 64, at 170. The reason he specifies for this is that these religions admit of "an account of free faith." Id. But it is quite unclear what this has to do with these religions being "reasonable" by Rawls's own explanation of the term. See id. at 48-54. Here, "reasonable persons . . . desire for its own sake a social world in which they, as free and equal, can cooperate with others on terms all can accept" and "insist that reciprocity should hold within that world." Id. at 50. One might think that "reasonable religions" should be held to the same standard, and also that "reasonable persons," not knowing their own sex behind the full or the

364

John Rawls

1556

FORD HAM LA W REVIEW

[Vol. 72

Yet the basic texts of Judaism, Christianity, and Islam are rife with sexism: the Torah/Bible reverses the reality of reproduction so that the first woman is made from a man, tells a history of the Jews from which women are virtually absent, and advises wives to obey their husbands. The Qur'an explicitly advocates beating women "from whom you fear disobedience," and suggests that some barrier be placed between the sexes, which has been interpreted in a myriad of ways, including heavy compulsory veiling and the "seclusion," or socially enforced imprisonment, of women in their homes, in much of the history of Muslim peoples.68 Reformed versions of all three dominant Western religions have, of course, acknowledged the equality of the sexes in numerous ways, including admitting women to their various ministries. However, the more orthodox (but by no means necessarily fundamentalist) versions of all three—including Orthodox Judaism, Catholicism and some Orthodox and Protestant branches of Christianity, and many variants of Islam—still discriminate against women and reinforce their subordination within religious practices, and within and outside the family, in numerous significant ways.69 In addition to indicating Rawls's acceptance of such beliefs and associated practices as "reasonable" comprehensive conceptions of the good, acceptable even in the just, well-ordered society of ideal theory, Political Liberalism exacerbated the problem feminists had pointed out regarding whether or not the family is part of the basic structure of society. While Rawls more or less reiterates the position taken in A Theory of Justice, stating that "the nature of the family" belongs to the basic structure, he also, seemingly paradoxically, states that the political is "distinct from... the personal and the familial, which are affectional... in ways the political is not."70 Philosopher John Exdell and I argued (in papers published at about the same time) that these aspects of Political Liberalism were not simply weaknesses and inconsistencies in its argument. They also endangered the very stability of the just society that the book aimed to ensure.71 It is clear that the even greater distinction between the political and the non-political that Rawls develops in Political Liberalism leads to partial veils of ignorance, might insist on this. 68. The quotation is from Surah 4:34: "[A]s for those women from whom you fear disobedience: admonish them, put them aside in their beds, and beat them.'* 69. Of course, women often (and men sometimes) resist. For one very obvious example where it seems both resist in huge numbers, witness the paucity of very large Catholic families in the United States. The "rhythm method" of contraception is not at all reliable, yet it is the only method of birth control that is supposedly practiced by Catholics who take the holy sacraments. 70. Rawls, Political Liberalism, supra note 64, at 137. 71. John Exdell, Feminism, Fundamentalism, and Liberal Legitimacy, 24 Can. J. Phil. 441 (1994); see also Okin, supra note 66.

John Rawls

2004]

JUSTICE & GENDER

365

1557

serious weaknesses or inconsistencies in the book's argument. As I have mentioned and will discuss further below, the unresolved problems about whether the family is part of the basic structure and whether and how the principles of justice are to apply to it are exacerbated. In addition, the political/non-political distinction, and the toleration of a wide range of religious and other cultural practices that it involves, lead Rawls to allow kinds of discrimination to be practiced against women that he disallows, if practiced against persons differentiated by racial or ethnic group. Rawls states clearly in Political Liberalism that the priority of right limits permissible comprehensive conceptions of the good and permissible ways of life, in a just society. Moreover, he specifies how it limits them. While "justice cannot draw the limit too narrowly," just institutions must "perinit[] but also sustain[] ways of life fully worthy of citizens' devoted allegiance."72 Admissible ideas of the good must "respect the limits of, and serve a role within, the political conception of justice."73 Specifically, "conception^] of the good requiring the repression or degradation of certain persons on, say, racial, or ethnic... grounds" are "in direct conflict with the principles of justice" and hence must be "discourage[d]... or even exclude[d]... altogether" in the just society.74 Unfortunately, Rawls does not spell out what it means "to exclude" or "not to permit" a conception of the good. He suggests that the discouraging of discriminatory conceptions of the good should be done "in ways consistent with liberty of conscience and freedom of speech," but he refrains from saying whether, or how, the exclusion of any that are impermissible can be accomplished in such a manner.75 This is somewhat puzzling. Could Rawls mean that groups within the just society are free to advocate or argue for racial or ethnic discrimination, though not to practice it? Whether or not this is his intention, what about sex discrimination? As I have mentioned, in seeming tension with the principles that appear to underlie his anti-discrimination stance, Rawls accepts as reasonable and therefore permissible all the main historical religions except for certain forms of fundamentalism. But many variants of these religions not only preach but also practice many forms of sex discrimination. As Exdell and I both asked, why are conceptions of the good that require the repression or degradation of women not susceptible to the same judgment and subject to the same treatment as those that are similarly racist or ethnically discriminatory? There are two likely reasons for Rawls's not applying his anti-caste principle to sex as well as to race and ethnicity.76 One is that applying 72. 73. 74. 75. 76.

Rawls, Political Liberalism, supra note 64, at 174. Id. at 176. Id. at 195-96. Id. at 195. In a recent article» I show that several other prominent contemporary political

366

John Rawls

1558

FORD HAM LA W REVIEW

[Vol. 72

it would rule out many prominent religions as beyond the pale of reasonableness. But what kind of a reason is thafl Why not instead argue that unless and until they reform themselves, as many variants of religion have already done, so as to accommodate sex equality, sexist religions too are to be discouraged or even excluded altogether from the just society, being no less inconsistent than racist or ethnically discriminatory ones are with the principles of justice? The other likely reason is that, in keeping with the long line of thinking about the public/private distinction that I briefly looked at above, he may have thought that the "affectional" nature of family life meant that women are less likely to be subjected to the degrees of degradation or repression that persons experience because of their race or ethnicity. But this way of thinking, too, has been exposed in the last few decades as largely based on myth. When beliefs about gender, marriage and family encourage the formation of families in which women, especially mothers, can easily be made vulnerable, oppressed and subordinated, many women do become vulnerable, oppressed and subordinated. This problem, in turn, intensifies the stability problem. How? Surely the stability of Rawls's just society must still rely on the moral development of its members—specifically, their development of a sense of justice—as well as on gaining the acceptance of its adult members? But whereas he, in Political Liberalism, seeks to render his just society more stable because more tolerant of a diversity of values, by reinforcing the distinction between the political and the nonpolitical and applying the principles of justice only to the former sphere of life, by doing so he greatly reduces the potential for the development of a sense of justice in families. As Exdell and I have both argued, the reduced scope of justice specified by Rawls for the sake of stability of the just society also decreased social stability by tolerating—accepting as reasonable—unjust, sexist family forms for the sake of religious pluralism. How, we asked, might children acquire the sense of justice that was needed for the society's stability within families that are not themselves regulated by the principles of justice, but might well instead become places where oppression is able to flourish.77 In the account of moral development offered in Political Liberalism^ Rawls completely omits the role of families as major influences during early childhood. But if the omission means that philosophers, including Joseph Raz, William Galston, and Chandran Kukathas, share Rawls's propensity to conceive of sex-based oppression as if it differs significantly in its seriousness from race-based oppression. See Susan Moller Okin, "Mistresses of Their Own Destiny": Group Rights, Gender, and Realistic Rights of Exit, 112 Ethics 205 (2002). On the anti-caste principle in U.S. constitutional law and its implications when taken seriously, see Cass Sunstein, The Partial Constitution 338-45, passim (1993). For discussions of gender as caste that are strongly influenced by Sunstein's ideas, see Okin, supra note 48, at 65-68, and Okin, supra note 66, at 39-43. 77. See Exdell, supra note 71; see also Okin, supra note 66.

John Rawls

2004]

JUSTICE & GENDER

367

1559

these passages of A Theory of Justice are meant to stand unchanged, then they are rendered even more problematic by the problems that I have outlined here.78 In a recent essay entitled Rawls and Feminism, in The Cambridge Companion to Rawls, Martha Nussbaum takes up such feminist criticisms of Political Liberalism, agreeing with some of them but taking issue with my claim that Rawls is inconsistent in restricting, discouraging, or even excluding comprehensive conceptions of the good that repress or degrade persons on ethnic or racist grounds, but at the same time allowing as acceptable and reasonable all of the major, frequently sexist, religions, excepting certain forms of fundamentalism. She claims that I am wrong about the fate of unreasonable comprehensive doctrines in a society based on Rawls's political liberalism. For Rawls, she claims, free speech protects the reasonable and the unreasonable alike, since "no political, religious, or philosophical speech can be censored, in Rawls's view, absent the existence of a grave constitutional crisis" in which free political institutions may be failing to preserve themselves. She also takes issue with my failure to "distinguish between doctrines holding that women should have unequal rights of citizenship and doctrines holding that they are metaphysically unequal or dissimilar in some other respect."79 Thus, she defends Rawls's distinction between the political and the non-political, interpreting it to mean, in this context, that as long as women's equal political rights are not questioned, women's equality in other respects is fair game for question or attack since comprehensive doctrines that subordinate women must be tolerated. It is unclear whence Nussbaum derives the extent and quasiabsolute status of freedom of speech she attributes to Rawls. As evidence for her statement about Rawls disallowing censorship of speech, she cites "p.343, etc." of Political Liberalism. The only passage on page 343 that seems pertinent is Rawls's statement that "[w]ithin our tradition there has been a consensus that the discussion of general political, religious, and philosophical doctrines can never be censored."80 On the following page, too, he mentions the "agreement that all general discussion of doctrine ... is fully protected."81 These statements, however, are statements about our political tradition, rather than direct statements by Rawls about his own beliefs. More importantly, they need to be understood in context; they are part of a 78. Rawls indicated to me in a conversation at Stanford in May 1993, that he intended his theory of early moral development in Part III oí A Theory of Justice to be read as unchanged by the arguments of Political Liberalism. 79. Martha C. Nussbaum, Rawls and Feminism, in The Cambridge Companion to Rawls 488,509 (Samuel Freeman éd., 2003). 80. Rawls, Political Liberalism, supra note 64, at 343. 81. /¿at344.

368

John Rawls

1560

FORD H AM LA W REVIEW

[Vol 72

lengthy section in which Rawls argues that it is inconsistent with liberal democracy to consider as crimes such speech as "seditious libel" or "defamation of the government," or to censor doctrines calling for the overthrow of the government by force—except in cases of such imminent danger to democratic political institutions as he considers never to have existed in the United States, even during the Civil War. Rawls does not explicitly endorse the "consensus" of "our tradition" he reports. But even if he did, to endorse freedom of speech for "the discussion of general political, religious, and philosophical doctrines," in the context of arguing against the criminalization of seditious libel against the government is hardly the same thing as to claim that "no political, religious, or philosophical speech can be censored/' as Nussbaum reports.82 So what did Rawls himself think about freedom of speech, beyond ruling out the criminalization of "seditious libel" and like offenses against governments? What did he think about freedom of speech that defames or degrades some categories of persons, for example, which is considerably more pertinent than the case of seditious libel to the issue Nussbaum is discussing? Considering both A Theory of Justice and Political Liberalism, it is very clear that Rawls considers free speech to be one among a number of basic liberties, which, "it is important to recognize, must be assessed as a whole, as one system."83 Taken together, they constitute "a fully adequate scheme" of liberties.84 While Rawls argues that liberty "can be limited only for the sake of liberty itself," he also says that "[c]learly when the liberties are left unrestricted they collide with one another."85 He states clearly numerous times that when any of these liberties conflicts with any other, all of them need to be considered and they need to be balanced, limited and adjusted against one another.86 Thus none of them has absolute protection. In a sentence familiar to all students of Rawls's work, he states: "Taking the two principles [of justice] together, the basic structure is to be arranged to maximize the worth to the least advantaged of the complete scheme of equal liberty shared by all."91 In the same section of Political Liberalism on which Nussbaum seeks to base her near-absolutist interpretation of his views about free speech, Rawls makes clear his own nuanced position on the subject. He subjects to ridicule—an unusual tactic for Rawls—Justice HolmesJs case of someone's falsely shouting "Fire!" in a crowded theater. This example is utterly "trivial," Rawls says, since it works 82. Nussbaum, supra note 79, at 509 (emphasis added). 83. Rawls, A Theory of Justice, supra note 6, § 32, at 203. 84. Rawls, Political Liberalism, supra note 64, at 356. 85. Id. 86. See, e.g., Rawls, A Theory of Justice, supra note 6, §§ 32, 39, at 203, 205, 244; Rawls, Political Liberalism, supra note 64, at 356. 87. Rawls, A Theory of Justice, supra note 6, § 32, at 205.

John Rawls

2004]

JUSTICE & GENDER

369

1561

only against "the view, defended by no one, that all speech of whatever kind is protected."88 This is very close to the "entrenched" position Nussbaum attributes to him. She states that he would rule proposals favoring serfdom or slavery off the political agenda in the sense that they would not be able to be voted on, but that he takes the position that "anyone who likes may make such proposals unconstrained."89 But Rawls makes it fairly clear that he holds no such view. Instead, he reminds us that the basic liberties are not absolute. They can be restricted in their content, though "only if this is necessary to prevent a greater and more significant loss ... to these liberties."90 The standard to which delegates and legislators must appeal in such cases, he argues, is "what best advances the rational interest of the representative equal citizen in a fully adequate scheme of basic liberties."91 He also argues in this same section that when any of the basic liberties comes into conflict with the fair value of political liberty, "here too the basic liberties must be considered together and weighed up against the threat their use might pose to this liberty, such that no one liberty can expect unconditional protection."92 This brings us to the second point Nussbaum seeks to make in defense of Rawls: that feminists such as myself have neglected the distinction political liberals make between challenges to women's political equality and challenges to women's equality in other respects—such as their metaphysical equality. The problem with this defense is twofold. First, Rawls's own prioritization of the fair value of political liberty along with the other basic liberties means that he must be concerned to protect more than women's formal political rights and legal equality. Indeed, anything, including any influential doctrine that contributes to women being represented politically far less than their proportion in the general population, becomes a matter of grave concern, since it endangers the basic liberties of women. Second, the whole point of my and Exdell's critiques of the political/non-political distinction is that it is a false dichotomy, I argued, along the lines of Marx in On the Jewish Question, that there is no way of separating out and isolating women's political equality

88. Rawls, Political Liberalism, supra note 64, at 345. 89. Nussbaum, Rawls and Feminism, supra note 79, at 509. 90. Rawls, Political Liberalism, supra note 64, at 356. 91. Id. 92. Id. Rawls's main account of the fair value or equal worth of political liberty, which runs parallel to his concept of fair equality of opportunity, is in Section 36 of A Theory of Justice, titled "Political Justice and the Constitution." It focuses on class inequality, but there is no reason why the standard it establishes should not apply to political inequality along race, ethnic or gender lines. Rawls says: "[I]deally, those similarly endowed and motivated should have roughly the same chance of attaining positions of political authority irrespective of their economic and social class." Rawls, A Theory of Justice, supra note 6, § 36, at 225.

370

John Rawls

1562

FORDHAM LAW REVIEW

[Vol. 72

from all the other aspects in which women are unequal in a sexist society.93 Thus, for example, were Rawls confronted with a situation in which a not insignificant religious group was promulgating the view that women have the souls of pigs (the kind of "metaphysical doctrine" about women's inequality that Nussbaum appears to want to protect, in the interest of religious freedom, and to seek Rawls's mantle of approval for so protecting), and in which such speech appeared to be distinctly affecting the fair value of women's political liberty, by preventing them from being taken seriously as political candidates, it would seem that Rawls might well argue for the suppression of such religious speech in favor of women's political liberty. Since he argues that "even in a well-ordered society under favorable circumstances liberty of thought and conscience is subject to reasonable regulations,"94 one might surmise that he would find such regulation reasonable in the rather unfavorable circumstances spelled out in my example. Nussbaum apparently accepts the highly dubious distinction between the political and the non-political that so much is made of in Political Liberalism. But of course "metaphysical" attacks on the full humanity of women are not distinct from "political" attacks on their equal citizenship.95 Both Rawls and Nussbaum are deluded on this score. But Rawls has the possible recourse of calling on his own earlier and clearer ideas about the need to balance and limit the various basic liberties from A Theory of Justice. Nussbaum has jumped to the conclusion that, as long as women's formal political and legal equality were not being directly attacked, for Rawls, freedom of speech, thought or conscience would trump the fair value of political liberty for women. But this seems highly dubious. The fair value of political liberty and the priority Rawls has clearly given it seem to restrict the permissibility of sexist comprehensive doctrines far more than he might himself have realized. This may have far-reaching implications for political liberalism, which cannot be pursued further here.

93. Okin, supra note 66. 94. RawJs, A Theory of Justice, supra note 6, § 39, at 244. 95. Examples from all over the world seem to bear this out almost daily. Two examples reported on the same day in late 2003: When E.U. President Berlusconi jokes, at a meeting including European parliamentarians of both sexes, that since the discussion of the matter on the agenda is not progressing, those present should instead discuss "women and football," is the political equality of the women present (and those they represent) preserved, or is it undermined? When male leaders in an Afghani loya jirga tell the female members of that council that they should moderate their demands because Islam considers a woman to be worth only half of a man, is their political equality threatened, or not?

JohnRawls

2004]

JUSTICE & GENDER

371

1563

V. RAWLS'S RECENTLY PUBLISHED RESPONSE TO FEMINIST

CRITIQUES

Finally in 1997, in an essay entitled The Idea of Public Reason Revisited, Rawls first published his response to feminist critics, which is also included in Justice as Fairness: A Restatement.96 In it, he calls it "a misconception" to think that "the principles of justice do not apply to the family and hence ... do not secure equal justice for women and their children."97 He writes: "If the so-called private sphere is alleged to be a space exempt from justice, then there is no such thing."98 He also states plainly that "the family is part of the basic structure, since one of its main roles is to be the basis of the orderly production and reproduction of society and its culture from one generation to the next" and he acknowledges (for the first time) that "reproductive labor is socially necessary labor."99 It seems as if Rawls was finally to attend to the justice or injustice of the gender structure, at least within families. But then he seems to take most of this back again. He says that the principles of justice "are to apply directly to [the basic] structure, but are not to apply directly to the internal life of the many associations within it, the family among them."100 What could this mean? If we substitute, here (in single inverted commas), Rawls's initial definition of the basic structure, it means that the principles of justice "are to apply directly to 'the way in which the major social institutions distribute fundamental rights and duties and determine the division of advantages from social cooperation', but are not to apply directly to the internal life of the many associations within [the basic structure], the family amongst them."101 This is more than a little puzzling. It is rendered even more puzzling when taken in conjunction with Rawls's statement that it is a misconception to think that the principles of justice do not apply to the family.102 First, Rawls has never previously suggested that the principles of justice are not, generally, to apply directly to the social institutions that make up the basic structure of society, as well as to the ways in which they distribute rights and duties and determine the division of advantages. But if the principles of justice are not to apply directly to institutions such as courts and constitutions, legislatures and laws, and even to duly regulated markets and systems of ownership, then one might well

96. Rawls, The Idea of Public Reason Revisited, supra note 62, at 595-601 ("On the Family as Part of the Basic Structure"). A similar section is included in Rawls, Justice as Fairness: A Restatement, supra note 63, at 162-69. 97. Rawls, The Idea of Public Reason Revisited, supra note 62, at 596. 98. Rawls, Justice as Fairness: A Restatement, supra note 63, at 161. 99. Rawls, The Idea of Public Reason Revisited, supra note 62, at 595-96. 100. Id. at 596. 101. Rawls, A Theory of Justice, supra note 6, § 2, at 6. 102. Rawls, The Idea of Public Reason Revisited, supra note 62, at 596.

372

John Rawls

1564

FORDHAM LA W REVIEW

[Vol. 72

ask what is the point of having them at all? We need to proceed with Rawls's argument, to make out what he could mean. Next, he repeats the argument, first made in Political Liberalism, that the family is similar in this respect to other associations such as churches and universities, business firms or labor unions. Like them, he says, it cannot violate the basic rights and freedoms of the equal citizens who are its members. But, like them, he implies, it is not itself subject to the principles of justice. Just as we do not require churches to be democratically governed, so we should not require families to be internally governed by the principles of justice; the family is "not peculiar in this respect." I think there is a lot of confusion here, which I have yet to completely sort out: Surely it is not as if families and universities should not be internally regulated by the principles of justice, but legislatures and courts should. Indeed it may make less sense to have these latter institutions directly ruled by such principles than to have families ruled by them. I can see no good reason, for example, to apply the difference principle to the property-holdings of the members of a legislative body. Nor, surely should the principle of equal "political liberty" or decision-making rights apply to all participants in a jury trial—where justice is better served if witnesses answer questions, judges sentence, and only jurors vote. On the other hand, neither can I see any good reason why decisions in families should not be made equally by their members (giving the same special treatment to children of different degrees of maturity as they are given regarding their public voting rights), or why the difference principle should not be applied within families.103 Perhaps, ironically, we may conclude that families are the quintessential place for justice, rather than a place where it is not needed or is impossible to apply, as has been more commonly thought. Not that families should be just just, as I have argued in response to Sandel; rather that justice is their primary, basic virtue, and that "nobler virtues" such as generosity and the willingness to sacrifice one's interest for those of others are unreliable, and even dangerous to some family members, unless built upon it.104 This is not the direction that Rawls takes, however. As in Political Liberalism, so in the last published works he draws a distinction between "the point of view of people as citizens and their point of 103. Robert Nozick cites what he regards as the inapplicability of the difference principle even within families as Aprima facie reason for its rejection as a principle for the wider social sphere. He argues in Anarchy, State, and Utopia that it is reasonable for parents to devote more educational resources to their most talented than to their least talented children. Though I cannot do it here, I think justice demands the reverse, or at least demands equal expenditures on both, except in such cases (for example, severe deprivation or living in the context of a "winner take all" occupational structure) in which the whole family's future survival is likely to depend on the earnings of the most talented of its members. 104. Okin, supra note 48, at 26-28.

JohnRawls

2004]

JUSTICE & GENDER

373

1565

view as members of families and of other associations."105 He claims that, from the latter point of view, wanting "a free and flourishing internal life appropriate to the association in question... [w]e wouldn't want political principles of justice—including principles of distributive justice—to apply directly to the internal life of the family.... Here those principles are out of place."106 What Rawls does insist on repeatedly, however, is that the principles of political justice impose external restraints on families, as on other associations. Thus, "[s]ince wives are equally citizens with their husbands, they have all the same basic rights, liberties, and opportunities as their husbands; and this, together with the correct application of the other principles of justice, suffices to secure their equality and independence."107 Later, Rawls spells out further what he refers to here as the correct application of the other principles of justice. He says: "A long and historic injustice to women is that they have borne, and continue to bear, an unjust share of the task of raising, nurturing, and caring for their children,"108 which can render them particularly vulnerable in the event of divorce. He notes that, if a basic cause of women's inequality as citizens is their greater share in nurturing and caring for children, "steps need to be taken either to equalize their share, or to compensate for it."109 In addition, while freedom of religion requires that "some traditional gendered division of labor within families" be allowed, it must be "fully voluntary and ... not result from or lead to injustice."110 While Rawls says that it is not for political philosophy to decide the specifics of this, he refers with seeming approval to the split paycheck idea, and to the equal sharing of assets in the event of divorce. He calls it "intolerably unjust" that a husband may leave his family taking his earning power with him. Given what Rawls had written earlier about the appropriateness of invoking "the same equality" invoked by Lincoln in his condemnation of slavery to the inequality and oppression of women, I have come to think of these suggestions of his as his version of "forty acres and a mule" for women.

VI. A BRIEF RESPONSE It is gratifying to feminist critics of Rawls that he eventually responded to our concerns about his theory of justice in some detail. Moreover, in doing so, he affirmed some of the suggestions we had come up with as to how to make families more just social institutions. However, several aspects of his response are still either puzzling or 105. 106. 107. 108. 109. 110.

Rawls, The Idea of Public Reason Revisited, supra note 62, at 597. Id, at 597-98. Id. at 597. Id. at 598. Mat600. Id. at 599.

374

John Rawls

1566

FORDHAM LAW REVIEW

[Vol. 72

unsatisfactory. First, he restates the idea that families are like other private associations, without addressing the trenchant critiques of this position that have been made. Families are not voluntary associations readily entered and exited. Families of origin are not entered at all voluntarily. And though families one forms are usually entered voluntarily, they are by no means always exited voluntarily; moreover, even when they are, typically such exit does not come without considerable struggle or loss. Divorce, even under the most favorable circumstances, could hardly be compared with graduating from college, or choosing to teach at a different university. Moreover families, unlike other private associations such as universities and churches belong, for excellent reasons, among the basic social institutions to which Rawls has said from the outset that his principles of justice are to apply. Indeed, there is no way of reconciling Rawls's latter-day position that families are like other, more voluntary associations with his own definition of the basic social institutions— those that affect their members' opportunities from birth, those that "have deep and long-term social effects and in fundamental ways shape citizens' character and aims, the kinds of persons they are and aspire to be."111 How could social institutions so defined not include families, which would thereby be differentiated clearly from the other, more voluntary, associations? Again, if the unknowns in the original position include one's sex, as Rawls has indicated since 1975, how could Rawls's "parties" not be seriously concerned with issues of justice internal to families? Surely they would want to ensure that, in the just society, public policy and institutions strongly fostered the equal division of unpaid labor within the home, and that women and men actually pursued this equal division so as not to disadvantage women both at home and in most other spheres of life? Would one not want to see justice within families, albeit in many respects probably not directly legally enforced, given very high priority in the well-ordered society? Moreover, Rawls has not responded to the important concern that has been raised repeatedly since Kearns's paper in 1983 about the internal inconsistency of a theory of justice that depends heavily for the moral development and socialization of its citizens on an institution that is not itself internally regulated by the principles of justice.112 As I have asked frequently both here and elsewhere, how could the social institution in which, as Rawls acknowledges, small children's first inklings of justice emerge in the context of their love and trust for those who care for them, forming the basis for moral 111. Rawls, Political Liberalism, supra note 64, at 68. 112. See the response to this concern in S.A. Lloyd, Situating a Feminist Criticism of John Rawls's Political Liberalism, 28 Lpy. L.A. L. Rev. 1319 (1995). Rawls cites this, among other feminist responses to his work, but does not indicate whether he concurs with its argument.

John Rawls

2004]

JUSTICE & GENDER

375

1567

development, not itself be based on internal justice?113 While he refers several times in "The Idea of Public Reason Revisited" to the role of the family in nurturing and developing citizens with a sense of justice, his notion that families are not special (or "peculiar" as he puts it) but, rather, similar to other social associations such as universities and trade unions seems completely to neglect the crucial function of families in promoting a sense of justice in the young.114 Finally, a point that warrants further development than I can devote to it here: Rawls simply states, without argument, that "[w]e wouldn't want" families to be regulated internally by principles of distributive justice. This view has also been voiced by some other very influential philosophers and political theorists, including Michael Sandel and Allan Bloom.115 But having spent much time thinking about justice and its applicability or lack of applicability to families, some of us are not sure that this is at all evident. We still ask: "Why not?"

113. My most sustained discussion of this is in Political Liberalism, Justice, and Gender, supra note 67, at 32-39, where I ask how the "political virtues" Rawls argues for in Political Liberalism can be acquired in unjust gender-structured families, bringing in some recent findings about moral development and about actual families of various types. 114. Gerald Cohen's Where the Action Is: The Site of Distributive Justice, supra note 58, draws an interesting and important analogy between issues of justice within the family and other voluntary aspects of egalitarianism in a just society, but is not concerned with the special role of the family in moral education which, as feminists have argued, makes its internal justice particularly compelling. See also Andrew William's response, Incentives, Inequality, and Publicity, 27 PhiL & Pub. Aff. 225 (1998). 115. Sandel's claim that families are "better than just" is a pivotal piece of the argument he makes against Rawls's claim that justice is the primary moral virtue, which he presents as a case against liberal accounts of justice in general. See Sandel, supra note 25, at 30-35. Bloom's claims that families are unjust, but naturally and necessarily so, because of the respective "natures" of women and men (especially men), are heavily based on Rousseau's anti-feminism, and are of considerable current influence because of the weight of neo-conservatism within the George W. Bush administration. See Allan Bloom, The Closing of the American Mind (1987); see also Jean Jacques Rousseau, Introduction to Emile (Allan Bloom trans., 1979). For a critique of both claims, see Okin, supra note 48, at ch. 2.

This page intentionally left blank

Part III A Political Liberalism

This page intentionally left blank

[14] A MORE DEMOCRATIC LIBERALISM Joshua Cohen * POLITICAL LIBERALISM. ByJohnRawls. New York: Columbia University Press. 1993. Pp. xxxiv, 401. $29.95. [W]hat do we do when we find the truth? . . . When men learned the Earth was round, did they allow their geographers to continue to teach that it was flat? . . . If you would see the monuments of a society that has come to consider the truths that Jesus Christ taught us as one among an indefinite variety of moral codes by which to live, look around you. Amen, and Happy Easter.1

I.

SOCIAL UNITY AND MORAL PLURALISM

When Peter Laslett published his first collection of essays on Philosophy, Politics and Society in 1956, he reported that "[f]or the moment, anyway, political philosophy is dead."2 As the book reviews in this annual Survey indicate, things have changed. Political philosophy is back, and its revival owes much to John Rawls's^ Theory of Justice (Theory).3 Published more than twenty years ago, Theory remains the starting point for contemporary work on justice. This fact by itself is sufficient to make the appearance of Rawls's second book, Political Liberalism (Liberalism),* an important event. But the intellectual importance of Liberalism reaches well beyond * Professor of Philosophy and Political Science, Massachusetts Institute of Technology. B.A. 1973, M.A. 1973, Yale; Ph.D. (Philosophy) 1979, Harvard. — Ed. I am grateful to Frank Michelman and Michael Sandel for discussions of drafts of their reviews of Political Liberalism. to John Rawls for countless discussions on the themes of his book and for helpful comments on an earlier draft of this review, and to participants in political theory seminars at Yale University and Wesleyan University for their suggestions. 1. Pat Buchanan, Tolerance and Truth at Easter, ARIZ. REPUBLIC, Apr. 3, 1994, at £5. 2. Peter Laslett, Introduction to PHILOSOPHY, POLITICS AND SOCIETY at vu, vu (Peter Laslett ed. 1956). 3. JOHN RAWLS, A THEORY OF JUSTICE (1971). John Rawls is James Bryant Conant University Professor Emeritus, Harvard University. 4. Political Liberalism is presented as a series of eight lectures, which descend in complex ways from earlier lectures and papers. The first five lectures are revisions of previously published articles, but the revisions are substantial even when — as with lectures 4 and 5 — the titles have not been changed. Lecture 6 is a significantly modified version of material presented in public lectures but never before published. Lectures 7 and 8 were published previously and are reproduced without modification. Pp. xii-xiv. Apart from lectures 7 and 8, then, it is a mistake to identify the views advanced in Political Liberalism with positions taken in earlier versions of the lectures.

380

John Rawls

1504

Michigan Law Review

[Vol. 92:1503

the biography of its author and the recent history of political philosophy. Rawls's book is a deep and original examination of a fundamental problem of modern politics. Modern societies are marked by manifest ethical, religious, and philosophical disagreements among citizens.5 Moreover, the disagreements are of a special kind. Although citizens commonly regard the moral, religious, and philosophical views of others as false, they need not regard others as unreasonable for endorsing those views.6 Because human reason appears not to converge on a single moral outlook, we seem to face "a plurality of reasonable yet incompatible comprehensive doctrines" (p. xvi). What are the implications of these doctrinal conflicts — this "fact of reasonable pluralism" (p. xvii) — for our understanding of the requirements of justice and the possibility of a just society?7 A. Democratic Toleration Liberalism addresses this question against the background of the account of justice as fairness advanced in Theory. In Theory, Rawls proposed an ideal of a well-ordered, democratic society featuring consensus on a conception of justice rooted in the value of fair cooperation among citizens as free and equal persons.8 But Theory, Rawls now thinks, did not take the fact of reasonable pluralism seriously enough. The presentation suggested that justice as fairness depends on a comprehensive liberal philosophy of life — that only people who endorse a view of our nature and of the human good that emphasizes independence, choice, and self-mastery have good reason to endorse justice as fairness.9 Liberalism asks, then, whether justice as fairness can be freed from this dependence. Can views that disagree about moral fundamentals — some of which reject a comprehensive liberal philosophy of life — nevertheless agree on a political conception of justice rooted in "values of equal political and civil liberty; fair equality of opportunity; . . . economic reciprocity; [and] the social bases of mutual respect between 5. \ say "manifest" because I do not suppose that any society is morally or religiously homogeneous, however much its institutions may suppress the expression of diSerences by limiting expressive liberty, establishing compulsory forms of worship, or narrowly circumscribing associative liberty. 6. I will say more about the distinction between reasonable and unreasonable later. See infra section IV.C. It will suffice here to note the familiar logical distinction between is true and is reasonable: inconsistent views cannot both be true, but they can both be reasonable. 7. Versions of this question are posed at pp. xviii, xxv, 4, and 133. Rawls does not suppose that the fact of reasonable pluralism taken on its own leads us to a particular conception of justice. The problem of Liberalism is generated instead by an apparent tension between the fact of reasonable pluralism and the ideal of a well-ordered society featuring consensus on a conception of justice that articulates such fundamental political values as fairness, equality, and liberty. 8. RAWLS, supra note 3, at 12-13. 9. On the idea of a comprehensive moral conception, see p. 13. For the concern that Theory endorses such a conception, see pp. xvi-xvii; see also infra section II.B.

John Rawls

May 1994J

A More Democratic Liberalism

381

1505

citizens" (p. 139)? Or does the fact of reasonable pluralism imply that we ought to give up on the idea of a consensus of justice, that democratic politics can never be more than a combination of individual calculation, group bargaining, and assertions of discrete collective identities — when democracy works well — and deceit, manipulation, and naked force — when democracy works badly? In a world full of cruelty, depravity, and grief, we ought not to dismiss the virtues of a politics of group bargaining within a framework of rules that win general compliance — "a mere modus vivendi" (p. 145). Still, Liberalism defends the possibility of doing better: of achieving a consensus on political justice under conditions of fundamental moral, religious, and philosophical disagreement. The key to that possibility is that political values — for example, the value of fair cooperation among citizens on a footing of mutual respect — are extremely important values and can be acknowledged as such by conflicting moral conceptions, by views that disagree with one another about ultimate values and about the best way to live.10 To be sure, those views will explain the importance of political values in very different terms:11 for example, as rooted in autonomy,12 or self-realization,13 or human happiness properly understood,14 or the appropriate response to life's challenges,15 or the value of individuality,16 or the equality of human beings as God's creatures.17 These competing explanations of the political values will in turn manifest themselves in conflicting views about individual conduct and personal virtue. Still, an affirmation of the importance of political values is not the unique property of a particular moral outlook. For this reason, the different moral views that flourish in a society governed by a conception of justice rooted in the ideal of fair cooperation on a footing of 10. See pp. 139, 155-57, 168-69, 208-09, 217-19. 11. Some views may treat fairness itself as a fundamental value and not as an implication of some deeper moral value. See the "third view" at p. 145. 12. See IMMANUEL KANT, METAPHYSICAL ELEMENTS OF JUSTICE (John Ladd trans., Bobbs-Merrill Co. 1965) (1797); JOSEPH RAZ, THE MORALITY OF FREEDOM (1986). 13. See T.H. GREEN, LECTURES ON THE PRINCIPLES OF POLITICAL OBLIGATION (Ann Arbor Paperbacks 1967) (1895); T.H. GREEN, PROLEGOMENA TO ETHICS (A.C. Bradley éd., 5th ed. 1906). 14. See JOHN S. MILL, UTILITARIANISM (1861), reprinted in UTILITARIANISM, LIBERTY, REPRESENTATIVE GOVERNMENT 1 (H.B. Acton éd., E.P. Dutton & Co. 1972); JOHN S. MILL, ON LIBERTY (1859) [hereinafter MILL, ON LIBERTY], reprinted in UTILITARIANISM, LIBERTY, REPRESENTATIVE GOVERNMENT, supra, at 65. 15. See Ronald Dworkin, Foundations of Liberal Equality, in 11 THE TANNER LECTURES ON HUMAN VALUES 1 (Grethe B. Peterson éd., 1990) [hereinafter TANNER LECTURES]. 16. See STUART HAMPSHIRE, INNOCENCE AND EXPERIENCE 114, 117-18, 124-36 (1989). Hampshire also explains the value of fair political process in terms of its role in preventing such great evils as "murder and the destruction of life, imprisonment, enslavement, starvation, poverty, physical pain and torture, homelessness, friendlessness." Id. at 90. 17. See JOHN LOCKE, The Second Treatise of Government, in Two TREATISES OF GOVERNMENT § 5 (Peter Laslett éd., Cambridge Univ. Press rev. ed. 1960) (1690).

382

John Rawls

1506

Michigan Law Review

[Vol. 92:1503

mutual respect may each have good and sufficient reason to support that conception as the correct account of justice and not simply as a suitable accommodation to conditions of disagreement. Citizens who endorse different moral axioms may still arrive at the same theorems about political justice, and some people may simply endorse a view of justice without resting that endorsement on a more comprehensive moral theory.18 In such a society, we have an "overlapping consensus" on a "political conception of justice."19 Citizens achieve social unity because they all accept that conception and so agree to conduct the fundamentals of political argument on the shared ground that the conception makes available and to set aside for political purposes their deep, ultimate, and persistent disagreements about what we are like, what the world is like, and how best to face its demands. This account of the combination of unity and pluralism rests on a new interpretation of the ideal of toleration — call it "democratic toleration"20 — paralleling the new interpretation of the social contract advanced in Theory. In Theory» Rawls proposed "to generalize and carry to a higher order of abstraction the traditional theory of the social contract."21 The combination of social unity and moral pluralism captured in Liberalism's idea of overlapping consensus generalizes and carries to a higher order of abstraction the conventional idea of toleration. Conventionally understood, toleration is a substantive political principle condemning the imposition of an authoritative form of religious worship or, in a more expansive version, an authoritative form of personal morality.22 Aiming to provide a conception of toleration better suited to "the historical and social circumstances of a democratic society" (p. 154), Rawls's political liberalism deepens the idea of toleration and "applies the principle of toleration to philosophy itself" (pp. 10, 154). That is, in addition to accepting the substantive requirement of toleration, Liberalism presents toleration as a condition on political justification, at least when the question concerns "constitutional essentials" and "basic questions of justice."23 Given the plurality of incompatible yet reasonable views held by equal citizens in a democratic society, the ideal of fair cooperation recommends that we free the vocabulary and premises of political justification from dependence on 18. This possibility plays an important role in Liberalism. See pp. 155-56. 19. On overlapping consensus» see pp. 132-72; on the idea of a political conception of justice, see pp. 11-15, 174-75. 20. Rawls rejects perfectionism in the name of "democracy in judging each other's aims." RAWLS, supra note 3, at 442. 21. Id. at viii; see also p. xv. 22. On the central role of religious toleration in understanding the value of toleration, see SUSAN MENDUS, TOLERATION AND THE LIMITS OF LIBERALISM 6-8 (1989). 23. See pp. 137, 227-30.

John Rawls

May 1994]

A More Democratic Liberalism

383

1507

any one view. Put otherwise, Rawls suggests that when we understand political power as "the power of free and equal citizens as a collective body" (p. 136) and take account of the fact of reasonable pluralism, we will want to be sure that political argument on fundamentals proceeds on grounds that are acceptable to citizens generally, not in the terms provided by a particular philosophical or religious tradition (pp. 136-68, 216-18). To be sure, it may be impossible to gain support for a conception of justice from all views. But perhaps support for the conception, and a willingness to conduct public political argument in its terms, will come from the "reasonable comprehensive doctrines*' (p. 59) held by reasonable citizens: the views held by people who are concerned to cooperate on terms that others accept and who recognize that reason itself does not select a single comprehensive view.24 The central line of thought in Liberalism, then, is that we can achieve the good of consensus on justice without comprehensive moral agreement;25 the absence of comprehensive agreement does not reduce politics to calculations of individual advantage, interest-group bargaining, or the self-affirmation of discrete collective identities. Instead, because political values are highly important values and are recognized as such within a wide range of moral conceptions, consensus on a conception of justice is possible under conditions of reasonable pluralism and must accommodate those conditions if it is to suit the equal citizens of a democratic society. B. Reconciliation Without Metaphysics Rawls's project in Liberalism bears certain important similarities to Hegel's in his Philosophy of Right»26 and it will be instructive to sketch both the commonalities and the différences between their projects. In his political theory, Hegel aimed to reformulate a classical ideal of political society, which supposed that citizens share an understanding of justice and the human good,27 in light of the post-Reformation idea of unbridgeable differences among citizens on fundamentals. How is it possible, Hegel asked, to achieve the good of shared commitments in the face of apparently ultimate differences in interest and outlook that are so much the focus of the energies of modern civil society? How, in Hegel's terms, can we give stable expression to both the uni24. On reasonable comprehensive doctrines, see pp. 58-66. 25. But cf. ALISDAIR MAC!NTYRE, AFTER VIRTUE 227-37 (1981). 26. G.W.F. HEGEL, ELEMENTS OF THE PHILOSOPHY OF RIGHT (Allen Wood éd., H.B. Nisbet trans., Cambridge Univ. Press 1991) (1821). 27. For a statement of this aspect of the classical ideal of political society, see ARISTOTLE, POLITICS 1280b23-1281a3 (T.A. Sinclair trans., Penguin Books, rev. ed. 1981).

384

John Rawls

1508

Michigan Law Review

[Vol. 92:1503

versal and particular aspects of our nature?28 Hegel and Rawls share broadly similar questions and both endorse the hopeful possibility of reconciling apparently competing demands of unity and difference. Their proposals about how to achieve that reconciliation differ profoundly, however, both in substance and in the insight about the reconciliation they expect philosophy to provide.29 Hegel located his answer within a generally antidualistic, logicometaphysical theory. His philosophical system revealed our nature as free beings,30 showed how our differences are less fundamental than we are prephilosophically inclined to think, linked the expression of our free nature to the institutions of a state whose aim is the realization of the good — understood as the expression of our nature31 — and showed how that expression and those institutions were the natural upshot of historical evolution.32 According to Rawls, evaluative theories are matters of reasonable disagreement, and for that reason we ought not to build a conception of political justice around the view of the good advanced within any one such theory. Moreover, the reconciliation of social unity and moral pluralism cannot proceed on the terrain of metaphysics. Because there are ultimate, reasonable disagreements about metaphysical doctrines, a general philosophical argument against dualisms, for example, cannot provide part of the case for overcoming the specific tension between pluralism and social unity.33 Political philosophy, if it seeks to operate on the shared ground available to equal citizens in a pluralistic public, cannot rest on a metaphysical theory of our true nature, nor can it provide any assurances, grounded in such a theory, about the ultimate expression of that nature in history.34 Its aims must be less ambitious, focused on clarifying how social unity is possible under pluralistic conditions. Such clarification will not yield the 28. On the role of the modern state in achieving this stable expression, see HEGEL, supra note 26, § 260. 29. For discussion of the idea of reconciliation in Hegel's political philosophy, see MICHAEL O. HARDIMON, HEGEL'S SOCIAL PHILOSOPHY: THE PROJECT OF RECONCILIATION (1994). 30. HEGEL, supra note 26, § 4. 31. "[The good is] realized freedom, the absolute and ultimate end of the world." Id. § 129. 32. G.W.F. HEGEL, THE PHILOSOPHY OF HISTORY (J. Sibree trans., Dover Publications 1956) (1899). 33. In an earlier version of some of the material published in Liberalism, Rawls indicated that "one of Hegel's aims was to overcome the many dualisms which he thought disfigured Kant's transcendental idealism," that Dewey "shared this emphasis throughout his work," and that "there are a number of affinities between justice as fairness and Dewey's moral theory which are explained by the common aim of overcoming the dualisms in Kant's doctrine." John Rawls, Kantian Constructivism in Moral Theory, 77 J. PHIL. 515, 516 (1980). My point is not to deny this common aim. I want only to emphasize that the presentation of justice as fairness as a political conception implies that its resolution of the apparent tension between social unity and moral pluralism cannot draw on a general antidualistic metaphysical view. 34. Later I will discuss some reasons for operating on shared grounds. See infra notes 11416 and accompanying text.

John Rawls

May 1994]

A More Democratic Liberalism

385

1509

assurances of unity associated with a historical theodicy;35 at best it will lead to an understanding of why the hope for reconciliation is not unreasonable. Once we understand how the stable combination of shared principles and conflicting faiths that defines an overlapping consensus is possible, then we can see — Rawls thinks — that it is reasonable to adhere to the ideal; the conditions of its possibility are not so demanding as to condemn it. In this way Liberalism offers a "defense of reasonable faith in the possibility of a just constitutional regime/'36 It argues for the reasonableness of that faith by revealing the commitments it requires as minimally demanding, emphasizing in particular that people within different moral and religious traditions can reasonably endorse those commitments. If Liberalism is right, then it is possible to combine fundamental moral pluralism — to take seriously one sort of difference — with consensus on a conception of justice suited to the equal citizens of a democratic society. But while philosophy can provide that service in a democratic society — that defense of reasonable faith — it can deliver no greater assurance of the rationality of what is actual. C. Consensus? Really? Those are the aims of Liberalism. They are likely to meet with skeptical response. The idea of combining disagreement on fundamentals with consensus on political principles suited to free and equal citizens may strike us as nice work if you can get it. In particular, it is natural to suspect that the demands of consensus are less minimal and the faith in its possibility correspondingly less reasonable than Rawls claims. There are at least four reasons for skepticism about the ideal of consensus, and I will discuss them in detail in Part IV of this review.37 As background for that discussion, I want first to explore more fully Rawls's new view, tracing the route from Theory to Liberalism — in Part II — and outlining the strategy of Liberalism itself— in Part III. Before getting to the route and the strategy, however, I want to enter a caveat. 35. History, Hegel says, is the "true theodicy." HEGEL, supra note 32, at 457. 36. P. 172; see also p. 101. The idea of philosophy as a defense of reasonable faith derives from Kant. See pp. 100-01, 172. On the background of Kant's idea of reasonable faith in Rousseau, see DIETER HENRICH, AESTHETIC JUDGMENT AND THE MORAL IMAGE OF THE WORLD: STUDIES IN KANT 10-28 (1992). 37. See infra Part IV. One basis of skepticism that I will not explore below endorses the possibility of combining political consensus and moral pluralism, but only if the political consensus is confined to questions of just procedure. I explore and criticize this view in Joshua Cohen, Pluralism and Proceduralism, 69 CHL-KENT L. REV. (forthcoming Summer 1994),

386

John Rawls

1510

Michigan Law Review

[Vol. 92:1503

D. A Different Book Liberalism is a very abstract book, in ways that contrast sharply with Theory. Much of the excitement of Theory derived from its claim to argue from relatively weak, abstractly stated assumptions to powerful, controversial, substantive claims about justice. Here was an egalitarian and liberal account of justice, concerned both with the protection of basic civil and political liberties and with assuring a distribution of resources that would enable people to make fair use of those liberties, and supported by premises arguably much less controversial than its conclusions.38 Moreover, Theory's many polemical edges helped to sharpen its central claims. Utilitarianism had dominated the field of systematic moral and political philosophy, and Theory aimed to displace it.39 In addition, Theory proposed an alternative to the ideal of natural liberty — sharp libertarian limits on the legitimate actions of the state — and to a liberal pluralism that would ensure fair process but would leave questions of substantive justice to bargaining in political and economic markets.40 To be sure, Rawls devoted stretches of Theory to the nature of justification, rationality, and goodness. But the discussion of these matters was never longer — or shorter — than necessary, and one felt that the discussion was never very far from first-order issues of justice. By contrast, Rawls's presentation of political liberalism puts substantive questions of justice aside. Here, Rawls does not focus on the content of justice but on whether justice as fairness can provide shared political ground given conflicting comprehensive moralities. Moreover, Liberalism lacks the well-defined opponents of Theory. To be sure, Rawls contrasts the ideal of overlapping consensus on a political conception of justice with the communitarian aspiration to achieve social unity through a shared conception of human nature and the human good.41 But communitarianism lacks the sharp definition of utilitarianism, libertarianism, or liberal pluralism, contributing to the relentlessly abstract character of Rawls's presentation. 38. A central claim in Theory is that we will be led to surprising, egalitarian conclusions about the limits of legitimate socioeconomic inequality by reasoning from the same fundamental ideas — about the equality of moral persons and our basic interests — that support familiar and settled convictions about the injustice of religious intolerance and racial discrimination. See RAWLS, supra note 3, at 19-20, 150-83. To make his case, Rawls gathers the less controversial claims and convictions together in the original position, thus requiring our reasoning about socioeconomic issues to conform to principles and ideas to which convictions about fairness and basic liberties already commit us. See Joshua Cohen, Democratic Equality, 99 ETHICS 727 (1989). 39. RAWLS, supra note 3, at vii-viii. 40. On natural liberty and liberal pluralism, see id at 65-75. For an argument against natural liberty and liberal pluralism, see BRIAN BARRY, THEORIES OF JUSTICE 217-34 (1989). I contrast liberal pluralism with Rawls's view in Cohen, supra note 37. 41. See pp. 42-43, 146, 201.

John Rawls

May 1994]

A More Democratic Liberalism

387

1511

Because it pays less attention to substantive issues of political justice and lacks such sharply defined opponents, Liberalism is unlikely to generate either the excitement of Theory or the same interdisciplinary ferment. But these are caveats, not criticisms. Liberalism is a book of very great depth and importance. In due course it will likely change the shape of political philosophy, sharpening political philosophy's autonomy by increasing its distance from moral philosophy, and perhaps will have similarly salutary effects on political argument itself. II. LIBERALISM: A PHILOSOPHY OF LIFE? Liberalism, Rawls says, addresses "a serious problem internal to justice as fairness" (p. xv) — the view presented in Theory. In general terms, the problem arises from a lack of realism engendered by inattention to the fact of reasonable pluralism (pp. xv-xviii). More particularly, the difficulty emerges in the account of stability advanced in Part Three of Theory. To locate the difficulty more precisely, and to see why it is so troubling, I will first sketch three main elements of Theory, then present an objection that many commentators have raised about the main line of argument in Theory, and finally restate that difficulty as a tension internal to justice as fairness. A. Three Elements of Theory Theory presents, first, an attractive ideal of a just society — a wellordered, democratic society, featuring a consensus on norms of justice. The content of the consensus is given by two principles: [First Principle:] Each person has an equal right to a fully adequate scheme of equal basic liberties which is compatible with a similar scheme of liberties for all. [Second Principle:] Social and economic inequalities are to satisfy two conditions. First, they must be attached to offices and positions open to all under conditions of fair equality of opportunity; and second, they must be to the greatest benefit of the least advantaged members of society.42

A society satisfying these principles achieves, Rawls proposes, some measure of "reconciliation of liberty and equality/'43 Suppose that the real value of the freedom guaranteed to a person by the protection of basic liberties is fixed by that person's command of resources, rather than by her position relative to others.44 Then the two 42. I take the formulation of these principles, first stated in Theory, from Liberalism. P. 291. 43. RAWLS, supra note 3, at 204. 44. Rawls does not think that the worth of political liberty to a person is fixed by that person's absolute command of resources. Because the political process has "limited space," the value of political liberty also depends on relative position. See pp. 328-29. For this reason, Rawls imposes a special requirement of the "fair value** of political liberty: roughly, that people in different social positions have equal chances to hold office and influence the political process. See pp. 327-31, 356-63; RAWLS, supra note 3, at 224-27. For a discussion of relative positions, see id. at 530-41.

388

John Rawls

1512

Michigan Law Review

[Vol. 92:1503

principles together require that a society "maximize the worth to the least advantaged of the complete scheme of equal liberty shared by all/*45 This requirement of maximizing the minimum worth of liberty, Rawls says, "defines the end of social justice/*46 Second, Rawls offers a contractual defense of this egalitarian-liberal conception of justice. Carrying the social contract idea "to a higher order of abstraction" (p. xv), he argues that the two principles would be chosen in an initial situation of choice — the "original position"47 — in which the parties are assumed not to know anything particular about themselves — about their position in the distribution of alienable resources, their position in the distribution of native endowments, and the determinate aims, attachments, or views of the world that comprise their conception of the good.48 Required to choose under conditions of severe ignorance, they are uncertain of the effects of their choice on their own lives. Concerned to assure that they can live with that choice wherever they end up, the parties would choose to provide themselves, Rawls argues, with the strong downside protection assured by the two principles.49 Third, Rawls proposes that the various constraints on knowledge imposed in the original position represent requirements that strike us, on reflection, as reasonable to impose on norms of justice or on their justification.50 Concerns about fairness, for example, and a conception of individuals as equal moral persons with a conception of the good and the capacity for a sense of justice fuel these constraints.51 B. Original Position: A Liberal Philosophy of Life? None of these three central elements of justice as fairness has won general acceptance.52 But criticisms of Rawls's claims about the reasonableness of the conditions imposed in the original position have been especially sharp among moral and political philosophers. Though the details of the criticisms take many forms, the central ob45. RAWLS, supra note 3, at 205. 46. Id. In Liberalism, Rawls says, less strongly, that maximizing the minimum worth of liberty "defines one of the central aims of political and social justice." P. 326. 47. RAWLS, supra note 3, at 17-22. 48. Id at 136-42. 49. Id. at 150-57, 175-83. 50. See id. at 18, 587. For a complete list of passages in Theory that state the idea of the original position as expressing reasonable requirements on arguments for principles, see p. 25 n.28. 51. "If the original position is to yield agreements that are just, the parties must be fairly situated and treated equally as moral persons." RAWLS, supra note 3, at 141; see also pp. 23-27. 52. For criticisms of the principles themselves, see DAVID GAUTHIER, MORALS BY AGREEMENT (1986); ROBERT NOZICK, ANARCHY, STATE, AND UTOPIA (1974); G.A. Cohen, Incentives, Equality, and Community, in 13 TANNER LECTURES, supra note 15, at 261. On the argument from the original position, see John C. Harsanyi, Can the Maximin Principle Serve as a Basis for Morality? A Critique of John Rawls's Theory, 69 AM. POL. Sci. REV. 594 (1975).

John Rawls

May 1994]

A More Democratic Liberalism

389

1513

jection to Rawls's construction is that the design of the original position presupposes a particular conception of the good. It does not, contrary to Rawls's claims, provide a reasonable device for addressing controversies about justice among people with différent conceptions of the good, because it will only be found attractive by people drawn to a liberal philosophy of life — one that holds that individual independence, choice, and self-mastery are the fundamental values that ought to govern our lives. Critics have localized the offending bias in different places. Thomas Nagel criticized Rawls's assumption that all the parties in the original position want "primary goods" — in particular, income and wealth — as unfairly biased in favor of individualistic conceptions of the good.53 Brian Barry objected to the individualism implicit in Rawls's contractual method of justification, which proceeds from individual judgments about what is best for me, all else equal, to judgments about how society ought to be arranged.54 According to Michael Sandel, Rawls assumed a liberal philosophy of life when he required that we place our conceptions of the good behind a veil of ignorance.55 To see the force of these criticisms, consider Sandel's objection. Reasoning from behind the veil of ignorance requires that we evaluate norms of justice without reference to our own conception of the good. It is a puzzling idea. Why, and how, are we to reason about justice without drawing on our views about the proper conduct and ends of human life? If we hold the sincere conviction that a life of self-realization is a better life — if we think that such a life is genuinely better, not simply the life that we prefer — then what reason could there be for bracketing that conviction when we assess principles of justice? One reason for such bracketing is that we cannot agree on terms of 53. See Thomas Nagel, Rawls on Justice, 82 PHIL. REV. 220 (1973). Rawls replies at pp. 195-200. 54. BRIAN BARRY, THE LIBERAL THEORY OF JUSTICE 116-27 (1975). For Barry's statement of the liberal philosophy of life, see id at 126-27. 55. MICHAEL J. SANDEL, LIBERALISM AND THE LIMITS OF JUSTICE (1982). Sandel sketches the liberal philosophy of life — a conception of *'[t]he deontological universe and the independent self that moves within it." Id. at 177. William Galston also criticizes Rawls for failing to acknowledge his reliance on a conception of the good. See WILLIAM A. GALSTON, LIBERAL PURPOSES 118-62 (1991). Galston argues, however, that such reliance is no embarrassment On the contrary, liberalism must openly avow its dependence on a view of the good, albeit a "deliberately thin** view, "a kind of minimal perfectionism." Id, at 177. Galston's view is puzzling. It is not controversial that some account of the good is required for an account of justice. See pp. 173-211; RAWLS, supra note 3, at 395-99. Moreover, Galston's account of the good is itself constrained by a concern "to provide a shared basis for public policy." GALSTON, supra, at 178. This constraint suggests that Galston's account of the good may not comprise part of a comprehensive perfectionist conception but may instead be part of a political conception of the good in the sense defined by Rawls at pp. 174-76. I say that Galston's account "may be" political because it is not clear what he means by a "shared basis of public policy" or how the concern to provide such a basis — as distinct from concerns within an account of the good — constrains the role of ideas of the good in his presentation of liberalism.

390

John Rawls

1514

Michigan Law Review

[Vol. 92:1503

cooperation for a pluralistic society if the rationale for those terms premises a particular conception of the good. Peaceful cooperation requires agreement, and agreement requires that citizens put aside "the contingencies that set them in opposition."56 But Rawls's reasons in Theory are not simply a matter of securing social peace. He argues instead that fairness to citizens as moral persons requires that we not rely on any particular conception of the good in justifying principles that all will have to live by. Instead, fairness demands that "ft]he arbitrariness of the world . . . be corrected for by adjusting the circumstances of the initial contractual situation/*57 But why is it unfair to people as moral persons to treat them in accordance with principles of justice chosen on the basis of an account of the best life? Why correct for the "arbitrariness of the world" by abstracting from convictions about the best life? Why not correct for that arbitrariness by encouraging everyone to endorse the truth about the best life? To be sure, conceptions of the good sometimes set people in opposition; but why are they "contingencies"? According to Sandel's objection, Rawls's answer to these questions itself relies in the end on a particular account of the best life and a particular view of the person that goes with that account. We will only take an interest in what is chosen behind the veil of ignorance if we deny that our fundamental aims and attachments are good indicators of who and what we are. Moreover, we will be drawn to that denial only if we regard ourselves as, at bottom, agents unencumbered by fundamental attachments to our actual ends, as essentially choosers of values rather than as carriers and renewers of the values of particular traditions and communities — only if we are attracted to the idea that our basic allegiances themselves are elements of the arbitrariness of the world and that the unchosen life is not worth leading.58 This will do as a statement of comprehensive liberalism, and we can understand a theory of justice built on these foundations as presenting the political implications of such a liberal outlook. But, 56. RAWLS, supra note 3, at 137 (emphasis added). 57. Id. at 141. 58. Sandel identifies two key assumptions in Theory: that we are essentially choosers — the priority of the self with respect to its ends — and that we are not essentially members of a community — "[t]he priority of plurality over unity." SANDEL, supra note 55, at 50-59. Notice that it is possible to deny the first proposition — thus affirming that our identity is fixed by our ends — without denying the second — that is, without affirming that we are essentially members of a community. I might regard myself as standing in an essentially personal relationship with God and as bound by obligations arising from that relationship, or as a locus of artistic creativity, or as essentially a seeker of truth. In each case, I might treat my relations with others as instrumental for those deeper purposes, rejecting the ideal of community. To put the point in historical terms, both Hegel and Nietzsche rejected the conception of the self as essentially a chooser of ends. But, not to put too fine a point on it, they had very different views about community. For criticisms of the conception of the self as chooser, see HEGEL, supra note 26, §§ 15-20, 105-141; and FRIEDRICH NIETZSCHE, ON THE GENEALOGY OF MORALS 44-46 (Walter Kaugmann éd., Walter Kaufmann & RJ. Hollingdale trans., Vintage Books 1969) (1887) (reprinted with ECCE HOMO).

John Rawls

A More Democratic Liberalism

May 1994]

391

1515

according to the criticisms, the original position's dependence on such specific commitments disqualifies it from serving as a shared or neutral basis for settling on principles of justice in a democratic society whose equal citizens disagree sharply about liberal ideals of autonomy and individuality. Sandel goes further. He thinks that Rawls's implicit commitment to a conception of the self as an essentially unencumbered chooser of ends is not merely morally sectarian; it is also inconsistent with Rawls's avowed aim of avoiding obscure and controversial Kantian metaphysical commitments59 and with our experience of both ourselves and our connections with our commitments.60 In short, Sandel is concerned not simply to demonstrate Rawls's own reliance on a view of the good but also to undermine the liberal conception of justice by exploding the views of the good and the self on which it depends.61 These further points are not, however, essential for our current purpose. C.

The Internal Problem: Congruence and Stability

Earlier, I mentioned Rawls's claim that Liberalism addresses a problem "internal" to justice as fairness.62 Thus far, however, I have presented an objection to the original position that might be thought to operate externally. I propose now to show how claims about the objectionable dependence of the original position on a particular philosophy of life can be turned into the internal tension in justice as fairness — the problem in Theory's account of stability — that Liberalism aims to address. In characterizing the ideal of a well-ordered society and presenting an account of its stability, Rawls makes essential use of the idea of normative consensus.63 In a well-ordered society, "[e]veryone has a similar sense of justice and in this respect a well-ordered society is homogeneous. Political argument appeals to this moral consensus."64 Moreover, this shared sense of justice plays a "fundamental role" in ensuring that "the basic structure is stable with respect to justice."65 To be sure, some idea of agreement figures in any contractual theory of justice. But the "moral consensus" Rawls refers to is not sim59. SANDEL, supra note 55, at 94-95. 60. Id. at 179. 61. The project of undermining liberalism by excavating and exploding its psychological and metaphysical commitments traces back to HEGEL, supra note 26. The most ambitious modern effort along these lines is ROBERTO M. UNGER, KNOWLEDGE AND POLITICS (1975). 62. See the introduction to Part II of this review, supra. 63. The following discussion draws on Joshua Cohen, Moral Pluralism and Political Consensus, in THE IDEA OF DEMOCRACY 270 (David Copp et al. eds., 1993). 64. RAWLS, supra note 3, at 263. On the role of consensus in the ideal of a well-ordered society, see p. 35. See also RAWLS, supra note 3, at 5, 453 -58. 65. RAWLS, supra note 3, at 458.

392

John Rawls

1516

Michigan Law Review

[Vol. 92:1503

ply an ex ante agreement on institutions and relations of authority of a kind associated with Hobbesian and Lockean social contracts.66 Closer in this respect to Rousseau» Rawls supposes that citizens in a just political society share a conception of justice and that politics is openly guided by that conception.67 Justice as fairness aims to specify the appropriate content for such a conception, the content of the general will for a society of free and equal persons. This emphasis on the role of consensus in the ideal of a wellordered society is understandable. A moral consensus on political fundamentals is a basic good for at least four reasons. First, for any conception of justice, the existence of a moral consensus on it increases the likelihood that social order will stably conform to the conception.68 Second, a moral consensus promotes a variety of specific values of considerable importance. Assuming that norms of justice are not motivationally inert, consensus on them increases social trust and harmony, supports social peace, simplifies decisionmaking, reduces monitoring and enforcement costs by encouraging a willingness to cooperate, and — if public debate and decisions reflect the consensus — reduces alienation from public choices because citizens embrace the norms and ideals that guide those choices. Third, a consensus on norms of justice provides a way to reconcile the ideal of an association whose members are politically independent and self-governing with an acknowledgment of the central role of social and political arrangements in shaping the self-conceptions of citizens, constraining their actions, channeling their choices, and determining the outcomes of those choices.69 When a consensus on norms and values underlies and explains collective decisions, citizens whose lives are governed by those decisions might nonetheless be said to be independent and self-governing. Each endorses the considerations that produce the decisions as genuinely moral reasons and af66. See THOMAS HOBBES, LEVIATHAN 120-29 (Richard Tuck éd., Cambridge Univ. Press 1991) (1651); LOCKE, supra note 17, at 374-77, 395-400. 67. On the role of a shared conception in Rousseau, see Joshua Cohen, Reflections on Rousseau: Autonomy and Democracy, 15 PHIL. & PUB. AFF. 275 (1986). 68. See Rawls's "third general fact" at p. 38. 69. See the discussion of full autonomy at pp. 77-78. Rawls distinguishes there between endorsing full autonomy as a political value and affirming autonomy as a comprehensive moral value, to be realized in all aspects of life and conduct. The concern to reconcile self-government with interdependence is central to Rousseau's project, though Rousseau's own presentation suggests that he thinks of self-government or moral liberty as a comprehensive moral value, tied to an account of our true nature. On moral liberty, see JEAN-JACQUES ROUSSEAU, ON THE SOCIAL CONTRACT (1762) [hereinafter ROUSSEAU, SOCIAL CONTRACT], reprinted in BASIC POLITICAL WRITINGS OF JEAN-JACQUES ROUSSEAU 139, 144-46 (Donald A. Cress trans. & éd., 1987); on our nature as free beings, see JEAN-JACQUES ROUSSEAU, Discourse on the Origin and the Foundations of Inequality Among Men, in THE FIRST AND SECOND DISCOURSES 137, 189-90 (Victor Gourevitch ed. & trans., Harper & Row 1986).

John Rawls

May 1994]

A More Democratic Liberalism

393

1517

firms their implementation.70 Finally, under conditions of political consensus, citizens achieve a form of mutual respect. Each offers as reasons for a decision only considerations that others who are subject to political power take as reasons, and state power is exercised only within the bounds set by these reasons.71 The force of this point as a basis for mutual respect is increased by recalling the distinction I noted earlier between a unanimous, ex ante agreement and an ex post consensus on norms of justice that frame political debate.72 In a Hobbesian contract of subordination, everyone agrees to submit to a common agent, accepting the will and judgment of that agent as authoritative.73 Nothing in the content of the agreement — nothing manifest in political experience itself — directly expresses mutual respect.74 With a political consensus, by contrast, the authorization of power proceeds in terms that all citizens accept ex post — in accordance with reasons that are shared and therefore accepted by all who are subject to the power. That does provide a basis for mutual respect. Consensus, then, has its virtues. But not every consensus is attractive. Those attractions depend on the content of the consensus and on the conditions under which it is sustained. Suppose, for example, that a moral consensus is attractive because it provides a way to make selfgovernment — or association on terms of mutual respect — consistent with the unavoidable chains of political connection (see the third and fourth reasons stated above). TTien the consensus must be freely sustained and not simply a form of enforced homogeneity. A consensus is free only if it is arrived at under conditions that ensure the possibility of individual reflection and public deliberation — for example, conditions that protect expressive and associative liberties. Here we arrive at the internal problem of Liberalism. Assurances of expressive and associative liberties — necessary if the consensus that defines a well-ordered society is to be free and attractive — are bound to be associated with moral, religious, and philosophical pluralism.75 But can the value of substantive consensus on justice survive such pluralism? Let us say that a society is liberal only if it strongly protects expressive and associative liberties. Then, to restate the question: Can there be political consensus and social unity, given the inevi70. We also need to add that everyone believes with good reason that the decisions express the shared norms and values. 71. See the discussion of legitimacy at pp. 136-37, 216-19. 72. See supra notes 66-67 and accompanying text. 73. See HOBBES, supra note 66, at 120-21. 74. But see ROUSSEAU, SOCIAL CONTRACT, supra note 69, at 197 ("Once the populace is legitimately assembled as a sovereign body . . . the person of the humblest citizen is as sacred and inviolable as that of the first magistrate."). 75. See the "first general fact" at p. 36.

394

John Rawls

1518

Michigan Law Review

[Vol. 92:1503

table pluralism of a liberal society?76 Why, in particular, ought we to expect — as Rawls suggests — that the members of a well-ordered society regulated by Rawls's principles of justice will find the conditions imposed on the original position reasonable? According to the criticisms I referred to earlier, the original position assumes a liberal philosophy of life and presents the political extension of that philosophy.77 If these criticisms are right, then the comprehensive views that some members of a just society find attractive will likely lead them to reject the original position. The discussion of stability in Part Three of Theory suggests that the criticisms are right. Because it does, Rawls concludes that his account of the stability of a well-ordered society is in trouble: that it is "not consistent with the view as a whole" (p. xvi). In Part Three, Rawls advances a two-stage case for the stability of a society regulated by his principles of justice.78 The first stage focuses on the acquisition of a sense of justice — "an effective desire to apply and to act from the principles of justice [the two principles chosen in the original position] and so from the point of view of justice."79 Rawls sketches how the members of a just society could be expected, through membership in a series of institutions — from family, to the associations of civil society, to citizenship in the state — to acquire an understanding of and an effective desire to act from a sense of justice to which Rawls's principles give content.80 The second stage shifts attention from the acquisition of a sense of justice to the congruence of that sense with a person's conception of the good. Here Rawls argues that the members of a just society would, with reason, regard the regulation of their conduct by their sense of justice — as given by the two principles — as itself good for them: that is, they would find their sense of justice congruent with their good, rather than regarding it as an unwelcome constraint on the pursuit of their good. If this claim about the good of a sense of justice is right, then we have an important force for stability in a just society.81 Moral pluralism causes troubles for this happy picture. Consider one of the arguments for congruence: "acting justly is something we want to do as free and equal rational beings. The desire to act justly 76. I do not mean to suggest that other societies are not pluralistic. See supra note 5. 77. See supra section II.B. In his Tanner Lecture on "liberal equality," Ronald Dworkin defends a version of liberalism on the grounds of its continuity with a more comprehensive liberal outlook on life. See Dworkin, supra note 15, at 20-22. 78. Rawls presents the first part of the case in RAWLS, supra note 3, at 462-96, and the second part in id. at 513-77. 79. Id. at 567. 80. Id. at 462-96. See also Hegel's theory of the formation of the will through the various spheres of ethical life — family, civil society, and state. HEGEL, supra note 26, §§ 142-329. 81. See RAWLS, supra note 3, at 499, 501.

John Rawls

May 1994]

A More Democratic Liberalism

395

1519

and the desire to express our nature as free moral persons turn out to specify what is practically speaking the same desire/'82 The claim that these desires have the same content rests on the argument from the original position. Or, as Rawls indicates elsewhere, the "sentiment of justice" is — for anyone who "understands and accepts the contract doctrine" — the very same desire as the desire to act on principles that would be chosen "in an initial situation which gives everyone equal representation as a moral person," and also the same as the desire "to act in accordance with principles that express men's nature as free and equal rational beings."83 In the original position, we are represented as free moral persons, so to act from the principles chosen there is to express our nature as free and not to "give way to the contingencies and accidents of the world."84 Moreover, the argument from the original position not only selects principles of justice but also requires that those principles take priority in regulating our conduct. To express "our freedom from contingency and happenstance,"85 then, we need more than a sense of justice given content by the principles chosen in the original position. We must also give priority to our sense of justice, assigning it an authoritative role in the regulation of conduct. A central element in the case for congruence and stability, then, is that members of a well-ordered society will develop a conception of their nature as free beings, will regard the expression of that free nature in their own conduct as a fundamental good, and will understand — because of their "lucid grasp of the public conception of justice upon which their relations are founded"86 — that such expression requires acting from the principles of justice that would be chosen in the original position, giving those principles a special regulative role.87 The case for the two principles, then, depends upon the case for stability; the case for stability depends in part upon the case for congruence; and the case for congruence depends upon an account of our "nature as free moral persons"88 and the desire to express our nature as free.89 But this line of dependence strongly suggests that the argument for congruence, and so the case for stability, depends upon a set 82. Id. at 572 (citation omitted). 83. Id. at 478. 84. Id. at 575. 85. Id. at 574. 86. Id. at 572. 87. The condition of "full publicity," defined at pp. 66-67, requires public availability of the conception of justice and the full rationale for it. 88. RAWLS, supra note 3, at 572. 89. Rawls ties this argument to the Kantian interpretation of justice as fairness. See id. The argument is one of four he offers in support of congruence. It might be interpreted as an argument addressed to those who endorse a comprehensive Kantian view, rather than as one of four arguments that citizens generally will find persuasive. But Theory clearly offers it as the latter.

396

John Rawls

1520

Michigan Law Review

[Vol. 92:1503

of moral commitments and self-understandings that some members of a well-ordered society will reasonably reject.90 For example, some citizens may think that their nature consists in the possession of various natural, human powers, that the human good consists in a perfection that fully realizes those powers, and that the requirements of morality set out the conditions for such perfection. Others may think of themselves as creatures of a God who imposes obligations that bind their moral freedom. Such citizens accept moralities that are, to use Kant's term, heteronomous. They, too, wish to express their nature and not to give way to the contingencies and accidents of the world. But it is unclear why they should find the original position a plausible way to specify the content of their expression. With Locke, they may suppose that their fundamental powers are the capacity to understand and to act from the Creator's requirements and that they express their nature by acting from those requirements.91 To be sure, adherents of such a view might reject the imposition of a religious establishment and affirm the importance of the free exercise of religion. But they would do so because forced religious practice does not fulfill basic religious duties and so provides no route to salvation,92 rather than because a regime of religious toleration expresses their "nature" as free moral persons. They do not acknowledge themselves to have such a nature. Some people, then, may reject the characterization of our nature as free; they will be drawn neither to the reasonableness of the original position as a rendering of their nature, nor to acting from the principles selected there because such action expresses their nature.93 Thus, Rawls concludes that the conception of a well-ordered society presented in Theory "is unrealistic . . . because it is inconsistent with realizing its own principles under the best of foreseeable conditions" (p. xvii). Under the best of foreseeable conditions, a society that satisfies the two principles will be a society in which some citizens reject the conception of our nature used in Theory to underwrite the original position and the account of congruence. "The account of the stability 90. I have concentrated on the problem for congruence. The account of acquisition, however, faces a parallel difficulty. An account of the acquisition of a desire to act on principles must explain why that desire, which is not instrumental, does not reflect a strange affection for rules. In Theory. Rawls responds to this concern by explaining that moral principles can "engage our affections*' in part because acting on them expresses our "nature as free and equal rational beings." Id. at 476. But this explanation leaves us with a gap in the account of acquisition in the case of those citizens who do not see their nature in such terms. In this connection, see pp. 82-86 on principle-dependent and conception-dependent desires. 91. See LOCKE, supra note 17, at 310-11. 92. See JOHN LOCKE, A LETTER CONCERNING TOLERATION 18-20 (Patrick Romanell éd., William Popple trans., 1955) (1689). 93. They may, of course, be attracted to those principles and to the original position itself for other reasons. See infra notes 110-12 and accompanying text.

John Rawls

May 1994]

A More Democratic Liberalism

397

1521

of a well-ordered society in Part III is therefore also unrealistic . . ." (p. xvii). How, then, is it possible to achieve consensus on a conception of justice suited to a democratic society of equal citizens and to reap the benefits of that consensus, given the pluralism of comprehensive moralities that inevitably marks such a society? More particularly, can the presentation and defense of a conception of justice for a democratic society be freed from the unacceptably narrow premises of a comprehensive moral liberalism? That is the question of political liberalism. III.

POLITICAL LIBERALISM

Rawls's answer to the question of political liberalism contains two parts. The idea of a political conception of justice plays a central role in the first part; the idea of an overlapping consensus is the key to the second. A. A Political Conception of Justice Given the plurality of comprehensive moralities, the claim that consensus is possible faces a threshold problem. A conception of justice can win general acceptance only if it can be suitably formulated. Its formulation must be understandable to citizens with competing views of the good and must not itself preclude acceptance by some citizens.94 Some conceptions of justice would, however, on their face, be unacceptable to some citizens — for example, if the conceptions appeal to values that are not implicated in public institutions or that reasonable people might reject. Suppose an account of justice requires a distribution of resources that ensures equal pleasure, or suppose it mandates a distribution that enables each citizen to come equally close to achieving his aims. Both views face troubles because citizens reasonably disagree about the relative value of pleasure and of relative achievement. So these conceptions would be, on their face, unacceptable.95 Rawls calls a view that is suitably formulated a "political conception of justice" (p. 11). Three features — each necessary if the conception is plausibly to provide the focus of agreement, given the fact of reasonable pluralism — define such a conception:96 it must have limited scope, extending only to issues about the basic structure of society 94. This condition is necessary but not sufficient, because a view that is formulated without reference to any comprehensive moral view may nevertheless be attractive only to those who hold a particular view. See infra notes 109-12 and accompanying text. 95. For discussion and criticism of these two versions of equality of welfare, see Ronald Dworkin, What Is Equality? Part I: Equality of Welfare, 10 PHIL. & PUB. AFF. 185, 204-09, 220-24(1981). 96. See pp. 11-15.

398

John Rawls

1522

Michigan Law Review

[Vol. 92:1503

and not to norms of personal conduct or ideals of life; it must draw on ideas familiar to citizens from the political culture of a democracy, not on ideas belonging exclusively to particular traditions of moral thought that are not available to all; and it must be presented as freestanding, not as depending for formulation or justification on its roots in a comprehensive morality.97 In short, a political conception of justice is formulated as autonomous from comprehensive conceptions of the good with respect to scope, content, and justification. Each of these three forms of autonomy should contribute to the possibility of its general acceptance. To see how these kinds of autonomy help to address the problems about the original position and veil of ignorance I sketched earlier, consider the aspect of the political conception that Rawls refers to as a "political conception of the person" (pp. 18-20, 29-35, 48-54, 86-88). The original position isolates certain features of people as relevant to its problem of justice, setting aside other features as irrelevant — and so to be excluded by the veil of ignorance. The relevant features include certain basic moral powers: the capacities for a conception of the good — to form, pursue, and revise such a conception — and for a sense of justice.98 The irrelevant features include sex, race, natural abilities, and determinate conceptions of the good. As my earlier discussion of the original position indicates,99 Theory was not entirely clear about the basis of this distinction between relevant and irrelevant characteristics. This lack of clarity contributed to the impression that justice as fairness was the political expression of a comprehensive moral liberalism. Thus, Rawls often referred to the morally relevant or irrelevant as if to say that the distinction derives from a comprehensive moral doctrine.100 Sometimes he referred to the irrelevant characteristics as "contingencies," as though to suggest a metaphysical foundation for the distinction.101 Sometimes — as I in97. A political conception is presented as freestanding in the way that logic or number theory is. A presentation of logical laws — for example, the law of excluded middle — proceeds without tying the laws to a theory of truth or issues in the theory of meaning; a presentation of number theory proceeds without reference to questions about the ontological status of numbers. 98. Pp. 18-20; RAWLS, supra note 3, at 561. 99. See supra notes 79-93 and accompanying text. 100. For example, Rawls refers to the constraints in the original position as "conditions that are widely recognized as fitting to impose on the adoption of moral principles." RAWLS, supra note 3, at 584. Similarly, the criticisms of natural liberty and liberal equality refer to social circumstances and natural assets as features that are "arbitrary from a moral point of view." Id. at 72, 74-75. 101. I say "sometimes" because some passages in Theory strongly suggest that apparently metaphysical notions should be interpreted morally. Take, for example, the following remark: "Our moral sentiments display an independence from the accidental circumstances of our world, the meaning of this independence being given by the description of the original position and its Kantian interpretation.11 Id. at 475 (emphasis added). To say that the description of the original position gives the meaning of independence is to say that independence is a matter of the irrelevance for moral purposes of certain features of the person, rather than a matter of the metaphysical contingency of those features.

John Rawls

May 1994]

A More Democratic Liberalism

399

1523

dicated in my discussion of congruence and stability102 — he suggested that the distinction is rooted in an account of "our nature," permitting both metaphysical and moral interpretations.103 Liberalism draws the distinction between relevance and irrelevance in the same place: the power to form, pursue, and revise a conception of the good and the power to form and act from a sense of justice are relevant; and sex, race, natural abilities, and determinate conceptions of the good are irrelevant (pp. 29-35). But the point of the distinction, according to Liberalism, is to present a conception of the person that will play a role in a political conception of justice, and so Liberalism underscores that the conception of the person is itself political in each of the three ways noted earlier: scope, content, and justification. Thus, irrelevant should not be understood absolutely, metaphysically, or in terms of a general moral view, but only as implying that a feature of a person is not important for the purposes of political argument — in particular, not important for political argument aimed at specifying the requirements of justice for a society in which members are understood as free and equal. Contingent ought similarly to be given a nonmetaphysical rendering, as implying that a feature is not relevant to political argument. We can, then, determine which features are "irrelevant, politically speaking, and hence [to be] placed behind the veil of ignorance" (p. 79) by systematizing and extending reasonably familiar ideas about the justification of political arrangements in a democratic society. This basis is appropriate for the distinction given the question that justice as fairness sets out to resolve: "What is the most appropriate conception of justice for specifying the terms of social cooperation between citizens regarded as free and equal, and as normal and fully cooperating members of society over a complete life?" (p. 20). The conception of citizens as free and equal represents a familiar element of the political culture of democratic societies. The problem is to determine more precisely what that political conception involves and to address a longstanding controversy about what account of justice is best suited to citizens as free and equal.104 Thus, we look to settled ideals and convictions about basic democratic institutions, and to settled understandings about the justification of public norms in a democratic society, and then draw the relevantirrelevant distinction by reference to the characteristics of persons that play a role in those ideas, convictions, and understandings. One may then call the irrelevant features "contingencies," but with no intention to affirm — or to deny — that an individual could exist without the feature in question, or to say — or to deny — anything about the 102. See supra notes 78-93 and accompanying text. 103. RAWLS, supra note 3, at 251-57, 572. 104. See pp. 20, 22, 26, 34-35.

400

John Rawls

1524

Michigan Law Review

[Vol. 92:1503

importance of irrelevant features in other settings. They are simply unimportant for the purposes at hand, whatever their metaphysical standing and however important they may be for other purposes, including other ethical purposes. To be more specific, arguments aimed at establishing that certain properties are contingent (irrelevant to the problem of political justification) and that others are aspects of our essential nature (important to that problem) proceed along at least two main lines. The first seeks to show that current ideals — for example, of fairness, religious toleration, and racial and sexual equality — and patterns of political argument — for example, on constitutional matters — treat certain facts as irrelevant. For instance, it is widely agreed that we ought to protect certain basic rights — expression, political participation, conscience, and equal treatment — without regard to social background, sex, or race. Furthermore, social class ought not to restrict opportunity. These are clear cases of unfairness. So in reasonably settled understandings of justice, we treat facts about class, sex, and race as contingencies — matters that are irrelevant to argument about the justice of basic institutions. Similarly, the constitutional treatment of religious and political ideals suggests the irrelevance of conceptions of the good to such argument. For example, conversion, sin, and religious laxity are not civil offenses. Whatever its implications for a person's self-conception, being "born again" has no civil consequences; being born again does not, for example, absolve a person of contractual obligations undertaken prior to that rebirth or give a person who is reborn on election day a right to a second vote. Furthermore, in the case of political ideals, endorsing the legitimacy of the political order is not — in principle, at least — a precondition for equal political rights, a point underscored by conventional hostility to regulating expression by virtue of its content and, more particularly, its viewpoint.105 A second strategy is to show that certain features of people are themselves so dependent on concededly irrelevant facts that to permit them to play a role in political justification would be tantamount to allowing the irrelevant facts to play a role. So they too should be treated as irrelevant. The development of abilities and talents, for ex105. I assume this hostility to be widely shared, even by people who do not think that content or viewpoint regulation is always impermissible. Justice Marshall provided a classic statement of the general concern about content regulation in Police Department v. Mosley, 408 U.S. 92, 95 (1972) ("[A]bove all else, the First Amendment means that government has no power to restrict expression because of its message, its ideas, its subject matter, or its content.")- On viewpoint discrimination, see Texas v. Johnson, 491 U.S. 397, 414 (1989). For discussion of content and viewpoint regulation, see John Hart Ely, Flag Desecration: A Case Study in the Roles of Categorization and Balancing in First Amendment Analysis, 88 HARV. L. REV. 1482, 1482-508 (1975); T.M. Scanlon, Jr., Content Regulation Reconsidered, in DEMOCRACY AND THE MASS MEDIA 331 (Judith Lichtenberg éd., 1990); Geoffrey R. Stone, Content-Neutral Restrictions, 54 U. CHI. L. REV. 46 (1987); and Geoffrey R. Stone, Restrictions of Speech Because of Its Content: The Peculiar Case of Subject-Matter Restrictions, 46 U. CHI. L. REV. 81 (1978).

John Rawls

May 1994]

A More Democratic Liberalism

401

1525

ample, seems closely linked to the social circumstances and aspirations that the entrenched forms of argument fix as contingencies. So talents and abilities ought to be treated as contingencies and not appealed to as fundamental reasons for differential advantages. I would need to say much more about these matters in order to evaluate Rawls's distinction between relevant and irrelevant, and the associated political conception of the person. I have provided only an outline of the rationale for the distinction. But its force — and limits — as a response to the original position's difficulties should now be clear. According to the objection, the original position rests on a liberal philosophy of life that places especially great weight on the importance of choice and that sees the self as, in its fundamental nature, a chooser of its own ends. Rawls's claim that "the self is prior to the ends which are affirmed by it"106 suggests a commitment to such a philosophy. But the political conception of the person offers a restrictive interpretation of this priority. It neither affirms nor denies that people could, as a metaphysical matter, exist without their aims as pure choosers of ends, as "Kantian transcendent or disembodied subject[s]" who are "shorn of empirically-identifiable characteristics";107 or that citizens can imagine their own lives continuing with their final aims different from what they now are; or that they would actually be the same persons if their final aims were radically altered; or that, as an ethical matter, the aims of citizens are worth pursuing only if chosen by them. Instead, the political conception ties both the content of and the rationale for the alleged priority to the aims of a theory of justice for a democratic society and to the public availability of the idea of citizens as equals. According to the political conception, citizens are prior to their ends in that no particular ends are mandatory from a public point of view, and citizens must be assured favorable conditions for reflecting on and revising their aims, should they wish. For example, obligations that a person has by virtue of her conception of the good do not have public standing as obligations. Moreover, civil standing does not alter with shifts in fundamental aims, no matter how much a person's selfconception is bound up with those aims. This is not to say, however, that all obligations are matters of self-legislation, or that fundamental values are a product of choice, or that they are only worth pursuing if they are such a product. The political conception of the person does not state a position on these matters.108 That conception is simply a statement about how citizens should be represented for the purposes of 106. 107. 108. laws do

RAWLS, supra note 3, at 560. SANDEL, supra note 55, at 95. The political conception does not take a position in the way that statements of logical not, on their face, take a position about the nature of meaning.

402

John Rawls

Michigan Law Review

1526

[Vol. 92:1503

political argument. For this reason, nothing in the very statement of the political conception of the person conflicts with comprehensive moralities that are not organized around the ideal of autonomy or around the thought that we are, by our nature, free beings. B.

Overlapping Consensus

Suppose this enterprise of reinterpretation succeeds — that a liberal conception can be formulated as a freestanding political doctrine, facially independent of any comprehensive moral conceptions. Providing this formulation would help in securing social unity under conditions of moral pluralism. It would overcome the threshold problem that I disclosed earlier.109 But it would not suffice to defeat the objections to or the associated internal troubles for Rawls's view. The objection to the original position was not that its very statement reveals it to be part of a liberal philosophy of life but rather that citizens will be drawn to it — will find it a reasonable device for settling on principles of justice — only if they endorse such a philosophy. So, too, even if the formulation of a political conception is freed from objectionable sectarianism, it may still win support only from adherents to a single comprehensive doctrine or a narrow range of such doctrines. Consider an analogy: logical laws can be formulated in a freestanding way, independent of controversies in the theory of meaning. Still, certain logical laws — such as the law of excluded middle — will arguably be found compelling only by people who hold particular views in the theory of meaning — for example, that we can understand the meaning of statements whose truth or falsity transcends our recognitional capacities.110 Take the claim that people with different conceptions of the good have the capacity to choose and revise their conception, as well as a fundamental interest in circumstances that enable them to revise it should they wish. This claim is an element of the political conception of the person, and it is one of the aspects of the person known behind the veil of ignorance. In presenting a political conception of the person, Rawls shows that endorsing this claim does not consist in believing that reflectively held convictions are uniquely worthy of our full allegiance or that we are essentially choosers of ends rather than servants of God; by formulating the political conception of the person as a freestanding view, he shows that that conception does not imply any particular nonpolitical view of the person, for the content of the political conception is very different from the content of any such view. Nevertheless, it may be true that we only have good reason to accept the political conception and the associated account of justice if we endorse a comprehensive liberal philosophy of life. 109. See supra notes 94-95 and accompanying text. 110. See MICHAEL DUMMETT, THE LOGICAL BASIS OF METAPHYSICS 184-99 (1991).

John Rawls

May 1994]

A More Democratic Liberalism

403

1527

Therefore, we may be misled when Rawls says that "accepting the political conception does not presuppose accepting any particular comprehensive . . . doctrine; rather, the political conception presents itself as a reasonable conception for the basic structure alone" (p. 175; emphasis added). Even if the conception presents itself as political, accepting it may still presuppose accepting a comprehensive view if a single view provides the only reasons for accepting the political conception. Here, then, we need the idea of an overlapping consensus: the idea that all people can — for the different reasons provided by their own reasonable comprehensive moral views — think that the same conception of justice is correct and not merely an accommodation required to ensure a stable peace under conditions of moral pluralism. Rawls imagines, for example, an overlapping consensus composed of four views, each of which is reasonable and each of which provides a rationale for political liberalism: one rooted in a Kantian morality of autonomy, another in utilitarianism, and a third in a religious conception that endorses free faith, while the fourth treats political liberalism as one part of a pluralistic ethical view — a part that needs to be adjusted to the other parts, though it is not derived from them. Consider, for example, the political conception of citizens as free. How might these four views endorse the idea that citizens are free as a shared basis for political argument? One aspect of political liberalism — captured in the veil of ignorance — is that citizens have the capacity to revise their aims and an interest in favorable conditions for such revision should they wish to pursue it, but that for the purposes of an account of justice the determinate aims of citizens are irrelevant. The Kantian view accepts this aspect of political liberalism because the Kantian conceives of the reflective choice of ends as a feature of an autonomous life and holds that the protection of citizens who wish to pursue such choice is required by respect for their dignity as autonomous. The utilitarian might endorse the interest in revising aims as fundamental because true happiness — whether consisting of pleasurable feelings or the satisfaction of rational desires — depends on the possibility for such revision.111 The conception of free faith also endorses this interest because of its connections with the appropriate fulfillment of religious obligations: that such fulfillment must reflect genuine "inward persuasion of the mind."112 In short, each view accepts, for its own reasons, a conception of persons and their basic interests that provides shared ground in political argument. But an overlapping consensus on a conception of justice cannot be sustained simply by the existence of points of agreement, for points of 111. See, e.g.t MILL, ON LIBERTY, supra note 14, at 116-17. Similar considerations would support a case for the interest within a view emphasizing self-realization. 112. LOCKE, supra note 92, at 18.

404

John Rawls

1528

Michigan Law Review

[Vol. 92:1503

disagreement among reasonable views are bound also to exist. Each view implies that the others are a mixture of truths and falsehoods. Why, then, should citizens who endorse a particular moral view — who believe it to be true — not hold that political power ought to be used to advance the values of that view?113 Why should they endorse as correct a view of justice that is confined to shared ground and accept that public discussion must provide justification according to that view? Three considerations explain this restraint.114 1. It is worth emphasizing again that citizens who hold competing comprehensive views may nevertheless agree that the values incorporated within the political conception are important values and that the norms and principles included in it provide genuine reasons. From within each comprehensive view, the political conception states nothing but the truth, even if not the whole truth. As my example about the interest in favorable conditions for revising aims indicates, adherents to different moral conceptions do not think that the political conception reflects a compromise required to ensure a stable peace. Instead they believe that the conception expresses a correct account of basic political interests. 2. In accepting as correct a conception of justice that does not include the whole truth, by their lights, citizens acknowledge both the reasonableness of at least some of the views that conflict with their own and the unreasonableness of imposing arrangements whose justification depends on aspects of their own view that others reasonably reject. The Kantian, for example, rejects the utilitarian conception of the good as the satisfaction of rational desires, but he can understand the utilitarian view as an application of theoretical and practical reason, appreciate the considerations that lead to that view, and see how its endorsement is compatible with a willingness to cooperate on terms that others can accept. So the Kantian's endorsement of a political conception that contains only part of the truth — that takes political autonomy rather than moral autonomy as a fundamental value in political argument115 — is not simply a compromise required by the existence of other views. Instead, the Kantian thinks it would be wrong to impose institutions and policies justified by a political conception that is rejected by others who are themselves fully reasonable. 113. It might be said that holding a moral view is a matter of having pro-attitudes rather than beliefs that are apt to be true or false. For a sketch of the difficulties in sustaining this position, see Paul Horwich, Gibbard's Theory of Norms, 22 PHIL. & PUB. AFF. 67 (1993) (book review). But see Michael Smith, Why Expressivists About Value Should Love Minimalism About Truth, 54 ANALYSIS 1 (1994), and the reply by Horwich in Paul Horwich, The Essence ofExpressivism, 54 ANALYSIS 19 (1994). 114. Rawls mentions the first two considerations at pp. 127-28 (referring to Cohen, supra note 63). 115. See supra note 69.

John Rawls

May 1994]

A More Democratic Liberalism

405

1529

3. As the second point suggests, the key to the possibility of overlapping consensus is that a conception of justice articulates values of great importance and that the existence of a shared political conception itself constitutes an important good. I suggested a case for these claims in my earlier remarks about the good of consensus on a conception of justice.116 Suppose that case is correct and that the political consensus does articulate important values. Suppose, too, that different, conflicting comprehensive moral conceptions agree on a conception of justice. Then adherents to those moral conceptions will be able to say — each from her own standpoint — that it is normally best to uphold institutions satisfying the conception of justice, even when policies selected by the institutions are inconsistent with her particular moral conception. These conflicting moral views will also agree that it is normally best to conduct public discussion about political fundamentals in terms of the values and principles of the political conception rather than to appeal to a particular comprehensive moral view that others reasonably reject. Much here rests on "normally." Views that form an overlapping consensus will rarely, if ever, hold that political values are ultimate. For that reason, there may well be occasions when a comprehensive moral view supports the conclusion that the stakes are too high and that political values must give way. Adherents to such a view may be optimistic and see deep disagreement as an occasion for a high-stakes effort to persuade others to drop their ultimate convictions; more likely, however, they will think that the time for debate has ended. Because political values are not widely regarded as ultimate values, this kind of breakdown is always possible. To that extent the bases of civic unity are fragile: such fragility is the inevitable result of the pluralism of comprehensive moralities. Despite this fragility, one can hope that civic breakdown will not occur. More immediately, the existence of cases in which it does occur, together with the fact that we all have more to say than we are prepared to say in politics, does not imply that consensus is impossible or unattractive, or that operating on the shared ground of a political conception of justice is merely a compromise dictated by circumstance.

IV. CONSENSUS? I said earlier that the idea of consensus is likely to elicit a skeptical response,117 and I want now to explore some of the sources of that skepticism. I will consider four objections to the idea of an overlapping consensus.118 Because I find the idea of consensus attractive, I 116. See supra notes 67-74 and accompanying text. 117. See supra note 37 and accompanying text. 118. The first objection I will consider overlaps with the fourth objection discussed by Rawls

406

John Rawls

1530

Michigan Law Review

[Vol. 92:1503

will present replies to each of the objections. The four objections form a natural sequence, beginning from the thought that it is simply naive to expect consensus in a large-scale political society. The second and third objections present different variants of a common concern: that the case for consensus reveals that it can be achieved only through an objectionable exclusion of views that fall outside the consensus. The fourth objection accepts the possibility of consensus but argues that an overlapping consensus truncates political argument; by effectively taking comprehensive moral views as given, overlapping consensus forestalls the deeper agreement that might emerge from a more vigilant political criticism. A. Hopelessly Naive Consider the depth and extent of disagreement on any important political issue: from abortion and taxes to health care reform and trade policy. Against this background of disagreement the idea of consensus may strike us as hopelessly naive. This objection gains added force from Rawls's rejection of the possibility of comprehensive moral agreement. If we are prepared to exclude convergence on morality quite generally — to affirm the fact of reasonable pluralism as a "permanent feature of the public culture of democracy" (p. 36) — why should we find agreement on a political conception of justice plausible? It will not suffice to say that political agreement is more plausible than comprehensive moral agreement because matters of political justice are a proper subset of moral issues, and agreement on a proper subset is more likely than agreement on the wider set itself. Issues about abortion are a subset of the moral, but I think most of us would be nearly as surprised by consensus on the morality of abortion as by consensus about morality in general. Moreover, it is not enough simply to point to impossibility of agreement on a political conception of justice among people who have different comprehensive views. That possibility is established by the coherence of the idea of an overlapping consensus. But the coherence of that idea does not suffice to show that it is any more realistic than agreement on comprehensive moral views, which is also possible. To answer these doubts, we need a mechanism — a social or political process that might produce convergence on political values but that does not similarly generate consensus on comprehensive moral values. The right place to look for such a mechanism is at the level of shared institutions, as they might plausibly play an educative role with respect to political ideas, but not with respect to comprehensive moral at pp. 158-68, though my reply differs from Rawls's in important details. The other three objections I will discuss differ from those Rawls considers at pp. 145-58.

John Rawls

May 1994]

A More Democratic Liberalism

407

1531

conceptions.119 Before explaining this role, however, I need to make two background points. First, it is worth emphasizing that we are concerned with agreement on conceptions of justice, not with a convergence of interests. Of course, if people are moved principally by interests, then the absence of such convergence may imply that agreement on justice is not a matter of great moment. Still, the immediate issue is convergence on justice — which, after all, seems less hopeless than an absence of conflicts of interest. Second, the agreement on justice will be limited in various ways; it will not extend to all judgments of policy or even to all fundamentals that might possibly arise. In overlapping consensus, agreement on procedures and basic protections — in Rawls's terms, on constitutional essentials and matters of basic justice — suffices to make the remaining disagreements less important or less immediate. Even with these two points of clarification, it may still seem unrealistic to expect agreement on matters of basic justice, given persisting différences in moral outlook. But perhaps we can address this concern about realism if we keep in mind the institutional aspect of the acquisition of political ideas and values. Although it is implausible to expect agreement on a conception of justice to result from a convergence of practical reasoning conducted within different, independent moral traditions, it is not so implausible to expect such agreement to emerge from the acquisition of ideas and principles embodied in shared institutions.120 As I indicated in the earlier discussion of stability, Rawls's views about the development of moral-political understandings are deeply institutional.121 The acquisition of conceptions of justice proceeds via participation in institutions of various kinds — families, associations, the state. The formation of moral-political ideas and sensibilities also proceeds less by reasoning or explicit instruction — which may be important in the formation of comprehensive moral views — than by mastering ideas and principles that are expressed in and serve to interpret these institutions. The underlying idea — which traces to Rousseauean and Hegelian theories of will formation — is that people living within institutions and a political culture shaped by 119. This distinction is implicit, I believe, in Rawls's remarks on the "wide role" of a political conception "as educator." When a political conception is fully public, citizens "are presented with a way of regarding themselves [as free and equal] that otherwise they would most likely never be able to entertain." P. 71. 120. I do not mean to deny that convergence of independent traditions is a possibility; my point is that an account of political consensus should not depend on it. Bernard Williams has argued that if there were moral consensus it could not be explained by the (perspective-independent) truth of the moral beliefs on which different traditions converged. See BERNARD WILLIAMS, ETHICS AND THE LIMITS OF PHILOSOPHY 132-55 (1985). Rawls's account of the possibility of consensus on a conception of justice does not require that the truth of the conception explains the agreement on it. 121. See pp. 158-68; RAWLS, supra note 3, at 462-79.

408

John Rawls

1532

Michigan Law Review

[Vol. 92:1503

certain ideas and principles are likely to come to understand those ideas and principles and to develop some attachment to them.122 Take, for example, an aspect of the political conception of the person — the (political) idea that citizens are equals in possessing to a sufficient extent the capacity for a conception of the good and for a sense of justice. This idea is manifest in various ways in the practices and traditions of interpretation and public discussion associated with citizenship in a democracy: for example, equality before the law, or equal civil and political rights. Moreover, a stable democratic political process, in which individuals and parties seek to win support for their projects from other citizens, puts some pressure on views to endorse the idea of citizens as equals.123 We can understand how citizens quite generally might acquire an understanding of one another as moral equals by holding the position of citizen and living in a political culture in which ideas of equality associated with that position play a central role in political discourse.124 The different comprehensive views that accept this political understanding of equality will have different ways of fitting it into their broader conceptions. Some will accept political equality as following from their more fundamental moral or religious convictions; others will accept political equality as an important, nonderivative value. But what keeps the expectation of general agreement from being hopelessly naive is the plausible thought that citizens who grow up within a reasonably stable democracy will find this (self-)conception familiar and attractive: the political ideas "expressed" in common, public institutions and appealed to in the culture to justify those institutions will shape citizens* moral-political education. Of course, the acquisition of moral ideas does not proceed exclusively through institutions. So citizens will need to find or to make a place within their comprehensive views for the political ideas and selfconceptions they acquire through institutions: to find a way to combine, for example, a conception of human beings as servants of God bound by natural duties with a political conception of citizens as free, equal, and self-governing. Many views — religious, moral, philosophical — have sufficient internal flexibility or openness to make such ac122. See HEGEL, supra note 26, §§ 142-329; Joshua Cohen, Autonomy and Authority: Rousseau on Democracy 113-19 (Mar. 1993) (unpublished manuscript, on file with author). On institutional forms and the acquisition of self-conceptions, see JOHN S. MILL, REPRESENTATIVE GOVERNMENT (1861), reprinted in UTILITARIANISM, LIBERTY, REPRESENTATIVE GOVERNMENT, supra note 14, at 171, 185-202; Karl Marx, The German Ideology: Part /, in THE MARXENGELS READER 146 (Robert C. Tucker ed. & S. Ryazanskaya trans., 2d ed. 1978). 123. For further discussion, see Cohen, supra note 37. 124. Consider in this connection the virtually unanimous popular endorsement of political equality and equality of opportunity indicated in HERBERT McCLOSKY & JOHN ZALLER, THE AMERICAN ETHOS: PUBLIC ATTITUDES TOWARD CAPITALISM AND DEMOCRACY 74 & tbl. 35, 83 tbl. 3-9 (1984).

John Rawls

May 1994]

A More Democratic Liberalism

409

1533

commodations possible.125 But because political values are a subset of moral values, we have no reason to expect the accommodation of shared political values to produce a more comprehensive agreement that extends to moral values generally; no institutional mechanism in a democratic society imposes pressure to overcome fundamental differences among moral, religious, and philosophical traditions. The pressure of the shared institutions in forging political agreement ends even as considerable disagreement remains. To be sure, this explanation provides only the barest sketch of a reply to the objection about realism, but it makes an essential point that is commonly overlooked when political philosophy is understood simply as applied moral philosophy. Political ideas are institutionalized in a democratic society in ways that comprehensive moral — or religious or philosophical — ideas are not. More precisely, comprehensive ideas are institutionalized — if at all — in more particular social associations that are not shared: different churches, for example, advance different comprehensive views. So citizens acquire conflicting comprehensive views through such associations. Political ideas, by contrast, are acquired in part through shared associations. So an account of how consensus might emerge on a political conception of justice among citizens living in a political society can draw upon resources unavailable to an account of a more comprehensive moral consensus. Of course no political mechanism can guarantee agreement: the development of an overlapping consensus requires, as I mentioned, that separate traditions are each able to accommodate the political values within their view, and nothing guarantees that they are able to do so. But we are not looking for a guarantee;126 we only need a mechanism that might plausibly produce convergence of political values even under conditions of moral pluralism. Finally, given the institutional explanation, it is not surprising that the political consensus is itself limited, being principally a matter of agreement on basic political values — such as fairness, equality of citizens, and liberty, for example — rather than an agreement on a definite conception of justice. For no definite conception — no specific interpretation and balancing of the basic political values — is institutionally expressed in the way that the basic values themselves are. Of course there may be an optimal way to articulate and combine those values, and then the underlying agreement may recommend a specific conception.127 But that is a matter for further argument — for political philosophy. It is not a conclusion that is manifest from the values themselves or from their institutional articulation. 125. See pp. 159-61. 126. Recall the contrast I drew earlier between Hegel and Rawls, supra section I.B. 127. The claim that there is such an optimal way provides the basis of Rawls's argument for justice as fairness. See p. 9.

410

John Rawls

Michigan Law Review

1534 B.

[Vol. 92:1503

Unattractively Explained

Let us suppose that this explanation of the difference between the expectations of political and comprehensive moral consensus can be sustained. Then, a second objection seems natural: that the institutional explanation limits the attractiveness of the consensus it explains. An attractive explanation would see political consensus as emerging from a convergence of argument within conflicting moral and religious traditions, or perhaps from unconstrained practical discourse among adherents of separate traditions.128 In either case, political consensus would reflect the operation of reason, driving separate moral positions to common political conclusions. By contrast, the explanation I have just sketched traces the emergence and reproduction of political consensus to shared background institutions. Through these institutions, citizens acquire moral-political ideas — including ideas of person and society. Moreover, the role of the institutions is crucial, because the content of a political conception for a democratic society does not rely only on practical reason; rather, it draws also on "political conceptions of society and person** — in particular, the idea of citizens "regarded as free and equal in virtue of their possessing the two moral powers to the requisite degree/*129 Thus, the political conception of justice expresses an ideal of political deliberation and justification in a democratic society, not a more generic conception of justification through reason.130 So it is especially implausible to think that the political conception might arise simply from the work of practical reason within and among traditions. Precisely this implausibility, however, may make an overlapping consensus seem less a result of free reflection than a product of the institutional constraints under which political argument proceeds. This criticism rests on an exaggerated distinction between institutional constraint and free reflection. Recall the background assumptions: the deliberative liberties are in place — and have a fair value — and the society features a range of comprehensive views, which provide intellectual and practical elaborations of different moral, religious, and philosophical traditions. Suppose now that as a consequence of democratic institutions and the position of equal citizen within these institutions, the members of such a society acquire a shared understanding of the equality of moral persons. Suppose, too, 128. I believe that Stuart Hampshire attributes such a view to Rawls when he suggests that Rawls endorses a "myth of reason'* whose roots lie in the Platonic conception of the soul. Hampshire neglects the institutional explanation of consensus. See Stuart Hampshire, Liberalism: The New Twist, N.Y. REV. BOOKS, Aug. 12, 1993, at 43 (reviewing Liberalism). 129. P. 109; see also pp. xx, 107-10. 130. See also Joshua Cohen, Deliberation and Democratic Legitimacy, in THE GOOD POLITY: NORMATIVE ANALYSIS OF THE STATE 17 (Alan Hamlin & Philip Pettit eds., 1989). I emphasize there that a conception of reasons suited to the ideal of deliberative democracy reflects an ideal of free deliberation among equals. Id. at 22-23.

JohnRawls

May 1994]

A More Democratic Liberalism

411

1535

that citizens adjust their comprehensive views — if they have them — to accommodate this shared understanding. For example, they adjust their conceptions of flourishing and true happiness to the many directions in which citizens develop and pursue their native abilities; they adjust their conception of the conditions required for salvation to accord with the circumstances of a political society that includes citizens of different faiths; and they adjust their views of the "nature" and "proper conduct" of men and women to take account of the equality of men and women as moral persons. Under these conditions, we face strong pressure to regard the acquisition of shared ideas and the adjustment of comprehensive views as a matter of learning rather than mere inculcation via institutional constraint: how, we may ask, does the inculcation work, given a background of deliberative liberties with a fair value? Why are the shared ideas that emerge resilient in the face of challenge? Of course, we can only presume learning. Someone may be able to show how the agreement reflects power, limited information, confusion and weakness born of moral cacophony, or a deep disparity between the apparent logic of institutions and their real operation. But the presumption is significant and imposes a serious burden on those who would treat the agreement merely as a product of inculcation and constraint. Consider again the political conception of the person: in particular, the idea of the equality of citizens as rooted in their possession of a capacity both for a sense of justice and for a conception of the good. Assume that people brought up in a just, democratic society find this conception compelling, and that this is so whether their comprehensive views are secular — perfectionist, utilitarian, Kantian — or religious.131 Suppose further that considerations within their own comprehensive views support the conception of citizens as moral equals. But suppose also that citizens reflect on the fact that their traditionstfvould likely have evolved differently under different institutional conditions; had their traditions not been subjected to these particular institutions, the traditions would not now provide the resources to support the political conception. If, for example, these same citizens had been raised in a more hierarchical society, their conceptions of flourishing, salvation, and gender might not be so egalitarian. How, they might ask, could the fact that a conception of justice is rooted in the political conception of the person give any special weight to the conception of justice, given the historically contingent attractiveness of the conception of the person? The problem with this objection is that it neglects the content of 131. To be sure, important historical strands of these views have rejected the political conception of equality. But we have already rejected the idea that the political conception must emerge from the separate elaboration of competing traditions.

412

John Rawls

1536

Michigan Law Review

[Vol. 92:1503

the institutional conditions under which the political conception of the person emerges. Recall that we are assuming that the deliberative liberties of citizens are secure and that citizens have a fair chance to exercise those liberties. Though the political conception of the person does not arise through reasoning that proceeds outside an institutional setting, it must successfully withstand pressures arising from the institutionalization of deliberation itself, from freedom of expression and association, and from a fair distribution of resources.132 The attractions of the political conception of the person, then, are assumed to survive criticisms that might be directed against it. If they do survive, then how could the mere fact that people would find other views attractive under different circumstances provide a reason for rejecting the views that they do hold? The fact that citizens' views are in part institutionally explained should not lead us to think that an allegiance to them is merely a product of political circumstance rather than free reflection, given the specificity of the institutions and their role in protecting public deliberation. C Objectionably Exclusionary The third objection begins from the observation that the difficulty of achieving consensus depends on the range of positions among which agreement is sought. As this range narrows, the likelihood of agreement increases. But at the same time, concern intensifies that this narrowing requires arbitrary and exclusionary restrictions on the set of relevant alternatives. Such restrictions would of course diminish the interest of the agreement. Let us bring this observation a little closer to the ground: Rawls tells us that an adequate conception of justice must be able to win the support of "reasonable citizens who affirm reasonable comprehensive doctrines."133 Other views likely exist and ought not to be suppressed: "That there are doctrines that reject one or more democratic freedoms is itself a permanent fact of life . . ." (p. 64 n.19). But the fact that certain doctrines do not accept the political conception of justice as the correct account — the fact that they do not compose part of the overlapping consensus — raises no troubles, Rawls claims, for the justification of the political conception. If a political conception is rejected by unreasonable comprehensive views, the legitimacy of the exercise of power through institutions justified by that conception is not undermined. Reasonable comprehensive doctrines "are the doctrines that reasonable citizens affirm and that political liberalism must address" (p. 36). The difficulty should now be clear: although confining the range of relevant conceptions to reasonable views increases the likelihood of 132. On institutionalizing deliberation, see Cohen, supra note 130, at 26-32. 133. P. 36; see Cohen, supra note 63, at 281-85.

John Raw I s

May 1994]

A More Democratic Liberalism

413

1537

agreement, it also prompts concern that the label unreasonable will be used to exclude views arbitrarily — simply to ensure agreement or to silence dissent. We may state the objection as follows: If unreasonable simply amounts in the end to an abstract abbreviation for "disagrees with the dominant political conception of justice," then of course all reasonable views will support the political conception. But then the idea that an adequate conception must win the support of reasonable citizens who affirm reasonable doctrines will be of uncertain interest. If, however, reasonable is defined independently from acceptance of the political conception — say, in terms of a willingness to entertain and respond to objections — then reasonable citizens will likely affirm reasonable views that reject the political conception. To respond, I should first note that even if acceptance of a particular political conception of justice in part constituted "reasonableness," the idea of an overlapping consensus would still be of interest. Given the fact of reasonable pluralism, a political conception that could be supported on the basis of premises provided by a variety of conflicting comprehensive moral conceptions would still be desirable. Because such conceptions would be reasonable in part because of their support for the political conception, we could not construe support from competing reasonable conceptions as providing an entirely independent check on the acceptability of the conception of justice. Still, this constitutive interpretation of reasonable would permit us to make a case for the thesis that consensus on a political conception of justice is compatible with moral pluralism — that it does not require agreement on a comprehensive conception of the good. Although reasonable person is a normative notion, the constitutive interpretation of reasonable is not right. Instead, persons count as reasonable only if they are concerned to live on terms that are acceptable to others who share that same concern (pp. 48-54). In addition, they must acknowledge the "burdens of judgment": the conditions that cause disagreement among persons who affirm the importance of cooperating on terms that others can accept — that is, among persons who are reasonable in the first sense (pp. 54-56). Thus, reasonableness is defined abstractly and not — as with constitutive interpretation — in terms of the acceptance of a particular political conception. It more or less directly follows from these two features of reasonableness, however, that reasonable citizens will endorse certain basic liberties (pp. 58-61): how else could they show that they wish to live according to principles that they can justify to others, given disagreements with others that reflect the burdens of judgment? But doesn't this characterization of reasonable show that the restriction of the overlapping consensus to reasonable views endorsed by reasonable citizens is arbitrarily exclusionary? Perhaps the arbitrariness is not as transparent as the constitutive interpretation suggests. Still, the restriction may seem to provide license to define away dis-

414

John Rawls

1538

Michigan Law Review

[Vol. 92:1503

senting views as unreasonable and to exclude them from public discussion, while celebrating public consensus among the reasonable. Three points suggest otherwise. First, we need to distinguish between tolerating a view and ensuring that it forms part of the overlapping consensus. It is no crime to be unreasonable — to favor institutions and policies that cannot be justified to others — or to express an unreasonable view, nor does the endorsement of such a view have any bearing on basic rights.134 The basis for such rights as expression and association is independent of the content of one's views. Insofar as unreasonable views are "excluded," then, that exclusion is of a special kind. Second, it is a mistake to suppose that, as a general matter, dissenting views turn out unreasonable according to the account provided earlier. Consider, for example, dissident movements on the left in the recent history of this country. Why would anyone think that antiintervention movements, or movements for civil rights, racial equality, women's equality, economic justice, and gay and lesbian rights, are or were unreasonable? All these movements appeal, as a general matter, to political values in the democratic tradition. They struggle against the injustice of circumstances in which life chances are fixed by race, class, gender, or sexual orientation. Critics of these movements may disagree with the ways they have articulated democratic values, but we expect reasonable people to disagree.135 As an example of a view that is at least in part unreasonable, Rawls mentions — plausibly, I think — the position that would deny to a woman "a duly qualified right to decide whether or not to end her pregnancy during the first trimester."136 The case for the unreasonableness of this denial proceeds implicitly in two steps. First, Rawls supposes that any reasonable view will endorse and seek to accommodate three political values as relevant to addressing the issue of reproductive choice: "the due respect for human life, the ordered 134. Rawls says: "That there are views that reject one or more of the democratic freedoms is itself a permanent fact of life, or seems so. This gives us the practical task of containing them — like war and disease — so that they do not overturn political justice." P. 64 n.19. This remark does not imply that we may do whatever we judge appropriate for containing objectionable views, any more than we can fight a disease by simply quarantining people who are sick. On tolerating the intolerant, see RAWLS, supra note 3, at 216-21; on the right of subversive advocacy, see pp. 340-56. 135. Consider, to take just one example, proposals to regulate pornography in order to ensure sexual equality. See, e.g., CATHARINE A. MACKINNON, ONLY WORDS (1993) (reviewed in this issue — Ed.). These proposals appeal to political values. They do not reject the value of liberty generally, or freedom of expression in particular. Instead, they offer a particular way to combine freedom of expression and equality. Although I do not agree with these proposals, it is simply wrong to argue that they reject the value of freedom of expression or that the arguments for them rely on a particular comprehensive view. See Joshua Cohen, Freedom of Expression, 22 PHIL. & PUB. AFF. 207 (1993); Joshua Cohen, Pornography: Left (Apr. 1994) (unpublished manuscript, on file with author). 136. P. 243 n.32. This right is much weaker than the right upheld in Roe v. Wade, 410 U.S. 113 (1973), which is not confined to the first trimester.

John Raw I s

May 1994]

A More Democratic Liberalism

415

1539

reproduction of political society over time . . . and finally the equality of women as equal citizens" (p. 243 n.32). Second, he claims that any "reasonable balance" of these values wiU support the "duly qualified right" (p. 243 n.32). To deny the right is either to deny, at the first step, that the equality of women is an important political value, or to claim, at the second step, that one of the other values — say, the due respect for human life — overrides the value of the equality of women, even if we confine our attention to the early stages of pregnancy. Assume that the case for denying the right accepts the equality of women and is based on the value of due respect for human life. What prevents someone who accepts the three values from rejecting the duly qualified right as inconsistent with the due respect for human life? The problem is that people reasonably disagree about the precise content of the value of "due respect for human life." Given the complexities of the question of the status of the fetus, the conscientious rejection by many citizens of the claim that due respect for human life requires that we treat the fetus as a human person in the first trimester, the weight of the equality of women as a political value, and the importance of justification to others when such weighty values are at stake, how could it be reasonable to urge the state to endorse and to enforce the view that due respect for human life bars first-trimester abortions? Someone who rejects first-trimester abortions may reply that when it comes to preventing the murder of innocent babies, being right is more important than being reasonable. But that reply concedes the point about reasonableness, which is the only issue I am now addressing. Coming now to the third point about the exclusionary character of the notion of reasonableness: it is not arbitrary to worry only about ensuring support from the reasonable conceptions endorsed by reasonable citizens and therefore to exclude unreasonable views from an overlapping consensus. Such views do not aim to find terms that can be justified to others, and to that extent they deny the values of selfgovernment and cooperation on terms of mutual respect. Moreover, one of the reasons for seeking common ground among conflicting views in the first place — for rejecting the appeal to the truth of our own view — is that we regard it as unreasonable to impose political power on others in the name of values that they reasonably reject — even if those values are correct. So the rationale for an overlapping consensus commits us to regarding views unconcerned with common ground as unreasonable. To permit those views to shape the content of a conception of justice is to permit the content of justice to be determined by the power of those views to make themselves heard. But no attractive conception can be built around such an accommodation to power.

416

John Rawls

1540

Michigan Law Review

[Vol. 92:1503

D. Overly Accommodating The final line of criticism I wish to explore accepts the ideal of political consensus but urges that an overlapping consensus is too limited. There are several variants of this concern, but I will focus here on one that takes Rawls's idea of public reason as its immediate target (pp. 212-54). According to the idea of public reason, we should set aside comprehensive conceptions of the good in certain political settings — when discussing constitutional essentials and matters of basic justice (pp. 227-30) — and conduct political argument on the shared ground provided by political values. The criticism I have in mind rejects these limits of public reason because the constraints they impose on political deliberation prevent us from achieving a deeper level of political agreement than the idea of an overlapping consensus promises.137 To be sure, Rawls describes several exceptions to the requirement of respecting the limits of public reason — several cases in which it is permissible to appeal to a wider range of moral values than those within a political conception of justice (pp. 247-52). But limits remain. None of the exceptions mentioned in Liberalism — and none added in a recent essay modifying Liberalism's account of public reason138 — would permit citizens, in the normal course of political argument, to bring the comprehensive views of others to the surface for the purpose of criticizing those views and the political implications that flow from them. Nor does the Rawlsian view encourage or require citizens to express their comprehensive conceptions in the course of political debate with a view to opening those conceptions up to the challenge of public discourse. The account of public reason may seem, then, to undervalue the importance of forms of critical discourse that do not respect the distinction between moral and political argument and as a result to truncate politics and practical reason. This tendency might seem objectionable for two reasons. 137. "[Liberalism] forgets the possibility that when politics goes well, we can know a good in common that we cannot know alone." SANDEL, supra note 54, at 183. See also the illuminating remarks by Seyla Benhabib on the limits of liberal and discursive models of the public space in SEYLA BENHABIB, Models of Public Space: Hannah Arendt, the Liberal Tradition and Ju'rgen Habermas, in SITUATING THE SELF 89 (1992). Benhabib explores feminist criticisms of "overly rigid boundaries ... between matters of justice and those of the good life, public interests versus private needs, privately held values and publicly shared norms." Id. at 111. In the end, however, I am not sure how far her own view differs from Rawls's. Here I will note just one reason. Benhabib uses the term political discourse in a very expansive way. See id. at 104. So her concern to open up public, political discourse to more comprehensive views — both matters of justice and those of the good life — reflects her idea that such discourse "can be realized in the social and cultural spheres as well." Id. Political discourse covers debates in "cultural journals" about sexual and racial stereotyping, for example. Id. As I explain in the text, see infra text accompanying note 140, Rawls uses the terms political and public more narrowly. So he agrees that the limits of public reason do not apply to political discourse, understood in such a capacious way. See pp. 214-15. 138. See John Rawls, The Idea of Public Reason: Further Considerations (Jan. 3, 1994) (unpublished manuscript, on file with author).

John Raw I s

May 1994]

A More Democratic Liberalism

417

1541

First, actual conceptions of the good may reflect traditions of injustice. A consensus that assumes such conceptions without challenging them — putting them behind a veil of ignorance, at the basis of an overlapping consensus, or off the political agenda — is for that reason less compelling as an account of ideal justice. According to the objection, if we wish to link justice and consensus, we need a consensus that emerges from unconstrained discussion, in which we may call on people to articulate their comprehensive conception of the good, which otheis may then challenge. Second, constricting the arena of public discussion — limiting its scope to what can now be shared — perhaps excludes constructive possibilities of consensus and community that might emerge from challenging received moral traditions. Opening up the public arena by dropping the limits of public reason allows deeper challenges to existing conceptions of the good, thus permitting a more expansive consensus to emerge, if only as an ideal of reason. To clarify the point of the objection, it may help to distinguish two conceptions of the aim of critical discourse. On one view, the point is to expose unreflective assumptions, thereby freeing ourselves from illusions and a false sense of coherence and necessity. This first understanding neither expects nor hopes that such a critique will generate a new and deeper consensus in which all previous views are understood as partial versions of the truth.139 According to an alternative conception, critique serves as an instrument of reasonable consensus. Instead of taking differences as fundamental and given, it invites a more searching public debate about hidden interests, suppressed alternatives, and moral disagreements with an eye to transcending current conflicts. Here I am concerned only with the second line of thought: with the rejection of the limits of public reason in the name of possibilities of more comprehensive agreement, and a corresponding rejection of overlapping consensus for its relaxed accommodation of de facto conceptions of the good. There are two responses. First, as a matter of clarification: to affirm the limits of public reason is not to deny the importance of a more comprehensive critical discourse, in which conceptions of the good — even if reasonable — are subject to challenge, unmasking, irony, and ridicule. Protection of freedom of expression always permits such discourse, and in some settings — even political settings — it may be entirely appropriate as a way to clarify views, to change minds, and perhaps to establish deeper mutual understanding. The question is whether comprehensive critical discourse is appro139. In the legal academy, Duncan Kennedy is the great exponent of this first form of critique. See. e.g., DUNCAN KENNEDY, SEXY DRESSING ETC. (1993).

418

JohnRawls

1542

Michigan Law Review

[Vol. 92:1503

priate140 in deliberative settings that are concerned with establishing the basic terms of political cooperation in a democratic society and sanctioning the exercise of power to enforce those terms. The idea of the limits of public reason is that "political values alone are to settle such fundamental questions as: who has the right to vote, or what religions are to be tolerated, or who is to be assured fair equality of opportunity, or to hold property" (p. 214). Whatever the benefits of more comprehensive critical discourse in such settings, there is likely to be a cost. Critical discourse is likely to impede cooperation on terms of mutual respect, particularly when the views at issue are acknowledged — as I am supposing they are — to be both fundamental and reasonable. But "[m]any if not most political questions do not concern those fundamental matters" (p. 214). Accordingly, the case for limits on argument in the conduct of debate about issues such as trade policy is correspondingly weaker. Second, given reasonable disagreements, the basis for expecting that a more comprehensive critical discourse will lead to a deeper consensus is unclear, which implies that the benefits are also unclear. It appears that "difference" is a fundamental fact, as fundamental as our commonalities. People disagree deeply, and political reason appears insufficient to resolve these differences. Putting aside comprehensive metaphysical theories according to which we all are the manifestations of spirit, or religious views accessible through faith, what reason could there be for denying that there are such rationally irresolvable disagreements? Everything points to the permanence of moral disagreement, and nothing points against it: there is thtfact of disagreement and the absence of any apparent tendency to comprehensive convergence; we have no theory of the operations of practical reason that would lead us to expect convergence on comprehensive moralities; and there is no mechanism of the kind I sketched earlier in the case of political values141 that might produce agreement on comprehensive views. One might argue that differences are not so deep because adherents of comprehensive moral conceptions believe their conceptions to be true and think they can withstand rational criticism. This observation suggests a fundamental common interest — in the truth, in living according to the best conception, or in living according to a view that can stand up to rational criticism — that lies deeper than any of our substantive disagreements about which conception is in fact true or best and therefore ought to guide conduct. The availability of such abstract characterizations of common interests that underlie moral disagreements is of considerable impor140. The issue is not whether critical discourse ought to be legally permissible. The legal right must be established because of the requirement of equal basic liberties. See p. 337. 141. See supra notes 120-26 and accompanying text.

John Rawls

May 1994]

A More Democratic Liberalism

419

1543

tance and may help to secure mutual understanding and respect. It may be important for me to view people who believe that the best life is a life that comports with God's prescriptions as having the same abstract, fundamental interest as I do — an interest in knowing what is true and in living the best life — even if I cannot imagine myself believing what they believe or conducting myself as they do. We all know how complex evaluative questions are, and we can understand how people conscientiously aiming at the same target might end up in very different places. Finding deep commonalities of interest within moral differences is, then, a significant value. Nevertheless, the availability of such common ground gives us no reason for expecting a more substantial convergence on comprehensive moralities. The interests are too abstract to provide a basis for such an expectation. People with conflicting religious convictions might acknowledge one another as sharing an abstract common interest in believing the truth and in conforming their conduct to their understanding of the truth. This point of agreement might, in turn, be important in ensuring mutual respect among people with conflicting religious convictions. It provides minimal leverage, however, in resolving religious disagreement, and thus very little reason for expecting people's religious convictions to converge. Why should comprehensive moralities be any different? V. DEMOCRATIC TOLERATION AND LIBERAL UNIVERSALISM Early in this review, I described Political Liberalism as a deep and original book. I want to conclude by returning to the sources of that depth and originality, indicating their continuity with Theory. There is of course no originality in the thought that people with different views of life can live together in a political society, and there is some evidence — relatively little, unfortunately — that toleration is a practical possibility. But the defense of toleration, when it does not appeal principally to the very great practical advantages of toleration, commonly proceeds in an "exclusivist" way. What I mean is that the defense of the claim that a political society ought to permit different outlooks on life to flourish within it commonly proceeds from the perspective of one of those outlooks.142 John Locke's defense of religious toleration, for example, seems to depend for its force on a Protestant view of salvation.143 Or consider John Stuart Mill's endorsement of individuality in On Liberty, his 142. There are some exceptions. See, e.g., BRUCE A. ACKERMAN, SOCIAL JUSTICE IN THE LIBERAL STATE (1980). Ackerman emphasizes the independence of political argument from moral argument and also the many routes to liberal political arguments. See id. at 355-59. But his discussion of "four of the main highways to the liberal state" suggests that his liberalism is a partially comprehensive doctrine. See id. at 359-69. I am indebted to John Rawls for a discussion of this issue. 143. See LOCKE, supra note 92, at 17-20.

420

John Rawls

1544

Michigan Law Review

[Vol. 92:1503

powerful defense of a society featuring "different experiments of living."144 In the course of that defense, Mill urges that "[i]t may be better to be a John Knox than of an Alcibiades, but it is better to be a Pericles than either/'145 Presumably Mill thought that at least some experiments in living would proceed more in the tradition of Alcibiades and Knox than of Pericles. The non-Periclean experiments should certainly be tolerated; Mill summarizes their toleration in his "harm principle."146 His reasons for such toleration, however, reflect the Periclean perspective; they draw on a conception of human excellence with roots in the "Greek ideal of self-development/'147 Mill reveals the depth of these roots when he urges that "developed human beings are of some use to the undeveloped" and that "those who do not desire liberty, and would not avail themselves of it" may nevertheless be won to the cause of liberty because they might "in some intelligible manner [be] rewarded for allowing other people to make use of it without hindrance."148 Rawls proposes something different, which I referred to earlier as "democratic toleration." By requiring toleration as a condition for acceptable public justification, he aims to free the defense of diverse experiments of living from the outlook of one such experiment. More broadly speaking, Rawls wishes to free the democratic ideal of a shared arena of public deliberation among equal citizens from dependence on the particular ethical outlook of any subset of the public. Whether he succeeds in this enterprise is another matter, though I find the case compelling for reasons I have already presented. The point I wish to stress here is that in advancing a democratic conception of toleration, Rawls presents a sustained response to an important line of criticism of classical liberal ideas of citizen, person, reason, and public. According to the criticism, the superficial and abstract universalism of these ideas masks a much deeper parochialism. Rawls's conception of an overlapping consensus on a political conception of justice suggests a way to present those ideals as genuinely shared ground. To be sure, liberal political thought has always been self-consciously universalistic, speaking in the name of all human beings, and urging the protection of the rights and interests of all, regardless of race, class, sex, religion, or any other of the particularisms that distinguish and divide us. But critics of liberalism have vigilantly revealed the hidden (and not-so-hidden) exclusions — of, for example, class, 144. MILL, ON LIBERTY, supra note 14, at 115. 145. Id. at 120. 146. Id. at 72-73, 114, 132, 149-50. 147. Id. at 120. For interesting suggestions about the connections of this feature of Mill's view with his affection for colonialism, see Bhikhu Parekh, Superior People: The Narrowness of Liberalism from Mill to Rawls, TIMES LITERARY SUPPLEMENT (London), Feb. 25, 1994, at 11. 148. MILL, ON LIBERTY, supra note 14, at 122. For example, the undeveloped "might possibly learn something'* from the developed. Id.

John Raw I s

May 1994]

A More Democratic Liberalism

421

1545

race, and gender — that compromise liberalism's defining promise: its capacity to say "all" without quite meaning it.149 Some critics have argued that its promise is essentially compromised. For them, liberal universalism is unavoidably exclusive; its fundamental categories, such as citizen, person, public, and rights, cannot be extended to include all people without losing their definition. These critics argue, for example, that the idea of a public sphere takes shape from its opposition to a private sphere and that the distinction between public and private stands in the way of the equality of women;150 or that the abstractions that define liberal universalism require that we neglect the more concrete differences — such as class and natural endowment — that shape actual lives.151 Liberals, of course, deny that the project of liberal universalism is hopelessly compromised and that abstraction is the enemy of equality and inclusion. But denial is one thing; it is quite another to make a constructive case that liberalism can deliver more fully on the universalistic promise of its classical proponents and to abandon key elements of liberalism to ensure that delivery. Consider in this light Rawls's project in Theory and Liberalism. Theory took seriously the egalitarian critique of liberalism: the charge that the defense of liberty is a defense of the privileges of people with the wealth or status needed to make effective use of their liberty. In response, Rawls moved the idea of the social contract to a higher order of abstraction, presenting it as an agreement among free and equal persons, not among property owners, or among men, or among individuals with definite conceptions of their own advantage.152 Through this abstract reinterpretation of the social contract, Rawls made a compelling case for the view that the best version of liberalism is more egalitarian and inclusive than had traditionally been thought. In short, Rawls gave us a more genuinely universalistic liberalism, committed to "democratic equality"153 and less susceptible to charges of 149. For representative examples of such criticisms in the case of Lockean liberalism, see C.B. MACPHERSON, THE POLITICAL THEORY OF POSSESSIVE INDIVIDUALISM (1962); CAROLE PATEMAN, Feminist Critiques of the Public/Private Dichotomy, in THE DISORDER OF WOMEN 118 (1989); Joshua Cohen, Structure, Choice, and Legitimacy: Locke's Theory of the State, 15 PHIL. & PUB. AFF. 301 (1986); and Uday S. Mehta, Liberal Strategies of Exclusion, 18 POL. & SOCY. 427 (1990). 150. See BENHABIB, supra note 137, at 107-13; CATHARINE A. MACKINNON, TOWARD A FEMINIST THEORY OF THE STATE 157-70 (1989); PATEMAN, supra note 149, at 119-24; Nancy Fraser, Rethinking the Public Sphere: A Contribution to the Critique of Actually Existing Democracy, in HABERMAS AND THE PUBLIC SPHERE 109 (Craig Calhoun éd., 1992). 151. See Karl Marx, Critique of the Gotha Program, in THE MARX-ENGELS READER, supra note 122, at 525, 530*31; Karl Marx, On the Jewish Question, in THE MARX-ENGELS READER, supra note 122, at 26. For contemporary discussion of this issue as it arises in the context of distributive ethics, see AMARTYA SEN, INEQUALITY REEXAMINED (1992). 152. On the Lockean contract as an agreement among property owners, see Cohen, supra note 149. 153. Democratic equality is Rawls's term for the conception of fair distribution that includes the difference principle. See RAWLS, supra note 3, at 75-83; Cohen, supra note 38, at 727-31.

422

John Rawls

1546

Michigan Law Review

[Vol. 92:1503

class exclusion. Although Liberalism is not so concerned with the class question, it, too, aims at a more genuinely universalistic liberalism. Generalizing and deepening the ideal of toleration — by carrying it to a higher order of abstraction — Rawls offers a democratic liberalism less susceptible to charges of moral parochialism, sectarianism, and elitism and more suited to "the historical and social circumstances of a democratic society" (p. 154). By "applying] the principles of toleration to philosophy itself," political liberalism leaves it "to citizens themselves to settle the questions of religion, philosophy, and morals in accordance with views they freely affirm" (p. 154). Here we come to the heart of Rawls's work and the basis of his permanent contribution to political philosophy: he offers us a new version of democratic liberalism, marked by a commitment to liberalism's universalistic promise and a willingness to pursue that commitment by transforming those aspects of liberal thought that are condemned by its own high aspirations. Consider the common ground of Theory and Liberalism from a different angle. In his Gettysburg Address, Lincoln said that the United States was "conceived in Liberty, and dedicated to the proposition that all men are created equal," and he wondered whether a political society with such abstract devotions could "long endure."154 Perhaps such a society would be unable to make good on the promise of liberty and equality; perhaps dedication to an idea and a proposition would provide too thin a basis for stable social unity. Theory and Liberalism are the product of a life's engagement with these concerns. Theory gives us an account of what the promise of liberty and equality demands and a measure of how far we are from keeping that promise. Liberalism offers hope and a warning: the hope that we can achieve social unity in a democracy through shared commitment to abstract principles, and the warning that any political bonds thicker than these155 would, by excluding some citizens, represent yet another failure to endure. 154. Abraham Lincoln, Gettysburg Address (Nov. 19, 1863), in ABRAHAM LINCOLN, SPEECHES AND WRITINGS, 1859-1865, at 536, 536 (Don E. Fehrenbacher éd., 1989). 155. For an example of thicker bonds, see the quotation from Pat Buchanan that begins this article. See supra text accompany note 1.

[15]

DISAGREEMENTS ABOUT JUSTICE BY

JEREMY WALDRON

/. What is the relation between philosophical disagreement about the good and disagreements about justice in a pluralistic society? By "philosophical disagreements about the good," I mean differences between various views about ultimate value, various conceptions of the. nature and meaning of life. I include, most prominently, religious disagreements, but also disagreements among various secular conceptions of the good. What is the relation between differences of these kinds and the disagreements we have in politics (and in political philosophy) concerning the fundamental principles of justice? Here are a couple of models, a couple of ways of thinking about the relation between disagreements of the two kinds: (1) In the first model, each conception of the good is associated with or generates a particular vision of the just society. Catholics, for example, have a particular conception of the good, and for many that conception issues in a particular vision of law and justice, expounded (say) in the jurisprudence of Thomas Aquinas. Muslims proclaim a comprehensive religious vision, and this generates for them a particular vision of the wellordered society, a conception of law and justice which they refer to as Shari'a. Also, on this model, someone who holds a secular view of human needs may develop a conception of justice which corresponds to that view; his convictions about justice may be expected to differ somewhat from convictions of justice held by people who base them on different visions of human fulfilment. Disagreement about justice is what disagreement about the good amounts to in the social and political sphere.

424

John Rawls DISAGREEMENTS ABOUT JUSTICE

373

(2) In the second model, particular theories of justice are not seen as tied to or generated by particular conceptions of the good. Instead, they stand apart from competing religious and philosophical conceptions. They present themselves as solutions to the various problems which disagreement about the good generates in society. Conceptions of justice, on this second model, are viewed as rival attempts to specify a quite separate set of principles for the basic structure of a society whose members disagree about the good. The rivalry between competing conceptions of justice is seen as independent of (and cutting across) the rivalry between competing conceptions of the good. Thus among Catholics there are socialists and libertarians, who, although they agree about ultimate values, disagree fundamentally about the principles that should govern the economic structure of a modern plural society. In the debate about justice, a Catholic socialist may have more in common with someone who is a Marxist-Leninist and thus an atheist, than he has with a fellow Catholic who is politically conservative. The disagreement between socialist Catholics and politically conservative Catholics—or between socialists and political conservatives generally—is motivated quite separately, on the second model, from disagreements about the good. Which (if either) of these models is implied or assumed in the arguments of John Rawls's book Political Liberalism! The answer seems obvious: the second model is the one that corresponds to Rawls's view of the matter. The answer seems obvious, in part because Rawls explicitly rejects the first model as a way of characterizing his own theory, justice as fairness (JAF). JAF is not the upshot of any particular conception of the good; it is presented as a "free-standing" conception (p. 12),1 intended to represent the terms of an "overlapping consensus" among the many ethical and religious conceptions that compete for adherents in society. And Rawls presents this not only as a claim about JAF; he maintains that any view of the kind he is defending must reject the first model as a way of specifying the agenda for discussions of justice. [T]he problem of political liberalism is: How is it possible that there may exist over time a stable and just society of free and equal citizens profoundly divided by reasonable though incompatible religious, philosophical, and moral doctrines? This is a problem of political justice, not a problem about the highest good. (p. xxv)

Thus Rawls's approach to justice seems to fit the second model inasmuch as it defines a task or an agenda in whose performance people may be at

John Rawls 374

PACIFIC PHILOSOPHICAL QUARTERLY

odds in a way that (as I said) cuts cross the disagreements which they have about the good.

IL On the other hand, fitting Rawls's account into the second model presupposes that he actually contemplates disagreements about justice in a well-ordered society. In fact, it is not at all clear that he does. I mean that last statement to be as tentative as it sounds. Rawls of course does not deny that people disagree about what justice requires.2 But he does not say much about these disagreements in his own discussion. Compared with what he says about ethical, religious, and philosophical disagreements, Rawls's treatment of disagreements about justice is really quite insignificant. To see this, imagine the following way in which the second of our models could be defended. One might offer an explanation of people's disagreements about justice that differed from the explanation one offered of their disagreements about the good. The dissonance between these explanations would then yield a prediction that the two types of disagreement could be expected to cut across one another. Now Rawls devotes a lot of attention to the etiology of disagreement in society. He asks: Why does not our conscientious attempt to reason with one another lead to reasonable agreement? It seems to do so in natural science, at least in the long run. (p. 55)

Rawls uses the phrase "the burdens of judgment" as a way of articulating his answer to this question. The burdens of judgment are "the many hazards involved in the correct (and conscientious) exercise of our powers of reason and judgment in the ordinary course of political life" (p. 56). For example, he says that, on any plausible account, human life engages multiple values and it is natural that people will disagree about how to balance or prioritize them. Also, on any plausible account, people's respective positions, perspectives and experiences in life will give them different bases from which to make these delicate judgments. These differences of experience and position combine with the evident complexity of the issues being addressed, meaning that reasonable persons may disagree not openly about what the world is like but about the relevance and weight to be accorded the various facts and insights that they have at their disposal. Together factors like these make disagreement in good faith not only possible but predictable.

425

426

John Raw I s DISAGREEMENTS ABOUT JUSTICE

375

Different conceptions of the world can reasonably be elaborated from different standpoints and diversity arises in part from our distinct perspectives. It is unrealistic ... to suppose that all our differences are rooted solely in ignorance and perversity, or else in the rivalries for power, status, or economic gain. (p. 58)

Thus, Rawls concludes, "many of our most important judgments are made under conditions where it is not to be expected that conscientious persons with full powers of reason, even after free discussion, will arrive at the same conclusion" (p. 58). This account, as it stands, could characterize political as well as ethical and religious disagreements; it could characterize differences about justice as well as differences about the good.3 However, Rawls quickly orients the burdens-of-judgment argument purely in the direction of disagreements about the good. "The evident consequence of the burdens of judgment," he says, "is that reasonable persons do not all affirm the same comprehensive doctrine" (p. 60). Nowhere, as far as I can tell, does he infer the equally evident conclusion that reasonable persons cannot be expected to agree about the proper balance to be assigned in social life to their respective comprehensive conceptions. Nowhere does he infer that reasonable people might be expected to disagree fundamentally about the basic terms and principles of their association.4 One of the things Rawls takes from the burdens-of-judgment argument is that "a public and shared basis of justification that applies to comprehensive doctrines is lacking in the public culture of a democratic society" (p. 61). This provides positive evidence for attributing to Rawls the belief that issues of justice are not subject to the burdens of judgment. The argument to that effect goes as follows. According to Rawls, issues of justice are to be dealt with on the basis of public reason: "As far as possible, the knowledge and ways of reasoning that ground our affirming the principles of justice ... are to rest on the plain truths now widely accepted, or available, to citizens generally" (p. 225). To apply the burdens-of-judgment idea in this area would suggest that some of the reasons which people appeal to in articulating their views about justice are not, and cannot be, widely shared in this sense. Thus, if he is to sustain the idea of public reason as a basis for argument about justice, Rawls must deny that the burdens of judgment affect such argumentation. If the burdens of judgment preclude the use of public reason in a given area, it would seem to follow that the burdens of judgment do not apply in areas where the idea of public reason is appropriate.5

JohnRawls 376

PACIFIC PHILOSOPHICAL QUARTERLY

This leaves us with the rather uncongenial conclusion that there is no such thing as reasonable disagreement in politics. The burdens of judgment explain how reasonable disagreement is possible. But the ideal of public reason seems to presuppose that the burdens of judgment do not apply to the issues that are under discussion in politics.

///. I am reluctant to attribute this conclusion to Rawls. Surely, there would be no need for public reason if there were not disagreement about justice. If they did not disagree about justice, what would people have to reason or argue (or vote) about in a democratic society? If, however, there is disagreement about justice, and people make proper use of public reason to articulate and resolve it, it would seem churlish to deny that such disagreement was—initially, at any rate—reasonable. As far as I can tell, Rawls says that the idea of public reason is incompatible at most with the existence of reasonable disagreement about the fundamentals of justice. It is not incompatible with reasonable disagreement about the way details are worked out. Accepting the idea of public reason and its principle of legitimacy emphatically does not mean ... accepting a particular liberal conception of justice down to the last detail of the principles defining its content. We may differ about these principles and still agree in accepting a conception's more general features, (p. 226)

But while he denies that the specific content of JAF is definitive of public reason, he is unabashed about offering the general principles of JAF— "the values expressed by the principles and guidelines that would be agreed on in the original position" (p. 227)—as a criterion of whether political argument is being conducted in accordance with the idea of public reason or not. We should pause to consider how remarkable this view is. In the world we know, people disagree radically about justice. Their disagreement is not just about details but about fundamentals. There are places where Rawls acknowledges this diversity. He says near the beginning of Political Liberalism that "[w]e turn to political philosophy when our shared political understandings ... break down" (p. 44). Still, in the passages we have been discussing, his view seems to be that these differences are to be aired and debated only within the medium of "principles and guidelines ... agreed on in the original position" (p. 227). In fact, there is barely more than a handful of academic political philosophers who accept the original position idea as Rawls expounds it or his view of the principles and guidelines that would be accepted therein. It seems odd to select this

427

428

John Rawls DISAGREEMENTS ABOUT JUSTICE

377

extraordinarily controversial conception as the basis of one's view of public reason, i.e. as the basis of one's normative view about the terms in which citizens properly conduct and attempt to resolve their disagreements with one another about justice. More abstractly, it is surely a mistake to identify the norms framing the public debate about justice with values and principles which are constitutive (even if only broadly constitutive, let alone constitutive in detail) of a particular position in that debate. In the passage we have been discussing, Rawls acknowledges the controversial nature of his suggestion about the content of the criterion which determines whether people are arguing in accordance with the ideal of public reason or not. Many will prefer another criterion ... It is inevitable and often desirable that citizens have different views as to the most appropriate political conception; for the public political culture is bound to contain different fundamental ideas that can be developed in different ways, (p. 227)

Quite so. Why, then, in the face of such controversy select one of the participants as the criterion to set the terms in which the controversy will subsequently be conducted? Rawls's answer is: "An orderly contest between [different fundamental political ideas] over time is a reliable way to find out which one, if any, is most reasonable" (p. 227). As I understand it, the idea is as follows. Ex ante, there may be many apparently reasonable approaches to justice, Rawls's approach among them. A process of argument must therefore take place in political philosophy to sort out which one, if any, is acceptable as a fundamental conception for a well-ordered society. Part of what we are trying to sort out in that argument, is which one of the competing approaches to justice would be acceptable as a basis for public reason.6 If we come up with an answer, then we can say ex post that the other views are unreasonable, because they have failed as candidates for criterion of public reason. Ex ante, we can talk about reasonable disagreement concerning political fundamentals; but having settled on a view about fundamentals we are no longer in a position to talk in that way. Ex post, the only reasonable disagreements that remain are disagreements about the working out of the details of the conception that the first phase of argument has yielded. I believe this is Rawls's position. I hope he will show me that it is not. For it seems to be open to a serious objection.

John Rawls 378

PACIFIC PHILOSOPHICAL QUARTERLY

IV. Before I outline that objection,7 let me back up a little. The view I have just been unravelling is not very different from the line which Rawls took in chapter four of A Theory of Justice* There he noted—quite properly— that "the question whether legislation is just or unjust, especially in connection with economic and social polices, is commonly subject to reasonable differences of opinion."9 The existence of such disagreement necessitates constitutional procedures (such as voting) that enable decisions to be taken and carried out even when the members of the society are divided as to what the decisions should be. Now of course people disagree also about these procedural arrangements. Thus, Rawls says, the public-minded citizen addressing the justice of his society's structure faces questions of two kinds: First of all, he must judge the justice of legislation and social policies. But he also knows that his opinions will not always coincide with those of others, since men's judgments and beliefs are likely to differ especially when their interests are engaged. Therefore secondly, a citizen must decide which constitutional arrangements are just for reconciling conflicting opinions of justice. We may think of the political process as a machine which makes social decisions when the views of representatives and their constituents are fed into it. A citizen will regard some ways of designing this machine as more just than others. So a complete conception of justice is not only able to assess laws and policies but it can also rank procedures for selecting which political opinion is to be enacted into law.10

How and on what basis are proposals about political procedure to be evaluated? Rawls's answer in A Theory of Justice is that people "are to choose the most effective just constitution, the constitution that... is best calculated to lead to just and effective legislation."11 Of course, we cannot guarantee that the results of any procedure will be just: "Clearly any feasible political procedure may yield an unjust outcome."12 At best, the situation is one of imperfect procedural justice. The task is to find a constitution that will maximize the prospect that legislative decisions will be good ones. But how can citizens agree on issues of constitutional choice given that they disagree about the telos of such choice. A libertarian will seek procedures that maximize the prospect of legislation that is just by his own free market standards, while a social democrat will seek procedures that maximize the prospects for legislation embodying collective and egalitarian concern. This indicates to Rawls that disagreements about justice must be dealt with first, before issues of constitutional design are even addressed. Thus what he says about constitutional choice, i.e. about the choice of legislative process, is predicated on the assumption that we already know and already agree about the basic principles of social justice.

429

430

John Rawls DISAGREEMENTS ABOUT JUSTICE

379

In framing a just constitution I assume that the two principles of justice already chosen define an independent standard of the desired outcome. If there is no such standard, the problem of constitutional design is not well posed, for this decision is made by running through the feasible just constitutions (given, say, by enumeration on the basis of social theory) looking for the one that in the existing circumstances will most probably result in effective and just social arrangements.13

Rawls concedes that people do have "reasonable differences of opinion" about whether legislation is just; but he characterizes those as disagreements about how to apply Rawlsian principles in a complex world, not disagreements about which principles to apply. Thus the approach in A Theory of Justice is almost exactly the same as that in Political Liberalism, where Rawls suggests that political differences are to be played out in a medium of public reason that not only accommodates, but is defined in terms of, "principles and guidelines that would be agreed to in the original position" (p. 227). Both books assume that there will come a point at which reasonable politics can presuppose that participants are agreed, at least at a general level, about which principles of justice they are to apply.

V. Suppose, for a moment, that we buy this. Obviously, it would be a mistake to infer anything from it for the problems about politics, procedure and constitutional choice that we face—we, in the real world where people do not at all agree about the fundamentals of justice. To think—as we have to, in the real world—about the politics of a society whose members differ radically and in principle about what justice requires is—in Rawls's terms—to move from ideal (or strict compliance) theory to non-ideal (partial compliance) theory.14 Both books are about "a well-ordered society"—"a society in which everyone accepts, and knows that everyone else accepts, the very same principles of justice" (p. 35). The application of their arguments to a society like ours, in which people do not accept, nor do they think of themselves as accepting, the same principles of justice—a society whose politics is dedicated quite explicitly to grappling with fundamental ¿//¿agreements about justice—is quite problematic. I wonder whether, on reflection, Rawls is happy with the width of this gap between the politics of a well-ordered society and the politics of a society like ours. There are two signs that he is not. First, he himself seems quite willing to draw conclusions about American constitutional politics and law from his arguments about public reason etc. Though these inferences should be regarded as quite reckless

John Raw I s 380

PACIFIC PHILOSOPHICAL QUARTERLY

given the divergence between the politics of a well-ordered society and the politics which the U.S. Supreme Court actually has to address,15 Rawls seems unabashed about drawing them. Secondly, Rawls is in general much less comfortable in his later work with the characterization of a well-ordered society which he gave in A Theory of Justice. He now denies what he then maintained—that in a well-ordered society people affirm the same conception of justice on the same moral and philosophical grounds. The point of Political Liberalism is to argue that in a well-ordered society people can (and probably must) affirm the same conception of justice on different moral and political grounds (pp. xvi-xvii). I wonder whether Rawls might be willing to yield a further concession. Maybe what is definitive of political philosophy— the philosophy of politics, after all—is that it asks how a society can be well-ordered in its procedures for debate and decision-making when its citizens disagree, not only about the good, but also about justice. The first concession, Rawls say, was motivated by the need to come to terms with the circumstances of "[a] modern democratic society" (p. xvi). But "pluralism of comprehensive religious, philosophical, and moral doctrines" (p. xvi) is not the only pluralism with which we have to deal in a modern democratic society. We also have to deal with justice-pluralism and justice-disagreement. Maybe political philosophy should be required to come to terms with that circumstance also.

VI. Against all this and in defense of Political Liberalism, it may be argued that Rawls cannot simply confront disagreements about justice as a spectator—carefully noting the diversity of views, the extent of disagreement, etc. For he is a theorist of justice. He engages in these disagreements as a participant, and as an uncompromising opponent of conceptions other than his own. He surely cannot be required to make room, in his own normative conception of a well-ordered society, for other views about justice that are incompatible with his, views which from his point of view have simply got things wrong. We may put this defense in terms that Ronald Dworkin once used to explain why the utilitarian calculus should exclude "external" preferences held by people as the upshot of political convictions incompatible with the grounds of utilitarianism itself. Dworkin said: Utilitarianism must claim (as ... any political theory must claim) truth for itself, and therefore must claim the falsity of any theory that contradicts it. It must itself occupy, that is, all the logical space that its content requires. ... Suppose the community contains a Nazi, for example, whose set of preferences includes the preference that Aryans have more and

431

432

John Rawls DISAGREEMENTS ABOUT JUSTICE

3 81

Jews less of their preferences fulfilled just because of who they are. A neutral utilitarian cannot say that there is no reason in political morality, for rejecting or dishonouring that preference, for not dismissing it as simply wrong, for not striving to fulfil it with all the dedication that officials devote to fulfiling any other sort of preference. For utilitarianism itself supplies such a reason: its most fundamental tenet is people's preferences should be weighed on an equal basis in the same scales, that the Nazi theory of justice is profoundly wrong, and that officials should oppose the Nazi theory and strive to defeat rather than fulfil it. A neutral utilitarian is in fact barred, for reasons of consistency, from taking the same politically neutral attitude to the Nazi's political preference that he takes to other sorts of preferences.16

We may not agree with this as an analysis of utilitarianism,17 but the methodological point seems a fair one. If JAF is offered as a theory of justice, its principles must do all the work there is to be done by principles of justice. As a conception of justice, JAF is not required to be fair to, or to accommodate, its rivals. If a well-ordered society is a just society— and if Rawls is right about justice—then a well-ordered society will be one in which the principles he defends, and not any others, hold sway. Moreover, this is not a trivial matter of a philosopher wanting to stake out his own convictions over as much of the field as possible. Justice is important: in a way, nothing is more important in the basic structuring of society. To form the belief that justice requires X rather than Y, is to form the belief that nothing less than X will do and that a compromise or accommodation with Y or anything else would be pernicious. Of course a given belief about what justice requires may be mistaken. But what one would be mistaken about, is what categorically and uncompromisingly ought to be done, so far as the basic structure of society is concerned. That is what one might say in defense of Rawls's approach—an approach which characterizes a well-ordered society purely in terms of the principles of justice he thinks correct, without reference to the competition between them and what he regards as their fundamentally mistaken rivals. In Dworkin's terms, he is simply claiming truth for his own theory, and thus the falsity of any theory that contradicts it. He is— necessarily—occupying all the logical space that the content of his conception requires.18 I have no objection to this as a way of thinking about justice. But I have misgivings about it as a way of thinking about politics, certainly as a way of thinking about the politics of justice.19 What is normally understood by politics is that it is an arena in which the members of some group debate and find ways of reaching decisions on various issues in spite of the fact that they disagree about the values and principles that the merits of those issues engage. (The existence of disagreement and the felt need for decision notwithstanding disagreement are, if you like, the elementary "circumstance of politics"—as moderate

John Rawls 382

PACIFIC PHILOSOPHICAL QUARTERLY

scarcity, reasonable pluralism and limited altruism are among the "circumstances of justice.")20 The empirical science of politics is the study of the ways in which this deliberation and decision-making actually take place. Normative reflection on politics—and ultimately political philosophy—amounts to reflection on the values and principles that are implicated in processes of deliberation and decision-making. To imagine that deliberative politics (or any form of peaceful politics) is possible is to imagine that people can agree on some of these procedural points even though they disagree on the merits of the issues that the procedures are, so to speak, housing. It is to imagine, in other words, that the procedural issues and the substantive issues are in some sense separable. Now I can certainly think about politics as the partisan of a particular conception of justice competing uncompromisingly with its rivals in the political arena. But I cannot do so if my thinking about political and constitutional procedure is conducted entirely in the shadow of my substantive convictions. For me to think about politics, there must be limits on the "logical space" that my substantive views occupy. To think about politics, I must be willing at least part of the time to view even my own uncompromising convictions about justice as merely one set of convictions among others. I must be willing to address, in a relatively neutral way, the question of what is to be done about the fact that people like me disagree with others in society about justice. That concession can be a demanding one. It is not just a matter of organizing neutral rules for a debating society. Political debate must issue in decision. To engage in politics is to subscribe to procedural principles (majority-rule is an example) that might yield outcomes at odds with my own substantive convictions, outcomes that my own substantive convictions condemn. How could one contemplate this on a matter of justice! If justice is the first virtue of social institutions, how could there be any political virtue or principle that required one to support injustice simply on the ground that the other side won a vote in parliament, congress or court? If justice is the first virtue of social institutions, then surely nothing—including political procedures, including voting— is more important. This response—which I have heard form others but which I am not attributing to Rawls—rests on an impoverished conception of the dimensions of political importance. To say that view Y, which we think unjust, should prevail because it has greater political support, is not to rank political or procedural considerations ahead of justice on some single dimension of moral importance. It is to come at the issue from a different direction. When we say that a view we think incorrect should prevail on political grounds, we approach it not in terms of intrinsic importance or priority, but in light of the basic circumstance of politics—the circumstance that even on the matters we think most important, a common decision

433

434

John Rawls DISAGREEMENTS ABOUT JUSTICE

383

may be necessary despite the existence of disagreement about what that decision should be.21 The problem defined by that circumstance is the problem of selecting a substantive principle of justice to act on (together) when we disagree about which principles are true or reasonable and which not. To say then that justice is subordinated to procedural values in political decision-making would be to beg the question of which of the positions competing for political support is to be counted as just.

VIL I began by looking at the relation between disagreements about justice and disagreements about the good. Two models were suggested: on the first, disagreements about justice are the upshot of disagreements about the good; on the second model, the two sorts of disagreements are independent of one another. We have seen that Rawls is probably not, after all, committed to the second model, at least not as a model of a well-ordered society. For on his account, a well-ordered society may exhibit disagreements about the good but it will not exhibit disagreements about the fundamentals of justice. What about the kind of society with which we are familiar, in which both types of disagreement are present. Which of our two initial models is more accurate with regard to the less-than-well-ordered society like our own? Here, I do not think Rawls would want to rule out the first model, or at least some version of it. If a religious or philosophical tradition has nurtured a rich and resourceful conception of the good, it would be odd to expect its priests, ideologues, or philosophers not to have developed that conception also in a social or political direction. Social and political concerns after all are among the most pressing concerns we have: it would be odd if a tradition had views about what made life worth living but no views about the basis on which we ought to live our lives together. This seems indicated, too, by Rawls's use of the term "comprehensive" doctrine or conception to describe views about religion and value that compete in society. "Comprehensive" seems to indicate an ambition on the part of such conceptions to answer all the big questions, from which questions about justice and the basic organization of society can hardly be excluded. It is significant that one of Rawls's most prominent examples of a theory of the good is utilitarianism (p. 13): and utilitarianism certainly implies (some would say, it just is) a view about justice. But if the first model explains even a part of people's disagreements about justice, then there is an interesting consequence for Rawls's account of the transition (so to speak) from a less-than-well-ordered society, in which people disagree about the fundamentals of justice, to a well-ordered

John Rawls 384

PACIFIC PHILOSOPHICAL QUARTERLY

society in which a particular conception of justice—say, JAF—is enshrined as a framework for public reason. That consequence has to do with Rawls's distinction in Political Liberalism between an overlapping consensus and a modus vivendi. Both overlapping consensus and modus vivendi are models of the many-to-one relation between conceptions of the good and a single favored conception of justice.22 Though Rawls is adamant that they are "quite different" from one another (p. 147), he conjectures that "an initial acquiescence in a liberal conception of justice as a mere modus vivendi could change over time ... into an overlapping consensus" (p. 168). One way of understanding that change is as follows. So long as each conception of the good generates its own conception of justice—i.e. so long as each conception of the good is truly "comprehensive" in the way I described a paragraph or two ago—it is impossible for competing conceptions of the good to be related to a single conception of justice (such as JAF) in the strong moral relation that Rawls refers to as "overlapping consensus." Why? Because each conception of the good generates a direct competitor to the conception of justice which is putatively the recipient of allegiance in overlapping consensus. Thus, among a range of rival conceptions of the good, Gl5 G2, ...Gn, we are likely to find that G¡ —> J l5 G2 —> J2, and so on (where J l5 J2, ...Jn are rival conceptions of justice). Each of these pairs—Gi/J i9 G2/J2, etc—defines a comprehensive philosophical conception C l9 C2, ...Cn. Now suppose there is a liberal conception of justice, JAF, which could in principle represent a genuine overlapping consensus as between G b G2, etc. From the point of view of J b J2, etc, allegiance to JAF will be, at best, a mere modus vivendi. For Jl and J2 contradict JAF: they are rivals to JAF as they are to each other. They cannot offer JAF the sincere moral supporting that an overlapping consensus presupposes without compromising their own claims about justice. Moreover, inasmuch as Ji is an integral part of Cb J2 an integral part of C2, etc., the comprehensive conceptions Cb C2, ...Cn will not be able to support JAF in overlapping consensus either. However, if JAF can secure itself for a period of time as a modus vivendi among Q, C2, ...Cn? it may cause each of the justice-components of those comprehensive conceptions to gradually lose ground, even within its general conception. J1? for example, may gradually come to seem redundant even to the followers of Q as its work qua conception of justice is done—albeit as a modus vivendi—by the liberal conception of justice JAF. This opens up the possibility that over time, Q will decompose into its constituent parts—G{ and ^—with J! being quietly dropped, and the conception of the good, G l5 being left to forge a genuine moral allegiance to JAF. And similarly, for many of the other comprehensive conceptions.

435

436

John Rawls DISAGREEMENTS ABOUT JUSTICE

385

If this process captures anything along the lines of what Rawls has in mind, it might also provide an additional basis for characterizing certain comprehensive conceptions of the good as unreasonable, from the point of view of JAF. "Unreasonable" has two meanings in Political Liberalism. On the one hand, as we have seen, it refers to a conception whose divergence from other conceptions is not intelligible in light of the burdens of judgment. On the other hand, it refers to a conception which makes claims for itself and its adherents without regard to a fair balance between its claims and those made in behalf of other conceptions (pp. 49ff). I have argued elsewhere that these definitions of "unreasonable" can come apart.23 Here, though, I want to suggest that one way of defining "unreasonable," in the second sense, is that it characterizes any comprehensive conception, Q, that has lived with other conceptions in modus vivendi for some time without being willing to abandon (or allow to wither away) its own tendentious theory of justice, Jj.

IX. What I have just set out are some ideas on how we might think about the relation between disagreements about the good and disagreements about justice towards the end of the era or phase in which disagreement about justice remains a reasonable possibility. However, these fascinating speculations about the withering away of reasonable disagreement about justice in a well-ordered society should not blind us to the fact that full-blooded disagreement about justice remains the most striking condition of our own politics. In these circumstances, we should be very careful about inferring anything for our politics—including our constitutional jurisprudence—from the purely theoretical possibility of a well-ordered society as John Rawls understands it. Because it may encourage or license such inferences, the argument in Political Liberalism needs to be hedged around with serious reservations and qualifications. Students of political philosophy need to be made aware of how much distance there is between the sort of theorizing about justice that goes on in works like Political Liberalism and the sort of theorizing that would be necessary if we were really to try making sense of the place that politics, process and constitution have in a society like our own.

Notes All page references in the text are to John Rawls, Political Liberalism (New York: Columbia University Press, 1993). 2 See, e.g., Political Liberalism, op. cit., p. 134. 1

John Rawls 386 3

PACIFIC PHILOSOPHICAL QUARTERLY

As Rawls initially presents it, the idea of the burdens of judgment does have a political orientation. He says the burdens affect not only our estimation of the place that various ends and values have in our own way of life (p. 56), but also our assessment of the claims others might make against us. The latter is certainly an issue for the theory of justice. 4 An arguable exception is a passage towards the end of Lecture IV ("The Idea of an Overlapping Consensus") in which Rawls hazards the suggestion that "different social and economic interests may be assumed to support different liberal conceptions" and to "give rise to ideals and principles markedly different from those of justice as fairness" (p. 167). He does not however dwell on this possibility. 5 That is, Rawls seems to be committed to saying: For all areas of dispute x, the existence of the burdens of judgment in x -> the lack of public reason in x. From which it would follow: For all areas of dispute x, the existence of public reason in x -> the absence of the burdens of judgment in x. 6 That is not all we are trying to sort out, of course, For of course our most important question is: which is the better conception to govern the basic structure of society? However, I believe it is part of Rawls's general commitment to liberal principles (such as publicity or transparency) that the conception which governs the basic structure and the conception which sets the terms for public reason should be one and the same. 7 See section VI. below. 8 John Rawls, A Theory of Justice (Cambridge: Harvard University Press, 1971). 9 Ibid., p. 198-99. 10 Ibid., p. 196. 11 Ibid., p. 197. (The ellipsis indicates the presence of an additional criterion of constitutional choice: the chosen constitution must not only be effective in yielding just choices, but it must also in itself do justice to those procedural rights—such as equal political participation—embodied in the first of the two substantive principles of JAF.) 12 Ibid., p. 198. 13 Idem. 14 Cf. ibid., pp. 8-9. 15 For an argument that American justifications of the judicial review of legislation commonly rest on a grave underestimation of the extent and character of political disagreement about rights and justice, see Jeremy Waldron, 'A Right-Based Critique of Constitutional Rights,' Oxford Journal of Legal Studies, 13 (1993), pp. 18-51. 16 Ronald Dworkin, 'Rights as Trumps,' in Jeremy Waldron (ed.) Theories of Rights (Oxford: Oxford University Press, 1984), pp. 155-56. 17 Rawls himself believed that it is an objection to utilitarianism that it is precisely not scrupulous in this way about which preferences it accommodates. See Rawls, A Theory of Justice, op. cit., pp. 30-31. 18 Rawls insists throughout Political Liberalism that the appropriate thing to say in accepting JAF is not that it is "true" (and other views "false") but that it is "reasonable." (See Political Liberalism, pp. xx and 116.) The view attributed to him in the text is then the following: one cannot be expected to accommodate views that one regards as ««reasonable within a framework defined by the only set of principles that can be regarded as reasonable. 19 It might also be a grave mistake in thinking about law. If I understand him correctly in Law's Empire, Ronald Dworkin argues that one does not take law seriously unless one is willing to contemplate the possibility that some past political decisions have been based on views about justice other than one's own. To take law seriously is to be willing, in some circumstances, to keep faith with.the tenor of past injustice. This is what law-as-integrity means and requires. When Dworkin says that "[W]e are all in politics together for better or worse, [and] that no one may be sacrificed, like wounded left on the battlefield, to the

437

438

John Rawls DISAGREEMENTS ABOUT JUSTICE

387

crusade for justice overall,'* it is clear in context that "no one" means "no view about how society should be organized," See Ronald Dworkin, Law's Empire (Cambridge: Harvard University Press, 1986), p, 213, This is intriguing, in contrast to the view of Dworkin's excerpted above in the text, In that passage, however, Dworkin is talking about the internal workings of utilitarianism as a theory of justice. In the passage presently under discussion— in this footnote—Dworkin is talking about law and legality, which for him deal with aspects of a society's life that arise from the fact that people disagree about justice. See also Jeremy Waldron, "The Circumstances of Integrity," forthcoming in Ronald Dworkin's Political Philosophy, edited by Justine Burley (Basil Blackwell, 1995). 20 See Rawls, Political Liberalism, p, 66. See also Rawls, A Theory of Justice, pp. 126-30. I have said a little more about this idea of "the circumstances of politics" in "Freeman's Defence of Judicial Review," Law and Philosophy, 13 (1994), 27, at pp, 34-35. 21 In much the same way, Rawls is not required to say that justice matters more than the values implicated in one's favorite comprehensive conception; instead he should say that in thinking about justice we are proceeding along a different dimension of priority, (See Rawls, Political Liberalism, op. cit., pp, 156-58.) 22 1 should clarify this point. Modus vivendi and overlapping consensus are competing models of the many-to-one relation that might exist as between various conceptions of the good and a single conception of justice. The models set out at the start of this paper, by contrast, are models of the many-to-w¿wy relation between various conceptions of the good and various conceptions (plural) of justice. 23 See Jeremy Waldron, "Justice Revisited" (A Review of Rawls, Political Liberalism), The Times Literary Supplement, 18 June 1993, pp. 5-6, Militant Islam provides an example of a comprehensive conception whose claims are (arguably) not unreasonable in light of the burdens of judgment but quite unreasonable in the sense of openness to accommodation with other conceptions.

[16] THE MORAL BASIS OF POLITICAL LIBERALISM* Charles Larmore

T

he exchange of essays that Jürgen Habermas and John Rawls published several years ago in this JOURNAL was a long-awaited event.1 At last, the two greatest political philosophers of our time turned to comment directly on each other's conceptions of political legitimacy and social justice, the nature of citizenship, and the goals of political argument. We now know what Habermas and Rawls each believe, rightly or wrongly, are the strengths and weaknesses of the other's philosophy. On the whole, Habermas sought to accentuate their differences, while Rawls, though noting significant disagreements, tried to underscore the common elements in their views. I believe that in reality the two thinkers are far closer to one another than Habermas realizes, but that their proximity is not of the form that Rawls imagines. To me it seems that their fundamental point of convergence involves a common failing. Each of them, aiming at a similarly "freestanding" or "autonomous" conception of political life, misses its underlying moral foundations. In Political Liberalism, Rawls2 declares that a political conception is freestanding, if it looks only to the principles that should govern the political life of society. It does not present itself as applying to the political realm a comprehensive doctrine about the ends of life (PL 12). Nothing, of course, prevents such a political conception from being inte-

* I thank Daniel Brudney, David Estlund, Alessandro Ferrara, Rainer Forst, and Andrew Wallace, as well as the members of my seminar at the University of Chicago, for their help on earlier versions of this paper. 1 Habermas, "Reconciliation through the Public Use of Reason: Remarks on John Rawls's Political Liberalism" this JOURNAL, xcn, 3 (March 1995): 109-31; and Rawls, "Reply to Habermas," this JOURNAL, xcn, 3 (March 1995): 132-80. Subsequent references to these essays will be given in the text, with the page number preceded by the abbreviation 'JP'. 2 New York: Columbia, 1993. References to this book will be given in the text, with the page number preceded by the abbreviation 'PL'.

440

John Rawls 600

THE JOURNAL OF PHILOSOPHY

grated into an encompassing moral, religious, or metaphysical view of the world. Indeed, Rawls believes that his political liberalism, aiming as it does to formulate liberal principles in this freestanding way, is not fully justified unless it forms a common part or "overlapping consensus" of the different visions of the good life which are prominent in modern society. But at the same time, the basic principles on which political association relies must be ones that can be framed and legitimated without appeal to those broader considerations. I share Rawls's conviction that a liberal conception of political association should be freestanding in this sense. For liberal thought is best understood, I believe, as responding to an essential ingredient of our modern self-understanding. On fundamental issues about the meaning of life, we have come to expect that reasonable people tend naturally to disagree with one another. In early modern times, in the course of a century of religious wars, many people came to the painful realization that even with the best will in the world they would continue to differ about the nature and obligations of the true faith. Since then, this conviction has widened in scope. In a free and open discussion about the fulfilled life, the human good, or the nature of self-realization, it seems that the more we talk, the more we disagree (sometimes even with ourselves). On these matters, being reasonable—that is, thinking and conversing in good faith and applying, as best as we can, the general capacities of reason that pertain to every domain of inquiry—tends not to produce agreement but to spark controversy. Taking this experience to heart, liberal thinkers have concluded that political association should no longer undertake to express and foster a conception of the ultimate ends of human existence. Instead, it must seek its principles in a minimal morality, which reasonable people can share despite their expectably divergent religious and ethical convictions. Only so can the political world, governed as it must be by coercive rules, sanctioned by state power, come to more than just the rule of force. Only so can it enjoy the sort of transparency in which citizens recognize their political principles as the expression of their own will. Central though it is to modern experience, the phenomenon of reasonable disagreement is not easy to explain, for it runs contrary to one of the deepest preconceptions of our philosophical tradition. Why should reason, on questions of supreme importance, work not to bring us together but to drive us apart? Historical contingencies surely play an important role. There is the great variety of life experiences created by modern Western society, with all its complex divisions of labor and rich heritage of many cultural traditions. No doubt, too, the liberal

John Rawls MORAL BASIS OF POLITICAL LIBERALISM

441 601

practices of toleration and public discussion have fostered this social and cultural heterogeneity, multiplying the very circumstances to which they respond. But we might also wonder whether reason does not naturally breed discord on difficult questions. To the extent that earlier societies achieved agreement about the good, or seem to have done so, does this not reflect the limits they placed on open discussion? And though the modern sciences do display widespread convergence of opinion about complex matters, might this not be because they subject the observation of nature to forms of reasoning, based on controlled experiment and measurement, which are geared precisely toward making agreement achievable? Explaining the disharmony of reason is bound to be a speculative endeavor. Drawing out its consequences, particularly in the political realm, is more straightforward. In general, we may say, the aim of political order has always involved restraining the struggle for advantage and the violence of emotion for the sake of achieving a common good. In the past, however, this common good was typically understood as involving a comprehensive vision of human purposes, a vision which belonged to a society's most treasured knowledge, interpreted and propagated by a clerisy (often a priestly class) and sanctioned by state authority. As the modern appreciation of reasonable disagreement has grown, the fundamental tasks of political association have come to appear more complex. No longer are interest and passion alone the chief objects of concern, for the proliferation of reasonable views of the good life represents a political problem in its own right. Although people may be looking to their own interest or may be carried away by the passions of conviction when they wish to impose on others their conception of the good, they may also be acting in the disinterested belief that political life would be better if organized along its lines. It has become a distinctively modern question to ask on what basis reasonable people, thus divided, can live together nonetheless in political community. I should point out that in referring as I have to reasonableness, I do not mean quite what Rawls himself does by this term. He means by it the moral commitment to seek and to abide by fair principles of cooperation, particularly in view of the extent to which people tend naturally to disagree in their comprehensive conceptions of human flourishing; thus he distinguishes the reasonable from the rational, the latter designating the intelligent pursuit of one's own ends, apart from considerations of fairness (PL 48-58). Here, as in previous writings,31 3

As in my book, The Morals of Modernity (New York: Cambridge, 1996).

442

John Rawls 602

THE JOURNAL OF PHILOSOPHY

use the term more abstractly to mean the free and open exercise of the basic capacities of reason. Reasonable people according to my usage are therefore rational, but they may be reasonable in Rawls's sense as well, if they are concerned to reason about how best to be fair (though their commitment to fairness will not follow from being reasonable alone). It follows that reasonableness for Rawls denotes a moral response to the predicament, the profusion of disagreement about the human good, that arises from the more inclusive disposition I call being reasonable. In itself, the difference is only terminological. I do not deny that only in virtue of certain moral principles can liberalism give the weight it does to what I have defined as reasonable disagreement. So far am I from denying it that this moral basis forms the principal subject of my essay. Liberalism, I have said, entails focusing on what reasonable people can still share, despite their differences about the good life. But we would badly misunderstand its nature, if we supposed its guiding principles to consist simply in whatever might turn out to be common ground among reasonable people otherwise divided by their convictions. More fundamental than the political principles on which they will agree is the very commitment to organize political life along these lines, to seek principles that can be the object of reasonable agreement. This commitment forms the moral core of liberal thought, and it embodies a principle of respect for persons.4 Such is the main point I want to establish, as I look at the recent exchange between Habermas and Rawls and at the central question of their debate: What can it mean for political association to be founded on freestanding principles designed to abstract from the ongoing disagreements about the nature of the human good? I. CLASSICAL AND POLITICAL LIBERALISM

First, some historical background is necessary. We need to understand why this question should assume the greatest importance for liberal thought today. From John Locke's time to our own, liberal thinkers have generally presented their political philosophy in terms of a full-scale individualism, urging a critical detachment toward inherited forms of belief and 4 Despite its philosophical salience, this moral commitment does not, I believe, provide a sufficient basis for a liberal political order. It cannot explain why people who hold this commitment should think of themselves as a people, bound together in a common political destiny which distinguishes them from others. This element in their self-understanding arises from a shared historical experience, particularly the memory of past conflicts which were fueled by the effort to impose faiths or other comprehensive ideals on one another. For some more on this point, see my Morals of Modernity, pp. 141-44.

John Rawls MORAL BASIS OF POLITICAL LIBERALISM

443 603

cultural traditions. This is not surprising. Individualism has formed a pervasive current in our culture. It grows out of basic features of modern society, particularly the market institutions of a capitalist economy. It has also drawn encouragement from the modern experience of reasonable disagreement, as people have concluded that they must therefore determine on their own their vision of the human good; and that conclusion has seemed to define the basis on which they must understand the terms of their collective life. Thus, the classical liberalism of Locke, Immanuel Kant, andj. S. Mill argued that the principles of political life should forego appeal to ideals of the good because thereby they express the individualist spirit which ought to shape the whole of our life. Classical liberals differed, of course, in important ways concerning the precise nature of the individualism they embraced. But they agreed on the fundamental idea that our allegiance to any substantial view of the good life—to any concrete way of life involving a specific structure of purposes, significances, and activities (such as a life shaped by certain cultural traditions, or devoted to a particular religion)—should always be a contingent one, révisable on reflection. Such forms of life can be truly valuable, they believed, only if we understand them as ones we choose, or would choose, from a position of critical reflection. Most importantly, they were at one in defending their political principles within the framework of this general philosophy of individualism. Our status as political subjects or citizens should be independent of whatever view of the human good we affirm, because in that way political principles respect—as Locke, Kant, and Mill would have said—the fallibilist, autonomous, or experimental attitude which we as persons should maintain at the deepest level of our self-understanding. Things, however, have not stayed so simple. The individualist view of life has itself become an object of reasonable disagreement. Especially in the wake of the Romantic movement, there has arisen a new sense of the significance of belonging, an appreciation of tradition to which the premium that individualism places upon critical reflection appears to embody a kind of moral blindness.5 In reality, a reflective attitude toward received forms of life is one value among others. To give it supreme authority may therefore block recognition of much else that is also of value. Thus it has been held that we can share in the good that some ways of life offer, only if we do not think of our allegiance to them as elective, as a matter of decision, 5 For a more detailed account of this Romantic theme, see my book, The Romantic Legacy (New York: Columbia, 1996), chapter 2, as well as The Morals of Modernity, pp. 127-34.

444

John Rawls 604

THE JOURNAL OF PHILOSOPHY

but regard it instead as constitutive of what we hold to be valuable, as rooted in a feeling of belonging. The importance of common customs, ties of place and language, and religious faith can lie in their shaping the very understanding of value on the basis of which we make the choices we do. At the very least, is it not difficult to imagine our deepest moral commitments as ones that, from a critical distance, we choose or would choose to affirm? To stand back in reflection from them would seem to require stripping ourselves of the resources necessary to guide any sort of moral evaluation. Ought they not to be understood, therefore, as the inherited basis of deliberation, rather than as objects of choice themselves? The praise of belonging, though often presenting its critique of individualism as an opposition to the Enlightenment, does not involve a reversion to premodern forms of thought. The forms of life with which it encourages identification were seldom tradition-minded themselves, seeking instead legitimation in some transcendent source, such as the voice of Reason or the will of God. Paradoxical as it may sound, traditionalism is a modern innovation, and for that reason we can expect it to prove an abiding feature of our culture. Today, despite the enormous influence they continue to wield, individualist modes of thought have turned eminently controversial. No doubt they have always been contested, but now their philosophical difficulties are open to view. However we settle to our own satisfaction the merits of individualism and the sense of tradition as conceptions of the good and the right, we cannot deny that on this question, reasonable people continue to disagree. In this regard, then, liberalism faces a new challenge. Should it keep its classical commitment to an individualist view of life? Or, taking instead its bearings from that modern experience of reasonable disagreement that was in the beginning its guide, should it seek a reformulation broad enough to appeal to those who are otherwise divided by this controversy? The second path is the one pursued by political liberalism, as Rawls and I understand it. The aim is to fix the principles of political association in terms independent, not just of religious convictions and substantive notions of the good life, but of comprehensive moral conceptions, too, to the extent that espousing individualist ideals or their rejection, they have become an object of reasonable disagreement in their own right. So understood, political liberalism does not represent a radical departure from the motivations of its classical forebears. The underlying continuity is the focus on the phenomenon of reasonable disagreement, and the differences stem from experience, as we have learned how pervasive

John Rawls MORAL BASIS OF POLITICAL LIBERALISM

445 605

this phenomenon actually is. There may not therefore be much of a puzzle as to why political liberalism did not emerge earlier. It took the Romantic critique of individualism and its absorption into our culture to make plain how controversial reasonable people can find the overall moral vision to which classical liberalism has appealed.6 II. POLITICAL LEGITIMACY AND MORAL RESPECT

Historical experience provides the impetus for liberalism to shed its individualist philosophy. Nonetheless, in assuming the form of what Rawls calls a freestanding political conception, liberal doctrine is heeding more than just the widening scope of reasonable disagreement. It is also drawing upon certain moral convictions, which imply that this is the proper route to take. After all, we may intelligibly ask why liberalism's response to this controversy should be a reformulation of its principles. Why should liberalism become political, in the sense that Rawls and I intend? Why should liberal thinkers not instead dig in their heels and, observing correctly that no political conception can accommodate every point of view, maintain that liberalism stands or falls with a general commitment to individualism? The answer must be that the essential convictions of liberal thought lie at a more fundamental moral level. So we need to bring out what these convictions are. Rawls remarks that liberalism seeks the form of a freestanding political ideal, because it "applies the principle of toleration to philosophy itself." By this he means that, because its goal is a conception of justice to which reasonable people can agree, this conception must "be, as far as possible, independent of the opposing and conflicting philosophical and religious doctrines that citizens affirm" (PL 9-10). In other words, the reason why liberalism must abandon its classical appeal to an overall individualism and become a strictly political doctrine lies in the idea that basic political principles should be acceptable to those whom they are to bind. This idea Rawls has come to call the liberal principle of legitimacy: "Our exercise of political power is fully proper," he explains, "only when it is exercised in accordance with a constitution the essentials of which all citizens as free and equal may reasonably be expected to endorse in the light of principles and ideals acceptable to their common human reason" (PL

6 In giving such weight to this historical perspective, I may differ from Rawls, who finds puzzling the late development of political liberalism. See his "Reply to Habermas," TP 133, footnote 1.

446

John Rawls 606

THE JOURNAL OF PHILOSOPHY

137) .7 This principle, I agree, reflects the abiding moral heart of liberal thought.8 But I do not think that Rawls has brought out as distinctly as he might the nature of this moral foundation and the precise position it occupies in political liberalism. To see better what the liberal principle of legitimacy involves, let us ask why we believe, if we do, that the fundamental terms of political life should be the object of reasonable agreement. Answering this question will prove essential, in fact, to determining what reasonable agreement should mean in this context. Do we accept the liberal principle of legitimacy, because we think that in general people are bound only by moral rules they could not reasonably reject? Whether correct or not, this view of morality is far too controversial for the present purpose. As an explanation of the grounds of moral obligation,9 it risks circularity: Do not the reasons we have to accept the fundamental rules of morality involve the fact that it is simply wrong not to observe them? Moreover, in this form the view seems difficult to detach from the broader moral ideal of individual autonomy or self-determination toward which political liberalism must strive to be neutral. And if the validity of moral rules is instead supposed, not to be explained by reasonable agreement, but only to imply it,10 the view still remains too controversial. For there is more than a little sense in holding that a person stands under the basic obligations of morality even if (perhaps through bad living) he has come to see the world in a way that can only be indifferent to them. In any case, general conceptions of moral obligation are not, at least directly, the source of our conviction that political principles must be able to meet with the reasonable agreement of the citizens they are to govern. That conviction reflects instead the distinctive feature of political principles which sets them off from the other moral rules to which we may believe people are subject. Moral principles fall into two groups. With some we believe people can be rightly forced to comply, but others we do not regard as valid objects of enforcement, whatever dis7

For other references to this principle, see PL 139-40, 143, 217, 225-26. Following Rawls, I also agree that this principle is meant to govern chiefly the choice of basic, constitutional principles. These principles will themselves allow for many decisions to be made according to less demanding rules, such as majority voting, in which indeed appeals to controversial ideals of the good may sometimes be appropriate. 9 Such a conception of morality is presented by T. M. Scanlon in What We Owe to Each Other (Cambridge: Harvard, 1998). 10 This appears to be the basis on which Gerald Gaus argues that political principles require public justification—-Justificatory Liberalism (New York: Oxford, 1996), pp. 121ff. 8

John Rawls MORAL BASIS OF POLITICAL LIBERALISM

447 607

approval or even outrage we may feel when they are violated. The first group alone has the status of political principles. For an association is political insofar as it relies upon the supposedly legitimate use of force to secure compliance with its rules. It is this coercive character of political principles which we have in mind, when we hold with the assurance that we do, and whatever our allegiance to the general ideal of judging people only by rules they could find acceptable, that such principles must be the object of reasonable agreement. Our belief is that only so can the use of force to implement these principles be justified. This, incidentally, is why political liberalism, though claiming to be a strictly political conception, does not imagine that the political domain forms a prepackaged sector of society, inherently distinct from the other areas of social life.11 It is we who decide what will count as political, depending on what elements of social behavior we think should be subject to coercion. With all these remarks Rawls himself would concur, for at one point he notes that the liberal idea of legitimacy rests on the fact that political power is coercive power (PL 139). But our analysis of this core liberal commitment must go deeper still. We need to make clear why it is that the validity of coercive principles should depend upon reasonable agreement. I believe that the source of this conviction is a principle of respect for persons. Let me explain.12 Observe first that the use or threat of force cannot be deemed wrong in itself, for then political association would be impossible. What we must regard as improper is rather to seek compliance by force alone, without requiring reasonable agreement about the rules to be enforced. For consider the basic fact that persons are beings capable of thinking and acting on the basis of reasons. If we try to bring about conformity to a rule of conduct solely by the threat of force, we shall be treating persons merely as means, as objects of coercion, and not also as ends, engaging directly their distinctive capacity as persons. True, they cannot be moved by threats except by seeing that they have good reason to fear what we may do. But we shall be appealing to their ability to act on reasons simply in order to achieve the goals of compliance—the establish-

11 Habermas makes this criticism of Rawls's political liberalism, wrongly it seems to me, in "Reconciliation through the Public Use of Reason," JP 129. 12 Here, I rely on the argument laid out in The Morals of Modernity, pp. 136-41.

448

John Rawls 608

THE JOURNAL OF PHILOSOPHY

ment of public order, perhaps also the reformation of people's character. We shall not be engaging their distinctive capacity as persons in the same way we engage our own, making the acceptability of the principle depend on their reason just as we believe it draws upon our own. Thus, to respect another person as an end is to require that coercive or political principles be as justifiable to that person as they presumably are to us. This is certainly not the only sense we can give to the rich moral notion of respect.13 But it is the one which liberals must regard as relevant from a political point of view. (In this regard, the extent to which it avoids the peculiarities of the Kantian conception of respect is an advantage.) III. THE MORAL FOOTING OF A FREESTANDING POLITICS

Liberalism, formulated as a strictly political doctrine, rests therefore on this moral foundation. It forms a freestanding conception in regard to comprehensive moral visions of the good life, but it cannot coherently claim to be freestanding with respect to morality altogether. In particular, we would be wrong to suppose that the moral principle of respect for persons has the political significance it does because reasonable people share a commitment to it. On the contrary, the idea of respect is what directs us to seek the principles of our political life in the area of reasonable agreement. Respect for persons lies at the heart of political liberalism, not because looking for common ground we find it there, but because it is what impels us to look for common ground at all. These reflections lead, I believe, to a very important result. As citizens of a political association organized in accord with the liberal principle of legitimacy, we cannot regard the norm of respect as having the same sort of validity as the constitutional principles by which we live. Those principles are legitimate in virtue of being the object of reasonable agreement, and thus their authority is strictly political, deriving from our collective will as citizens. Or at least so they are un-

13 Thus, as William Galston points out—Liberal Purposes (New York: Cambridge, 1991), p. 109—we respect a person in a different sense when explaining to him fully our reasons for the principle by which we judge his conduct, whether or not he can appreciate those reasons. My point is not to settle what respect "really" means, for it really means a great many different things. My concern is with what it ought to mean when figuring at the foundations of liberal thought. Also, I agree with Rawls—A Theory of Justice (Cambridge: Harvard, 1971), pp. 585-86—that the relevant notion of respect is best circumscribed by referring to characteristic liberal principles (such as the liberal idea of legitimacy). But our aim ought then to be understood as determining the sense of respect for persons on which those principles rest.

John Rawls MORAL BASIS OF POLITICAL LIBERALISM

449 609

derstood for political purposes (for when we refer to our global visions of the good, we may also attribute to them an independent status) . But even in our capacity simply as citizens—relying solely on the commitments we share as members of a political community—we cannot take this view of the principle of respect. It must instead be understood as having more than just political authority. We must consider respect for persons as a norm binding on us independent of our will as citizens, enjoying a moral authority that we have not fashioned ourselves. For only so can we make sense of why we are moved to give our political life the consensual shape it has. It is not clear to me that Rawls understands his own thought in this way. To be sure, he thinks of his political liberalism as a "moral conception" (PL lln). In holding that liberal principles should be formulated as a strictly political conception, he does not mean to suggest any contrast between the political and the moral, as though liberal doctrine did not form a normative conception, as it manifestly does, consisting in certain ideals, standards, and values. But would he be willing to admit that, as citizens reasoning from the standpoint of this political conception, we must acknowledge a moral authority higher than the political principles we give ourselves? No unambiguous answer is available. About the liberal principle of legitimacy, Rawls says that it has the same basis as his two principles of justice: it would be chosen in what he calls the "original position," being indeed bound up in those very two principles (PL 137n, 225-26). In other words, Rawls seems to regard that principle of legitimacy as one whose validity, at least from the political point of view, depends on our collective will as citizens, whereas I of course have been arguing for just the opposite conclusion. Yet Rawls also notes that the original position, which is but a device for representing the basis on which we understand ourselves as choosing or imposing on ourselves political principles, incorporates certain values in the conditions under which such a choice is to be imagined taking place, values which thus are not themselves the object of choice (for example, PL 103). Although we do not, Rawls stipulates, imagine the parties choosing in the original position as endowed with any moral sensibility (they are merely rational, engaged in the efficient pursuit of their ends), the fact that we place certain conditions on their choice—they are to be supposed ignorant of their own talents or conception of the good life, for example—reflects a moral commitment of our own, namely, a commitment to what he calls reasonableness, the readiness to seek fair principles of cooperation (PL 305). Political principles apparently have then a moral basis that we

450

John Rawls 610

THE JOURNAL OF PHILOSOPHY

cannot conceive as rooted like them in our political will. Moreover, Rawls describes this moral commitment of reasonableness as embodying the resolve to propose political norms justifiable to all (PL 49, 50); and this brings us right back to the liberal principle of legitimacy. That principle cannot therefore have the same status as the two principles of justice; and, as it expresses in effect the idea of respect for persons, Rawls would apparently agree after all that this idea must have a moral authority for citizens that is independent of their political will. Yet Rawls never makes this point explicitly. On the contrary, his stated view is that because of its freestanding character, political liberalism is "doctrinally autonomous" (PL 98-99). At least in our role as citizens (if not in our comprehensive views of the world), we are to regard our political principles as deriving their validity from our political will. These principles cannot accordingly be understood, he holds, as "moral requirements externally imposed" (PL 98). This phrase can have various meanings, however. If it means that basic political principles are not to be imposed upon a citizenry by some external agency—by an enlightened monarch, for example—then I agree.14 But if instead it means that citizens should not regard their political principles as drawing upon moral requirements whose validity is external to their collective will, then I believe Rawls goes wrong. Political liberalism makes sense only in the light of an acknowledgement of such a higher moral authority. In fact, the idea of respect plays this foundational role in several ways. As we have seen, it forms the basis for believing that political principles should be the object of reasonable agreement. But, in addition, it serves to define the very nature of the agreement to be sought. To explain this point, I begin by observing that consensus in this case is clearly a normative notion if only because it refers to reasonable instead of simply actual agreement. Political life is to be based on principles that citizens, despite their various moral, religious, and metaphysical beliefs, can see reason to accept, exercising (as I have said) the basic capacities of reason applicable to any domain of inquiry.15 But reasonableness (so understood) is not the

14

In this connection, see Rawls, "Reply to Habermas," JP 162-63. The reasonable so understood does not entail uncritical deference to common sense and to the formal fallacies of reasoning which it may harbor or to the refusal to admit the existence of reasonable disagreement about the good which it may show. Contrary to Gaus—-Justificatory Liberalism, pp. 3-5, 131-36—political liberalism as I conceive it (or, I believe, Rawls as well) is not populist in this sense. 15

John Rawls MORAL BASIS OF POLITICAL LIBERALISM

451 611

sole normative criterion defining the sort of consensus in question. Because the idea of respect directs us to look for common ground in the first place, it should be understood as a further condition which acceptable principles must satisfy. The terms of political association are to be judged by reference to what citizens would accept, were they reasonable and committed to the principle of equal respect for persons. The notion of agreement to which political liberalism appeals is therefore an idealization. It comes into play only within the bounds set by these two norms, the one epistemic, the other moral. This conclusion is, again, one which Rawls should be willing to admit. His liberal principle of legitimacy ties consensus to precisely these two assumptions when referring to what citizens in the light of "their common human reason" may be "reasonably expected to endorse." (Recall that his notion of the reasonable embodies the principle of respect.) Yet he is not as clear as he should be about the moral foundations of this idea of consensus. IV. WHAT HABERMAS AND RAWLS SHARE

Habermas, too, aims to develop an autonomous conception of the guiding principles of modern democracy. This ambition inspires the defense of radical democracy in his book, Faktizitát und Geltung: Beitráge zur Diskurstheorie des Rechts und des demokratischen Rechtsstaates.™ The democratic ideal as Habermas understands it has nothing to do with the communitarian dream that our political life might once again draw its bearings from a common vision of the human good. In this respect, it shares the outlook of political liberalism. But Habermas also opposes it to what he calls the liberal idea that individual rights set limits to the exercise of democratic selfrule. Liberal thinkers err, he believes, in thus holding political association accountable to moral norms supposedly given in advance. For they thereby fail to heed the intellectual conditions of modern times. In an age where religious and metaphysical worldviews have lost their authority, we can have reason to consider ourselves subject to political principles only if we are able at the same time to see ourselves as the authors of these principles:

16 Suhrkamp: Frankfurt, 1992, p. 13; English translation by William Rehg: Between Facts and Norms: Contributions to a Discourse Theory of Law and Democracy (Cambridge: MIT, 1996), p. xlii. Subsequent references to this work will be given in the text, with 'FG' preceding the page in the German original and 'BFN' the corresponding page in the translation. In quotations, I have changed the English translation as I saw fit.

452

John Rawls 612

THE JOURNAL OF PHILOSOPHY

Without religious or metaphysical support, the coercive law designed for legal behavior can preserve its socially integrative force only insofar as the addressees (Adressaten) of legal norms may at the same time understand themselves, in their collectivity, as the rational authors ( Urheber) of those norms (BFN 33; FG 51-52).

In Habermas's view, this kind of collective autonomy (to be distinguished, obviously, from the ideal of individual autonomy, though the two are structurally similar) constitutes the source of the basic norms shaping political association. We miss the true character of the principles of modern democracy, he claims, if we adhere to the "classical" (that is, premodern) notion of a normative hierarchy whose pinnacle is occupied by independent moral principles to which political association must defer (FG 137; BFN 106, 449). Because liberal thought, subordinating democratic self-rule to individual rights, continues to rely on this hierarchical model, Habermas believes that we must move beyond a liberal conception of political life. We must be resolutely modern, recognizing that political principles have their basis only in the autonomous will of citizens who are the full authors of the rules that bind them. Habermas often describes his position as one in which fundamental individual rights, no longer serving to check democratic self-rule, turn out to be co-original (gleichursprünglich) with it. In other words, the modern and ancient ideals of liberty, as those two principles have come to be called, are not so open to conflict as commonly supposed.17 Sometimes Habermas's idea seems to be that the two mutually support one another. Just as self-government can serve to protect individual rights, so these rights themselves provide the necessary means for the exercise of popular sovereignty—the right to free expression, for example, making available the wide range of information and points of view that permit wise political decisions (FG 155, 161; BFN 122, 127-28). But, in reaching for the idea of co-originality, Habermas has clearly a deeper point in mind as well. It is that the two principles have a common origin, and that source is, he believes, the autonomy by which a community gives shape to its political life. The result is that, despite his talk of mutual implication, Habermas unmistakably makes democratic self-rule a principle prior in status to that of individual rights. The claim that rights and self-rule are co-original is 17 See Habermas, Die Einbeziehung des Anderen (Frankfurt: Suhrkamp, 1996), p. 298; English translation by Ciaran Cronin and Pablo De Greif: The Inclusion of the Other: Studies in Political Theory (Cambridge: MIT, 1991).

John Rawls MORAL BASIS OF POLITICAL LIBERALISM

453 613

misleading, in light of what he takes to be the only plausible origin of political norms in the modern age. In reality, popular sovereignty functions for Habermas as the ultimate basis on which our political life should be organized, and so as the true source of individual rights, their ultimate justification lying in the way they embody the principle of self-rule (cf. FG 134-35; BFN 104). The reason is plain: only so can we escape letting the principles of political association draw upon moral values having an independent validity. (I come back to the details of Habermas's argument in section vil.) A striking similarity thus emerges between Habermas's conception of autonomy and Rawls's demand that political principles be freestanding. Indeed, as I have remarked, Rawls himself calls a freestanding political conception "doctrinally autonomous" and explains this term in much the same way. A political conception is autonomous, he writes, if "the political values of justice and public reason (expressed by their principles) are not simply represented as moral requirements externally imposed" (PL 98). In their companion essays in this JOURNAL, both philosophers emphasize this point of convergence (JP 127, 150). But it is also obvious that Habermas's idea of political autonomy is the very notion that I have criticized in my discussion of Rawls. Political association, so we have seen him claim, must not be understood as drawing its principles from a moral source superior to its collective will. In his hands, this conception fares no better than it does in Rawls's. Habermas, too, misses the moral basis that supports the democratic ideal of self-rule. To make clear where precisely his mistake lies, I shall first look at some disagreements between the two philosophers. In their debate in this JOURNAL, they both noted important differences in their handling of this common idea. Rawls rejected the manner in which Habermas arrives at his conception of political autonomy. Habermas, in turn, charged that Rawls strays from a consistent development of such a conception. Understanding these points of divergence will put us in a better position to see why, contrary to them both, modern democracy rests upon independent moral foundations. V. METAPHYSICS AND POLITICS

As Rawls observes (JP 135-38), political autonomy in Habermas's eyes stems from a broader philosophical position described as postmetaphysical In our time, Habermas declares, metaphysical theories asserting the existence of entities that are neither physical nor psychological in nature ("ideal entities" he calls them) as well as religious conceptions that see the world as the work of God and the

454

John Rawls 614

THE JOURNAL OF PHILOSOPHY

vehicle of providence have lost their plausibility. His point is not simply that they no longer enjoy the society-wide authority they once possessed, having become the object of apparently irresoluble disagreement. For Habermas, the age of metaphysical and religious worldviews is over in the deeper sense that, properly speaking, rational belief in their truth is no longer possible. Reason itself, he believes, now presents itself as essentially finite, fallible, procedural, and oriented toward intersubjective agreement. It consists in thinking or acting on the basis of reasons we take to be valid, not because we imagine them guaranteed by an objective order of the world, but because we suppose that they would command the assent of others under appropriately ideal conditions. We are to locate the norms of reason, no longer in an ideal realm independent of us, but instead in the idealizations we necessarily make in speaking with one another.18 A battery of philosophical arguments, assembled under the title of discourse theory, aims to bring out the necessity of this postmetaphysical shift in our conception of reason and to explain the idealized form of discussion (Diskurs) to which reason thus refers. Habermas claims that it is in virtue of this general standpoint that the idea of political autonomy recommends itself. We are to establish the terms of political association by applying this general account of reason to the fundamental task of politics, which is to determine the scope of the rule of law. Thus, once reason can invoke no higher authority than agreement under ideal conditions, the principles of conduct that are to bind us must be such that we can see ourselves to be their source. This is the meaning of the passage I began by quoting in section rv. Rawls takes his distance from this approach, and rightly so. Habermas's argument has the significant disadvantage that it relies on a comprehensive vision of our place in the world and the nature of reason's authority. His postmetaphysical point of view shares, as Rawls insightfully notes, the ambition of G. W. F. Hegel's logic, providing "a philosophical analysis of the presuppositions of rational discourse...which includes within itself all the allegedly substantial elements of religious and metaphysical doctrines" (JP 137). Such doctrines will scarcely recognize, much less accept, the form into which they are thus translated, for it is one in which, as Habermas himself remarks, "worldviews are measured more by the authenticity of the life styles they shape than by the truth of the 18

There is a succinct and lucid account of this line of thought in FG 24-37; BFN9-21.

John Rawls MORAL BASIS OF POLITICAL LIBERALISM

455 615

statements they admit" (JP 126). This idea of postmetaphysical reason can be expected therefore to provoke controversies of precisely the sort that we must look beyond, so Rawls believes (and I agree), if we are to discover acceptable terms of political association. After all, the notion of ideal discussion to which Habermas appeals is not self-explanatory, and the transcendent interpretation he favors, defining ideal conditions as he does by reference, not to the ("local") standards of belief we currently espouse, but to the best standards we ever could have (FG 30-31, 36, 202, 566; BFN 15, 20, 163-64), is neither the only possibility nor one likely to strike many people as anything short of metaphysical. In truth, the vitality or obsolescence of metaphysical and religious worldviews is an issue on which reasonable people tend naturally to disagree. It is fair to observe that today none of these conceptions can prove authoritative for society as a whole. Yet to go further and say, as Habermas does, that they can no longer figure as objects of rational belief, but only as life styles, means adopting a point of view that itself is party to these disputes. Habermas has misidentified the feature of modern experience that is crucially relevant to the basis of political association. The decisive element is not the waning of metaphysical and religious worldviews (though that has occurred). It is instead the recognition that such worldviews, as well as the recurrent postmetaphysical efforts to do without them, are an enduring object of reasonable disagreement. As a matter of fact, Habermas's idea of postmetaphysical reason seems to me to go wrong in two important respects.19 First, I believe that our conception of the world must have room for ideal entities, for only if reasons exist (a reason being itself neither physical nor psychological in character) can there be such a thing as normative knowledge—knowledge of how we ought to act, but also more fundamentally knowledge of how we ought to think. Second, I do not think that the notion of ideal discussion forms a substantial part of the theory of justified belief. We may say, if we wish, that justified beliefs are ones to which we assume that all would agree in an "ideal" discussion. But we can have a grasp of such ideal conditions only if we suppose them to embody our best standards of warranted belief, and so we might as well say, more directly, that the beliefs are justi-

19 For more on these two points, see my Morals of Modernity, chapter 5 and pp. 206-10.

456

John Rawls 616

THE JOURNAL OF PHILOSOPHY

fied if they satisfy these standards.20 In Habermas's account of justified belief, it is true, ideal discussion plays a role that resists elimination of this sort, since it is assumed to transcend our current standards. But in consequence it loses any discernible content. Yet I should emphasize that the main point in the present context is not whether Habermas's conception of reason is correct, but whether it can serve as an appropriate basis for establishing the terms of our political life. These two questions must be distinguished, once we are convinced that political association turns crucially on finding principles reasonable people can accept despite their disagreements about fundamental matters of life and value. In seeking a solution to this political problem, we cannot call upon all that we ourselves may reasonably believe to be true about such matters. Habermas gives little attention to the difference between these two perspectives. That no doubt is why he runs them together. The result is an account of the foundations of political life which, whatever its interest, and whatever the ultimate validity of the conception of reason on which it relies, is destined to be an object of reasonable disagreement. In this regard, Habermas's political theory falls short of what we need. VI. HABERMAS'S IDEAL OF POLITICAL AUTONOMY

Habermas himself locates the crucial difference in their accounts of political autonomy elsewhere. In his view, Rawls fails to adhere consistently to this principle in working out his theory of justice, reverting to the liberal assumption that there exist individual rights prior to democratic self-rule and defining its scope. This backsliding, he charges in the exchange in this JOURNAL, arises from the "two-stage character of his theory" (JP 128). While Rawls's idea of the original position, in which rational parties choose principles of justice, captures something like collective autonomy, it yields a conception of the just society in which "liberal rights...constrain democratic selflegislation" (JP 128-29). Citizens cannot then "reignite the radical democratic embers of the original position in the civic life of their society" (JP128). Behind this objection there may well lie, as Rawls suspects (JP 160), some sympathy with the quixotic idea (once famously voiced by Thomas Jefferson) that every generation should give itself its own

There is a similar criticism in Albrecht Wellmer, Ethik una Dialog (Frankfurt: Suhrkamp, 1986), pp. 70-80; English translation by David Midgley included in The Persistence of Modernity: Essays on Aesthetics, Ethics, and Postmodernism (Cambridge: MIT, 1991). 20

John Rawls MORAL BASIS OF POLITICAL LIBERALISM

457 617

constitution, unfettered by the past decisions of those now dead. But I believe that the gist of Habermas's complaint has little to do with a call for permanent revolution. It concerns instead the understanding we should have of the basis of our political association. His view is that citizens must be able to see all their political principles, even those establishing individual rights, as rooted in their autonomous political will. Such a self-understanding is blocked by a conception that gives rights, as he believes Rawls's theory does, a status prior to the democratic process. This is the point at which liberalism and Habermas's own theory of radical democracy supposedly part ways. What is the force of Habermas's criticism? I am inclined to think that he terms inconsistency what I have called ambiguity. Habermas objects that Rawls fails to carry through consistently an autonomous or freestanding conception of political life. But perhaps another way to put the point is to say, as I have argued, that Rawls's idea of a freestanding political conception is ambiguous. Sometimes he implies and sometimes denies that his strictly political doctrine rests on the basis of a moral principle of equal respect. To the extent that he admits this truth, his understanding of a freestanding political conception does not really correspond to what Habermas calls autonomy, which excludes this sort of normative hierarchy. In any case, it is clear that Habermas and I take opposite views of these tensions in Rawls's thought. But there is a further complication. Habermas's own political theory does not hold consistently to the conditions he assigns to the notion of autonomy. In fact, his conception of political autonomy makes sense only on the assumption that self-rule rests on an underlying, moral norm of respect. Nor is that astonishing. For only on such a moral basis can popular sovereignty take the form that we prize as modern democracy. Moreover, this moral foundation involves, contrary to Habermas's stated intention, an idea of individual rights that precedes and defines the exercise of self-rule. In the end, therefore, the sharp opposition he constructs between liberalism and his own conception of radical democracy, and thus the criticism he makes of Rawls, lose their raison d'être. Habermas, I believe, misunderstands himself. It is to developing these points that I now turn. To begin, let us examine more closely the way Habermas understands political autonomy. In explaining how the autonomy characteristic of modern democracy must proceed without relying upon any antecedent moral norms, Habermas appeals to what he calls discourse principle D:

458

John Rawls 618

THE JOURNAL OF PHILOSOPHY

Just those action norms are valid to which all possibly affected persons could agree as participants in rational discourses (Diskurseri) (BFN 107; FG138).

Modern democracy has its basis in the collective autonomy of its citizens, Habermas declares, because in determining the principles by which they will govern themselves, they defer to no pregiven norms other than just this principle D. "Nothing is given prior to (vorgegeberi) the citizens' practice of self-determination (Selbstbestimmungspraxis) other than the discourse principle" (BFN 127-28; FG 161). In contrast to liberal theories, which appeal to antecedent moral norms (in the form of individual rights) to limit the scope of self-rule, radical democracy, as Habermas conceives it, looks to no higher authority than its own self-legislative activity as governed by principle D. Rights guaranteeing individual freedoms (such as speech and assembly) have their basis, he claims, in the application of the principle to the very form of a legal system, whereas the rights of political participation serve to make possible the institutionalization of this sort of self-rule (FG 154-62; BFN 121-28) .21 My concern lies, not with the details of this "logical genesis of rights," but with the status of D itself. It is crucial for Habermas's conception of radical democracy, as he makes clear (FG 138; BFN 107), that principle D possess no moral content of its own.22 Morality and democracy, he holds, involve distinct and equally basic applications of principle D, the difference consisting in whether the principle is to shape the interactions between human beings as such or is to take a legal form binding for a specific political community. Certainly, there is a significant overlap between the norms thus generated in these two domains. But in Habermas's view, this does not change the fact that the terms of political life have their justification solely in the autonomous will of the citizens, as expressed according to principle D, and not in any moral principles that citizens must recognize as setting limits to their will. This is the point, however, where Habermas displays an incomplete understanding of his very devotion to democratic self-rule. Observe that popular sovereignty itself may take different forms, depending on how the popular will is supposedly determined and given authoritative expression. Even modern dictatorships, since un-

See also Habermas, Die Einbeziehung des Anderen, pp. 299-301. As Habermas observes (FG 140; BFN 108), this thesis of Faktizitat und Geltung represents a departure from his earlier writings. I think it is a wrong move. 21

22

John Rawls MORAL BASIS OF POLITICAL LIBERALISM

459 619

like premodern monarchies they disclaim dynastic or divine legitimacy, are a phenomenon of the modern democratic age: they too draw their authority from the will of the people, as they choose to understand it.23 When conceived as heeding a principle such as D, popular sovereignty takes on a form distinctive precisely by virtue of its moral character, for all that Habermas may protest to the contrary. Z), he admits, possesses "normative" content, even if it is not a moral principle (FG 138; BFN 107). That is surely a feeble distinction. D has moral content, and we can bring it to light by asking the simple question: Why should we believe, as this principle requires, that norms of action must be rationally acceptable to all whom they are to bind? Whence arises the authority of principle D itself? VII. DEMOCRACY AND LIBERALISM

Habermas himself never provides an adequate answer to this question. He suggests that such a principle is inevitable, once we realize that in our age metaphysical and religious conceptions, characteristically inclined to invoke principles of conduct of which we are not the source, have lost their rational warrant. But this response will not work on a variety of counts. As I have already indicated, it is rash to imagine that reason today stands uncontroversially opposed to belief in the truth of metaphysical or religious worldviews. A more accurate diagnosis of our situation is that such visions can no longer enjoy reasonable agreement within society as a whole. We might well take this fact to entail that metaphysical and religious conceptions should not figure among the principles that bind us together politically. But note that such a conclusion is valid, only if we already accept some principle such as D; it cannot serve to justify that principle. Observe further that, if we refuse a political role to metaphysical and religious worldviews on some other grounds—because, for example, we believe, as Habermas does, that they can no longer be rationally held to be true, we shall not thereby have reason to adopt principle D. Why should we not suppose that without appeal to such ultimate truths the problem of political association is insoluble? Or why should we not organize our political life around some principle other than D, such as the maximization of the general welfare? Sometimes Habermas claims that the principle enjoys a privileged status because it inheres in the very idea of discussing with others the

23 This point was rightly made, however disingenuously for his own purposes, by Carl Schmitt, The Crisis of Parliamentary Democracy (Cambridge: MIT, 1985), p. 28 (German original: Die geistesgeschichtliche Lage des heutigen Parlamentarismus (Berlin: Duncker and Humblot, 1923 and 1926)).

460

John Rawls 620

THE JOURNAL OF PHILOSOPHY

norms to which our actions should conform, that idea being the only common resource we can count on in an age of controversy.24 This suggestion continues Habermas's longstanding interest in basing morality on "universal pragmatics." I doubt, however, that so much can be gotten for so little. D is too substantial a principle, and precisely because of its significant moral content, to be implicit in the mere idea of practical discussion. A good way to see this is to reflect that, taken as a general principle applicable to all domains of life, D is not even obviously correct. Of course, the very meaning of D is obscure so long as the notion of "rational discourse" is left undefined, and Habermas's own wish to understand that notion in abstraction from any specific standards of belief does not help. But however we explicate the idea of rational agreement, principle D in its general form faces the sort of objections I mentioned in section II. Thus, many people believe indeed that we should not judge others by rules that on due consideration they would not themselves accept. But many, too, believe the opposite. They are convinced that certain fundamental norms of conduct, possessing an objective validity, can serve as a basis for judging others, whether or not they meet with agreement. I myself subscribe to such a view. But what should be clear is that the general validity of a principle such as D is a matter on which reasonable people disagree. Equally plain is the fact that this disagreement turns on different moral convictions about the conditions under which we may judge others morally and no doubt, too, on different appreciations of the moral ideal of individual autonomy. This suffices to show that contrary to Habermas, D, taken as a general principle, has a moral content and a controversial one at that. But let us go further and look at D in the political role it is supposed to play. Recall a point made earlier: norms of conduct are of two kinds. There are, on the one hand, norms that we invoke solely to judge the conduct of others, and, on the other, norms that we believe should be backed up by the threat and use of force to ensure compliance. So let us consider principle D as applied simply to norms of the second kind, that is, to political principles. Now its complexion is rather different. It need not be rejected by someone who believes that we may rightly judge others in the light of norms that they themselves would not accept. For it requires only that the prin-

24 See, for example, most recently Habermas, Die Einbeziehung des Anderen, pp. 58-59.

John Rawls MORAL BASIS OF POLITICAL LIBERALISM

461 621

ciples of our political life, being coercive, be rationally transparent to those whom they are to bind. In this limited, political form, D thus amounts to the conviction that Rawls calls the liberal principle of legitimacy. As a result, however, it cannot be right for Habermas to regard D as independent of any antecedent moral commitments. For, as I have argued, the liberal principle of legitimacy embodies a moral principle of respect for persons. This moral foundation is what gives D the political authority it enjoys. Thus, Habermas's conception of democracy fails to be autonomous in the absolute way he desires. If we believe our political life should be organized by some principle such as D, that is only because we embrace the moral principle of equal respect for persons. The sort of normative hierarchy that Habermas dismisses as premodern shapes in fact his own thinking. Self-government, so far as it admits only those political principles which can be the object of reasonable agreement, rests on an independent moral basis. We cannot regard respect for one another as just one more political principle among others whose validity derives from our political will. On the contrary, this principle justifies the ideal of self-government and defines for us what it means. Reasonable agreement must in this context be understood (as I argued in section ill) as circumscribed by the moral principle of respect. Comprehensive moral views of the ends of life being eminently controversial, our political life cannot seek its basis in them. About this point Habermas is right. But we cannot suppose, as he does, that our political life should be similarly freestanding or autonomous with respect to morality altogether. Citizens can understand themselves as the collective author of their political principles, only if they see themselves as already bound by the claims of respect for persons. This moral principle refers, moreover, to an individual right, even if one more fundamental than the political rights that usually figure among explicit constitutional guarantees. Every person has the right, it claims, to be bound only by political principles whose justification he can rationally accept. As a result, Habermas's position comes down to one in which, contrary to his own account, an individual right sets limits to democratic self-rule, determining as it does the sort of expression of the popular will that shall count as democratic. The familiar constitutional rights of free-expression, property, and political participation, though no doubt serving to promote the goal of democratic self-rule, also have an independent rationale. They draw upon that most fundamental of individual rights, which is the right to equal respect.

462

John Rawls 622

THE JOURNAL OF PHILOSOPHY

In the Postscript to the English translation ofFaktizitát una Geltung, Habermas has replied to an earlier version of this criticism.25 To understand the normative basis of modern democracy, he argues, we must distinguish the horizontal sociation of citizens, in which they constitute their collective will and in so doing mutually accord rights to one another, from the subsequent step in which they set up rights to protect individuals from the political power so constituted. Therefore, individual rights against the state "are not originary but rather emerge from a transformation of individual liberties that were at first reciprocally granted" (BFN 457)—that is, established in the very formation of their collective will. I am not sure how neatly these two steps can be distinguished from one another, but I shall let that go. My concern lies with the first step. Observe that Habermas himself describes the horizontal sociation, which he holds to be foundational, as a process where citizens "recognizing one another as equals, mutually accord rights to one another" (BFN 457). Whence, I ask, comes their recognition of one another as equals? As Habermas's very words imply, this is not a commitment they acquire in virtue of such association, but rather one which defines the sort of association they understand themselves to be forming. This commitment amounts in fact to precisely what I have been calling the principle of respect. And that principle embodies an individual right—a right requiring that the terms of political association to be developed be as transparent to one's own reason as to that of others. Such a right does not arise through being mutually granted by citizens as they form their collective will. It has an independent authority which individuals must acknowledge if they are to form a democratic will in the way Habermas envisions. I see no reason, therefore, to change my conclusion that Habermas's notion of radical democracy is not really so radical that it differs materially from the idea of liberal democracy to which he imagines himself opposed. Failing to note the moral basis of his own commitment to democratic self-rule, he slips past the fact that he too assumes, if only implicitly, the antecedent authority of individual rights. VIII. FREEDOM AND MORALITY

Modern democracy is government by self-rule, animated by the ideal that the basic rules of political life should draw their legitimacy from being the object of reasonable agreement among those whom they 25 I first presented this criticism in a review of Faktizitàt una Geltung, "Die Wurzeln radikaler Demokratie," Deutsche Zeitschrift fur Philosophie, XLI (1993): 321-27. I developed it further in Morals of Modernity, chapter 10.

John Rawls MORAL BASIS OF POLITICAL LIBERALISM

463 623

are to bind. But we misunderstand the nature of the democratic ideal, I have insisted, if we suppose that for it, the collective will of the citizens constitutes the ultimate source of authority. Looking more closely at the notion of the collective will employed, we see that it is conceived in moral terms. Popular sovereignty can be understood as manifesting itself through reasonable agreement, only if it is defined as heeding the obligation of respect for persons. Democracy is thus a moral conception, and not just in the trivial sense that the principles and values by which a democratic people organize their political life are recognizably moral in character. More profoundly, democracy involves commitment to a moral principle that citizens must see as binding on them independent of their democratic will. Respect for persons is what gives their democratic will the normative shape it has. I have claimed that Habermas misses this moral basis of modern democracy and that Rawls, too, though recognizing it in part, fails to bring it out as distinctly as he should. My point, as I hope is plain, is not to reject the project of working out a conception of liberal democracy that is politically freestanding. That project I share. But I believe its moral presuppositions should be made explicit. Why, we may ask, should there be such reluctance to admit these assumptions? One reason is not far to seek. Liberal thinkers who continue to sympathize with earlier individualist versions of liberal democracy do not hesitate to expound the moral foundations of this ideal as they see them.26 But both Habermas and Rawls, in their different ways, aim to work out a political conception that does not depend upon comprehensive moral views, especially individualist ones, about the ends of life. In the effort to devise a political conception that is thus freestanding, it is only too easy to slide over the moral commitments propelling this very enterprise. I do think it important that the principle of respect does not express or entail a comprehensive moral philosophy. It has its place in a great many otherwise disparate ideas of the human good. Particularly significant is the fact that it can figure in those conceptions of life which refuse to accord supreme value to critical reflection and call instead upon forms of moral allegiance that are rooted in a sentiment of belonging. We may feel that our deepest commitments are constitutive rather than elective, fashioning our very sense of the

26 See, for example, Jeremy Waldron, Liberal Rights (New York: Cambridge, 1993), pp. 56-57, 163-68.

464

John Rawls 624

THE JOURNAL OF PHILOSOPHY

persons we are and the choices we can understand ourselves as making, and at the same time believe that political principles, relying as they do upon coercion, must be reasonably acceptable to all whom they are to bind, even to those who differ with us on the relative importance of critical reflection. There exist, of course, conceptions of the human good which reject a principle of respect for persons. No doubt, many of them also desire to limit critical reflection and preach instead obedience to traditional, often religious authorities. But these two views do not go necessarily together. In fact, individualist views of life, too, when they exalt the will to power of creative individuals and heap contempt upon the bulk of mankind, can prove illiberal. A reverence for tradition is not inherently hostile to liberal ideals and more than the cultivation of individuality is intrinsically friendly to them. But it also bears noting that reasonable people, exercising the basic capacities of reason and conversing with others in good faith, are not obliged on this basis alone to believe that the use of force should remain within the bounds of consensus. Being reasonable (in my sense of the term) does not entail the principle of respect for persons. If the moral basis of liberal democracy stands free of any comprehensive conception of the good, it does not on that account become inherently universal. Its appeal can extend only so far as people happen to be committed to the principle of respect.27 We ourselves may find this commitment so obvious that, like Habermas and Rawls, we fail to recognize or mention it. We may simply look through it, as we reflect upon the constitutional rules a democratic people should give themselves to govern their political life. But in reality, this commitment forms the specific moral "we" we are, even as we remain divided by important differences about other fundamental matters. Self-understanding, here as elsewhere, requires that we turn our mind to what lies so close to us as to go unseen. Then we can comprehend why this moral outlook, transparent though it may be to us, is not universal and has met indeed with sincere rejection from others. Yet I wonder whether there is not another reason why Habermas and Rawls do not make explicit the moral basis of their conceptions of liberal democracy. Perhaps in their case, as certainly in the case of others, another contemporary ideal has come to stand 27 I agree therefore with Stanley Fish—"Mission Impossible: Settling the Just Bounds Between Church and State," Columbia Law Review, xcvii, 8 (December 1997): 2255-333—that liberalism, like any political doctrine, must involve exclusion. All the same, a distinctive virtue of political liberalism is the breadth of its inclusiveness.

John Rawls

465

MORAL BASIS OF POLITICAL LIBERALISM

625

in the way of seeing our deepest commitments for what they are. In our time, freedom of self-determination, as both an individual and a collective value, enjoys a tremendous prestige. It commands so ready an allegiance that all other values can seem of subordinate importance. Individuals can be bound, it is said, only by the rules they give themselves. So, too, we continually hear that, in their collective capacity as members of a political community, citizens are to determine themselves the principles by which they will live.28 I doubt that the moral life in general can make much sense if its roots are sought in freedom as a supreme value.29 I am certain, as I have argued in this essay, that our commitment to democracy or political self-determination cannot be understood except by appeal to a higher moral authority, which is the obligation to respect one another as persons. CHARLES LARMORE

University of Chicago

28 Cf. Habermas, Die Einbeziehung des Anderen, p. 301: "Menschenrechte môgen moralisch noch so gut begrundet werden kônnen, sie dürfen aber einem Souveran [here, the democratic will, C.L.] nicht paternalistisch ubergestulpt werden.... Dieser Idee widerspràche es, wenn der demokratische Verfassungsgesetzgeber die Menschenrechte als so etwas wie moralische Tatsachen schon vorfinden wûrde, um sie nur noch zu positivieren." 29 por reflections along these lines, see The Morals of Modernity, pp. 87-88, and "L'autonomie de la morale," Philosophiques (Québec), xxiv, 2 (automne 1997): 313-28.

This page intentionally left blank

[17] What is reasonableness? James W. Boettcher Abstract The concept of reasonableness is essential to John Rawls's political liberalism, and especially to its main ideas of public reason and liberal legitimacy. Yet the somewhat ambiguous account of reasonableness in Political Liberalism has led to concerns that the Rawlsian distinction between the reasonable and the unreasonable is arbitrary and ultimately indefensible. This paper attempts to advance a more convincing interpretation of reasonableness. I argue that the reasonable applies first to citizens, who then play an important role in determining which comprehensive doctrines and political conceptions of justice are reasonable. In addition, while Rawls fails to specify explicitly the meaning of the reasonable in his standard of political justification (i.e. the liberal principle of legitimacy based on the criterion of reciprocity), I offer an interpretation of what it means for citizens to present reasonable claims and arguments to one another in public reason. Key words reasonable

legitimacy • liberalism • public reason • John Rawls •

Readers of John Rawls's Political Liberalism are often struck by the curious fact that one of that text's defining ideas, the idea of the reasonable, is so frustratingly difficult to define.1 Rawls's writings, beginning especially with the essays published in the 1980s, are replete with references to reasonableness. The terms 'reasonable' and 'unreasonable' are applied repeatedly by Rawls to persons, judgments, institutions, conceptions of justice, religious and philosophical views, social conditions such as pluralism, and forms of agreement and disagreement. We might be tempted to identify reasonableness with the very foundation of Political Liberalism, though invoking the notion of a foundation for Rawls's view would be somewhat misleading. Just as in A Theory of Justice, the methodology of Political Liberalism avoids reliance on a single, foundational value or ideal, proceeding instead with a series of

468

John Rawls

598 Philosophy & Social Criticism 30 (5-6) related, fundamental ideas that are capable of harmonizing with our considered convictions in reflective equilibrium. Yet, these ideas and convictions are held together, I believe, by the idea of the reasonable, which, like 'justice' in Theory and 'decency' in The Law of Peoples., occupies a privileged position in the conceptual apparatus of Political Liberalism. Consider political liberalism's ideas of public reasoning and liberal legitimacy. These ideas represent the main innovations of Rawls's political turn, as they enable us to understand how citizens and officials of a pluralistic constitutional democracy should conduct themselves in the public political forum, and how their decisions might be politically justified. The idea of public reason and the liberal principle of legitimacy are also defined directly in terms of reasonable citizens who justify their decisions with reasonable claims and arguments. Rawls begins The Idea of Public Reason Revisited', for example, by suggesting that 'in public reason comprehensive doctrines of truth and right be replaced by an idea of the politically reasonable addressed to citizens as citizens' (Rawls, Collected Papers [hereafter CP], p. 574; cf. Rawls, The Law of Peoples [hereafter LP], p. 55). He goes on to explain that citizens and officials adhere to the requirements of public reason, and also aim at legitimate decision-making, by seeking political justifications which satisfy the following 'criterion of reciprocity': Our exercise of political power is proper only when we sincerely believe that the reasons we would offer for our political actions - were we to state them as government officials - are sufficient, and we also reasonably think that other citizens might reasonably accept those reasons. This criterion applies on two levels: one is to the constitutional structure itself, the other is to particular statutes and laws enacted in accordance with that structure. To be reasonable, political conceptions must justify constitutions that satisfy this principle. (CP, pp. 578-9; emphasis added)

Along with the rather conspicuous appeals to 'reasonably' and 'reasonable' in this passage, the two paragraphs leading up to this formulation of the criterion of reciprocity include nine additional references to reasonableness. Yet, despite the obvious importance of the reasonable in political liberalism, Rawls has been reluctant to provide a straightforward exposition of its meaning. He remarks at a number of points that we all know what it means to say that a person's position is rational, given the circumstances, but unreasonable nevertheless (Rawls, Political Liberalism [hereafter PL], p. 48; Rawls, Justice as Fairness [hereafter/P], p. 7; Rawls, Lectures on the History of Moral Philosophy [hereafter HMP], p. 164). And he sometimes suggests that it is unnecessary to define reasonableness 'directly' and 'explicitly' (PL, p. 48; /P, p. 82). Instead,

John Rawls

469

599 Boettcher: What is reasonableness? Rawls's strategy has been to specify the meaning(s) of the reasonable by examining how the term might be used to characterize different subjectmatter (PL, p. 48; JT, p. 82). Thus 'Lecture IF of Political Liberalism spells out the distinction between the reasonable and the rational, and also includes an account of reasonable persons and reasonable comprehensive doctrines. Elsewhere Rawls applies the reasonable to political conceptions of justice, as well as to other ideas of political liberalism. There are at least two problems with Rawls's account of the reasonable. First, this account does not adequately explain how the application of reasonableness is extended from one concept or subject-matter to the next. What do these various applications of the reasonable share? What are the main substantive ideas that support appeals to reasonableness? Is there one use or specification of the reasonable that is primary or most basic? Second, Rawls fails to elucidate what would seem to be one of political liberalism's most important applications of the reasonable, namely, the concept of a reasonable claim or argument. Just to repeat, the ideas of public reason and liberal legitimacy, based on the criterion of reciprocity, require citizens and officials to locate reasons that they reasonably think that other citizens might reasonably accept. This requirement, like the requirement sometimes to exercise restraint in the appeal to comprehensive doctrines in public reasoning, is characterized by Rawls as a moral duty of democratic citizenship. If indeed citizens and officials are morally bound by the so-called duty of civility, they quite understandably will want to know which reasons, claims and arguments are reasonable, or which are reasonably acceptable to others. Not surprisingly, critics have been quick to raise questions about reasonableness. In his debate with Rawls, Jürgen Habermas suggests that the idea of the reasonable is too weak to function as a normative validity-claim, and that political liberalism does not provide the necessary moral grounding for the distinction between reasonable and unreasonable.2 Gerald Gaus contests the boundaries of the 'domain of the political' that Rawls associates with reasonableness, and Gaus also rejects standards of political justification that are based on the notion of a reasonable agent or a reasonable belief.3 Other critics have expressed concern that the idea of the reasonable is designed to exclude unfavorable or illiberal persons or points of view from the process of political justification. Thus Marilyn Friedman maintains that by limiting the 'legitimation pool' to reasonable citizens, Rawls 'rigs the election in advance', while Miriam Galston concludes that 'under the rubric of "unreasonableness" ', political liberalism simply defines certain comprehensive doctrines 'out of existence'.4 These are serious charges, and even Rawls concedes at one point that his understanding of reasonable comprehensive doctrines 'runs the

470

John Rawls

600

Philosophy & Social Criticism 30 (5-6) risk of being arbitrary and exclusive' unless it is clearly based on 'aspects of the reasonable itself (PL, p. 59). But the important question remains: What is the meaning of 'the reasonable itself? Answering this question is of the highest importance for political liberalism, since otherwise its central ideas of public reasoning and liberal legitimacy would be impracticable, if not incoherent. Specifically, absent the constraint of the reasonable, the criterion of reciprocity would render all exercises of political power illegitimate insofar as we anticipate any objections to them, no matter how outrageous or one-sided these objections may seem. Furthermore, one of the purposes of a standard of legitimacy, or what I also refer to as 'political justification', is to determine how political power may be exercised non-arbitrarily. Yet a standard of political justification that depends in an essential way on what amounts to an arbitrary distinction between the reasonable and the unreasonable would seem to be self-defeating. In what follows I do not respond directly to Rawls's many critics. However, I offer an interpretation of reasonableness that, if successful, might serve as the first step in a more complete defense of political liberalism from charges of arbitrariness and exclusion. As I hope to show, the many uses of the reasonable are based on the same, fundamental moral-political ideas. However, reasonableness must not be understood as a norm that allows Rawls or any other theorist to circumvent public political deliberation by determining exactly which laws and policies should be adopted in particular cases. After a brief discussion of the development of Rawls's idea of the reasonable in section 1,1 examine the normative content of reasonableness, which is based on political liberalism's fundamental ideas of society, the person and human judgment (section 2). I then explain how this content leads Rawls to propose criteria of reasonableness for political conceptions of justice and comprehensive religious and philosophical doctrines (section 3). I argue that, in addition to the general criteria proposed by Rawls, citizens and officials (hereafter called 'citizens'5) must play an important role in determining which particular doctrines and conceptions of justice are reasonable. In section 4,1 apply the idea of the reasonable to the claims and arguments presented by citizens in public reason. One of the main goals of this section, and of the paper as a whole, is to demonstrate how citizens can honor the idea of public reason despite their ongoing disagreement about which claims are true and which political conception of justice is valid, or most reasonable.

1 The development of the Rawlsian idea of reasonableness References to the reasonable can be traced all the way back to Rawls's 1951 essay 'Outline of a Decision Procedure for Ethics' where justifiable

John Rawls

471

601 Boettcher: What is reasonableness? ethical principles must be understood as reasonable according to the considered judgment of a competent moral judge (CP, pp. 10-12). The term also appears at a number of points in A Theory of Justice.6 But it is not until 'Kantian Constructivism in Moral Theory' (1980; CP, pp. 303-58) that Rawls begins to use 'reasonable' as a term of art with a specific moral-political meaning. Rawls realized that readers of Theory may be misled by the claim, which he has subsequently disavowed, that his account of justice falls under the rubric of a theory of rational choice (Rawls, A Theory of Justice [hereafter T/], p. 16). In particular, many readers have mistakenly conflated the description of the reasoning within the original position with an account of the practical reason or motivation of persons generally. Yet for Rawls the perspective of the parties in the original position must be distinguished from the perspectives of citizens and theorists (PL, p. 28). As citizens, persons have two basic moral powers, and two corresponding higher-order interests in realizing those powers. The 'Rational' corresponds to the power to define and pursue a conception of the good. Thus it applies primarily to the ends and interests adopted by an agent, and to the principles that an agent applies in seeking to satisfy those ends and interests. Rational agents are governed by familiar principles like choosing the most effective means to ends, organizing ends coherently, and assigning greater weight to the more probable consequences of an action. In justice as fairness, the moral power of the rational is represented in the original position by the reasoning of the parties. But a second moral power, the power to possess and act on an effective sense of justice, is also represented by the original position. This power is expressed not by the parties' reasoning, but by the various constraints within which their reasoning is conducted, e.g. the publicity condition, the veil of ignorance, and the symmetry of their situations. In 'Kantian Constructivism', Rawls associates these features of the original position with the capacity to honor fair terms of cooperation, or what he calls 'the Reasonable' (CP, pp. 316-17). The design of the original position actually incorporates the priority of the reasonable over the rational, since the parties' rational deliberations are constrained by devices that reflect a concern for the fairness of their eventual choice. Since its initial introduction in 'Kantian Constructivism', the idea of reasonableness has gradually taken on a greater role within Rawls's thought, and it would be a mistake to assume that the only purpose, or even the main purpose, of the reasonable is to clear up misconceptions about the original position. One only has to recall that, according to political liberalism, reasonable citizens may adhere to the idea of public reason while at the same time affirming conceptions of justice other than Rawls's favored conception of justice, justice as fairness, with its device of the original position (CP, p. 578). In short, justice as fairness is a view about what justice is; political liberalism is a view about public

472

John Rawls

602

Philosophy & Social Criticism 30 (5-6) reasoning and liberal legitimacy which admits justice as fairness as one reasonable political conception of justice among others.7 And so it is political liberalism that requires a robust notion of reasonableness. In Political Liberalism, Rawls acknowledges two main influences on his understanding of the reasonable (PL, pp. 48-9, n. I). 8 First, and not surprisingly, Rawls notes that the distinction between the reasonable and the rational is ultimately inspired by Kant. This suggestion, mentioned only in a footnote to Political Liberalism, can be better understood by considering the interpretation of Kant presented in Rawls's lecture course on moral philosophy, recently published as Lectures on the History of Moral Philosophy. In the Lectures, Rawls takes the Kantian term vernünftig to mean both reasonable and rational. The phrase 'reasonable and rational', Rawls explains, captures the fullness of Kantian practical reason, whereby agents reason practically by means of both categorical and hypothetical imperatives (HMP, pp. 164-5). Indeed in summarizing Kant's view of the categorical imperative, Rawls maintains that it permits us to act on a maxim only if, as sincere reasonable and rational agents, we can regard ourselves as members of the adjusted social world associated with the universalization of such a maxim, and also will and affirm that world. We must avoid presuming, however, that reasonableness has exactly the same meaning in political liberalism as it does in Rawls's interpretation of Kant's moral theory. In Political Liberalism Rawls notes that his own use of the reasonable is more narrowly restricted to certain political ideas and ideals. And elsewhere Rawls argues that the content of reasonableness is not derived from a philosophical theory of practical reason, observing instead that it is political liberalism which 'specifies the idea of the reasonable' (The Law of Peoples [hereafter LP], p. 87; cf., CP, p. 613).9 Thus political reasonableness is not equivalent to the capacity for pure practical reason, as Kant understands it. Nevertheless the general sense of the reasonable described in Rawls's Lectures remains consistent with the specification of reasonableness in political liberalism. To be reasonable, Rawls claims in the Lectures, is to be 'judicious' and ready to listen to the reasons offered by others (HMP, p. 164). Both in the moral life and in the politics of public reason, reasonable people avoid 'pushing a hard and unfair bargain' simply because they are in a position to do so. The second influence on Rawls is a 1953 essay by W. M. Sibley entitled 'The Rational Versus the Reasonable'.10 The main purpose of Sibley's essay is to distinguish, at least for the purposes of moral theory, the 'proper meaning' of the reasonable from that of the rational. According to Sibley, a rational agent is an agent who (1) makes informed choices about ends in light of an understanding of the relation between various ends and preferences, (2) selects the most effective means to

John Rawls

473

603 Boettcher: What is reasonableness? those ends, and (3) tends to act in accordance with knowledge about the best means to preferred ends. Rationality requires an agent to consider the interests of others, but only to the extent that the others' interests may affect the chances of the agent to realize his or her own interests. Suppose, for example, that my apartment complex requires dog-owners to dispose of their dogs' waste. If one of my interests is to avoid personal inconvenience, there is nothing inherently irrational, according to Sibley's sense of the rational, about my regular failure to pick up after my dog. Rationality only requires that I balance the end of convenience against other concerns, such as not paying the fines imposed by the management office or not losing the esteem of my friends and neighbors. Hence, in Sibley's terms, I may act rationally in this example by obeying the rules only when I believe that others are observing my actions. Rationality alone requires us to act intelligently, but it is also consistent with egoism. Reasonableness, however, is not simply the exercise of rational and intelligent judgment. According to Sibley, the moral sense of being 'reasonable' implies a willingness to consider our actions from a common standpoint and in light of the interests of others. Returning to the above example, suppose that my neighbors, some of whom are children, use the common area for recreation and play. From a common perspective, then, my deliberate failure to pick up after my dog is plainly unreasonable. My decision is not justifiable from a common point of view. Indeed Sibley also associates the reasonable with the categorical imperative, claiming that the demand for impartial or objective judgment in Kant's moral theory 'expresses] the essence of "reasonableness" '-11 However, according to Sibley, reasonableness is not just a matter of judgment but also a matter of motivation. A reasonable agent, in addition to having the capacity for reasoning in common with others, must also have a desire to do so. A reasonable agent is willing to treat the effects of her actions on the welfare of others as relevant to her decision. It is in this sense that Sibley refers to reasonableness as a 'distinctive disposition', that cannot be derived from rationality as such.12 To be sure, Rawls is concerned with how citizens can act reasonably in a political sense, and not with the more general 'disposition to act morally', as Sibley puts it.13 For Rawls reasonableness is essentially a political standard. But just what kind of standard is it? Who determines its criteria and content? How do we know when a person is acting unreasonably? I hope to answer these questions by examining how reasonableness may be used to characterize both the dispositions and beliefs of persons, and by clarifying the relationship between a reasonable citizen and the kind of claims and arguments that might comprise a reasonable citizen's political justifications and public reasoning.

474

John Rawls

604

Philosophy & Social Criticism 30 (5-6) 2 The normative content of reasonableness In Political Liberalism, reasonableness is introduced first as a kind of disposition, a characteristic of citizens. It is, as Habermas puts it, a predicate that is 'extended from the attitudes to the beliefs of reasonable persons'.14 As we have seen, reasonableness corresponds to the moral power to possess and act on a sense of justice. We can act reasonably because of our sense of justice; however, the reasonable itself is not defined by Rawls as a moral power. Instead Rawls begins 'Lecture IF of Political Liberalism by associating it with the attitudes and capacities of persons, referring to reasonableness as both a 'virtue' and a 'disposition' (PL, pp. 48-9, n. 1). The first feature of this disposition is the willingness to propose and honor fair terms of cooperation, the most central of political liberalism's fundamental ideas. A reasonable citizen is willing to propose terms that she believes others could accept, and to abide by those terms, provided that others are willing to do so as well. Fair terms of cooperation minimally involve relations between citizens that are more than mutual advantage but less than the altruistic promotion of the general good. They correspond to an idea of reciprocity: All persons involved are 'to benefit in an appropriate way as assessed by suitable benchmark of comparison' (PL, p. 16). A 'suitable benchmark' cannot simply be derived from existing social conditions, whatever they may be; rather, this benchmark must be based on a moral standard, such as the original position, which treats persons with equal concern and respect.15 Those persons who have benefited from unjust conditions might not gain, judging from their previous standpoint, from the transition to a more just social order. In aiming at fair terms of cooperation, reasonable citizens propose principles, laws and institutions that are consistent with a consideration of others as free and equal. In other words, a reasonable disposition is one by means of which a citizen treats others as free and equal. Persons are equal as citizens on the basis of their having the capacity for the two moral powers, at least to a minimal degree.16 Implicit in political liberalism is also the notion that persons are equal in the sense of being equally entitled to reasons for the exercise of coercive power. Or, as Rawls suggests in an early article, persons are viewed as having a 'right to equal consideration and respect in the design of their society' (CP, p. 259).17 Persons are free as citizens in three ways (PL, pp. 29-34). First, they are free to exercise their moral powers, to embrace and revise a conception of the good, and to do so without change to their public political identity (as citizens). So a religious conversion or the adoption of a new vocation or life-project would not alter the identity of a person qua citizen. Second, free citizens are the 'self-authenticating sources of valid claims' (PL, p. 32). That is, their claims are worthy of consideration because

John Rawls

475

605 Boettcher: What is reasonableness? they are freely endorsed, and not just insofar as they derive from established social roles or obligations to society. For political liberalism, as for republicanism, the antithesis of the free citizen is the slave, a person whose claims count for nothing and who is 'socially dead' (PL, p. 33).18 Finally, citizens are free in the sense that they are responsible for their ends, and are capable of adjusting these ends in light of the demands of justice. The privileges of political freedom are accompanied by corresponding obligations of citizenship. It is worth noting that in describing these aspects of a reasonable citizen, Rawls draws directly on the ideas that also support the construction of a political conception of justice. To put it in the simplest terms, a reasonable citizen is a person whose attitudes are informed by the fundamental building blocks of political liberalism itself. The final feature of a reasonable disposition is the willingness to recognize and accept the consequences of what Rawls calls the burdens of judgment. The burdens of judgment contribute to an explanation of the fact of reasonable pluralism. They imply that sincere and generally reliable reasoners sharing a common human reason will nevertheless often fail to reach agreement about many religious, philosophical and even political issues. Our practical reason is burdened, as it were, with a number of imperfections, including but not limited to the following:19 1 2 3 4 5 6

difficulties in assessing complex evidence disagreement about the weight of various considerations indeterminacy in interpreting vague concepts or in deciding hard cases divergent background experiences informing and affecting our judgment deciding between genuine, competing normative considerations that fall on both sides of an issue setting priorities among values when a decision to favor one value comes at the expense of another.

The Rawlsian burdens of judgment do not depend on a philosophical skepticism about value judgments defended by some liberal theorists.20 Rather, they imply that citizens should remain aware of the fact of reasonable pluralism, and should adopt an attitude of fallibilism toward their own claims and commitments. Being reasonable means not only recognizing but even expecting disagreement about important questions, including the questions of justice and constitutional interpretation that primarily comprise the domain of public reason. And when the source of our disagreement can be located in the burdens of judgment, disagreement is reasonable disagreement, even when we are confident about the validity of our own claims. In other words, we do not presume from the fact of disagreement alone that others must be

476

John Raw I s

606

Philosophy & Social Criticism 30 (5-6) unreliable reasoners, stupid, selfish, willful, or blinded by ideology and prejudice. This is not to say that reasonableness requires that we avoid altogether such ad bominem explanations. Critics sometimes worry that political liberalism demands civility in the public sphere even when a more radical or confrontational kind of discourse is needed.21 Sometimes people are willful or prejudiced or ideologically motivated, and a political theory with any critical teeth ought to enable us to call them what they are, especially when the exercise of political power is at stake. Rawls never denies this. His argument is rather that we should avoid either automatically presuming this kind of explanation or asserting it without supporting reasons, since each person can always accuse the other of familiar vices and shortcomings, no matter what his view. As Rawls observes in The Domain of the Political and Overlapping Consensus', 'the disposition to make such accusations without compelling grounds is plainly unreasonable, and often a declaration of intellectual war' (CP, p. 478). In cases of disagreement, reasonable persons give one another the benefit of the doubt. In short, reasonableness consists in the willingness to propose and honor fair terms of cooperation, to treat others as free and equal citizens capable of and interested in exercising the moral powers, and to recognize the burdens of judgment (Rawls, 'Reply to Habermas' [hereafter RH], p. 134). On my interpretation, reasonableness is primarily a disposition of citizens: it is an attitude, a way of taking or treating something or someone, which a citizen has a characteristic tendency to hold and express in political judgment and action. Reasonable persons take others to be politically free and equal and deserving of fair terms of social cooperation. In attributing freedom, equality and the basic moral powers to others, a reasonable citizen thereby takes on a series of commitments. She commits herself to beliefs and actions compatible with a respect for the other's basic moral powers. To take an obvious example, treating others as free and equal politically means rejecting practices linking citizenship to religious affiliation or social caste. Denying full citizenship to persons of a certain religious faith is incompatible with recognizing them as political equals. Less obvious examples abound, however, and citizens may even sometimes fail to appreciate the extent of the commitments implied by their own attitudes and beliefs. Thus, the reasonable is a substantive idea, referring to political liberalism's burdens of judgment and its fundamental ideas of the person and society. This is one way in which certain attitudes, as expressions of reasonableness, depend on beliefs about other persons and social institutions. Yet the link between reasonable citizens and reasonable beliefs should be understood in a second way as well. At times, Rawls uses the reasonable as a predicate applied directly to certain beliefs and

John Rawls

477

607 Boettcher: What is reasonableness? doctrines. For example, political liberalism refers to justice as fairness not as the true political conception of justice, but as a reasonable one or, at best, the most reasonable one. How is reasonableness, as a predicate characterizing conceptions of justice and other beliefs, assigned? I submit that the answer to this question is twofold. On the one hand, Rawls provides general criteria of reasonableness for different subjectmatter, namely, comprehensive doctrines and political conceptions of justice. But on the other hand, the specific content of reasonable doctrines, conceptions of justice, and even political claims and arguments must also be worked out and interpreted by citizens themselves. Both Rawls's specification of the reasonable and the judgments of citizens are governed by the basic normative content of reasonableness: terms of social cooperation must be fair as judged by a suitable benchmark of comparison, other citizens must be treated as free, equal and capable of and interested in exercising their basic moral powers, and claims and arguments must be subject to the burdens of judgment.

3 Reasonable comprehensive doctrines and political conceptions Reasonable comprehensive doctrines Political liberalism seeks overlapping consensus only among reasonable comprehensive doctrines. What is a reasonable comprehensive doctrine? In Political Liberalism, Rawls describes such doctrines by means of three general characteristics (PL, p. 59). First, as an exercise of both theoretical and practical reason, reasonable comprehensive doctrines coherently address the major philosophical, theological and moral concerns of human beings. Second, each doctrine organizes and weighs values, including non-political values, in a particular fashion. Finally, comprehensive doctrines are usually tied to a tradition of thought so that they tend to develop slowly, remaining stable over time. These characteristics suggest that for Rawls a 'comprehensive doctrine' is not equivalent to an individual's organized set of beliefs, whatever they happen to be. Rather a comprehensive doctrine is a view that is shared by a number of persons over time and tied to a recognizable tradition of thought. Beyond these three characteristics, though, reasonableness is also a moral-political standard by means of which various doctrines are assessed. It is unreasonable to use political power to enforce our comprehensive beliefs on others, and, in this sense, a reasonable comprehensive doctrine involves a principle of toleration (/P, p. 191). It should be noted that this substantive dimension of reasonableness does not immediately follow from Political Liberalism's general description of a

478

John Rawls

608

Philosophy & Social Criticism 30 (5-6) reasonable comprehensive doctrine. Miriam Galston criticizes Rawls in this regard for simply stipulating a substantive condition that a doctrine must meet in order to qualify as 'reasonable'.22 By invoking a principle of toleration, she argues, Rawls shifts to a more popular sense of the term 'unreasonable', but fails to connect adequately this popular sense of the term with his formal definition of a reasonable comprehensive doctrine. As I see it, Rawls is not guilty of such a philosophical sleight-ofhand. To be sure, the stipulation that reasonable doctrines must include the virtue of toleration does not tell us which views should or should not be tolerated. We might still ask, which doctrines are the unreasonable ones? But before addressing this question, we should first ask just what kind of answer to it would be required by political liberalism. Some doctrines are plainly unreasonable. Yet, in many cases, there is no bright line that can be drawn for the purpose of definitively and finally separating reasonable from unreasonable views. Nor do we need such a line. The specific content of reasonable philosophical or religious doctrines need not be determined in advance by political liberalism. Instead, according to Rawls, reasonableness is conferred upon a comprehensive doctrine insofar as a citizen can affirm it 'in a reasonable way' (PL, p. 60, n. 14). This means that reasonable doctrines are comprised of beliefs not inconsistent with the ongoing willingness to accept a fair distribution of social benefits and burdens, to regard others as free and equal and to recognize the burdens of judgment (/P, p. 191). After all, comprehensive doctrines are often quite flexible, and the same doctrine, whether religious or secular, may be cited in support of reasonable or unreasonable political aims. Moreover, as Erin Kelly and Lionel McPherson have argued, a person might be politically reasonable despite affirming a doctrine that is philosophically unsound.23 Thus, for most doctrines, their reasonableness or unreasonableness cannot be established a priori, once and for all. Everything depends on whether citizens can affirm the doctrine in question, or an interpretation of it, while at the same time respecting the moral powers of other citizens. Likewise, Rawls claims that a doctrine is unreasonable when it is affirmed in an unreasonable way, or when it authorizes the suppression of other, not unreasonable views. In explaining the unreasonableness of comprehensive doctrines, Rawls again emphasizes the importance of citizens taking one another to be reasonable or unreasonable in their affirmation of a comprehensive doctrine. When we insist on imposing our comprehensive beliefs politically, he avers, 'others, who are themselves reasonable, must count us unreasonable' (PL, p. 61; emphasis added). It is the task of citizens themselves to recognize one another as reasonable or unreasonable in particular cases. And, beyond the appeal to political liberalism's fundamental ideas of the person, society and

John Rawls

479

609 Boettcher: What is reasonableness? judgment, there is no precise formula for identifying the point at which others may regard our comprehensive doctrines as unreasonable on the whole. As Rawls notes, although a comprehensive doctrine may lead to an unreasonable conclusion on one question or another, it may still remain on balance reasonable (PL, p. 244, n. 32). Moreover it is always possible that a citizen will affirm an '[in principle] reasonable doctrine in an unreasonable way' (PL, p. 60, n. 14). Thus Rawls provides a specification of a reasonable comprehensive doctrine, without, however, offering anything like a deduction of its possible normative content. As I have already observed, Rawls suggests at one point that certain claims in Political Liberalism are misleading insofar as they imply that 'the content of the reasonable and the rational is derived from the principles of practical reason' (LP, p. 86, n. 33). Rawls's considered view is that the 'idea of the politically reasonable is sufficient unto itself for the purposes of public reason' (CP, p. 613). This does not mean that the criteria of reasonableness are the same for every subject-matter (i.e. for persons, doctrines, or political conceptions). Just to repeat, on the basis of political liberalism's fundamental ideas, the relevant criteria are to be worked out and explained for each subjectmatter to which the idea of the reasonable is applied (PL, p. 94; LP, p. 87). Discussing the quite different notion of 'decency', but referring by analogy to reasonableness, Rawls observes: 'We give it meaning by how we use it' (LP, p. 67). My suggestion is that each application of the reasonable to the beliefs that comprise comprehensive doctrines, political conceptions of justice, or other claims and arguments, must also depend, even if only indirectly, on the attitudes and judgments of reasonable citizens. In other words, while a general framework of reasonableness is suggested by political liberalism's fundamental ideas, when it comes to the details of which particular comprehensive doctrines or doctrinal claims are reasonable, we as citizens jointly give meaning to reasonableness by how we use it.

Reasonable political conceptions of justice According to political liberalism, public reasoning about constitutional essentials and matters of basic justice should be based on the values and norms of a political conception of justice. The content of public reason is not limited to a single view such as justice as fairness. Nor does it include all accounts of justice that might be categorized as political^ i.e. as independent of comprehensive doctrines and applied directly to the basic structure of society. Rather it is the family of reasonable political conceptions that serves as the content of public reason. But which political conceptions of justice are reasonable? Rawls maintains that a reasonable political conception of justice must satisfy three criteria. It

480

John Rawls

610

Philosophy & Social Criticism 30 (5-6) must (1) include a list of basic rights, liberties and opportunities, (2) prioritize these rights, liberties and opportunities over concerns for the general good and perfectionist values, and (3) provide 'measures ensuring for all citizens adequate all-purpose means to make effective use of their freedoms' (CP, pp. 581-3).24 For example, Rawls's two principles of justice, along with their priority rules, qualify as a reasonable political conception according to these standards. Yet, we might still ask why Rawls proposes these criteria as the test of a political conception's reasonableness. The answer to this question brings us back to the fundamental ideas of society and the person. Citizens are politically free and equal to the extent that their powers of choice, action and self-determination are legally protected in the form of a basic system of rights which includes positive political liberties as well as negative liberties. Once citizens regard one another as equal in possessing the two moral powers, basic rights, liberties and opportunities are necessary in order to secure fair terms of cooperation, especially in light of citizens' incommensurable conceptions of the good (PL, p. 303). Individual citizens and communities and associations of citizens must be free to organize their lives and projects in accordance with such a conception, without thereby being persecuted by a cultural majority or having to sacrifice basic legal protections and opportunities. Thus a reasonable political conception must not only protect the liberties associated with the idea of citizenship and the democratic procedures necessary for resolving political disputes. It must also include substantive rights such as freedom of conscience which are not immediately associated with democratic political procedures and participation.25 Moreover, rights and opportunities ultimately fail to protect the moral powers of persons qua citizens if the basic rights or opportunities of some citizens can be sacrificed for the sake of a general good other than justice itself, such as economic growth or efficiency. Hence the need for priority rules that comport with criterion (2).26 Finally, basic rights, liberties and opportunities also fail to protect our moral powers on an equal basis if they are merely formal. A right to political participation, for example, is essentially meaningless when actual opportunities for political influence depend almost entirely on the lobbying power that derives from private and corporate wealth controlled by a few. Thus the worth of liberty must be protected by measures securing the all-purpose means for the use of liberty. These measures include an equitable distribution of primary goods and a guarantee of the fair value of political liberties (PL, pp. 324-31). Yet, this framework of basic rights, liberties and opportunities still leaves open the question of which specific conceptions of justice are reasonable, all things considered, and which is most reasonable. After

John Rawls

481

611 Boettcher: What is reasonableness? all, citizens must still make a number of determinations about how to secure, interpret, organize and weigh various rights, liberties and opportunities. Furthermore, in summarizing his view of the family of reasonable political conceptions, Rawls introduces the criterion of reciprocity as the 'limiting feature' of this family, suggesting that reasonable political conceptions justify constitutions that meet this criterion (CP, p. 581). But the criterion of reciprocity already includes an appeal to reasonableness. Therefore, just as with respect to comprehensive doctrines, the determination of which political conceptions are reasonable, and which is most reasonable, depends in part on the judgments of citizens. As Rawls notes, the balance of political values that our nonpolitical beliefs support must not only be reasonable. This balance must also be cone that can be seen to be reasonable by other citizens' (PL, p. 243). This last point also serves as an important reminder that a political conception is not reasonable just because a citizen convinces herself that it is so. Of course each citizen must at some point make a judgment about which conception is most reasonable; this is part of the firstperson dimension of public reasoning and political justification. At the same time, '[n]ot any value is reasonably said . . . to be a political value; and not any balance of political values is reasonable' (PL, p. 227). It is difficult to imagine how a political conception that arbitrarily denies basic rights to a class of citizens could be endorsed by those same citizens acting as free and equal members of the political community (JF, p. 192). Seeking a fair distribution of benefits and burdens and designating others as free and equal through the terms of a political conception represents the first moment of establishing that conception's reasonableness. As I hope to show in the next section, this conception must also be one which others can in principle acknowledge as reasonable.

4 Reasonableness and public reasoning What about the claims and arguments that are introduced into the public political forum by citizens engaging in public reason? Unlike comprehensive doctrines and political conceptions of justice, Rawls does not specify criteria for reasonable claims and arguments. In fact, Rawls offers no explanation of what constitutes a reasonable claim or argument other than the suggestion that claims and arguments in public reason should somehow be based on the values of a reasonable political conception of justice. Yet this suggestion is only somewhat helpful. Claims and arguments in public reason will often involve more than an appeal to the basic values of a political conception. They might also

482

John Rawls

612 Philosophy & Social Criticism 30 (5-6) refer to other ideals or social goals and to empirical, scientific and common-sense judgments. Debates on affirmative action, for example, might involve questions of efficacy and feasibility, as well as a consideration of the value of diversity, which must be weighed alongside the values of a political conception. Moreover, even when basic political values such as liberty, equality and opportunity are at issue, debates in public reason will often aim at determining which interpretation of these values is reasonable, or most reasonable. Reasonable citizens will presumably endorse political conceptions that specify a central range of application for each form of liberty (PL, p. 297f.). But, especially insofar as they are expected to endorse competing political conceptions that will order political values differently, citizens must still resolve difficult questions of interpretation and application. Indeed I see no way for political liberalism to avoid the problem of how a standard of reasonableness is to be applied directly to claims and arguments in public reason. Recall that these claims and arguments are governed by the criterion of reciprocity, and that the main formulation of the criterion of reciprocity already includes a double reference to the notion of reasonableness: citizens must reasonably think that others might reasonably accept their reasons for a political choice.27 As I see it, this means that citizens must seek claims and arguments that are both reasonable and capable of being widely appreciated as reasonable. Thus the meaning of the reasonable must be specified from the perspective of the reason-giving agent and from the perspective of the addressees of the agent's claims and arguments. A first question is how a standard of reasonableness should apply to the claims and arguments of a reason-giving agent, i.e. a citizen engaging in public reason. T. M. Scanlon has argued that '[a] claim about what it is reasonable for a person to do presupposes a certain body of information and a certain range of reasons which are taken to be relevant, and goes on to make a claim about what these reasons, properly understood, in fact support'.28 Adapting Scanlon's definition, we can say that for Rawls, a reasonable argument will draw on and make a claim about a certain body of information or range of reasons in a particular case. The body of information and range of reasons presupposed by claims and arguments in public reason are by now quite familiar: (1) the search for fair terms of cooperation as judged by a suitable benchmark of comparison, (2) the freedom and equality of persons who are taken to possess basic moral powers and (3) the burdens of judgment. As reason-giving agents, reasonable citizens assume that these ideas are or ought to be shared by all other citizens, and that they limit the permissible premises and conclusions of a reasonable argument. An argument in public reason that treats historically arbitrary inequalities as an appropriate benchmark of comparison is

John Rawls

483

613 Boettcher: What is reasonableness? unreasonable. So too is an argument that is based on intolerance toward members of a minority religious group, or on the notion that one's comprehensive philosophical or religious doctrine provides infallible guidance in political life. Following Sibley, we can also say that reasonable claims and arguments should demonstrate a citizen's willingness to consider political issues from a common standpoint which acknowledges the interests of other citizens. In public reason, such claims and arguments should draw as much as possible on the values of a political conception of justice that the reason-giving agent regards as the best interpretation of basic political values. An interpretation of political values, in the form of a pro tanto justified political conception, should be 'suitably complete', providing answers to most questions concerning constitutional essentials and matters of basic justice (PL, p. 241; RH, pp. 142-3). Thus reasonable political arguments should not be based on ad hoc appeals to claims about which the reason-giving agent is skeptical, but which happen to support his agenda in a particular case.29 Finally, and especially in light of the difficulties associated with the burdens of judgment, citizens should propose and accept what they take to be strong claims and sound arguments. Citizens should attempt to avoid fallacies and errors in judgment, and they should subject all relevant facts, evidence and arguments to critical scrutiny. Public reason is, after all, a form of reasoning. And all forms of reasoning include 'common elements' such as 'the concept of judgment, principles of inference, and rules of evidence, and much else, otherwise they would not be ways of reasoning but perhaps rhetoric or means of persuasion' (PL, p. 220). While the reasonable is not primarily an epistemological standard, reasonable political claims and arguments, like all claims and arguments, should be evaluated in terms of logical and epistemic standards. Which claims and arguments can be widely appreciated as reasonable? The focus of this question is again on the first-person dimension of political justification, i.e. the case of a citizen deciding on a political justification and/or presenting it to others in public reason. It is a separate question whether any political justification is likely to meet with the actual assent of all reasonable citizens.30 There are many occasions when agents suspect that their arguments and political justifications will in fact be rejected by others. But that is not the point. In political deliberation, a good citizen considers the interests and perspectives of others. In addition to identifying valid claims and arguments that, according to the reason-giving agent, satisfy the basic criteria of reasonableness (1-3), the agent should also believe that a reasonable addressee could understand that the argument satisfies these criteria. The agent hopes that other citizens, who are similarly disposed

484

John Rawls

614 Philosophy & Social Criticism 30 (5-6) to seeking fair terms of cooperation, could appreciate that the argument is at least minimally consistent with the fundamental ideas of political liberalism. She should also believe that the argument could be acceptable in this way to others who are free to deliberate on the basis of their own interests and commitments. Thus an agent attempts to determine what others would accept provided that they were in a suitable position to do so, i.e. a choice situation in which they were seeking fair terms of cooperation, aware of their interests, sufficiently informed about the relevant issues and free of intimidation, manipulation or coercion. Even though there are often multiple reasonable arguments for a given matter of law or policy, a reason-giving agent will likely believe that her own claims and arguments are not just reasonable, but true. And it is important to point out that the burdens of judgment do not direct citizens to give up the search for sound arguments based on true claims. But they do imply that, with respect to difficult and complex issues, what one takes to be the truth may turn out to be incorrect. Thus a citizen's claims and arguments should fall under what Rawls calls the 'precepts of reasonable discussion' (CP, pp. 478-9).31 These precepts instruct citizens to avoid unwarranted accusations, to expect sincere disagreements and to begin deliberations by crediting others with good faith reasoning. They also remind citizens that their judgments are fallible, and open to revision and correction. Reasonable citizens listen to others, and demonstrate a willingness to learn about the issues and sometimes change their views. Finally, a reasonable citizen relies generally on methods of reasoning and inquiry that are accessible to others. The aim is not simply to find arguments that are sound, but arguments that other generally reliable reasoners can recognize as sound and reasonable. Citizens should appeal to reasons that are based on common sense, logical consistency, strong evidence or established methods of science and inquiry. This is not to say that a claim based on complicated scientific evidence is unacceptable as a public reason simply because it is difficult to understand. But in general the guidelines of public reason should be based on forms of argument that are accessible to democratic citizens: This helps to insure that public reasoning can publicly be seen to be - as it should be - correct and reasonably reliable in its own terms' (PL, p. 162). In sum, in offering what she considers to be the most reasonable argument, a reason-giving agent hopes to combine valid reasoning with the best interpretation of a pro tanto justified political conception and the most accurate survey of the facts and circumstances relevant to the case at hand. In seeking an argument that could be accepted by others as reasonable, the agent should also consider whether an addressee, similarly committed to seeking fair terms of cooperation, could

John Rawls

485

615 Boettcher: What is reasonableness? recognize that the agent's claims and arguments are consistent with the burdens of judgment and political liberalism's fundamental ideas of society and the person. In other words, in advancing a political justification, a citizen provides what, from her own perspective, is the most reasonable claim or argument and what, from the perspective of an addressee, may be considered at least reasonable. What Rawls at one point observes about political conceptions should be applied generally to claims and arguments in public reason: The criterion of reciprocity requires that when those terms are proposed as the most reasonable terms of fair cooperation, those proposing them must also think it at least reasonable for others to accept them, as free and equal citizens, and not as dominated or manipulated, or under the pressure of an inferior political or social position. (CP, p. 578)

Consider, for example, proposals in the USA for the public funding of religious schools.32 The idea of the reasonable as such is not a substitute for the reasoning of citizens, legislators and judges who must resolve this issue, but it can play a role in guiding their public political deliberation. It would be unreasonable for a proponent of funding to offer as her sole political justification the goal of religious salvation, the promotion of a particular faith, or the ideal of a religiously homogenous social order. Opposing citizens would not only reject her justifications as inadequate. They would also fail to understand how such justifications are consistent with a respect for their status as free citizens, capable of pursuing other religious and non-religious conceptions of the good and entitled to a political order that protects these conceptions with legal guarantees and safeguards. From their perspective, the reason-giving agent seems to be insisting on her beliefs just because they are her beliefs (PL, p. 61 ).33 This insistence is tantamount to a rejection of the reasonableness of the comprehensive doctrines of at least some other citizens (PL, p. 128). The same conclusion would be reached about an opponent of public funding whose justification consisted of the view that religion is an inherently irrational institution that should be discouraged by the state. In keeping with the ideal of reasonableness, however, an agent might sincerely believe that there are adequate public reasons in support of public funding for religious schools. She may argue that public funding promotes equality of opportunity by providing a better education to disadvantaged students who are particularly ill served by failing public schools. Or she may argue that some religious schools promote core liberal-democratic values, such as tolerance and fair-mindedness. Or she may propose that religious parents of school-age children deserve public assistance as a matter of fairness or right. In the case of these justifications, opposing citizens can at least recognize the public grounds of

486

John Rawls

616 Philosophy & Social Criticism 30 (5-6) the agent's argument, even when they find its conclusion to be incorrect.34 These justifications seem implicitly to acknowledge the reasonableness of the comprehensive perspectives of opposing citizens (PL, pp. 127-8).35 Thus political liberalism seems to require its reasonable citizens to maintain multiple interpretations of basic political values. On the one hand, they endorse a particular conception of justice with its ordered constellation of political values as most reasonable, i.e. as the best interpretation of political liberalism's underlying ideas of society and the person. On the other hand, they attempt to understand which rival conceptions of justice and constellations of political values meet the threshold of reasonableness and are at least consistent with these underlying ideas and with the burdens of judgment. As rational, citizens naturally want their claims and arguments to be accepted by others. But, as reasonable, they also want their claims and arguments, along with their underlying dispositions, to be acknowledged as reasonable by other citizens who are similarly committed to seeking fair terms of cooperation. This special concern with the judgments of our reasonable interlocutors is a familiar one. In cases when others disagree with us we want to believe that they have at least approached the matter reasonably, that is, in a way that demonstrates a fair consideration of our interests, respects our moral powers, and acknowledges the burdens of judgment. In fact we are more inclined to reconsider our own views when we believe that those who disagree with us are reasonable, and we are more likely to dismiss interlocutors whose objections strike us as unreasonable. Public reasoning is primarily a justificatory practice that aims at the main goal of arriving at legitimate political decisions based on the right kind of reasons. In addition to this goal, public reason also contributes to the goals of stability and social unity by promoting sentiments of civic friendship. By adhering to the requirements of public reason, citizens are also attempting to reassure one another of their commitment to reasonableness. Yet, because the sources of disagreement on any given issue are far from transparent, it will often be difficult for a citizen to determine if another citizen's claim is simply incorrect, or both incorrect and unreasonable. This is why it is important for citizens to attempt to acknowledge that other citizens, with whom they disagree about particular claims, arguments and decisions, nevertheless may be committed to reasonable comprehensive doctrines that support reasonable political conceptions of justice. In addition to honoring the obligations of public reason, what steps might citizens take in order to reassure one another of their reasonableness? As I see it, a discussion of this question ultimately requires a separate treatment of the virtues of liberal-democratic citizenship. But, before concluding, I want to suggest that such reassurance would seem

John Rawls

487

617 Boettcher: What is reasonableness? to depend on citizens achieving a better understanding of how competing political conceptions of justice organize basic political values and how various comprehensive doctrines justify political conceptions. In order to understand these differences, citizens should cultivate virtues besides the familiar liberal virtue of toleration. They must actually engage one another, and attempt to know more about the doctrines and conceptions of justice that flourish in their society. Public reasoning, which implies restraint in the appeal to doctrinal justifications in certain cases, must often be supplemented by other forms of discourse that would allow citizens to bring their non-political views into the public political forum and demonstrate to one another their commitment to reasonable political aims. This suggestion might come as a surprise to those critics who have wrongly interpreted Rawls as trivializing religious belief or presenting an exclusivist view of public reason that would relegate all comprehensive discourse to the 'private' sphere.36 As it turns out, the public presentation of comprehensive doctrines is entirely consistent with the idea of public reason, and especially with the forms of discourse associated with what Rawls calls the 'wide view of public political culture' (CP, pp. 591-4). In demonstrating their commitment to public reason, while also introducing relevant aspects of their comprehensive views, 'citizens who hold different doctrines are reassured, and this strengthens the ties of civic friendship' (CP, p. 594).

5 Conclusion I have argued that the reasonableness of political conceptions of justice and comprehensive doctrines is dependent on the judgments of reasonable citizens, and that the reasonableness of citizens is based on dispositions oriented by political liberalism's fundamental ideas of society, the person and human judgment. I have also attempted to explain what it means for a citizen to satisfy the criterion of reciprocity by advancing reasonable claims and arguments in public reason. Each citizen offers a political justification that is, as he or she sees it, the best interpretation of political liberalism's fundamental ideas and also capable of being appreciated by other reasonable citizens as at least consistent with these ideas. The idea of public reason and the liberal principle of legitimacy may aim ideally at agreement. But their principal function is to show us how we can act together politically in light of ongoing disagreement. Absent the unlikely prospect of overlapping consensus on a single political conception of justice, citizens will continue to disagree not only about theological issues and questions of first philosophy but also about

488

John Rawls

618 Philosophy & Social Criticism 30 (5-6) constitutional essentials, matters of basic justice, and other issues of law and policy. Far from being a weakness of Rawls's view, as some have suggested, disagreement about matters of justice is precisely what the main ideas of political liberalism, especially the idea of the reasonable, help us to understand.37 Adhering to the idea of public reason, citizens attempt to recognize other claims and arguments, and ultimately one another, as reasonable, while nevertheless disagreeing about what justice entails. Department of Philosophy, St Joseph's University, Philadelphia, PA, USA

Notes The author would like to thank David Rasmussen for valuable discussion. This manuscript was submitted in August 2002. 1 John Rawls, Political Liberalism, 2nd edn (New York: Columbia University Press, 1996) [abbreviated hereafter as PL]. Other references to Rawls's writings will be abbreviated as follows: A Theory of Justice, 1st edn (Cambridge, MA: Harvard University Press, 1971) as T/; 'Reply to Habermas'Jowrad/ of Philosophy 92, 3 (March 1995): 131-80 as RH-, The Law of Peoples (Cambridge, MA: Harvard University Press, 1999) as LP; Collected Papers, ed. Samuel Freeman (Cambridge, MA: Harvard University Press, 1999) as CP; Lectures on the History of Moral Philosophy, ed. Barbara Herman (Cambridge, MA: Harvard University Press, 2000) as HMP; and, Justice as Fairness: A Restatement, ed. Erin Kelly (Cambridge, MA: Harvard University Press, 2001) as/jp. 2 See Jürgen Habermas, The Inclusion of the Other, ed. Ciaran Cronin and Pablo De Greiff (Cambridge, MA: MIT Press, 1998), chapters 2-3. For related criticism, see David Estlund, The Insularity of the Reasonable: Why Political Liberalism Must Admit of Truth', Ethics 108 (January 1998): 252-75. 3 See Gerald Gaus, Justificatory Liberalism (New York: Oxford University Press, 1996); Gerald Gaus, 'Reason, Justification, and Consensus: Why Democracy Can't Have It All', in Deliberative Democracy, ed. James Bohman and William Rehg (Cambridge, MA: MIT Press, 1997), pp. 205-42; and Gerald Gaus, 'Reasonable Pluralism and the Domain of the Political: How the Weaknesses of John Rawls's Political Liberalism Can Be Overcome by a Justificatory Liberalism', Inquiry 42 (1999): 259-84. For related criticism, see also Bruce Brower, 'The Limits of Public Reason', The Journal of Philosophy XCI (January 1994): 5-26 and Joseph Raz, 'Disagreement in Polities', The American Journal of Jurisprudence 43 (1998): 25-52. 4 Marilyn Friedman, 'John Rawls and the Political Coercion of Unreasonable

John Rawls

489

619 Boettcher: What is reasonableness?

5

6

7

8

9 10

11 12 13 14 15

People', in The Idea of Political Liberalism, ed. Victoria Davion and Clark Wolf (New York: Rowman oc Littlefield, 2000), pp. 16-33, and Miriam Galston, 'Rawlsian Dualism and the Autonomy of Political Thought', Columbia Law Review 94, 6 (October 1994): 1842-59. See also the related criticism of Chantai Mouffe, 'Political Liberalism: Neutrality and the Political', Ratio Juris 7, 3 (December 1994): 314-24 and Stanley Fish, The Trouble with Principle (Cambridge, MA: Harvard University Press, 1999). Throughout this essay, I follow Rawls in using the term 'citizen' as an inclusive term that refers to all of the subjects of public reason. I leave aside the issue of how the idea of public reason might apply in different ways to democratic citizens and to various officials (PL, pp. 215-16, 231-40; CP, pp. 576-7). For an interesting discussion of this issue, see Kent Greenawalt, Private Consciences and Public Reasons (New York: Oxford University Press, 1995), especially chapters 11-15. For an instructive discussion of the 'non-technical' usage of the reasonable in Theory., as well as the development of the idea of the reasonable in political liberalism, see Jon Mandle, The Reasonable in Justice as Fairness', Canadian Journal of Philosophy 29, 1 (March 1999): 75-108. To be sure, Rawls continues to endorse justice as fairness, and refers to it at a number of points in Political Liberalism and related essays. It should also be noted that the relationship between justice as fairness and political liberalism is far more complicated than I have indicated, especially insofar as many of the ideas of political liberalism originated with Rawls's attempt to correct the account of stability in Theory. Without examining this relationship in more detail, I shall assume that the main ideas of political liberalism now constitute a view that might be understood independently of Rawls's own theory of justice (even if Rawls's theory of justice now requires some of the ideas of political liberalism). Rawls admits as much at CP, p. 582, n. 27. See my 'Political Liberalism Without a Theory of Justice', Contemporary Philosophy XXV, 1 and 2 (2003): 3-8. In a subsequent footnote, Rawls connects his account of the reasonable to T. M. Scanlon's principle of moral motivation (PL, p. 49, n. 2). In developing an interpretation of reasonable claims and arguments below, I draw on Scanlon's view. See also JF, p. 196: 'the reasonable generates itself and answers itself in kind. It is not alone among dispositions is this respect [sic]. What makes it unique is its tie with reason.' W. M. Sibley, 'The Rational Versus the Reasonable', Philosophical Review 62,4 (October 1953): 554-60. Rawls briefly discusses this essay in Political Liberalism, observing that Sibley's account of the reasonable is 'broader but consistent' with his own (PL, p. 49, n. 1; cf. JF9 p. 7, n. 6). Sibley, 'The Rational Versus the Reasonable', p. 558. ibid. ibid., p. 560. Habermas, The Inclusion of the Other, p. 88 [original emphasis]. Other standards besides the original position might also specify fair terms of cooperation, provided that these standards adequately represent citizens as free and equal persons.

490

John Rawls

620

Philosophy & Social Criticism 30 (5-6) 16 A similar basis for equality is discussed at T/, pp. 504-12. Rawls claims that 'the capacity for moral personality is a sufficient condition for being entitled to equal justice. Nothing beyond the essential minimum is required' (T/, pp. 505-6). He also argues that the equality of persons is presumed to be the norm; deviations from it must be justified. See also the discussion of equality at/F, p. 132. 17 This understanding of equality is found in 'A Kantian Conception of Equality5 (1975; CP, pp. 254-66). 18 For a full account of the republican view of freedom as freedom from domination and servitude, see especially Phillip Pettit, Republicanism (New York: Oxford University Press, 1997). Pettit considers his interpretation of freedom to represent a distinct alternative to the liberal interpretation, though this contrast, I would argue, depends on a rather selective and uncharitable interpretation of liberalism. For a discussion of these issues, see David M. Rasmussen, 'Accommodating Republicanism', Denver Law Review 76, 4 (1999): 955-60 and Sebastiano Maffettone, 'Liberalism and Its Critique: Is the Therapy Worse than the Disease?', Philosophy ó1 Social Criticism 26, 3 (May 2000): 1-37. 19 In what follows I paraphrase PL, pp. 56-7. For an earlier account of these burdens, then referred to as the 'burdens of reason', see CP, pp. 475-8. See also /F, pp. 35-6. 20 A stronger version of the burdens of judgment might be based on skepticism about certain truth-claims or conceptions of the good. For such a view, see Brian Barry, Justice as Impartiality: A Treatise on Social Justice, Volume II (New York: Oxford University Press, 1995). 21 See, for example, Jack Knight and James Johnson, 'What Sort of Political Equality Does Deliberative Democracy Require?', in Deliberative Democracy, ed. James Bohman and William Rehg (Cambridge, MA: MIT Press, 1997), pp. 279-319, especially pp. 284-5. See also Sheldon Wolin, 'The Liberal/Democratic Divide: on Rawls's Political Liberalism9, Political Theory 24, 1 (February 1996): 97-120. 22 Galston, 'Rawlsian Dualism', pp. 1847-8. For additional questions about reasonable comprehensive doctrines, see Brian Barry, 'John Rawls and the Search for Stability', Ethics 105 (July 1995): 874-915. 23 Erin Kelly and Lionel McPherson, 'On Tolerating the Unreasonable', The Journal of Political Philosophy 9, 1 (March 2001): 38-55. 24 At times these criteria are said to characterize liberal political conceptions of justice (PL, pp. 6,156,223). Reasonable political conceptions are, in this sense, liberal political conceptions. 25 On this point, see Joshua Cohen, 'Pluralism and Proceduralism', ChicagoKent Law Review 69, 3 (October 1994): 589-618. 26 Rawls also argues at length that, from the perspective of the original position, the basic liberties and their priority are required by each moral power. See PL, pp. 310-23. 27 This double reference is not accidental. In the 'Introduction to the Paperback Edition' of Political Liberalism Rawls specifically calls attention to the fact that ' "reasonably" occurs at both ends' of his more recent formulations of the criterion of reciprocity. See PL, p. xliv, n. 9.

John Rawls

491

621 Boettcher: What is reasonableness? 28 T. M. Scanlon, What We Owe to Each Other (Cambridge, MA: Harvard University Press, 1998), p. 192. Of course a reasonable claim is not equivalent to a claim about what is reasonable for an agent to do. Here I focus mainly on Scanlon's suggestion that reasonableness be associated with a body of information and range of reasons. I am indebted to Scanlon and Rasmussen for a valuable discussion of Scanlon's view of reasonableness. 29 It is possible for a citizen to support a policy on the basis of one political justification, and also appeal to another political justification in public reason, provided that the citizen's appeal to the alternate justification satisfies a sincerity condition. An examination of the details of such a sincerity condition will have to wait for another occasion. 30 The liberal principle of legitimacy based on the criterion of reciprocity is a first-person standard that guides the deliberation of citizens and officials as they formulate and present political justifications to one another. But it may also be understood as a standard that enables us to judge collective decisions as politically justified, or legitimate. Determining how to interpret the criterion of reciprocity as an inter subjective standard of legitimacy for collective decisions raises a number of questions that I cannot address here. 31 The precepts of reasonable discussion are introduced by Rawls in 'The Domain of the Political and Overlapping Consensus' (1989) (CP, pp. 478-9). 32 See Alan Wolfe (ed.) School Choice: The Moral Debate (Princeton, NJ: Princeton University Press, 2003). Rawls discusses the related issue of school prayer at CP, p. 601 f. 33 See also Thomas Nagel, 'Moral Conflict and Political Legitimacy', Philosophy a1 Public Affairs 17 (Summer 1987): 215-40. 34 Of course, under the 'wide view' of public reason with its 'proviso', it is legitimate for a citizen to offer both public and comprehensive reasons in support of the same decision. For a discussion of the proviso, see CP, p. 591 and the 'Introduction to the Paperback Edition' at PL, pp. li—Hi. 35 On the idea of acknowledging the reasonableness of other comprehensive views, see also Joshua Cohen, 'Moral Pluralism and Political Consensus', in The Idea of Democracy, ed. David Copp, Jean Hampton, and John E. Roemer (New York: Cambridge University Press, 1993), pp. 270-91. 36 I examine this charge and related issues in my 'Public Reason and Religion', in The Legacy of John Rawls, ed. Thorn Brooks and Fabian Freyenhagen (New York: Continuum, forthcoming). 37 This variety of criticism appears in Michael Sandel's review essay, 'Political Liberalism', Harvard Law Review 107 (1994): 1765-94.

This page intentionally left blank

[18] RELIGIOUS CITIZENS WITHIN THE LIMITS OF PUBLIC REASON Philip L. Quinn

E

was a surprising discovery of modern liberal politics. As John Rawls puts it, he success of liberal constitutionalism came as a discovery of new social possibility: the possibility of a reasonably harmonious and stable pluralist society."1 There was, he thinks, no way of knowing in advance of that possibility. Moreover, this pluralism is itself reasonable. Modern democratic societies are characterized, according to Rawls, "not simply by a pluralism of comprehensive religious, philosophical and moral doctrines but by a pluralism of incompatible yet reasonable comprehensive doctrines.*'2 So liberals should now start from the assumption that such pluralism is the normal result of the exercise of human reason within the context of the free institutions of a democratic society. The assumption is supported by the modern history of secular philosophical ethics. As Robert M. Adams notes, "nothing in the history of modern secular ethical theory gives reason to expect that general agreement on a single comprehensive ethical theory will ever be achieved — or that, if achieved, it would long endure in a climate of free inquiry."3 And, needless to say, the prospects for agreement on a single comprehensive religious doctrine are, at best, equally dim. I share with Adams the view that the discovery of this new social possibility was "an important moral discovery," though, as he notes, Rawls is "more explicit about discovery of possibility than about discovery of value."4 In particular, it seems to me that liberal societies in which there is a pluralism of reasonable comprehensive religious doctrines realize some very significant values. Among them is the availability to those who live within such societies of a variety of sources of insight into human spirituality and a transcendent dimension of reality as well as ample social space for criticism and reform of religious traditions. Of course, like Rawls, who attributes this thought to Isaiah Berlin, I believe "there is no social world without loss."5 And, like Adams, I think that "some social arrangements seem to have few advantages, but all have disadvantages."61 can appreciate the values, especially those connected with social and cultural integration, that can be realized in a religiously homogeneous society but are bound to be lost, at least to some extent, in a religiously pluralistic society such as ours. But if I had a choice about which sort of society I would inhabit, I would opt for a religiously pluralistic society. So I am a supporter of the religious arrangements commonly found under modern liberal constitutional regimes. Believing what I do about loss, however, I consider it incumbent on me to be attentive to the costs such regimes impose on

494

John Rawls religion and their religious citizens. A big question for me therefore is how to arrange such costs so that they are not an excessive burden on religious citizens. In my opinion, the version of political liberalism developed in the past two decades by Rawls is the best attempt so far to work out in detail a liberal political theory according to which the costs liberal regimes should impose on religious citizens are, though real, close to minimal. An important innovation in that theory is its distinctive idea of public reason, which is explained in Lecture VI of the book Political Liberalism and then refined in the article "The Idea of Public Reason Revisited." Can religious citizens live comfortably within the limits of public reason and should they be willing to do so? Some philosophers favor negative answers to these questions. In a pair of powerfully argued book chapters, "Why We Should Reject What Liberalism Tells Us" and "The Role of Religion in Decision and Discussion of Political Issues," Nicholas Wolterstorff has directed critical fire against Rawls's version of the liberal position. I think much of this critical fire misses the mark, and I argue for that conclusion in this paper. Its first part is devoted to explaining the Rawlsian idea of public reason; its second part defends that idea against some of Wolterstorff s critical assaults. I. THE RAWLSIAN IDEA OF PUBLIC REASON EXPLAINED For Rawls, public reason is the reason of democratic citizens, who share the status of equal citizenship, on the subject of the good of the public. It is to be contrasted with the reasons of rulers in aristocratic and autocratic regimes and with the nonpublic reasons, within a democratic society, of churches, universities and many other associations within civil society. Two structural features Rawls attributes to it deserve to be emphasized. The first is the fundamental political questions to which it applies; they are questions of constitutional essentials and matters of basic justice. Constitutional essentials concern the general structure of government and political processes as well as the equal basic political rights and liberties of citizens. Matters of basic justice have to do with the basic structure of society and, in particular, how it ought to respond to social and economic inequalities. The second feature is the persons to whom the idea of public reason applies; they are persons 'John Rawls, Political Liberalism (New York: Columbia University Press, 1993), xxv. (Hereafter cited as PL,) 2 PL, xvi. 3 Robert M. Adams, "Religious Ethics in a Pluralistic Society," Prospects for a Common Morality, eds. Gene Outka and John P. Reeder, Jr. (Princeton: Princeton University Press, 1993), 97. (Hereafter cited as "REPS") 4 Robert M. Adams, Finite and Infinite Goods: A Frameworkfor Ethics (New York and

106

Oxford: Oxford University Press, 1999), 368. (Hereafter cited as FIG.) *PLy 197. 6/r/G, 334. 7 John Rawls, "The Idea of Public Reason Revisited," The University of Chicago Law Review 64 (1997), 767. (Hereafter cited as "IPRR.") 8 PL, 217. 9 "IPRR," 769. l °Ibid.,169. n PL, 218.

John Rawls

495

who conduct discussions of the fundamental political questions in the public political forum. Rawls specifies this forum quite narrowly. He says it "may be divided into three parts: the discourse of judges in their decisions, and especially of the judges of a supreme court; the discourse of government officials, especially chief executives and legislators; and finally, the discourse of candidates for public office and their campaign managers, especially in their public oratory, party platforms, and political statements."7 To get around the problem of drawing a line separating candidates and those who run their campaigns from other politically engaged citizens, Rawls by stipulation makes candidates and those who run their campaigns responsible for what others say or do on behalf of candidates. It is thus important to be clear at the outset that the idea of public reason is meant to apply only to questions of constitutional essentials and matters of basic justice and only to discussions of them in the public political forum, narrowly circumscribed. Public reason helps to specify an ideal conception of citizenship for a constitutional democratic regime. Rawls thinks it is a normative ideal. Thus it imposes on citizens "a moral, not a legal duty — the duty of civility — to be able to explain to one another on those fundamental questions how the principles and policies they advocate and vote for can be supported by the political values of public reason."8 How do citizens live up to the ideal and fulfill the duty? Judges, chief executives, legislators and candidates for public office do so when they "act from and follow the idea of public reason and explain to other citizens their reasons for supporting fundamental political positions in terms of the political conception of justice they regard as the most reasonable."9 And how do citizens who are not government officials comply with the duty of civility? Recently Rawls has said that they "fulfill their duty of civility and support the idea of public reason by doing what they can to hold government officials to it."10 But earlier he had presented his ideal of democratic politics as asking more than that. One of the tasks of all democratic citizens is that they should at least try to "be ready to explain the basis of their actions to one another in terms each could reasonably expect that others might endorse as consistent with their freedom and equality."11 It is not clear, at least to me, whether he then thought that performing this task is required by the duty of civility and, if he did, whether he has since backed off from that view. What does seem clear is that the duty of civility requires government officials and political candidates to use the political conception of justice they find most reasonable to explain to their fellow citizens why they support the positions they do on constitutional essentials and matters of basic justice and requires other citizens to try to hold government officials and political candidates to conducting themselves in this manner. And it also seems clear that Rawls once thought that citizens who are not government Religious Citizens Within the Limits of Public Reason Philip L. Quinn 107

496

John Rawls

officials or political candidates should, in order to realize fully the ideal of public reason, do what they can to make themselves ready to explain to one another their positions and actions regarding constitutional essentials and matters of basic justice in terms each could reasonably expect that the others might endorse as consistent with their freedom and equality. What kinds of normative constraints do the ideal of public reason and the duty of civility it imposes actually involve? In order to address this question, it is important to understand Rawls's idea of a political conception of justice and how he specifies the reasonable. Let us begin with the idea of a political conception of justice. While a political conception is a moral conception, it is restricted in some significant ways. According to Rawls, political conceptions have the following three features: "First, their principles apply to basic political and social institutions (the basic structure of society); second, they can be presented independently from comprehensive doctrines of any kind (although they may, of course, be supported by a reasonable overlapping consensus of such doctrines); and finally, they can be worked out from fundamental ideas seen as implicit in the public political culture of a constitutional regime, such as the conceptions of citizens as free and equal persons, and of society as a fair system of cooperation."12 The first feature insures that political conceptions are limited in scope; they do not apply to every issue where justice is at stake. The second guarantees that they are freestanding relative to comprehensive philosophical and religious doctrines in the sense that presentations of them need not appeal to such comprehensive doctrines. And the third specifies a source for political conceptions that is generally accessible to and widely shared by citizens of a constitutional democracy. Rawlsian political liberalism is, of course, particularly concerned with liberal political conceptions of justice. Liberal political conceptions are political conceptions with the following additional features: "First, a list of certain basic rights, liberties, and opportunities (such as those familiar from constitutional regimes); Second, an assignment of special priority to those rights, liberties, and opportunities, especially with respect to the claims of the general good and perfectionist values; and Third, measures ensuring for all citizens adequate allpurpose means to make effective use of their freedoms."13 These features help to configure distinctively liberal substantive principles of justice for political conceptions. In addition to their substantive principles of justice, liberal political conceptions also contain guidelines for inquiry that "specify ways of reasoning and criteria for the kinds of information relevant for political questions."14 The substantive principles of justice and the guidelines for inquiry in turn specify political values of two sorts. The substantive principles give rise to values of political justice;

12

"IPRR," 776. lbid., 774. 14 PL, 223. 15 "IPRR," 777. I3

108

16/Wrf., 773. Ibid., 774. ^Ibid., 775. 19PL, 49.

]1

John Rawls

497

they include equal political and civil liberty and equality of opportunity. The guidelines generate values of public reason, which include political virtues such as reasonableness and a readiness to honor the duty of civility, Rawls also insists that liberal political conceptions should be complete. For him, this means that "each conception should express principles, standards, and ideals, along with guidelines of inquiry, such that the values specified by it can be suitably ordered or otherwise united so that those values alone give a reasonable answer to all, or to nearby all, questions involving constitutional essentials and matters of basic justice."15 It is worth calling attention at this point to the fact that Rawls has deliberately characterized liberal political conceptions of justice rather abstractly. It follows directly that "the content of public reason is given by a family of political conceptions of justice, and not by a single one."16 Even the liberal political conception Rawls himself prefers, justice as fairness, which makes use of the representational device of the original position in specifying its substantive principles of justice, is only one liberal political conception among many. To be sure, such conceptions are united in their endorsement of the general ideas of citizens as free and equal persons and of society as a fair system of cooperation over time. However, as Rawls hastens to add, "since these ideas can be interpreted in various ways, we get different formulations of the principles of justice and different contents of public reason."17 Rawlsian political liberalism does not aspire to fix public reason in the form of just one liberal political conception. Even if one or a few conceptions were, in the course of a society's historical development, to become dominant, "the forms of permissible public reason are always several."18 In short, public reason does not demand of citizens of a constitutional democracy, even at the level of principle, unanimity on political questions, even when constitutional essentials or matters of basic justice are at stake. As we have seen, the idea of the reasonable enters into Rawls's characterization of the duty of civility as it applies to government officials and political candidates, into his description of a task he once thought the ideal of public reason sets for democratic citizens who are not government officials or political candidates, into his specification of the political values of public reason, and into his definition of completeness for liberal political conceptions of justice. So let us now turn to what he has to say about this idea. Rawls contrasts the reasonable with the rational, and he does not try to define the reasonable directly. Instead, he specifies two of its aspects as virtues of persons. The first is this: "Persons are reasonable in one basic aspect when, among equals say, they are ready to propose principles and standards as fair terms of cooperation and to abide by them willingly, given the assurance that others will likewise do so "19 Proposing principles and standards as fair terms Religious Citizens Within the Limits of Public Reason Philip L. Quinn

109

498

John Rawls

of cooperation involves taking them to be reasonable for everyone to accept. The second aspect is "the willingness to recognize the burdens of judgment and to accept their consequences for the use of public reason in directing the legitimate exercise of political power in a constitutional regime."20 The burdens of judgment are factors that contribute to explaining reasonable disagreement under conditions of free inquiry. According to Rawls, they include conflicting and complex evidence, disagreement about the weights to assign to various considerations, conceptual vagueness or indeterminacy, the influence of differing total life-experiences on how we balance values, difficulties in coming up with an overall assessment of competing normative considerations, and being forced to select among values that compete for room in a social space that cannot fully realize them all. Accepting the consequences of these burdens, and others like them, involves acknowledging that reasonable people can disagree about how to order or balance political values when they are applied to cases even if they agree about which political values are applicable. As Rawls observes, liberal political conceptions differ not only about substantive principles but also "in how they order, or balance, political principles and values even when they specify the same ones."21 What is more, even within a single liberal political conception, there can be reasonable disagreement about how to order or balance its political values in particular cases. Thus it is only to be expected that "reasonable political conceptions of justice do not always lead to the same conclusion; nor do citizens holding the same conception always agree on particular issues."22 In other words, public reason itself often allows more than one reasonable answer to a particular question. What are we to do in such situations? Rawls is aware that some will be tempted to say that, since public reason has failed to resolve the question, "citizens may legitimately invoke principles appealing to nonpolitical values to resolve it in a way they find satisfactory "23 But he thinks the ideal of public reason urges us not to proceed in this way on questions of constitutional essentials and basic justice. We should instead stick to considering the question exclusively in terms of what we regard as our political conception. In other words, "we should sincerely think that our view of the matter is based on political values everyone can reasonably be expected to endorse."24 The question can then be resolved by a vote. When we vote, however, the ideal of public reason is sustained if the question has been debated by appeal to political values and citizens vote on the basis of political values they sincerely believe everyone might reasonably endorse. In such cases, "what public reason asks is that citizens be able to explain their vote to one another in terms of a reasonable balance of public political values, it being understood by everyone that of course the plurality of reasonable ™Ibid., 54. 2I "IPRR," 774. *Ibid., 798. 23 PL, 240.

110

uibid., 241. *Ibid., 243. 26 "IPRR," 770.

John Rawls

499

comprehensive doctrines held by citizens is thought by them to provide further and often transcendent backing for those values."25 So even in situations of reasonable disagreement in which questions are to be settled by a vote the ideal of public reason has some bite. It proposes to democratic citizens a task of abstinence; they should abstain from going outside the realm of political values in search of a basis for discussion and voting. It also proposes to them a task of articulateness; they should be prepared to explain their votes to each other in terms of a reasonable balance of political values. Perhaps an example will serve to make it clear that the task of abstinence is not trivial. Consider the issue of capital punishment. It is obviously a matter of basic justice, and so the ideal of public reason urges us to consider it exclusively within our political conceptions. Suppose the main political values that bear on the issue are respect for human life, public safety and compensation for victims of serious crimes or their relatives. I myself think that the most reasonable balance of these values, one that gives very great weight to respect for human life, supports opposition to capital punishment. In his comments on this paper at the fourth Henle Conference, Jeffrie G. Murphy constructed an argument for opposing America's harsh system of criminal punishment, particularly the death penalty, that appeals to comprehensive Christian doctrine. No doubt some Christians will believe that Murphy's argument, or something like it, articulates their deepest reasons for opposing capital punishment. Even if they agree with my judgment about the most reasonable balance of political values, it may seem to them to provide only relatively shallow reasons for opposition to capital punishment. I sympathize with this point of view. So I am prepared to grant that performing the task of abstinence will in some cases involve appealing only to shallow reasons even though one has deeper reasons for one's position. For Rawls, of course, this is simply a price that must be paid if, given the fact of reasonable pluralism, we are to realize the ideal of living politically with one another in the light of reasons all might reasonably be expected to endorse. Yet it is a real price, and it may not always be a small one. I think the deepest consideration Rawls mobilizes in support of this way of responding to reasonable disagreement about questions of constitutional essentials and matters of basic justice is an appeal to a criterion of reciprocity. When citizens propose fair terms of cooperation to one another according to what they regard as the most reasonable political conception, he tells us, the criterion of reciprocity requires that they "must also think it at least reasonable for others to accept them, as free and equal citizens, and not as dominated or manipulated, or under the pressure of an inferior political or social position."26 To be sure, citizens will disagree about which political conception is the most reasonable, and so public reason can Religious Citizens Within the Limits of Public Reason Philip L. Quinn 111

500

John Rawls lead to a stand-off on particular issues. Yet Rawls hopes that citizens will agree that several political conceptions are reasonable, even if some of them are barely so. When citizens appeal to the political conceptions they regard as most reasonable and a stand-off results, they can think it reasonable that others, who disagree with them about what is most reasonable, can at least agree that their proposals are reasonable. If citizens appeal beyond such political conceptions to other parts of their comprehensive doctrines as a basis for their proposals, they will not, according to Rawls, reasonably think others could reasonably accept those proposals. Hence when stand-offs occur and are to be resolved by vote, "citizens must vote for the ordering of political values they sincerely think the most reasonable. Otherwise they fail to exercise political power in ways that satisfy the criterion of reciprocity."27 And to violate reciprocity is to flout the ideal of public reason. Or at least this is so under what we might think of as the best-case scenario. In Political Liberalism, Rawls distinguishes between the exclusive view and the inclusive view of public reason. According to the exclusive view, "reasons given explicitly in terms of comprehensive doctrines are never to be introduced into public reason."28 Rawls suggests that the exclusive view would be appropriate for the highly idealized case of a more or less well ordered society untroubled by deep disputes. To claim that a society is well ordered is to say three things: first, it is a society in which everyone accepts, and knows that everyone else accepts, the same principles of justice; second, its basic structure is publicly known, or with good reason believed, to satisfy these principles; and third, "its citizens have a normally effective sense of justice and so they generally comply with society's basic institutions, which they regard as just."29 Hence in a more or less well ordered society there is just one political conception. Rawls supposes that "in this case the values of the political conception are familiar and citizens honor the ideal of public reason most clearly by appealing to those values."30 Even in a more or less well ordered society there is, of course, room for reasonable disagreement, as a consequence of the burdens of judgment, about how to apply the shared principles and balance the values of the political conception in particular cases. But, absent deep disputes, there is no need for citizens to go outside the political conception they share to locate a balance of values they consider most reasonable and also sincerely regard as reasonable for their fellow citizens to accept. So the exclusive view, which in effect has been the focus of my discussion thus far, seems well suited for the highly idealized case of a more or less well ordered society not stirred by deep disputes. Rawls, however, acknowledges that the inclusive view is better suited to encouraging citizens to honor the ideal of public reason and to securing its social

"Ibid., 797. 2 *PL,247. »/fó¿, 35. wibid., 248.

112

*Ibid.t 247. 327Wtf.,251. 33/Wi, 253.

John Rawls

501

conditions in the longer run in less idealized cases. On that view, citizens are allowed, in some circumstances, "to present what they regard as the basis of political values rooted in their comprehensive doctrine, provided they do this in ways that strengthen the ideal of public reason itself (my emphasis).31 By way of illustration, Rawls discusses two cases. The first involves a nearly well ordered society in which there is a serious dispute about the application of one of the shared principles of justice. Suppose religious groups dispute about whether the principle of fair equality of opportunity, as applied to education, supports state aid to parochial schools. Rawls imagines that religious leaders might, by introducing portions of their comprehensive doctrines into the public forum, show how those doctrines affirm the values of the shared political conception. If the religious leaders succeeded in doing this, they would, Rawls thinks, strengthen the ideal of public reason and thereby satisfy the proviso of the inclusive view. His second case involves a society that is not well ordered in which there is a profound division about constitutional essentials. He asks us to consider the historical examples of the nineteenth-century abolitionists and the civil rights movement led by Martin Luther King, Jr. His claim is that they did not go against the ideal of public reason, even though they appealed to comprehensive doctrines, "provided they thought, or on reflection would have thought (as they certainly could have thought), that the comprehensive reasons they appealed to were required to give sufficient strength to the political conception to be subsequently realized."32 Though this remark is a bit cryptic, I take it the idea behind it is that the abolitionists and civil rights leaders would have satisfied the proviso of the inclusive view if they had thought that invoking comprehensive doctrines was the best way, or perhaps the only way, open to them to contribute to bringing about a society in which the ideal of public reason eventually could be honored. And as we move away from highly idealized cases, it seems that Rawls is willing to relax, to some extent, other constraints such full satisfaction of the criterion of reciprocity, ideally, public reason asks that the balance of political values we consider most reasonable is one we sincerely believe others can also see as reasonable, if only barely so. But Rawls allows that the best we can do on some questions may be to strike a balance we think "can be seen as at least not unreasonable in this sense: that those who oppose it can nevertheless understand how reasonable persons can affirm it."33 The general idea, I presume, is that when, in unfavorable circumstances, full compliance with the ideal of public reason is not feasible, our departures from full compliance should be both reluctant and minimal. In "The Idea of Public Reason Revisited," Rawls signals a change of mind when he replaces the inclusive view with the wide view of public political culture. Religious Citizens Within the Limits of Public Reason Philip L. Quinn

113

John Rawls

502

Like the inclusive view, the wide view is specified in terms of a proviso, but its proviso is more permissive than that of the inclusive view. According to the wide view, "reasonable comprehensive doctrines, religious or nonreligious, may be introduced in public political discussion at any time, provided that in due course proper political reasons — and not reasons given solely by comprehensive doctrines — are presented that are sufficient to support whatever the comprehensive doctrines are said to support" (my emphasis).34 Rawls is aware that the wide view's new proviso will give rise to questions. When does it need to be satisfied, and who is obliged to satisfy it? However he professes not to be able provide much help in answering such questions. He thinks the details of how to satisfy it "must be worked out in practice and cannot feasibly be governed by a clear family of rules given in advance."35 An advantage of the wide view is that it enables Rawls to simplify his treatment of the abolitionists and the leaders of the civil rights movement. He no longer needs to speculate about the truth of counterfactuals concerning what they would have thought on reflection. Instead he simply notes that the wide view's proviso "was fulfilled in their cases, however much they emphasized the religious roots of their doctrines, because these doctrines supported basic constitutional values — as they themselves asserted — and so supported reasonable conceptions of political justice."36 Even if they had not thought that the measures they advocated were supported by a reasonable balance of political values, we see that they are so supported, and on that account the new proviso is satisfied in their cases. I conjecture that Rawls moved from the inclusive view to the wide view in part because he wanted to address the concerns of some religious citizens that his political liberalism is biased toward the secular and unnecessarily exclusive of the religious. But there is only so far he can move while remaining wedded to the core idea of liberalism. For Rawls, I think, this is its principle of legitimacy. According to that principle, our exercise of coercive political power "is proper and hence justifiable only when it is exercised in accordance with a constitution the essentials of which all citizens may reasonably be expected to endorse in the light of principles and ideals acceptable to them as reasonable and rational."37 So even when the proviso of the wide view is in place, ultimate political justification must be in terms of a reasonable balance of the values of a political conception of justice and so must be conducted within the limits of public reason. As Rawls puts it, "this justification is still given in terms of a family of reasonable political conceptions of justice."38 Nor does a core idea of this sort seem to be distinctive of Rawlsian political liberalism. Something similar is at the heart of Charles Larmore's liberalism of equal

^"BPRR," 783-784. Ibid., 784. 36 /¿>z¿, 785-786. 37 />L, 217. 38 "IPRR," 784. 35

114

39Oiarles Larmore, The Morals of Modernity (Cambridge: Cambridge University Press, 1996), 137. 40pL> 226. «I'TPRR," 777-778.

John Rawls

503

respect. As Larmore sees it, "To respect another person as an end is to insist that coercive or political principles be as justifiable to that person as they are to us. Equal respect involves treating in this way all persons to which such principles are to apply."39 The common idea is that liberal justification of at least the fundamental principles of the exercise of coercive political power must be justification to all persons to whom those principles are to be applied. Rawls is candid about the ways in which the ideal of public reason is supposed to constrain discussion in the public forum. It asks citizens to submit to the discipline of working up and being prepared to defend a political conception of justice. For Rawls, "the point of the ideal of public reason is that citizens are to conduct their fundamental discussions within the framework of what each regards as a political conception of justice based on values that the others can reasonably be expected to endorse and each is, in good faith, prepared to defend that conception so understood."40 It also asks citizens to submit to the discipline of using their political conceptions as a filter on arguments that makes the route from their comprehensive doctrines to ultimate political justification inevitably indirect. According to Rawls, "What we cannot do in public reason is to proceed directly from our comprehensive doctrine, or a part thereof, to one or several political principles and values, and the particular institutions they support. Instead, we are required first to work to the basic ideas of a complete political conception and from there to elaborate its principles and ideals, and to use the arguments they provide."41 We must proceed by way of political conceptions because in public reason we aim at justification in terms of values all can reasonably be expected to endorse. And we must proceed by way of complete political conceptions because in public reason we aim at a unified system of answers to the full range of fundamental questions of constitutional essentials and basic justice. Neither of these requests is trivial. Bearing the yoke of the discipline they call for will not be equally easy for all citizens and cannot be expected always to be easy for each citizen. Is this too much to ask of the religious citizens of a liberal democracy? Nicholas Wolterstorff thinks so. I now turn to an examination of his views. II. THE RAWLSIAN IDEA CRITICIZED AM) DEFENDED Let me make two preliminary remarks before I begin explaining and criticizing WolterstofFs arguments. The first is that he himself is committed to some liberal political values; the final paragraph of his book, John Locke and the Ethics of Belief, makes this very clear. It says this: Yet we must live together. It is to politics and not to epistemology that we shall have to look for an answer as to how to do that. "Liberal" poliReligious Citizens Within the Limits of Public Reason Philip L. Quinn 115

504

John Rawls

tics has fallen on bad days recently. But to its animating vision of a society in which persons of diverse traditions live together injustice and friendship, conversing with each other and slowly altering their traditions in response to their conversation — to that, there is no viable alternative.42 So Wolterstorff does not reject liberalism's animating vision; he thinks there is no viable alternative to endeavoring to realize and live in accord with it. His quarrel is with what some liberal theorists tell us about what this project requires of us. It is also worth noting for the record that the Rawls Wolterstorff attacks is not exactly the same as the Rawls I shall try to defend. The two book chapters in which Wolterstorff criticizes Rawls, both of which were published in 1997, target the views of Political Liberalism and do not take into account the shifts of emphasis, clarifications, elaborations and even changes of mind contained in "The Idea of Public Reason Revisited," which was also published in 1997. My exposition of Rawls, though it points out some salient differences between the views of Political Liberalism and those of "The Idea of Public Reason Revisited," draws freely on both sources. Hence it presents a picture that is not a perfect match with the one likely to emerge from reading Political Liberalism on its own. Yet it is this picture I want to defend. And I hope to show that it contains the features of Rawlsian liberal theory that Wolterstorff finds generally and most deeply objectionable. We can gain entry into the general perspective from which Wolterstorff approaches liberal theorists if we start by considering an image he uses to paint in broad brushstrokes an aspect of their ideas he rejects. He distinguishes between a parliamentary session and a quaker meeting. In a parliamentary session, after appropriate discussion has occurred, a vote is taken, and the will of a suitable majority constitutes the decision of the body. By contrast, in a quaker meeting, as Wolterstorff is conceiving it, a proposal is regarded as adopted by the body only if it is supported by a consensus, where a consensus involves positive agreement by all and not just the mere acquiescence, on the part of some, represented by failure to dissent. Criticizing Robert Audi and other unnamed liberal theorists, Wolterstorff accuses them of seeing "the concept of liberal democracy as incorporating the ideal of all the citizens together constituting a quaker meeting when it comes to the choosing of laws and governmental policies that are in any way coercive."43 Of course, as Wolterstorff acknowledges, such theorists realize that sometimes, on Nicholas Wolterstorff, John Locke and the Ethics of Belief (Cambridge: Cambridge University Press, 1996), 246. 43 Robert Audi and Nicholas Wolterstorff, Religion in the Public Square: The Place of Religious Convictions in Political Debate (Lanham, MD: Rowman and Littlefield, 1997), 152. (Hereafter cited as RPS.) 42

116

""REPS," 112. Nicholas Wolterstorff, "Why We Should Reject What Liberalism Tells Us," Religion and Contemporary Liberalism, ed. Paul J. Weithman (Notre Dame, IN: University of Notre Dame Press, 1997), 171. (Hereafter cited as "WWSR.") *RPS, 92. 45

John Rawls

505

account of persisting disagreement, the citizens will have no recourse but to constitute themselves as a parliamentary session. But when they do, their society, by falling short of the ideal, either is to that extent not a liberal democracy or is a malformed liberal democracy. Nor is Wolterstorff the first to raise a worry of this sort about Rawlsian liberalism. Writing slightly before Political Liberalism's publication, Robert M. Adams had said that it seemed to him that "Rawls underemphasizes the combative aspects of a democratic polity and tends to overestimate the level of theoretical agreement in political ethics needed for an attainably just society."44 Whatever may be the case for Audi and others, it is surely no part of Rawlsian political liberalism to conceive of the citizens of a liberal democracy as ideally constituting a quaker meeting whenever it comes to choosing coercive laws and policies. As we have seen, the ideal of public reason applies only to questions of constitutional essentials and basic justice. It says nothing against adopting coercive laws and policies on other matters in parliamentary session. What is more, even in the extremely idealized case of a society more or less well ordered by a single political conception, Rawls recognizes, as his example of a serious dispute over government aid of parochial schools makes clear, that there may be disagreement about the application of shared political values in particular situations. In light of the burdens of judgment, reasonable citizens will acknowledge that public reason does not require of them a resolution of such disputes by unanimous agreement. And in less idealized cases, in which public reason takes on several of its permissible forms and society contains a family of political conceptions, there may be reasonable disagreement in particular instances about exactly which political values are applicable and how the applicable political values are to be ordered or balanced. In short, even when constitutional essentials and matters of basic justice are at stake, Rawls leaves room for many issues that it is beyond the scope of public reason to adjudicate in terms of the unanimous agreement of the model of the quaker meeting. Wolterstorff s striking image of the liberal polity as a quaker meeting would, then, be misleading if it were applied to Rawlsian political liberalism. Wolterstorff does not explicitly make such an application. However, he does say things about Rawls that strongly suggest an interpretation which distorts in a similar way. He asserts, for example, that Rawls "employs a consensus populi strategy."45 Though he admits that Rawls does not use the words "consensuspopuli'' he takes it that Rawls's "suggestion, at bottom, is that, in a liberal democracy, the consensus populi ought to be used to form the political basis of the discussions and decisions of the citizens."46 Wolterstorff apparently thinks that this is the suggestion because of the way Rawls specifies the second and third features of political conceptions. As previously noted, political conceptions are supposed to be worked up from ideas Religious Citizens Within the Limits of Public Reason Philip L. Quinn 117

506

John Rawls

implicit in the public political culture of a constitutional regime and freestanding in the sense of being able to be presented independently of comprehensive doctrines. Wolterstorff seems to think Rawls wants political conceptions to have these features in order to guarantee or at least make it likely that there will be a consensus populi on a single favored political conception. He rightly observes that such an outcome is not likely. As he puts it, "no matter what principles of justice a particular political theorist may propose, the reasonable thing for her to expect, given any plausible understanding whatsoever of 'reasonable and rational,' is not that all reasonable and rational citizens would accept those principles, but rather that not all of them would do so."47 But, not only does Rawls not deny this point, he insists that even within the limits of public reason there will be different formulations of the principles of justice. It seems to me Rawls is correct to insist that the content of public reason cannot be fixed in the form of a single liberal political conception of justice; the fundamental ideals of the public political culture from which it derives can reasonably be interpreted in different ways and are likely to continue to permit such diverse interpretations as long as free inquiry endures. When Wolterstorff notices that Rawls does not think that public reason requires us all to agree on a single political conception of justice, he regards this as remarkable. He takes Rawls to be making, in the interests of realism, the concession that "we do not need, for political life in a liberal democracy, principles endorsed by all reasonable and rational citizens."48 And he concludes that such unanimous endorsement of principles and so public reason itself must be ideals, the difficulty being that they may never be anything more. Of course, it may be that Rawls thinks that it would be ideal in some sense if we were all to endorse the same principles of justice, but public reason does not demand this of us. And he does subscribe to an ideal of public reason, but it can be satisfied even in the absence of unanimous endorsement of principles of justice. Yet it does not occur to Wolterstorff to reconsider his claim that Rawls employs a consensus populi strategy. This is unfortunate. Had he done so, he might have seen that Rawlsian political liberalism does not insist on and, in fact, does not even aspire to consensus populi about the content of public reason or at the level of principles of justice. Perhaps Rawls does indeed aspire to a level of agreement in political morality that is neither feasible nor desirable. But his ideal of public reason does not require anyone to aspire to the level of agreement of Wolterstorff s imaginary quaker meeting. Nor does public reason aim at a consensus populi on a single political conception or a single set of principles of justice. On these issues Wolterstorff's critique sees Rawlsian political liberalism through a distorting lens. Rawls simply does not 47

/¿>/d.,99. **Ibid., 101. 49 Ibid., 105. 50 Ibid., 105.

118

5i«\vwSR," 166. Ibid., 176. 53/¿,¿a.f 176.

52

John Rawls

507

aspire to agreement in political morality to the extent Wolterstorff thinks he does, much less to the level Wolterstorff attributes to Audi and other unnamed liberal theorists. It seems clear to me, however, that Wolterstorff's main quarrel with Rawls does not lie in the area of how much agreement it is reasonable to expect or aspire to in a liberal society. Even if Wolterstorff were to concede that Rawls, perhaps unlike some other liberal theorists, does not make the unreasonable or unrealistic demands implicit in the quaker meeting image or the consensus populi strategy, he would still object to conducting political argument within the limits of Rawlsian public reason because he thinks its limits impose an unfair burden on religious citizens of certain sorts. Where the shoe really pinches, for Wolterstorff, is where it excludes, as he sees it, the reasons of comprehensive religious doctrines from the public forum. I read him as having two arguments for the unfairness of the exclusion. The first focuses on the idea of the free exercise of religion, and so I shall call it the free exercise argument. The second emphasizes the concept of living a religiously integrated existence, and so I shall call it the integrity argument. The free exercise argument comes in two versions. The premise of the first concerns the religious convictions of some religious citizens. Wolterstorff says: "It belongs to the religious convictions of a good many religious people in our society that they ought to base their decisions concerning fundamental issues of justice on their religious convictions."49 The conclusion he draws from this premise is this: "Accordingly, to require of them that they not base their decisions and discussions concerning political issues on their religion is to infringe, inequitably, on the free exercise of their religion."50 In order to understand the second version of the free exercise argument, we need to grasp what Wolterstorff describes as liberalism's independent-basis principle. It is the principle that citizens of a liberal democracy "are to base their political debate in the public space, and their political decisions, on the principles yielded by some source independent of any and all of the religious perspectives to be found in the society."51 The premise of second version is closely related to the conclusion of the first. Wolterstorff claims: "Should someone try to stop me from voting, and acting politically, on the basis of my religious convictions, that would violate the free exercise of my religion."52 From this claim he infers immediately the following conclusion: "Accordingly, if honoring the freedom and equality of citizens did require adherence to the independent-basis principle, then honoring the freedom and equality of citizens would also require non-adherence."53 What are we to make of this chain of arguments? I think both of them prove to be, upon examination, bad arguments. Consider the first version. Its premise is undoubtedly true. There are religious citizens whose religious convictions include the belief that they should base their political Religious Citizens Within the Limits of Public Reason Philip L. Quinn

119

508

John Rawls

responses to fundamental questions of justice on their religion. Its conclusion, however, is ambiguous. If there were a legal requirement that they not do so, it would clearly infringe on the free exercise of their religion. But Rawlsian political liberalism proposes no such legal duty. As previously noted, its duty of civility is a moral rather than a legal duty. Rawls is explicit about how this distinction bears on constitutionally protected liberties. He emphasizes that the duty of civility is not a legal duty because "in that case it would be incompatible with freedom of speech."54 Similarly, the duty of civility would be incompatible with the free exercise of religion if it were a legal duty. So if the conclusion of the argument is to make contact with Rawls, it must be understood to say that morally requiring religious citizens not to base their political activity on their religion is an infringement of the free exercise of their religion. But thus interpreted the conclusion is false, and the argument is unsound. Morally prohibiting religious citizens from basing their political activity on religious convictions would no more infringe on their free exercise of religion than would morally prohibiting them from treating homosexuals with a lack of respect for religious reasons. In both cases, the free exercise clause constitutionally protects conduct that violates moral requirements, and as long as religious citizens are legally free to violate such requirements there has been no infringement of the free exercise of their religion. Turn now to the second version. Something closely akin to its premise is true. Someone who tried to stop Wolterstorff from acting politically on the basis of his religious convictions would, at least if the attempt succeeded, violate the free exercise of his religion, assuming of course that his religiously based conduct was not contrary to legitimate law. However, nothing in Rawls's political liberalism justifies anyone in making an attempt of this sort, and so it is not easy to see how this premise is supposed to support a conclusion applicable to his view. Indeed I do not even find it easy to make sense of the conclusion's claim that if honoring the freedom and equality of citizens required adherence to the independent-basis principle, it would also require non-adherence. Evidently Wolterstorff thinks the independentbasis principle is self-undermining or something of that sort, but some work is needed to get at his reasons for thinking so. I presume he thinks Rawls is committed to the conclusion's antecedent. It is plausible to attribute Rawls the thought that we are morally required to adhere to the principle in order to honor the freedom and equality of our fellow citizens because, if we did not, we would act politically on a basis they could not reasonably endorse consistent with their freedom and equality. So suppose the conclusion's antecedent is true on this interpretation. On this assumption, the conclusion's consequent should be taken to assert that we are also morally required not to adhere to the principle in order to honor the freedom and equality of our fellow citizens. The only reason I can come up with that might be offered in 54

"IPRR," 769. "WWSR," 165.

55

120

^Ibid., 176-177.

John Rawls

509

support of this assertion and invokes the free exercise of religion is the claim that we would violate the free exercise of religion by some of our fellow citizens and so fan to honor their freedom and equality if we adhered to the principle. But this claim is false. If we adhered to the principle, we would be committed to thinking that religious citizens who violated it when they exercised their religious freedom were doing something contrary to duty. However we would not be violating their free exercise unless we also went on to prevent them from acting in this way. Hence I can see no good reason deriving from the free exercise of religion that supports the conclusion's consequent. What is more, its consequent is pretty clearly false. We do not violate the religious freedom of fellow citizens and thus fail to honor their freedom and equality when we adhere to a principle according to which some legally permissible exercises of that freedom are contrary to moral duty, and so we are not morally required not to adhere to such a principle in order to honor their freedom and equality. Therefore, on an interpretation of the argument's conclusion which at least has the merit of making the idea of free exercise of religion cited in its premise relevant to the conclusion, the conclusion's antecedent is true while its consequent is false. Thus the conclusion as a whole is false, and the argument is unsound. The integrity argument is meant by Wolterstorff to challenge a wide range of liberal positions. It is directed against the independent-basis principle as well as against the separation principle, according to which "government is to do nothing to advance or hinder any religion."55 Because Rawls does endorse a version of the independent-basis principle when he includes the duty of civility within public reason, the argument applies to his view. Wolterstorff sees a common pattern in the liberal's impression that these two principles deal fairly with religion. His argument, which is worth quoting at length, continues as follows: That common pattern is this: The liberal assumes that requiring religious persons to debate and act politically for reasons other than religious reasons is not in violation of their religious convictions; likewise he assumes that an educational program which makes no reference to religion is not in violation of any parent's religious convictions. He assumes, in other words, that though religious people may not be in the habit of dividing their life into a religious component and a nonreligious component, and though some might be unhappy doing so, nonetheless, their doing so would not be in violation of anybody's religion. But he's wrong about this. It's when we bring into the picture persons for whom it is a matter of religious conviction that they ought to strive for a religiously integrated existence — it's then, especially, though not only then, that the unfairness of liberalism to religion comes to light.56 Religious Citizens Within the Limits of Public Reason Philip L. Quinn

121

510

JohnRawls

Wolterstorff is obviously quite fond of this argument. It appears almost word-forword identically in both the book chapters in which he attacks liberalism!57 I think Wolterstorff is on to something important when he makes this argument, though it will take some work to get clear about exactly what it is. Wolterstorff seems to believe the argument shows that liberalism is unfair to religion as such or to religious people generally, but it is not strong enough to do that. While there are, to be sure, some religious people for whom it is a matter of religious conviction that they ought to strive for a religiously integrated'existence, there are others for whom it is not. Some people who are religiously devout are content to separate their lives into religious and non-religious spheres, acting politically for reasons other than religious reasons does not violate their religious convictions. So even if the argument is successful, its conclusion applies only to some and not to all religious citizens. In addition, there are non-religious people whose comprehensive doctrines make it a matter of conviction for them that they ought to strive for lives integrated by those doctrines. Examples include some secular Millian liberals and Marxist socialists. Hence if the argument succeeds, a parallel argument will show that liberalism is also unfair to non-religious citizens of these kinds. In other words, the question we ought to be asking is not whether liberalism is unfair to religion but whether it is unfair to a group of people, some of whose members are religious while others are not, all of whom as a matter of conviction want to live in ways tightly integrated around their comprehensive doctrines. For the sake of convenience, let us call the members of this group "the integralists." It seems to me appropriate to ask whether Rawlsian political liberalism in particular is unfair to the integralists. As previously noted, public reason asks citizens generally to work up and be prepared to defend a political conception of justice; according to Rawls, "it is central to political liberalism that free and equal citizens affirm both a comprehensive doctrine and a political conception."58 What is more, it also asks them to argue politically from their political conceptions rather than directly from their comprehensive doctrines. Of course it asks these two things of all free and equal citizens. But compliance is likely to be a heavier burden for the integralists than for others, because they would prefer to argue politically directly from their comprehensive doctrines and so will be skeptical of the value of working up a political conception that is to be interposed, so to speak, between their comprehensive doctrines and the public political forum. They may also fear, with some justification, that people of their sort are bound to decline in numbers and influence in a society in which the ideal of public reason becomes more fully realized over time.

Compare RPS, 116. "IPRR," 800. 59 PL, 197. «Mid., 198. 61 Philip L. Quinn, "Political Libearlisms and 57

58

122

Their Exclusions of the Religious," Religion and Contemporary Liberalism, ed. Paul J. Weithman (Notre Dame, IN: University of Notre Dame Press, 1997), 156. 62 "IPRR," 803.

John Rawls

511

It should be granted to Wolterstorff, then, that political liberalism does impose special burdens on the integralists. It may even in the long run, to the extent that it comes to prevail in a society, doom integralist forms of life, religious and non-religious alike, to extinction. But should it also be granted that Rawlsian political liberalism is, on that account, unfair to the integralists? I think not. Along with Rawls, I endorse Berlin's view that there is no social world without loss. The fact, should it turn out to be a fact, that the integralists cannot, as a matter of social necessity, reproduce themselves over time in a society well ordered by political liberalism would no doubt be cause for regret. However, as Rawls insists, such social necessities "are not to be taken for arbitrary bias or injustice."59 Even the relatively capacious culture and institutions of liberal democracy are bound to prove uncongenial to some valuable forms of life, and integralism of various stripes may be among them. Obviously we must, with Rawls, reject the notion that "only unworthy forms of life lose out in a just constitutional regime."60 Once that has been done, there remains no good reason to think that religion as such or integralism in general has been unfairly treated if integralist forms of life, religious and otherwise, fail to pass the historical test of enduring and gaining adherents in a liberal democratic society. So I conclude that Wolterstorff has not shown that Rawlsian political liberalism is unfair to integralist religion, much less to religion as such. As I see it, therefore, Wolterstorff has established neither that Rawlsian political liberalism demands too much agreement in their views from those who live under a regime ordered by it nor that it is unfair to religious citizens. In other words, he has not shown that it asks too much of the religious citizens of a liberal democracy. But it may ask too much even if he has not proved that it does. Does it? I am on record as having argued that an inclusivist ideal of public political discourse is more attractive, all things considered, than the ideal of public reason proposed by Rawls.61 In concluding, let me return briefly to this issue. It must be acknowledged that Rawls's position has become more inclusive as a result of the substitution of the wide view of public political culture for the inclusive view of public reason. Does his ideal remain, by my lights, too restrictive? This is not an easy question to answer. Rawls himself modestly concludes: "I do not know how to prove that public reason is not too restrictive, or whether its forms are properly described. I suspect it cannot be done."62 He thinks this will not be a serious problem if, as he believes, most cases fit the framework of public reason and those that do not have special features that enable us to understand why they cause difficulty and show us how to cope with them as they arise. And I, of course, do not know how to prove that public reason is too restrictive and suspect that cannot be done either. Nor can I show that we will be unable to deal with the hard Religious Citizens Within the Limits of Public Reason Philip L. Quinn 123

512

John Rawls

cases in the way Rawls hopes we can. It seems clear to me, moreover, that it would be foolish to abandon the ideal of public reason merely because it is restrictive to some extent. It would be highly desirable if exercises of political coercive power on fundamental questions were able to be justified to all reasonable citizens of a polity in terms of a balance of political values they could reasonably endorse as consistent with their freedom and equality even if some citizens regarded other balances of political values as more reasonable to endorse. Surely there will be some restrictions on political discourse and voting whose acceptance would be a price well worth paying in order achieve such a desirable state of affairs. Perhaps, given prevailing circumstances in the United States here and now, it is unrealistic to expect the ideal of public reason to be attained. It is evidently not always a serious criticism of an ideal that it is unrealistic. However I think it is a serious criticism of an unrealistic ideal that imposes moral duties. Clearly Rawls conceives of his ideal of public reason as imposing a moral duty, the duty of civility. As Paul Weithman has pointed out, we should be puzzled about how moral ideals can impose moral requirements.63 According to a somewhat different conception of moral ideals, they urge or advise conduct that is morally good but above and beyond the call of duty. On this conception, I am prepared to subscribe to the ideal of public reason, understood as including the wide view of public political culture, despite its restrictions, as a political ideal in our present circumstances. Under different and presently non-actual circumstances, for example, circumstances including the existence of an overlapping consensus on a family of reasonable political conceptions of justice by all the reasonable comprehensive doctrines present in our society, I would be prepared to subscribe to the ideal of public reason conceived in the way Rawls thinks of it, as imposing a duty of civility. I take these two commitments to be a way of subscribing to articles of a liberal political faith. But I am not persuaded that the ideal of public reason, interpreted as adapted to present circumstances in the United States, imposes moral duties such as the duty of civility. So, to ask the Kantian question implicit in the title of my paper, should religious citizens here and now live within the limits of Rawlsian public reason? I think it would be, morally speaking, very good indeed if they did, and I would encourage them to try to do so. I do not, however, think they presently act contrary to duty or are guilty of wrongdoing if, chafing at even the modest restrictions implied by the wide view, they choose not to live within those limits.64

63 Paul J. Weithman, "Citizenship, Reflective Endorsement and Political Autonomy," The Modern Schoolman, 78(2001): 135-49.

124

M l am grateful to Jeffrie G. Murphy, who was my commentator at the fourth Henle Conference, and to members of the conference's audience for stimulating discussion.

[19] The completeness of public reason Micah Schwartzman

University of Virginia School of Law, USA

abstract

A common objection to the idea of public reason is that it cannot resolve fundamental political issues because it excludes too many moral considerations from the political domain. Following an important but often overlooked distinction drawn by Gerald Gaus, there are two ways to understand this objection. First, public reason is often said to be inconclusive because it fails to generate agreement on fundamental political issues. Second, and more radically, some critics have claimed that public reason is indeterminate because it cannot provide any citizen with sufficient reason(s) for making important political decisions. Against the first of these objections, I argue that the purpose of public reason is not to end reasonable disagreement. Rather, it is to provide a suitable framework of values and principles within which citizens may resolve their moral and political differences. Against the second objection, I argue, first, that the indeterminacy of public reason is much less common than its inconclusiveness; and, second, that there are second-order decision-making strategies that may enable citizens to cope with cases of indeterminacy. The incompleteness of public reason, whether it takes the form of inconclusiveness or indeterminacy, is not a reason for citizens to abandon their commitment to public justification.

keywords

public reason, public justification, political liberalism, Rawls, Gaus

introduction Political liberals believe that in a society well-ordered according to an ideal of public reason, citizens justify their political decisions by appealing to considerations that they sincerely believe others can reasonably accept. They recognize a

DOI: 10.1177/1470594X04042963 Micah Schwartzman recently received his doctorate from The University of Oxford and is currently studying law at the University of Virginia, 580 Massie Road, Charlottesville, Virginia 22903-1789. [email: [email protected]]

514

John Rawls politics, philosophy & economics 3(2)

'duty of civility' that requires them to refrain from appealing to arguments that can only be supported by nonpublic reasons drawn from their comprehensive ethical, philosophical and religious views. Those who act on sectarian grounds without providing adequate public reasons are liable to criticism for failing to respect the interest that others have in being governed according to principles and policies that they can reasonably accept.1 Against this view, critics have argued that public reason cannot resolve fundamental moral and political issues because it excludes too many considerations from the political domain. The claim is that the content of public reason is not substantial enough to settle controversial matters such as abortion, stem-cell research, human cloning, animal rights, the right to die, the morality of homosexual relationships, and so on. To decide these sorts of issues, citizens have no choice but to appeal to nonpublic values drawn from their comprehensive doctrines. Therefore, nonpublic reason(s) must be allowed to play a greater role in political justification than would otherwise be permitted by the idea of public reason.2 The force of this objection is that it denies the completeness of public reason. The idea of public reason derives its content from liberal political conceptions of justice that are supposed to be autonomous or independent of any particular comprehensive doctrine. The values, principles and methods of reasoning organized by a political conception must be sufficiently robust to give answers to most, if not all, important political issues. As John Rawls says, The significance of completeness lies in the fact that unless a political conception is complete, it is not an adequate framework of thought in the light of which the discussion of fundamental political questions can be carried out.'3 Citizens must be able to deliberate about controversial matters in terms of public reasons alone. Otherwise, they will have to go beyond the limits of public reason to decide important political issues. This would be damaging to liberal political legitimacy because it would mean that, in many cases, political decisions would be made on grounds that some reasonable citizens would have no reason to accept. Although it has been elaborated at great length and by a number of different critics, I believe that this objection is ultimately unpersuasive. Public reason may be incomplete to some extent, but not to the degree that some critics have suggested. Furthermore, even if it could be shown that public reason fails to answer important political questions, this would not give citizens a reason to appeal to their comprehensive doctrines. The idea of public reason is supported by great political values that should not be abandoned the moment people run into difficulty resolving controversial issues. Responding adequately to the incompleteness of public reason does not require sacrificing the liberal commitment to public justification. This article proceeds in the following way: in the next section, I distinguish between two types of incompleteness. Public reason might be incomplete either because it fails to bring citizens into agreement about how to decide an issue, or 192

John Rawls

515

Schwartzman: The completeness of public reason

because it does not provide sufficient reason(s) to make a decision in the first place. I shall follow Gerald Gaus in referring to the first type of incompleteness as inconclusiveness and to the second as indeterminacy. In section two, I argue that the inconclusiveness of public reason is a permanent feature of liberal politics. But this does not mean that public reason is indeterminate. On the contrary, our working assumption should be that most issues can be decided within the limits of public reason. Section three canvasses five decision-making strategies that can be used to resolve cases of incompleteness if and when they occur. My conclusion will be that citizens should rarely, if ever, appeal solely to their comprehensive doctrines for the purpose of political justification. Before continuing, let me say a preliminary word about the scope of the argument that follows. It is not, nor is it intended to be, a complete defence of the idea of public reason. This article addresses a particular objection to public reason as part of an effort to develop a stronger and more sophisticated conception of public justification. There are, of course, many other objections to public reason not considered here. For example, some have argued that public reason excludes people arbitrarily by requiring them to affirm controversial moral and epistemological doctrines;4 that it rejects comprehensive ethical and religious views as sectarian while ignoring deep disagreements about the reasonableness of particular conceptions of justice and legitimacy;5 that it needlessly and unfairly excludes the convictions of religious believers;6 that it relies on implausible psychological assumptions about the ability of citizens to reason independently of their comprehensive views;7 that it stifles social innovation by suppressing novel political arguments;8 and that it succumbs to conservatism by drawing its values and principles from the established political culture.9 A full defence of public reason would have to supply clear and convincing responses to each of these objections. Although I believe that such a defence is possible, the purpose of this article is narrower. If public reason is incomplete, then it fails on its own terms. My aim is to show that it does not.

1. Two types of incompleteness Public reason is said to be incomplete if it does not provide citizens with the means by which to answer fundamental political questions. There are two ways in which this can happen. Public reason may be incomplete if citizens have competing but inconclusive public justifications for a given political decision. Alternatively, public reason may be incomplete because it is indeterminate. This occurs when citizens do not have sufficient public reason to decide between mutually exclusive political outcomes. The purpose of this section is to clarify the distinction between inconclusiveness and indeterminacy. I begin by describing each of them in turn.

193

516

JohnRawls politics, philosophy & economics 3(2)

1.1. Inconclusiveness Public reason may be incomplete because it fails to generate convergence among reasonable people on a single political outcome. This happens when citizens justify their political decisions on the basis of public reasons but disagree with one another about which of their positions is most reasonable.10 Suppose that A supports policy PI, and B supports P2, and that both A and B justify their positions on the basis of what they sincerely believe are reasonable balances of political values (Ri and R2, respectively). Thus, A offers the justification RI -^ P^ and B offers R2 —> P2. Now if PI and P2 are mutually exclusive, and if neither A nor B can prove to the other that his or her position is the most reasonable, then they have reached an impasse within the limits of public reason. Each of them believes that public reason provides a determinate conclusion, but they are unable to reach agreement on a specific political outcome. Perhaps a more concrete example will help to illustrate this point. In a muchdiscussed footnote in Political Liberalism, Rawls describes how a reasonable balance of political values might support a qualified right to abortion. He appeals to 'three important political values: the due respect for human life, the ordered reproduction of political society over time . .. and finally the equality of women as equal citizens' and suggests that 'any reasonable balance of these three values will give a woman a duly qualified right to decide whether or not to end her pregnancy during the first trimester. The reason for this is that at this early stage of pregnancy the political value of the equality of women is overriding, and this right is required to give it substance and force.'11 Rawls has been criticized for failing to explain why he thinks that the equal treatment of women overrides the other political values that he mentions. Some have argued that those who do not share Rawls's intuitions about this issue could build their case against a right to abortion on the same political values that he invokes by claiming that respect for human life should take priority over the value of treating women equally.12 Assuming this claim is as cogent as its alternative,13 it would then be possible to argue the issue of abortion in either direction by appealing to a reasonable balance of political values. The failure of convergence in this example results from the inconclusiveness of conflicting public justifications. Following Gerald Gaus, we can say that a public justification is inconclusive if it is undefeated but not victorious.14 Briefly, a justification is undefeated if it can be defended against arguments that would either undermine or rebut its conclusions. Let me illustrate this point by continuing the above example: suppose that A proposes the justification R\ -> PI, where RI, a balance of political values in which the equal treatment of women overrides the value of respect for human life, justifies PI, a policy of permitting a right to abortion. There are two ways that one might attempt to defeat this justification. First, one might object that A is mistaken about what follows from the reasonable balance of political values that she believes justifies her position. The value of treating women as equal citizens might be overriding in this case, 194

JohnRawls

517

Schwartzman: The completeness of public reason

but it must also be shown that this value requires granting women a right to abortion. If such a right does not follow from the value of treating women equally, then A's inference from R\ to PI has been undermined. Alternatively, one might agree that R\ supports PI but argue that there is nevertheless a weighty, countervailing reason (or balance of reasons) to reject PI. In the absence of other considerations, A may be justified in believing that the value of treating women equally supports a qualified right to abortion. However, if the value of respect for human life outweighs the value of the equality of women, then R\ —> P\ would be defeated not because it has been shown to rest on an inferential mistake but because there is a countervailing reason that rebuts R\. Thus, R\ —> PI can be defeated by criticisms that challenge either the validity of its inferential structure or the soundness of the reasons that support its conclusions, or both. If A's justification withstands such criticisms, it is undefeated but not necessarily victorious. This is possible because defeat and victory do not provide a complete range of descriptions for the epistemic status of a proffered public justification. Gaus argues that an undefeated justification must meet two further conditions for it to be victorious. First, others must have the opportunity to examine, clarify and object to the proposed justification. The argument for this publicity condition is that public scrutiny improves the epistemic quality of political justifications. Arguments for political principles are generally quite complicated. People may have very different ways of thinking about such principles, and a diversity of perspectives may help to reveal considerations that might not have been apparent to individuals reasoning alone.15 Second, to be victorious, a public justification must meet a higher burden of proof than one which is merely undefeated. Under some circumstances, A may be entitled to accept PI on the basis of reasons that are sufficient but not decisive. However, if A demands that others comply with PI, she must demonstrate that it would be ««reasonable for anyone to disagree with the justification that she has offered for her position. An argument to that effect must meet a fairly high threshold of epistemic credibility. Gaus has suggested that we use the standard of being beyond a reasonable doubt.16 If we accept his proposal, as I think we should, an undefeated public justification would have to meet this burden of proof under the conditions of actual publicity before it could be said to be victorious. The impasse between A and B in the example above can be described as a failure of public reason to provide a victorious public justification for either P\ or PI. Given that both A and B present their (presumably undefeated) arguments in terms of a reasonable balance of political values, neither can claim to have demonstrated that the alternative is unreasonable. The most that A (or B) can say is that she sincerely believes that hers is the most (but not the only) reasonable position.

195

518

JohnRawls politics, philosophy & economics 3(2)

1.2. indeterminacy Public reason may also be incomplete if it fails to provide a sufficient reason for selecting between two (or more) responses to some issue. This is what I shall refer to as indeterminacy. There are a number of reasons why public reason might be indeterminate. First, there may be vagueness in the concepts, values and principles that are supposed to determine the resolution of a particular issue. Citizens are sometimes required to make decisions about matters that have multiple dimensions of evaluation, none of which may be especially well defined. A lack of precision in the criteria that are supposed to determine a case may make it difficult if not impossible to reach a reasoned conclusion.17 Second, public reason would also be indeterminate if it finds that two options are of equal value. In this case, citizens would not have sufficient public grounds on which to choose between them. A third and more controversial explanation of indeterminacy rests on the claim that two options are incomparable. This is the case when, given some criteria of evaluation, neither option is better than or equal in value to the other. If PI and P2 are incomparable, nothing positive can be said about the relation between them.18 Again, there would be no way to decide which policy is preferable on the basis of public reason alone. To see the difference between indeterminacy and inconclusiveness, consider again A's claim that R\ —» P\ is an inconclusive justification for a right to abortion. Her argument does not provide decisive grounds for supporting a right to abortion, but it does have the status of a determinate, undefeated public justification. Now it might be objected that PI cannot be established based on public reason alone. When asked to provide an argument for the way in which she has ordered the political values of respect for human life and the equality of women (/?i), A may have no further public reason(s) to support her position. If her argument cannot be defended by anything other than an appeal to her comprehensive view, then it would seem that public reason has not in fact provided her with a basis on which to determine the issue. Furthermore, if the same holds true for £, who opposes a right to abortion, then neither A nor B has a public reason for the manner in which they have ordered the relevant political values.19 What appeared only a moment ago to be a failure of public convergence may actually turn out to be a case of indeterminacy. Public reason would be indeterminate between their positions because it does not provide sufficient reason(s) for choosing between them. Suppose for the moment that neither A nor B can provide any further justification to break the 'stand-off between them. Philip Quinn has suggested that their situation can be described in one of two ways: One could say that public reason in this case yields no reasonable balance of the values for human life and equality of women because its resources are too weak to determine any reasonable balance. Or one could say that public reason permits two reasonable balances of those values because its resources are too weak to single out just one 196

John Rawls

519

Schwartzman: The completeness of public reason

reasonable balance. In either case, public reason does not determine a unique reasonable balance of the values in question and so is incomplete to that extent.20

The conclusion that Quinn draws is correct, but his account of incompleteness is inadequate because it does not discriminate between two different types of indeterminacy. In the first of his descriptions, public reason is indeterminate because it fails to select any reasonable balance of political values. But there is a crucial difference between having no reasonable options and having insufficient reason to decide between two options that are equally reasonable. While it would be illegitimate for citizens to impose either policy under the first description, they might legitimately choose one or the other as presented under the second. The options under the second description are an instance of what Gaus has usefully described as nested indeterminacy. Public reason may provide a reason to select a policy from a set of exclusive disjunctions {P\ u P^ u PZ, . . . Pn} without specifying a uniquely or maximally reasonable choice from within that set.21 For example, A and B may agree that their dispute over the right to abortion should be settled by appealing only to those justifications that are based on some balance of the political values of respect for human life and the equality of women. We can represent this set of permissible justifications as [R\ -> PI u R2 -> P2], where the two justifications support mutually exclusive policies by giving different priorities to the two controlling values. If this restriction on the domain of justification is itself conclusively justified, then an argument appealing to an additional value R^ (e.g. the sacredness of ensoulment) would be excluded as unreasonable. However, from within the set of permissible justifications, there may be no public reason for preferring one option over the other. Quinn is right that, to that extent, public reason would be incomplete. The problem with his account is that it does not distinguish between cases of indeterminacy in which any choice is unreasonable and cases of nested indeterminacy in which there are a number of reasonable options but no reason to prefer any one of them. Again, this distinction is important because it would be legitimate to select any of the nested options, whereas, in a pure case of indeterminacy, the options fail to meet the threshold of reasonableness.22 Importantly, Gaus has shown that this analysis of nested decision-making also applies to inconclusive justifications. For example, liberal institutions protect a set of basic individual rights that includes the freedoms of speech, press, association, religious worship, and so on.23 These freedoms are thought to be victoriously justified, so that anyone who denies them, at least at such a high level of abstraction, is unreasonable.24 Of course, this does not preclude a wide range of reasonable disagreement about how these freedoms should be interpreted. In the United States, many, if not most, citizens believe that religious freedom is incompatible with state-sponsored religion. The constitutional arguments that support freedom of religion are believed to have decisively defeated establishment, although they remain inconclusive with regard to a host of other questions. 197

John Rawls

520

politics, philosophy & economics 3(2)

A commitment to nonestablishment is compatible with reasonable disagreement over interpretations of the First Amendment free exercise and establishment clauses.25 One can disavow the idea of establishing a national church and still argue in favour of federal funding for faith-based charities, prayer in public schools, and voucher programs for religious education. The general point here is that public reason may nest inconclusive justifications, e.g. arguments about whether the establishment clause permits the funding of parochial education, by allowing competing interpretations of a victoriously justified principle, such as the principle of religious toleration that informs the First Amendment.26 Finally, it is worth noting that, depending on the form it takes, incompleteness can have both interpersonal and mirapersonal dimensions. We have seen that when A and B support competing policies based on inconclusive justifications, public reason fails to bring about agreement between them. By contrast, cases of indeterminacy are distinctly mirapersonal. If a person is justified in believing that public reason does not provide sufficient grounds to select between two options, then, as David Reidy says, '[T]he "stand-off within public reason here is not among citizens each of whom is able to reach and vote on his or her determinate and reasoned judgement, but rather within citizens unable to do so (within public reason).'27 To be sure, failures of public convergence also depend, in part, on assessments that take place within a person's practical reasoning. For A to recognize the reasonableness of justifications that compete with her own, she must assess the status of her own arguments in light of those presented by others. She must decide whether her arguments have been adequately tested in public and whether they are sufficiently robust to warrant the use of political power. For practical purposes, however, there is a difference in the location of conflict in cases of inconclusiveness as opposed to indeterminacy. The former occurs between citizens; the latter within them.28 My purpose up to this point has been to describe the ways in which public reason might be incomplete. The next section considers whether, and to what extent, public reason is inconclusive and/or indeterminate. I argue that reasonable disagreement based on inconclusiveness is something to be expected within the normal politics of a liberal democratic society. The same, however, does not seem to be true of indeterminacy. I shall argue that public reason is, in principle, capable of delivering determinate although sometimes inconclusive responses to fundamental political issues.

2. The case for incompleteness Political liberalism tolerates more reasonable disagreement than many of its critics have acknowledged. Public reason may be inconclusive about a wide range of fundamental political issues including matters of basic justice. Yet claims about indeterminacy are much more controversial. It is one thing to show that public reason leaves people in a position to disagree about important issues, 198

John Raw I s

521

Schwartzman: The completeness of public reason

but it seems much more objectionable to think that people will have no, or insufficient, reason by which to decide fundamental issues. I shall begin with arguments for inconclusiveness and then ask how much room is left for more radical claims about indeterminacy.

2.1. Arguments for inconclusiveness

The possibility of inconclusiveness depends on whether the alternatives of victory and defeat exhaust the possible outcomes of public justification. If we follow Gaus in believing that they do not, then there is good reason for thinking that public reason will often be inconclusive. The argument for the possibility of inconclusiveness is similar to the argument based on the 'burdens of judgement' that Rawls gives for the possibility of reasonable disagreement. Recall that reasonable disagreement, which takes the form of conflicting inconclusive justifications, cannot be the result of ignorance, irrationality or self-interest. Although various forms of irrationality and impermissible partiality are often implicated in actual disagreements, they carry no weight in the process of justification. The reason for this is that justification is a matter of giving people good reasons. Therefore, if inconclusive justifications are to be possible not only in theory but also in practice, there must be an explanation for disagreement that does not rely on alleged failures of rationality or impartiality. Fortunately, there is such an explanation. Long-standing moral and political controversies are often best explained by the complexity of the evidence, the need to balance and sometimes trade off values that are vague or otherwise difficult to interpret, and the differences in life experience which inform the judgements that people make about such issues.29 Furthermore, because individuals have limited time to process relevant information, it is far too demanding to require that they act only on the best of all possible reasons. Even if it were possible, the psychological costs and difficulties30 of reasoning in this way would be overwhelming. As Gaus explains, 'Belief systems are vast and complex; our standard epistemological situation is an overabundance, not a paucity of reasons. . . . Given our inability to scan our belief systems for all possible inconsistencies and all justified inferences, it is entirely unsurprising that people arrive at opposing conclusions.'31 The epistemic limitations intrinsic to practical reasoning explain why it is acceptable for people to form judgements on grounds that are sufficiently credible but that do not exhaust all of the reasons that might, in principle, be available to them.32 Having clarified the theoretical and practical bases of inconclusive justification, it hardly seems necessary to argue for their application to fundamental political issues. Yet the idea of public reason is often criticized for failing to recognize that reasonable disagreement applies not only to comprehensive doctrines but also to matters of basic justice. In particular, Rawls is frequently accused of attempting to insulate his preferred theory of political right, namely, justice as fairness, from the burdens of judgement. For example, Jeremy Waldron argues that Rawls is committed to the view that 'a well-ordered society may 199

John Rawls

522

politics, philosophy & economics 3(2)

exhibit disagreements about the good but it will not exhibit disagreements about the fundamentals of justice'.33 But the attribution of this view to Rawls, and to political liberalism more generally, is mistaken. This is perhaps understandable, as the idea of a well-ordered society can be misleading. Recall that a well-ordered society is one in which '(1) everyone accepts and knows that others accept the same principles of justice, and (2) the basic social institutions generally satisfy and are generally known to satisfy these principles'.34 On its face, this definition does not seem to allow any room for disagreement about principles of justice. But to assume that the idea of public reason excludes disagreement because the definition of a well-ordered society requires agreement about justice is to misunderstand the purpose of having an idea of a well-ordered society. As Rawls explains, the idea functions as a 'very considerable idealization' which allows us to test whether a conception of justice 'can serve as the publicly recognized and mutually acknowledged conception of justice when society is viewed as a system of social co-operation between free and equal citizens from one generation to the next'.35 A conception of justice that fails to meet this condition is unacceptable from the point of view of public reason. Nevertheless, there may be a plurality of conceptions of justice that can serve as the basis of a well-ordered society. Indeed, Rawls is explicit about the fact that people may reasonably disagree about (1) which principles of justice are the most reasonable, and (2) how to determine whether the most reasonable principles have been satisfied. As evidence for the first of these points, consider what Rawls says about the content of public reason: It is crucial that public reason is not specified by any one political conception of justice, certainly not by justice as fairness alone. Rather, its contents — the principles, ideals and standards that may be appealed to — are those of a family of reasonable political conceptions of justice and this family changes over time. . . . The content of public reason is not fixed, any more than it is defined by any one reasonable political conception.36

A reasonable political conception of justice must treat people as free and equal citizens. But since these ideas can be interpreted reasonably in different ways, Rawls explains that 'we get different formulations of the principles of justice and different contents of public reason. Political conceptions differ also in how they order, or balance, political principles and values even when they specify the same ones. . . . Political liberalism, then, does not try to fix public reason once and for all in the form of one favored political conception of justice.'37 This is not to say that public reason imposes no limits on disagreement. Liberal political legitimacy requires that conceptions of justice protect the basic rights, liberties and opportunities of individuals. Principles of justice that cannot be interpreted in ways consistent with liberal freedoms are inadmissible because they cannot serve as the basis of a well-ordered society. They do not appear in the nested set of public political conceptions about which citizens may reasonably disagree. 200

John Rawls

523

Schwartzman: The completeness of public reason

The second point is that political liberalism acknowledges that even when people agree about which principles of justice are most reasonable, they may disagree about when those principles have been fully satisfied. This is the reason that Rawls gives for not requiring the constitutionalization of the difference principle within justice as fairness. He writes that: whether the aims of the principles covering social and economic inequalities are realized is far more difficult to ascertain [than whether constitutional essentials have been realized]. These matters are nearly always open to wide differences of reasonable opinion; they rest on complicated inferences and intuitive judgments that require us to assess complex social and economic information about topics poorly understood.38

It would be difficult to find a better example of the application of the burdens of judgement. Agreement on a conception of justice such as justice as fairness does not preclude disagreement about what is required by that conception and how institutions might be structured to meet its demands.39 These two examples of disagreement — the first about which principles of justice are most reasonable, the second about what those principles require — are both cases of nested inconclusiveness. To put things in Gaus's terms, political liberalism takes as victoriously justified a principle of legitimacy that requires citizens to affirm some liberal political conception of justice. Since there may be more than one such conception, people may reasonably disagree about which is best suited to govern their society. Although political liberals have sometimes been reticent to say it, reasonable disagreement is a constitutive part of ordinary politics. There is no reason to think that the idea of public reason will exclude controversy about fundamental moral and political issues. Its purpose is not to suppress such conflict but to permit it to take place within the bounds of a legitimate democratic process. Within the constraints imposed by constitutional principles that are conclusively justified, disagreements will abound. Gaus nicely summarizes the dual structure of liberal politics when he writes that: Constitutional politics concerns what is conclusively justified, and so determines the agenda of normal politics. Widespread consensus is to be expected. Normal politics is about fundamental matters, but is not a search for consensus; it is the confrontation of undefeated, unvictorious judgments about the demands of basic principles.40

Within the constraints of a legitimate constitution, we should expect reasonable disagreement to be a permanent feature of politics in a liberal democratic society. At times Rawls seems to endorse a similar theory of constitutional politics. Yet it is not entirely clear whether he would agree with Gaus's picture. Rawls follows Ackerman in distinguishing between 'the higher law of the people' and the 'ordinary law of legislative bodies'.41 The higher law of the constitution must accord with the conception of political legitimacy expressed by the idea of public reason. Rawls tells us that the principle of legitimacy 'applies on two levels: one 201

524

John Rawls politics, philosophy & economics 3(2)

is to the constitutional structure itself, the other is to particular statutes and laws enacted in accordance with that structure. To be reasonable, political conceptions must justify only constitutions that satisfy this principle [of legitimacy]'.42 A conception is reasonable only if it justifies a constitution regulated by the principle of legitimacy. However, public reason does not (necessarily) apply to ordinary legislation, provided that such legislation is enacted within the constraints of higher-order constitutional principles. As Rawls writes in Political Liberalism: Citizens and legislators may properly vote their more comprehensive views when constitutional essentials and basic justice are not at stake; they need not justify by public reason why they vote as they do or make their grounds consistent and fit them into a coherent constitutional view over their whole range of decisions.43

Elsewhere, I have criticized the distinction between constitutional essentials and ordinary legislation for being unfaithful to the liberal aspiration of political reconciliation.44 When citizens make political decisions that cannot be justified to others, they transgress an ethic of civility in which politics is conducted according to reasons that everyone should, in principle, be able to accept. But perhaps this much might be said in Rawls's defence: The political values that give content to public reason are not sufficient to provide determinate conclusions about ordinary legislation concerning issues such as tax law, environmental regulation, public health standards, education reform, defence policy, budget resolutions, and so on. Political values will certainly have some bearing on these issues, but it is unrealistic to expect that they will pick out unique solutions in every case. However, as long as citizens endorse a constitutional structure that accords with the liberal principle of legitimacy and interpret that structure based on public reason, they may vote and act politically for nonpublic reasons without jeopardising the legitimacy of the system.45

It is difficult to say whether Rawls would accept this view. At times he seems to hint that public reason may apply across the board but that he has simply not yet dealt with the question of non-essentials.46 If that is the case, then perhaps there is no real disagreement here. But if Rawls does maintain that nonpublic reasons are permissible at the level of ordinary politics, then he seems to have mis-described his own theory as a form of constitutional dualism. It would be more accurate to think of his version of political liberalism as tripartite in structure. First, at the highest level, citizens are expected to give priority to the constraints imposed by the liberal principle of legitimacy. They must affirm some political conception of justice that protects the rights of citizens as free and equals. This requirement can be thought of as establishing the boundaries for a second level of constitutional politics. Within these boundaries, there may be significant disagreement about which political conception is most reasonable. Different political conceptions may yield competing interpretations of constitutional principles. As we have already seen, Rawls believes that 'this is the normal case: unanimity is not to be expected. Reasonable political conceptions of justice do not always lead to the 202

John Rawls

525

Schwartzman: The completeness of public reason

same conclusion, nor do citizens holding the same conception always agree on particular issues.'47 Finally, beneath the level of constitutional politics, citizens are not beholden to one another to provide justifications based on reasonable balances of political values. They can decide everyday political issues by appealing directly to nonpublic values drawn from their comprehensive views. This account of ordinary politics — whether we think of it as part of a dualist or tripartite view — has been interpreted by some as a serious concession to perfectionist criticisms of political liberalism.48 Rawls seems to have admitted that it is legitimate for citizens to legislate their conceptions of the good, at least when constitutional essentials are not at stake. But this view seems to me mistaken. Rawls does not restrict the application of the idea of public reason to constitutional matters for perfectionist reasons. He does so to avoid worries about the extent of its (in)completeness. In the next section, I try to allay some of these concerns by arguing that, while public reason is often inconclusive, the case for its indeterminacy is significantly weaker. 2.2. Arguments for (and against) indeterminacy Suppose we accept that public reason is inconclusive about a wide range of political issues. Must we also accept that it is radically indeterminate? Recall that public reason is indeterminate if it does not provide citizens with sufficient reason to select between multiple options to resolve a political issue. This section considers whether public reason is incomplete in this way. Those who argue that public reason is indeterminate usually agree that political values have at least some bearing on controversies about fundamental political issues. A commitment to public justification may not be compatible with certain political positions, especially on matters that are the subject of intense religious dispute such as abortion and homosexual marriage.49 For example, it seems impossible to justify laws that prohibit abortion in cases of rape or incest, or where the mother's life is at risk, on the basis of any reasonable balance of political values.50 The idea of public reason appears to rule out the sorts of arguments that are used to defend such restrictions. Consider the case of religious believers who justify absolute prohibitions of abortion because they are committed to the idea of ensoulment at conception. Kent Greenawalt has written that, The very notion that at a specific time God gives each human a soul may influence someone to look for a critical either-or point, a point at which a shift takes place from virtually no moral status to moral status equivalent to that of a full human being'.51 For those who believe that this point is fixed at conception, it may be difficult to see why the physical development of the fetus should have any bearing on its moral status. Furthermore, as Greenawalt continues, 'the ensoulment idea is unfavorable to the claim that the pregnant woman's interests override those of the fetus. Not only does it make it harder to say that the fetus with some worth has less inherent worth than that of a human being after birth, it tends to undercut the attractiveness of the argument that the woman may extin203

John Rawls

526

politics, philosophy & economics 3(2)

guish the fetus in "defense" of her own body.'52 For this reason, the idea of ensoulment makes it especially difficult to justify even a severely qualified right to abortion. Perhaps it is possible that less severe restrictions can be adequately (inconclusively) justified according to publicly accessible reasons.53 But it should be clear that a commitment to public justification excludes the kind of religiously based reasoning that might lead to the absolute prohibition of abortion. The point of this example is simply to show that public reason is not fully indeterminate on even the most controversial of questions. It may rule out certain options, while remaining incomplete with regard to others. Those who criticize public reason for its incompleteness need not challenge this conclusion. The important point for them is that public reason is incapable of giving determinate answers within the range of options that it allows. The question, then, is whether there are any compelling arguments to support this claim. What reasons are there for thinking that public reason is indeterminate? The first argument to consider is based on an appeal to the burdens of judgement. The third limitation of practical reasoning that Rawls lists in his description of the burdens is that: To some extent all our concepts, and not only our moral and political concepts, are vague and subject to hard cases; and this indeterminacy means that we must rely on judgment and interpretation (and on judgments about interpretations) within some range (not sharply specifiable) where reasonable persons may differ.54

This observation about the vagueness and imprecision of abstract concepts seems particularly apt with regard to the values that Rawls identifies as distinctively political: 'a more perfect union, justice, domestic tranquillity, the common defense, the general welfare, and the blessings of liberty for ourselves and our posterity',55 as well as values that are subsumed by these such as 'equal political and civil liberty; equality of opportunity; the values of social equality and economic reciprocity' and others.56 When citizens are confronted with hard cases that require them to make fine-grained decisions about how to apply such values, it may be difficult for them to reach conclusions about what to do. Second, the problem is compounded by the fact that political values are not only vague, but they often seem to conflict with one another. David Reidy writes that: on a great many political issues, some fundamental, citizens will have no sense of how competing or conflicting values are to be ordered within their political conception of justice. More importantly, they will find within their political conception of justice no criterion or criteria for ordering competing or conflicting values or ends in new and difficult cases.57

Thus, even if people can sharpen the content of political values to the point where they feel confident in applying them, they must still confront the problem of how to weigh such values when they conflict. 204

JohnRawls

527

Schwartzman: The completeness of public reason

A third argument for indeterminacy is that public reason is simply silent about some questions. Some of the most troublesome and controversial political issues turn on whether an entity is entitled to a certain level of moral consideration. Disputes about issues such as animal rights, abortion, human cloning and stemcell research are what Greenawalt refers to as 'borderline status cases'. These are issues in which public reason seems to avoid the central question of who, or what, counts as a moral subject. Again, in the abortion case, comprehensive doctrines may have very determinate things to say about whether the fetus is a person, and, if so, what level of moral and legal protection is owed to it. By contrast, public reason seems to 'run out' when it comes to answering this question. As Greenawalt says, 'It is not that the arguments . . . are fairly evenly balanced, rather they seem not really to get to the heart of the problem. Some critical and very important judgment is needed, and the arguments seem of only slight help.'58 A fourth and related argument is that any determinate resolution of a borderline status case is in fact the work of intuitive judgements that rely on comprehensive doctrines. Indeed, Greenawalt's discussion of ensoulment is, in part, an effort to explain how religious convictions inform some of the public justifications for restrictions on abortion. As he says, 'even if the point of ensoulment is decided on the basis of purely naturalistic reasoning the whole perspective that surrounds that notion will subtly influence how the moral status of the fetus and the permissibility of abortion are regarded'.59 If A believes that God gives human beings a soul at conception, she may be inclined to accept naturalist arguments about the arbitrariness of assigning moral status at any point after conception. Moreover, the idea of ensoulment may lead her to an either/or view of moral personality, so that the fetus is either a full moral person or not one at all. The objection, then, is that when A weighs the value of respect for human life against other political values, this way of thinking about the status of the fetus will be responsible for her determination of which value is assigned priority. As Reidy asks, 'Can there be any doubt that she will be drawing entirely, if indirectly, on non-public reason(s) to order the relevant politics values?'60 These four objections — (1) to the vagueness and abstractness of political values, (2) to the lack of principles for ordering them when they conflict, (3) to the inability of public reason to reach questions about moral status, and (4) to the implicit reliance on comprehensive doctrines to settle cases of indeterminacy — contribute to a deeply sceptical view of the possibility of reasoning about controversial issues within the bounds of public reason alone. Even if the use of public reason excludes some positions as clearly unreasonable, it will not provide citizens with enough guidance to decide among the remaining possibilities. The scope of nested indeterminacy in borderline status cases is too broad for public reason to do any real work. This is a formidable challenge to the completeness of public reason. But I do not think these objections are decisive. The main problem is that they beg the question against those who argue that public reason is inconclusive about funda205

John Rawls

525

politics, philosophy & economics 3(2)

mental matters. For starters, the burdens of judgement are intended not as an explanation of indeterminacy but as an account of why people reasonably disagree. The vagueness of abstract concepts contributes to divergences of opinion because imprecise concepts can be given conflicting interpretations. Furthermore, what Rawls refers to as the 'indeterminacy' of moral and political concepts is compatible with determinate but inconclusive judgements about the issues to which they apply. A may believe that the value of respect for human life is vague, amorphous, or too thin to provide a complete basis for political action. But this does not mean that she cannot make use of an interpretation of that concept as part of a more complex and determinate justification. It may be helpful here to repeat what Rawls says in describing the burden of vague concepts, which is that 'this indeterminacy means that we must rely on judgment and interpretation (and on judgments about interpretations) within some range (not sharply specifiable) where reasonable persons may differ'. A concept may itself be indeterminate but nevertheless play a partial and supportive role within an argument that can be accepted as one among many reasonable or inconclusive justifications. A similar criticism applies to the objection that public reason provides no criteria for ordering conflicting values. Political conceptions of justice often contain priority rules that rank principles for the purpose of evaluating possible social and political outcomes. To establish that such principles are indeterminate would require showing that those who believe in their determinacy are mistaken. Rawls is surely right to say that this 'cannot be decided in the abstract independent of actual cases. We need such cases carefully spelled out to clarify how we should view them. For how to think about a kind of case depends not on general considerations alone but on our formulating relevant political values we may not have imagined before we reflect about particular instances.'61 The ambition of a project that took this task seriously would be phenomenal. On any given issue, one would have to refute, conclusively, every claim for which public reason provides a determinate justification. Anything less would allow for reasonable disagreement about whether public reason is indeterminate or merely inconclusive. This point is worth emphasizing. If A claims that public reason is indeterminate between PI and P^, she is taking a particular stand in the debate over which is preferable. Against those who think that public reason favours one or the other, she is saying that it cannot decide between them. But this requires as much argumentative support as any of the other positions in the debate.62 The question of who has the burden of proof in establishing claims of indeterminacy is a difficult one. But three further points should tip the scale towards a presumption in favour of determinacy. The first is that the political landscape of a modern democratic society is never static. It is constantly raising issues that call for innovation and creativity in moral reasoning. It would be premature to conclude that public reason cannot give determinate answers to new and difficult questions. David Richards has pointed out that critics of public reason often appeal to novel cases to advance their argument.63 Recent developments in 206

John Rawls

529

Schwartzman: The completeness of public reason

genetics and cloning are the most obvious examples. It may be true that public reason is currently perplexed by such matters. But that is to be expected. It is a commonplace that moral and political thought lags behind the development of new technologies and the ethical questions raised by them. We demand too much of public reason if we expect it to have answers waiting in the wings. Indeed, many comprehensive doctrines, including some of the most venerable religious traditions, are often similarly bewildered by the rapidity of scientific progress and the many moral and political questions that follow from it. Turning away from public reason is no guarantee of finding determinate answers. Second, one of the arguments that supports the inconclusiveness of public reason is that our epistemic capacity to process relevant information is limited. In addition to carefully spelling out actual cases, those who claim that public reason is indeterminate must also consider the possibility that they have not, and perhaps could not, address all of the reasons that pertain to an important and complicated issue. Gaus is right to emphasize that 'our methodological assumption ought to be that the available reasons rarely run out, and so indeterminacy is rare' ,64 This is not to say that indeterminacy is impossible; rather, we should assume that public justification is (inconclusively) determinate until we can be shown otherwise. Third, the view that public reason is generally indeterminate is potentially selffulfilling. People who are inclined to invoke their comprehensive doctrines whenever public reason is indeterminate have an incentive to cut short the search for a publicly justifiable conclusion.65 Working on the assumption that public reasons rarely run out is a way of countering the tendency to find more indeterminacy than is actually out there. Finally, consider the objection that comprehensive views lead to determinate but distorted public justifications. The wide view of public reason acknowledges that citizens will often have independent sources of motivation for their political decisions. There is nothing illegitimate about this so long as they can supply adequate public reasons that they sincerely believe justify their views within the political domain. As Rawls says, The ordering [of political values] is not distorted by those [comprehensive] doctrines provided that public reason sees the ordering as reasonable.'66 If A believes that abortion is wrong for nonpublic reasons drawn from her comprehensive view, she must have sufficient public reason(s) to support any political action taken on the basis of that view. But it is no objection to the idea of public reason that A will be predisposed to give favourable consideration to some public arguments rather than others. As long as the order of values she defends can be seen as reasonable from the point of view of public reason, this is not in itself objectionable. Reidy has objected to this reply because he thinks that public reason has no way to determine when A's reliance on nonpolitical values is unreasonable. The argument for this is that public reason is insufficiently determinate to evaluate the ordering of political values that A claims publicly justify her position. There are 207

530

John Rawls politics, philosophy & economics 3(2)

no independent criteria that can adjudicate between priority rules when orderings of political values conflict.67 The problem with this argument is that it presupposes that the case for indeterminacy has already been established. If it could be shown that priority rules for ordering conflicting values are not justifiable according to public reason, then Reidy's argument would have greater force. But this is precisely the sort of claim that is undermined by the arguments presented above. To sum up, public reason can be incomplete in two ways: (1) reasonable people may disagree about fundamental political issues because their public justifications are inconclusive, and (2) public reason may be indeterminate if it fails to select a uniquely reasonable outcome between two (or more) political options. I have argued that political liberalism allows for a wider scope of reasonable disagreement than many have previously assumed. Reasonable disagreement occurs when citizens offer justifications that have not been publicly tested or that have failed to meet a threshold of epistemic credibility high enough to warrant the exercise of coercive power by the state. In cases of nested inconclusiveness, however, the use of state power may be justified on the basis of inconclusive arguments about how to interpret and apply conclusively justified constitutional principles. For example, citizens may agree to the broad outlines of a liberal conception of justice, including the provision of 'adequate all-purpose means' to guarantee that basic rights are not merely hollow formalities. But even if there is agreement at this level, people will no doubt continue to disagree about what constitutes 'adequate' means and how best to arrange political institutions so as to provide for them. Such reasonable disagreement is to be expected as a permanent feature of liberal politics. Should we expect the same to be true of indeterminacy? The appeal of claims about indeterminacy stems largely from the abstractness of political values, their tendency to come into conflict, and the apparent silence of public reason on central issues in borderline status cases. These considerations, however, are insufficient to show that public reason is indeterminate rather than inconclusive. Of course nothing I have said rules out the possibility of reasonable disagreement about this claim as well, but the burden of proof should rest on those who believe that public reason is unable to decide fundamental political issues. Having argued that public reason is to some extent incomplete, I now want to ask how citizens should respond to cases where their justifications are either inconclusive or indeterminate. I argue that citizens should not abrogate their duty of civility even when it is difficult or impossible for them to decide how to resolve political issues. The next section attempts to show how citizens can respond to incompleteness without sacrificing the weighty moral and political values that support their commitment to public justification.

208

John Rawls

531

Schwartzman: The completeness of public reason

3. Towards completeness The objection to public reason that I have been considering says that citizens may legitimately appeal to nonpublic reasons when public reason is incomplete. If citizens ought to abide by the duty of civility, it must be possible to resolve political issues on the basis of public reason alone. If it turns out that public reason is incapable of providing citizens with guidance about such matters, then they cannot be morally required to refrain from invoking whatever reasons they sincerely believe justify a determinate outcome. The problem with this argument is that, even if we assume that public reason is incomplete with regard to some political issues, it does not follow that citizens must appeal to nonpublic reasons to resolve them. There are at least five decision-making strategies for coping with incompleteness that forestall or make unnecessary the need to go beyond the limits of public reason. Each of these strategies deserves more attention than can be provided here, but even a brief overview should be sufficient to demonstrate the various ways in which decisions might be reached within public reason despite claims of incompleteness. (i). intrapersonal delegation Rather than appeal to nonpublic reasons to resolve an issue, citizens may opt for what Cass Sunstein and Edna Ullmann-Margalit have called 'intrapersonal delegation'.68 Delegation is usually thought of as an mterpersonal act that occurs when A relinquishes control over some decision and authorizes B to make it for her. /«impersonal delegation occurs when A decides to postpone making a decision, in effect delegating the decision to her future self. It makes sense to put off a first-order political decision if, in the future, citizens expect to 'have more information, suffer less or not at all from cognitive difficulties, bias, or motivational problems, or be in a better position to assume the relevant responsibility' .69 Delaying a decision may allow citizens to pursue additional inquiries, evaluate new evidence and engage in further political deliberation. Moreover, this strategy seems especially well suited to issues that pose novel questions or that are sensitive to rapidly developing areas of expertise, such as cloning, biomédical engineering, or perhaps the latest advances in information technology. Intrapersonal delegation is not an excuse for procrastination or dithering. If there is no reason to expect new and probative evidence to be forthcoming, then citizens should decide on the strength of the existing balance of public reasons. Yet, as we have already seen, under most circumstances the problem is not that relevant considerations 'run out', so to speak. It is rather that politics imposes severe constraints on the resources (e.g. time and money) that citizens can devote to any particular issue. Indeed, in light of the costs of inquiry and deliberation, it is notoriously difficult to determine what is a sufficient, let alone optimal, basis for rational decision-making.70 Nevertheless, in some cases, it may be reasonable 209

532

John Rawls politics, philosophy & economics 3(2)

to table a decision temporarily rather than force a conclusion by appealing to nonpublic reason. One might object that all this talk about intrapersonal delegation is naïve. Very few people have enough control over the political agenda to delay a decision about an important issue, and even those who do exert some influence over agenda-setting may be forced to expedite their decisions. For example, consider the relatively recent decision made by Britain's highest court in which doctors were ordered to separate two sisters who were born conjoined.71 Had the court delayed its decision to deliberate at length about the morality of killing one twin to save the other, especially against the wishes of their Roman Catholic parents, it would have made a defacto choice permitting the deaths of both twins. There is often no real alternative but to decide hard cases in a timely manner. The question, then, is how should officials or citizens proceed when public reason is incomplete and ex ante deliberation is no longer a reasonable option? (ii). Deference to others It is probably too much to expect that every citizen will always be able to decide a given political issue solely on the basis of public reason. But when some citizens are able to do so, others have at least a prima facie reason to recognize and accept the public justifications that have been offered to them. Against this view, Reidy has objected that, 'Some [citizens] may find themselves able to reach a determinate and reasoned resolution . . . within public reason alone. But many, perhaps a majority will not.'72 However, it is entirely unclear why numbers should matter at this stage of public justification. If A believes P\ only on the basis of nonpublic reason but acknowledges that B' s support for a competing position P2 can be justified by public reason, then A should accept P2 as the only reasonable position. It makes no difference whether A is one person or a million. Of course Reidy goes on to argue that the majority should not concede that public reason is in fact determinate. But his argument is only sustainable if there is a reasonable explanation for why some citizens succeed in justifying their views on the basis of public reason while others do not. Barring some further objection against those who claim that public reason is complete, citizens who are unable to justify a decision based on public reason should defer to the reasonable claims of others. (iii). Moral accommodation When citizens reasonably disagree about some issue, they should try to accommodate one another within the bounds of public reason. The subject of moral accommodation has received extensive discussion in the philosophical literature,73 but a couple of points here should suffice. First, citizens should economize their disagreements by narrowing as much as possible their major points of contention.74 They can do this by searching for areas of convergence, extending these where possible, and isolating their remaining differences. A successful economy 210

John Rawls

533

Schwartzman: The completeness of public reason

of moral disagreement is obviously only a partial response to moral conflict. It may help to lessen tensions, but serious disagreements will often remain. Thus, and secondly, the strategy of accommodation must explain how citizens should proceed even after extensive moral deliberation has failed to produce agreement. The most obvious move is for citizens to negotiate a principled compromise. If A and B disagree about whether the state should adopt PI or P2, perhaps they can agree to P^ as an acceptable alternative, even if neither views it as an optimal outcome.75 It is worth emphasizing that compromises of this kind are not merely prudential concessions. To be able to live with others on terms of mutual respect is a significant moral achievement, one that sometimes justifies settling on second-best solutions with which everyone can reasonably agree.

(iv). Democratic procedures When the time for deliberation has been exhausted, and if no compromise has been reached, citizens in a liberal democratic society resolve their reasonable disagreements by turning to legitimate political procedures. A complete theory of public justification would give an account of which political institutions are best suited for this purpose.76 All I can do here is suggest, once again, that, rather than impose their nonpublic reasons on others, citizens can choose to submit their disputes to various forms of procedural adjudication. For example, a simple majority rule can be used to decide a contest between competing perspectives. As long as citizens vote for what they believe is most publicly justifiable, they can be satisfied that their collective decision is legitimate. Rawls puts this point well in A Theory of Justice when he writes that, 'if the law actually voted for is ... within the range of those that could reasonably be favored by rational legislators conscientiously trying to follow the principles of justice, then the decision of the majority is practically authoritative, though not definitive'.77 The decision-making procedure, whether it be by simple majority or some other standard, is used to determine a legitimate outcome. Its purpose is to break a deadlock between those who disagree about which outcome is most reasonable. For this reason, those who find themselves in the minority need not agree with the outcome of the vote, although they should recognize its legitimacy.78

(v). Arbitrary decision procedures

Democratic procedures can be employed to adjudicate reasonable disagreements, but they are useless in the face of indeterminacy. It would be pointless to call for a vote if citizens do not have sufficient public reason to decide the issue before them. Some additional consideration would be needed to tip the scales, which is why it is tempting to appeal to nonpublic reasons. Andrew Williams has suggested that a randomizing device, such as tossing a coin or holding a lottery, could be used to resolve such cases of indeterminacy.79 At this point, however, it will no doubt be objected that an ideal of public reason which requires flipping a 211

534

John Rawls politics, philosophy & economics 3(2)

coin to decide fundamental political issues, such as the permissibility of abortion or the ethics of cloning, is highly implausible, if not altogether absurd. There are at least three explanations for why we might be apprehensive about 'picking' a solution arbitrarily rather than 'choosing' one on the basis of public reasons.80 The first of these explanations turns on a general claim about human psychology, which is that people have a strong desire to act for good reasons. Rather than admit that reason does not determine an issue, people will often search indefinitely for new evidence, smuggle in extraneous considerations, or give more weight than is deserved to what they claim are decisive factors. In a slightly different but related context, Jon Elster has diagnosed this as 'an addiction to reason\ a special form of irrationality which he calls 'hyperrationality', or the refusal to admit that reason does not always issue determinate conclusions.81 On this view, if people are honest with themselves, they will recognize that indeterminacy is a pervasive phenomenon. Moreover, they will see that the most appropriate response is 'the conscious use of chance to make decisions when rational argument fails. Although the bleakness of this vision may disturb us, it is preferable to a life built on the comforting falsehood that we can always know what to do.'82 It may be tempting to contrast this view with the methodological assumption that public reason is determinate unless proven otherwise. But it is not entirely clear that the two are in tension. Elster believes that it is rational to adopt lotteries because he thinks that arguments can be produced which demonstrate indeterminacy in some cases, notably in child custody disputes.83 It would take me too far afield to discuss Elster's position in the detail that it deserves, but his argument is not necessarily inconsistent with the view defended here. If someone does manage to prove that public reason is indeterminate on some issue, citizens would then have reason to accept an arbitrary decision procedure to resolve the matter. Indeed, as Elster shows, some important political decisions seem to satisfy these conditions and are regularly made by lottery.84 Second, and pace Elster's analysis, our aversion to the thought of using lotteries (or other randomizing devices) to resolve most fundamental political issues (e.g. abortion, cloning, animal rights or homosexual marriage) may simply bring us back to the view that public reason is, in fact, rarely indeterminate. We baulk at using arbitrary decision procedures because they are unresponsive to reasons for action. And where we think that public reason gives determinate if inconclusive reasons, it would be inappropriate to leave such matters to chance. Better to employ any of the decision-making strategies described above than to adopt a strategy that ignores undefeated justifications. A third explanation for the view that political indeterminacies should not be resolved arbitrarily is an objection which says that it would be better to appeal to nonpublic reason(s) than to no reason at all. When public reason is indeterminate and nonpublic reason calls for a specific outcome, the two sets of reasons conflict. Citizens must then decide whether the values embodied by the ideal of 212

John Rawls

535

Schwartzman: The completeness of public reason

public reason are weighty enough to require that political outcomes be determined by chance. Furthermore, if there is no public reason to prefer a specific outcome, then why not let nonpublic reasons act as tie-breakers?85 The answer to this objection is that using nonpublic reasons to break deadlocks is a secondorder decision-making strategy that contravenes values central to the ideal of public reason. Here an example may help to clarify things. Consider a medical ethics board charged with the responsibility of allocating organs for transplantation. Suppose that, in a particular round of allocation, two candidates for a transplant rank equally according to all of the criteria that can be justified according to public reason. After weighing medical need, life expectancy (or duration of benefit), change in quality of life, time spent on waiting-lists, personal responsibility for contributing to the need for a transplant, etc., the board finds that it has no public reason to prefer either candidate. Would it then be acceptable for board members to appeal to nonpublic reasons? Would it be acceptable, for example, for board members who are religious to choose a believer over an atheist, or perhaps the righteous over the sinful? The answer, I think, is that appeals to nonpublic reasons should be rejected in favour of a fair, second-order decisionmaking procedure, such as a lottery. The reason is that, even in cases of firstorder indeterminacy, decisions should be made in ways that satisfy the political values of fairness and legitimacy. As Williams has argued, this way of responding to indeterminacy makes sense if we realize that the ideal of public reason embodies an exclusionary reason requiring citizens to refrain from relying on nonpublic reasons.86 An exclusionary reason is a higher-order reason for not acting on the basis of some set of valid first-order reasons.87 For example, in the paradigm case of authoritative directives, if A has conclusive reason to accept B' s (practical) authority, and B orders A to (£, then A has a reason to $ regardless of whatever other reasons A has to $ or not to (/>. Acceptance of B's authority entails a commitment not to act on reasons excluded by B9 s orders. Now suppose that A has strong first-order reasons not to $. This fact would certainly be relevant in evaluating her decision and whatever subsequent action she takes on the basis of it. As Raz explains, we are often willing to judge someone less harshly when they have acted on the basis of valid first-order reasons, even though those reasons have been excluded by some authority.88 For example, if a soldier violates orders because he has determined correctly that they would result in needless endangerment, he may receive less reproach than a soldier who violates them out of mere caprice. We can applaud the first soldier for acting wisely, while at the same time admonishing him for not obeying orders; whereas we will have nothing but opprobrium for the soldier who disobeys without reason.89 Similarly, if A is convinced that public reason is indeterminate with regard to abortion rights, she may have reservations about setting aside valid nonpublic reasons that support a determinate conclusion. That she has mixed feelings in this case is evidence of her belief in a second-order reason that does not override her 213

536

John Rawls politics, philosophy & economics 3(2)

first-order nonpublic reasons but that nevertheless requires her to set them aside. Her nonpublic reasons remain valid even though she has an exclusionary reason to refrain from relying on them as the basis of her political position. Furthermore, if she does base her decision on her comprehensive view, it will not be because public reason could not have resolved the matter. It will be because she has decided that it is more important to act on her personal convictions than to meet her duty of civility. From the point of view of public reason, however, this would make her liable to criticism for flouting her second-order commitment to providing others with justifications that they can reasonably accept. I have argued that citizens can employ a range of second-order decisionmaking strategies to cope with various forms of incompleteness. In each case, there is an alternative to appealing to nonpublic reasons. Therefore, the failure of public reason to provide determinate first-order conclusions is not, in itself, an argument for widening the scope of political justification to include nonpublic reasons.

Conclusion Reasonable disagreement is a much more common feature of liberal politics than indeterminacy. Yet neither poses insurmountable problems for an adequate theory of public reason. Many have underestimated the resources available to citizens who are committed to an ideal of public justification. This is due, in part, to the neglect of second-order reasons and strategies of decision-making as features of practical reasoning. Their complex structure has only been the subject of thorough investigation for a relatively short time. And where they have been analysed, it is mainly for application in legal reasoning, or else in economics and decision theory. But the central insights of nested justification, second-order decisions, and exclusionary reasons can be used to give a more sophisticated account of public justification within political liberalism. An ideal of public reason that has absorbed these innovations is in a stronger position to respond to the criticism that it cannot resolve fundamental political issues. It is a mistake to think that public reason will settle the most intractable of our moral controversies. But it is an even bigger mistake to think that this is a reason to abandon a politics in which decisions are made within the limits of public reason alone. notes

Earlier versions of this article were presented at the Nuffield Political Theory Workshop and the Oxford Political Philosophy Research Seminar. I would very much like to thank Adam Swift, George Klosko, Jonathan Quong and Andrew Williams for their comments and support. 1. For this ideal of public reason, see John Rawls, 'The Idea of Public Reason Revisited', in Collected Papers, edited by Samuel Freeman (Cambridge, MA: 214

John Rawls

537

Schwartzman: The completeness of public reason

2.

3. 4.

5.

6.

7. 8.

9.

Harvard University Press, 1999), pp. 573-615 [hereafter cited as 'Public Reason Revisited']; Political Liberalism (New York: Columbia University Press, 1996), especially 'Introduction to the Paperback Edition', and Lecture IV [hereafter cited as PL]', and Justice as Fairness: A Restatement (Cambridge, MA: Belknap Press, 2001), especially §9, p. 26. For examples of this line of argument, see Kent Greenawalt, Religious Convictions and Political Choice (Oxford: Oxford University Press, 1988), especially Chs 6-8; Peter de Marneffe, 'Rawls's Idea of Public Reason', Pacific Philosophical Quarterly 75 (1994): 232-250; Philip L. Quinn, 'Political Liberalisms and their Exclusion of the Religious', in Religion and Contemporary Liberalism, edited by Paul J. Weithman (Notre Dame: Notre Dame University Press, 1997): 138-161; Fred M. Frohock, 'The Boundaries of Public Reason', American Political Science Review 91 (1997): 833-844; and David A. Reidy, 'Rawls's Wide View of Public Reason: Not Wide Enough', Res Publica 6 (2000): 49-72; and John Horton, 'Rawls, Public Reason and the Limits of Liberal Justification', Contemporary Political Theory 2 (2003): 5-23. Rawls, 'Public Reason Revisited', p. 585; PL, p. 225. See Lief Wenar, 'Political Liberalism: An Internal Critique', Ethics (October 1995): 32-62; Erin Kelly and Lionel McPherson, 'On Tolerating the Unreasonable', Journal of Political Philosophy 9 (2001): 38-55. For a reply to this sort of objection, see Larry Krasnoff, 'Consensus, Stability, and Normativity in Rawls's Political Liberalism', Journal of Philosophy 95 (1998): 269-292, especially p. 279, n. 8 (answering Wenar). For an excellent response to Kelly and McPherson, see Jonathan Quong, The Rights of Unreasonable Citizens' (forthcoming in Journal of Political Philosophy). See, e.g. Michael Sandel, 'A Response to Rawls' Political Liberalism', in Liberalism and the Limits of Justice, 2nd edn (Cambridge: Cambridge University Press, 1998), pp. 202-210. See also, Stephen Mulhall and Adam Swift, Liberals and Communitarians 2nd edn (Oxford: Blackwells, 1996), pp. 234-235; Simon Clarke, 'Contractarianism, Liberal Neutrality and Epistemology', Political Studies 47 (September 1999): 627-642; and Joseph Chan, 'Legitimacy, Unanimity and Perfectionism', Philosophy & Public Affairs 29 (Winter 2000): 5-42. The literature here is vast and continually expanding. See, e.g. Nicholas Wolterstorff, 'The Role of Religion in Decision and Discussion of Political Issues', in Robert Audi and Nicholas Wolterstorff, Religion in the Public Square: The Place of Religious Convictions in Political Debate (New York: Roman & Littlefield, 1997), pp. 67-120; Paul J. Weithman, Religion and the Obligations of Citizenship (Cambridge: Cambridge University Press, 2002); and Christopher Eberle, Religious Convictions in Liberal Politics (Cambridge: Cambridge University Press, 2002). Kent Greenawalt, Religious Convictions and Political Choice, Ch. 8, esp. pp. 154f; Greenawalt, Private Consciences and Public Reasons (Oxford: Oxford University Press, 1995). Jeremy Waldron, 'Religious Contributions in Public Deliberation', San Diego Law Review 30 (Fall 1990): 817-848. For a direct reply to Waldron, see Lawrence Solum, 'Novel Public Reasons', Loyola of Los Angeles Law Review 29 (1996): 1459-1485. Bruce Ackerman, 'Political Liberalisms', The Journal of Philosophy (1994): 364-386. 215

538

John Rawls politics, philosophy & economics 3(2)

10. See Rawls, PL, Lecture VI, §7.1, pp. 240-241. 11. Rawls, PL, p. 243. In the 'Introduction to the Paperback Edition', Rawls says that he did not intend this passage to be an argument for a qualified right to abortion, although it does represent his opinion. The express purpose of the original footnote was to demonstrate how a comprehensive doctrine might 'run afoul of public reason'. See his comments in PL, p. Ivf., n. 31 ; or 'Public Reason Revisited', p. 605, n. 80. 12. As Philip Quinn writes, 'Some who agree with Rawls about the values at stake will claim that a fetal right to life is required to give substance and force to that value'. See Quinn, 'Political Liberalisms and their Exclusion of the Religious', p. 150; also de Marneffe, 'Rawls' Idea of Public Reason', p. 234f. 13. This is obviously an enormous assumption. However, my purpose here is not to argue the issue of abortion but simply to illustrate one type of incompleteness. For contrasting perspectives on the question of whether a reasonable balance of political values must support at least a qualified right to abortion, see Judith Jarvis Thomson, 'Abortion: Whose Right?', Boston Review (Summer 1995) and Peter de Marneffe, 'Rawls's Idea of Public Reason', especially sees. 3-4, pp. 239-243. 14. The following discussion of inconclusiveness owes much to Gaus's account of public justification in Justificatory Liberalism (Oxford: Oxford University Press, 1996), especially sees. 9.3-9.4, pp. 144-158 [hereafter cited as JL\. 15. See, Gaus, 7L, p. 148. Cf. John Rawls, A Theory of Justice (Cambridge, MA: Belknap Press, 1971), pp. 358f., revised edition Oxford University Press, 1999, p. 315 (hereafter cited as 77, with rev. ed. page numbers in brackets). 16. The argument for this, which I cannot defend here, is the idea that, generally speaking, individuals have authority over determinations about what they have reason to believe. If political principles are to be imposed on them, despite their disagreement, those principles must be justified in a way that overcomes the presumption in favour of their individual epistemic authority. See Gaus, 7L, sec. 9.3.3, pp. 149-152. For a similar argument, see Colin Bird, 'Mutual Respect and Neutral Justification', Ethics 107 (October 1996): 62-96, pp. 78f. 17. I return to this point in sec. 2.2 below. 18. See Joseph Raz, Morality of Freedom (Oxford: Clarendon Press, 1986), Ch. 13, especially pp. 333f. For criticism of Raz and other incomparabilists, see Ruth Chang, éd., 'Introduction', Incommensurability, Incomparability, and Practical Reason (Cambridge, MA: Harvard University Press, 1997): 1-34. For the view that vagueness is the only plausible explanation for incomparability, see John Broome, 'Is Incommensurability Vagueness?' in Chang, Ch. 4, pp. 67-89. 19. There may of course be other political values that are applicable to the issue of abortion, but I leave these aside to simplify discussion. 20. Quinn, 'Political Liberalisms and their Exclusions of the Religious', p. 150. 21. Gaus, 7L, p. 156. 22. How to decide between two maximally reasonable options is a problem I return to in sec. 3. 23. See Rawls, TJ, p. 61 [53]; also, Brian Barry, 'How Not to Defend Liberal Institutions', in Liberalism and the Good, edited by Bruce R. Douglas et al. (New York: Routledge, 1990): 3-58, p. 44. 24. See Gaus, JL, Ch. 10, especially sec. 10.2. 25. For different perspectives on the jurisprudence of the First Amendment religion 216

John Rawls

539

Schwartzman: The completeness of public reason

26. 27. 28.

29. 30. 31. 32. 33.

34. 35. 36.

37. 38.

39.

clauses, see David Richards, Toleration and the Constitution (Oxford: Oxford University Press, 1986), Ch. 5; Michael Perry, Religion in Politics (Oxford: Oxford University Press, 1997), Ch. 1; Amy Gutmann, 'Religion and the State in the United States: A Defense of Two-Way Protection', in Obligations of Citizenship and Demands of Faith, edited by Nancy Rosenblum (Princeton: Princeton University Press, 2000): 127-164. Gaus, JL, p. 158. Reidy, 'Rawls's Wide View of Public Reason', p. 65. There may also be permutations of these two types of incompleteness, for example, where a person is inconclusively justified in believing that public reason results in nested indeterminacy. Although I do not pursue this possibility in depth, it may have important implications for establishing the burden of proof in arguments about incompleteness. See sec. 3.2 for further discussion. Rawls, PL, pp. 54-58. For the distinction between cost and difficulty, see Cohen, If You're an Egalitarian, How Come You're So Rich? (Cambridge, MA: Harvard University Press, 2000), p. 171. Gaus,/L, p. 155. For the idea of a 'sufficiently credible belief, see Gerald Gaus, 'The Rational, the Reasonable, and Justification', Journal of Political Philosophy 3 (1995): 232-256, pp. 250-252. Jeremy Waldron, Law and Disagreement (Oxford: Oxford University Press, 1999), p. 161; also Thomas Christiano, The Significance of Public Deliberation', in Deliberative Democracy, edited by James Bohman and William Rehg (Cambridge, MA: MIT, 1999): 243-277, p. 276, n. 7; Chan, 'Legitimacy, Unanimity, and Perfectionism', p. 7, n. 7; Roberto Alejandro, The Limits ofRawlsian Justice (Baltimore: Johns Hopkins University Press, 1998), p. 134. Rawls, 77, p. 5 [4]; Rawls, PL, Lecture I, §6.1, pp. 35f. Rawls, Justice as Fairness, §3.2, p. 9; cf. Rawls, PL, pp. 35f. Rawls, PL, pp. liif. Rawls writes, 'Accepting the idea of public reason and its principle of legitimacy emphatically does not mean, then, accepting a particular conception of justice down to the last details of the principles defining its content. We may differ about these principles and still agree in accepting a conception's more general features. We agree that citizens share in political power as free and equal, and that as reasonable and rational they have a duty of civility to appeal to public reason, yet we differ as to which principles are the most reasonable basis of public justification. The view I have called "justice is fairness" is but one example of a liberal political conception; its specific content is not definitive of such a view'. PL, p. 226. Rawls, 'Public Reason Revisited', p. 582. Rawls, PL, p. 230. This is consistent with what Rawls says in Part II of 77, where he writes, 'Now the question of whether legislation is just or unjust, especially in connection with economic and social policies, is commonly subject to reasonable differences of opinion'. 77, § 31, pp. 198f. [174]. These points are explored more fully in an excellent discussion by David Estlund, 'The Survival of Egalitarian Justice in John Rawls's Political Liberalism', Journal of Political Philosophy 4 (1996): 68-78. 217

540

John Rawls politics, philosophy & economics 3(2)

40. 41. 42. 43. 44. 45.

46. 47. 48. 49. 50. 51. 52. 53.

54. 55. 56. 57. 58. 59. 60. 218

Gaus, JL, pp. 232f. Rawls, PL, pp. 232-233. Rawls, 'Public Reason Revisited', p. 579. Rawls, PL, p. 235. Micah Schwartzman, Towards a Defence of Public Reason' (D.Phil, thesis, University of Oxford, 2003), pp. 57-58. For a recent example of this sort of reply, see T.M. Scanlon, 'Rawls on Justification', in The Cambridge Companion to Rawls, edited by Samuel Freeman (Cambridge: Cambridge University Press, 2003), p. 163. For an important response to Scanlon, see Jonathan Quong, 'The Scope of Public Reason' (forthcoming in Political Studies). Rawls, PL, p. 215; for discussion, see Stephen Mulhall and Adam Swift, Liberals and Communitarians, p. 227. Rawls, PL, Ivi. See, for example, Simon Caney, 'Antiperfectionism and Rawlsian Liberalism', Political Studies 43 (1995): 248-264, p. 250. For criticism of attempts by John Finnis and others to argue on nonreligious grounds against the morality of homosexual relationships, see Perry, Religion and Politics, pp. 82-96. See Judith Jarvis Thomson, 'A Defense of Abortion', Philosophy & Public Affairs 1 (1971): 47-66. Greenawalt, Religious Convictions and Political Choice, pp. 154f. Ibid. Robert George, who argues that the prohibition of abortion can be justified on publicly accessible grounds, thinks that restrictions on abortion in cases of rape and incest are a 'close call'. For what it is worth, I think his arguments for this view, which depend on extending the rule of double effect to cover abortion in these (but not other) cases, are dubious to say the least. For George's view, see his essay on 'Law, Democracy, and Moral Disagreement: Reciprocity, Slavery, and Abortion', in Deliberative Politics: Essays on Democracy and Disagreement, edited by Stephen Macedo (Oxford: Oxford University Press, 1999): 184-197, esp. p. 196, n. 21. For criticism of George's claim (following Finnis) that, aside from cases of rape and incest, the general issue of abortion is 'not even a close call', see the reply by Gutmann and Thomson in the same volume, at pp. 268-270; also Stephen Macedo, 'In Defense of Liberal Public Reason: Are Slavery and Abortion Hard Cases?', in Natural Law and Public Reason, edited by Robert P. George and Christopher Wolfe (Washington, DC: Georgetown University Press, 2000): 31-45. A fairly balanced discussion of these issues can be found in Amy Gutmann and Dennis Thompson, Democracy and Disagreement (Cambridge, MA: Harvard University Press, 1996), pp. 73-77, 85-90. Rawls, PL, p. 56. Rawls, 'Public Reason Revisited', p. 584. Rawls, PL, pp. 224, 245. Reidy, 'Rawls's Wide View of Public Reason', p. 65. Greenawalt, Religious Convictions and Political Choice, p. 147. Ibid., p. 155. Reidy, 'Rawls's Wide View of Public Reason', p. 67.

John Rawls

541

Schwartzman: The completeness of public reason 61. Rawls, PL, p. lui. 62. Here I follow Ronald Dworkin's criticism of claims about indeterminacy in 'Objectivity and Truth: You'd Better Believe It', Philosophy & Public Affairs 25 (1996): 87-139. 63. David Richards, 'Book Review', Georgia Law Review 23 (1989): 1189-1200, p. 1192. 64. Gaus, /L, p. 225. 65. Robert Audi, 'Religion and the Ethics of Political Participation', Ethics 100 (1990): 386-397, pp. 395f. 66. Rawls, 'Public Reason Revisited', p. 585. 67. Reidy, 'Rawls's Wide View of Public Reason', p. 68. 68. Cass Sunstein and Edna Ullmann-Margalit, 'Second-Order Decisions', Ethics 110 (1999): 5-31, sec. ELD. 69. Ibid. 70. See Jon Elster, Sour Grapes: Studies in the Subversion of Rationality (Cambridge: Cambridge University Press, 1983), pp. 16-18; Shaun Hargreaves Heap et al., The Theory of Choice (Oxford: Blackwell, 1992), pp. 14-21. 71. Re A (Children) (Conjoined Twins) [2000] 4 All ER 961. 72. Reidy, 'Rawls's Wide View of Public Reason', p. 64. 73. See Amy Gutmann and Dennis Thompson, 'Moral Conflict and Political Consensus', Ethics 101 (1990): 64-88; David B. Wong, 'Coping with Moral Conflict and Ambiguity', Ethics 102 (1992): 763-784; Frederick Schick, 'Coping with Conflict', Journal of Philosophy 85 (1988): 362-375. 74. Gutmann and Thompson, 'Moral Conflict and Political Consensus', pp. 82f. 75. Although for criticism of this suggestion, see Gaus, JL, p. 184. 76. The institutional implications of public justification are discussed in Rawls, PL, Part III; Brian Barry, Justice as Impartiality (New York: Oxford University Press, 1995), Ch. 4; and perhaps most systematically in Gaus, JL, Part III. For an attempt to justify democracy on the basis of its unique ability to create incentives for public reasoning, see Richard Vernon, Political Morality: A Theory of Liberal Democracy (London: Continuum, 2001), especially Chs 2-4. 77. Rawls, TJ, p. 362 [318]. 78. Cf. Rawls, PL, pp. lv-lvii, 241. 79. Andrew Williams, 'The Alleged Incompleteness of Public Reason', Res Publica (2000): 199-211. 80. For the distinction between 'picking' and 'choosing', see Edna Ullman-Margalit and Sidney Morganbesser, 'Picking and Choosing', Social Research 44 (1977): 757-783. 81. Elster, Solomonic Judgements (Cambridge: Cambridge University Press, 1989), p. 116 (italics in the original). 82. Ibid., p. 121. 83. Ibid., Ch. 3. 84. For examples, see Solomonic Judgements, Ch. 2. 85. Anthony Laden defends this view in his book Reasonably Radical (Ithaca: Cornell University Press, 2001), p. 123. Laden writes, 'By hypothesis, there are no decisive public reasons that reject the outcome, and so I do not have substantive grounds for raising a deliberative challenge. This leaves the possibility that I might object to a 219

542

John Rawls politics, philosophy & economics 3(2)

86. 87. 88. 89.

220

decision procedure that allowed people to rely on nonpublic reasons as tiebreakers. Since the alternative in this narrow range of cases is that people vote hypocritically or arbitrarily, it seems as if I have good public reasons to allow some nonpublic reasons to play this kind of limited role. At the same time, I have reason to reject allowing just any kind of nonpublic reasons to play this role. Reasons of narrow self-interest, for instance, might still be unacceptable'. I have three objections to Laden's argument. First, he does not present an argument against deciding arbitrarily; nor does he consider that there might be public reasons for endorsing a random decision procedure. Second, the possibility of deciding arbitrarily means that citizens do not have to be hypocritical. They do not have to appeal to nonpublic reasons and then pretend they voted based on public reasons. They can defend their decisions in good conscience by saying that the fairest or most legitimate way to resolve the indeterminacy was by deciding arbitrarily. Third, Laden does not explain what kinds of nonpublic reasons count as legitimate tie-breakers. Without some test to determine which nonpublic reasons are acceptable, I believe his view is open to objections based on examples similar to the one about organ allocation presented in the text below. Here I follow Williams, 'The Alleged Incompleteness of Public Reason', p. 210. For discussion, see Joseph Raz, Practical Reason and Norms 2nd edn (Oxford: Oxford University Press, 1990), especially sees. 1.2, 2.2, and the 'Postscript to the Second Edition'. Ibid., p. 45. Ibid., p. 43.

Part IV A Liberal Foreign Policy

This page intentionally left blank

[20] Critical Notice KOK-CHOR TAN JOHN RAWLS, The Law of Peoples: With the 'Idea of Public Reason Revisited'. Cambridge, MA: Harvard University Press 1999. Pp. vii + 199.1 I

Taking their inspiration from John Rawls's A Theory of Justice, liberals like Charles Beitz and Thomas Pogge have argued that Rawls's arguments for social and economic equality should apply also to the global context.2 Just as Rawls considers a person's race, gender, talents, wealth and other natural and social particularities about her to be 'arbitrary from a moral point of view/3 so too, they argue, are factors like a person's nationality and citizenship morally arbitrary. And as the effects of these contingencies on a person's life chances in the domestic sphere are to be

1 Many thanks to Karen Detiefsen, David Dyzenhaus, Will Kymlicka, Cheryl Misak and two referees of this journal for their very helpful comments and criticisms on an earlier draft. I thank also members of the Political Philosophy Reading Group, Queen's University — in particular, Philippe Constantineau, Jackie Davis, Alistair Macleod, Christine Sypnowich — for a lively discussion on the book under review here. Finally, my gratitude to Erin Kelly whose thoughtful comments on a different (unpublished) paper at the American Philosophical Association Eastern Division 1998 Meeting alerted me to some of the issues that I am now addressing, and to Thomas Pogge for helpful and generous discussion on this topic on several different occasions. This paper is part of a larger postdoctoral project that is being funded by the Social Sciences and Humanities Research Council of Canada. 2 John Rawls, A Theory of Justice (Cambridge, MA: Harvard University Press 1971). Charles Beitz, Political Theory and International Relations, 2nd ed. (Princeton: Princeton University Press 1999); Thomas Pogge, Realizing Rawls (Ithaca: Cornell University Press 1989).

3 Rawls, Theory of Justice, 15

546

John Rawls 114 Kok-ChorTan

nullified by certain distributive principles of justice, so too should the effects of global contingencies be mitigated by certain global distributive principles. Thus, Rawls's two principles of justice — in particular the second principle governing social and economic equality — should apply between individuals across societies and not just within a single society.4 But in his own extended commentary on international relations, Rawls explicitly rejects the concept of global distributive justice. This view was first presented in his essay "The Law of Peoples/ and is further developed in his latest book of the same title.5 In this critical notice, I wish to evaluate Rawls's reasons for rejecting the idea of global distributive justice. An important contrast between Rawls's Law of Peoples and the views of liberals like Beitz and Pogge, as we will see, is that the former is avowedly non-cosmopolitan. I will argue, however, that a liberal Law of Peoples ought to endorse the cosmopolitan ideal. To see better where the dispute between Rawls and the proponents of cosmopolitan justice lies, let me provide a very quick overview of Rawls's international project. The fundamental aim of The Law of Peoples is to examine how the content of a theory of international justice 'might be developed out of a liberal idea of justice similar to, but more general than, the idea [of] justice as fairness' (3).6 This 'globalizing' project proceeds in three stages. The first stage extends the 'social contract idea to the society of liberal peoples' (4-5). This first stage allows us to identify the international principles that would be agreed to by representatives

4 To recall, the two principles are: 1) Each person is to have an equal right to the most extensive basic liberty compatible with a similar liberty for others; 2) Social and economic inequalities are to be arranged so that they are both (a) to the greatest benefit of the least advantaged, consistent with the just savings principles, and (b) attached to offices and positions open to all under conditions of fair equality of opportunity (Theory of Justice, 302). Principle 2(a) is labelled the difference principle. 5 Rawls, "The Law of Peoples' in On Human Rights, Stephen Shute and Susan Hurley, eds. (New York: Basic Books 1993), 41-82,220-30; and The Law of Peoples (Cambridge, MA: Harvard University Press 1999). References to The Law of Peoples will be subsequently noted in parenthesis in the text. 6 One issue I will leave aside in this discussion is Rawls's assumption that peoples are more or less coextensive with states as these are currently demarcated, as his remarks on 38-9 suggest. But for the purpose of consistency, I will use 'peoples' or 'societies' (interchangeably as Rawls does) instead of 'states/ To be sure, Rawls took pains to dissociate peoples from states; but his main intention here is to distinguish peoples from 'political states as traditionally conceived' (25) in the realist tradition, i.e., as political entities motivated primarily by self-interests and power, rather than to question existing political boundaries (39; see 25-8).

John Raw I s Critical Notice of John Rawls The Law of Peoples 115

of liberal societies at a global original position.7 Among the principles affirmed here include the duty of humanitarian assistance (37).8 The second stage, which concerns one of the central themes of The Law of Peoples, aims to show how and why representatives of certain nonlibeml but well-ordered societies would also endorse the same set of principles. These are nonliberal societies in that they do not endorse the standard range of liberal democratic rights, like the freedom of expression and association, religious equality, the right to equal political participation, and so on. That is, individuals in nonliberal societies are 'not regarded as free and equal citizens, nor as separate individuals deserving equal representation' (71; see 71-5). Yet these societies honor basic human rights (e.g., right to life and security, and subsistence) and are respectful of other peoples (64-7) as required by the Law of Peoples. Consequently, these nonliberal 'decent peoples/ as Rawls calls them, qualify as 'societies in good standing/ and are, therefore, to be tolerated by liberal societies. This means that liberal societies are 'to recognize these nonliberal societies as equal participating members in good standing of the Society of Peoples/ and not just 'refrain from exercising political sanctions — military, economic, or diplomatic — to make a people change

7 The original position, as we may recall, is 'a device of representation' where representatives of rational but reasonable individuals deliberate on the appropriate principles of justice for the basic structure of their society. To ensure that this hypothetical deliberation is fair and equal, parties to the original position go behind a Veil of ignorance/ That is, they are asked to imagine that they do not know their actual status and station in society; their talents and conceptions of the good; nor the wealth and the extent of their territory and its population in the case of the global original position (32-3). In this way, no one party could insist on terms biased in her favor according to her own social standing. See Theory of Justice, 17ff. An important difference in the global original position is that the parties to the deliberation are representatives of peoples rather than of individuals. The significance of this anti-individualist shift will be discussed in due course. 8 The eight principles (abridged) are: 1) Peoples are free and independent. 2) Peoples are to observe treaties. 3) Peoples are equal and are parties to the agreements binding on them. 4) Peoples have a duty of non-intervention. 5) Peoples have the right of self-defense, but not the right to wage war other than for self-defense. 6) Peoples are to honor human rights. 7) Peoples are to observe justice in war. 8) Peoples have a duty to assist peoples lacking the resources to sustain just regimes. These principles are not exhaustive and more may be added (37).

547

548

John Rawls 116 Kok-ChorTan

its ways' (59). Nonliberal peoples are tolerated as a matter of liberal principle, and not merely accommodated on account of practicality. This last point is important. The Law of Peoples wants to achieve a global stability with respect to justice, and not a stability as a modus vivendi (i.e., a stability as a balance of forces) (12-13; 44-5). The two-stage procedure described above is thus crucial to Rawls's project because it attempts to show that the global principles proposed by liberal peoples are also principles that can be independently adopted by decent nonliberal peoples; that it is not the case that liberal peoples have tailored their global principles specifically in view of accommodating nonliberal peoples or existing global institutional arrangements. Whether Rawls succeeds in meeting his stated goal is a question we will take up below. These two stages complete the ideal theory part of the Law of Peoples. The aim of ideal theory is to identify the principles that should govern the relationship between societies with the requisite political and economic conditions to be well-ordered and to comply with the Law of Peoples. In this ideal condition, the goals of justice and stability for the right reasons between societies can be achieved. But how about societies without the economic resources to support well-ordered institutions, or societies that blatantly refuse to comply with the principles of the Law of Peoples? These difficulties stemming from 'the highly nonideal conditions of our world with its great injustices and widespread social evils' (89) is the concern of the third stage of Rawls's project. The nonideal theory aspect of the Law of Peoples thus addresses (i) the problem of noncompliance, as when 'outlaw' societies 'refuse to comply with a reasonable Law of Peoples' (90), and (ii) the problem of unfavorable conditions, where burdened' societies lack the basic resources to become well-ordered. A complete Law of Peoples has to confront these nonideal cases, and offer guidance on how well-ordered peoples may defend themselves against outlaw regimes and help bring on reform within these regimes in the long run (92-3); and how they may assist burdened societies and help bring them 'into the Society of well-ordered Peoples' (106). From his discussions concerning burdened societies and his proposal of a duty of assistance, it is clear that Rawls does not advocate an isolationist foreign policy which holds the fate of these societies to be a matter of indifference for liberal and decent peoples. He maintains that better-off societies have a duty of humanitarian assistance towards burdened societies in order to help them achieve the requisite level of economic and social development to become well-ordered. Yet Rawls also stresses that this duty of humanitarian assistance is distinct from, and does not entail, a duty of distributive justice. So while a duty of humanitarian assistance is required by the Law of Peoples as part of its nonideal theory, a distributive principle has no place at all here.

John Rawls Critical Notice of John Rawls The Law of Peoples

II

549 117

Humanitarian Duty and Duty of Justice

Why does Rawls reject the concept of global distributive justice? He gives us two main arguments, as I see it. The first is that since a duty of humanitarian assistance is already required by the Law of Peoples as part of nonideal theory, global principles of distributive justice would be redundant. The other is that global distributive principles, moreover, would have unacceptable results. I will explain and assess these arguments in turn.9 Consider the redundancy argument. In our nonideal world, with its gross injustices, vast inequality and abject poverty, the Law of Peoples, as we have seen, recognizes that 'well-ordered peoples have a duty to assist burdened societies/ to bring them into the Society of well-ordered peoples (ibid., 106). Yet, the argument goes, these 'goals of attaining liberal or decent institutions, securing human rights, and meeting basic needs ... are [sufficiently] covered by the duty of assistance' (ibid., 116). Thus a global distributive principle serves no additional purpose in this regard. But this argument misses one important difference between duties of humanity and duties of justice, a difference that is more than semantic. If we concede that rich countries have only a duty of humanity to poorer countries, we are conceding also that the existing baseline resource and wealth distribution is a just one, that the global basic institutions organized around and legitimizing the prevailing allocation of wealth and resources are acceptable. Duties to assist each other, on this account, are duties that take place within a just institutional framework. In other words, duties of humanity speak to how states should interact with one another, and while certain institutional mechanisms may be required to facilitate some of this interaction, the global basic structure (e.g., the norms governing the allocation and ownership of resources and wealth), within which such interactions occur, is taken as a given. By contrast, duties of justice speak directly to the basic structure; justice is concerned with the baseline distribution of wealth and resources, and the basic institutions and principles that legitimize and rationalize this distribution. To put it perspicuously, while duties of humanity aim to redistribute wealth, duties of justice aim to identify what counts as a just distribution in the first place. The aim of justice, properly speaking, is not to transfer wealth as such (i.e., by taking them from their rightful owners and reallocating them to others), but rather to establish the criteria of

9 This section is adapted from Chap. 7 of my Toleration, Diversityf and Global Justice (University Park, PA: Penn State University Press 2000).

John Rawls

550 118 Kok-ChorTan

rightful ownership, to redefine 'what justly belongs to a country/10 Duties of justice, then, would call on us to reconceive our present global basic structure; whereas duties of humanity take this to be more or less sound, and only exhort countries to do more within this given framework. One could say with Brian Barry that justice is prior to humanity in that 'We cannot sensibly talk about humanity unless we have a baseline set by justice. To talk about what I ought, as matter of humanity, to do with what is mine makes no sense until we have established what is mine in the first place' (Barry, 249). Hence the long-term aims of humanity and justice are quite distinct, but not just in their objective or duration, as Rawls notes, but also in their subject. The former calls for greater humanitarianism between countries within an existing institutional framework, whereas the latter calls for a critical evaluation of that framework. This is an important difference. As many egalitarians have long argued, the root cause of global poverty and inequality lies with our global arrangements and institutions, and so to seriously tackle global poverty would require reforming these institutions and arrangements.11 Moreover, this difference in focus has important and more immediate consequences for foreign policy. If foreign aid is considered a matter of humanity, it could be subject to conditions imposed by donor countries (it is their resources they are giving up, after all, on this view). But if we treat foreign aid as a matter of justice, it would not be vulnerable to such demands in principle, for any resource transfer is, on this view, strictly speaking, not a redistribution in the sense of taking something from its rightful owner and giving it to the more needy, but a correction of an initial unjust distribution. Thus treating global inequality as a matter of humanity, as Barry points out, obscures the basic point, 'that if some share of resources is justly owed to a country, then it is (even before it

10 Brian Barry, 'Humanity and Justice in Global Perspective/ in Ethics, Economics and the Law, J. Roland Pennock and John W. Chapman, eds. (New York: New York University Press 1982) 219-52, at 248. See also Pogge, Realizing Rawls, 16-17,68-70. 11 I will not argue for this point here, but see, for some examples, Frank Cunningham, 'Democracy, Socialism, and the Globe/ in The Real World of Democracy Revisited (Atlantic Highlands, NJ: Humanities Press 1994) 137-54, at 143-5; Kai Nielsen, 'Global Justice, Capitalism and the Third World/ injustice and Economic Distribution, 2nd ed. John Arthur and William H. Shaw, eds. (Englewood, NJ: Prentice Hall 1991) 228-41, at 229-32; Onora O'Neill, Faces of Hunger: An Essay on Poverty, Development and Justice (London: George Allen and Unwin 1986); Pogge, 'Economic Justice and National Borders/ Revision 22.2 (1999) 27-34; and Henry Shue, Basic Rights: Subsistence, Affluence and US Foreign Policy, 2nd ed. (Princeton: Princeton University Press 1996).

John Rawls Critical Notice of John Rawls The Law of Peoples 119

has been actually transferred) as much that country's as it is now normally thought that what a country normally produces belongs to that country/12 So it makes an immense difference whether wealth redistribution between countries is conceived as a matter of humanity or justice. Far from being superfluous, treating duties between countries as a matter of justice highlights for us the proper locus of our concern (i.e., institutions and their underlying norms) and reminds us that the crucial issue is ultimately the question of rightful ownership rather than that of humanitarian contribution. Even if we were to confine ourselves to the nonideal case of burdened societies, it makes an important normative difference whether we think we are assisting only out of humanitarian concern, or whether we are assisting because we recognize the fact of prevailing injustices in our global arrangement. None of the above denies that humanitarian assistance is important as well. But as long as humanitarianism takes place within the present global arrangement, it serves only to treat the symptoms of injustice rather than tackle the underlying cause of it. Humanitarian assistance applies as long as there are burdened societies, but principles of justice would push us to assess the framework within which such assistance is being rendered. It is a familiar fact that Rawls's domestic egalitarianism is directed at the basic institutions of society; his second principle provides liberals with a basis against which to assess and critique these institutions, and to reject institutional arrangements which perpetuate and legitimize inequality of opportunity between citizens. So it seems that to be consistent with his basic philosophical ideals, Rawls too should hold up the basic structure of the Society of peoples against his principles of justice, instead of taking that as a given. So Rawls clearly is aware of the important differences between humanitarian duties and duties of justice. His supposition seems, then, to be that the global distribution of resources and wealth is not an issue of justice. The present global distribution of wealth and resources is taken

12 Barry, 248. Of course this does not mean that we are never in a position to withhold that which rightly belongs to another — recall Plato's example of not returning a dagger to its rightful owner who has gone mad (The Republic 331c). Barry notes that it is as acceptable to withhold resources justly owed to a country that is violating basic human rights or spending vast quantities of money on arms, as it is 'to refuse to pay debts to it or to freeze its asset7 (Barry, 248). And certainly such disincentives could be used as a tool for human rights reform, if properly applied. The crucial point here is that treating inequality as a matter of justice shifts the burden of proof away from 'recipient' countries and onto the 'donor' country.

551

John Rawls

552 120 Kok-ChorTan

to be an acceptable starting point for him, and not as arbitrary as some of his own followers have thought. But why does Rawls think this? An examination of Rawls's second argument for rejecting global distributive principles may shed some light on this question. Rawls's belief here is that global distributive principles, unlike domestic distributive principles, would have unacceptable results. As he tells us, a duty of humanitarian assistance is a 'principle of transition ... [it] holds [only] until all societies have achieved just liberal or decent basic institutions. [It is] defined by a target beyond which [it] no longer hold[s]' (Law of Peoples, 118). That is, the duty of assistance is satisfied once all societies have attained the basic developmental level sufficient for establishing and maintaining decent institutions. By contrast, distributive 'principles do not have a defined goal, aim, or cut-off point, beyond which aid may cease' (ibid., 106). So while a duty of humanity would work towards improving the situation of societies 'burdened' by unfavorable circumstances, such assistance is not required as part of ideal theory in which all societies are assumed to have attained the basic developmental level requisite for a decent society. A principle of distributive justice, on the other hand, is an integral part of ideal theory, and so would apply as long as there are inequalities between societies, even 'after the duty of assistance is fully satisfied' (ibid., 117). But, and here is the crux of the argument, this would have unacceptable results, argues Rawls, for we would then not be able to discriminate between societies which through foresight and prudence have increased their wealth, from societies which through neglect and imprudence have squandered theirs (ibid., 117); or societies which have managed to curb their population growth and are therefore better able to optimize their resources, from societies which have neglected to control their population and hence are worse-off as a result (ibid., 117-18). A global egalitarian principle would insist, in both of these cases, that resources be transferred from the wealthier societies to the poorer ones, even though both may have started with an equal amount of wealth and resources. And this is unacceptable for it would involve penalizing some societies for their sound domestic policies in order to compensate other societies for their careless policies. In short, while a duty of assistance is in force only within nonideal theory and would cease once no peoples are so burdened as to be unable to establish well-ordered institutions, a distributive principle falls under ideal theory and continues to apply as long as inequality between different societies persists. And it is this fact — that distributive principles would insist on redistribution as long as there is inequality between peoples no matter what the cause of this inequality — that Rawls takes exception to.

John Rawls Critical Notice of John Rawls The Law of Peoples 121

Implicit in Rawls's argument here, evidently, is the distinction between inequality as a result of choice and inequality due to circumstance. Just as a domestic distributive scheme ought not to compensate individuals for their poor choices by taking from those who have made good choices, neither, too, ought a global scheme to compensate societies for their poor governance by penalizing other societies for their good governance. The aim of distributive justice is to compensate persons for inequality due to circumstance outside their control and not to subsidize them for their (poor) choices.13 It appears, then, that Rawls thinks that in the global context, inequality in circumstance is sufficiently redressed by the duty of humanity. A distributive principle, on the other hand, would be insensitive to the choice/circumstance distinction; it would treat citizens of well-managed economies unfairly by transferring their gains to citizens of poorly-managed economies continuously as long as global inequality remains. One crucial premise of this argument is that the reasons for a country's failure to adopt sound social and economic policies are largely internal, and hence freely adopted by governments of worse-off countries. Rawls draws attention to various domestic factors that determine a society's economic and social performance, including its political culture and virtues (including here a respect for basic human rights), its civic society and 'its members' probity and industriousness' and its population policy (108-11). Yet this premise, which Thomas Pogge has labelled 'explanatory nationalism/ is highly questionable as a matter of fact.14 Explanatory nationalism 'present[s] poverty as a set of national phenomena explicable mainly as a result of bad domestic policies and institutions that stifle (or fail to stimulate) national economic growth and engender national economic injustice' (ibid., 497). But, as Pogge notes, this explanation leave[s] open important questions, such as why national factors (institutions, officials, policies, culture, natural environment, level of technical and economic development) have these effects rather than others' by ignoring the causal effects of global factors (e.g., trade practices, consumption patterns of affluent countries, international law, etc.) on a nation's domestic policies and their outcomes (ibid., 498-9). But even if we were to accept explanatory nationalism, it is not so obvious that global distributive principles would violate the choice/cir-

13 See Kymlicka, Contemporary Political Philosophy (New York: Oxford University Press 1990), 73-6. 14 Pogge, 'The Bounds of Nationalism/ in Rethinking Nationalism, Jocelyne Couture, Kai Nielsen, and Michel Seymour, eds. (Calgary: University of Calgary Press 1998) 463-504, at 497-502.

553

554

John Rawls 122 Kok-ChorTan

cumstance distinction. As Charles Beitz points out, the domestic equivalent to the case of citizens faring poorly due to the bad policies of their own governments is not that of an individual having to bear the consequences of her own bad choices, but, rather, more like that of children who have to suffer for the poor choices of their parents. And in such a case, we would not say that 'the offspring are responsible for their own condition ... [and] considerations about responsibility do not diminish the weight of the ethical concern about the well-being of the offspring/15 Similarly, individual citizens of poor countries need not have freely consented to their countries' policies — indeed, they likely would not have had the option if they belong to hierarchical and nondemocratic societies, or if these were policies implemented before their time (as the case might well be, given the intergenerational implications of social policies like population control). Thus their disadvantages are due more to circumstance than choice, albeit circumstances of the society that they happen to find themselves in, and distributive principles that aim to compensate individuals for these disadvantages cannot be said to offend against the choice/circumstance distinction. If justice is individual choice-sensitive and circumstance-insensitive, then one cannot accept that global distributive principles neglect this distinction. The problem, then, with Rawls's second argument is that while the choice-circumstance distinction is applied individualistically in the domestic case, it is applied communally (to a people as one entity) in the global context. Citizens of disadvantaged countries are collectively held accountable for their country's unsound domestic policies, even when a majority of them had no part in the making of these polices. And this is clearly inconsistent with Rawls's own moral individualism.16 On Rawls's own reasoning, a person born into a society with poor population control and economic policies cannot be said to deserve her fate any more than another born into more favorable circumstances deserves her. These are mere accidents of birth, and are as morally arbitrary as is being born into wealth or poverty in the domestic context.

15 Beitz, 'Social and Cosmopolitan Liberalism/ 527. My argument here follows closely Beitz's discussion on 526-8. 16 Recall: We want to account for the social values, for the intrinsic good of institutional, community, and associative activities, by a conception of justice that in its theoretical basis is individualistic. For reasons of clarity among others, we do not want to rely on an undefined concept of community, or to suppose that society is an organic whole with a life of its own distinct from and superior to that of all its members in their relations with one another' (A Theory of Justice, 264).

John Rawls Critical Notice of John Rawls The Law of Peoples 123

We may put the above point in a somewhat different way: while Rawls's moral individualism sets firm limits on the extent to which collective decision may affect individual well-being in his domestic conception of justice, there seems to be no similar limitations in his international theory. Collective national-decisions are regulated and constrained by principles of justice that take the individual to be the basic reference point in the domestic context; yet in moving to the international context, the same restriction no longer applies.17 It is true that questions of inefficiency and waste do arise when channeling resources to poorly planned economies (so we may have to think carefully how we are to tackle these problems efficiently and with minimum waste); but this does not undermine what justice as a matter principle demands.18 Distributive principles are still in force; how they are to be affected or realized is a separate question, a question of policy or strategy, and it is only at this level that a country's domestic policy becomes a relevant factor of consideration.19 Ill

Liberalism and Cosmopolitan Justice

The aim of my above argument is not to show that Rawls's Law of Peoples is objectionable because it is not sufficiently individualistic—that would be no objection for one of Rawls's propositions is that the basic units of the Law of Peoples are peoples or societies, and not individuals. My point, rather, is that the choice/circumstance distinction I take to be implicit in

17 This was suggested to me by Will Kymlicka. 18 See Beitz, 'Social and Cosmopolitan Liberalism/ 528; also Pogge, 'An Egalitarian Law of Peoples/ Philosophy and Public Affairs 23/3 (1994): 195-224,202-5. 19 It might be pointed out that in Rawls's view (if not in all defensible conceptions of justice), efficiency is one of the factors (even if it is not among the most fundamental) to be taken into consideration when deliberating about justice, and so my claim that we can separate efficiency from principle is mistaken. But this objection commits a category mistake by conflating two different accounts of efficiency. To explain: a complete presentation of a theory of justice has to first (a) identify a set of principles that fulfills, inter alia, some stipulated efficiency criterion (e.g., that it does not stunt individual incentive or ambition); then, it has to (b) ask how we can most efficiently realize these principles. The former poses a conceptual question, the latter, a strategic one. Both these efficiency considerations are, of course, highly relevant to the pursuit of justice; but the former bears directly on what our principles should look like, the latter does not. So, unless one wishes to assert that corrupt third-world governments and wastage in cross-border resource transfers are inevitable facts of our world, there is no reason why such contingent and alterable facts alone should compel us to reconceive what justice requires.

555

John Rawls

556 124 Kok-ChorTan

Rawls's argument (i.ev that global distributive principles would have unacceptable results) makes sense only when applied individualistically, but not when it is applied to a society as a whole. At any rate, Rawls explicitly retracts the individualism that informs his domestic theory of justice. As he contrasts the Law of Peoples with what he calls the cosmopolitan view: "The ultimate concern of a cosmopolitan view is the well-being of individuals and not the justice of societies.... What is important to the Law of Peoples is the justice and stability for the right reasons of liberal and decent societies' (119-20). Rawls's rejection of cosmopolitanism, to my mind, reveals a fundamental shift in his political philosophy. Starting from purportedly liberal individualistic grounds, Rawls arrives at an international theory that is more aligned with those of 'communitarians/ like David Miller and Michael Walzer, who have very different philosophical starting points.20 Compare, for example, Rawls's opposition to global distributive justice with Miller's: To respect the self-determination of other nations also involves treating them as responsible for decisions they may make about resource use, economic growth, environmental protection, and so forth' (Miller, 108). Why does Rawls reject the cosmopolitan ideal in his international theory? After all, the cosmopolitan view seems to be more consistent with his famous phrase that individuals are 'self-originating sources of valid claims/21 And, as mentioned, Rawls's own followers have long argued that the cosmopolitan view follows naturally from Rawls's own fundamental commitments.22

20 David Miller, On Nationality (Oxford: Oxford University Press 1995), esp. Chaps 3 and 4; and Michael Walzer, Spheres of Justice (New York: Basic Books 1983), and Walzer, 'The Moral Standing of States: A Response to Four Critics/ Philosophy and Public Affairs 9.3 (1980) 209-29. See also other 'communitarians' like Michael Sandel, Democracy's Discontent (Cambridge, MA: Harvard University Press 1996), 338ff; and Alasdair Maclntyre, 'Is Patriotism a Virtue?' The Lindley Lectures. Dept. of Philosophy, University of Kansas 1984. 21 Rawls, 'Kantian Constructivism in Moral Theory,' Journal of Philosophy 77 (1980) 515-72, at 543 22 Thus Pogge: 'Taken seriously, Rawls's conception of justice will make the life prospects of the globally least advantaged the primary standard for assessing our social institutions' ('Rawls and Global Justice/ The Canadian Journal of Philosophy 18.2 [1988] 227-56, at 233); and Beitz: 'It seems obvious that an international difference principle applies to persons in the sense that it is the globally least advantaged representative person (or group of persons) whose position is to be maximized' (Political Theory and International Relations, 152).

John Rawls Critical Notice of John Rawls The Law of Peoples 125

Rawls's reason for this shift from his domestic individualism to his international communitarianism is that a Law of Peoples founded on the ideal of individuals as free and equal would make the basis of that Law 'too narrow/23 In other words, to insist on an international theory of justice premised on the cosmopolitan ideal that individuals are ultimate, is to propose a conception of justice that nonliberal societies could reasonably object to. It would amount in effect 'to saying that all persons are to have the equal liberal rights of citizens in a constitutional democracy ... that only a liberal democratic society can be acceptable' (82-3). And this, Rawls says, 'would fail to express due toleration for other acceptable ways (if such as there are, as I assume) of ordering society' (59). It is for this reason that Rawls rejects the proposal that there be a single global original position procedure where individuals are represented, and opts instead for a two-stage procedure in which only representatives of societies are convened at the second global stage (82-3; 30-5). A global original position would have to assume that all individuals 'have the equal liberal rights of citizens in a constitutional democracy' (82), and this we should not do. This toleration for nonliberal ways of organizing society, Rawls argues, stems from a principle basic to political liberalism, that 'a liberal society is to respect its citizens' comprehensive doctrines — religious, philosophical, and moral — provided that these doctrines are pursued in ways compatible with a reasonable political conception of justice and its public reason7 (59). Thus, likewise, liberal societies are to tolerate nonliberal societies so long as these are decent, i.ev capable of conforming to the principles of the Law of Peoples. And as mentioned, one of the central aims of The Law of Peoples is to show how and why certain nonliberal peoples can nonetheless endorse the principles of the Law of Peoples, and why they are therefore reasonable societies or societies in good standing and to be tolerated by liberal peoples. The idea of public reason central to Rawls's political liberalism is therefore extended to the international context in the following way: 'in proposing a principle to regulate the mutual relations between peoples, a [liberal] people or their representatives must think not only that it is

23 The phrase in quotations is from the essay "The Law of Peoples/ 65. David Dyzenhaus has argued that this rejection of 'individualist liberalism' is already evident in Rawls's move from the comprehensive liberalism of-A Theory of Justice to the political liberalism defended in Political Liberalism. David Dyzenhaus, 'Critical Notice: Charles Larmore, The Morality of Modernity/ Canadian Journal of Philosophy 28.2 (1998) 269-86, at 280. If this is right, our criticism of Rawls's Law of Peoples amounts ultimately to a criticism of Rawls's political liberalism. I pursue this point in Toleration, Diversity, and Global Justice.

557

John Rawls

555 126 Kok-ChorTan

reasonable for them to propose it, but also that it is reasonable for other peoples to accept it' (57).24 So while it would not be unreasonable, but indeed 'is a consequence of liberalism and decency' (81), to criticize, and even intervene against in grave cases, violations of basic human rights in outlaw societies, it would be unreasonable to demand that all societies adopt liberal democratic institutions. This idea that the liberal concept of toleration extends also to nonliberal societies — in spite of their restrictions on important liberal freedoms like freedom of association, expression, right to democratic participation and so on — is a point of much contention in Rawls's international theory.25 But I will leave this matter aside and focus primarily on the implications of Rawls's account of toleration for distributive justice. Is it true that 'political liberalism would fail to express due toleration for other acceptable ways ... of ordering society' (59) if it insisted on some liberal conception of distributive justice (e.g., Rawls's own second principle) as part of an international theory? In his first presentation of the Law of Peoples, Rawls writes that nonliberal societies would reject any liberal distributive principles between societies because they reject liberalism. In his own words: Tor their part, the hierarchical societies reject all liberal principles of domestic justice. We cannot suppose that they will find such principles [e.g., the difference principle] acceptable in dealing with other peoples.'26 But this argument is too quickly made. It is not clear at all why the rejection of liberal principles has to be an all-or-nothing affair. Just because nonliberal societies reject (as they likely would) liberal principles pertaining to the full range of liberal civil and political rights, it does not follow at all that they win also reject liberal principles pertaining to economic and social rights. There is no reason why a society which does not accept as relevant, say, the ideals of free association and expression, cannot nonetheless endorse global principles that will distribute re-

24 The Law of Peoples includes a reprint of Rawls's most recent account on the idea of public reason, 'The Idea of Public Reason Revisited/ previously published in the University of Chicago Law Review 64 (1997). 25 See, e.g., Fernando Tesón, 'The Rawlsian Theory of International Law/ Ethics and International Affairs 9 (1995) 79-99; Stanley Hoffmann, 'Dreams of a Just World/ New York Review of Books (2 Nov. 1995) 52-7.1 discuss this also in 'Liberal Toleration in Rawls's Law of Peoples/ Ethics 108.2 (1998) 276-95; and also Toleration, Diversity, and Global Justice, Chaps. 2 and 4. 26 'Law of Peoples/ 75, my emphasis.

John Rawls Critical Notice of John Rawls The Law of Peoples 127

sources more equally between societies.27 Adopting principles of justice to regulate distribution between peoples, unlike the ideals of civil and political rights, need not have any direct liberalizing implication for the domestic politics of nonliberal peoples. To put it simply, nonliberal societies can accept, as principles governing the relations between societies, liberal principles of economic justice even as they reject liberal principles of political justice.28 Indeed given his own claim that 'a people sincerely affirming a nonliberal idea of justice may still reasonably think its society should be treated equally in a reasonably just Law of Peoples' (70), Rawls has to agree that it is entirely consistent for nonliberal peoples to endorse liberal egalitarian ideals to regulate relations between societies even as they reject the same ideal for their own domestic institutions. As a matter of fact, given that in the real world it is nonliberal societies that tend to be the less well-off ones, and hence would be the main beneficiaries of global redistribution, it seems all the more likely that nonliberal peoples would whole-heartedly embrace such a distributive ideal between societies.29 In short, the worry that imposing a liberal conception of distributive justice to regulate relations between societies would be an imposition on nonliberal societies — and hence a violation of the liberal principle of toleration — is unfounded both conceptually and empirically. But while the above argument will take us beyond the duty of humanitarianism to a duty of distributive justice between societies, it still does not go far enough for cosmopolitans. It only accounts for equality between societies, but remains indifferent to inequality within society. Hence it is possible within this conception of international justice that

27 This is the case in the real world: nonliberal developing countries want more financial assistance from the developed world (economic equality) while resisting pressures from the developed world that they liberalize their political institutions. A global distributive scheme on a Rawlsian concept would be readily adopted by these nonliberal countries. (If there should be difficulties here, it would be that Rawlsian distributive principles do not go far enough.) 28 These labels are Beitz's. See 'Social and Cosmopolitan Liberalism/ 515. 29 This observation was made by Thomas Pogge, 'An Egalitarian Law of the Peoples/ 218.1 do not imply here that nonliberal states may pick and choose aspects of liberal theory that they find useful and discard aspects that they find troublesome. My point here is that nonliberal states, in general, will not find liberal distributive principles (to regulate relations between countries) to be burdensome and unreasonable, as Rawls thinks. That nonliberal states are also expected to endorse basic civil and political liberties is a position I argued for in 'Liberal Toleration in Rawls's Law of Peoples/

559

560

John Rawls 128 Kok-ChorTan

resources and wealth equally distributed between societies are not in turn redistributed equally between persons within hierarchical societies. So a cosmopolitan justice requires more than distributive equality between societies; it also calls for distributive equality within societies. Adapting Rawls's difference principle, for instance, a cosmopolitan view would hold that our social institutions are to maximize the life prospects of the globally worst-off individuals no matter where they reside. This means that a cosmopolitan theory of global justice has to go beyond regulating the relationship between societies, and has to impose certain egalitarian demands within societies as well. Yet the Law of Peoples objects to this on the ground that it would undermine liberalism's principle of toleration by requiring that all societies conform to liberal egalitarian standards (by adopting, for instance, Rawls's second principle, or some other distributive principle premised on the equal moral worth of individuals). But can egalitarian liberals consistently accept nonegalitarian ways of ordering societies? As mentioned, Rawls argues that liberalism has to be accepting of well-ordered though nonliberal modes of ordering society. This, he says, is analogous to the political liberal ideal of tolerating nonliberal but reasonable philosophical, moral or religious comprehensive views within a democratic liberal society. The Law of Peoples, he says, extends this conception of toleration to the international plane. But it seems to me that the analogy between the domestic and the international spheres does not hold: while political liberalism tolerates nonliberal philosophical, moral and religious outlooks, it does not, and cannot, tolerate challenges to liberal political ideals. As Rawls himself points out, 'comprehensive doctrines that cannot support... a democratic society are not reasonable' (172-3; see also 178-9).30 That is to say, the scope of liberal toleration does not and cannot extend to alternatives to liberalism itself. A political philosophy, for reasons of consistency, must take a stance against competing political philosophies. Ronald Dworkin puts this point across neatly: any political theory must 'claim truth for itself, and therefore must claim the falsity of any theory that contradicts it. It must itself occupy ... all the logical space that its content requires.'31 To hold the counter-intuitive view that liberalism must also tolerate nonliberal

30 See also Rawls, Political Liberalism (New York: Columbia University Press 1992), 152-3. 31 Ronald Dworkin, A Matter of Principle (Cambridge, MA: Harvard University Press 1985), 361

John Rawls Critical Notice of John Rawls The Law of Peoples 129

politics calls to mind Robert Frost's well-known caricature of the liberal as a person who is unable to take sides in her own quarrel.32 If it is correct that the scope of liberal toleration does not extend to nonliberal ways of ordering politics, it is hard to see why it should do so when we move to the global context. To be consistent with its own fundamental commitments, a liberal Law of Peoples has to take a stance against nonliberal hierarchical societies. While hierarchical societies may find such critical judgments on their domestic institutions an imposition, this is not an unreasonable imposition from the liberal point of view. A Law of Peoples that claims to be 'an extension of a liberal conception of justice for a domestic regime to a Society of Peoples' (9, my emphasis) has to remain steadfast in its commitment to liberalism.

IV

Justice and Global Institutions

It might be said here that the absence of an enforceable international law is the central reason why Rawls thinks liberal principles cannot be extended globally. Naturally, the objection cannot mean merely that there is a current lack of the appropriate institutions, for this would be contradictory to the Rawlsian idea of justice, and hence an argument that Rawls himself would not accept.33 For Rawls, justice informs and constrains our institutions, not the other way around. For instance, concerning the case of outlaw societies, Rawls writes: 'The Society of Peoples needs to develop new institutions and practices under the Law of Peoples to constrain outlaw states when they appear' (48, my emphasis). To limit the Law of Peoples against existing institutional schemes is to render it 'political in the wrong way,' contradicting thus Rawls's expressed goal of achieving stability with respect to justice (44-5) J34 Given Rawls's stated goal and his idea of justice, the suggestion that it is because of the absence of institutional enforcement mechanisms that has compelled him not to

32 My argument here is indebted to David Dyzenhaus, 'Liberalism After the Fall/ Philosophy and Social Criticism 22.3 (1996) 9-37. For Frost's exact quotation, see The Macmillan Dictionary of Political Quotations (Toronto: Macmillan 1993), 372. 33 Thus Rawls writes in A Theory of Justice: 'From the standpoint of justice as fairness, a fundamental natural duty is the duty of justice. This duty requires us to support and to comply with just institutions that exist and apply to us. It also constrains us to further just arrangements not yet established, at least when this can be done without too much cost to ourselves' (115, my emphasis).

34 See also Rawls, Political Liberalism, 146-8.

561

John Rawls

562 130 Kok-ChorTan

extend liberal principles globally amounts, in fact, to an objection to, rather than a defense of, Rawls. The above objection, however, could be interpreted to mean that there is simply no feasible global institutions, short of enacting a world-state, that can support cosmopolitan distributive principles; and the untenability and undesirability of a world-state has been famously pointed out by Kant a long ago.36 If it is true that the principles of cosmopolitan justice can be realized only by certain forms of institutional arrangements that are either unattainable in the real world, or undesirable even if attainable (or both), then we will have to seriously re-evaluate the cosmopolitan idea. But this claim, that either we establish a world-state or we reject cosmopolitan distributive principles, presents a false dilemma. Various plausible institutional means of regulating global distribution have been proposed that do not invoke the idea of a world-government.37 At the very least, these alternative (non-statist) means of implementing and regulating global principles should be given due consideration, before we may rule them out as impracticable. Rawls himself believes that alternative means of enforcing the principles of his Law of Peoples are available. He thinks that global institutions other than a world-state, 'such as the United Nations ideally conceived' (36), can have the requisite authority and capacity to express and enforce these principles. More crucially, at no point in his rejection of global distributive justice does Rawls tie that idea to the concept of a world government. He rejects global distributive justice not because he thinks that this idea is inevitably dependent on some account of a world-state, but for the reasons considered earlier.38

35 For one such objection, see Hoffmann, 54. 36 Kant, 'Perpetual Peace/ in Kant's Political Writings, 2nd ed. Hans Reiss, trans, and ed. (Cambridge: Cambridge University Press 1991), 113. Rawls supports Kant's view on 35-6. 37 See, for example, Pogge's proposal for a Global Resource Tax (CRT) and his outline as to how this could be put into effect without a world-state. Pogge, 'An Egalitarian Law of Peoples/ Other long-term means of redistributing wealth globally that have been seriously advanced in global forums, again without the presupposition of a world-state, include the Tobin Tax (that will both tax and discourage short term international financial speculation) and the recently proposed 'bit tax' (to be imposed on internet transmission). 38 The cosmopolitan theorists with whom Rawls is dealing have themselves explicitly rejected the idea of world-state. See Pogge, 'Moral Progress/ in Steven Luper-Foy, Problems of International Justice (Boulder: Westview Press 1988). The problem of

John Rawls Critical Notice of John Rawls The Law of Peoples 131

In short, to put too much weight on the problem of enforceable international law obscures (by pushing a step-back) the real dispute between Rawls and the cosmopolitans. The fundamental reason why Rawls thinks the cosmopolitan idea to be unacceptable is the fact of reasonable pluralism.40 In Rawls's view, justice among peoples should differ from justice within a state, not ultimately because of the absence of an enforceable international law, as one might suggest, but because he thinks the scope of reasonable pluralism ought to be broadened in the global context. And I have tried to show above why this relaxing of the criteria of reasonableness is objectionable from a liberal point of view. V

Conclusion

Given existing institutional shortcomings, and entrenched habits and customs, it is true that liberals are sometimes limited in how they can go about supporting reforms globally and within hierarchical societies. And for the sake of peace, liberal societies may sometimes be forced to mute their global commitments. A liberal Law of Peoples may, therefore, be compelled to accommodate nonliberal states for practical reasons. What is important, though, is that it should regard this as an accommodation for the purpose of a modus vivendi, a nonideal scenario, rather than as a matter of toleration as required by ideal theory, as argued for by Rawls.41 And this is not merely an academic quibble: The aim of an ideal

world-state is thus not relevant to the dispute between Rawls and these cosmopolitans. It is important, in this regard, not to commit the mistake of associating cosmopolitanism with world-statism. I will not pursue this point here but see Beitz, Political Theory and International Relations, 182-3; 199-200. 39 This is not to say that the issue of international law is not an important one in its own right; on the contrary, a complete defense of cosmopolitan justice has to certainly confront this issue. My point here, to repeat, is that this is not central to the dispute between Rawls and the cosmopolitans. 40 To see more clearly why the question of enforceable international law is not the fundamental issue, we may ask this question: would Rawls propose a fully liberal theory of international justice were it possible for liberal states to effectively enforce liberal principles globally, everything else being equal? The answer would still be in the negative, given Rawls's (more fundamental) concern about reasonable pluralism. I thank Thomas Pogge for a helpful discussion on this point and the more general points in this section. 41 Hoffmann, 54. See also Thomas Nagel, 'Justice, Justice, Shalt Thou Pursue. A Review of Rawls's Collected Papers, The Law of Peoples, and A Theory of Justice, Revised Edition/ The New Republic (25 Oct. 1999), 41.

563

564

John Rawls 132 Kok-ChorTan

theory of justice is to provide us with a standard to aspire to. As Rawls himself has put it in A Theory of Justice, ideal theory 'presents a conception of a just society that we are to achieve if we can/ A Law of Peoples that regards hierarchical decent societies as societies in good-standing in the Society of Peoples, and that tolerates great inequality between states as part of its ideal theory, sets our sights too low. Achieving justice in our less than ideal world is, no doubt, an enormous challenge. But to set for ourselves a lesser goal because of this is to too easily give up our hope for a just world. Received: February, 2000 Revised: April, 2000

42 Theory of Justice, 246

KOK-CHOR TAN Queen's University Kingston, ON Canada K7L 3N6

[21] THE LAW OF PEOPLES, SOCIAL COOPERATION, HUMAN RIGHTS, AND DISTRIBUTIVE JUSTICE* BY SAMUEL FREEMAN I. INTRODUCTION My aim in this essay is to discuss, and defend against some frequent objections, John Rawls's rejection of a global principle of distributive justice. As is well known, Rawls's A Theory of Justice argues for a principle of distributive justice, the difference principle, that is to be applied within different societies but not among them.1 According to A Theory of Justice, each society has the duty to set up its economic and legal institutions in such a way that they make the least advantaged among its own members better off than the least advantaged would be if that society were structured according to any other distribution principle. But each society does not have a duty to structure its system so as to maximize the position of the least advantaged in the world at large. Though it is a universal principle that is to apply severally, or within every society, the difference principle is not global in reach, applying jointly to all societies simultaneously. To critics of many political persuasions, this seems a peculiar position. Why should principles of justice be domestically rather than globally applied? Rawls's position in A Theory of Justice becomes even more complicated in Political Liberalism and The Law of Peoples, where he is guided by questions of political legitimacy, and feasibility (or "stability") of liberal regimes.2 In Political Liberalism and later works,3 Rawls appears to give up on the idea that a well-ordered society of justice as fairness is feasible (such a well-ordçred society is one where every rational and reasonable citizen affirms, for moral reasons, justice as fairness, including the difference principle). The best we can expect of this world, he now seems to claim, * I am grateful to K. C. Tan for many helpful discussions and criticisms of this essay. I am also grateful to the other contributors to this volume for their comments, and to Ellen Paul for her many helpful suggestions in preparing the final version of this essay. 1 John Rawls, A Theory of Justice (Cambridge, MA: Harvard University Press, 1971; revised edition, 1999), cited as TJ in the text. References will be made to the 1999 revised edition. 2 John Rawls, Political Liberalism (New York: Columbia University Press, 1993; paperback edition, 1996); John Rawls, The Law of Peoples (Cambridge, MA: Harvard University Press, 1999), cited as PL and LP respectively in the text. 3 Here I have in mind particularly Rawls's last paper, "The Idea of Public Reason Revisited/' in John Rawls, Collected Papers, ed. Samuel Freeman (Cambridge, MA: Harvard University Press, 1999).

John Rawls

566 30

SAMUEL FREEMAN

are liberal and democratic societies in which all citizens recognize and accept the basic liberties and their priority, and the duty of society to provide a social minimum adequate to the exercise of the basic liberties. The social minimum need not be defined by the difference principle for liberal societies' economic systems to be politically legitimate, and their laws worthy of respect. It is this conception of the legitimacy of liberal regimes and of their economic distributions that underlies Rawls's account in The Law of Peoples of a "reasonably just Society of Peoples/' A just Society of Peoples is not (necessarily) a world in which all, or even any, of its membernations structure their economies according to the difference principle. So long as a society provides for the basic needs of its citizens, respects their human rights, and is regulated by a common-good conception of justice, it is "decent" and has political legitimacy within the Society of Peoples, and is to be tolerated and respected by other peoples. There is no specific principle of distributive justice that must be met, either domestically or globally, within a just Society of Peoples. The difference principle simply drops out of the picture in Rawls's account of the Law of Peoples. Oddly, perhaps, even within the Law of Nations that Rawls outlines in A Theory of Justice, Rawls held (or would have held) the same position. That is, even though, according to A Theory of Justice, each nation has a duty to realize the principles of justice in its basic institutions, the domestic justice of all member nations is not a condition of the justice of the well-ordered Society of Peoples. Moreover, Rawls still believed, when he wrote Political Liberalism and The Law of Peoples, that the difference principle defined the conditions of distributive justice for any and all societies; or at least it was for Rawls the ideal that societies should strive for. (This was part of his comprehensive liberal view.) However, being a just member of the Society of Peoples never meant for Rawls, even in A Theory of Justice, that a society must be fully just, especially not in its economic distributions of income and wealth among its members. For Rawls, it is simply not the role of peoples, individually or collectively, to enforce distributive justice anywhere except among their own peoples. There are then two points of contention. First, Rawls, early and late, rejects the idea of a global principle of distributive justice, a principle of justice that is global in reach. The reference point for assessment of judgments of distributive justice is the "basic structure" of particular societies. This means we cannot know whether a person has his or her fair share of resources without knowing (at least) which society he or she is a member of, and details about the basic institutions of that society. Second, Rawls holds that a society can be in good standing in a "reasonably just Society of Peoples" without complying with requirements of distributive justice, or for that matter, even respecting basic liberties, so long as it respects

John Rawls THE LAW OF PEOPLES

567 31

human rights.4 How can the Society of Peoples be just if all its members are not just? Many different kinds of objections have been raised against Rawls's position. Utilitarians (R. M. Hare, Peter Singer, et al.), libertarians (Robert Nozick et al.), and others who reject the entire Rawlsian framework (Joseph Raz, John Finnis, et al.) claim that Rawls starts out by focusing on the wrong values and ideas. Then there are other critics who are more sympathetic to Rawls, insofar as they accept many of his ideas and the principles of justice. Their primary objection, in effect, is that Rawls goes astray by limiting the reach of the difference principle to the basic structure of society, thereby making membership within a particular society a condition for assessing claims of distributive justice. These critics contend that whatever reasons there are for applying a principle of distributive justice internally within a society must also justify its application to individuals all over the world. I want to limit my discussion here mainly to objections by these "Rawlsian cosmopolitans," as I call them. I cannot here defend the entire Rawlsian framework against its external critics. I do hope to argue, however, that the position is not guilty of some blatant or even subtle inconsistency, as Rawlsian cosmopolitans seem to suggest. Finally, by way of introduction, many of the criticisms from all sides brought against Rawls assume current conditions, with all their injustices. For example, some critics claim that Rawls's Law of Peoples would allow for exploitation by private corporations of helpless people, or would permit conditions of peonage, apartheid, etc. Some also claim that it would allow for a government to impose egregiously unjust conditions on its people, such as ethnic cleansing, and anything short of slavery and forced servitude.5 The problem with these sorts of objections is that they fail to recognize that the Law of Peoples is formulated to apply to ideal conditions (of a sort), among "well-ordered societies" all of whom are members of a "Society of well-ordered Peoples" (PL, 17-19). Well-ordered societies (both liberal and "decent") are societies where reasonable and 4 For Rawls, the human rights all persons have under the Law of Peoples are a subset of the basic liberties all societies should provide. "Among the human rights are the right to life (to the means of subsistence and security); to liberty (to freedom from slavery, serfdom, and forced occupation, and to a sufficient measure of liberty of conscience to insure freedom of religion and thought); to property (personal property); and to formal equality as expressed by the rules of natural justice (that is, that similar cases be treated similarly" Rawls, The Law of Peoples, 65. 5 See Simon Caney, "Cosmopolitanism and the Law of Peoples/' The Journal of Political Philosophy 10, no. 1 (March 2002): 95-123, at 102, where he says: "Rawls's schema, thus, allows racial discrimination, the political exclusion of ethnic minorities, the forcible removal of members of some ethnic communities (that is, ethnic cleansing), the reduction of some to just above subsistence whilst other members of that society luxuriate in opulent splendor, and the perpetuation of grossly unequal opportunities and political power." None of these accusations is correct, for these practices are unjust, according to Rawls's principles of justice, wherever they occur. Moreover, some of them violate human rights, and all are difficult, if not impossible, to reconcile with any reasonable common-good conception, which is a condition of decency of peoples.

568

John Rawls 32

SAMUEL FREEMAN

rational members generally accept the governing principles of justice and terms of cooperation of their society, and rely upon these principles and terms in their public reasoning about justice. Moreover, these terms of cooperation comply with a conception of the common good according to which all members of society are benefited, and the members of society generally endorse this conception of the common good. (In a liberal society, the conception of the common good is largely defined by a liberal conception of justice; in "decent hierarchical societies" [PL, 71], it is defined by some nonliberal moral conception.) These conditions virtually guarantee that apartheid, ethnic cleansing, and other egregious forms of discrimination will not be practiced in well-ordered decent societies. If Rawls's account of human rights does not prohibit such unjust practices (though I believe it does), then the implications of any common-good conception of justice conjoined with the general acceptability of terms of cooperation within decent societies should prohibit them. (I assume here that those who suffer from apartheid and ethnic cleansing would not endorse the treatment they endure. Even if they did, however, it is impossible to conceive of a feasible conception of the common good that incorporates terms of apartheid and ethnic cleansing directed toward those who are supposed to benefit from these very same terms.) One question I consider is whether these conditions and restrictions will be adequate to prevent the economic exploitation of poorer peoples by richer peoples or multinational corporations. I contend that just as liberal societies will not permit economic exploitation of their own citizens since it presumably makes them worse off than many alternative terms of cooperation, so a decent society that domestically enforces a common-good conception of justice will not allow foreign or multinational corporations to take advantage of its members in exploitative ways. Moreover, there is plenty of room within Rawls's Law of Peoples for the Society of Peoples to limit exploitative economic dealings between private corporations and the political representatives of poorer peoples. Just because Rawls does not provide a principle of global distributive justice does not mean that unmitigated laissez-faire is the general rule of economic interaction within the Society of Peoples. The Law of Peoples, then, is designed to apply in the first instance to hypothetical conditions, among well-ordered liberal and decent societies each of which has concern for the well-being of its own people and seeks their common good, and respects others as free and equal peoples. Rawls did not envision the Law of Peoples as the sole element of the terms of cooperation that apply among peoples "in our world as it is with its extreme injustices, crippling poverty, and inequalities/76 Under current conditions, we are in the realm of nonideal theory and partial compliance. Just as Rawls regarded preferential treatment as permissible as a means of 6

Rawls, The Law of Peoples, 117.

John Rawls THE LAW OF PEOPLES

569 33

transition to a just liberal society, but as inappropriate under conditions of a well-ordered liberal society that is just, so too nations in the contemporary world are in less than ideal conditions and may require special remedies as a means to establishing a well-ordered Society of Peoples. Similar considerations apply to Allen Buchanan's claim that Rawls's Law of Peoples suffers from a failure to address questions of secession. The Law of Peoples is not a general theory of global justice that is designed to address all the problems that arise in the contemporary world. Rather, it is set forth as part of political liberalism, to provide the principles of foreign policy for a well-ordered liberal society (LP, 9-10). Questions of secession simply do not arise within this setting; they are, again, problems that arise within nonideal theory. It is no more a problem with Rawls's Law of Peoples that it fails to address issues of secession, than is its failure to address many other problems that arise within nonideal theory (for example, the problem of resolving boundary disputes, or a formula for deciding war reparations). Of course, it would have been wonderful had Rawls been able to tell us what he thought about rights of secession before he died, just as it would have been wonderful had he addressed many other political and moral issues. But Rawls's Law of Peoples cannot be dismissed or criticized for failing to address problems not within its intended purview. II. SOCIAL COOPERATION AND SOCIAL JUSTICE The use of the term "distributive justice" to connote standards for assessing the distribution of income and wealth is relatively recent; it seems to have evolved out of the socialist critique of capitalism.7 In Anarchy, State, and Utopia, Robert Nozick says that the term "distributive justice" is not neutral, since it suggests a central distribution mechanism that doles out a supply of resources to people and redistributes resources among people when their individual choices fail to match some principle.8 Still, Nozick must have found the idea useful, since he entitles the central chapter in his book, covering eighty-three pages, "Distributive Justice." I will use the term "distributive justice" as neutrally as I can: namely, to designate moral standards for assessing ongoing methods of distribution of rights to income and wealth that are implicit in any economic system. It is the sense in which I believe that both Nozick and Rawls used the term. A principle or conception of distributive justice applies to all existing income and wealth within an ongoing system of production, exchange, 7 The idea of "fair distribution" was current among nineteenth-century non-Marxian socialists. Marx himself disdained the French socialists' idea of fair distribution, along with the idea of "equal right/' calling them "obsolete verbal rubbish." See Karl Marx, "Critique of the Gotha Program," in Karl Marx: Selected Writings, ed. David McLellan (Oxford: Oxford University Press, 1977), 569. Aristotle distinguished distributive justice from commutative justice, but the term was not used in the contemporary sense. For a discussion, see Samuel Fleishacker, Distributive Justice (Cambridge, MA: Harvard University Press, 2005). 8 Robert Nozick, Anarchy, State, and Utopia (New York: Basic Books, 1974), 149.

John Rawls

570 34

SAMUEL FREEMAN

and consumption. It specifies standards for deciding how and whether income and wealth are justly distributed among individuals—who should have rights to what and in exchange for what (if any) contributions. The difference principle, the principle of utility, and the equality principle are all distributive principles insofar as they cover all the economic resources that exist in any economic system. So too are Nozick's libertarian entitlement principles, according to which (very roughly) rights in things are fairly distributed only if they are transferred by market exchanges, gifts, bequests, gambling, or some other voluntary mode of transfer by the person who holds those rights. Such common-sense precepts as "To each according to effort," "To each according to contribution," and "To each according to need," when left unsupplemented, may not fit this definition of distributive principles. (For example, what if all needs are satisfied, or all efforts and contributions are rewarded, and there is still a remaining surplus?) Still, I presume these precepts can be made to fit the definition easily enough so long as we supplement each precept with some other (for example, the surplus remaining after all needs, efforts, or contributions are met might then be equally distributed, or distributed to maximize utility, etc.). On this understanding, a people's duty to assist its own and other peoples so that their basic needs are met is not (by itself) a principle of distributive justice. For this duty, once satisfied, extends no further and establishes no further claim within the ongoing system of economic production and exchange. The same is true of duties to meet special needs of the handicapped; by themselves, these duties are not principles of distributive justice. In contrast, a principle of restricted utility which says "A society has a duty to provide a social minimum for all its citizens sufficient to meet their basic needs, and once basic needs are met, the economy is to be designed to distribute income and wealth so as to maximize utility" (or alternatively "to satisfy the principle of efficiency") is a distributive principle; for it applies to and enables an assessment of the justice of the distribution of all wealth within the economic system. Rather than providing a specific global distribution principle, Rawls provides an account of human rights, coupled with the claim that the basic needs of all individuals in the world are to be met, partly as a matter of their human rights. This provides the basis for a duty of peoples to assist "burdened peoples" who are unable to meet the basic needs of all their members. The "target" of the duty of assistance is the capacity of a people to be economically independent so that they may at least meet all citizens' basic needs, and become bona fide members of the Society of Peoples.9 At a minimum, a person's basic needs are those that need to be 9 Principle number 8 of the Law of Peoples says: "Peoples have a duty to assist other peoples living under unfavorable conditions that prevent their having a just or decent political and social regime" (LP, 37). As a gloss upon this principle, Rawls says: "Certain

John Rawls THE LAW OF PEOPLES

571 35

met to enable him or her to effectively exercise human rights. The human rights that Rawls mentions (initially) are: "the right to life (to the means of subsistence and security); to liberty (to freedom from slavery, serfdom, and forced occupation, and to a sufficient measure of liberty of conscience to ensure freedom of religion and thought); to property (personal property); and to formal equality as expressed by the rules of natural justice (that is, that similar cases be treated similarly)" (LP, 65).10 Here it is important to note that by "basic needs" Rawls is not just talking about subsistence needs that are protected by human rights—or what is needed so that people do not starve or perish from disease. In addition to human rights, Rawls intends that people should be able to "take advantage of the rights, liberties, and opportunities of their society," which would require institutional rights and liberties, and economic means, that go beyond what is needed to exercise one's human rights. Given the role of Rawls's idea of human rights in determining the extent of the duty of assistance, it is important to emphasize the basis for Rawls's account of human rights. Rawls is often criticized for not including certain liberal and democratic rights among the human rights: primarily equal rights of political participation, freedom of speech and expression, and freedom of association, all of which are among the basic liberties that are part of Rawls's first principle of justice.11 Many critics believe that Rawls's list of human rights, in addition to being truncated, is arbitrarily drawn, with no solid basis. Here one can add to this list of complaints that, in the case of the basic liberties, Rawls appealed to a conception of persons as free and equal democratic citizens with the two moral powers.12 The basic liberties were regarded as fundamentally necessary to the exercise and development of the moral powers, and to enable democratic citizens to freely pursue a wide variety of permissible plans of life.13 In the case of human rights, however, Rawls's list seems to

provisions will be included for mutual assistance among peoples in times of famine and drought, and, insofar as it is possible, provisions for ensuring that in all reasonable liberal (and decent) societies people's basic needs are met. [Note:] By basic needs I mean roughly those that must be met if citizens are to be in a position to take advantage of the rights, liberties, and opportunities of their society. These needs include economic means as well as institutional rights and freedoms" (LP, 38). 10 Evidently, Rawls does not mean this list to be exclusive, for he says, "Among the human rights are .. ."—after which follows the quotation in the text above. 11 See Charles Beitz, "Rawls's Law of Peoples," Ethics 110 (July 2000): 669-96, at 683-86; and Allen Buchanan, "Rawls's Law of Peoples: Rules for a Vanished Westphalian World," Ethics 110 (July 2000): 697-721, at 718-19 (on Rawls's "rather lean set of individual rights"). 12 The two moral powers are a capacity for a sense of justice (to understand, apply, and act from principles of justice) and a capacity for a rational conception of the good (to form, revise, and pursue a rational conception of the good). See Rawls, Political Liberalism, 19, 81, 103-4. Rawls calls these powers, respectively, the capacities to be reasonable and to be rational. They are, in effect, the capacities for practical reasoning as applied to matters of justice, which, Rawls believes, are necessary for social cooperation. 13 See Rawls, Political Liberalism, lecture 8, "The Basic Liberties and Their Priority."

John Rawls

572 36

SAMUEL FREEMAN

be without any such foundation. Isn't Rawls's account of human rights then simply an appeal to unfounded intuitions? Rawls apparently does not provide a conception of the person to ground his account of human rights, and more generally the Law of Peoples, because there is no shared conception of the person that will provide a basis for the public reason of the Society of Peoples. Decent societies, in particular, rely upon comprehensive religious and philosophical doctrines that will conflict with almost any conception of the person that might be acceptable to liberal peoples and other decent peoples. This does not mean, however, that Rawls's list of human rights is conjured out of thin air. It is a list of rights that, he says, is not distinctly liberal, and that all liberal and decent peoples can agree to on the basis of their liberal and decent comprehensive views. Moreover, the list does not depend upon a particular religious or philosophical doctrine, not even (Rawls contends) upon liberalism. Rather, it has a substantial foundation that is part of the Law of Peoples itself. What is the basis, then, for Rawls's account?14 Oddly, critics of Rawls rarely, if ever, discuss the substantial basis Rawls provides for his account of human rights: "What have come to be called human rights are recognized as necessary conditions of any system of social cooperation. When they are regularly violated, we have command by force, a slave system, and no cooperation of any kind" (LP, 68). Rawls has nothing more to say about social cooperation in The Law of Peoples, but in Political Liberalism (PL, 16) he has this to say: First, social cooperation is by its nature voluntary, and involves an absence of forced servitude and other conditions that would prevent us from holding people responsible for their conduct. Second, social cooperation is to be distinguished from efficiently coordinated behavior, where people are working in a group, but their behavior is regulated simply for the sake of effectively achieving purposes that none of them may endorse. Prisoners in a work gang are not engaged in social cooperation, though they might work quite efficiently as a group (repairing roads, picking up trash, etc.). By contrast with efficiently coordinated behavior, the idea of social cooperation for Rawls assumes that each person has an idea of his or her good, and is benefited in some way that he or she would acknowledge by engaging in cooperation with others. Presumably, if a person's good were 14 Thomas Pogge conjectures that Rawls's account of human rights is based in a concern to accommodate nonliberal peoples' rejection of liberal rights, in the hope that they can at least accept a smaller list of human rights. See Thomas Pogge, "An Egalitarian Law of Peoples/' Philosophy and Public Affairs 23, no. 3 (Summer 1994): 195-224. This implies that the Law of Peoples is a modus vivendi between liberal and decent peoples, a claim that Rawls denies. Pogge's interpretation discounts the centrality of social and political cooperation to Rawls's account of human rights, and also to his account of distributive justice. Pogge's interpretation also ignores the fact that human rights are agreed to among liberal peoples themselves, as being among the conditions whose violation is necessary to justify intervention in the affairs of other liberal peoples. If respect for human rights is adequate for the ideal case of a Society of Liberal Peoples, then how can it be a compromise designed to accommodate decent peoples?

John Rawls THE LAW OF PEOPLES

573 37

not in any way furthered by interaction, then his or her actions would not be rational or even voluntary. Finally, for Rawls social cooperation also involves an idea of reciprocity and fair terms of cooperation, which provide a sense of what is "reasonable." These are norms that members of the cooperating group rely upon and use to guide their conduct and regulate the distribution of benefits and burdens among themselves. Moreover, the members mutually recognize these fair terms and refer to them not only to regulate but to criticize and assess one another's conduct. These terms are, in this regard, public standards, not just strategic norms. For Rawls, then, social cooperation incorporates a distinctly moral component—a notion of fair terms of cooperation, understood as "reciprocity," which provide standards of reasonableness. Rawls distinguishes reciprocity from mutual advantage, which can be explicated entirely in terms of a person's good and what is rational for her to do (the third aspect of social cooperation mentioned above). People who are engaged in social cooperation normally are not focused exclusively on their own good, in the sense that they are ready to take advantage of others and free-ride whenever circumstances permit. This does not mean that people have to be altruistic to cooperate with others; but they do normally have a sense of fairness or justice—a settled disposition to comply with terms of cooperation, and do their part, even on occasions when it is not to their benefit (so long as others manifest a like disposition). For Rawls, one thing that distinguishes "decent peoples" from "states" (as traditionally understood) is that they have moral motives and a sense of justice as a people, that enable them willingly to comply with their duties under the Law of Peoples, and not take advantage of weaker peoples whenever it seems favorable to do so.15 It is this sense of social cooperation that Rawls seems to rely upon in drawing up his list of human rights. The right to life, freedom from involuntary servitude, the right to hold and use at least some personal property, and other human rights that Rawls mentions are minimal reasonable terms necessary to social cooperation. The right to vote and the right to run for office, however central to democratic societies, are not necessary for social cooperation as such; other methods of decision making are compatible with social cooperation. Historically, most people in most societies have not enjoyed democratic rights, and even in societies where they do, these rights often willingly go unexercised. This option is not true of human rights generally. To contend that democratic rights of political participation are on a par with, and just as important as, the right to life, freedom from involuntary servitude, the right to hold personal property, and other human rights that Rawls mentions, is implausible and unreasonable. And while it may not be as unreasonable to say that liberal 15 Cf. Rawls, The Law of Peoples, 17,28-29: "A difference between liberal peoples and states is that just liberal peoples limit their basic interests as required by the reasonable" (ibid., 29).

John Rawls

574 38

SAMUEL FREEMAN

freedom of association and freedom of speech are equally fundamental to social cooperation, it is still far-fetched and unconvincing. Some degree of freedom of speech and of association surely must be a human right, and can be included under what Rawls calls the "right to liberty/' For example, fundamentalist Muslim societies that punish women for just talking with men who are not family members surely violate human rights. But Rawls denies that among the human rights must be included liberal freedoms of speech and association, with all that they include (for example, the right to defile or destroy national or sacred symbols, or enjoy pornography, or freedom of same-sex relations). To hold otherwise is not to take the idea of human rights seriously. The idea of social cooperation also is central to Rawls's account of social justice. It underlies his distinction between "domestic justice" and the Law of Peoples. Moreover, the idea of social cooperation informs Rawls's account of the difference principle. What makes social cooperation possible for Rawls are the basic institutions that constitute "the basic structure of society"16 Here it is crucial that, for Rawls, political cooperation under the terms of a political constitution, including the legal system that it regulates, is a central aspect of the basic structure of society. For Rawls, political cooperation is part of social cooperation; it makes social cooperation according to the terms of other institutions (particularly economic institutions) possible and is necessary to those institutions. (One important example: The institution of property is presupposed by economic cooperation; property is largely a legal institution and cannot exist—except perhaps in primitive form—in the absence of political cooperation according to the terms of a political constitution and a legal system.) When Rawls says that the political constitution is part of the basic structure, he does not just mean the procedures that specify how laws are enacted and that define offices and positions of political authority. He means more or less the entire legal system, including most public and private law, that is the product of the constitution in this procedural sense. Modern legal systems, such as the federal system in the United States, are made up of countless acts of legislation, administration, judicial precedent, and other legal rulings that are issued by the multiple legal bodies with lawmaking authority. An economic system that is regulated by the legal norms that are issued by the political constitution is also part of the basic structure. Here, of course, the legal norms of property, contract, commercial law, intangibles, and so on that are essential for eco16 The primary basic institutions that constitute the basic structure of society are the following: the political constitution; the legal system of trials and other legal procedures it supports; the institution of property; markets and the myriad laws and conventions making economic production, exchange, and consumption possible; and the institution of the family, which enables a society to raise and educate children and reproduce itself as an ongoing system over time. See Rawls, A Theory of Justice, 6-7; and Rawls, Justice as fairness: A Restatement, ed. Erin Kelly (Cambridge, MA: Harvard University Press, 2001), sections 4,15, and 16.

John Rawls THE LAW OF PEOPLES

575 39

nomic production and exchange are to be included in the basic structure. What makes possible the incredibly complicated system of legal norms that underlie economic production, exchange, and consumption is a unified political system that specifies these norms and revises them to meet changing conditions. Nothing comparable to the basic structure of society exists on the global level. Moreover, if Rawls is correct—if a stable "world state" that assumes the primary functions of governments is not feasible—then nothing comparable to the basic structure of society can ever stably endure on a global level (see LP, 36, 48). This means that social cooperation is and must remain distinct from the kinds of relations that hold between different societies individuated by their own separate political systems. This does not mean that different societies do not cooperate; of course they do. But they do not engage in social cooperation in Rawls's sense of cooperation framed within the basic structure of society. Cooperation among peoples (to use Rawls's terms) is a qualitatively different kind of cooperation from social cooperation, and it has its own distinctive fair terms of cooperation. These terms ideally are the Law of Peoples, the terms of cooperation for the Society of Peoples. Some of Rawls's critics confidently claim that "there is a global basic structure,"17 and argue that for this reason there must be principles of global distributive justice. This simply begs the question, however. Rawls does not need to deny a "global basic structure" in some sense, but clearly he would contend that it is very different from the basic structure of society. For Rawls, the global basic structure would just be the set of institutions that are needed to give effect to the Law of Peoples. Rather than "global basic structure" Rawls refers to "the basic structure of the Society of Peoples" (LP, 61). Whatever we call it, the important point is that these global institutions are very different—qualitatively different—from the basic structure of a society that makes social cooperation possible. It is only because there are societies, with their distinct basic structures of political, social, and economic institutions, that we can take seriously and regard as feasible any kind of "global basic structure." Global cooperation and global institutions are supervenient upon social cooperation and basic social institutions. Why is the basic structure of society qualitatively different from the basic structure of the Society of Peoples, and from any other realistic and stable "global basic structure"? The Society of Peoples is not a political society, and thus has no original political jurisdiction or effective basic political power. (Here I use the terms "political society" and "political power" in John Locke's sense to include the idea of political authority, or having the right to rule.) The effective political power and jurisdiction that global 17 Allen Buchanan asserts this and sees the existence of a global basic structure as sufficient grounds for a global distribution principle. See Buchanan, "Rawls's Law of Peoples/' 705. As I argue in the text, this conclusion does not follow.

John Rawls

576 40

SAMUEL FREEMAN

institutions exercise are possessed by these institutions only to the degree that they have been granted such power and jurisdiction by independent peoples. Global political authority (such as it is) then exists only as a result of the legal acts of independent peoples—"legal" insofar as these acts are authorized by their own constitutions. So long as independent peoples can withdraw or revise the grant of political power transferred to global institutions (granted there may be significant costs in doing so), global political power remains supervenient upon the political power of independent peoples. This is what it means to say that global political authority and jurisdiction is not "basic" or "original." Basic political authority resides only within the basic structure of societies. This is not a necessary truth, but it is a significant empirical truth, assuming that a world government would not be capable of any enduring stability (LP, 36). Basic political authority resembles the idea of political sovereignty, but it does not carry the connotations of absolute political power that reside with the idea of sovereignty. Of course, no political power is absolute de jure, not even the power of peoples. The legitimacy of all political power (not to mention its justice) is subject to respect for human rights and the other conditions that Rawls imposes upon a decent (hierarchical) society. Thus, social cooperation, in Rawls's sense, is not the only kind of cooperation. There is cooperation among peoples, and there are also different kinds of cooperation that exist within families, universities, churches, and other associations and groups. Naturally, all cooperation is social in a sense, but cooperation among members of a society—that is, among people who share the same basic structure—is a distinct kind that Rawls calls "social cooperation." Then, to avoid confusion, he refers to the cooperation within groups in society as "associational cooperation." (Here we might provide each of these many different forms of cooperation with a distinct name, to distinguish them all from social cooperation—familial cooperation, religious cooperation, etc.) For each of these different kinds of cooperation, there are rules that specify the fair terms of cooperation among members of that cooperative institution. These are the rules of "local justice" for lesser associations that exist within the basic structure of society. To be just or fair, however, these rules need not all be the same. (For example, children should not have equal say with their parents; and employees need not have equal sway along with managers in order for the work rules internally regulating firms to be fair.) Rawls says, "Justice as fairness starts with domestic justice—the justice of the basic structure. From there it works outward to the Law of Peoples, and inward to local justice."18 What is distinctive about the basic structure of society is not simply that it exercises profound effects on individuals' aims, characters, and future prospects. The same can be said, of course, about the institution of the family, or the religious institutions that 18

Rawls, Justice as Fairness: A Restatement, 11.

John Rawls THE LAW OF PEOPLES

577 41

provide moral structure for many people's lives. It is the purportedly profound effects that global institutions and economic relations exercise on people's lives that advocates of a global difference principle or other global distribution principle point to when they contend that there is nothing distinctive about the basic structure of a society. Rawls need not deny the significant (though not equally profound) effects of global cooperation to claim that domestic justice, the justice of the basic structure, concerns a different kind of institutional cooperation than does global justice, and that it therefore warrants its own principles of justice. Nor does Rawls even need to deny that there are global principles of justice that regulate global cooperation in order to claim that domestic justice, the justice of the basic structure, has a kind of priority over other forms of justice. But it is part of Rawls's "political constructivism" that terms of social cooperation must be worked out first, independently of the principles of justice of other institutions, whereas the principles of global and local associational justice presuppose the principles of justice for the basic structure. Principles of social (or domestic) justice constrain or limit, but do not entirely determine, principles of local justice. That is, we cannot fully specify the rights and obligations family members have to one another until we first work out the principles of domestic justice. (For parents must respect the rights of their children, spouses have certain duties of support owed not only to their children but to each other, etc.) We can, however, specify the rights, duties, and obligations that citizens of a democratic society owe to one another without first working out the terms of the Law of Peoples. What we cannot do is decide the duties that different peoples owe to one another as peoples, and how they are to act toward one another as peoples in their political relations, before the principles of domestic justice are decided. This is part of what Rawls means by "political constructivism": the principles that appropriately regulate social and political relations depend upon the kinds of institutions or practices involved, and are "constructed" on the basis of ideas that are central to the functioning of those institutions or practices and people's awareness of them. Political constructivism is, I believe, integral to Rawls's rejection of cosmopolitanism and a global principle of distributive justice. Cosmopolitanism, as a view about distributive justice, claims that "the content of social justice cannot be arrived at by considering the individual society as a closed system in isolation from all others."19 As such, cosmopolitanism is or at least involves an epistemological/methodological claim that denies the possibility of political constructivism. Now what ultimately underlies Rawls's political constructivism are his views regarding moral justification. The justification of principles of justice ultimately involves bringing into a "wide reflective equilibrium" the considered convictions of justice 19 See Samuel Scheffler, "Conceptions of Cosmopolitanism/' in his Boundaries and Allegiances (Oxford: Oxford University Press, 2001), 116.

John Rawls

575 42

SAMUEL FREEMAN

that we share regarding social institutions or practices. Practices and institutions each have their own rules that are constitutive of the practice or institution. So long as we do not question the existence of the institution itself, it is the role of a conception of justice to provide principles for the regulation of these rules constituting practices or institutions. Of course, it may be that as a matter of justice itself (the justice of society, or global justice), certain practices or institutions should not exist or should be radically revised. (Many people have raised questions regarding the institutions of marriage and the family, for example, arguing that for reasons of social justice they should be radically revised, and in the case of civil marriage, even eliminated.) Anarchists, of course, argue that the state, and political society among people, is inherently unjust and should not exist. Some cosmopolitans contend that the state or political society as traditionally conceived (as an independent and autonomous entity, with exclusive control over a territory) should not exist, and that justice requires a world government of some kind. In response, Rawls's account of the Law of Peoples is based upon certain empirical assumptions and theoretical commitments that, taken together, require the existence of separate societies, each regulated by its own basic structure. Rawls assumes that justice and social cooperation too (at least under modern conditions) are not possible without governments and complicated legal systems, and that what social justice involves, in large part, are principles for structuring and defining the powers of political institutions. He also assumes that a politically autonomous worldstate is Utopian (i.e., a state that is capable of serving the functions that different peoples with their governments controlling their own territories now perform). For Rawls, this means that attempts to provide principles of justice on the assumption of conditions that could obtain only within (or that optimally obtain only within) a world-state are misguided, since in an important sense these principles are not compatible with human nature, the human good, or the possibilities of stable human society. What does this mean? On the face of it, the practical impossibility of a world government might not seem to pose any problem in formulating global principles of justice. For example, a utilitarian would say that a just global distribution is one that maximizes the sum of global happiness. The utilitarian then might contend: "The fact that a world government is not feasible is not relevant to the justification of the principle of utility as a standard for global distributive justice (and everything else). It is, rather, a pragmatic consideration relevant to the application of the principle of utility to decide how institutions should be feasibly designed to achieve global justice. But the fact that one or another institution is not feasible has no bearing on standards for distributive justice, and whether they are to apply locally or are global in reach. Philosophical conceptions of justice should not be made hostage to contingencies, but are a priori, the product of (if you will) 'pure reason/ "

John Rawls THE LAW OF PEOPLES

579 43

The role of empirical considerations and natural regularities in the justification of a conception of justice is a far more complicated question than can be addressed here.20 This much should be said: For Rawls, our considered convictions of justice arise within the practices and institutions we live with, and are attuned to the structure and demands of those institutions. Our considered judgments regarding individual liberties and just distributions originate within the framework provided by the basic structure of society. Moreover, they primarily apply to those institutions (for example, what constitutional rights should people have? how should the system of property and taxation be structured? and so on). Our considered judgments regarding global distributive justice are more tentative and much less secure, since there are few global institutions that give rise to them or anchor them. For many reasonable persons, it is hard to know where to even begin in considering these issues. Among our more abstract considered convictions of justice that a conception of justice must accommodate is, Rawls assumes, that principles of justice that regulate social and political relations should be publicly knowable and reasonably acceptable to people whose lives and relations these principles regulate. This is part of what it is to be a free person for Rawls—to know, understand, and be able to reason about the bases of social and political relations, and not be under any illusions about them. For Rawls, this suggests that principles of justice should be capable of serving as principles of practical reasoning and justification among people who conceive of themselves as free and responsible agents. It is also a requirement of democracy for Rawls that citizens be in a position to know the bases of their political relations; hence, a publicity condition is incorporated into political liberalism. To serve this public role, however, principles of justice have to be generally acceptable to reasonable people as a public basis for deliberation and justification. This is necessary to citizens' and a democratic people's political autonomy. Moreover, the institutions that these principles presuppose and support, if they are to be publicly acceptable among free persons, must be feasible and capable of enduring over time; and to meet these requirements, they must gain citizens' willing support. As Rawls says, institutions should be "stable for the right reasons," that is, acceptable to citizens on the basis of their sense of justice. Now assuming that what we are looking for is a conception of justice that (1) fits with our considered convictions of justice, as reasonable persons, (2) is publicly acceptable to free and equal persons who are also reasonable, and can serve them as a basis for public justification, and (3) is feasible and will endure over time and across generations while serving 20 G. A. Cohen challenges Rawls's reliance on facts in justifying the principles of justice. See G. A. Cohen, "Facts and Principles/7 Philosophy and Public Affairs 31, no. 3 (Summer 2003): 211-45.

580

John Rawls 44

SAMUEL FREEMAN

this public role, then it should follow that (4) this public conception must be one that takes into account and is responsive to the permanent facts about human beings and their living together in social groups, including how they conceive of their good. To put the point another way: Assuming that we are concerned with achieving the freedom and equality of real persons in the world, given their nature and limitations as human beings, a conception of justice must be responsive to facts about human nature if it is to be stable by engaging people's sense of justice. Consider now the fact that human beings are sociable creatures who develop within and are profoundly affected by the basic institutions of their society. Consider also that a world government and basic structure that includes all the world within one society, and is capable of enduring and remaining stable on reasonable terms over time, is not empirically feasible. For a variety of reasons, the existence of a number of different societies, each with their own political institutions, is a permanent fact about social life. If this is true—if there is no escaping the fact of independent societies each with their own basic structure—and if what we seek is a publicly acceptable conception of justice, then there is no escaping the need for an independent conception of social justice that applies domestically to regulate the basic structure of society. This is required, not simply because of considerations of stability of social groups, but because a conception of justice, for Rawls, has the role of providing a basis for public justification among people who regard themselves as free and as equals. (From the fact that a world-state is not feasible, it should also follow that there is no global conception of distributive justice that is acceptable as a basis for public justification among free and equal persons that is feasible and can remain stable across generations. But this separate issue will be addressed in the next section.) Assuming all this is true (of course, many will contest it), the following problem arises: Once the conception of justice for the basic structure of society is in place, there is a further need for an account of how different societies are to interrelate with one another. An account of international justice, or "the Law of Peoples," is in this regard an extension (not a precondition) of an account of social justice, and presupposes it. This claim was already implicit in A Theory of Justice, where Rawls remarked on the need "to extend the theory of justice to the law of nations" in order to guide the foreign policy of a nation regulated by the principles of justice (TJ, 331-33). This claim is only slightly modified in The Law of Peoples, where Rawls says: I emphasize that, in developing the Law of Peoples within a liberal conception of justice, we work out the ideals and principles of the foreign policy of a reasonably just liberal people. This concern with the foreign policy of a liberal people is implicit throughout. (LP, 9-10, emphasis in original; cf. 83)

John Rawls THE LAW OF PEOPLES

581 45

This is what Rawls means when he says the Law of Peoples arises "within political liberalism" (LP, 9). Since the Law of Peoples basically concerns the foreign policy of liberal peoples, questions of global justice are already confined to a narrow range of issues, relatively speaking, within the Law of Peoples. Many questions that occupy advocates of cosmopolitan justice do not even arise, having been preempted, in effect, by the problems that the Law of Peoples is designed to address. In the following section, I indicate in more detail how the question of global distributive justice is among these preempted issues. III. COSMOPOLITAN OBJECTIONS TO THE LAW OF PEOPLES I turn now to more specific objections to Rawls's account in The Law of Peoples, especially those relevant to distributive justice. Rawlsian cosmopolitans (as I call them) criticize Rawls for failing to take into account a number of considerations in designing the Law of Peoples. Three of the main criticisms they make are the following: 1. With regard to Rawls's account of human rights: The parties to the original position among liberal peoples, where parties are representatives of (liberal) peoples, should carefully consider the list of human rights, and when they do, they will agree that the liberal rights that are a part of Rawls's first principle of justice, and are generally endorsed in all liberal societies, should be regarded as human rights enforceable over all the world by the Law of Peoples.21 After all, what's good for the (liberal) goose should be good for the (nonliberal) gander. If so, then it is not at all clear that liberal societies will agree to tolerate decent (but nonliberal) societies in the way Rawls suggests. 2. Regarding natural resources: The representatives to the original position for liberal and decent peoples naturally should be concerned about the level of resources that their people respectively control. Not knowing this level, since they are behind the veil of ignorance, they should all insist that resource-poor nations be provided with resources from resource-rich nations. After all, as Rawls says in relation to domestic justice, no one deserves the resources he is born with, so no one deserves to be born rich or poor. By the same token, resource-rich nations do not deserve the natural resources that happen to be deposited by nature on their territory. These resources should be redistributed to resource-poor nations (or at least subject to a resource tax upon extraction) until those nations receive their fair share, and the rational representatives of liberal and decent peoples will insist on as much in the interest of those they represent.22 21 See, for example, Pogge, "An Egalitarian Law of Peoples/' 214-16; K. C. Tan, Toleration, Diversity, and Global Justice (State College, PA: Perm State Press, 2000), 28, 79-80. 22 See Pogge, "An Egalitarian Law of Peoples/7 199ff.; and Charles Beitz, Political Theory and International Relations (Princeton, NJ: Princeton University Press, 1979; 2d edition, 1999), 138, 141.

John Rawls

552 46

SAMUEL FREEMAN

3. Regarding distributive justice: Since representatives of peoples behind the veil of ignorance are ignorant not only of natural resources, but also of the level of talent, knowledge, technology, capital, and culture their people enjoy, they should want to protect themselves not just against resource poverty, but also against occupying globally less-advantaged positions. As a result, and for the same reasons as the parties in the domestic original position, they should choose a principle of global distributive justice. Here, some critics argue that for the same reasons that parties agree to it in the domestic original position, the global principle of distributive justice should be the difference principle. (Charles Beitz, Thomas Pogge, Brian Barry, David Richards, K. C. Tan, and others make this argument.) Others contend that Rawls's argument for the difference principle is mistaken and the global distribution principle should be another principle of justice. (Allen Buchanan makes this argument.) A third group contends that the difference principle should domestically apply within nations, subject to the requirements of a different global distribution principle which determines the share that each nation receives and to which they are to apply the domestic difference principle. (Pogge and Tan have suggested this position in conversation.) One feature that is common to all three of these main criticisms is that they seem to assume that Rawls has made a mistake in applying the terms of his own argument from the original position. I do not think this assumption is warranted. In this section, I address the first two criticisms. In Section IV, I turn to the criticism regarding Rawls's rejection of a global difference principle or other global distribution principle. A. Why liberal rights are not incorporated into the Law of Peoples

1. Human rights. In response to the first main criticism, the claim that liberal rights should be human rights, I have already discussed in Section II the connection Rawls forges between human rights and the minimally necessary conditions for social cooperation. To defend his distinction between human rights and liberal rights against the cosmopolitan argument that it would not be accepted by liberal peoples within the original position, it is important, first, to see how Rawls sets up the original position in his discussion of the Law of Peoples. For simplicity's sake, I will call this the "original position among (liberal or decent) peoples," to be distinguished from the "global original position" among representatives of all world-inhabitants that is argued for by Rawlsian cosmopolitans. The first original position among peoples consists of representatives of liberal peoples (not representatives, one for each person in the world, like a global original position), with all peoples regarded as equals regardless of the size of their populations. These representatives of liberal peoples are subject to a veil of ignorance; they do not know specific

John Rawls THE LAW OF PEOPLES

583 47

contemporary or historical facts about their own and other societies (their population, resources, wealth, and so on). They do know, however, that they are regulated by a conception of liberal justice—if not specifically justice as fairness then some other liberal conception that guarantees the basic liberties and their priority, and an adequate social minimum. Importantly, and as in the case of Rawls's domestic original position, the representatives of liberal peoples are concerned solely with promoting the "fundamental interests" of the individual society that each one represents. Unlike the parties to the domestic original position, however, these representatives are not moved only by a purely rational motive, to procure a greater share of the primary goods. Rather, their main aim is to obtain terms of cooperation among peoples that best guarantee liberal justice within their own society and among their own people (LP, 33, 40). But while the parties are moved by a concern for justice for their own peoples, they are not moved by benevolence toward other peoples or even by a concern that liberal justice be done to them for its own sake. They are mutually indifferent in this regard. In this respect, the original position among liberal peoples does not differ from the domestic original position.23 23 Here there is an important point of interpretation. Earlier in section 2.3 of The Law of Peoples, Rawls sets forth four fundamental interests of liberal peoples: "They seek [1] to protect their territory, [2] to ensure the security and safety of their citizens, and [3] to preserve their free political institutions and the liberties and free culture of their civil society. Beyond those interests, a liberal people [4] tries to assure reasonable justice for all its citizens and for all people" (LP, 29). In section 3.3 Rawls adds a further interest, [5] "a people's proper self-respect of themselves as a people" (LP, 34). It is important to my argument above (and, I believe, to Rawls's argument as well) that a distinction be drawn between the fundamental interests of liberal peoples, and the motivations of their representatives in the original position. It is a fundamental interest of liberal peoples "to assure reasonable justice ... for all people"; this is what it means for a people to "have a moral nature." But as for the representatives of peoples in the original position, they are not moved by this moral motive. Like the parties to the domestic original position, they are "modeled as rational" in Rawls's sense (LP, 32,33), which means they are not morally motivated and are indifferent toward the interests of other parties and peoples they do not represent (except insofar as it promotes the fundamental interests of their own people). This motivational assumption of mutual indifference of the parties is necessary for the structure of the original position, in order for it to do the work Rawls assigns to it: namely, to regulate rational judgments (regarding the interests of those one represents) by reasonable constraints (the veil of ignorance and other constraints of right). Nevertheless, the fact that the parties' representatives are modeled as purely rational and indifferent to one another and other peoples does not by any means imply that liberal peoples themselves are purely rational and indifferent. Indeed, Rawls is careful to emphasize in section 9 that what primarily distinguishes peoples from states is that peoples have a moral nature; thus, they are not moved solely by their own interests and are not indifferent to one another, but do seek "to assure reasonable justice ... for all peoples" (LP, 29). The reason this is an important point of interpretation is that, if the parties to the original position were motivated to do justice, not just for their own people, but for all peoples, then this would open the way for the argument that the representatives of liberal peoples should also be concerned that nonliberal peoples accept liberal justice, including all the liberal basic liberties. If so, then liberal peoples would not have reason to tolerate nonliberal decent peoples, at least not for the reasons Rawls suggests. K. C. Tan, among others, has made this argument, and my remarks here are intended as a response to him. See Tan, Toleration, Diversity, and Global Justice, chap. 2.

584

John Rawls 48

SAMUEL FREEMAN

Given this concern for liberal justice among their own people, and assuming that other liberal peoples already recognize the basic liberties and their priority, there is no reason for a liberal people to agree to enforce liberal rights against other liberal peoples. To begin with, other liberal peoples already accept and domestically enforce requirements of liberal justice—this is what entitles them to take part in the agreement among liberal peoples. Moreover, even if an injustice is done by another liberal government against one of its own people (suppose, for example, that a liberal people falsely convicts a member of a minority group that is widely disliked), this poses no danger of injustice to other liberal peoples. Given their interests in liberal justice in their own societies, there is insufficient reason for the representatives of liberal peoples to incorporate liberal basic liberties into the Law of Peoples that regulates their foreign relations with one another. What about liberal peoples' relations with nonliberal peoples? Here Rawls invokes the original position a third time, and imagines an agreement, not among liberal peoples and decent peoples, but only among decent hierarchical peoples themselves. They too, Rawls contends, would agree to the same Law of Peoples as would liberal peoples, including respect for everyone's human rights. For this reason, Rawls concludes that liberal peoples should tolerate decent hierarchical peoples and accept them as equal members in the Society of Peoples (LP, 84). The apparent reason for this claim is that liberal peoples have nothing to fear from a people if the latter endorses the Law of Peoples. For a people that endorses this Law respects other peoples' integrity and political autonomy; moreover, it also respects the human rights of its own members and of other individuals, and seeks to realize a common-good conception of justice among its own people. Decent peoples, then, do not present a threat to liberal peoples, or to anyone else, so there is no reason for the representatives of liberal peoples to refuse to tolerate decent peoples and recognize them as equals. Here enters the objection (posed by Beitz, Tan, and others) that if a liberal people is genuinely concerned about enforcing liberal rights among its own citizens, then it should also be concerned about enforcing the same rights among nonliberal peoples. As we have just seen, however, the representatives of liberal peoples in the original position have no motivation to enforce liberal rights among nonliberal peoples. This is not their assigned role. This is not to say that as individuals, representatives of liberal peoples do not care about the extension of liberal rights—they may care quite a lot that all the world adopt liberalism. But in their capacity as the legal representatives of their peoples' own interests, they are not motivated to enforce liberal rights around the world. Instead, they seek a Law of Peoples that secures the conditions of liberal justice for their own people. It is not their responsibility, charge, or jurisdiction to agree on a cosmopolitan conception of justice providing liberal rights for all peoples (any more than it is any other trustee's or legal representative's charge or

John Rawls THE LAW OF PEOPLES

585 49

responsibility to concern himself with the rights and interests of third parties or the general population as a whole). So long as nonliberal peoples can accept and respect the Law of Peoples, and thus respect the integrity of other peoples and their human rights, there is no need for them also to accept liberal rights for their own people; such acceptance is not necessary to secure the conditions of liberal justice for a liberal people. The general point is that there is no room in the original position for the argument that nonliberal people should not be tolerated because they do not accept liberal rights. That question simply does not arise. Of course, here it might be objected that this is arbitrary, due simply to the artifice of the original position and how the interests of the parties are defined by Rawls. But that is a different objection from the one we have been considering, namely, that Rawls's argument from the original position fails and is inconsistent since it fails to address certain legitimate concerns the parties may have. In response to this defense of Rawls, it may be said that if the representatives in the original position were only concerned about justice among their own peoples, why would they agree to tolerate only decent peoples? For there are many "outlaw" and "burdened" states whose rulers do not respect the human rights of their own people, but nonetheless present no danger to liberal peoples (for example, the rulers of many African states). Why not tolerate these "harmless outlaws" too, so long as they present no threat to one's own security?24 This suggests that Rawls, by his own arguments, puts himself into the position of having to accept a realist foreign policy, in spite of his efforts to do otherwise, and must tolerate gross injustices and violations of human rights by some nations so long as they do not jeopardize other nations7 security. One reply to this argument is that liberal people do indeed have a good deal to fear from a state that has no respect for human rights or other principles of the Law of Peoples.25 One only has to look at the dislocation of individuals and the disruption and war among neighboring nations that is caused by violations of human rights in Africa today. Behind the veil of ignorance, a representative does not know whether a neighboring country is liberal, decent, or an outlaw regime, nor does he know its relative size or strength in comparison with other peoples. Moreover, putting the original position and the motivations of its parties aside, the fact is that liberal peoples themselves do value human rights for their own sake, and seek "reasonable justice . . . for all peoples" (LP, 29). Peoples as peoples have a moral nature, and as such they have a sense of justice and are concerned with respect for human rights for their own 24

Charles Beitz raises considerations along these lines in "Rawls's Law of Peoples/' 685. Rawls says, "Liberal and decent peoples have extremely good reasons for their attitude. Outlaw states are aggressive and dangerous; all peoples are safer and more secure if such states change, or are forced to change, their ways. Otherwise they deeply affect the international climate of power and violence" (LP, 81). 25

586

John Rawls 50

SAMUEL FREEMAN

sake. (They are also concerned with respect for liberal rights for their own sake, with the important qualification that people who enjoy them conceive of themselves as free and as equal citizens.) The objection stems from focusing on the motivations of the parties to the original position. Just because, from the point of view of the parties to the original position, respect for human rights by other peoples is important for instrumental reasons, does not mean that the Law of Peoples regards human rights purely instrumentally or from a self-interested perspective. (The parallel here is with citizens in a well-ordered society of justice as fairness, who value justice for its own sake, and their representatives in the original position, who are not motivated by considerations of justice in the original position.) To say that Rawls's justification of human rights is purely "instrumental" is to fail to see that the parties in the original position are only part of a larger argument. They are not real people, but merely embody rational considerations regarding the good of those they represent, whose interests are then subjected to the moral constraints of the original position, and the moral nature of free and equal people. Here it bears emphasizing that the parties' toleration of nonliberal decent peoples, and their failure to include liberal rights in the list of human rights that the Law of Peoples protects, does not mean that liberal citizens or liberal societies do not regard liberal rights as universally applicable or as an ideal all societies ought to aspire to. Clearly, most liberal citizens do, as do most liberal government officials. Nevertheless, as a people (if not individually), they also respect well-ordered decent peoples as free and equal peoples, and as politically autonomous, capable of self-determination, and capable eventually of coming to an acceptance of liberal rights themselves. Recall that in a well-ordered decent hierarchical society, all reasonable and rational members of society accept its nonliberal terms of cooperation, and likely accept too the nonliberal comprehensive doctrine that is used to justify these terms of cooperation. To regard all liberal rights as human rights and insist they should be enforced by the Law of Peoples is to impose upon nonliberal but decent peoples, for reasons they cannot accept, terms of cooperation that are universally at odds with the moral and political views of nearly everyone in that society. There is no justification within the public reason of the Society of Peoples for such measures. Moreover, such measures would fail to respect decent peoples as politically autonomous, both as individuals and as a people. Finally, there is the practical consideration that nonliberal decent peoples will not come to accept and endorse liberal rights under coercive terms, or if they are made to feel that they and their own comprehensive views are disdained and not worthy of respect by liberal peoples. 2. The duty of assistance. Since liberal peoples have reason to care about reasonably just institutions and practices, including human rights, for their own sake, and not simply because they are themselves benefited, we can see why Rawls argues for a duty of assistance for burdened peoples.

John Rawls THE LAW OF PEOPLES

587 51

To begin with, for Rawls, the human right to life includes a "basic right" to the means of subsistence as well as security (LP, 65). Subsistence includes "minimum economic security" (LP, 65 n. 1). To justify a basic right to minimum economic security, Rawls says that "the sensible and rational exercise of all liberties, of whatever kind, as well as the intelligent use of property, always implies having general all-purpose economic means" (LP, 65). The question is, who has the duty to see to it that this right is satisfied? Clearly, the government that is the agent responsible for a people has the primary duty to provide the means of subsistence for its members.26 But when a government is incapable or refuses to provide economic means sufficient to meet its members' subsistence needs (and basic needs as well?), it falls to the Society of Peoples to fill this duty. The eighth Law of Peoples says: "Peoples have a duty to assist other peoples living under unfavorable conditions that prevent their having a just or decent political and social regime" (LP, 37). Here it appears that the duty of assistance requires more than simply providing for the subsistence needs of a burdened people, which is a human right they have. For Rawls's gloss on the eighth Law of Peoples refers to "basic needs," which is clearly a broader category than subsistence needs since it includes "those [needs] that must be met if citizens are to be in a position to take advantage of the rights, liberties, and opportunities of their society. These needs include economic means as well as institutional rights and freedoms" (LP, 38 n. 47). The duty of assistance is keyed to this broader concept of basic needs, and this suggests that the duty of assistance extends beyond meeting a burdened peoples' subsistence needs.27 The

26

Rawls says that for a government to allow its people to starve when starvation is preventable reflects a "lack of concern for human rights'' (LP, 109). 27 It is fairly clear that Rawls sees the duty of assistance as providing more than basic means of subsistence for people. He says that the aim of the duty of assistance "is to help a people manage their own affairs reasonably and rationally" (LP, 111), and to help them "to be able to determine the path of their own future for themselves" (LP, 118). "When the duty of assistance is fulfilled, and each people has its own liberal or decent government... each people adjusts the significance and importance of the wealth of its own society for itself. If it is not satisfied, it can continue to increase savings, or, if that is not feasible, borrow from other members of the Society of Peoples" (LP, 114). Either alternative assumes that a people is in a position to create wealth well above what is needed to provide means of subsistence for all its members. Here again, the argument cannot be made that Rawls's reliance on a duty of assistance and his rejection of a global distributive principle allows corrupt governments to borrow money and saddle their people with poverty for generations. This criticism, based on contemporary practices by corrupt regimes and their relations with affluent nations, refuses to acknowledge that the Law of Peoples is formulated for the ideal feasible case of well-ordered decent societies governed by a common-good conception of justice. Nothing in the Law of Peoples implies that peoples now should tolerate the exploitation of less-developed peoples either by their own governments or by multinational corporations. Quite the contrary: since the aim of the Law of Peoples is that all peoples should be members of a well-ordered Society of Peoples, the implication is that historical practices of exploitation of one people by another or by private interests should be prohibited, since such practices impede the development and independence of burdened peoples.

588

John Rawls 52

SAMUEL FREEMAN

implication is that while it is not a human right to have all one's "basic needs'' met, as defined, still the Society of Peoples has a duty of assistance to meet basic needs until burdened peoples can provide for all their members themselves and become self-sustaining members of the Society of Peoples. Rawls says that the "target" of assistance is to enable "burdened societies to be able to manage their own affairs reasonably and rationally and eventually to become members of the Society of wellordered Peoples" (LP, 111). The "final aim of assistance" is "freedom and equality for the formerly burdened societies." Particularly important, as we will see, is his claim that the purpose of the duty of assistance is to "assure the essentials of political autonomy" or the capacity of a people "to determine the path of their own future for themselves" (LP, 118). This duty of assistance can be quite extensive, then, especially if it is made contingent upon both a society's capacity for political autonomy and also the culture of particular societies and the resources needed to take advantage of the opportunities they offer. It would suggest, for example, a duty to provide in some way for the educational needs of a burdened people, so that they can find employment and be economically self-sufficient, and can actively participate in the life of their culture. Rawls has too little to say here. He does say that what the representatives of peoples discuss in the original position, instead of competing conceptions of the Law of Peoples, are different formulations and interpretations of the eight principles of the Law of Peoples (LP, 40). This seems to leave it open for further determination that the duty of assistance might impose rather exacting demands upon members of the Society of Peoples, to assist burdened peoples. This is an important point in responding to the argument against Rawls for his failure to provide for a principle of global distributive justice (a subject I will return to in Section IV). B. A resource redistribution principle?

Still, the duty of assistance has both a "target" and a "cutoff," and does not amount to an open-ended principle of distributive justice (as Rawls uses that term). Once peoples' basic needs are met and they are selfsustaining members of the Society of Peoples, the duty of assistance to other peoples is fully satisfied by other members of the Society of Peoples. There is no further duty arising from a principle of distributive justice to continually provide once-burdened societies with resources. The implication of this lack of a distribution principle is that no real limit is imposed on the degree of inequality that can exist among peoples (other than the limit implied by the duty of assistance). Here it is objected (by Beitz, Pogge, Martha Nussbaum, and others) that Rawls is guilty of (yet another) inconsistency in his argument. In his argument for principles of domestic justice, Rawls makes much of the fact that people do not deserve either the greater natural talents or the social

John Rawls THE LAW OF PEOPLES

589 53

position with which they are born.28 But if no person deserves his starting position in life, or to be born with greater (or less) natural and social advantages than anyone else, then surely it must also be true that no nation or people deserves to be "bom" with greater or less natural resources than other peoples either. Charles Beitz says, "Like talents, resource endowments are arbitrary in the sense that they are not morally deserved/'29 But the natural resources a people controls contribute decisively to the level of income and wealth its members enjoy and their comparative (dis)advantages. In his earlier work, Beitz argues that the parties to an international original position would insist on a resource redistribution principle; they would do this, not simply because it is fair, but to protect their own interests. "Not knowing the resource endowments of their own societies, the parties would agree on a resource redistribution principle that would give each society a fair chance to develop just political institutions and an economy capable of satisfying its members7 basic needs/'30 Is Beitz's claim true when applied to the original position among peoples that Rawls sets forth in his discussion of the Law of Peoples, which involves separate agreements among liberal peoples and among decent peoples? Why wouldn't liberal peoples (and then decent peoples) agree at least among themselves to redistribute resources, in order to protect themselves from the eventuality that they may represent resource-poor peoples? Here, surely, we have a different case from the preceding one, where liberal peoples had insufficient reason to agree to enforce liberal rights globally. If we focus just on the artifice of the original position, I think it must be said once again (though here the argument is not as conclusive) that the original position is structured so that the question of redistributing resources does not arise. The purpose of the agreement within the original position is to establish principles of foreign policy among liberal peoples; its purpose is not to arrive at principles of compensation or at a conception of cosmopolitan justice. Of course, the parties might be concerned with the level of natural resources that they have. But would their concern lead them to insist upon redistributing natural resources among Here is it important to emphasize that Rawls does not say that people do not deserve their natural talents. This is a common misreading, and it leads to much criticism of Rawls (e.g., claims to the effect that Rawls rejects natural rights to one's own talents and abilities). What Rawls in fact says is that we do not deserve to be "better endowed" than others (TJ, 13), and that "[i]t seems to be one of the fixed points of our considered judgments that no one deserves his place in the distribution of natural endowments, any more than one deserves his initial starting place in society" (TJ, 89, emphasis added; see also 274, where Rawls repeats this claim). He means here simply that people do not deserve the differences in natural talent or social position they are born with. No one deserves to be born smarter or richer than anyone else. This seems obvious to Rawls, a "fixed point in our considered judgments" (TJ, 274). It is a very different claim from the one often falsely attributed to Rawls, namely, that no one deserves his or her natural talents. 29 Beitz, Political Theory and International Relations. 139. 30 Ibid., 141. 28

John Rawls

590 54

SAMUEL FREEMAN

peoples as a condition of cooperating as peoples? (Or, we might ask, would it lead them to insist upon a global resource tax, as Pogge argues?) Here they would need to take into account the burdens and the degree of interference with their independence that may be involved in redistributing natural resources discovered in their territory. (Is it to be done in kind? If so, who is responsible for extracting resources? If instead of going to the enormous trouble of redistributing resources in kind, poor peoples are to be compensated monetarily, then where is this money to come from and what is the periodic rate of payment? Are the peoples of resource-rich countries to be taxed, and can they afford this? Or is a resource-rich country to flood the world market with its resources—thus driving down the price—to raise money to pay compensation to resource-poor nations?) Many questions like these arise, which might give representatives of peoples pause in considering the wisdom of a resource redistribution principle.31 Just as the representatives to Rawls's domestic original position do not agree to a resource redistribution principle per se, or a principle that compensates people proportionately for the degree of their disadvantage, but instead agree to a principle of distributive justice that structures the economy so as to maximally benefit the worst off, so it would seem to be more rational for the parties to the extended original position to simply agree to a principle of distributive justice to deal with issues of resource inequality. Nevertheless, even dealing with the problem that way might be questionable, for reasons Rawls himself suggests. Rawls gives short shrift to the resource redistribution argument. Basically, he denies its claim that the natural resources a people controls determine its members' income and wealth. Pointing to the people of Japan and other resource-poor but still well-to-do peoples (and also to the people of Argentina and other resource-rich but still poor peoples), he says that how a nation fares has much more to do with its political culture than with the natural resources it controls within its boundaries: There is no society anywhere in the world—except for marginal cases— with resources so scarce that it could not, were it reasonably and rationally organized and governed, become well-ordered.... [T]he crucial element in how a country fares is its political culture—its members7 political and civic virtues—and not the level of its resources. (LP, 108, 117) Given the centrality of political culture to a society's well-being, Rawls says, "the arbitrariness of natural resources causes no difficulty" (LP, 117). 31 Here it might be said that these are simply questions of the application of a global resource principle. However, if we accept the publicity condition on principles, then I think that more is involved here than simply the application of a vague principle. Why should representatives of peoples want to accept a principle that inevitablv will give rise to interminable disputes among peoples?

John Rawls THE LAW OF PEOPLES

591 55

I think an important assumption underlying Rawls's argument here is that, on Rawls's view, the duty of assistance should be set at a level that assures political autonomy, the capacity of a people "to be able to determine the path of their own future for themselves" (LP, 118). It is only once the political autonomy of a people is procured that the level of wealth a country enjoys can become largely a matter of its political culture (as opposed to being a matter of the natural resources it has on hand). (See LP, 117-18.) Here, again, it is important to emphasize that Rawls's argument applies to well-ordered societies in a well-ordered Society of Peoples. It is under those circumstances, where there is an absence of internal political corruption and external exploitation by other peoples and by multinational business interests that a people's level of well-being will largely be decided by its political culture, and the natural resources it controls will not be such a significant factor. We can assume that the parties to Rawls's original position have this general knowledge. Thus, if Rawls's argument is convincing, then it is questionable whether it is rational for the representatives of peoples to be worried about instituting a resource redistribution principle or about compensating the resource-poor at all, especially given all the potential problems that arise with administering such a principle. One response to Rawls's argument is this: Why should the inhabitants of poor countries have to take responsibility for the miserable political and social culture they are born into any more than they should have to take responsibility for the level of resources they are born with? From the perspective of individuals, it is just as arbitrary for one to be born into a culturally impoverished country as it is for one to be born into a resourcepoor country. This raises the question of the need for a global distribution principle (as opposed to a global resource redistribution principle), a subject that will be considered in the next section. For his part, Beitz in his response to Rawls says that the question of the degree to which political culture, natural resource endowments and the lack thereof, technology and human capital, and other factors contribute to economic backwardness is in dispute; moreover, it may not even be an intelligible question, given developing societies' enmeshment in a world economy.32 I would argue, however, that to refute Rawls's position, one would have to make a more convincing empirical case that political culture cannot be relatively independent and self-determining even under ideal conditions. In any case, Beitz in later works seems to drop his earlier insistence on a resource redistribution principle, and resolves problems of differences in natural resource endowment with a global principle of distributive justice. Here, again, it has to be kept in mind that Rawls is focused on the possibility of political autonomy and a people's control over its political and social culture under conditions of well-ordered societies, all of which 32

Beitz, "Rawls's Law of Peoples/' 690.

John Rawls

592 56

SAMUEL FREEMAN

are members of the Society of Peoples. The importance of this point will be emphasized again in the next section in my discussion of a global principle of distributive justice. But the main point I want to make here is that these criticisms of Rawls's position—for failure to include a resource redistribution principle and perhaps a principle of distributive justice also—indicate the degree to which the cosmopolitan position depends (frequently, if not always) upon what is often called "luck egalitarianism," or the idea that it is the role of a conception of justice to correct for and equalize the effects of natural and social chances and accidents. The degree to which many of Rawls's cosmopolitan critics are luck egalitarians is evident when they suggest that Rawls is inconsistent in this argument.33 They claim that if Rawls is bothered by the luck of the draw in the natural and social lottery in the domestic case, he should also be bothered by the luck of the draw among peoples in the "resource lottery." This contention underlies many arguments for the resource redistribution principle, and for a global distribution principle. For Beitz and others who advocate a resource redistribution principle, being engaged in a cooperative endeavor is not a condition for the application of this principle. Indeed, a cooperation requirement would be contrary to the purpose of the principle, since the opportunity for cooperation is itself nearly as dependent upon arbitrary factual contingencies as other facts that influence distributions of benefits. Why should the natural resources a country has at its disposal depend upon whether it cooperates (economically, politically, etc.) with other nations? After all, the lack of natural resources has a direct bearing upon an impoverished people's inability to cooperate. Luck egalitarians, who include the Rawlsian cosmopolitans, often refer to Rawls's claim that no one deserves greater natural talents or his starting position in society, and that the outcome of the natural lottery "is arbitrary from a moral perspective" (see TJ, sec. 12), and contend that there is an inconsistency in Rawls's argument. They use the same idea to support their position that the consequences of chance should be equalized or at least neutralized in the distribution of natural resources among peoples. Here it is relevant that, when Rawls says in A Theory of Justice that no one deserves his place in the distribution of native endowments (TJ, 89) and that this natural distribution is morally arbitrary (TJ, 64), he is making this point within the context of an argument for principles of justice that apply to societies' basic structures, as ongoing socially cooperative endeavors. The inference he draws is not that a society should seek to equalize distributions that are the consequences of chance. It is, rather, that the natural (or social) endowments one is born with should not be allowed to determine one's place in the distribution of income and wealth; instead, some other principle should. This principle is the difference principle, as Rawls goes on to argue. The difference principle is not 33

Cf. Beitz, Political Theory and International Relations, 137-38.

John Rawls THE LAW OF PEOPLES

593 57

a luck egalitarian principle, or, as Rawls says, it "is not the principle of redress77 (T], 86-87). That is not its point. It does not try to equalize the results of birth, social class, and other contingencies, or compensate the disadvantaged for their unfortunate circumstances. Rather, it is a principle that "distributes" the benefits and burdens that result from natural and social differences, regardless of whether they are the product of chance, so as to maximally benefit the least advantaged. As I discussed earlier, the difference principle presupposes, and is designed to apply to, the basic institutions constituting the basic structure of society. Rawls's point is not that luck or natural facts should never determine or affect distributions of income and wealth of any kind. (How could that be possible? No matter what we do, morally arbitrary natural facts are going to affect distributions of assets in some way.) It is rather that, within socially cooperative frameworks, the distribution of natural assets should not be allowed to determine the distribution of income and wealth, unless it also maximally benefits the leastadvantaged members of society. It does not follow that one may generalize this point, and apply it globally, where there is an absence of socially cooperative frameworks and a shared basic structure. Such a move rejects Rawls's position that social cooperation and the special political and social relationships of a shared basic structure matter to distributive justice. I conclude that Rawls is not guilty of the inconsistency he is accused of on this point. The luck egalitarianism presupposed by cosmopolitans invokes larger issues that cannot be discussed here. Briefly, however, a problem with the luck egalitarian position—that the results of natural and social contingencies should be equalized or at least neutralized—is that it has no clear stopping place in the following sense. Luck egalitarians want to draw a sharp distinction between events that result from chance and those that result from people's choices. Differences in distributions should only reflect people's free choices, after everyone has been compensated (or taxed) for the unfavorable (or favorable) circumstances he or she starts out with. But our capacities for choice, our occasions for choice, and the alternatives for choice we confront are also influenced by chance circumstances (by our natural talents, our upbringing, others' talents, upbringing, and choices, our social connections, etc.); so it is hard to say when a choice is not itself the result of "brute luck." Moreover, for the strict determinist, nothing is left undetermined by nature; thus, our choices are themselves as much a matter of our circumstances as are our starting positions in life. The point is not that the choice/chance distinction does not have its place within our practices of holding people responsible and in deciding particular questions regarding fair distributions. Rather, it is that, because the distinction is so uncertain in so many cases, it is an unsuitable foundation for a conception of distributive justice.34 34 On this issue, see Samuel Scheffler, "What Is Egalitarianism?" Philosophy and Public Affairs 31, no. 1 (Winter 2003): 5-39. For a more general discussion, see Susan Hurley, Justice, Luck, and Knowledge (Cambridge, MA: Harvard University Press, 2003).

594

John Rawls 58

SAMUEL FREEMAN

IV. PROBLEMS WITH A GLOBAL DISTRIBUTION PRINCIPLE The argument is frequently made that Rawls's difference principle, or some other principle, should globally apply to determine the distribution of income and wealth.35 In this section, I discuss some potential problems with the idea of a global difference principle and, more generally, some reasons why Rawls rejects the idea of global distributive justice. It is sometimes suggested that, in rejecting a global distribution principle, Rawls wrongly assumes that laissez-faire among peoples is the default position for global economic distribution. Thomas Pogge, for example, says that Rawls in effect gives a Nozickian reply to arguments for a global difference principle. For Rawls, "It is somehow natural or neutral to arrange the world economy so that each society has absolute control over, and unlimited ownership of, all natural resources within its territory/'36 However, Pogge claims, Rawls gives no reason for this assumption, nor does he show why the opposite assumption should not be made, namely, that the difference principle or some other distributive principle should be the default position and should apply globally. I would argue, however, that Rawls does not assume that the difference principle globally applies for the simple reason that he believes that claims of distributive justice are already settled at the domestic level. For reasons I will discuss momentarily, it would make no sense to argue that the difference principle should apply a second time, before or after it is domestically applied. Moreover, the disanalogies with Nozick's libertarianism are too numerous for Pogge's comparison to be of critical value. Rawls does not assume absolute property rights or laissez-faire economic relations among peoples. For Rawls, the duty of assistance to burdened societies is a condition upon a society's use of its resources; by contrast, in Nozick's libertarianism there is no political duty to assist anyone in need. Moreover, on Rawls's view, the Society of Peoples has requisite authority to restrict or regulate nations' and corporations' detrimental uses of resources in ways that would not be recognized by libertarians. Finally, the analogy between a people's independent control of a territory and an 35

For example, Thomas Pogge says: "Taken seriously, Rawls's conception of justice will make the life prospects of the globally least advantaged the primary standard for assessing our social institutions/' Thomas Pogge, "Rawls and Global Justice/7 The Canadian Journal of Philosophy 18, no. 2 (1988): 227-56, at 233. And Charles Beitz has said: "It seems obvious that an international difference principle applies to persons in the sense that it is the globally least advantaged representative person (or group of persons) whose position is to be maximized/' Beitz, Political Theory and International Relations, 152. T. M. Scanlon also once suggested that the difference principle should apply globally. See T. M. Scanlon, "Rawls' Theory of Justice," in Reading Rawls, ed. Norman Daniels (Palo Alto, CA: Stanford University Press, 1989), 202. 36 Pogge, "An Egalitarian Law of Peoples," 212-13. See also Pogge, "Rawls and Global Justice," where he says that Rawls leaves international economic interactions up to "libertarian rule-making" (250) and that "[t]he economic order of Rawls's utopia ... is shaped by free bargaining" (252).

John Rawls THE LAW OF PEOPLES

595 59

individual's rights over property (however extensive) is unfitting. By exercising political jurisdiction over a territory, a people establishes a system of property, which is a complicated system of rules and interdependent expectations. By contrast, individuals do not establish systems of property; rather, they hold and use possessions within property systems, subject to the systems' legal rules and expectations. A people's control over a territory is not a kind of property; it is the condition for the existence of the social institution of property. More generally, it is a condition for social and political cooperation and the very existence of a political people. Why does Rawls reject the global application of the difference principle? There are several reasons: 1. Rawls regards the difference principle as a principle of reciprocity, designed to apply under conditions of social cooperation to the basic structure of society, where the members of society are regarded as engaging in a complex web of political and social institutions that make up the basic structure. People engaged in social cooperation in a common basic structure of institutions are confronted with a crucial question: How are the terms of cooperation to be structured among themselves as they each pursue their individual purposes and conceptions of the good? The difference principle is designed to express the idea of reciprocity from a benchmark of equality that (for Rawls) defines the terms of social cooperation among free and equal democratic citizens. Assuming that the members of a democratic society are engaged in a common social and political endeavor, advances in the position of those better off should not at any point come at the expense of the worse off; rather, the worse off should consistently benefit from changes in the terms of cooperation that benefit the more advantaged members of society.37 To say that the difference principle should apply globally, and regardless of the kind of cooperation that exists among people, implies that there is nothing special about social and political cooperation within the basic institutions of society; moreover, it implies that democratic social and political cooperation is of no consequence to questions of distributive justice. This is explicit in the objection of those luck egalitarians who argue that the difference principle should apply globally since people cannot control which society they are born into or whom they are destined to cooperate with. The luck egalitarian will say: "If cooperation and whom you cooperate with are just as arbitrary as are the talents and social position you are born with, then the fact of cooperation should not act as a limit upon the application of the difference principle."38 But if society 37 1 refer here to Rawls's graphical depiction of the difference principle in A Theory of Justice, sec. 13, figure 6, and in Justice as Fairness: A Restatement, sec. 18, p. 62, figure 1. 38 As K. C. Tan argues: On Rawls's account of distributive justice, "Citizens of disadvantaged countries are collectively held accountable for their country's unsound domestic policies, even when a majority of them had no part in the making of these policies. And this is

John Rawls

596 60

SAMUEL FREEMAN

is to be possible at all among individuals, there must be special terms of cooperation that apply to members of the group that do not apply to those who are not members. Among these, Rawls contends, are principles of distributive justice based on the idea of reciprocity. 2. Luck egalitarianism is one basis for a global distribution principle. Another potential basis is more Rawlsian: it is the claim that global cooperation and a global basic structure exist, particularly in economic relations, and that by Rawls's own criteria a global distribution principle is appropriate in order to reward all those poorer peoples who do their part in global economic production. Allen Buchanan, among others, argues that "it is unjustifiable to ignore the global basic structure in a theory of international law," and that if Rawls had recognized the influence of the global basic structure on individuals' and peoples' prospects, he would have accepted a global distribution principle. Constituting the global basic structure for Buchanan are the following institutions: [Rjegional and international economic agreements (including the General Agreement on Tariffs and Trade, North American Free Trade Agreement, and various European Union treaties), international financial regimes (including the International Monetary Fund, the World Bank, and various treaties governing currency exchange mechanisms), an increasingly global system of private property rights, including intellectual property rights that are of growing importance as technology spreads across the globe, and a set of international and regional legal institutions and agencies that play an important role in determining the character of all the preceding elements of the global basic structure.39 If this is what Buchanan means by "global basic structure," it is incorrect to suggest that Rawls ignores it. For Rawls, institutions like these are an integral part of the Society of Peoples, but he refuses to regard them as a global basic structure of the kind that warrants principles of distributive justice. Buchanan and others put greater emphasis on these international institutions since they believe they affect people's future prospects just as a domestic basic structure does.40 But there is really no comparison between the basic structure of society and the effect of global institutions. The difference is not simply a matter of the (far) greater degree to which clearly inconsistent with Rawls's own moral individualism. On Rawls's own reasoning, a person born into a society with poor population control and economic policies cannot be said to deserve her fate any more than another bom into more favorable circumstances deserves her[s]. These are mere accidents of birth, and are as morally arbitrary as is being born into wealth or poverty in the domestic context." K. C Tan, "Critical Notice of John Rawls, The Law of Peoples," Canadian Journal of Philosophy 31 (March 2001): 113-32, at 122. 39 Buchanan, "Rawls's Law of Peoples/' 706. 40 Ibid.

John Rawls THE LAW OF PEOPLES

597 61

domestic institutions affect people's lives. Nor is it simply that international institutions are supervenient upon national ones: they presuppose the complicated basic institutions (the many systems of property and contract, for example) of the basic structures of the many societies whose practices they regulate. Rather, it is also that these international institutions are the product of independent peoples' exercise of their original political jurisdiction as members of the Society of Peoples, which they agree to in order to maintain their own basic structure of society, over which they exercise political autonomy. Simply put, there is no global basic structure mainly because there is no world-state, with all it would entail. (For example, as I discuss below, since there is no world-state, there is no independent global property system to apply a principle of distributive justice to, such as the difference principle; international property conventions presuppose and are confined by the terms of the rules of property systems of politically independent peoples.) There is, then, a fundamental difference between Rawls's and his cosmopolitan critics7 assumptions regarding the conditions of social cooperation and distributive justice. 3. Rawls regards the difference principle as a political principle in the sense that it is to guide legislators in defining and regulating the uses of property, setting commercial policies, specifying schemes of taxation, regulating securities and negotiable instruments, defining conditions for copyrights, patents, royalties, and other forms of intellectual property, and establishing the other indefinitely many laws and regulations that structure an economy. (Compare Rawls's remarks regarding application of the difference principle at the legislative rather than the constitutional stage [PL, 229-30].) Strictly applied, the difference principle would require legislators and other officials to consider the effect of laws and regulations upon the prospects of the worst-off members of society. While they may not need to scrutinize each decision to determine if it maximizes the prospects of the worst off, governing officials should at least determine that the least advantaged are not disadvantaged indirectly by decisions made, and that any benefits created by political and economic policies also redound to an appropriate degree to the benefit of the worst off. There are an enormous number of laws, regulations, legal precedents, and conventions that structure property and economic systems (literally millions in the U.S. federal system). In the absence of political authority and political cooperation, it is hard to see how the difference principle could be applied to influence, much less determine, these innumerable laws, rules, and social practices. The primary practical problem with the cosmopolitan suggestion that the difference principle should be globally applied is that, in the absence of a world-state, there is no political agent with authority to apply it on a global level. Here it might be suggested that each government representing the world's many peoples should apply the difference principle, if not jointly then at least severally. This

John Rawls

595 62

SAMUEL FREEMAN

would require that each nation calculate the effects of its many decisions upon the worst-off members of the world. This is not feasible, nor would it have the desired (or even desirable) effects. The coordination problems of many nations separately trying to tailor their many decisions to affect peoples in distant lands over whom they have no political authority seem insurmountable in the absence of a world-state. This does not mean that, in making their laws regulating economic institutions, societies should not take into account the adverse effects of their economic policies and property norms upon other peoples, particularly less-advantaged peoples. But this would not be equivalent to instituting the difference principle on a global scale, since it would not require governments to choose only those policies that maximally benefit the (world's) least advantaged. Moreover, there is no need for a global difference principle, or a global distribution principle of any other sort, to induce governments to take into account and restrict the adverse effects of their policies on the world's least-advantaged nations. The duty of assistance already implies a duty to consider and rectify adverse effects of economic policies upon burdened peoples. Moreover, in reference to the "guidelines for setting up cooperative organizations," "standards of fairness for trade," and "provisions for mutual assistance," Rawls states: "Should these cooperative organizations have unjustified distributive effects, these would have to be corrected in the basic structure of the Society of Peoples" (LP, 115). Presumably he means here the "Confederation of Peoples" and international economic institutions with regulatory oversight duties over economic agents as they engage in trade, international investment, and other economic transactions (see LP, 42-43, sec. 4.5 on cooperative organizations). Nothing Rawls says implies that standards for fairness of trade and "distributive effects" are to be decided (as Pogge and other critics suggest) at the global level by a doctrine of unmitigated laissez-faire or libertarian entitlement principles. On the contrary, the implication is that economic relations among peoples are to be regulated with the aim of rendering them all independent and self-sustaining members of the Society of Peoples, each of whose individual members' basic needs are met so that all persons are able to actively participate in their particular society and take advantage of the rights, liberties, and opportunities it offers them (cf. LP, 38 n. 47). 4. It has been suggested (by Pogge, K. C. Tan, et al.) that since there is global economic cooperation, the difference principle should doubly apply, both at the global and at the national level. The thought here seems to be that just as the members of a family can distribute their resources according to principles of local justice once they obtain their fair share under the difference principle in their (liberal) society, so too a society can apply the difference principle to distribute its fair share of the global product once the global difference principle has been satisfied. But the difference principle is not like a principle that applies to a fixed allocation of goods—it's

John Rawls THE LAW OF PEOPLES

599 63

not as if the total global product is like a big cake that we can keep slicing into shares at the global, domestic, and then familial levels. The difference principle can apply only once to structure economic and property institutions, either globally or domestically. It cannot apply to both. (Among other reasons, we can seek to maximize the position of the globally least advantaged, or the domestically least advantaged, but not both, for we can maximize only one thing.) In this way, the analogy between the dual application of the difference principle at the global and domestic levels differs from the dual application of the difference principle first domestically and then within the family. The economy and the family are two different kinds of social institutions, whereas domestic and global economic relations are not. It is not clear what sense could be given to the idea of first structuring economic institutions so that they maximize the position of the worst off in the world, then structuring domestic economic and legal institutions so that they maximize the prospects of the domestically worst off within each society. 5. Not only is there the problem (mentioned in item 3 above) of who is to apply a global difference principle to achieve its intended effects, but there is also a problem regarding just what a global difference principle is supposed to apply to. In the absence of a world-state, there is no global legal system regulating relations among individuals. To take one important example, one role of the difference principle is to specify rights and permissible uses of property interests, and to regulate transactions involving property. However, there is no global property system to apply a global difference principle to. There are economic relations among peoples and members of different peoples. But even here, in the case of international trade, it is (in the absence of treaties setting out specific rules) the property and contract laws of one or another country that apply to regulate economic transactions and to decide disputes about rights when they break down. There is very little international property law that applies to international trade, and the international law that does exist is largely the result of treaties among peoples, and is dependent in many ways upon existing laws in various countries. 6. Let us suppose, however, that we can make sense of the idea of a global difference principle, independent of a basic structure of society and a world-state (and leaving aside questions of how it is to be implemented among nations, and whether it can practicably serve as a public conception of justice that guides the actions of independent peoples or individuals). Rawls's explicit reason for rejecting a global difference principle, or any other global principle of distributive justice, is that since a global distribution principle would continuously apply to all wealth without a cut-off point, it would be unfair to politically independent peoples. He gives two examples, both of which assume the ideal case of well-ordered societies that are members of the Society of well-ordered Peoples. The first example involves two societies beginning with the same level of

600

John Rawls 64

SAMUEL FREEMAN

wealth, one of which saves and invests its resources in industrialization and over time becomes wealthier, the other of which prefers to remain "a more pastoral and leisurely society" of modest means. It would be "unacceptable/' Rawls says, to tax the incremental wealth of the richer society and redistribute it to the poorer nation. The second example runs parallel to the first, but assumes a rather high rate of population growth. One society undertakes population control measures to restrain the high rate of growth and achieves zero-growth, while the other society, for religious and cultural reasons "freely held by its women" does not. (Rawls's example here presupposes "the elements of equal justice for women as required by a well-ordered society" [IP, 118].) Over time, the per-capita income of the society practicing population control is higher. Again, it "seems unacceptable" to tax the wealth of the richer nation and redistribute it to the poorer nation that has freely chosen to maintain its population at higher levels for religious reasons (LP, 117-18). Importantly, Rawls's rejection of a global distribution principle rests upon the assumption that it is possible for a well-ordered people to exercise political autonomy, that its members are economically selfsufficient (relatively speaking) and not subject to manipulation by external forces beyond their control, and that they can control their level of wealth through savings, investment, population control, and other measures. (See The Law of Peoples, 117 and 118, where Rawls indicates that well-ordered liberal and decent peoples are "free and responsible and able to make their own decisions," and "able to determine the path of their own future for themselves.") Rawls's critics seem to question this crucial assumption of the independence of peoples under ideal conditions of a well-ordered Society of Peoples. It has been said, for example, that the inequalities of resources that Rawls's account allows for will inevitably lead to political corruption and the exploitation of less-developed peoples by richer peoples. "In a world with large international inequalities, the domestic institutions of the poorer societies are vulnerable to being corrupted by powerful political and economic interests abroad."41 Again, it is important to recall that the Law of Peoples is drawn up for the ideal case of well-ordered societies joined into a well-ordered Society of Peoples. The Law of Peoples includes a duty of peoples to provide for material and other conditions that enable all peoples to be politically autonomous and independent. These and other requirements should protect less-advantaged peoples from "being corrupted by powerful political and economic interests abroad." If critics claim that such corruption is nonetheless inevitable so long as inequality exists, then clearly Rawls is 41 See Pogge, "An Egalitarian Law of Peoples/' 213. Pogge further argues against the inequality of wealth allowed by Rawls's account, saying that "[rjelative poverty breeds corruptibility and corruption/7 and that "[i]t is entirely unrealistic to expect that such foreignsponsored corruption can be eradicated without reducing the enormous differentials in per capita GNP" (ibid., 213, 214).

John Rawls THE LAW OF PEOPLES

601 65

more sanguine here than his critics are regarding human nature and the possibilities for a "realistic utopia" of politically autonomous and independent well-ordered peoples. Let us put this dispute aside, since it involves largely empirical conjectures about the capacity for political autonomy and the workings of an economy, both domestically and worldwide, under conditions of a well-ordered Society of Peoples. What I want to focus on instead is Rawls's claim that the argument for a global distribution principle made by cosmopolitans is grounded in concern for "the well-being of individuals and not the justice of societies." What are the origins of this claim? I believe we can understand its origins by examining the two examples I discussed earlier, involving rich and poor societies. A global distribution principle would require that wealth be transferred from richer peoples to poorer peoples in these two examples, even though the relative levels of wealth were (we are assuming) the result of each well-ordered people's free decisions. The insistence that such redistribution should nonetheless be effected can only be based in an ultimate concern for the well-being of individuals, independent of the choices made by their own political culture (and even by themselves, assuming they agree with the decisions of their political culture). This goes beyond my earlier suggestion, that cosmopolitan accounts of distributive justice are ultimately guided by a kind of luck egalitarianism, or the idea that distributions of income and wealth should not in any way reflect outcomes due to chance. It says that transfers of wealth should occur from richer to poorer without regard to either chance or choice, and should be decided purely on the basis of comparative welfare. Rawls's rejection of welfarism is integral to his rejection of a global distribution principle. In the domestic case, the end of social justice is not individual welfare, but the freedom and equality of citizens. Similarly, in the international case, the end of the Law of Peoples is not the total welfare of a people. (It is not even the welfare of its least-advantaged individuals, though all individuals' basic needs are to be met so that they can participate in the social and political life of their culture.) The end of the Law of Peoples is equal political autonomy, or "the freedom and equality of a people as members of the Society of well-ordered Peoples" (cf. LP, 118). Essential to this is that a society should be in a position to meet the basic needs of all its members so that they can participate in the social and political life of their culture. Recall that this is the basis for the duty of assistance, as opposed to a principle of distributive justice. Here again, however, cosmopolitans may object that, if not welfare, then at least the freedom and equality of individuals, and not of peoples, should be the aim of an account of international justice. But Rawls focuses on peoples rather than individuals in the global case, we have seen, because of the priority he assigns to the basic structure of society and the central role that political cooperation, political culture, and political autonomy

602

John Rawls 66

SAMUEL FREEMAN

play in his account of social justice.42 This focus is precisely the result of his concern for the freedom and equality of individuals, which is in the background throughout in The Law of Peoples. (Recall that the book's purpose is to "work out the ideals and principles of foreign policy of a just liberal people" [LP, 10, emphasis in original].) A condition of the freedom and equality of individuals, as Rawls conceives these basic democratic values, is politically autonomous citizenship within the basic structure of a democratic society that itself exercises political autonomy (that is, its citizens are "able to make their own decisions" and "able to determine the path of their own future for themselves" [LP, 118]). In the end, Rawls's rejection of a global distribution principle does not rest simply upon the assumption that a people can exercise political autonomy, that its members can be economically self-sufficient (relatively speaking) and not subject to manipulation by external forces beyond their control, and that they can control their level of wealth through savings, investment, population control, and other measures. It also rests upon his ideal conception of the freedom and equality of democratic citizens, and the social and political conditions that must hold if that ideal of the person is to be realized. 7. Finally, there is an issue that requires more discussion than I can give it here:43 Rawls envisions the difference principle as a principle that structures property institutions so as to encourage (when conjoined with fair equality of opportunity) widespread ownership and control of the means of production, either in a "property-owning democracy" or in a liberal socialist economy. Like John Stuart Mill, Rawls believed that for workers to have realistically available to them only the option of a wage relationship with capitalist employers undermines individuals' freedom and independence, blunts their characters and imaginations, diminishes mutual respect among members of different income classes, and leads to the eventual loss of self-respect among working people. For this and other reasons, Rawls was attracted by such ideas as a "share economy" (where workers have part ownership of private capital), workers' cooperatives, public provision of capital to encourage workers to become independent economic agents or to start up small businesses, and other measures for the widespread distribution of control over the means of production.44 Because there is no global basic structure, advocates of a global difference principle are required to envision the difference principle as a re42 Cf. Rawls's claim: "It is surely a good for individuals and associations to be attached to their particular culture and to take part in its common public and civic life.... This is no small thing. It argues for preserving significant room for the idea of a people's selfdetermination" (LP, 111). 43 For a more in-depth discussion of this issue, see my "Distributive Justice and the Law of Peoples/' in Envisioning a New International Order: Essays on Rawls's ''The Law of Peoples/' ed. Rex Martin and David Reidy (Oxford: Basil Blackwell, 2006). 44 See, for example, Justice as Fairness: A Restatement, 176, where Rawls endorses Mill's idea of worker-owned cooperatives as part of a property-owning democracy. On Mill, see also LP, 107n.

John Rawls THE LAW OF PEOPLES

603 67

allocation principle, where the income and wealth of more-advantaged societies are reallocated to less-advantaged peoples. But since this principle does not apply to any substantial basic structure to shape property and other economic relations, and is not conjoined with a principle of fair equality of opportunity, the allocative model of the global difference principle can do little to further Rawls's primary aims. Rawls writes: The intent [of the difference principle] is not simply to assist those who lose out through accident or misfortune (although that must be done), but rather to put all citizens in a position to manage their own affairs on a footing of a suitable degree of social and political cooperation. ... The least advantaged are not, if all goes well, the unfortunate and unlucky—objects of our charity and compassion, much less our pity—but those to whom reciprocity is owed as a matter of political justice among those who are free and equal citizens along with everyone else.45 This is not to say that the difference principle, when applied domestically, does not also have an allocative role (primarily in the form of income supplements for workers who earn too little to achieve economic independence [TJ, 252]). As Rawls makes clear, however, the difference principle specifically, and distributive justice more generally, should not be confused with measures for alleviating poverty or misfortune; nor is its purpose to assist those with special needs or handicaps, or to compensate the unfortunate for bad luck, natural inequalities, and other accidents of fortune. Any number of principles, domestic and global, can provide a decent social or global minimum and serve the role of poverty alleviation and meeting special needs. There is no need to appeal to a dysfunctional "global difference principle" for these purposes. Rawls's duty of assistance to meet basic needs is already sufficient to serve the role of addressing global poverty and special needs. The general point, then, is that Rawls does not regard distributive justice in terms of the alleviation of poverty or misfortunes; rather, he transforms the issue from a narrow question of the allocation of a fixed product of wealth in order to address a larger set of issues. Distributive justice is made part of the larger question about how to fairly structure economic and property relations among productive, socially cooperative agents, who regard themselves as free and equal, and each of whom does his or her fair share in creating the social product. In effect, Rawls incorporates the question of distributive justice into the tradition of Mill and Marx, where the primary focus is on how to fairly structure production relations in a way that affirms the dignity, freedom, and equality of socially productive agents. The robust conception of reciprocity implicit in the 45

Rawls, Justice as Fairness: A Restatement, 139.

604

John Rawls 68

SAMUEL FREEMAN

difference principle is a response to this general issue. It is not the proper response to the problem of global poverty, or to the other alleviatory issues I have mentioned (meeting handicaps and special needs, redressing misfortune, etc.). These are specific problems to address in nonideal theory, by reference to moral duties of assistance, mutual aid, and so on, and are to be determined by citizens7 democratic deliberations, on the basis of their knowledge of available resources once the demands of distributive justice are in place and satisfied. These alleviatory problems of nonideal theory raise issues that are separate from the question in ideal theory of determining appropriate standards for just distributions among socially productive democratic citizens who are cooperative members of a well-ordered society. V. CONCLUSION A central theme running through this essay is the centrality of political cooperation and political autonomy to Rawls's account of distributive justice, human rights, and the Law of Peoples. Political autonomy is essential to his idea of social cooperation and the basic structure of society. It accounts for his position regarding distributive justice as a requirement of domestic rather than global justice. Finally, political autonomy provides the basis for his account of the Law of Peoples. We would, perhaps, be making too strong a claim if we said that, for Rawls, political autonomy of individuals and well-ordered societies would not be possible if there were a global principle of distributive justice. But given the centrality of the difference principle to his account of domestic justice and the task of democratic legislation, the claim is not too far off the mark. For this reason, I conclude that the dilution of political autonomy of a (democratic) people that is required by a global distribution principle entails the dilution of the ideal of free and equal democratic citizens around which Rawls's account of justice and political liberalism is constructed. Philosophy and Law, University of Pennsylvania

Name Index Ackerman, Bruce 523 Adams, Robert M. 493, 505 Alcibiades 420 Anderson, Elizabeth 309, 311, 322, 325, 327, 338, 342 Aquinas, Thomas 423 Aristotle 121, 138, 146, 152, 155 Arneson, Richard xx, 169-72, 182-3, 187-9, 192, 234, 315-17 Arrow, Kenneth 172-3, 194 Astell, Mary 359 Audi, Robert 504-5, 507 Barry, Brian 194, 346, 389, 550, 582 Beauvoir, Simone de 345 Beitz, Charles xxiv, 545-6, 554, 582, 584, 588-9, 591-2 Benhabib, Seyla 352 Bentham, Jeremy 216, 219 Berlin, Isaiah xiv, 345, 493, 511 Blincoe, Lawrence J. 227 Blackburn, Simon 63-4, 67 Blaug, Ricardo 194 Bloom, Allan 375 Boettcher, James xxiii Bradley, F.H. 133, 153 Buchanan, Allen xxiv, 569, 582, 596 Cohen, G.A. xvii, xx, 8, 16-21, 31, 35, 108, 169-72, 182-3, 185-90, 192, 234, 282-3, 298, 322-3, 338, 342 Crowley, Brian Lee 248 Daniels, Norman xix-xx, 49, 65-6, 194 DeMarco, Joseph 194 Descartes, René 158 Dworkin, Ronald M. 27-8, 38-40, 54-5, 61, 65, 89, 150, 153, 169-70, 181, 183, 185-6, 234, 241, 312-13, 316, 323-4, 338-42, 346, 431-2, 560 Eberle, Chris xxiii Elster, Jon 534

English, Jane 355 Exdell, John 364-6, 369 Finnis, John 567 Freeman, Samuel xxiv Frege, Friedrich 52 Friedan, Betty 345 Friedman, Marilyn 469 Frost, Robert 561 Galileo 200 Galston, Miriam 469, 478 Galston, William 108-9, 111 Gaus, Gerald 469, 513, 515-17, 519, 521, 523, 529 Gauthier, David 108 Gilligan, Carol 351-2 Gödel, Kurt 52 Gouges, Olympe de 359 Greenawalt, Kent 525, 527 Habermas, Jürgen xxii, 234, 250-51, 439, 442, 451-64, 469, 474, 476 Hare, R.M. 567 Hart, H.L.A. xiv, 91, 194 Hegel, Georg W.F. xi, xii, xix, 133-58, 251, 383-4, 454 Hobbes, Thomas xi, 136, 144, 151 Holmes, Justice 368 Hospers, John 119 Hume, David 120, 158 James, Aaron xvii-xviii Jefferson, Thomas 456 Kant, Immanuel xi, xix, 47, 136, 143-4, 146-7, 154, 203, 205-7, 352, 396, 443, 472-3, 562 Kearns, Deborah 356, 358, 361, 374 Keat, Russell 194 Kekes, John 108, 119-22 Kelly, Erin 478 King Jr. Martin Luther 501 Knox, John 420 Kohlberg, Lawrence 351-2

606

John Rawls

Kronman, Anthony 12 Kymlicka, Will xxi, 312, 314 Larmore, Charles xxii, 502-3 Laslett, Peter 379 Lincoln, Abraham 373, 422 Locke, Johnxi, 137, 144, 151, 155, 396, 419, 442-3, 503, 575 Lukes, Steven 346 McClain, Linda 350-51, 353-4 MacIntyre,A. 134, 139, 148, 154 McPherson, Lionel 478 McTaggart, J.M.E. 133 Malcolm, Norman xiii Mandle, Jon xviii Martin, Rex xxi, 261, 264, 271-3, 275, 279-81 Marx, Karl xi, 137, 156-7, 239, 242, 369, 603 Mill, John Stuart xi, 137, 215, 359-60, 419-20, 443, 602-3 Miller, David 194, 556 Millett, Kate 345 Moore, G.E. 133, 158 Murphy, Liam xvi-xvii Murphy, Jeffrie G. 499 Nagel, Thomas 10-11, 17-18, 22, 36-7, 238, 241-3, 389 Neurath, Otto 143 Nielsen, K. 45, 49, 51 Nozick, Robert xi, xvii, 8, 78, 108, 279, 567, 569-70, 594 Nussbaum, Martha 350, 354, 367-70, 588 Okin, Susan Moller xxi, 20 Olsen, Fran 3 59-60 Parfit, Derek 15 Pateman, Carole 359 Patón, H.J. 206 Pericles 420 Pogge, Thomas xx, xxiv, 23-8, 33, 40, 114, 545-6, 553, 582, 588, 590, 594, 598 Quinn, Philip L. xxiii, 518-19 Rae, Douglas 273 Rakowski, Eric 324 Rawls, John passim Raz, Joseph xvii, 235-6, 246-7, 273-8, 535, 567

Reidy, David 520, 526-7, 529-30, 532 Richards, David 528, 582 Richmond, Samuel 194 Roemer, John 234 Ross,W.D. 119 Rousseau, Jean-Jacques xi, xii, 137, 359, 392 Russell, Bertrand 133 Sandel, Michael 108-10, 113-14, 116, 134, 1435, 148-9, 154, 351-2, 372, 375, 389-91 Scanlon, T.M. 98, 482 Schwartz, Adina 238-43 Schwarzenbach, Sibyl xix Schwartzman, Micah xxiii Sen,Amartyaxx, 170-75, 179-80, 188-9, 192, 234, 280-81 Shaw, George Bernard 359 Sheffler, Samuel xx Sher, George 108 Shue, Henry 194 Sibley, W.M. 472-3, 483 Sidgwick, Henry xi, xii, 216 Singer, Peter 567 Smith, Adam 155 Socrates 121, 140, 200 Stace, Walter xiii, xiv Stout, Jeffrey xxiii Sunstein, Cass 531 Tan, Kok-Chorxxiv, 582, 584, 598 Taylor, C. 134, 146, 149, 152, 154 Taylor, Robert xx Tesón, Fernando xxiv Tomasi, John 218 Ullmann-Margalit, Edna 531 Urmson, J.O. xiv Waldron, Jeremy xxii, 116, 521 Wall, Steven xxi Walzer, Michael 134, 139, 148, 556 Weithman, Paul 512 Williams, Andrew D. xxi, 533, 535 Wolff, Jonathan 325 Wollstonecraft, Mary 359-60 Wolterstorff, Nicholas xxiii, 494, 503-11 Young, Iris Marion 352 Zaitchik, Alan 108, 110