PERFECTIONISM A N D NEUTRALITY
This Page Intentionally Left Blank
PERFECTIONISM AND N E U T R A L I T Y Essays in L...
51 downloads
1229 Views
19MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
PERFECTIONISM A N D NEUTRALITY
This Page Intentionally Left Blank
PERFECTIONISM AND N E U T R A L I T Y Essays in Liberal Theory
Edited by Steven Wall and George Klosko
-
ROWMAN & LITTLEFIELD PUBLISHERS, INC.
Lanham
*
Boulder
New York
Oxford
ROWMAN & LITTLEFIELD PUBLISHERS, INC. Published in the United States of America by Rowman & Littlefield Publishers, Inc. A Member of the Rowman & Littlefield Publishing Group 4501 Forbes Boulevard, Suite 200, Lanham, Maryland 20706 www.rowmanlittlefield.com PO Box 317 Oxford OX2 9RU, UK Copyright 0 2003 by Rowman & Littlefield Publishers, Inc.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior permission of the publisher. British Library Cataloguing in Publication Information Available Library of Congress Cataloging-in-PublicationData Perfectionism and neutrality : essays in liberal theory / edited by George Klosko and Steven Wall. p. cm. Includes bibliographical references and index. paper)-ISBN 0-7425-0844-7 (pbk. : alk. paper) ISBN 0-7425-0843-9 (a. 1. Political ethics. 2. Liberalism-Moral and ethical aspects. I. Klosko,George. 11. Wall, Steven, 1967BJ55 .P33 2003 172'.2-dc2 1
2002036604
Printed in the United States of America
WMThe paper used in this publication meets the minimum requirements of American National Standard for Information Sciences-Permanence Materials, ANSUNIS0 239.48- 1992.
of Paper for Printed Library
Contents
vii
Preface Introduction
1
Part One: Selections from Published Works A. Neutrality “Liber&m,” in Public and Private Morality, selections Ronald Dworkin Social Justice in the Liberal State, selections Bruce Ackerman Patterns of Moral Complexity, selections Charles Larmore Liberal Virtues, selections Stephen Macedo Political Liberalism, selections John Rawls Equality and Partiality, selections Thomas Nagel
B. Perfectionism 7 Equality, Liberty and Perfectionism, selections Vinit Haksar 8 The Morality of Freedom, selections Joseph Raz -v-
31 41 47 61 67 77
91 103
Contents
vi
9 Perfectionism, selections
Thomas Hurka
125
Part Two: Original Papers 10 “Liberal Neutrality: A Compelling and Radical Principle” Gerald F. Gaus 11 “Reasonable Rejection and Neutrality of Justification”
137
12
191
13 14 15
George Klosko “Liberal Neutrality on the Good An Autopsy” Richard J. Arneson “Freedom of Expression in the Non-Neutral State” George Sher “The Structure of Perfectionist Toleration” Steven Wall “Religion, Law, and Politics: Arenas of Neutrality” Kent Greenawalt
167
219 23 1 257
Further Reading
28 1
Index
285
About the Contributors
289
Preface
0
VER THE LAST QUARTER-CENTURY, the concepts of neutrality and perfec-
tionism have been central to liberal political theory. Claims that the state should not act on controversial views of the good, but should be neutral among them, are strongly defended by some theorists. They insist that state neutrality, once it is adequately understood, is a central ideal of liberal politics. Supporters of perfectionism, by contrast, argue that governments have a duty to act on the basis of sound views, whether they be controversial or not. They insist that perfectionist politics is not inherently opposed to liberal values. In this volume, we provide both a comprehensive introduction to the subject and new contributions by prominent theorists. The volume is divided into two parts. In part I, we collect the most influential contributions to the debate published to date. Detailed discussion of these pieces and their place in the neutrality-perfectionism controversy are provided in the introduction. In part 11,we present six original chapters, each of which addresses key aspects of the debate as it currently stands. By discussing a range of subjects from different points of view, these chapters clarify the main issues that continue to divide participants in this controversy. Acknowledgments
The idea for this volume originated in a panel on liberal neutrality at the 2000 meeting of the American Political Science Association, in Washington, D.C. In -Vll-
...
Vlll
Preface
addition to the editors, the participants were William Galston, Anne Manuel, and George Sher. We are grateful for their participation and for subsequent conversations and advice. We wish to thank Richard J, Arneson, Gerald F. Gaus, Kent Greenawalt, and George Sher for their fine contributions to this volume. We acknowledge our gratitude to the authors and to their publishers. The selections we include are drawn from the following works: Ronald Dworkin’s “Liberalism,”in Public and Private Morality; Bruce Ackerman’s Social Justice in the Liberal State; Charles Larmore’s Patterns of Moral Complexity; Stephen Macedo’s Liberal Virtues; John Rawls’s Political Liberalism; Thomas Nagel’s Equality and Partiality; Vinit Haksar’s Equality, Liberty and Perfectionism; Joseph Raz’s The Morality of Freedom; and Thomas Hurka’s Perfectionism.
Introduction
R
interest in the role of “neutrality” and “perfectionism” in liberal political theory. Theorists generally use the former term to refer to the moral principle that, in a pluralistic society, the state should not take sides between different citizens’ moral, religious, and philosophical views, or, as this is generally referred to, between their conceptions of the good. The neutrality principle was first clearly enunciated in the 1970s in important works by Ronald Dworkin, Bruce Ackerman, and other writers. Great claims have been made for the idea. According to Dworkin, neutrality is the “nerve of liberalism”; and in his pioneering work on the subject, Ackerman describes it as “the organizing principle of liberal thought.” Perfectionism, by contrast, is the idea that the state should favor particular moral ideals. The state should not strive to be neutral between conceptions of the good, but should promote valid or sound conceptions of the good and discourage worthless ones. Perfectionism arose largely in reaction to claims made for neutrality, with its proponents arguing either that neutrality is an unworkable principle and/or that the state has a duty to promote good activities and discourage bad ones. Important recent proponents of perfectionism include Joseph Raz, Thomas Hurka, and George Sher. By way of introducing this collection of chapters, we will review the history of neutrality and perfectionism and lay out the important theoretical issues that bear on the debate between them. ECENT YEARS HAVE SEEN CONSIDERABLE
-1-
2
Introduction
Liberal Neutrality: A Historical Overview On a general level, “neutrality” is a familiar idea that is applicable in many different contexts. The concept gains its sense from an agent confronting a dispute situation involving at least two other agents or entities.’ Thus, in international relations, Switzerland was neutral during World War 11, because it did not take sides in the conflict between the Allies and their foes. In the context that concerns us, the modern state is faced with the decision of whether it should strive to be neutral between the different conceptions of the good2 that are (or may be) held by persons within its jurisdiction. The interest in state neutrality arises from the circumstances of contemporary pluralistic societies. Citizens in these societies hold different moral and religious views. In response to this fact, some have insisted that the state should not favor the views of certain citizens over others. To a certain extent, this is clearly implied by the important liberal norm of equal treatment. Just as one group of people should not be treated better than others because of their race, ethnic background, or the social class into which they are born, so too, it has been thought, individuals should not receive preferential treatment because of their moral or religious views. Understanding what it means for the state to be neutral between the moral and religious views of its citizens, however, has proved to be a difficult task. Different theorists have characterized state neutrality in different terms. There is no single conception of state neutrality that commands assent among proponents of the idea. Accordingly, defenses of state neutrality are in part efforts to articulate the concept in a satisfactory way and in part efforts to explain why the concept so articulated should play a prominent role in political philosophy. In the next section, we will outline some of the important theoretical decisions that must be made by anyone who wishes to defend the concept of state neutrality. But first in this section, we will briefly review the historical development of this concept. This overview should provide a sense of the diversity of ideas associated with state neutrality as well as introduce the early papers on the topic that are reprinted in this volume. Neutrality emerged as an important topic in political philosophy in the 1970s with the publication of John Rawls’s A Theory ofJustice. Although in this book Rawls neither explicitly uses the term “neutrality” nor articulates a principle of state neutrality, he argues for principles of justice that are chosen under circumstances that are designed to rule out perfectionist conceptions of justice. The principles are chosen by representative individuals in a hypothetical “original position,” behind a “veil of ignorance.” The parties in the original position are deprived of particular knowledge about
Introduction
3
themselves concerning their position in society, their race and sex, and their talents and abilities. In addition, Rawls argues that they should not know the content of their conception of the good. Excluding knowledge of conceptions of the good in this way, in effect, carries with it a commitment to state neutrality since the parties in the original position have no motivation to opt for principles of justice that would favor some conceptions of the good over others. This commitment to state neutrality is evinced in Rawls’s explicit rejection of perfectionist public policies. State action designed to promote excellence is not permitted by his principles of justice. As he writes: “[Tlhe principles of justice do not permit subsidizing universities and institutes, or opera and the theater, on the grounds that these institutions are intrinsically valuable, and that those who engage in them are to be supported even at some significant expense to others who do not receive compensating benefit^."^ If any such state action is justifiable, Rawls maintains, the funds that support it must be unanimously (or nearly unanimously) approved. This strict requirement rules out almost all perfectionist policies, since in a pluralistic society there is little hope of securing unanimity on any proposed public policy. The trend toward neutrality is also evident in Robert Nozick’s similarly important work Anarchy, State and Utopia, which was published three years after A Theory oflustice. Nozick insists that a state that respects the rights of its citizens must be “scrupulously neutral” between them. However, it was with the publication of Dworkin’s essay “Liberalism” in 1978 that the idea began to gain widespread attention. Although his main concern in this essay is not liberal political philosophy per se, but rather, more narrowly, the political views espoused by those who can be identified as “liberals,” Dworkin gives a clear account of state neutrality. He identifies as the “constitutive political morality” of liberalism the requirement that “government must be neutral on what might be called the question of the good life”: “Since the citizens of a society differ in their conceptions [of what gives value to life], the government does not treat them as equals if it prefers one conception to another, either because the officials believe that one is intrinsically superior, or because one is held by the more numerous or more powerful group”4 ( 191). Thus, according to Dworkin, a government must be neutral between the conceptions of the good of its citizens if is to treat them as equals. Ackerman builds on the connection between equality and neutrality in his influential book Social Justice in the Liberal State. Ackerman claims that central to the liberal tradition is “an effort to define and justify broad constraints” on justifications for the use of power. Neutrality is just such a constraint. It rules out all appeals to “privileged moral authority.” More
4
Introduction
precisely, the neutrality constraint holds that “no reason is a good reason if it requires the power holder to assert: (a) that his conception of the good is better than that asserted by any of his fellow citizens, or (b) that, regardless of his conception of the good, he is intrinsically superior to one or more of his fellow citi~ensl’~ Interestingly, Ackerman refuses to be pinned down on how this constraint is best defended. He claims that there are a number of good reasons in support of neutrality and does not choose between them. Indeed, according to Ackerman, efforts to defend neutrality that begin by assuming the truth of one moral outlook-such as the moral outlook of neo-Kantian egalitarianismbetray the very neutrality they are striving to defend. The task for the neutrality theorist is to identify and call attention to the different “argumentative paths” that lead to neutrality without assuming that any one of them is intrinsically better than the others. While differing in significant ways from Ackerman’s approach, a similar concern with the proper justification of neutrality is present in Charles Larmore’s important book Patterns of MoraI Complexity. Larmore describes the ideal of state neutrality as “the distinctive liberal notion,” “the central ideal of the modern liberal state.”6 According to his construal, the requirement is that the state “not seek to promote any particular conception of the good life” because of its intrinsic superiority. Importantly, Larmore argues that neutrality does not pertain to the results of state action, but rather to the considerations that are relied on to justify political decisions. Like Ackerman, Larmore is concerned that nonneutral justifications of neutrality betray the ideal. Accordingly, he rejects defenses of neutrality based on skepticism and a strong conception of autonomy as unacceptably controversial. In their place, he attempts to provide “a neutral justification of neutrality.” At root, the key to his own justification is a norm of equal respect. In order to be faithful to the Kantian injunction to treat people as ends and not merely as means, Larmore argues that we must recognize that they have their own points of view and perspectives on the world. When our actions affect them, we should acknowledge their perspectives and so provide reasons for our actions that are acceptable from their points of view. Larmore’s effort to provide a neutral justification for neutrality illustrates the connection between state neutrality and what is now commonly termed “public justification.’’This is the idea that the state’s use of coercive authority must be able to be justified openly and publicly to all its subjects. In a society in which people subscribe to many different conceptions of the good, this is
Introduction
5
possible only if the state acts on principles that are not bound to the ideals of any particular group, but can be defended on grounds that are neutral between them. The requirement of public justification thus carries with it a commitment to state neutrality. This connection between public justification and state neutrality is picked up and substantially developed in the recent works of Rawls and Thomas Nagel. For example, in Political Liberalism, Rawls departs from some of the views he expressed in A Theory ofJustice and illustrates the connection between public justification and state neutrality by introducing a distinction between a comprehensive doctrine and a political conception of justice. A comprehensive doctrine “includes conceptions of what is of value in human life, as well as ideals of personal virtue and character, that are to inform much of our nonpolitical conduct.” Such doctrines are controversial. Reasonable people can and do disagree over which ones are sound. A political conception of justice, by contrast, is noncontroversial in the sense that it is worked up from shared fundamental ideas embedded in the public political culture of a democratic society. As such, a political conception of justice is independent of any particular comprehensive doctrine. For this reason, Rawls argues that it is possible for a society that is governed by a political conception of justice to be neutral in the sense that its major institutions are not designed to favor any particular comprehensive doctrine. Likewise, Nagel in Equality and Partiality insists that a liberal political society is one in which the coercive power of the state is justifiable to each person who is subject to it. This requires the liberal to draw a distinction between the “values a person can appeal to in leading his own life and those he can appeal to in justifylng the exercise of political power.”’ The former values must be excluded from justifications for the exercise of political power if those justifications are to be acceptable to people who adhere to a range of different moral and religious views. Of course, as Nagel acknowledges, liberal principles are not compatible with every actual religious and moral view, and the liberal aspiration to justify the coercive power of the state to everyone who is subject to it may not be achievable in our present circumstances. Nonetheless, the liberal is committed to a particular ideal of unanimity in justifymg the exercise of political power. This is “a unanimity which could be achieved among persons in many respects as they are, provided they were also reasonable and committed within reason to modifying their claims, requirements and motives in a direction which makes a common framework of justification possible.”s If such a common framework of justification were indeed possible, then the liberal state could be justified in a way that is neutral between the different religious and moral views that divide its citizens. By emphasizing the need to develop such a framework, Nagel follows Rawls in situating the commitment to state
6
Introduction
neutrality within a larger contractualist approach to moral and political philosophy. The Neutrality Principle We have been tracing the history of the idea of neutrality as it emerged in liberal political philosophy over the past three decades. Not surprisingly, as this idea gained currency in philosophical discussion, it provoked a barrage of criticism. We shall discuss some of the more important criticisms that have been directed against it shortly; but first we need to identify more precisely the different formulations of the neutrality principle that have been proposed and the different justifications that have been offered in its defense. As mentioned earlier, the neutrality principle is commonly understood to be the moral principle that in a pluralistic society the state should not favor or take sides between different citizens’moral, religious, and philosophical views, or as this is generally referred to, their conceptions of the good. However, stated in these general terms, the principle is subject to a variety of rival interpretations. Two issues in particular need to be clarified. The first issue concerns the scope of the neutrality principle. The term “state” can refer to the fundamental constitutional structure of a society (roughly what Rawls calls the “constitutional essentials”), or it can refer simply to the government of such a society. With this in mind, it is useful to distinguish a comprehensive neutrality principle from a narrow one. The narrow principle holds that it is only the constitutional structure or basic framework of the state that should be neutral between different citizens’ conceptions of the good. So long as this structure is neutral, state policies need not be. By contrast, the comprehensive principle holds that both the constitutional structure and state policies should be neutral with respect to citizens’ conceptions of the good. The motivation for the narrow principle is that the constitutional structure of a society has an enormous impact on the ability of persons to pursue their conception of the good. If it were nonneutral, then the fact that other, less fundamental state policies are neutral would provide little comfort to those whose conception of the good was disfavored. The converse, however, is not true. One might think, as several contemporary proponents of the neutrality principle appear to believe, that if the constitutional structure of the state is neutral then it is not necessary for less fundamental state policies to be neutral. To demand state neutrality here is to demand too much. Still, it is not uncommon for defenders of state neutrality to insist that all state policies be neutral. Restricting the scope of the neutrality principle to the constitutional structure, they argue, is an arbitrary limitation. They point out
Introduction
7
that even when the constitutional structure of a state is neutral it can still exercise immense power to advantage some and disadvantage other conceptions of the Proponents of state neutrality, then, divide between those who embrace the comprehensive neutrality principle and those who think that narrow neutrality is sufficient. One further point about the scope of the neutrality principle should be mentioned. Many writers on neutrality stress the coercive nature of the state. But while the state is a coercive institution, it is not true that every exercise of political power is an exercise of coercion. This raises an important question. Should the neutrality principle apply only to the coercive exercise of political power or should it apply to noncoercive exercises as well? The question is pressing, since (as we explain later on) some contemporary perfectionists believe that state coercion is not in general an effective means for promoting the good. Accordingly, the policies favored by these perfectionists may be acceptable to those who defend the neutrality principle, but limit its application to the coercive exercise of political power. The second issue that needs clarification concerns how to understand the idea that the state should not favor any particular conception of the good. Depending on how the verb “favor” is construed, we will get different formulations of the neutrality principle. Here, a key question is whether to favor a conception of the good the state must intend or aim to do so. For example, suppose the state introduced and enforced a requirement that all schools teach English and suppose that this had the effect of making it more difficult for non-English-speaking citizens to pursue their favored conception of the good. If the state did not intend this nonneutral effect, and if it had good neutral reasons for enforcing the requirement, then is it right to say that it did not favor the conceptions of the good of its English-speaking citizens? Many think that the answer is “yes.”’o But others disagree. Why, they ask, should we be so concerned with the intentions or aims of state officials? Not only are these intentions and aims often unknown, but also state officials often have a variety of motives for supporting the policies they support. In the light of this, some writers have suggested that we shift the focus from the intentions of state officials to the consequences of the policies they support.’ The difficulty of ascertaining the actual aims or intentions of state officials has led others to argue that a state is neutral if and only if the justifications that it offers in defense of what it does do not claim or entail that some conceptions of the good are superior to others. On this understanding, the requirement that the state not favor any conception of the good simply rules out various arguments or considerations for justifying state action. Since justifications for state action are public and open to inspection, we need not worry about discovering the true aims behind what the state is doing. Thus, while
8
Introduction
some defenders of neutrality have claimed that the state must not aim to favor any particular conception of the good, we do well to construe this as the claim that the state should not justify what it does by invoking controversial claims about the worth or value of different conceptions of the good. We can call this view “neutrality of justification.” It formulates the neutrality principle as follows:
Neutrality of justification: The state should not aim to do anything to promote any particular conception of the good, or give greater assistance to those who pursue it, unless a plausible neutral justification can be given for the state’s action. Neutrality of justification has been the dominant formulation of the neutrality principle. Most of the writers discussed in the previous section have endorsed it in one version or another, and several of the new chapters in this book proceed on the assumption that it is clearly the correct formulation. However, an important alternative formulation of the neutrality principle is possible and it should not be dismissed without argument. This view holds that the state can favor a conception of the good even if it does not aim to do so and even if a plausible neutral justification can be given for what it does. If this is right, then neutral justifications for state action may be necessary, but are not sufficient, for state neutrality. We can call this view “neutrality of effect.” It formulates the neutrality principle as follows:
Neutrality of effect: The state should not do anything that has the effectwhether intended or not-of promoting any particular conception of the good, or of providing greater assistance to those who pursue it. If not interpreted carefully, neutrality of effect is clearly an impracticable principle. It is impossible for a state to ensure that each person subject to its authority has an equal chance to pursue and realize his or her conception of the good. There is, however, a way of interpreting the principle that makes it workable. The proponent of neutrality of effect can hold that if the state acts in a way that has the effect of disadvantaging a particular conception of the good, then the complaint that this would be objectionably nonneutral cannot be defeated simply by establishing that a plausible neutral justification can be given for the state’s action. Perhaps, those who have been disadvantaged by the nonneutral effects of the state’s action are entitled to compensation. If so, then this is something that is captured by neutrality of effect, but not by neutrality of justification. The difference between the two formula-
Introduction
9
tions of the neutrality principle, then, boils down to whether or not the nonneutral effects of the state’s neutrally justified actions should in themselves generate a claim to compensation on the part of those who have been disadvantaged by these effects. The possibility that the neutrality principle might require this kind of compensation raises an interesting, and insufficiently discussed, question. Might the neutrality principle require a government to take measures designed to favor a particular conception of the good if those measures were necessary in order to compensate for nonneutral effects that disadvantaged that conception of the good and were the result of the maintenance of a neutrally justified constitutional structure or a neutrally justified governmental policy? If the answer is yes, then difficult questions arise about the kind of compensation that would be appropriate. A state might grant a group special rights or exemptions if doing so would help it prosper and if the group’s ability to survive had been set back by prior neutrally justified state action. In other cases, monetary compensation might be appropriate to compensate persons for disadvantages they confront in pursuing their conception of the good. Proponents of neutrality of effect would need to develop a theory of compensation if their understanding of the neutrality principle were to be workable; but, in principle, there is no reason to think such a theory could not be developed. Our discussion of this issue has been abstract. But the difficulty of identifying the correct formulation of the neutrality principle also can be glimpsed by considering briefly how judges have attempted to apply the idea in practice. Neutrality is a familiar idea in American constitutional law, although in this area, the idea has been formulated less sharply than in the philosophical literature. As interpreted in a series of U.S. Supreme Court decisions, the guarantees of religious freedom enshrined in the First Amendment were taken to mean that the government must be “neutral” between religions. Neutrality between religions-and between religion and nonreligion-is clearly proclaimed in a series of cases. For instance, a strong statement is found in Justice Thomas Clark’s opinion in a 1963 case that concerned mandatory Bible reading in public schools.l 2 Defending the “inviolable citadel of the individual heart and mind,” Clark wrote: “We have come to recognize through bitter experience that it is not within the power of government to invade that citadel, whether its purpose or effect be to aid or oppose, to advance or retard. In the relationship between man and religion, the state is firmly committed to a position of neutrality.”13 Justice Clark‘s mention of the state’s “purpose or effect” appears to encompass both neutrality of justification and neutrality of effect. Many other decisions argue similarly. According to the now widely accepted “threepronged test of constitutionality of government accommodations to religion, for a measure to pass muster: (1) there must be a primary secular legislative
10
Introduction
purpose; (2) its effect must be neither to advance nor to inhibit religion; and (3) its result must not be an “excessive entanglement” of government and religion. The first and the second points appear to coincide with neutrality in the two senses we have distinguished. Thus, neutral justification is not enough. The effects of a given law must also be taken into account.14But the courts have not explained in a systematic way how we are to know when an effect of a measure advances or inhibits religion and they have not considered the possibility of compensating for these effects. It remains unclear, then, which formulation of the neutrality principle the courts have adopted. A final matter is worthy of discussion. We have said nothing so far about the kind of moral requirement generated by the neutrality principle. Assuming that one could identify its correct formulation and properly fix its scope, one still would need to understand the nature and strength of the requirement that it imposes on the state. Most writers have spoken of the neutrality principle as a constraint on legitimate state action. Neutrality is not simply one consideration among many that the state should consider before undertaking a course of action, but rather it is a constraint on how the state may justifiably act at all. But this is not entirely satisfactory, since constraints can be strong or weak. At the limit, we can view the neutrality principle as generating an absolute constraint on state action that no other consideration could override. This understanding of the constraint has the virtue of clarity, but it is not very plausible. As Larmore observes, neutrality “is not the liberal’s only desideratum.”15 It may need to be compromised for the sake of other values. Yet, if one says this, then what is to guide one in the compromise?If neutrality can justifiably be set aside for the sake of some other liberal desideratum, then it is not in general true that a legitimate liberal state must be a neutral state. So long as the values that weigh against neutrality are weighty enough, then the liberal state may legitimately depart from its commitment to neutrality. Given the importance of this issue, it is surprising that proponents of the neutrality principle have not had more to say about the strength of the constraint it imposes. Perhaps, however, precision is not to be expected in this area. Rawls claims, for example, that the values of a political conception of justice-which, as we have seen, include a commitment to the neutrality principle-“are very great values and hence not easily overriddenl’16This suggests that the moral constraint imposed by the neutrality principle is very strong, but overrideable in exceptional circumstances. This understanding coheres well with the statements quoted earlier to the effect that neutrality is “the nerve of liberalism” or “the fundamental liberal principle.” It is worth noting, however, that it is possible-even if no proponent of state neutrality has adopted this positionto accept the neutrality principle, but insist that the constraint it imposes on the state is easily overridden. Such a view would not be very interesting; but its possibility shows that proponents of state neutrality must do more than defend the
Introduction
11
neutrality principle. They must establish that the constraint it imposes on the state is stringent. Justifying Neutrality The scope, formulation, and strength of the neutrality principle cannot be defined without reference to the values that justify it. This naturally leads to the question of how the principle is best justified. As we already have seen, some proponents of the principle believe that it must be justified in a way that does not betray its spirit. The justification for the neutrality principle, they argue, must itself be neutral. Such a neutral justification for neutrality might be accomplished in one of two ways. Following Ackerman, one might try to show that there is a plurality of “argumentative paths” that all lead to the same neutrality principle. For this reason, one need not choose between them. One can remain neutral on the question of which particular argumentative path best justifies the principle. We can call this the “ecumenical approach” to justifylng neutrality. Alternatively, following Larmore, one might try to show that the neutrality principle can be derived from certain minimal substantive moral considerations that are common ground between those who adhere to different conceptions of the good. Considerations relating to political stability and the norm of equal respect have been invoked to play this role. We can call this the “deductive approach.” Not every proponent of state neutrality, however, believes that the neutrality principle can and should be given a neutral justification. In a perceptive critique of the ecumenical approach, Jeremy Waldron argues that different justificatory grounds support different conceptions of neutrality, which may not sit well together. He makes the point that the meaning of a concept depends on its underlying rationale. In applying a concept like neutrality in troublesome cases, one must appeal to this rationale. Describing Ackerman’s strategy as “promiscuous across an array of different justifications,”Waldron writes: “The proponent of liberal neutrality cannot afford to be negligent about the task of justifying the position she wants to embrace . . .because justifying it is part and parcel of the task of articulating it.”” If Waldron is right, then neutrality must be defended on particular grounds, which in turn will result in the establishment of a particular conception of neutrality, In a similar vein, but for different reasons, others have argued that the deductive approach also fails to provide a neutral justification for neutrality. The neutrality principle cannot be derived from minimal substantive moral considerations. In pluralistic societies, these critics maintain, there is simply insufficient common moral ground from which to derive the neutrality principle. Efforts to do so invariably end up smuggling controversial moral claims into the justification.
12
Introduction
In assessing such criticisms, it is important to tread carefully. No clearheaded proponent of state neutrality has ever claimed that the neutrality principle is itself morally neutral. As a moral principle, it is plainly incompatible with moral ideals that reject it. This should surprise no one. The aspiration to provide a neutral justification for neutrality is not the obviously misguided hope that one can derive a moral principle from nonmoral considerations. Rather, the idea behind the aspiration is that because state neutrality is a response to the fact of ethical disagreement about the good life we should not try to justify it in a way that appeals directly to the truth of some controversial conception of the good life. We should instead-as much as it is possible to do so-strive to justify neutrality by appealing only to shared ideals and values. Yet, even understood in this more modest way, the attempt to provide a neutral justification for the neutrality principle has looked unpromising to many. This has suggested to some proponents of the principle that it is best justified in a nonneutral manner. For example, in his recent writings, Dworkin now attempts to derive the neutrality principle from a particular (and controversial) conception of the good life. This conception, which Dworkin terms “the challenge model of ethics,” holds that a good life consists in the skillful managing of a challenge. As Dworkin puts it, “a good life, on this view, has the kind of value a brilliant dive has and retains when the ripples have died away? This conception of a good life is controversial, but it is compatible with a wide range of different substantive conceptions of the good. There is a plurality of ways in which one can lead a life that skillfully manages the challenges one assigns to oneself or views oneself as confronting. The details of Dworkin’s “model of ethics,” however, are not our concern here.19 What is of interest is the justificatory strategy that his argument instantiates. The neutrality principle is understood to be a “theorem” that is derived from a particular understanding of the good life. Moreover, this understanding of the good life is not taken to be shared or noncontroversial or reasonably acceptable to all persons. Working with the terminology previously introduced, Dworkin’s argument is an instance of the deductive approach to justifying neutrality, but one in which the neutrality principle is derived not from shared, minimal moral commitments, but from robust, controversial considerations. Diagram 1 summarizes the different strategies for justifying the neutrality principle that we have been discussing. Ecumenical Approach
Deductive Approach
/ \
From shared moral commitments Diagram 1
From controversial moral commitments
Introduction
13
Some writers have attempted to combine elements of both approaches. In Political Liberalism, for example, Rawls not only attempts to derive a political conception of justice from the shared ideas embedded in the public political culture of a democratic society (an instance of the deductive approach from shared moral commitments), but also labors to show that this conception can be accepted for different reasons by persons with different philosophical, moral, and religious commitments (an instance of the ecumenical approach). Here, we will not attempt to determine which of these justificatory strategies has the most promise. We distinguish them to give a sense of the complexity of arguments for state neutrality. As noted earlier, the understanding of the scope, formulation, and stringency of the neutrality principle will depend in no small measure on which justificatory strategy is adopted. The proponent of the neutrality principle must therefore strive to achieve consistency between the particular understanding of the principle he or she proposes and the underlying values he or she identifies as providing its rationale. Perfectionism and Politics Having clarified state neutrality, we now turn to perfectionism in politics. Perfectionism is sometimes identified with the view that the state should “maximize the achievement of human excellence in art, science and culture.”20This characterization, however, is doubly misleading. First, it identifies perfectionism with a particular substantive account of excellence: the development of art, science, and culture. But different perfectionist theories have understood excellence in different terms. Second, it identifies perfectionism with a maximizing morality; but perfectionists need not think that the state should strive to maximize the good. We need, therefore, a more general characterization of perfectionism. All perfectionists hold that the state should favor valuable conceptions of the good. More precisely, they hold that the state should promote excellence and/or assist its citizens in their efforts to lead worthwhile lives, even if doing so requires it to undertake political action that is reasonably controversial. Understood in these general terms, perfectionism is not a novel view. Indeed, as one commentator writes, “if one takes a long view of the development of Western political thought perfectionism seems to be the standard view of the state.”21Notwithstanding this point, many contemporary writers sympathetic to perfectionism have defined themselves largely in opposition to state neutrality. Such a posture is evident in the work of Vinit Haksar, an early perfectionist critic of Rawls. According to Haksar, state neutrality should be rejected because
14
Introduction
it is inconsistent with moral equality and because it exaggerates the significance of disagreement in determining which principles of political morality are appropriate. He presses the first point by calling attention to the fact that moral egalitarians need to identify the properties of beings that entitle them to moral respect. Such an exercise, he insists, is necessarily perfectionist, since it involves ranking some beings as having more intrinsic worth than others. He presses the second point by claiming that Rawls excludes perfectionist standards from his theory of justice on the grounds that the application of such standards leads to disagreement. But this fails to distinguish perfectionist standards from nonperfectionist standards that are also subject to disagreement. Haksar concludes, “[Iln any case, it is important to realize that the extent to which a principle commands general acceptance is only one of the determinants of how good and important the principle is as a moral and political principle.”22 Haksar’s suspicion that antiperfectionist liberalism rests on perfectionist presuppositions that it neither acknowledges nor defends was a common theme in many of the early criticisms of state neutrality. Critics claimed that proponents of state neutrality violate its own constraints by appealing to particular values, including the values of a liberal society. Democracy, a respect for individual rights, and a measure of social justice carry with them particular value commitments that are not neutral between all conceptions of the good. As Stephen Macedo stresses, “Liberal principles and goals shape our lives pervasively, deeply, and relentle~sly.”~~ Even if these principles do not directly determine our choices, they certainly limit the range from which we can choose. Given this, the professed neutrality of a society governed by the neutrality principle appeared to many to be a myth. Despite the rhetorical force of this early critique of state neutrality, perfectionism came into its own only with the publication of Raz’s The Mordity of Freedom. Importantly, Raz called attention to the logical gap between moral pluralism and state neutrality. The former view “asserts the existence of a multitude of incompatible but morally valuable forms of life,” but it “allows that certain conceptions of the good are worthless and demeaning, and that political action may and should be taken to eradicate or at least curtail them.”24 Since the recognition of moral pluralism often had been thought to provide support for state neutrality, Raz’s articulation of a perfectionist political morality that was compatible with moral pluralism posed a strong challenge to defenders of state neutrality. Raz also called attention to the logical gap between perfectionist political action and the coercive imposition of a controversial way of life. The latter is but one way by which the state might promote the good; but much perfectionist political action need be neither coercive nor controversial. As Raz observes, perfectionist political action may consist in noncoer-
Introduction
15
cively encouraging valuable conduct and discouraging disvaluable conduct. And since the justification of coercive perfectionist measures raises special concerns, it is a mistake to identify it too closely with perfectionism in general. In this way, by outlining an attractive account of perfectionism, one that was compatible with moral pluralism and sensitive to the dangers of coercively promoting the good, Raz showed how perfectionism in politics could be reconciled with many of the commitments that motivate support for state neutrality. In particular, Raz’s work suggested that the contrast between a neutral state and a state that enforced a controversial mode of life was misleading. The perfectionist state promotes the good, but this does not imply that it must be grounded in a unitary comprehensive conception of the good life.25 Raz’s defense of perfectionism in politics was followed by a number of book-length studies defending versions of perfectionism. These included Thomas Hurka’s Perfectionism, Sher’s Beyond Neutralitx and Steven Wall’s Liberalism, Perfectionism and Restraint. Hurka presented an account of perfectionism that built on the work of many writers in the Western tradition of philosophy including Aristotle, Spinoza, Kant, Marx, and T. H. Green. This view identifies the best human life with the life that best develops capacities or properties that are essential to human nature. Although Hurka’s central concern was the articulation and defense of this moral theory, he also argued that it provided support for perfectionist political measures designed to help people discover and appreciate valuable options and strengthen their tendencies to engage in them. On Hurka’s view, it is a mistake to assume that in general “human beings left on their own will always choose what is best.”26Accordingly, the state, assuming that it is competent to do so, should assist them in their efforts to lead valuable lives. Likewise, Sher and Wall both argued that the case for excluding perfectionist considerations from public policy had not been established. Appeals to the need for political stability, to the value of autonomy, or to the principle of equal respect for persons all fail to show that perfectionist political action would be illegitimate. Moreover, on their view, it is not only permissible for the state to rely on perfectionist considerations, but in a well-ordered state these considerations would inform a wide range of public policies. As Sher puts it, perfectionist considerations are “often relevant to decisions about public assistance, educational policy, the criminal and civil justice system, the prison system, city planning and land use, transportation policy, the tax code, support for cultural institutions, regulation of the entertainment industry, investment incentives, and the structure of institutions such as the military-to name just a few of the more obvious candidate^."^^ This does not mean that perfectionists must defend indiscriminately all efforts by the state to promote the good and discourage the bad. As Wall notes,
16
Introduction
under some circumstances, the best way for the state to promote the good is for it to do very little.28This will be the case whenever there is ample reason to believe that those in positions of political authority do not have the competence to promote the good. In these circumstances, perfectionists will join proponents of state neutrality in rejecting perfectionist political action based on misguided evaluative beliefs. This point is often misunderstood. To make it clearer, it will be helpful to distinguish two levels of perfectionist political theory,29One level, which we can call “philosophical perfectionism,” concerns the ultimate standards of political evaluation. The philosophical perfectionist holds that in assessing political and social institutions we must consider the extent to which they promote valuable ways of life and discourage worthless or empty ones. The other level, which we can call “state perfectionism,” holds that the state should aim to favor valuable ways of life over disvaluable ones. Philosophical perfectionists will tend to affirm state perfectionism, but they need not affirm it uncritically. If they encounter a state whose officials hold what they consider to be mistaken beliefs about the good, then they will surely reject state perfectionism for this state.30Indeed, clearheaded philosophical perfectionists might adopt a policy of selective support for the principle of state neutrality-whenever a given state is not up to the job of promoting the good, then it is a good thing for it to strive to comply with the principle of state neutrality. Some critics of perfectionism, however, have claimed that no state is up to the job when it comes to supporting valuable options and discouraging disvaluable ones. The reason for this is that those who gain positions of political authority typically are not the best judges of the good; and even when they are good judges, power tends to corrupt their judgment.31If this were right, then all philosophical perfectionists would have reason to embrace some version of the principle of state neutrality, We cannot assess the plausibility of this kind of skepticism here. But it is worth noting that the considerations often invoked to call into question the state’s competence to promote the good would also call into question its competence to pursue other political goals, such as social justice.32The proponent of state neutrality who also wants the state to take an active role in the promotion and enforcement of social justice, therefore, should be wary about pressing this particular line of argument against those who endorse state perfectionism. One final distinction is helpful in clarifylng the role of perfectionism in politics. Some have suggested that all justified political action is perfe~tionist.~~ Call this the “strong thesis.” Others have claimed that perfectionist considerations can legitimately inform political action.34Call this the “weak thesis.” On the weak thesis, it is possible that nonperfectionist considerations-that is, considerations not related to the promotion of excellence or the protection and
Introduction
17
promotion of good human lives-might compete with and might need to be balanced against perfectionist considerations in justifying political action. The plausibility of these two views cannot be assessed in the abstract. Different perfectionist theories understand the good differently, and so the division between perfectionist and nonperfectionist considerations will differ, depending on the theory in question. Still, the distinction is worth noting because it underscores an important point. Whether or not the strong thesis is true, perfectionism as such is committed only to the weak thesis. Liberal Perfectionism What, if any, connection is there between perfectionism and liberalism? An often-voiced worry about perfectionism is that it justifies too much state coercion. Indeed, many versions of perfectionism drawn from the history of political thought have paid little heed to the modern value of individual freedom. And if the state is given the task of using its vast resources of power to guide citizens toward valuable options and away from disvaluable ones, then does this not, in and of itself, pose a significant threat to individual freedom? Contemporary defenders of perfectionism have responded to this worry in one of two ways. Drawing on the distinction between the strong and the weak thesis, they claim that perfectionist political action can and should be tempered by nonperfectionist consideration^.^^ For example, if there are strong nonperfectionist reasons for minimizing coercion, then perfectionists (who accept the weak thesis) might limit their support for perfectionist political action to noncoercive measures. Alternatively, perfectionists have sought to identify perfectionist reasons for limiting the power of the state. This second line of response is the more satisfying one. For if it can be carried through successfully, then perfectionism, properly understood, will turn out to be supportive of, rather than hostile to, individual freedom. With this in mind, we shall stipulate that liberal perfectionism is a version of perfectionism that identifies perfectionist reasons for favoring limited government. So understood, many contemporary defenders of perfectionism are liberal perfectionists. In all likelihood, the perfectionist case for limited government will not rest on any one master argument or single value.36If there are good perfectionist reasons for favoring limited government, then this (likely) will be true because a plurality of different, but compatible, perfectionist considerations come together to support it. Liberal perfectionists have identified a number of such considerations. We shall discuss three of them here. To begin with, as we have already noted in discussing Raz’s work, perfectionism is compatible with the doctrine of value pluralism. If there exists a
18
Introduction
plurality of fully good ways of life, then perfectionists need not hold that the state should promote a single conception of the good life. They can argue that the state best promotes the good by providing conditions under which its citizens can pursue and realize a wide range of valuable activities and goods. This shows that it is a mistake to think that perfectionists must be in the business of harnessing the coercive power of the state to advance a unitary substantive conception of the good life. Nonetheless, the truth of value pluralism, by itself, does not provide good reason to support limited government. Even if value pluralism were true, the state might attempt to guide its subjects into pursuing options that best suited their talents, irrespective of their desires to pursue these options. Plato’s idea that we should organize political and social life in such a way that each person does the job that best suits his nature is not refuted by the truth of value pluralism. This brings us to a second consideration liberal perfectionists often press in the service of limited government. The modern ideal of personal autonomy itself can be understood to be an important perfectionist The autonomous person is one who leads his life on his own terms, freely making choices from a wide and varied range of options. If the realization of this ideal were an important good, then when the state promotes valuable options and discourages worthless options it should not do so in ways that substantially undermine the autonomy of its citizens. This appears to ground some important limits to state power. For example, it would explain why it would be wrong for the state to compel its citizens to enter into the state’s specific occupations rather than letting them choose for themselves; for doing so would substantially obstruct their ability to lead their lives on their own terms. The perfectionist appeal to autonomy, however, must be qualified in two respects. First, autonomy is just one perfectionist good. It does not-at least not plausibly-always take priority over other goods. Second, the realization of autonomy may be only conditionally valuable. According to some perfectionists, an autonomous life is valuable only if it consists in engagement with valuable pursuits and relationship^.^^ If this is correct, then when the state closes off worthless options, while it might limit the autonomy of some of its citizens, it would not necessarily limit their capacity to lead a valuable autonomous life. And since, on this view, the latter, not the former, is what is valuable, the appeal to autonomy would not provide good perfectionist reasons for not closing off these options. The significance of these two qualifications perhaps can be best appreciated by considering the traditional issue of “morals legislation.” It is widely thought that laws that criminalize self-regarding immoral actions violate the autonomy of those they constrain. Suppose, for example, the state criminalizes
Introduction
19
prostitution or the use of dangerous recreational drugs. Then, on this widely held view, whatever benefits might result from such laws must be balanced against the costs in terms of autonomy that the laws impose. But if autonomy is just one good among others, and if it is only conditionally valuable, then it will not provide a strong reason for opposing such laws. The option to engage in prostitution or to consume dangerous recreational drugs may be a worthless option. If so, it will not be one that contributes to a person’s ability to lead a valuable autonomous life. This has led some perfectionists to reject the liberal case against morals legi ~ l a t i o n But . ~ ~ others have attempted to identify supplementary reasons for opposing such laws. Some have claimed, for example, that virtue cannot be promoted by coercion. In this vein, it is often said that state power extends only to external conduct; but virtue and morality require an “inner perfection.” The virtuous person not only acts rightly, but rightly for the right reasons. And this is not something that the state can compel. Other perfectionists have argued that coercive efforts to discourage immoral or worthless conduct have indiscriminate effects on the autonomy of those they target. For example, Raz claims that “there is no practical way of ensuring that the coercion will restrict the victim’s choice of repugnant options but will not interfere with other choices.”40If this were true, then even if the restriction of the bad options did not objectionably compromise autonomy, the state’s efforts to restrict these options might have effects that would do so. These considerations provide some reason for favoring noncoercive over coercive forms of state perfectionism. Still, it may not be true that noncoercive efforts to promote the good always will be as effective as coercive efforts. If all else is equal, then perfectionists have reason to favor noncoercive over coercive forms of state action. But of course, all else is seldom equal. In some contexts, coercive perfectionist political action may be the only effective way to sustain a decent and healthy environment in which citizens are more likely to lead good lives.*l This brings us to a third and final consideration sometimes invoked by liberal perfectionists in defense of limited government. Perfectionist political action attempts to guide citizens toward valuable options and away from worthless ones. But this kind of political action can, and often does, have important effects on how citizens understand themselves and their relationship to the political societies in which they live. These effects are significant, for the character of the political relationship can have great value. As we have seen, concerns about this relationship have led some antiperfectionists to insist that the state’s exercise of coercive power must be reasonably justifiable to all its citizens. But some perfectionists have also expressed concern about the character of this relationship. They have argued that all citizens have an interest in being full members of the political societies in which they
20
Introduction
An example illustrates the idea. Suppose a state were to promote atheism on the grounds that religious belief obstructs the ability of its citizens to lead good rational lives. Even if the state’s beliefs were true on this matter, its promotion of atheism likely would make it impossible for religious citizens to identify with the state. They would need to choose between their religious commitments and their political identity, and many would reasonably conclude that they could not view themselves as full members of the political society that governed them. This, in turn, might make their lives much less good; for if being a full member of one’s political society were an important ingredient of a good life, then the state’s promotion of atheism would set back their ability to lead a good life. Assuming that this line of thought were sound, it would have important implications for state perfectionism. If an otherwise valid perfectionist measure were to make it impossible for some citizens reasonably to view themselves as full members of their society, then the measure might be unjustified. Whatever benefits it brought would need to be balanced against the costs it imposed. Since these costs concern the interest that citizens have in not being alienated from the political institutions that apply to them, we can refer to them as “alienation costs.” Policies that make it reasonable for some to view themselves as second-class citizens would have high alienation costs. By focusing on these costs, perfectionists may be able to identify perfectionist reasons for restraint in pursuing perfectionist political goals. To be sure, this line of thought has its limits. Laws that criminalize pedophilia may make it impossible for some pedophiles to view themselves as full members of the political societies in which they live; but clearly, the benefits such laws provide in terms of the protection they give children outweigh any alienation costs. Still, in other cases, and in particular in cases where harm to third parties is not an issue, the need to avoid alienating citizens from their political institutions may provide a significant limit on the power of the state to promote the good. A strong emphasis on alienation costs brings liberal perfectionism much closer to state neutrality. Like the proponent of state neutrality, the perfectionist may worry that aggressive efforts by the state to promote the good will end up alienating many citizens. And he may believe that citizens who are alienated from the political societies in which they live will not be able to lead fully good lives. Yet, even if all this were accepted, perfectionism would still diverge from state neutrality. For not every perfectionist measure brings with it significant alienation costs, and those measures that do may yield benefits that justify the costs. The considerations discussed in this section do not provide a limit on state power as definite and clear as traditional liberal principles, such as the harm
Introduction
21
principle or the consent principle of political legitimacy. For this reason, the worry that perfectionism stands in tension with limited government will likely persist in some quarters. But the considerations do suggest how versions of perfectionism can be recognizably liberal. This should not be surprising, given that a number of important thinkers in the liberal tradition of political thought from von Humboldt to Mill to Green have been perfectionists. New Papers As this brief overview of the perfectionism-neutrality debate in contemporary political theory suggests, a number of issues remain unresolved and call for further investigation. Unclarity on these issues can generate considerable confusion over what exactly is at stake in the debate. Critics of state neutrality have all too often rejected overly simple versions of the neutrality principle, whereas critics of perfectionism have all too often painted it in colors that make it appear hostile to pluralism, toleration, and freedom. Progress in the debate requires moving beyond these crude characterizations. With this goal in mind, three questions, in particular, need further investigation: 1. What is the most promising formulation and justification of the neu-
trality principle?
2. To what extent is perfectionism compatible with important liberal val-
ues and principles?
3. And what kinds of institutions and policies would be ruled out by a
plausible neutrality principle, but favored by plausible versions of perfectionism?
In different ways, each of the new papers included in this volume attempts to answer one or more of these questions. The first two papers, those by Gerald F. Gaus and George Klosko, defend the neutrality principle. Both contributors present versions of this principle that are intended to overcome many of the standard objections to it. After noting the existence of a plurality of rival conceptions of state neutrality, Gaus argues that we can identify the appropriate conception by deriving it from foundational claims about the nature of moral reasoning. Morality requires “taking up a point of view that addresses reasons to all.” From this, Gaus derives a basic coercion-limiting principle. It holds that governments ought never to act without impartial justification, where impartial justification is understood in terms of providing reasons that would be acknowledged as justificatory by every fully rational citizen who is coerced by the
22
Introduction
government’s action. Properly understood, this principle, Gaus further argues, supports a version of the neutrality principle. The reason for this is that appeals to claims about the good life or about human perfection “rarely if ever” provide the kind of reasons that could justify state coercion. Many recent defenders of state neutrality have labored to show that the principle is, in fact, consistent with much of what modern liberal states do. An interesting and important feature of Gaus’s defense of state neutrality is that it has radical implications for public policy. Beyond laws that define property rights, and thus make political society possible, there is very little state coercion that would pass his neutrality test. Thus, Gaus’s version of the neutrality principle situates it within the classical liberal tradition of political thought, which views the justification of state coercion as the central issue in political philosophy. In a similar spirit, Klosko argues that the principle of state neutrality should be formulated to exclude certain kinds of reasons or considerations in the justification of state policy. He claims that the principle is “an inescapably liberal idea,” one that is based on particular value claims. Nonetheless, the principle is appropriate for broadly liberal societies in which there is substantial agreement on certain core liberal values such as respect for rights and democratic government. In these societies, Klosko argues, the neutrality principle functions to identify standards of proper argument in the justification of public policies. These standards rule out appeal to controversial conceptions of the good and, more broadly, forms of reasoning that are not publicly accessible and widely shared. A key feature of Klosko’s defense of the neutrality principle is his development of the idea of “reasonable rejection.” This idea distinguishes value claims that can be properly appealed to in justifications of state policy from those that cannot. The failure to draw this distinction clearly, Klosko argues, has contributed to the sense that either state neutrality rests on value skepticism or it rules out almost all state policies. To counter these criticisms, Klosko first outlines and clarifies the idea of reasonable rejection and then applies it to a range of public policies from school prayer to the treatment of animals to economic policy. Understanding how the neutrality principle should function, Klosko maintains, requires examining how it can be applied in concrete policy situations such as these. Not all the new contributors to this volume, however, share Gaus’s and Klosko’s confidence in the neutrality principle. In his contribution, Richard J. Arneson presents a spirited attack on it. He observes that this principle is generally understood to apply to claims about the good and not to claims about the right. This means that advocates of the principle must explain what it is about the good that renders it an unsuitable basis for state policy. Arneson
Introduction
23
then considers and rebuts a range of arguments that have been offered to explain why the state should not attempt to promote the good. Of particular interest here is his discussion of Dworkin’s recent argument in defense of state neutrality-an argument that appeals to the idea that a good life is a skillful performance that is produced in response to the challenge that one confronts. The central problem with this argument, Arneson contends, is that we have standards for assessing the goodness of lives that are independent of any account of success in responding to the challenges of one’s situation. These standards refer to objectively valuable goods such as friendship, scientific understanding, and meaningful work. Once this is adequately understood, the door is open for state efforts to promote these goods, even when doing so is controversial. Arneson also takes aim at recent efforts to soften the impact of the neutrality principle by restricting its application to the constitutional structure (what we earlier called narrow neutrality). This, he argues, is an unstable position; for if there are good reasons for neutrality with respect to constitutional essentials, they should apply to other state policies as well. Likewise, Arneson contends that it is a mistake to think that the neutrality principle applies only to the justification of the democratic process and not to the policies that might result from this process. Neutrality is best understood as “a constituent of a substantive conception of the just state.” Arneson thus is in considerable agreement with Gaus that, properly understood, the neutrality principle has radical implications for state policy, ruling out much of what liberal states do. But unlike Gaus, Arneson believes that this counts as a strike against the principle. George Sher also rejects the neutrality principle. His paper addresses a specific, but important, question: Is state perfectionism compatible with a strong commitment to free speech? The question is pressing, since it is conceivable that in order to promote the good effectively the state will need to suppress certain ideas. If this were indeed the case, then one would need to choose between state perfectionism and a strong commitment to free speech. Moreover, as Sher explains, a number of the most influential arguments for free speech tell against perfectionism (or could be developed in ways that tell against perfectionism). These include the consequentialist arguments of Mill, the autonomy-based argument of Thomas Scanlon, and the political argument of Alexander Meiklejohn. To reconcile state perfectionism with a strong commitment to free speech, Sher thinks that we must explain why the good arguments for free speech do not also call into question the state’s efforts to restrict disvaluable activities. The way to do this, he argues, is to identify the relevant differences between speech and action. Sher identifies two such differences. First, “foolish actions
24
Introduction
have costs that false speech does not”; and, second, “access to uninhibited discussion contributes far more to our ability to make informed decisions than does access to diverse lifestyles.” These differences suggest how state perfectionism and free speech can be reconciled; but interestingly, as Sher notes, the reconciliation may not be as complete as some would like. The highlighted differencesbetween speech and action apply to propositional speech. They do not necessarily obtain with respect to nonpropositional forms of expression. Just as Sher worries about the compatibility of perfectionism and free speech, others have worried about whether perfectionism sits well with a firm commitment to toleration. In his contribution to this volume, Wall asks whether there are distinctively perfectionist reasons to tolerate ways of life or social practices that are rightly disesteemed. He presents and develops a perfectionist account of toleration that relates judgments of toleration to judgments of objective value. The key to his account is the somewhat paradoxical idea that respecting the bad can be a condition for respecting the good. Wall presents two main arguments that delineate the structure of this account of toleration. The first argument identifies a relatively narrow range of objects that should be tolerated because they are inseparably intertwined with ways of life that are on balance valuable and worthy of respect. If successful, this argument dissolves the paradox in the idea that respecting the bad can be a condition for respecting the good. The second argument seeks to broaden the range of objects that warrant toleration by identifying dispositions of conduct and habits of thought that need to prevail if the good is to be adequately respected in societies that contain significant social diversity. As Wall stresses, these arguments do not provide a complete account of toleration. The arguments presuppose an account of objective value that must be filled in by a theory of the good. Still, despite this limit, the arguments reveal how there can be distinctive perfectionist reasons for tolerating the bad. Far from being hostile to toleration, then, perfectionism, as Wall understands it, can generate its own reasons for not suppressing the bad. The final paper is by Kent Greenawalt. He attempts to illuminate the idea of neutrality by considering its application to issues in the American law of church and state. This idea has loomed large in recent constitutional decisions that establish how a liberal state should deal with religion; but there is not, as Greenawalt observes, a single conception of neutrality at work in this context. Different understandings of religious neutrality, resting on rival background beliefs and values, have been formulated. This truth, however, should not lead us to dismiss neutrality as unimportant or empty. Greenawalt points out that some critics of neutrality have assumed unreasonably that a genuine account of religious neutrality would be one that all citizens would perceive as neutral. Clearly, no such account is possible. But from this, Greenawalt contends, it
Introduction
25
does not follow that neutrality cannot play an important role in justifying the state’s involvement with religion in a number of important areas, such as exempting claims of religious conscience from ordinary laws or granting aid to religious organizations. Greenawalt next considers the role that religious ideas should play in political discussion. Here, he introduces two important distinctions: that between advocacy and justification and that between public officials and ordinary citizens. The core of the position he defends is that if religious ideas should be excluded from American political discussion, this exclusion should apply only to the public statements that officials make in the course of defending their policies. It should not extend to the grounds on which they reach their decisions. Nor should it apply to the political advocacy of ordinary citizens. Greenawalt concludes by noting that his position on the role that religion should play in American political discussion, even if sound, does not apply to all liberal political orders. The shape neutrality should take and the importance it has vary with the history and culture of different societies. Thus, Greenawalt’s defense of neutrality is unusually modest. Unlike those who have elevated it to a central norm of political morality, Greenawalt depicts neutrality as an ideal that must be filled in by reference to particular background beliefs and purposes, and one whose evaluative significance varies from place to place.
Notes 1. Alan Montefiore, ed., Neutrality and Impartiality (Cambridge: Cambridge University Press, 1975), 5. 2. Among proponents of state neutrality, there is no clear and widely accepted understanding of what the phrase “conception of the good” refers to. While most everyone agrees that conceptions of the good include controversial ideals about what constitutes a good life, there remains considerable disagreement over what else is included. For helpful discussion of this idea, see George Sher, Beyond Neutrality: Perfectionism and Politics (Cambridge: Cambridge University Press, 1997), 37-44; and Joseph Chan, “Legitimacy, Unanimity, and Perfectionism,” Philosophy and Public Affairs 29, no. 1 (winter 2000): 11-14. 3. John Rawls, A Theory of Justice (Cambridge, Mass.: Harvard University Press, 1971), 291-92. 4. Ronald Dworkin, “Liberalism,” in Public and Private Morality, ed. Stuart Hampshire (Cambridge: Cambridge University Press, 1978), 127. 5 . Bruce Ackerman, SocialJustice in the Liberal State (New Haven, Conn.: Yale University Press, 1980), 11. 6. Charles Larmore, Patterns ofMoral Complexity (Cambridge: Cambridge University Press, 1987), 42.
26
Introduction
7. Thomas Nagel, Equality and Partiality (Oxford: Oxford University Press, 1991), 156. 8. Ibid., 34. 9. See Lawrence Solum, “Constructing an Ideal of Public Reason,” San Diego Law Review 30 (1993): 729-62. 10. Will Kymlicka, Contemporary Political Philosophy, 2nd ed. (New York Oxford University Press, 2002), 344. 11. See Robert E. Goodin and Andrew Reeve, “Do Neutral Institutions Add up to a Neutral State?” in Liberal Neutrality, ed. Robert E. Goodin and Andrew Reeve (London: Routledge, 1989); and Steven Wall, “Neutrality and Responsibility,” Journal of Philosophy 98, no. 8 (August 2001): 389-410. 12. See Abington School District v. Schempp, 374 U.S. 203. 13. Ibid., at 266. 14. First advanced by Chief Justice Warren Burger in Walz v. Tax Commission of the City of New York (397 U.S. 664 [ 19701, at 674-75). It is notable that U.S. courts had propounded neutrality of religion much earlier. In the opinion just cited, Justice Clark notes that Alphonso Taft, father of President and Chief Justice William Taft, had defended the idea almost a hundred years earlier: “The government is neutral, and, while protecting all, it prefers none, and it disparages none” (371 U.S. 203 [at 214-151; the Taft opinion is reported as unpublished, in Board of Education of Cincinnati v. Minor, 203 Ohio St. 211,253 [1872]). 15. Larmore, Patterns ofMoral Complexity, 67. 16. John Rawls, Political Liberalism (New York: Columbia University Press, 1993), 139. 17. Jeremy Waldron, “Legislation and Moral Neutrality,” in Liberal Neutrality, ed. Robert E. Goodin and Andrew Reeve (London: Routledge, 1989), 69. 18. Ronald Dworkin, “Foundations of Liberal Equality,” in Equal Freedom (Ann Arbor: University of Michigan Press, 1995), 195. 19. For discussion and criticism, see Patrick Neal, “Dworkin on the Foundations of Liberal Equality,” Legal Theory 1, no. 2 (summer 1995): 205-26; and Arneson’s chapter in this volume. 20. Rawls, Theory of Justice, 325. 2 1. Chan, “Legitimacy, Unanimity, and Perfectionism,” 5. 22. Vinit Haksar, Equality, Liberty and Perfectionism (Oxford: Oxford University Press, 1979), 288. 23. Stephen Macedo, Liberal Virtues: Citizenship, Virtue and Community in Liberal Constitutionalism (Oxford: Clarendon, 1990), 288. 24. Joseph Raz, The Morality of Freedom (Oxford Oxford University Press, 1986), 133. 25. On this point, see Sher, Beyond Neutrality, 1. 26. Thomas Hurka, Perfectionism (Oxford: Oxford University Press, 1993), 160. 27. Sher, Beyond Neutrality, 246. 28. Steven Wall, Liberalism, Perfectionism and Restraint (Cambridge: Cambridge University Press, 1998), 15. 29. Here, we draw on Hurka, Perfectionism, 162-63.
Introduction
27
30. Thomas Hurka, “Indirect Perfectionism: Kymlicka on Indirect Perfectionism,” Journal of Political Philosophy 3, no. 1 (March 1995): 51. 31. Ackerman, Social Justice in the Liberal State, 362-63. 32. See Simon Caney, “Consequentialist Defences of Liberal Neutrality,”Philosophical Quarterly 41, no. 165 (January 1991): 460-63. 33. See Raz, Morality of Freedom, 133. 34. See Sher, Beyond Neutrality, 246. 35. Chan, “Legitimacy, Unanimity, and Perfectionism,” 15. 36. Joseph Raz, “Liberty and Trust,” in Natural Law, Liberalism and Morality, ed. Robert George (Oxford Oxford University Press, 1996), 113-29. 37. This point is well developed in the selection from Hurka reprinted in this volume. 38. See Raz, Morality of Freedom, 378-81. 39. See Robert George, Making Men Moral (Oxford Oxford University Press, 1993). 40. Raz, Morality of Freedom, 419. For a critical discussion of this argument, see Wojciech Sadurski, “Joseph Raz on Liberal Neutrality and the Harm Principle,” Oxford Journal $Legal Studies 10, no. 1 (spring 1990). 41. George, Making Men Moral. 42. See Raz, “Liberty and Trust,” 122-28; see also Robert Adams, Finite and Infinite Goods (Oxford: Oxford University Press, 2001), 332-40.
This Page Intentionally Left Blank
I SELECTIONS FROM PUBLISHED WORKS
A Neutrality
This Page Intentionally Left Blank
1 “Liberalism,” in Public and Private Morality Ronald Dworkin
Introduction It is often said that in regard to clashes between the values of liberty and equality, liberals favor equality while conservatives prefer liberty. In the selection excerpted here, Dworkin demurs. He argues that liberals and conservatives differ primarily over competing conceptions of what equality requires. Liberals believe in treating people as equals by not favoring some people’s conceptions of the good over those of others. Conservatives have a robust conception of virtue, and hold that good government consists in treating people as if they were “desirous of leading the life that is in fact good.” Dworkin considers the institutions that follow from his conception of liberalism, bearing on both questions of distribution and individual rights. The connection drawn here between liberalism and neutrality had significant influence on subsequent discussions, even though Dworkin’s primary concern is to distinguish liberal political positions from conservative ones by tracing them back to fundamental philosophical differences.
Ronald Dworkin, “Liberalism,”in Public and Private Morality, ed. Stuart Hampshire (Cambridge: Cambridge University Press, 1978), 127, 128-30, 132-38. Copyright 0 1978 by Cambridge University Press. Reprinted by permission of Cambridge University Press.
- 31 -
32
Ronald Dworkin
Mr”
AT DOES IT MEAN FOR the government to treat its citizens as equals? That is, I think, the same question as the question of what it means for the government to treat all its citizens as free, or as independent, or with equal dignity. In any case, it is a question that has been central to political theory at least since Kant. It may be answered in two fundamentally different ways. The first supposes that government must be neutral on what might be called the question of the good life. The second supposes that government cannot be neutral on that question, because it cannot treat its citizens as equal human beings without a theory of what human beings ought to be. I must explain that distinction further. Each person follows a more-or-less articulate conception of what gives value to life. The scholar who values a life of contemplation has such a conception; so does the television-watching, beer-drinking citizen who is fond of saying “This is the life,” though he has thought less about the issue and is less able to describe or defend his conception. The first theory of equality supposes that political decisions must be, so far as is possible, independent of any particular conception of the good life, or of what gives value to life. Since the citizens of a society differ in their conceptions, the government does not treat them as equals if it prefers one conception to another, either because the officials believe that one is intrinsically superior, or because one is held by the more numerous or more powerful group. The second theory argues, on the contrary, that the content of equal treatment cannot be independent of some theory about the good for man or the good of life, because treating a person as an equal means treating him the way the good or truly wise person would wish to be treated. Good government consists in fostering or at least recognizing good lives; treatment as an equal consists in treating each person as if he were desirous of leading the life that is in fact good, at least so far as this is possible.
I now define a liberal as someone who holds the first, or liberal, theory of what equality requires. Suppose that a liberal is asked to found a new state. He is required to dictate its constitution and fundamental institutions. He must propose a general theory of political distribution, that is, a theory of how whatever the community has to assign, by way of goods or resources or opportunities, should be assigned. He will arrive initially at something like this principle of rough equality: resources and opportunities should be distributed, so far as possible, equally, so that roughly the same share of whatever is available is devoted to satisfying the ambitions of each. Any other general aim of distribution will assume either that the fate of some people should be of greater concern
“Liberalism,”in Public and Private Morality
33
than that of others, or that the ambitions or talents of some are more worthy, and should be supported more generously on that account. Someone may object that this principle of rough equality is unfair because it ignores the fact that people have different tastes, and that some of these are more expensive to satisfy than others, so that, for example, the man who prefers champagne will need more funds if he is not to be frustrated than the man satisfied with beer. But the liberal may reply that tastes as to which people differ are, by and large, not afflictions, like diseases, but are rather cultivated, in accordance with each person’s theory of what his life should be like.’ The most effective neutrality, therefore, requires that the same share be devoted to each, so that the choice between expensive and less expensive tastes can be made by each person for himself, with no sense that his overall share will be enlarged by choosing a more expensive life, or that, whatever he chooses, his choice will subsidize those who have chosen more expensively.2 But what does the principle of rough equality of distribution require in practice? If all resources were distributed directly by the government through grants of food, housing, and so forth; if every opportunity citizens have were provided directly by the government through the provisions of civil and criminal law; if every citizen had exactly the same talents; if every citizen started his life with no more than what any other citizen had at the start; and if every citizen had exactly the same theory of the good life and hence exactly the same scheme of preferences as every other citizen, including preferences between productive activity of different forms and leisure, then the principle of rough equality of treatment could be satisfied simply by equal distributions of everything to be distributed and by civil and criminal laws of universal application. Government would arrange for production that maximized the mix of goods, including jobs and leisure, that everyone favored, distributing the product equally. Of course, none of these conditions of similarity holds. But the moral relevance of different sorts of diversity are very different, as may be shown by the following exercise. Suppose all the conditions of similarity I mentioned did hold except the last: citizens have different theories of the good and hence different preferences. They therefore disagree about what product the raw materials and labor and savings of the community should be used to produce, and about which activities should be prohibited or regulated so as to make others possible or easier. The liberal, as lawgiver, now needs mechanisms to satisfy the principles of equal treatment in spite of these disagreements. He will decide that there are no better mechanisms available, as general political institutions, than the two main institutions of our own political economy: the economic market, for decisions about what goods shall be produced and how they shall be distributed, and representative democracy, for
34
Ronald Dworkin
collective decisions about what conduct shall be prohibited or regulated so that other conduct might be made possible or convenient. Each of these familiar institutions may be expected to provide a more egalitarian division than any other general arrangement.
* * * In a society in which people differed only in preferences, then, a market would be favored for its egalitarian consequences. Inequality of monetary wealth would be the consequence only of the fact that some preferences are more expensive than others, including the preference for leisure time rather than the most lucrative productive activity. But we must now return to the real world. In the actual society for which the liberal must construct political institutions, there are all the other differences. Talents are not distributed equally, so the decision of one person to work in a factory rather than a law firm, or not to work at all, will be governed in large part by his abilities rather than his preferences for work or between work and leisure. The institutions of wealth, which allow people to dispose of what they receive by gift, means that children of the successful will start with more wealth than the children of the unsuccessful. Some people have special needs, because they are handicapped; their handicap will not only disable them from the most productive and lucrative employment, but will incapacitate them from using the proceeds of whatever employment they find as efficiently, so that they will need more than those who are not handicapped to satisfy identical ambitions. These inequalities will have great, often catastrophic, effects on the distribution that a market economy will provide. But, unlike differences in preferences, the differences these inequalities make are indefensible according to the liberal conception of equality. It is obviously obnoxious to the liberal conception, for example, that someone should have more of what the community as a whole has to distribute because he or his father had superior skill or luck. The liberal lawgiver therefore faces a difficult task. His conception of equality requires an economic system that produces certain inequalities (those that reflect the true differential costs of goods and opportunities) but not others (those that follow from differences in ability, inheritance, and so on). The market produces both the required and the forbidden inequalities, and there is no alternative system that can be relied upon to produce the former without the latter. The liberal must be tempted, therefore, to a reform of the market through a scheme of redistribution that leaves its pricing system relatively intact but sharply limits, at least, the inequalities in welfare that his initial principle prohibits. No solutions will seem perfect. The liberal may find the best answer in
“Liberalism,”in Public and Private Morality
35
a scheme of welfare rights financed through redistributive income and inheritance taxes of the conventional sort, which redistributes just to the Rawlsian point, that is, to the point at which the worst-off group would be harmed rather than benefited by further transfers. In that case, he will remain a reluctant capitalist, believing that a market economy so reformed is superior, from the standpoint of his conception of equality, to any practical socialist alternative. Or he may believe that the redistribution that is possible in a capitalist economy will be so inadequate, or will be purchased at the cost of such inefficiency, that it is better to proceed in a more radical way, by substituting socialist for market decisions over a large part of the economy, and then relying on the political process to insure that prices are set in a manner at least roughly consistent with his conception of equality. In that case he will be a reluctant socialist, who acknowledges the egalitarian defects of socialism but counts them as less severe than the practical alternatives. In either case, he chooses a mixed economic system-either redistributive capitalism or limited socialism-not in order to compromise antagonistic ideals of efficiency and equality, but to achieve the best practical realization of the demands of equality itself. Let us assume that in this manner the liberal either refines or partially retracts his original selection of a market economy. He must now consider the second of the two familiar institutions he first selected, which is representative democracy. Democracy is justified because it enforces the right of each person to respect and concern as an individual; but in practice the decisions of a democratic majority may often violate that right, according to the liberal theory of what the right requires. Suppose a legislature elected by a majority decides to make criminal some act (like speaking in favor of an unpopular political position, or participating in eccentric sexual practices), not because the act deprives others of opportunities they want, but because the majority disapproves of those views or that sexual morality. The political decision, in other words, reflects not just some accommodation of the personal preferences of everyone, in such a way as to make the opportunities of all as nearly equal as may be, but the domination of one set of external preferences, that is, preferences people have about what others shall do or have.3 The decision invades rather than enforces the right of citizens to be treated as equals. How can the liberal protect citizens against that sort of violation of their fundamental right? It will not do for the liberal simply to instruct legislators, in some constitutional exhortation, to disregard the external preferences of their constituents. Citizens will vote these preferences in electing their representatives, and a legislator who chooses to ignore them will not survive. In any case, it is sometimes impossible to distinguish, even by introspection, the external and personal components of a political position: this is the case, for
36
Ronald Dworkin
example, with associational preferences, which are the preferences some people have for opportunities, like the opportunity to attend public schoolsbut only with others of the same “background.” The liberal, therefore, needs a scheme of civil rights whose effect will be to determine those political decisions that are antecedently likely to reflect strong external preferences and to remove those decisions from majoritarian political institutions altogether. The scheme of rights necessary to do this will depend on general facts about the prejudices and other external preferences of the majority at any given time, and different liberals will disagree about what is needed at any particular time.4 But the rights encoded in the Bill of Rights of the United States Constitution, as interpreted (on the whole) by the Supreme Court, are those that a substantial number of liberals would think reasonably well suited to what the United States now requires (thought most would think that the protection of the individual in certain important areas, including sexual publication and practice, are much too weak). The main parts of the criminal law, however, present a special problem not easily met by a scheme of civil rights that disable the legislature from taking certain political decisions. The liberal knows that many of the most important decisions required by an effective criminal law are not made by legislators at all, but by prosecutors deciding whom to prosecute for what crime, and by juries and judges deciding whom to convict and what sentences to impose. He also knows that these decisions are antecedently very likely to be corrupted by the external preferences of those who make these decisions because those they judge, typically, have attitudes and ways of life very different from their own. The liberal does not have available, as protection against these decisions, any strategy comparable to the strategy of civil rights that merely remove a decision from an institution. Decisions to prosecute, convict, and sentence must be made by someone. But he has available, in the notion of procedural rights, a different device to protect equality in a different way. He will insist that criminal procedure be structured to achieve a margin of safety in decisions, so that the process is biased strongly against the conviction of the innocent. It would be a mistake to suppose that the liberal thinks that these procedural rights will improve the accuracy of the criminal process, that is, the probability that any particular decision about guilt or innocence will be the right one. Procedural rights intervene in the process, even at the cost of inaccuracy, to compensate in a rough way for the antecedent risk that a criminal process, especially if it is largely administered by one class against another, will be corrupted by the impact of external preferences that cannot be eliminated directly. This is only the briefest sketch of how various substantive and procedural civil rights follow from the liberal’s initial conception of equality; it is meant to suggest, rather than
“Liberalism,”in Public and Private Morality
37
demonstrate, the more precise argument that would be available for more particular rights. So the liberal, drawn to the economic market and to political democracy for distinctly egalitarian reasons, finds that these institutions will produce inegalitarian results unless he adds to his scheme different sorts of individual rights. These rights will function as trump cards held by individuals; they will enable individuals to resist particular decisions in spite of the fact that these decisions are or would be reached through the normal workings of general institutions that are not themselves challenged. The ultimate justification for these rights is that they are necessary to protect equal concern and respect; but they are not to be understood as representing equality in contrast to some other goal or principle served by democracy or the economic market. The familiar idea, for example, that rights of redistribution are justified by an ideal of equality that overrides the efficiency ideals of the market in certain cases, has no place in liberal theory. For the liberal, rights are justified, not by some principle in competition with an independent justification of the political and economic institutions they qualify, but in order to make more perfect the only justification on which these other institutions may themselves rely. If the liberal arguments for a particular right are sound, then the right is an unqualified improvement in political morality, not a necessary but regrettable compromise of some other independent goal, like economic efficiency. I said that the conservative holds one among a number of possible alternatives to the liberal conception of equality. Each of these alternatives shares the opinion that treating a person with respect requires treating him as the good man would wish to be treated. The conservative supposes that the good man would wish to be treated in accordance with the principles of a special sort of society,which I shall call the virtuous society. A virtuous society has these general features. Its members share a sound conception of virtue, that is, of the qualities and dispositions people should strive to have and exhibit. They share this conception of virtue not only privately, as individuals, but publicly: they believe their community, in its social and political activity, exhibits virtues, and that they have a responsibility, as citizens, to promote these virtues. In that sense they treat the lives of other members of their community as part of their own lives. The conservative position is not the only position that relies on this ideal of the virtuous society (some forms of socialism rely on it as well). But the conservative is distinct in believing that his own society, with its present institutions, is a virtuous society for the special reason that its history and common experience are better guides to sound virtue than any nonhistorical and therefore abstract deduction of virtue from first principles could provide.
38
Ronald Dworkin
Suppose a conservative is asked to draft a constitution for a society generally like ours, which he believes to be virtuous. Like the liberal, he will see great merit in the familiar institutions of political democracy and an economic market. The appeal of these institutions will be very different for the conservative, however. The economic market, in practice, assigns greater rewards to those who, because they have the virtues of talent and industry, supply more of what is wanted by the other members of the virtuous society; and that is, for the conservative, the paradigm of fairness in distribution. Political democracy distributes opportunities, through the provisions of the civil and criminal law, as the citizens of a virtuous society wish it to be distributed, and that process will provide more scope for virtuous activity and less for vice than any less democratic technique. Democracy has a further advantage, moreover, that no other technique could have. It allows the community to use the processes of legislation to reaffirm, as a community, its public conception of virtue. The appeal of the familiar institutions to the conservative is, therefore, very different from their appeal to the liberal. Since the conservative and the liberal both find the familiar institutions useful, though for different reasons, the existence of these institutions, as institutions, will not necessarily be a point of controversy between them. But they will disagree sharply over which corrective devices, in the form of individual rights, are necessary in order to maintain justice, and the disagreement will not be a matter of degree. The liberal, as I said, finds the market defective principally because it allows morally irrelevant differences, like differences in talent, to affect distribution, and he therefore considers that those who have less talent, as the market judges talent, have a right to some form of redistribution in the name of justice. But the conservative prizes just the feature of the market that puts a premium on talents prized in the community, because these are, in a virtuous community, virtues. So he will find no genuine merit, but only expediency, in the idea of redistribution. He will allow room for the virtue of charity, for it is a virtue that is part of the public catalog; but he will prefer private charity to public, because it is a purer expression of that virtue. He may accept public charity as well, particularly when it seems necessary to retain the political allegiance of those who would otherwise suffer too much to tolerate a capitalist society at all. But public charity, justified either on grounds of virtue or expediency, will seem to the conservative a compromise with the primary justification of the market, rather than, as redistribution seems to the liberal, an improvement in that primary justification. Nor will the conservative find the same defects in representative democracy that the liberal finds there. The conservative will not aim to exclude moralis-
“Liberalism,”in Public and Private Morality
39
tic or other external preferences from the democratic process by any scheme of civil rights; on the contrary, it is the pride of democracy, for him, that external preferences are legislated into a public morality. But the conservative will find different defects in democracy, and he will contemplate a different scheme of rights to diminish the injustice they work. Notes 1. See Ronald Dworkin, Taking Rights Seriously (Cambridge, Mass.: Harvard University Press, 1977), chapter 12. 2. Ibid., 227. 3. See T. M. Scanlon, “Preference and Urgency,” Journal ofPkilosopky 72 (1975). 4. A very different objection calls attention to the fact that some people are afflicted with incapacities like blindness or mental disease, so that they require more resources to satisfy the same scheme of preferences. That is a more appealing objection to my principle of rough equality of treatment, but it calls, not for choosing a different basic principle of distribution, but for corrections in the application of the principle like those I considered later.
This Page Intentionally Left Blank
2 Social justice in the Liberal State Bruce Ackerman
Introduction In this excerpt, Ackerman argues that claims to exercise power over other people must be able to be defended in open dialogue. Not every defense, however, is admissible. Ackerman seeks to articulate “a conversational constraint” (what he refers to here as Z) that will identify types of reasons that are inadmissible in the justification of power claims. The neutrality principle emerges as the best interpretation of this conversational constraint. It is, Ackerman claims, “the organizing principle of liberal thought.” This constraint holds that claims to superior power cannot be justified by assertions that one’s conception of the good is better than those of other people or that one is oneself intrinsically superior to them. Ackerman further holds that the neutrality principle can be justified in many different ways. Just as a liberal political system allows people to live according to different conceptions of the good, so too liberal political philosophy encourages a number of different argumentative paths to neutrality.
Bruce Ackerman, SocialJustice in the Liberal State (New Haven, Conn.:Yale University Press, 1980), 10-12,355-57. Copyright 0 1980 by Yale University Press. Reprinted by permission of Yale University Press. -41
-
42
Bruce Ackerman
W
HEN OTHERS HAVE SOUTH TO GIVE LIBERALISM SYSTEMATIC FORM, they have tuned to other ideas-most notably contract or utility-to serve as their organizing principle. In contrast, I hope to convince you that the idea of constrained conversation provides a far more satisfactory key to the liberal enterprise. And once the lock is turned, the liberal tradition will reveal unsuspected resources of methodological rigor and substantive depth.
Neutrality and Convergence My particular Z taps the liberal’s opposition to paternalism. The germ of the idea is that nobody has the right to vindicate political authority by asserting a privileged insight into the moral universe which is denied the rest of us. A power structure is illegitimate if it can be justified only through a conversation in which some person (or group) must assert that he is (or they are) the privileged moral authority: Neutrality.No reason is a good reason if it requires the power holder to assert: (a) that his conception of the good is better than that asserted by any of his fellow citizens, or (b) that, regardless of his conception of the good, he is intrinsically superior to one or more of his fellow citizens. I will defer important questions of interpretation to emphasize the main point-which is the way Neutrality promises to satisfy the two conditions we require for a potent Z. Since the breadth of the exclusion imposed by Neutrality is obvious, the crucial question is whether the formulation suffers the defects of this virtue: Does its very breadth make it impossible to generate arguments that will justify its acceptance as a fundamental constraint on power talk? Not at all. It is downright easy to think of several weighty arguments in support of Neutrality. The first is a skeptical argument: While everybody has an opinion about the good life, none can be known to be superior to any other. It follows that anyone who asserts that either he or his aims are intrinsically superior doesn’t know what he’s talking about. Yet this is precisely the move barred by Neutrality. But there is no need to be a skeptic before you can reason your way to Neutrality. Even if you think you can know something about the good life, there are several good reasons for imposing liberal constraints on political conversation. Most obviously, you might think that you can only learn anything true about the good when you are free to experiment in life without some authoritative teacher intervening whenever he thinks you’re going wrong. And if you think this, Neutrality seems made to order. But, once again, this view is only one of many that will provide a plausible path to Neutrality. Even if you don’t think you need to experiment, you may adopt a conception of the good that
Social Justice in the Liberal State
43
gives a central place to autonomous deliberation and deny that it is possible to force a person to be good. On this view, the intrusion of non-Neutral argument into power talk will seem self-defeating at best-since it threatens to divert people from the true means of cultivating a truly good life. Assume, finally, that you think you know what the good life is and that it is of a kind that can be forced on others; then the only question is whether the right people will be doing the forcing. A single glance at the world suggests that this is no trivial problem. People adept in gaining power are hardly known for their depth of moral insight; the very effort to engross power corrupts-at least if your theory of the good embraces any number of familiar moral ideals. Not that it is absolutely impossible to reason yourself to a rejection of Neutrality. Plato began systematic political philosophy with such a dream; medieval churchmen thought there were good reasons to confide ultimate secular authority to the pope. Only they recognized-as modern totalitarians do not-the depth of the reconceptualization required before a breach of Neutrality can be given a coherent justification. It is not enough to reject one or another of the basic arguments that lead to a reasoned commitment to Neutrality; one must reject all of them. And to do this does not require a superficial change of political opinions but a transformation of one’s entire view of the world-both as to the nature of human values and the extent to which the powerful can be trusted to lead their brethren to the promised land. In proposing Neutrality, then, I do not imagine I am defending an embattled citadel on the fringe of modern civilization. Instead, I am pointing to a place well within the cultural interior that can be reached by countless pathways of argument coming from very different directions. As time passes, some paths are abandoned while others are worn smooth; yet the exciting work on the frontier cannot blind us to the hold that the center has upon us.. . . Beyond Monologue Consider, then, how dull a world it would be if we spent all of our time talking about the question of legitimacy that is the idCe fme of liberal political philosophy. Rather than our exclusive interest in life, legitimacy talk is but a narrow stream feeding a large sea of talk dealing with a bewildering set of questions of infinite variety: How does a good person live his life? What is happening around me? What can possibly happen in the future? What was the best joke I’ve ever heard? And so forth. Moreover, just as we can philosophize about the question of legitimacy, so too we may expose any of these other questions to sustained questioning. Hence in addition to political philosophy, there are countless other philosophical domains, whose relative importance to one another is itself a controversial
44
Bruce Ackerrnan
matter. Suppose, for example, that you do not begin-as I have done-with the question of legitimacy, but with some other question of greater interest to you-say, the question of truth or beauty or the way a person ought to live his life. Having come to an answer (however tentative) to one or more of these questions, you may find that your answer implies a particular solution to the question of legitimacy. Thus, once Plato answered questions as to the nature of truth, beauty, and the good, the question of legitimacy seemed to him child’s play. It could be answered by making a few deductions from the answers he had given to these other questions. While many of Plato’s great successors disagreed with his substantive answers, they shared his belief that the question of legitimacy depended on the successful solution of some other philosophical puzzlement. Thus, Augustine or Kant thought Plato’s answers-indeed, his basic questions-faulty; nonetheless, once they answered the right questions, they thought that the question of legitimacy could be resolved in a relatively straightforward manner. With apologies, I shall call this the deductionist approach to the question of legitimacy-denoting the belief that the first principles of political philosophy are really “second” principles that may be deduced from the conclusions reached in some other “higher” domain of philosophical discourse. Now this deductionist strategy surely does avoid the problems of intuitionism. Once the move is taken, it is easy-all too easy-to explain why you should purge yourself of all political prejudices that cannot be defended through a Neutral dialogue: YOU: Why should our power struggle be regulated by Neutral dialogue? I: Here. Read this (tendering my favorite philosopher). It contains the truth about humanity’s ultimate predicament. YOU: And why is this truth relevant to my question? I: Because once you glimpse the truth, it will logically compel you to accept Neutral dialogue. YOU: But a commitment to Neutral dialogue offends some of my intuitive judgments about politics. I: Once you have read the truth, you will see it is your job to purify yourself of all inconsistent prejudices. Read then, and the truth will make you free. This single conversation threatens the thousands of words that have preceded it. After consistently avoiding any claim of privilege access to the ultimate truth, I would have succeeded in grounding my entire structure on the very conversational move I hoped to avoid. If the deductionist strategy is the only alternative, no wonder that liberals tend toward intuitionism, whatever its ultimate weaknesses.
Social Justice in the Liberal State
45
I should emphasize, moreover, that the deductionist strategy is not made any less suspect-only more insidious-if your favorite philosopher happens to be somebody like Kant or Hume rather than Plato or Augustine. Of course, if I were forced down the deductionist path, I would prefer Kant to Plato as I sought to convince you of the metaphysical truth of liberalism. Nonetheless, in making this move, it remains true that I have made liberalism a hostage of a particular metaphysical system. Yet it is the essence of liberalism to deny people the right to declare that their particular metaphysic and epistemology contains the truth, the whole truth, and nothing but the truth. If intuitionism is question begging, deductionism begs the peculiar difficulties involved in providing a liberal answer. How, then, to avoid the impasse? By returning to the last dialogic script and considering whether there are any moves that “I” left unexplored in my previous encounter with You. And when this is done, I think we can find a way to talk our way out of the impasse:
YOU: Why should our power struggle be regulated by Neutral dialogue? I: A good question. Would you mind telling me why you find this conversational constraint unduly confining? At first, bouncing the conversational ball back to you may seem a cheap trick. What is to prevent you from simply refusing the invitation to play the conversational game beyond this point? While this is indeed the ultimate question, I want to put it in brackets for a time until the more affirmative aspects of this conversational turn appear to view. After all, there is no reason to think that most people will refuse to play the game at this point; moreover, the person I am most interested in, at the moment, is you, the reader of this book. Can you put into words the reasons why you find the idea of Neutral conversation troubling? If you can, then I have something that we can work with in a way that does not offend liberalism’s distrust of deductionism. Instead of launching into my own monologue on the nature of things, I can respond to your efforts to verbalize the particular reasons you chafe under the constraint of Neutrality. Once you verbalize the reasons you are troubled, perhaps I can show you that they are not as persuasive as you might have thought: that even within your own view of the world, you should find ultimately unpersuasive the reasons you can give for your initial disquiet. If I am successful in this project, our dialogue will not conclude with my shoving my own words down your throat. Instead, it will end with you yourselfrecognizing that the world, as you understand it, makes Neutral dialogue the most sensible way of regulating our power struggle.
This Page Intentionally Left Blank
Patterns of Moral Complexity Charles Larrnore
Introduction Larmore attributes the need for state neutrality to the ineradicable pluralism of contemporary liberal societies. He claims that when people disagree about some point at issue, they should retreat to neutral ground and find a basis for common agreement. Larmore holds that a wide range of governmental functions can be justified on such grounds, not only protection of life and property, but also promotion of freedom and economic equality. The neutral state is not necessarily the minimal state. Among Larmore’s most important contributions are his arguments concerning the proper justification of the neutrality principle. Rejecting attempts to ground it on skepticism, the value of experimentation, or individual autonomy, Larmore presents what he calls “a neutral justification of political neutrality.” This justification appeals to “a universal norm of rational dialogue,” which is in turn based on a requirement of mutual respect. As Larmore explains, to respect another person is to recognize that he, too, has a distinctive perspective on the world. State actions that affect him must be justifiable from within his own perspective.
Charles Larmore, Patterns of M o m Complexity (Cambridge: Cambridge University Press, 1987), 50-55, 5 9 4 1 , 6 4 6 8 . Copyright 0 1987 by Cambridge University Press. Reprinted by permission of Cambridge University Press. -47-
Charles Larmore
48
What Is Neutrality?
T
HE IDEAL OF NEUTRALITY CAN best be
understood as a response to the variety of conceptions of the good life. In modern times we have come to recognize a multiplicity of ways in which a fulfilled life can be lived, without any perceptible hierarchy among them. And we have also been forced to acknowledge that even where we do believe that we have discerned the superiority of some ways of life to others, reasonable people may often not share our view. Pluralism and reasonable disagreement have become for modern thought ineliminable features of the idea of the good life. Political liberalism has been the doctrine that consequently the state should be neutral. The state should not seek to promote any particular conception of the good life because of its presumed intrinsic superiority-that is, because it is supposedly a truer conception. (A liberal state may naturally restrict certain ideals for extrinsic reasons because, for example, they threaten the lives of others.) One of the most important questions of liberal theory concerns the precise way in which pluralism and disagreement should be seen as justifying the neutrality of the state. But before looking at this question, I want first to explain a little more carefully just what such neutrality means. It is a general truth that what the state does, the decisions it makes and the policies it pursues, will generally benefit some people more than others, and so some conceptions of the good life will fare better than others. This, I believe, is unavoidable. At the very least, those who desire a life of theft will find it rough going. And to take a more serious example, ways of life that depend upon close and exclusive bonds of language and culture-the French in Canada or the Bretons in France-may lose, within a liberal society also tolerating quite different and more open ways of life, some of the authority and cohesion that they would have if they formed complete societies unto themselves. Ideals of the good life that are open in principle to all, whatever one’s cultural inheritance (e.g., professional ideals), may well prove stronger under a liberal state. There is no getting around the fact that goods are not all fully compatible, that gains generally entail losses. But all of this does not impugn the neutrality of the liberal state. Its neutrality is not meant to be one of outcome, but rather one of procedure. That is, political neutrality consists in a constraint on what factors can be invoked to justify a political decision. Such a decision can count as neutral only if it can be justified without appealing to the presumed intrinsic superiority of any particular conception of the good life. So long as a government conforms its decisions to this constraint, therefore, it will be acting neutrally. There is no independently describable condition of society to be called “neutral” that the ideal of political neutrality requires a government to promote or maintain.
Patterns of Moral Complexity
49
Neutrality understood procedurally leaves open to a large extent the goals that the liberal state ought to pursue. Of course, some ends (e.g., the establishment of a state religion) are impermissible, because there can be no neutrally justifiable decision to pursue them. But any goals for whose pursuit there exists a neutral justification are ones that a liberal state may pursue. I do not think that the protection of life and property are the only goals that will satisfy this condition. The ideal of neutrality would not prevent a state from undertaking to ensure a particular pattern of wealth distribution, so long as the desirability of this pattern does not presuppose the superiority of some views of human flourishing over others held in society. (As I shall argue later, Rawls’s original position is best understood as a position of neutrality, so one might think here of his argument for the difference principle.) So a neutral state need not necessarily be a “minimal state.” Nineteenth-century liberals who demanded a minimal “nightwatchman” state (the Manchester school, for example) may have had many reasons-some intellectually responsible, some not-for this view, but their best reason was the idea that the free market is the most efficient and neutral means of producing and distributing wealth and resources. A contemporary liberal who desires a more active interventionist role for the state is best understood as still sharing their ideal of liberal neutrality, but as disagreeing with them about the nature of the free market. The theoretical possibility that a liberal state can be one that intervenes in social life (subject to the constraint of neutrality) is also, of course, a historical fact. The development of liberalism over the past century has involved both an increase in individual freedom and an increase in state power-and this should be paradoxical only to those who believe that power is a zero-sum game. This state intervention has not been equal, however, in all areas of social life: The liberal state has undertaken to manage the economy and redistribute wealth, but it has not interfered with the membership rules of churches, for example. How can the ideal of neutrality explain this fact? In order to answer this question, I must begin by emphasizing that for the liberal, neutrality is a political ideal. The state’s policies and decisions must be neutrally justifiable, but the liberal does not require that other institutions in society operate in the same spirit. Churches and firms, for example, may pursue goals (salvation, profits) that they assume to be ideals intrinsically superior to others. In other words, neutrality as a political ideal governs the public relations between persons and the state, and not the private relations between persons and other institutions. This is a very important feature of liberalism, and in Chapter 5 I shall discuss how antiliberal critics have generally failed to appreciate it. But here I am using this point to a different purpose. How can a neutral state justifiably treat firms and churches differently, taxing firms for the purposes of income redistribution and imposing on them fair
50
Charles Larmore
hiring procedures, while allowing churches to maintain restrictive membership rules? The answer lies in the fact that economic efficiency has been a neutrally justifiable goal in modern western societies, while personal salvation has not been one. To promote economic efficiency (in abstraction from how this may conflict with other goals) does not suppose the intrinsic superiority of some ideals of good life over others held in these societies.With the recognition that the free market is not adequately efficient, the liberal state could come justifiably to intervene in economic activity. Since, however, the state must regulate the distribution of wealth, in order to enhance efficiency, it should not allow distribution to be determined on nonneutral grounds. It cannot justifiably assume that the interests of the wealthy are intrinsically superior to those of the poor. This seems to me the way in which liberals ought to think about the justifiability of progressive income taxation and welfare legislation. I am not claiming that this form of justification was the most prominent in the recent rise of liberal state intervention; nor do I believe that all aspects of that intervention can be so justified or that all justifiable forms of such intervention have been tried. Specific questions of this sort are not my concern here. What I wish to emphasize is the general reason why a liberal state could intervene in one area of social life and not in another. It is not that the area of intervention does not itself exemplify neutrality. For a liberal neutrality is a political, not a general social ideal. The liberal state can intervene in an area of social life only if the state has a neutrally justifiable goal that requires that intervention, and only to the extent required by its pursuit of that goal can it justifiably institutionalize neutrality in that area. Liberalism as a political doctrine has generally been associated with the ideal of freedom. I think that this has been a mixed blessing because of all the different things that “freedom” can mean. One distinct advantage of making neutrality the primary ideal of liberalism is that it explains what freedom has generally meant for the political liberal. By denying the state any right to foster or implement any conception of the good life that some people reject, neutrality emphasizes the equal freedom that all persons should have to pursue their conception of the good life. This fact shows not only how the liberal state will differ from states whose ambition it is to promote some conception of the good life, but also that, for the purposes of liberal political theory, “freedom” has a sharply circumscribed sense. It covers only the right of the person not to face neutrally unjustifiable interference by the state. It conforms, in other words, to what Isaiah Berlin has called the “negative” conception of freedom, that we find in the writing of classical liberals such as Locke, Kant, Constant, Tocqueville, and Mill. Now the crucial point to recognize is that lib-
Patterns of Moral Complexity
51
erals need regard negative freedom only as a political ideal. They need not deny that as an ideal for the good life, freedom should mean more than the absence of a certain sort of governmental interference; positively, freedom must include, perhaps, self-realization or the availability of “meaningful” choices. However, just as ideals of the good life are a permanent object of dispute, so too are more substantial ideals of freedom, and so it cannot be the concern of the state to promote these. Berlin himself made it fairly clear that negative liberty is a political ideal and need not exhaust everything we mean by freedom, inasmuch as he refused to deny that positive liberty is also a value. The subsequent criticisms of Berlin’s work, however, have largely lost any sense of the need to determine what ought to be politically relevant in our manifold idea of freedom. There remains one last but important observation that I must make about the liberal ideal of political neutrality. It would be misunderstood, if it were thought to minimize the significant role that public discussion (or “Offentlichkeit”) should play in a liberal political culture. In particular, the ideal of political neutrality does not deny that such discussion should encompass not only determining what are the probable consequences of alternative decisions and whether certain decisions can be neutrally justified, but also clarifymg one’s notion of the good life and trying to convince others of the superiority of various aspects of one’s view of human flourishing. This ideal demands only that so long as some view about the good life remains disputed, no decision of the state can be justified on the basis of its supposed intrinsic superiority or inferiority. Why Neutrality?
The fact that different people hold different conceptions of the good life does not strictly entail the demand for a neutral state. Other responses are possible. The government might look to means of repression, or set up a lottery to pick the one ideal to be politically favored. Nonetheless, pluralism and disagreement about the good life do make political neutrality reasonable, and the nature of its reasonableness is what I wish to examine. In the Preface I discussed one of the central ambiguities of the liberal tradition. Liberalism has always urged toleration for the diversity of ideals and forms of life, but almost as often it has sought to justify this position by appealing to some particular and controversial view of human flourishing. Such justifications are not improper for those who can accept their premises. It cannot be denied that these have been very influential rationales for a liberal political order. Indeed, one of liberalism’s distinct strengths is the fact that there are so many different arguments appealing to different interests that converge in its favor.
52
Charles Larmore
But a liberalism come of age cannot rest content with these arguments. Its fundamental justification must be one that forgoes any appeal to the ideals whose controversial character sets the problem, after all, for political liberalism. It must be acceptable by reasonable people having different views of the good life, not just by those who share, for example, Mill’s ideal of the person. Thus we can distinguish in general two different ways in which pluralism and disagreement with regard to the good life can be made to justify political neutrality. The first will invoke some view of human flourishing that can best be promoted if government maintains a neutral posture toward the variety of human aspirations. The most familiar arguments of this sort appeal to the values of skepticism, experimentation, or individual autonomy, respectively: 1. When ideals clash, some people conclude that there is no reason to prefer any of them, and so no government should seek to institutionalize them. 2 . Other people believe that the best way to arrive at a firm conception of the good life is to participate in a number of different forms of life, comparing them and rejecting those that give us less of a sense of fulfillment. Such experimentalism would be hampered if government set out to foster only some of these ideals of the good life. 3. Others again hold that people cannot properly understand what it is to have a flourishing life unless they have worked out their ideal for themselves, making their own mistakes and learning from them. Hence, no government should undertake to make these decisions for people.
Each of these three arguments is well represented in the history of liberal thought and in our general culture. The skeptical argument underlies Voltaire’s plea for religious toleration. The appeal to experimentalism dominates Mill’s essay O n Liberty. And the insistence upon the value of autonomy, however differently construed, unites the thought of Mill and Kant. All three make up the core of the reasons that Ackerman has recently offered for neutrality.’ Each of these arguments, however, will prove unacceptable to those holding different but still reasonable views about the human good-to those, that is, who do not find skepticism the proper response to deep disagreement or who do not find the values of experiment and individual autonomy so important. The fact that a conviction of mine about the meaning of life is controversial, rejected even by others whom I consider reasonable, may not offer me a sufficient reason to suspend belief in it, if it continues to make sense of my experience.An experimental spirit may be alien and destructive to some forms of life, for example, religious orthodoxies, whose claim is that in them one is to be brought up from infancy, acquiring habits and expectations that are to last a lifetime. Au-
Patterns of Moral Complexity
53
tonomy, too, is a value that plays a role in some conceptions of the good, but not so prominently in all, I will have more to say about this in Chapter 5. These classical liberal arguments for neutrality will convince, therefore, only those who believe some things about the nature of human flourishing that others will not accept, and for reasons that are not without any merit. Thus when Ackerman asserts that liberal theory contains “an insistence that the forms of social life be rooted in the self-conscious value affirmations of autonomous individuals,” he is invoking one of liberalism’s most influential arguments for political neutrality, but he has not himself assumed a neutral position with respect to controversial ideals of the good life.2 Controversy about ideals of the good life and the demand that the state remain neutral toward them have been the central ingredients of the liberal vision of politics. This means that if liberals are to follow fully the spirit of liberalism they must also devise a neutral justification ofpolitical neutrality. This is a second sort of justification, one which is not easily to be found in the liberal tradition, but an imperative one for liberals to work out. In this section I shall outline how such a neutral justification would proceed, and then in subsequent sections I shall explore in more detail various aspects of the argument. The neutral justification of political neutrality is based upon what I believe is a universal norm of rational dialogue. When two people disagree about some specific point, but wish to continue talking about the more general problem they wish to solve, each should prescind from the beliefs that the other rejects, (1) in order to construct an argument on the basis of his other beliefs that will convince the other of the truth of the disputed belief, or (2) in order to shift to another aspect of the problem, where the possibilities of agreement seem greater. In the face of disagreement, those who wish to continue the conversation should retreat to neutral ground, with the hope either of resolving the dispute or of bypassing it. Thus abstracting from a controversial belief does not imply that one believes it any less, that one has had reason to become skeptical toward it. One can remain as convinced of its truth as before, but for the purposes of the conversation one sets it aside. Observe also that the aim of this maneuver is not to provide an opportunity for experiments in living nor to enhance the autonomy of the participants. The goal for the sake of which neutrality forms a response to disagreement is scarcely so controversial. One wants to keep the conversation going, in order to achieve some reasoned agreement about how to solve the problem at hand. Now it cannot be supposed that social life, and political activity in particular, is only conversation. Die Weltgeschichte ist kein Weltseminar. Nor can it be assumed that all conversations aim at reasoned agreement: Often they are also occasions for self-display,or means to dominate others, or simply ways to pass
54
Charles Larmore
the time. Nonetheless, conversation is not foreign to the political realm. The question of what principles ought to be the basis of the state’s decisions is the object of debate in almost every society. Conflicting views of the good life, in the absence of any recognized constraints, will produce conflicting answers to how this question should be resolved. The liberal will claim that the state’s decisions should accordingly be justifiable on ground neutral with respect to these conflicting views. And this neutrality of the state can itself be justified neutrally, if the norm of rational conversation already mentioned is applied to that political debate. Let us suppose that the participants in that debate accept that norm. Then each would retreat to neutral ground in order either to convince others of the truth of that disputed aspect of his own ideal of the good life, or to elaborate principles of state action upon this neutral basis itself, without resolving that dispute. To a large extent these conflicts about the nature of the good life will remain unresolved, so neutrally justifiable political principles will generally have to be found in this second, more abstract way. In either case, political principles will have to be justifiable neutrally with respect to controversial views of the good life. In this way the norm of rational conversation would serve to shape a political culture in which the public could continue to discuss disputed views about the good life with the hope of expanding the scope of agreement, but in which it would also agree that the state’s decisions cannot be justified by an appeal to the intrinsic superiority of any such view that remains disputed. Now clearly this argument is not morally neutral. It relies upon a commitment to converse rationally about what ought to be collectively binding political principles, and, as I shall show later, several other normative commitments are involved as well. But this is not a weakness, for the argument does not aim at complete moral neutrality. It intends to be neutral only with regard to controversial conceptions of the good life and not to all values or norms whatsoever. Although, as I shall have occasion to observe later, it is not completely neutral in this regard either, it is very nearly so, and certainly neutral enough for practical purposes. Most importantly, the argument is neutral with regard to those controversial ideals of the person (skepticism, experimentalism, autonomy) that earlier liberals often invoked, as well as to most other views of the good life. So it will be possible to reject those so-called liberal ideals of the person, and still regard neutrality as the fundamental political value. Equal Respect There is another unavoidable question that needs to be answered about this neutral justification of political neutrality. Different individuals with different
Patterns of Moral Complexity
55
conceptions of the good life can be understood as initially placing different demands upon what ought to be collectivelybinding political principles. In the face of this disagreement, I have said that a common norm of rational argument requires that if the individuals still want to talk about what political principles to establish, they retreat to neutral ground. But, we may ask, why should any of them feel obliged to continue the conversation?Why should they not resort to other means (force, deceit) of establishing political principles? In part this is not a very serious question. When those with whom we are in disagreement have views for which we nonetheless feel some sympathy, or possess along with others who share their views some significant amount of power, the reasons for continuing the conversation are obvious. Some sense of community and a desire for civil peace will suffice in these cases to keep the conversation going. But the question does take on a real importance when we consider why we should continue to talk with those who neither stand at least close to us in their views nor have enough power to make us pay attention. Why should we continue the conversation with them? After all, continuing the conversation with those who are similar to us or who are powerful will in the end commit us to finding political principles of already considerable neutrality. Why should we be ready to increase their neutrality for the sake of those who are so unlike us and so powerless? Let me pause to set down two conditions that an adequate answer to this question must fulfill. First, and most obviously, this justification of political neutrality can remain neutral only if the reasons for continuing the conversation with these people are neutral ones in the sense I described earlierneutral with respect to controversial ideals of the good life. For many fanatics and would-be martyrs, however, civil peace is not so important. Here, then, is a limit to the neutrality of my argument for political neutrality, but it is not, I think, a very grave one. Why must a political value be made justifiable to those who are scarcely interested in rational debate about justification anyway? A liberal political system need not feel obliged to reason with fanatics; it must simply take the necessary precautions to guard against them. Our present question is whether there exist similarly neutral reasons for continuing to talk with those who lack the power to call on our desire for civil peace and whose ideals seem quite foreign to ours. Once again, I am not denying that some conceptions of the good life would encourage further talk with them, and that such justifications are fine for those persuaded by them. But my object is a different sort of argument. The second condition is that the reasons for continuing the conversation cannot be expected to follow from the notion of rationality alone. To an important extent, my neutral justification of political neutrality has relied upon what I have called a norm of rational conversation. But there are limits to what
56
Charles Larmore
the bare idea of rationality can establish. Even if it constrains how a conversation should develop, it cannot alone justify that the conversation be undertaken. Thus, the reasons for continuing the conversation in the face of disagreement will have to embody more substantive commitments than just a commitment to reason. I have already invoked two such further commitments: sympathy with those whose ideals are similar to ours, and a desire for civil peace. These are rather substantive commitments, although they are largely neutral with respect to controversial ideals of the good life. The problem at hand is whether there is some additional commitment, of a similar neutrality, that would justify our continuing the conversation with those who are strange and weak. It must be a commitment that is morally more substantive than the bare idea of rationality, but that is not part of any disputed notion of the good life. The answer I propose is that the neutral reason for continuing the conversation with them must lie in the wish to show everyone equal respect3
* * * To have respect for a person is to view him as capable of elaborating beliefs that we would r e ~ p e c t . ~ Now there are two significant differences between respect for beliefs and respect for persons. Respect for beliefs is simply a belief about their justifiability within the other’s perspective; so it is not itself an obligation we could bear toward others. Respect for persons, however, is not just a belief that others have a capacity for developing beliefs justifiable within their own perspective, and it cannot be simply deduced (even by “analysis of concepts”) from that fact; it is an obligation to treat others in a certain way because of that fact (in what way, I shall come to shortly). Furthermore, respect for beliefs is not something that anyone can be expected to accord equally to all. Some beliefs deserve it, others do not. By contrast, a capacity for working out a coherent view of the world is one that everyone (except some of the clinically insane) possesses. So respect for persons, as an attitude involving recognition of this capacity, is something that we can show equally to others. Of course, some people have this capacity to a greater degree than others do, but respect is something that others as persons are due just by virtue of having that capacity, so it should be given equally to aL5 What is precisely then the obligation of equal respect that we should show others in virtue of their having the capacity to work out their own view of the world? Whatever we do that affects another is something with which he must deal from within his own perspective. When he demands that we justify our action to him, he is recognizing that we, too, have a perspective on the world
Patterns of Moral Complexity
57
in which presumably our action makes sense, and indicating his willingness to discuss it rationally with us (of course, his notion of rational discussion may differ substantially from ours, as I noted earlier). The obligation of equal respect consists in our being obligated to treat another as he is treating us-to use his having a perspective on the world as a reason for discussing the merits of our action rationally with him (in the light of how we understand a rational discussion). This is the way in which equal respect involves mutual respect, as Hegel understood. I have discussed respect at such length for two reasons. First, previous treatments of respect have usually been rather vague, passing over the difference and connection between the two sorts of respect, as well as the rationale for the obligation of equal respect. Just as importantly, we can now see why the obligation of equal respect will lead one to continue the conversation about political principles even with those who lack other means (similarity or power) of interesting one in this. To show another equal respect is to treat his demand for justification as part of a rational discussion one must have with him. Even if two people disagree significantly about what would be the ideal conditions under which this discussion should proceed, they will both agree that this disagreement is reason to retreat to neutral ground and proceed from there. Does equal respect, so understood, satisfy the two conditions I laid down at the beginning? It is certainly not an obligation we should accept just by virtue of being rational beings, for there is nothing contrary to reason in refusing to treat others as they treat us. So the second condition is satisfied. What about the first? Is equal respect a neutral reason, in the sense of being impartial between conflicting views of the good life? I believe that it is largely neutral. Respect for persons is an attitude that we can adopt, just about however much we may disagree with them about the nature of the good life. It certainly seems clear that when equal respect requires us to treat acceptingly those who disagree with us, it is not for the obviously nonneutral reasons that classical liberals have often given for adopting the stance of political neutrality. Others are due equal respect by virtue of their capacity for working out a coherent view of the world and indeed of the good life, whether or not they exercise the capacity autonomously and experimentally, or through the uncritical acceptance of traditions and forms of life. Equal respect does not demand that we approve of people working out their ideals autonomously or in an experimentalist spirit. We could repudiate the value of autonomy or experimentalism and take steps to discourage it in others, while still consistently feeling obliged to explain the merits of these actions to those who took exception to them. I do not mean to suggest that a commitment to equal respect has been a perennial feature of human culture. Clearly it has not been such, and there are
58
Charles Larmore
surely specific historical conditions that make this commitment relatively widespread today. I wish to insist only that it is compatible with a very great variety of ideals of the good life, including those that were dominant at times when, as a matter of fact, the norm of equal respect was not so widely shared. For example, the belief that some particular hereditary class produces those most fit to rule does not exclude equal respect, for one could feel obliged (as some such aristocrats were) to justify this belief to the other classes. However, some views of the good life-some especially virulent forms of racism, for example-must reject the obligation of equal respect in order to remain consistent. But once again I do not believe that this forms a significant limit to the neutrality of my argument. Liberals need not have an argument to convince people of this sort, only safeguards for preventing them from acquiring political power. After all, such people seem little interested in rational argument. Let me close, nonetheless, on a more modest note. Whatever the limits to its neutrality, equal respect is neutral with respect to the ideals of skepticism, experimentalism, and autonomy. Consequently, the justification of political neutrality that I have developed is more neutral than the classical justification that appealed to those ideals, and so-most importantly-it can work for those who do not share those ideals. As I shall show in the next chapters, the controversial character of these ideals has been a central reason for opposition to liberalism. My neutral justification of liberalism seems the best way for liberals to meet this form of criticism. Practical Limits This completes the neutral justification of political neutrality. A commitment to treating others with equal respect forms the ultimate reason why in the face of disagreement we should keep the conversation going, and to do that, of course, we must retreat to neutral ground. There remains, however, one final objection to this argument for the distinctive features of a liberal political order. Even if this argument is faultless, some will complain, the obvious fact is that “neutral ground” offers too weak a basis for deriving any political principles that assign basic liberties and distribute wealth. Neutrality is powerless, the objection claims, as a method for solving the fundamental problem of politics. As a result, actual political systems that call themselves liberal must be either confused or disingenuous. This objection is serious, but it should not ultimately discomfit the liberal. We should first observe that political neutrality, as I have described it, is a relative matter. It does not require that the state be neutral with respect to all
Patterns of Moral Complexity
59
conceptions of the good life, but only with respect to those actually disputed in the society. Where everyone agrees about some element of human flourishing, the liberal should have no reason to deny it a role in shaping political principles. Nonetheless, the objection could be reformulated to say that in any society such as ours where disagreement about the good life extends so deeply, even a relative neutrality will lack the content necessary for elaborating political principles. The most direct way to deal with this objection, of course, would be to examine how far this is true of particular societies. But I believe that the proper response should remain at a higher level of generality. There is no reason, I think, why a liberal should have to dispute the fact that full neutrality in a modern society may prove too empty to generate any substantive political principles. Neutrality, however, is not the liberal’s only desideratum. There is also the need for some decision to be made about what should be the principles governing basic liberties and distribution. So the liberal must be willing to consider tradeoffs between these two goals. If full neutrality makes such a decision impossible, then neutrality should be made more restrictive until a decision does become possible. The important point for the liberal is that such tradeoffs can still be devised in the spirit of neutrality. Guiding them will be a principle embodying a higher neutrality; namely, that one should institute only the least abridgement of neutrality necessary for making a decision possible. Thus, if a certain view about the good life with which some people disagree is nonetheless admitted by the liberal as part of the basis for deciding about substantive political principles, this will not be because he assumes it to be true. Instead, this will be because he believes that admitting it will constitute the least restriction of neutrality necessary for arriving at an adequate basis for the decisions needing to be made. In what, however, does the “least restriction” consist? There appear to be two different dimensions along which a minimum, in the spirit of neutrality, could be sought: 1. One could admit beliefs that are the least central to anyone’s idea of the
good life, or
2. One could admit beliefs that the least number of people do not hold.
The first dimension is the more neutral alternative, but in many cases it may prove impossible to implement, and then one must resort to the second. Political neutrality, therefore, can be both a justifiable and a practicable ideal. My chief concern in this chapter has been to show how this ideal can be justified neutrally, without need for appeal to any disputed notion of the good
60
Charles Larmore
life. I must insist once again that I do not hold a neutral justification to be the only justification for neutrality. But I do consider it the only mode of justification fully within the liberal spirit. For the fact that lies at the heart of liberalism is that reasonable people differ a n d disagree about the nature of the good life.
Notes 1. See Bruce Ackerman, Social Justice in the Liberal State (New Haven, Conn.: Yale University Press), p. llff. Ackerman’s fourth and remaining argument is that even if one thinks one knows what the good life is and that it is of a kind that can be forced on others, one cannot be sure that the right people will be doing the forcing. This seems to me rather weak. Often one could be sure of that, and even if there was some doubt, one might reasonably be willing to take the risk; furthermore, laws and regulations could be instituted to lessen the likelihood of abuse of power. 2. Ibid., p. 196. Ackerman seems aware of this fact, since he has more recently argued that “it would be a categorical mistake to imagine that there could be a Neutral justification for the practice of Neutral justification.’’ (Idem, “What Is Neutral about Neutrality?” Ethics 93 (1983): 387.) With this I disagree. 3. In “Liberalism,”113-43, Dworkin also connects the idea of neutrality with that of equal respect, although without any argument why “the government does not treat them [citizens] as equals if it prefers one conception of the good life to another” (p. 127).That is what my neutral justification of political neutrality tries to explain. 4. The relation between these two notions of respect has received far less attention than it deserves. For discussions of the distinction itself that are similar, though not identical, to mine, see Stephen Darwall, “Two Kinds of Respect,”Ethics 88 (1977): 36-49; and Ernst Tugendhat, Probleme der Ethik, Reclam 1984. 5. On the connection between respect for persons and equal respect, see Bernard Williams, Problems of the Self (Cambridge: Cambridge University Press, 1973),234ff.
4 Liberal Virtues Stephen Macedo
Introduction Critics of liberalism frequently view it in exclusively negative terms, as tolerating everything and standing for nothing. Against this, Macedo expounds a robust conception of liberalism, “an attractive vision of a distinctively liberal community and liberal virtues.” He argues that particular values and institutions are central to the liberal state and pervasively affect its citizens. Because of these moral commitments, claims of neutrality are misguided. In particular, Macedo criticizes Larmore’s defense of neutrality based on a norm of rational dialogue. Reasonable conversation is not neutral, Macedo argues. On closer examination, one can detect distinctively liberal values and virtues shining through.
Stephen Macedo, Liberal Virtues: Citizenship, Virtue and Community in Liberal Constitutionalism (Oxford: Clarendon, 1990), 256-63. Copyright 0 1990 by Stephen Macedo. Reprinted from Liberal Virtues: Citizenship, Virtue and Community in Liberal Constitutionalism by Stephen Macedo (1990) by permission of Oxford University Press.
-61
-
Stephen Macedo
62
Pluralism and Liberal Unity
L
conflict as the basic fact of social life, taking their cues from Hobbes. Liberal toleration of individual liberty is a necessity, we are told, because people disagree about what is good in life: people disagree about their religious commitments and other goals and ideals. And from the purportedly basic fact of disagreement, liberal political imperatives are characterized in negative terms: avoiding injustice and tyranny, keeping rules of law purposeless or non-instrumental. Or liberals emphasize the fundamental importance of keeping government ‘neutral’or impartial between parties who disagree about what is good in life (as liberals like Dworkin, Ackerman, and Larmore have put it).’ Or, sometimes, liberalism is taken to stand for nothing more definite than toleration and a ‘spirit of accommodation’. While the permanent fact of pluralism is the heart of the liberal political problem, conflict and disagreement are not the basic facts of liberal social life. It is wrong to think that liberal law is, can be, or even ought to be, in any strong sense purposeless, non-instrumental, or neutral with regard to conceptions of the good life, either at the level of society as a whole or at the level of individual lifeplans. Likewise, if liberalism stands for mere toleration or an indiscriminate spirit of accommodation, then it stands for everything, and it takes a stand for nothing. These errors are related to the vulnerability of liberalism, as it is commonly understood, to the communitarian critique of liberalism. A political theory that appears to rest on the centrality of disagreement or conflict or self-interest, and that characterizes its basic political values in negative terms or in terms of neutrality or mere accommodation, has a kind of hardheaded quality to it; it modestly avoids reliance on contestable ideals. But to sap liberal politics of positive values in any of these ways is to underdescribe what liberalism stands for, even in its minimal cases, and it leads us to misunderstand what it means to be a liberal. When liberalism is underdescribed in these ways, a great gap looms between hard-headed, non-ideal cases of liberalism, and positive liberal ideals. When we underdescribe and misdescribe the authoritative public values of a liberal regime we also fail to see the extent to which liberalism shapes the lives of liberal citizens. Even in its unidealized forms liberalism cannot really be neutral among public values-it stands for the supreme worth of certain public values: of individual liberty and responsibility, of tolerance of change and diversity, and of respect for the rights of those who themselves respect liberal values.2 Liberalism requires the support of positive values to explain why respect is owed to persons and to justify their having the rights they do. Liberalism stands, above all, for the positive value of freedom, freedom to devise, criticize, revise, and IBERALS OFTEN REGARD ‘DISAGREEMENT’ OR
Liberal Virtues
63
pursue a plan of life, and it calls upon people to respect the rights of others whether or not they share the same goals and ideals. Liberalism, as I argued in Chapter 2, rules out certain conceptions of the good life altogether: any that entail the violation of liberal rights. And liberalism positively requires that everyone’s scheme of values include certain features: respect for the equal rights of others, a willingness to persuade rather than coerce, the subordination of personal plans, projects, and desires to impersonal rules of law, and a contribution to the provision of public goods. The colouring of liberal values splashes pervasively over the vast canvass of a pluralistic liberal society. Some things are excluded completely, and everything is limited and conditioned. Besides a set of positive values, liberalism represents a set of political institutions: a system of legal rules establishing order and giving substance and specificity to the ideal of equal freedom, representative institutions, and courts of law to adjudicate disputes, test the reasonableness of government actions, and enforce constitutional limitations even against the people’s representatives. All liberal governments provide certain public goods-at a bare minimum, the apparatus of national defence, courts and police for justice (they are governments after all). In addition, encouragements to commercial activity, public schools, and poverty relief are almost always provided by liberal governments nowadays. These are not ‘neutral’goods, but nearly all liberals would agree that some range of such goods should be provided by a coercive state. (And liberal societies also have semi-public institutions that help define and enforce liberal values, such as universities and the professions of journalism and law.) So liberals have positive political values and political institutions and practices designed to embody and sustain these values. What liberals often fail to recognize is that these values, institutions, and practices exert a pervasive effect on the lives and the character of liberal citizens. Because liberalism is not neutral (either in its direct or its indirect consequences), and because liberal citizens may (and should) affirm and act from liberal values, we can speak of a distinctive liberal ‘character’. And yet, one might object, decidedly non-liberal groups do survive within polities structured by liberal values, such as the small, but not insignificant, groups of American Nazis. But let us consider what is required of groups like the Nazis if they are to live peaceably in a liberal regime. They must, first of all, respect the rights of those they hate or else suffer at the hands of the law. They must, that is, respect the property, the political rights, and freedoms of Jewish Americans. They may, occasionally, march in Jewish communities, but they must get permits, keep order, and otherwise respect the peace and quiet of these neighborhoods. They can gather in uniforms, with broadsheets, slogans, music, and other paraphernalia, in legally rented private halls, as long as they do not make too much noise. Nazis must pay taxes to support the liberal
64
Stephen Macedo
institutions they detest, including public schools. The liberal polity requires that the Nazis be law-abiding Nazis and that is not easy. They cannot be ‘gungho’ Nazis, in fact they cannot be Nazis at all but only play at it.3 That a liberal society makes life hard on Nazis is not a matter for regret. But far less disagreeable and perhaps even admirable groups may find an open, pluralistic environment less than hospitable. Those who favor simple ways may be disorientated by the pace of change and movement. The devout are liable to object to the materialism and license of a liberal society. Among those prepared to go along with the liberal settlement not all will readily embrace liberal virtues such as autonomy. The Mirage of Liberal Neutrality It is, as I have already argued, impossible to sustain the claim that liberalism does not have important and non-neutral effects on which ways of life flourish and gain adherents in a liberal regime. More sophisticated versions of the neutrality thesis do not, however, argue for neutrality of consequences. Liberal neutrality, Larmore argues, ‘is not meant to be one of outcome, but rather one of procedure. That is, political neutrality consists in a constraint on what factors can be invoked to justify a political de~ision’.~ The sophisticated neutralist would carefully restrict the kinds of reasons that are morally admissible grounds for government action, requiring that, ‘The state should not seek to promote any particular conception of the good life because of its presumed intrinsic superiority-that is, because it is supposedly a truer con~eption’.~ The liberal state may restrict ideals of life only ‘for extrinsic reasons because, for example, they threaten the lives of others? The neutral state may not seek to ‘foster or implement any conception of the good life that some people reject’, for only then can it respect ‘the equal freedom that all persons should have to pursue their conception of the good life’.7 Governmental neutrality expresses equal respect for moral persons who are defined by their reflective capacity for a conception of the good life and not by the particular conception they choose. A neutral justification for neutrality can be found, Larmore claims, in ‘a universal norm of rational dialogue’.8 When reasonable people disagree each should, out of concern with keeping the conversation going, ‘prescind from the beliefs the other rejects’, and ‘retreat to neutral ground, with the hope either of resolving the dispute or of bypassing What moves both sides is simply a wish to keep the conversation going until they can find grounds for agreement that neither could reasonably reject. The consistent neutralist will insist on neutral reasons ‘all the way down’, as it were, or all the way through the reasoned conversation of politics. But
Liberal Virtues
65
how neutral is reasonable conversation? Not very: some people will find head-bashing easier and more satisfying than reason-giving. And so, Larmore’s ideal of reasonable conversation purports to be ‘neutral only with regard to controversial conceptions’, and not even ‘completely neutral in this regard either, [but] it is very nearly so, and certainly neutral enough for practical purposes’.1° Neutral enough for government work turns out to be not very neutral at all. Liberal ‘neutrality’ stands for mutal respect among people committed at base to values that are by no means neutral. Larmore would not have us respect everyone, but only those who are reasonable. Among those who will not be respected are ‘fanatics and would-be martyrs’ for whom ‘civil peace is not so important’, racists, and all those who reject ‘the obligation of mutual respect’.” The bounds and nature of neutral respect are defined by liberal political norms and substantive criteria of reasonableness, criteria that are by no means neutral with respect to ideals of life. None of this is surprising since, as Larmore admits, ‘neutrality’is ‘too empty to generate any substantive political prinicples’.12 The closer we get, the faster the mirage of neutrality vanishes. Larmore finally advocates only what he calls ‘the spirit of neutrality’,or ‘a higher neutrality; namely, that one should institute only the least abridgment of neutrality necessary for making a decision possible’.” Larmore’s neutrality, like any political neutrality, is selective, and selective neutrality is not very neutral. The commitment to political neutrality depends on the acceptance of a conception of social life in which citizens take seriously a public project of justification: these citizens wish to conduct a reasonable conversation about political morality with other people who share a basic interest in such a conversation, or in a process of public justification. It seems to me right to think of this commitment as the moral core of liberalism but wrong to think that this core is in any important sense morally neutral. The popularity of neutrality-talk is the consequence, I think, of two factors: a laudable concern to respect reasonable people who disagree widely over their conceptions of the good life, and a desire to facilitate agreement by describing liberalism modestly and uncontroversially. Neutrality builds on principles that are central to liberalism, but from them it erects an excessively strong ban on judgments about human ideals. Liberals properly deploy reasons that can be widely seen to be reasonable, and liberals believe in respect for all those who pass the threshold requirements of reasonableness. Liberals resist paternalism, and minimize interference with people’s choices. These do not, however, add up to neutrality. Liberal restrictions on the reasons that can be offered to support government actions are not strict enough to constitute a commitment to neutrality. Liberal ideals and virtues
66
Stephen Macedo
remain in the background, never quite out of view. Autonomy is, from a liberal perspective, not simply one ideal of life among others (though it will sometimes be treated as such). It is not to be regretted that liberal institutions help, in modest and gentle ways, to promote the ideal of autonomy, for that capacity helps make people more competent as liberal citizens and better able to flourish as persons in a liberal society. Notes 1. Ronald Dworkin, “Liberalism,” 191; Bruce Ackerman, Social Justice in the Liberal State, 11; Charles Larmore, Patterns of Moral Complexity, passim. 2. See Kant, The Metaphysics of Morals, Part 2: “The Metaphysical Principles of Virtue,” in Kant’s Ethical Philosophy, trans. James W. Ellington, intro. Warner A. Wick (Indianapolis: Hackett, 1986), and Judith Shklar, Ordinary Vices (Cambridge, Mass.: Harvard University Press, 1984), 240115. 3. Groups like the Nazis do raise serious problems, regarding the education of their children for instance. My point here is that liberal regimes importantly structure the lives and activities of such groups. 4. Larmore, Patterns of Moral Complexity, 44. 5. Ibid., 43. 6. Ibid. 7. Ibid., 46. 8. Ibid., 53. 9. Ibid. 10. Ibid., 55. 11. Ibid., 60. 12. Ibid., 67. 13. Ibid., 68.
5 Political Liberalism John Rawls
Introduction In his writings since A Theory of Justice (1971), Rawls has sought to show how his conception of justice is compatible with the ineradicable pluralism of contemporary liberal democratic societies. Central to this project is the idea that all citizens should be able to accept the fundamental principles that regulate the basic structure of their society. To explain how this is possible, Rawls introduces a distinction between comprehensive views and a political conception of justice. The former includes “conceptions of what is of value in human life” and “ideals of personal virtue and character.” The latter refers to a conception of justice that is worked up from certain fundamental intuitive ideas that are present in the public political culture of liberal democracies. Rawls contends that only a political conception of justice, one that is independent from any comprehensive view, can secure the allegiance of all reasonable citizens in a liberal democratic society. In the selection included here, Rawls addresses the connection between a political conception of justice and state neutrality. He argues that a political conception of justice, such as justice as fairness, is not strictly neutral among all conceptions of the good. It will disadvantage some conceptions by excluding them altogether and disadvantage others by not allowing conditions that would enable them to survive. Notwithstanding this, a political conception of justice does aspire to a kind of
- 67 -
68
lohn Rawls
neutrality. It strives to be neutral among comprehensive views that can flourish in a democratic society in which basic individual rights are respected. John Rawls, Political Liberalism (New York Columbia University Press, 1993), 191-200. Copyright 0 1993 by Columbia University Press. Reprinted by permission of Columbia University Press.
Political Liberalism
69
Permissible Conceptions of the Good and Political Virtues 1. Historically one common theme of liberal thought is that the state must not favor any comprehensive doctrines and their associated conception of the good. But it is equally a common theme of critics of liberalism that it fails to do this and is, in fact, arbitrarily biased in favor of one or another form of individualism. As I noted at the outset, the assertion of the priority of right may seem to leave justice as fairness (as a form of political liberalism) open to a similar objection. Thus, in discussing the next two ideas-those of permissible conceptions of the good (those permitted by the principles of justice) and of the political virtues-I shall use the familiar idea of neutrality as a way of introducing the main problems. I believe, however, that the term neutrality is unfortunate; some of its connotations are highly misleading, others suggest altogether impracticable principles. But with due precautions taken, and using it only as a stage piece, as it were, we may clarify how the priority of right connects with the above two ideas of the good. 2 . Neutrality can be defined in quite different ways.' One way is procedural, for example, by reference to a procedure that can be legitimated, or justified, without appealing to any moral values at all. Or if this seems impossible, since showing something justified appears to involve an appeal to some values, a neutral procedure may be said to be one justified by an appeal to neutral values, that is, to values such as impartiality, consistency in application of general principles to all reasonably related cases (compare to: cases similar in relevant respects are to be treated similarly),2and equal opportunity for the contending parties to present their claims. These are values that regulate fair procedure for adjudicating, or arbitrating, between parties whose claims are in conflict. The specification of a neutral procedure may also draw on values that underlie the principles of free rational discussion between reasonable persons fully capable of thought and judgment, and concerned to find the truth or to reach reasonable agreement based on the best available inf~rmation.~ 3. Justice as fairness is not procedurally neutral. Clearly its principles of justice are substantive and express far more than procedural values, and so do its political conceptions of society and person, which are represented in the original position. As a political conception it aims to be the focus of an overlapping consensus. That is, the view as a whole hopes to articulate a public basis of justification for the basic structure of a constitutional regime working from fundamental intuitive ideas implicit in the public political culture and abstracting from comprehensive religious, philosophical, and moral doctrines. It seeks common ground-or if one prefers, neutral ground-given the fact of pluralism. This common ground is the political conception itself as the focus
70
John Rawls
of an overlapping consensus. But common ground, so defined, is not procedurally neutral ground. A very different way of defining neutrality is in terms of the aims of basic institutions and public policy with respect to comprehensive doctrines and their associated conceptions of the good. Here neutrality of aim as opposed to neutrality of procedure means that those institutions and policies are neutral in the sense that they can be endorsed by citizens generally as within the scope of a public political conception. Thus, neutrality might mean for example. a. that the state is to ensure for all citizens equal opportunity to advance any conception of the good they freely affirm; b. that the state is not to do anything intended to favor or promote any particular comprehensive doctrine rather than another, or to give greater assistance to those who pursue it;4 c. that the state is not to do anything that makes it more likely that individuals accept any particular conception rather than another unless steps are taken to cancel, or to compensate for, the effects of policies that do this.5 The priority of right excludes the first meaning of neutrality of aim, for it allows that only permissible conceptions (those that respect the principles of justice) can be pursued. The first meaning can be amended to allow for this; and as so amended, the state is to secure equal opportunity to advance any permissible conception. In this case, depending on the meaning of equal opportunity, justice as fairness may be neutral in aim. As for the second meaning, it is satisfied in virtue of the features of a political conception expressing the priority of right: so long as the basic structure is regulated by such a view, its institutions are not intended to favor any comprehensive doctrine. In regard to the third meaning, however, it is surely impossible for the basic structure of a just constitutional regime not to have important effects and influences as to which comprehensive doctrines endure and gain adherents over time; and it is futile to try to counteract these effects and influences, or even to ascertain for political purposes how deep and pervasive they are. We must accept the facts of commonsense political sociology. To summarize: we may distinguish procedural neutrality from neutrality of aim; but the latter is not to be confused with neutrality of effect or influence. As a political conception for the basic structure justice as fairness as a whole tries to provide common ground as the focus of an overlapping consensus. It also hopes to satisfy neutrality of aim in the sense that basic institutions and public policy are not to be designed to favor any particular comprehensive doctrine? Neutrality of effect or influence political liberalism abandons as impracticable, and since this idea is strongly suggested by the term itself, this is a reason for avoiding it.7
Political Liberalism
71
4. Even though political liberalism seeks common ground and is neutral in aim, it is important to emphasize that it may still affirm the superiority of certain forms of moral character and encourage certain moral virtues. Thus, justice as fairness includes an account of certain political virtues-the virtues of fair social cooperation such as the virtues of civility and tolerance, of reasonableness and the sense of fairness. The crucial point is that admitting these virtues into a political conception does not lead to the perfectionist state of a comprehensive doctrine. We can see this once we are clear about the idea of a political conception of justice. As I have said (in §I), ideas of the good may be freely introduced as needed to complement the political conception of justice, so long as they are political ideas, that is, so long as they belong to a reasonable political conception of justice for a constitutional regime. This allows us to assume that they are shared by citizens and do not depend on any particular comprehensive doctrine. Since the ideals connected with the political virtues are tied to the principles of political justice and to the forms of judgment and conduct essential to sustain fair social cooperation over time, those ideals and virtues are compatible with political liberalism. They characterize the ideal of a good citizen of a democratic state-a role specified by its political institutions. In this way the political virtues must be distinguished from the virtues that characterize ways of life belonging to comprehensive religious and philosophical doctrines, as well as from the virtues falling under various associational ideals (the ideals of churches and universities, occupations and vocations, club and teams) and of those appropriate to roles in family life and to the relations between individuals. Thus, if a constitutional regime takes certain steps to strengthen the virtues of toleration and mutual trust, say by discouraging various kinds of religious and racial discrimination (in ways consistent with liberty of conscience and freedom of speech), it does not thereby become a perfectionist state of the kind found in Plato or Aristotle, nor does it establish a particular religion as in the Catholic and Protestant states of the early modern period. Rather, it is taking reasonable measures to strengthen the forms of thought and feeling that sustain fair social cooperation between its citizens regarded as free and equal. This is very different from the state’s advancing a particular comprehensive doctrine in its own name.8
Is Justice as Fairness Fair to Conceptions of the Good? 1. The principles of any reasonable political conception must impose restrictions on permissible comprehensive views, and the basic institutions
72
John Rawls
those principles require inevitably encourage some ways of life and discourage others, or even exclude them altogether. Thus, the question arises over how the basic structure (realized by a political conception) encourages and discourages certain comprehensive doctrines and ways of life associated with them and whether how it does this is just. Considering this question will explain the sense in which the state, at least as concerns constitutional essentials, is not to do anything intended to favor any particular comprehensive view.9At this point the contrast between political and comprehensive liberalism becomes clear and fundamental.l0 The encouraging or discouraging of comprehensive doctrines comes about for at least two reasons: their associated ways of life may be in direct conflict with the principles of justice; or else they may be admissible but fail to gain adherents under the political and social conditions of a just constitutional regime. The first case is illustrated by a conception of the good requiring the repression or degradation of certain persons on, say, racial, or ethnic, or perfectionist grounds, for example, slavery in ancient Athens, or in the antebellum South. Examples of the second case may be certain forms of religion. Suppose that a particular religion, and the conception of the good belonging to it, can survive only if it controls the machinery of state and is able to practice effective intolerance. This religion will cease to exist in the well-ordered society of political liberalism. Let us assume there are such cases; and that some other comprehensive doctrines may endure but always among relatively small segments of society. 2. The question is this: if some conceptions will die out and others survive only barely in a just constitutional regime, does this by itself imply that its political conception of justice is not fair to them? Is the political conception arbitrarily biased against these views or better, is it just or unjust to the persons whose conceptions they are, or might be? Without further explanation, it would not appear to be unfair to them, for social influences favoring some doctrines over others cannot be avoided by any view of political justice. No society can include within itself all forms of life. We may indeed lament the limited space, as it were, of social worlds, and of ours in particular; and we may regret some of the inevitable effects of our culture and social structure. As Berlin has long maintained (it is one of his fundamental themes), there is no social world without loss: that is, no social world that does not exclude some ways of life that realize in special ways certain fundamental values. The nature of its culture and institutions proves too uncongenial." But these social necessities are not to be taken for arbitrary bias or injustice. The objection must go further and hold that the well-ordered society of political liberalism fails to establish, in ways that existing circumstances allowcircumstances that include the fact of reasonable pluralism-a just basic struc-
Political Liberalism
73
ture within which permissible forms of life have a fair opportunity to maintain themselves and to gain adherents over generations. But if a comprehensive conception of the good is unable to endure in a society securing the familiar equal basic liberties and mutual toleration, there is no way to preserve it consistent with democratic values as expressed by the idea of society as a fair system of cooperation among citizens viewed as free and equal. This raises, but does not of course settle, the question of whether the corresponding way of life is viable under other historical conditions, and whether its passing is to be regretted.12 Historical experience shows that many ways of life pass the test of enduring and gaining adherents over time in a democratic society; and if numbers are not the measure of success-and why should they be?-many pass that test with equal success: different groups with distinctive traditions and ways of life find different comprehensive views fully worthy of their allegiance. Thus, whether political liberalism is arbitrarily biased against certain conceptions and in favor of others turns on whether, given the fact of reasonable pluralism and the other historical conditions of the modern world, realizing its principles in institutions specifies fair background conditions for different and even antagonistic conceptions of the good to be affirmed and pursued. Political liberalism is unjustly biased against certain comprehensive conceptions only if, say, individualistic ones alone can endure in a liberal society, or they so predominate that associations affirming values of religion or community cannot flourish, and moreover the conditions leading to this outcome are themselves unjust, in view of present and foreseeable circumstances. 3. An example may clarify this point: various religious sects oppose the culture of the modern world and wish to lead their common life apart from its unwanted influence. A problem now arises about their children’s education and the requirements the state can impose. The liberalisms of Kant and Mill may lead to requirements designed to foster the values of autonomy and individuality as ideals to govern much if not all of 1ife:But political liberalism has a different aim and requires far less. It will ask that children’s education include such things as knowledge of their constitutional and civic rights so that, for example, they know that liberty of conscience exists in their society and that apostasy is not a legal crime, all this to insure that their continued membership when they come of age is not based simply on ignorance of their basic rights or fear of punishment for offenses that do not exist. Moreover, their education should also prepare them to be fully cooperating members of society and enable them to be self-supporting; it should also encourage the political virtues so that they want to honor the fair terms of social cooperation in their relations with the rest of society. Here it may be objected that requiring children to understand the political conception in these ways is in effect, though not in intention, to educate them
74
John Rawls
to a comprehensive liberal conception. Doing the one may lead to the other, if only because once we know the one, we may of our own accord go on to the other. It must be granted that this may indeed happen in the case of some. And certainly there is some resemblance between the values of political liberalism and the values of the comprehensive liberalisms of Kant and Mill. But the only way this objection can be answered is to set out carefully the great differences in both scope and generality between political and comprehensive liberalism. The unavoidable consequences of reasonable requirements for children’s education may have to be accepted, often with regret. I would hope the exposition of political liberalism in these lectures provides a sufficient reply to the objection. 4.Beyond the requirements already described, justice as fairness does not seek to cultivate the distinctive virtues and values of the liberalisms of autonomy and individuality, or indeed of any other comprehensive doctrine. For in that case it ceases to be a form of political liberalism. Justice as fairness honors, as far as it can, the claims of those who wish to withdraw from the modern world in accordance with the injunctions of their religion, provided only that they acknowledge the principles of the political conception of justice and appreciate its political ideals of person and society. Observe here that we try to answer the question of children’s education entirely within the political conception. Society’s concern with their education lies in their role as future citizens, and so in such essential things as their acquiring the capacity to understand the public culture and to participate in its institutions, in their being economically independent and self-supporting members of society over a complete life, and in their developing the political virtues, all this from within a political point of view. Notes 1. This does not deny that there may be, depending on the comprehensive view, perfectly good reasons in other kinds of cases. 2. A number of these I discuss in the text. One I don’t take up is William Galston’s view that some forms of liberalism are neutral in the sense that they use no ideas of the good at all except ones that are purely instrumental (neutralmeans, as it were).See his “Defending Liberalism,”in American Political Science Review 72 (September 1982): 622. Contrary to his suggestion, justice as fairness is not neutral in this way, as will become clear, if it is not clear already. 3. Thus Herbert Wechsler, in his well-known discussion of principled judicial decisions (he is concerned mainly with decisions of the Supreme Court), thinks of neutral principles as those general principles that we are persuaded apply not only to the present case but to all reasonably foreseeable related cases likely to arise given the con-
Political Liberalism
75
stitution and the existing political structure. Neutral principles transcend the case at hand and must be defensible as widely applicable. Weschler says little about the derivation of such principles from the constitution itself or from precedent. See “Towards Neutral Principles of Constitutional Law,” in Principles, Politics, and Fundamental Law (Cambridge, Mass.: Harvard University Press, 1961). 4. For this kind of view, see the instructive discussion of Charles Larmore, Patterns of Moral Complexity, 53-59. He speaks of the “neutral justification of political neutrality as based on a universal norm of rational dialogue” (p. 53), and he draws on (while modifying) the important ideas of Jurgen Habermas. See the latter’s Legitimation Crisis, translated by Thomas McCarthy (Boston: Beacon Press, 1976), pt. 111; and “Discursethik-Notizen zu einem Begrundungsprogramm,” in Moralbewusstsein lrnd kommunikatives Handeln (Frankfurt: Suhrkamp, 1983), 53-125. 5. This is the meaning of neutrality in Ronald Dworkin’s essay“Libera1ism.” See A Matter of Principle (Cambridge, Mass.: Harvard University Press, 1985), 191ff. 6. This statement of the three versions of neutrality draws on Joseph Raz’s formulations in his The Morality of Freedom (Oxford Clarendon Press, 1986), 114ff. 7. This distinction between neutrality of procedure and neutrality of outcome is adapted from Larmore, Patterns of Moral Complexity, 4247. 8. Some may feel that abandoning neutrality of effect or influence as impractical may lead to an excessively secular public life and background culture. How far this is a worry should be considered in light of what is said in the next section, especially in 3 4 about religious sects that oppose the culture of the modern world and about trying to answer the question of requirements on children’s education within the limits of the political conception. I am indebted to Dennis Thompson for the need to call attention to this connection. 9. Keep in mind here that the political virtues are identified and justified by the need for certain qualities of character in the citizens of a just and stable constitutional regime. This is not to deny that these same characteristics, or similar ones, might not also be nonpolitical virtues insofar as they are valued for other reasons within various particular comprehensive views. 10. This was the second meaning of neutrality of aim noted in the previous section; it is satisfied by a political conception. 11. The next several paragraphs are adapted from my reply in “Fairness to Goodness,” Philosophical Review 82 (April 1975): 548-51, to an objection raised by Thomas Nagel in his review, same journal, 82 (April 1973):226-29. In an instructive discussion that I shan’t attempt to summarize here, Nagel argues that the set up of the original position in Theory, although it is ostensibly neutral between different conceptions of the good, is not actually so. He thinks this is because the suppression of knowledge (by the veil of ignorance) required to bring about unanimity is not equally fair to all parties. The reason is that primary goods, on which the parties base their selection of principles of justice, are not equally valuable in pursuit of all conceptions of the good. Moreover, he says the well-ordered society of justice as fairness has a strong individualistic bias, and one that is arbitrary because objectivity between conceptions of the good is not established. The reply in the text above supplements that in “Fairness to Goodness” in two ways. It makes clear, first, that the conception of the person used
76
John Rawls
in arrriving at a workable list of primary goods is a political conception; and second, that justice as fairness itself is a political conception of justice. Once we understand justice as fairness and the conceptions that belong to it in this way, we can make a more forceful reply to Nagel’s objection, provided of course it is accepted that neutrality of influence is impracticable. 12. For a clear presentation of Berlin’s view, see his essay “The Pursuit of the Ideal,” in The Crooked Timber ofHurnanity (New York Knopf, 1991), esp. 11-19. Thus, he says on p. 13: “Some among the Great Goods cannot live together. That is a conceptual truth. We are doomed to choose, and every choice may entail an irreparable loss.” See also “Two Concepts of Liberty” (1958), reprinted in Four Essays on Liberty (New York Oxford University Press, 1969), 167ff. For Berlin the realm of values is objective, but values clash and the full range of values is too extensive to fit into any one social world; not only are they incompatible with one another, imposing conflicting requirements on institutions; but there exists no family of workable institutions that can allow sufficient space for them all. That there is no social world without loss is rooted in the nature of values and the world, and much human tragedy reflects that. A just liberal society may have far more space than other social worlds but it can never be without loss. The basic error is to think that because values are objective and hence truly values, they must be compatible. In the realm of values, as opposed to the world of fact, not all truths can fit into one social world. 13. In the passage from “Fairness to Goodness,” which Galston criticizes at p. 627, I should have mentioned and endorsed Berlin’s view as indicated in the text. Indeed we may often want to say that the passing of certain forms of life is to be lamented. What is said in that passage is not, I think, inconsistent with political liberalism, but it is seriously lacking by not also emphasizing Berlin’s view. It should have gone on explicitly to reject the idea, which Galston rightly rejects, that only unworthy forms of life lose out in a just constitutional regime. That optimistic view is mistaken. It may still be objected by those who affirm the conceptions that cannot flourish that political liberalism does not allow sufficient space for them. But there is no criterion for what counts as sufficient space except that of a reasonable and defensible political conception of justice itself. The idea of sufficient space is metaphorical and has no meaning beyond that shown in the range of comprehensive doctrines which the principles of such a conception permit and which citizens can affirm as worthy of their full allegiance. The objection might still be that the political conception fails to identify the right space, but there is simply the question of which is the most reasonable political conception. 14. And that of Raz in his The Morality ofFreedom, esp. chaps. 14 and 15, to mention a contemporary example.
Equality and Partiality Thomas Nagel
Introduction In recent years, a number of writers have defended state neutrality by appealing to a contractualist norm of reasonable acceptability. According to this norm, state coercion is justifiable only if it can be based on grounds that all reasonable citizens could not reasonably reject. In this excerpt, Nagel discusses the kind of toleration that follows from a serious commitment to this norm. He contends that in order to treat people as ends, not merely as means, one must not force them to serve values that they themselves cannot recognize. This remains true even when one is justified in believing that the values are sound. Liberal toleration thus calls for “a higher order impartiality” with respect to fundamental disagreements over the meaning and value of life. Nagel defends this position by claiming that “one may be justified in holding a belief about something of fundamental importance without having to conclude that those who do not share it are irrational or unreasonable.”
Thomas Nagel, Equality and Partiality (Oxford: Oxford University Press, 1991), 154-68. Copyright 0 1991 by Thomas Nagel. Reprinted by permission of Oxford University Press, Inc. -77-
Thomas Nagel
78
Toleration
T
HE MOST INTELLECTUALLY DIFFICULT PROBLEM regarding an
acceptable partition of motives arises not from conflicts of interest but from conflicts over what is truly valuable. Members of a society all motivated by an impartial regard for one another will be led into conflict by that very motive if they disagree about what the good life consists in, hence what they should want impartially for everyone. Anyone with a particular conviction about the good for human beings will naturally be inclined to get the power of the state behind it, not only for his own sake but out of concern for others. Those who disagree will want the state to promote other ends. Such disagreements can be much more bitter and intractable than mere conflicts of interest, and the question is whether there is any method of handling them at a higher level which all reasonable persons ought to accept, so that they cannot object to the particular result even if it goes against them. Simple impartiality will not generate a solution, since the conflicting positions already embody it, and the difference is over what it means. Some of these disagreements can be handled within the ordinary life of politics, where arguments over ends are part of the process of gaining majority support for particular policies. But there are other disagreements so deep and so acute that it is not possible to devise a method of fighting them out politically whose results could command the reasonable acceptance of the losers. Conspicuous among them are religious differences, but other convictions about the ultimate meaning of life or the sources of its value should be included as well. For this type of case another device is needed to ensure legitimacy-the exclusion of certain values from the admissible grounds for the application of coercive state power. We must agree to refrain from limiting people’s liberty by state action in the name of the values that are deeply inadmissible in a certain way from their point of view. This adds something to the purely instrumental theory of toleration. I would like to begin by discussing the principle in rather abstract terms, leaving the consideration of specific policies till later. What I shall call liberal toleration, using “liberal” in its American sense, depends on the acceptance of an impartiality of a higher order than that which leads us to recognize the equal value of everyone’s life. Impartiality among persons could take its content from a specific conception of the good, which others might not share. But the higher-order impartiality I am thinking of operates precisely on the conflicts between different first-order impartialities informed by conflicting conceptions of the good. Opposed ideas of the good, and therefore of what is impartially desirable for everyone-and not only op-
Equality and Partiality
79
posed personal interests-are counted among the conflicts with which a legitimate political system must deal, and with respect to which it must try to be fair among its citizens. The opposite view is that in political justification one must rely on one’s convictions about what human good consists in, even if it is nothing but the satisfaction of individual preferences-and that it is enough to see that the system treats people impartially with respect to that good, whatever it is. The intellectual problem is how to make sense of a supposed higher-order impartiality, either logically or morally. It seems to require us to subordinate our concern for people’s good to something else, but it is obscure both what that is and why it should carry such weight. Haven’t we gone as far as necessary (and perhaps even as far as possible) in taking up other people’s point of view when we have accepted the impartial component of our own moral position? The motive for higher-order impartiality is far more obscure than the motive for wanting everyone to have a good life. It is so obscure that critics of the liberal position on toleration often doubt that its professions of impartiality are made in good faith. Part of the problem is that liberals ask of everyone a certain restraint in calling for the use of state power to further specific, controversial moral or religious conceptionsbut the results of that restraint appear with suspicious frequency to favor precisely the controversial moral conceptions that liberals usually hold. For example, those who argue against the restriction of pornography or homosexuality or contraception on the ground that the state should not attempt to enforce contested personal standards of morality often don’t think there is anything wrong with pornography, homosexuality, or contraception. They would be against such restrictions even if they believed it wus the state’s business to enforce personal morality, or if they believed that the state could legitimately be asked to prohibit anything simply on the ground that it was wrong. More generally, defenders of strong toleration tend to place a high value on individual freedom, and limitations on state interference based on a higherorder impartiality among values tend to promote the individual freedom to which they are partial. This leads to the suspicion that the escalation to a higher level of impartiality is a sham, and that all the pleas for toleration and restraint really disguise a campaign to put the state behind a secular, individualistic, and libertine morality-against religion and in favor of sex, roughly. Yet liberalism purports to be a view that justifies religious toleration not only to religious skeptics but to the devout, and sexual toleration not only to libertines but to those who believe extramarital sex is sinful. It distinguishes between the values a person can appeal to in conducting his own life and those he can appeal to in justifying the exercise of political power. What I want to know is whether a position of this type is coherent and defensible.
80
Thomas Nagel
The question is important even though this is only one of the arguments for toleration. A historically significant and politically more effective argument is that those who have the upper hand now may not hold it forever, and that out of prudence they should refrain from imposing a sectarian view on others in exchange for the assurance that they will be treated with similar restraint if they find themselves in the minority. This is an argument for political toleration and impartiality as a second-best solution, acceptable because the best solution-political imposition of your own world view without any risk of future suppression-is not available. Such a defense of toleration as a modus vivendi can be offered to holders of radically divergent moral and religious positions, but it is an instrumental argument, and does not present higher-order impartiality in the political sphere as a value in itself. It could not therefore be offered as a reason for toleration to those who felt certain that their domination of the society was completely secure. Another argument for toleration would be simply to deny the truth of those religious and moral beliefs which seem to generate reasons against itacknowledging that we are faced with a battle between world views, which must be fought out at ground level. But apart from the politically suicidal aspect of an attempt to defend toleration by attacking religion, this position fails to account for the reach of the argument that many political liberals wish to offer. Liberal toleration is not compatible with absolutely any set of particular values and beliefs, but there is a version of it which aspires to be acceptable to those who disagree deeply over many other matters of the first importanceincluding the value of individual autonomy. Even if he cannot convince all of those people, it is important to a liberal of this stripe that he should at least be able to convince himselfthat they have reason to accept certain principles of political toleration and impartiality-and that such acceptance would not require them to abandon their religious or moral views, because his principles don’t rest on the denial of those views. This would satisfy the condition of legitimacy as an ideal of possible unanimity at some sufficiently high level with respect to the way disagreements are handled. So my aim is to achieve a certain peace of mind. But I have to say that although I shall offer an answer to the problem of interpretation and justification that has been posed, the unanimity being sought at a higher level is in this case even more doubtful than usual. To restate the apparent paradox: Liberalism asks that citizens accept a certain restraint in calling on the power of the state to enforce some of their most deeply held convictions against others who do not accept them, and holds that the legitimate exercise of political power must be justified on more restricted grounds-grounds which belong in some sense to a common or public domain.
Equality and Partiality
81
But it is not clear why this restricted form of justification should be the standard of political legitimacy at all. To put the argument against: Why should I care what others with whom I disagree think about the grounds on which state power is exercised?Why shouldn’t I discount their rejection if it is based on religious or moral or cultural values that I believe to be mistaken?Isn’t that being too impartial, giving too much authority to those whose values conflict with mine-betraying my own values, in fact?’ If I believe something, I believe it to be true, yet here I am asked to refrain from acting on that belief in deference to beliefs I think are false. It is unclear what possible moral motivation I could have for doing that. Impartiality among persons is one thing, but impartiality among conceptions of the good is quite another. True justice ought to consist of giving everyone the best possible chance of salvation, for example, or of a good life. In other words, we have to start from the values that we ourselves accept in deciding how state power may legitimately be used. And it might be added, aren’t we doing that anyway, if we adopt the liberal standard of impartiality? Not everyone believes that political legitimacy depends on this condition, and if we impose political institutions on others in our society because they do meet it (and block the imposition of institutions that do not), why aren’t we being just as partial to our own values as someone who imposes a state religion? It has to be explained why this is a form of impartiality at all. To answer these questions we have to identify the moral conception involved and see whether it has the authority to override those more particular moral conceptions that divide us-and if so, to what extent or in what respects. Rawls has said that if liberalism had to depend on a commitment to comprehensive moral ideals of autonomy and individuality, it would become just “another sectarian doctrine.”2 The question is whether its claim to be something else has any foundation. I believe that underlying the constraints of liberal toleration is the Kantian requirement of unanimity, which I have invoked throughout, usually referring to the first formulation of the categorical imperative, in terms of universalizability. Here I wish to emphasize the second formulation: that one should treat humanity never merely as a means, but always also as an end. On one reading of this principle, it implies that if you force someone to serve an end that he cannot be given adequate reason to share, you are treating him as a mere means-even if the end is his own good, as you see it but he d ~ e s n ’ tIn .~ view of the coercive character of the state, the requirement becomes a condition of political legitimacy. The problem is to interpret it, so that it rules out neither too much nor too little. The position I want to defend depends on a four-fold classification of grounds for coercion: (1) grounds which the victim would acknowledge as
82
Thomas Nagel
valid; ( 2 ) grounds which the victim does not acknowledge, but which are nevertheless admissible because he is grossly unreasonable or irrational not to acknowledge them; (3) grounds which the victim does not acknowledge, without being irrational, but which are admissible under a higher-order principle which he does acknowledge, or would be unreasonable not to; and (4) grounds which the victim does not acknowledge-either reasonably or even somewhat unreasonably-and which are such that he cannot be required to accept a higher-order principle admitting them into political justification even if most others disagree with him. It is type (4) that determines the extent of toleration essential to a legitimate system, and the problem is to explain why its supposed instances do not fall under ( 2 ) or (3) instead. To illustrate: Type (1) is exemplified by Hobbesian coercion, where each of us wants to be forced to do something as part of a practice whereby everyone else is forced to do the same, with results that benefit us all in a way that would not be possible unless we could be assured of widespread compliance. This is not really forcing people to do what they don’t want to do, but rather enabling them to do what they want to do by forcing them to do it. Type ( 2 ) is exemplified by the enforcement of criminal law against the willfully antisocial, and also by very basic forms of paternalism. In both cases lack of concern about the harms being prevented is unreasonable or irrational. Someone forcibly prevented from committing armed robbery or from drinking lye during a psychotic episode is not being coerced on grounds that he cannot be given sufficient reason to share: He just doesn’t recognize their sufficiency. Type (3) is exemplified by public policies based on judgments over which reasonable persons can disagree, but where it is also reasonable to agree to allow policy to be determined by a political process in which opposing points of view are represented, and given the chance to prevail. Many disagreements over what is good and bad fall into this category, including most of the values and the priorities among them that enter into debates about economic policy, criminal law, and other matters. Type (4), if it exists, is exemplified most clearly by the political enforcement of religious, sexual, or cultural orthodoxy. The liberal case for toleration depends on showing that such grounds for state coercion cannot be put under either type ( 2 ) or type (3), and that they therefore fail the Kantian condition of possible unanimity. This means that they are not just values on which persons who are not grossly unreasonable or irrational may disagree, but values of such a type that a person cannot reasonably be expected to agree to a system which authorizes the use of political power to enforce or promote values opposed to his own, just because the majority accepts them.
Equality and Partiality
83
This depends on the assumption that one may be justified in holding a belief about something of fundamental importance without having to conclude that those who do not share it are irrational or unreasonable, even though they have been presented with the same reasons or evidence that one has found compelling. It is not easy to say what distinguishes cases like this from others in which the recalcitrance of those who are not convinced can be dismissed as unreasonable. It seems to me clear that as things now are, those who do not accept the truth of a particular religion (or of atheism) ought not to be judged unreasonable by those who do, and that anyone who today is unconvinced by the germ theory of disease must be judged irrational. This is to reject the position that it is reasonable to believe something only on grounds which make it unreasonable or irrational not to believe it. That would make any reasonable belief an adequate ground for coercion, because those who did not accept it after being given those grounds could be coerced on grounds that it was unreasonable for them to reject. Likewise it would be unreasonable to hold any belief that others could reasonably reject and that was therefore not a possible basis for coercion. I just don’t think belief is like that: There is a substantial middle ground between what it is unreasonable to believe and what it is unreasonable not to believe (where belief and non-belief are taken as exhaustiveso non-belief is not disbelief). Belief is reasonable when grounded on inconclusive evidence plus judgment. In such a case one usually acknowledges the possibility of some further standard to which impersonal appeal can be made, even though it cannot settle existing disagreements at the moment. But even without such a standard, belief may not be unreasonable. In any case, it would be absurd to claim that individuals should decide what to do in their own lives only on grounds which they believe it would be unreasonable for anyone else to reject, including grounds having to do with fundamental values they believe to be objectivelycorrect. But in the political realm we have to find a more objective form of justification. If those whom we propose to subject to political coercion cannot be expected to accept the values we wish to further by it, we will be justified only if there is another description of the grounds of coercion that they can be required to accept. Sometimes, “You lost the election” will serve that purpose, but not always. It depends on whether the issue is one which it is reasonable to require everyone to put to decision by vote. This is really a problem of how to interpret the familiar role-reversal argument in ethics: “HOWwould you like it if someone did that to you?” That argument invites the further question, “How would I like it if someone did what to me?” Since there is more than one true description of every action, the selection of the morally operative one is crucial. If someone believes that by restricting freedom of worship he is saving innocent people from the risks of eternal damnation to which they are exposed by deviation from the true faith,
84
Thomas Nagel
then under that description he presumably would want others to do the same for him, if he were in spiritual danger. But under the description “restricting freedom of worship,” he wouldn’t want others to do it to him, since in light of the fact that his is the true faith, this would be to hinder his path to salvation. For purposes of political argument we have to exclude the description of what is done in the contested terms of a particular faith, and find instead a way of applying the role-reversal argument in terms of descriptions and values that must be accepted by all reasonable parties, as a basis for regulating or handling those disagreements that reason cannot eliminate. Legitimate government would be impossible if it were never legitimate to impose a policy on those who reasonably rejected the values on which it was based. It is not in general a valid role-reversal argument to ask, “How would you like it if someone did something to you that you reasonably didn’t want him to do?” The reply is that you might not like it, but might nevertheless be prepared to accept it, depending on the nature of their reasons and the institutions or procedures under which they were empowered to act on them in opposition to your preferences. The legitimacy of democratic government depends on its ensuring that we can all countenance, even if we don’t like, what it may impose on us against our wills. But this means that it is legitimate only if those impositions that we should not be asked to countenance are kept beyond its power. Why should I not accept the efforts of others, if they can muster the necessary political strength, to ensure my eternal salvation as they understand it by preventing the spread of heresy and atheism? Why is this so different from accepting what I may believe to be deeply misguided policies on public health, national defense, or education, if they are democratically a d ~ p t e d ? ~ The answer here cannot be just in terms of the priority of my interest in basic personal autonomy over other people’s interest in promoting what they regard as a desirable moral environment, since we are dealing here not with a conflict of interest but with a conflict over what my most fundamental interests are. Those who wish to limit my religious freedom are doing so, in the case under consideration, with my own best interests in mind.5 They believe that eternal salvation has priority in any person’s good even over basic personal autonomy, and if I shared their views I would have to agree with them. The question is why my conception of my good should block the use of their conflicting conception in the justification of political control over me in this case, but not in others. I think the problem is that there is no higher-order value of democratic control or pursuit of the good abstractly conceived which is capable of commanding the acceptance by reasonable persons of constraints on the pursuit of their most central aims of self-realization-except for the need to respect
Equality and Partiality
85
this same limit in others. Ethics does not license an unmediated universal altruism, precisely because that leads in ordinary circumstances of disagreement over the nature of the good to inevitable conflict rather than possible unanimity. Mere altruism, surprisingly enough, does not provide a common standpoint from which everyone can reach the same conclusions-and that is the essence of the contractarian or Kantian idea of legitimacy. Altruism by itself generates as many conflicting standpoints as there are conceptions of the good. And where no common standpoint is available at any level to authorize the collective determination by democratic procedures of policies about which individuals find themselves in radical disagreement because of incompatible values, it is best, if possible, to remove those subjects from the reach of political action. In some cases, such as national defense, a common standpoint can be found despite extreme disagreement, because everyone recognizes that some unified policy is absolutely necessary, and we all have to take the risk that the actual policy decided on will be abhorrent to us. But that is not true of religion and other basic choices regarding what life is about and how it is to be led. There the argument of necessity does not supply a common standpoint capable of containing the centrifugal force of diametrically opposed values, and legitimacy requires that individuals be left free, consistent with the equal freedom of others, to follow their own paths. We have to answer the question, why this position is not equivalent to the public adoption of an individualistic conception of the good, according to which each person’s good consists in his being able to satisfy his preferences or pursue his own freely chosen aims. That of course would be unacceptable to those who hold more unified conceptions: It would be just “another sectarian doctrine,” the dread bourgeois individualism, perhaps. The answer is that the consequences of making such a conception the basis of policy would be quite different from the position of liberal toleration. If a state really set out to promote the good for everyone so interpreted, it would have to try to discourage forms of life that thwarted individual self-expression and inculcated obedience to authority or to divine law, or subordination of personal aims to the collective goals of an organic community. This might argue for restrictions on religious education, for example, or on the formation of private communistic associations. The true liberal position, by contrast, is committed to refusing to use the power of the state to impose paternalistically on its citizens a good life individualistically conceived. Even if one does not like the results, the requirements of legitimacy dominate the desire to benefit others by your own lights, whatever they may be. The consequences of such a position are complex, for there are several ways in which state action may serve a conception of the good, and they will not all
86
Thomas Nagel
be equally unacceptable to those who do not share it. (1) A state might force people to live in accordance with the conception, or prohibit them from living in ways it condemns. (2) A state might support the realization of the preferred conception, by education or resource allocation, thus involving all citizens and taxpayers indirectly in its service. (3) A state might adopt policies for other reasons which have the effect of making it easier for one conception to be realized than another, thus leading to growth in adherence to the one as opposed to the other. Clearly the first way is the most illegitimate; examples are restrictions on the free exercise of religion, on the basic style of personal life, or on private sexual conduct. The second is less of an assault on individuals who do not share the dominant values, but still of questionable legitimacy; the clearest example would be public support of an established church. The third is in some degree unavoidable; liberal toleration, for example, though not motivated by the aim of promoting secularism and discouraging religious orthodoxy, may have these effects nevertheless. Hence it will not be neutral in efect among conceptions of the good, though it is based on impartiality among those conceptions, and avoids appealing to any of them to justify the use of coercive state power. Impartiality of this kind has to be distinguished from the kind of intrusive even-handedness which would require the use of state power to ensure that the society gives them all an equal chance to flourish.6 This therefore is another area in which the distinction between positive and negative responsibility has moral significance with regard to state action. The state is positively responsible for the difficulties of a particular religion if it either suppresses that religion or actively supports another. Citizens who adhere to the disfavored religion clearly have no reason to authorize such policies. But they do not have similar reasons to refuse acceptance of a system which simply fails to prevent the decline of their religious community. There is much more to be said about all this. For example there may be cases of state action not intentionally aimed against a particular community of conviction which nevertheless foreseeably damages it to such an extent and so directly that positive responsibility must be acknowledged, and the action’s legitimacy rendered doubtful. But I have attempted only a rough sketch of the liberal position here. Let me make a further comment about he positive-negative responsibility distinction. The denial of the moral import of that distinction in evaluating a society’s policies of taxation and distribution played a significant role in the discussion of socio-economic equality. Here, by contrast, it has been invoked as relevant to the difference between state support for a religion and other
Equality and Partiality
87
types of impact of state action on the success of a religion. The reason is that the role of the state as actor on behalf of its citizens assumes prominence when the action is based on commitment to values in direct contradiction to the deepest convictions of some citizens about the meaning of life. That, I think, is deeply offensive and unacceptable, and forfeits the state’s claim to represent them in a way in which the promotion of other values some of them do not share does not. The objections to it do not apply to the promotion of controversial aesthetic values, for example. There are also other cases where positive responsibility is particularly important, some of which are discussed in the previous chapter. How the state treats people in its police function should also be constrained by the fact that it is acting in the name of its citizens, and there are things it should not do because there are things they should not do. The constraints of personal morality, which prohibit certain forms of direct harm to others, even as a means to valuable ends, apply in this way to the state as well as to individuals, and the distinction between positive and negative responsibility for such harm applies here. The state also has a stronger negative responsibility for the failure to prevent such harms than individuals have. But this does not obliterate the positive-negative distinction. As was noted earlier, a state is not justified in employing brutal treatment of criminal suspects to fight crime, just because the overall number of brutal acts would be reduced thereby. To return to the subject of toleration: There remains the question how much disagreement and of what kind the principle of higher-order unanimity can contain. How effectively can the desire to find a common standpoint of justification curb the desire to pursue the transcendent good as one understands it? After all, if someone is willing to commit his own life to a particular conception, and convinced that the alternative is catastrophic, then it may be hard to resist imposing his opinion on others who, understandably but erroneously, fail to accept it. This tendency may be reinforced by the inclination to find them irrational, and thus subject to paternalistic coercion even under the liberal standard. But even without this, it may be difficult to subordinate a concern for their good as he sees it to a requirement of Kantian respect, if he is really convinced that Kantian respect will allow them to doom themselves. Such a conflict may bring the commitment to legitimacy to the breaking point, but in that respect toleration is not different from other questions of political theory. There are some conceptions of the good and of morality which are incompatible with the ideal of reasonable unanimity that is the heart of the Kantian position. The most one can hope is that it will be able to contain most of the disagreement that divides democratic societies, and that the forms of fanaticism which it cannot accommodate will gradually die out.
88
Thomas Nagel
Notes 1. Robert Frost defined a liberal as someone who couldn’t take his own side in an argument. 2. John Rawls, “Justice as Fairness: Political not Metaphysical,”Philosophy and Public Affairs 14 (1985): 246. 3. I. Kant, Foundations of the Metaphysics of Morals (1789, Prussian Academy Edition, vol. 4, translated by H. J. Patton in The Moral Law (London: Hutchinson, 1948), 429-30. See Onora O’Neill, “Between Consenting Adults,” Philosophy and Public Affairs 14 (1985): 261-63; and Christine M. Korsgaard, “The Right to Lie: Kant on Dealing with Evil,” Philosophy and Public Affairs 15 (1986): 330-34. 4. I offered one answer to this question in “Moral Conflict and Political Legitimacy,” Philosophy and Public Affairs 16 (1987) but while I still believe in the conclusion, I no longer think that “epistemological” argument works. I was finally persuaded by arguments that had been urged against me for some time by Lawrence Sager in discussionarguments also presented by Joseph Raz in “Facing Diversity: The Case of Epistemic Abstinence,” Philosophy and Public Affairs 19 (1990). 5. But in real life one must always be skeptical about this. See John Locke, A Letter Concerning ToZeration (1689), edited by J. H. Tully (Indianapolis: Hackett, 1983): Let them not call in the Magistrate’sAuthority to the aid of their Eloquence, or Learning; lest, perhaps, whilst they pretend only Love for the Truth, this their intemperate Zeal, breathing nothing but Fire and Sword, betray their Ambition, and shew that what they desire is Temporal Dominion. For it will be very difficult to persuade men of Sense, that he, who with dry Eyes, and satisfaction of mind, can deliver his Brother unto the Executioner, to be burnt alive, does sincerely and heartily concern himself to save that Brother from the Flames of Hell in the World to come. (p. 35) 6. This point is made by John Rawls in “The Priority of Right and Ideas of the Good,” Philosophy and PubZic Affairs 17 (1988).
I SELECTIONS FROM PUBLISHED WORKS
B Perfectionism
This Page Intentionally Left Blank
7 EqualitF Liberty and Perfectionism Vinit Haksar
Introduction
Egalitarianism and perfectionism are often thought to stand in tension with each other. If some forms of life are superior to others, and if the state favors them over inferior forms, then will it treat all its citizens with equal respect? In this selection, Haksar addresses this question. He draws a distinction between the view that some human beings have greater intrinsic worth than others and the view that some forms of life are intrinsically superior to others. Haksar notes that one can affirm the latter view without also affirming the former, and he argues that only the former view conflicts with the requirement that the state treat all its citizens with equal respect. But while Haksar contends that the state should discourage inferior forms of life, he maintains that it does not follow from this that it should not tolerate them. He proposes a system of “toleration without equal liberties.” Such a system is based on the idea that the state has reason to tolerate inferior forms of life, even while it denies them equal status with superior forms of life.
Vinit Haksar, Equality, Liberty and Perfectionism (Oxford Oxford University Press, 1979), 284-97. Copyright 0 1979 by Vinit Haksar. Reprinted from Equality, Liberty and Perfectionism by Vinit Haksar (1979) by permission of Oxford University Press.
- 91 -
Vinit Haksar
92
A
LTHOUGH THE VIEW THAT SOME human
beings have greater intrinsic worth than others is incompatible with the doctrine of equal respect and consideration, the view that some forms of life (such as the pursuit of truth and beauty) are intrinsically superior to other forms of life (such as a life devoted to the eating of one’s excrement), and that the government should take such differences of worth into account, is compatible with the doctrine of equal respect and consideration. Those who believe in the first view normally believe in the second view, but the converse is not true. One can quite consistently believe that some forms of human life are inherently superior to other forms of human life, without believing that the person who practices the latter form of life has inferior worth or deserves less (intrinsic) consideration than the former. But does not the liberal doctrine of equality commit the state to seeing things from the point of view of the individuals concerned, and how can the state do this, if it treats some forms of life as inherently more worthy than others? Will not such perfectionism commit the state when dealing with an individual to using its own conception of the good instead of the agent’s conception of the good, thus undermining the whole basis of liberalism? I think such objections are based on a confusion. The view that some forms of life are inherently superior to others, and that political principles should take this into account, does not commit one to a policy of paternalism with regard to those who practice the inferior forms of life. We can give inferior status to the way of life of the person who goes in for bestiality or smoking or for eating his excrement, not because we think that by doing so we shall force him to be better but, for instance, in order to provide a decent and morally healthy environment for the coming generation or even for the adult members of society who want to protect themselves from the temptations of the inferior forms of life (see chapter 12). Of course if and when the state’s rationale for regarding some forms of life as inferior is that it regards as inferior those who practice such forms of life, then such a state would be violating the egalitarian doctrine. Now in a sense the person who is practicing the inferior form of life will get a worse deal under the scheme that I am suggesting than the person who practices better forms of life. But this does not go against the fundamental equality of respect owed to all human beings, any more than does the view that a kidney machine should be given to a young man rather than a senile old man (see chapter 9). Later on in this chapter it is argued that the doctine of toleration and unequal liberties is in harmony with the egalitarian approach. Of course there may be considerable disagreement over perfectionist judgments, about whether for instance human beings flourish more under a system of individual families or communal families, whether enforced sex to
Equality, Liberty and Perfectionism
93
benefit the worst off is degrading. But this does not undermine the point that egalitarianism needs to be supplemented by perfectionist considerations, which include considerations about human well-being, human dignity, etc. The existence of disagreement among human beings over perfectionist values does, however, seem to create a problem about whose judgment regarding such values should carry the day. But such problems are not peculiar to perfectionist theories. Even if one takes a purely want-regarding line, there is the problem about who has the authority to impose the recommended policy, and about whose views should carry the day on what the want-regarding approach implies in practice. Even if one takes a non-perfectionist approach like that of B. F. Skinner, who wants to jettison all ideas of autonomy and dignity and concentrate on scientific means of making human beings conform to desirable goals, there is still the problem about whose views about what ought to be done should carry the day. Skinner is sometimes accused of being dictatorial, of trying to impose his own solution of behavior conditioning on an unwilling population. He can reply to such accusations by making a distinction between recommending a policy that should be adopted by legislators and dictatorially imposing a policy. He could admit that he has no right, no authority, to impose his views by dictatorial means, but this does not prevent his views from being correct and from being deserving of implementation by the state through democratic channels. Similarly, the person who believes in perfectionism as a political principle is not committed to being a dictator. The view that bestiality is degrading and should be given an inferior status compared to conventional sexual practices, does not commit one to abolishing liberal democracy and becoming a dictator. One could, as a good democrat, try to convince the democratically elected legislators of the correctness of one’s views. Elsewhere I have argued that the failure to appreciate such points has led to the spurious paradox of democracy.’ Some philosophers are so obsessed with avoiding disagreement that they urge us to adopt some kind of negative utilitarianism as our main political principle. On this view the state should not try to promote positive goals such as happiness about which people disagree, for even among people who pursue happiness, individuals pursue different things under the name of happiness, and doctrines such as that of the greatest happiness of the greatest number give an illusory and spurious unity to people’s diverse goals; instead the state should concentrate on reducing evils that are generally regarded as evils. Thus negative utilitarians stress that there is general agreement that suffering is an evil and so the state should be encouraged to reduce suffering among its citizens. But even negative utilitarianism gives only the illusion of getting general agreement. Even in the case of reduction of suffering there is enormous
94
Vinit Haksar
disagreement about what compulsory methods of reducing suffering are legitimate (is it for instance legitimate to kill some innocent people if there is a net reduction of suffering?) and how much weight to attach to the reduction of suffering compared to other legitimate goals. Moreover, as R. N. Smart has pointed out, negative utilitarianism would imply that we should approve of a tyrannical and benevolent person who blows up the world as a means to preventing infinite future misery. Bernard Gert believes that his own negative approach can avoid such pitfalls. He suggests that there are other evils besides suffering; for instance, he stresses, death is generally regarded as an evil. But I do not see how this helps. By blowing up the world we could abolish death! Admittedly, if we take a contractarian approach, then we could argue that the contractarian parties would not like to be killed themselves in order to abolish deaths in the future. But then we have seen in chapter 2 the limitation of the contractarian approach; for instance, we saw that it cannot deal with the problem of people in weak bargaining positions, such as future generations, children, and weak foreigners. It might be thought that both perfectionism and non-perfectionist utilitarianism of the kind that Skinner recommends get us involved in controversial judgments about what values the state should promote and about the most effective means of promoting them. For people disagree a lot about the good. And it might be suggested that the way to keep the state neutral and above controversies is to adopt something like Nozick‘s minimal state. But it seems to me that there can be much disagreement about what rights individuals have. For instance, there can be much disagreement about whether or not your property was legitimately acquired and consequently there will be disagreement about whether you have rights over your property. Even with regard to your bodily parts, there can be dispute about whether you have an absolute right. Thus it may be objected that you have fattened as a result of exploiting poor people in the past, and so the healthy blood, kidneys, and so forth that you have are not wholly owned by you. And, as we saw earlier, there can be dispute about whether property rights are absolute or defeasible and whether human beings have a right to food, medicine, and so forth that can sometimes trump other people’s property rights. Again, there can be dispute about whether or not the minimal state should respect animal ‘rights’, the ‘rights’of fetuses, of infants, and so forth. So how can we keep clear of controversies and disagreement? So the argument that perfectionism should be ruled out as a political principle because it involves the use of controversial, non-neutral principles is too strong. Often we do make judgments to the effect that some forms of life (such as bestiality) are inherently less desirable than other forms of life. Even Rawls admits that comparisons of intrinsic value can be made and that the ends of
Equality, Liberty and Perfectionism
95
some people have more intrinsic value than the ends of others2 But he complains that people disagree so much about the application of the perfectionist ~ t a n d a r d .But ~ how does he get from this to the conclusion that perfectionism should not be allowed as a political principle4 or to the conclusion that it should never be allowed to override any of the contractarian considerations? There can also be much disagreement about what is required by the contractarian model.5Whether they ought to or not, many people will disagree with Rawls, as I do, about how to construct the model, about what principles follow from it, about their priorities and about their application. Rawls himself admits that the infractions of his difference principle are difficult to ascertaim6Why then should the perfectionist principle be rejected as a political principle, while the difference principle is accepted? In any case, it is important to realize that the extent to which a principle commands general acceptance is only one of the determinants of how good and important the principle is as a moral and political principle. Toleration without Equal Liberties Rawls thinks that the contractarian parties would rule out perfectionism as a political principle, for since they have disparate goals and aims they cannot be expected to agree on which of these goals should be ranked higher and which lower. Their concern for their integrity would not allow any of them to consent to anything less than equal liberties. Rawls believes that it would be wrong for the parties to take a gamble and opt for unequal liberties in the hope that their own religions and moral views come out on top. To take such gambles may show an irresponsibility towards one’s religious and moral interests, for if we lose the gamble, such views will not fare too well. Moreover, how would one justify such gambles to one’s descendants? Suppose our descendants belong to the religion that has been given inferior status, will not our decision to gamble in the original position appear grossly irresponsible to them? Even if they can break the law, they may have to put up with the humiliation of being given inferior status, unless they have been suitably conditioned and reconciled to their inferior status. Such Rawlsian arguments at best show that it is irrational for the parties in the original position to gamble with religious and moral liberties; but suppose the parties opt for unequal liberties not because of a gamble but for better reasons. For instance, because they believe that some forms of life are better than others, that the worship of false gods needs to be discouraged because it impoverishes and alienates the individual. If that is their reason for opting for unequal liberties, there is nothing for them to be ashamed of, and
96
Vinit Haksar
they could attempt to justify their decision to their descendants. Compare: a person in the original position may opt for a system that permits life imprisonment for murderers, even though there is a chance that he or his descendants may turn out to be a murderer; or again he may opt for a system that bans the use of LSD even though it may turn out that when the veil of ignorance is removed, he or his descendant may long for LSD. It has been argued in this book that perfectionism is presupposed by egalitarianism, and also that perfectionist judgments are required if we are to derive a liberal society from egalitarian premises. If this argument is sound, it is absurd for an egalitarian liberal to ban the admission of perfectionist considerations into his theory of justice. And once one admits perfectionist considerations it is difficult to see what is wrong with giving some forms of life (such as life devoted to bestiality and the eating of one’s excrement) lower status than other forms of life. It is true that the doctrine of equal respect and consideration gives equal intrinsic weight to every human being’s interests. But a person’s interest has to be understood dynamically by seeing what is in his interest taking his life as a whole and not just by seeing what is in his interest at a particular stage in his life. And the doctrine of equal respect commits us to balancing his rights against the rights of other individuals, such as children who need a morally healthy environment and adults who want to protect themselves from their own weaknesses (see chapter 12); an acceptance of a right-based account is consistent with the view that rights have to be balanced against other rights, though it is not consistent with the view that rights can be sacrificed for ordinary utilitarian reason^.^ Admittedly, an individual, especially an autonomous individual, might find that at some stage in his life he wants to embrace a form of life that under a system of unequal liberties gets only inferior status. An autonomous individual may be a Christian this year, but next year he may want to become a Zen Buddhist and the year after he may want to become, say an atheist. So it might be suggested that an autonomous individual would need to keep his options open and would prefer to live in a system of equal liberties rather than in a system of unequal liberties where he may find himself practicing a form of life that is given only inferior status. Some such argument for equal liberties is found in the writings of Rawls and T. Scanlon. What are we to make of this argument? Now such considerations have to be balanced against other considerations. When we are looking at an individual’s interests we must, as I have stressed, go by what is in his interest taking his life as a whole, and not just be concentrating only on some stage of his life, such as his life after he has become an adult. Now if we take an individual who is just born and if we admit that forms of life are inherently superior to others, then it seems that it could be
Equality, Liberty and Perfectionism
97
in the interest of this person to be brought up in a society where inferior forms of life such as bestiality and smoking cigarettes are given inferior status and are discouraged but tolerated in private, rather than to be brought up in a society where there is a greater chance of his getting ‘hooked’ on such inferior forms of life. For the better forms of life are like a tender plant that need cultivation. What I would suggest is a system of toleration but not of equal liberties for different forms of life. This suggestion tries to get, as far as possible, the best of both worlds. Admittedly, an individual taken at random may get ‘hooked’ on an inferior form of life, such as bestiality, and so if we have his interests in mind, we must tolerate his way of life. Yet we must also help individuals to try to avoid the inferior forms of life and encourage them to go in for one of the non-inferior forms of life, and this may be better done in a system of unequal liberties where the inferior forms of life are given lower status than in a system of equal liberties. So under the compromise I am suggesting, the state, by giving lower status to inferior forms of life, should discourage people from embracing such forms of life, but also it should tolerate those who have embraced such forms of life. My suggestion of unequal liberties with toleration is quite compatible with respect for an individual’s autonomy. An individual in order to exercise his autonomy does not have to live in a society where all major options are equally encouraged or discouraged. An individual does not lose his autonomy because the legal system forbids murder and theft. And there is no reason why he should lose his autonomy if, say, bestiality is given lower status than conventional sexual practices. Moreover, the fact that some forms of life are to be given inferior status is consistent with having equal liberties between the noninferior forms of life; so there can be plenty of scope for people to exercise their autonomy in choosing their way of life. I have commended unequal liberties only in the sense which would imply that some forms of life would get lower status than others. This view does not imply that human beings who practice such forms of life should be regarded as inferior; it does not imply that such people should have any less of the worthwhile liberties such as the liberty to vote or freedom of discussion. It is important to stress that a system which gives some forms of life lower status than others is quite consistent with people being allowed full liberty to discuss and criticize and try to change the ranking between different forms of life. In extreme situations they may even be justified in using disobedience as a means of getting greater recognition for their form of life and altering the majority’s moral perceptions.8 So there is plenty of scope for exercising one’s autonomy under such a system. People are engaged in a constant debate about not only the correct application of political principles but even about
98
Vinit Haksar
their very foundations. The debate is not just at the theoretical level; discussion, debates, protests, disobedience, are all parts of the process by which people sometimes cajole and haggle with each other, and sometimes even convert others to their point of view, or themselves get converted to another point of view. One of the fears that some liberals such as Rawls have in admitting perfectionist considerations into political arguments is that they fear that this will lead to the banning of degrading forms of life, even in private, ‘if only for the sake of the individuals in question irrespective of their wi~hes’.~ But my view is that if perfectionism is combined with egalitarianism it will not imply that people who practice degrading forms of life in private should be punished. The doctrine of transitivity of ends in themselves ensures that people are not expelled from the egalitarian club merely because they happen to practice inferior forms of life. Earlier in chapter 9, it was pointed out that even murderers are not expelled from the egalitarian club. Now if individuals who practice inferior, degrading forms of life are not expelled from the egalitarian club, then their interests must be given equal intrinsic weight to the interests of other people. So an individual’s well-being must not be sacrificed lightly, even though he is practicing degrading forms of life. Of course, sometimes, as in the case of pedophiles, we may prevent them practicing their form of life even in private; for their form of life involves violating the rights of children-the example of pedophilia again shows the importance of perfectionist considerations, for we have to appeal to such considerations when we maintain that an adult person who makes love to a child violates the child’s right or his dignity. But when a person is indulging in his degrading form of life in private, without violating anyone else’s rights, then it seems we should tolerate him rather than punish him. It might be objected that a perfectionist approach would tend to encourage the state to be intolerant towards those who practice degrading forms of life, because on the perfectionist view a person’s good and his well-being is understood in terms of some objective Good rather than in terms of the form of life that the agent may have got ‘hooked’ on. So should the perfectionist state not try to forcibly liberate him, for his own good, from his inferior form of life? One way of answering this objection would be to contend that there is value in the free choice of a way of life. As Mill said, a person’s ‘own mode of laying out his existence is the best, not because it is the best in itself, but because it is his own mode’.1° The great value that Mill places on an adult autonomously choosing his own form of life could be quite consistent with some forms of life being intrinsically superior to others. It is quite consistent to say that, other things being equal, a form of life A is superior to a form of life B, but if other things are not the same, if for instance a person has autonomously
Equality, Liberty and Perfectionism
99
chosen form of life B, then this situation is preferable to the situation where the form of life A is imposed on him. There is the complication that Mill at times implied that people would not autonomously choose the lower form of life. Indeed, part of his criterion for saying that some forms of life are superior to others was that if people had a free or autonomous choice between the superior and the inferior form of life, they would choose the former. If this is so, then the choice of inferior forms of life must be a non-autonomous choice. So might the state not be justified in liberating such people (for their own good) by force? There are several reasons for saying that the answer to the above question is No. First, as we have seen, forms of life are like a tender plant that needs cultivation in the early stages. If the better forms of life have not been properly cultivated in the early stages of a person’s life, then it is likely that the person’s potential for such forms of life is destroyed. If Mill is right in this, it would follow that it is dangerous to try to forcibly alter an adult person’s form of life in the hope that he may after all still be able to pursue the higher forms of life. Why forcibly destroy his inferior form of life, if he is no longer capable of leading the higher form of life? True, had he been brought up in a proper environment, perhaps he would never have got addicted to the inferior form of life; but once an adult person has embraced an inferior form of life he is likely to be miserable if he is forcibly prevented, for instance by threats of punishment, from indulging in his form of life. As for brainwashing him, it is possible that, as techniques of brainwashing improve, the person brainwashed may be less miserable at the end of the brainwashing than when he was indulging in his own mode of existence. But brainwashing of adults for their own good without their consent raises several dangers and problems. Secondly, one can point out that even if the adult person who leads an inferior form of life did not autonomously choose his way of life, but has followed it, for instance, because it was the only real option in the social circles where he was brought up, it does not follow that his present judgment about his own good is worse than the judgment of the state about his good. The man who, through getting addicted to an inferior form of life, has lost his wonderful potential, does not necessarily lose his ability to judge what is now good for him. It is true that in some cases of addiction, such as addiction to hard drugs, a person’s powers of judgment about his good is seriously impaired, but in the case of many other inferior forms of life, such as bestiality, it is not obvious why this should be so. In general, as Mill pointed out, the state has less interest in an adult individual’s welfare than he does himself and is likely in general to make less competent judgments about which form of life is suited to him. And there is the danger that the state could abuse its powers if it was allowed to liberate people for their own good.
100
Vinit Haksar
Finally, even when more good than harm were to come from the state trying to forcibly liberate its citizens, there is the problem (at any rate for nonutilitarians) that the state may not have the right to interfere with the adult individual for his own good. There are also utilitarian arguments for toleration, such as those found in Mill, On Liberty Thus Mill points out how society could gain by tolerating and sometimes even encouraging unconventional experiments in living. Some, though not most, of the experiments in living may well get adopted by others and be an improvement on customary practices. Moreover, even the people who continue to hold to established practices will be better off if they can choose their form of life in preference to the alternative experiments in living, rather than if they just followed the customary practices mechanically without any new challenges from alternative life styles. It might be suggested that on a right-based account such utilitarian advantages are a desirable and important by-product of toleration, but they are not the main justification for toleration. The main reason for tolerating such forms of life is to be found by appealing to the right to respect and consideration of individuals who practice such forms of life. However, there is the point that the good utilitarian effects of toleration can help to show that such toleration does not harm the rights of the coming generation. And this can be important on a right-based view; for on such a view an individual’s right can be trumped by a competing stronger right of another individual. And so some people might argue against tolerating inferior forms of life even in private on the grounds that such toleration could lead to the violation of the right of the young to grow up in a morally healthy environment. A society that bans such forms of life even in private might, by deterring individuals from practicing such forms of life, reduce their incidence, and this could (so the argument runs) lead to a morally healthier environment for the young to grow up in. For one cannot in practice wholly isolate what children are tempted to do from what adults in their society do, and are allowed to do Now it seems to me that such arguments have some prima facie force but one could reject them by pointing out the good effects of toleration. If such good effects are real, then the overall effect of toleration may well be to produce a morally healthy environment in which children grow up. For instance, if there is intolerance they may grow up in a smug, dogmatic, and bigoted atmosphere, which is not good for them. Such considerations show that there is a case for tolerating inferior forms of life, even from the point of view of the coming generation, but they do not show that such forms of life should be given equal status to the ordinary ways of life. To give an equal status to all forms of life is to refuse to learn from experience that some forms of life are
Equality, Liberty and Perfectionism
101
prima facie more suited to human beings, including children. In chapter 11,
section 4 we saw that Mill is also committed to some such view. Against tolerating inferior forms of life even in private, it is sometimes argued that if a majority feels disgust at a practice then it has a right to ban its practice even in private. But it seems to me that on a right-based approach the majority is not entitled to tyrannize the minority in this way. The interests of people who practice degrading forms of life must on a right-based egalitarian approach carry equal weight to the interests of people who lead more conventional lives. If we use the test that an individual's right to his well-being can only be trumped by a competing right of at least comparable strength that another individual has, then it can be seen that we should not punish people for practicing their forms of life in private merely because of the feelings of disgust that many conventional people feel at the thought of such practices taking place in private. To be deprived of being able to indulge in one's form of life even in private poses too severe a psychological hardship on such people, whereas the disgust that many conventional men feel at the thought that degrading forms of life are taking place in private does not or need not upset their well-being and it must not be allowed to trump the right of the minority to choose their form of life in private, any more than the hurt feelings that rejected suitors have provide a ground for trumping a woman's right to choose her sexual partner. Of course, there may be unusual cases where a person may have a neurotic and excessive reaction to the thought of people being allowed to practice degrading forms of life in private. Now such abnormal people should try to get cured rather than impose their way of life on others. However, in order to show why such reactions are abnormal we may need to appeal to perfectionist considerations and argue that it is better for a person not to be bigoted and intolerant of what other people do in private. Another way of tackling the problem would be to appeal to the idea that external preferences of people must not be allowed to trump an individual's rights-though this would still leave us with the problem of what rights an individual has, which cannot be solved without appealing to perfectionist considerations. We can now appreciate the difference between the view that some forms of life are superior to other forms of life (and that political principles should take this into account) and the view that some human beings are intrinsically superior to others (and that political principles should take this into account). The latter view, but not the former, is incompatible with egalitarianism, and would lead to intolerance and sacrifice of the interests of the inferior for the sake of the superior. My view that some forms of life should get lower status than others is not quite as outrageous and eccentric as it may appear at first. It harmonizes with what happens in liberal democracies. For instance, even when there is equality
102
Vinit Haksar
between all religions, the equality does not automatically extend to all groups that claim to be religious. Thus, in the United States there are groups that go in for drugs such as LSD as a means of attaining spiritual salvation, yet such groups are not automatically given all the legal privileges and rights that are extended to all recognized religious groups. Again, there are forms of life, such as bestiality, nudism, eating one’s own excrement, and so forth, which get only an inferior status in our society compared to the conventional forms of life. It is quite plausible to assert that some of the unconventional forms of life that are given inferior status in our society should have equal status with the conventional forms of life. But in the case of any particular form of life that has an inferior status the case for upgrading it is not an automatic one but has to be made by appealing to perfectionist considerations. For instance, whether homosexuality ought to be given equal status to heterosexuality would depend (at least partly) upon whether homosexuality is as fulfilling and worthwhile as heterosexuality. One must not bypass such perfectionist considerations and simply argue that all forms of life (that are not obviously antisocial) should get equal status. Notes 1. See Haksar, “The Alleged Paradox of Democracy.” 2. A Theory of Justice, 328. 3. Ibid., 331. 4. Ibid., 329. 5. Cf. Ibid., 147. 6. Ibid., 372. 7. See Dworkin, Taking Rights Seriously, chapter VII. 8. See Haksar, ‘Rawls and Gandhi on Civil Disobedience’, Inquiry, xix (1976), 151-92. 9. Rawls, A Theory of Justice, 331. 10. On Liberty, chapter 111, para. 14.
The Morality of Freedom Joseph Raz
Introduction The publication of Raz’s The Morality of Freedom was a watershed event in the perfectionism-neutrality debate. In this excerpt from his book, Raz presents a sustained critique of the antiperfectionist idea that the promotion of the good is not a legitimate concern of governments. In particular, Raz takes aim at the principle of state neutrality as it is articulated in the writings of Rawls and Nozick. After rejecting this principle, Raz argues that perfectionists can affirm “moral pluralism.” This is the view that there exists a multitude of incompatible forms of the good and that individuals should be free to choose between them. Raz claims that many writers have assumed a tight connection between moral pluralism and state neutrality, but that no such connection exists. Moral pluralism does not rule out state perfectionism. Raz further argues that supporting valuable forms of the good is a collective endeavor, not something that can be left to the uncoordinated decisions of individuals. Antiperfectionism in practice, he concludes, would “undermine the chances of survival of many cherished aspects of our culture.” Joseph Raz, The Morality of Freedom (Oxford Oxford University Press, 1986), 110-33, 161-62. This selection originally appeared as “Liberalism,Autonomy and the Politics of Neutral Concern,” in Midwest Studies in PhiZosophy, vol. vii, edited by French, Uehling, and Wettstein (Minneapolis: University of Minnesota Press, 1982), 89-120. Copyright 0 1982 by University of Minnesota Press. Reprinted by permission of University of Minnesota Press. - 103 -
104
Joseph Raz
. . . [the citizen’s] allegiance (as other individuals do not)’, writes Robert Nozick, ‘therefore scrupulously must be neutral between its citizens.” Writers of an egalitarian liberal persuasion often join libertarian liberals in applauding the same sentiments, different though their interpretation of them usually is. The anti-perfectionist principle claims that implementation and promotion of ideals of the good life, though worthy in themselves, are not a legitimate matter for governmental action. The doctrine of political neutrality seeks to implement it through a policy of neutrality. Government action should be neutral regarding ideals of the good life.2 Such a doctrine is a doctrine of restraint. Doctrines of governmental restraint are those which deny the government’s right to pursue certain valuable goals, or require it to maintain undisturbed a certain state of affairs, even though it could, if it were to try, improve it. As we shall see, many but not all liberal theories of political freedom are based on the endorsement of one or another of several principles of restraint, of which the principle of neutrality is one. Principles of restraint restrict the pursuit of good or valuable goals, they exclude action for valid, sound reasons for action, or they enjoin the government to preserve a state of affairs which there are good reasons to change. There is no need for a special principle to require action for good reasons, or to maintain valuable states of affairs. The doctrine of political neutrality is a doctrine of restraint for it advocates neutrality between valid and invalid ideals of the good. It does not demand that the government shall avoid promoting unacceptable ideals. Rather, it commands the government to make sure that its actions do not help acceptable ideals more than unacceptable ones, to see to it that its actions will not hinder the cause of false ideals more than they do that of true ones. The fact that anti-perfectionism is based on restraint, on not doing as much good as one can, lends it a slightly paradoxical air. But there is no denying its great intuitive appeal. At the intuitive level anti-perfectionism responds to a widespread distrust of concentrated power and of bureaucracies. Any political pursuit of ideals of the good is likely to be botched and distorted. The best intentions and the wisest council are likely to misfire if entrusted to the care of the machinery of state action. Beyond that there is the deep-felt conviction that it is not within the rights of any person to use the machinery of state in order to force his conception of the good life on other adult persons. Later in the book we will consider alternative responses based on these intuitions. Here the mention of but one example will suffice. A right of conscientious objection to military service, and perhaps to some other legal duties, combines a political pursuit of goals which may be justified by their contribution to a valid conception of the good with a refusal to ram that conception
A
STATE OR GOVERNMENT THAT CLAIMS.
The Morality of Freedom
105
down the throats of people who are deeply opposed to it. In other words, one alternative to anti-perfectionism is restrictions on the choice of means through which perfectionist ideals are pursued. In a sense anti-perfectionism is merely a more radical restriction of the employment of means through which one may pursue conceptions of the good. It denies the appropriateness of using any political means to pursue such ends. I will first examine the meaning of political neutrality (Section One) and its possibility (Section Two). Consideration of some of the justifications advanced for it will lead to casting doubts on its cogency. The argument will not be concluded until the examination of the principle of the exclusion of ideals from politics in the next chapter. Forms of Neutrality
Nozick, in the quotation given above, writes of neutrality between individuals. I referred to neutrality between ideals or conceptions of the good. The two ideas are closely related. Discrimination between individuals consists in making it easier for some than for others to realize their ideals of the good. But it is important to distinguish between two principles of political neutrality: Principles of Scope A Neutrality concerning each person’s chances of implementing the ideal of the good he happens to have. B: Neutrality as in A, but also regarding the likelihood that a person will adopt one conception of the good rather than another.
B is the more radical principle, and in the absence of any special reason to prefer A, and given that writers supporting neutrality say little that bears on the issue, I will assume that the doctrine of neutrality advocates neutrality as in B. Another ambiguity concerns the level of neutrality. Is politics generally to be neutral between conceptions of the good, or does neutrality apply to the constitution only? The first, stricter, doctrine of neutrality allows that individuals may act to implement their ideals in their lives, and in the life of their community, but only to the extent that they can do so by non-political means. They may found voluntary associations, buy neighborhoods and impose restrictive covenants on the use of properties in them, secure conditions of employment, endow charitable institutions, and promote their conception of the ideal community in many ways. Such activities, when successful, will make the pursuit of some ideals of the good easier, and that of others more difficult.
106
Joseph Raz
But no planning regulation may be passed, no law affecting the education of the young or the use of the highway, and so on, which will make it easier to realize one conception of the good or more difficult to pursue another. The second, less strict, doctrine allows the use of politics in the pursuit of ideals. It insists, however, that that will be done in a constitutional framework which allows everyone an equal chance to endorse any conception of the good and to realize it. The constitution is neutral. But the law is not necessarily so, and many people born today will find their chances to pursue different conceptions of the good affected by law as a result of the activities of our and previous generations. Again it seems best to assume the stricter doctrine as background for our discussion. It is important to realize what is involved in talking of principles of neutrality. ‘To be neutral . . . is to do one’s best to help or to hinder the various parties concerned in an equal degree13 This is the primary sense of neutrality, which I will call principled neutrality. In this sense one is neutral only if one can affect the fortunes of the parties and if one helps or hinders them to an equal degree and one does so because one believes that there are reasons for so acting which essentially depend on the fact that the action has an equal effect on the fortunes of the parties. One secondary sense of neutrality regards persons as neutral if they can affect the fortunes of the parties and if they affect the fortunes of all the parties equally regardless of their reasons for so doing. When ‘neutral’ is used in this sense I refer to it as by-product neutrality, for here neutrality may well be an accidental by-product of the agent’s action and not its intended outcome. Principles of neutrality state that there are reasons to be neutral. They are satisfied by any behavior that affects the fortunes of the parties in equal degree (by-product neutrality). They are followed by people acting neutrally in a principled way, i.e., because they believe that there is a reason not to help or hinder one side more than the other. Our interest is in political theories that require neutrality, i.e., that are followed by acting neutrally in a principled way. Some theories may be such that behavior that follows them is also neutral as a by-product. But those are of no special interest from our point of view. Neutrality is sometimes conceived of as being necessarily at least prima facie desirable. If so, the principles of neutral political concern are at least prima facie valid. But the definition of neutrality adopted above is not committed to such a view, which is rooted in the confused notion that to act neutrally is to act fairly. Montefiore gives a familiar example: Two children may each appeal to their father to intervene. . . in some dispute between them. Their father may know that if he simply ‘refuses to intervene’ the
The Morality of Freedom
107
older one, stronger and more resourceful, is bound to come out on top . . . In other words, the decision to remain neutral, according to the terms of our present definition, would amount to a decision to allow the naturally strong child to prevail. But this may look like a very odd form of neutrality to the weaker child.4 Does this show that the father would not be neutral if he remained aloof from the quarrel? Even if (and I do not accept this at all) neutrality could only be justified as a means to a fair contest, it should not be identified with action securing a fair contest. An attorney representing a client before a court helps to make the trial fair, but the attorney is not neutral. In his situation to act neutrally is to act unfairly. All that Montefiore’s example shows is that there are circumstances in which it is unfair to act neutrally, where there are not even prima facie reasons to be neutral. The example is of a case where the father should not, not where he cannot, remain neutral. The question of the justification of neutral political concern invites moral and political argument and cannot be settled by the inherent appeal of neutrality as such. Supporters of the doctrine of neutral political concern differ on many issues. But all of them endorse one or the other of the principles of political neutrality mentioned above and seek to implement it by some variant of the following principles of restraint which limit the political relevance of ideals of the good. Interpretations of Political Neutrality 1. No political action may be undertaken or justified on the ground that it promotes an ideal of the good nor on the ground that it enables individuals to pursue an ideal of the good. 2. No political action may be undertaken if it makes a difference to the likelihood that a person will endorse one conception of the good or another, or to his chances of realizing his conception of the good, unless other actions are undertaken which cancel out such effects. 3. One of the main goals of governmental authority, which is lexically prior to any other, is to ensure for all persons an equal ability to pursue in their lives and promote in their societies an ideal of the good of their choosing.
The first principle is offered by R. Nozick as an interpretation of neutrality. To the argument that his State is not neutral, for its political organization, its property and contract laws, favor the implementation of some conception of the good, Nozick replies: Not every enforcement of a prohibition which differentially benefits people makes the state non-neutral. . . . Would a prohibition against rape be non-neutral? It
108
Joseph Raz
would, by hypothesis, differentiallybenefit people; but for potential rapists to complain that the prohibition was non-neutral between the sexes. . .would be absurd. There is an independent reason for prohibiting rape. . ..That a prohibition thus independently justifiable works out to affect different people differently is no reason to condemn it as non-neutral, provided it was instituted or continues for (something like) the reasons which justify it . . . similarly with the prohibitions and enforcements of the minimal state.5
Nozick is certainly raising a point which supporters of the other versions of neutrality have to contend with. According to their interpretation of neutrality, he is implying, prohibiting rape is not neutral. But they have a ready answer to this challenge. Supporters of the third principle may claim that the reason rape is prohibited is that its perpetration deprives the raped of their chances to live according to their conception of the good. The limitation imposed on would-be rapists does no more than establish equal chances to pursue one’s conception of the good. It limits people’s ability to pursue some conceptions of the good, but only in order to equalize the opportunity to do so overall. Supporters of the second of the principles above will have a slightly different answer. Conceding that the prohibition of rape is itself non-neutral, they will then require the government to take action to compensate would-be rapists by improving their chances to realize other aspects of their conception of the good. Notice that the thought that in principle a would-be rapist should be compensated for giving up the right to rape is implicit in the answer of the supporters of the third principle as well. They claim that the prohibition in fact merely equalizes people’s ability to pursue their own conception of the good. If it turns out that this is not the case, if it is discovered that the prohibition denies the would-be rapists more of a chance to pursue the good life than it gives their possible victims, then one may have to adjust other features of the political framework to make sure that this does not result in inequality of ability to pursue one’s conception of the good. Many readers may join Nozick in regarding this as a reductio ad absurdum of the second and third principles stated above. We will return to their examination below. Whatever the force of Nozick‘s implied challenge to alternative interpretations of neutrality, the question remains whether his reply succeeds in its primary aim of defending his own. Nozick‘s case rests on the view that so long as one is not acting for the reason that one’s action will favor one of the parties or hinder the other, but for a valid independent reason, then one’s neutrality is intact. If so then on the assumption, surely acceptable to Nozick as to most people, that the prospect of a profit is a valid reason for most commercial activities it follows that selling arms to one of the combatants for profit does not jeopardize one’s neutrality.
The Morality of Freedom
109
One is not free to deny that when one’s action actually helps one of the parties then the profit is not a valid reason, for that is to abandon the case and adopt the view of neutrality espoused above according to which neutrality is determined by the consequences of one’s action. Nozick‘s State is not neutral, and his principle (principle 1 above) is not a principle of neutrality, but it shares with the doctrine of neutrality an anti-perfectionist bias and will therefore be examined in the next chapter. It may be thought that the same verdict applies to the third principle as well. Ensuring to all an equal ability to realize their conception of the good is more likely to require acting in a non-neutral way, acting to improve the ability of some at the expense of others. The principle is an anti-perfectionist principle but can it be regarded as a principle of neutrality? I think that this objection is mistaken and that the third principle is a principle of neutrality. I will call it the principle of comprehensive (political) neutrality to distinguish it from the second principle which will be called the principle of narrow (political) neutrality. We can best examine their character as two competing principles of neutrality in the context of objections to political neutrality on the ground that it is impossible or even incoherent. The Impossibility of Strict Political Neutrality The most serious attempt to specify and defend a doctrine rather like our principle of comprehensive neutrality is that of J. Rawls. While our discussion in this chapter is of the doctrine of neutrality as such, Rawls’ treatment of it will serve to illustrate the problems involved. Rawls believes that principles chosen rationally but in ignorance of one’s actual situation in society, one’s natural endowments, and one’s conception of the good, are bound to be neutral principles. He believes that under these conditions of choice, which he describes as the original position, there is only one set of principles which can be rationally chosen to govern a society enjoying favorable social and economic conditions. They are the principle of equal liberty, assuring everyone equal measure of an enumerated list of basic liberties (freedom of expression, religion, etc.) and the difference principle according to which everyone should enjoy equal allocation of the other primary goods (i.e. those desirable whatever one’s conception of the good may be, e.g. wealth, income, opportunities, status) except in so far as deviation from strict equality would improve the prospects of the worst-off group in society. Rawls’ theory deviates from comprehensive neutrality in requiring equal ability to pursue ideals of the good only in so far as the ability depends on the principle of equal liberty. For the rest, the difference principle allows deviations from an
110
Joseph Raz
equal distribution of primary goods where the worst-off group will benefit from them. Finally, Rawls is only to a limited extent concerned with correcting inequalities in the ability to promote the good which are due to one’s natural endowments (the difference principle is about social, not natural, primary goods). T. Nagel has criticized Rawls’ theory of justice on the ground that it purports to be neutral between different conceptions of the good, but is not. If Rawls’ theory is based on a doctrine of neutrality it is a doctrine of comprehensive neutrality. Hence the relevance of the objection to our purpose. Nagel explains his position as follows: It is a fundamental feature of Rawls’ conception of the fairness of the original position that it should not permit the choice of principles of justice to depend on a particular conception of the good over which the parties may differ. The construction does not, I think, accomplish this, and there are reasons to believe that it cannot be successfully carried out. Any hypothetical choice situation which requires agreement among the parties will have to impose strong restrictions on the grounds of choice, and these restrictions can be justified only in terms of a conception of the good. It is one of those cases in which there is no neutrality to be had, because neutrality needs as much justification as any other position. (pp. 8-9)6
The specific point that Nagel is making is that there is no way of justifying the conditions of choice in the original position except from the point of view of a certain conception of the good. But later on he makes it clear that this is so, at least in part, because the supposedly inevitable outcome of that choice is not really neutral: The original position seems to presuppose not just a neutral theory of the good, but a liberal, individualistic conception according to which the best that can be wished for someone is the unimpeded pursuit of his own path, provided it does not interfere with the rights of others. The view is persuasively developed in the later portions of the book but without a sense of its controversial character. Among different life plans of this general type the construction is neutral. But given that many conceptions of the good do not fit into the individualistic pattern, how can this be described as a fair choice situation for principles of justice? (P. 10)
One should be careful not to misinterpret this point. Rawls often writes of individuals’ conceptions of the good as if they are their views of the good life for themselves. But he is also aware that they may be conceptions of the good life for people generally and for society as a whole. Individuals may use their primary goods to promote non-individualistic conceptions of the
The Morality of Freedom
111
good. To use Nagel’s own words, individuals may use their primary goods to implement views that hold a good life to be readily achievable only in certain well-defined types of social structure, or only in a society that works concertedly for the realization of certain higher human capacities and the suppression of baser ones, or only given certain types of economic relations among men. (p. 9)
The individualistic bias that Rawls is accused of by Nagel is not that he rules out such conceptions but that he is not neutral regarding them because he makes their successful pursuit more difficult than that on individualistic conceptions of the good. Nagel’s reason for alleging the existence of the bias is that ‘the primary goods are not equally valuable in pursuit of all conceptions of the good’ (p. 9). They serve individualistic conceptions well enough but ‘they are less useful in implementing’ non-individualistic conceptions. This point is valid. Rawls is surprisingly brief and inexplicit on the issue. He seems to take assessment of wealth to be ~nproblematic.~ If some market mechanism, actual or hypothetical, is assumed, the value of primary goods is the function of supply and demand where the demand is partly determined by the usefulness of the goods in the implementation of conceptions of the good which are actually pursued in that society and by the number of those pursuing different conceptions. Relative to any such evaluation of primary goods some conceptions of the good will be harder to implement, i.e., will require primary goods of greater value to realize, than others. These need not be nonindividualistic conceptions. All conceptions involving the cultivation and satisfaction of the so-called expensive tastes are harder to satisfy, and the Rawlsian theory can be said to discriminate against them. Nonindividualistic conceptions are likely to be among the expensive tastes since their realization depends on the cooperation of others, and they will take some convincing to come round to the agent’s point of view. Alternatively, it will take a lot of resources to buy their consent to his point of view. This consideration points to the fact that the very restrictions imposed on societies by the Rawlsian principles of justice make the implementation of some conceptions of the good more difficult and their pursuit by individuals less attractive than that of others. Furthermore, the implementation of some conceptions of the good is incompatible with the principles of justice and is ruled out altogether. Neutrality, however, can be a matter of degree. One can deviate from complete neutrality to a greater or lesser extent. Can Rawls’ principles be defended on the basis that they approximate complete neutrality better than any alternative? The argument remains to be made. Whatever our judgment of Rawls’ principles on that account, at least if political neutrality is a coherent and
112
Joseph Raz
desirable ideal then the impossibility of complete adherence to it need not undermine its force as a political doctrine. So let us turn to two arguments challenging the validity of the idea in a way which makes even approximation to complete neutrality a chimerical notion. Neutrality is concerned only with the degree to which the parties are helped or hindered. It is silent concerning acts which neither help nor hinder. This may lead some to suppose that any attribution or commendation of neutrality assumes that whereas one is morally responsible (i.e., accountable) for what one does, one is not morally responsible for what one does not do, or some similar distinction. Those who reject this view may conclude that neutrality is impossible, or even incoherent. The conclusion is, however, unwarranted. No such assumption underlies the attributions of neutrality. Commendations of neutrality, while consistent with such a view, do not depend on it. They may be defended simply because in certain circumstances not helping is to be preferred to helping. It is true, though, that valuing neutrality presupposes a distinction between not helping and hindering. Consider a country that has no commercial or other relations with either of two warring parties. This was true of Uruguay in relation to the war between Somalia and Ethiopia. It may nevertheless be true that such a country may have been able to establish links with either party. Would we say that Uruguay was not neutral unless the help that it could have and did not give Ethiopia was equal to the help that it could have and did not give Somalia? This will not be the case if, for example, Uruguay could have supplied the parties with a commodity that, though useful to both, was in short supply in one country but not in the other. Should we then say that Uruguay is not neutral unless it starts providing the country suffering from the shortage in that commodity? If by not helping it Uruguay is hindering it, then this conclusion is forced on us. But according to the common understanding of neutrality, Uruguay would have been breaking its neutrality if in the circumstances described it would have started supplying one of the parties with militarily useful materials after the outbreak of hostilities. It follows that the distinction between helping and hindering is crucial to an understanding of neutrality, as is the distinction between hindering and not helping. But it does not follow that this distinction is always of moral significance (neutrality is not always defensible). Nor does it follow that the distinction can always be drawn. Neutrality is possible in some cases, but it may be impossible in others. A second argument designed to show that neutrality is chimerical claims that whether or not a person acts neutrally depends on the base line relative to which his behavior is judged, and that there are always different base lines leading to conflicting judgments and no rational grounds to prefer one to the
The Morality of Freedom
113
others. Imagine that the Reds are fighting the Blues. We have no commercial or other relations with the Blues, but we supply the Reds with essential food which helps them maintain their war effort. If we want to be neutral, should we continue normal supplies to the Reds or should they be discontinued? If we continue supplying the Reds, we will be helping them more than the Blues. If we discontinue supplies, we will be hindering the Reds more than the Blues. (I am assuming that even if similar supplies to the Blues will help them, continuing not to help them is not hindering them.) It may be said that this is just one of the cases where it is impossible to be neutral. Without confusing not helping and hindering, such cases cannot be multiplied. They form a special class where, in the circumstances of the case, not helping is hindering. But the case invites a more radical rebuttal. In it two standards of neutrality conflict. The basic idea is simple. Neutrality is neutrality between parties in relation to some issue regarding which the success of one sets the other back. Various aspects of the parties’ life, resources, and activities will be helpful to them in the conflict, but many of these are resources and activities that they will have possessed or engaged in or wished to possess or to engage in in any case, even if they did not take part in the contest. Some of the activities and resources are such that the parties engage or wish to engage in them or possess them only because of the conflict. We could therefore distinguish between comprehensive and narrow neutrality. Comprehensive neutrality consists in helping or hindering the parties in equal degree in all matters relevant to the conflict between them. Narrow neutrality consists in helping or hindering them to an equal degree in those activities and regarding those resources that they would wish neither to engage in nor to acquire but for the conflict. ‘Neutrality’is used in ordinary discourse to indicate sometimes narrow and sometimes comprehensive neutrality. Sometimes various intermediate courses of conduct are seen as required by neutrality. This reflects the fact that several kinds of considerations may lead to different and incompatible policies all of which are commonly regarded as policies of neutrality, because all of them demonstrate an even-handed treatment of the parties either by not helping one more than the other, or by not helping one more than the other to take special measures to improve his position in the conflict, and so on. The difference between military equipment and food supplies illustrates the point. To be comprehensively neutral one should supply the Reds in our example neither with arms nor with food. But one is narrowly neutral even if one provides the Reds with normal food supplies.Arms on the other hand are needed specifically for military use, and continued supply of arms to one side is incompatible with neutrality.
114
Joseph Raz
Several subsidiary points require further argument. First, some arms are necessary for police action, some may be solicited and supplied in anticipation of a different possible conflict. This may show that unequal supply of arms can be compatible with narrow (though clearly not with comprehensive) neutrality. Second, some feel that no supply of arms to the combatants, however even-handed, is compatible with neutrality. My response is, in brief, that nosupply is not more neutral than even-handed supply. If it is thought to be preferable, this is for reasons having nothing to do with neutrality, such as a desire to bring the war to a speedy end or not to allow people to profit from wars. Third, my example assumed that we are regular suppliers of food to the Reds. This assures us that the continued supply is of normal quantities required anyway, but it also makes discontinuation of supply a positive harm rather than a refusal to help. Thus a person who has no regular trading relations with the parties can provide the Reds with their ordinary requirement of food while remaining narrowly neutral, or refuse to sell them anything that improves their position and be comprehensively neutral. We, being their regular suppliers, do not have the option of action that neither helps nor hinders. Finally, it may be claimed that real neutrality is comprehensive neutrality, narrow neutrality being compromised neutrality which is sometimes all that is possible. My point is merely that even if this is so, the fact remains that narrow neutrality is often all that is meant by ‘neutrality’. This brings us back, at long last, to the choice between the narrow and the comprehensive principles of neutrality broached in the last section. The conflict in which the state is supposed to be neutral is about the ability of people to choose and successfully pursue conceptions of the good (and these include ideals of the good society or world). It is therefore a comprehensive conflict. There is nothing outside it which can be useful for it but is not specifically necessary for it. The whole of life, so to speak, is involved in the pursuit of the good life. Can one be narrowly neutral in a comprehensive conflict? Furthermore, within the range of duties which the State owes its citizens, failure to help is hindrance. If I owe a client a duty to increase the value of his investment portfolio then if I fail to take action which could enhance its value I am positively harming my client’s interests. A surgeon who fails to take action to help the recovery of a patient prostrate in front of him in the operating theatre is hindering his recovery. So within the range of the state’s responsibility to its subjects failing to help is hindering. If the state is subjected to a requirement of comprehensive neutrality and if its duties to its citizens are very wide-ranging then the principle of comprehensive neutrality is a principle of neutrality indeed. On those assumptions the state can be neutral only if it creates conditions of equal opportunities for people to choose any conception of the good, with an equal prospect of real-
The Morality of Freedom
115
izing it. Egalitarian supporters of the neutrality principle such as Rawls endorse a variant of the principle of comprehensive neutrality. Libertarian supporters of neutral political concern gravitate towards the narrow principle. The difference is hardly surprising. It reflects the underlying, and all too often unexpressed, differing assumptions about the responsibilities of states and the limits of their authority. So much for clarifying the doctrine of neutrality. In evaluating its credentials our prime example will again be Rawls’ theory. Neutrality and the Social Role of Justice Since in the original position no one knows his own moral ideals, his own conception of the good, Rawls concludes that no perfectionist standards will be adopted in it. The principles of justice adopted in the original position are neutral between different conceptions of the good. To vindicate Rawls’ position one requires convincing reasons first for excluding moral and religious beliefs from the information available behind the veil of ignorance, and second for accepting that neutral or maximally neutral principles will be chosen in these circumstances. A Theory of Justice contains hardly any explicit argument for the exclusion of moral and religious beliefs from the original position. Such argument as there is turns on the need to secure unanimity, the need to have, in the original position, one viewpoint which can be the ‘standpoint of one person selected at random’ which excludes bargaining and guarantees unanimity. Therefore, only non-controversial information can be available. Otherwise there will be different standpoints defined by different informational bases, and it will be impossible to guarantee agreement. The obvious reply to this argument is that we need a reason to accept a decision reached behind this veil of ignorance, and the claim that no decision would be reached behind a differently constructed veil of ignorance is not such a reason unless it has already been shown, as it has not in fact, that we are bound by the results of some veil of ignorance, whatever it may be. The argument showing that no moral ideal of the good will be chosen as a principle of justice in the original position is the argument that Rawls uses for rejecting perfectionist principles: For while the persons in the original position take no interest in one another’s interests, they know that they have (or may have) certain moral and religious interests and other cultural ends which they cannot put in jeopardy. Moreover, they are assumed to be committed to different conceptions of the good and they think that they are entitled to press their claims on one another to further their separate aims. The parties do not share a conception of the good by reference to which the fruition of their powers or even the satisfaction of their desires can be
116
Joseph Raz
evaluated. They do not have an agreed criterion of perfection that can be used as a principle for choosing between instructions. To acknowledge any such standard would be, in effect, to accept a principle that might lead to a lesser religious or other liberty, if not to a loss of freedom altogether to advance many of one’s spiritual ends. If the standard of excellence is reasonably clear, the parties have no way of knowing that their claims may not fall before the higher social goal of maximizing perfection. Thus it seems that the only understanding that the person in the original position can reach is that everyone should have the greatest equal liberty consistent with a similar liberty for others. They cannot risk their freedom by authorizing a standard of value to define what is to be maximized by a teleological principle of justice. (pp. 327-8) Rawls, however, claims more than this argument establishes. He claims not only that the parties to the original position will avoid choosing any particular perfectionist principle as a constituent of their doctrine of justice, but that they will not even accept a doctrine of justice including an agreed process for determining which perfectionist principle should be implemented in the state. Given their concern that only well-founded ideals of the good, and not any ideal one believes in, should be implemented, and given the general knowledge of human fallibility, it is not really surprising that the original position does not yield a commitment to any particular ideal. But it may yield an agreement to establish a constitutional framework most likely to lead to the pursuit of well-founded ideals, given the information available at any given time.* Ignorance of one’s particular moral beliefs will not exclude this possibility, since the parties in the original position know that they have moral ideals. They accept, in other words, ‘a natural duty’ to pursue the best-founded moral ideal. This argument presupposes that moral ideals are based on rational considerations, and Rawls is anxious not to deny this: ‘in making this argument,’ he says, referring to the passage quote above, ‘I have not contended that the criteria of excellence lack a rational basis in everyday life’ (p. 328). An agreement on a method for choosing between perfectionist principles cannot be ruled out on the grounds that the methods of evaluating different ideals are themselves subject to evaluative controversy. They are not more controversial nor more evaluative than some of the psychological facts available to the parties, such as the Aristotelian principle and the considerations concerning self-respect on which the priority of liberty is based (cf. Sections 65,67, and 82). Nor has Rawls provided any argument to show that they are. In his Dewey lectures Rawls regards the exclusion of beliefs about ideals of the good from the original position as part of an exclusion of all controversial information and forms of reasoning. Moral and religious disagreements are seen as endemic to a democratic society, and the validity of the Rawlsian conception of justice is limited to such a society9The justification for excluding controver-
The Morality of Freedom
117
sial truths from the original position lies in the social role of justice, which is ‘to enable all members of society to make mutually acceptable to one another their shared institutions and basic arrangements, by citing what are publicly recognized as sufficient reasons.’1° The interest of subjecting one’s society and one’s life to such principles of justice is assumed to be everyone’s highest interest.” It is of the utmost importance that the Rawlsian method of argument requiring unanimity behind the veil of ignorance should not be defended on the ground that otherwise the resulting principles would not be fair because unanimity is a condition of fairness, or by any other moral argument. Rather, the reason must be that principles not generated by the Rawlsian method will fail to fulfill the social role of a doctrine of justice. The essential point is that a conception of justice fulfills its social role provided that citizens equally conscientious and sharing roughly the same beliefs find that, by affirming the framework of deliberation set up by it, they are normally led to a sufficient convergence of opinion. Thus a conception of justice is framed to meet the practical requirements of social life and to yield a public basis in the light of which citizens can justify to one another their common institutions. Such a conception need be only precise enough to achieve this result.’* Even if one accepts this view of the social role of a doctrine of justice, and gives it the priority Rawls assigns to it, his conclusions are not supported by his arguments for at least three reasons. First, even though different people differ in their conception of the good, it does not follow that in a given culture (and Rawls’ theory claims validity only within our culture) there are no common elements in their varying conceptions of the good. Such common elements need not be excluded by the veil of ignorance since their presence does not jeopardize the social role of the doctrine of justice. Second, in particular it is possible that there is a wide measure of agreement concerning the modes of reasoning by which ideals of the good are to be evaluated. Third, and most important, is Rawls’blindness to the possibility that the social role of a doctrine of justice may be met by consensus concerning the second best, given that an ideal constitution is not feasible. Rawls assumes without argument that the social role can be fulfilled only by a perfect doctrine of justice, i.e., one which establishes a perfect government for his actual society. But there is no reason why the doctrine of justice actually reflected by the constitutional arrangements of a state may not be reached as a result of people realizing that their different ideals of the good, each leading to a different doctrine of justice, cannot be implemented because of the widespread disagreement in society concerning their value. Each will then agree, as a second best,
118
Joseph Raz
to a doctrine of justice to which all or nearly all members of society could agree (as a second best) if they argue rationally each from his different conception of the good provided they are united in ranking the need for a public conception of justice very highly in their order of priorities. Some people may feel that Rawls’ original position is already a description of a process of compromise, of settling for the second best. This, one may feel, is implied by the need to disregard one’s conception of the good, thus jeopardizing one’s chances of implementing it. When so regarded the objection will be rephrased there is no reason to think that people who know they differ in their conceptions of the good but do not know how will reach a compromise, whereas those who also know how will not. In both cases one needs a common belief in the social role of justice and in its priority to be assured of a compromise. And in both cases if these conditions obtain some compromise will be reached, or at least there is a good chance that it will. A doctrine of justice reached in this alternative way is likely to differ from Rawls’ two principles. It will fulfill the social role of justice, but it will be agreed upon by a process of reasoning quite different from his. First, different ideals of the good far from being excluded will form the starting points of the argument about a doctrine of justice. Second, and as a result, supporters of different conceptions of the good will follow different routes in arguing for the doctrine of justice. There will be unanimity in the conclusion but (given the different starting point) no unanimity on the route to it. Third, the common feature of most routes will be the reliance on a rational reconstruction of a process of bargaining by which the common overriding goal of reaching an agreement leads the parties to compromise by accepting a less than perfect doctrine as the optimally realizable second best. The resulting compromise will be fair by Rawls’ own standards, in that each one of the parties conforms, while bargaining, to the principles of his morality. By the moral standards of some of the bargainers the claims of some of the others may be immoral. But this does not undermine the fairness of the outcome of the bargaining as long as all acknowledge the Rawlsain assumption of the overriding goal of achieving a common agreement concerning a doctrine of justice. This overriding goal entails that the best constitutional arrangements each person can reach while acting in conformity with his own moral ideals are morally valid, and since the commonly agreed upon arrangements are so regarded by everyone, they are morally binding on all. For reasons which are fairly obvious this procedure of bargaining may in some societies lead to the endorsement of highly wicked principles as that society’s doctrine of justice. This should be regarded as a reductio ad absurdurn of the presupposition of Rawls’ theory of justice and especially the assumption of the overriding goal to reach a doctrine of justice at whatever
The Morality of Freedom
119
cost to one’s conception of the good. I am not advocating the bargaining procedure described above. My purpose was merely to show that, though it is a perfectionist procedure, it is not ruled out by Rawls’ arguments against perfectionism; and to suggest that the assumption that he relies upon against perfectionism leads to strongly counter-intuitive results. From Neutrality to Pluralism Rawls’ Kantian constructivist approach to ethics suggests an argument for political neutrality which is in part independent of the one criticized above and which turns on autonomy: Kant held, I believe, that a person is acting autonomously when the principles of his action are chosen by him as the most adequate possible expression of his nature as a free and equal rational being. The principles he acts upon are not adopted because of his social position or natural endowments, or in view of the particular kind of society in which he lives or the specific things that he happens to want. To act on such principles is to act heteronomously. Now the veil of ignorance deprives the persons in the original position of the knowledge that would enable them to choose heteronomous prin~ip1es.l~
Rawls applies the Kantian insight only to the choice of the principles of justice. But it seems that if it is valid for that purpose, it must be valid for morality generally. In A Theory ofJustice Rawls does not use the Kantian insight as an argument for the acceptability of his theory. He merely points out that his theory is consistent with it. But it ‘Kantian Constructivism in Moral Theory’ the Kantian insight is used to defend his theory against various criticisms by providing it with an epistemological foundation. Applied to moral considerations generally, the Kantian insight yields a new argument for excluding the parties’ moral and religious beliefs behind the veil of ignorance. While not explicitly using this argument, Rawls comes close to doing so in saying that the parties in the original position do not agree on what the moral facts are, as if there already were such facts. It is not that, being situated impartially, they have a clear and undistorted view of a prior and independent moral order. Rather (for constructivism), there is no such order, and therefore no such facts apart from the procedure of construction as a whole; the facts are identified by the principles that res~1t.l~
In developing the Kantian insight along Rawlsian lines, several points must be borne in mind. First, the fundamental idea which enjoys universal validity is that morality is the free expression of a person’s rational nature. The claim
120
Joseph Raz
that principles chosen in the original position express this nature depends on a certain conception of the person which is among the deep common presuppositions of our culture but no more.15 Second, Rawls is anxious to make clear that the choice of principles because they express human nature ‘is not a socalled “radical” choice: that is, a choice not based on reasons. . . . The notion of radical choice . . . finds no place in justice as fairness.’16 Furthermore: ‘The ideals of the person and of social cooperation . . . are not ideals that, at some moment in life, citizens are said simply to choose. One is to imagine that, for the most part, they find on examination that they hold these ideals, that they have taken them in part from the culture of their society.’17The conception of morality as an expression of the rational nature of people is consistent with the view that people’s nature is socially determined, thus rendering the concrete manifestation of morality equally socially determined. Third, Rawls’ conception of the person does not lead to unanimity of moral views. Given this conception of morality, it is evident that the parties to the original position know nothing of their moral beliefs. They are not excluded, as is knowledge of the particular circumstances of each participant’s life, in order to ensure that choice in the original position represents people’s nature. They are in fact not excluded at all. The parties to the original position have no knowledge of their moral beliefs because they have as yet no moral beliefs and because whatever moral beliefs they should have depend on the outcome of their deliberations in the original position and cannot affect it. Thus extending the Kantian insight beyond political morality to morality as a whole explains the elimination of moral beliefs and ideals of the good from behind the veil of ignorance. But this argument by itself does not justify political neutrality. The argument for neutrality still rests on the further assumption that it is in people’s highest interest to adopt principles fulfilling the social role of justice. Without its support we still have no reason to believe that any agreement will be reached behind the veil of ignorance. Rawls shows neither that this assumption follows from the Kantian insight nor that it leads to neutral political concern. The counter-argument of Section Four above still holds good. This reconstruction of Rawls’ argument for the doctrine of neutral political concern attempts to found it on the notion of autonomy through the notion of moral self-determination. The intuitive idea behind it seems to be this: since morality is an expression of one’s rational nature, it is essentially self-determined. Given the social determination of the concept of a person and the absence of unanimity in the outcome of moral deliberation, the only proper course seems to be to endorse constitutional arrangements neutral between conceptions of the good in order to enable all individuals to develop and pursue their own conception of the good. Since no conception of the
The Morality of Freedom
121
good which expresses the rational nature of the person upholding it is better than any other, the constitutional arrangements should be neutral between them. The role of the state is to enable all persons to express their nature and pursue their own autonomously conceived conception of the good and plan of life. As noted, this intuitive idea relies on a plausible-looking but unfounded belief in the acceptance of a need for unanimously approved principles of justice as in everyone’s highest interest. This is enough to reject it. Furthermore, this way of expressing the intuitive idea is unlikely to gain Rawls’ approval. It advocates not neutral political concern as a principle of restraint but neutrality between those conceptions of the good which greatly value an autonomous development of one’s life in accordance with one’s rational nature. It is in fact not a doctrine of neutrality but of moral pluralism. Moral pluralism is often thought to necessitate neutral political concern. It will be the main task of Part Four of the book to explore the rationality of perfectionist moral pluralism, i.e. of pluralism of many forms of the good which are admitted to be so many valuable expressions of people’s nature, but pluralism which allows that certain conceptions of the good are worthless and demeaning, and that political action may and should be taken to eradicate or at least curtail them. For the present all we need do is point to the logical gap between pluralism and neutrality. Moral pluralism asserts the existence of a multitude of incompatible but morally valuable forms of life. It is coupled with an advocacy of autonomy. It naturally combines with the view that individuals should develop freely to find for themselves the form of the good which they wish to pursue in their life. Both combined lead to political conclusions which are in some ways akin to those of Rawls: political action should be concerned with providing individuals with the means by which they can develop, which enable them to choose and attempt to realize their own conception of the good. But there is nothing here which speaks for neutrality. For it is the goal of all political action to enable individuals to pursue valid conceptions of the good and to discourage evil or empty ones. Or at least that is the argument of Part Four. Our provisional conclusion can be no more than that Rawls and others arguing for neutrality in similar ways fail to establish their case, and that sometimes they assume too quick or simple a connection between neutrality and personal autonomy.
The most deeply rooted confusion leading to the intuitive appeal of the antiperfectionist is in the thought that anti-perfectionism is necessary to prevent
122
Joseph Raz
people from imposing their favored style of life on others. The confusion is both practical and moral. On the practical side it assumes that perfectionist action is aimed by one group at another, attempting to bring it to conform with its habits and way of life. This need not be the case. Perfectionist political action may be taken in support of social institutions which enjoy unanimous support in the community, in order to give them formal recognition, bring legal and administrative arrangements into line with them, facilitate their use by members of the community who wish to do so, and encourage the transmission of belief in their value to future generations. In many countries this is the significance of the legal recognition of monogamous marriage and prohibition of polygamy. Furthermore, not all perfectionist action is a coercive imposition of a style of life. Much of it could be encouraging and facilitating action of the desired kind, or discouraging undesired modes of behavior. Conferring honors on creative and performing artists, giving grants or loans to people who start community centers, taxing one kind of leisure activity, e.g., hunting, more heavily than others, are all cases in which political action in pursuit of conceptions of the good falls far short of the threatening popular image of imprisoning people who follow their religion, express their views in public, grow long hair, or consume harmless drugs. Finally, the view we are discussing assumes a rigoristic moral outlook, that is one allowing for only one morally approved style of life. That is why it is suspected that if some people pursue a different style of life from that practiced by those with political power they will be persecuted. Perfectionism is, however, compatible with moral pluralism, which allows that there are many morally valuable forms of life which are incompatible with each other. That possibility will be examined in detail in Part Five. If a plurality of incompatible, even rival, forms of life is valuable then perfectionism would not lead to the suppression of forms of life which are not practiced by those in power. Even if the anti-perfectionist worry about people imposing their conceptions of the good on others suffers from exaggeration for the reasons just described, is there not a simple argument supporting its conclusion? Perfectionism assumes that some people have greater insight into moral truth than others. But if one assumes that all stand an equal chance of erring in moral matters should we not let all adult persons conduct themselves by their own lights? Whatever else can be said about this argument one point is decisive. Supporting valuable forms of life is a social rather than an individual matter. Monogamy, assuming that it is the only morally valuable form of marriage, cannot be practiced by an individual. It requires a culture which recognizes it, and which supports it through the public’s attitude and
The Morality of Freedom
123
through its formal institutions. Much more will be said on this point later in the book. T h e short summary is that perfectionist ideals require public action for their viability. Anti-perfectionism in practice would lead not merely to a political stand-off from support for valuable conceptions of the good. It would undermine the chances of survival of many cherished aspects of our culture. The sources of the appeal of anti-perfectionism are sound. It stems from concern for the dignity a n d integrity of individuals a n d a revulsion from letting one section of the community impose its favored way of life on the rest. These concerns are real a n d important. They do not, however, justify antiperfectionism.
Notes 1. R. Nozick, Anarchy, State, and Utopia (New York: Basic Books, 1974), 33 (italics in the original). 2. This interpretation of the doctrine is silent on whether individual political action (voting in elections etc.) may rightly aim at the promotion of some conception of the good. Little attention has been paid to this as a separate issue in ‘neutralist’writings. 3. A. Montefiore in the book he edited, Neutrality and Impartiality (Cambridge: Cambridge University Press, 1975), 5. 4. Ibid., 7 5. R. Nozick, Anarchy, State, and Utopia, 272-73. 6. T. Nagel, “Rawls on Justice,” Philosophical Review 82 (1973): 220, reprinted in Reading Rawls, ed. N. Daniels (Oxford Oxford University Press, 1975), to which all page references refer. 7. J. Rawls, A Theory ofJustice (Oxford Oxford University Press, 1971), 93-94. 8. This point raises a wider issue of some importance to the evaluation of Rawls’ procedure. Should one assume that the parties take notice in the original position of their own fallibility, and agree on constitutional arrangements that will be selfcorrecting if it turns out that the fundamental beliefs concerning human nature, on which their substantive principles of justice are based, turn out to be wrong, or not? One might think that they need not. We could always engage in the Rawlsian form of argument and apply it to the new information once it becomes available. This seems to be the way Rawls treats the problem (though he is sensitive to fallibility concerning applied principles (see pp. 195-97)). But quite apart from the fact that by this one assumes that in the original position the parties discount a general fact of human nature, namely its fallibility, this procedure will sanction non-adaptable constitutionsa highly counter-intuitive conclusion. 9. Rawls, “Kantian Constructivism in Moral Theory,” Journal of Philosophy 77 (1980): 54042. 10. Ibid., 517 and cf. 561.
124
Joseph Raz
Ibid., 525. Ibid., 561. Rawls, A Theory offustice, 252. Rawls, “Kantian Constructivism and Moral Theory,” 568. Cf. Ibid., 520. It seems to me that Rawls is ambiguous on the crucial issue of whether the reflective equilibrium argument supports the fundamental Kantian insight or presupposes it. I assume here that there is a sense in which the latter is the case, though the full story is more complex and cannot be explored here. 16. Ibid., 568. 17. Ibid., 568-69. 11. 12. 13. 14. 15.
Perfectionism Thomas Hurka
Introduction In this selection, Hurka argues that perfectionism can affirm principled limits to state interference in a citizen’s self-regarding actions. It can do so, he argues, because autonomy is itself a perfection. Free choice from many life options exercises deliberative rationality by organizing an agent’s intentions as they are realized in the world. Yet, while autonomy is plausibly a perfection, Hurka contends that it is merely one among others. Accordingly, it cannot ground absolute limits to state interference. Nonetheless, the perfectionist value of autonomy does show that state interference in self-regarding actions always brings some cost. After justifying autonomy’s claim to be a perfectionist good, Hurka argues that while autonomy provides support for limits on the state’s power, it does not provide support for the principle of state neutrality. He suggests a number of ways by which the state can promote valuable activities while adequately respecting the liberty of its citizens.
Thomas Hurka, Perfectionism (Oxford: Oxford University Press, 1993), 148-52, 158-60. Copyright 0 1993 by Thomas Hurka. Reprinted by permission of Oxford University Press, Inc. - 125 -
Thomas Hurka
126
Autonomy as a Perfection
P
often focused on the harm principle, or liberty principle. Stated initially by Humboldt and John Stuart Mill, this principle says that the state must never coerce its citizens except to prevent them from harming others or interfering with others’ liberty. This principle defines classical liberalism and is what perfectionism is most often said to violate. It is therefore our subject in this chapter. We will ask whether perfectionism can affirm a version of the liberty principle or, on the contrary, favors state interference with citizens’ self-regarding action. HILOSOPHICALDISCUSSIONSOF LIBERTY HAVE
1.1
Let us start by considering perfectionism just in the broad sense, without foundation in human nature. Broad perfectionism can most easily affirm a liberty principle by treating autonomy, or free choice from many life options, as itself an intrinsic good. If self-determination is itself a perfection, any restrictions on it are prima facie objectionable. Humboldt and Mill both take this broadly perfectionist line. Humboldt says it is part of human perfection that one “strives to develop himself from his own inmost nature, and for his own sake.”’ For him the chief requirements for perfection are therefore “freedom” and, what he calls “intimately connected” with freedom, a “variety of situations.”*Mill says the qualities that are “the distinctive endowment of a human being” are exercised “only in making a ~hoice.”~ To encourage such choices he recommends, again, no interference in the private realm and “experiments of l i ~ i n g to ” ~make different life options vivid for all. Humboldt and Mill do not affirm just some liberty principle; they affirm an absolute principle. They say that in actual conditions the state must never interfere in citizens’ private lives. Can this absolute constraint be derived from a perfectionist valuing of autonomy? Even considering only the value of autonomy, it cannot. Sometimes restricting a person’s autonomy now will do more to increase her autonomy in the future, by giving her more options in the future or a greater capacity to choose autonomously among them. Humboldt and Mill may believe that this situation arises infrequently and that governments are especially inept at recognizing it. Nonetheless, it is one bar to a simple argument from the value of autonomy to an absolute liberty principle. A more serious impediment comes from the recognition of perfections other than autonomy. No plausible value theory can treat free choice as the only intrinsic good. It must acknowledge some other goods, so that, for ex-
Perfectionism
127
ample, freely chosen creativity is better than freely chosen idleness, and autonomous knowledge is better than autonomous ignorance. Does this not reopen the door to interference? If there are goods other than autonomy, may sufficient increases in them not outweigh any loss in autonomy? They may do so if autonomy is just one good among others, to be weighed against them in standard fashion. But there are other possibilities. A broad perfectionism can treat autonomy as lexically prior to other perfections, so that even small losses in it outweigh large gains in those others. More moderately, perfectionism can treat some minimum of autonomy, one involving reasonable self-determination, as lexically prior, so that losses that take one below this minimum (although not other losses of autonomy) outweigh large gains in other perfection^.^ Finally, the theory can treat autonomy, not as a good among others, but as a condition of goods. Then states and activities that would have value if they were chosen freely have none if they are coerced. Autonomy is not a competitor with other goods, but a condition of their worth. These three views all limit the possibility that losses in autonomy will be outweighed by increases in other goods, but only the first-full lexical prioritydoes so entirely. The second and third views exclude the most serious restriction on liberty: forcing people into a single best activity. Imagine that ten selfregarding acts are ranked in value from 1 to 10. An absolute liberty principle condemns laws that forbid just the tenth-ranked option, but these two views do not. Such laws leave people free to choose among the nine remaining options, thus leaving them reasonable autonomy and any autonomy that could plausibly be a condition of goods. Their loss of liberty is not of the kind that, according to the two views, has special political significance. Whatever their precise implications, none of the three views is plausible as a view about the good. Imagine that Mozart was as a young boy forced into music, so his life lacked reasonable autonomy and the autonomy that may be a condition of goods. All three views imply that Mozart’s life, despite its great musical achievements, contained less perfection than if he had been given freedom in his youth and had autonomously chosen a life of suntanning. This claim is not one that, if we think seriously about the good, we can make. Even if autonomy has some value, it cannot have so much as to outweigh all Mozart’s music. A plausible broad perfectionism, then, can treat autonomy only as one good among others, which may sometimes be outweighed. It therefore cannot endorse an absolute liberty principle, but it can endorse a non-absolute principle. If free choice is intrinsically good, any restriction on it threatens some perfectionist cost and is therefore prima facie objectionable. Other things equal, the state should not interfere with liberty except to protect the greater liberty of others.
128
Thomas Hurka
1.2
It is important that autonomy can be added to a list of merely broad perfections, but it would be more impressive to justify its place there. In our Aristotelian perfectionism, we would need to show that autonomy develops human nature by exercising sophisticated rationality. Can autonomy be valued as an exercise of reason? There is an obvious connection between autonomy and rational deliberation. Someone with many life options can reflect upon these options and in so doing exercise his rational powers. He can weigh their respective merits and defects and reach a reasoned judgment about them. This connection is not, however, quite what we want. As described previously, deliberation is an intellectual activity that does not require options actually to be open. Someone who has no free choice-a slave at the extreme-can stiU evaluate possibilities. He can ask which career would be best for him if he were able to choose careers and can arrive, in principle at least, at the same intellectual conclusion as someone who has many careers open. Of course, a slave is unlikely to go through this process. People rarely deliberate about options they cannot choose, and this is an important instrumental argument for liberty. By giving its citizens many alternatives, a free society encourages reasoning that would have no practical point if options were closed. But the argument is instrumental and does not yet give us rational values present in the act of choice itself. To locate these values, we must examine choice more closely. Imagine that one person chooses life a from among ten life options, while another person has only life a available. It may be true of each that she has made a the case and is in that sense responsible for it. But there is an important difference between them. The first or autonomous person has also made nine alternatives to a not the case: If her options included b, c, and d, she is responsible for notb, not-c, and not-d. The second person did not have this further effect. Her non-realization of b, c, and d is due, not to her, but to nature or to whatever person limited her options. This difference is important because practical perfection involves agency: It involves expressing intentions in the world and determining what it does and does not contain. The autonomous agent, by virtue of her autonomy, more fully realizes this ideal. When she makes choices, she has two effects: realizing some options and blocking others. She has a more extensive causal efficacy than someone who lack options, and a higher score on the practical dimension of number. By letting her determine what she does not do as well as what she does, her autonomy makes her more widely active and more practically efficacious! To have value, this efficacy requires more than just the availability of options. A person must know about the options or she cannot intend their non-
Perfectionism
129
realization. Moreover, she must make in the fullest sense a choice among them: a choice that is for one option and against others, so her rejection of the others appears in her mind. This does not always happen in intentional action. Someone who is driven by an obsession may know that alternatives to her act are available, but her intentions do not reflect this. They go blindly for b, if b is what she does, without preferring b to other options. (A strong claim is that the obsessed person intends only b, without rejecting anything; a weaker claim is that she rejects only the vague alternative not-b. Either way there is not the rejection of individually discriminated options that on my view increases agency.) The same is true of someone who is weak-willed. If he acted on an allthings-considered desire for a, he would prefer a to other options. But he does not act weakly in preference to other options; he just acts weakly If autonomy involves successfully realizing many intentions, it requires choice in the fullest sense: a simultaneous realization of some possibilities and rejection of others, with one’s rejection of the others reflected in one’s will. Autonomy realizes agency even more so when it follows deliberation. (We can now take up the earlier suggestion about practical reasoning.) Someone who reflects on her options may find that a has the most of desirable property F, b has defect G, and so on. When she chooses, therefore, she can intend not only a but the-act-with-the-most-F,and in rejecting b she can reject the-act-with-G. Her deliberative knowledge, if it guides her choice, can give her more intentions in and around her options than if she picked blindly among them. There is a further effect. If she has deliberated, she can choose a and reject b as means to a single goal, perhaps getting the-most-F-without-G. Her positive and negative intentions can form a unified means-end structure, and the more complex this structure is, the more it embodies dominance. Then we have not just autonomy but deliberated autonomy: free choice from a wide range of options that reflects practical reasoning about them. This kind of autonomy realizes deliberative rationality, but in a more than intellectual way. The elements it organizes are not beliefs, which are available to a slave, but intentions realized in the world. It therefore presupposes the simpler autonomy that consists in any free choice among options. Its foundation is a set of acceptances and rejections that converge on one goal, which is possible only for an agent who has many options to accept or reject. 1.3
An objection may be raised against this account of autonomy. If it is good to achieve additional ends, it may be argued, we should, when doing something trivial such as lifting a fork, think of the many things we are not doing and consciously reject them. We should think that we are not lifting our knife, using our other hand, and so on. Is this not absurd?
130
Thomas Hurka
The reference to consciousness here is a red herring because ends can be willed or rejected without conscious awareness. Even apart from this point, the objection fails by ignoring the dimension of quality. The alternatives we could reject when we lift a fork are highly particular and, as such, of minimal worth. So if autonomously manipulating cutlery distracts us from greater goods, as it surely will if it is conscious, it is on balance wrong. The point is like one about knowing the number of redheads in Beiseker, Alberta. In itself this knowledge has (minimal) value, but someone who takes the time to acquire it will hardly be maximizing his excellence. In the same way, some gains in autonomy are too trivial to be worth seeking. What matters in Aristotelian perfectionism is the autonomous choice of encompassing ends, ones that shape a day, month, or life. In this kind of choice, we both select and reject goals of high quality and thereby realize substantial agency. This last point is a further implication of dominance. Earlier this measure preferred autonomous choices that follow deliberation; here it and extent prefer choices, whether deliberated or not, that are of important or organizing ends. This implication fits the views of Humboldt and Mill, who care most about the freedom to fE the general shape of one’s life or one’s general guiding ends. There are doubtless instrumental reasons for this preference; any extrinsic benefits of free choice are surely greatest when the choice is most far-reaching. But there is also an argument about intrinsic value. Autonomous choice seems most valuable, as autonomous choice, when it is of important, organized ends. Our generality measures explain why. There is also a connection with Humboldt’s and Mill‘s interest in “a variety of situations” and “experiments of living.”As we have seen, autonomy requires knowledge of available options, but it seems fullest when this knowledge is most extensive, when the options are grasped not vaguely but in articulated detail. Again our generality measures explain why. Someone who lives in a varied society can choose among life forms he knows at first hand and, when he chooses, accept and reject not just vague possibilities but connected series of acts. Within a large hierarchy uniting choices and rejections, he can have many smaller hierarchies representing all the worked-out options he declines. There is therefore an Aristotelian account of autonomy, which values most highly those choices that follow deliberation, are of organizing ends, and involve articulated knowledge. But the account has an important limitation. Because it values autonomy for the same reason it values other goods, as a realization of rationality, it cannot plausibly treat autonomy as special among goods, for example, as lexically prior to or a condition of goods. This limitation, however, is not troubling. The stronger valuings of autonomy are not intuitively plausible, and a perfectionism that rejects them still has at least one argument against any illiberality: If autonomy is one intrinsic good, interfering
Perfectionism
131
with citizens’ private lives always threatens some perfectionist harm. This argument may not always be decisive, especially against the milder illiberality of merely forbidding one worst option. But its non-absolute character is less important when we realize that it does not operate alone. Beyond the autonomy view is a further perfectionist argument for liberty, one that would be available even if autonomy were not directly valuable or free choice itself a good.
Liberty versus Neutrality 2.1
Although perfectionism supports liberty, it does not go beyond it to support the stronger ideal of state neutrality. According to this ideal, the state must not only not coerce citizens to make them better, it must never aim, coercively or otherwise, to promote one set of values over others. It must be neutral about the good, never having as its justification for acting that some ways of life are intrinsically preferable to other^.^ Neutrality is not a traditional liberal ideal, for it is rejected by Mill: He thinks a person’s choosing badly, although no reason to coerce her, does justify “remonstrating” and “reasoning” with her.8 Nor is neutrality supported by our perfectionist arguments. These arguments tell strongly against coercing citizens into the good, but they do not have the same force against noncoercively promoting the good. Some such promotion can have bad effects: Offering rewards for excellence can induce bad motives just as much as can threatening punishments. But many of its forms do not diminish citizens’ autonomy, reduce their routes to excellence, or undermine valuable inner states. There is much non-coercive promotion of the good that perfectionism approves. The state can, first, promote perfection through its education system. Its schools can teach students about the natural world and the history of their culture, in part because knowing these subjects is intrinsically good. They can also introduce students to literature, music, and athletics. The schools’ efforts here will not be undiscriminating; there will not be courses exposing students to drug-taking or professional wrestling. The education system will lay the foundations for valuable activities, not for ones of minimal worth. The state can continue by educating adults. The Canadian government sponsors advertisements encouraging citizens to participate in sports and other physical activity, in part because such activity is intrinsically good.
132
Thomas Hurka
Similar publicity for other perfections could, if free of non-rational persuasion, lead people in valuable directions without bad effects. The state can also subsidize valuable activities. Some people’s best talent may be for a rare perfection, one that few people practice. Their limited numbers may make this perfection more expensive than others that are intrinsically no better, but whose many practitioners permit economies of scale to be realized. Here the state can subsidize the rare activity, giving those drawn to it the same chance at a good life as the majority. Of course, the costs of subsidization must not be too great, but this condition is often met. When it is, subsidization encourages perfection in the minority, by making a good option more available, but does not coerce them into anything. The state may have other reasons for subsidization. There may be valuable activities that people cannot appreciate when they are young and will not try as adults if the initial costs of doing so are high. Here state aid, by lowering these costs, can help more people discover a valuable option. The state may also subsidize activities whose value people already know. It may believe that, although citizens have some tendency to seek out valuable activities, they have other, less desirable impulses, for example, to engage in consumption or passive amusement. It may fear that if the best activities cost their full market price, many who would choose them at lower prices will prefer something less valuable. It may therefore subsidize the activities to help more people’s better tendencies actually guide their conduct. These various justifications come together in a policy of government support for the arts, for example, for performing arts companies that tour across a country. By supporting these companies and not professional wrestling, the government affirms its belief that the arts are more valuable than professional wrestling and invites citizens to reflect whether this does not match their own convictions. Its support may permit some arts, for example, opera, to survive that on their own would be unsustainable, and takes them to places, such as smaller towns, that could not otherwise experience them. Finally, by lowering ticket prices, state support increases the number of people who attend the performing arts, both for the first time and regularly. The support does not force anyone to attend; they can watch wrestling if they prefer. It merely makes a valuable option more available and easier to choose from valuable motives. 2.2
This non-coercive promotion of the good would be less necessary if some optimistic assumptions about humans, for example, a strong natural desire doctrine, were true. If all humans wanted above all to develop their natures, there would be less need for publicity encouraging them to do so. There would
Perfectionism
133
also be less need for subsidies: Not only would the market cost of good activities be lower but also people would be willing to pay more for them. In chapter 3 we rejected the desire doctrine as unrealistically optimistic, and this has been reflected in our discussion of neutrality. Not believing that humans left on their own will always choose what is best, we have approved some state aid to help them do so. At the same time, our position does not go to the opposite extreme and see humans as depraved. This was implicit in our composite case for liberty. That forbidding bad options prevents people’s autonomously choosing against them would not be a relevant point if those who had the options could not resist them but would succumb willy-nilly to temptation. Similarly, the argument that coercion induced bad motives assumes that people who are not coerced can sometimes choose from good motives, an assumption I would affirm. In considering the tendency doctrines, we endorsed a moderate position like Green’s: Humans have some tendency to the good and in favorable conditions can follow it, but they also have other desires that they can need help in resisting. It is therefore fitting that on a political question we also reach a position like Green’s: generally favorable to liberty but rejecting the stronger ideal of state neutrality.
Notes 1. Wilhelm von Humboldt, The Limits of State Action, trans. J. W. Burrow (Cambridge: Cambridge University Press, 1969),19. 2. Ibid., 16. 3. John Stuart Mill, On Liberty, in Utilitarianism, Liberty and Representative Government (London: J. M. Dent, 1910), 116. 4. Ibid., 115. 5. James Griffin, Well-Being: Its Meaning, Measurement and Moral Importance (Oxford: Clarendon Press, 1986),24345. 6. Although autonomy realizes agency, it is not identical to agency because it involves only intentions about one’s own life. Determining what is and is not true about other people or the material world involves a form of agency distinct from autonomycall it external power. 7. John Rawls, A Theory of Justice (Cambridge, Mass.: Harvard University Press, 1971);Ronald Dworkin, “Liberalism:’ in Public and Private Morality, ed. Stuart Hampshire (Cambridge: Cambridge University Press, 1980); and Bruce Ackerman, Social Justice and the Liberal State (New Haven, Conn.: Yale University Press, 1980). 8. Mill, On Liberty, 73, also 133. Humboldt does endorse neutrality (The Limits of State Action, 25), but only by stretching his arguments beyond plausibility.
This Page Intentionally Left Blank
II ORIGINAL PAPERS
This Page Intentionally Left Blank
10 Liberal Neutrality: A Compelling and Radical Principle” “
Gerald F. Gaus
In this paper, Gaus defends a radical neutrality principle. His principle precludes most contemporary legislation in liberal states. Although the principle would have radical consequences for social policy, Gaus argues that it can be derived from basic and intuitively compelling claims about moral reasoning. Central to his argument is the claim that coercive actions by the state must be impartially justifiable. To pass this test, the reasons put forth in defense of these actions must be justifiable to all fully rational beings. Since there is general disagreement among fully rational beings about claims about the good life, Gaus argues that reasons grounded in such claims will not pass this test. Moreover, while many particular goods and values pass the test of impartial justification, Gaus holds that the same cannot be said of rankings or justifications of trade-offs between them. Since such trade-offs are integral to most state action, the neutrality principle rules out most state action, including efforts to make “others more perfect in our own eyes.”
- 137-
Gerald F. Gaus
138
1. Introduction
C
contemporary political philosophy, the light-to-heat ratio of discussions of neutrality has been somewhat dismal. Although most political philosophers seem to know whether they are for it or against it, there is considerable confusion about what “it” is. To be sure, some of this ambiguity has been noted, and at least partially dealt with, in the literature. Neutrality understood as a constraint on the sorts of reasons that may be advanced to justify state action is regularly distinguished from “consequential neutrality”-that the effects of state policy must somehow be neutral.’ Yet, interpretations of neutrality are far more diverse than most analyses recognize.2Neutrality is sometimes understood as a doctrine about the intent or aim of legislation or legislators? the proper functions of the state: the prohibition of the state “taking a stand” on some i s s ~ e sthe , ~ prohibition of the state enforcing moral character,6 or the requirement that the state take a stance of impartial it^.^ Alternatively, neutrality can be understood as a requirement of a theory of justice rather than state action.* There are also differences about whether neutral states (or theories of justice or legislators) are supposed to be neutral between conceptions of the good? particular sets of ends,’*comprehensive doctrines and conceptions of the good,I1particular or substantive conceptions of the good,I2ways of life,13 final ends,14or controversial conceptions of the g00d.l~And it is unclear whether every principle of neutrality is inherently one of liberal neutrality, or whether liberal neutrality is a specific sort of neutral principle.I6 How to select one conception of neutrality over its competitors? One method-implicit in many defenses of neutrality-is to select a conception that seems intuitively plausible in the sense that we can imagine that the principle could be followed by a state that does most of what our current liberallike governments do. This conception of neutrality can then be offered as an articulation of the principle of legitimate state action underlying contemporary liberal-like states. In response, critics of neutrality can seek to show that the proposed conception cannot account for the state as we know it. I do not employ that method here. As I argue in sections 4 and 5, this approach is ideologically conservative, supposing that liberal neutrality must somehow justify current liberal-like regimes. The task of political philosophy is not to legitimate current regimes, but to examine the conditions under which political coercion can be justified. In any event, rather than presenting an independently intuitively plausible conception of neutrality and then seeking to defend it, I shall work the other way around. Section 2 derives a principle of neutrality, which I call “minimal neutrality,” from basic claims about the nature of morality and reasoning. The OMPARED TO OTHER DEBATES IN
“Liberal Neutrality: A Compelling and Radical Principle”
139
problem of selecting the appropriate conception of neutrality is thus solved by the derivation: the appropriate conception of neutrality is that which is derived from the foundational claims about moral reasoning. Because these basic claims are intuitively compelling, so too is the conception of neutrality to which they give rise. Liberal neutrality, I argue in section 3, can be derived from minimal neutrality by two additional claims. Although these claims are compelling, they are somewhat more contentious than are the claims on which minimal neutrality is based; still, I shall argue, it is difficult to reject them. Hence, liberal neutrality is also a compelling principle. Section 4 then argues that this favored conception of liberal neutrality is a radical principle, which shows much of what is currently done in the name of liberalism to be illegitimate. Section 5 concludes with a brief defense of radical political philosophy. 2. Minimal Neutrality 2.1
I begin with a moral judgment that is well-nigh universally accepted, and deeply compelling: (I) (a) It is prima facie morally wrong for Alfto coerce Betty, or to employ force against her. (b) With suficient justification, the use of coercion or force by Alfagainst Betty may be morally justified.
Claim (I) is certainly consistent with what Joel Feinberg calls the “presumption in favor of liberty”: “liberty should be the norm, coercion always needs some special ju~tification.”’~ And, as I have argued elsewhere, something like this presumption is foundational to liberalism.18 However, claim (I) is not a particularly liberal moral principle. It asserts nothing about a presumption in favor of liberty, only that, in the absence of adequate justification, it is morally wrong for Alf to coerce Betty. The core claim of (I) is that, other things being equal, the use of force or coercion against another is wrong. To show that other things are not equal, and so that the use of force and coercion is morally permissible, a moral justification is required.19 A complete exposition of claim (I) would have to identify the set of agents who can occupy Alf’s and Betty’s roles. Philosophers disagree about how the set is to be defined: is it to be restricted to humans, or does it include some nonhuman animals? Does it include all humans, or only those who have a set of capacities that qualify them as persons (say, self-consciousness or project
140
Gerald F. Gaus
pursuit)? I shall put aside these disputes and suppose that we can identify a set of moral persons: agents who possess a minimal ability to direct their actions according to moral requirements. Given that my aim is to derive neutrality from basic moral intuitions, it makes sense to restrict my focus to those who possess, in some minimal degree, the ability to direct their actions according to moral requirements. It is not part of my thesis that those who do not possess moral personality so conceived are outside the protection of morality; that is another question. I shall simply restrict my focus here to the set of moral persons; I take it that this includes all but the most severely injured, insane, or psychopathic personalities.20In any event, we shall see that this narrowing of the set of agents serves only to make the principle of neutrality less radical; those who seek to broaden the set will also even further radicalize the principle of neutrality. 2.2
I do not claim that it is impossible to deny (I), or that no one has ever done it. There are two paths to disputing claim (I). Neither, I think, has much plausibility. (i) Most obviously, one may reject clause (a) by insisting that it is morally permissible for Alf to employ coercion against Betty without justification because there is no constraint at all on the moral permissibility of Alf coercing Betty. This might be because, first, there is no constraint on anyone employing coercion on anyone else. Hobbes held that in the state of nature each has a “blameless liberty” to use his power and ability as he sees fit.21In this case, each would have a Holfeldian liberty to coerce the other: one may coerce another without offering any justification as one is not under any duty to refrain from coercing.22A related way of denying clause (a) is to assert that some people, such as Alf, do not have to justify their coercive acts, though perhaps others do. Perhaps, some naturally have a superior status over others such that they can coerce them simply to advance their own desires and purposes, and need offer no moral justification for doing so. Thus, a Superman destroyer of conventional egalitarian morality, asserting his will over others, I take it, would deny clause (Ia). Neither a morality inspired by Hobbes’s blameless liberty, nor a Nietzschean will to power is a compelling moral view; I shall not spend time here seeking to refute them.23 (ii) Clause (b), following W. D. Ross, is put in terms of prima facie wrongn e ~ sTo . ~say ~ that an act is prima facie wrong is to say that other things being equal, performing the action is wrong, that the act is presumptively wrong, or that it has a wrong-making property; the wrongness, however, is only presumptive, and good reasons can show that the act is morally permissible. An
“Liberal Neutrality: A Compelling and Radical Principle”
141
absolutist pacifist might deny clause (b) by insisting that the use of coercion can never be justified. Given claim (11) below (Q2.3), this implies that we ought not to defend ourselves against aggressors. Though some have advanced such views, the claim that one is not justified in employing coercion to repel aggression against oneself is highly counterintuitive; I will not pause to consider it further. The absolutist may reply, however, by redefining “coercion” such that, if an act is morally justified, it is not truly coercive. Coercive acts, it may be argued, are inherently acts of unjustified use of force; thus, if there are good moral reasons for the use of force, there is no coercion. That would render clause (Ib) self-contradictory. This reply is made more persuasive by an inspection of the philosophical literature on coercion, which reveals a dispute as to whether coercion is a descriptive term, or whether it presupposes a moral background of rights and duties.25Now we need not seek to define “coercion”here (indeed, it is doubtful that there are necessary and sufficient conditions for the proper use of “coercion”), but an interpretation of “coercion” that renders claim (I) not simply wrong, but senseless, is surely objectionable. The view being considered would render oxymoronic the idea of morally justified coercion-as soon as an act is justified, it would not be coercive (or an act of force). Surely, even if it is possible to hold that claim (I) is false, it is not meaningless. Still, should someone wish to dig his heels in on this point, we can reformulate claim (I) to accommodate it:
Claim (I’) (a) It is prima facie morally wrong for Alf to perform act @-an act of coercion or force against Betty (b) With suficient justification what would otherwise be considered a @ act, is shown to be a P ‘ act, which is morally justified.
2.3
Building on claim (I),we also ought to embrace:
(II) One ought not to engage in coercion or force ifone does not have an adequate moral justificationfor it. [Or,following claim If,one ought not to engage in acts that would be coercive ifone does not have adequate moral justification for performing them].26 Claim (11) does not say that acts are wrong if one does not have justification for them, though that is consistent with claim (11). It could be that the wrongness of acts is purely an objective matter. For example, a utilitarian might hold
142
Gerald F. Gaus
that an act is wrong if it does not maximize utility, and whether it has the property of maximizing utility does not depend on one’s reasons for doing it. Thus, wrongness does not depend on justification or one’s reasons for performing the act. But insofar as morality is practical-insofar as it guides one’s practice-one ought to do what one has the best moral reasons for doing, and so one ought to refrain from coercion if one does not have a moral justification for it. Let us call this the supposition of the “practical nature of morality.” Morality is about what one is to do and one’s reasons for doing it. To be a fully rational moral agent is to be guided by moral reasons in one’s practice. Thus, although a utilitarian such as J. J. C. Smart does not think that the rightness or wrongness of actions is determined by one’s reasons, Smart does think that when deliberating, one ought to do that which one has the best justification for thinking will make mankind happiest.27 Claim (11) might seem inconsistent with an externalist conception of moral obligation according to which one ought to do what is right, and that says nothing about what one has reasons to do.28On further reflection, though, we can see that it is not inconsistent. Even allowing that one ought to do what is morally right (independently of one’s reasons for thinking it is right), it still follows that a moral agent can only act on her best reasons; thus, from the point of view of her practice, she morally ought to do what she has the best reasons to do. To deny this would divorce “moral ought” from practice. Suppose that we ought to do what is right, but we also believe that our reasons are very bad indicators of what this might be, so that it does not follow that an agent ought to do what she has best reason to think morally right. If so, we could not use morality as a guide to our practice, as we would have no epistemic access to its instructions. It would speak, but we would be deaf to it. This is the view of morality rejected by claim (11),the practical nature of morality. Claim (11), then, is inconsistent with an esoteric conception of morality such as that discussed by Sidgwick, according to which only the elite is guided by good moral reasons while hoi polloi are best guided by bad reasons that lead them to do the correct action.29In a pretty straightforward sense, such a morality denies that hoi polloi ought to act on the best justifications or their best reasons. Rather than being treated as rational agents, such an esoteric morality manipulates the masses to do what they ought to do, though they know not why they ought to do it. Claim (11) then expresses an antielitist view of the moral life. 2.4
Claims (I) and (11) presuppose an account of moral justification. What types of reasons provide moral justification? To have a moral justification for
“Liberal Neutrality: A Compelling and Radical Principle”
143
@-ing (a coercive act) cannot be simply to have reasons that are good reasons for you to @, based on your desires, ends, and so on. A moral justification for @-ing requires impartial reasons: a moral reason is not simply a reason from your point of view, but from some shared or impartial point of view. “[Bly the moral point of view, we mean a point of view which furnishes a court of arbitration for conflicts of interest. Hence it cannot (logically) be identical with the point of view of any particular person or group of persons.”30All plausible moral views, then, must accept:
(HI) To morally justifi coercion requires impartial reasons for the coercion. It seems, though, that this drives us back to an intractable problem in ethics: seeking to characterize impartiality. Is an impartial morality one that treats each as one, and no one more than One that would be accepted by a view from nowhere?32One that advances the good of everyone?33One that would be adopted by an impartial spectator? We need not, however, settle on any specific conception of impartiality here. Rather, I wish to advance a constraint on all plausible theories of impartial morality, which I shall call the “public reason principle”:
(IV)A reason R is a moral, impartial, reason justifiing @ only if all fully rational moral agents coerced by @-ing would acknowledge R, when presented with it, as a justification for @-ing. This public reason principle rejects as genuinely impartial a view according to which one person may have a good moral reason for @-ing but some other fully rational moral agents are simply incapable of seeing this reason as a justification. Now this may seem vacuous: if R is a justification-a good reasonto @, then it must be the case that, by definition, all fully rational people acknowledge it as a reason. If I believe that R is a moral reason justifying @-ing, and you deny it, then I will simply conclude that you are not fully rational, for you have failed to see a good reason for what it is. What better indication could I have that you are not fully rational than you fail to see what is a good reason? Even thus interpreted, claim (IV) is not vacuous. If one accepts claim (IV), and also claims that one has advanced a justificatory reason R that some others refuse to acknowledge as such, then one is committed to judging them to be less than fully rational. Thus, one could not claim that some reason that was available to the faithful, but not to other fully rational moral agents, is justificatory. However, we possess a notion of the rational that is not simply derivative of our understanding of what is the best reason. A rational person takes into account all the relevant available evidence, takes care when evaluating it, is not
144
Gerald F. Gaus
subject to various distortions of deliberation or action (e.g., she is not under the influence of drugs or compulsions), and so on. It is a demanding ideal, much more demanding than being simply a reasonable person (although it does not require omniscience; rational people do not know all there is to be known). Nevertheless, we can apply it even when we do not know what is the best reason. If a person displays the virtues of rational deliberation and action and none of the vices then, given our understanding of a rational agent, she qualifies as such. Given this, if Alf is to satisfy claim (IV), when he advances a purported justificatory reason that Betty, another moral agent, refuses to acknowledge, he is committed to the further claim that there must be some rational failure of Betty’s that leads to this lack of appreciation of R: a cognitive flaw, processing error, inattention to sound information, or whatever. I am not saying that Alf must be able to show that such a flaw exists; after all, access to this information may be hard to come by. But he is committed to there being some such flaw if his claim about R is correct. Now, the less plausible it is to attribute such a flaw to Betty-the more all the evidence suggests that she is an excellent cognitive and practically moral rational agent-the less faith Alf will have in his claim that R is an impartial justificatory reason. If Betty is a fully rational moral person, and if she does not recognize R as a justificatory reason, then Alf cannot claim that it is an impartial reason that meets the public reason requirement. Claim (IV), then, is by no means vacuous. Some may reject claim (IV) on the grounds that reason R is impartial just in case R is true or correct, and its truth or correctness does not imply that all fully rational moral agents would acknowledge R. Consider, for example, a realist view according to which R is an impartial reason just in case R is entailed by a moral truth 1; and it is not the case that fully rational moral agents would necessarily both acknowledge T and that T implies R. Impartiality, on this view, is not what all rational moral persons would accept, but what is impartially true, though this truth is not available to all fully rational moral persons. The crux of claim (IV) is that this view is not sufficiently impartial. It does not represent a shared or common moral point of view: the claim that R is a reason is itself partial to a certain perspective, which is not shared by other fully rational moral agents. Suppose that the proponent of R is challenged by someone doubting whether it is really an impartial reason. In response, the advocate of R would claim that it is based on truth T, but now we see that this claim is one that depends on his perspective, for he admits that other fully rational moral agents would not acknowledge Tas a truth. Such an esoteric conception of moral truth, and so impartiality, is tantamount to an assertion that one’s perspective yields privileged insight into the moral truth. Claim (IV) requires that one treat all fully rational moral agents as fellow participants in moral discourse and practice: it thus specifies a minimally antielitist concep-
“LiberalNeutrality: A Compelling and Radical Principle”
145
tion of moral impartiality. Impartial morality speaks to all rational moral agents, and constitutes a common evaluative perspective. I have identified the relevant public as those affected by 0: only those who suffer coercion need acknowledge R as a good reason justifying @. If rational moral agents who are not themselves coerced would reject R, but all those who are coerced by @-ing do accept R, then @-ing is not morally wrong. Some perhaps are drawn to a stronger version of claim (IV) according to which all rational moral agents must acknowledge R: this would yield an even stronger version of the neutrality principle. I shall focus here on the more limited understanding of the public, and so a more circumscribed neutrality principle. The argument for the radical nature of liberal neutrality, though, is not weakened if one insists on a broader definition of the relevant public. 2.5
Together, claims (I) to (IV) support: (V)Alfought not to coerce or force Betty unless Alfhas an impartial reason justifying the coercion, a reason that as a fully rational moral agent, Betty would accept as justifving the coercion.
Although claim (V) follows from the intuitively compelling claims (I) to (IV), some contemporary philosophers appear to deny it. As I understand him, Steven Wall does so when claiming that “[s]ometimes people will rightly conclude that they must make demands-and attempt to enforce them-on others even when these demands cannot be justified at all.”34If we take this as meaning that Alf may permissibly coerce Betty because Alf believes that he has good moral reasons to do so, although they are not recognized as justificatory reasons “at all” by other fully rational moral agents, we must reject Wall’s position. Alf‘s claim that his reason is impartial would then be ill-founded; by his own admission, other fully rational moral agents do not acknowledge it as a reason. It is not an impartial reason that is confirmed from the moral point of view (see further 53.4). 2.6
Claim (V) is about what one individual actor (Alf) can do to others. Now liberals-and many others join them-also insist: (VI) Claim (V) applies to governments and their agents (qua agents); in particular it applies to them when they occupy the role of Alf:
146
Gerald F. Gaus
Some have argued that the normal moral restraints that apply to individual actors do not apply to the state. Bernard Bosanquet famously insists that “the State, as such, certainly cannot be guilty of personal immorality, and it is hard to see how it can commit theft or murder in the sense in which these are moral offenses.”35Even Bosanquet, however, insists that, while the state as such could not be guilty of immorality, “if an agent, even under the order of his executive superior, commits a breach of morality, bonafide in order to do what he conceives to be a public end desired by the State, he and his superior are certainly blamable.”36 However, there seems no good reason to follow Bosanquet’s Hegelianism and its allowance that, while its agents may be held accountable for personal immorality, the state as such cannot. Insofar as the state is an actor, it is subject to the restrictions of claim (V). The distinction, however, is not of crucial importance, as the state can only act through its agents: if they ought not to coerce without good reasons, the state ought not. 2.7
In order for claim (V) to be relevant to most action by government and its agents, it must be the case that most of their actions are coercive. Hence:
(VII) Government actions are coercive. So by (V)and (VI),government and its agents ought not to act unless they have impartial justijicatory reasons. Again, some deny this. Taking up on some themes in Rousseau’s political philosophy, it might be argued that (a) laws that express the general will express what people really wish to do, and (b) one cannot be coerced into doing what one really wants to do. Even if we accept part (a), (b) does not follow, and certainly Rousseau didn’t accept (b). After all, it was Rousseau who insisted that one could be “forced to be free”; that coercion is consistent with freedom would not show that it was not really coercive.37 J. S. Mill suggests a more modest way to reject claim (VII). Mill distinguished authoritative from nonauthoritative interventions by government; while the former takes the form of a command backed by enforcement, the latter gives advice and information, or establishes an agency to deal with a problem while allowing others to compete. Although the government threatening drug users with prison sentences is indeed an act of coercion, drug education programs, on this view, would not be coercive. George Sher makes much of this point. In order to promote certain aims, governments might offer rewards, engage in economic policies that favor the aim, fund educational programs, and so If these policies are not coercive, then they do not fall under claim (I).
“Liberal Neutrality: A Compelling and Radical Principle”
147
What is sometimes called the “libertarian” response must be right here: each of the supposedly noncoercive measures is only possible because of a prior act of coercion, be it threats associated with the tax code, threats that back up banking regulations (relevant, say, to setting interest rates), and so on. The action “conduct an educational program” presupposes the action “raise via taxation the revenues to conduct the program.” Assuming the former is impossible without the latter, it is inappropriate to separately evaluate them. An act that depends on having certain resources or powers cannot be evaluated without consideration of the legitimacy of obtaining those resources or powers. Here at least is a case where the dictum that “He who wills the end must also will the means” is appropriate: to insist that there is nothing coercive about the end, when the only way to achieve the end is through a coercive means, is disingenuous. The object of evaluation should be the complex act (e.g., raise revenues through taxation and spend them on an educational program). That the complex should be the focus of evaluation is by no means simply a libertarian view: the American Civil Liberties Union sues public authorities that use tax money to advertise religion (say, by using public workers to erect signs saying “Jesus is Lord”).39The idea is that this is not a mere educational measure that does not impose burdens on some; it does impose burdens that must be born by dissenting citizens because of threats of punishment by the Internal Revenue Service. It can, though, plausibly be maintained that some state actions are less coercive than others. Just as a threat of a short prison sentence is less coercive than threat of a long one, and threat of a small fine is less coercive than threat of a moderate jail term, so too the coercion involved in an extra 1 percent marginal tax rate is typically less coercive than the threat of jail.40Thus, if we concern ourselves with the strength of the justifications required to legitimate the coercion, then a distinction between stronger and milder forms of coercion will be relevant. But that distinction is not relevant to claim (VII), which concerns the set of actions that require justification. 2.8
Again, though, many resist this. They insist that what the government does should be evaluated independently of the means required to do it. Although, as I have said, this seems quite wrong, to some extent we can put this dispute aside as there is another way to arrive at a principle akin to claim (VII). On the modern view of the state, it is not the private domain of some political actors, nor is the public weal identified with the private good of any person or section of the public. In contrast to private actors, public authorities cannot with legitimacy act on the basis of whim, personal aims, or sectarian
148
Gerald F. Gaus
advantage. Such action is corrupt. Public officials occupy a position of trustees acting in the interests of the citizens. Consequently, they are only authorized to act for good reasons: everything they do must be done for good reasons. Public officials require a warrant or authorization to act, and they are always subject to a demand that they justify their actions as falling under that a~thorization.~~ Moreover, given the modern democratic view of the state, its members are citizens, not subjects. An implication of this is that what constitutes a public reason must be one that fully rational citizens could recognize as such. In a Hobbesian state, each becomes a subject, and public reason is simply the reason of the sovereign.42In a democratic state, however, reasons that do not speak to the public at large are not truly public: if the state is the trustee of the public, then its justifications must be directed at what fully rational trustors understand as good reasons. If justifications are offered that are not good reasons to a fully rational person, then the government is not acting as a trustee in relation to him; it treats him as a subject to whom it is not responsible. This line of reasoning-which I will not develop further here-points to a principle along the lines of: (VII’) Governments ought never to act without justifjing reasons that all fully rational citizens would acknowledge as such. 2.9
Claims (V) to (VII/VII’) lead to the conclusion that governments ought not to act unless they possess a reason that would be acknowledged as justifying the action by every fully rational moral pers~n/citizen~~ coerced by the action. We are thus led to a “minimal principle of neutrality”:
(VIII) A government ought never to act without impartialjustification; R is an impartial justificatory reason for action @ only if it would be acknowledged as justificatory by every fully rational citizen coerced by the government’s @-ing. Let us say that a purported justificatory reason R is rationally rejectable if and only if a moral agent could reject it without thereby showing herself to be less than fully rational. It thus follows that minimal neutrality requires that a government and its officials are justified in @-ing,only if they possess reasons for it that are not rationally rejectable. Neutrality is often understood as prohibiting government actions that can be justified only by appeal to a controversial conception of the good. The min-
“Liberal Neutrality: A Compelling and Radical Principle”
149
imal principle of neutrality does not go that far, for it says nothing about whether reasons based on the good are rationally rejectable (see 53.4). But insofar as any reason is rationally rejectable, the minimal principle of neutrality deems that reason unacceptable as a justification of government action. Government thus must be neutral with respect to citizens’ reasons that are rationally rejectable by others.
3. Liberal Neutrality 3.1
Our very concept of morality and moral reasoning, then, leads us to the “minimal neutrality principle.” Although some nihilistic and elitist conceptions of morality reject it, they are implausible views. The minimal neutrality principle, however, allows government acts and policies that run afoul of “neutralist” liberalism as it is typically understood. T. H. Green held that each person’s “real self” sought self-perfection and development; thus, a policy that aimed at developing the perfection of each could be justified to all, since all fully rational citizens would act on their real selves, as opposed to their actual selves.44On the basis of this reasoning, Green and his followers advocated paternalistic policies, such as regulating alcohol sales for the good of the working class. More generally, contemporary perfectionists argue for government programs that either discourage people from seeking goods that they would not want if they appreciated and acted on good reasons (cigarettes and other drugs) or else endeavor to provide people with goods that they would want if they were enlightened or could properly appreciate good reasons (e.g., public broadcasting or art galleries). However, much more worrying perfectionist policies might conform to the minimal neutrality principle. Just as it can be held that drugs interfere with selfdevelopment and perfection, it can, with at least equal plausibility, be held that superstitious beliefs, such as those associated with organized religion, thwart people’s development, especially vulnerable children. Thus, a perfectionist might propose a policy that would tax religious observances or a regulation that would allow only adults into church services. More mildly, mandatory atheism classes might be included in all school curricula. Since fully rational citizens would not have such superstitious beliefs, they would tend to approve of such policies as ways to promote development, either of others or of themselves in their weaker moments. I must confess they strike me as excellent perfectionist policies; that contemporary perfectionists do not recommend them is something of a testament to the antirationalistic nature of our times.
150
Gerald E Gaus
3.2
Although I think these would be excellent perfectionist policies, they surely would be illiberal. Liberalism emerged as a distinct political theory as a call for freedom of speech and of thought. As John Plamenatz observes, freedom of thought “is an idea which emerges slowly in the West in the course of the sixteenth and seventeenth centuries; and yet today, in the eyes of the liberal, it is this liberty which is most precious of all.”45Right from the outset, the liberal case for freedom of thought has derived from devotion to human reason.46Although an important ground for freedom of thought is as a means to produce truth-“Let her [i.e., truth] and falsehood grapple: who ever knew truth put to worse, in a free and open enc~unter?”~’-a more fundamental ground of freedom of thought is respect for the reasoning powers of each individual. Each has a right to believe as his reason dictates. Now typically this is understood as simply an external moral right not to be interfered with in certain activities related to belief formation and expression. But one of the puzzling aspects of freedom of thought is that in an important way merely external restraints typically do stop one from freely thinking. Brainwashing does occur, but rarely. Much more commonly one is punished for communicating or expressing one’s thoughts; rarely does this actually interfere with a person’s thinking what he wishes. If in fact it is actually difficult to interfere with a person’s thinking, why do people claim freedom of thought rather than simply freedom to speak, to publish, to worship, and so on? As long as we see freedom of thought as purely a moral right involving external actions, it will seem a misnomer; and it will not respect people’s reasoning, as much as their actions that express their views, and their rights to gain access to the views of others. The fundamental place of freedom of thought becomes apparent, though, if we do not insist on a sharp distinction between moral and epistemic rights, what one is warranted in doing and what one is warranted in believing. To respect a person’s reason and his freedom of thought is to grant to him a right to think as he pleases in the sense not only of external moral rights, but a claim that his deliberations properly determine what he ought to believe. When someone says he has a right to his opinion, he is not merely claiming that he cannot legitimately be forced to abandon it; he is insisting that his opinions properly track his own deliberations. That is, freedom of thought supposes the right to believe that which one has deliberated on and has determined to be well founded. Or, as I shall say to stress its epistemic aspect, one has some warrant to believe that to which one’s deliberations lead. I shall call this “respect for the people’s deliberations”:
[IX) That Betty’s deliberations lead her to conclude p rather than not-p, is itselfa warrant for Betty believing thatp.
“Liberal Neutrality: A Compelling and Radical Principle”
151
Claim (IX) is inconsistent with a certain sort of epistemology-which I shall call the “manual view”-according to which the warrant of a belief is entirely a matter of impersonal standards. On this impersonal view, ultimately the only warrant for Betty believing p is that p conforms to the correct epistemic standards as they are laid out in the best manual of epistemology. Thus, if the best manual of epistemology says that the reasons for p, given belief system B, support a level of confidence in p of, say, 0.6, then Betty and anyone else with B is warranted in attributing to p a level of confidence no less and no greater than 0.6. But clearly the manual view, while useful when we are trying to think through our commitments and beliefs, is not the way we decide whether a person’s beliefs are warranted. Warrant is to a great extent processdependent. That you have deliberated and concluded p itself confers warrant on you believing that p. Even an advocate of the manual view can attribute to actual deliberations the role of indicator of warrant, though not an independent source of warrant. Given the complexity and diversity of people’s systems of reasons and values, it is very hard to know just what level of confidence in belief in p is justified. Thus, the results of thinking through the justification of p on the basis of standards that approach those in the manual of epistemology is evidence of what the manual recommends. So even a proponent of the manual view can take warrant as process-dependent, though only because the process has evidential value of the upshot of the standards set out in the manual of epistemology regarding p. This justification of the process-dependency of warrant, to which I have been much attracted, has much to recommend it, but I now think that it is not the whole story. It falls short in two ways. Suppose that I am in general uncertain about what is the best manual: I cannot then base my evaluation of the process on whether it approaches the best manual, should the recommendations of my preferred manuals diverge. In this case, I am uncertain what is the belief recommended by the manual view, yet the mere fact that I have thought things through increases my warrant for the belief I settle on. Those who are devoted to rival manuals can converge in approving of beliefs that result from deliberation, and not just because these beliefs are apt to conform to their preferred manual. The second shortcoming of the purely evidential justification of processdependent warrant displays an interesting intersection between ethics and epistemology. To be a rational believer and agent is not simply to be someone who has rationally justified beliefs; it is to be a producer of such beliefs and actions based on them. Now respect for this status of others as producers of beliefs requires that the fact that a deliberator has herself actually produced p rather than not-p gives her some epistemic-moral claim to believe that p. Consider the
152
Gerald E Gaus
denial of this right. Suppose that as a result of your deliberation that p was the thing to believe, you had no more warrant to believe p than not-p, if another person, equally acquainted with your belief system and equally competent as a reasoner, concluded that not-p is what your system commits you to. Surely, this would undermine your status as the rightful producer of your own beliefs. Your deliberations are not merely, as it were, one consideration in favor of your believing that p. They cause you to believe it.48One’s status as a consumer of beliefs cannot be abstracted from one’s status as a producer of beliefs: What one is warranted in consuming depends to a great extent on what one has produced. It does not follow from this that whatever beliefs, reasons, and so on one has produced one is warranted in believing. The manual view has this much to say for it: We do have access to some impersonal standards. We also have reason to think that some manuals are better than others: As I have argued elsewhere, manuals that allow for some variation in standards from one system to another are to be preferred to manuals that insist on all the same standards for all systems of reasons, beliefs, and values.49Still, sometimes we can be extremely confident that basic standards have been flouted by a believer, and so we can with justification say that she is not warranted in believing the upshot of her own deliberations. Claim (IX) is consistent with this use of the manual of epistemology to dispute a person’s warrant for her belief. 3.3
This analysis, though, would seem to have an unwelcomed implication: If freedom of conscience is grounded on claim (IX), and the process-dependent warrant for believing is not absolute-it can be overturned by manifest violations of epistemic standards-it would seem to follow that freedom of conscience is also not absolute. A person whose beliefs typically and manifestly violate basic epistemic norms is not warranted in those beliefs, and if freedom of thought is bound up with epistemic warrant at all, this would seem to imply that the justification for freedom of thought does not apply to this person. Three replies are in order. (i) I have not claimed that the entire case for freedom of thought rests on respect for the reasoning of others. As Mill shows in the second chapter of On Liberty, we can have strong grounds for insisting on the liberty to entertain and express patently absurd claims. Simply having false views to argue against may sharpen our own understanding. It is also surely true that granting the state authority to determine which views are sound and which are absurd would threaten the freedom of all. (ii) Insofar as our devotion to freedom of thought derives from respecting the reasoning and deliberations of others, those who are systematically irra-
“Liberal Neutrality: A Compelling and Radical Principle”
153
tional are indeed not the sorts of reasoners who merit respect. The insane, the deeply emotionally disturbed, the senile, and those in the grips of delusions really are harder cases for freedom of thought than are competent deliberators. So, too, I think, are those who adopt manifestly irrational views such as flat-earthers: if people typically used their freedom of thought to arrive at loony conclusions, freedom of thought could not be based on our respect for people’s reasoning. And we certainly would not hold freedom of thought in such high regard. The liberal tradition in politics is not irrationalist or even skeptical about the ability of reason to arrive at truth. Liberals have been impressed by the ways that our reason leads us to different conclusions, but this has not led them to depreciate reason, or like some postmodernists, to call the very idea of reason itself into question. These contemporary defenders of freedom of thought who insist that it is in no way grounded on the respect for our reasoning powers are departing from, not articulating, the liberal tradition. (iii) Those who insist that the case for freedom of thought has nothing to do with respecting the reasoning of others, and so applies as much to the insane or the manifestly absurd as to the reasoned conclusion, nevertheless ought to accept claim (IX) as a principle of warranted belief. One can reject my thesis that claim (1X)-respect for people’s deliberations-is a ground for freedom of thought while still embracing claim (IX) as an epistemological principle. (Likewise, those who reject [ 1x1 as an epistemological principle might nevertheless endorse a version of it on purely moral grounds.) 3.4
Claim (IX) makes it much more difficult to provide impartial justificatory reasons, as required by claim (VIII). If one important warrant for a person believing not-R is that not-R is the result of her own deliberations, then there is a real barrier to the state justifying its actions by showing that there exists some reason R that all fully rational moral agents will acknowledge. In short, if we add to the idea of a publicly justified morality (claim V) that applies to all the state’s actions (claim VIII) a further requirement that each person’s reasoning merits some respect (claim IX), we generate a neutrality principle according to which many reasons that might prima facie appear to justify state action are excluded. Now, in order to arrive at liberal neutrality as it is usually understood we must also add:
(X)(a) Reasons that presuppose values, claims about the good life, or about human perfection rarely if ever can justifi coercion by the state. (b) Reasons that presuppose basic rules of justice can justifi coercion by the state.
154
Gerald E Gaus
Claim (Xa) articulates the often-repeated claim that the state must be neutral between conceptions of the good. However, unlike many accounts of liberal neutrality, it is not incumbent on us to explicate the idea of a “conception of the good,” “comprehensive doctrine,” “notion of the good life,” and so on. These ideas are not foundational to my account (hence they come at the end). The important point is that reasons that rational moral agents or citizens can reject as justificatory are excluded; the questions of what precisely these reasons are is not basic to the defense of liberal neutrality. There is, though, powerful reason to think that few reasons based on values or goods will qualify. To see why, consider a philosophically well-grounded perfectionist theory, such as those proffered by Sher or R ~ Z . ~For O any such theory, there is a wellgrounded philosophical rejection, for example, a well-developed subjectivist account of value. Now suppose that a perfectionist state inspired by Sher or Raz offers a justification of its action by appeal to one of these perfectionist theories, call this R. A subjectivist citizen can point out that her account of value denies R. Her theory functions as a defeater doctrine: whether true or false, she is warranted in believing it, and so appeal to reasons such as R that are inconsistent with it fail to qualify as moral justifications that respect the reasoning of each. Whether or not perfectionism is the correct philosophical doctrine is beside the point. The question is whether perfectionism provides the basis for an impartial reason that respects the reasoning of each: a doctrine can be true and still fail to qualify. Perfectionists such as Wall, as I have said ($2.5), find this implausible. How can a moral agent not act on the best reasons as he sees them? How can morality itself dictate that a person not act on what he thinks are the best moral reasons? As Wall understands it, a theory that restrains citizens from relying on their best reasons in this way privileges civility (not imposing reasons on others) over content (the best reasons). While it might be plausible for civility to sometimes override content, surely, Wall points out, it is not plausible to say that it always overrides it.51This, though, wrongly characterizes the issue. If reasons are not impartial, they are not moral reasons. Morality requires taking up a point of view that addresses reasons to all. So, the fact that other reasoners fail to accept your reasons impugns the impartiality of those reasons; you are asserting the specialness of your reasoning over fully rational others on issues that involve their lives ($2.4). All this is implied by the minimal neutrality principle. Liberal neutrality adds an extra constraint: We must be extremely cautious in claiming that others are in error about what they ought to believe, for to a significant extent the warrant for a person believing something is process dependent (claim IX). Given this, the perfectionist himself ought to conclude that while he has a plausible theory of good moral reasons, the case for these reasons is too weak to make out the claim that they really are impar-
“Liberal Neutrality: A Compelling and Radical Principle”
155
tial reasons suited for justifying coercion. Thus, liberals must reject Wall’s claim that if you have “sound political views . ..the fact that some may reasonably reject” them does not itself show that you should not enforce them.52 Moral theory does not yield applied ethics in the way so many contemporary philosophers would have us think. According to what might be called the “textbook view of applied ethics,” one does moral theory in part one, deciding which is the best moral theory. Having done that, in part two one proceeds to apply it, and this not simply in the sense that one guides one’s own actions by it, but one employs it to force others to conform to it. Having decided that perfectionism was correct in part one (hence one knows the correct “content”), one proceeds to advocate perfectionist coercive policies in part two (even if this requires overriding “civility”). But part one had a number of competing theories, which have by no means been decisively defeated. Given this, one’s claim that one is giving impartial reasons, which all fully rational moral agents must acknowledge, is false despite one’s plausible claim that if everyone were as enlightened as you are they would embrace these reasons. Claim (Xa) does not insist that all reasons based on conceptions of the good fail to qualify as justificatory reasons in the requisite sense. I, indeed, think that some notions of autonomy, which are implied by the very idea of morality, provide such reasons. And Rawls may be correct that a list of basic goods can function as public reasons. The problem is not whether a reason concerns “the good,” but whether it is such that all rational citizens must acknowledge it as justificatory. 3.5
Now, it may well seem that the argument for (Xa) shows that (Xb) is false. If we exclude as genuinely justificatory those reasons that are based on value, because they would with warrant be rejected by some reasoners, then surely we cannot justify principles of justice, for they, too, can with warrant be rejected by some. Observe first that this seems like the right sort of challenge for a liberal theory. As Robert Nozick, James M. Buchanan, and John A. Simmons all realize, liberalism is at the edge of anarchy, in the sense that liberal principles, if carried to their extreme, appear to lead to anarchism.53If liberty is so fundamental, justifying any limitation on it may be a real challenge. A colleague of mine once suggested that an anarchist is a “200-proof liberal.” While I think this is wrong-an anarchist is not a pure liberal-it does indicate the direction in which liberalism is pulled. Because of this, I would venture, one indication of whether an account is genuinely liberal is how much of a challenge anarchism poses.54 If anarchism never seems a threat at all, it may well be
156
Gerald F. Gaus
because the account is not liberal at all. Today, this is perhaps forgotten: Theories that call themselves liberal are often modified forms of socialism or collectivism; carried to their extremes, these theories lead to statism. Can the challenge be met, or is liberal neutrality really anarchist neutrality? The case against anarchism is itself an important part of liberalism, and it certainly cannot be given, or indeed, really summarized here.55We can, though, point to two lines of argument that a defender of liberalism ought to employ in fending off the anarchist challenge. (i) First, it is important that impartial reasons have not been characterized as reasons that are in fact accepted by all. Although one’s warrant for believing is partially process dependent, it is not wholly so. A strong enough case may provide an impartial justification even in the face of actual disagreement. Thus, for example, those who reject all coercive states on the grounds that, in the absence of the state, people would all voluntarily do the right thing and contribute to the public welfare are making claims that fly in the face of everything we know about human psychology and collective action problems.56 They are unwarranted beliefs, and so cannot block a public justification. (ii) The basic liberal claim, then, is that the total absence of a coercive state is impartially demonstrably worse than a limited state that enforces personal rights and some system of property rights. This, we might say, is the master argument of the social contract tradition, and there is little point in repeating its familiar conclusions. What ought to be pointed out is that the argument does not depend on a decisive argument for a single best system. That is, a typical objection to a social contractualist account is that, although it shows that some limited states are better than anarchy, it cannot show that everyone has decisive reason to embrace the same regime. Thus, it is said, the argument is inconclusive. But a justification of a regime need not proceed by showing that there are direct decisive reasons to embrace this particular regime in all its particularity. The most plausible line of argument is to show (a) that there is a set of regimes that everyone has decisive reason to prefer to anarchy and (b) that there is some sort of procedure that can be justified (say, a democratic one) for selecting within the set57(see further 94.2).
4. Radical Neutrality
4.1
To paraphrase Nozick, so strong and far-reaching is liberal neutrality that it Alraises the question of what, if anything, the state and its officials may though, I have argued, the liberal tradition, and especially contractualism,
“Liberal Neutrality: A Compelling and Radical Principle”
157
shows that some state action can be justified, it is doubtful that much in the way of public policy survives the neutrality test. Daniel M. Weinstock apparently disagrees. He believes that “a public conception of the good” can be justified. “[Slubstantive values can constitute part of a conception of public reason when they either are or could be shared by different contending conceptions of the Weinstock argues that a value can be part of public reason when one of three conditions are met: (a) if some value V is shared by all conceptions of the good, it can be supported by the state; (b) if some value V is presupposed by all conceptions of the good, in the sense that all conceptions of the good require V for their realization, V can be appealed to; and (c) there may be empirical evidence that some actions are required to promote everyone’s values. “For example, non-coercive measures promoting physical exercise can ultimately be grounded in the empirically verifiable view that moderate amounts of exercise contribute causally both to the length and to the quality of life, as can the state’s mildly coercive policies aimed at making smoking less attractive.”60 Weinstock‘s three basic claims are correct. Insofar as all fully rational citizens have warrant for sharing a value, or warrant for believing that some good is necessary for all values, or for believing empirical propositions that have manifest implications for their values, appealing to such values and claims in justifying state action does not violate liberal neutrality. If the claims are not rationally controversial, they do not run afoul of liberal neutrality. However, Weinstock makes much too quick a jump from the claim that (a) value Vis embraced by every fully rational moral person/citizen to (b) Vprovides a reason R that justifies policy P that all fully rational moral persons/citizens would accept. The crucial problem is the rankings of values. R is not an impartial justification for P unless it would be accepted by all fully rational moral persons/citizens. Now although a rational Betty might reject R as a reason because it appeals to a value that she does not share, another reason that she will reject it is that it appeals to a ranking of values, or a trade-off of values, that she does not share. According to Milton Rokeach, a psychologist, Americans agree in affirming a set of thirty-six values; what they differ on is “the way they organize them to form value hierarchies or priorities.”61If so, our main disagreements about the good are not about what is of value, but the relative importance of values. After all, what is a ranking of values but a “conception of the good”? Liberal neutrality requires justification by impartial reasons, and reasons that presuppose a controversial value ranking do not qualify. Consequently, even if everyone agrees that smoking causes cancer, rational people clearly do disagree about whether the pleasures are worth the risk of death. Given that rational people weigh the relative values of pleasure
158
Gerald F. Gaus
and safety differently, coercive acts that can only be justified on the grounds that the pleasure does not outweigh the risk to health fail to provide a neutral case. Thus, although the badness of ill health caused by smoking can, as Weinstock says, be invoked in a neutral justification, that its badness outweighs the goodness of the pleasure of smoking cannot; and without that, no state policies discouraging smoking will be justified. This has direct relevance to the U.S. drug policy, which is based on certain middle-class value rankings, and which results in policies that place inordinate costs on the poor.62 This same problem applies to the goods that we all share. There is a long tradition in moral and political philosophy of arguing that, since we all agree on what is good, these are impartial considerations. According to Kurt Baier, for example: I take a person’s good, what is for his good, to be roughly things such as health, wealth, power, longevity, having friends, about which experience suggests that they are a good thing from anyone’s point of view. . . .We are quite confident in saying that everyone prefers being healthy to being sick, that everyone is better able to attain his objectives whatever they may be if he is healthy rather than sick and so on. Nevertheless, even in the case of health what makes it important, really worth caring about, being of value to one, is the central and almost indispensable role it plays in virtually everything one wants to do.63
Surely we all-or at least very nearly all-can agree to advance these values or goods. Again, though, to publicly justify a state action or policy it is not sufficient to show that we all accept these values or goods, but that there is a tradeoff rate or ranking that all fully rational citizens would accept. But one of the ongoing disputes in our society is, for example, the proper way to trade off our interest in making our own choices (the value of liberty) and the good of being healthy. Consider the case of mandatory motorcycle helmet laws. The point of helmet laws is to advance motorcyclists’ good by keeping them alive and healthy. From the perspective of that ingredient of their good, it seems to many uncontroversial that motorcyclists have reason to embrace such laws. But many motorcyclists prefer the freedom to choose-to run the risk of a serious accident and brain damage, to ride with the wind in the hair-and so resent these laws because they take away their freedom of choice. Now, many legislators appear to assert that there is a trade-off rate that all fully rational citizens would embrace: it is unreasonable to reject mandatory helmet laws because the health gains outweigh the losses to liberty. To many motorcyclists, however, this is simply the trade-off rate of middle-class, middle-aged people, who are risk-adverse. Is it not clear that it is the motorcyclists who are being unreasonable?
“Liberal Neutrality: A Compelling and Radical Principle”
159
4.2
Weinstock suggests another way to avoid the apparent radical implications of liberal neutrality: it can be applied at different levels.64Rawls, for example, appears to apply it only to constitutional matters; or it might be applied only to matters of basic justice. It seems that the more we restrict the scope of neutrality, the less radical the principle will be. Suppose, however, that we have a neutral justification of a constitution and are now advocating policy l? There are two possibilities. It might be that the constitution, ex hypothesi neutrally justified, authorizes l? In this case, P does have a neutral justification insofar as it is justified through the constitution that is neutrally justified. The issue here, though, is how extensive a constitution could be so neutrally justified. It seems doubtful indeed that, for example, a constitution that allows the government to use taxation to discourage smoking is capable of neutral public justification. Given what has been said in section 4.1, there is a strong presumption that all coerced citizens could not be given impartial reasons of the requisite sort for granting the state authority over that matter. The other alternative is that the constitution is neutrally justified, but only the constitution; there is no indirect neutral justification via the constitution for policy l? Policy Psimply is another matter. In this case, it looks as if Pviolates claim (I): it is a coercive act without justification. Neutrality cannot be restricted to a certain “level” since claim (I) is a fully general principle, applying to all coercive acts. Witness how questions as to the proper specification or the liberal neutrality, or debates about its proper application or scope, are avoided by the derivation of liberal neutrality from more fundamental moral principles. The favored interpretation is determined by the derivation: the scope of the principle is determined by its reliance on claim (I). 4.3
It should be noted that although liberal neutrality is a radical principle and severely limits what policies a state can pursue, it is not a libertarian principle as that is normally understood. In any political society, some set of laws defining property rights must be publicly justified. Property rights involve coercion: the holder of the right can force others not to use, transfer, or destroy that over which he has property rights. Thus, they stand in need of justification. Yet, without a distinction between mine and thine, peaceful and cooperative social life is not possible. Thus, some system of property rights is necessary. Now, the favored impartial justification of property is a matter of dispute. So-called left-libertarians such as Hillel SteineF5insist that egalitarian distributions are required to justify
160
Gerald E Gaus
individual property rights, while “traditional”libertarians insist that redistributions violate rights to property. This dispute, over what is required for a justified property regime, is about what is required for liberal neutrality. There is a conclusive case for property rights, but not a reasonable dispute about the impartially justified system. Doesn’t this leave liberal neutrality open to precisely the same objection as that leveled against Weinstock in section 4.1? I argued earlier that, while we all may agree that Vis a value, because we rationally disagree on the relative value of Vand, say, W no appeal to the superiority of Vover W can provide an impartial reason of the required sort. Similarly, one may say that while everyone agrees that some system of property rights is better than none, no argument for a specific system of property can function as an impartial reason of the requisite sort, as it will be rationally rejected either by a left or a right libertarian. The difference between the smoking case and the property rights case lies in where the option “no policy at all in this matter” is on each rational citizen’s rankings. Suppose that all rational citizens endorse or prefer systems of property {P,, P2, P3} over no system of property rights, but prefer no system of property rights over {P4,P,}. If so, there is a liberal neutralist justification for selecting from the set of {P,, Pz, P3};in addition, of course, we require a selection procedure that also can be so justified.66If we can do that, then we will have a fully justified system of property rights despite the disagreements between right and left libertarians on the relative merits of {PI,P2,P,}. The contrast to most public policy cases is that, over a very wide range of political issues, it would seem that for each and every proposal P in the set of options, a number of citizens rank P as inferior to “no law at all on this matter.” No law at all will be preferred, first, by those who prefer no law to any policy, and so rank P and all other policies behind no policy at all (e.g., classical liberals regarding pornography regulation). Second, P will also be ranked worse than no policy at all by those who prefer some other policy P“ to no policy at all, but prefer no policy at all to P7 Thus, on issues where some rational citizens fall into one of these two groups, no public justification of P can be advanced. Of course, if impartial justice requires P or a policy like P, then we move to a case much more like that of property rights. It is very likely, though, that once we take account of comparative judgments, liberal neutrality precludes most contemporary legislation; it certainly excludes most of what “liberal perfectionists” would have the state do. 5. Conclusion: In Defense of Liberal Radicalism
Liberal neutrality is a radical principle. Some think that radical principles are implausible just because they are radical. Referring to a radical understanding
“Liberal Neutrality: A Compelling and Radical Principle”
161
of liberal neutrality according to which the state must be neutral between “all matters relating to the human good,” Weinstock insists that this leads to “complete state inaction, something which no liberal thinker, not even libertarians like Robert Nozick, ever advocated.”68Even speed limits, Weinstock argues, would be prohibited, since they “place an obstacle in the way of those who value thrills over safety.”69Such a radical conception of neutrality, says Weinstock, “would be absurdly self-st~ltifying.”~~ Now, I have already argued that the social contract tradition provides an extended reply to the anarchist challenge, one that I believe is clearly successful. Impartial reasons can be provided to justify at least some state activities; so whether or not Weinstock‘s criticisms apply to the specific interpretation of neutrality he is considering, I believe that there are good grounds for maintaining that anarchism does not follow from our radical interpretation of liberal neutrality. In concluding, however, I wish to focus on the implied premise of Weinstock‘s argument: A principle that might show the state to be morally illegitimate is absurd, and so a conclusion in favor of anarchism functions as a reductio ad absurdum. Liberals, Weinstock insists, should not be “self-stultifying.” Weinstock‘s position presupposes what liberals must question: that state coercion is justified. If we embrace Weinstock‘s assumption, a conclusion that state coercion cannot be justified, it simply shows that our premises are wrong: Since we know that coercion by the state is justified, any principle that indicates that it is not is ips0 facto absurd. I see no compelling reason to accept what amounts to a presumption in favor of state force. Such a presumption, of course, is part and parcel of conservative thinking, but the liberal tradition in political philosophy takes the state, and all coercion, as requiring justification. To appropriate a contemporary if not pellucid term, the state is problematic. That is why the classic social contract theories begin with the state of nature and seek to justify the creation of a state. Weinstock‘s presumption reverses this: We start out with a presumption that a large range of state activities are justified and use this as benchmark for testing our principles. Some may justify this manifestly conservative approach to justification by appeal to reflective equilibrium; indeed, in contemporary political philosophy reflective equilibrium is almost an orthodoxy. Careful consideration of reflective equilibrium, I have argued elsewhere, shows it to be fraught with diffkulties, especially as a method of moral decision making and argument, as opposed to a standard of correct belief.71Even if we embrace reflective equilibrium, however, and so hold that acceptable moral and political principles must accommodate certain considered judgments, this hardly shows that the accepted contemporary understanding of the limits of state action qualifies as such a touchstone judgment. Supposing it is true that “daily politics is
162
Gerald E Gaus
irretrievably perfectionist,” this would by no means show that antiperfectionism is absurd or misguided.72 The liberal tradition denies that the justification of organized coercion is a settled conviction that can overturn fundamental moral principles. If compelling moral claims such as (I) to (X) imply that most state coercion is unjust, then the loser is state coercion, not these fundamental moral convictions. Liberal moral principles are indeed “self-stultifying’’ when what is being stultified is unjustified coercion of some by others. Morality stultifies a host of things that we may wish to do, including making others more perfect in o u r own eyes.
Notes 1. See Charles Larmore, Patterns of Moral Complexity (Cambridge: Cambridge University Press, 1987), 43ff; Will Kymlicka, “Liberal Individualism and Liberal Neutrality,” Ethics 99 (July 1989): 833-905; and Simon Caney, “Consequentialist Defences of Liberal Neutrality,” Philosophical Quarterly 41, no. 165 (January 1991): 457-77. 2. The most careful discussion is George Sher, Beyond Neutrality: Perfectionism and Politics (Cambridge: Cambridge University Press, 1997), chapter 2. 3. See Jeremy Waldron, “Legislation and Moral Neutrality,” in Liberal Rights: Collected Papers 1981-1 991 (Cambridge: Cambridge University Press, 1993), 149ff. 4. Peter Jones, “The Ideal of the Neutral State,” in Liberal Neutrality, ed. Robert E. Goodin and Andrew Reeve (London: Routledge, 1989), 9; and Govert den Hartogh, “The Limits of Liberal Neutrality,” Philosophica 56 (1995): 61. 5. Colin M. MacLeod, “Liberal Neutrality or Liberal Tolerance?”Law and Philosophy 16 (September 1997): 532; and Paul Rosenberg, “Liberal Neutralism and the Social-Democratic Project,” Critical Review 8 (1994): 2 18. 6. Wojciech Sadurski, “Theory of Punishment, Social Justice and Liberal Neutrality,” Law and Philosophy 7 (1988-1989): 371. 7. See Jones, “Ideal of the Neutral State,” 9. 8. See Philippe van Parijs, Real Freedom for All (Oxford Oxford University Press, 1995), 28. 9. Bruce Ackerman, SocialJustice in the Liberal State (New Haven, Conn.: Yale University Press, 1980), 11. 10. Jones, “Ideal of the Neutral State,” 9. 11. See John Rawls, Justice As Fairness: A Restatement, ed. Erwin Kelly (Cambridge, Mass.: Belknap, 2001), 152. 12. See Brian Barry, Justice As Impartiality (Oxford Clarendon, 1995), 139-45; Ronald Dworkin, “Liberalism,” in A Matter of Principle (Cambridge, Mass.: Harvard University Press, 1985), 191; and Roger Paden, “Democracy and Liberal Neutrality,” Contemporary Philosophy 14, no. 1: 17-20. 13. Kymlicka, “Liberal Individualism and Liberal Neutrality,” 886.
“LiberalNeutrality: A Compelling and Radical Principle”
163
14. J. Donald Moon, Constructing Community: Moral Pluralism and Tragic Conflicts (Princeton, N.J.: Princeton University Press, 1993), 55. 15. Larmore, Patterns of Moral Complexity, 53ff. 16. See Barry, Justice As Impartiality, 125E and Robert E. Goodin and Andrew Reeve, “Liberalism and Neutrality,” in Liberal Neutrality, ed. Robert E. Goodin and Andrew Reeve (London: Routledge, 1989), 1-8. 17. Joel Feinberg, Harm to Others (New York Oxford University Press, 1984), 9. 18. Gerald F. Gaus, Justificatory Liberalism (Oxford Oxford University Press, 1996), 164ff;and Gerald F. Gaus, Social Philosophy (Armonk, N.Y.: Sharpe, 1999), 117ff. 19. In order to simplify, I will henceforth refer to “coercion” rather than the more cumbersome “force and coercion.” 20. On psychopathy see Gerald F. Gaus, Value and Justification (Cambridge: Cambridge University Press, 1990), 292-330. Compare John Rawls, Political Liberalism (New York Columbia University Press, 1996), 51. 21. Thomas Hobbes, The Elements of Law, Natural and Politic, ed. Ferdinand Tonnies (London: Cass, 1969), 71. 22. According to Wesley Hohfeld, Alf has a liberty to 0 if and only if Betty has no claim against Alf that he not a.But that Betty has no claim that Alf refrain from 0ing does not imply that she has a duty to let him 0; thus, just as he is at liberty to 0, she is at liberty to anti-0. For Hohfeld, when we talk about a person having a right to do something, we sometimes mean that she is merely at liberty to do so; she has no duty to refrain. But merely to have a liberty to do something does not imply that you have a claim that others not interfere. The classic example is the liberty of two pedestrians to pick up a dollar bill lying on the sidewalk. Neither has a duty to refrain from picking it up, but neither has a claim on the other to stand aside and let the other pick it up. Such “naked liberties” often characterize competitions; people have the liberty to win, but no one has a claim to win. For Hohfeld’s classic analysis, see his “Some Fundamental Legal Conceptions As Applied in Judicial Reasoning,” Yale Law Review 23 (1913): 16-59; see also H. L. A. Hart, “Legal Rights,” in Essays on Bentham: Studies in Jurisprudence and Political Theory (Oxford Clarendon, 1982), 171-72. 23. I have done so in Gaus, Value and Justification. 24. W. D. Ross, The Right and the Good (Oxford: Clarendon, 1930), chapter 2. 25. See Alan Wertheimer, Coercion (Princeton, N.J.: Princeton University Press, 1987). 26. I shall not henceforth note the necessary changes if one accepts claim (1’) rather than (I). 27. J. J. C. Smart, “An Outline of a System of Utilitarian Ethics,” inUtilitarianism: For and Against, by J. J. C. Smart and Bernard Williams (Cambridge: Cambridge University Press, 1973), 30. 28. The classic paper on this issue is, of course, William K. Frankena, “Obligation and Motivation in Recent Moral Philosophy,” in Perspectives on Morality: Essays by William K. Frankena, ed. K. E. Goodpaster (Notre Dame, Ind.: University of Notre Dame Press, 1976), 49-73; see also Philippa Foot, Virtues and Vices (Berkeley: University of California Press, 1978), 157-73. 29. Henry Sidgwick, The Methods of Ethics, 7th ed. (Chicago: University of Chicago Press, 1962), 489ff.
164
Gerald F. Gaus
30. Kurt Baier, The Moral Point of View (New York Random House, 1965), 96, emphasis in the original. 31. See R. M. Hare, Freedom and Reason (Oxford: Clarendon, 1963), 118. 32. This, of course, refers to Thomas Nagel, The View from Nowhere (New York Oxford University Press, 1986), chapter 9. 33. See Baier, Moral Point of View, 106. 34. Steven Wall, Liberalism, Perfectionism and Restraint (Cambridge: Cambridge University Press, 1998), 118. This same view is suggested by Moon, Constructing Community, 63. 35. Bernard Bosanquet, “The Philosophical Theory of the State,”in The Philosophical Theory of the State and Related Essays, ed. Gerald F. Gaus and William Sweet (Indianapolis: St. Augustine, 2001), 285. 36. Ibid., 284. 37. Jean Jacques Rousseau, The Social Contract, trans. Maurice Cranston (Harmondsworth, UK: Penguin, 1968), book 1, chapter 7. 38. Sher, Beyond Neutrality, 34-37. 39. Times-Picayune (New Orleans), 24 February 2002, 1. 40. See Daniel M. Weinstock, “Neutralizing Perfection: Hurka on Liberal Neutrality,” Dialogue 38 (1999): 45-62. 41. See S. I. Benn and G. F. Gaus, “The Public and Private: Concepts and Action,” in Public and Private in Social Lqe, ed. S. I. Benn and G. F. Gaus (New York St. Martin’s, 1983), 3-27. 42. See David Gauthier, “Public Reason,” Social Philosophy and Policy 12 (winter 1995): 19-20, reprinted in Fred D’Agostino and Gerald F. Gaus, eds., Public Reason (Aldershot, UK Ashgate, 1998),43-66. 43. Claim (VII) points to the class of moral persons and claim (VII’)to the class of citizens. I will employ these specifications of the class interchangeably, though, of course, they are not equivalent. 44. See T. H. Green, “On the Different Senses of ‘Freedom’As Applied to the Will and the Moral Progress of Man,” in Green’s Lectures on the Principles of Political Obligation, ed. Paul Harris and John Morrow (Cambridge: Cambridge University Press, 1986). 45. John Plamenatz, Man and Society, vol. 1 (London: Longman’s, 1973), 46. 46. See Gerald F. Gaus, Contemporary Theories of Liberalism: Public Reason As a Post-Enlightenment Project (London: Sage, forthcoming), chapter 1. 47. John Milton quoted in D. J. Manning, Liberalism (London: Dent, 1976), 46. Compare John Stuart Mill, On Liberty, reprinted in John Gray, ed., On Liberty and Other Essays (New York Oxford University Press, 1991), chapter 2. 48. See Gaus, Justificatory Liberalism, chapter 2. 49. Ibid., chapters 3-4. 50. See Sher, Beyond Neutrality; and Joseph Raz, The Morality of Freedom (Oxford Oxford University Press, 1986). 5 1. Wall, Liberalism, Perfectionism, and Restraint, 8Off. 52. Ibid., 101. 53. See Robert Nozick, Anarchy, State and Utopia (New York Basic Books, 1974); James M. Buchanan, The Limits of Liberty: Between Anarchy and Leviathan (Chicago:
“Liberal Neutrality: A Compelling and Radical Principle”
165
University of Chicago Press, 1975);and John A. Simmons, On the Edge ofAnarchy: Locke, Consent and the Limits of Society (Princeton, N.J.: Princeton University Press, 1993). 54. See Eric Mack and Gerald F. Gaus, “Classical Liberalism and Libertarianism,” in Handbook of Political Theory, ed. Gerald F. Gaus and Chandran Kukathas (London: Sage, forthcoming). 55. A case is provided in Gaus, Justificatory Liberalism, 180ff. 56. See Michael Bakunin, “Statism and Anarchy,” in Bakunin on Anarchy, ed. Sam Dolgoff (New York Vintage, 1871), 323-51. 57. See Gaus, Justificatory Liberalism, part 111. 58. Nozick, Anarchy, State and Utopia, ix. 59. Weinstock, “Neutralizing Perfection,” 55. 60. Ibid. 61. See Milton Rokeach, The Nature of Human Values (New York Free Press, 1973), 110; and Milton Rokeach, “From Individual to Institutional Values,” in Understanding Values (London: Collier Macmillan, 1979), 208. 62. See J. Donald Moon, “Drugs and Democracy,” in Drugs and the Limits of Liberalism, ed. Pablo De Greiff (Ithaca, N.Y.: Cornell University Press, 1999), 133-55. 63. Kurt Baier, “Comments,” in Reason, Ethics, and Society: Themesfrom Kurt Baier, with His Responses, ed. J. B. Schneewind (Chicago: Open Court, 1996), 216. 64. Weinstock, “Neutralizing Perfection,” 54. 65. Hillel Steiner, An Essay on Rights (Cambridge, Mass.: Blackwell, 1994). 66. I argue that constitutional democracy is such a procedure in Gaus, Justificatory Liberalism, part 111. 67. I explore this problem in much more depth in Gerald F. Gaus, “The Legal Coordination Game,” American Philosophical Association’s Newsletter on Philosophy and Law (forthcoming). 68. Weinstock, “Neutralizing Perfection,” 47. 69. Ibid. Again, we see how confusions result from separating the justification of raising revenues from the justification of expenditures (52.7). Weinstock supposes that there simply are roads out there, and then there is an entirely distinct issue of whether there will be speed limits. Surely, we have to ask whether the provision of roads without speed limits can be justified to all taxpayers. I assume it is manifest that the thrill seekers would not be able to get others to pay for their public racetracks, and probably would be unable to pay for them themselves. No doubt, the others could well afford to pay for a system of roads without including thrill seekers, should the thrill seekers really prefer not using roads to paying for roads with speed limits. Of course, private roads would solve this difficulty. Witness again the conservative horizons of the analysis. 70. Ibid. 71. See Gaus, Justifcatory Liberalism, 101-9. 72. den Hartogh, “Limits of Liberal Neutrality,” 59.
This Page Intentionally Left Blank
11 “Reasonable Rejection and Neutrality of Justification” George Klosko
Klosko presents and defends a moderate principle of state neutrality, one that seeks to overcome the main objections that have been pressed against stronger accounts of the principle. He views neutrality as a constraint on government actions justified on grounds that rest on particular conceptions of the good. So formulated, neutrality is able to address the common criticisms that it rests impermissibly on value skepticism and that it bars governments from performing many actions that are widely expected of them. If conceptions of the good are distinguished from other, nonsectarian value claims, it can be seen that neutrality bars actions based on the former, though not the latter. Klosko argues that appeal to nonsectarian values can justify many of government’s traditional functions, while the principle of state neutrality bars government actions that rest on values or ideals that can be reasonably rejected by some citizens.
- 167-
George Klorko
168
A
LTHOUGH NEUTRALITYIS WIDELY RECOGNIZED as a central liberal principle, it
s also widely criticized. In this paper, I attempt to develop the core of a defensible neutrality claim, employing the notion of arguments that cannot be “reasonably rejected.”’ In the literature, neutrality is generally filled in as between different peoples’ conceptions of the good. Exactly what we mean by “conceptions of the good” will be discussed later on. For now, we can say that these are overall views of what is of value in life-whether rooted in what we would call moral, religious, philosophical, or other beliefs. So construed, conceptions of the good are analogous to John Rawls’s “comprehensive views.”2 Neutrality is also filled in as “neutrality of justification,” as opposed to neutrality of consequence^.^ The governing idea is that people should not be subject to state coercion unless there are good reasons for it, reasons that should be convincing to them. This paper’s central claim is that neutrality of justification should be filled in as: governments should not base their actions on grounds that can be reasonably rejected. The particular force of neutrality centers on two kinds of nonneutral reasons: (1) certain kinds of value claims and (2) forms of reasoning that objectionably rest on particular comprehensive views. Arguments can be reasonably rejected for other reasons as well, for example, if their logic or factual basis is weak. The particular focus of neutrality is rejection of arguments because of their dependence on comprehensive views. A given value claim or mode of reasoning objectionably rests on a specific comprehensive view if it cannot be supported without reference to the latter, as will be explained. These contentions will be examined in succeeding sections, while I will argue from this perspective to defuse two important criticisms of neutrality. These concern the claim that neutrality collapses into objectionable value skepticism and that, devoid of substantive values, neutrality rules out necessary state policies. In addressing the second objection, I will sort out value claims that neutrality does and does not allow, and so argue that acceptably neutral justifications can be based on specific values. What is more, I will argue that neutral argument is in its essence argument rooted in such values. Showing that neutrality as construed in the following sections is able to deflect these two lines of criticism should go a long way toward establishing its defensibility. Before proceeding, I should note that my purposes are limited in two ways. First, although the objections I address are important, there are other problems with neutrality with which I do not deal. Although I believe they, too, can be overcome, showing this cannot be undertaken here.4 The second way is the nature of the principle I will present. To some extent, I believe, the implications of neutrality of justification have not been fully worked out because of difficulties in drawing clear borderlines between reasons neutrality does and
“Reasonable Rejection and Neutrality of Justification”
169
does not allow. Although I will present guidelines, they will be rough, with the details to be filled in according to the circumstances of different cases. This will oftentimes be a messy, controversial process. But this is not to say that such borderlines do not exist. Although there will be close or controversial cases, others will be clear cut, falling obviously on one or the other side of the dividing line. The force of neutrality is clearly seen in such cases. Although the version of neutrality formulated in this paper might appear to be less robust than what is widely seen in the literature, I believe it is in extension more or less equivalent to common versions of the principle and fulfills the moral role generally assigned to them. I
Defense of neutrality of justification requires clarification of its sense and focus. There is no inherent essence of neutrality. The way we work out its details depends on our understanding of its role in determining social policies and the values we wish to promote. I believe that the most defensible form is an expression of a particular value commitment, part of what it means to treat people with adequate r e ~ p e c t .Like ~ the basic principle that government should be neutral between religious groups (and between religion and nonreligion), neutrality between conceptions of the good is necessary to avoid particular forms of oppression. Just as it is wrong for the religious views of one group in a diverse society to be coercively imposed on nonmembers, so it is wrong for a specific comprehensive view to be similarly imposed. Not interfering with religious autonomy is central to treating people with respect and is of course a basic liberal value. Neutrality between conceptions of the good should be understood along similar lines.6 For example, assume that adherents of a specific religion believe that sexual activities between members of the same sex are strictly forbidden in the Bible and so are a violation of God’s law. Let us also assume that, under the influence of this sect and its teaching, government passes a law criminalizing such activities. If Jones, who is not an adherent of the sect, does not accept its interpretation of God’s law and objects to the law, there is little chance that he will be satisfied with the considerations advanced in its defense. For government to impose such a policy on Jones and like-minded others on the basis of such grounds would be an act of oppression, treating Jones without due respect. Contrast this case with a government policy requiring fluoridation of water. Although Green objects to this, proponents of the policy have strong arguments in its favor. They appeal to a kind of reasoning-scientific-that she
170
George Klosko
cannot reasonably reject (although I do not rule out the possibility of scientific evidence against fluoridation). It is possible that Green could object to considerations of science; she might prefer to make public policies on the basis of biblical authority or other similar sources. But a counter of this sort would not constitute a defensible rejection of the arguments for fl~oridation.~ To substantiate these claims and to fill in the idea of “reasonable rejection,” we require standards of proper argument. Once again, how we construe these depends on the goals and values we wish public policies to realize. Neutrality is bound up with preference for modes of argument that are publicly accessible and reliable. Pride of place is therefore given to empirical science, common modes of logical inference, and other forms of argument that are demonstrable and replicable.8 In a diverse society, inhabitants, large numbers of them, could well prefer other grounds of justification, notably religious authority and interpretations of sacred texts.9 Given such circumstances, neutrality will favor modes of argument that are publicly accessible, with the full realization that these are not “neutral” between competing groups. Neutrality is an inescapably liberal idea, in keeping with the movement against arguments from authority and intellectual obscurantism that has characterized liberal thought since its inception.’O In addition to epistemic standards, the degree of popular adherence a given argument form enjoys can be relevant. There is no clear formula here, but all things being equal, neutrality favors forms of reasoning that are widely subscribed to over those that are not. In many cases, epistemic and popular considerations will run in tandem. For example, a public policy directed at sterilizing drinking water will be supported by scientific reasoning, which is also presumably accepted by a large majority of the population. At the other extreme, combatting a cholera epidemic through prayer might be supported by a specific interpretation of the Bible, with such reasoning subscribed to by a small percentage of the population. In cases of these two kinds, there is little doubt what neutrality requires. In other cases, however, matters are less clear cut. If the epistemic and popular standards conflict, the epistemic will ordinarily take precedence. This is seen in the currently salient issue of high school biology curricula. Proponents of science want evolution to be taught, while adherents of different religions want creationism. What neutrality requires is clear. To allow creationism would amount to allowing particular religious groups to impose their own views on nonmembers.” This would be no less true if, as is the case, proponents of creationism were a substantial majority of the population.12 The fact that a given religion encompasses a large majority of society does not license imposing it on nonmembers. As the examples illustrate, what constitutes a defensible argument under neutrality must be understood in relation to a set of background beliefs and
“ReasonableRejection and Neutrality of Justification”
171
assumptions. l3 Especially important assumptions include the superiority of some forms of reasoning to others, even when substantial portions of the population believe otherwise. Additional important assumptions include the great worth of values such as freedom, rights, equality, and human dignity. In liberal societies, these are generally subscribed to although the details of specific values must be worked out in particular cases. Neutrality is of a piece with such values, justified by them, and also contributing a measure of justificatory force to them. With suitable epistemological and axiological background assumptions in place, reasons that contravene them can be reasonably rejected. A main reason why the idea of neutrality is so controversial and the concept has been so difficult to formulate clearly is the need to specify the background assumptions. To a large extent, as it seems to me, debate about neutrality has been about these, while it seems clear that they are linked to specific sociocultural conditions and must change along with changes in the latter. In section 111, I will briefly discuss problems in specifying particular value assumptions. In spite of such difficulties, I should distinguish problems in formulating an acceptable neutrality principle from problems in applying it. Once again, although precise borderlines are controversial and difficult to draw-as the U.S. Supreme Court is constantly at pains identifymg the borderline between church and state-at any given time a relatively uncontroversial core of the concept should be possible to identify. A final point is that neutrality should be interpreted along the lines of “wide” public reason, rather than a more exclusive form.I4It is notable that in debates on important public issues speakers regularly appeal to religious and other comprehensive views and are not ordinarily condemned for doing so. Prominent examples are debates over the abolition of slavery and the struggle for civil rights. Neutrality does not require that such considerations not be voiced, but only that public policies not be based on them. In a given policy debate, neutrality allows Smith to appeal to nonneutral reasons, as she pleases, as long as she also presents acceptably neutral ones that are adequate to support her position.15Neutrality of justification is just that. It is the requirement that public policies be justified on neutral grounds. Its primary intention is to constrain the grounds on which policies are made, not the kinds of considerations to which policy makers give voice in debates or discussions of public issues. This account of the neutrality principle is obviously only a sketch. Filling it in would involve numerous complex issues, while it is not clear how much more can be said in the abstract, in remove from specific circumstances. Policy decisions will be especially troublesome when science and common sense cannot resolve contested points and competing justificatory grounds
172
George Klosko
are similarly plausible (or implausible) and necessarily include controversial philosophical premises. For instance, thorny issues are encountered in questions that involve the status of animals and proper attitudes toward the natural world.“j Some guidance for these cases will be provided. However, troublesome borderline issues do not lessen the importance of neutrality in many other, clearer cases.
I1 As I have noted, the significance of neutrality stems from the pluralism of contemporary societies and accompanying disagreements over conceptions of the good. Obviously, if all inhabitants agreed on a single conception, there would be no need for the state to be neutral between them.” Since neutrality operates only in contexts in which no single conception of the good can be established, and the most direct way to argue that no conception is defensible is from a commitment to value skepticism, neutrality’s proponents have been criticized for being skeptics. However, value skepticism is a controversial epistemological position that can be reasonably rejected. Embracing it would transform neutrality into a sectarian position, dependent on what is in effect a controversial conception of the good.ls Thus, the challenge is to explain the continuing importance of neutrality without falling into skepticism. Consider the possibility that a specific conception of the good, C, could be established, so that it could not be reasonably rejected. Under those circumstances, all competing conceptions of the good could be reasonably rejected, and so the state would be required to act on C, making considerations of neutrality irrelevant. Thus, it could be argued that, unless we can invoke value skepticism to rule out this possibility, neutrality will be indefensible. To examine this line of argument, we will look briefly at sophisticated accounts of the relationship between neutrality and skepticism in the works of Brian Barry and John Rawls. l9 Barry’s defense of neutrality depends on a skeptical premise. Central to his theory of “justice as impartiality” is the contractarian requirement that principles of justice be agreed on by citizens who wish to reach agreement with one another, on terms that cannot be reasonably rejected. (11, 10) But even if no conception of the good can be demonstrated, the possibility remains that Jones could still view his conception as established and believe it should be imposed on other people, rather than conducting their affairs on principles mutually agreed to. To get around this possibility, Barry argues that “no conception of the good can justifiably be held with a degree of certainty that warrants its imposition on those who reject it” (169). He refers to this contention
“ReasonableRejection and Neutrality of Justification”
173
as “skepticism,” though “uncertainty” is probably a more accurate term, since Barry does not claim that all moral values are unable to be substantiated, which is what is generally meant by “skepticism.”20Barry defends his skeptical premise on empirical grounds, the long history of disagreements over the good, which show no sign of coming to an end: “It is hard not to be impressed by the fact that so many people have devoted so much effort over so many centuries to a matter of the greatest moment with so little success in the way of securing rational conviction among those not initially predisposed in favor of their conclusions” (JI, 171). Although Barry is correct that combining the requirement of agreement and skepticism would establish the need for neutrality, there are problems with his skeptical premise.21To begin with, it is not clear that the historical evidence establishes the impossibility of substantiating conceptions of the good. The fact that people have subscribed to different conceptions is compatible with moral pluralism, that there are multiple conceptions of the good, which are not necessarily incapable of being substantiated. More important, it is unsettling that Barry’s skepticism applies to some areas but not others. Although no conception of the good can be held with the requisite certainty, the same does not hold in regard to opinions concerning justice, notably “justice as impartiality” and the set of assumptions on which it rests. Against Barry, one could argue that, if a long history of disagreements demonstrates the impossibility of substantiating conceptions of the good, the similar history of disagreements about conceptions of justice should have similar implications in regard to them-including Barry’s contractarian theory.22Thus, although Barry’s commitment to skepticism is necessary for his defense of neutrality, successful defense of his theory of justice as impartiality requires suitable replies to a set of troubling questions. Another means through which neutrality can be justified is presented by Rawls in Political Liberalism. Rejecting skepticism as a controversial moral view (PL, 152), Rawls argues according to what he calls the “burdens of judgment” (54-58). These are aspects of human reason that lead to reasonable disagreements and so to the pluralism of contemporary societies. Part of being reasonable, according to Rawls, is to recognize the burdens and so the inescapability of moral disagreement. Because of the burdens, general agreement on comprehensive moral views can be secured only through the oppressive use of state power (37). Because of the intractability of moral disagreement, Rawls argues that his preferred principles of justice should be in effect neutral. To use his term, they should be “freestanding,” not based on specific comprehensive views (10,39-40). Although I find Rawls’s account of the burden of judgment plausible, there are problems with its ability to ground a view of neutrality. Most significant is
174
George Klosko
that it does not necessarily lead to n e ~ t r a l i t yIf . ~a~ given conception of the good could be substantiated, the burdens of judgment would provide no reason for public policy to be neutral in regard to it. In order for Rawls to rule out this possibility, he would have to introduce a skeptical premise, along the lines of Barry’s. But as we have seen, invocation of skepticism raises problems of its own. I believe that a neutrality principle can be defended more successfully on grounds similar to but slightly different from Barry’s. The problems with Barry’s skepticism can be circumvented through appeal to what we can call “situational irresol~ability.”~~ The main idea here is that questions of neutrality arise in definite policy situations. Since neutrality presupposes disagreement on conceptions of the good, a neutrality principle will be in effect only when all competing conceptions of the good appealed to in a given policy context can be reasonably rejected.
Situational irresolvability: In regard to a given issue, I, no conception of the good underlying a proposed policy cannot be reasonably rejected. Defense of situational irresolvability is, as in Barry’s account, based primarily on experience. In many policy areas, proponents espouse competing conceptions of the good that they are unable to establish and are able to raise telling objections against those of others. In such circumstances, the appearance of intractable disagreement should be taken at face value, although without drawing wider epistemological implications. The main force of situational irresolvability is to place the burden of justification. If Jones, disputing the appearance of intractable disagreement, argues that a given conception of the good can be established, then it is up to him to establish this.25Situational irresolvability is in effect until a satisfactory argument that cannot be reasonably rejected is produced. When situational irresolvability is in effect, disagreements over conceptions of the good cannot be overcome, and so government should be neutral between them. This gives us a weaker conception of neutrality than that defended by Barry, as it does not rule out the possibility that some conception of the good might one day be established. This possibility could be ruled out through the introduction of a skeptical premise, but this is of course undesirable in light of what we have seen. However, in the range of disagreements to which neutrality is generally applied, the practical implications of situational irresolvability will differ little from those stemming from Barry’s epistemological skepticism. If Jones believes that his conception of the good cannot be reasonably rejected, he must meet the heavy burden of justification posited by situational irresolvability and demonstrate
“Reasonable Rejection and Neutrality of Justification”
175
this.26 It is likely that in many of the policy contexts in which neutrality principles are ordinarily defended, value disagreements will resist solutions, and neutrality will obtain. Thus, as noted earlier, neutrality on this formulation will be roughly extensionally equivalent to neutrality based on skepticism, but far more defen~ible.’~
I11 In this section, I move from epistemological dimensions of neutrality to criticisms based on its value claims. It can be argued that neutrality disqualifies too much. Because of the nature of value claims, it is not clear that any can be established so as not to be subject to reasonable rejection. The implication is that in many public policy situations, government will be unable to establish a basis on which to act. I will examine two versions of this objection, pertaining to value claims and other public policy considerations, the former in this section and the latter in the next. In regard to the contention that value claims can generally be reasonably rejected, one possible result is a reductio ad absurdum, along the following lines: (a) Neutrality restricts value claims that can be reasonably rejected in policy deliberations (b) Political decision makers ordinarily do and must appeal to claims concerning different goods in order to address numerous issues28 (c) Because the restrictions in (a) are sharply at odds with the conduct of decision makers and would make it impossible for them to address these issues, neutrality must be rejected An argument along these lines is presented by Richard Kraut, according to whom neutrality “provides too weak a foundation for the kind of free and diverse society we would all like to live in.”29As Kraut construes neutrality, it rules out the kinds of moral judgments that are necessary to support public school classes in such subjects as drama, music, literature, and history ( 3 2 2 ) . Similarly, the state cannot support measures to help its citizens develop their autonomy and independence, nor can it support democratic institutions, which are “best interpreted as a way of supporting a substantive conception of the g o o d (324-28). In order to counter such an argument, we must look more carefully at the value claims that can be reasonably rejected. The question is whether we can identify a suitable class of value claims that are not subject to reasonable rejection.
176
George Klosko
I believe we are able to identify an appropriate class and so we can avoid the reductio. The point to note is that, as generally presented (including in the discussion earlier), neutrality rules out appeal to conceptions ofthe good. And so we must attempt to distinguish between conceptions of the good and other value claims. Although the phrase “conceptions of the g o o d is omnipresent in the literature, different thinkers use it to mean different things.30While there is considerable lack of clarity in the literature, the general sense is that “conceptions of the good” are abstract, developed sets of views on complex, controversial moral issues. In a diverse society, such viewpoints are generally subject to dispute, and so neutrality between conceptions of the good is to avoid governmental appeal to controversial moral views. Barry describes conceptions of the good as “complex systems of belief” that are “open to rational argument” (JI, 167). His usage is similar to that of Rawls, who describes a moral view as “comprehensive”if it contains answers to a full range of difficult and controversial questions, for example, what is of value in human life, ideals of personal character, and other subjects (PL, 13). Rawls’s overlapping consensus is neutral in the relevant sense in excluding appeals to comprehensive views. The basic thrust of a defensible neutrality principle is along similar lines. As a principle restricting sectarian arguments, it will rule out appeals to moral conceptions that depend on sectarian assertions. Thus, we can reserve the term “conceptions of the good” for such moral conceptions. In this sense, conceptions of the good are similar to religious outlooks, and we could characterize religious views as conceptions of the good. Defined in this way,“conceptionsof the g o o d are more complex than many other moral judgments and more subject to controversy. However, although highly controversial value and factual claims must be ruled out, many moral claims cannot be reasonably rejected and so are acceptable under neutrality. These include appeals to commonly recognized values, experiences, traditions, the practices of different cultures, and the relationships of certain traits or entities to basic human capacities. Innumerable simple moral judgments can be defended on nonsectarian grounds, for example: unnecessary suffering is wrong; all things being equal, promises should be kept; all things being equal, it is wrong to kill another human being; and access to art, music, and unspoiled natural areas and opportunities for leisure time improve the quality of human lives. Similarly, the value of many governmental activities that promote uncontroversial goods should be recognized for example, providing a peaceful and stable environment; warding off foreign aggressors; and actions to protect public health and to limit environmental pollution. Regardless of how they characterize neutrality, supporters of the notion generally accept the legitimacy of governmental activities along these lines, thereby recognizing that certain
“ReasonableRejection and Neutrality of Justijication”
177
values are permitted by their neutrality principles. The set of acceptable values in regard to a specific society constitutes the necessary background to a neutrality principle. Identifying these with precision will inevitably be a difficult process. In a pluralistic society, inhabitants will likely disagree about the controversiality of specific value claims as well as, perhaps, applicable standards of controversiality. But once again, clear parameters can be identified. In established liberal societies, value claims that cannot be reasonably rejected include the worth of certain rights that are so deeply ingrained in public culture and so widely accepted that they are virtually beyond question. The same is true of democracy as a form of g0vernment.j’ For our purposes, what is distinctive of such value claims is that they do not rest on single sets of controversial premises, but can be argued for from numerous different perspectives. They are compatible with and recognized by most if not all conceptions of the good in pluralistic liberal societies. It is of course possible to contest these values, for example, to argue that music is not valuable for human beings. But because of the obvious pleasure music provides in virtually all, if not all, human cultures, it is likely that such an argument would be made from a distinct, sectarian standpoint-relying, for example, on the fact that music is sinful, the Devil’s work. For most, if not all, of the value claims noted here, although it is possible to reject them, it is unlikely that one could do so defensibly. Rather, to reject the value of health or public safety, preservation of natural recreation areas, or a clean environment, one would have to argue from a distinctive moral standpoint-from a particular conception of the good. I do not contend that all claims about commonly recognized goods are beyond controversy. But it seems clear that many important claims have strong nonsectarian support. Granted, distinctions between conceptions of the good and nonsectarian value claims, policy justifications that satisfy neutrality will have two main features. First, policies must pursue nonsectarian values. Their goals must be able to be explained without reference to specific conceptions of the good. Second, reasoning for particular steps to achieve the goals should not depend on sectarian modes of argument, as discussed briefly earlier. If these two conditions are satisfied, justifications of public policies should be able to satisfy neutrality. Satisfaction of these two conditions is not enough, however, to justify government actions. Not only must acceptable reasons be given as to why government should provide particular values, but an inescapable feature of political life is choice between competing values. A policy in favor of a given goal will necessarily come at the expense of others, and so trade-offs must be justified. That pursuit of a given value-including values that are nonsectarian-is at the expense of others is readily seen. To take a familiar example, equality and
178
George Klosko
liberty are both important liberal values, but they often conflict. For example, redistributive policies that raise the standard of living of some citizens and lessen economic inequalities will reduce other people’s freedom to do what they wish with their property. Competing values can be balanced in many different ways, and as long as the values in question are nonsectarian, neutrality does not rule them out. Judgments that a given government should pursue policy X rather than Y can of course be reasonably rejected. Even if X is intended to realize values that are nonsectarian, one can contend that the government should pursue Y instead. If Y is too nonsectarian, why doesn’t opting for X violate neutrality? Although more sophisticated than the reductio argument presented earlier, this line of argument would delegitimize virtually all government actions and so have similar consequences. The response depends on distinguishing rejection of arguments because they rest on particular conceptions of the good and because of policy preferences. As noted at the beginning of this paper, arguments can be reasonably rejected for many different kinds of reasons. What is distinctive of neutrality is rejection because of dependence on particular comprehensive views. In the case in point, preference for X will satisfy neutrality if X’s rationale, that is, the values the government seeks to realize in implementing it, are nonsectarian and so cannot be reasonably rejected (and the designated means to them are also nonsectarian). Preference for Y would satisfy neutrality if the same conditions held. In such cases, it is a major function of democratic decision procedures to choose between them. Neutrality itself does not settle such disagreements, although it makes a contribution. In eliminating appeals to conceptions of the good and ruling out sectarian modes of reasoning, neutrality narrows the grounds of disagreement. It is possible to criticize the position developed here as disingenuous. If a given policy dispute is settled by weighing certain values over others, then a particular conception of the good must be involved. Central to any conception of the good is prioritization of certain values over others, and so a specific set of policy priorities must reflect some conception of the good. But to this, too, there is a response. Because neutrality is of justification, the fact that people are motivated in their choice of values by their conceptions of the good is not inconsistent with neutrality. Nor is it forbidden that government policy priorities reflect some conceptions more than others. Neutrality requires only that public policies be intended to realize nonsectarian values and that the relevant means be similarly defensible. Many public policy positions can be justified on such grounds, with disagreements on policy preferences resolved by democratic procedures. Once again, as we have seen throughout this paper, it seems that the most severe problems with neutrality arise in applying these principles to the com-
“Reasonable Rejection and Neutrality of Justijication”
179
plex circumstances of contemporary life. But whatever the difficulties with borderline issues, there are many cases the principle will clearly allow and others it will rule out. Consider, for example, a mandatory seat belt law. This can be justified as intended to prevent injuries and preserve lives, values that are uncontroversial, while the measures imposed can be defended on nonsectarian grounds. Not everyone, of course, will agree with such a law. In an ultimate sense, a given person’s preference for either injury prevention or liberty is grounded in a particular conception of the good. But such relationships do not cause problems for neutrality when, as in this case, preferences for either side can be justified by reasons common to many comprehensive views. Something similar is true with other issues, for example, restrictions on pollution, laws against drunken driving, or funding for music education in public schools. In these and countless other cases, both the goals of the policies in question and the steps recommended can be nonsectarian, while setting priorities among such values is a central task of democratic politics. While cases along these lines should be acceptable, other policy trade-offs will be ruled out. Among interesting cases are when valuation of given goods increases, sectarian premises can be required to support what can be viewed as extreme trade-offs. We can introduce an additional requirement of policy making. Like all government policies, those in accord with neutrality must be justified and so must satisfy a criterion of what we can call adequacy. Because values require trade-offs with others, as the degree or amount of a given value increases, its cost in terms of other, competing values also increases, as does the justificatory burden of its proponents. Adequate arguments are those able to satisfy this justificatory burden. A familiar example of a stricture along these lines is the public law doctrine that important rights can be abridged only for compelling state interests3*For example, in a liberal society, an artist’s preference to forego material enjoyment for her art is ordinarily a matter that concerns her alone and need not be justified to other people. But strong arguments will be needed to support public policies based on similar trade-offs. It seems likely that, in order to develop a convincing case for enormous public expenditures on the arts or highly intrusive arts education, the artist will have to appeal to controversial value claims. For example, she might argue according to Marx’s view that man’s nature, his species being, is to be freely creative in an aesthetic sense, and so that promotion of these capacities should be a top priority in society.33Neutrality, of course, rules out the use of such controversial premises. Full explication of the idea of adequacy is not possible here and must be the subject of another paper.34But a few basic points can be made. To begin with, different conceptions of a given value can be distinguished, including
180
George Klosko
distinctions between conceptions that are acceptably neutral and others that rest impermissibly on comprehensive views. For example, in regard to values associated with art, one can envision a continuum, extending from noncontroversial values at one pole to values rooted in specific comprehensive views at the other. At the noncontroversial end would be values associated with common experiences of art, for example, the enjoyment derived from what are regarded as great masterpieces, the value of learning about art, and perhaps learning how to create it, for example, to paint or to work at a craft. At the other pole would be more distinctive values, for example, those associated with justifications of specific civilizations on the basis of their art, without regard to their other features. Another example is the imperative to create that is bound up with Marx’s view. Precise dividing lines between neutral and nonneutral values are inevitably messy and controversial. Inhabitants of pluralistic societies may well have different ideas concerning borderlines and where they should be located. Once again, I believe that it is in questions such as these rather than in justifying the principle per se that the most severe problems with neutrality lie. Neutrality, of course, requires that a given public policy be justified by values from the neutral side of the continuum. The adequacy requirement entails that the specific conception of the value in question should be sufficient to support the trade-offs it requires. In practice, people are likely to disagree about specific trade-offs, and so such disputes will have to be resolved by democratic procedures. Adequacy requires that the trade-offs from which procedures select must be plausible. Explication of adequacy points to a particular form of intellectual honesty on which neutrality depends. Although not falling under neutrality proper, the requirement here is that one should clearly connect one’s support of given policies with specific conceptions of values in question and avoid false claims in regard to both factual matters and value trade-offs. These requirements fall within the rough category of Consider defense of school prayer. Assume that Jones supports school prayer for religious reasons: he believes it will contribute to saving souls. Neutrality clearly excludes such an argument. But it is still open to Jones to come up with other grounds. He could perhaps argue that introducing prayer might improve student conduct: attendance might increase; there might be fewer fights; and drug use and teenage pregnancy might decline. All of these purported effects are desirable policy goals that do not depend on particular comprehensive views, and so neutrality allows Jones to appeal to them, but not, we should note, if he believes that either the facts do not hold or the value trade-offs are not reasonable. Because the purported factual connections are almost certainly false, the costs of imposing prayer on unwilling students are not justified. However, Jones could deceptively claim that the facts hold and support prayer on this basis. Even though such contentions
“Reasonable Rejection and Neutrality of Justijication”
181
are not ruled out by neutrality, they are excluded by Jones’s general duty of civility not to deceive his fellow citizens. If Jones actually believed the factual claims and accepted the trade-offs in question, that would be another matter. But for him to advance this argument while aware of the problems would be intellectual fraud.36If a majority of citizens in Jones’sjurisdiction accepted the argument, mandatory prayer could well be implemented, again, without violating neutrality, although the policy would be wrong and oppressive for other, closely related reasons. As noted earlier, in order to work acceptably in practice, neutrality requires certain qualities of the populace-in this case, a certain level of intellectual sophistication. In the kind of society in which the necessary background requirements were met, such arguments for school prayer would be unlikely to convince a majority of citizens and so unlikely to be endorsed democratically. Along similar lines, civility also requires that, if one is advancing arguments that are intended to pass muster under neutrality, one should argue according to the specific conception of the value in question that is actually the goal of one’s policy position.37For example, assume that Jones supports enormous government spending on art because of particular features of his comprehensive view; for example, he believes that art is the supreme achievement of civilization, thereby justifying enormous sacrifices. Neutrality excludes this view, and we will also assume that acceptably neutral values associated with art do not plausibly support the desired level of spending. Even though Jones’s particular values are excluded, neutrality does not prevent him from contending that neutral values of art do in fact justify the spending. He could perhaps argue that art education is so important that it justifies extensive cuts in other government spending on public schools or social welfare programs. But clearly, an argument along these lines would be dishonest. For Jones to advance it while believing it is false would violate his overall duty of civility, if not exactly the neutrality principle. If Jones actually accepted the trade-off in question, that would again be another matter. But with an acceptably neutral conception of art clearly indicated, it would be difficult for him to do so, and such an argument would be unlikely to convince many of his fellow citizens. As a final example, consider the treatment of animals. Within certain parameters, governmental policies to ensure humane treatment should be consistent with all or virtually all conceptions of the good in liberal societies. Such policies include measures requiring that laboratory animals be subjected to as little pain as possible and that animals be raised and slaughtered humanely. I take it that the value of reducing needless suffering among animals is noncontroversial. This value can, of course, be rejected by adherents of particular conceptions. But it seems likely that such grounds for rejection will be sectarian and so themselves rejectable. Along similar lines, policies that place far
182
George Klosko
greater emphasis on, say, animal rights are more difficult to justify. For example, policies banning all medical experiments on animals or mandating vegetarianism will require views of the significance of animals that will be sectarian and difficult to defend. I think it unlikely that the desire to limit suffering alone would be widely viewed as supporting complete bans on medical experiments and on eating meat. It is of course possible that citizens will claim that these trade-offs are justified, although, again, without appeal to some additional values connected with animals, the bans are unlikely to be viewed as justified. Neutrality, of course, makes it difficult to invoke the additional values, while duties of intellectual honesty and civility require proponents on all sides of the debate to indicate clearly the particular conceptions of their values to which they appeal and the precise nature of the trade-offs they support. Once again, in narrowing the range of values to which participants can appeal, and so in calling attention to the relevant trade-offs, neutrality would facilitate resolution of disputes. Thus, I conclude that if enough citizens support the neutrality principle and possess the necessary intellectual sophistication, their deliberations can be acceptably neutral. Neutrality will not settle debates, although it will circumscribe the considerations participants can justifiably raise. It will eliminate arguments from values that depend on particular comprehensive views, and so will allow a clearer grasp of what is at stake. As long as participants argue from acceptable values, neutrality will contribute markedly to acceptable resolution of disputes.
IV Because neutrality pertains only to values and modes of argument rooted in comprehensive views, it cannot be criticized as ruling out too much, in a different respect. A critic might contend that claims essential to virtually all public policy debates are able to be reasonably rejected. For instance, proponents of lower income tax rates argue that these would dramatically stimulate the economy and allow people to keep more of their own money. Opponents claim that stimulus effects will be much weaker and that lower rates will exacerbate economic inequality, which is undesirable. Granted the distinction between acceptable and unacceptable value claims, arguments from inequality and allowing people to keep more of their own money would appear to be admissible. But the crucial predictions of economic consequences could still appear to be in jeopardy. Predictions about the likely consequences of a tax cut, like other claims concerning economic matters, are clearly bound up with particular compre-
“Reasonable Rejection and Neutrality of Just$cation”
183
hensive views. In the United States at the present time, there are conservative economists and liberal economists, while the two groups routinely reject forecasts of the other camp and develop strong reasons for doing so. Because of clear connections between economic beliefs and overall comprehensive views, it could be argued that neutrality requires that the former should be barred from policy debates. Such a consequence would be unacceptable on its face and so lead us to reject neutrality. For the sake of argument, I will concede close connections between economic and overall comprehensive views. There is strong evidence that conservatives and liberals view the world-or at least important aspects of it-in sharply different ways.38 Their economic convictions are part and parcel of their overall views, “constrained by the latter and can be accurately predicted from knowledge of other aspects of the However, economic forecasts and similar claims are allowable under neutrality because of the distinction between justification of a given claim and its motivation. Although Green might hold his specific economic views in some sense because of his comprehensive view, there are generally preferred modes of argument for economic and similar claims. Although economic claims are oftentimes far from certain, they are essentially empirical and should be based on the best available analytical techniques, whether scientific or more loose appeals to experien~e.~~ Economists generally agree on what constitutes relevant evidence and the kinds of considerations that would resolve contested issues. In order to be taken seriously, economic predictions should be supported by the best evidence available, though such evidence is frequently unable to resolve disagreements. Given the complex nature of the factors involved and the inherent difficulty of conducting controlled experiments in the social sciences, disputes might never be resolved. But in principle, were the necessary evidence collected and crucial experiments conducted, we would know about the consequences of particular policies. In such cases, predictions based on reliable evidence could not be reasonably rejected-as, to use Thomas Nagel’s example, predictions based on the germ theory of disease cannot be.41In such cases, appeal to techniques other than those recognized could be reasonably rejected. The principles appealed to here are extensions of those presented earlier. Roughly, if there are clearly preferable methods of inquiry in a given area and general agreement on the kinds of considerations that are relevant to resolving disputes, neutrality allows arguments based on these considerations. Because preferred analytical techniques are not rooted in particular comprehensive views but are neutral between them, specific claims supported by the techniques are acceptable, even if they are subject to disagreement by people with different comprehensive views who share the same paradigm of inquiry.42
184
George Klosko
Once again, it could be argued that the distinctions here are disingenuous. If conservative and liberal economists both hold their views and disagree with each other because of their comprehensive views, then what is at issue between them are their different comprehensive views. But once again, the response is that the participants are able to ground their preferred policies in a mode of argument that is recognized as acceptably neutral, to which they both subscribe. Moreover, as in the cases discussed earlier, we can presume that they share nonsectarian central policy goals, for example, increased economic growth and greater general prosperity and contentment throughout society, and so disagree mainly about means. Combine pursuit of nonsectarian ends and agreement on modes of argument applicable to means and the resulting arguments should satisfy neutrality. Although the policy views of disputants in economics and other areas could well be constrained by other aspects of their comprehensive views, their situation is importantly different from that in regard to conceptions of the good. No accumulation of empirical facts could prove the superiority of Kantianism over utilitarianism, liberalism over conservatism. Among proponents of these different views, there is no agreement on the kinds of evidence that would constitute decisive support for one position over others. Because more than factual considerations are involved, reasonable rejectability in regard to such views is qualitatively different than in regard to essentially factual issues encountered in many areas of public policy, regardless of disputants’ underlying motivation. V
To close, I will briefly address the possible criticism that the defense of neutrality outlined here concedes too much. A conception of neutrality that allows government to weigh different values and decide among them cannot be called ‘[neutral”in any meaningful sense. I believe there is a response. As we have seen, neutrality is principally intended to exclude certain kinds of reasons from public deliberations, and such a view provides the logical working out of neutrality of justification. Moreover, it is important to recognize that the view developed here is still “neutral” in perhaps the central sense of the term. As noted earlier, in the literature, [‘neutrality’’is generally taken to mean neutrality between conceptions of the good. Theorists fill this in to mean not being able to claim the superiority of one conception of the good over others. According to Charles Larmore, “political neutrality consists in a constraint on factors that can be invoked to justify a political decision. Such a decision can count as neutral only if it can be justified without appealing to the presumed
“ReasonableRejection and Neutrality of Justijkation”
185
intrinsic superiority of any particular conception of the good life.” Central to Bruce Ackerman’s neutrality principle is forbidding a power holder to assert “that his conception of the good is better than that asserted by any of his fellow citizens.” In Barry’s words: “[N]obody is to be allowed to assert the superiority of his own conception of the good over those of other people.”43 Although the sense of neutrality defended in this paper allows participants in public policy debates to claim value for many things, it constrains appeal to sectarian views to defend them. The value claims that neutrality allows are common to-neutral between-the diverse conceptions of the good found in contemporary societies. Governmental promotion of these values is not premised o n assertions of the superiority of specific conceptions of the good, and so is consistent with the constraint that Ackerman and others require.44For these reasons, the policy results of the view presented here would resemble those yielded by other conceptions of neutrality, although without, as I maintain, objectionable philosophical features of the latterj5
Notes 1. Formulation in terms of reasonable rejection is owed to Thomas Scanlon,“Contractualism and Utilitarianism,” in Utilitarianism and Beyond, ed. A. Sen and B. Williams (Cambridge:Cambridge University Press, 1982);and Thomas Scanlon, What We Owe to Each Other (Cambridge,Mass.: Harvard University Press, 1999). 2. John Rawls, Political Liberalism (New York: Columbia University Press, 1993), 13. 3. For discussion, see Jeremy Waldron, “Legislation and Moral Neutrality,”in Liberal Neutrality, ed. Robert E. Goodin and Andrew Reeve (London: Routledge, 1989); and George Klosko, Democratic Procedures and Liberal Consensus (Oxford Oxford University Press, ZOOO), 14-18. 4. For these criticisms of neutrality, see William Galston, Liberal Purposes (Cambridge: Cambridge University Press, 1991), chapters 4-5; Stephen Macedo, “The Politics of Justification,” Political Theory 18 (1990): 280-304; Stephen Macedo, Liberal Virtues: Citizenship, Virtue and Community in Liberal Constitutionalism (Oxford Clarendon, 1990), chapter 7; and Michael Sandel, Democracy’s Discontent: America in Search of a Public Philosophy (Cambridge, Mass.: Harvard University Press, 1996). Brief responses to these lines of criticism are found in Klosko, Democratic Procedures, 13-18. 5. Conceptions of neutrality based on respect are criticized by George Sher (Beyond Neutrality: Perfectionism and Politics [Cambridge: Cambridge University Press, 19971, chapter 4) and Steven Wall (Liberalism, Perfectionism and Restraint [Cambridge:Cambridge University Press, 1998],83-91). Because the conceptionsof neutrality they discuss are far stronger than that presented here, their arguments do not tell against it. A different conception of respect is advanced by Galston and other scholars. To quote Galston: “[Wle show others respect when we offer them, as explanation,what we take
186
George Klosko
to be our true and best reasons for acting as we do” (Liberal Purposes, 109). This conception of respect cannot be discussed in this chapter. Briefly, if adherents of a given conception of the good are a majority in a given society, this conception would allow them to impose their views on nonadherents. This line of argument is exploredalong with other implications of this conception of respect-in George Klosko, “Two Concepts of Respect” (forthcoming). 6. For this reason, all social policies should be subject to the neutrality requirement, not only matters concerning constitutional essentials and basic justice, as Rawls (Political Liberalism, 215,235) argues; see also Brian Barry, Justice As Impartiality (Oxford Clarendon, 1995), 143. However, the wrongfulness of disrespect in regard to specific governmental actions varies with the type of actions; see Robert Audi, Religious Commitment and Secular Reasons (Cambridge: Cambridge University Press, 2000). 7. In many disagreements over social policies, proponents of both sides put forth arguments that the other can reasonably reject, for reasons closely related to their opposed comprehensive views. Reasons why such cases do not fall under neutrality will be discussed in section IV. 8. Rawls, Political Liberalism, lect. 6, especially 223-27; and John Rawls, “The Idea of Public Reason Revisited,” University of Chicago Law Review 64 (1997): 773-80. We can posit a rough continuum between kinds of arguments of different levels of acceptability. Moving down from those that are most publicly accessible, a weaker category would include other empirical considerations, for example, evidence based on what a given person has seen or heard. Less satisfactory are modes of argument that depend on particular forms of authority or particular personal experiences. Common members of these classes are appeals to religious authorities and personal revelation. In addition to superiority according to generally accepted epistemic standards, the preferred argument types can be justified far more easily to the bulk of a liberal population. For discussion, see Kent Greenawalt, Religious Convictions and Political Choice (New York Oxford University Press, 1988); and Kent Greenawalt, Private Consciences and Public Reasons (Oxford Oxford University Press, 1995). 9. For relevant percentages of the population in American society, see Klosko, Democratic Procedures, chapter 4. 10. See Jeremy Waldron, “The Theoretical Foundations of Liberalism,” Philosophical Quarterly 37 (1987). 11. This is the position of the U.S. Supreme Court in Edwards v. Aguillard, 482 U.S. 578 (1987); see also Epperson v. Arkansas, 393 U.S. 97 (1968). 12. Edward Larson, Summer for the Gods (Cambridge, Mass.: Harvard University Press, 1997), 258, 304n27. For overall levels of support for biblical authority, see Klosko, Democratic Procedures, 106-8. 13. As with Rawls’s “political liberalism,” defense of neutrality presupposes a basically liberal society, with appropriate public culture. 14. For “wide” public reason, see Rawls, “Idea of Public Reason Revisited,” 783-87. For “inclusive” public reason, see Lawrence Solum, “Inclusive Public Reason,” Pacific Philosophical Quarterly 75 (1994): 217-31. 15. Audi, Religious Commitment, 70-73; Rawls, “Idea of Public Reason Revisited,” 783-87. Smith‘s conduct would also be acceptable under neutrality if she believed
“ReasonableRejection and Neutrality of Justification”
187
there were adequate neutral reasons for the policy in question, even though she did not present them in public debate. 16. See Greenawalt, Religious Convictions. 17. Because I assume a liberal society, I do not discuss issues encountered if inhabitants support an authoritatively grounded conception of the good, especially one that is highly intolerant or otherwise conflicts with liberal values. 18. Rawls, Political Liberalism, 150. 19. Barry, Justice As Impartiality (cited hereafter as JI, generally in the text); Rawls, Political Liberalism (cited as PL, generally in the text). 20. Simon Clarke, “Contractarianism,Liberal Neutrality and Epistemology,” Political Studies 47 (1999): 634. 21. In the following paragraphs, I draw on points presented by Simon Caney, “Impartiality and Liberal Neutrality,” Utilitas 8 ( 1996): 276-80; and Clarke, “Contractarianism,”634-37. They present other arguments, in addition to the ones discussed here. 22. It is notable that, in a symposium on JI, Barry presents no real response to his critics in regard to the different statuses of beliefs about justice and the good, aside from promising to demonstrate this in the third volume of his treatise (Brian Barry, “Contractual Justice: A Modest Defense,” Utilitas 8 [ 19961: 375-76). 23. Wall, Liberalism, chapter 3. I should note that Rawls is not deeply concerned with establishing the validity of a neutrality principle, and so would probably not be troubled by this conclusion. 24. I am indebted to Steven Wall for suggesting this locution. 25. We can distinguish two burdens of justification. (1) The proponent of neutrality is committed to presenting acceptable grounds for distinguishing between neutral and comprehensiveviews-what Rawls calls “political”and “comprehensive reasons.” This distinction is to be made on the basis of considerations concerning the scope of different views and epistemological considerations in regard to their support. (1) The critic of situational irresolvability in a given area must demonstrate the force of his arguments and that they fall on the neutral side of the distinction in (1). 26. It is possible that Jones could argue for forcible imposition of some policy, based on his conception of the good, while conceding that his conception could be reasonably rejected. While I do not necessarily rule this possibility out, once again, Jones would have to meet a severe burden of proof to defend his position. For points in the paragraph, I am indebted to discussions with Wall. 27. I should note that situational irresolvability does not leave the argument of this chapter vulnerable to criticisms that beset Barry’s view. The fact that convictions concerning justice seem to be as controversial as those concerning the good is a problem for Barry, not because of its implications for neutrality, but because he wishes to establish a theory of justice in spite of neutrality. For the purposes of this chapter, there is no drawback to recognizing that situational irresolvability applies to views concerning both justice and the good. Moreover, as we will see in the next section, situational irresolvability allows acceptance of a wide range of value claims, and so applies far more narrowly than Barry’s epistemological skepticism. 28. Representative issues include the following: whether the government can support museums, symphonies, and national parks, whether public schools can teach art
188
George KZosko
and music, whether the state can take actions to develop citizens’ autonomy, and many others. See note 29 for those discussed by Richard Kraut. 29. Richard Kraut, “Politics, Neutrality and the Good,” Social Philosophy and Policy 16, no. 1 (1999): 317 (parentheticalpage references in this paragraph are to this work). Similar arguments are presented by Michael Sandel, “A Response to Rawls’ Political Liberalism,”in Liberalism and the Limits oflustice (Cambridge: Cambridge University Press, 1998), 196-202; and David Reidy, “Rawls’s Wide View of Public Reason: Not Wide Enough,” Res Publica 6 (2000): 49-72. 30. See Sher, Beyond Neutrality, 37-44. A valuable discussion is found in Joseph Chan, “Legitimacy, Unanimity, and Perfectionism,” Philosophy and Public Affairs 29, no. 1 (winter 2000): 10-14. As I use the term, “conceptionsof the good” correspond to what Chan identifies as “comprehensive”conceptions. 31. For support of rights and democracy, on an abstract level, and implications, see Klosko, Democratic Procedures, chapters 2 , 5 . 32. For a discussion, see Gerald Gunther, Individual Rights in Constitutional Law, 4th ed. (Mineola, N.Y.: Foundations, 1986), 20, 640, 711; and Henry Abraham, Freedom and the Courts, 5th ed. (Oxford Oxford University Press, 1988), chapter 2. For lists of interests that have been viewed as compelling and not compelling in U.S. Supreme Court decisions, see George Dent, “Religious Children, Secular Schools,” Southern California Law Review 61 (1988): 902-4. 33. Karl Marx, “Economic and Philosophic Manuscripts of 1844,” in The MarxEngels Reader, 2nd ed., ed. Robert Tucker (New York Norton, 1978), 74-76. 34. George Klosko, “The Politics of Public Reason” (forthcoming). 35. Rawls, “Idea of Public Reason Revisited,” 770-71. Rawls relates duties under public reason to ideals of civility, reciprocity, and civic friendship. For a somewhat different natural duty of civility, see Rawls, A Theory oflustice (Cambridge, Mass.: Harvard University Press, 1971), 355. On these points, I have benefited from discussion with Micah Schwartzman. 36. The requirement here is similar to Audi’s “principle of secular motivation” in Religious Conviction, 96-100. It is somewhat narrower than Audi’s principle, in that the requirement is that acceptable arguments must be sincerely viewed as such, whether or not Jones is in fact motivated by them. 37. According to wide public reason, not all arguments advanced must be neutral, although there must be adequate neutral grounds for the position one defends. The requirements under consideration here pertain only to the reasons that are intended to satisfy the neutrality requirement. 38. Herbert McCloskey and John Zaller, The American Ethos (Cambridge, Mass.: Harvard University Press, 1984); and Klosko, Democratic Procedures, chapter 6. 39. For “constraint,” see Philip Converse, “The Nature of Belief Systems in Mass Publics,” in Ideology and Discontent, ed. David Apter (Glencoe, Ill.: Free Press, 1964). 40. One could describe many economic predictions as “quasi empirical.”Although they are factual claims about the economy (or have the appearance of factual claims), in many cases they are not subject to empirical testing, for reasons along the lines of those referred to by Rawls as burdens of judgment (PL, 54-58).
“ReasonableRejection and Neutrality of Justification”
189
41. Thomas Nagel, Equality and Partiality (Oxford: Oxford University Press, 1991), 161. 42. Compare Galston’s arguments against the neutrality of scientific reasoning (Liberal Purposes, chapter 5). 43. Charles Larmore, Patterns of Moral Complexity (Cambridge: Cambridge University Press, 1987),44; Bruce Ackerman, SociaZJuusticein the Liberal State (New Haven, Conn.: Yale University Press, 1980), 11; Barry, JI, 160; see also Joseph Raz, The Morality of Freedom (Oxford: Oxford University Press, 1986), 110-1 1. 44. The position developed in this chapter is similar to “moderate perfectionism,” in Chan’s sense (“Legitimacy,Unanimity, and Perfectionism”),perhaps indicating that the labels “neutrality” and “perfectionism”have become outmoded. But in defending perfectionism and criticizing neutrality, Chan does not consider the implications of what I take to be the core of neutrality: exclusion of controversial reasons. 45. Earlier versions of this paper were presented at the 2000 meeting of the American Political Science Association, at the University of Colorado, and to the University of Virginia Philosophy Department. I am grateful to the following for helpful comments and discussions: Ernie Alleva, Colin Bird, Talbot Brewer, Richard Dagger, David Mapel, George Sher, Micah Schwartzman,Michael Tooley, and Steven Wall.
This Page Intentionally Left Blank
12 "Liberal Neutrality on the Good: An Autopsy" Richard J. Arneson
In this paper, Arneson carefully examines a range of arguments that have been advanced to justify a neutrality principle. After making clear what he takes to be the heart of state neutrality, Arneson discusses what he calls the responsibility and self-defeating arguments, the views of Ronald Dworkin, defenses of neutrality from a skeptical standpoint, and other arguments in the literature. Arneson contends that none of these arguments succeeds and that the neutrality principle lacks a compelling justification. He concludes by arguing that objective human goods can be identified and should be promoted by the state, even if they are controversial.
- 191 -
192
S
Richard J. Arneson
HOULD GOVERNMENT BE NEUTRAL “on
the question of the good life, or of what gives value to life”?’ Some political theorists propose that governmental neutrality is a core commitment of any liberalism worth the name and a requirement of justice. For them, neutrality is the appropriate generalization of the ideal of religious tolerance. The state should be neutral in matters of religion, and neutral also in all controversies concerning the nature of the good or the ways in which it is valuable and worthwhile to live. Opposed theorists find neutrality on the good to be not only wrong-headed but obviously and definitively wrong-headed. A reviewer of a recent book attacking the neutrality doctrine chimes in with the confident belief that “the period of neutralist liberalism is now over.”2 The dispute about neutrality on the good breaks down into several different disputes, because what some authors defend under the heading of neutrality is not the same as what others attack. An interpretation of the neutrality ideal must answer at least the following questions: (1) Is it the political government alone that is bound by the neutrality ideal, or is there a version of neutrality that applies to civil society institutions and individual persons as well? ( 2 ) How should one draw the line between matters that fall under the concept of the good, concerning which some entities are supposed to be neutral, and matters that do not fall under the good and hence are not bound by the neutrality constraint? (3) What is a neutral policy or framework? Before tackling these questions, I want to emphasize the limited character of this dispute. The issue is sometimes framed as one between liberals and their opponents. Liberals are thought to be for wide individual freedom of action and limited governmental interference in the conduct of individual lives, and neutrality is thought to be part of the moral argument that justifies these stances. Ronald Dworkin once ventured the interesting suggestion that liberalism is distinguished from rival left-wing and right-wing doctrines by upholding a conception of equality that is neutral on the good, whereas conservatives and radicals alike favor a nonneutral conception of equality and aspire to use state power to establish a virtuous society where the true good prevails3 There is no point in quarreling over the use of the term “liberal,” which has meant many different things to different people. Dworkin is free to use the term as he likes. But the obvious point should be noted that political views that uphold wide individual freedom of action and limited governmental interference in the conduct of individual lives can appeal to people who are not at all attracted by neutrality or neutral conceptions of equality. John Stuart Mill in a famous essay argued that the goal of maximizing human goodwhen human good is rightly understood-requires strict conformity to his liberty principle that tightly limits the proper grounds for coercive interference in individual liberty of action:
“LiberalNeutrality on the Good: An Autopsy”
193
Mill’s utilitarianism is one example of a nonneutral political morality, one that proceeds by taking a stand on the nature of human good, that in its nonneutral way supports wide individual freedom of action and limited governmental interference in the conduct of individual lives and various other policies often associated with the ideals of liberalism. There are other examples of nonneutral political moralities that support liberal policies. Within the nonneutralist camp, this brand of liberalism has its critics. In this chapter, I want to stand aside from these intramural disputes. My aim is to attack the neutrality doctrine. 1. The Idea of Neutrality
Discussions of the neutrality doctrine usually distinguish three versions of neutrality and repudiate one of them as not belonging to a defensible neutrality ideal.5 The three versions are: 1. Neutrality of aim requires that no actions or policies pursued by the state
should aim at promoting one controversial way of life or conception of the good over others. 2. Neutrality of effect requires that the policies pursued by the state should not bring it about that any controversial way of life or conception of the good is advantaged over others. Nor should state policy bring it about that adherents of any controversial way of life or conception of the good are advantaged by comparison with adherents of other ways and conceptions. 3. Neutrality of justification requires that any policies pursued by the state should be justified independently of any appeal to the supposed superiority of any way of life or conception of the good over others. Once these distinctions among types of neutrality are made, it is immediately clear that nobody who wants to defend neutrality on the good would really want to defend neutrality of effect.6 To see this point, consider that basic religious tolerance straightforwardly violates neutrality of effect. If the state guarantees that all its members are free to practice the religion of their choice and proselytize freely on behalf of any religious belief they care to defend, the effects of this policy of religious toleration will be nonneutral. Some religious doctrines are implausible and cannot withstand public scrutiny and open debate. If the implausible doctrine sect must defend its creed in freewheeling debate, the sect loses adherents, but if religious proselytizing were prohibited,
194
Richard I. Arneson
the sect might well thrive, or at least retain members for many generations. Sects espousing plausible doctrines (doctrines that will find willing adherents in free and open religious debate) will be advantaged in the sense of gaining more adherents under a regime of freedom of religion than under a nopoaching regime in which adherents of rival sects are not permitted to attempt to persuade others to convert to their own sect. But most friends of the neutrality ideal would regard neutrality in any relevant sense as satisfied and not violated by the regime of freedom of religion, including religious free speech, no matter what sects gain and what sects lose under this regime.7 Beyond this agreement, one finds the ideal of neutrality bifurcating. One ideal is neutrality of justification alone, interpreted to require that the ultimate justification for all state policies should avoid reliance on controversial ideas of the good. This ideal allows as consistent with neutrality that the state might conceivably be justified in pursuing policies that are nonneutral in aim. To take a simple example, one might adopt a policy that aims to promote Roman Catholicism over other religions in a case where this policy has a sound neutral justification, say the achievement of civil peace. A quite different neutrality ideal results if the neutrality advocate starts with the idea that given the correct understanding of human good, along with other elements of a reasonable political morality, the adoption of policies by the state that are nonneutral in either justification or aim is unjustifiable. On this view, neutrality is identified with neutrality of aim plus neutrality of justification qualified by the idea that the adoption of this package ideal is itself ultimately justifiable partly on the basis of an ideal of human good. One might put the point in this way: Given the nature of human good, the state should never aim to promote any controversial ideals of the good, and its policies (except for this very doctrine being proposed) should not be such as to be justifiable only by appeal to the claim that some ideal of the good is superior to any other. 2. The State Should Be Neutral Concerning ControversialIdeas of the Good
Neutrality of justification is usually held to apply to the state and not to private citizens or other nonstate agencies. The question arises: Why should the neutrality doctrine take this particular shape? The state routinely coerces its subjects and typically claims the legitimate authority to act in the name of all members of the nation. Some theorists claim that these attributes of the state constrain what can qualify as morally acceptable state policies and that neutrality is among these constraints. Perhaps, we need not be concerned to show that the neutrality applies to the state
“LiberalNeutrality on the Good: An Autopsy”
195
and only to the state. If advocates of neutrality can demonstrate that it should constrain state policy, it may not be problematic that the appropriate scope of neutrality might extend farther. As characterized so far, neutrality of justification exhibits an odd asymmetry. We are told that government should not pursue policies that are justifiable only by appeal to controversial conceptions of the good. If this is sound advice, one wonders whether it applies more broadly. In the same spirit of neutrality, should government refrain from pursuing policies justifiable only by appeal to controversial conceptions of the right? For that matter, a generalized neutrality policy would forbid a government from pursuing policies justifiable only by appeal to controversial empirical claims such as the theory of evolution or the theory that AIDS is a sexually transmitted disease. The last extension clearly would go too far for neutrality advocates. But then why should neutrality have just the scope that its advocates claim for it? This chapter canvasses two distinct responses. One is that the nature of the good happens to require neutrality. This would be so whether or not claims about the good were controversial. Sections 3 and 4 pursue this response. A quite different response is that “principles of justice [political morality] are objective and interpersonally recognizable in a way that conceptions of the good are not.”8 Sections 5 to 8 explore this response. A further preliminary question is what qualifies a conception of the good as controversial. Neutrality theorists sometimes treat this as a descriptive issue: A conception is controversial just in case people actually dispute it. With controversiality interpreted in this way, the neutrality doctrine would then hold that the policies pursued by the government ought to be actually acceptable to all citizens. Charles Larmore indicates that this is how he construes the ideal of neutrality that he espouses. Political neutrality, he writes, “does not require that the state should be neutral with respect to all conceptions of the good life, but only with respect to those actually disputed in the s~ciety.”~ If everyone happens to agree that some way of life or conception of the good is superior to others, the neutrality ideal allows state policy to be based on this shared judgment. This does not seem to be an attractive interpretation of neutrality. It is both too strong and too weak. Too weak Even if everyone in society happens to agree that Roman Catholicism is correct, basing state policy on the judgment that Catholicism is superior to other religions and worldviews or making Catholicism the established state religion seems to be a bad idea, and not just because someone in the future might well disagree with Catholicism. Too strong: A claim can be controversial just because people actually dispute it, even though there are no good grounds for disputing it. Why does an otherwise acceptable governmental policy become unacceptable just because
196
Richard J. Arneson
someone or some number of citizens object to it even if they have no good reasons to object to it? One reason that state establishment of Roman Catholicism should strike us as a bad idea even in a society composed entirely of devout Roman Catholics is that there are many good reasons to raise skeptical doubts about the claimed superiority of Catholicism over other versions of Christianity, the claimed superiority of Christianity over other religions, and for that matter about the claimed superiority of religious as opposed to nonreligious worldviews. State coercion deployed to prop up a religion to which reasonable people might well object is unacceptable even if no one actually objects. The point could be made by stating that if we are trying to interpret a norm against coercive state imposition of what is controversial as sympatheticallyas possible, whether a claim or doctrine is controversial should be a normative issue, not a sociological issue. The question should not be whether the claim is actually disputed but whether it ought to be disputed, whether there are sufficiently strong grounds that warrant raising objections. 3. The Nature of the Good Rules Out State Action to Gain It
In a statement that may merit canonical status, Dworkin writes that “the government must be neutral on what might be called the question of the good life.” He glosses this remark as follows: “[Plolitical decisions must be, so far as is possible, independent of any particular conception of the good life, or of what gives value to life.”’O No mention is made here of what is controversial or should be controversial. Instead, the contrast is between claims about the nature of the good life and other moral claims. In particular, let us roughly distinguish claims about what we owe to others and are owed by them by way of conduct and terms of interaction and claims about what is valuable or worth seeking in human lives. In a slogan, we distinguish claims about what is good and claims about what is right (moral, fair, and just), and neutrality is supposed to be an ideal that applies to the former category of claims only. But the basis for this restriction on the scope of neutrality is not that all claims about the good are uncertain or undecidable and (some) claims about what is right are objectively decidable. Rather, the idea is that something about the good renders it an unsuitable basis for determining state policy and an improper goal for state policy. There are at least two claims to consider that connect the appropriateness of state neutrality with the nature of the good. One claim is that each individual has a special and nondelegable moral responsibility to decide what is to count as a good life for herself and to try to bring it about by her own agency that her
“Liberal Neutrality on the Good: An Autopsy”
197
life qualifies as a good life by this self-chosen standard. Some actions that would violate the neutrality ideal wrongfully interfere with the special responsibility that each individual has to pursue her own good and are improper for that reason. If you have a special nondelegable moral responsibility to mow the grass on your parents’ lawn, then no one else should do it for you and no one should interfere with your carrying out this responsibility, other things being equal. Each person’s relation to her own good is like that, so there is to this extent a moral underpinning for state neutrality. The other claim to consider is that for each person the good life has to be an achievement and cannot be a benefit that is conferred on her. State action that violates the neutrality ideal renders what would be the attainment of the good life for an individual an achievement of the state rather than an achievement of the individual, and such state action is thus always self-defeating, as though one tried to help one’s son walk on his own by “helping” him in ways that guarantee that his movements will not qualify as walking on his own. It may be worth pausing to notice that the two arguments for state neutrality just given do not rely on anything like skepticism about the possibility of knowledge about the good. Nor is there any reliance on any claim to the effect that only the individual can ever be in a position to know what is the good life for that very individual. To use a very simple example, suppose that there is just one life choice that Smith might make-to be a rancher or a music composer, and there is no doubt that the life of a music composer is superior to the life of a rancher. The responsibility claim is that Smith has a special nondelegable responsibility to make her life a good one and this implies that Smith alone should make the choice to be a rancher or composer and that it would be wrong for anyone to use state power to preempt her choice, manipulate it, coerce it, or otherwise interfere with her freedom to choose. This moral requirement could hold even if such state action would produce the better state of affairs, the one in which Smith becomes a composer and lives better. The self-defeating claim is that the nature of the good is such that Smith‘s life goes better if she freely chooses it-even if she chooses badly-than her life would go if she were prevented from choosing freely (or simply did not choose freely), regardless if the forced choice or preempted choice leads to a way of life that, taken by itself in abstraction from consideration of the process by which she got to it, would be better. In other words, the self-defeating claim is that interference with an individual’s living her life as she chooses in order to improve the quality of that life is inherently self-defeating. This articulation of the state neutrality ideal has a certain charm. The case for neutrality is interwoven with traditional liberal views concerning the sovereign moral urgency of allowing individuals to lead their own lives as they
198
Richard J. Arneson
choose within constraints set by rights of others reasonably understood. Neutrality in this vision is the comrade of antipaternalism. The responsibility argument and the self-defeating argument can be made, and have been made, in more sophisticated ways than the crude rendering that I have given. But my rendition contains flaws that (so I would claim) plague any version one might devise, however sophisticated. The arguments do not provide sound support for neutrality of justification. Suppose a government deems opera to be a valuable enterprise and provides subsidies to opera productions in order to provide every citizen a cheap opportunity to become a member of the audience of a live first-rate production of a fine opera. The justification of this policy is the judgment that the classic repertoire of operas is splendid music and theatre and is superior in quality to other forms of entertainment that are likely to be popular among citizens. The policy then straightforwardly violates the neutrality ideal. But if we try to wheel out the responsibility and self-defeating arguments to defend the neutrality ideal in this application, we are stymied. All that the state has done is offer a sample of a good to the individual. This action, reasonable or not, could not be thought to trench on the responsibility the individual has to choose her own conception of the good and fashion a life plan to achieve it. The state provides an opportunity to undergo a certain experience (and let us say provides an accurate description of it so that the offerees have a reasonably clear idea what it is they are accepting or rejecting) and does not force anyone to have that experience. The opportunity enables one who takes it to form in this regard better informed preferences than one would have had absent the experience. Remaining free to take or leave the offered experience, and, if she takes it, to form her own judgment on it and to be moved or left cold by the experience, it is hard to see how the individual's responsibility for her own good is anything but intact. One might object that the nonneutral state that subsidizes opera productions out of the conviction that this form of entertainment is more excellent than others must fund the scheme somehow, there being no free lunch, and hence must ultimately take money from its citizens' pockets to pay for the scheme. This is money the citizen might have used for any purpose she chooses, including sampling experiences that would tend to form her preferences in ways that diverge from state priorities. Hence, there is nonneutral tampering with the individual's pursuit of her own good in her own way. I confess this seems to me to stretch the idea of interference beyond good sense. And anyway, we can imagine that the state finds itself with a nonconvertible windfall, so it can either subsidize opera (at no cost to taxpayers) or not but cannot put the windfall to alternative uses, and decides to subsidize opera, the justification being its (correct, we can assume) judgment of the superiority of
“LiberalNeutrality on the Good: An Autopsy”
199
the operatic conception of the good to others. Neutrality is still violated in the example but the responsibility argument provides no support at all for neutrality in this application. Nor does the self-defeating argument help the neutralist cause. Whatever view we should take of the best process for forming values, ambitions, and preferences, simply offering the opportunity to someone to sample a good she might find to be valuable cannot fall in the category of acts that aim to enhance another’s good but are self-undermining. This example might be claimed to be of the wrong sort to challenge neutrality. The claim would be that a neutral justification can be given of state policies that aim to encourage people to adopt a conception of the good deemed especially valuable (or to reject a conception of the good deemed especially lacking in value) and do this merely by offering individuals the opportunity to experience the favored conception. Such policies merely expand people’s freedom, it might be thought. One might wonder whether expanding freedom on some particular construal of freedom is itself neutral, but just assume for the sake of the argument that some neutral justification for this could be found. Still, the line of thought to this point does not justify state policy that expands individuals’ freedom by subsidizing experiences of opera or Shakespearean drama rather than some of the many forms of low-grade and nearly worthless activities that for some reasons are not fashionable in society at present. The justification for expanding freedom by subsidizing opera rather than cockfighting or roller derby must appeal to the claimed superiority of opera. Another possibility worth discussing is that the responsibility argument and the self-defeating argument, even if they do not amount to a defense of neutrality as so far conceived, might carve out a position that is defensible and in the neighborhood of neutrality. This possibility seems unlikely. Suppose we accept that each person has a special responsibility to fashion for herself a life that has meaning and worth. This leaves it entirely open that society, the rest of us who could do something to improve that person’s life, might have a backup responsibility in this regard. For example, even if we think that addictive recreational drugs might play a positive role in the lives of some people, for most of us, the use of such drugs is bad for us, and we have a duty to avoid them. I have a duty to myself to avoid addiction to hard drugs, but it is reasonable to suppose that society has a duty to assist me to avoid addiction, and a coercive ban on the recreational use of certain addictive drugs might well be, all things considered, a sound policy. Such a policy would be justified in part by the judgment that addiction to certain drugs worsens the lives of most addicts.
200
Richard J. Arneson
I do not mean to enter into the complex issue of whether a coercive ban on the use of certain recreational drugs would be good law in the actual circumstances of a particular society. My point is that it might well be, and that the assertion of the special responsibility of each individual to fashion her own good life does not tell against the proposal. Left to my own devices, I might make bad choices and lead a worthless life. Nudged by some coercive laws and a social culture that discourages some of the worst life choices and encourages a range of better choices, I would live better. The mere fact that I would benefit from state action that promotes the good does not suffice to justify such state action. To decide that question, one would have to weigh the overall costs and benefits-some state policies that would benefit me and others like me might exact too high a cost from other people and be indefensible all things considered. But just as the special responsibility of parents to care for their own children is fully compatible with there being a strong obligation on the part of society to provide for the care and nurturance of children if parents fail in their responsibilities, and equally compatible with there being a strong obligation on the part of society to supplement the efforts even of responsible parents to ensure that all children enjoy a satisfactory level of care, the special responsibility of the individual to fashion her own good life is fully compatible with the responsibility of other people to boost her prospects of attaining the good. The self-defeating argument, once placed under scrutiny, fares no better than the responsibility argument. It does not get us anywhere near the neutrality doctrine. Many elements of the good life of an individual require intentional pursuit of that element by that very individual. But nothing so far rules out the possibility that nonneutral state prompting might promote individuals’ intentionally acting in ways that achieve important goods to a greater extent than would have occurred absent the state prompting. The good of skiing is not achieved unless putative skiers intentionally move their limbs in certain ways, but a state policy that subsidizes skiing and so encourages people to ski rather than to watch television might still succeed in its own terms. The attainment of goods of friendship and love involve the formation of subtle mental states and delicate patterns of reciprocal behavior, so that a third party can do little directly to orchestrate the actions that transform acquaintances into friends and lovers. But a third party can be a matchmaker, or more generally can act so as to place people in situations that render the formation of friendship and love more likely. A society, dismayed by the outcomes of individual romantic choice, might deliberately introduce the custom of arranged marriages and arranged friendships. This might well be bad policy for many reasons, but there is no reason to think it is inherently self-defeating. Genuine love can develop between Harry and Sally even if
“Liberal Neutrality on the Good:An Autopsy”
20 1
they are thrown together against their will by practices fostered by nonneutral state policy.” 4. Dworkin’s Views
In many situations, it makes sense to think that X is valuable but that forcing or manipulating a person to get X does not enhance the person’s life. In part, the disvalue of the forcing or manipulating outweighs whatever value X has. Also, the forcing or manipulating taints X and lowers its value, perhaps to nothing or less than nothing. Dworkin tries to extend these claims. He writes, “my life cannot be better for me in virtue of some feature or component I think has no value.”‘*Call this the “endorsement constraint.” Nothing is noninstrumentally good for a person unless that very person endorses it.” By itself, the endorsement constraint does not rule out strong paternalism, restricting a person’s liberty against her will for her own good in order to attain a goal the person has not adopted. I might value X, but not enough to warrant adopting X as a goal, and someone might think I am underestimating the value of X and force it on me. The endorsement constraint is satisfied here. Also, by itself, the endorsement constraint does not rule out any paternalism that coerces one to attain something for one’s own life that one now finds valueless provided the coercion brings it about that one comes to value the imposed good. The endorsement constraint can be interpreted so it is quite weak, if any endorsement however faint counts as endorsement and if any endorsement of a feature or component of one’s life at any time in one’s life suffices to satisfy the constraint. The “any-time any-amount endorsement constraint” would allow that if I believed for one moment as a small child that birthday parties have slight value, then bringing it about that I get a birthday party any time in my life can enhance the value of my life even if I loathe birthday parties passionately and steadily after that one childhood moment. To restore the endorsement constraint as a normative bulwark against paternalism, first note that endorsement must be a genuine doing of the endorsing individual, not a state that one falls into through trickery or coercion or manipulation. If Simon Legree paternalistically forces Tom to eat strawberry jam and intimidates him into liking it, the endorsement constraint is not satisfied. Next, require that the endorsement must be simultaneous or retrospective. One might also construe endorsement as a sliding scale constraint, so that no feature of one’s life can have value unless one values it, and no feature can have greater value than the value the person herself attributes to it.
202
Richard J. Arneson
Dworkin pushes the endorsement constraint idea by asserting “the priority of ethical integrit~.”’~ Suppose that the ways that a particular person might live can be ranked from best to worst objectively. The priority of ethical integrity insists that the value of living any of these lives is only fully achieved if the life satisfies integrity. Moreover, no life a person might live that fails to achieve integrity, no matter how wonderful it might be in all other respects, is a better life than any life she might live that does achieve integrity. In the overall ranking of ways a person might live her life, the value of ethical integrity takes strict priority. Dworkin asserts, “Someone has achieved ethical integrity, we may say, when he lives out of the conviction that his life, in its central features, is an appropriate one, that no other life he might live would be a plainly better response to the parameters of his ethical situation rightly judged.”15 When someone contemplates acting to enhance the quality of another person’s life, that person’s settled convictions about what sort of life would be best for her to live strongly limit the possibilities that coercion or manipulation or inducement might improve the person’s life. The priority of ethical integrity proposes that one cannot improve a person’s life by bringing it about, that she leads a life that lacks ethical integrity if the life she would have lived absent one’s interference would have satisfied ethical integrity. Dworkin situates the endorsement constraint and the priority of ethical integrity within a discussion of a way of conceiving of the good life that he calls the “challenge model” and believes to be widely embraced in contemporary societies.16The challenge model regards a good life as a skillful or admirable performance that is produced in response to the challenge posed by the fact that one has a life to live. Of course, an individual could respond in an admirable way to the challenge posed by horrific life conditions without living what anyone would regard as a life that was good for him. So the account is qualified by the stipulation that a good life is an admirable performance that is a response to a challenge constituted by favorable life conditions. Dworkin introduces an important distinction between two kinds of life conditions, “limits” and parameter^."^^ For an example of the distinction at work, suppose you are born with enormous potential artistic talent and a strong propensity to form artistic ambitions. These life circumstances might plausibly be viewed not as features that limit the extent to which one can achieve a good life but rather as features that partly define the challenge of one’s life, responding well to which would be, for one, living well. In contrast, such life circumstances as that one suffers from chronic allergies that make it difficult to practice art and that one is impoverished and has no money for paint brush and easel are limits. A further complexity in Dworkin’s challenge model of the good life is that to some degree it is up to the individual to decide which features of her life are to count as challenge-defining parameters and which are
“Liberal Neutrality on the Good: An Autopsy”
203
to count as limits-aids or obstacles that facilitate or hinder her achievement of her good. Living well is in part doing well at this fundamental task of interpreting and thereby constituting the nature of the challenge posed by one’s life circumstances. Another bit of complexity is that some parameters are normative, not simply descriptive. Some parameters just occur or not. It may be partly definitive of my life challenge that I am a Roman Catholic boy from Minnesota, but I could just as well have been a Lutheran boy from Iowa. But there are some features or possible features of an individual’s life that ought to be parameters of it and others that ought not to be parameters. The bearing of this discussion of challenge on the priority of ethical integrity and the impossibility of (certain kinds of) successful paternalistic coercion is supposed to be straightforward. The priority of ethical integrity does not straightforwardly rule out coercive paternalism because it is conceivable that one might force someone to sample a good that they then come to appreciate and to value for its merits, this good then being integrated in a life of integrity that is objectively better than what the individual would have had without suffering the coercion. Also, in some cases, a person is not leading a life of ethical integrity, and would not do so absent paternalistic coercion, so the priority of ethical integrity does not in that sort of case rule out paternalism as inadmissible. It might even turn out that via coercive paternalism a person comes to be pushed toward a way of life that she comes to value and affirm as best for her, whereas without the paternalism she would have led her life drifting from one set of goals to another without really affirming and endorsing the goals she seeks. Nonetheless, the priority of ethical integrity united to the challenge model of the good life is a strong bar to many kinds of what might have seemed attractive paternalism. If Smith values the life of religious monasticism, one cannot improve her life by coercing or manipulating her into a life of politics that she continues to regard as less good than the monastic alternative. Dworkin’s accounts of the challenge model, the endorsement constraint, and the priority of ethical integrity raise many questions. One might for the sake of the argument go along with much that Dworkin says but still resist the claim that a life that lacks ethical integrity cannot be better than one that has it. Recall that Dworkin makes no appeal to the claim that what is really good for a person is subjectively fmed by her preferences and attitudes. An advisor who judges that Smith would lead a better life if she were wholeheartedly devoted to politics than she would if she were wholeheartedly devoted to monasticism may be correct, Dworkin concedes. So it might seem that if we load up one life with large amounts of objectively great goods, but stipulate that the person leading the life underestimates these goods and thinks an alternative life she might have lived would have been better, and contrast it with
204
Richard I. Arneson
another life for that person in which through overestimation of the quality of the goods she gets, the person achieves ethical integrity, but imagines the actual character of the life as progressively worse, at some point we should say that the achievement of ethical integrity in the second life is outweighed by the superior quality of the goods achieved in the first, so the first life is overall better, even though it lacks ethical integrity. This line of doubt can perhaps be stretched further. Dworkin presents ethical integrity as though one either has it or lacks it, period, but one might insist that ethical integrity as Dworkin conceives it must come in degrees. One can be more or less confident that the life one leads is better than any alternative one might have led. One can believe that the life one leads is better by a long or by a short chalk mark than the next-best life one might have led. If one does not believe the life one is leading is the best among the alternatives, one might be more or less confident of this judgment, and one might believe one’s actual life falls short of the best by a smaller or greater amount. One can still assert that these various judgments of degree are cut by a sharp line that separates having ethical integrity and lacking it. One might say that if one is just barely more inclined than not to accept the judgment that the life one actually is leading is the best one could lead, one achieves ethical integrity. But then the objection suggests itself: Why should the slight difference between suspended belief and a bare tilt toward acceptance of the belief that constitutes ethical integrity make the enormous ethical significance that Dworkin attributes to it? A lot of weight, so far as determining the quality of a person’s life is concerned, is being made to rest on not much. A closely related objection would claim that even if ethical integrity is conceded to be important for the good life, it should not be deemed to have lexical priority over all other constituents of the good life. Dworkin illustrates how (as he sees it) the challenge model of the good life supports the priority of ethical integrity by imagining bad defenses of coercive critical paternalism that try to launch themselves within the challenge framework. For example, one might propose paternalistic legal prohibition of some bad ways of life, in the expectation that with these bad ways no longer being options, people will pick among the remaining better options and will by and large be attracted to the ones they select by a sound appreciation of their good qualities. If this desirable scenario unfolds, the advocate of paternalism holds, paternalism will be compatible with the challenge model of ethics, the endorsement constraint, and the priority of ethical integrity. Dworkin demurs. His ground for objecting is that we bowdlerize the challenge a person faces in his life when we narrow the range of options among which he chooses just in order to prevent his selection of a bad option. This manipulation and narrowing of the choice set of competent adults must make the
“LiberalNeutrality on the Good: An Autopsy”
205
challenge they face a worse one, so we cannot really be improving their lives as the paternalist advocate had hoped. Dworkin puts a further rejoinder in the mouth of the paternalist: If we can bring it about that a person has a better chance of choosing and achieving a truly valuable life if we put in place restrictions on his choice set, why isn’t that improving the quality of the challenge posed by his life? Dworkin responds, “That reply misunderstands the challenge model profoundly. . . . It assumes that we have some standard of what a good life is that transcends the question of what circumstances are appropriate for people deciding how to live, and so can be used in answering that latter question.” But, says Dworkin, the assumption is false. “On the challenge view, living well is responding appropriately to circumstances rightly judged, and that means that the direction of argument must go in the other way.”’* For the case at hand, that means we would have to have “some independent reason for thinking it is better for people to choose in ignorance of lives other people disapprove,” and we conspicuously have no such independent reasons. These objections and replies provided by Dworkin not only improve our understanding of his challenge model of ethics, but also give us good reasons to reject it, so any support that might challenge the endorsement constraint or the priority of ethical integrity will finally evaporate under scrutiny. The problem is that we do have a standard of what makes for a good life that is independent of any specification of a desirable challenge for life circumstances to pose for an individual. A life that has lots of pleasure, especially when this comes by way of enjoyment of what is truly excellent, a life that includes sustained and deep relationships of friendship and love, a life that includes significant achievement in art or culture or systematic scientific understanding, a life that includes significant and sustained meaningful and interesting work-these features of a life inherently make it a better one for the one who lives it. One responds well to the challenge posed by one’s life, so far as prudence (gaining a life that is good for oneself) goes, insofar as one brings it about that one gets more rather than less of these and whatever other objectively valuable goods there are. 5. Skepticism about Knowledge As the Basis for Neutrality
Dworkin’s conception of the neutrality ideal (and related ideals) takes a stand on the nature of the good and deduces that given the nature of the good, a policy of state neutrality concerning it is morally required. Another line of thought that leads to state neutrality begins with claims about moral epistemology.Whatever the nature of human good, the relevant
206
Richard J. Arneson
starting point for thinking about state policy on the good is that claims to knowledge in this domain are inherently uncertain and controversial. The point does not have to be that all claims about what is intrinsically good have this status, just that some do. A popular version of the neutrality constraint construes it as prohibiting only pursuit by the state of policies justifiable (if at all) and only by appeal to a claim that some controversial conception of the good is superior to another. According to this version of neutrality, if in a particular society it is uncontroversial that ice cream is good, state policies favoring the production and consumption of ice cream do not qualify as objectionably nonneutral. It is not plausible to suppose that in the general case, it is wrong for individuals to act in ways whose justification is controversial. Some controversial claims are true, and others are reasonable to believe in the circumstances even if untrue. Faced with a grim epidemic, it may be empirically unclear what strategies of response have a good chance of success, so any state health policy to combat the epidemic will be controversial. But if state officials act for the best as they see it after reasonable attempts to figure out what will ensue if the various candidate epidemic-fighting strategies are pursued, state policy is not faulty just on the ground that it is controversial. Defenders of the neutrality constraint do not rest their case on the claim that no one should ever do what is controversial. The claim they defend is more limited. The neutrality constraint actually defended rules out not any policy based on a controversial justification but those that appeal to controversial conceptions of the good, and moreover not such policies pursued by anyone but only such policies pursued by the state. The neutrality constraint bars the pursuit by the state of policies justifiable only by appeal to the claim that some controversial conception of the good is superior to another. Among defenders of the neutrality constraint, something approaching a consensus has formed around a further limitation on the intended scope of neutrality. The constraint applies not to each and every policy the state pursues, but only to constitutional essentials and basic justice, or the principles that regulate the basic structure of society. Defenders of neutrality deliberately leave open the possibility that the legislation and regulatory policies of a state might be nonneutral but not objectionably nonneutral because they do not concern the basic setup of social arrangements. In this spirit, John Rawls writes, “Our exercise of political power is fully proper 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 rea~on.”’~ Brian Barry asserts, “[Nlobody is to be allowed to assert the superiority of his own conception of the good over those of other people as a rea-
“LiberalNeutrality on the Good: An Autopsy”
207
son for building into the framework for social cooperation special advantages for it.” Explaining the same idea, he states, “at the point where basic principles and rules are being drawn up, no conception of the good should be given a privileged position.”20The contrast being drawn by these authors is between basic rules of society, which are to be justified by neutral principles, and nonbasic rules of society, which need not be. One might also draw the contrast between the constitution that sets the political decision-making rules and particular policies adopted under these rules. Only the former must satisfy the neutrality constraint. Or rather, within a neutral constitutional framework laws are passed by a majority that might suppose its enactments are justified by the superiority of its conception of the good, but what renders these laws acceptable to all is that they became law by way of a fair procedure that is neutrally justifiable. Barry provides a clear statement of the claim of moral epistemology that underpins liberal neutrality: “[Nlo conception of the good can justifiably be held with a degree of certainty that warrants its imposition on those who reject it.”21The position then is that at the level of constitutional essentials, the principles that justify the basic political setup should themselves have a neutral justification-a justification that does not appeal to any claim that one controversial conception of the good is superior to another. The appeal to one’s own controversial conception of the good may be adequate to justify how one conducts one’s private life and also one’s pursuit of public policies (under a neutral constitution) that dovetail with that conception, but it can never justify the basic terms of social cooperation. For the moment, leave aside the limitation of the neutrality constraint to the determination of constitutional essentials. This limitation represents a significant and not clearly motivated concession by neutrality advocates to their opponents. I take up neutrality on constitutional essentials in the next section. The present question is whether skepticism or uncertainty that one’s current conception of the good is correct could justify the position that appeals to such a conception to justify one’s own conduct of life can be acceptable but that appeals to it to justify that coercive uses of state power are morally unacceptable. Disagreement about the good is rife in modern societies. Intelligent, “noncrazy” people show no discernible tendency over time to converge in beliefs about religion or more generally what is valuable and worthwhile in human life. At least, this is so in societies that allow wide freedom of discussion and self-regarding action. Given these evident facts, any individual should be doubtful that her own conception of the good is really superior to what others profess. In this situation, acting on one’s own conception can still be reasonable but imposing it on others coercively must be unreasonable. So goes the argument.
208
Richard J. Arneson
The difficulty with this argument is that shifting from the private action context to public action simply increases the stakes. The possible costs and benefits of action and also of inaction are usually magnified when the issue is public policy rather than how one individual should conduct her own life. But the argument under review does not appeal to blanket skepticism about knowledge of the good or claim that in this domain there is no such thing as knowledge. But if, even knowing that my current beliefs about the good are fallible, it can still be rational all things considered to act on them, then there will be situations in which just that modest degree of confidence in the truth of one’s beliefs about the good will justify coercive uses of state power (if one could get one’s hands on state power). Compare uncertain factual belief. My engineering knowledge is weak, let us say, but the value to me of having a bridge over a small stream on my property renders it reasonable to try to build a bridge despite knowing that errors in my understanding might result in a bridge that collapses. But then there will be some balance of expected costs and benefits such that my fallible engineering knowledge justifies me in extending a public highway via a bridge across a large river despite the risk that the bridge will be badly constructed and collapse owing to some mistaken engineering beliefs I hold. Moving from the private decision to the public decision magnifies the stakes, but this increases the possible costs and benefits of both inaction and action and so cannot uniquely counsel inaction. The same goes for fallible and uncertain ethical beliefs. In the face of pluralism of belief, according to the neutrality advocate, it is supposed to be the case that it can be reasonable for an individual to conduct her life according to her own view of the good even though others disagree, but never reasonable to impose her view on others who disagree.22If one is nonnegligently ignorant either of the existence of others who disagree with one’s views or the grounds for their disagreement, one will not be epistemically at fault in ignoring the challenge their views pose to one’s own. But once one is aware, or should be, that others disagree for certain reasons, then one cannot proceed as though there is no challenge. Either it is the case that the others’ objections to one’s views neutralize or defeat them or not. If the latter, then one is reasonable to continue to believe in the superiority of one’s own views. But then one’s own views, rationally deemed superior, are available to justify both the conduct of one’s life and the choice of public policy. If the former, then I do not see how a rational person could continue to insist on the superiority of her own views. If any reasons I have to support Christianity are countered by equally strong reasons another can give to support Hinduism, then I cannot continue to believe Christianity is a uniquely correct doctrine. I should give up my belief in the unique correctness of Christianity and believe
“Liberal Neutrality on the Good: An Autopsy”
209
that either Christianity or Hinduism might be correct or that neither one is. If the arguments for my own view are tied or bettered by arguments for opposing views, I have no basis for trying to bring it about that state power is put behind my view and against the others. But equally I have no basis for regarding my own view as superior and acting on it when what is at issue is how to conduct my own life. 6 . Neutrality on Constitutional Essentials
Some theorists claim that neutrality on the good becomes more plausible and attractive if the scope of required neutrality is limited to basic matters-to constitutional essentials or to matters of basic justice.23Is this so? In the absence of an account that specifies what is constitutionally essential or falls under the heading of basic justice, it is hard to answer this question decisively. But there are grounds that tilt toward a negative answer. Neutrality on constitutional essentials is an unstable position. If there are good grounds for neutrality, they extend beyond constitutional matters, and if there are good grounds for adopting nonneutral policies, they press beyond legislative policies to constitutional essentials or alternatively beyond policies concerning nonbasic matters to policies affecting basic justice. To illustrate this possibility, suppose that a majority of religious fundamentalists might vote for a local public school curriculum that entrenches a conservative view of the value of sexual expression. On this view, sex outside of marriage is valueless and so is nonheterosexual sexual activity. If the content of a local public school curriculum is regarded as left open by constitutional essentials and basic justice, then the neutrality doctrine, applying only to the essentials and the basics, does not condemn this public policy specifying the school curriculum. But if neutrality ever looks plausible, this sort of case should be its home ground. In the example imagined, some citizens are putting state power behind a conception of the good that is sectarian in the sense that many other citizens will reject it and be perfectly reasonable in doing so. Surely, this sort of thing should qualify as objectionably nonneutral if anything does. The neutrality advocate who favors restricting its scope to the essentials and the basics may have in mind the thought that beyond this restricted sphere it is acceptable for those who favor different views of the good to compete in the democratic process to try to enact policies they favor. Those who win will have won in a fair democratic competition. But this appeal to democratic process is a red herring. Let us assume that democracy suitably defined is a fair political procedure and that democratically
210
Richard I. Arneson
enacted policies are procedurally fair. This leaves it entirely open whether democratically enacted state policies are substantively fair or just. Neutrality is supposed to be a constituent of a substantive conception of the just state. Some who endorse restricting the scope of neutrality to the essentials and basics suggest that no coherent system of legislation could conform to neut r a l i t ~On . ~ ~this view, restricted neutrality is achievable; neutrality with expanded scope is not. But this claim is false. Neutrality with expanded scope is surely feasible. A government might limit itself to the enforcement of Lockean libertarian rights, which are taken to be morally fundamental. Such a government would not be in the business of pursuing policies that could only be justified by the claim that some controversial conception of the good is superior to another. It might or might not be desirable for a government to refrain from basing its policies on controversial views about the good, but it is feasible. Barry writes that “decisions about what the publicly run schools are going to teach must obviously involve a view about the value of learning some things rather than others,” and that for this reason “it would be absurd to suggest that there is some way of determining a curriculum that is neutral between all conceptions of the good.”25The setting of a public school curriculum is proposed as just one example of the impossibility of requiring that not only constitutional essentials but all legislation and public policy be neutral on the good. But what Barry suggests is absurd seems to me entirely possible. Given that what we have in mind is not neutrality of effect but neutrality of justification, we can fix a school curriculum by appealing only to neutral conceptions of people’s individual rights coupled with uncontroversial ideas of the good. If everyone agrees that basic literacy and mathematical competency are good, we can appeal to the idea that it is fair that every person have fair opportunity to attain some reasonable threshold level of literacy and mathematical competence, and run public schools on this basis. Barry might be intending to assert not that strict adherence to the neutrality ideal in the formation of public policy would be impossible but that its predictable results would be very undesirable. With this claim, I would of course have no quarrel. The question would then be whether this is the opening wedge toward admitting that the neutrality ideal should be rejected. We return to the question of why neutrality on the good might be deemed desirable, but only when limited to the essentials and basics. Suppose that constitutional essentials include a specially protected set of rights that are regarded as morally important, likely to be threatened by majority rule in some circumstances when the rights pinch majority interests, and effectively protectable if insulated to some extent against the immediate will of the majority of voters. I see no principled objection against including among the set of
“LiberalNeutrality on the Good: An Autopsy”
211
constitutionally protected rights rights defined in terms of controversial conceptions of the good. For example, each citizen might be guaranteed a right to education for autonomy: Each citizen has a right to an education that induces in her the ability and disposition to live according to values endorsed after reflective scrutiny. For another example, a constitution might declare that apart from instances of unrelievable pain, anguish, or loss of intelligent consciousness, continued human life is always a good opportunity for an individual, even if the individual thinks otherwise, so there should be no general legal entitlement to suicide at the individual’s discretion. I do not claim that it would necessarily be a good idea to entrench autonomy or the value of life as a constitutional right. A constitution is supposed to be an instrument for securing several moral aims over the long run in the context of a particular people or group of peoples with a particular history. The appropriate aims that a constitution should serve might also vary with the historical context. What provisions should go into a constitution in a particular setting will obviously depend on complex arguments that vary from case to case. The exploration of this topic lies beyond the scope of this chapter. My suggestion is merely that if it is thought legitimate that a constitution should give special protection to certain individual rights that unchecked majority rule (or unchecked minority rule for that matter) is likely to threaten, then some of these individual rights may be rights to controversial aspects of good. Nothing per se in the shift from ordinary legislative and administrative politics to constitutional politics supplies any reason to eschew controversial claims about human good. 7. The Principle of Liberal Legitimacy
Neutrality of justification can be regarded as a special case of a broader principle that has been called the principle of “liberal legitimacy? A legitimate government is one that does not act toward its members in ways that can be justified (if at all) only by appeal to principles that someone could reasonably reject. The policies pursued by a legitimate government have justifications available to any reasonable member of society. Neutrality of justification is weaker than liberal legitimacy. Some policies forbidden by the latter are not forbidden by the former. For example, a government might impose on its citizens a conception of distributive justice that is controversial and reasonably rejectable. But the appeal is to a controversial conception of justice or fairness, not to a controversial conception of the good, so neutrality of justification does not object. However, any policy that is condemned by neutrality of justification will also be condemned by liberal legitimacy.
212
Richard 1. Arneson
At least, this is so if the terms “controversial”in the statement of neutrality of justification and “reasonably rejectable” in the statement of liberal legitimacy are interpreted as equivalent in extension, so whatever counts as controversial also counts as reasonably rejectable. One can obviously set the reasonable rejectability standard high or low. At one limit, all candidate principles and norms are reasonably rejectable unless they are best or tied for best along the dimension of being supported by morally relevant reasons. If one sets the bar of reasonable rejectability lower, some of these principles that are best supported by reasons will still qualify as reasonably rejectable because someone who is reasonable enough though not fully reasonable would reject them. The weaker the standard of reasonable rejectability, the more demanding is the principle that one should not impose policies on people that are not justified by principles they cannot reasonably reject. A principle is weakly reasonably rejectable when it can be rejected by a person who is being somewhat reasonable, but making some mistakes, reasoning in confused ways, or failing to appreciate some reasons at their true value. But so far as I can see the only version of the liberal legitimacy principle that is acceptable is one that incorporates the notion of maximally reasonable rejectability. If, all things considered, principle X is best supported by relevant moral considerations, why is it wrong to act on the basis of X just because some people would reject it for less than fully adequate reasons? Moreover, those who propose the liberal legitimacy principle incorporating some notion of weakly reasonable rejectability do not attend to the possibility that there are no principles and policies that pass the test of liberal legitimacy so construed. Maybe every principle is rejectable from some weakly reasonable perspective. People sometimes write as though it were obvious that some “lowest common denominator” principles will be acceptable to anyone who is at all reasonable, but this neglects the possibility that acting on the basis of putative lowest common denominator principles is itself objectionable from some weakly reasonable standpoints. The notion of being reasonable and acting on reasons is inherently a maximizing It is not fully reasonable to be moved by some reasons while ignoring or misunderstanding stronger ones. A fully reasonable agent identifies all relevant reasons for action and assigns each its correct weight and acts on the basis of the resultant-what it is most reasonable to do all things considered. Hence, the liberal legitimacy norm either is unacceptable or reduces to the prescription that one should be (maximally) reasonable. This dismissal of liberal legitimacy may be too swift. A strand in liberal theories holds that a legitimate government is one that secures the actual consent of the governed-not their hypothetical ideally rational consent. Expressing
“Liberal Neutrality on the Good: An Autopsy”
213
this idea, Jeremy Waldron asserts that “the liberal individual confronts his social order now, demanding respect for the existing capacities of his autonomy, his reason, and his agency.”28William Nelson interprets political liberalism as committed to the ideal that “principles are adequate only if they should be accepted by all reasonable persons on the basis of beliefs and values they already hold” (rather than on the basis of beliefs and values they would hold after ideal critical reflection with full i n f o r m a t i ~ n )Larmore .~~ suggests that liberal legitimacy interpreted in this way is supported by a Kantian ideal of respect for persons. 8. Equal Respect for Persons
The connection between equal respect for persons and liberal legitimacy is supposed to run as follows: 1. If one uses force or the threat of force against another person in ways (in
accordance with principles) that she could reasonably reject, one uses the other person as a mere means, and fails to treat her with the respect that is due to all persons. 2. One ought always to treat persons with the respect that is due to all persons. 3. One ought never to use force or the threat of force against another person in ways (in accordance with principles) that she could reasonably reject. In this argument, the notion of what one could reasonably reject is to be interpreted as weak reasonable rejectability. Notice that the argument condemns all forcing of persons in ways they could weakly reasonably reject, and not merely forcing people through the agency of the state. The problem with the argument from equal respect is that on no plausible interpretation of the norm of respecting other persons will it turn out to be the case that imposing on persons coercively in the name of principles they “reasonably” reject has to be failing to treat them with the equal respect owed to all persons. Premise 1 in the above argument is incorrect. On nonmaximal notions of reasonable rejectability, one may “reasonably” reject a principle by making a mistake in reasoning or failing to weigh properly the force of the relevant reasons. But one does not exhibit disrespect for a person by treating her in accordance with principles that she actually rejects, but that she would accept if she were fully rational. The principles that fully rational persons would accept may be imputed to all rational agents including
214
Richard
I. Arneson
imperfectly rational agents. When I exercise my practical reason, I seek to find what is truly rational, not just what looks to me to be rational. Interacting with others, I have a right to be treated according to rational principles, the principles most strongly supported by practical reason. If my actual exercises of practical reason go off the track, and I end up affirming, for example, racist principles, you do not treat me with disrespect by treating me in accordance with rational nonracist principles that I actually reject but would accept if I were fully rational. The norm of treating persons with equal respect tends to be associated with the Kantian principle that one should treat humanity, whether in oneself or another, always as an end and never merely as a means. The idea of humanity at the center of this principle is the idea of rational agency capacity, the power to appreciate reasons and to be moved to action by one’s judgment of the balance of reasons impinging on one’s decision. To respect a person is to abide by the principle of humanity and thus to respect her rational agency capacity. But again, the question arises, what does respect for rational agency capacity require? Notice that treating people with respect for their rational agency cannot require treating them always only according to principles they actually accept, because one might face a situation where one’s action will inevitably affect two people and they accept conflicting principles, so that treating A according to the principles she accepts is ips0 facto treating B according to principles he does not accept. But one might hold that one should so far as is possible always treat people only according to principles they actually accept. Consider cases in which an agent must act in some way or other that will affect persons A and B and all parties correctly agree that the case falls under moral duty: the agent is duty-bound to treat A and B in a particular way. A, B, and the agent disagree as to what way this is, what moral duty requires in these circumstances. Suppose the agent is correct in her judgment of what duty requires, and A and B are both incorrect. Then it would not really accord with the will of A and B if the agent alters her course of action away from what duty requires and toward what A and B incorrectly think duty requires. As moral agents, A and B want action to be done that actually accords with moral requirements, not action that accords with their opinions about what is morally required if those opinions are wrong. (Suppose this is not so. A and B want above all to be treated according to morality as they see it, regardless of what moral requirements as they bear on the case really specify. Then what A and B want is not morally acceptable and failing to treat them as they want, in case morality as they see it diverges from objective moral requirements, is not plausibly regarded as failing to accord them the respect that morality demands.)
“LiberalNeutrality on the Good: An Autopsy”
215
9. Conclusion
I have argued that sound conceptions of human good, even if they are controversial within a society, contribute nonredundantly to the justification of the state policies and constitutional essentials that ought to be followed in any given society. This argument might reduce to paper shuffling for all practical purposes if no sound yet controversial conceptions of human good can be identified. Identifying such ideas of human good and defending them is the topic for another chapter. In conclusion, I simply want to point toward the kinds of claims that would fill this bill. Advocates of the neutrality doctrine often seem to have in mind state establishment of religion as the paradigm of nonneutral state policy. In this way of thinking, conceptions of good tend to be either controversial religious doctrines such as Roman Catholicism or Hinduism or controversial philosophical theses about the nature of good such as hedonism (pleasure and pleasure alone is intrinsically good) or narrow perfectionism (what is intrinsically good is developing to a high degree the essential human capacities, the properties that make humans human). But the opponent of neutrality can readily concede that there are many claims about human good that are contentious and uncertain and unlikely to figure in any successful argument concerning what state policy should be. To defend nonneutrality in practice one must only hold that there are some sound yet controversial conceptions of good that should figure in justification. The most likely candidates are common-sense and perhaps humdrum notions. Consider this list of candidate human goods: 1. 2. 3. 4. 5. 6. 7. 8.
Pleasurable experience and especially enjoyment of the excellent Satisfaction of reasonable life aims Relationships of friendship and love Intellectual and cultural achievement Meaningful work Athletic excellence Living one’s life according to autonomously embraced values and norms Systematic understanding of the causal structure of the world
The claim is that these are objective human goods and that the more of them an individual gains over the course of her life, the better her life goes for her.30 The state ought to promote these goods, or rather some function of these goods that political morality selects as fair and right. Moreover, the full moral justification of the policies the state ought to pursue will incorporate
216
Richard J. Arneson
the premise that these putative goods are genuine, intrinsic goods. Hence, state policy should be nonneutral. It will be objected that the conception just described is too banal and indeterminate. At this high level of abstraction, it is uncontroversial that the items o n the list are genuinely good. But at more fine-grained levels of description, these putative goods become genuinely controversial. To serve as nontrivial guides to policy, one would need to have on hand principles that weight these goods against each other and assign them comparative values.31But reasonable people will disagree about these matters. In response: I submit that something in the neighborhood of the proposed list of goods is sound yet controversial, and that even in the absence of quantitative comparative ranking principles, the conception amounts to a substantive component of the justification of morally acceptable state policies. Regarding comparison and measurement, one should expect to find partial commensurability. For many possible mixtures of goods, there may be no fact of the matter as to which package is best. But acceptance of partial commensurability is not acceptance of neutrality; quite the contrary.
Notes 1. Ronald Dworkin, “Liberalism,”in A Matter of Principle (Cambridge,Mass.: Harvard University Press, 1985), 181-204. 2. Thomas Hurka, review of Beyond Neutrality: Perfectionism and Politics, by George Sher, Ethics 109, no. 1 (October 1998): 190. The last sentence of the book reads: “As one finishes his [Sher’s] book it is hard not to believe that the period of neutralist liberalism is now over.” 3. Dworkin, “Liberalism,”191. 4. John Stuart Mill, “On Liberty,” in Collected Works, ed. J. M. Robson, vol. 17 (Toronto: University of Toronto Press, 1963). John Skorupski defends a nineteenthcentury, broadly Millian conception of the relation between the theory of the good and political morality against contemporary doctrines of neutrality on the good in his Ethical Explorations (Oxford Oxford University Press, 1999),chapters 9-1 1. 5. The versions are distinguished in Charles Larmore, Patterns of Moral Complexity (Cambridge: Cambridge University Press, 1987); see also Richard Arneson, “Neutrality and Utility,” Canadian Journal of Philosophy 20, no. 2 (June 1990):215-40. 6. Joseph Raz tends to think that neutrality should incorporate neutrality of effect, but he is a critic not a supporter of neutrality. See Joseph Raz, “Facing Diversity: The Case of Epistemic Abstinence,” Philosophy and Public Afluirs 19, no. 1 (1990): 3-46. 7. Although neutrality of effect is certainly not the sort of neutrality proneutralists favor and should not be the target of antineutralist attack, theorists often slip into the posture of assuming that neutrality of effect is the core commitment that neutrality
“LiberalNeutrality on the Good: An Autopsy”
217
advocates should make. For example, in his book on multicultural citizenship Will Kymlicka argues that the policy of separation of ethnicity and state, modeled on the policy of separation of church and state, is incoherent. His idea is that while it is possible for the state to be neutral on matters of religion, it cannot in a similar way be neutral in matters of ethnicity, because government business must be conducted in some language or other, and any choice inevitably nonneutrally favors the ethnic members of society who share that language and disfavors those ethnicities who speak other languages. When the government carries out its ordinary functions, it inevitably gives a boost to some group of language users in the society, so in this way (and perhaps in others as well) the state cannot be neutral on ethnicity as it can perhaps be neutral on religious matters. What sort of neutrality am I talking about here? Suppose Canada contains French and English speakers and a decision is made that government business will be conducted solely in English, the language of the majority. The grounds for the decision are to save the administrative costs of conducting public business simultaneously in two languages and to encourage all Canadian citizens to learn English so that there is a language, shared by all, in which national democratic discussion can be conducted. This may be good or bad, fair or unfair, policy-that is not at issue. The stated policy on language does not aim to promote English, so neutrality of aim is not violated. Even more clearly, the justifications for the policy do not appeal to any judgment that one ethnicity or culture or linguistic group in society is superior to any others, so neutrality of justification is not violated. If one is to claim that the stated policy is nonneutral, one must have in mind neutrality of effect as the pertinent neutrality ideal. 8. This quote is from Thomas Nagel’s characterization of Rawls’s view in “Rawls on Justice,” in Reading Rawls: Critical Studies of ‘2 Theory of Justice,” ed. Norman Daniels (New York Basic, 1975), 9. 9. Larmore, Patterns of Moral Complexity, 67. 10. Dworkin, “Liberalism,” 191. 11. The responsibility argument and the self-defeating argument are criticized in George Sher, Beyond Neutrality: Perfectionism and Politics (Cambridge: Cambridge University Press, 1997), chapters 3-4; see also Steven Wall, Liberalism, Perfectionism and Restraint (Cambridge: Cambridge University Press, 1998), chapters 6-8. 12. Ronald Dworkin, Sovereign Virtue: The Theory and Practice of Equality (Cambridge, Mass.: Harvard University Press, ZOOO), 268. 13. Good criticisms of the endorsement constraint deployed as an argument for liberal neutrality on the good are made in Thomas Hurka, “Indirect Perfectionism: Kymlicka on Liberal Neutrality,” Journal of Political Philosophy 3, no. 1 (March 1995): 36-57. I discuss the endorsement constraint in Richard Ameson, “Human Flourishing versus Desire Satisfaction,” Social Philosophy and Policy 16, no. 1 (Winter 1999): 113-42. Dworkin’s views on this topic appeared earlier in his “Foundations of Liberal Equality,” in The Tanner Lectures on Human Values, vol. 11, ed. Grethe B. Peterson (Salt Lake City: University of Utah Press, 1990), 1-1 19. For criticism, see Richard Ameson, “Cracked Foundations of Liberal Equality” (forthcoming). 14. Dworkin, Sovereign Virtue, 270. 15. Ibid.
218
Richard J. Arneson
16. On the challenge model of ethics, see Dworkin, Sovereign Virtue, 250-54. 17. Ibid., 260-63. 18. Ibid., 273. 19. John Rawls, Political Liberalism, rev. ed. (New York Columbia University Press, 1996), 137. 20. Brian Barry, Justice As Impartiality (Oxford: Clarendon, 1995), 160 Charles Larmore, “The Moral Basis of Political Liberalism,” Journal OfPhilosophy 96, no. 12 (December 1999): 599-625; and Charles Larmore, The Morals ofModernity (Cambridge: Cambridge University Press, 1996), chapters 6-7. 21. Barry, Justice As Impartiality, 169. For criticism, see Richard Arneson, “The Priority of the Right over the Good Rides Again,” Ethics 108, no. 1 (October 1997): 169-96. 22. For a discussion, see Richard Arneson, “Rawls versus Utilitarianism in the Light of Political Liberalism,” in The Idea ofa Political Liberalism: Essays on Rawls, ed. Clark Wolf and Victoria Davion (Lanham, Md.: Rowman & Littlefield, 2000). 23. For an argument for this view, see Peter de Marneffe, “Liberalism, Liberty and Neutrality,” Philosophy and Public Afairs 19, no. 3 (summer 1990): 253-74. 24. See Barry, Justice As Impartiality, 161. 25. Ibid. 26. For criticism of the liberal legitimacy principle deployed to support the neutrality doctrine, see Wall, Liberalism, Perfectionism and Restraint, chapters 3-5; Sher, Beyond Neutrality, chapter 6; and Joseph Chan, “Legitimacy, Unanimity, and Perfectionism,” Philosophy and Public Afairs 29, no. 1 (winter 2000): 5-42. 27. A maximizing ideal does not presuppose full commensurability. With respect to a standard of choice and a set of alternatives, one maximizes if one chooses an alternative that is not worse than any other alternative according to the standard. One can do this even if some choice is no better than some other, not worse than this other, and not exactly as good as the other. 28. Jeremy Waldron, “Theoretical Foundations of Liberalism,” in Liberal Rights: Collected Papers 1981-1991 (Cambridge: Cambridge University Press, 1993), 62. 29. William Nelson, “Liberal Theories and Their Critics,” in The Blackwell Guide to Social and Political Philosophy, ed. Robert L. Simon (Malden, Mass.: Blackwell, 2002), 212. 30. Similar lists of objective goods are proposed by Derek Parfit, “Appendix: What Makes Someone’s Life Go Best?” in Reasons and Persons (Oxford: Oxford University Press, 1984); Sher, Beyond Neutrality, chapter 9; and Martha Nussbaum, Women and Human Development: The Capabilities Approach (Cambridge: Cambridge University Press, 2000), 78-80. (Nussbaum proposes a list of freedoms or capabilities for the good, which should be secured for all individuals. Whether or not the capabilities are exercised or fulfilled is not, according to her, the business of the state.) 31. Chan has an interesting discussion of this issue in “Legitimacy, Unanimity, and Perfectionism,” 10-20.
13 “Freedom of Expression in the Non-Neutral State’’ George Sher
In this paper, Sher seeks to reconcile state perfectionism with a strong commitment to free speech. The reconciliation is needed, since it is possible that in order to promote the good effectively a state may need to suppress certain dangerous ideas. Moreover, as Sher explains, a number of the most influential arguments for free speech can be extended so that they tell (or appear to tell) against state perfectionism. In response, Sher argues that there are important disanalogies between speech and action that explain why these arguments support free speech, but do not undermine state perfectionism. Principal among them is the idea that we learn more from what others say than from what they do and that restrictions on speech undermine our autonomy to a greater extent than restrictions on actions.
- 219 -
220
George Sher
I
N THIS PAPER, I WILL discuss the
relation between two principles which figure prominently in much recent political thought. The two principles are, first, that governments may legitimately use their powers, including their formidable power to coerce, to promote activities and traits that are good; and, second, that citizens should enjoy (nearly) absolute liberty to say, and to hear said, whatever they choose. Although the two principles are obviously in some tension-it is at least conceivable that in order to promote the good effectively, a government may have to suppress certain seductive but dangerous ideas-I am quite sympathetic to each. Thus, in what follows, I want to explore the prospects for reconciling them. Because the first principle endorses what is in one sense a perfectionist state, I shall refer to it as P. The second, which calls for freedom of expression, will be designated FE.
I The first point that needs to be made is that P and FE are not flatly inconsistent. They do not contradict each other because P is a claim about the legitimacy of certain reusons for governmental action while FE is a claim about the illegitimacy of a certain type of government act. Because P and FE are thus cross-cutting, we can without contradiction accept them both. To do so, we need only insist that what is objectionable about the use of censorship to promote the good is not its goal but only its method of achieving that goal. Nevertheless, despite their formal consistency, P and FE do not sit well together; for once we acknowledge that the state may legitimately interfere with nonexpressive activities to promote the good, we are bound to wonder why expressive activities should enjoy a more protected status. This question is hard to answer because many prominent arguments against censorship either have in fact been taken to tell also against P or else plausibly could be so taken. That 5ome important arguments for FE are also arguments against P is hardly a surprise. Ronald Dworkin, for example, is well known for defending FE by appealing to a right to equal concern and respect that requires “that government. . . be neutral on what might be called the question of the good life”’-a requirement that is itself a simple denial of P. What is more striking, however, is that the same point holds for many other important defenses of FE. To bring this out, I shall briefly review, in order, the consequentialist arguments of John Stuart Mill, the autonomy-based argument of Thomas Scanlon, and the explicitly political argument of Alexander Meiklejohn. 1. Consider, first, Mill’s famous defense of FE in On Liberty. In that work, Mill argues that we all benefit from exposure to the widest possible range of opinions, and that we benefit most from exposure to the very opinions that
“Freedom of Expression in the Non-Neutral State”
22 1
are most unpopular and with which we most strongly disagree. If an unpopular opinion is true and we accept it, we gain new knowledge; if it is false and we reject it, we gain from having to think through our opposition to it; if it contains an element of truth and we come to recognize this, we gain a more complete picture of the whole. There are of course also less happy outcomesif an unpopular opinion is false and we come to accept it, we end up knowing less than we did at the outset-but Mill appears to view such outcomes as relatively infrequent. He also appears to believe that governments (or those who make decisions on their behalf) are not well positioned to know which opinions, if disseminated, will give rise to false beliefs. For both reasons, Mill rejects selective censorship in favor of a blanket prohibition on the regulation of expressive content. And, for precisely parallel reasons, he also argues that we must reject P. Mill’s parallel argument against P asserts that we all learn from exposure to the widest possible range of ways of living; that we learn the most from those ways of living that we find most unfamiliar and alien; and that these benefits accrue whether the unfamiliar alternatives are wise and constructive or foolish and self-defeating. As Mill himself puts the argument: [I]f [the individual] refrains from molesting others in what concerns them, and merely acts according to his own inclination and judgment in things which concern himself, the same reasons which show that opinions should be free prove also that he should be allowed, without molestation, to carry his opinions into practice at his own cost. That mankind are not infallible;that their truths, for the most part, are only half-truths; that unity of opinion, unless resulting from the fullest and freest comparison of opposite opinions, is not desirable, and diversity not an evil, but a good, until mankind are much more capable than at present of recognizing all sides of the truth, are principles applicable to men’s modes of action not less than to their opiniom2
As the first two clauses of this eloquent passage indicate, Mill does not take the educative effects of different “experiments in living” to justify the state in tolerating activities that harm unwilling others. However, he does take those educative effects to justify a blanket prohibition on governmental attempts to regulate citizens’ lives for other reasons-for example, to prevent them from harming themselves, from living badly, or from influencing impressionable others. Thus, in Mill’s own view, the same considerations that tell for FE tell equally against P. Whether Mill is right to say this, and if so what follows about the prospects for reconciling P and FE, are questions to which I will return. 2. Thomas Scanlon defends FE in his important essay “A Theory of Freedom of Expre~sion.”~ Although Scanlon’s version of FE is similar to Mill’sScanlon, indeed, calls it “the Millian Principle”-the normative grounds on
222
George Sher
which he argues for it are different. Instead of invoking the benefits we all gain from exposure to diverse views, Scanlon appeals to the principle that “the powers of a state are limited to those that citizens could recognize while still regarding themselves as equal, autonomous rational agent^."^ This recognizably Kantian principle is said to imply that the state may not interfere with speech with the aim of preventing citizens from forming beliefs about what is worth doing that (in its view) are false or will lead to harmful action^.^ The reason the Kantian principle has this implication is that anyone who authorized such censorship would thereby consent to being deprived of information that he needs to exercise his own independent judgment. In so doing, he would be making a decision that was inconsistent with his own status as an autonomous rational agent. Unlike Mill, Scanlon does not argue that considerations analogous to the ones he advances in support of FE can also be used to discredit P. However, the fact that Scanlon does not make such an argument does not mean that one cannot be made. Moreover, when we inquire about the possibilities, we find not only that this is possible, but that it can be done in two quite different ways. The first way of extending Scanlon’s argument to tell against P is through a fuller specification of the prerequisites for autonomous agency. Although Scanlon asserts only that no one can function as an autonomous agent without being able to form his own beliefs, it is equally impossible to function autonomously if one lacks the ability to act on one’s beliefs. Because full autonomy requires independence of action as well as thought, it is undermined by the destruction of the necessary conditions for either. Hence, just as someone who authorized his government to regulate its citizens’ beliefs would thereby endorse the abridgment of a necessary condition for his own autonomy, so too would someone who authorized the state to interfere with citizens’ purely personal choices. If rejecting FE would violate the Kantian principle, then so too, albeit for a somewhat different reason, would accepting P. The second way of extending Scanlon’s argument for FE to tell against P is to conjoin his own premise, that having adequate information is a prerequisite for exercising one’s autonomy, with Mill’s claim that exposure to many different ways of living is immensely informative. If Mill’s claim is even roughly right, then anyone who authorized his government to prevent its citizens from living what it considers degraded or unworthy lives would thereby relinquish access to much of the information that he could gain through exposure to diverse “experiments in living.” To agree to be deprived of this information would be to acquiesce in becoming less well equipped to decide among-or, in many cases, even to recognize-the available life options. And, for this reason, Scanlon’s Kantian principle may be said to tell against P not
“Freedom of Expression in the Non-Neutral State”
223
only because agreeing to P would mean relinquishing one’s freedom to act on one’s independent judgments, but also because agreeing to P-like agreeing to censorship-would compromise one’s very ability to reach independent judgments. 3. Scanlon’s defense of FE rests on a principle of political morality that is not tied to any single form of government. By contrast, the third defense of FE that I will consider-the one proposed by Meiklejohn6-is more restricted in its scope. Meiklejohn’s aim is not to justify freedom of expression in the abstract but only to reconstruct the rationale for its protection that is implicit in the First Amendment of the U.S. Constitution. According to Meiklejohn, the fundamental reason our Constitution protects our ability to speak freely is that nothing less than the freedom to discuss all subjects openly can secure the conditions under which the citizens of a democracy like ours can be truly selfgoverning. Meiklejohn’s argument for this contention is straightforward. In a democracy, the ultimate authority for every decision made by government rests with the citizens themselves. Thus, the necessary conditions for a democracy’s effective functioning include whatever conditions are necessary for effective decision making by its citizens. Moreover, for familiar reasons, the latter necessary conditions include open and vigorous debate: When men govern themselves, it is they-and no one else-who must pass judgment upon unwisdom and unfairness and danger. And that means that unwise ideas must have a hearing as well as wise ones, unfair as well as fair, dangerous as well as safe, un-American as well as American. . . .The principle of the freedom of speech springs from the necessities of the program of self-government. It is not a Law of Nature or of Reason in the abstract. It is a deduction from the basic American agreement that public issues shall be decided by universal ~uffrage.~ Like Scanlon, Meiklejohn takes effective decision making to require that those making the decisions be adequately informed; like Mill, he takes their being adequately informed to require that they be exposed to the full range of opinions; unlike either, he invokes both requirements with an eye to grounding FE in the demands of one particular political system. How, exactly, does Meiklejohn’s defense of FE bear on P? In my own book Beyond Neutrality, I advance a similar argument with the aim of reconciling P and FE. I argue, in particular, that if we adopt P, and thus reject the neutralist view that governments may not base their actions on any particular conceptions of the good, then we must replace neutralism with a weaker principle, dubbed “M” for “moderately restrictive,” which “permits us to base political decisions only on conceptions of the good that we have critically scrutinized.”* Because we live in a democracy in which everyone has a voice in the state’s
224
George Sher
decisions, M requires that all citizens be able critically to evaluate each of the many different conceptions of the good on which their government might act. That in turn requires “that everyone have the freest access both to everything that might be said for and against any conception of the good and to all empirical information that might be relevant to such debates. . . . Precisely because it leaves all questions about the good on the public agenda, M requires an especially uncompromising commitment to freedom of thought and expres~ion.”~ By thus appealing to the prerequisites for democratic deliberation, I try in effect to derive FE from P. But the argument I propose is inadequate as it stands. The problem, by now predictable, is that if Mill is right to say that we learn as much from exposure to diverse ways of living as from exposure to diverse ideas, then the same reasoning that shows that democracies cannot deliberate adequately about the good unless their citizens are exposed to the full range of opinions will also have the less welcome implication that democracies cannot deliberate adequately about the good unless their citizens are exposed to the full range of lifestyles. This last implication is troubling because allowing citizens to be exposed to the full range of lifestyles requires tolerating the full range of lifestyles, and that in turn requires rejecting P. Thus, given Meiklejohn’s argument together with Mill’s premise about the educative effects of exposure to diverse experiments in living, we seem compelled to conclude that if a democratic state accepts P, then its citizens must draw on knowledge which they can only acquire if the state rejects P. Far from enabling us to reconcile FE with P, the demands of democratic decision making may therefore actually render P self-defeating. I1
So far, we have seen that three influential arguments for FE can be extended to tell against P. We have also seen that two of the three extended arguments (and one of two variants of the third) turn on an alleged parallel between what we learn from exposure to diverse ideas and what we learn from exposure to diverse ways of living. Thus, to assess those arguments, we must now ask whether that parallel can be sustained. Is there really nothing special about speech?Are governmental attempts to promote the good really as detrimental to informed deliberation as is interference with the free flow of written and spoken ideas? That depends, in part, on how intrusive those attempts to promote the good are. A government like the Taliban, which until recently sought to control all aspects of life along the lines prescribed by a single extreme interpre-
“Freedomof Expression in the Non-Neutral State”
225
tation of a single religion, is bound to confine its citizens to a very narrow spectrum of experience. Because of this, its policies are indeed likely to have the stifling effect on originality against which Mill warned. However, precisely because the Taliban’s policies were so extreme, we cannot assume that what holds for them must also hold for policies adopted against a background of freedom and aimed at restricting only the most egregiously disvaluable or self-destructive of activities. Moreover, when we confine our attention to perfectionist policies of this more modest kind-the only kind that seems possible within a pluralistic democracy-we quickly see that Mill’s point is badly overstated. For, whatever else is true, the ability to deliberate effectively cannot possibly require that one previously has observed the effects of every course of action one might now take. No such requirement can apply because no one has the time or the resources to acquire this broad a range of experience. Given the obvious facts of scarcity, we must, and standardly do, base our assessments of competing lifestyles primarily on other, less direct sources of information -sources that prominently include a historical record that codifies the results of many previous “experiments in living.” Because this record is easily available, the need for new experiments is largely vitiated by our extensive repository of information about old ones. We hardly lack data about the effects of heroin use or the benefits or drawbacks of easy access to divorce: experiments yielding such data have already been performed many times. Because each agent’s options will anyhow outrun his experience,a few more restrictions on what agents can observe seem unlikely to have much impact on the quality of their decisions. Because of this, we need not worry that the arguments that seek to ground FE in the requirements of informed decision making will also show that we must reject P. Indeed, in light of what has been said, the more pressing question is whether those arguments really even support FE. If a certain amount of governmental interference with purely personal behavior is compatible with informed decision making, then won’t a certain amount of governmental interference with speech be compatible with it as well? And, if so, then isn’t the proper conclusion not merely that the arguments for FE do not compel us to reject P, but also that they do not even compel us to accept FE? The answer, I think, is that this is not the proper conclusion, but to see why, we must compare speech and action more carefully than we have yet done. Despite Mill’s attempt to assimilate them, speech and action remain disanalogous in two important respects: first, because foolish actions have costs that false speech does not, and, second, because access to uninhibited discussion contributes far more to our ability to make informed decisions than does access to diverse lifestyles. Once these differences are made clear, we will be able
226
George Sher
to see why each of the cited arguments tells far more strongly for FE than it does against P. Consider a first cost. When someone adopts a lifestyle that destroys his health, causes him to pass up important opportunities, or otherwise worsens his life, any lessons that others learn from his errors are learned at his expense. Given its high human toll, this form of education is hardly optimal. By contrast, the mere act of uttering a proposition-of saying or writing something-is in itself harmless, so knowledge gained by hearing what others have to say is generally not gained at anyone’s expense. Because the lessons learned through observation are thus much costlier than the ones we learn from discussion-because talk is cheap-the value of the former must also be higher if their cost is to be justified. But, far from learning more from what others do than from what they say, we seem in fact to learn far less. One reason we learn less is that experience does not wear its meaning on its sleeve. The truism that observation is only informative when interpreted is relevant not only because interpretation is linguistic through and through, but also because interpretation draws on the full range of linguistic resources. As soon as we move beyond certain obvious claims about harm-that promiscuity promotes AIDS, that drug users become addicts, and so on-our attempts to make sense of what we observe seem bound to take us into realms of imaginative reconstruction that are more akin to literature than to social or natural science. To evaluate any important experiment in living-think, for example, of open marriage, home schooling, telecommuting, the assignment of combat roles to women, or the euthanatizing of the terminally ill-we must look not only beyond anecdotal evidence, but also beyond the endless, earnestly compiled statistics about divorce rates, test scores, productivity, and the like. We must, in addition, articulate to ourselves both how each innovation alters the texture of human experience and how it will transform the structure of meanings in terms of which we understand the other parts of our lives. Because articulations of meaning are delicate, elusive, and indefinitely far-ranging-because they draw on the skills of Henry no less than William James-the ability to produce them is especially vulnerable to limitations on what can be said. And there is also a further reason to hold that we learn far less from what others do than from what they say: namely, that language touches on a vastly greater range of subjects. As Mill’s own phrase suggests, “experiments in living” are primarily tests of unfamiliar activities and social arrangements. Thus, what we gain from them is mainly an enlarged awareness of the range and usual consequences of the options available to us. However, this awareness, though far from trivial, is only a small part of what we need to know in order to engage effectively in either public or private deliberation. We also need,
“Freedom of Expression in the Non-Neutral State”
227
among other things, much specific factual information (about the aims and attitudes of those with whom we must interact, the social and legal background against which we must choose, the resources on which we can draw, and so on); much empirical information of a more abstract nature (about facts and principles of politics, economics, science, technology, and much else); a complex battery of inferential and calculative skills; a firm grasp of what is morally required and what is humanly important; and simple good judgment and a sense of proportion. Because this combination of knowledge and skills requires both a grasp of many abstract propositions and facility at manipulating such propositions, it can only be acquired through the exchange of ideas clothed in language. Given our inability to predict which assertions will provide key pieces of information or forestall crippling misunderstandings, there is again no reasonable prospect of segregating essential from inessential speech on the basis of content. Hence, for this reason, too, restrictions on what can be said seem far more damaging to the prospects for effective deliberation than restrictions on how citizens may live. I11
All in all, the demands of informed deliberation tell far more strongly for FE than they do against P. Because of this, we need not worry either that accepting FE on the basis of those demands requires that we reject P or that accepting P despite those demands requires that we conclude that they fail to support FE. However, while this conclusion may be sufficient to block most anti-P variants of the arguments of Mill, Scanlon, and Meiklejohn, there is one such variant-the first extended Scanlonian argument-which does not rest its case against P on the demands of informed deliberation, and so is not affected by anything said so far. By considering that variant separately, we may hope both to tie up an important loose end and to bring out a further important difference between speech and other forms of behavior. The remaining extended Scanlonian argument, it will be recalled, takes as its point of departure the observation that full autonomy requires independence of action as well as thought. To be autonomous, an agent must be in a position not only to reach his own judgments about what he ought to do, but also to put those judgments into effect. This is said to show that agents who regard themselves as autonomous are no more able to authorize perfectionist restrictions on their conduct than to authorize censorship of what they may hear; for just as authorizing censorship would mean relinquishing access to information one may need to exercise one’s autonomy, authorizing perfectionist restrictions on conduct would mean relinquishing access to options
228
George Sher
one may need to act autonomously. Thus, invoking the Kantian principle that the state may not legitimately exercise powers that citizens who regard themselves as autonomous could not authorize, the extended Scanlonian argument concludes that perfectionist restrictions on conduct are no more legitimate than restrictions on speech. But, on closer inspection, this last attempt to transform an argument for FE into an argument against P is no more successful than its predecessors; for just as the previous arguments went wrong by assuming that we learn as much from what others do as from what they say, the current argument goes wrong by assuming that autonomy is as easily undermined by the absence of specific options as by the absence of specific information. To see what is wrong with this last assumption, we need only remind ourselves of why autonomy is often undermined by the absence of specific information. The reason, surely, is that the knowledge an agent needs to act autonomously is always specific to his choice-situation. To act autonomously is to exercise control over one’s life-to make one’s own choices for one’s own reasons and in one’s own way-and so it requires knowledge of the available options. This does not mean that an agent cannot act autonomously unless he knows everything there is to know about his options-that would set the standards for autonomy impossibly high-but it does mean both that he needs to know a good deal about them and that whatever information he needs must somehow be relevant to them. It is precisely our inability to anticipate what information will be relevant to our options that leads Scanlon to insist that someone who views himself as autonomous can never reasonably agree to censorship. But the options an agent needs to act autonomously are not in any comparable way situation-dependent. Because every choice is made from a menu of options, there is no room for a distinction between a person’s options and his choice-situation. For this reason, there is also no room for the claim that because a person is in a given choice-situation, he needs certain options in order to act autonomously. This of course does not mean that there cannot be situation-independent reasons to view sets of options as inadequate for autonomy: as Joseph Raz observes, an agent may lack autonomy either because he has too few options or because his options are all short-term and inconsequential.1° However, because size and significance are very general features of option-sets, these requirements can be satisfied in many ways. For this reason, Raz is surely right to claim that “while autonomy requires the availability of an adequate range of options, it does not require the presence of any particular option among them.”” Because each agent needs some particular information but no particular options in order to act autonomously, it follows that an agent’s autonomy is
“Freedom of Expression in the Non-Neutral Stute”
229
far more likely to be undermined by his lack of access to some particular item of information than by his inability to exercise some particular option. Given the complexity and unpredictability of the relations between the utterances to which an agent is exposed and what he subsequently knows, this means that even a highly selective policy of censorship can be expected to deprive some agents of some of the information they need in order to act autonomously, but that no comparably selective perfectionist policy can be expected to have any comparable effect on any agent’s autonomy. That in turn means that even if Scanlon is right to insist that no one who regards himself as autonomous can consistently accept any form of state censorship, there is no comparable reason to believe that no one who regards himself as autonomous can consistently accept any perfectionist restrictions on his conduct. This is not to deny that many who regard themselves as autonomous would in fact reject any perfectionist restrictions on their conduct; but it is to insist that this stance is not forced on them by the requirements of autonomy, but is rather a reflection of their substantive preferences or normative commitments.
IV This completes my examination of the possibility that a defense of FE of the kind advanced by Mill, Scanlon, or Meiklejohn might be extended to discredit P. I have tried to show that despite initial appearances, none of the extended arguments show that FE and P cannot be reconciled. Because there are indefinitely many other possible ways of transforming these (and other) defenses of FE into attacks on P, the reconciliation is not final-what reconciliation is?-but amity has at least been temporarily restored. Because the informed-deliberation defenses for FE appear to tell only for freedom of propositional speech, the reconciliation that has been achieved may not extend to versions of FE that are robust enough to protect the various forms of artistic expression that are not propositional; but that is a limitation I am simply willing to accept. In this respect, too, speech may be special.12
Notes 1. Ronald Dworkin, A Mutter of Principle (Cambridge,Mass.: Harvard University Press, 1985), 191. Dworkin says that the right supports liberty of expression in Taking Rights Seriously (Cambridge,Mass.: Harvard University Press, 1977), 277. He discusses various other questions associated with free speech in chapters 17-19 of A Mutter of Principle.
230
George Sher
2. John Stuart Mill, On Liberty (Indianapolis: Hackett, 1978), 53-54. 3. Thomas Scanlon, “A Theory of Freedom of Expression,” Philosophy and Public Affairs 1, no. 2 (winter 1972): 204-26. 4. Ibid., 215. 5. In Scanlon’s own words, what he says is that “[tlhere are certain harms which, although they would not occur but for certain acts of expression, nonetheless cannot be taken as part of a justification for legal restrictions on these acts. These harms are: (a) harms to certain individuals which consist in their coming to have false beliefs as a result of those acts of expression; (b) harmful consequences of acts performed as a result of those acts of expression, where the connection between the acts of expression and the subsequent harmful acts consists merely in the fact that the act of expression led the agents to believe (or increased their tendency to believe) these acts to be worth performing” (Ibid., 213). 6. See Alexander Meiklejohn, Political Freedom (New York Harper, 1948,1960). 7. Ibid.,27. 8. George Sher, Beyond Neutrality: Perfectionism and Politics (Cambridge: Cambridge University Press, 1997), 138. 9. Ibid. 10. Joseph Raz, The Morality of Freedom (Oxford Oxford University Press, 1986), chapter 15. 11. Ibid., 410. 12. I want to thank William Nelson and Alan Wertheimer for helpful comments on an early draft and Emily Fox Gordon for assistance throughout. I owe a special debt to Nomy Arpaly, who at a crucial juncture yanked the paper’s structure straight.
14 “The Structure of Perfectionist Toleration” Steven Wall
In this paper, Wall presents and defends a perfectionist account of toleration. This account relates judgments of toleration to judgments of objective value. As such, it has a distinctive structure, one that distinguishes it from skeptical, subjectivist, and contractualist accounts of toleration. The key to Wall’s account of toleration is the somewhat paradoxical idea that respecting the bad can be a condition for respecting the good. Wall illustrates and defends this idea with two main arguments. The first argument identifies a relatively narrow range of objects that warrant toleration because they are inseparably intertwined with practices or ways of life that are valuable and worthy of respect. The second argument seeks to broaden the range of objects that warrant toleration by identifying habits of thought and dispositions of conduct that need to prevail if the good is to be adequately respected in societies marked by significant social diversity. Taken together, the two arguments dissolve the paradox in the idea that respecting the bad can be a condition of respecting the good. After defending these arguments from possible objections, Wall concludes with a discussion of the limits of perfectionist toleration.
- 231 -
Steven Wall
232
T
OLERATION IS NECESSARY WHEN PERSONS (or
groups of persons) with conflicting beliefs and ideals must live together. In these circumstances,which I shall refer to as the “circumstances of social diversity,”’ each person or group must exercise a measure of restraint toward others if social cooperation is to be achieved and violence is to be avoided. Since the benefits of social cooperation and the avoidance of violence are great goods, toleration is valuable. It is a practice for securing or bringing about these goods. This, in very rough outline, is the classical liberal argument for toleration.2 There is much that this argument gets right; but rather than ask how far it can take us, I want to address a different question. Does toleration have any value over and above the instrumental value assigned to it by this argument? This is an important question, for while the classical liberal argument clearly establishes that toleration has value, its value is conditional on the circumstances of social diversity obtaining. If, in a diverse society, one group gains in strength and numbers, then at some point it may be able to dispense with toleration altogether and enforce its preferred way of life on others. Once the circumstances of social diversity cease to obtain, then toleration no longer will be necessary for securing peaceful social cooperation. More precisely formulated, then, my questions in this paper are: Would toleration have value even if the circumstances of social diversity ceased to or did not obtain? And, if so, what would account for this value? I shall argue that we should answer the first question in the affirmative, and I shall develop a perfectionist account of toleration to explain why we should do SO.^ This account of toleration will establish that toleration has value even when the circumstances of social diversity do not obtain, but it will allow that the purview of justified toleration is conditioned by facts of social diversity. The Dificulty of Toleration The nature and contours of toleration are not uncontested. The understanding of the concept that I will be working with in this paper certainly is not free from contention, but it does not differ radically from how toleration has been generally understood in the recent philosophical literat~re.~ Still, to forestall possible misunderstandings, a few stipulative remarks are in order. The object of toleration, whether it be persons, social practices, behaviors, or beliefs, is an object that is disliked or disapproved To cover both dislike and disapproval, I shall say simply that the object of toleration must be disesteemed. The tolerant person then is one who is disposed not to interfere with an object that he disesteems. Moreover, his disposition not to interfere is a moral disposition. Someone who does not persecute another solely because
“The Structure of Perfectionist Toleration”
233
she believes the persecution would be ineffective is not showing tolerance6toward the other. The tolerant person, in other words, is distinguishable from the efficient perse~utor.~ The former, but not the latter, checks his desire to interfere because he believes he has a moral reason for doing s ~ . ~ This brings into view what is sometimes called “the paradox of t~leration.”~ Tolerance is a morally motivated disposition to refrain from interfering with objects that one disesteems.1° The difficulty is in seeing how this could be a virtue when the object that one disesteems is an object that one rightly disesteems. It is sometimes suggested that value pluralism, or the recognition of value pluralism, can help with the difficulty. For once we recognize that there is not one, but a plurality of fully good ways of life, then we should be able to see that we have reason to respect ways of life (and the forms of behavior that constitute them) that are incompatible with our own. But this thought, appealing as it is, conflicts with the thought that tolerance is called for when one rightly disesteems an object of toleration. For once I recognize that a way of life that I previously held in contempt is in fact one of the many fully good ways of living, then while I may not affirm it, I should cease to disesteem it. In short, the recognition of value pluralism should reduce the number and range of objects that it is reasonable for a person to disesteem. Far from resolving the paradox of toleration, then, the appeal to value pluralism threatens to displace the need for it. Still, even after the implications of the recognition of value pluralism have been fully taken into account, there remains space for toleration. After all, even if one grants that there exist a plurality of fully good ways of life, one need not (and should not) go on to say that nothing-no way of life, no conception of the good, or no form of behavior-is disvaluable and therefore rightly disesteemed. The need for toleration, accordingly, will resurface when one is faced with the prospect of interfering with one of these disvaluable objects. So precisely where toleration is needed, the difficulty of toleration remains with us-how can it be virtuous not to interfere with an object that one rightly disesteems? The Idea of Independent Value As I have already noted, part of the answer to the difficulty lies with the instrumental value of toleration. But this is only part of the answer. Toleration can be valued independently of the consequences it causally brings about. I shall refer to this kind of value as “independent value.” Since independent value is a term of art, it requires clarification. Rather than invoking the more familiar idea of intrinsic value-which has been understood in different ways by different writers-I propose the following formulation for
234
Steven Wall
independent value: a disposition, activity, relationship, object, or event has independent value if it is valuable irrespective of the consequences it brings about or if it is reasonable to value it for its own sake. Both ideas in this formulation have at times been identified with intrinsic value.” But independent value, as I shall understand it here, includes both of them. The disjuncts in the formulation are subtly, but importantly, different. To illustrate the differences between them, it will be helpful to consider a couple of examples. Consider first the example of religious worship. Assume that religious worship is valuable if and only if it is freely engaged in. On this assumption, if religious worship is valuable, then the freedom to engage in it is also valuable, for the latter is a condition of the value of the former. But freedom’s value in this context is not well described in instrumentalist terms. It is not the case that the freedom to engage in religious worship is valuable as a means to realizing the value of religious worship. Rather, its value is a condition-an essential constitutive condition-of the value of religious worship.12This is an instance of independent value on the first disjunct of the previous formulation. Consider now a second example. Assume that a man’s relationship with his boss is valuable to him solely because it will advance his career. Assume further that it is true, and that he knows that it is true, that the best way for him to have the kind of relationship with his boss that will help him is for him to value this relationship for its own sake. Given these facts, while it is true of this relationship that it is not valuable irrespective of its consequences, it is reasonable for John to value the relationship for its own sake.13 The value of this relationship is an instance of independent value on the second disjunct of the previous formulation. How, then, could toleration have independent value? If the disposition to be tolerant were a condition of showing or manifesting respect for objects of value, and if this kind of respect were itself valuable, then the disposition to be tolerant would be valuable. And it would be valuable not as a means to bringing about the respect in question, but rather as a condition for its existence. Likewise, if the disposition to be tolerant were one that it was reasonable for a person to have for its own sake, then even if the disposition were not valuable irrespective of its consequences,then it would still have independent value. Both possibilities, as we shall see, express a more general idea. Toleration is valuable as a way of respecting thegood. Stating this idea in these terms, however, raises the difficulty of toleration in a particularly stark form. How can tolerance be characterized as a disposition to manifest respect for objects of value when tolerance is required precisely when one is confronted with objects that one rightly disesteems? I shall argue that the way out of the difficulty is to view tolerance as a disposition for indirectly manifesting respect for objects of value. Just as some-
“The Structure of Perfectionist Toleration”
235
times one best pursues the good by not aiming directly at it, so too one may best manifest respect for objects of value by not aiming to respect only objects that one deems to be valuable. Before developing this line of argument, however, I want to pause and discuss briefly a different argument for the independent value of toleration. This argument will strike some as the natural way to account for the independent value of toleration. As a way of providing motivation for my own account, I want to call attention to some difficulties it confronts. Autonomy-Derived Toleration The difficulty of toleration centers on the question: How can it be virtuous not to interfere with an object that one rightly disesteems? To this question, an obvious answer suggests itself: We have reason to respect the freedom of people to live as they please, even if they are making poor choices. This is a coherent position, since one can consistently hold both that someone is acting wrongly and that that person has a moral right to do Consider someone who is contemplating suicide for trivial reasons. One can coherently maintain that what the person is contemplating is seriously wrong and that no one has the right to interfere with what he is planning to do. On the assumption that we ought to take the autonomy of persons seriously, a substantial measure of toleration can be justified. In tolerating the wrongdoing of others, we manifest respect for their autonomy. Let us call this “autonomy-derived toleration.” Practicing toleration, on this view, is not first and foremost a means for furthering the autonomy of persons (although it may have this result), but rather a way of respecting it. For this reason, autonomy-derived toleration has independent value. Once again, an example should help to make this clearer. Consider a heroin user. In certain contexts, if we do not interfere with her choices, then her autonomy will be diminished. If we value toleration solely as a means for causally promoting her autonomy, then, at least in these contexts, we have no autonomy-based reason to tolerate her behavior. However, if we value toleration as a way of manifesting respect for her autonomy, then we may have reason to tolerate her choices even when we know that they will have the effect of undermining her autonomy. This view of autonomy assigns independent value to toleration-value over and above the instrumental value it may have in promoting autonomy, My intention is not to refute this account of toleration. Nor do I think that it contains no measure of truth; but I do want to raise a few doubts about its plausibility. According to it, we have reason to tolerate objects that are rightly disesteemed. In doing so, we manifest respect for the autonomy of persons.
236
Steven Wall
But, as the previous discussion suggests, we can value the autonomy of persons even when we do not tolerate the choices they make that we believe are rightly disesteemed. Indeed, our very concern for their autonomy may motivate us to interfere. The point here is simply that valuing autonomy does not commit one to autonomy-derived toleration. Accordingly, a defense of autonomy-derived toleration must do more than point to the value of autonomy. It must show that a proper regard for its value commits one to autonomy-derived toleration. This is harder to demonstrate than it may seem. Autonomy can be understood as an ideal to be promoted or as a right (or set of rights) to be respected. If it is understood as an ideal to be promoted, then it is natural to think of it as one element in a good life. On this view, the autonomy of persons should be promoted because it will, other things being equal, make their lives go better.15 But this view of autonomy substantially restricts the reach of autonomy-derived toleration. For if persons are engaging in an objectively worthless way of life, then respecting their freedom to do so will not likely make their lives go better, particularly if there is reason to believe that interfering with them will help them to pursue a more valuable way of life. Autonomy, understood in these terms, will therefore ground toleration only in those instances where respecting a person’s choice to engage in an objectively worthless way of life is more important-in terms of his own well-being-than in getting him to pursue a more valuable way of life. However, if autonomy is understood as a right (or a set of rights), and if it is further said that respecting this right is always more important than promoting a person’s good, then autonomy-derived toleration will extend even to choices by persons to destroy or ruin their lives. So, in order to get a robust account of toleration, the proponent of autonomy-derived toleration must defend the rights-based understanding of autonomy over the ideal-based understanding. Moreover, she will need to hold that respecting autonomy always (or almost always) takes precedence over promoting or safeguarding the good of persons.16 These claims are not easy to defend. But the proponent of autonomyderived toleration must confront another set of issues as well. Properly understood, autonomy-derived toleration should extend only to ways of life that are autonomously engaged in. These ways of life need not be autonomously chosen. Nor need they be ways of life that exhibit a high degree of independence or active decision making. One can autonomously engage in a way of life, such as joining the army or living in a monastic order, in which many of the decisions about one’s life are not subject to choice. One’s engagement, however, must be one’s own. One must, at least at some point, voluntarily affirm the way of life one is engaging in for it to be an au-
“The Structure of Perfectionist Toleration”
237
tonomous way of life. This gives rise to two problems. First, autonomyderived toleration will not extend to nonautonomous ways of life.I7 This means that if there are cases in which one should tolerate nonautonomous ways of life, then autonomy-derived toleration will not be able to explain why this is so. Second, in order for autonomy-derived toleration to provide practical guidance on what should be tolerated, it will need to be supplemented with an account of voluntary choice or affirmation. The notion of a perfectly voluntary choice is notoriously elusive, and this gives rise to a host of difficulties. I believe we have no cogent and perspicuous account of voluntary choice that can provide clear practical guidance as to which ways of life are sufficiently voluntary to warrant respect on the autonomy-derived account of toleration. l8 I said earlier that I did not intend to refute autonomy-derived toleration. And I certainly have not done so. My remarks are meant to suggest only that this account of toleration faces a number of difficult problems, problems over and above justifying autonomy’s status as a fundamental value. To summarize, they include: 1. Defending the rights-based understanding over the ideal-based understanding of autonomy 2. Defending the relatively stringent priority of respecting the autonomy of persons over promoting their good 3. Proposing and defending a cogent and practically workable account of voluntary choice or affirmation
Perhaps these problems can be successfully met.I9 But, at the very least, they suggest why it is worthwhile to search for an alternative account of the independent value of toleration. Toleration and the Good The key to the account of toleration I shall present is the admittedly paradoxical idea that respecting the bad can be a condition for respecting the good. If sense can be made of this idea, then tolerating objects that are rightly disesteemed may be justified because it is necessary to respect the good. Such a conclusion would provide perfectionist reasons for holding that toleration has independent value. A perfectionist account of toleration relates judgments of toleration to judgments of objective value. Such an account has a distinctive structure, one
Steven Wall
238
that distinguishes it from skeptical, subjectivist, and social contract accounts of toleration. The account of toleration I will present has this structure, but I will not have much to say about which social practices and which ways of life are objectively valuable and why they are objectively valuable. For this reason, my account should be acceptable to those who adhere to different views about the content of the good. A complete perfectionist account of toleration ultimately must propose and defend an account of objective value. But prior to this, perfectionism needs an account of the structure of toleration. This structure will explain how there can be sound reasons for tolerating objects that are rightly disesteemed. The arguments presented here delineate that structure. My account of toleration consists of two main arguments. Both arguments are premised on the normative principle to respect the good. This principle admits of two interpretations. On the first interpretation, the principle holds that we have reason to respect the good, unless not doing so would lead to greater good being promoted. This makes the principle consequentialist; but insofar as we must respect the good in order to promote it, it grounds injunctions to respect the good. By contrast, on the second interpretation, the principle holds that, at least in some cases, we have reason to respect the good, even if not doing so would lead to greater good being promoted. I favor the second interpretation, but will not defend it here.20The two arguments that I shall present are compatible with either interpretation. The first argument identifies a relatively narrow range of objects that should be tolerated because they are inseparably intertwined with practices or ways of life that are valuable and worthy of respect. If successful, this argument dissolves the paradox in the idea that respecting the bad can be a condition for respecting the good. The second argument seeks to broaden the range of objects of toleration justified by the first argument. It does so by identifying dispositions that need to prevail if the good is to be adequately respected in societies that contain significant social diversity. The Basic Argument The first argument I will refer to as the “basic argument.” Stepwise, it proceeds as follows. I. We have reason to respect (i.e., to not hinder, suppress, or destroy) practices or ways of life that are on balance valuable. 11. A practice or way of life is on balance valuable if its valuable features or properties are evaluatively more important than any disvaluable features or properties it may have.
“The Structure of Perfectionist Toleration”
239
111. Some on-balance valuable practices or ways of life are inseparably intertwined with features or properties that are rightly disesteemed. IV. Therefore, we have reason to respect features or properties that are rightly disesteemed if they are inseparably intertwined with practices or ways of life that are on balance valuable. This is a valid argument. But are its premises true? Premise (I) is relatively uncontroversial, and so I shall pass over it. Premises (11) and (111), however, require clarification and defense. Premise (11) holds that some features or properties of a practice or way of life can be “evaluatively more important” than others. We need to explain what this means. Suppose a way of life has two properties, A and B. If property A is evaluatively more important than property B, then its value or disvalue dominates over the value or disvalue of property A. The notion of the value of one property dominating over the value of another is not itself altogether clear. It may suggest a quantitative measure. If the magnitude of the value of property A is greater than the magnitude of the value of property B, then the value of the former will dominate over the value of the later. But the notion of dominance is not filly captured by quantitative measurements. In many cases, it may not be possible to measure the value of different properties along a common scale. In these cases, however, it can still be true that the value of one property dominates over the value of another. For example, sometimes we hold that the value of one property is incommensurably greater than the value of another.21 Likewise, the evaluative importance of the value of a property may depend, in part at least, on the degree to which it is unique. By this I mean the following: if X is a valuable property that is instantiated in only one extant way of life, then X will acquire greater evaluative importance than it would if many Suppose, for example, that a given way of extant ways of life instantiated life realizes a distinctive kind of valuable solidarity between its members. Then, on the view I am proposing, the fact that this solidarity exists only in this one way of life is relevant to assessing its evaluative importance. We may rightly consider this solidarity a realization of a human perfection and we may rightly think that it is important for this perfection to be realized somewhere in the world, even if we do not think that it is particularly important that it be realized in lots of places in the There are likely other factors relevant to assessments of evaluative importance. But the main idea should be reasonably clear. If the valuable properties of a way of life are evaluatively more important than its disvaluable properties, then it is a good thing that it continues to exist. If they are not, then it is not a good thing that it continues to exist.24That is why it is appropriate to say
240
Steven Wall
that a way of life is on balance valuable if its valuable properties are evaluatively more important than its disvaluable ones. It is natural to object that if it is good for a way of life that is on balance valuable to continue to exist, it would be even better for it to continue to exist without its disvaluable properties. Consider the way of life that realizes the distinctive solidarity I mentioned earlier. Suppose this way of life also contains relatively strict gender roles. Now, even if this way of life were on balance valuable, it is reasonable to think that it would be even more valuable if its practices were transformed to overcome this kind of gender discrimination. But while this is reasonable, it might be the case that the strict gender roles are a constituent element of the distinctively valuable solidarity. This brings us to the crucial idea expressed in premise (111) that some disvaluable objects can be “inseparably intertwined” with valuable ones. This idea itself admits of a strong and a weak interpretation. On the strong interpretation, if a disvaluable property is inseparably intertwined with a valuable property, then it will be impossible to remove the disvaluable property without destroying the valuable one. If the distinctive solidarity of a group could not exist without the existence of strict gender roles, then the latter property would be strongly inseparably intertwined with the former. By contrast, on the weak interpretation, if a disvaluable property is inseparably intertwined with a valuable property, then while it may be possible to remove the former without destroying the latter, in practice it will be extremely difficult to do so. Perhaps, for example, the solidarity of the group I have been discussing could be preserved even if its practices were changed so as to eliminate gender discrimination. But even if this were possible, it could remain true that the transformation of these practices would require a long and gradual process of change. Here, the strict gender roles would be weakly inseparably intertwined with the distinctively valuable solidarity. This would mean that until the culture could be transformed those who wished to preserve its valuable properties would need to respect its strict gender divisions. Having now clarified the ambiguous phrases in premises (11) and (111),their truth (or at least plausibility) should be apparent. Nearly all, if not all, cultures and ways of life that we are familiar with contain both valuable and disvaluable properties. Many of them, however, are on balance valuable because their valuable properties dominate over their disvaluable ones. This is sufficient to establish the second premise. It is less clear that many on-balance valuable practices and ways of life contain disvaluable properties that are strongly inseparably intertwined with their valuable properties. Some will reject the idea that a bad property could ever be a constituent element of a larger whole that is good. But this idea is not as strange as it may first appear to be. When reflecting on the virtues of individual persons, it is sometimes clear that the
“The Structure of Perfectionist Toleration”
24 1
virtue of a person is the flip side of a certain deficiency or vice.25For example, a loyal person, by virtue of his loyalty to his friends, may be less inclined to treat strangers fairly when their interests conflict with those of his friends.26 If this holds true for persons, then it is reasonable to think it could hold true for social practices or ways of life. Indeed, my example concerning solidarity and gender division suggests how this could be true. Still, some will reject this possibility. They will claim that the disvaluable properties can be (in principle) always separated from the valuable ones. Yet, even these people should allow that disvaluable properties can be weakly inseparably intertwined with valuable properties. So even they should be able to accept the basic argument. This brings me to a final point. The conclusion of the argument I have presented holds that persons have reason to respect objects that are rightly disesteemed if they are inseparably intertwined with practices or ways of life that are on balance valuable. It is important here to note that to “respect”an object one does not have to value it intrinsically or celebrate its existence for its own sake. By respect, I mean a disposition not to interfere with the object in a way that would damage, suppress, or destroy it. Clearly, if one judges a property to be disvaluable, and if one perceives it to be only weakly inseparably intertwined with the valuable properties of an on-balance valuable way of life, then one will not value it for its own sake. One may think, however, that it would be wrong to damage, suppress, or destroy it. Indeed, one might think that the disvaluable property could only be eliminated from the way of life by the members of the way of life themselves taking steps to eliminate it gradually. One might welcome such a change while at the same time holding that it would be wrong for one to interfere. If so, one might appropriately view one’s disposition to tolerate the disvaluable property as independently valuable, for this disposition manifests respect for the on-balance valuable way of life of which the disvaluable property is but one constituent element. The basic argument, therefore, establishes how we can have reason to tolerate practices that we rightly disesteem. The reasons are derived from our concern to respect the good. The argument shows how (in certain contexts) a condition of respecting the good is a willingness to respect the bad. In this way, the basic argument dissolves the paradox of toleration. The Extended Argument Notwithstanding its merits, the basic argument gives rise to at least two important worries. First, it establishes the reasonableness of toleration at the cost of significantly restricting its scope. The only objects that warrant toleration turn out to be objects that are inseparably intertwined with valuable objects. This implies that people have no reason to tolerate practices and ways of life
242
Steven Wall
that either possess no valuable properties or possess disvaluable properties that dominate over their valuable ones.27Second, the basic argument presupposes a standpoint from which judgments of value can be made. But toleration becomes a pressing matter when different social groups must live together and these groups naturally will assess the value of practices and ways of life differently. In the circumstances of social diversity, there is no common or shared evaluative standpoint for assessing the value of objects of toleration. These worries are not without merit; but they should not be misunderstood. The value of toleration provides persons with moral reasons to show restraint in the face of objects they rightly disesteem. These reasons must be grasped from the first-person standpoint. This is because the problem of explaining the value of toleration is (on the perfectionist account I am proposing), first and foremost, the problem of explaining how a person with sound evaluative beliefs could have reason to respect objects that he rightly disesteems. Strictly speaking, then, it is entirely proper for an account of toleration to begin by focusing on the reasons that persons with sound evaluative beliefs have to respect objects that they rightly disesteem, even if their beliefs are not widely shared in their society. Nonetheless, the worries in question do gesture toward an important feature of toleration of which I have not yet taken adequate account. This is the idea that toleration is more than an individual virtue. It is not just something that we do on our own, but also it can be something that we do together. This idea brings us to the second argument in my account of toleration. This argument broadens the range of toleration by explaining the reasonableness of indirectly respecting the good by complying with a societal ethic that, if generally observed by others, would result in maximal respect for the good in one’s society. The argument holds that, in the circumstances of social diversity, the best way to respect the good is to support certain habits of thought and dispositions of conduct that could be accepted by those who disagree over the content of the good. I shall refer to it as the “extended argument.” Like the basic argument, it can be set out in a series of steps. I. Members of a society have reason to promote and sustain a societal ethic of restraint toward objects that are disesteemed (henceforth to be referred to simply as the “optimal societal ethic” [OSE]). 11. The OSE is an ethic that, if generally observed in the circumstances for which it is proposed, will result in the good being respected as much as it can be in those circumstances. 111. Adherence to the OSE will sometimes require members of a society to respect objects that they rightly disesteem when there is no other reason for them to do so.
“The Structure of Perfectionist Toleration”
243
IV Therefore, members of a society can have reasons to show respect for objects that they rightly disesteem, which are derived entirely from the need to promote and sustain the OSE. This argument consists of premises that require clarification and defense. But if the argument is sound, then it will justify a further measure of toleration for those who live in the circumstances of social diversity. Specifically, for any given person, the argument broadens the range of objects that warrant toleration depending on the strength of the reasons that he has to promote and sustain the OSE in the society in which he lives. The crucial premises in the argument are (I) and (111). Premise (11) clarifies the idea of an OSE, but the truth of premises (I) and (111) yields the conclusion reached in (IV). Why, then, should we accept these premises? I need to say more about the idea of an OSE. An ethic of toleration consists of a set of habits of thought and dispositions of conduct that lead those who have them to show restraint in the face of objects that they disesteem. An OSE of toleration is an ethic of toleration that, if it were internalized by a substantial majority2*of the members of a given society,29would result in the good being respected in that society better than it could be respected if any other societal ethic of toleration were internali~ed.~~ Thus, we can view an ethic of toleration as a device for respecting the good in a society in which there is disagreement over the content of the good. This device would be unnecessary in a society in which there was widespread agreement on the good. It becomes necessary only when the circumstances of social diversity obtain. For this reason, the OSE must be tailored to people as they are. A societal ethic of toleration cannot require that all persons to whom it applies adopt the same beliefs about the good or about which social practices are valuable. While it prescribes habits of thought and dispositions of conduct, the OSE must take as given the fact that those to whom it applies disagree about the good. We might call this the “realism constraint.” Consider now a society composed of different social groups, each of which has a different view about which social practices are valuable and worthy of support. Suppose that in this society there is a range of shared judgments that cut across the differences between the groups, but that there is no consensus on all judgments of value. Now, given these circumstances, one possibility is for the society to have no societal ethic of toleration. Each group could simply work out its own stance on toleration, respecting those practices and ways of life that it held to be valuable and withholding respect from those that it held to be disvaluable. On this possibility, while the differing social groups might share a commitment to certain political and legal institutions, they
244
Steven Wall
would not share a social ethic of toleration. However, it is very likely that this possibility is not optimal. It is very likely that there is some ethic of toleration that would both be capable of being internalized by the different social groups and, if it were generally internalized, would result in the good being better respected than it would be without it. This ethic would respect the realism constraint by not demanding that the different social groups abandon their differences and affirm the same evaluative beliefs. Assuming such an ethic were possible, it would be an optimal societal ethic of toleration. But optimal with respect to whom? Would the OSE need to be seen to be optimal by each social group in the society? Since toleration on the account I am proposing is valuable as a way of showing respect for the good, the OSE must be optimal from the standpoint of those with sound evaluative beliefs. If it were not, then it would not be an optimal ethic for respecting the good. But, at the same time, part of what makes an ethic optimal is that it can be internalized by different social groups in the society. Suppose, for example, that in a given society there is one group whose judgments about what is valuable and what is not perfectly track the truth. This group might propose that an OSE for its society would be simply its own ethic of toleration. This is the ethic of toleration that it would favor if it did not have to take into account any facts about social diversity. That is, it might propose that all other groups simply accept its account of what is worthy of respect and what is not. But it is extremely unlikely that this proposal could be internalized by a sufficient number of members from the other social groups; and, if it could not be, then it would not be an OSE. Its own ethic of toleration, however, could and should serve as the benchmark for assessing possible societal ethics. As a rough approximation, we can say that the OSE for this society would be the societal ethic closest to this group’s own ethic of toleration that could be internalized by a sufficient number of the members of the society to make it effective. Assuming such an ethic were possible, persons in this society would have reason to promote and sustain it. Their reason for doing so would be based on the fact that, in their circumstances, supporting this ethic would be the best way to respect the good.31This is why we should accept premise (I). It also suggests why we should accept premise (111). For, in these circumstances, those who had sound views about what was valuable would have reason to tolerate some objects that they rightly disesteemed for the reason that doing so was necessary to support and sustain the OSE. Since we should accept premises (I) and (111),we should accept the conclusion of the extended argument. My discussion of the extended argument has been fairly abstract. This is appropriate, since I am presently attempting to outline the structure of perfectionist toleration. But before considering objections to the argument, it may
“TheStructure of Perfectionist Toleration”
245
be helpful if I relate it to a concrete case. Consider then the traditional issue of religious toleration. Those with sound evaluative beliefs who live in societies marked by religious pluralism will, in all likelihood, have strong reason to support an ethic of restraint toward all the major religious groups in their society. Doing so will require tolerating some religious practices that are rightly disesteemed, for this may be the best way to ensure that worthwhile religions are respected. Such an ethic, of course, could allow for exceptions. Perhaps, one of the religious groups in the society adheres to practices that are particularly offensive. These practices could be excluded from the ethic of restraint. But exceptions are costly. Too many exceptions will undermine the ethic. In order for the ethic to be effective, it must be capable of being internalized by members throughout the society. This means that the ethic must not only be acceptable to a substantial number of the members of the society, but also that it should not become so complicated that it is difficult to internalize or transmit to the young. Under circumstances of significant religious diversity, these considerations tell strongly in favor of simple codes that do not attempt to make fine distinctions between worthy and unworthy religious practices. Thus, under these circumstances, the need to sustain an optimal ethic of toleration would provide good reason to tolerate a range of religious practices that might rightly be disesteemed. Two Objections The extended argument, at least in the circumstances of social diversity, widens the range of objects that warrant toleration. Like the basic argument, it justifies toleration on the grounds that respecting the bad can be a condition of respecting the good. The argument, however, will likely provoke a number of objections. To further clarify it, I now will discuss and respond to two of them. These are (1) the claim that the extended argument does not show that toleration has independent value, but rather only instrumental value; and (2) the claim that this argument unacceptably makes the scope of toleration depend on the power relations that obtain between different social groups in a given set of circumstances. Let me begin with the first objection, since it is likely to have occurred to the reader during the discussion of the extended argument. I have claimed that independent value can be contrasted with instrumental value in the following way: if something is independently valuable, then it is valuable independently of the consequences that it causally brings about or it is reasonable to value it for its own sake. One way in which a disposition can have independent value is
246
Steven Wall
by appropriately manifesting respect for what is valuable. In discussing the basic argument, I claimed that tolerating the disvaluable properties of an onbalance valuable way of life that are inseparably intertwined with the valuable properties of that way of life can be an appropriate way of manifesting respect for what is good. But even if I am correct in claiming this, it might now be wondered how a disposition to support and sustain the OSE could be a way of manifesting respect for the good. For, it may be said, is not the OSE just a pragmatic device, one that is valuable because adhering to it will have better consequences than not adhering to it? The first thing to be said in response to this objection is that the OSE is not an explicit code. As I have described it, it is a set of habits of thought and dispositions of conduct. These habits and dispositions are those that, if they were to prevail in a given society, would result in the good in the society being respected as well as possible given the kind of society it is. Speaking broadly, the content of these habits and dispositions could be either instrumentalist or noninstrumentalist. If they were instrumentalist, then those who possessed them would value them as a means to the end of respecting the good in their society. But if they were noninstrumentalist, then those who possessed them would value them for their own sake. It should now be clear that the objection we are considering assumes that the OSE is a set of instrumentalist habits and dispositions. This is why the objection concludes that the extended argument fails to show that toleration has independent value. Against this assumption, a strong case can be made that the habits of thought and dispositions of conduct that would comprise the OSE for a society that contained significant social diversity would be noninstrumentalist in content. Given the diversity of assessments of value in such a society, the ethic will be more stable and more likely to function well if it is valued for its own sake. If everyone who adheres to the ethic values it (only) instrumentally, then everyone will be inclined to depart from it when doing so looks to be a more effective means for respecting what they take to be correct values. This is a recipe for the disintegration of the OSE. Far better, then, if each person were to value the ethic for its own sake. Still, it may be objected, this does not show that adherence to the OSE provides all persons with noninstrumental reasons for valuing toleration. To see this, assume (artificially) that we can distinguish sharply those with sound values from those who have unsound values.32With respect to the former, they have reason to value the OSE for its own sake, for if they do not value it in this way, then their adherence to it will be unsteady. And if their adherence to it is unsteady and publicly seen to be unsteady, then this will weaken or undermine general compliance with the ethic. This means that, in practice, they have reason to bracket or forget about the costs of t ~ l e r a t i o nDoing . ~ ~ so will
“The Structure of Perfectionist Toleration”
247
strengthen the OSE, general compliance with which is the best way to respect the good in the circumstances in which they find themselves. But what about the latter group (i.e., those with unsound values)? Given their mistaken beliefs about what is good and worthy of respect, they will best respect the good if they adhere faithfully to the OSE and value it for its own sake. But what do they have to do so? After all, the OSE has not been formulated to be a device for respecting their values. This point is well taken, but it does not defeat the argument in question. I have been trying to explain how toleration can have independent value. To do this, we do not need to show that all persons have reason to value toleration for its own sake. If the extended argument shows how those with sound values have good reason to value the OSE for its own sake, then this will be sufficient. It would establish that toleration has independent value. Of course, as I have pointed out, for the OSE to be an optimal social ethic it must be capable of being accepted or internalized by a significant number of people with mistaken evaluative beliefs. Doing so, however, does not require that they have a reason to value it for its own sake. Here, instrumental considerations as well as those derived from indifference or skepticism may be sufficient to motivate support for the ethic. So far I have said very little about the content of the OSE. The reason for this is that its content will vary from society to society, depending on the social groups that are present. This brings me to the second objection mentioned earlier. This objection holds that the extended argument makes the scope of justified toleration depend on the power relations that obtain between different social groups in any given society. Toleration provides persons with moral reasons to exercise restraint. But how can these reasons be dependent on underlying power relations? The reply to the objection is to concede its main charge, but deny that this is a weakness of the argument. To appreciate the force of the reply, however, it is necessary to distinguish the perfectionist account of toleration I am defending from social contract or contractualist accounts of t ~ l e r a t i o n Ac.~~ cording to the latter, the scope of justified toleration is determined by what cannot be reasonably rejected by persons who live together in political society. On this understanding of toleration, changes in the relative power of different social groups should not affect the content of justified toleration. Contractualist toleration is not a mere modus vivendi, but a set of practices that can elicit the approval of all reasonable persons who are subject to them.36So, if the account of toleration presented here were a version of contractualist toleration, then the objection would be fatal. But the extended argument does not rest on the contractualist norm of reasonable rejectability. An OSE need not be one that is reasonably acceptable to all.
248
Steven Wall
Does this mean that the measure of toleration justified by the extended argument is indistinguishable from that which is justified by a modus vivendi arrangement? Possibly, but not necessarily. As I have conceded, the content of the OSE for a given society will vary depending on the nature and strength of different social groups present in it. This can best be seen by considering the artificial society alluded to earlier, which is composed of one group with sound values and another group with unsound values. As the numbers and strength of the group with sound values increase, the OSE will more closely approximate its own ethic of toleration. The converse,however, is not true. If the numbers and strength of the group with unsound values increase, then it will not be true (at least not necessarily) that the OSE will more closely approximate its own ethic of toleration. The reason for this asymmetry is that the OSE is the ethic that best respects the good in the circumstances for which it is proposed. The beliefs and judgments of those with unsound values affect its content only to the extent that the ethic must be sufficiently acceptable to them in order for it to be effective. Whether or not a group’s practices or way of life receive protection from the OSE will not turn solely on its relative strength vis-a-vis other groups. It will also turn on the extent to which its way of life contributes to the good. To be sure, in some circumstances, given the underlying power relations in the society and given the beliefs of the members of different social groups, the content of the OSE for the society might be identical with the content of a modus vivendi ethic that reflected the bargaining power of the different groups. But this would be a coincidence. The justification for the OSE, and the reason why it gives people grounds for valuing toleration for its own sake, is that adhering to and complying with it is the best way to respect the good in the circumstances in which they find themselves. Since these circumstances reflect the power relations that obtain between the different social groups in the society, the content of the OSE will reflect these differences. This does not, however, transform the OSE into a mere pragmatic agreement between conflicting groups. It remains a device or a mechanism for optimally respecting the good. Reconciliation
The extended argument addresses the two worries the basic argument gave rise to. In the circumstances of social diversity, it justifies a broad measure of toleration and it does so by viewing toleration as a social practice and not merely as an individual virtue. I have suggested that this argument complements the basic argument. It broadens the range of objects that warrant tol-
“The Structure of Perfectionist Toleration”
249
eration on the basic argument. But now a new worry comes to the surface. Might the basic argument and the extended argument conflict with, rather than complement, one another? Can we affirm both arguments, or must we choose between them? Two points speak to this issue. The first point is that the basic argument is logically prior to the extended argument. In order to formulate an OSE for a particular society, one must identify the social practices and ways of life that are proper objects of toleration. This will require one to refer to the basic argument, for one will need to determine which of the practices and ways of life in the society in question are on balance valuable. It follows that even if persons should scrupulously comply with the OSE for their society, this would not render the basic argument otiose. The second point is that the two arguments need not conflict since they have different force in different circumstances. A full account of toleration should tell us what measure of toleration is justified in both a society in which the circumstances of social diversity obtain and one in which these circumstances do not obtain. The perfectionist account of toleration presented here does precisely this. The basic argument would have a great deal of force in a society in which the facts of social diversity did not obtain, whereas in a society in which these facts did obtain both the basic argument and the extended argument would need to be taken into account. Reflection on these two points reveals how the two arguments I have presented fit together. The basic argument, as its name implies, is prior to the extended argument. It provides individuals with reasons to tolerate objects that they rightly disesteem. The extended argument, in contrast, views toleration as a social practice that is particularly relevant to societies marked by significant social diversity. It provides persons who live in these societies with indirect reasons to tolerate objects that they rightly disesteem. The Limits of Toleration A final objection remains to be discussed. An account of toleration should not only justify toleration where it is appropriate, but also not condemn interference where it is appropriate. In defending a perfectionist account of toleration, I have put the focus on the value of respecting the good. But does (or could) this account demand too much restraint? In its orientation to respect the good, does perfectionist toleration fail to come to grips with the fact that some things are simply intolerable? Practices such as slavery or torture, for example, surely lie beyond the pale of justified toleration.
250
Steven Wall
To see why this issue is pressing, note how both the basic and the extended arguments shift the focus from an isolated practice that is rightly disesteemed to some larger more valuable whole. The basic argument does so by shifting the focus from a disvaluable practice to a larger way of life in which it is embedded that is on balance valuable. The extended argument does so by shifting the focus from a disvaluable practice or way of life to a larger societal ethic that, when generally adhered to, best respects the good. The objection is that in its concern to manifest respect for the good, the account of toleration I have proposed neglects or discounts the importance of suppressing the bad. A couple of replies can be made to the objection. First, as the larger wholes within which the rightly disesteemed practices are situated must themselves be on balance valuable, the disvalue or badness of the practices in question will have (perforce) been taken into account in arriving at judgments of justified toleration. If there exist circumstances in which persons are justified in tolerating slavery or torture, then-on the account presented here-this must be because it is possible for the disvalue of such practices to be counterbalanced by the good of some larger whole in which they are situated. Second, the perfectionist account of toleration I have advanced is premised on the normative principle to respect the good. But I have not claimed that this principle is the only normative principle. It is possible that other normative principles can compete and conflict with this normative principle. If so, then even if my account of toleration is correct, then it will identify only pro tanto, not necessarily conclusive, reasons for restraint.37 To illustrate this last point, consider the normative principle that holds that individual persons should not be treated unfairly. This principle can conflict with the principle to respect the good. Many examples bear this out, but a particularly interesting example is provided by Thomas Nagel. In defending the legitimacy of at least some unfair inequalities in wealth, Nagel suggests that permitting such inequalities, at least in our present economic circumstances, may be necessary to respect excellence. For without these inequalities,“the enjoyment of life at its upper boundaries” would not be possible. As Nagel sees it, the wonderful things that comprise life at its upper boundaries are rare perfectionist goods and “no egalitarianism can be right which would permit [them] to disappear just because not everyone can have them.”38 We do not have to accept the particular claims Nagel advances here in order to appreciate the point he calls attention to. Honoring the injunction to respect the good may require us to countenance the unfair treatment of individual persons.39I am suggesting that the converse is also true. Honoring the injunction to treat individual persons fairly may require us to countenance inadequate respect for the good. If so, then the pro tanto reason to tolerate an
“The Structure of Perfectionist Toleration”
25 1
object that one rightly disesteems may be overridden by a stronger reason to treat all persons fairly. With this in mind, return to the “intolerable”practices mentioned earlier. If it is possible that slavery and torture warrant toleration on the account I have proposed, then it would not automatically follow that it would be morally right to tolerate them. This automatically follows only if toleration overrides all conflicting values. But this is unlikely, given that the value of toleration derives from the normative principle to respect the good and that this normative principle is not the only normative principle. Conclusion
My argument for the independent value of toleration is now complete. I have claimed that tolerance is a virtue because it is (or can be) an appropriate way of manifesting respect for the good. This is possible since to respect the good it will sometimes be necessary to restrain the inclination to interfere with or suppress the bad. As we have seen, however, the scope of valuable toleration cannot be stated in the abstract. Toleration can have value even when the circumstances of social diversity do not obtain, but the scope of warranted toleration in a society depends on the nature and extent of social diversity present in that society. This is true for two reasons. First, societies differ in the number and range of on-balance valuable ways of life they contain; and, second, societies each have their own OSE depending on the social facts that obtain in them. By way of closing, I want to return to the limits to the argument I have presented. In grounding the independent value of toleration in the normative principle to respect the good, my argument for toleration has a perfectionist structure. But I have said very little here about the content of justified perfectionism. Rather than defend a specific account of the good, I wanted to show how toleration can have independent value on the assumption that some ways of life and social practices are objectively better than others. Since this assumption is shared by all perfectionists, my argument for toleration does not rest on any particular view about the content of the good. Still, as I mentioned earlier, a complete perfectionist argument for toleration eventually must identify and defend a theory of the good. Likewise, I have not defended perfectionism against subjectivist and skeptical objections. Those who deny either that some ways of life are objectively better than others or that we could come to know which ones were objectively better have been given no reason to accept my account of toleration. Notwithstanding these (severe) limits, my argument points toward an important and surprising
252
Steven Wall
conclusion. One important reason why we should tolerate the bad, when it is appropriate for us to do so, is that by tolerating the bad we may respect the
good.
Notes 1. To be more precise: the circumstances of social diversity obtain in a given society to the extent that each of the following three conditions is satisfied-(1) there exists a plurality of distinct social groups, (2) the social groups adhere to conflicting beliefs about how social life should be organized, and (3) the social groups are roughly equal in size and strength. 2. What I am calling the classical liberal argument is found in the works of Hobbes, Locke, and Spinoza. In calling it “the classical liberal argument,” I do not mean to imply that it is the only argument for toleration offered by these writers. It clearly is not. 3. By perfectionism, I mean the view that some ways of life, social practices, and forms of behavior are objectively better than others because they either (1) possess greater intrinsic value or (2) contribute more to the development or perfection of human nature. For accounts of perfectionism along these lines, see Vinit Haksar, Equalitp Liberty and Perfectionism (Oxford Oxford University Press, 1979), 3 4 , and George Sher, Beyond Neutrality: Perfectionism and Politics (Cambridge: Cambridge University Press, 1997), 8-10. 4. For an overview of how the concept of toleration has been understood in recent philosophical discussion, see the collected papers in David Heyd’s Toleration (Princeton, N.J.: Princeton University Press, 1996). 5. As will become apparent, the main objects of toleration that I am concerned with in this chapter are social practices and, more generally, ways of life. 6. In this chapter, I use the terms “tolerance”and “toleration” interchangeably. 7. I draw here on the discussion in Steven Wall, Liberalism, Perfectionism and Restraint (Cambridge: Cambridge University Press, 1998), 63-67. 8. I do not mean to imply that this belief must enter overtly into the deliberations of the tolerant person. In general, there are many ways in which moral considerations can impinge on the deliberations of a person who acts appropriately from the standpoint of morality. I shall pass over the numerous complications raised by this, but for a discussion,see Samuel Scheffler’s Human Morality (Oxford Oxford University Press, 1992), 29-51. 9. See Susan Mendus, Liberalism and the Limits of Toleration (London: Macmillan, 1989), 18-19. 10. Bernard Williams distinguishes the virtue from the practice of toleration. He
suggests that “we can have practices of toleration underlaid by skepticism or indifference or, again, by an understood balance of power.” He may be right, but I am interested in discussing practices of toleration that are themselves morally motivated or supported by moral reasons. These strike me as the genuine practices of toleration. I
“TheStructure of Perfectionist Toleration”
253
shall therefore not distinguish sharply between the virtue and the practice of toleration. See Bernard Williams, “Toleration: An Impossible Virtue,” in Toleration, ed. David Heyd (Princeton, N.J.: Princeton University Press, 1996). 11. See Christine Korsgaard, “Two Distinctions in Goodness,” Philosophical Review 92 (1983). 12. See Locke’s claim that only the free acceptance of a religious practice can bring about the goods of that practice; namely, salvation. John Locke, A Letter Concerning Toleration, ed. James Tully (Indianapolis: Hackett, 1983), 38. 13. To value the relationship for its own sake, the person in this example will need to bracket or forget the reasons why the relationship is valuable. As will become apparent later, this notion of bracketing plays a role in accounting for the independent value of toleration. 14. See Jeremy Waldron, “A Right to Do Wrong,” in Liberal Rights: Collected Papers 1981-1991 (Cambridge: Cambridge University Press, 1993). 15. For a defense of this view of autonomy, see Joseph Raz, The Morality of Freedom (Oxford Oxford University Press, 1986). 16. The idea here can be expressed by appealing to Robert Nozick‘s idea of “side constraints.” The proponent of autonomy-derived toleration, I am suggesting, needs to understand a person’s autonomy as consisting of a set of rights that others must respect. These rights constrain the efforts of others to promote their good. The robustness of the account of toleration that results will be a function of the stringency of these constraints. 17. See Mendus, Toleration and the Limits of Liberalism. 18. For the best attempt to provide such an account of which I am aware, see Joel Feinberg, Harm to Self(0xford: Oxford University Press, 1986), 98-171. 19. If so, then this will pose no problem for the argument of this chapter. Autonomyderived toleration need not conflict with the account of toleration that I am proposing. 20. On this second interpretation, we can be required to respect an excellence or a perfection, even if violating it would produce a greater excellence or perfection or a greater sum of excellence overall. Such a view is committed to the idea that some kinds of goods command respect for their own sakes. 21. See Raz, Morality of Freedom, 345-53. 22. There is an asymmetry here between valuable and disvaluable properties. The fact that a disvaluable property is only instantiated in one way of life does not affect its evaluative importance. 23. These claims have an affinity with Thomas Nagel’s discussion of perfectionist value in “The Fragmentation of Value,” in Mortal Questions (Cambridge: Cambridge University Press, 1979); see also Thomas Hurka, Perfectionism (Oxford: Oxford University Press, 1993), 79-83. 24. What if the valuable and disvaluable properties are exactly equal in evaluative importance? Then, we should say, it is neither good nor bad for the way of life to continue to exist. 25. This point is suggested by Raz, Morality of Freedom. Raz does not claim that the flip side of a given virtue may be a particular vice, but he does write: “The reason people lack certain virtues or accomplishments may be, and often is, that they possess other incompatible virtues and accomplishments” (402).
254
Steven Wall
26. This, of course, runs counter to a long and venerable tradition of thought that holds that the virtues are unitary. I believe this view is mistaken; but I will not be able to argue against it in this chapter. For a discussion, see Isaiah Berlin, Four Essays on Liberty (Oxford Oxford University Press, 1969); Stuart Hampshire, Morality and Conflict (Oxford Blackwell, 1983); and Michael Stocker, Plural and Conflicting Values (Oxford Oxford University Press, 1989). 27. They might, of course, have instrumental reasons or reasons derived from the value of autonomy to tolerate such objects. But we should put such reasons to one side for the purpose of evaluating the perfectionist argument for toleration that I am proposing. 28. Admittedly, this reference to “a substantial majority” of the members of a society is vague. I am assuming that to be effective an ethic of toleration need not be internalized by all the members of a society. Moreover, insisting that it should be would threaten to make the OSE utopian. But I doubt that we can state with precision the degree to which the ethic must be internalized in order to be effective. 29. Here, I will not offer an analysis of the concept of a society. For present purposes, Rawls’s account may suffice: a society is “a more or less self-sufficient association of persons who in their relations to one another recognize certain rules of conduct as binding and who for the most part act in accordance with them.” See John Rawls, A Theory ofJuustice (Cambridge, Mass.: Harvard University Press, 1971), 8. For a more precise account, see David Copp, Morality, Norrnativity and Society (Oxford Oxford University Press, 1995), 124-43. 30. This formulation is actually too strong. There may exist more than one ethic of toleration that would have unsurpassed value in respecting the good if it were internalized by the members of a given society. If so, an OSE would be any one of the ethics that had unsurpassed value. But, as taking this into account would greatly complicate my discussion, I shall set this possibility aside. 3 1. Here, I pass over a very important complication. The reasons to tolerate that are derived from the need to promote and sustain the OSE will differ in weight depending on whether the ethic is established or not established. If it is, the reasons are reasons to sustain it. If it is not, the reasons are reasons to bring it into existence. But if the prospects for establishing any effective societal ethic of toleration are bleak, then the reasons in question lose much of their force. 32. This is an artificial assumption since it is unrealistic to think that members of a society divide neatly into those with sound values and those with unsound values. In reality, most members of societies have some true and some false beliefs about what is valuable. 33. For a related idea, see David Lewis, “Mill and Milquetoast,” in Papers in Ethics and Social Philosophy (Cambridge: Cambridge University Press, 2000), 183-85. 34. In one sense, they do have a “reason” to do so; but, given their mistaken evaluative beliefs, it is not a reason that they can identify with and endorse. 35. See Thomas Nagel, Equality and Partiality (Oxford Oxford University Press, 1991), 154-68; and John Rawls, Political Liberalism (New York Columbia University Press, 1993).
“The Structure of Perfectionist Toleration”
255
36. This view of toleration is discussed in detail and is subject to criticism in Wall, Liberalism, Perfectionism and Restraint, 63-104. 37. To be convincing, this point needs considerable development. It assumes that the normative principle to respect the good does not subsume all other normative principles. At the same time, it allows that if a social practice or way of life violates a valid normative principle, then this makes it less valuable. The point, then, is that the weight or importance of some normative principles may not be fully accounted for by the normative principle to respect the good. But I shall not be able to spell out the full implications this point would have for the account of toleration presented here. 38. Nagel, Equality and Partiality, 138. It is clear from Nagel’s discussion that he believes that these inequalities are unfair if considered strictly from the standpoint of social justice. 39. In this context, consider the defense of slavery that holds that, in certain economic and social circumstances, in order for some to be free others must be slaves. On this view, the leisure and wealth necessary for the fully good life are only possible given that others are unjustly bound to physical labor. See Hannah Arendt, On Revolution (New York: Viking, 1963), 59-114.
This Page Intentionally Left Blank
15 “Religion, law, and Politics: Arenas of Neutrality” Kent Greenawalt
As in many of his previous, well-known studies, Greenawalt works to disentangle complex aspects of the relationship between religion and liberal political theory. He argues that, properly understood, neutrality can contribute to resolving constitutional and philosophical issues, although by itself it cannot provide solutions. Greenawalt responds to skeptical attacks on the concept of neutrality by theorists who claim that neutrality is essentially empty, that it is a rhetorical ploy used to lay claim to the moral high ground. Greenawalt further considers the role that religion should play in political debate in a liberal democracy like the United States. He rejects two extreme positions on this matter: an “exclusivist” position, according to which all religious claims should be barred from political debate; and an “inclusivist position,” according to which all religious arguments should be welcomed into political debate. Greenawalt presents and defends an intermediate position that seeks to take account of the different ways in which religious conviction can inform political discussion and the different roles that religious argument can play in public policy debates.
- 257 -
Kent Greenawalt
258
I. Introduction
T
HIS PAPER RELATES THE CONCEPT of
neutrality to central issues in the American law of church and state and to recent discussions about the role of religious ideas in the political discourse of liberal democracies. My central thesis is that “neutrality” can be of modest assistance in resolving questions of constitutional law and political philosophy,’ although it cannot produce solutions on its own. Notions of neutrality, if they are to be practically useful, must be filled in according to sets of “background beliefs,” but that does not mean neutrality can be filled with any content whatsoever. One must undertake initial analysis to decide whether, in context, neutral treatment of people, ideas, or organizations seems called for. Once one reaches that (tentative) decision, neutrality will foreclose some forms of treatment; other competing forms will be neutral in different respects. One’s final determination about what to do will depend on background beliefs. These will be employed in part not only to resolve which kind of neutrality is most important, but also to give effect to values not easily captured by the term “neutrality.”2 I begin the paper with brief theoretical observations about the concept of neutrality, focused on two critics of the idea that neutrality should guide relations between religious institutions and the government. I tie these observations to debates about how judges should interpret the Free Exercise and Establishment Clauses, indicating various senses of neutrality, but without undertaking a normative appraisal of how the First Amendment should be interpreted. Then I turn to the problem of religion and politics: I summarize views I have developed over the past two decades and explain how these views illustrate my theoretical generalization^.^ 11. How Far Does Neutrality Get You?
For nearly four decades, “neutrality” has been a major concept in the constitutional law of church and state, though different Supreme Court justices and academic commentators have had contrasting ideas about what neutrality entails. And one way of understanding claims about constraints on the proper grounds for political decisions is in terms of carrying the state’s “neutral” treatment of religion into a realm of political philosophy that reaches beyond constitutional law and beyond religious perspectives to all worldviews. An unsophisticated observer might suppose that neutrality is a very powerful concept in settling both how a liberal democratic state should deal with religion and how religious convictions should (or should not) influence its laws and policies. But skeptics rejoin that “neutrality” is essentially empty; it can be
“Religion, Law, and Politics: Arenas of Neutrality”
259
filled in with virtually any content and thus is no help whatsoever, except as a rhetorical ploy to invoke the high ground of fairness. Stephen Smith4 and Stanley Fish5 are two prominent skeptics who target claims of political theory as well as the law of church and state.6 Both assert that no theory of religious freedom can stand on its own, independent of assumptions about human nature, political organization, and religious truth. Both discuss John Locke,’ and we can understand their claims easily in respect to his theory. In his 1689 “A Letter Concerning Toleration,”8 Locke argued that people cannot be coerced into believing things, that Christian belief can be religiously efficacious only if it arises voluntarily, that many subjects that have divided Christians are really matters of indifference, and that government should concern itself only with protecting life, liberty, health, and outward possessions, not with the “care of ~ouls.’’~ Accordingly, people should have freedom to believe what they choose, to express those beliefs, and to practice religion as they see fit.1° Laws cannot be aimed against religious practices, such as animal sacrifice, but religious practices that happen to violate laws adopted for independent reasons can be stopped. Thus, if a law forbidding the killing of calves is adopted because “some extraordinary murrain” (pestilence or disease) has destroyed the stock of cattle, officials rightly enforce the law against members of a religious group that sacrifices animals. As Fish and Smith point out, Locke’s theory depends on highly debatable premises: about the effectivenessof coercion to affect belief;” about the need for voluntary acceptance of religious truth; about the unimportance of precise nuances of doctrines and forms of worship; about the state’s responsibility to protect natural rights rather than seeing that people lead good lives; and perhaps on an implicit assumption that external actions are not central for true religion.12 Fish and Smith argue that Locke’s general approach to religious liberty depends on numerous background beliefs with which other people disagree, and that Locke can mount no convincing set of arguments to lead opponents to surrender their background beliefs in favor of his. What is true for Locke, say Fish and Smith, is true for any theory of religious freedom. Their position is well captured by Fish‘s response to a passage of Jeremy Waldron’s, suggesting that insofar as Locke’s position is developed on the basis of Christian premises, it is “insufficiently general to be philosophically intere~ting.”’~ Fish answers that one cannot “devise principles of such generality that they speak to the intuitions of all persons however situated. . . and are therefore acceptable to all persons, even those whose interests might be disadvantaged by their realization in public lifel’14 Fish and Smith are right that no theory of religious freedom can satisfy everyone, regardless of background belief. A person who is convinced (1) that
260
Kent Greenawalt
only true believers will be saved and others condemned to everlasting fire, (2) that coercion can induce true, saving belief, and (3) that governments are well situated to identify religious truth, will be unlikely to endorse as ideal anything like American notions of free exercise and nonestabli~hment.’~ Indeed, this is one obvious lesson of the revulsion extreme Islamic fundamentalists have to the values of our society. What follows from the impossibility of satisfjmg everyone? May a society nonetheless have a justified approach to religious liberty, based either on a consensus or on sound principles? Many societies may enjoy very broad agreement among people with widely different background beliefs about basic relations of church and state. If, let us say, 98 percent of the population happen to agree on these matters, perhaps they may proceed accordingly,even if no one can present a convincing argument to persuade the other 2 percent.16 Fish and Smith, though for somewhat different reasons, would regard agreement among the 98 percent as failing to provide a just resolution for many issues. The agreement is irrelevant to Fish’s fundamental point, which is that the 2 percent end up as losers for reasons that cannot be expected to persuade them.” Smith would look at American society (and other Western societies) and say that any such broad agreement will be abstract, too circumscribed in substantive content, to provide a cogent theory of religious freedom for resolving constitutional cases. Judges resolving actual cases, and left with a shared public understanding that is too thin to yield coherent principles, will need to refer to their own background beliefs or to other criteria, such as the original intent of those who adopted the Constitution or doctrines settled by judicial precedents. Theories of religious freedom depend on background beliefs. What follows? Some of what Smith says and much of what Fish says follows does not. Here, we reach crucial points about neutrality and fairness. Smith and Fish are right that “neutrality” cannot be self-defining. One must determine neutrality of what kind for whom in respect to what aspects of which subjects; and the path one chooses will depend on background beliefs. Thus, one person might urge that neutrality as to religion would exclude creationism from public schools, if religious conviction is the underlying basis for belief in creationism. Another might object that neutrality requires that students be exposed to creationism if they are exposed to the competing theory of evolution, a theory that rejects the religious understandings of many citizens.’* In the abstract, one cannot say which approach is more neutral; judgment on that score requires filling in a number of assumption^.'^ Fish says, “[Nleutrality, like any other abstraction, has meaning only with some set of background conditions; as a rule or measure it will always reflect decisions and distinctions it cannot
“Religion, Law, and Politics: Arenas of Neutrality”
261
recognize because it unfolds and has applications within them.”20Smith remarks, “[Tlhe ideal of religious neutrality is simply not coherent.”21On the closely related topic of fairness, Fish tells us that the hope of liberals-“to establish a form of political organization that is fair to all parties-cannot possibly be realized.”22 Let us grant that no schema can be established that all citizens will actually perceive to be neutral or even that they should find acceptable, consonant with their own background beliefs. It neither follows that “neutrality” is empty nor that no schema is fairer or more importantly neutral than any other. Once one specifies a subject matter, “neutrality” has some minimum content. Neutral treatment of individuals or groups implies some form of equal treatment.23The idea that a government should be “neutral as to religion” has a fair amount of descriptive content. Not every conceivable approach to how the state should treat religion is well cast as filling in the “empty” content of neutrality; some positions self-consciously reject neutrality in favor of other values. Thus, two people might agree that one government is more “neutral” toward religions than another, yet disagree sharply over which approach is preferable. Philosophically inclined members of the Taliban might well acknowledge that the American government is more neutral about religion than was their regime, but they would reject neutrality as an ideal in favor of adherence to the true Another point is more important, however. If one set of background beliefs is best, then the ideal church-state relations for any particular society are what that set of beliefs recommends for a society that is so c ~ n s t i t u t e dIf. ~that ~ ideal was implemented, if a society had the best form of church-state relations possible, and that form achieved neutrality in some sense, then everyone could be treated fairly, and the best form of neutrality would be realized, although not everyone could be expected to recognize that. If a version of liberal political philosophy requires that everyone perceive that they are being treated fairly, or that they be presented with arguments that should convince them (with the background beliefs they have) that they are being treated fairly, we should recognize that that aspiration of fairness is impossible to achieve. But if that form of fairness is impossible to achieve, should we not conclude that real fairness does not require that everyone recognize it as such?26I understand, for example, that many people have background beliefs according to which controversial forms of affirmativeaction are unfair; but I believe nonetheless that they are fair. If it is impossible for people to agree on what is fair, we should not make such an agreement a requisite for what actual fairness entails, as Fish may What I have said about neutrality and fairness, of course, hinges on the assumption that some sets of background beliefs really are better than others.
262
Kent Greenawalt
These background beliefs might be what Rawls calls comprehensive views, but they could also include ideas of political philosophy that are detached from specific comprehensive views. Fish suggests powerful doubt that any set of background beliefs is really better, or identifiably better, than any other. In a revealing, if elusive, footnote, he tells us, “I shall several times deny the availability of any mechanism or calculus for determining the truth about a matter. This does not mean that I believe there to be no such truth, only that I believe there to be no way of flushing it out so that everyone, however situated, will assent to it.”28 According to the most modest version of this passage, relevant truth cannot be flushed out so that everyone will assent to it. That seems right, at least in regard to many claims of People certainly do not give up talking about truth if this condition cannot be met. Most of us are highly confident about some scientific and historical facts (such as the killing of millions of Jews and other civilians in death camps during World War II), although we acknowledge that not everyone can be expected to assent to those What are we to make of Fish‘s own thesis in his essay? It contains a set of claims about what is impossible. These seem to be about what is true philosophically,although Fish cannot expect everyone to assent,31given the background beliefs many people have that are fundamentally at odds with his claims. If we lower our sights and accept that many strong arguments about what is true and not true may be valid although these arguments will not (and even hypothetically could not, under varied conditions) convince everyone, we revert to the possibility that one set of background beliefs might be better than any other and that human beings may have a degree of capacity to identify sound background beliefs. We might wonder if political, moral, and religious positions are somehow different from science and history, not susceptible to discovery, not true or false, even according to a less ambitious criterion.32Once we understand the question as whether, and how far, human beings can grasp religious, moral, and political truth, we quickly see that the answer cannot be yanked out of a hat of conceptual deconstruction. It requires serious analysis of various assertions about truth or validity in these domains. Where does all this leave us with respect to church-state relations? Fish suggests that principles are inevitably tools of rhetoric, not devices to discern In discussing the views of David M. Smolin, who acknowledges that he would like to convert others to his Christian views but wants to do so “fairly,”Fish admires his candor but expresses puzzlement at his failure to take the final step of dispensing with illusory constraints of fairness.34 Fish‘s treatment of Smolin reveals strikingly how unpersuasive is his dismissal of ideas of fairness. Smolin may believe that God wants true religion to
“Religion, Law, and Politics: Arenas of Neutrality”
263
flourish in an atmosphere of freedom. A government fully responsive to Smolin’s concerns would support religion in ways that secular liberals would reject; but Smolin does not want the law to silence dissenting religious voices. Smolin wants to spread his beliefs but according to principles of government that he deems fair to opposing views. Fish regards this as peculiar, showing a kind of failure by Smolin to carry out his full convictions. But notions of fairness might well comprise part of Smolin’s overall background beliefs.35Ideas of fairness are part of many people’s background beliefs, and principles of fairness might well comprise part of the best set of beliefs, perhaps connecting to his religious ideas about human freedom. We would be foolish to simply throw out ideas of fairness and neutrality or to suppose they are only rhetorical tools in people’s pursuit of more concrete or substantive objectives. 111. Religion Clause Issues Ideas of neutrality figure significantly in adjudication under the Free Exercise and Establishment Clauses, but does the concept of neutrality make a difference?Let us look briefly at five areas of potential state involvement with religion: singling out one religion for favorable or unfavorable treatment; preferring religion in general; exempting claims of religious conscience from ordinary laws; resolving interchurch disputes over property; and granting aid to religious organizations. We see that “neutrality”excludes some possibilities, but leaves open other competing options. A. Singling Out for Favorable or Unfavorable Treatment? Perhaps the least controversial aspect of the Free Exercise and Establishment Clauses is that they allow neither the favoring nor the persecution of any particular religion. If the government promoted the doctrines of the Presbyterian church or made grants to it that were not available to other religious groups, it would impermissibly establish Presbyterianism. And, to take the facts of a case that reached the Supreme Court, if a city forbids animal sacrifice, a practice engaged in by only one church within its boundaries, and it allows other killings of animals that are indistinguishable in terms of human health and animal suffering, it has violated the Free Exercise Clause.36 On these issues, a principle of state neutrality regarding religion is nearly dispositive. If the state should be neutral about religion or among religions, teaching the doctrines of a particular religion or aiding just that religion is out. (However, if one thought the crucial neutrality were one of justification, not actual treatment, one might defend an established church on the neutral
264
Kent Greenawalt
principle of aiding the religion of most citizens, to provide secular benefits.) Also forbidden is restricting a religion because most people reject its ideas or dislike its practices (which cause no more secular harm than other practices they tolerate). Saying that neutrality about religion does not permit the favoring or the persecution of one religious body does not, of course, resolve whether such neutrality is desirable. One needs arguments, based on background beliefs that may include an interpretation of our national history, of why neutrality to this degree is called for. But the United States enjoys a very broad consensus that the government should be neutral about religion in this sense.37
B. Preferring Religion in General? In its development of Establishment Clause doctrine, the Supreme Court has declared that the government may not prefer religion in general to nonreligious beliefs, organizations, and practices. Thus, even if a public school could come up with a prayer or another form of devotion that would please members of every religion (probably an impossibility), it could not carry on that practice. The idea of neutrality between religion and nonreligion is much more controversial than the idea of neutrality among religions. Nonetheless, if one believes that the Establishment Clause should be understood to prohibit favoritism toward religion, certain things follow. The government should not teach broad religious ideas as true in the public schools, nor should it give money to religious organizations that run hospitals and refuse money to nonprofit nonreligious organizations that run hospitals. For other issues, the implications of neutrality between religion and nonreligion will be less clear. What should schools do about teaching evolution and an alternative that suggests divine creation?38The standard neo-Darwinian version of evolutionary theory includes no premise about divine creative activity; that theory is sharply at odds with the religious convictions of persons who accept the account of origins in Genesis as literally true, and of persons who believe that, though Genesis is not literally accurate, God actively intervened at various stages in the development of complex life. If evolution is taught as true, is that consonant with neutrality between religion and nonreligion or should schools, instead, either ignore evolution or teach creationism or “intelligent design” as an alternative? On this subject, the basic idea of neutrality does not get us very far. As Smith points out, it could be viewed as unneutral to teach evolution as but it could also be viewed as unneutral to teach material just because people believe it reflects religious truth.
“Religion,Law, and Politics: Arenas of Neutrality”
265
C. Exempting Religious Claims of Conscience? By far the most controversial aspect of modern free exercise law concerns religious exemptions. Should people have a free exercise right to be free of legal obligations validly imposed on others? For example, should religious objectors be excused from jury service or military duty; should Amish parents be able to withdraw their children from school after the eighth grade; should members of a Native American church be allowed to use peyote in their worship services?Since 1990, the prevailing approach of the Supreme Court, with various qualifications, is this: if a law is neutral, and generally applicable, a person whose conduct is religiously motivated has no constitutional claim to violate the law.40It is easy to see how this posture could be viewed as neutral between religion and nonreligion. As in Locke’s example of killing cattle, the people with religious reasons to engage in conduct are treated like nonreligious people who want to engage in conduct for their own secular reasons. But consider the following argument. People with strong religious convictions may have uniquely powerful reasons not to comply with certain legal requirements, and indeed one might even consider them “unable to perform” in a sense. A political majority will not impose requirements that are obnoxious to a great many citizens-no American state is going to forbid the use of wine in Christian communion, for example-but it may be insensitive to small religious minorities. If one wants a genuine substantive neutrality among religions and between religion and nonreligion, some exceptions to laws should be recognized con~titutionally.~~ These will put religious minorities on a more nearly equal footing with major religions and with those whose reasons for action are nonreligious. Now, what I have said so far fails to grapple with three critical problems. What should be done about nonreligious conscience? How does one distinguish between needed exemptions that relieve burdens and favoritism that violates the Establishment Clause? In light of all the adjudicative difficulties, may not the best resolution be that reached by the Court-not requiring exemptions but allowing legislatures and administrative bodies to grant them? In a full evaluation, each of these issues would have to be resolved. But they do not affect my major point-the inconclusiveness of the notion of neutrality-which Locke’s original example well illustrates. In one sense, it is neutral to subject everyone to the ban on killing cattle. But if some people have a religious conviction that they must kill a few animals for sacrifice, if that conviction is a much stronger perceived reason than anyone else has for killing a few cattle, if allowing an exemption would not undermine the basic prohibition, and if the legislature would never have enacted the ban at all had most people felt that their reasons for killing cattle were as strong as
266
Kent Greenawalt
the perceived reasons of those who engage in animal sacrifice, one might understand fairness and genuine neutrality to demand an exemption. The no-exemption and exemption approaches each realize different concrete notions of neutrality; perhaps, they reflect disagreement about the best way to carry out some yet more fundamental idea of equal treatment or neutrality. One must engage the arguments about why exemptions are or are not fair, and the further arguments about whether they should be created by legislators or courts. Neutrality, by itself, does not take us very far. D. Resolving Interchurch Disputes over Property For nearly four decades, the Supreme Court has said that courts may not decide internal disputes within religious organizations based on which faction is truer to the group’s religious traditions-a counsel of avoidance that harks back to a post-Civil War decision.42To resolve disputes about which faction controls property, courts may accept the decision of the highest body in a hierarchical church, or they may to adhere to neutral principles-examining relevant documents in a manner that requires no reliance on religious understandings. Here, “neutral” does not function as an independent idea; it means simply ordinary secular, rather than religious, reasons. Yet, a more substantial idea of neutrality does rest in the background-the idea that secular courts should remain neutral, in the sense of not resolving disputes about religious understanding, because they are not appropriate arbiters of such claims. E. Aid to Religious Organizations
The most complex ideas of “neutrality” are found in connection with aid to religious organizations, such as schools and hospitals, that provide secular services.43In some Establishment Clause decisions, the Supreme Court has spoken of neutrality as the right balance between aiding and handicapping religion. In other decisions, the justices have referred to neutral services, such as education, in the sense of secular or nonreligious services. But the dominant use of “neutral” in recent decisions is giving the same treatment to religious and nonreligious organizations that provide the same services. Four justices now believe that if the government treats all organizations equally without respect to religion, and does not explicitly aid religious functions, it may provide aid on an equal basis, even if the predictable effect is significantly to assist the religious aims of most of the providing 0rganizations.4~ Thus, if the great majority of private schools receiving grants are religious, and the assistance is so great that its effect will undoubtedly be to help religious aspects of the school programs, nevertheless, the aid is permissible, so
“Religion,Law, and Politics: Arenas of Neutrality”
267
long as the category of grantees is not drawn in religious terms, the reasons for aid are not religious, and the aid is not given for religious endeavors. This approach is definitely one exemplification of what neutrality entails, and it has considerable intuitive appeal. But the competing position, which has so far prevailed in school aid cases, may also be cast in terms of neutrality. The fundamental principle that the government should not aid or inhibit religion is a principle of neutrality; that principle entails that government, even when its purpose is appropriate, should not provide assistance whose effect will substantially assist religious endeavors. The concept of neutrality provides no decisive resolution. However, i f one reduces neutrality here to equal treatment and takes neutrality as the controlling value, neutrality seems to support the position of the four justices. The opposing position is best understood as relying on values of governmental nonassistance to religion and noninvolvement in religion that do not connect so closely to neutrality. (Since this essay was written, the Supreme Court has upheld substantial aid to religious schools that is conveyed through the independent choices of parents under a voucher program in which religious and nonreligious schools are treated equally.) This last observation suggests a more general caution. In many circumstances, among the various values that are relevant to a decision, some will fit better than others under the rubric of neutrality. Competing advocates will not see the issue as starkly requiring a choice between neutrality and a competing approach, as with the choice whether to have an established state relig i ~ n Rather, . ~ ~ either choice could be defended as appropriately neutral. But many of the relevant reasons are not best seen as manifestations of neutrality. Thus, we might conclude that the best reasons overall point one way and the reasons that best reflect neutrality values point the other way. We must be careful not to assume that if both sides agree that neutrality in some sense should be achieved, it follows that the best solution is the one that seems most cogent if one focuses exclusively on realizing neutrality. IV. Religious Ideas in Politics
In addressing the subject of religious convictions and political discourse, I restrict myself largely to developing my own views-the reader will need to look elsewhere for an analytical examination of competing position^.^^ In contrast with my brief account of legal problems under the religion clauses, I do draw normative conclusions, which I relate to the general themes about neutrality. The fundamental question I consider is whether, when they resolve political issues within liberal democracies, citizens and officials properly rely on religious
268
Kent Greenawalt
conceptions that inform moral judgments, which in turn bear on political resolutions. The modern social problem around which this question dominantly swirls is the legal status of abortion, but religious understandings may also bear on the treatment of animals and the environment, on aid to the poor and disabled, on capital punishment, and on a host of other political problems. One issue about the legitimacy of particular political outcomes, and of the systems in which these outcomes occur, is whether the grounds on which issues are debated and resolved are appropriate. Should religious arguments figure like any others in public debate or should people keep them out of the political forum? A similar question can be put about nonreligious comprehensive views about human life, such as “greatest happiness utilitarianism” and Kantianism, but I shall concentrate on the more stark and controversial question about religion. The specific issue about use of religious grounds connects to a more sweeping question about political philosophy. How far can conclusions about employing religious conceptions be provided universally, or at least for all liberal democracies; how far do conclusions depend on particular cultures and political orders? My position about the use of religious grounds falls between two extremes: it is intermediate, nuanced in terms of our culture and history and realistic about human capacities. I first summarize more extreme positions, sharpening the crucial issues by clarifymg what is not in dispute. After this summary and clearing away, I present my own approach, before offering some concluding observations about doing political theory and about “neutrality.” A. Exclusive and Inclusive Positions People who challenge the injection of religion into politics adopt what we may call an “exclusive”position. Religion should be excluded from politics. In the politics of pluralist liberal democracies, decisions (they claim) should be made on grounds that are shared premises of that form of government and on forms of justification and ways of determining facts that are accessible to all citizens. Whatever is the exact mix of the rational, nonrational, and irrational in religious understandings, no religious perspective is shared by all citizens and no perspective rests on methods of justification and determining facts that are accessible in the required way. To some extent, religious belief depends on faith, personal experience, and distinctive tradition; adherents of one religion cannot present “logical” arguments that alone will persuade outsiders to their views. Religious belief and practice is fine for individuals and communities of faith; and religious perspectives may enrich our cultural understandings. But when citizens are coerced, the state acts unfairly unless it has reasons that have force for all citizens. Religious reasons do not fall into this category. They do not
“Religion, Law, and Politics: Arenas of Neutrality”
269
belong in democratic politics. This is a matter of fairness, and also of political stability. Neither citizens nor officials should present religious reasons in public debate; neither group should rely on such reasons. Here are some brief clarifications about this “exclusive”position. First, it concerns politics, not broader public culture. It need not assert that religion belongs in a private, wholly nonpublic sphere; it need assert only that religion should be kept out of politics. Second, no one claims that people making political decisions will be wholly uninfluenced by religious understandings, any more than people are wholly uninfluenced by their early childhood experiences. Any claim of that sort would be naive. People should discuss political issues in public without reliance on religious premises and they should try to make up their minds accordingly, as they should try to decide independent of personal grudges they may carry. Third, the claim is not that religion is foolish superstition and therefore deserves no place in our political life. Of course, all arguments based on foolish superstition should be avoided, but if that were the basis for excluding religion, the exclusion would have to rest on a persuasive argument that religion is foolish superstition. Whatever they may think about religion, proponents of exclusion base their position on premises of democratic government, not on the intrinsic foolishness of religion. More to the point, more than 90 percent of our citizens identify themselves as religious; one can’t reasonably suppose they should avoid religion in politics because religion is foolish. Finally, we have to be careful about what the “exclusive” position entails. What is being urged is “self-exclusion.”No one wants people punished or silenced for making religious arguments; indeed, guarantees of free speech and free exercise protect such arguments. The proposal is that people should refrain from making religious arguments because they do not fit with how liberal democracies should work. The external sanctions, if any, would be that other people would regard arguments in religious terms as improper and express their disapproval, and that courts might consider laws based on religious grounds as invalid. The “inclusive” position is that citizens and officials should be able to rely on whatever sources of understanding seem to them most reliable and illuminating. If a respected religious authority like the pope, a divinely inspired text, or one’s personal sense of how God relates to human beings suggests that we should help those who are less fortunate, why should that not count for our position on welfare reform and medical insurance? People do not feel whole if they try to divorce their deepest sources of insight from their political stances. Moreover, shared premises and methods of justification are too thin to resolve many political issues; they just do not settle enough in a society as diverse and divided as our own. Fairness consists not in exclusion, even selfexclusion, but in everyone’s relying on what each thinks is most convincing. A full airing of all those views will enrich everyone’s understanding. A healthy
270
Kent Greenawalt
democracy will not be unstable if religious arguments are part of political discourse. For the inclusive position, I want to make only one clarification. Someone need not claim that every ground for a political position is appropriate. Certain grounds may be contrary to premises of liberal democracy. We now suppose that racism and other denials of equal worth fall into this category. But religion has never been so regarded in our country. From the beginning, religious belief and practice have been thought fully compatible with the underpinnings of our political order. In the course of presenting two straightforward competing positions, I have indicated five matters that are not at issue, but are sometimes confused with what is genuinely debatable. What is not at issue is whether (1) some justifications do not belong in liberal politics because they conflict with basic premises of our form of government, (2) religion is foolish superstition, (3) religious belief and practice should be relegated to a private realm, (4) people will be influenced in politics to some degree by religious understandings, and (5) people should be forbidden to express political positions they base on religious premises.
B. “Detached” Political Philosophy and Impositions Two other clarifications concern the underlying basis for analysis and the problem of impositions. Should one start with a particular comprehensive view, say Roman Catholicism, orthodox Judaism, liberal Protestantism, or Kantianism, and see what implications follow for how religion should figure in politics, or should one try to do what may be called “detached” political philosophy, not relying on any particular comprehensive view? Both exercises are valuable, but I am undertaking the latter, in the hope that the analysis will appeal to those who hold different comprehensive views. My second clarification is a bit more complicated. It is about the connection between what legislation does and the grounds for its adoption. Some laws are bad because they impose unacceptably on others, whatever the underlying reasons of those who adopt them. To take an extreme case, imagine that the government forbids all religious practice except that according to Islam or Catholicism. Its reason has nothing to do with religious truth; it wants to promote political unity. Nevertheless, the law would be an unacceptable imposition of religion. I believe the same is true about class prayer in public schools; it is wrong because of what it does, whatever reasons officials may have for supporting it. Many proponents of the inclusive position about religious grounds may also support measures that impose on individuals in this way. One thinks of the Moral Majority and the Christian Coalition. This leads
“Religion, Law, and Politics:Arenas of Neutrality”
271
some opponents to suppose that if religious impositions are bad, religious grounds must also be bad. But this is a fallacy. There are many pieces of legislation, such as abolition of capital punishment and forms of assistance to the poor, that do not impose religious views in this manner, yet might be based on religious grounds. Such legislation fails not only to impose a religious view on nonbelievers, but also to promote a religious view.47 Saying whether a law imposes in the way I have suggested is not always easy. Laws against same-sex marriages, for example, are difficult to classify. What I am claiming is that religious grounds might figure in the adoption of many laws that do not impose. These are the instances that concern me.
C. Intermediate Positions There is something strongly attractive in each of the two basic positions about using religious grounds. Indeed, the attractions of one are a powerful obstacle to wholesale acceptance of the other. Once we examine matters more closely, we should begin to doubt that one answer is right for all liberal democracies. We should also see that the strength of the arguments for the two basic positions varies with different sorts of people and different aspects of participation in political life. Before I sketch my own position, I want to mention two other kinds of intermediate positions. One is that everything depends on the kind of religious understanding involved. Views that are not too dogmatic and sectarian, that are open to competing points of view, play a useful role in politics. Views that are narrow and dogmatic, that leave nothing for dialogue with others, do not belong in the politics of a liberal democracy. I think this approach founders on the problem that what is at issue is self-exclusion. People are hesitant to label their own particular views as narrow, dogmatic, and inaccessible; and even if they do, they are likely to think it is unfair for them to refrain from relying on religion when their looser, more liberal, perhaps heretical, brothers and sisters can do so. A different intermediate position, that of Rawls, distinguishes ordinary political issues from constitutional issues and issues of basic justice:* For ordinary issues, religious grounds are appropriate; but for constitutional issues, people should rely on reasons that would have persuasive force for all reasonable citizens. This position faces a kind of technical difficulty. Ordinary issues are deeply intertwined with constitutional issues and issues of basic justice. For example, the status of an embryo shortly after conception turns out to be crucial for abortion, a constitutional issue, and for stem-cell research, an “ordinary” issue. Telling people they can rely on religious grounds for deciding that status for stem-cell research but not abortion would not be a counsel
272
Kent Greenawalt
most people could grasp and accept. More fundamentally, the grounds for any such sharp distinction between the two kinds of issues are not convincing. D. Crucial Distinctions My own position relies heavily on distinctions between advocacy and justification, on the one hand, and grounds of judgment, on the other, and between officials and ordinary citizens. When we think about how we make up our minds and how we discuss issues, we see that monitoring our discourse is a lot easier than restricting our bases for decision. Moreover, other people hear our discourse; they cannot know our full grounds of decision. These truths have great importance. Most people would be hard put to try to carry out a program of excluding their deepest religious convictions from their political judgments. They could not disentangle what they believe because of underlying religious convictions from what they would believe if they relied only on premises of liberal democracy and shared techniques of understanding. Speaking without reference to religious convictions is not difficult. Members of our law faculty share an assumption that school problems are to be resolved in terms of values that are not explicitly connected to particular comprehensive views. I have yet to hear a specifically Jewish, Christian, atheist, or Benthamite argument for a faculty decision. Yet, when decisions involve the point of legal education, I wonder if colleagues try rigorously to remove the threads of their religious understandings about the nature of society and education for a profession. People can tell easily whether arguments are being made from explicit religious premises; they will know if restraint on their part matched. If they try to purge their silent deliberations of religious influence, they cannot be sure if others are similarly motivated. And once someone realizes how difficult this purging exercise is, he will question the success of others, even if he thinks they are trying. Such uncertainties are a poor basis for restraint that should be reciprocal. Consider some differences between officials and ordinary citizens. Officials have a lot more to do with the law that gets made and applied than do citizens; there are many more citizens than officials. Officials are accustomed to making judgments and offering reasons that do not include all that is relevant in their personal lives. Citizens are not familiar with practicing such restraint. Perhaps a highly educated, participating citizenry could learn to draw distinctions between what matters for most aspects of life and what matters for politics. But that is not our citizenry. When officials practice some restraint, that impinges much less on a population’s religious liberty than when citizens do
“Religion,Law, and Politics: Arenas of Neutrality”
273
so. Official restraint more greatly affects the quality of political life. These basic distinctions-between advocacy and judgment and between officials and citizens-suggest that if any self-exclusion is justified, it is self-exclusion for officials in their public statements. That indeed is the core of my position. E. Officials
Among officials, we can divide roughly between those who apply law and those who make law or exercise ordinary discretionaryjudgment. Among those who apply law, judges and quasi-judicial officials often provide reasoned justifications for their decisions. At this stage of American history, one does not find explicitly religious grounding in opinions, even when courts reach beyond standard legal sources to comment on the social benefits or harms of a possible ruling. By an explicitly religious grounding, I mean reasoning in this form: “Given a true religious proposition, these conclusions about social good follow.” Some examination of religious sources might be acceptable to show the community’s attitudes toward a practice or its deep moral assumptions, and judges might employ familiar religious stories to illustrate a point; but none of these is a reliance on religious grounds in the sense that I mean. Although judicial opinions are rarely completely candid about the strength of competing arguments, one expects judges to rely on arguments they believe should have force for all judges. In our culture, this excludes arguments based on particular religious premises. When we turn to legislators, we may start with the proposition that if an explicit religious grounding were placed in the preamble to a statute, that might be viewed as a promotion of religion that would violate the Establishment Clause. Members of Congress typically do not make religious arguments on the floor of Congress or before their constituents. There is, however, no accepted understanding that they should avoid giving any weight to their own religious convictions and to those of constituents in the formulation of their positions. I believe legislators should give greater weight to reasons that are generally available than to those they understand are not; but some reliance on religious and similar reasons is appropriate, especially since the generally available reasons are radically indecisive about some crucial social problems. I believe that present conventions about national legislators reflect a sound accommodation of the needs of a religiously diverse citizenry with the inclination of legislators to bring all they believe to bear on political problems. If legislators rely on religious understandings more than their public advocacy reflects, are they not lacking in candor? Does restraint impoverish discourse and leave voters less well informed than they might be? Realism counsels that much of what legislators say is far from fully candid, so self-restraint about
Kent Greenawalt
274
religious grounds is hardly a major contributor to lack of candor. In any event, the value of self-restraint overrides this drawback and whatever reduction in information voters suffer. Legislators should not deny religious bases that motivate them, but they should not develop public arguments in these terms.
F. Citizens Because citizens are not used to practicing self-restraint of this kind, and because most citizens have little involvement in the political process, I do not think they should regard themselves as constrained to avoid relying on religious grounds or to avoid stating those grounds. Some citizens, however, such as university and corporation presidents, and individuals consistently engaged in political life, have a much more public role. For them, something like the constraints for legislators is appropriate. Religious leaders and organizations have a special place. They properly develop religious grounds as these relate to political problems, and they also properly take part in direct efforts to win support for particular positions. I think it is usually unfortunate when religious leaders endorse parties or candidates and when people who become officials in high governmental positions continue to hold themselves out as religious leaders. G . Summary
When we examine our political practices, we see that our society has some loose, somewhat controversial, conventions about the place of religion in political life. I think those conventions represent an appropriate approach within a liberal democracy, one that is well suited for our society at this time. It is regrettable that within the larger culture a kind of divide exists between serious religion, which affects many people and influences political life, and a popular and intellectual nonreligious culture. More dialogue in this larger culture about a religious or spiritual dimension of life would be healthy. But a substantially increased injection of religious premises into discussions of particular political issues would not. V. Concluding Thoughts about Political Philosophy and Neutrality Much of the theorizing about this and related subjects has been cast in terms of liberal democracies in general, or as what the Establishment Clause of our Constitution actually requires. Neither of these approaches answers the most central practical questions.
“Religion, Law, and Politics: Arenas of Neutrality”
275
The Establishment Clause, in its direct force, has modest implications here. It is mainly about what laws do, not why they are enacted. What of liberal democracies and theories of legitimacy? Democratic theorists argue persuasively that in a liberal society people will adopt many different comprehensive views. This condition will not change. The history of Western liberal democracies, forged out of religious division, shows that differences in religious views can be a source of intense conflict; but we can imagine that people of various religious views seek to learn from one another and trust one another’s social judgments. These people might welcome religious perspectives in political discourse. On the other hand, one would not recommend an explicitly religious politics as the most fruitful approach for a newly constituted Northern Ireland or for the fragile political union that may be formed in Bosnia. Much depends on history, culture, the religious and other comprehensive views that people hold, and their degree of mutual tolerance and respect. If I am right about this, specific principles of self-restraint must be offered for particular political orders, not in gross. If this is true about religious discourse and public reason, it is also true about many other practical issues to which political philosophers speak. The United States is a country of great diversity in culture and religion. The percentage of our people who are neither Christian nor Jewish increases steadily, with immigration policies that no longer discriminate egregiously against Asians. Outright religious conflict is rare, but religious differences remain a source of distrust and tension. Religious convictions are intense and widespread enough to influence politics and to disturb people with their influence. This is why some restraint is needed. Both the exclusive position and the inclusive position can be cast in terms of neutrality. The exclusive position exemplifies neutrality by saying that laws should be based on reasons that are in some sense accessible to everyone. The inclusive position exemplifies neutrality by allowing people to rely on whatever reasons they find most persuasive. The concept of neutrality cannot adjudicate between the two positions; each realizes one important value of neutrality and fairness. One can decide among these positions and various intermediate ones only on the basis of a complex evaluation of political life in particular liberal democracies. “Neutrality” is not exactly empty here; it points to values to be realized, but genuine, valuable forms of neutrality conflict with one another. To resolve this conflict, to decide how best to serve an overall ideal of neutrality or fairness, one must engage a contextualized inquiry, bringing to bear background values and factual beliefs. As my own exercise establishes, these background beliefs could be ones about political life that are independent of any particular comprehensive views. An intermediate position like that I have
276
Kent Greenawalt
sketched could be the most sound, even though it translates less simply into being characterized as “neutral” than either of the extremes.
Notes 1. Two clarifications are in order. This paper is not about “neutral principles” of adjudication, in the sense of “principled” decision making, a subject I explore in Kent Greenawalt, “The Enduring Significance of Neutral Principles,” Columbia Law Review 78 (1978): 82. What I say about constitutional law has relevance for common law and statutory interpretation and for matters of legislative choice; but I do not draw out the connections. 2. One possibility is that, on examination, values not well captured by the term “neutrality” will seem very important. Giving these effect might lead one to a resolution that is less easily called “neutral” than a competitor or even to a resolution that is not comfortably called neutral at all. 3. They illustrate the generalizations in showing how they fit them well. A critic is free, of course, to dispute both the generalizations and my views about religion and politics. 4. Stephen D. Smith, Foreordained FaiZure (New York: Oxford University Press, 1995). 5. Stanley Fish, “Mission Impossible: Settling the Just Bounds between Church and State,” Columbia Law Review 97, no. 8 (1997): 2255. 6. Smith‘s focus is more directly on law than is Fish’s, but when Smith analyzes possible development of the religious clauses according to theory, he directly engages broad theories about religious freedom. See Smith, Foreordained Failure, 55-1 17. 7. Ibid., 64-68; and Fish, “Mission Impossible,” 2258-69. 8. Reprinted in John Horton and Susan Mendus, eds., John Locke, ‘2 Letter Concerning Toleration” in Focus, Routledge Philosophers in Focus Series (London: Routledge, 1991). 9. However, Locke suggests that though the magistrate may not coerce about religion, he can persuade. 10. There are certain apparent limits to freedom-atheists cannot be trusted to keep promises, and perhaps Roman Catholics need not be tolerated because they owe allegiance to a foreign power-but the general stance is liberty of religious exercise. 11. Most obviously, if adults are effectively coerced not to express a particular belief, children are not likely to arrive at that belief. 12. Let me pause a moment over this last point. Fish claims that Locke’s conclusion about the law protecting calves shows that he does not value highly forms of worship that might run up against prohibitions adopted for reasons not related to religion. “Not only does Locke . . . equate” the harms of economic disadvantage for the ordinary farmer and inability to engage in the religious ritual, he “denies the severity of the religious harm by identifying true and essential religious practice with the internal notions of mind and heart.” See Fish, “Mission Impossible,” 2266-67. If Locke saw exter-
“Religion,Law, and Politics:Arenas of Neutrality”
277
nal actions as a vital aspect of religious practice, Fish implies that he would have considered whether an exemption should be given for practitioners of animal sacrifice. I am not so certain. Locke may have thought that the law could not tolerate a regime of exemptions or that no appropriate religious practices would ever raise this problem. Or, concentrating on laws directed against religious practices themselves, he may have conceded the example involving a general law that is not directed at religion without carefully considering whether an exemption would be preferable. 13. Jeremy Waldron, “Locke: Toleration and the Rationality of Persecution,”in John Locke, ‘2Letter Concerning Toleration” in Focus, Routledge Philosophers in Focus Series, ed. John Horton and Susan Mendus (London: Routledge, 1991), 98,99; and Fish, “Mission Impossible,” 2270-71. 14. Fish, “Mission Impossible,” 2271. 15. However, persons with these convictions, recognizing they are in a small minority, might accept religious liberty and nonestablishment as the best they can hope for in a society in which their religious opinions are not widely shared. 16. A crucial question in this regard is whether people can agree not only on some basic approaches, such as not establishing a state church and not restricting advocacy of religious ideas, but also that resolutions of debated issues can be reached without people repairing to their own background beliefs, relying instead on shared common grounds (i.e., shared by the 98 percent). I am skeptical about this possibility, and this skepticism is part of what I find less than persuasive in John Rawls’s idea of an overlapping consensus that will support decisions made by reference to principles of liberal democracy. See Kent Greenawalt, Private Consciences and Public Reasons (Oxford Oxford University Press, 1995), 106-20; Kent Greenawalt, “Some Problems with Public Reason in John Rawls’s Political Liberalism,” Loyola Law Review 28 (1995): 1303-17. 17. However, even if the 2 percent want a theocratic state, they may well prefer modern religious freedom to a theocracy of a majority religion sharply opposed to their own. See note 15. 18. Smith, Foreordained Failure, 82-83, criticizes the assumption in Epperson v. Arkansas, 39 U.S. 97 (1968) (a decision holding invalid a law forbidding the teaching of evolution) that teaching of a secular theory about the creation of life is neutral. 19. I tackle these problems in Kent Greenawalt, “Teaching about Religion and in Establishing Religious Ideas: Evolution, Creationism, and Intelligent Design” (forthcoming). 20. Fish, “Mission Impossible,” 2226. I am more hopeful than Fish that if people engage in close examination of their own ideas, they will be able to “recognize”crucial “decisions and distinctions.” 21. Smith, Foreordained Failure, 94. 22. Fish, “Mission Impossible,” 2256. 23. On the subject of whether “equality” is empty, my views are in Kent Greenawalt, “How Empty Is the Idea of Equality?” Columbia Law Review 83 (1983): 1167; and Kent Greenawalt, “‘Prescriptive Equality’: Two Steps Forward,” Harvard Law Review 1 10 (1997): 1265. 24. Of course, they could say that the ideal is to treat every true religion neutrally and every false religion neutrally, but to treat these two categories differently.
278
Kent Greenawalt
25. Some background beliefs may suggest that the government’s treatment of religion should not depend on the particular culture and prevailing religious beliefs within a society; others would treat those as relevant, either because the ideal government for a society would depend substantially on its culture and religion or because a tiny religious minority could not expect in fact to rule a country, even if its rule would otherwise be desirable. As part IV of this paper reflects, I think culture and religious beliefs may matter for what relations should be between government and religion. 26. This is not the occasion to explore the history of liberal political philosophy, but that history is rich, and many liberal theorists have not subscribed to the version Fish attacks. 27. Possibly, he means only that this particular liberal aspiration to fairness is impossible to attain. 28. Fish, “Mission Impossible,” 2257n4. 29. There are many truths that everyone does take for granted: for example, that 2 + 2 = 4, and that outdoors it is brighter when the sun shines than in the middle of the night. 30. Perhaps, my example is ill chosen. One might think that given a minimal level of intelligence and access to information, no one could reasonably deny the truth of the death camps. But what are we to say of someone who confidently asserts that Holocaust history is the product of conspiratorial control of media outlets? 31. I am offering here a variation of the common argument made against claims that truth is subjective or relative. One could respond that the claim “truth is subjective” is the one truth claim that is not subjective, and perhaps Fish can develop a similar rejoinder. 32. I once offered as a position that admits of a judgment of truth that is accessible to common human understanding: “unrestricted governance by sadists is undesirable.” See Greenawalt, Private Consciences and Public Reasons, 27. The idea was that if governing sadists used their own citizenry to satisfy their sadistic inclinations, that would be bad even for sadists (except those few in power). Fish generously calls this observation “as safe as it is unhelpful.” See Fish, “Mission Impossible,” 2262. What I think it establishes in the realm of political and moral philosophy is that there is at least some room for reasoned consideration of what is right and wrong that rises above particular cultural presuppositions and personal beliefs. And if that is possible for “safe” conclusions, it may be possible for less safe ones. 33. See Fish, “Mission Impossible,” 2319-22. He also suggests that recognition of this fact will not change our behavior. Ibid., 2325-33. But if we are persuaded by Fish, we will understand that what really counts are our objectives, the ones that we use principles to try to achieve. If we think other people are also persuaded to this understanding, might we not self-consciously calculate that contentions about principle will be less (or conceivably more) effective in achieving our objectives than we believed when we, and others, thought principles had intrinsic power? Even if we are the only ones persuaded by Fish, we might come to view principles in a different way, deciding to talk more directly about objectives, or perhaps being more selfconsciously manipulative. There is no conceptual reason to suppose our behavior would be completely unaffected by the insight Fish gives us; whether we would be af-
“Religion,Law, and Politics: Arenas of Neutrality”
279
fected is an empirical question, and Fish offers no empirical basis for his skepticism about that. 34. Ibid., 2330-32; and David M. Smolin, “Regulating Religious and Cultural Conflict in a Postmodern America: A Response to Professor Perry,” Iowa Law Review 76 (1991): 1067 (book review). 35. Imagine a tennis player who is playing in the finals of a tournament she desperately wants to win. Her opponent hits a ball that the referee calls out, but she sees clearly that the ball is in. According to her beliefs, winning is important but one should do so fairly, and she thinks (as many others do not) that taking advantage of an obviously mistaken call is unfair. She asks the referee to have the point played over (or she loses the next point on purpose to counterbalance the unfair advantage). There is nothing irrational or peculiar in her behavior. Similarly, many people think their pursuit of political and moral objectives should be constrained by notions of fairness. If all this is so, why should we assume principles of fairness are merely manipulative to other objectives? 36. City ofHialeah v. Church of the Lukumi Babalu Aye, 508 U.S. 520 (1993). I do not mean to imply that the result would be different if more than one church engaged in animal sacrifice. What is crucial is that the legislation revealed objection to this religious practice rather than a policy to protect animals or human health. 37. Some people do believe that Christianity should be favored over other religions, but even they do not think the government should favor any particular branch of Christianity. 38. I explore this problem in depth in Greenawalt,“Teaching about Religion and in Establishing Religious Ideas.” 39. See Smith, Foreordained Failure, 82-83. 40. Employment Division v. Smith, 494 U.S. 872 (1990). 41. Douglas Laycock advocates a view of “substantive neutrality” according to which “the religion clauses require government to minimize the extent to which it either encourages or discourages religious belief or disbelief, practice or nonpractice, observance or unobservance.” See Douglas Laycock, “Formal, Substantive, and Disaggregated Neutrality toward Religion,” De Paul Law Review 39 (1990): 993, 1001. Smith suggests that Laycock’s position rests on background beliefs that religious beliefs and practices are valuable, but only if voluntarily chosen. Smith, Foreordained Failure, 8 1. 42. See Kent Greenawalt, “Hands Off! Civil Court Involvement in Conflicts over Religious Property,” Columbia Law Review 98 (1998): 1843. 43. See Mitchell v. Helms, 530 U.S. 793,878-84 (2000) (Justice David H. Souter dissenting). 44. Mitchell v. Helms, 530 U.S. 793, 801-36 (2000) (plurality opinion of Justice Clarence Thomas). 45. A choice to have a state religion could be neutral in either of two senses. It could be neutral in purpose in regard to religious reasons, if it were based only on secular benefits of having a state religion. It could be neutral in underlying purpose about particular religions if the fundamental principle were to reflect the religion of a majority of the population.
280
Kent Greenawalt
46. My perspective on some of these positions is in Greenawalt,Private Consciences and Public Reasons. 47. Of course, it is possible to argue that any reliance on religious grounds is a kind of imposition; but that is a different kind of imposition. On that view, the same law might or might not be an imposition depending on the reasons it was adopted. 48. See especially John Rawls, Political Liberalism (NewYork Columbia University Press, 1993), 212-54.
Further Reading
Ackerman, Bruce. Social Justice and the Liberal State. New Haven, Conn.: Yale University Press, 1980. .“What Is Neutral about Neutrality?”Ethics 93, no. 2 (1983). .“Why Dialogue?”Journal ofPhilosophy 76 (1989). Adams, Robert. Finite and Infinite Goods. Oxford Oxford University Press, 1999. Alexander, Larry. “Liberalism,Religion and the Unity of Epistemology.” San Diego Law Review 30, no. 4 (1993). Alexander, Larry, and Maimon Schwarzschild.“Liberalism,Neutrality and Equality of Welfare vs. Equality of Resources.” Philosophy and Public Aficairs 16, no. 1 (1987). Ameson, Richard. “Neutrality and Utility.” Canadian Journal of Philosophy 20, no. 2 (June 1990). .“Perfectionismand Politics.” Ethics 11 1, no. 1 (2000). Audi, Robert. Religious Commitment and Secular Reasons. Cambridge: Cambridge University Press, 2000. .“The Separation of Church and State and the Obligations of Citizenship.”Philosophy andPublicAficairs 18, no. 3 (1989). Barry, Brian. “Contractual Justice:A Modest Defense.” Utilitas 8 (1996). -. “How Not to Defend Liberal Institutions.” In Liberalism and the Good, ed. B. Douglas et al. New York Routledge, 1990. .Justice As Impartiality. 2 vols. Oxford Clarendon, 1995. Bellamy, Richard, and Martin Hollis. Pluralism and Liberal Neutrality. New York: Frank Cass, 1999. Bird, Colin. “Mutual Respect and Neutral Justification.”Ethics 107, no. 1 (1996). Brighouse, Harry. “Neutrality,Publicity and State Funding of the Arts.” Philosophy and Public Aficairs 24 (1995). Caney, Simon.“Anti-Perfectionismand Rawlsian Liberalism.”Political Studies 43 ( 1995). -281
-
282
Further Reading
. “Consequentialist Defences of Liberal Neutrality.” Philosophical Quarterly 41, no. 165 (January 1991). . “Impartiality and Liberal Neutrality.” Utilitas 8 (1996). Chan, Joseph. “Legitimacy, Unanimity, and Perfectionism.” Philosophy and Public Affairs 29, no. 1 (Winter 2000). Clarke, Simon. “Contractarianism, Liberal Neutrality and Epistemology.” Political Studies 47 (1999). Cohen, Joshua. “Moral Pluralism and Political Consensus.” In The Idea of Democracy, ed. David Copp et al. Cambridge: Cambridge University Press, 1993. de Marneffe, Peter. “Liberalism, Liberty and Neutrality.” Philosophy and Public Affairs 19, no. 3 (Summer 1990). den Hartogh, Govert. “The Limits of Liberal Neutrality.” Philosophica 56 (1995). Dworkin, Ronald. “Foundations of Liberal Equality.” In Equal Freedom. Ann Arbor: University of Michigan Press, 1995. -. “Foundations of Liberal Equality.” In The Tanner Lectures on Human Values, vol. 11. Ed. Grethe B. Peterson. Salt Lake City: University of Utah Press, 1990. .“Liberalism.”In Public and Private Morality, ed. Stuart Hampshire. Cambridge: Cambridge University Press, 1978. -. A Matter of Principle. Cambridge, Mass.: Harvard University Press, 1985. Fish, Stanley. “Mission Impossible: Settling the Just Bounds between Church and State.” Columbia Law Review 97, no. 8 (1997). Galston, William. “Defending Liberalism.” American Political Science Review 76 ( 1982). .Liberal Purposes. Cambridge: Cambridge University Press, 1991. Gaus, Gerald F. Justificatory Liberalism. Oxford Oxford University Press, 1996. George, Robert. Making Men Moral. Oxford Oxford University Press, 1993. Goodin, Robert E., and Andrew Reeve. “Do Neutral Institutions Add up to a Neutral State?” In Liberal Neutrality, ed. Robert E. Goodin and Andrew Reeve. London: Routledge, 1989. Greenawalt, Kent. Private Consciences and Public Reasons. Oxford: Oxford University Press, 1995. Greene, Abner. “Government of the Good.” Vanderbilt Law Review 53, no. 1 (2000). Haksar, Vinit. Equality, Liberty and Pe$ectionism. Oxford: Oxford University Press, 1979. Hampton, Jean. “Should Political Philosophy Be Done without Metaphysics?”Ethics 99, no. 4 (1989). Hurka, Thomas. “Indirect Perfectionism: Kymlicka on Liberal Neutrality.” Journal of Political Philosophy 3, no. 1 (March 1995). .“Perfectionism.”In The Encyclopedia ofEthics, ed. Lawrence Becker. New York Garland, 1992. . Perfectionism. Oxford: Oxford University Press, 1993. Husak, Douglas. “Liberal Neutrality, Autonomy and Drug Prohibitions.” Philosophy and Public Affairs 29, no. 1 (2000). Jones, Peter. “The Ideal of the Neutral State.” In Liberal Neutrality, ed. Robert E. Goodin and Andrew Reeve. London: Routledge, 1989. Klosko, George. Democratic Procedures and Liberal Consensus. Oxford Oxford University Press, 2000.
Further Reading
283
Kraut, Richard. “Politics, Neutrality and the Good.” Social Philosophy and Policy 16, no. 1 (1999).
Kymlicka, Will. “Liberal Individualism and Liberal Neutrality.” Ethics 99 (July 1989). Larmore, Charles. “The Moral Basis of Political Liberalism.” Journal of Philosophy 96, no. 12 (December 1999) . Patterns ofMorul Complexity. Cambridge: Cambridge University Press, 1987. . “Political Liberalism.” Political Theory 18, no. 3 (1990). Lloyd-Thomas, David. In Defence of Liberalism. Oxford Blackwell, 1988. Macedo, Stephen. “Liberal Civic Education and Religious Fundamentalism: The Case of God v. John Rawls?”Ethics 105, no. 3 (1995). .Liberal Virtues: Citizenship, Virtue and Community in Liberal Constitutionalism. Oxford Clarendon, 1990. MacLeod, Colin M. “Liberal Neutrality or Liberal Tolerance?”Law and Philosophy 16 (September 1997). Mason, Andrew. “Autonomy, Liberalism and State Neutrality.” Philosophical Quarterly 40 (1990).
McCabe, David. “Knowing about the Good A Problem with Anti-Perfectionism.” Ethics 110 (2000). . “Liberal Education Is Moral Education.” Social Theory and Practice 21, no. 1 (1995).
Miller, David. Market, State and Community. New York Oxford University Press, 1990.
Montefiore, Alan, ed. Neutrality and Impartiality. Cambridge: Cambridge University Press, 1975. Nagel, Thomas. Equality and Partiality. Oxford: Oxford University Press, 1991. .“Moral Conflict and Political Legitimacy.” Philosophy and Public Affairs 16, no. 3 (1987).
Neal, Patrick. Liberalism and Its Discontents. New York: New York University Press, 1997.
Newey, Glen. “Metaphysics Postponed Liberalism, Pluralism, and Neutrality.” Political Studies 42 (1997). Nozick, Robert. Anarchy, State and Utopia. New York Basic, 1974. Nussbaum, Martha. “Human Functioning and Social Justice.” Political Theory 20 (1992).
Paris, David. “The Theoretical Mystique: Neutrality, Plurality and the Defense of Liberalism.” American Journal ofPolitica1 Science 31, no. 4 (1987). Price, Terry. “Epistemological Restraint Revisited.” Journal of Political Philosophy 8, no. 1 (2000).
Rasmussen, Douglas. “Human Flourishing and the Appeal to Human Nature.” Social Philosophy and Policy 16, no. 1 (1999). Rawls, John. “The Idea of Public Reason Revisited.” University of Chicago Law Review 64 (1997).
.Justice As Fairness:A Restatement. Ed. Erwin Kelly. Cambridge, Mass.: Belknap,
2001.
. Political Liberalism. New York Columbia University Press, 1993.
284
-.
Further Reading “Reply to Habermas.” Journal ofPhilosophy 92 (1995).
. A Theory offustice. Cambridge, Mass.: Harvard University Press, 1971.
Raz, Joseph. “Facing Diversity: The Case of Epistemic Abstinence.” Philosophy and PublicAfairs 19, no. 1 (1990). . “Facing Up: A Reply to Critics.” Southern California Law Review 62 (1989). .“Liberalism, Skepticism and Democracy.” Iowa Law Review 74, no. 3 (1989). . “Liberty and Trust.” In Natural Law, Liberalism and Morality, ed. Robert George. Oxford Oxford University Press, 1996. . The Morality of Freedom. Oxford: Oxford University Press, 1986. Sadurski, Wojciech. “Joseph Raz on Liberal Neutrality and the Harm Principle.” Oxford Journal ofLegal Studies 10, no. 1 (Spring 1990). .“Theory of Punishment, Social Justice and Liberal Neutrality.” Law and Philosophy 7 (1988-1989). Sandel, Michael. Democracy’s Discontent: America in Search of a Public Philosophy. Cambridge, Mass.: Harvard University Press, 1996. . Liberalism and the Limits of Justice. Cambridge: Cambridge University Press, 1982. Scheffler, Samuel. “The Appeal of Political Liberalism.” Ethics 105, no. 1 (1994). Schwartz, Adina. “Moral Neutrality and Primary Goods.” Ethics 83, no. 4 (1973). Sher, George. Beyond Neutrality: Perfectionism and Politics. Cambridge: Cambridge University Press, 1997. Solum, Lawrence. “Constructing an Ideal of Public Reason.” Sun Diego Law Review 30 (1993). Tomasi, John. Liberalism beyond Justice. Princeton, N.J.: Princeton University Press, 2000. Waldron, Jeremy. “Autonomyand Perfectionism in Raz’s Morality of Freedom.” Southern California Law Review 62 (1989). .“Legislation and Moral Neutrality.” In Liberal Neutrality, ed. Robert E. Goodin and Andrew Reeve. London: Routledge, 1989. . “The Theoretical Foundations of Liberalism.” Philosophical Quarterly 37 (1987). Wall, Steven. Liberalism, Perfectionism and Restraint. Cambridge: Cambridge University Press, 1998. .“Neutrality and Responsibility.” Journal of Philosophy 98, no. 8 (August 2001). Weinstock, Daniel M. “Neutralizing Perfection: Hurka on Liberal Neutrality.”Dialogue 38 (1999). Weithman, Paul. Religion and Contemporary Liberalism. Notre Dame, Ind.: University of Notre Dame Press, 1997. Wenar, Leif. “Political Liberalism: An Internal Critique.” Ethics 106, no. 1 (1995).
Index
Abington School District v. Schempp, 9, 26n14 Ackerman, B., 1,4-5, 11,4145,52,53, 62,185 anarchism, 155-56, 161 Aristotle, 15, 71 Arneson, R., 22-23,191-218 Audi, R., 188n36 Augustine, St., 44,45 autonomy, 18-19,43,52, 53,54,57,58, 66,73,81,84,96-97,98-99, 120, 121, 125-31,222,227-29,235-37
Baier, K., 158 Barry, B., 172-73, 174, 176, 185, 187n27, 206-07,210 Berlin, I., 72 Board of Education of Cincinatti v. Minor, 26n14 Bosanquet, B., 146 Buchanan, J., 155 burdens of judgment, 173-74,188n40 Burger, W., Chief Justice, 26n14 censorship, 220-21,222,223,227,229 Chan, J., 21,25n2,189n44
civil peace, 55-56 civility, 180-8 1, 182 Clark, T., Justice, 9, 26n14 coercion, 19,78,79,81-82,83,86, 122, 131-32,133,138-43,14548,155, 156,157,159,161-62,213; justification of, 196-201,20245; non-coercive pursuit of good, 19, 122-23,131-33,146-47 compensation, 8-9 comprehensive views. See conceptions of the good conceptions of the good, 1,2,5,6,7,8, 9,11-13,14,15,17-18,25n2,32-33, 41-42,48-55,58-59,62,65,67-68,
70,71,71-72, 73, 74, 78,81,92-102, 104-23,153-54,155,157,167-89, 192-216,223-24,262,268,270,275 consequential neutrality. See neutrality, of effect conservatism, 37-3 9 Constant, B., 50 constitutional essentials, 6,23, 72, 159, 20647,209-11,271-72 constitutional law, 9-10,24-25 contractualism, 77,94
- 285 -
286
Index
dialogue, rational, 53-54,55-60,61, 64-65 diversity. See pluralism Dworkin, R., 1,3, 12,23,31-39,62, 191, 192,196,201-05,220,229nl equality, 31-39 Establishment Clause (of First Amendment), 258,263-64,265-66, 269,273,274-75 exclusive position, on religion and politics, 268-69,275 experimentation, as value, 52,53,54,57, 58,100 experiments in living, 126 130,221,222, 224,225,226 Feinberg, J., 139 Fish, S., 259-63,276n12,277n20, 278n32,278n33 forms of life. See conceptions of the good Free Exercise Clause (of First Amendment), 258,263,265,269 freedom of expression, 219-30 Galston, W., 185n4,185n5,189n42 Gaus, G., 21-22,137-65 Green, T. H., 15,21, 133, 149 Greenawalt, K., 24-25,257-80 Haksar, V., 13-14,91-102 Hobbes, T., 80, 140, 148,252112 Hohfeld, W., 140,163n22 Humboldt, W., 21,126, 130 Hume, D., 45 Hurka, T., 1, 15, 125-33 impartiality, 77-87 inclusive position, on religion and politics, 268,270,275 integrity, 202-05 intermediate position, on religion and politics, 268,271,275-76 James, H., 226 James, W., 226
justification: moral, 142-43; of neutrality (see neutrality); public, 4-5,65,80-81,82,170,186n8 Kant, I., 32,44,45,50,52,73,74,81,85, 87,119,213,214,222,227 Klosko, G., 21-22,167-89 Kraut, R., 175 Kymlicka, W., 216n7 Larmore, C., 4,10,47-60,61,62,64,65, 184,195,213 Laycock, D., 279n41 liberal character, 63 liberal legitimacy, 2 11-1 3,2 18n26 liberal perfectionism. See perfectionism liberal virtues, 62,63,65 liberalism, passim: comprehensive, 72, 73-74,81; political, 48,67-74 liberty, equal, 95 liberty principle, 126-27 Locke, J., 50,252n2,253n12,259,265, 2761112 Macedo, S., 14,61-66 Marx, K., 15 Meiklejohn,A., 23,220,223,227,229 Mill, J. S., 21,23, 50, 52,73, 74,98-100, 101, 126, 130, 131, 146, 192, 193, 220-21,223,224,225,226,227,229 Montefiore, A., 106 moral point of view, 142-43 “morals” legislation, 18-19 Nagel, T., 5,77-87, 110, 111, 183,250 Nelson, W., 213 neutrality, passim: of aim, 70, 193-94; between religion and non-religion, 264, 265; comprehensive, 6-7; of consequences (see neutrality, of effect); of effect, 8-9, 70, 86, 138, 168,193-94,21OI216n7; history of, 2-6; of justification, 8-9,48-49, 138,167-89,193-94,195,210,
21 1-13; justification of, 11-13,
Index 21-22,42-45,51-54,191-218;
-,
“deductive approach,” 11-13; -, “ecumenical approach,” 11-13; -, neutral justification of, 11-12, 53-60,61,64-65; minimal, 139-49, 154; narrow, 6-7; of outcome, 48; principle, 6-1 1; -, scope of, 6-7; of procedure, 48-49,64,69-70; radical, 156-60; of religion, 9, 257-80 Nietzsche, F., 140 non-neutral state, 219-30 Nozick, R., 3, 103, 104, 105, 107-09, 155, 156,161 original position, 2-3, 109, 115-16, 118, 119 perfectionism, passim: alienation costs of, 20; Aristotelian, 128, 130; liberal, 17-21; philosophical, 16; state, 16; strong, 16-17; weak, 16-17 Plamenatz, J., 150 Plato, 43,44,45, 71 pluralism, 2, 11, 14, 17-18,48,51,62, 67,69,172,173,207-08,24245, 246-47,24849,251,275; moral, 103, 121-23 “political” conceptions, 5,67-68; of justice, 67-74 political liberalism. See liberalism public policy disagreements, 156-60, 177-82,182-84 public reason, 143, 171, 186118; wide, 171 Rawls, J., 2-3,5, 10, 13, 15,35,49, 67-74,81,96,98,103,109-11, 115-19,119-21,159,168,172,
173-74,176,186n8,206,262,271, 277n16 Raz, J., 1, 14-15, 17-18, 19, 103-24, 154, 216n6,228,253n25 reasonable rejection, 77,83,84, 149, 155, 167-89,211-12,213,247 reasons, impartial, 143-45, 154-60, 161
287
reflective equilibrium, 161 religion: neutral treatment of, 257-80; and politics, 24,268-276; -, exclusive position, 268-69,275; -, inclusive position, 268,270,275; -, intermediate position, 268,271, 275-76 respect, 54-58,62,63,65,87,91-92, 100,169,185n5,213-14,220,235, 237-45,246-47,250-51,252; for the good, 237-45,24647,250-51, 252 Rokeach, M., 157 Ross, W. D., 140 Rousseau, J.-J., 146 Scanlon,T., 23,96,220,221-23,227-29, 230n5 Sher, G., 1, 15,23-24,25n2, 146, 154, 219-30 Simmons, A. J., 155 “situational irresolvability,” 174-75, 187n27 skepticism, 42,52,54,58, 167, 168, 172-75,197,20549 Smart, J. J. C., 142 Smith, S., 259-61,264 Smolin, D., 262-63 social diversity, circumstances of, 232, 252n1 speech, free, 23-24 Spinoza, B., 15,252n2 Steiner, H., 159 Supreme Court, United States, 9-10, 171,186nll,258,263-67 Taft, A., Justice, 26n14 Taliban, 224-25 Tocqueville,A. de., 50 toleration, 24,61,62,77-87,91,95-102, 224,231-55; autonomy based, 235-37; paradox of, 233; perfectionist, 23 1-55 value claims, non-sectarian, 167, 175-85, 215,2181130
288 values, liberal, 62,63,65 veil of ignorance, 2-3 virtues: liberal, 61,64; political, 71 virtuous society, 37 Voltaire, 52 Waldron, J., 11,213,259
Index
Wall, S., 15-16,24, 145, 152, 154-55, 23 1-55 Wabv. Tax Commission of the City of New York,261114 Weinstock, D., 157-58, 159, 160, 161, 165n69 Williams, B., 252n10
About the Contributors
Bruce Ackerman is professor of law and political science at Yale Law School. Richard J. Arneson is professor of philosophy at the University of California at San Diego. Ronald Dworkin is professor of law at New York University. Gerald E Gaus is professor of philosophy at Tulane University. Kent Greenawalt is professor of law at Columbia University. Vinit Haksar is an honorary fellow and a former reader in the Department of Philosophy at the University of Edinburgh. Thomas Hurka is professor of philosophy at the University of Toronto. George KZosko is professor of politics at the University of Virginia. Charles Larmore is professor of philosophy and political science at the University of Chicago. Stephen Macedo is professor of politics and director of the University Center for Human Values at Princeton University.
- 289 -
290
About the Contributors
Thomas Nagel is professor of philosophy and law at New York University. John Raw& was professor of philosophy at Harvard University. Joseph Raz is professor of philosophy of law and a fellow of Balliol College, Oxford. He is also visiting professor at Columbia University. George Sher is professor of philosophy at Rice University. Steven Wall is an associate professor of philosophy at Bowling Green State University.