Truth, Autonomy, and Speech: Feminist Theory and the First Amendment 9780814784723

Citation preview

Truth, Autonomy, and Speech

CRITICAL AMERICA General Editors: Richard Delgado and Jean Stefancic White by Law: The Legal Construction of Race

To Be an American: Cultural Pluralism and the Rhetoric of Assimilation

Ian F. Haney López

Bill Ong Hing

Cultivating Intelligence: Power, Law, and the Politics of Teaching

Negrophobia and Reasonable Racism: The Hidden Costs of Being Black in America

Louise Harmon and Deborah W. Post

Privilege Revealed: How Invisible Preference Undermines America Stephanie M. Wildmand with Margalynne Armstrong, Adrienne D. Davis, and Trina Grillo

Jody David Armour

Black and Brown in America: The Case for Cooperation Bill Piatt

Black Rage Confronts the Law Paul Harris

Does the Law Morally Bind the Poor? or What Good’s the Constitution When You Can’t Afford a Loaf of Bread?

Selling Words: Free Speech in a Commercial Culture

R. George Wright

R. George Wright

Hybrid: Bisexuals, Multiracials, and Other Misfits under American Law

The Color of Crime: Racial Hoaxes, White Fear, Black Protectionism, Police Harassment, and Other Macroaggressions

Ruth Colker

Critical Race Feminism: A Reader Edited by Adrien Katherine Wing

Immigrants Out! The New Nativism and the Anti-Immigrant Impulse in the United States Edited by Juan F. Perea

Taxing America Edited by Karen B. Brown and Mary Louise Fellows

Notes of a Racial Caste Baby: Color Blindness and the End of Affirmative Action Bryan K. Fair

Please Don’t Wish Me a Merry Christmas: A Critical History of the Separation of Church and State Stephen M. Feldman

Katheryn K. Russell

The Smart Culture: Society, Intelligence, and Law Robert L. Hayman, Jr.

Was Blind, But Now I See: White Race Consciousness and the Law Barbara J. Flagg

The Gender Line: Men, Women, and the Law Nancy Levit

Heretics in the Temple: Americans Who Reject the Nation’s Legal Faith David Ray Papke

The Empire Strikes Back: Outsiders and the Struggle over Legal Education Arthur Austin

Interracial Justice: Conflict and Reconciliation in Post–Civil Rights America

Why Lawsuits Are Good for America: Disciplined Democracy, Big Business, and the Common Law

Eric K. Yamamoto

Carl T. Bogus

Black Men on Race, Gender, and Sexuality: A Critical Reader

How the Left Can Win Arguments and Influence People: A Tactical Manual for Pragmatic Progressives

Edited by Devon Carbado

When Sorry Isn’t Enough: The Controversy over Apologies and Reparations for Human Injustice Edited by Roy L. Brooks

John K. Wilson

Aftermath: The Clinton Impeachment and the Presidency in the Age of Political Spectacle

Disoriented: Asian Americans, Law, and the Nation State

Edited by Leonard V. Kaplan and Beverly I. Moran

Robert S. Chang

Getting over Equality: A Critical Diagnosis of Religious Freedom in America

Rape and the Culture of the Courtroom Andrew E. Taslitz

The Passions of Law Edited by Susan A. Bandes

Global Critical Race Feminism: An International Reader Edited by Adrien Katherine Wing

Law and Religion: Critical Essays

Steven D. Smith

Critical Race Narratives: A Study of Race, Rhetoric, and Injury Carl Gutiérrez-Jones

Social Scientists for Social Justice: Making the Case against Segregation John P. Jackson, Jr.

Edited by Stephen M. Feldman

Victims in the War on Crime: The Use and Abuse of Victims’ Rights

Changing Race: Latinos, the Census, and the History of Ethnicity

Markus Dirk Dubber

Clara E. Rodríguez

From the Ground Up: Environmental Racism and the Rise of the Environmental Justice Movement Luke Cole and Sheila Foster

Nothing but the Truth: Why Trial Lawyers Don’t, Can’t, and Shouldn’t Have to Tell the Whole Truth Steven Lubet

Critical Race Theory: An Introduction Richard Delgado and Jean Stefancic

Playing It Safe: How the Supreme Court Sidesteps Hard Cases Lisa A. Kloppenberg

Original Sin: Clarence Thomas and the Failure of the Constitutional Conservatives Samuel A. Marcosson

Policing Hatred: Law Enforcement, Civil Rights, and Hate Crime Jeannine Bell

Destructive Messages: How Hate Speech Paves the Way for Harmful Social Movements Alexander Tsesis

Moral Imperialism: A Critical Anthology Edited by Berta Esperanza Hernández-Truyol

In the Silicon Valley of Dreams: Environmental Injustice, Immigrant Workers, and the High-Tech Global Economy David N. Pellow and Lisa Sun-Hee Park

Success without Victory: Lost Legal Battles and the Long Road to Justice in America Jules Lobel

Greasers and Gringos: Latinos, Law, and the American Imagination Steven W. Bender

Mixed Race America and the Law: A Reader Kevin R. Johnson

Critical Race Feminism: A Reader, Second Edition Edited by Adrien Katherine Wing

Saving Our Children from the First Amendment Kevin W. Saunders

Elusive Citizenship: Immigration, Asian Americans, and the Paradox of Civil Rights John S. W. Park

Murder and the Reasonable Man: Passion and Fear in the Criminal Courtroom Cynthia K. Lee

Truth, Autonomy, and Speech: Feminist Theory and the First Amendment Susan H. Williams

Truth, Autonomy, and Speech Feminist Theory and the First Amendment Susan H. Williams

a NEW YORK UNIVERSITY PRESS New York and London

new york university press New York and London www.nyupress.org © 2004 by New York University All rights reserved Library of Congress Cataloging-in-Publication Data Williams, Susan Hoffman, 1960– Truth, autonomy, and speech : feminist theory and the First Amendment / Susan H. Williams. p. cm. Includes bibliographical references and index. ISBN 0–8147–9359–2 (cloth : alk. paper) 1. Freedom of speech—United States. 2. Feminist theory—United States. I. Title. KF4772.W55 2004 342.7308'53—dc22 2003026706 Earlier versions of some of the arguments in this book appeared in the following journal articles, and they appear here with the permission of the copyright holder: Susan H. Williams, Feminist Legal Epistemology, 8 Berkeley Women’s Law Journal 63–105 (1993) (© by the Regents of the University of California); Susan H. Williams, Truth, Speech, and Ethics, Genders (Nov. 1999) (electronically published, see http://www.genders.org/); Susan H. Williams, A Feminist Reassessment of Civil Society, 72 Indiana Law Journal 417–47 (1997).

New York University Press books are printed on acid-free paper, and their binding materials are chosen for strength and durability. Manufactured in the United States of America 10 9 8 7 6 5 4 3 2 1

For David, Benjamin, and Sarah

Contents

Acknowledgments Introduction

I 1

2

II

Foundations

xiii 1 11

The Theoretical Foundations of the Law of Free Expression: Truth and Autonomy A. Truth B. Autonomy C. Democracy

13 13 18 24

Philosophical Foundations: Cartesianism and Liberalism A. Truth B. Autonomy

32 32 41

Without Foundations

49

3

The Feminist Critique A. Truth B. Autonomy C. Connections to Feminist Themes

51 51 58 68

4

The Feminist Dilemma A. The Threats Posed by the Social Constructionist Critiques 1. Truth, Judgment, and Justification 2. Autonomy and the Nature of the Subject 3. Autonomy and Victimhood 4. Playfulness and Responsibility

71 71 71 76 78 81

ix

x | Contents

B. Feminist Strategies for Reconceiving Social Constructionism 1. Standpoint Epistemologies 2. Pragmatism 3. An Independent Moral Standard

III

Reconstruction

82 82 86 88 93

5

Reconstructing Truth A. The Functions of Truth 1. Shared Reality 2. Deep Critique 3. Decisions That “Work” 4. Connection to Nonhuman Reality B. A Relational Model of Truth 1. A Non-Cartesian Model 2. Connection to Feminist Themes a. Inclusion b. Relationality c. Blending Morality and Epistemology 3. The Functions of Truth a. Shared Reality and Deep Critique b. “Working” c. Connection to Nonhuman Reality C. Conclusion

95 95 96 99 100 103 105 105 113 113 116 116 118 118 122 126 129

6

Reconstructing Autonomy A. The Functions of Autonomy 1. Self-Respect, Self-Trust, and Self-Esteem 2. Character and Integrity 3. Personal Responsibility 4. Social Change and Democratic Politics B. A Narrative Model of Autonomy 1. A Brief Description 2. A Relational, Nonliberal Model 3. Connection to Feminist Themes 4. The Functions of Autonomy a. The Possibility and Meaning of Self-Direction in a Narrative Model of Autonomy

130 131 131 135 137 143 148 148 150 156 158 159

Contents | xi

b. The Specific Functions of Autonomy 1. Self-Respect, Self-Trust, and Self-Esteem 2. Character and Integrity 3. Responsibility 4. Social Change and Democratic Politics C. Conclusion

IV

Speech

164 164 165 167 169 172 173

7

Free Speech and Truth A. Shared Reality and Deep Critique B. “Working” C. Connection to Nonhuman Reality D. Conclusion

175 176 189 195 197

8

Free Speech and Narrative Autonomy A. Individual Speech as an Exercise of Narrative Autonomy 1. Symbolic Speech 2. Living One’s Story 3. The Nature of the Right B. Systems of Speech 1. Speech Systems and the Opportunity for Exercising Autonomy 2. Speech Systems and the Development of the Capacities Necessary for Narrative Autonomy C. Autonomy, Speech, and Democracy D. Conclusion

199

Conclusion: The Enlightenment Vision and the Interpretive Approach A. The Enlightenment Vision B. The Interpretive Approach C. The Interpretive Response to the Enlightenment Fears D. Speech E. Interdependence and Commitment

200 200 204 206 209 209

221 224 229

231 231 236 240 241 242

xii | Contents

Notes

245

Index

295

About the Author

317

Acknowledgments

I would like to thank the Indiana University School of Law–Bloomington for summer grants supporting this research. I would also like to thank my colleagues, in both law and gender studies, who were so generous with their time and interest when I presented parts of the book at several faculty colloquia at IU, particularly Judith Allen, Jeannine Bell, Roger Dworkin, Don Gjerdingen, Carol Greenhouse, Karen Hanson, and Lauren Robel. I am also grateful for the comments and criticisms offered by fellow panelists and participants at two meetings of the Law, Culture, and Humanities Association during which I presented early versions of certain chapters. In particular, Ed Baker, Vince Blasi, Susan Brison, Judy Failer, and Steve Heyman offered valuable insights and raised important concerns. Special thanks to Steve Heyman and Kate Baldwin, whose friendship, support, and intellectual companionship I have relied upon through so many conversations over so many years. I am also greatly indebted to the students in my seminar who offered thoughtful responses to the manuscript, and to the many research assistants who have contributed to this project, including Jessica Barth, Sean Bockover, Amanda Couture, Shannon McClellan, Kristy Murphy, and Christopher Saporita. And I am also grateful to my patient and extremely competent secretary, Sharon Nejfelt. Special thanks are due to former dean Alfred Aman Jr. and present dean Lauren Robel of the Indiana University School of Law–Bloomington. Their energy and enthusiasm have opened possibilities for me, as for so many others. Their friendship has kept me going in moments of discouragement. And their love for the academic enterprise, and for the community whose life it constitutes, has made them the kinds of deans that most academics can only dream of having. My deepest thanks go to my husband, David, and my children, Ben and Sarah, to whom this book is dedicated. Their loving support, both

xiii

xiv | Acknowledgments

practical and emotional, made it possible to write the book. David is and has always been my best and closest colleague. The breadth of his wisdom and the depth of his spirit are still a wonder to me after eighteen years. This book is about sustaining hope in the face of fear and finding strength in vulnerability. David’s courage and love have allowed me to see that such hope and strength are possible. In his love, and in Sarah’s and Ben’s faces, I find that ultimately sustaining joy on which all else depends.

Introduction

There are certain aspects of law—particularly constitutional law—that play a significant role in our cultural imagination. Regardless of whether or not these legal doctrines make much difference in terms of deciding cases or structuring legal relations, they make an enormous difference in the way we understand ourselves and our relationships to each other and to our government. They matter to us deeply because they represent our deepest commitments and capture our deepest fears. The First Amendment’s protection for freedom of speech is one of these cultural icons. Freedom of speech is an ideal that is close to the hearts of many Americans. It is connected to the hopes and dreams that shape our national identity, including individualism, progress, and liberty. But our commitment to freedom of speech presently rests on shaky foundations. We are using our free speech principle not so much to pursue our dreams as to evade our nightmares. We have justified freedom of speech by reference to particular ideas of truth and individual autonomy. Those ideas, however, are siren songs that lead us to sacrifice other values in our lives in the pursuit of an illusion of safety and certainty that we can never attain. We must reimagine truth and autonomy so that they can become vehicles for the hopes and dreams of an age that must live without such certainty. Through a critical examination and reconstruction of the justifications for free speech, this book will explore this corner of our cultural landscape and offer a vision for its renewal. In order to unravel the cultural meanings of the First Amendment, it is helpful to examine the justifications that have been offered for our extraordinary protection for freedom of speech. Part i: Foundations sketches the architecture of these justifications. Chapter 1 outlines two of the most common and respected of such frameworks: the truth theory and the autonomy theory. In its simplest form, the truth theory asserts that freedom

1

2 | Introduction

of speech is one of the best means of assuring human discovery of the truth. Because the discovery of truth is of such great social and individual importance, protecting free speech in order to achieve this goal is justified despite the admittedly serious costs that speech sometimes generates. The autonomy theory, in its simplest form, holds that free speech is one of the more important mechanisms for the exercise of individual autonomy. Since the purpose of government, and the foundation for its legitimacy, both include the protection of individual autonomy, the government is disabled from interfering with freedom of speech. These justifications help to explain why “speech is special.” The first chapter will provide a brief history of the development of each theory, a summary of the primary arguments for and against them, and a review of the evidence of their impact on the Supreme Court’s decisions. The chapter will then argue that another very common theory offered in justification of free speech—the democracy theory—should be understood as a variation on the truth and autonomy theories. The democracy theory is really a conjoining of two different arguments about the connection between speech and democracy. The first is a programmatic argument about the practical effects of speech on democratic decision making: when we hear all the speech available, we have more of the information we need to make good decisions in the democratic process. The second argument is about the relationship between speech and full participation as a democratic citizen: speech is a primary mechanism through which we participate fully in creating and shaping the public opinion that guides a democracy. Understood in this way, the democracy theory reduces to a special case of each of the other two theories. The first argument is a particular application of the general truth theory to the specific case of information to be used in democratic decision making. The second argument rests on a vision of democratic politics as one—particularly important—arena for the exercise of autonomy and speech as the primary mechanism for that exercise. Thus, chapter 1 will conclude that we can focus our attention on the truth and autonomy justifications and treat the democracy theories as variants of each. Upon closer examination, it becomes apparent that these justifications incorporate and rely upon particular conceptions of truth and autonomy that are central to our philosophical tradition and to our culture more generally. Chapter 2 examines these philosophical foundations in greater detail. The conception of truth at work in the truth theory of free speech rests on a particular epistemology, associated with the Enlightenment and

Introduction | 3

the work of Descartes. In this familiar view, truth is understood as objective and universal. Truth is not dependent on the perspective—let alone the identity—of the observer; it is verifiable by anyone using the appropriately rational techniques. The relationship between the knower and the thing known is a relationship between subject and object: the knower as a subject regards the known as an object of knowledge. It is truth understood in this way that is supposed to be the product of free speech. Similarly, the conception of autonomy in First Amendment theory rests on a particular form of liberal moral and political philosophy, with its foundations in the work of Locke and Kant. In this philosophy, autonomy is understood as a characteristic of individual human beings, a capacity of will that is independent of our relationships with other persons or our social or cultural context. We are autonomous when our actions— and perhaps even our values and characters—are internally generated rather than dictated by forces outside ourselves. It is this conception of autonomy that is supposed to be promoted and preserved by the protection of freedom of speech. These traditional, one might even say conventional, conceptions of truth and autonomy have come under attack recently from a variety of directions. The conception of truth has been criticized by pragmatists; the conception of autonomy has been criticized by communitarians; and both have been challenged by postmodernists. One of the most powerful of these critiques has been offered by feminist theorists. Feminists in a wide range of disciplines have argued that these conceptions of truth and autonomy are incoherent, impossible, and dangerous. The feminist critique is the subject of part ii: Without Foundations. Chapter 3 outlines the feminist argument against the traditional models. The feminist critique of truth—like those associated with some forms of postmodernism and pragmatism—grows out of a social constructionist challenge to Cartesianism. Social constructionism suggests that the search for objectivity (a “God’s eye view”) is both futile and misguided; it represents a refusal to accept the fundamental embeddedness of the human condition. This critique challenges the individualism, rationalism, and universalism of the traditional model. The feminist critique of the traditional model of autonomy is parallel. Here again, social constructionism calls into question the conception of the self on which the model is based. This argument suggests that the search for free agency or autonomy—an “uncaused cause”—is impossible and undesirable; it represents a refusal to accept the fundamental interconnectedness and dependency

4 | Introduction

of the human condition. Feminist theorists have demonstrated that these conceptions of truth and autonomy both suffer from a moral problem in addition to these conceptual flaws: these models incorporate and perpetuate gender hierarchy. Truth and autonomy—understood in the traditional ways—are biased toward male experience and away from the experience of many women. Indeed, these conceptions form part of the traditional justification for the subjugation of women. These critiques raise serious questions about the continued vitality of the truth and autonomy justifications for freedom of speech. If truth and autonomy are conceptually incoherent, practically impossible, and/or tools of gender oppression, then why should we want them? And if we don’t want them, then is there any reason to protect freedom of speech? In a world where we can no longer believe in these conceptions of truth and autonomy, how can we understand the function and cultural meaning of free speech? This book will offer answers to these questions. These answers begin from the assumption that truth and autonomy cannot simply be abandoned, despite the failures of their present, common conceptions. First, these concepts are critical to the feminist project of achieving greater freedom and equality for women. Feminist theorists, while rejecting the traditional models of truth and autonomy, cannot happily accept some of the consequences associated with social constructionism, particularly in its postmodernist forms. For example, some have seen judgmental relativism as a consequence of social constructionism. But feminists, who support a program of practical political reform, cannot simply dispense with the concept of truth in favor of a thoroughgoing relativism. We need to be able to assert that women really are oppressed, and not merely that this perspective is one valid point of view among many. Similarly, the destruction of the concept of the “self” and the undermining of categories for social analysis like “women” are both detrimental from a feminist viewpoint. Feminists, who are seeking change in a sexist society, cannot abandon the notion of individual agency or autonomy in favor of a complete, causal determinism. We need to understand how, even within a sexist society, individuals and groups can act self-consciously to bring about change. Chapter 4 describes the dilemma faced by feminist theorists who find themselves unable either to accept the concepts of truth and autonomy or to do without them. Moreover, the concepts of truth and autonomy are also critical to the projects of epistemology and moral and political theory more generally.

Introduction | 5

These concepts do real work in our struggle to understand the nature of human knowledge and to explore and justify our moral and political systems. The traditional models of truth and autonomy may be incapable of performing these tasks, but that does not mean that we should simply give up on the enterprise completely. Instead, feminist theory must undertake to reconceive both truth and autonomy in ways that preserve the important functions of these concepts without perpetuating the incoherent and damaging elements of the traditional versions. The first step in reconstructing these concepts is to think as rigorously and precisely as possible about what work they do. The second step is to design a revised model of each concept that can perform that work without compromising the insights of the feminist critique. Those projects occupy part iii: Reconstruction. Chapter 5 offers an analysis of why we value truth. What is the function of the concept of truth, both for social groups and for individuals? What is it that we would feel ourselves unable to do, that we want and need to do, if we did not have it? The concept of truth serves several functions, all of which are grounded in a recognition that reality is something we share with other human beings and that the process of creating and discovering reality generates certain obligations and responsibilities to those with whom we share it. First, the concept of truth allows us to construct a shared social reality and it facilitates joint decision making within a social context. Truth allows us to make claims on others to recognize and respond to the same reality we perceive. Second, the concept of truth provides grounds for a deep criticism of social conventions. Third, the concept of truth allows us to insist upon and recognize decisions that “work” as opposed to those that are simply of symbolic value. And finally, for many people, the concept of truth provides a foundation for their connection to a stable, nonhuman reality, such as nature or God. Chapter 5 then begins the process of constructing a model of truth based on a particular notion of interpretive agency and responsibility. In this relational model, truth is a product of a particular stance, not of detachment but of engagement, that accepts responsibility for the joint creation of reality. This model grows out of the social constructionist critique and fulfills the functions identified earlier in the chapter, but avoids the pitfalls discussed in chapter 2. The relational model of truth also grows out of concerns that are common to feminist analysis across a wide range of disciplines, such as the focus on excluded groups, the relational

6 | Introduction

nature of knowledge and identity, and the focus on agency and responsibility. In its explicit connections to issues of agency and moral responsibility, the relational model of truth begins the work of healing the rift between epistemology and moral/political theory, a rift that is one of the legacies of the traditional models. Chapter 6 continues this analysis and reconstruction by turning to the concept of autonomy. There are several functions served by the concept of autonomy, all of which are grounded in our sense that we have an active role in the creation of ourselves and our world. That active role is extremely important because it provides, first, a basis for self-respect, selftrust, and self-esteem; second, a basis for a conception of character that allows for the possibility of integrity; third, a foundation for moral responsibility and accountability to a community with shared norms; and fourth, a basis for self-conscious personal and social change and for a meaningful vision of democracy. This chapter then turns to the reconstruction of autonomy. The model of autonomy proposed here is a narrative model: autonomy is the ability to tell and retell one’s own story. Autonomy is therefore centered in the process of narration and interpretation rather than in acts of choice. Both creativity and responsibility are crucial to the exercise of narrative autonomy, as is participation in a community of shared normative meanings. Indeed, the centrality of a communal framework helps to explain the connection between autonomy and certain forms of democracy and lends support to the democracy variant of autonomy theory discussed in chapter 1. This chapter will outline the advantages, and limitations, of the narrative model of autonomy in fulfilling the functions of autonomy identified earlier and the connections between this model and important themes in feminist scholarship. The new conceptions of truth and autonomy can then form the foundation for a new approach to the theory and doctrine of freedom of speech. Part iv: Speech turns back to the theory of free expression. Both truth and autonomy can still provide justifications for a free speech principle, but the new models will alter the shape of the principle that emerges. Chapter 7 outlines a relational truth theory of speech. If we see truth as engaged and tied to moral responsibility, speech is still one of the most useful mechanisms for generating truth. Speech both creates and embodies the orientation toward others that is central to the search for such an engaged truth: an orientation of respect, responsibility, and commitment

Introduction | 7

to a search for reasons. A relational truth theory of speech suggests new approaches to many First Amendment issues, including the assessment of government purposes in regulating speech, the restriction of speech on government property (i.e., the “public forum” issues), and the need to open opportunities for challenges to power across both public and private institutions. Chapter 8 outlines a narrative autonomy theory of free speech. If we see autonomy as concerned with narrative or interpretive agency, speech is still deserving of protection because it is the primary mechanism for the exercise of interpretive creativity. Moreover, a narrative model of autonomy suggests that we must assess government regulation of speech systems not only for their impact on individual speakers but also for their contribution to the systemic concerns of a relational theory. Speech systems provide opportunities for the exercise of autonomy and also encourage the development of the many capacities necessary for narrative autonomy. These systemic concerns deserve a place in First Amendment analysis if the goal of free speech is to promote narrative autonomy. Finally, a narrative model of autonomy offers both support and some redirection for democracy theories of free speech. Areas of First Amendment doctrine that might be altered by a move toward narrative autonomy include campaign finance reform, the treatment of government subsidized speech, and symbolic speech. Thus, under either a relational truth or a narrative autonomy theory, speech is still special, but for different reasons and in different ways than we previously imagined. The conclusion will offer a broader view of the intellectual landscape in which the previous chapters operate. Seen from this vantage point, the traditional conceptions of truth and autonomy that were the subjects of critique reveal an underlying similarity. They are both preoccupied with issues of certainty and vulnerability; they attempt to mark out a territory that will be insulated from the vicissitudes that plague human life. Truth offers us a secure foundation for our knowledge of reality, a foundation subject neither to the limitations suffered by actual, concrete human beings nor to the moral and political beliefs of individuals and communities that have proven so changeable and contentious. Autonomy offers us a stable foundation for our identity as persons. This foundation secures our identity and responsibility against erosion by social forces, erosion that might lead either to the disintegration of identity and responsibility or to their expansion beyond comfortable bounds. Our traditional conceptions of truth and autonomy are responses to the perceived need for

8 | Introduction

such secure and insulated foundations. The mainstream traditions in both epistemology and moral/political theory are responding to similar perceived problems with similar solutions. Not surprisingly, the reconceptualizations of truth and autonomy offered in chapters 5 and 6 are also fundamentally connected. They are, indeed, simply different aspects of a single vision centered on a notion of interpretive agency and its attendant responsibility. Moreover, in this vision, human problems cannot be neatly divided into two separate and independent realms of epistemology and moral/political theory. Epistemology and moral/political theory appear in this approach as deeply interconnected, so that the most fundamental issues in each implicate and rely upon assumptions about the other. In considerations of epistemic responsibility, for example, it is impossible to escape concerns about what is right or good, while in considerations of moral responsibility, it is impossible to escape concerns about how and what one can know. These fundamental issues are a hybrid that can best be understood through categories that cross these disciplinary boundaries. Agency and responsibility—two concepts central to the interpretive model—are examples of categories that meld epistemological and moral/political concerns. The interpretive model does not degenerate into judgmental relativism and simple causal determinism, but it also does not reinstate the illusion of certainty and safety of the older conceptions of truth and autonomy. In place of insularity, the interpretive model offers interdependence. In place of certainty, the interpretive model offers commitment. This is a book about feminist theory, but it is also about the value of feminist theory to legal analysis. Legal scholars operate under the— largely salutary—discipline of knowing that their ideas could be used to decide issues of great importance in actual people’s lives. The practical impact of legal doctrine serves wonderfully to focus the mind on the issues at hand. But sometimes this focus obscures the larger intellectual landscape: the cultural battles being waged within the doctrine, the philosophical assumptions on which legal rules and concepts rest. Feminist theory offers law a lens to illuminate those assumptions and lines of conflict. Indeed, the particular assumptions and conflicts that feminist theory highlights are those that are otherwise least likely to be recognized because they blend so well with widely accepted background social and philosophical understandings. In those areas of law where the greatest attention has been paid to theory—such as the First Amendment—feminism can provide a particular

Introduction | 9

service. In addition to the usual tasks of showing how specific legal rules harm women, or fail to take into account our experiences, or contribute to the maintenance of gender hierarchy, feminism can engage with the theoretical foundations of the law. Through an analysis of the theory, feminism can show that even the ideals and conceptual categories of the law are gendered. In free speech theory, such an analysis leads us to the models of truth and autonomy underlying the most prominent theories of free speech. A feminist perspective on those models allows us to see that they are conceptually incoherent, practically impossible, and deeply implicated in gender hierarchy. These failures have their genesis in the compulsion for certainty and security at the heart of both of these traditional models. When we build our theories on the fear of uncertainty and vulnerability, we find that we have succeeded only in defining our problems in ways that make them permanently insoluble. Moreover, we purchase our illusory certainty at the cost of very real hierarchy. But the feminist analysis also highlights the legitimate needs that are filled by the concepts of truth and autonomy. Truth captures our commitment to create a joint reality, and the responsibility that such a commitment entails. Autonomy captures our dedication to having an active role in the creation of ourselves and our communities, and the responsibility that such a role entails. These insights are crucial, not only to the feminist project of gaining equality for women but also to the projects of epistemology and moral/political theory more generally. Feminist theory can, and should, take the next step and offer a positive program in addition to the critique. The legal and philosophical culture in which we find ourselves and with which we work has virtues as well as vices. It is neither prudent nor just to allow ideas like “truth” and “autonomy” to be held hostage by those who purvey the illusion of certainty and pander to the fear of vulnerability. We must salvage the best of the existing theory but place it in a new framework, a framework that grows out of women’s experiences and viewpoints but that offers liberation to all people. This book is an effort to contribute to that liberatory project.

part i

Foundations

This first part sketches the architecture of free speech theory. Chapter 1 outlines the truth and autonomy theories of free speech. This chapter provides a brief description of the development of each theory, a summary of the primary arguments for and against them, and a review of their impact on the Supreme Court’s decisions under the free speech clause of the First Amendment. Chapter 1 will also argue that the other common free speech justification—the democracy theory—is best understood as a variation on these two primary theories. Chapter 2 provides a more detailed examination of the models of truth and autonomy underlying each theory. First, I describe the Cartesian model of truth and trace its roots in epistemology and its implications for our understanding of ourselves and our knowledge. Then, I explore the liberal model of autonomy, including the moral and political theory from which it is derived and the conception of the self on which it rests. By the end of part 1, the foundations of free speech theory will be uncovered and charted, allowing us to see the vision that presently serves as our justification for freedom of speech and its connection to larger cultural movements and ideas.

11

2 Philosophical Foundations Cartesianism and Liberalism

The truth and autonomy theories of free speech are small parts of the larger philosophical edifice of Enlightenment thought. In order to see the connections between the two theories, it is helpful to explore the philosophical assumptions on which they are constructed. This chapter will outline the models of truth and autonomy at work in the theories and trace the evidence for each model in the philosophical and legal materials. The primary point in this exposition is to demonstrate that, in their present forms, the truth and autonomy theories of free speech rely upon a Cartesian model of truth and a liberal model of autonomy and to highlight the parallel features and concerns of those models.

A. Truth The truth theory of free speech rests on a particular epistemological foundation that shapes the nature of its strengths and weaknesses. Since the Enlightenment, epistemology, both in and outside of law, has been developing along lines consistent enough to generate a mainstream tradition. Indeed, this epistemological stance has come to be so widely accepted and so much a part of many of our social institutions that it is almost invisible to us. As a result, lawyers and judges operate largely unself-consciously within this epistemology, but philosophers have examined its assumptions in great detail. This mainstream tradition is Cartesianism. Cartesianism may best be understood as a family of theories. There are many issues either unresolved or contested within the range of Cartesian theories, but the many disagreements that exist have been fought out largely within the assumptions of Cartesianism, rather than as challenges

32

Philosophical Foundations: Cartesianism and Liberalism | 33

to those assumptions. Thus, it is reasonable to describe Cartesianism as the mainstream tradition in epistemology. In Cartesian epistemology, truth is a quality of beliefs, preferably beliefs that can be expressed in the form of propositions. A true belief accurately describes reality. Reality, in turn, is understood as objective; that is, as existing independent of human perception or understanding of it.1 This objective reality is available to individual knowers through the use of their reason, sometimes combined with their sense perception. The knowledge attained is universally true, rather than true merely for a particular person in a particular time and place. This view gives rise to a series of dichotomies that have formed a mainstay of philosophical speculation and have had a dramatic effect on our cultural imagination. The cumulative effect of the Cartesian vision and its associated dichotomies is a model of knowledge in which to know something is to exercise power over it, to dominate or control it. Feminist philosophers Alison Jagger and Susan Bordo have compiled a useful list of the assumptions that make up the Cartesian model.2 The first assumption associated with Cartesianism is that reality has an objective nature, that is, a nature independent of human understandings of it. Reality is simply “out there” and its character is unaffected by whether we recognize or understand it. Second, this objective reality is, at least in principle, accessible to human knowledge. This position, when combined with the first assumption, gives rise to the dominant theory about the nature of truth in Western philosophy and culture: the correspondence theory of truth. The correspondence theory holds that a proposition is true if and only if it accurately describes reality.3 Truth is only about such objective realities. Truth is, therefore, fundamentally independent of moral and political goals, beliefs, or values. For example, there is a fact of the matter—a truth—about whether there is a table in front of me and that fact is unaffected by whether or not I believe in equality or happen to find myself in a democratic society. Thus, one manifestation of objectivity is the famous fact/value distinction. Third, Cartesianism assumes that people approach the task of gaining knowledge individually rather than as socially constituted members of particular groups. In the form most relevant to this argument, this “epistemological individualism” means that the tools or characteristics necessary for the pursuit of knowledge exist in individual human beings considered independently of the particular social context that they may

34 | Philosophical Foundations: Cartesianism and Liberalism

inhabit. For example, people’s exercise of their sensory organs arguably can be understood without reference to their particular social context, while their aesthetic sense arguably cannot. As a result, sense data would qualify as facts to be known, while aesthetic judgments are seen as matters of taste rather than as matters of knowledge. Fourth, Cartesianism exhibits a “rationalist bias.” That is, Cartesianism assumes that the primary faculty through which human beings gain knowledge is their reason.4 In some types of Cartesianism, this faculty is strongly supplemented by the use of the senses, while in other types reason is held to be quite efficacious on its own.5 But even where sense data are considered necessary additions, reason is the faculty used to assess that data and to acquire knowledge from it. Reason is understood as a human faculty fundamentally unconnected to a particular social or physical context. What one can see may depend on where one stands, but what one can logically deduce from a given bit of information does not. Reason is instrumental and abstract. This conception of reason leads us to the fifth aspect of Cartesianism: universalism. The knowledge attained through the proper use of these faculties is true for all people; it is universal or neutral.6 “Differences in the situations of human beings, rather than being recognized as providing alternative perspectives on reality, are seen as conquerable impediments to a neutral, ‘objective’ view of things.”7 There cannot be multiple, valid truths; on any given issue, there is only one truth.8 All other perspectives are more or less false due to a greater or lesser degree of failure properly to exercise the capacities for reasoning or gathering sense data. Finally, this view of truth contains a particular understanding of language as well. Language is the vehicle through which truth is represented. Words paint a picture that corresponds to the external reality, or fails to do so. Language is representational rather than performative. These Cartesian assumptions are closely associated with a series of dichotomies that permeate not only the philosophical tradition but also popular culture. The dichotomies include mind/body, culture/nature, universal/particular, reason/emotion, and objective/subjective. It is not difficult to see how the Cartesian assumptions are related to these dichotomies. The rationalist bias obviously places a premium on distinguishing reason from other faculties, like emotion. Similarly, the belief in an external reality independent of human understanding and in the need for knowledge to be universal and neutral leads naturally to the di-

Philosophical Foundations: Cartesianism and Liberalism | 35

chotomies between objective and subjective beliefs or perspectives and between universality and particularity. The mind/body dichotomy is explained by the belief that knowledge, because it is pursued through reason, is understood as an attribute of the mind rather than of the body. The body, on the other hand, is the site of all the innumerable particularities that mire us in emotion and subjectivity. Finally, the culture/nature dichotomy often functions as a summation of all of the previous dichotomies. Nature represents all that is physical, moved by emotion or instinct rather than by reason, sunk in subjectivity and particularity. Culture is the triumph of mind and reason, imposing objective and universal constraints—perhaps most clearly, although not exclusively, in the form of law—over these forces of chaos, danger, and ignorance. Nature may be the nonhuman world—the resources and raw materials, along with the plagues and natural disasters—against and over which man stands as the representative of culture. But nature may also be people: the “barbarian” hordes of another nation, the subset of our own population in need of control (e.g., women, the poor, minorities), or even the part of each individual that sometimes threatens to overwhelm his reason. In other words, the nature/culture distinction does not, as it might at first appear, mark the boundary between human beings and the rest of existence. It constructs, instead, the boundary between the orderly and productive realm in which reason and objectivity rule and the confused, inarticulate, and possibly dangerous area beyond the wall, which has yet to be subdued. Human beings can, and do, live on both sides of that wall. These dichotomies, then, spell out with greater specificity some of the implications of the Cartesian assumptions. They point to the values enshrined by that view of knowledge and to the dangers or failures to be avoided. Because of their extremely deep and pervasive role in Western culture, the dichotomies also illuminate how extensively Cartesian assumptions have broken the bounds of professional philosophy and permeated popular ideas. One important result of these assumptions and dichotomies is the connection forged in Cartesianism between knowledge and power. There are at least two ways in which the process of acquiring knowledge is related to control or power. The first, and often unrecognized, connection is that control is a necessary aspect of the knower’s relation to himself. The activity of gathering knowledge requires the knower to exercise control over those parts of himself and his environment that might

36 | Philosophical Foundations: Cartesianism and Liberalism

interfere with his access to an objective reality. He must restrain his emotions so that they do not cloud his reason, and he must reduce his particularities—as a judge, for example, assumes an institutional role that distances him from personal experiences that might affect his view of the issues before him—so that he can reach a more neutral or universally valid answer. The second sense in which Cartesianism links knowledge and control concerns the relationship between the knower and the known: the knower exercises power or control over the known. The external world, the things to be known, are constructed on an analogy to the part of the self to be subdued. Those things are conceived as passive, not in the sense of being inactive, but in the sense of being reactive rather than self-initiating. They are, therefore, subject to prediction and often to manipulation. Reason, which is the key to autonomy and freedom from determinist particularity, is available to the knower but not to the known. Perhaps the impact of the Cartesian assumptions and dichotomies can be summarized by saying that they posit the thing known as an object rather than a subject in its own right.9 This result frees the knower to exercise power or control over the thing known for his own ends, rather than to consider and respect its ends or establish a community of shared ends, as one would hope he would feel required to do with another subject. Since other people can be among the things known, it is possible, of course, that a person could act as the object of knowledge at one time and as a knower herself at another time. It is even possible that she could act as knower and known simultaneously. What is not possible is that she should be seen by her knower as standing in a relationship in which knowledge, and therefore power, flows in both directions. That is, to the extent that someone or something is the object of knowledge, she/it will be treated as an object. So, to summarize: in the Cartesian model, truth is objective, universal, and singular. It consists of representations of an external reality that is independent of human moral and political values, indeed, independent of human understanding generally. Reason is the primary faculty through which human beings access this truth. And language—particularly propositional speech understood as representation—is the primary vehicle through which this truth is comprehended and expressed. This system of epistemological assumptions gives rise to a series of dichotomies and supports a model of knowledge as power. It generates a distant and hierarchical relationship between the knower and the known and promises

Philosophical Foundations: Cartesianism and Liberalism | 37

knowledge with a high degree of certainty and generality. In exchange, it exacts from the knower a sense of deep fragmentation both in himself and in the world more generally. This Cartesian model of truth lies at the foundation of the truth theory of free speech. It is because speech is useful—perhaps even essential—to the discovery and spread of this type of truth that the truth theory argues it must be protected. I will briefly canvass some of the evidence for this model of truth in the writings of John Stuart Mill and in the free speech decisions by the Supreme Court. Although my interpretation of these materials is hardly uncontroverted, it is, I believe, an interpretation so established and common as not to require extended documentation.10 In his defense of freedom of speech in On Liberty, Mill relies implicitly on an objective sense of truth. His faith in an objective reality is the foundation for his argument that truth’s real advantage over falsehood is that, regardless of how often it is suppressed, there will always be people to rediscover it until it finds a setting in which it can be freely aired.11 It is because people can access an independent reality—despite their contrary culture—that they continue to rediscover these unpopular truths.12 Mill also embraces reason—understood as the process of argumentation that challenges and corrects our views—as the primary mechanism through which truth is apprehended.13 He is also quite clear about his faith that truth is universal, in the sense that there is on each issue a single truth that is in principle accessible to all. “As mankind improve, the number of doctrines no longer disputed or doubted will be constantly on the increase: and the well-being of mankind may almost be measured by the number and gravity of the truths which have reached the point of being uncontested.”14 Finally, Mill’s entire argument rests on the assumption that truth will take the form of propositions—claims to describe reality through language—that can then be contradicted, disputed, and subjected to logical analysis. Although he recognizes a wider range of subjects for such claims than many other Cartesians, he shares the Cartesian assumption about the form of knowledge. The Supreme Court has referred explicitly to Mill as a source for its own reliance on truth as a justification for freedom of speech. Although Supreme Court opinions rarely provide systematic expositions of underlying legal theory, the Court has relied upon the truth theory often enough to supply some evidence of the nature of the truth it has in mind. First, many members of the Court have believed that the truth with which the First Amendment is concerned is to be pursued through reason. Speech is

38 | Philosophical Foundations: Cartesianism and Liberalism

often held to have lower value when it is, as the Chaplinsky Court stated, “no essential part of any exposition of ideas, and of such slight social value as a step to truth.”15 Justice Brandeis made the connection to reason even clearer in his Whitney concurrence, when he attributed to the framers a “[b]elie[f] in the power of reason as applied through public discussion.”16 The Court also seems to assume that the truth we are pursuing concerns an objective reality. The evidence on this point is quite indirect because the assumption of an objective reality is so much a part of our background culture that it is seldom even recognized directly, let alone defended in an explicit way. As a result, the Court’s references, described below, are not inconsistent with other possible models of truth, but they do fit most comfortably within the common view of truth as representations of an objective reality. For example, the Court has held that a statement must be “provable as false before there can be liability under state defamation law.”17 The Court recognizes that not all statements concern “facts” that can be proven true or false. In its reliance on a version of the fact/value distinction, the Court is implicitly incorporating a view of truth as concerned with the representation of an objective reality that is, at least in principle, “provable.” Indeed, the Court itself refers to the type of evidence to be used in such a proof as “objective.”18 The use of the word “provable” also suggests both a rationalist bias (proven how? through reason) and universalism (proven to whom? any reasonable person). The universalist view is perhaps most forcefully stated by Justice Holmes in his assertion that “the best test of truth is the power of the thought to get itself accepted in the competition of the market.”19 The idea that truth will win in a free market of ideas rests on the assumption that people are capable of recognizing and accepting truth regardless of their social circumstances. The results of the market will, therefore, reflect the underlying objective reality rather than merely the social forces pressing toward one position or another. The marketplace metaphor captures the faith that all (or at least most) people are able to recognize such objective truths. The difficulty, of course, is that Justice Holmes himself probably did not see his market metaphor this way. Given his pragmatic and skeptical leanings, it is far more likely that Holmes intended a less objective meaning of truth and a less universal level of recognition.20 His own position on the value of free speech may have had much more to do with democracy than with truth.21

Philosophical Foundations: Cartesianism and Liberalism | 39

Despite its creator’s probable views, however, the marketplace metaphor has come to be the symbol for the faith in an objective reality, the human capacity to comprehend it, and the centrality of free speech to this enterprise. For example, in Milkovitch, the Court, in the course of asserting that the First Amendment does not make a distinction between opinion and fact, quoted with approval from a lower court opinion by Judge Friendly in which he argued that the marketplace of ideas metaphor “points strongly to the view that the ‘opinions’ held to be constitutionally protected were the sort of thing that could be corrected by discussion.”22 As the Court put it, “expressions of ‘opinion’ may often imply an assertion of objective fact.”23 It is their connection to such objective facts that makes such expressions capable of correction by discussion and that justifies the marketplace approach. Finally, the Court’s approach operates on the assumption that truth will take the form of propositional language: claims to represent reality. It is these claims that are the competitors in the marketplace of ideas. Learned Hand identified this “assumption—itself an orthodoxy, and the one permissible exception [to the rule against orthodoxies]—that truth will be most likely to emerge if no limitations are imposed upon utterances that can with any plausibility be regarded as efforts to present grounds for accepting or rejecting propositions whose truth the utterer asserts or denies.”24 This preference for truths expressed through representational language is perhaps clearest in the Court’s discomfort with nonpropositional forms of communication. The Court’s devaluation of symbolic speech—treating many regulations that impact the message of such speech as though they were no more than time, place, and manner regulations—is some evidence of this discomfort.25 This lack of sympathy is undoubtedly related to the rationalist bias: such nonpropositional forms of speech seem to appeal to faculties other than reason and are more difficult to assess or counter with rational argument. Many members of the Court seem to feel that nonverbal or nonrepresentational forms of expression simply do not advance the search for truth in a meaningful way.26 My assessment of the Court’s opinions focuses on the Court’s faith in the Cartesian program for human knowledge. A powerful argument can be made, however, that the Court has given up the hope for such knowledge in favor of a skeptical view resting on the fallibility of human judgment and the subjectivity of at least some types of knowledge.27 I believe that this alternative perspective, however, simply provides additional

40 | Philosophical Foundations: Cartesianism and Liberalism

support for my claim that the Court has adopted a basically Cartesian view of the nature of truth and knowledge. Skepticism—if that is the Court’s position—shares the picture of truth I have outlined above; it simply does not share the optimism about it. The skeptic doubts either that human beings could ever attain such truth, or perhaps even that it could exist. But he does not doubt that this is the meaning of truth; that is why he finds that its impossibility undermines all knowledge claims.28 As critics of Cartesianism have pointed out, skepticism is the flip side of Cartesian faith; it is not a different view of truth but simply a pessimistic version of the same view.29 Therefore, if the Court has embraced skepticism, that would provide support for my claim that it is relying upon a Cartesian view of truth.30 Thus, hints of all of these elements of Cartesianism can be found in the opinions of the Supreme Court interpreting the free speech clause of the First Amendment. They are, however, often no more than hints because the Court was not engaged in the self-conscious development of an epistemological foundation for free speech theory. The Court was occupied with the business of deciding cases and interpreting the Constitution, and one must read between the lines to find the epistemological assumptions that guided the justices in those pursuits. Nonetheless, I believe that it is a fair characterization of the Court’s position to say that it is using what might be called “naive Cartesianism”: an acceptance of the general epistemological framework shared by its society without much attention to its possible deficiencies. That framework includes the objectivism, rationalism, universalism, and linguistic representationalism of the Cartesian model of truth, and all of those elements make an appearance in the truth theory of free speech, as it is presented in the opinions of the Supreme Court. Underlying the truth theory of free speech, then, is a very particular model of truth, with highly systematic assumptions and implications. That model, in its less systematic forms, is one shared by Western culture generally. In its more systematic forms, the model has been the subject of criticism, defense, and revision by philosophers and others. Thus, the Cartesian model has powerfully shaped our intellectual tradition, our popular culture, and our legal theory—all of which form the foundation for the truth theory of free speech.

Philosophical Foundations: Cartesianism and Liberalism | 41

B. Autonomy The autonomy theory of free speech rests on a closely related foundation. The traditional notion of autonomy, like that of truth, encompasses both a popular (and somewhat exaggerated) version and more nuanced versions developed by philosophers. In this chapter I will sketch the broad outlines that are shared by many of the variations, both popular and scholarly. In the next section of the book, when I take up the criticisms of this model, I will develop the qualifications and modifications offered by some philosophers. The word “autonomy” means “self-rule,” and the heart of the concept concerns notions of self-determination. As Joel Feinberg has noted, there are at least four uses of “autonomy” in the philosophical literature. “It can refer either to the capacity to govern oneself . . . ; or to the actual condition of self-government and its associated virtues; or to an ideal of character derived from that conception; or . . . to the sovereign authority to govern oneself. . . .”31 While these different uses of the concept do affect the shape of its description, there are certain themes that are consistent throughout most of the versions of autonomy. This section will trace those themes. The traditional conception of autonomy is deeply embedded in liberal moral and political theory, with foundations in the work of Immanuel Kant, John Locke, and John Stuart Mill. This liberal conception of autonomy focuses on choice: self-determination is understood as exercised through the process of choosing.32 The central distinction on which the traditional concept rests is between choices originating within the person, which may be autonomous, and choices made in response to external compulsion, which are not.33 The idea is that certain kinds of choices are legitimately the agent’s “own” and that only such choices count as autonomous. The agent who makes such choices is generally conceived as a fundamentally unified self, capable of knowing and reflecting on his concerns and desires. The process of reflection tends to be focused in many accounts on reason and rationality. And the capacities necessary for such reflection—while they may be historically the products of socialization— are understood in a fundamentally individualistic way. The first aspect of the traditional conception of autonomy is that it is focused on the capacity for choice. The autonomous agent is a chooser, and it is his choices that are to be understood as meeting or failing to meet the requirements of autonomy. “An autonomous person is one who

42 | Philosophical Foundations: Cartesianism and Liberalism

chooses for herself what to think and what to do. The values, standards, principles and motives which determine her choices are, in an important sense, her own values, standards, principles and motives.”34 Autonomy, in this traditional model, is centrally concerned with the way one exercises the capacity for choice. What distinguishes autonomous choices, in this conception, is that they are in some sense the agent’s own. The precise meaning of this ownership is the subject of much philosophical speculation, but the most common formulation requires that the causes for the choice belong to the agent rather than to forces outside the agent. The opposite of autonomous choice, in this sense, is coercion. These conceptions of autonomy focus on the process of choice rather than the substance of the choices made in determining whether the choice was autonomous.35 What is it about the process that makes some choices autonomous? Some theories identify the historical genesis of the causes as the critical issue: choices that result from certain types of illegitimate interference with the agent, such as brainwashing, are not autonomous. Other theories focus on the agent’s own sense of identification with the desires and motives that move him to action: only if the agent forms a “secondorder” desire to claim the underlying motives as his own is the choice autonomous.36 I will examine these alternatives in greater detail in the following section of the book, but the central point for present purposes is that these models propose that some act internal to the individual chooser is the foundation of autonomy. Such process-focused models have struggled with two major difficulties: the metaphysical problem of free will and the psychological problem of socialization.37 The metaphysical problem concerns how to understand the possibility of free will or, conversely, how to make such a focus on choice consistent with some versions of causal determinism.38 The socialization problem concerns how to understand the agent’s motives and values—the most common internal causes of action—as “her own” in light of the prevalence of socialization. The socialization problem is best understood as a particular case of the metaphysical issue of determinism. If a person’s motives are themselves the causal products of forces of socialization over which she has no control, then in what sense is the action that proceeds from those motives autonomously her own? The problem lies in the need to break the deterministic causal chain in order to make room for some act internal to the agent on which autonomy might rest.

Philosophical Foundations: Cartesianism and Liberalism | 43

In many forms of autonomy theory, the primary act that plays this role is an act of reason. This focus on reason takes its most extreme form in the work of Kant, where all motives and concerns other than reason are regarded as heteronomous, and only choices guided by reason qualify as autonomous.39 Even in more modern theories, however, the emphasis is often placed on rational assessment and evaluation of one’s motives and values as the primary path to autonomy.40 Reason is what frees us from determinism and returns control over our choices to our own will. There is a particular conception of the person implicit in these models of autonomy. First, the person is seen primarily as an agent: a doer and a chooser. It is this role that confers the moral status of personhood—that makes one, in Kant’s terms, an “end in itself.” Second, the person is essentially self-transparent; that is, people are effectively able to look into themselves and see what their motives and values are in order to reflect and act upon them. Knowledge about other people’s minds has long been a thorny philosophical problem, but the assumption is that, in general, no such barriers interfere with a person’s knowledge of her own mind. We may sometimes struggle to overcome desires in ourselves that we do not like, but recognizing and identifying our desires is, in normal cases, a fundamentally straightforward process. Moreover, it is a process in which no other person could possibly be in a better situation to know one’s mind than oneself. Third, the person is fundamentally unified. We may sometimes have conflicting desires, of course, but we must possess essentially unified systems of values and desires if we are ever to rely on such systems to generate our choices. Indeed, the process of autonomous reflection can be understood as a form of internal housekeeping, in which we bring a greater and greater degree of order and coherence to our motivational structure. Many autonomy theories therefore identify personal integration as a hallmark of autonomy.41 Finally, the focus on choice, will, and reason leads to a conception of the person with a pronounced individualism. In some sense, personal autonomy is an inherently individualistic concept: it refers to the autonomy of individual human beings rather than to the autonomy of social groups. But the individualism of traditional autonomy theories goes beyond this rather trivial sense. Many modern autonomy theorists do recognize the contribution made by social relationships to the formation of autonomy: we learn to be rational through social interaction.42 But such theorists generally see the exercise of autonomy itself as a fundamentally individual

44 | Philosophical Foundations: Cartesianism and Liberalism

activity. That is, once a person has been trained through her social relationships to reflect rationally, the capacities that she uses to act autonomously are all comprehensible without reference to her continuing social relationships. She could, in theory, go on being autonomous all alone on a desert island. Indeed, if she could have been trained to rationality by a computer, rather than by human teachers, she could have been autonomous without ever having known another human being.43 The capacity of will, for example, is “in no way depend[ent] on the existence or character of either our relations with other persons or our social situation.”44 Thus, the mainstream conception of autonomy is individualistic in the sense that it contains within it no necessary reference to or reliance upon the social relations in which persons are embedded.45 Thus, the liberal model of autonomy posits an agent who is essentially unified and capable of relatively transparent self-knowledge. This agent exercises his autonomy through his role as chooser and doer: it is through the activities of choice and action that autonomy is actualized. The capacities most central to this exercise of autonomy are related to reasoning. Moreover, these capacities and the autonomy they facilitate are understood in a fundamentally individualistic way: the social relations and context of the agent are no part of the necessary description or understanding of the concept of autonomy. The liberal conception of autonomy provides the foundation for the autonomy theories of free speech. This family of theories focuses on the ways in which speech is connected to autonomy. There are three forms of autonomy theory. In the first form, speech is protected because it is instrumentally valuable to autonomy. This value may lie in the information that speech makes available to an agent, allowing her to make autonomous choices more effectively, or it may lie in the way in which the agent’s participation in speech develops the personal capacities—like reason and self-knowledge—necessary for autonomous action. In the second form of autonomy theory, speech is seen as itself an exercise of personal autonomy and is protected for that reason. And in the third form of theory, a social contract argument is added to the second form: the basic respect for individual autonomy on which the social contract rests requires that the government protect exercises of that autonomy through speech. In each of these forms of autonomy theory, the liberal model of autonomy is at work. First, the self-realization form of autonomy theory focuses on the usefulness of speech, either in providing information that aids the process of

Philosophical Foundations: Cartesianism and Liberalism | 45

choice or in developing the personal capacities of critical reflection and rationality. Martin Redish, whose approach combines both of these aspects of self-realization, relies on a liberal model of autonomy in defining and defending his thesis that freedom of speech serves the value of self-realization. In his argument about the usefulness of speech in providing information, he posits the individual chooser as the moral agent. He assumes that this agent is interested in information primarily for the purpose of making choices, and he further assumes that this agent is capable of the relatively transparent self-knowledge and the relatively unified sense of self and values that would allow him to use that information to generate rational choices.46 In his description of the other aspect of self-realization—the way in which speech helps to develop human capacities—Redish is not very explicit about the particular capacities that are promoted by speech. Nonetheless, the structure of his argument suggests that he is focusing primarily on the liberal capacities of reasoned reflection. In particular, he argues that many of these capacities can be developed even in the areas of individual or “private” self-government, and not only in communal settings where deliberation is aimed at the common good. The “intellectual” capacities to which he refers here are plainly the ones associated with reason, as opposed to the “moral” capacities that some more republican theorists had in mind.47 Similarly, David Richards, another autonomy theorist who is a bit more specific on this subject, identifies rationality, self-determination, and integrity as among the capacities promoted by freedom of speech.48 These plainly grow out of the focus on reason, choice, and a unified self that ground the liberal model of autonomy. In the second form of autonomy theory, speech is seen as a direct exercise of personal autonomy and as valuable for that reason. The type of autonomy implicit in this argument is consistent with the liberal model outlined above. So, for example, David Richards describes the right to freedom of speech as “a further elaboration of the equal respect that grounds the inalienable right to conscience. For the same reason that the conscience of every person stands equal before the law, the communicative integrity of the exercise of each person’s rational and reasonable powers enjoys a comparable protection.”49 This conception of speech as an exercise of autonomy—in the sense of freedom of conscience—focuses on reason as a central aspect of autonomy. Finally, Richards’s approach to constitutional interpretation provides an example of the third form of autonomy theory. Richards is explicitly

46 | Philosophical Foundations: Cartesianism and Liberalism

working within the liberal, social contract tradition in arguing for a general freedom of conscience that includes free speech. He cites to John Rawls and describes autonomy in traditional liberal terms as “a capacity to formulate and act on higher-order plans of action, which take as their self-critical object one’s life and the way it is lived. . . .”50 His exposition of this concept focuses on the role of will and on the capacity for self-criticism and change within individuals.51 He argues that autonomy requires both the capacity for “rational self-direction” and freedom from external coercion.52 In addition, he argues that autonomy requires freedom from the kind of social conditioning that impairs one’s ability to critically assess social roles, values, and expectations.53 He is quite explicit about his debt to Kant, Locke, and Rawls and his desire to work out the implications of this liberal social contract tradition for constitutional interpretation.54 Perhaps the most influential and thoroughly developed version of autonomy theory in First Amendment scholarship is C. Edwin Baker’s liberty approach. Baker’s theory exhibits all three forms of autonomy arguments. He argues that speech is helpful in developing capacities for selfrealization and self-determination;55 that speech is itself an exercise of autonomy;56 and that the nature of political obligation requires that government respect the autonomy of its citizens, including protecting their freedom of expression.57 The model of autonomy at work in these arguments, while fundamentally liberal, also includes important modifications that anticipate some of the feminist critiques in the next chapter. First, the model of autonomy in Baker’s liberty theory includes the focus on choice that is central to the liberal model. Baker describes his liberty theory as “emphasiz[ing] the speech’s source in the self and mak[ing] the choice of the speech by the self the crucial factor in justifying protection.”58 One of the categories of unprotected speech is speech that is “unchosen”: that is, it is not the product of choice by an autonomous subject.59 This focus on choice is deepened by the principle used to limit the area of protected expression: it cannot be coercive.60 Coercion, of course, is the opposite of choice. Coercion is connected to overbearing the will of another.61 By defining coercion narrowly, Baker ensures that a broad area will be categorized as chosen. For example, he rejects “harms that result because the listener adopts certain perceptions or attitudes” as a basis for regulating speech and argues that “respecting the listener’s integrity as an individual normally requires treating the listener as responsible for her conduct unless she has been coerced or forced into the activity.”62 This

Philosophical Foundations: Cartesianism and Liberalism | 47

focus on responsibility arising from freedom, or autonomy, where freedom is contrasted with coercion, again relies upon the liberal model of autonomy. The emphasis on choice coexists, however, with a sophisticated and nuanced view of the role of social forces in constructing both reality and identity, a view that reduces the individualism of the model. Baker’s epistemology is explicitly social constructionist, in line with the feminist critique I will outline in the next chapter. He asserts that truth is not objective and our values are deeply implicated in the search for knowledge.63 Similarly, he acknowledges that “personhood and much of the specific content of a person’s identity is a social or collective achievement.”64 In his later work, this social constructionist view is even deeper and richer, helping to provide a powerful account of meaning as a social artifact and of the expressivist view of law.65 But, in the face of such social construction, Baker emphasizes “the normative claims of autonomy” and insists that only a collective order that “allows a member to affirm, to accept, and thus also to deny or reject this identity” is deserving of compliance.66 Thus, individual choice is assigned the central normative role despite the introduction of significant nonindividualist elements.67 Baker’s approach does, however, substantially modify the rationalism of the liberal model. Baker emphasizes that many of the ways in which speech serves the values of self-realization and self-determination are not matters of rational argument. He offers a detailed and illuminating list of some of the less commonly recognized functions of speech, and the result is to substantially reduce the rationalist focus.68 Thus, while Baker’s liberty theory is fairly characterized as liberal, it includes important steps toward recognizing and accommodating the feminist criticisms of liberalism that I will explore in the next chapter. The liberal model of autonomy is also implicit in the references to autonomy theory in Supreme Court opinions. Choice, reason, and individualism are all central to the justices’ conception of autonomy. For example, Justices White, Brennan, and Marshall, in rejecting a speech claim by a corporation, pointed out that such speech fails to fulfill the constitutional value of self-realization because it “do[es] not represent a manifestation of individual freedom or choice.” Similarly, the Court’s hostility to paternalism rests on an objection to government efforts to manipulate citizens’ choices by keeping them in ignorance of relevant information.69 This antipathy is so strong that the Court has even extended the protection against such regulation to the less protected category of commercial

48 | Philosophical Foundations: Cartesianism and Liberalism

speech.70 The government’s goal in such regulations, even if it is a sincere effort to serve the interests of consumers, is suspect because it fails sufficiently to respect the autonomy of the listeners. Indeed, one might even see the instances in which the Court has refused to adopt the autonomy theory as evidence that the model of autonomy on which it is relying is basically a liberal one. The two areas in which an autonomy theory would push towards greater protection than the Court has been willing to endorse are obscenity and symbolic speech. In both of these areas, many versions of the autonomy theory would see the speech at issue as implicating the autonomy of speaker or listener or both. But the Court has been consistently unwilling to expand protection in these areas in the ways that these autonomy theories suggest. One reason often given by the justices for this reluctance is that the speech at issue in these areas does not appeal to the rational faculties of the listeners, is “no essential part of any exposition of ideas”71—indeed, functions as no more than “an inarticulate grunt.”72 If autonomy is understood in terms of the liberal model, in which rational self-reflection is the central activity on which autonomy rests, then speech that does not engage the rational faculties of the listeners would make little contribution to their autonomy. Thus, the autonomy theories of free speech rest on the liberal model of autonomy. This model has, like the Cartesian model of truth, permeated both our general culture and our legal theory. It offers a vision of individuals who possess rational faculties allowing them to assess their choices. It posits that autonomy consists in the exercise of such capacities for reasoned reflection over choice. This model sees individuals as fundamentally unified selves, capable of reasonably transparent self-knowledge. It focuses on the interference with autonomy posed by external coercion, whether in the form of physical coercion or in the form of pervasive socialization. And the liberal model assumes that the process of autonomous choice can be conceptualized, understood, and carried out without reference to the social relations or cultural context of the agent. The autonomy theory of free speech is constructed on this liberal foundation.

part ii

Without Foundations

The intellectual edifice described in part 1 is built on the foundation of Cartesian and liberal thought. Together, these visions of truth and autonomy form the heart of a comprehensive view that shapes how we—the inheritors of the Western tradition—see ourselves and our world. It is a powerful vision and it is a ubiquitous one, but it has not gone unchallenged. Some of the challenges have been directed at truth, and some at autonomy; rarely have the critiques been brought together. Chapter 3 will summarize each of the two critiques separately and then offer some thoughts about the connections between them. These critiques, while posing a serious challenge to the Enlightenment vision, are problematic for feminism as well. In chapter 4, I will explore some of the difficulties raised by these critiques. If the critiques are understood to lead to the complete abandonment of truth and autonomy, then they pose severe challenges to some of the goals of feminism. The purpose of part ii, then, is to demonstrate why we must give up the foundation described in part 1, but why we must, nonetheless, not give up truth and autonomy.

49

3 The Feminist Critique

The recent challenges to the traditional conceptions of truth and autonomy have come from a number of different directions, including hermeneutics, pragmatism, postmodernism, communitarianism, and feminism. This section will focus on the challenges posed by feminist theorists. While all of these movements share many insights, the emphases and underlying concerns differ somewhat in each. It is my hope and belief that one could reach many of the same results I will suggest by traveling one of these other paths as well, but the path I will chart here begins from the concerns and methods of feminism. There are, of course, many forms of feminism and they differ in important ways. Despite their many differences, however, feminist theoretical frameworks across a wide range of positions share certain themes on which I will rely for this critique. One such theme is the belief that women are and have been systematically denied equal treatment and respect or oppressed by men and that this situation is wrong and should be corrected.1 A second theme, shared by many but not all feminist positions, is a recognition of social constructionism.2 Feminists also largely agree on the importance of listening to and taking seriously women’s own accounts of their experience.3 The criticisms of Enlightenment thought that I will describe build on these widely accepted themes.4

A. Truth In the context of the epistemological argument, the recognition of social constructionism takes the form of the social construction of knowledge. The central claim is that knowledge making is an activity that takes place only within, and is deeply shaped by, a cultural context.5 The feminist critique of Cartesianism links the social constructionist insight to the

51

52 | The Feminist Critique

commitment to end patriarchy by attempting to show how an epistemology that sees truth as objective, rationalistic, universal, and propositional/representational works systematically to support gender hierarchy. The social constructionist critique begins by undermining the claim that truth is objective, in the sense of reflecting a reality independent of human perception of it. Instead, the critique argues that truth is deeply, necessarily shaped by the social and personal context of the observer. There are at least three ways in which the process of acquiring knowledge is shaped by the knower’s context. First, the very facts that are taken by Cartesianism to be the materials out of which reason constructs knowledge are shaped by culture. “One thing is clear: making facts is a cultural enterprise.”6 Our experience does not come to us in prearranged bundles; rather, facts are made by a process of selection from experience. What we notice and the way we organize our experiences are both constrained by the conceptual categories that our culture makes available to us.7 Moreover, the orienting assumptions or cultural categories that we use to organize our experience are largely, if not totally, immune to challenge by the data or facts that result from our investigations under them. “[O]nce an investigator has adopted a given ontology, this system of orientation determines what is counted as an event; data cannot correct or falsify the ontology because all data collected within the perspective can be understood only in its terms.”8 To add to the difficulty, the process of acculturation that creates these categories and assumptions need not be identical for every member of a culture. A society may give different conceptual tools to different groups. For example, women may be taught to notice and identify subtle changes in emotional states while men generally are not. In addition, the more particular context surrounding an individual—family, neighborhood, religious association, ethnic group, and so on—may add to or alter the cultural impact of the larger society on that individual’s way of knowing. As a result, each person has a potentially unique collection of conceptual categories available for understanding his or her experience.9 This first sense in which knowledge is culturally constructed—the argument that there are no “brute facts” to which human beings have access independent of their culturally contingent conceptual categories— grows out of the denial of epistemological individualism and objectivism, but it also raises serious difficulties for the correspondence theory of truth. Even if some reality exists that is independent of human understandings of it, people do not have access to it. The kind of unfiltered, di-

The Feminist Critique | 53

rect knowledge promised by the Cartesian model is simply not available to human beings. The second sense in which knowledge is socially constructed is that in order to define data and to analyze them, interpreters must make value choices. Experience always underdetermines the data that it is used to construct and the data underdetermine the theories that they are used to construct. To choose between the interpretations that could be used to explain experience or data, one must rely—explicitly or implicitly—on a value judgment. In traditional science, some value judgments are explicitly acknowledged. For example, when choosing between two theories, both of which could explain the available data, scientists will prefer the theory that is simpler and more elegant. This is by no means an inevitable choice: one could choose instead (or in addition) to favor the theory that was most general (or most specific), most likely to produce human control over the phenomena at issue (or most likely to produce human respect for those phenomena). Some of the value choices at work in science are, however, less explicit and even more problematic. For instance, scientists long explained the “data” about differences between women’s and men’s analytical abilities by theorizing about biological differences between the sexes. Other kinds of explanations that are now preferred by many scientists (such as socialization) were ignored because the political and moral values of patriarchy were better served by biological explanations.10 Finally, value judgments affect not only the construction of facts and the interpretation of facts but also the choice of which issues or questions are worthy of investigation. There is no such thing as a problem in need of study without people who have the problem: “[A] problem is always a problem for someone.”11 Which problems are studied will depend on whose perspectives, concerns, and needs are considered most important by society. Perhaps the argument about the cultural construction of knowledge could be summarized by saying that interpretation is a necessary part of knowledge creation at every stage in the process and that interpretation is inescapably cultural and evaluative in nature. The value judgments inherent in these epistemological choices have important moral and political implications.12 Different value choices may have a significant impact on the social or political status of various groups of people, on the rights and responsibilities that society is understood to owe to its members, and on the social institutions and mechanisms that are seen as most appropriate or effective for promoting those rights and

54 | The Feminist Critique

responsibilities. Because of these moral and political implications, the social constructionist critique argues that these value choices are subject to moral and political criticism and justification. The problems one chooses to investigate, the way one describes the relevant data, and the interpretations one places on those data can all be criticized on the grounds that they are shaped by value choices that are morally or politically objectionable. Science, in other words, cannot be separated from morality and politics. It is, instead, always permeated by them. The refusal to recognize that connection does not eliminate politics and morals from science, but it may serve to disguise the particular value choices being made and thereby immunize them from criticism. One important consequence of recognizing the role of moral and political values in knowledge creation is that the knower becomes personally responsible for the knowledge that she creates. She is implicated— morally and politically—in those knowledge claims. She must accept responsibility for the values she used in generating and defending them. She may make demands on others to justify the values behind their own claims and she must be prepared to respond to such demands placed on her. She must recognize the process of making knowledge claims as itself a moral and political act for which she bears personal responsibility. The interrelation of science, on the one hand, and morality and politics, on the other, strikes a blow at yet another aspect of Cartesianism: its rationalist bias. Reason was distinguished precisely by its distance from emotions and values, and the objectivity and universality that such distance made possible.13 The social constructionist critique argues, however, that it is never possible to neutralize emotions and values since they are an integral part of interpretation and interpretation is an inescapable aspect of defining a problem, describing the relevant facts, and theorizing about those facts. Thus, reason, understood in some instrumental and nonsubstantive way, is simply insufficient (with or without the addition of sense perception) as a foundation for knowledge. Emotion and value— with all of their subjectivity and particularity and cultural contingency— are also necessary. The denial of individualism, objectivism, and rationalism leads directly to the denial of universalism as well. If our ability to theorize and even our ability to perceive are so thoroughly dependent on the emotions, values, and cultural conditioning that vary dramatically from one person to another, then it is foolish to think that we can find answers, or even define questions, that are universal or neutral across people. Such neutrality is

The Feminist Critique | 55

unattainable, certainly in every case, perhaps in any case. As a result, we may be left with not one but many equally valid interpretations of reality. Perspectives or points of view must be recognized as potentially valid alternatives, rather than as barriers or failures to be overcome.14 Knowledge may be personal or social; it is not necessarily universal. This critique also has important implications for the relationship between the knower and the known. No longer is the relationship a distant one, in which knower and known are separated from each other by a huge gulf of reason and objectivity, in which knowledge flows in only one direction, and in which the known is seen as an object to be manipulated or controlled. Instead, there is no rigid separation of the knower and the known. Obviously, such separation is not possible when the very “facts” about the known are themselves permeated by the identity and context of the knower. In addition, the known, whether it is another human being or some nonhuman aspect of nature, is seen as active and complex rather than as passive and needing control. In other words, the process of influence and even of knowledge runs in both directions: just as the known is defined by the characteristics of the knower, the knower’s very identity is altered by the process of coming to understand the known. “The reconstruction of knowledge is inseparable from the reconstruction of ourselves.”15 Once we abandon the dichotomies of Cartesianism and give up our claims of access to an objective reality, a broader range of approaches to truth opens up to us. In particular, truth need not take the form of propositions that make provable claims to represent reality. This impoverished vision of truth was a product of Cartesianism. When the reality that truth addresses is contextual and constructed, when the tools one can use to seek understanding of that reality are not limited to an instrumental form of reason, then the scope and form of truth expands. No area of human concern is necessarily excluded (e.g., aesthetics, morality). Nor can representations generally, and propositions in particular, claim any privilege as the ideal forms of knowledge. In sum, the social constructionist critique of Cartesianism begins by arguing that knowledge is a human, and therefore social, artifact, deeply shaped by the particular social context in which it is created. The argument then goes on to generate conclusions that contradict every premise of Cartesian epistemology. The vision of knowledge that emerges is one in which the knower and the known are intimately connected, indeed, mutually defining, and exist only within a particular cultural context.16

56 | The Feminist Critique

As I have just described it, the social constructionist critique is in no sense unique to feminism. Indeed, some critique along these lines is common to many different schools of thought, including some forms of Marxism, pragmatism, poststructuralism, and postmodernism. It is, nonetheless, appropriate to identify the critique as feminist (in addition to these other labels). The distinctively feminist version of the critique includes a focus on the relationship between the Cartesian premises, on the one hand, and the meaning of gender and the oppression of women, on the other hand. The feminist argument is that Cartesian epistemology has been used as a foundation for defining the difference between the genders and justifying the oppression of women. “Gender” is used here, as it generally is in the feminist literature, not to describe biological differences but to describe the socially constructed identities of male and female. This is not to deny that there are biological differences between the sexes,17 but it is to deny that those differences (in isolation from culture) determine in any significant way the highly detailed, complex, and deeply significant gender identities that exist in our society. Indeed, the attempt to separate biology and culture may be futile and misguided: they stand in a dialectical relationship in which each influences the other.18 The concept of gender is intended to continually remind us to ask about the cultural meanings and the social structures implicated in this system. Cartesianism is one of those structures. The difference between the genders is constructed out of the dichotomies that form the popular representation of Cartesianism. Who knows, how one knows, and what is known are all related to gender. The Cartesian knower is culturally male. The characteristics associated with the knower—objectivity, reason, intellect, universality—are traditionally associated with men. The thing known is defined as culturally female. The characteristics associated with the object of knowledge—particularity, emotion, physicality—are traditionally associated with women. Indeed, nature itself, as the quintessential object of knowledge, is commonly understood as feminine and women are understood as more closely connected to nature than men.19 Even the process of acquiring knowledge is culturally masculine. The relation between knower and known is one of separation, a relation of objectivity. Such separation is required for the autonomy of the knowing subject. And masculinity is, of course, defined importantly in terms of autonomy and separation. As the next part of this chapter will describe, the dominant model of autonomy

The Feminist Critique | 57

is deeply gendered, and its gender is male. Femininity has, conversely, been traditionally defined in terms of connection and dependence, characteristics that make the acquisition of knowledge by women perhaps impossible, certainly unfeminine. In other words, the gender dichotomy male/female is parallel to and defined in terms of the other dichotomies of Cartesianism. The Cartesian premises and dichotomies do not merely define men and women as different; they also form the foundation for a justification of gender hierarchy. The hierarchy is produced by the valuing of one side of these dichotomies and the devaluation of the other that results from associating only one side with the production of knowledge. The side associated with men is privileged, thus helping to justify the social privilege of the persons who are assumed to best exemplify those characteristics. Gender hierarchy also results from the connection between knowledge and power in the Cartesian view. Knowers exercise power or control over the known. Since knowers are, almost by definition, male, it is men who are thereby authorized to exercise this power. Since women are, almost by definition, part of the natural world to be known, they are among the fit objects of such control. There is extensive argument among feminist theorists over the question whether this gender hierarchy is a necessary aspect of a Cartesian epistemology or whether it is only an historical accident that the two ideologies have become joined in this way.20 For the purposes of my argument, it does not matter which position one adopts on this issue. Even if the connection is a matter of historical happenstance, it is too strong and deep to now be severed. Whether or not it was necessary that it be so, it is now the case that part of the meaning and function of Cartesian epistemology is to reinforce existing gender hierarchy in a deep and damaging way. It is particularly damaging because Cartesianism provides a foundation for so many diverse social practices, thus supporting the inequality and oppression of women in a wide range of circumstances. The links between knowledge and masculinity and power contribute to the unequal education of girls,21 to the discomfort with women holding important political and economic positions,22 even perhaps to the use and abuse of women in sexual relationships.23 The feminist version of the social constructionist critique, then, goes beyond the more general version by pointing out how a particular cultural category—gender—both supports and is supported by the particular epistemological assumptions of Cartesianism. The feminist critique

58 | The Feminist Critique

does not simply argue that value judgments and social goals are generally implicit in epistemological choices; it demonstrates how a particular set of values and goals—those of gender distinction and gender domination—are implicit in a particular epistemology. So, the social constructionist argument forms the foundation of a powerful feminist critique of the mainstream tradition in epistemology. This critique suggests that every aspect of the Cartesian view of truth is mistaken. Truth is contextual and relational rather than objective. It can and must be pursued through means that recognize emotion and value as well as reason and that generate responsibility on the part of the knower whose values are implicated. Truth may be plural or perspectival rather than universal. And truth may take many forms other than propositions purporting to represent an external reality. With the tools provided by social constructionism, feminists have also uncovered the deep connections between our culture’s understanding of the human relationship to reality, on the one hand, and our culture’s commitment to gender difference and gender oppression, on the other hand. In short, the Cartesian version of truth encodes and contributes to gender hierarchy. Cartesianism, then, is both epistemologically flawed and morally objectionable.

B. Autonomy The feminist critique of liberal autonomy arises from the same social constructionist vision as the critique of Cartesianism. In addition, it uses as its materials the lives and experiences of women in challenging the traditional model of autonomy. Every aspect of the liberal model—the rationalism, the individualism, the unified and transparent self, and the focus on choice—has been scrutinized and found wanting by feminist theorists. The conclusion of this critique is that the liberal conception of autonomy is inaccurate and incoherent as a model for actual human beings and that it grows out of an illusion of independence that is deeply gendered and implicated in the reproduction of gender hierarchy. The feminist critique has often been directed against the substantive image of the autonomous man so prevalent in both our popular culture and our philosophical tradition. Thus, Lorraine Code finds that this conception of autonomy posits a person who is “self-sufficient, independent, and self-reliant, a self-realizing individual who directs his efforts toward maximizing his personal gains. His independence is under constant threat

The Feminist Critique | 59

from other (equally self-serving) individuals: hence he devises rules to protect himself from intrusion.”24 Similarly, Seyla Benhabib sees in the state-of-nature metaphor “a vision of the autonomous self: this is a narcissist who sees the world in his own image, who has no awareness of the limits of his own desires and passions; and who cannot see himself through the eyes of another.”25 And Jennifer Nedelsky charges that, on the contemporary liberal conception of autonomy, “[t]he most perfectly autonomous man is thus the most perfectly isolated.”26 These theorists go on to demonstrate that this substantive image of autonomy is not only undesirable but also deeply gendered. As Benhabib puts it, “this is a strange world; it is one in which individuals are grown up before they have been born; in which boys are men before they have been children; a world where neither mother, nor sister, nor wife exist. . . . The point is that in this universe the experience of the early modern female has no place.”27 Carol Pateman has forcefully argued that this vision is itself the foundation for gender hierarchy. She contends that in our classic social contract stories, “[t]he new civil society created through the original contract is a patriarchal social order. . . . The original pact is a sexual as well as a social contract: it is sexual in the sense of patriarchal— that is, the contract establishes men’s political right over women—and also sexual in the sense of establishing orderly access by men to women’s bodies.”28 And the parties to these contracts are the isolated, autonomous individuals of liberal theory.29 These criticisms are a powerful indictment of both popular culture and some of the classic texts and ideas in the liberal philosophical tradition. But, as Marilyn Friedman has pointed out, the criticisms are less directly relevant to contemporary philosophical accounts of autonomy, which tend to adopt a procedural rather than a substantive approach.30 These philosophical accounts generally eschew any endorsement of isolation, or narcissism, or self-sufficiency (let alone sexual domination) as integral to autonomy. Instead, they focus on the process through which an individual comes to have the particular values he or she holds rather than on the substance of those values. That process is generally conceived as one in which the person comes to claim the values as her own, and in which she is not subject to certain kinds of illegitimate manipulation. The social constructionist vision underlying the feminist arguments against substantive autonomy does, however, have implications for the procedural model of autonomy as well. In particular, this vision begins by denying the continuing individualism of the procedural models and goes

60 | The Feminist Critique

on to challenge each of the other aspects described in chapter 2. I will outline each of these challenges. First, the feminist critique argues that autonomy is fundamentally relational. It is relational in several senses: causally, substantively, and conceptually. The liberal, procedural model of autonomy can accommodate the first of these senses, and it recognizes the second to some extent (albeit incompletely), but it fails adequately to include the third. Autonomy is causally relational in the sense that the capacities required for the exercise of autonomy are themselves the product of social relationships. That is, autonomy is created, as a matter of personal history, through the relationships of individuals. For example, some basic sense of self-esteem and self-trust appears to be necessary to the exercise of autonomy. Without self-esteem—understood as a belief in one’s own value—no one would attempt to project a life plan and to assess choices in light of it. Self-esteem is also a necessary condition for self-trust. Selftrust can be understood as a willingness to rely on oneself (even if it causes vulnerability).31 Defined in this way, self-trust requires not only that one value oneself but also that one believe in one’s own competence or efficacy.32 Without such a belief, a person might be unwilling or unable to undertake action to fulfill her life plan. But the development of self-esteem and self-trust in children is closely tied to certain relationships. Children learn that they are competent and worthwhile when their caretakers respond to their needs and signals. Indeed, it is only through a relationship to another person that the child could learn these lessons. It is precisely because the caretaker is not simply an extension of the infant, the infant needs something from the caretaker, and the caretaker responds to those needs that the infant experiences agency or autonomy. If the caretaker were not separate from the infant—if she were simply a part of the infant or completely controlled by the infant—then the experience of response by the caretaker would not represent an extension of the infant’s sense of agency beyond herself and into the world. In other words, the infant’s dependence is not an obstacle to the experience of autonomy but rather a precondition for it: it is the recognition by another person with whom you need to interact that provides the experience of autonomy.33 We might extrapolate from this argument to the supposition that it is the continuing recognition and responsiveness of other people that sustain these beliefs in self-worth and competence (and the capacities for

The Feminist Critique | 61

self-esteem and self-trust that are built on them) in adults as well. If we do not continue to elicit the responsiveness of others to our needs and signals, then we lose our sense of worth and of competence or efficacy in the world. The capacities required for autonomy are causally relational, then, in the sense that they require certain sorts of social relations for their creation and maintenance in individual persons. Many contemporary liberal theorists are comfortable acknowledging the causal sense in which autonomy is relational. In particular, they recognize that the capacities for rational reflection that are so central to many of their models are themselves the product of social relationships.34 The acknowledgment may be somewhat grudging, hedged with assertions or assumptions about how such capacities may exist to some extent even in the absence of socialization.35 And the mainstream theorists have, for the most part, notably failed to consider the rather obvious question of what social relationships and conditions might be necessary to promote and maintain autonomy, given this causal dependence.36 Nonetheless, the theories tend not to find the fact that autonomy is causally dependent on social relations threatening to the basic liberal model they are proposing. The second sense in which autonomy is relational is that the very substance out of which our autonomous choices and values are made is given to us in social relationships rather than invented by us. The values, ideals, categories, and identities among which and according to which we choose are social products and are passed on to us through our relationships with others in our culture. Thus, I choose a certain action because I am a “good mother” or a “free spirit” or a “Babbitt.” Even when we define ourselves in opposition to a given social category, we are defining ourselves in relation to it: “transcending or defying roles is a stance taken in a systematic and patterned relationship to those roles, and is to that extent governed by them.”37 The particular choices I make are “comprehensible only with reference to shared social norms, values, and concepts.”38 Again, many mainstream theorists have recognized this relational aspect of autonomy, although they have generally seen it as a challenge to their approach. This claim about the socially constructed nature of our values and identities is, in fact, the socialization problem recognized by many liberal theorists.39 They have offered a variety of responses to this claim, not denying its truth but seeking to cabin its consequences for their theories. For example, Joel Feinberg emphatically asserts,

62 | The Feminist Critique

No individual person selects his country, his language, his social community and traditions. No individual invents afresh his tools, his technology, his public institutions and procedures. And yet to be a human being is to be a part of a community, to speak a language, to take one’s place in an already functioning group way of life. We come into awareness of ourselves as part of ongoing social processes. Their fruits and instruments, precedents and records, wisdom and follies accumulate through the centuries and leave indelible marks on all the individuals who are a part of them. And all individuals are a part of these social histories.40

Nonetheless, Feinberg claims that the choices based on such socialized values and identities can be properly understood as one’s own, as autonomous, in normal cases. In the continuous development of the relative-adult out of the relativechild there is no point before which the child himself has no part in his own shaping, and no point after which he is the solely responsible maker of his own character and life plan. . . . The extent of the child’s role in his own shaping is, instead, a process of continuous growth already begun at birth. . . . These contributions are significant even though the child is in large part (especially in the earliest years) the product of external influences over which he has no control, and his original motivational structure is something he just finds himself with, not something he consciously creates.41

As Robert Young points out, this reliance on the (ever-increasing) role of the individual in shaping his own identity is not an entirely satisfactory answer to the problem of socialization. The fact that we “contribute to the making of ourselves . . . does not of itself undermine the claim that we desire and value what we are conditioned to desire and value.”42 Young suggests that the real issue is not how we come to have our particular values and desires but “whether or not on reflection one desires to have such desires.”43 Here too, however, the socialization problem reappears at the next level: perhaps our second-order desires—what we want to want— are also a product of socialization, along with the very capacity to form such second-order desires at all. Young’s response to this reiteration of the socialization problem is to point out that socialization is complex and not all of it qualifies as coercion.44 As with Feinberg’s approach, however, this response is not a sufficient answer. Even if socialization is not coercive, it

The Feminist Critique | 63

is not clear why the desires—first- or second-order—that result from it should be considered autonomous. As Young recognizes, coercion is not the only thing that undermines autonomy: internal restraints, such as neuroses and weakness of will, can also interfere with autonomy.45 If choices that result from these internal restraints are not autonomous, even though they are uncoerced, then why are choices that result from socialization autonomous? Thus, while the liberal autonomy theorists have struggled to recognize the substantively relational aspect of autonomy—the fact that the very materials out of which our choices are made are given to us through our social relationships—they have failed to fully integrate that fact into their theories. The liberal model, with its focus on the internal, individual act of choice or will, sits uncomfortably with the recognition of such socialization. The effort to integrate this recognition often results in simply pushing the moment of undetermined choice further back (to the choice to reflect on one’s desires) or further up (to a second-order level of desire) without an adequate explanation of why the problem of socialization does not simply reappear at that moment. Finally, and most fundamentally, the liberal models of autonomy are individualist in that they fail to recognize the extent to which the concept of autonomy is inherently relational. Each of the aspects of the self included in the liberal model suffers from this implicit individualism: the transparent self, the unified self, the rationalist self, and the focus on choice. In each of these aspects, the liberal models have failed to recognize the relational elements that, once included, transform the concept. The self is not as transparent as the liberal model suggests, and the process of self-knowledge is a social process, one that takes place within relationships and is dependent upon them. As Michael Sandel put it, “Where seeking my good is bound up with exploring my identity and interpreting my life history, the knowledge I seek is less transparent to me and less opaque to others.”46 Once we acknowledge fully the substantively relational aspect of our choices, it becomes possible and sensible that someone else could know me better than I know myself because the materials out of which that self is constructed are, at least in part, shared. In addition, once we acknowledge the causal sense in which our capacity for autonomy is relational, we can see that my own ability to know myself is dependent on there being others to reflect back to me the knowledge that I think I see. As we experience ourselves changing, we need those around us to confirm our developing self-knowledge and the new

64 | The Feminist Critique

identities that come with it. Susan Brison’s description of the need of trauma survivors to have someone listen to their stories is true in a less dramatic way for all of us: “not to be heard means that the self . . . does not exist. . . . [T]he . . . self needs to be known and acknowledged in order to exist.”47 The unity of the self in the liberal model also represents a failure to fully incorporate the relational aspects of autonomy. The traditional view of integrity as a kind of internal coherence of will is defined without reference to social relationships. Integrity is understood as logical coherence or consistency among the desires—first and perhaps second order—that provide the moving force behind our autonomy. But if one takes seriously the degree to which the fabric of our identities is woven out of our relationships with others, then the idea of integrity takes on a very different meaning. As Margaret Urban Walker has argued, integrity might then be seen as a kind of reliability, a form of taking responsibility on which both we ourselves and others can count. Integrity would be the willingness to account for our choices, the ability to stand by them even under pressure, and the acceptance of responsibility for our behavior even when things don’t go the way we hoped or expected and we can’t control the consequences.48 In this richer form, integrity is relational both in its nature and in its value. Integrity is about the accounts we give and the actions we take in relation to others: it is reliability to them that is part of what is at issue. In other words, integrity is practiced and demonstrated in and through relationships. Moreover, the value of integrity understood this way is also relational: my own integrity is valuable not just to me but to others who rely upon me as well. The rationalism of the liberal model also fails to adequately account for the relational aspects of autonomy. If reason is understood in a strictly instrumental way—as means/end rationality—then it is obviously insufficient to ground the act of choice on which the liberal model depends. How would one use such a limited type of reason to form a second-order desire about which of one’s first-order desires one should embrace? Indeed, it is precisely the emptiness of this conception of reason that leaves liberal autonomy theorists in the difficult position of having to accept that one’s desires (first or second order) might be given by one’s socialization (that is, the substantive sense of relationality discussed above). But reason can be understood in a richer and more substantive sense: the sense of giving good reasons for something. In that form, reason has

The Feminist Critique | 65

powerful relational elements. First, the content of such reasons is largely socially constructed: what counts as a good reason is culturally dependent and varies even within a single culture from one activity or social group to another. One consequence of this recognition is that there may be multiple, valid forms of reasoning. Indeed, feminist research in moral psychology suggests that at least some women use a form of reasoning systematically different from that described by psychologists who have primarily studied men.49 The understanding of the substantively relational aspect of reasoning helps to make sense of this finding and to support the claims of feminists that such alternative forms of reasoning are also entitled to respect. Second, not only are the contents of reasons socially constructed but also the process of giving reasons is a social practice, the rules and patterns of which are learned, practiced, and challenged within relationships. Whether or not one could learn to reason alone, one cannot go on reasoning alone indefinitely. In other words, people need certain kinds of social support in order to engage in the practice of reasoning. They need partners to accompany them, at least sometimes, and to offer them the social resources that allow them to engage in the practice, such as time, trust, and comprehension or understanding. Understood in this way, the process of rational reflection engages a much larger collection of capacities than it might have appeared in the liberal model, including many that would be conventionally understood as emotional rather than “rational” in the narrow sense. Finally, the focus on choice in the liberal model is a distortion that results from the individualism of this approach. The fixed attention to the interior act of choice or will makes sense only if we are not paying attention to the surrounding web of relationships in which the chooser is embedded. The point in choice is that it is internal: in some sense, it belongs to the agent rather than to anyone else. The point in choice is that it is not causally determined by something outside the agent. The point in choice is to ground autonomy on something that is insulated from such external forces. Once we include the relational aspects in our concept of autonomy, however, we must abandon such an illusion of insularity. It is not that we do not make choices—the experience of choice is still a real and significant aspect of our moral lives—but that choice, understood in this relational context, cannot be the foundation for autonomy in the sense envisioned by liberal theory. We might see the locus of autonomy, then,

66 | The Feminist Critique

not in choice but in a different place: the place where we construct the narratives of our lives out of which our choices and actions grow. Narratives tie us to relationships and communities; narratives are constructed always in dialogue with others, real or imaginary. As I will explore in chapter 6, it is such narratives that form the basis for our autonomy. So, the feminist critique demonstrates that the liberal model of autonomy is individualist in several senses. Although the contemporary liberal theorists acknowledge some of the ways in which autonomy is relational, they fail to adequately account for the implications of these facts in their conceptualization of autonomy. Every aspect of their model—from the rationalism to the unified and transparent self to the focus on choice—requires modification once the relational aspects of autonomy are taken seriously. The feminist argument is that the conceptual individualism of the liberal model makes it incapable of accounting for these relational aspects of autonomy. In addition to these conceptual problems with the liberal model of autonomy, feminists have also identified a moral and political problem: this model of autonomy encapsulates and perpetuates gender hierarchy. The traditional model of autonomy encapsulates gender hierarchy because it is based on an experience that is far more typical of men than of women in our society. As a result, it develops a model that purports to apply to all human beings but that largely ignores the experiences of many women. Moreover, this difference in experience is not merely happenstance; it is a systematic result of gender structures. Both conceptions of femininity and masculinity and the gendered life patterns that are enforced by social institutions lead to systematically different experiences for men and women—and the liberal model of autonomy is tied to masculinity and to male life patterns. As a result, women, predictably, often fail to meet the standards for autonomy, understood in this traditional way, and the standard, predictably, fails to recognize or value experiences of autonomy that are more typical of many women. The purportedly universal standard of autonomy perpetuates gender hierarchy because it is sometimes used to show that women are less autonomous than men, and therefore less deserving of the respect and freedom that is accorded to autonomous agents. I will briefly explore each of these ways in which the liberal model of autonomy is connected to gender hierarchy. First, the liberal model of autonomy is grounded in an experience of separation and fundamental independence that is deeply gendered. The individualism of the model, described and analyzed above, posits a self

The Feminist Critique | 67

that is essentially separate from its connections and relations to others. The characteristic emphasis on rights and justice in liberal political theory is a natural outgrowth of this vision of the self.50 There is persuasive evidence, however, that this experience of self is much more common for men than for women in our society. Carol Gilligan’s pathbreaking work on moral reasoning, and the many studies that have followed, indicate that men are more likely than women to experience their separateness as central to their moral identities and orientations. Women, on the other hand, are more likely to begin from an assumption of connection in which the “foreground” is occupied by relationships rather than by separate individuals.51 While there is controversy over this claim, there is also much evidence to support it as a descriptive generalization. What is less clear is the cause or source of such different senses of self. Some feminists suggest that the foundation is in gender role conditioning,52 others point to the development of gender identity in children in a society where their primary caretakers are almost always women,53 while still others suggest that the biological differences between women and men—including the female capacity to gestate, give birth to, and breastfeed an infant—contribute to the more connected sense of identity shared by many women.54 There is also the concomitant disagreement over whether these differences are necessary or contingent. If they are the product of either social conditioning or social practices like female parenting, then it would be possible to alter them by changing those conditions. If they are the result of biology, then it may be less likely that men would develop a more connected sense of identity, or women a more separate one, as long as women continue to give birth to children. For the purposes of my argument, however, it does not matter which source one ascribes to these differences nor whether they are seen as inevitable or not. As long as the generalizations about separation and connection are accurate descriptions of the experiences of many women and men in our present culture, the continued emphasis on liberal autonomy will serve to systematically misdescribe women. Autonomy—which was intended to be a concept applicable to all human beings—will actually be based on an experience typical of (some) men but not of most women. The problem with such a biased view of autonomy is not just that it fails as an accurate description for at least half of the people to whom it purportedly applies. The problem is that this failure is converted from a failure of the concept into a failure of the women themselves: because

68 | The Feminist Critique

they do not fit the concept, the women are seen as less autonomous.55 They fail to understand moral reasoning, or they suffer from educational and occupational deficits that reduce their ability to reason morally.56 And, most dangerous of all, once women are believed to be less autonomous, they may also be held to be less deserving of the respect due to autonomous persons. While such conclusions are rarely drawn by contemporary moral theorists, they are the norm rather than the exception in the literature that makes up the liberal canon in philosophy.57 And, as long as the model of autonomy that dominates that tradition continues to shape our conception of autonomy, that model will continue to generate the perception that women are less autonomous and the lingering doubts about their moral agency that such a perception entails. Thus, the liberal model of autonomy is both internally flawed, because of its excessive individualism, and morally suspect, because of its contribution to gender hierarchy.

C. Connections to Feminist Themes The feminist critiques of Cartesian truth and liberal autonomy are exactly parallel. They both draw on a social constructionist vision to undermine the central claims of each model, and then each points out the ways in which the traditional conceptions rely upon and reinforce gender roles tied to male dominance. In addition, the two critiques are closely related to several themes that run through much feminist theory in a variety of disciplines. These themes help to shape the reconstructions of truth and autonomy that I will explore in the next section of the book, so it is useful to see how they also undergird the critiques. One of the central themes of feminism is concern for those excluded. This concern has its foundation in the exclusion of women, but is not limited to that. The concern for those excluded is reflected in the critique of Cartesianism in the denial of a singular truth and a narrow conception of reason as the primary, if not exclusive, path to that truth. In particular, feminists have been at pains to demonstrate that the exclusion of women and other social groups is not simply a result of mistaken stereotypes or dislike on the part of individuals, but is the product of systems of exclusion that are deeply embedded in a wide range of social practices and institutions. The social constructionist analysis of Cartesianism is one example of such a demonstration.

The Feminist Critique | 69

Concern for the excluded is also at work in the critique of liberal autonomy. For example, the feminist critique points out that autonomy is learned through certain social relations and sustained by certain social conditions, such as the recognition and responsiveness of others to our needs. These conditions are not equally distributed in our society. It is clear that imbalances in recognition occur across lines of race, class, and disability as well as gender. The focus on the need for social relationships and conditions to support autonomy is, thus, tied to the concern for bringing in those who have been traditionally excluded from social power and recognition. Second, the relational nature of knowledge that is exposed by the argument from social construction is consistent with a theme common to much feminist theory. Contrary to the sharp division made by Cartesianism between the knower and the known, feminist theorists have insisted that knowledge is about the relationship between knowers and what they know.58 Such a relational view is consistent with trends in feminist moral and political theory as well as in feminist psychology.59 The critique of liberal autonomy shares this focus on relationship as the basic unit of analysis. In its rejection of the individualism of the liberal model, the feminist critique relies on feminist work in other fields— most notably psychology—that demonstrates the ways in which our identities and capacities are fundamentally relational. Much of this work focuses on the experience of women, but it suggests that a relational model is a more accurate and coherent one for all people. Finally, in both the epistemological and the moral context, the feminist critiques raise echoes of the other area of philosophy. In the critique of Cartesianism, for example, the feminist argument focuses on issues of agency and responsibility, traditionally conceived as moral issues. The breakdown of the Cartesian dream of objectivity gives rise to these concerns about the active role of the knower in shaping knowledge and the forms of responsibility implicit in that role. Similarly, in the critique of liberal autonomy, the feminist argument focuses our attention on the ways in which issues of knowledge are implicated in our moral categories. The feminist critique questions the liberal view of identity in which an agent is able to truly know herself, in a way that is in principle inaccessible to anyone else, and raises the possibility that autonomy might inhere in acts of cognition—such as telling one’s story—as well as in acts of volition. The feminist critique, in other words, highlights the areas of overlap between the moral and epistemological issues. This focus

70 | The Feminist Critique

on the ultimate connections between issues of knowledge and issues of moral personality is another broad theme in feminist theory. The traditional concepts of truth and autonomy were undoubtedly liberating in the social and historical context in which they arose, but they have become barriers to our ability to engage with the issues and needs of our own time. In particular, truth and autonomy have become rigid attempts to guarantee safety and certainty in a time that requires us to embrace greater fluidity and vulnerability. And truth and autonomy have become bastions of patriarchy in a time that is searching for a way to make real its growing commitment to gender equality. When we examine these foundations of our legal and philosophical culture, we find that they are too weak to deliver the security they promise and yet, at the same time, they are powerful engines for gender inequality. Having seen these concepts in this light, we may feel ourselves to be left without such foundations, with no choice but to abandon truth and autonomy altogether. The next chapter will explore the dilemma posed for feminism by such a move.

4 The Feminist Dilemma

In light of the need for a new direction, one might reasonably suggest that feminists should simply jettison the concepts of truth and autonomy and go on without them. For all of its value as a critique, however, the social constructionist position, when purged of truth and autonomy, poses some serious threats to feminist projects, both theoretical and political. The absence of truth threatens to leave feminists without a standard by which to justify their criticism of gender oppression. The absence of autonomy threatens to so “decenter” the subject or self that the concept “woman” will become useless as the basis either for theory or for political organization. And the critique of autonomy also may leave us with a self-image of victimization that is neither appealing to all the women who need the movement nor effective as a basis for political organizing. This chapter will explore these threats to feminism posed by undermining truth and autonomy, and review several efforts by feminist theorists to respond to them.

A. The Threats Posed by the Social Constructionist Critiques 1. Truth, Judgment, and Justification In order to see how social constructionism threatens to destroy standards of justification, it is useful to imagine the two extreme cases of cultural coherence. First, if the cultural context is conceived as monolithic and consistent (either generally or for any given individual), then it is difficult to see how an individual could achieve sufficient critical distance from her context to recognize elements of it that may be pernicious. Second, if the cultural context is conceived as multifaceted and potentially inconsistent,

71

72 | The Feminist Dilemma

then it is difficult to see what standard an individual could use to choose between conflicting values or interpretations.1 Assume for a moment that we have a society with a unified, consistent culture and that it is that culture that guides the process of knowing (of defining questions, of recognizing and interpreting data, etc). How would it be possible for us to escape from that culture sufficiently to criticize it, to recognize its destructive elements, and to see the possibility of an alternative toward which our society might move? If the very questions we ask and the data we perceive are shaped by those cultural assumptions, how can we ever detach from them sufficiently to compare them to the alternatives and find them wanting? This is the problem that Margaret Radin has called “bad coherence.”2 A social constructionist might respond that we never can escape totally, and we can never be certain that we have escaped. Even when we feel that we have broken free of our context, we should always be vigilant to search out the ways in which our cultural assumptions continue to shape our responses. Feminists, in particular, are quite adamant about the need for this type of scrutiny, of ourselves as well as others. The concepts in which feminist ideas must be expressed reproduce the very cultural categories that feminism means to challenge.3 As Audre Lord so eloquently put it, how can the master’s tools be used to dismantle the master’s house?4 But while total or confident escape may be impossible, feminists in such a consistent culture would have to believe in the possibility of some degree of extracontextual criticism. If we are truly and completely trapped within our cultural assumptions, then feminism would never be possible in a sexist society. But feminists want to assert both that feminism is possible and that our society is sexist. How can that be? How do we acquire the ability to see the world through a feminist lens if our culture consistently uses a sexist lens? “How is it possible to have an engaged truth that does not simply reiterate its determinants?”5 How can a critical distance be created?6 Various feminists have suggested several answers to that question. Perhaps the most common answer is “experience.” “[Feminism’s] project is to uncover and claim as valid the experience of women, the major content of which is the devaluation of women’s experience.”7 It is precisely the lack of fit between what we actually experience and what our culture tells us we should (or do) experience that creates the possibility of escape. Alison Jagger has described what she calls “outlaw emotions”: emotions

The Feminist Dilemma | 73

that are not socially sanctioned but that are experienced because one’s social situation makes the cost of the approved emotion too high.8 For example, there are times when it is simply too demeaning to feel flattered by sexual harassment, and then we feel the outlaw emotions of anger and fear. Outlaw emotions are not only politically subversive but also epistemologically subversive because they are incompatible with dominant perceptions and values.9 They create, in other words, a crack in the cultural edifice. It is true that these emotions and experiences must then be expressed in the conceptual categories provided by our culture and that process can cause them to be distorted. Some writers have suggested that this is why feminist ideas may best be expressed through forms—like poetry, metaphor, myth, and fantasy—that allow for more complex and creative meanings to be generated.10 Feminists must, in a sense, write a new language to name their experience, and through naming to make that experience real and visible for the first time.11 This process of creating new meaning through existing language is difficult enough, but it is not the major problem with this approach. The larger issue is how we could come to have the outlaw experiences and emotions in the first place in such a monolithic culture. According to the social constructionist theory, experience, including emotional experience, does not come in prepackaged bundles. We recognize only those experiences that our culture gives us the conceptual tools to recognize. How then can we recognize these experiences when our culture is telling us that they don’t exist? In other words, why are we not all victims of false consciousness, perhaps feeling anger (for example) but suppressing it and not recognizing it as such when our culture tells us anger is not an appropriate response?12 Alison Jagger acknowledges that these outlaw emotions stand in a dialectical relationship with critical theory; each gives rise to the other.13 That is, the critical theory allows us to recognize, name, and thereby make real our outlaw emotions and the outlaw emotions show us the gap between our experience and our cultural assumptions that gives rise to critical theory. The difficulty here is that outlaw emotions cannot be used to explain the way critical theory comes into existence if their own existence presupposes the availability of some degree of critical theory. In other words, these experiences and emotions cannot create feminism in a society that is coherently and thoroughly sexist because, unless some feminist conceptual categories are available

74 | The Feminist Dilemma

already, those experiences and emotions will not even be perceived to exist. To say that these experiences could be perceived in the absence of such concepts is to reject the basic premise of social constructionism itself: that knowledge and experience are always mediated by cultural conceptual categories. If women can recognize sexual harassment as harassment in a culture that sees sexual advances as flattering rather than demeaning or threatening, then why couldn’t scientists recognize when they are more directly experiencing an objective reality, rather than one distorted by cultural assumptions? A social constructionist cannot escape this difficulty by positing a basis for experience free from cultural categories without risking the whole critique of Cartesianism. This difficulty with the critique is worth exploring because social constructionists, including feminists, sometimes describe the existing culture as though it were a consistent monolith, at least on certain issues such as sexism and patriarchy.14 In fact, however, contemporary American culture is anything but monolithic. We are presented with a complex and confusing collection of cultural categories, including the influences of many other cultures and many previous eras, some of which are in tension with or direct contradiction to each other. And this complexity and contradiction exists in gender issues as well. In light of this situation, it is not at all implausible that the seeds of feminism, along with other dissenting positions, can be found in the culture itself, alongside the sexism it opposes. This cultural complexity solves the problem of how we can recognize experience that contradicts some of our cultural categories: we recognize it because it is consistent with, or even required by, other conceptual categories that are also part of our inconsistent culture. Such an inconsistent culture raises, however, another problem. If the culture allows for different, even contradictory, answers on questions as fundamental as the nature of gender identities and relations, how can we settle disputes between those who adopt these differing views? Both in relations between private individuals (as in a dispute between a husband and wife over how much housework and child care the husband ought to do) and in the design of public policies and programs (as in the decision whether to punish sexual harassment as gender discrimination), competing visions would lead to very different results. Where can we find a standard with which to decide between them if our own culture, the only standard available to us under social constructionism, includes both?

The Feminist Dilemma | 75

The response of some feminists seems to be that such value conflicts are indeed unresolvable and will (and should) remain so. One branch of feminist theory seems, in fact, to have adopted the irreducibility of multiple viewpoints as a value in itself.15 It is, however, one thing to accept, even celebrate, descriptive relativism—the claim that we all see the world in different ways—and quite another to accept moral or decisional relativism—the claim that there is no standard to judge between our various visions on moral or other grounds. Feminism, because it must at a minimum reject certain types of oppression of women, cannot accept moral relativism of this kind. It cannot be neutral as between different value systems, some of which oppress women and others of which do not.16 Seyla Benhabib put the matter this way: we cannot ignore the fact that such conflicts pose “moral as well as cognitive problems, or that the question of validity inevitably confronts us, and that we cannot extricate ourselves from an answer by gazing in wonderment at the plurality of language games and life-forms.”17 Thus, social constructionism seems to degenerate either into simple conventionalism or into a kind of cultural relativism.18 Conventionalism is, in a sexist society, obviously inadequate for feminism. Relativism, whatever one may think of it in general, is also insufficient to meet the purposes of feminism. To be an acceptable basis for a feminist epistemology, social constructionism must be modified or supplemented to provide some standard for criticism, of both our culture generally and the viewpoints of particular persons or groups within it.19 Social constructivism falls into this dilemma in part because it ultimately fails to escape the Cartesian framework it seeks to challenge. The Cartesian knower has been described as trying to perform the “God trick” by achieving a “view from nowhere,”20 because he seeks to distance himself from any culturally contingent location. In attempting to refute the possibility of such a transcendent position, the social constructionist has described and celebrated a multiplicity of valid (though partial) viewpoints. As Susan Bordo has pointed out, the question remains, however, how the human knower is to negotiate this infinitely perspectival, destabilized world. Deconstructionism answers with constant vigilant suspicion of all determinate readings of culture and a partner aesthetic of ceaseless textual play as an alternative ideal. Here is where deconstruction may slip into its own fantasy of escape from human locatedness—by supposing that the critic can become

76 | The Feminist Dilemma

wholly protean by adopting endlessly shifting, seemingly inexhaustible vantage points, none of which are “owned” by either the critic or the author of a text under examination.21

The injunction to celebrate and participate in multiple perspectives, with no prescription for choice among them, denies the “somewhere” that we all do and must inhabit. “[T]he philosopher’s fantasy of transcendence has not yet been abandoned. . . . [I]nstead, it has been replaced by a new postmodern configuration of detachment, a new imagination of disembodiment: a dream of being everywhere.”22 Rather than the view from nowhere, we have the view from everywhere. This lapse into transcendence is caused by the failure of social constructionism to escape from the underlying Cartesian view of objectivity. Having accepted the skeptical dilemma around which mainstream epistemology is constructed—that there must be only one truth or no truth at all—and having rejected the Cartesian attempt to build a foundation for a single truth, social constructionists find themselves with no way to distinguish between truth and falsity in the multiplicity of viewpoints.23 They have challenged the Cartesian notion that truth must be transcendent and unitary, but only in order to say that such a truth is impossible, not to provide an alternative account of a situated truth that might justify a choice between viewpoints. As a consequence, they are left with no truth at all and so are insufficient to the feminist task of identifying and resisting the oppression of women.

2. Autonomy and the Nature of the Subject A second danger to feminism posed by social constructionism concerns the challenge it represents to the Enlightenment concept of the subject or self. Social constructionism challenges the Cartesian and liberal conceptions of the self by showing that subjects are deeply dependent on the social relations in which they exist, and often deeply fragmented by cultural tensions and contradictions.24 The self, on this social constructionist view, is neither transparent, nor unified, nor clearly bounded from the others who form the relations in which it exists. Social constructionism’s success in blurring the boundaries between inside and outside may, however, call into question the coherence of “the subject” as a concept at all.25 The “[p]ostmodernists ask us to cease thinking of ourselves as having

The Feminist Dilemma | 77

identities and to begin understanding ourselves as sites for competing cultural interpretations.”26 Such disintegration of the self creates certain difficulties for feminism.27 From a theoretical point of view, “[i]f there is no subject, who is left to emancipate?”28 The very idea of gender oppression requires that we be able to identify a subject—“women”—that has both epistemological and moral significance.29 But if the permeability of identity boundaries leads to their complete collapse, then the subject can no longer function as a meaningful unit for the purposes of either knowledge claims or moral claims.30 Moreover, the concept of gender oppression requires that we be able to identify power arrangements that qualify as “oppression.” The disintegration of the self, however, leads social constructionism to blur the lines of power, seeing power as everywhere, as a discursive background, rather than as a force used by some particular people against others.31 When there are no meaningful subjects left, power becomes an attribute of social institutions rather than of persons. Given the continuing pattern of patriarchal violence against women, however, feminists cannot afford to dispense with the notion of power as violence against a subject by other subjects.32 The directionality of power—its use by some particular people against others—is as significant for feminist purposes as its background pervasiveness. Nor can feminists manage without the subject on a more practical level. Both the small-scale consciousness raising and the large-scale political organizing that form the framework of a feminist movement depend on a notion of the subject. One of the most pernicious and damaging aspects of gender oppression—and one particularly apparent from the perspective of legal theory—has been the denial to women of the status of subjects or persons.33 The reclamation of a self within, a self denied by patriarchy, is one of the central achievements of the consciousness-raising method.34 Indeed, the rise of social constructionism and postmodernism at precisely this moment in history may appear highly suspicious: just when previously silenced persons have begun to speak for themselves, the concept of the subject and the possibility of truth come under fire.35 Moreover, political progress seems to depend on a meaningful subject. “To the extent that feminist politics is bound up with a specific constituency or subject, namely women, the postmodern prohibition against subject-centered inquiry and theory undermines the legitimacy of a

78 | The Feminist Dilemma

broad-based organized movement dedicated to articulating and implementing the goals of such a constituency.”36 A political movement must be organized around something. Even if it avoids the potential essentialism of an identity politics, it requires the sorts of subjects that can have interests, or rights, or values that are both identifiable and morally significant. Participants in the movement must see themselves as such subjects in order to motivate action, and other members of the polity must also see them that way if they are to have any impact. The successful deconstruction of subjectivity would undermine such political action. Finally, in order for that political organizing to be possible, people must be more than mere reflections of their social conditioning. Without such a potential, feminism would be impossible in a sexist society. If people are constituted by their social conditioning, so that even their fundamental sense of self is determined by their culture, and if sexism is a deep, even foundational, aspect of that culture, how can women ever hope to change their condition? If the critique of autonomy goes so far as to make any form of agency or resistance impossible, then feminism itself would become incoherent, futile, and useless.

3. Autonomy and Victimhood The collapse of the traditional model of autonomy under the pressure of the social constructionist critique can lead to another problem for feminism: the dominance of a victim perspective. When all sense of agency is lost, and people come to see themselves as the passive locations in which social forces collide, they may adopt a view of their oppression as simply a matter of victimhood. But this view is a problem for feminist political movement because it is unacceptable to many women and may be, in itself, a barrier to certain forms of action and understanding. First, the victim perspective is implicated in the continuing exclusion of women of color from feminist movement. Feminists of color have been pointing out with increasing frequency that the view of women as simply the victims of society, shaped rather than shaping, has the effect of systematically excluding them. As bell hooks has written, [W]omen’s liberationists [made] shared victimization the basis for woman bonding. This meant that women had to conceive of themselves as “victims” in order to feel that feminist movement was relevant to their lives. Bonding as victims created a situation in which assertive, self-

The Feminist Dilemma | 79

affirming women were often seen as having no place in feminist movement. It was this logic that led white women activists (along with black men) to suggest that black women were so “strong” they did not need to be active in feminist movement.37

There are very good reasons why this image of passive victimization is particularly unacceptable to certain groups of women. Women from systematically oppressed groups must turn to a more active process of selfdefinition in order to achieve a meaningful sense of self-worth.38 Women who are subject to the most severe practical oppression may actually be less likely to adopt a victim stance than those who experience less onerous circumstances. “Women who are exploited and oppressed daily cannot afford to relinquish the belief that they exercise some measure of control, however relative, over their lives. They cannot afford to see themselves solely as ‘victims’ because their survival depends on continued exercise of whatever personal powers they possess.”39 If feminism as a political movement is to include these women as well, then it must develop a model of personhood that includes some capacity to redefine ourselves and make self-conscious change in the oppressive conditions that have shaped us. The language of victimization is, not surprisingly, problematic even for many more privileged women as well. For these women, the self-image of the victim is unacceptable as a personal matter. We do not recognize ourselves in these images and we find that this language does not help us to address the problems of sexism and gender identity with which we are struggling in our own lives. As Martha Minow has pointed out, one of the central difficulties with “victim talk” is that it reduces our attention to the particular, complex person by focusing us on a single slice of her identity and experience. “Any richer sense of the person undermines the claim of victimhood because victimhood depends on a reductive view of identity.”40 Moreover, the language and imagery of victimization runs the serious risk of being misappropriated and used against women. The passivity and helplessness of the victim identity is dangerously close to the traditional stereotypes of femininity. The use of victim language by women is sometimes used to revive such stereotypes to the detriment of those claiming victim status. For example, when feminist lawyers began offering evidence of “battered woman syndrome” in self-defense cases for women accused of attacking their batterers, they discovered that women who are

80 | The Feminist Dilemma

seen as too passive, helpless, and victimized are not likely to be given custody of their children. Victim language can be used as an excuse to rely upon traditional stereotypes of women.41 Although these sorts of complaints against the socially constructed identity of victimhood can take the form of the antifeminist diatribes of the popular press, they also come from women deeply committed to feminism as both a theory and a movement.42 As Adrienne Rich explains, “Pride is often born in the place where we refuse to be victims, where we experience our own humanity under pressure, where we understand that we are not the hateful projections of others but intrinsically ourselves.”43 Finally, the victim perspective is crippling to feminism because it limits what can be seen and what can be said in ways that are ultimately harmful to feminist goals. Susan Wendell offers a fascinating description of the “perspective of the victim,” in which “the victim recognizes the oppressor’s responsibility and assigns blame to the oppressor” rather than to herself.44 This perspective has distinct advantages over the perspective of the oppressor, which often precedes it, in which the victim is blamed (and often blames herself) for her own suffering. But it also has a major disadvantage: “it can easily obscure the actual or potential power and choices of victims.”45 We need to be able to see the many ways in which people in oppressive situations continue to exercise choice and responsibility. We need to see their heroism, their acts of resistance large and small, as well as their responsibility for the harm they may do even while they are oppressed. We need to see these things because if we don’t see them we risk falling into the despair of powerlessness. We become so mesmerized by our present and past victimization that we are incapable of imagining and taking action to create a different future. In addition, the perspective of the victim excuses us from scrutinizing our own participation in systems of oppression and thereby perpetuates them in much the same way that the perspective of the oppressor excuses by blaming the victim. The goals of feminism cannot be promoted without a recognition of the reality of victimization and the responsibility of oppressors, but neither can those goals be achieved without going beyond the perspective of the victim to the perspective of the “responsible actor” who is capable of some autonomous action.46

The Feminist Dilemma | 81

4. Playfulness and Responsibility Finally, social constructionism, at least in its poststructuralist and postmodernist forms, poses a more diffuse, atmospheric problem for feminism. These intellectual movements are sometimes characterized by a desire to “hav[e] it any way they want. They refuse to assume a shape for which they must take responsibility.”47 They “replace [Cartesian] metaphors of spectatorship with metaphors of dance; . . . relinquish . . . fantasies . . . soberly fixed on necessity and unity [in favor of] those that are intoxicated with possibility and plurality.”48 Such a constantly shifting, playful attitude can be exhilarating and liberating. It frees us to reimagine reality in endlessly new and creative ways. It challenges us to attempt to see the world from someone else’s perspective. It is undeniably an important part of any project that seeks, as feminism does, to be inclusionary and liberatory. But this attitude can also be a barrier, a barrier both to an adequate appreciation of difference and to a responsible stance toward the actual suffering of human beings. If human beings had a truly infinite capacity to assume the perspective of another, real difference would be impossible; we would be transparent to one another. It is our particularity, and the boundedness that it creates, that makes otherness possible. To refuse to recognize (indeed, even to celebrate) that boundedness is to refuse to recognize or celebrate difference. We must take seriously the ineradicability of otherness.49 Playfulness also fails to take seriously the real pain caused by certain perspectives and cultural contexts. We cannot simply reserve judgment, critiquing and deconstructing existing concepts and contexts but refusing to commit ourselves to some and not others. The attitude of playfulness is morally inappropriate, indeed irresponsible, when the cost of certain kinds of diversity is paid in suffering.50 If feminism is committed to taking women’s experiences seriously, and if women have experienced much of that suffering, feminism simply cannot adopt this attitude.51 Although I have described the difficulties with social constructionism separately, each is of course implicated in the others. Together they form a whole that threatens to rob feminism of its critical bite and its moral and political force. These are not superficial difficulties; they arise from the same basic components of social constructionism that account for its usefulness to feminism in critiquing Cartesianism and liberal autonomy.

82 | The Feminist Dilemma

The challenge, then, is to find a way to retain the useful aspects while reducing the impact of their dangerous implications.

B. Feminist Strategies for Reconceiving Social Constructionism Some feminist theorists, both in and outside of law, have accepted this challenge. Three of the most common strategies for reconceiving social constructionism are, first, to adopt a standpoint epistemology; second, to incorporate pragmatist methods and insights; and third, to attempt to define and justify a standard for judgment that is in some sense cross-cultural but still not transcendent. In my view, it has been difficult to pursue these strategies successfully because we lack a clear idea of how a solution could be both consistent with the basic premises of social constructionism and sufficient as a guide for resolving conflicting perspectives. While such a guide may never generate certainty, a sufficient standard must provide us with mechanisms for narrowing the legitimate alternatives to a manageable number and with a stock of arguments or concerns through which to evaluate those alternatives. I will suggest such standards and concerns in later chapters; this section will explain why I believe that the most common approaches taken by feminist theorists to this set of difficulties—while useful and important—require further elaboration in order to be sufficient.

1. Standpoint Epistemologies In order to deal with a heterogeneous culture, several writers have suggested a sort of procedural basis for deciding between competing viewpoints and values. They claim that women, along with other oppressed groups, have an “epistemic advantage” that gives them a better claim to see clearly than those who are higher up in the social hierarchy.52 The reason for this advantage is that oppressed groups must understand, and to some extent internalize, the dominant culture in order to survive, but they live simultaneously in a different subculture. In other words, members of oppressed groups find themselves in two different, and often contradictory, contexts. This disjunction gives them the distance from each context that allows them to adopt a critical perspective toward it. Such distance is either impossible or much more difficult for members of the dominant group because they live wholly within a majority culture that reflects and

The Feminist Dilemma | 83

legitimates their experience. It is this critical distance that constitutes the epistemic advantage of the oppressed.53 And it is this advantage that justifies preferring the viewpoint of the oppressed to the viewpoint of the dominant group. Such a means of deciding between competing claims of value or vision without reference to their particular content would, of course, be very useful. Moreover, if the epistemic advantage is simply a prima facie claim that brings to the center of our attention those perspectives that have been most ignored in a society, this approach has much to recommend it.54 Nonetheless, I think there are serious difficulties with a standpoint epistemology if it is actually used to resolve disputes rather than merely to frame them. The first set of difficulties consists of problems of application. First, it is not always easy to tell which group is oppressed and in relation to whom. If two oppressed groups—say, women generally on the one hand and working-class or poor men on the other hand—have competing values, how do we decide this issue? Do we ask who is more oppressed in general? Who is more oppressed on the particular issue in question? Do we look for some group that is at the very bottom of the hierarchy and have its views guide us about everything? Second, it is simplistic, and contradicted by their own accounts, to assert that the members of an oppressed group will all share the same vision or values on any given issue. Feminism has, indeed, been struggling for some time with the reality of diversity among women. When there is disagreement, which oppressed voice should guide us—the majority of the oppressed (because they better represent the whole group) or the minority (because they are even further oppressed and distanced from their context by being dissenters within their own group)? These practical difficulties are, moreover, compounded by conceptual ones. Being between two contexts or cultures is probably a characteristic of a majority of people in twenty-first-century America rather than a unique experience of some small, determinate group that could be called oppressed. In our extremely diverse society, almost everyone will be part of a subculture of one type or another. Not only gender and race but also religion, ethnicity, sexual orientation, economic class, physical challenges, occupation, and many less obvious characteristics can bring a person into contact with subcultures that differ, to varying degrees, from the dominant culture. It is true, of course, that members of groups with certain characteristics—a culture that covers many aspects of life rather than just

84 | The Feminist Dilemma

a few, and a culture that is despised by the majority—may have a more intense experience of conflict between the two contexts, which may generate more of a critical distance. But that issue is one of degree and will vary with the particular circumstances of an individual; no group can claim a special privilege that is in principle unavailable to many other people. Perhaps even more importantly, people who experience this conflict of contexts do not necessarily respond by creating a critical distance between themselves and each context. The conflict can lead instead to a rejection of one context and the wholesale adoption of the other, or to dichotomizing one’s life so that one culture applies without conflict to certain aspects (e.g., domestic life guided by the values of the “feminine” subculture) while the other culture applies to other aspects (e.g., work life guided by the values of the dominant masculinist culture).55 Neither of these strategies would necessarily contribute to the achievement of critical distance. Moreover, the critical distance that can be achieved if one remains simultaneously in contradictory cultures is purchased at a cost. Those who feel themselves to be shaped by a subculture that is despised by the majority often suffer very real scars from that experience. The “oppressed [may be] in fundamental ways damaged by their social experience.”56 Standpoint epistemology may perpetuate a fantasy of innocence, when in fact the oppressed often are not epistemologically or morally “innocent”: their own internal culture is implicated in and shaped by the systems of oppression.57 Another price commonly paid by those who straddle two cultures is a “sense of totally lacking roots or any space where one is at home in a relaxed manner.”58 This sense, like the damaging scars inflicted by majority disdain, might itself have certain systematic implications for one’s value choices. Those substantive implications may be good or bad, but they would have to be assessed independently in order to determine how they affect the usefulness of this approach. That assessment will require some moral standard that addresses the substantive question of how to choose between different values and visions. Thus, a standpoint epistemology faces both practical and conceptual difficulties and it seems incomplete without the addition of a substantive analysis. Sandra Harding, one of the most thoughtful proponents of a standpoint epistemology, recognizes these flaws and is attempting to construct a standpoint theory with a standard for a new, “strong” objectivity built into it.

The Feminist Dilemma | 85

A feminist standpoint epistemology requires strengthened standards of objectivity. The standpoint epistemologies call for recognition of a historical or sociological or cultural relativism—but not for a judgmental or epistemological relativism. They call for the acknowledgment that all human beliefs . . . are socially situated but they also require a critical evaluation to determine which social situations tend to generate the most objective knowledge claims.59

In light of this demand for objectivity, Harding distinguishes between a perspective—“which anyone can have simply by ‘opening one’s eyes’”— and a standpoint—which is “an achievement,”60 an “objective perspective from women’s lives that gives legitimacy to feminist knowledge.”61 And she endorses a distinction between “‘good’ and ‘bad’ belief formation and legitimation,” both of which are socially caused but only one of which leads to true or better knowledge.62 The inclusion of a sufficiently well-defined theory of objectivity would solve some of the problems noted above. It would specify which outlooks were entitled to this privilege—only standpoints, not mere perspectives— and it would explain why they were privileged: because they are more objective. Harding’s definition of objectivity, which centers on valuing and being able to assume (temporarily) the perspective of the Other,63 even helps to explain why the oppressed position of certain groups gives their members a greater claim to objectivity: members of oppressed groups learn, as a matter of survival, how to adopt the perspectives of others, so they are more likely to have an objective view. But Harding has not spelled out the nature of this situated objectivity sufficiently to resolve the difficulties posed by social constructionism. We need to know more about what “good” beliefs are or we will have no way to evaluate and choose between the various perspectives of others once we have tried them on. Harding says that good reasons “do not refer to transcendental, certain grounds for belief of the sort claimed by conventional epistemologies [nor do they] privilege what any group of actual, historical humans say about how they see the world. . . .”64 She suggests that they rest instead on some normative values.65 But the relation of such values to standpoint epistemology is somewhat problematic. If the values are determinate enough, then it is no longer clear what work “standpoints” do. If the goal is to identify beliefs that are justified by good reasons, rather than by social causes alone, then why shouldn’t we address that issue directly? Why do we need to privilege any

86 | The Feminist Dilemma

standpoint at the beginning? For example, if, as Harding suggests, one of these values is a respect for Otherness that attempts neither to deny the existence of the Other nor to exert unilateral control over the Other,66 then we can use that standard to measure beliefs directly. We do not need to look for groups whose experience is most likely to lead them to this value and then privilege their perspectives. A standpoint epistemology is a clumsy and less reliable path to that goal. This problem arises because it is simply not clear in what sense this value or standard (i.e., this objectivity) is situated. It is situatedness that the “standpoint” captures, and it is situatedness that a social constructionist view requires. The normative function of objectivity in Harding’s theory, on the other hand, requires that the standard be meaningfully independent of particular historical contexts. Until we have an adequate description of what it means for a standard to be both independent and situated, we will continue to oscillate between the Cartesian poles of universalism and normative relativism. In describing an alternative model of truth, in chapter 5, I hope to offer such a view.

2. Pragmatism Another strategy some feminists have adopted to deal with the difficulties of social constructionism is to embrace pragmatism. Although pragmatism is no more easily susceptible of definition than feminism, I believe it is fair to characterize it as an approach or practice that eschews abstract, general, and transcendent systems of thought in favor of thinking that is contextual and situated (always embodied in particular practices) and instrumental (meant to solve particular problems).67 While the emphasis on local specificity allies it with social constructionism, the instrumentalism of pragmatist thought keeps it firmly normative rather than skeptical.68 Our moral positions therefore “reflect not objective truth, but the grammar of what it means to be us,”69 where the boundaries of “us” is itself a normatively contested concept and our truths are always provisional.70 Some pragmatist feminists adopt this methodological formulation of pragmatism. Joan Williams, for example, suggests the types of questions we must ask ourselves in order pragmatically to assess systems of thought and the social practices in which they are embodied. The issues, she contends, are “what would life have to look like to make this a justifiable choice? . . . [and] do I want to change in the ways I would have to in order

The Feminist Dilemma | 87

to adopt this novel point of view?”71 This formulation of the issues addresses some of the difficulties with social constructionism. It recenters a concept of identity, while continuing to insist on a high degree of social construction of self. It does not require coherence in that identity, but it does require that we not be so totally fragmented and boundary-less that we are unable to talk about “who we are.” Williams is also very concerned to remedy the apparent lack of seriousness of some postmodern versions of social constructionism. She turns from metaphors of play to the metaphors of pragmatism precisely in order to rid her position of the amoral aestheticism that she recognizes as distasteful to many people.72 Williams’s formulation does not, however, address the first difficulty with social constructionism: the lapse into either relativism or conventionalism.73 She has moved the inquiry to a new level—the level of who we are and want to be rather than what we should do in a specific instance—but she has not specified the kinds of considerations that should be relevant or determinative when we attempt to answer those questions. Again, we are left with only our culture to rely upon for these images of identity; if that culture is consistent, then there is no room for a critical stance, and if it is inconsistent, there is nothing beyond it to which we can appeal in order to choose. In other words, the pragmatic standard of functionality—what “works”—must itself be filled in by culturally specific concepts. To a Christian Scientist, prayer works as healing and medicine does not because the goal is defined in spiritual as well as physical terms. Instrumentalism implies an end; the means are assessed by how well they serve that end. But pragmatism, defined in this methodological way, does not itself specify the end. It, therefore, encourages the right kind of moral responsibility but does not provide the epistemological tools with which to fulfill it. This type of methodological pragmatism restates, but does not resolve, the dilemma of social constructionism. Some pragmatist feminists have sought to fill this gap by specifying a normative standard by which to judge various social practices or identities. By far the most common contender for this position is some version of an “antisubordination” principle: a principle requiring us to recognize and redress imbalances of power and respect, particularly when they fall along the traditional lines of race, gender, and class.74 This approach is the substantive analogue of the more “procedural” attempt to find an epistemological privilege for the viewpoint of those who are oppressed, and it often accompanies a standpoint epistemology. The attempt explicitly

88 | The Feminist Dilemma

to provide such a standard represents the third strategy for dealing with the problems of social constructionism.

3. An Independent Moral Standard Pragmatists who use a principle like antisubordination as part of their approach are relying on a strategy of defining an independent moral principle with which to choose between norms and perspectives. The difficulty with such a principle is essentially a problem of description: what sort of thing is it? If the commitment to antisubordination is a transcendent principle, one that applies across cultures and to all contexts within cultures, the preeminent moral standard against which practices and other norms must be judged, then it seems to violate the very heart and soul of social constructionism. It is not a new epistemology, but simply a different contender within the old epistemological battles. And it is fundamentally unconnected to the methodology of pragmatism. If, on the other hand, antisubordination is a contextually derived and contextually bounded norm, then there is no reason to think that it should apply to every practice we wish to assess, let alone that it would be the most important norm for every practice. Antisubordination would, instead, be one consideration among many, the relevance and importance of which would have to be judged in particular instances, and how that judgment would be made is itself extremely unclear. In other words, in order for such a principle to do the work assigned it, it must neither be totally transcendent nor simply one context-specific norm among many. It must, that is, escape the Cartesian echo of one truth or none. The question is, of course, what is left? What else could it be? Katherine Bartlett has explored the nature of such an intermediate position.75 Bartlett attempts to design an epistemological stance that will maintain the cultural contingency of the social constructionist argument while salvaging some basis for truth or objectivity that would allow for critique. In her approach, called “positionality,” truth is always partial, incomplete, and founded on experience rather than unitary and transcendent.76 Nonetheless, it is possible to improve your knowledge by attempting to incorporate other perspectives by imaginatively understanding the viewpoints of others.77 Thus, effort, self-discipline, and self-criticism are essential to knowledge.78 This combination is extremely useful in sketching the outlines of a middle ground. By emphasizing that truth is founded on experience,

The Feminist Dilemma | 89

Bartlett provides grounds for asserting that it is “real”—“in the sense of produced by the actual experiences of individuals in their concrete social relationships”79—and “valid”—“truth claims are significant or ‘valid’ for those who experience that validity.”80 We are justified in relying on such truths as long as we remember that they are provisional and partial. We can hope to resolve disagreement over such truths only through struggles over social reality because there is no standard external to experience to which we can have recourse.81 But resolution, although never certain, is possible: “internal truths [are those that] make the most sense of experienced, social existence.”82 In a more recent article, Bartlett offers a view of the uses of tradition that builds on this epistemological middle ground. She describes tradition as multiplicitous and constantly in a process of recreation through interpretation. She then argues forcefully that feminists should appropriate and reinterpret tradition in the law, rather than advocate a model of change that is a simple rejection of tradition.83 I am in great sympathy with her suggestion that feminists should explore the positive aspects of our legal tradition and seek to build on them.84 Indeed, that project is one of the primary purposes of this book. Nonetheless, these descriptions of an intermediate epistemological stance are presently incomplete. At two different points, an important piece of the argument is in need of further elucidation. When searching out and imaginatively entering other perspectives, we are not required to incorporate all viewpoints, but simply to be open to their challenge.85 This limit is, of course, essential if positionality is not to degenerate into “a strategy of process and compromise that seeks to reconcile all competing interests.”86 But if we must make judgments about which perspectives to incorporate and which not, then by what standard are we to assess competing viewpoints once we have tried them on? Similarly, when we struggle collectively over descriptions of our tradition (or our present reality—which are never entirely separate), how are we, as a political community, to choose between competing and incommensurate—and equally “valid”—truths? In other words, as a matter of both individual and group decision making, what standards are available from this epistemological position? Bartlett has provided a useful sketch of the kind of stance we must assume when using or seeking our moral principles, but we also need to know more about what such principles themselves would be like. At some points, Bartlett seems to suggest that there just are certain values that we all can and should agree on that might guide at least some of these

90 | The Feminist Dilemma

choices.87 She never provides, however, a clear statement about the source or content of these values.88 Positionality and the creative reappropriation of tradition are important parts of the solution to the problems of social constructionism—along with pragmatism and standpoint theory— but they must be supplemented by an account of the sort of standards available to us. Many theorists have attempted to define such moral standards; ones that would not collapse into either transcendence or relativism.89 Some feminists have built on the work of Jurgen Habermas, which draws on a model of human communication to define methods of and preconditions for participation that, in turn, provide a foundation for moral and political judgments.90 Others have suggested a model of human flourishing,91 a traditionally feminine ethic of care,92 or a postmodern insight of respect for otherness93 as possible foundations for ethical and political values. These theorists are seeking to define a middle ground denied by Cartesianism. This is not a middle ground between Cartesianism and social constructionism. It is, instead, a middle ground between the poles of the various dichotomies within Cartesianism (e.g., objective/subjective, universal/particular). The dichotomies and Cartesianism itself deny the possibility of such a middle ground. Social constructionism can only help us to escape the Cartesian trap if it can explain or describe this excluded middle. It is the failure to do so that generates the echo of Cartesianism within social constructionist theory. Martha Nussbaum, for example, has recognized and addressed these issues. She has suggested that an Aristotelian theory of the virtues could be both particularist and objective,94 threading the middle path between such dichotomies. First she identifies certain areas of life and then defines virtues as choosing or acting properly in those areas.95 She acknowledges that the answer to the question “how does one act properly in this area?” will be highly context specific; indeed, there may be many equally valid answers for the same virtue in different contexts. But she insists that “[t]he fact that a good and virtuous decision is context-sensitive does not imply that it is right only relative to, or inside, a limited context, any more than the fact that a good navigational judgment is sensitive to particular weather conditions shows that it is correct only in a local or relational sense.”96 In later work, Nussbaum has pursued this middle course further. She argues, for example, that ethical truth can be understood on a therapeutic model that collapses neither into the Cartesian (and in her view, Pla-

The Feminist Dilemma | 91

tonic) vision of objective and universal truths nor into simple conventionalism. Ethics, she argues, can be seen instead as pursuing knowledge that, while fundamentally tied to the human condition, is nonetheless entitled to be called “true” because of its internal coherence, its correspondence with our best account of human nature, and its consistency with other areas of human knowledge.97 Nussbaum’s concern to preserve this middle ground has led her to be highly critical of postmodernist arguments that undermine the possibility of truth.98 She has identified her form of feminism as both humanist and liberal, arguing that a great deal of social construction can be recognized within these traditions and that some universalism is necessary to ground the critique of unjust social institutions.99 The willingness to embrace these labels distinguishes her from many others who are attempting to chart this middle course. Nonetheless, Nussbaum’s position illustrates certain characteristics that many feminist theories in this category have in common. The principles they suggest are not transcendent because they are tied to human experience, subject to constant revision in light of that experience, and susceptible to real cultural variation. They therefore meet the requirements of social constructionism. But they are “objective” enough to provide a basis for some critical bite because they are not simply culturally contingent; they rely on some regularity in human experience at least over some periods of time. There is, of course, no reason to believe that only one substantive principle will fill this middle ground: it may be that several or even many principles will serve the purpose of dissolving this dilemma. It is not necessary that we have some final, determinate method of resolving all moral or epistemological disagreements.100 But it is necessary that we have some way of narrowing the field of contending positions to a manageable number and that we have some idea of the considerations that are relevant in arguing about those contenders. The notion of a non-Cartesian middle ground could provide such parameters for debate if it were sufficiently specific. Such specificity is, however, essential. Without it, it is all too easy to slip back into the Cartesian trap of reading any such principle as either unattainably transcendent or uselessly relativistic. We need to know more about what the middle ground looks like, how it operates, in order to use it as an effective basis for an alternative epistemology. It is entirely possible—indeed, from a feminist perspective almost inevitable—that such

92 | The Feminist Dilemma

specificity will only be available locally rather than globally. That is, the standards for judgment will need to be defined for particular areas of inquiry rather than for all possible knowledge claims. Nussbaum has explored this type of judgment in detail in literary and philosophical contexts in her other work.101 One goal of this book is to continue such exploration in a particular area of law.

part iii

Reconstruction

As the previous chapter demonstrates, there are good reasons why feminists should not, perhaps even cannot, abandon the concepts of truth and autonomy. These concepts serve a variety of functions that are important and valid. This section will attempt to describe those functions, and then go on to outline social constructionist models of truth and autonomy that can fulfill those purposes. The goal is to demonstrate that feminists who are sympathetic to the social constructionist critique can, nonetheless, recognize the values of truth and autonomy and promote them in a form consistent with feminist commitments.

93

5 Reconstructing Truth

In this chapter, I will describe the functions of the concept of truth in our individual and collective lives. I will then argue that it is possible to conceive of truth in a way that avoids the dangers, incoherencies, and harms of the Cartesian model and that still allows the concept to fulfill the functions I have identified.

A. The Functions of Truth I intend this description of the functions of truth as an exercise in cultural anthropology. I am trying to describe how we actually use truth claims— what we would feel ourselves unable to do, that we want and need to do, if we could no longer make such claims. Part of the way we use such claims is, of course, to bolster the hierarchies generated by Cartesianism. I have omitted those uses that I have already criticized as illegitimate. Thus, the descriptive project is supplemented by a normative assessment. I hope to demonstrate that there are functions of truth in our present practices that are useful and important. In other words, this is a description of a normatively defined subset of existing social practices.1 The functions that I will describe are, I believe, distinct and independent, in the sense that, if one is found to be illegitimate or unjustified, the others will not necessarily be affected. Indeed, because this is an effort to describe concrete social practices rather than an attempt to define a coherent concept of truth, the various functions need not be joined by any common bond beyond our social usage of truth claims. Nonetheless, there may be a general category of which all of these functions are members: the category of claims that other people can and should recognize one’s view. By “recognize” I mean not merely that they should acknowledge that your view exists, nor even that they should admit your right to

95

96 | Reconstructing Truth

hold it, but that they should see the substance of the view as making a demand on them as well. It is the purpose of this section to describe the nature of those demands inherent in a truth claim.

1. Shared Reality Perhaps the most obvious social and political function of truth is to facilitate the creation of a shared reality that can ground joint action and decision making by groups. In the political realm, if anything like a democratic process is to be used for decision making—whether it is majoritarian or consensus based, direct or representative, and regardless of the fine points of political theory and voting procedures—then the members of the group must be able to recognize at least some shared realities. In social life more generally, the possibility of cooperative action for any group—a family, a business, a school—requires at least a minimally shared reality. There are at least two ways to see this connection between cooperative action and shared reality: one is in terms of the cognitive elements of cooperative decision making (i.e., language and conceptual categories) and the other is in terms of trust. First, cooperative decision making requires a shared reality because it depends upon collective use of language and conceptual categories. In other words, some minimal intersubjective knowledge must be possible to allow for the use of language, and language use is a precondition of perhaps all forms of human social organization.2 In the political realm in particular, if we see democracy only as a form of preference summing, in which fully formed individuals merely register their desires through a voting mechanism, this shared reality might be limited to some ability to understand the proposal under consideration and the procedures themselves. Even here, however, such a minimal shared reality is necessary if the process is not to degenerate into total chaos. In any richer account of democracy, which includes some role for deliberation, whether on means alone or also on ends, a correspondingly richer shared reality would be needed. The intersubjective knowledge claims possible in such a richer, shared reality might extend to moral and political claims along with more empirically oriented ones. Second, social cooperation requires some minimum amount of trust between the participants. Whatever else may be necessary to such trust—i.e., perhaps practical connections or emotional commitments—

Reconstructing Truth | 97

there is certainly an epistemological precondition. One cannot trust someone who does not share, at least to some extent, the same reality. One cannot trust someone who cannot or will not participate in the process of making and accepting claims of intersubjective knowledge. Part of this is simply a matter of sincerity: if your interlocutor is not participating in this process because he is not sincere, then he is manipulating you and treating you with insufficient respect for your subjecthood. This is no basis for trust. But sincerity is insufficient. Even a completely sincere person cannot be trusted if he is delusional. He does not share the same reality and, therefore, cannot be relied upon. Collective social action—including democratic political action—requires this reliance. Thus, one way of understanding the need for shared reality in order to support social cooperation is that such a shared reality is necessary to trust and trust is necessary to social cooperation, including in politics.3 A truth claim can be understood as a claim about such shared realities. It makes a demand for recognition by the listener that is not present where no truth is claimed. To call something true is to imply that it (generally) can and should be accepted as a belief by the listener. In other words, the function of truth here is to make claims for the possibility of intersubjective knowledge and the existence and desirability of a shared reality. A qualification is in order: knowledge production, making truth claims, is hardly the only way in which we build a shared reality. Many— perhaps all—of the activities in our lives contribute to the construction of the reality we share. We build a shared reality every time we smile at another person, or turn our back on her, but we generally do not understand such activities as efforts to seek truth or knowledge. In other words, the processes that we call truth making are only a subset of the many ways in which we construct a shared reality, but the construction of such reality is nonetheless one of the important functions of truth processes. There are large and profound questions here concerning the relationship between truth and knowledge that I do not intend to answer. At this stage in the argument, I am seeking to show only that we do use truth claims to construct a shared reality and that we find this an important practice that we do not want to abandon along with Cartesianism. An alternative model of truth should, therefore, accommodate this use of truth. A variety of positions on where we draw the lines around the categories

98 | Reconstructing Truth

of truth and knowledge are consistent with this claim and I do not intend to foreclose any of them here. There is one question, however, that I think I must answer in order to make my limited point: why is truth so important if there are other ways of generating shared reality? I am arguing that we should retain the concept of truth, even if we reject Cartesian versions of it, because it does important work, but why must we retain truth processes if we could achieve the shared reality necessary to joint decision making through other activities? The answer is this: truth processes are a particularly important means of creating shared reality because they involve relationships that reflect the moral foundations of the goal of a shared reality. Why do we care about the creation of shared reality? I have suggested that we care because we need to be able to make joint decisions and engage in cooperative social action, but the issue is, of course, much broader. The point in seeing shared reality as a goal is that we recognize our interdependence and the moral necessity for each of us to treat the others as joint participants in the project of building our collective life. The value of shared reality thus arises from a particular moral commitment: to go forward together on the basis of trust and understanding. Making truth claims is a form of reality building that relies upon and supports this commitment.4 In the process of making truth claims on each other, we are guided to treat each other as equal participants in the construction of social life.5 Some forms of creating shared reality violate this commitment by essentially forcing some participants to go forward together without the necessary trust or understanding. The kind of shared reality that operates in abusive families or hostage situations might be an example of this phenomenon.6 Truth making is a particularly important way of building a shared reality because it involves and generates the moral relationship that is at the heart of the reason why shared reality matters.7 If the critique of Cartesianism is taken to the extreme suggested by some postmodernist theorists, it might seem that no such intersubjective knowledge or shared reality is possible.8 Indeed, with the disintegration of the self, the “subjects” between whom such knowledge is shared become so fragmented that the very notion of intersubjectivity may be called into question. One can see social constructionism as faced with a dilemma here. If the social foundations of knowledge are seen as too static, then social constructionism becomes excessively conservative and cannot allow sufficient room for serious criticism and reform of society.

Reconstructing Truth | 99

If, on the other hand, the social foundations are seen as too fluid and unstable, then social constructionism becomes unable to support claims for intersubjectivity and community.9 It is the latter horn of the dilemma that causes this problem for democratic decision making and the underlying value of reciprocal moral relationships. The point I wish to emphasize here is simply that we cannot accept that horn: the need for an intersubjective knowledge that will allow for a shared reality is fundamental to democratic aspirations and, indeed, social cooperation more generally, and any form of social constructionism that makes it impossible—i.e., that demands that we completely abandon truth, understood in this way—is for that reason unacceptable.

2. Deep Critique Nor can we accept the other horn of the dilemma. Truth claims also function as the basis for an appeal to something larger and deeper that can be used to criticize social conventions or culturally sanctioned assumptions. Martha Nussbaum has pointed out that appeals to abstract values, for example, assert norms independent of a particular culture and can be especially important for oppressed groups.10 Just as the first use of truth—insisting on the possibility of a shared reality or intersubjective knowledge—rejected one horn of the social constructionist dilemma (total relativism), this use of truth—insisting on a grounds for deep and sweeping criticisms of culture—rejects the other horn (conventionalism). From a feminist perspective, finding stable grounds for such a critique is essential, because we wish to assert both that we live in a sexist culture and that feminism is possible. Some forms of social constructionism apparently accept the conventionalist position. Richard Rorty, for example, has been quite explicit about this acceptance.11 He has, as a result, been criticized by feminists, among others, for ignoring the impact of such conventionalism on groups suffering under systematic domination.12 There is no question that the ability to make claims about a larger or deeper reality in order to criticize one’s own society is a valuable tool, particularly to those in positions of subordination within that society. The question, of course, is whether such a tool can be fashioned from the materials of social constructionism. As with the other horn of the dilemma, I wish to suggest that a social constructionism that cannot sustain at least some version of this function of truth is, for that reason, unacceptable. Finding an adequate model or

100 | Reconstructing Truth

metaphor—a description of the position on this continuum that offers a workable balance between these two extremes, a way of thinking about one’s position that helps to retard the slide to either end—is one of the challenges of social constructionist epistemology and is part of the goal of the next section.

3. Decisions That “Work” In the social or political realm of group decision making, truth serves another function as well. Even if we are capable of reaching a joint decision, through democratic procedures and with adequate scope for criticism, such decision making loses a great deal of its value if the decisions reached through it don’t work.13 We want the process to generate decisions that move us closer to our goals and that produce as few unanticipated evils as possible. To claim that something is true is, in part, to argue that knowledge of it will help us to reach decisions that will work and, implicitly, that reaching such decisions is a valuable social enterprise. Such truth claims insist that we should care, not only about the symbolic import of our decisions, nor even just about the process through which they are reached, but also about their practical consequences. The central issue, of course, is determining what we mean when we say that something “works.” This use of truth may be the one that is closest to the hearts of many Cartesians because they understand “working” to mean manipulating an external and objective reality to meet our needs or desires. Thus, the truth that it is necessary to take into account in this process is knowledge about the nature of that external reality. For example, if I want to build a bridge that will allow me to cross a river, I must take into account the width of the river, the strength and flexibility of my materials, and the vertical distance the river may rise with flooding. If we hope to escape from Cartesianism, then we must come up with a meaning for “working” that does not rely on such objective knowledge of an independent, external reality but that nonetheless insists that practical consequences matter. I will draw the connections between this meaning of “working” and an alternative model of truth in the next section, but here I want to argue two preliminary points: first, that such a standard is conceivable in non-Cartesian terms; and second, that there are important values served by caring about such a non-Cartesian version of workability.

Reconstructing Truth | 101

A concept of workability need not be Cartesian. Indeed, the philosophical tradition most concerned with this notion is the very non-Cartesian tradition of pragmatism. The pragmatic approach suggests that truth is what works, and makes no claims for its relation to any reality independent of human understandings. It “works” if it helps us to cope with our environment, including the internal tensions we may feel as well as the social and political circumstances in which we find ourselves.14 It “pays to pursue” truth;15 truth is what it is good for us to believe.16 Feminism can add to this non-Cartesian account the stipulation that it only “works” if it does not result in systematic, illegitimate hierarchy, such as patriarchy. Another way to put this is that the “us” for whom it must work must centrally include those who are systematically oppressed by such hierarchies.17 Thus, the standard for the social constructionist concept of truth, understood as workability, is fundamentally ethical rather than simply epistemological,18 and it is thoroughly non-Cartesian. The importance of such a non-Cartesian concept of workability lies precisely in its pragmatism, that is, its focus on the concrete details of human experience. Perhaps an example will illustrate this best. Some years ago Newt Gingrich suggested that we should drastically curtail welfare and revive the practice of placing children whose parents have insufficient funds to raise them in orphanages and similar institutions. Assume that we shared Gingrich’s apparent belief that many poor parents show insufficient responsibility for their children, that they should refrain from bearing children that they cannot afford to raise, and that the government should not encourage their irresponsible behavior by helping them to support those children. If we shared these beliefs, then this proposal would have merit as a symbolic gesture through which the government expressed its disapproval of this behavior and its unwillingness to facilitate such irresponsibility. Even if we thought all of that was true, however, I suspect that many people would want to say, “But what about the children? What about the practical consequences to a child who is now faced with the prospect of living with her parents but without adequate food, shelter, education, and health care or procuring those necessities only at the cost of giving up the only loving family available to her?” We would want to deny that the symbolic gains—even if real—outweigh the practical costs. We would want to say, “Don’t sacrifice these children to your symbols. The truth is that your proposal would cause them great suffering.”

102 | Reconstructing Truth

The importance of this function of truth—the claim that, in making decisions, we must focus not only on the symbolic significance of our actions but also on the practical impact they will have—lies in its ethical dimension. It demands that we take seriously the pain and the happiness of others. As such, it represents a form of respect for them.19 This claim also insists that we recognize responsibility towards those others in making such decisions. In other words, it is because failure has consequences for other subjects who matter that we care whether or not a decision “works.” This focus on “working” helps to explain why, even if we wish to abandon Cartesianism, it would be foolish and self-defeating to abandon the practice of science. The many benefits of science—for example, medicines to cure disease and bridges that don’t fall down—can be recognized and appreciated without Cartesian assumptions if one includes the concern for the impact on other subjects as part of the function of truth. Indeed, I believe that it is precisely this ethical concern that gives force to the Cartesian commitment to this aspect of truth as well. If I used inadequate materials to build my bridge just because I wanted to make some symbolic point about rejecting the producers of the appropriate materials (even if that point was a valid one), I could be criticized for failing to take account of the practical consequence that people could be hurt by the failure of the bridge. Or if I made decisions about the use of certain medicines on the basis of a process—no matter how democratic—that did not take account of the impact of my choices on those suffering from disease, then, again, I could be criticized for ignoring the “truth.” This ethical understanding of what it means to “work” captures the real, although perhaps hidden, moral concerns behind even the Cartesian use of this meaning of truth. Understanding “working” in this way does, however, contradict Cartesian assumptions by confirming the extent to which knowledge claims are permeated by culture and context. For example, if it turned out that the bridge in question was one that I was building in my garden strictly for the purpose of my visual appreciation, then it is not at all clear that my choice did not work. If I could not have appreciated the bridge if it had been made of the symbolically hateful material, then it fulfills its purpose better being built the way it was, even if it is unable to support actual traffic. In other words, “working” is defined, not by reference to some external and objective reality, but by reference to ethical considerations applied within the social and individual context of

Reconstructing Truth | 103

the decision, a context always permeated by our purposes and conceptual frameworks. In constructing a non-Cartesian explanation of the “working” function of truth claims, I do not intend to deny the existence of the world or assert that whether or not something works is all in our heads. We experience the world as real and I am not suggesting that we should try to do otherwise. The world is just as real for us whether we imagine it in Cartesian terms or not. Indeed, what it means for the world to be real, in nonCartesian terms, is simply that we experience it as real, rather than as a dream, or an illusion, or one of the other categories of our experience that we distinguish from reality. In other words, the opposite of “real” is not “constructed”; it is “imaginary.” Both real and imaginary experience are constructed, but they are still different in important ways. And one of the differences between the real and the nonreal is that we experience the real world as affecting us in ways that we do not cause or control. That does not mean that our experience of the real world is unconstructed: it is deeply shaped by our cultural categories and expectations, but it is nonetheless not simply imagined by us. One of the ways in which we capture this fact that the world is real (i.e., nonimaginary) and that it has such effects on us is by reference to the moral concern about the impact on others that is described by our truth claims about “working.” In other words, the concern about working is comprehensible without a Cartesian understanding of an objective and external reality that is independent of our conceptual frameworks. If part of the function of truth is to focus our attention on “working,” then, to that extent, the concept is important and should not lightly be abandoned. An alternative model of truth should be able to make sense of this claim and allow for it to serve this function.

4. Connection to Nonhuman Reality In addition to these social or political functions of truth, the concept seems to play an important personal role in the lives of many people. This function is somewhat more difficult to describe than the previous ones, perhaps because it often takes a form that might be called spiritual. Truth can form the basis for an understanding of oneself as connected to reality in a stable way, so that reality is not set adrift or up for grabs, but anchored. And it can offer a foundation for that stability outside of

104 | Reconstructing Truth

oneself—a view of reality in which one is not the ultimate creator but simply a part of the larger whole of creation. The relationship to the elements of nature—not just other human beings—is central to this view: the sense of being part of, perhaps even in some ways on a par with, the rest of nature. Some versions of postmodernist theory have shown little sympathy for this function of truth. Richard Rorty, for example, has suggested that the need for transcendence can be eliminated by a “suitable moral education” that “tries to sublimate the desire to stand in suitably humble relations to nonhuman realities into a desire for free and open encounters between human beings, encounters culminating either in intersubjective agreement or in reciprocal tolerance.”20 In an extremely interesting article, Richard Thomas displays somewhat greater sympathy when he proposes that “[s]uch longings [for transcendence and foundations] are the stuff of religion, and the choice to be secular entails a loss that cannot be recouped through the law, or politics, or reason, or art.”21 While it may be true that we should not seek to found our politics on these sorts of truth claims, we may, nonetheless, wish to secure First Amendment protection for them as speech. In other words, if this is one of the sorts of truths that serve an important function for human beings, it should be part of the truth protected by the free speech clause.22 I would, therefore, like to tentatively suggest that a social constructionist revision of the concept of truth must attempt to recognize in some way the personal function of truth as a stable connection to a reality in which one is a part of the larger whole. The new concept of truth may not be able to provide a basis for every aspect of this function as experienced by all those who feel these longings, but it should be designed to provide as much as it can. Truth, then, serves several important functions in human experience. While the Cartesian model of truth should be rejected, for the reasons suggested by the feminist critique, we need to explore the possibility of alternative models that might allow us to continue to fulfill these needs. Such an alternative might then offer a new vision of the relationship between truth and speech.

Reconstructing Truth | 105

B. A Relational Model of Truth This section will offer an outline of an alternative vision of truth. First, I will describe this new model in order to demonstrate the senses in which it is non-Cartesian. Second, I will connect the relational model to the feminist themes that I identified in chapter 3 as lying behind the critique of Cartesianism. And finally, I will show how this alternative vision allows us—at least to some extent and in certain ways—to fulfill the functions of truth that I discussed previously.

1. A Non-Cartesian Model The relational model of truth is non-Cartesian because it rejects all of the central tenets of the Cartesian model of truth: objectivity, rationalism, universalism, and the representational view of truth. This new model sees truth as fundamentally contextual and normative. That is, knowers can engage in the activity of knowledge making only from a particular position or perspective, one that is radically shaped by their cultural context, including their normative concerns. Such contexts should not be seen as impediments to knowledge, but rather as the preconditions for it. To ask what we could know if we could escape all such perspectives is like asking what we could see if we could escape having any particular sort of eyes. Interpretation, with its contextual and evaluative character, is as essential to the process of making truth as the biomechanics of eyes are to the process of seeing. One result of this focus on the role of interpretation is that truth is seen as made by human actors rather than found. There is no passive receptivity to an external reality; rather, there is an active process of interpretation. Another result of this focus is that the traditional dichotomies— particularly any version of the fact/value distinction—are made untenable. There are no facts that are not permeated by values because all result from a process of interpretation that is inescapably evaluative in nature. The relational model also rejects universalism: truth may often be plural rather than singular. The identity, cultural and personal, of the knower is implicated at every point in the process of knowledge creation. As a result, different identities may generate different visions of the truth that are equally valid. The rationalist bias of Cartesianism also falls away. First, reason cannot be understood in the abstract and instrumental way that it

106 | Reconstructing Truth

traditionally has been. Reason, like truth, is highly contextual. Second, reason cannot be contrasted with emotion or politics simply because the latter are seen as evaluative in nature. The opposite of a reasoned judgment is not an emotional one, but an unreasonable one: emotional reactions are often quite reasonable.23 None of this means, however, that reason should lose its place among—maybe even in the forefront of—those human capacities that we use to make knowledge. Reason, like truth, is a category worth rehabilitating. When it is seen as contextual, evaluative, and inclusive (i.e., when its opposite is understood as unreasonableness), then reason can play an extremely important role in this alternative model of truth. Reason is the process through which we make sense of our experience. “Making sense” is, of course, a highly contextual matter: the sort of argument that can make sense of our experience in a physics laboratory (mass, velocity, vectors, etc.) is not necessarily the same sort of argument that can make sense of our experience watching a baseball game (strikes, balls, and base hits). To try to explain the latter in the same terms as the former would be to fail to provide good reasons for the relevant experience. The concept of reason captures this concern that some things (arguments, conceptual categories, etc.), in some contexts, help us to make sense of our experience better than others.24 Understood in this way, the concept of reason has a useful role to play in this alternative model of truth. Finally, the relational model sees language in a way quite different from the traditional Cartesian propositional, representational view. Because cultural understandings so deeply shape our truths, and because language is one of the primary vehicles through which such understandings are created, maintained, and transmitted, language is seen as constitutive rather than as representational. Such a view of language and the role of conceptual categories does not mean, of course, that we can make something real simply by saying so. The type of social constructionism on which I am relying does not imply that language is all there is to reality. We experience ourselves as sometimes running up against the world: I cannot make the world whatever I want it to be simply by thinking of it or describing it that way. Our experience never includes a reality that is simply separate from our conceptual frameworks, but that does not mean that our conceptual frameworks are all there is. Our experience of reality is conceptual all the way down, but it is not only conceptual. The important point is that we can never get

Reconstructing Truth | 107

behind or below the conceptual part. The reality that we experience is always already constructed. There is a range of possible metaphors through which we could try to understand the relationship between our conceptual categories and reality. At one end is the Cartesian view that the categories can (and should) be a transparent tracking of a reality totally external to them. My claim is that the other end of this range is not a picture in which there is nothing that counts as reality at all. That is a possible picture, of course, but it is not one that is required by social constructionism, and it fails to preserve our experience of reality. Rather, the other end of the spectrum is a view in which no part of reality is independent of our conceptual categories, but where the difference in our experience between the real and the imaginary is still meaningful. I will offer a metaphor to help capture this view shortly, but first I think it is important to say a little bit about the middle range of this continuum. Many people have wanted to incorporate a substantial amount of social constructionism in their view of reality without giving up entirely on the idea of an independent reality that is not reducible to our experience of it. They have suggested metaphors that involve seeing our conceptual and linguistic categories as a filter or screen through which we perceive, with greater or lesser distortion, the world out there. The idea is that we can never get out from behind the screen—our experience is always constructed—but we posit something independent of us that causes that experience: the world on the other side of the screen or filter. There are other possible metaphors with a similar structure, in which one element of the metaphor represents an external reality independent of our conceptual frameworks. These include the metaphor I offered at the beginning of this chapter, where our interpretive frameworks are imagined as the eyes that allow us to see a reality independent of us. The metaphor we use is important in shaping our understanding of the role of conceptual categories and language in knowledge. I would like to leave open the greatest range of such metaphorical views consistent with the rejection of Cartesianism. My goal here is twofold. First, I will indicate why I think some of these intermediate views—most notably the screen or filter metaphors—are not sufficiently distant from Cartesianism to withstand the feminist critique. The point is to mark the limits on the range of acceptable alternatives. Second, I would like to offer a metaphor to capture the view at what I have defined as the far end of the continuum:

108 | Reconstructing Truth

the view in which no element of the metaphor represents an external (to our experience) and independent (of our conceptual frameworks) reality but in which it is still possible to capture our sense that some things are real and others are not and that this distinction has meaning for us. I hope to demonstrate that this view is more congenial than many have assumed and, in particular, that one could move to this view without having to stop talking about reality altogether.25 Some metaphors, while they recognize social construction, are still too close to the Cartesian view to avoid the impact of the feminist critique. The screen or filter metaphor, for example, is problematic because it retains the Cartesian assumption that the point in the process of truth seeking is to get our picture to match the picture on the other side of the screen as closely as possible. And the picture on the other side, of course, is by definition independent of our conceptual categories, our moral and political commitments. The great danger in such a view is that it might lead to the denial that moral and political values are implicated in our knowledge claims and to a refusal to respond to moral and political challenges to those claims. The screen or filter metaphors suggest that we can get our screen clearer and clearer and that this is the goal. And that suggestion undermines the central benefit of a move to social constructionist epistemology from a feminist point of view: the ability to challenge knowledge claims on moral and political grounds. I should hasten to add that the existence of such a challenge doesn’t mean that the knowledge claims we normally regard as empirical or scientific would automatically lose. As I have described in the earlier section on “working,” the desire to understand the world well enough to have our decisions “work” has itself a deep moral grounding. The moral claim of the scientist who studies some disease may well be more pressing than the claim of someone who wants the process of study to be more democratic, at least if the present procedure better serves the goal of finding a cure. The point is that the scientist must engage the challenge on these moral grounds, rather than saying, “This is a value-neutral way of seeing an objective and external reality.” I will examine this concern at greater length in the subsection on working that follows. The metaphor of eyes, on the other hand, while retaining an element that represents an external, independent reality, does not imply that we could get our picture to match it. Indeed, it suggests that what we see will be determined by the types of eyes we have, and that the question of what that external world “looks like” in the abstract is meaningless: it can only

Reconstructing Truth | 109

look like anything to an observer with a particular sort of eyes. Once we recognize that the eyes that allow us to see anything are conceptual frameworks that include moral and political commitments and assumptions, it is always possible to challenge a knowledge claim on the basis that it would be better, morally and/or politically, to look with different eyes. As a result, this metaphor does not allow such an easy retreat to a claim that the view of the knower is justified simply because it is closer to what the external reality “really looks like.” In other words, metaphors that include an element that stands for an independent, external reality will be sufficient to meet the needs of the feminist critique of Cartesianism as long as they do not facilitate the avoidance of moral and political challenges to knowledge claims by measuring the truth of such claims against coherence with an external reality seen as independent of morality and politics. The external reality can appear in the metaphor, but it cannot do this work of justification independent of moral and political commitments. Within this limit, however, I mean to leave open the possibility that those who find these intermediate models appealing could adopt the argument of this book with respect to truth and speech without having to abandon those metaphors that include an external reality. The remaining question is, what is the alternative? Is there a metaphor that might capture a more thoroughgoing social constructionism and explain how it is possible to continue to talk about reality without including an external, independent world in the metaphor? I would like to suggest a metaphor of weaving. I hesitate to suggest this metaphor in light of the fact that weaving has become such a cliché, in feminist writing generally and in alternative epistemologies in particular. But, as a weaver myself, I find it useful and appealing. If others are put off by it, they are welcome to rely upon the description rather than the image. In this metaphor, the loom is our conceptual frameworks, the yarn is the impact of the world on us, and the finished cloth is the reality we actively construct. This metaphor helps to highlight several aspects of the situation that are important for understanding the constructed nature of reality. First, the yarn has the same metaphysical status as the cloth: that is, the yarn is itself a complex creation made through a process in which conceptual categories are already inherent in our experience of the world. The yarn is already constructed: the impact of the world on us is never independent of our conceptual categories. Nor, on the other hand is it simply reducible to them; that is the point in using a separate element (the

110 | Reconstructing Truth

loom) to represent the conceptual frameworks themselves. This understanding of the yarn helps to avoid the confusion in the screen analogy that comes from the fact that the “picture” we imagine on the other side of the screen is supposed to be actually on the other side: independent of our conceptual frameworks. Nothing in the weaving metaphor is simply and purely independent in this way. Conversely, the loom is not purely socially constructed. Our conceptual frameworks are not made in isolation from our experience of the world but in response to it. I mean to leave open the possibility, in other words, that there might be a regularity in conceptual frameworks across cultures that arises from regularities in human experience of the world. Conceptual frameworks cannot be just anything we want them to be: they must respond, in some way, to an experience that they help to construct but that is not simply reducible to them—just as a loom, while it determines the type of cloth that can be woven, must be built to use the types of yarn available and to result in usable cloth. Thus, both the loom and the yarn should be seen as both constructed and constrained. Why then adopt a metaphor that separates them, implying for each a different nature that I mean to deny? Despite the risks of this misunderstanding, I think it is useful to maintain two separate elements to emphasize that our experience of constraint by the world is more profound with respect to certain elements than others. Sometimes we have the sense of bumping into the world. While such experiences are never innocent of social construction, they are also meaningfully different from the experience of manipulating categories of conceptual analysis. The metaphor, then, is not intended to suggest that yarn and loom are different in nature; rather, they represent different blends that correspond to different experiences. What is the source of the constraint that we experience? Constraint should not be elided with physical reality. We must reject the dichotomy between physical reality and social reality. The reality that constrains us is likely to have both physical and social aspects and, indeed, it is not especially useful, or often even possible, to separate these elements. Thus, my experience of being unable to distinguish between two shades of blue may be a result of the biomechanics of my eyes, or of the lack of distinction in my culture for such shades, or some combination of the two. Physical capacities can be altered by social practices: for example, better nutrition made possible by agricultural advances leads to larger and stronger people. There is a dialectical relationship here that makes it difficult and often pointless to say that the constraint comes from a phys-

Reconstructing Truth | 111

ical rather than a social cause.26 We may, indeed, wish to distinguish between different types of constraint for other purposes—there are contexts in which the distinction would be more useful and coherent—but epistemologically the more salient distinction is between situations in which we feel such constraint and those in which we don’t, regardless of whether the source of that constraint is a human institution or not. It is this distinction that captures the difference between what is real, in the sense of not being simply reducible to how I think about it, and what is “all in my head.” Social institutions can be very real indeed.27 The distinction between the yarn and the loom is intended to point to this difference in the levels of constraint in our experience. Second, neither the yarn nor the cloth is a picture of reality. The cloth is reality, while the yarn is simply a less complex—and probably less useful—piece of reality. To put this another way, if we unravel the cloth to find out what the yarn is like, we will not thereby have come closer to reality. We will, instead, simply have reduced the complexity and richness of the reality we have to work with. This may sometimes be a useful process—when, for example, we need the simpler piece to do some work, as in isolating a virus to make a vaccine—but this process does not bring us closer to some unconstructed reality. Third, every aspect of the finished cloth is a result of the combination of the yarn and the contribution of the loom. The texture, even the color, comes not from either yarn or pattern alone but from the interaction of the two. There is no part or aspect that can be ascribed to only one. And finally, the metaphor captures the idea that we operate under constraints not of our own creation (the nature of the yarn, which is our felt experience of running up against the world, and also the limits of the conceptual categories we are given) but we work creatively with them to make something useful. And the weaving metaphor makes sense of the idea that improvements in our skill in working with our conceptual categories, or even designing new ones (i.e., a different kind of loom), can lead to a greater ability to produce useful and beautiful products. The purpose of knowledge production, in this metaphor, is not the creation of a clearer picture of an external reality. The purpose is, instead, the creation of more useful and/or beautiful cloth. The point in human knowledge is to promote human life, understood in the broadest and most ethical terms. Moreover, to fully capture the relational model of truth, we would need to add to the metaphor that the loom is constructed in such a way that the weaving can only be done in relationship with

112 | Reconstructing Truth

other weavers. As a result, part of what is necessary to the process is to establish the relationship with those other weavers required to create beautiful and useful cloth. And finally, we must add that we find ourselves at the loom with the weaving already in progress: we do not choose the loom or the yarn or the design of the project, although we can make modifications in all three as we go along. In other words, we are thrown into the process of making knowledge as an ongoing project rather than controlling it or choosing it. This metaphor fails in two important respects, which I would like to acknowledge. First, it presents the process of knowledge making as selfconscious. Weaving is an activity we decide to do and we generally know we are doing it when we do. Knowledge making, however, is more like breathing: it is often unconscious and more or less continuous. Second, the weaving metaphor presents the process of knowledge creation as essentially safe and leisurely. This is a serious misrepresentation. We make knowledge not because we want to but because we must. It is a necessary aspect of our lives and one that is often fraught with substantial danger and anxiety: there are serious consequences if the cloth we make is not useful to us. Sometimes, perhaps often, we struggle to achieve understanding, and we do so under conditions in which much turns on whether we are successful or not. I want to emphasize again that the relational model of truth does not mean that reality can be whatever we want it to be or say it is. This model asserts that reality is always constructed, but it does not seek to deny our experience of constraint in that process of construction. The advantage of this metaphor is that it seeks to understand that constraint not from some view external to our experience, in which an objective and external reality that exists independent of our conceptual categories would be visible, but from within human experience, where the constraint operates in and through a world that is always already constructed. I understand that there are important moral considerations that incline some people sympathetic toward social constructionism toward metaphors that retain a world external to our experience. The role played by the external world in the metaphor demands that we recognize nonhuman realities as significant and independent from us in morally important ways. Indeed, I think there are powerful arguments that metaphors that include an external world may better serve the fourth function of truth: providing a connection to a nonhuman reality. I, therefore, do not wish to suggest that the adoption of a social constructionist view requires

Reconstructing Truth | 113

one to move to the weaving metaphor I have just outlined, in which the external world does not appear as a separate element. The move to social constructionism does require one to adopt a metaphor in which the external world cannot do the work of justification that it did in Cartesianism and thereby short-circuit the moral and political challenges to knowledge claims. But the choice between the remaining metaphors is open. The point in the extended consideration of the weaving metaphor was to demonstrate that one of the common fears associated with this more thoroughgoing social constructionism is misplaced: adopting this view does not mean that one loses the ability to talk about reality in a way that is meaningful and related to our experience of how the world constrains us. I wanted to show that this view should be included as one possibility in the range of social constructionist positions on truth remaining after one accepts the critique of Cartesianism.

2. Connection to Feminist Themes The relational model of truth—which can be captured by this range of metaphors—also draws on and exemplifies several themes shared widely by feminist theories in a variety of disciplines. As I argued in chapter 3, these themes and concerns provide the groundwork for the feminist critique of Cartesianism.28 They also provide the supporting structure for this alternative model of truth.

a. inclusion The first of those themes is a concern for those—women and others— who have traditionally been excluded from the social process of knowledge making. This concern is incorporated into the relational model of truth in the form of a focus on the relationship between epistemology and politics. “[I]f rhetoric is the means by which reality, or plural realities, are constructed, then democratization of discursive or rhetorical opportunities is inseparable from the project of democratization itself.”29 The recognition of this connection between epistemology and politics leads to a conception of the process of knowledge making as itself a political process, and one in which democratic reforms are necessary in order to end the traditional exclusion of less powerful groups. So, for example, one feminist scholar has suggested that we should adopt a “representative democracy” model of science, in which experts act as our representatives, applying our values to issues before them and maintaining a

114 | Reconstructing Truth

constant dialogue with us.30 This may sound like a nightmare to some because they envision a collection of people, ignorant of the scientific background, telling scientists that they may not study something (or perhaps even worse, that when they do study it they must come to a particular conclusion about it). There are two responses to this concern. First, a social constructionist view does not require that we reargue the values underlying a particular social practice every time we engage in it. That would be a colossal waste of time and energy. In many situations, we simply go forward, on the basis of the values inherent in our practice, on the assumption that they are worthwhile, and without any direct examination of them. This is completely appropriate because, in many circumstances and for many purposes, there is no need to reexamine the underlying value structure: the practice is working for all those concerned with it. Such is certainly true of much scientific practice, along with many other areas of life, and the adoption of a relational model of truth would not change the fact that we do not critically examine everything we do. Such a model does require, however, that we cannot always go along without such examination. Sometimes, and particularly (although perhaps not exclusively) when the values underlying our practice are challenged, we must engage in that self-examination. Relational truth requires that we are always vulnerable to such challenge and we are always subject to the risk that we will need to engage in this dialogue, but it does not require that we will actually do it every minute of every day. Second, the nightmare vision here relies upon an implicit contrast between the practice of science, seen as neutral, and the moral and political concerns that will be used to cabin it. But the point in relational truth is that the practice of science is already based on such concerns. Some of those concerns are extremely important and legitimate (e.g., the promotion of human welfare), while others are less legitimate (e.g., the maintenance of gender or racial hierarchy). The demand for inclusion requires that scientists scrutinize their practices for such implicit values and be prepared to engage in a dialogue with those who would challenge those values or suggest that others should be more important. The “representative democracy” suggestion simply offers the view that, in the end, the values served by the practices of science ought to be those of the people as a whole, and not of scientists alone. The goal in such a dialogue about values is not some idealized speech situation in which we experience no hierarchy or power differentials. All discourse suffers from distortions, and the search for an undistorted dis-

Reconstructing Truth | 115

course is, in the end, not very different from the Cartesian search for a neutral observer.31 The point is, rather, to subject our process of knowledge production to scrutiny under the same democratic standards that we apply—or aspire to apply—to our political process. Indeed, the point is to recognize that the process of knowledge production is ultimately a part of the political process of determining the distribution of power and resources in society. If we are concerned about the exclusion or marginalization of certain groups from that political process generally, we should be equally concerned about it here. The argument I have just offered describes the way in which the relational model of truth opens epistemology up to the sorts of concerns about exclusion that are common to feminist theory. There is another argument that could be made about this connection, however, one suggesting that this alternative model of truth helps to justify or ground the concern about exclusion, rather than merely sharing it or extending it to a new area. Once truth is seen as contextual and perspectival, we must recognize our dependence on others, with different perspectives, to help us see beyond our own limited view.32 Indeed, when the standard for justification is no longer a correspondence to an external reality, but is instead defined by reference to concerns and assumptions—moral and otherwise—shared by a certain community, then the broader and more inclusive the audience a position can address, the better justified it is.33 This process may sound like the competition of the marketplace of ideas, but it is not. The views from different perspectives are not competitors for belief: each view does not seek to disprove the others and it is not assumed that one will come out the winner. Instead, they all offer a way to see a shared reality and, ideally, the view that will prove most justified is the one that is best able to accommodate a broad range of perspectives. Nor, however, is this a simple majoritarian process: our perspectives are changed through the experience of exchanging views, hopefully changed in ways that make such accommodation more possible and more likely. In addition, some views may not be properly accommodated at all, for example, racist and sexist views. The process of accommodation includes evaluation; it is not simply a summing up of the views with which we each begin. Thus, one of the conditions for believing that we have adequate grounds to consider something true is that we have listened to and considered the views offered from a broad variety of perspectives. Bringing in the excluded becomes a precondition for making knowledge well. This

116 | Reconstructing Truth

connection is one of the bridges tying moral and political theory to epistemology in this alternative vision.

b. relationality The second feminist theme that I discussed earlier—the focus on the relational nature of knowledge—is so basic to this new model of truth that I have called the model itself relational. The model of truth is relational in two senses: first, the knower and known shape each other through their relationship, and second, the process of knowledge making takes place within a relationship between members of a community. In this model, indeed, no knowledge is possible outside of relationship, and knowledge itself may be said to consist of certain relationships. The first sense in which the model is relational is that it focuses on the relationship between the knower and the known. This relationship highlights the fact that knowledge implicates the identity of the knower, both in shaping the thing known and in being shaped by it. As a result, when one accepts this responsibility, seeks out excluded perspectives, and tries to learn from them, one changes not just what one knows but also who one is. The relational aspect of this model of knowledge helps to explain how it is that the process of learning from others leads to the creation (not the discovery) of a new reality, including knowers with new identities, and therefore opens up the possibility of still other new realities that did not appear possible before. The second sense in which the model is relational concerns the need for knowledge making to take place within social relations among knowers. This relationship highlights the fact that knowledge is always contextual, situated, and shared. It emphasizes the often invisible, but always necessary, background of practices, institutions, and conceptual systems on which the process of knowledge making depends.

c. blending morality and epistemology The third theme common to much feminist theorizing involves breaking down the barriers between epistemology and moral and political theory. The relational model of truth participates in this process by focusing on the issues of agency and responsibility. These concerns are absolutely central to this alternative model of truth. Once one adopts a social constructionist view, the issue of agency becomes critical because one must explain the meaning and conditions for individual agency under social constructionism. The next chapter will address this issue more generally. Here I

Reconstructing Truth | 117

will focus on one particular aspect of the issue of agency in the context of social constructionist epistemology: the role of responsibility. Responsibility is a fundamental part of the process of making knowledge because of the inevitably normative and evaluative nature of the project. Truth is made through a process of selection and the selection necessarily involves judgment, including moral judgment. The knower must take personal responsibility for those judgments.34 This is not to say that these judgments are totally uncaused; they are, of course, deeply shaped by the context and life history of the knower. But persons are not simply the passive locations in which social forces collide. There is an active process of interpretation—of remaking of the cultural materials one is given—that is an inescapable part of knowledge making. Thus the centrality of responsibility in the epistemological context is closely tied to the possibility of agency in the moral/political context. To accept epistemic responsibility is to recognize that the frameworks one applies in the process of interpretation are subject to challenge and criticism, including moral criticism, and must be acknowledged, examined, and defended in those terms. Responsibility, in other words, is a relationship to other people—those who do or might engage in such a challenge.35 It is also, however, a relationship to oneself and to the frameworks and moral ideals one holds. One must engage in the practice of knowledge making in a way that does not violate those moral commitments. One important implication of this focus on agency and moral responsibility is that it helps to break down the assumption, commonly associated with a social constructionist view, that a particular experience leads inexorably to a certain identity and perspective.36 If interpretation is an active process, and if we must accept responsibility for its moral implications and assumptions, then we can and must learn from those with a different set of experiences. We cannot, perhaps, ever simply see the world the way they do, but we may see the world or ourselves differently if we actively use their insights as a lens on our own experience.37 Thus, the relational model of truth makes explicit and central the relationship between epistemology and both politics and morality. It causes us to shift our attention away from mechanisms for “accurate representation” and toward concerns about inclusion and moral responsibility. And it makes clear the possibility—indeed, the moral imperative—that we examine our own perspective critically and attempt to learn from the perspectives of others.

118 | Reconstructing Truth

3. The Functions of Truth With this sketch of an alternative model of truth available, we can now turn to consider whether and to what extent this model can serve the purposes of truth outlined in the previous section.

a. shared reality and deep critique The first function of truth is to serve as the basis for claims of intersubjective knowledge, claims concerning a shared reality that makes democratic decision making and cooperative social action possible. The second function of truth is to serve as the basis for a critique of existing social institutions and conventions, a critique that can be broad and deep rather than merely a matter of “tinkering.” I suggested in the previous section that these two functions marked the acceptable range on a particular continuum: a continuum running from a totally open cultural framework, which would allow for too little shared reality, to a totally closed one, which would allow for too little critique. The difficulty, of course, is that however one defines the cultural framework, there are constant pressures from both directions. For example, in working with the cultural category of gender, when one focuses on the force of its constraints on women’s lives, one can find oneself sliding toward a position in which the cultural category is so fundamental and so powerful that it is difficult to see how any challenge to it could possibly arise.38 If, on the other hand, one focuses on the incredible variation within the category and the ability of individuals to reinterpret it, then it becomes difficult to see gender as a coherent category at all or as a source of serious oppression.39 The challenge is to define or describe a position on that continuum that will retard the slide to either extreme. If knowledge is all contextual, one way to see this dilemma is as a problem about the size of the context. If we are looking for a context rich and powerful enough to generate a shared reality on a controversial issue, then we can find ourselves drawing the context ever smaller, in order to find enough common ground for a consensus. Such a shrinking context, however, can leave us with a radical relativism in which truth can be determined only from within a given viewpoint and viewpoints are—at least potentially—no bigger than a single person. Shared reality for the purposes of political decision making and cooperative action would then become difficult or impossible. On the other hand, if we are looking for a context big enough to include deep and broad challenges to cultural as-

Reconstructing Truth | 119

sumptions, then we may find ourselves expanding the boundaries of the context beyond the usual cultural limits. Such an expanding context, however, can leave us with too little in common to draw on for any kind of claim at all. Deep criticism would then become meaningless: it would be like criticizing humans for not being whales. Thus, the two functions of truth push in opposite directions, but each—if taken to an extreme—undermines its own goals. The need for a shared reality pushes us toward smaller contexts, but if they are too small they cannot sustain such a reality. The need for deep critique pushes us toward larger contexts, but if they are too large, they cannot sustain such a critique. When seen in this way, the problem seems to be finding the right size context: one that is small enough to generate a strong shared reality, but big enough to include the materials for deep criticism of its own assumptions. I want to suggest that what makes this problem so intractable is the assumption that contexts are given, as though there were a list of the possible contextual frames for a particular issue and we merely have to choose between them. For example, on issues of gender equality, one might see the possible frames as running from the level of each individual (“you see it your way and I see it mine”) to the ethnic and religious communities that have taken various positions on the issues (“you’ll never convince them because they are coming at the issue from within their religious framework”) to the national community that has made certain legal commitments (“it’s about the meaning of the equal protection clause”) to the historical tradition of which that national community is seen to be a part (“the history of Western culture”) to the history and/or present circumstances of human beings all over the world (“the human condition”). Although arguments are possible within each of these frameworks, nonetheless, the choice of a framework often seems to largely determine whether there is sufficient common ground for a joint decision and whether that common ground includes sufficient materials for a serious challenge to deep cultural assumptions. This phenomenon of assuming preexisting contextual frameworks contributes to the difficulty in two ways. First, it makes it seem as though the ability of any given framework to meet the needs for common ground or deep criticism is set in advance because the nature of the context— its boundaries and meaning—are given. Under the relational model of truth, this is simply wrong. Each context is created in the act of calling upon it, and its resources and possibilities are constantly subject to

120 | Reconstructing Truth

reinterpretation and revision. This does not mean that simply anything is possible in any framework, of course, but it does mean that we determine what is possible through the very process of defining it rather than knowing its limits in advance. It is the need to recreate one’s context through this process of interpretation that has led Katherine Bartlett to argue that tradition is not contrary to change but constituted by it.40 Second, this list-of-contexts phenomenon suggests that there is a finite number of contexts on which we could draw and they are determined in advance. Again, this is a mistake: not only the substance of a framework, but its very existence as such is a matter of constant creation and recreation. To take the example above of gender equality issues, feminism claims that one context from which we could see these issues is the experience of women, and one of feminism’s greatest achievements has been to make that context a reality through mechanisms like consciousness raising and political action. This context was created by the activity of feminists and wasn’t on anyone’s list before they acted. Thus, the list phenomenon is misleading both about the fluidity of the boundaries and meanings of existing contexts and about the possibility of new contexts. And it is misleading in a particular way: it fails to recognize the active, interpretive, and relational character of the activity of knowing. We are not at the mercy of some slippery slope down which we will necessarily slide into either too large or too small a context. Contexts don’t just sit there waiting for us; they have to be built and constantly rebuilt. And it is the connection to politics and morality that must guide us in interpreting and constructing contexts. Thus, if we find that there is not enough common ground for the social group that must make joint decisions, then we must build that common ground. The construction of such common ground is a large part of the purpose and appeal of the narrative turn in legal scholarship.41 The use of stories allows us to expand our experience without actually changing the circumstances of our lives. And that expanded experience may provide a foundation for a shared reality across significant lines of social difference, such as race, class, or gender.42 Or if we find, on the other hand, that we share a great deal but that there are not sufficient materials within our present context for deep criticism, then we must build a context that includes those materials. As I mentioned above, the recent history of consciousness raising provides a striking example of the process of using the materials of a shared reality— the experiences and unhappiness of many women in a patriarchal soci-

Reconstructing Truth | 121

ety—to generate new grounds for critique. Another example might be the borrowing of ideas from continental philosophy by American legal scholars and the effort to work them into forms that are comprehensible and useful within our existing legal culture.43 Such efforts at cultural construction are not guaranteed to be successful, and certainly not within any particular time frame. A useful cultural context is an achievement; it exists only in and through the creative activity of people. And it may take the efforts of many people and significant amounts of time before the context can effectively be drawn upon for the purposes of joint decision making and/or deep critique. Seeing the resolution of many social issues as a matter of cultural evolution may seem frustrating and hopeless because of the incredibly slow pace of such cultural change. For feminists thinking about issues of violence against women, for example, such arguments may seem to be a counsel of despair because of all of the women who will be harmed or die while we are waiting for the culture to shift. There are three possible responses to this very serious problem. The first response is that one should never underestimate the possibility of cultural revolution. Once the seeds have been sown—as they clearly have for feminism—it is sometimes possible for a single event to galvanize wide and deep cultural change. I have heard people describe the television coverage of Martin Luther King Jr.’s march from Selma to Montgomery as such an event. It seems that many people who simply did not see the evil of racism until that moment finally had a vision of its ugliness when they saw the fire hoses turned on peaceful protesters. Such an event is never the end of such change—as the continuing racism in this country amply demonstrates—but it may provide the impetus for a very large and quick jump forward. The second point is that this concern about the speed of change should lead us to pay very serious attention to the mechanisms for building cultural contexts. We must be sure that our channels for cultural construction are open and strong so that we can respond to these pressing needs as quickly and powerfully as possible. In chapter 7, I will present an argument for why speech is a particularly important mechanism for cultural construction and reinterpretation. If that argument is persuasive, then it suggests that the health of our public dialogue should be a central concern for those who wish to press for deep, and quick, cultural change. Finally, let me say that I have tremendous sympathy for this complaint about the social constructionist approach generally and this model of

122 | Reconstructing Truth

truth in particular. I believe the inability of the relational model to provide more hope for and help with fast change on pressing problems is indeed a weakness. But, I do not see any alternative that is superior on this score. Cartesianism may promise more in the way of grounds for fast change, but those promises have not been kept, and cannot be. And, in exchange for those empty promises, Cartesianism has exacted a tremendous price in inequality and separation. Overall, then, I believe that we are better off seeing truth in this contextual way, and recognizing contexts as cultural constructions, even if that leads to a certain amount of pessimism about the rate of change. We cannot insulate ourselves from all tragedy, and the desire to do so—though understandable—may itself lead to the loss of many valuable aspects of human life.44 Knowledge making, then, is similar to, even continuous with, political action. The focus of the relational model of truth on knowledge making as active, as political, and as involving moral responsibility is the key to escaping the idea that our contexts are always threatening either to disintegrate or explode. Obviously, this model of truth does not guarantee that we will have a context for any given issue that is both sufficient for joint decision making and open to deep criticism; our efforts may fail rather than succeed in any given instance. But it does guarantee that we will not find ourselves sliding uncontrollably toward one end or the other of this spectrum, trapped either in a context that cannot ground a shared reality or one that cannot sustain deep criticism. Contexts are not so given, and we are not so passive, as that picture would suggest. Thus, the challenge before us is not a conceptual one (i.e., how to imagine our situation so as to avoid this slide), but a moral, political, and practical one: deciding what sort of context would be good and just for the purposes of a particular issue and then taking the action necessary to create it. Seeing the goal in this way does not, of course, generate simple or determinate answers, but it does make it plain that answers are never simply foreclosed by our epistemological situation. Neither relativism nor conventionalism is ever inevitable.45 Truth—understood as either claims on a shared reality or claims to expand the context to include materials for a deep critique—is possible, meaningful, and desirable within this new model.

b. “working” The third function of truth is to provide a basis for arguments about whether a given course of action will “work.” At the heart of this func-

Reconstructing Truth | 123

tion of truth is a focus on the practical impact of our decisions on all the members of the relevant community. This focus is important because it represents an ethical commitment to take seriously the pain and happiness of others. As such, it is a form of respect for their personhood or subjecthood,46 a refusal to allow symbolic goals or procedural commitments to simply override their concerns. I hope it is fairly clear how this meaning of “working,” this function of truth claims, is consistent with, even required by, the relational model of truth offered here. The responsibility that is central to this model requires that we consider and respond to the challenges, moral and political, that others might bring against our decisions. This form of truth claim is precisely that sort of challenge. It insists that, from the perspective of this person or group of people, our decision looks like a failure to respect their personhood. It argues that we have defined the community in too narrow or exclusionary a way if it is not broad enough to include them as morally significant persons. The relational model of truth does not merely allow for the “working” type of truth claim; it places such claims at the center of our attention as raising precisely the issues of moral and political responsibility upon which epistemology rests. The relational model of truth insists that truth cannot simply be a matter of manipulating symbols, but must be based upon a consideration of moral and political obligations within a human community. Both my earlier discussion of the connection between epistemology and politics and this understanding of the truth function of “working” may raise similar concerns for some readers. Allowing—even demanding—that we subject our epistemological processes and their resulting truth claims to moral and political criticism may seem to open the door to forms of thought control that are truly terrifying. It suggests that one could criticize or refuse to fund or perhaps even punish criminally one who dared to make a truth claim that violated a deeply held moral or political commitment of the community. It is political correctness with a vengeance.47 This is a danger that we must take very seriously in light of the history of attempts by governments and powerful groups to control truth for political ends. My disagreement is not with the claim that such a danger exists, nor with the admonition that we must forcefully resist it, but with the assumption that a commitment to a Cartesian epistemology is either a necessary or a sufficient foundation for such resistance. First, I will argue that Cartesianism cannot provide a guarantee against such tyranny

124 | Reconstructing Truth

and, indeed, that it has as often been part of the problem as part of the solution. Second, I will argue that social constructionism—although certainly it provides no guarantee either—offers tools with which to resist such thought-control efforts. The general conclusion on which both arguments converge is that our only real protection against this danger lies in our moral and political commitments themselves and that an epistemology that guides us to address those issues directly and seriously is a better defense than one that distracts us from them. Even the most forceful critics of social constructionism generally do not claim that Cartesianism is a guarantee against efforts to control “truth” for political ends. Particularly if one includes the milder versions of such control—i.e., informal censure, tenure denials, funding choices— it is plain that within academic and government institutions outspokenly committed to a Cartesian view of objective truth, such efforts are commonplace. Even the most extreme forms—criminal punishment and civil liability—are not unheard of in societies that embrace an objective notion of truth.48 The claim is not that a commitment to Cartesianism will always forestall such thought control, but that it may do so and sometimes has. My response to this is twofold. First, I simply wish to agree. Cartesianism—understood most simply as a commitment to objective truth independent of political and moral values—is sometimes helpful in resisting this danger. I do not wish to deny the courage of those who have resisted such abuses in the name of truth, nor do I wish to deny the power of such resistance to thwart efforts at thought control. I do wish to point out, however, that Cartesianism also has a long history of acting as the tool of such repression rather than as a basis for resistance to it. The history of research on sex differences that has been shaped by and used as a justification for gender inequality is a prime example of the role of Cartesianism in enforcing moral and political values in scholarship.49 Challenges to the existing hierarchy have traditionally been met by the claim that the hierarchy is simply a reflection of an underlying objective reality of differences. Cartesianism has sometimes been part of the solution, but it has often been part of the problem too. Indeed, the reason that Cartesianism has been part of the problem is that it has facilitated the denial that moral and political values were at work in such research. Cartesianism has allowed scholars to hide behind neutral explanations that did not require them to acknowledge and defend their values openly. A commitment to Cartesianism does not empty

Reconstructing Truth | 125

truth seeking of its moral and political content—nothing can do that— but it does make that content harder to see and therefore more difficult to challenge. The reason Cartesianism has sometimes been helpful is that it makes such moral and political content—once exposed—grounds for holding the belief based on it to be presumptively invalid. It discredits knowledge based on such commitments. But such a strategy is much more often, and more effectively, used against those who would challenge the underlying moral commitments of their society—and whose “biases” therefore stand out in bold relief—than against those whose work assumes the accepted view and therefore appears neutral. Thus, Cartesianism’s capacity for criticism—which is real—has an inherently conservative bias. Not only is Cartesianism insufficient as a basis for resistance to the danger of thought control; it is also unnecessary. As I will argue in detail in chapter 7, the social constructionist model of truth provides a foundation for the protection of freedom of speech. The functions of truth in a relational model are served in important ways by the social practices of speech. In particular, speech is a primary—perhaps a necessary—mechanism for practicing the respect, responsibility, and reason that are central to the search for relational truth. There are strong reasons why, if we care about truth understood in a social constructionist way, we would protect speech. An opponent of social constructionism might reply that this epistemology has also been misused to support programs of thought control and can, therefore, claim no advantage over my description of Cartesianism. Certainly, if language creates reality, it should be no surprise when governments claim that they need broad powers over speech in order to ensure the moral and just society whose creation and maintenance was their purpose for existing. Once again, I must agree: social constructionism, like Cartesianism, offers both the possibility of repression and the possibility for resistance to repression. If the argument got me no further than this sort of moral equivalence, I would be satisfied. If proponents of Cartesianism would approach the claims of social constructionists without constantly raising the specter of totalitarianism, that would improve the dialogue in a meaningful way. In fact, however, I do not think there is a moral equivalence. Once one recognizes that both epistemologies are merely tools that can be used either to support or resist oppressive policies, the question immediately arises: what determines in which way they will be used? The answer is, of

126 | Reconstructing Truth

course, the political and moral commitments of the people using them. In other words, the only ground on which we can ultimately rely in efforts of resistance is composed of moral and political values.50 This does not mean, however, that all epistemologies are equal. If our ultimate assurance lies in our values, then an epistemology that directs our attention away from those values is less useful than one that makes those values a central concern. We must recognize when those values are at stake and rally in a clear and self-conscious way to defend them. Social constructionism, by highlighting rather than hiding the role of moral and political value judgments, facilitates such recognition and concerted action. This model of truth is, therefore, better protection against the evils that so justifiably worry some Cartesians.

c. connection to nonhuman reality Finally, the fourth function of truth that I discussed in the previous section is as a basis for a connection to a stable, nonhuman reality in which human beings are simply one constituent part. The truth claims that fulfill this function may often take a form that many would consider religious: claims about God or about the harmony and moral standing of the natural world. But religious claims—conventionally understood—are only a subset of the category I have in mind here. Truth claims based on a secular environmental ethic, or on a commitment to the rights of animals, could also serve the same function. Any claim that positions human beings in a reality that they alone do not control, a reality shared with other moral agents or entities, serves this function. This function poses the greatest challenge for a social constructionist vision of truth. In the relational model, truth is a human creation, deeply shaped by culture and personal context. A large part of the struggle to escape Cartesianism consists of the refusal to associate truth with a transcendent reality, at least if “transcendent” means “beyond human culture.” Social constructionism cannot take a position on what, if anything, is “really out there.” Indeed, from a social constructionist view, this is a meaningless question, since all knowledge making goes on only “in here”—within human cultures. The fact that the question is meaningless, that we cannot ever make knowledge claims about such a transcendent realm, does not mean, however, that a social constructionist view must assert that nothing exists except human beings and their experiences. Indeed, such an assertion would be simply another form of forbidden claim about what is (or is not) “re-

Reconstructing Truth | 127

ally out there.” The claim must be, instead, that all knowledge is relational in the sense that it is about the relationship between the human knower and whatever is known. This does not deny the existence of the thing known, but it does deny that knowledge, as we practice it, is directly about that. Thus, the social constructionist view does not deny the reality of nonhuman nature. Indeed, in one sense it is very respectful of nature by positing it as a partner in the knowledge-making process, a process that has profound effects on the life and identity of the human knower. But the social constructionist view does imply that what we know of nonhuman nature (and indeed of another human being as well) is always and only our own relationship to that subject. We can alter and multiply our perspectives, but we can never step outside of our perspective entirely. Thus, a social constructionist view does not guarantee that any particular claim about the human relationship to nonhuman nature is true, but it recognizes such categories of claims as legitimate. In other words, social constructionism does not require agnosticism about the issue of whether or not nonhuman entities exist. Quite the contrary, from a social constructionist perspective it is the skeptical claim that nothing else might exist outside one’s own mind that is foreclosed. It is foreclosed because it relies upon a suprahuman, supracultural notion of “what is out there” that social constructionism makes untenable. From a social constructionist point of view, nonhuman reality certainly exists, because all existence does or could mean is defined from within a human perspective. A claim that something exists is a claim that we stand in a certain kind of relationship to it and, as such, the claim may be true or false, but it is certainly legitimate and comprehensible. It does not become impossible to talk about our role in some reality large enough to include nonhuman moral elements just because we have adopted a social constructionist view of truth. We can speak about this subject only from within a human perspective, of course, but that does not mean we are silenced.51 The relational view of truth that I am proposing leaves room for this function of truth. Some readers may be concerned about the application of this approach to the subset of these claims that would conventionally be characterized as religious.52 A social constructionist view cannot accept the claim that we ever have unmediated access to a transcendent reality, and if that is the meaning of revelation and other sorts of mystical experiences, then they cannot be accommodated in this view. Such unmediated access is not,

128 | Reconstructing Truth

however, the only way to understand mystical experiences. One might see them, instead, as mediated, but by a different sort of framework than the usual epistemological ones—a framework that, while still deeply culturally shaped, gives freer rein to capacities of mind and spirit that are normally more restrained. Indeed, one common aspect of religion may be to provide such an alternative framework to facilitate such experiences.53 This understanding of religious experiences includes no particular limits on their content. This understanding also suggests that truth claims based on such experiences are no more or less grounded than the truth claims produced within other sorts of culturally specific frameworks. Our cultures give us a variety of lenses through which to operate, and each set of lenses opens up a different view to us. One such set of lenses might be called spiritual or religious. It is no less culturally generated, and no less able to ground truth claims, than any other set. What do we do, then, when truth claims generated from within such a religious framework conflict with truth claims generated from one of our more mundane sets of lenses? This is, I think, one of the foundational questions that any social organization must address. There is a variety of possible answers, ranging from theocracy to enforced secularism. Our Constitution has attempted to avoid both of these extremes. It requires the recognition, even accommodation, of religious claims to truth, but also limits their role in public policy.54 The exact implications of the First Amendment for the practical difficulties raised by such a balancing act are, of course, quite controversial.55 We must and should continue to argue about how and why certain frameworks can contribute to or be excluded from our joint decision-making processes. The important point is that the social constructionist view of truth does not in any way resolve such issues in advance. Seeing religious or spiritual frameworks as comparable to, although distinct from, other sorts of frameworks does not determine what role, if any, they should play. They have as much claim, and as little, as any other framework to create true knowledge. And they must be judged, for the purposes of inclusion in joint decision making, by the same ethical and political standards that guide the choice between other frameworks.56 Thus, both religious truth claims and the larger category of truth claims directed to our role in nonhuman nature are accommodated by the relational model of truth.

Reconstructing Truth | 129

C. Conclusion There is an alternative model of truth that avoids the difficulties of Cartesianism. This relational model relies upon and elucidates common feminist themes, and it serves the primary functions of truth claims. It achieves these goals by focusing on the contextual and relational nature of truth and on the moral and political responsibility of the knower, a responsibility that is grounded in her active, interpretive role. The social constructionist model of truth can provide a basis for both a shared reality and a deep critique. This model explains and justifies the function of truth as a claim to pay attention to the practical concerns of others. And this model allows for truth claims that address the human relationship to a nonhuman reality. As I will discuss in chapter 7, the relational model also provides a basis for the protection of freedom of speech.

6 Reconstructing Autonomy

The feminist critique of liberal autonomy explored in chapter 3 argued that the traditional model of autonomy is both conceptually flawed because of its overemphasis on individualism and morally suspect because of its role in gender hierarchy. In chapter 4 I suggested that, nonetheless, we cannot afford to allow our disappointment with this model to lead us to abandon the notion of autonomy altogether. In order to reconstruct autonomy, we need a much clearer sense of the work that a concept of autonomy must do. As with the discussion of truth in the previous chapter, I am attempting here a form of cultural analysis rather than traditional philosophy. What is the value to us of the practices through which we ascribe autonomy, to ourselves or others? This question about function is both descriptive and normative: it asks about the value of our current practices that we might wish to preserve. The goal is to set some standards by which to measure the success of an alternative model of autonomy. The new model should allow us to continue to use autonomy in these valuable ways while avoiding the problems of individualism and gender hierarchy inherent in the old model. This chapter will argue that there are at least four significant functions of the concept of autonomy. First, autonomy provides a foundation for a collection of important orientations towards ourselves, including self-respect, self-trust, and self-esteem. Thinking of ourselves as autonomous is necessary to feeling these ways about ourselves. Second, autonomy allows us to understand both ourselves and others in terms of a model of character that allows for the possibility of integrity. We are not disconnected collections of desires and actions, but persons with a consistent character structure over time, for whom issues of integrity and the moral evaluation of character make sense. Third, autonomy provides the foundation for our moral accountability to others in a community of shared

130

Reconstructing Autonomy | 131

norms. Autonomy is crucial to our understanding of individual responsibility, of the appropriateness and meaning of the social practices of praise and blame. Finally, autonomy is necessary to our capacity to challenge and resist social norms and conventions. Autonomy describes the locus of our individual agency and denies that we are simply the passive recipients of social forces beyond our control. These functions are legitimate and important. They may have been served in the past primarily through the problematic means of the liberal model of autonomy, but they are still significant for us. After I have described each function in detail, I will outline an alternative conception of autonomy and explore the ways in which it might fulfill these functions, while avoiding the difficulties of the liberal model.

A. The Functions of Autonomy In describing these functions, I am using a broad and necessarily vague notion of autonomy. Autonomy in this discussion simply means some form of self-direction and/or self-ownership. I realize that this definition is extremely imprecise. Such imprecision is, however, required by the nature of the project. The idea of autonomy used to outline the functions must be as broad and general as possible so as not to beg the question of what the ultimate model of autonomy should be. If I began by asking about the functions of a very specific model of autonomy, it would then be worth little to conclude that the same model served those functions. I must ask about the functions of the broad and general idea of autonomy and then show that the particular model I propose can serve those functions. As a result, I must rely on a general and imprecise idea of autonomy that would cover a broad range of models when I am describing the functions of the concept. The purpose of this broad idea of autonomy is simply to help us identify the social practices whose functions we will examine, while leaving open the possibility of a variety of more precise models.

1. Self-Respect, Self-Trust, and Self-Esteem In chapter 3 I argued that the exercise of autonomy depends upon the possession of a minimum of self-trust and self-esteem. The converse is also true: our conceptions of self-esteem, self-trust, and also self-respect

132 | Reconstructing Autonomy

depend upon and incorporate the assumption that we are fundamentally autonomous. This is not a logical necessity: perhaps other bases for selfesteem, self-trust, and self-respect are possible. But our present understanding of these important orientations towards oneself includes this reliance on autonomy. One of the functions of autonomy in our moral lives is, then, to support our self-esteem, self-trust, and self-respect. First, it is useful to define these interrelated concepts. Self-trust is the ability and inclination to rely upon oneself, even if it leads to vulnerability. This inclination requires that one see oneself as competent and efficacious in the world. Self-respect is the belief—some would say the justified belief—that one is worthwhile as a person and the disposition to act on that belief. This understanding of self-respect would include resisting violations of one’s human rights, being committed to one’s own projects and values, and maintaining one’s personal standards.1 Some theorists would end there, seeing self-respect as limited to a sense of worth based on personhood. Others would include also the sense of self-worth based on particular individual characteristics or accomplishments.2 The meaning of self-esteem seems to depend to some extent on which definition of self-respect one embraces. Those theorists who embrace only the more general, personhood-based version of self-respect might use selfesteem to describe the more individualized sense of self-worth. Self-esteem and self-respect would then be related but distinct concepts. For those theorists who embrace the more inclusive notion of self-respect, self-esteem might be a particular form of self-respect, the form based on more individualized assessment rather than on personhood generally. Finally, some theorists describe self-esteem very broadly as a sense of one’s own worth, on whatever basis.3 From this perspective, self-respect is a particular form of self-esteem. Regardless of terminology, we have at least two forms of self-worth at issue: a form based on one’s status as a person and all that such status entails, and a form based on a more individualized assessment of one’s own character and actions. Autonomy is important to both of these forms of self-worth and to self-trust. Before proceeding to assess the connections between these concepts and autonomy, I should note that all of these attitudes towards oneself are deeply shaped by social structures and institutions, such as patriarchy or a sex-segregated workforce. A person’s ability to trust herself, to respect herself as a person, or to experience self-esteem based on her characteristics or accomplishments is shaped by such social circumstances.4 Moreover, one of the effects of social hierarchies, like sexism or racism, is to

Reconstructing Autonomy | 133

systematically impair the ability of those on the bottom to develop these attitudes towards themselves and others in their group.5 The following analysis, in focusing on autonomy, is not intended to obscure the dependence of self-trust and self-esteem on social conditions. The argument here establishes the need for a notion of autonomy to support these valuable orientations towards oneself, but that notion need not be—indeed, cannot be—disconnected from social context. Self-trust requires one to believe in one’s own competency and be willing to rely upon it, even when such reliance leads to vulnerability. This relation to oneself rests on a notion of autonomy, understood as self-direction. If one experienced oneself as the unwilling and helpless pawn of forces beyond one’s control, without any capacity to direct oneself, then it would be impossible to trust oneself.6 We all make mistakes and we all fail sometimes, and self-trust does not and should not require us to believe ourselves infallible. But without some minimum sense of self-direction we cannot trust ourselves. Autonomy is also central to self-respect, both the form based on personhood and the form based on an evaluation of one’s individual character and actions. What does it mean to respect oneself as a person? The usual answers to this question focus on avoiding servility, resisting violations of one’s rights, and believing in the worth of one’s self and one’s plan of life.7 Thus, self-respect includes a variety of beliefs, emotional reactions, and dispositions to action.8 Robin Dillon has described in detail some of the elements of this form of self-respect based on personhood. Dillon argues that this “recognition self-respect” includes three elements. First, there is “interpersonal recognition self-respect,” which refers to one’s standing in the moral community as a person.9 This form of self-respect explains the need to resist violations of one’s rights that might lead to or represent an undermining of that standing. Second, there is “agentic recognition self-respect,” which refers to appropriately valuing one’s capacity for agency. This, in turn, includes several elements, including “developing, exercising, and protecting one’s capacity for agency by committing oneself to values worthy of one’s commitment,” “striving to be autonomous,” and “fulfilling responsibilities one has as a person and taking responsibility for oneself.”10 And finally, there is “personal recognition self-respect,” which requires that one have a personal ideal and live in accord with it.11 All three aspects of this form of self-respect rely on some notion of autonomy. Autonomy is directly necessary to the second aspect—agentic

134 | Reconstructing Autonomy

self-respect—because it is the foundation of our sense of agency. It is the ability to direct oneself—both at the level of committing oneself to values and at the level of acting on those commitments—on which our sense of individual agency rests. Whether one conceives of such commitments as a matter of choice or not, one must see them as authentically belonging to the self if the self is to be understood as an agent and not merely as a mechanism transmitting the power of other agents (such as social forces). The other two aspects of self-respect based on personhood also rely on an assumption of autonomy, although somewhat less directly. Interpersonal recognition self-respect requires one to stand in the community as a person among persons. But if one of the aspects of personhood is agency and if agency is dependent on some notion of autonomy, as the discussion above indicates, then part of being recognized as a person by others means being recognized as autonomous. Similarly, personal recognition self-respect requires that one have an ideal and live in accord with it. But both the construction of such an ideal and the activity of conforming one’s actions to it require the ability to direct oneself. Our sense of our place in a moral community and our sense of the pattern of our own moral lives are both dependent upon our sense of ourselves as autonomous agents. The second form of self-worth, in which we value ourselves because of our particular characters or actions, is also dependent upon a notion of autonomy. Evaluative self-respect, as Dillon calls it, is a matter of “earned worth, a positive appraisal of one’s quality as a person in light of the standards given in one’s self-ideal.”12 Because one needs to have a personal ideal and live up to it in order to have this sort of self-respect, and because one needs some notion of autonomy to make sense of the idea of holding an ideal as your own and living up to it, autonomy is necessary to evaluative self-respect as well. There is a collection of positive attitudes towards oneself—such as selftrust, self-respect, and self-esteem—that we find to be not simply useful, but necessary to a meaningful human life.13 But we have found that, in order to hold such attitudes, we must adopt a certain perspective towards ourselves: a perspective in which we appear as autonomous agents. So, whatever our view of the world and the issues of causal determinism when we adopt the perspective of an “observer,”14 we must also have available to us the perspective of the autonomous, that is, self-directing, agent.15 Allowing us access to this perspective is one of the primary and

Reconstructing Autonomy | 135

important functions of the concept of autonomy that any acceptable conception of autonomy must fulfill.

2. Character and Integrity The perspective of the autonomous agent is also crucial to our understanding of character and of the possibility of integrity. Character is a collection of personality traits, attitudes, and values held by an individual that is relatively stable over time. Character ties a life together, allowing us to see ourselves and others as more than simply arbitrary collections of behaviors and experiences. A sense of autonomy—understood as being the author of one’s own acts—is part of the foundation for character. Owen Flanagan offers a description of soldiers suffering from an identity crisis as observed by Erik Erikson: [T]hey normally experience themselves as the locus of a set of subjectively linked events, as a sort of conduit in which a certain bland and low-level sameness and continuity subsists. What they lack . . . is any sense of coherent and authoritative “me-ness,” of personal sameness— any sense that these subjectively linked events occurring to and in them constitute a person, a self, a life.16

For these soldiers, the problem is not that they cannot tell whether an act or emotion happened to them or to someone else: there is not, in other words, a complete breakdown of the boundaries of personal identity. The problem is that they cannot see why the simple fact of the location of that act or emotion within them makes it theirs in any meaningful sense. They are missing a sense of authorship or autonomy. This sense of autonomous agency is the glue that holds together the disparate elements of a life into a single person with a coherent character.17 Perhaps the connection is not logically necessary: perhaps one could have a different conception of character in which something else serves this function of gluing together the aspects of a life. But in our present moral universe, autonomy is one of the primary means through which we construct sufficient coherence to sustain a sense of character. For that reason, a new model of autonomy should ideally serve this function as well. Moreover, autonomy is also one of the important means through which we understand the possibility of change in our characters. Character change can, of course, come about through processes that are not

136 | Reconstructing Autonomy

autonomous; for example, someone subjected to a long period of sustained cruelty might become timid and passive. But we also understand our character as subject to change by ourselves. Changing one’s character is generally a gradual process, piecemeal and time consuming, often difficult; but we experience it as possible. “To view oneself as incapable of acting on this understanding [that one can change one’s own character] is to take a perspective on oneself that is ultimately incompatible with seeing action from the inside.”18 The concept of integrity is closely related to the issue of character. Integrity is traditionally understood as referring to the consistency and coherence of one’s character, including both one’s actions and one’s values. This idea of integrity is primarily formal in nature: it concerns the consistency of one’s behavior with one’s values (and the internal consistency of those values) rather than the substance or source of those values.19 One might question the significance of this formal and rationalistic conception of integrity,20 but whatever one thinks of it, it is clearly dependent upon an idea of autonomous agency and the value of that agency. Many feminists have, in fact, criticized this formal conception of integrity. They point out that many people find themselves in contexts in which they are subject to conflicting systems of norms and responsibilities and/or they experience fragmented or “intersectional” identities.21 Sometimes this experience is due to conditions of oppression, in which their identity as members of a socially oppressed group means that they are expected to follow norms that are in tension with the more general social norms: for example, when a woman is expected to subordinate her career to her husband’s in violation of a general social norm to seek success. Sometimes this fragmentation is simply the result of membership in overlapping but distinct moral communities, for example, when a person is both a lawyer and a member of a religious organization that prizes cooperative and nondominating methods over competitive and adversarial ones. In such situations, being true to oneself often requires not coherence and logical consistency, but a tolerance for tension and ambivalence. To have integrity here means neither ignoring the conflict nor (necessarily) suppressing one aspect of one’s identity, but living creatively with the tensions between them.22 In a later section of this chapter, I will consider alternative understandings of integrity and the ways in which a new model of autonomy might contribute to and support those forms of integrity. For the purposes of the present argument, however, the important point is that our current

Reconstructing Autonomy | 137

understandings of character and integrity rest in important ways on a conception of autonomy. It may well be the case that a new conception of autonomy would lead us to new understandings of character and integrity—indeed, I shall argue that this is so—but that simply emphasizes our need for a conception of autonomy to ground our conceptions of character and integrity. There is a close connection between issues of character and issues of responsibility. In many ways, character and responsibility are mutually self-defining. First, “judgments of character are parasitic upon judgments of responsibility. . . . Only those actions for which one is responsible can be taken to reveal one’s character.”23 Moreover, as we broaden and narrow the scope of responsibility, we alter the boundaries of identity itself: who I am is in part determined by the slice of the world for which I am responsible. The extent to which my identity is bound up with other parts of my reality—including other persons, individually or as social groups— is marked for me, in part, by the extent to which I am understood, and understand myself, to be responsible for them.24 Finally, responsibility is generally understood to rest, in part, on considerations of character. A judgment that a person is responsible for an act is often based upon the sense that the act reflects something about the agent’s character.25 Thus, the consideration of character leads naturally to the question about the relationship between autonomy and ascriptions of responsibility.

3. Personal Responsibility There is an enormous philosophical literature on responsibility. The primary focus of this literature is the nature and possibility of the type of freedom that is necessary to ground ascriptions of responsibility. In particular, philosophers argue about whether or not people need to be able to have done otherwise (referred to sometimes simply as free will, at other times more specifically as contra-causal freedom) in order to be held responsible for their acts.26 The major difficulty is the causal determinism inherent in a scientific view of reality. The question is whether such determinism undermines the foundations of responsibility.27 The first requirement in discussing responsibility is, therefore, to explain the relationship between autonomy and free will. Sometimes the terms are used interchangeably, such that autonomy is understood to be simply the capacity for free will, or the capacity for contra-causal freedom. On this understanding, autonomy would be an all-or-nothing

138 | Reconstructing Autonomy

affair—either one has this capacity or one does not—and would be true of all normal human beings.28 On other occasions, autonomy is used to describe the mechanism or process that allows for such freedom: for example, a process of rational reflection.29 On this understanding, autonomy might well admit of degrees—one could engage in this process to a greater or lesser extent—and might vary both across persons and for a single person over time (i.e., one might learn to be autonomous). I will use autonomy to refer to the sense that a choice or action—or one’s whole life—is one’s own.30 In this usage, autonomy is only indirectly related to free will or contra-causal freedom. This sort of free will might be one way of explaining why the choice or act is one’s own, but it is in no way required by this conception of autonomy: there are other ways of understanding such ownership that can also fulfill the functions served by the concept of autonomy. By using the word in this more general way, I hope to be able to explore the connections between autonomy and responsibility without having to engage the entire free will/determinism debate. There is a general consensus that responsibility rests on autonomy. For those who see autonomy as connected to rational reflection on one’s firstorder desires, responsibility is predicated on the capacity for (and perhaps the actual exercise of) autonomy.31 As Marcia Homiak has explained in her analysis of Aristotle, The ability to act from decision thus seems to be what distinguishes morally responsible agents, whose actions are appropriate candidates for moral praise and blame, from those who are not morally responsible. . . . [O]nly beings who have some control over who they are and over what they can do can fruitfully undertake the deliberation involved in forming decisions. Were a being simply the passive subject of her desires, such deliberation would have no useful practical role in her life.32

Thus, our autonomy grounds our responsibility. Because the connection between autonomy and responsibility is so widely accepted, even by those who disagree about the precise meaning of autonomy, it is rarely argued as a general matter. Instead, the connection is generally presented in order to demonstrate the advantages of a particular conception of autonomy in supporting our ascriptions of responsibility. In order to describe more precisely the relationships between responsibility and the general conception of autonomy (as having one’s

Reconstructing Autonomy | 139

choices, acts, life be one’s own), it is helpful to examine the functions of ascriptions of responsibility and the ways in which this general conception of autonomy supports them. Not every function will be involved in every ascription of responsibility, but the following list suggests some of the major purposes served by our practice of ascribing responsibility to people.33 In describing these purposes, I am using a fundamentally relational model of responsibility. As Paul Benson has argued, “being morally responsible involves being worthy of a certain social standing, that of an eligible participant in various kinds of moral exchange, such as offering reasons, seeking excuse, begging forgiveness, and so forth.”34 Rather than being understood as a characteristic of individuals, responsibility is seen as a relationship between people. Responsibility is relational in several different senses. First, responsibility is causally relational because we learn the practice from our relationships with others. Second, it is substantively relational because the norms that comprise the standards of evaluation must be “publicly shareable.” And third, responsibility is conceptually relational because it depends on there being others who could hold us accountable.35 Within this relational understanding of responsibility as a social practice, perhaps the most obvious purpose of ascribing responsibility is to change people’s behavior in the future. The person held responsible for a past bad act will, we hope, reflect and decide differently the next time such a situation arises.36 Moreover, the practice of assigning responsibility, of praising and blaming, shapes the behavior of other actors who are aware of it as well. It establishes social standards for behavior and invites people into the relationships in which those standards are created and maintained.37 Thus, at least one purpose of ascribing responsibility is to affect the future behavior of members of the community. One need not adopt a contra-causal notion of freedom to make sense of this purpose, but one does need a basic concept of autonomy. It may not matter whether the agent could have done differently in the past, in the act for which she is being judged responsible; but it certainly does matter that she has some capacity for agency, for decision making, for self-direction in response to the community’s norms and practices in order for the ascription of responsibility to affect her future behavior. We must be the authors of our actions in a meaningful way in order for this practice to make sense as a mechanism for affecting our future behavior.38

140 | Reconstructing Autonomy

Another purpose of the practices of blaming and praising is to establish and maintain the relationships of trust on which the community is based. We must be able to trust each other, in general, not to violate the shared norms of our community. The practice of ascribing responsibility allows the victim of a violation of that trust to register her hurt and outrage and to assert the fundamental relationship of obligation between herself and the agent of that violation. When such practices work, “they strengthen the common fabric of trust in people’s senses of responsibility.”39 As I argued above with respect to self-trust, however, trust requires a sense of the autonomy of the one trusted. Without a minimum degree of self-direction, an agent cannot be trusted. We might well sometimes feel confident in predicting or relying on certain future behaviors because we believe them to be so likely to occur (e.g., if I tickle my child, she will laugh), but we would not describe our confidence in such situations as trust in the person involved. Trust is a distinctive attitude towards another that requires more than merely the assurance that his or her behavior is causally reliable. It is part of the agent orientation described in relation to self-respect and self-trust, but, in this case, that orientation is directed toward someone other than oneself. Obviously, when applied to someone other than oneself, trust includes elements in addition to autonomy. Even if someone is autonomous, we still might not trust her if she is also malicious. Autonomy is not sufficient for trust, but it is necessary. Ascriptions of responsibility also function to mark the boundaries of the relevant moral community. There is an important sense in which someone who cannot be held responsible cannot be a full member of the community. At its strongest, this claim suggests that “to exempt someone from responsibility and blame for wrongdoing is simply to deny that person’s humanity.”40 Even if the person is not totally demoted to the status of an object, however, she is certainly not regarded as standing in the reciprocal relationship that is the foundation for the community.41 This is why it is sensible to claim that the ascription of responsibility is less a factual judgment about the agent, and more a performative act creating or reasserting a certain reciprocal relationship between the blamer and the blamed. The relationship between responsibility and autonomy is, of course, dialectical. Just as ascriptions of responsibility depend upon a conception of autonomy, so also a sense of one’s autonomy depends, in part, upon

Reconstructing Autonomy | 141

the practices of ascribing responsibility. As Stephen White has pointed out, we should ask ourselves “what would be lost to a person for whom, or a society for which, responsibility and blame were unintelligible?” He suggests that such a person or society would lose both the sense of authorship of action (the general idea of autonomy I have been using) and the sense that persons have a sphere of authority in which their own decisions count (a practical application of the idea of autonomy).42 The latter connection—to an area of decisional authority for the person—highlights the interrelationship of autonomy and responsibility with rights. The dominant understanding of rights in the legal literature is as protection for such an area of decisional authority for individuals.43 The standard justification for such protection relies upon the value of the individual autonomy that is thereby recognized and promoted.44 The former connection drawn by White—to the way in which responsibility supports a conception of autonomy as authorship—can be fleshed out by considering in greater detail the relationship between responsibility and identity. In the discussion of character, I argued that the range of phenomena in the world for which one is held responsible shapes the boundaries of one’s identity. I want to take this argument a step further now and suggest that the process of ascribing responsibility is one of the social practices through which we create and mark the boundaries of a sense of agency: in other words, it is not just the particular content of our identities that is dependent on this process, but the sense of having an identity as an agent at all. First, the experience of choice is dependent upon the social practices, including (but certainly not limited to) the ascription of responsibility, through which we establish limits on choice. If we understood ourselves to be faced with a limitless variety of possibilities, we would be incapacitated from choosing. “Indeed, choice is made possible only by the social relations and conventions that frame a finite set of alternatives.”45 The practice of ascribing responsibility, along with many other social practices, supports the social understandings about the normative and practical limits on our actions that circumscribes our options to a range in which choice becomes possible. Moreover, the issue is not simply the number of options, but their meanings. As Charles Taylor has argued, to define an identity requires a background of intelligibility about what matters. A view of choice in which choice itself confers worth implicitly denies this background and thereby undermines itself: choice becomes insignificant in the absence of

142 | Reconstructing Autonomy

any standards of value according to which one could choose.46 The practice of ascribing responsibility, to the extent that it establishes and maintains these background understandings about value and meaning, helps to provide the foundation for the experience of choice. The practice of ascribing responsibility also generates more specific limits that support the experience of agency. Agency is dependent on a sense of particular identity and that sense of identity is itself a product, in part, of our practices of ascribing responsibility. Agency is dependent on a sense of having a particular identity because, in order to be our own, our actions and choices must issue from that identity. To put this another way, having a character is necessary to the experience of choice. And, as I have argued above, the ascription of responsibility is one of the mechanisms through which we define the boundaries of our character.47 In addition, our practices of ascribing responsibility define the meaning of agency in our culture. Through the notions of excuse, ignorance, coercion, and so on, we define what people are expected to be able to do. To say that someone should not be held responsible for a particular act because, for example, he or she was coerced is to say, in part, that we do not expect people to exercise the type and degree of control that would have been necessary to resist that coercion.48 Thus, our practices of ascribing responsibility in part define the meaning of agency for us; they set our own expectations for ourselves and shape our experience of our own agency.49 Finally, to bring the argument full circle, when we internalize the practice of ascribing responsibility, we acquire the ability to be accountable to ourselves and thereby facilitate the experiences of self-trust, self-esteem, and integrity discussed in the earlier sections of this chapter. It is not just that we learn these standards and processes from specific other people, but that we go on applying them through a process of dialogue with others—real or imaginary—throughout our lives.50 Thus, the social practice of ascribing responsibility allows in a continuing way for the possibility of holding oneself accountable to oneself.51 That possibility, in turn, grounds the experiences of self-trust (I trust myself because I stand in the same reciprocal relationship of accountability to myself that I do to others in my moral community), self-esteem (I value myself because I am the sort of person who can be trusted in this way), and integrity (the process of holding myself accountable allows me to evaluate my actions to see if they are in accordance with my principles and thereby maintain my integrity).

Reconstructing Autonomy | 143

Responsibility and autonomy, then, are deeply implicated in each other. Our practices of ascribing responsibility depend upon certain understandings about the autonomy of responsible agents and, conversely, our understandings about autonomy are deeply shaped by and dependent upon our practices of assigning responsibility. One of the important functions of the concept of autonomy is to support this web of connections between responsibility, character, integrity, self-trust, and so on. Modifying the conception of autonomy will undoubtedly result in modifications in our understanding of all of these related concepts, but any alternative conception of autonomy must be designed to support useful versions of these important aspects of our moral life.

4. Social Change and Democratic Politics In discussing character, I suggested that some notion of self-authorship is necessary to allow for the possibility that we can change our characters. Similarly, self-authorship is an important part of the way we understand the possibility of social change and the meaning of politics. I have already discussed in chapter 4 why feminists need a conception of autonomy to allow them to argue that we can change the patriarchal social conditions under which we find ourselves and to avoid the politically and personally debilitating effects of women seeing ourselves as simply victims of social oppression. Obviously, if we are the helpless products of our social conditioning, we cannot hope to be the conscious agents of social change. Change may come, and it may even come through us, but it will not be under our direction unless we have some authorship of our own actions. We must “account for the control women exert over their lives under patriarchy, for their opposition to subordinating social norms and institutions, and for their capacity to bring about emancipatory change. The concept of autonomy promises to be helpful in these endeavors.”52 In this section, I would like to build on this argument by exploring some of the implications for our understandings of political movement and democratic politics. What would political movement or democratic politics mean in a world without autonomy? I will argue that the meaning and value of democratic politics is integrally related to understanding ourselves as autonomous. Any discussion of political movement in a feminist context must begin by considering consciousness raising. Through this process of sharing and reflecting on their experiences in a group, many women discovered that

144 | Reconstructing Autonomy

the issues and struggles they had assumed were their own individual problems were instead the products of social oppression that affected women more generally. The feminist mantra “the personal is political” is the fruit of this experience.53 The structure of this process relies upon a model of the participants as largely socially conditioned: it is, after all, their status as members of a particular social group that caused them to have similar experiences and even similar identities (personal characteristics, attitudes, and values). Nonetheless, both the process and the results of consciousness raising rely upon the participants understanding themselves as fundamentally autonomous: as the potentially self-directing authors of themselves and their acts. The process of consciousness raising assumes that it is possible for the participants not only to come to realize the impact of socialization on their identities and experiences but also to reimagine their relationships to the social structures at issue and take action based on those new conceptions of themselves and their lives.54 Consciousness raising is a fundamentally pragmatic process: its goal is to give women more control over their lives, not just to make them see that they have less control than they thought. It assumes that greater understanding about the depth and breadth of gender oppression will open up new avenues for self-direction, rather than merely incline the participants to despair. Consciousness raising assumes and relies upon the existence of some degree of personal autonomy. In this respect, consciousness raising is continuous with democratic politics more generally. Our understanding of and commitment to democratic politics rests on an assumption of individual autonomy. In order to see this connection, we need to distinguish, first, between competing conceptions of democracy and, second, between the various aspects of freedom that they serve. In modern political theory, conceptions of democracy are legion, but Iris Marion Young has suggested that we can organize them into two rough categories: aggregative models and discursive or deliberative models.55 Both models share certain features: a reliance on the rule of law, voting as the major decision-making procedure, and a belief in certain freedoms, including freedom of speech.56 The aggregative model, however, sees democracy as “a process of aggregating the preferences of citizens in choosing public officials and policies,”57 while the discursive or deliberative model sees democracy as “a form of practical reason” in which

Reconstructing Autonomy | 145

“open discussion and the exchange of views lead to agreed-upon policies.”58 Both of these major categories of democracy theories rely upon an assumption of the autonomy of citizens to make sense of the value and meaning of democracy. The aggregative model is the one with which most people are most familiar. In this model, we all come to the political process with our preferences in place and we register those preferences through voting. Because we each have one vote, the resulting social policies represent an equal respect for each of us and the best accommodation of our competing preferences. It is possible to give this aggregative model either a utilitarian or a deontological liberal spin, but in either case it relies upon an assumption of autonomy. In its liberal version, the connection to autonomy is direct and clear. As I suggested in the discussion of democracy theories of speech in chapter 1, in order to uncover this connection, one need only ask, why is democracy valuable? The answer is that this form of political system best preserves the freedom of citizens and respects their status as autonomous persons.59 The liberal aggregative model of democracy assumes that “individuals are . . . free agents in the sense of possessing free choice.”60 Indeed, it is the moral significance of their choices that makes it so important that government be designed to respect those choices and to provide the greatest possible freedom for each person to pursue his own vision of the good consistent with an equal freedom for all. Because the value of democracy in the liberal aggregative model is that it promotes and protects the autonomy of its citizens, this ideal of democracy depends directly upon the possibility and importance of autonomy. Perhaps this is an opportune time for a digression into the relationship between freedom and autonomy. Many formulations of democracy theory focus on freedom,61 while I have focused on autonomy in the description above. While closely related, freedom and autonomy are not identical, and their difference will become significant when we move in the next part of the book to the relationship between autonomy and speech. I am using autonomy to refer to self-direction and self-ownership: the sense in which one’s choices, actions, life, belong to oneself. Freedom is best seen as a much broader concept, including potentially both the absence of external constraints on action and the presence of means and opportunities for actions (i.e., both negative and positive aspects of freedom). We sometimes use the word “free” in a way that suggests that one

146 | Reconstructing Autonomy

could be free without being autonomous, as when we consider an act free with reference only to external conditions and without attention to the process of reflection or choice from which it issued. This usage is an artifact of the reliance on a limited form of negative freedom in which “free” fundamentally means “uncoerced.”62 I will, however, use freedom to refer to the entire collection of concepts: autonomy, negative liberty, and positive liberty. Autonomy is necessary for freedom, because no action—however uncoerced or socially supported—could count as free if it were not, in some minimal sense, one’s own. But autonomy is not sufficient for freedom: one whose autonomous choices are frustrated by coercion or lack of opportunity is not free.63 In this usage, it is most accurate to say that the liberal aggregative theories of democracy see democracy as a means of assuring freedom, where autonomy is a necessary (but not always sufficient) aspect of that freedom. The utilitarian version of aggregative democracy is also dependent upon an assumption of autonomy, although in a somewhat less direct way. The utilitarian argument suggests that the purpose of aggregative democracy is to maximize the satisfaction of preferences. When one asks why that goal is valuable, the immediate answer seems to be independent of autonomy: the satisfaction of preferences is valuable because it generates the greatest happiness for the greatest number. It looks as though it should not matter whether the preferences are formed through an autonomous process or not, as long as their satisfaction brings happiness to their holder.64 But, in fact, the process of maximizing happiness through satisfying preferences generally rests on an assumption that the citizens enjoy at least a minimal level of autonomy. If all preferences were understood as simply the creation of social forces beyond an individual’s control, then there would be no reason to believe that the satisfaction of our existing preferences would be most likely to lead to our happiness. It is at least as likely that the rearrangement of our social conditions to generate different preferences might lead to greater happiness in the long run. It is because the process of acquiring our preferences is understood to belong to us—and that process, moreover, is seen as valuable to us—that the prospect of replacing those preferences with others, not of our choosing, leads to a reduction in happiness. In other words, in order to be the basis for our happiness, our preferences must be seen as, to some minimal degree, our own. Some notion of autonomy is implicit in this utilitarian argument.65

Reconstructing Autonomy | 147

Finally, discursive or deliberative models of democracy also rely upon the autonomy of citizens to make sense of their commitment to democracy. These models often begin by expanding the conception of freedom that grounds democracy in the liberal models. Carol Gould, for example, argues that the liberal model is too restrictive in its view of freedom as essentially negative liberty. She suggests a model that includes “capacity for choice and its exercise [autonomy], absence of constraining conditions [negative liberty], and the availability of means [positive liberty].”66 The “self-development” that is constituted by these multiple requirements is the foundation for democracy, understood as “an equal right to participate in determining the course of . . . common activities.”67 Similarly, Iris Young sees democracy as fundamentally about promoting justice.68 Justice, in turn, involves self-determination: “being able to participate in determining one’s action and the condition of one’s action.”69 Self-determination is, of course, a form of autonomy, although in Young, as in many deliberative models, it is a more relational and less liberal form of autonomy. Since democracy is about promoting justice, and justice consists of the promotion of self-determination (as well as self-development), “[d]emocracy . . . is entailed by self-determination, though the value of self-determination does not reduce to democratic participation.”70 The connection between democracy and autonomy should be unsurprising. The meaning of political action generally and democratic politics in particular is dependent on our understanding of ourselves as in some sense self-directing: this is what democratic politics is all about. One of the major functions of our concept of autonomy is to provide us with this support for our understanding of democratic politics. Conversely, autonomy is itself dependent upon democratic politics. Politics may be one of the most important realms in which we experience ourselves as autonomous and in which we feel the need to have that autonomy recognized and reflected in our institutions and cultural forms. Without a democratic politics to back it up, any form of autonomy will be insufficient and ultimately ineffective. Thus, autonomy and democracy are interdependent. Any workable form of autonomy theory will need to support this web of commitments. Autonomy, then, serves several important functions in our moral, social, and political life. Autonomy helps to ground our senses of self-trust, self-respect, and self-esteem; it provides a foundation for our conception of character and integrity; and it mutually reinforces our understandings

148 | Reconstructing Autonomy

of responsibility and of democratic politics. Some of the limits and difficulties of our present approaches to these concepts are due to the limitations of the liberal model of autonomy that has dominated our theorizing. As a result, we will not necessarily want to maintain the connections between autonomy and each of these other categories in exactly the way they have developed thus far. Nonetheless, any alternative conception of autonomy ought to be able to make sense of our experiences of self-trust, character, responsibility, and democracy and, hopefully, offer us new insights about the relationship and meaning of these categories.

B. A Narrative Model of Autonomy This section will sketch the outlines of a narrative model of autonomy that can serve the functions described above while avoiding the problems associated with the liberal model. A variety of theorists, both feminists and others, have been developing the elements of relational models of autonomy, and this section will draw heavily on their work. As with truth, I will build the picture of this model in layers, beginning with a brief description and then moving on to how it incorporates the relational approach of the feminist critique; next explaining how it fulfills the functions identified in this chapter; and finally connecting it to feminist themes more broadly.

1. A Brief Description The narrative process is one in which a person “orders a sequence of events [or, I would add, people or things or concepts] for the purpose of revealing or creating meaning.”71 In order for that process to qualify as an exercise in autonomy, the meaning at issue must be personal: it must relate to the meaning of one’s own life. This does not require, of course, that such a meaning must concern nothing but oneself: the meaning of one’s life could be fundamentally connected to other people, or to God or a religious tradition, for example. But the story must be personal in that it must concern one’s own role in these larger entities. Thus, a person would review her “personal history, weighing the particulars of [her] past in terms of more general moral values, and discerning a course of action which expresses a commitment to these particulars.”72 The process thus

Reconstructing Autonomy | 149

involves a shifting of perspectives, in which the general is always understood through and in terms of the particulars of the life at issue. The activity of narrative construction—of interpretation and reinterpretation—begins, of course, from the materials at hand. A person works with her own experiences and with the stories, values, and concepts that are available to her in whatever culture(s) she inhabits. These materials are always, and from the beginning, both given and created. They are given in that they are shaped by forces beyond any individual’s control; they are created in that each new repetition of such cultural and personal artifacts is always a reinterpretation rather than merely a replication. A narrative model of autonomy identifies the central focus of autonomy not as an act of volition, but as an act of interpretation and narration. The primary experience and exercise of autonomy takes place not when we choose, but when we tell our own stories. If the underlying dimension in the liberal model of autonomy is space (because the self is understood in terms of boundaries), then the underlying dimension in the narrative model is time (because narrative exists and is created in time). Our identities and our lives are conceived in this model in narrative terms and the meaning of autonomy is that one’s identity and life are not “the object or medium of someone else’s speech, [but rather] the subject of one’s own.”73 This focus on telling one’s own story does not eliminate either the importance or the experience of choice. It is most often in response to the need to make choices that we engage in the self-conscious versions of telling our own stories. But that is certainly not the only time we feel the need to revise or retell our stories. A dramatic change in our lives, brought about with or without our choice, such as a natural disaster, an illness, or the birth or death of a family member, commonly causes us to return to the conscious process of telling our stories. Moreover, we often turn to this process because some new knowledge has made us feel the need for revisions of the story. Choice, then, is only one of many occasions for the exercise of narrative autonomy. In addition, choice is only one of many outcomes of the exercise of narrative autonomy. When we have told, or retold, our stories, we commonly find that they lead us to certain sorts of choices. But the stories also lead us to certain sorts of knowledge—of ourselves or others—as well as to certain sorts of relationships in the world: relationships that are experienced not always as the product of a choice, but also as discovered or

150 | Reconstructing Autonomy

created through this process of telling one’s story. Choice is unique neither as the occasion nor as the outcome of narrative autonomy. But can we be said to be autonomous if our subsequent choices are not consistent with the story we have told? That is, is choice perhaps not sufficient but necessary as an outcome of narrative autonomy? Again, choice plays an important role here, but not a unique one: our narrative autonomy would be just as much in question if our subsequent self-understandings were inconsistent with the story or if our relationships with the people in our lives contradicted it. The story we tell about ourselves creates implications for the broad range of our actions and reactions— emotional, cognitive, and volitional—and choice holds no privileged position: it is no more significant as an indication of our narrative autonomy than any other element. Indeed, under certain conditions—such as conditions of oppression in which an agent’s range of choices is severely limited—her choices may be a less reliable indicator of her autonomy than her self-understanding or her relationships. In this narrative model, autonomy is neither a preexisting condition to be assumed for all persons,74 nor is it an end state that can be taken for granted once achieved. Instead, it is a process, a process that must be continually ongoing in order for a person to be autonomous. The narrative model of autonomy suggests that autonomy can be achieved to differing degrees and in different areas of one’s life. We may, and do, tell different stories at different times and for different purposes and concerning different aspects of our lives or identities. Narrative autonomy does not require that we tell a single, overarching story.75 We also find it harder or easier to frame such stories and to embrace them as our own under different circumstances. Narrative autonomy, then, could be assessed for a particular time period and/or area of activity; it would rarely be useful to assess it for a person’s life as a whole.76

2. A Relational, Nonliberal Model The model of narrative autonomy is relational in all the senses required by the feminist critique of liberal autonomy. First, the narrative model emphasizes that autonomy is substantively relational. When autonomy is understood in terms of stories, then autonomy is dependent on social relations with other persons for its substance because the very characters and plot lines available to us in constructing our stories are culturally given. Even when we define ourselves in opposition to a certain social

Reconstructing Autonomy | 151

category, we are defining ourselves in relation to it.77 While this cultural “givenness” is true of all of our moral categories (as I argued in chapter 3), the focus on stories makes the cultural dependence clearer because we are so accustomed to the experience of discovering that the possible range of narrative elements, such as plot and character, differ from one culture to another. Such stories and their elements are never static, of course; they are constantly in the process of being reinvented and reinterpreted—indeed, narrative autonomy is one of the processes through which they are reinvented—but the cultural categories available at any given moment provide the materials on which narrative autonomy works. Narrative autonomy is also inherently relational. An exploration of the sense of self inherent in the narrative model illustrates that this model rejects the unified, transparent, and narrowly rationalistic self of the liberal tradition. In addition, the capacities necessary for the exercise of narrative autonomy are themselves essentially relational. First, the self that exercises narrative autonomy is not necessarily unified. Because our identities are permeated by our relation to a social world, we may find ourselves with overlapping and even contradictory identities if we inhabit a society with overlapping and contradictory roles or communities. Far from being a rare and troubling psychological problem, such complex identities appear both common and extremely productive from the perspective of a narrative model of autonomy. Diana Meyers has argued that such “intersectional” identities are a resource for autonomy: because we all belong to multiple groups, we can use the experience of distance from any one of them to gain critical insight and open possibilities for new interpretations.78 Narrative models also reject the transparent self of liberal theory. Once we recognize the substantively social sources of our stories, it becomes apparent that someone else who shares my culture might be able to understand me as well as, or even better than, I understand myself.79 Indeed, far from being automatic, self-knowledge, in a narrative model, is a difficult and important achievement, and self-ignorance is one of the greatest obstacles to autonomy.80 Self-knowledge is never just “a matter of easy or immediate introspection” because we can disavow things that we find within us and we must always interpret what we see.81 The process of coming to know ourselves, moreover, is a fundamentally social one: “we come both to know and to define ourselves in our interactions with others.”82 Thus, we need other people to engage in this essential

152 | Reconstructing Autonomy

process of self-knowledge, and they participate as potential sources of knowledge about us, rather than simply as sounding boards. Finally, the narrative model also rejects the narrow, instrumental model of reason in favor of a more substantive and more relational account. The liberal model of autonomy, focused on self-determining freedom, leads naturally to a reliance on instrumental reason as part of the mechanics of domination and control.83 Moreover, this narrow conception of reason is also connected to the ideal of the disengaged self, in which body, dialogue, emotion, and tradition are placed outside the self and only “pure, self-verifying rationality” is understood as internal.84 But reason can be understood in a broader way: as making sense of our experience. And in that broader conception, reason is essential to narrative autonomy because a large part of the process of narration involves making sense of our experience through stories. While means/end rationality is not irrelevant to this process, it is only one of the many forms of analysis that we use. Reasoning by analogy—so dear to the hearts of lawyers—and the use of metaphor are both important methods employed in the construction of stories.85 The analogies and metaphors available to us, like the characters and plot lines of our stories, are given by our cultures. These forms of reasoning, then, are always done within a particular cultural context and in relation to the other people who share it. Unlike means/end rationality, they are not even superficially conceivable in some ahistorical, noncontextual form. The relational forms of self-knowledge and rationality implicit in the narrative model are among the capacities necessary for the exercise of narrative autonomy. The other capacities often associated with narrative autonomy share this relational character. Related to the broader conception of reason is another capacity that is extremely useful in enabling narrative autonomy: the capacity to enjoy and value the exercise of our rational abilities. The process of telling our stories is not always a pleasant one. Sometimes we discover that the story that best makes sense of our lives is not the one we wanted to tell at all, or that we have been misunderstanding ourselves and others in ways that undermine our faith in our existing stories. In those difficult moments, we need a motive strong enough to keep us seeking narrative autonomy or we may simply give up. A variety of such motives is possible, but one of them is the pleasure in and commitment to exercising our broad range of rational capacities. Part of what keeps us going, in the face of the pains

Reconstructing Autonomy | 153

and disappointments of our efforts at autonomy, is the sense that we want to use our capacity for rationality.86 This desire to use our rational abilities is itself deeply cultured. First, as with metaphors and plot lines, the roster of available rational abilities and what they mean is culturally constructed and culturally specific. Second, the culture may not (as ours does not) distribute the resources necessary for the exercise of rationality equally to all people. Those resources include education, leisure time, and social recognition.87 Indeed, the desire itself may be distributed unevenly because of social stereotypes convincing people that their group (women of all races, men and women of color, working-class people, etc.) are less capable of and less interested in reasoning. Such a stunting of the desire to be rational can lead to limitations on narrative autonomy. Another capacity that is crucial to the exercise of narrative autonomy is imagination. Imagination is the capacity to conceive of alternative possibilities. Whatever story we begin with, imagination allows us to consider how it might be otherwise. This capacity is crucial to “self-understanding, self-reflection, and rational deliberation about the self.”88 We use our imagination to try on possibilities, to explore various new stories, and to offer new interpretations of old stories. Imagination has intrinsically relational elements. First, the cultural imagery available to us deeply shapes and can even limit the range of our own imaginings.89 Second, social relations and cultural conditions can affect the range of images available to us. Because social recognition is so important to self-worth, we feel tremendous pressure to identify with those cultural images that offer the hope for such recognition.90 And, of course, not all images within a given society are equally available to all people.91 Various other capacities contribute to narrative autonomy,92 but one that bears mention in a discussion of the relationality of the model is the capacity to connect to other people. We must have the interpersonal skills to connect to others in order to work together with them to affect the group identities that shape us. Our own individual stories are tied to the stories of the social groups to which we belong. We must be able to join with other members of those groups to tell those group stories in order to be autonomous.93 Moreover, we must also be able to understand and appreciate the normative standards of others. One’s ability to understand how others will evaluate her actions “affects the potential disclosive power of her acts and, therefore, her freedom” or autonomy.94 Narrative

154 | Reconstructing Autonomy

autonomy is in no way contrary to connection: rather, it is dependent upon it. Narrative autonomy is dependent on connection in yet another sense as well: it is created and sustained through specific social relationships. In other words, narrative autonomy is causally relational. We are taught to be autonomous by those persons who interact with us in ways that develop the necessary capacities,95 and we are sustained in our autonomy by those persons who continue to provide us with the opportunity to exercise those capacities. This view of the social creation of autonomy focuses on the particular individual relationships that affect a person’s development of specific capacities. For example, children need a caretaker who is responsive to their needs (to develop self-esteem and self-trust), and they need to be encouraged in imaginative play, and they need broad exposure to the images and patterns of reasoning available in their culture, and they need friends who will participate in their process of self-understanding. Mapping out the particular sorts of relationships that support each capacity is extremely important and could have enormous implications for areas as diverse as educational reform, welfare law, and child-custody determinations. It is a project that will require the interdisciplinary efforts of psychologists, sociologists, anthropologists, philosophers, and lawyers, among others. Taking a broader view, however, we should also inquire about the large-scale social structures that facilitate or inhibit the creation of these relationships for groups of people. Feminist philosophers exploring relational models of autonomy have suggested, for example, that structures of inequality—like racism and sexism—systematically reduce the resources available to oppressed people for developing autonomy.96 Such structures do not simply foreclose the possibility of autonomy for those oppressed: people can and do exercise autonomy under conditions of oppression.97 Indeed, a narrative model highlights the ability of people to exercise autonomy, however partial or episodic, even under conditions in which their choices are severely limited. Nonetheless, we must acknowledge the ways in which systems of social inequality interfere with the formation and exercise of narrative autonomy. Indeed, we need more complete descriptions of this process in order to make the case that such systems are, among their other immoralities, violations of personal autonomy. I will offer such a description of how systems of inequality can undermine autonomy in the context of speech in chapter 8.

Reconstructing Autonomy | 155

Finally, one question often asked of the “rational reflection” approach to autonomy is relevant to the narrative model as well: must the process be self-consciously engaged in by a person or is it enough if her behavior is consistent with the results of such a process? Do you actually have to tell your own story—to yourself or someone else—or is it enough that there is a story that could be told (perhaps by someone else) that would make sense of your actions as autonomous? The implicit criticism is that requiring someone to actually go through this process and tell the story is excessive. First, it is too linguistic: we often know things that we neither understand nor can articulate in language.98 Second, it seems to impose a highly cognitive model that may not capture the actual experience of many people who act in accordance with their own story without ever self-consciously telling it.99 My response to this criticism is to offer two qualifications and then a reaffirmation of the central point. The first qualification is to admit that stories can be told in ways other than through words. Any symbolic system of communication can be used for narrative purposes: dance, music, or visual images all can be used to tell stories. The narrative model does not privilege linguistic versions of stories over these other sorts, although it does require that whatever medium is used to communicate the story must offer a symbolic representation of meaning to the person telling the story and to a sufficient audience—real or imaginary—to function as the telling of a story and not only as the creation of beauty or the stimulation of a particular emotional response.100 The second qualification is to recognize that we do not and cannot tell stories constantly: much of our action on a narrative model will be implicitly autonomous rather than explicitly so. An autonomous person, on a narrative model, would often—perhaps usually—act spontaneously in ways that are consistent with her stories of herself without retelling those stories. Now the reaffirmation: I do not believe it is possible to call someone autonomous who experiences only this implicit type of autonomy and never engages in the self-conscious process of telling her own story, to herself or anyone else. A person who lived in unconscious harmony with a never-examined sense of self might be admirable in many ways—indeed, we might even say that she was a good person—but she would not be autonomous. The process of self-conscious reflection is essential to the possibility of conscious change, and the possibility of conscious change is,

156 | Reconstructing Autonomy

in turn, part of what makes autonomy so valuable both to individuals and to the social groups to which they belong. This requirement of self-conscious storytelling is not, however, as onerous as it might sound. Even when a person engages in telling her own story, she may do so in ways that are far less analytical and cerebral than in the liberal model of rational reflection (to which these sorts of criticisms are generally directed). While many analytical capacities and opportunities for education may be useful to improving and expanding the process, telling one’s own story requires no special intellectual skills. Narrative autonomy does not require one to identify second-order desires, to form a coherent life plan, or to assess the instrumental rationality of one’s first-order desires, as many liberal models of autonomy do. Narrative autonomy understands the process of self-knowledge and self-direction as centered on stories. Everyone tells stories.

3. Connection to Feminist Themes The three feminist themes identified in chapter 5 were a concern for those traditionally excluded, relationality, and the blending of moral and epistemological concerns and categories. The previous subsection provided an extended discussion of the ways in which narrative autonomy is relational. This section will briefly outline the ways in which this model of autonomy draws on the other two feminist themes. Feminism’s concern for those excluded begins, although it does not end, with women. The narrative model of autonomy, like other relational models, follows the same path. Relational models of autonomy generally arise from a serious and detailed examination of the experience of women. They build on the insight that women often experience themselves as exercising their autonomy within connections rather than in opposition to connection. They also build on women’s experiences of being denied recognition for autonomy and the meanings and harms of that status. Thus, the relational models start from the experience of those who have often been excluded from the status of autonomous individuals and build an understanding of autonomy that coheres with that experience. Indeed, the social constructionist approach that provides the foundation for narrative autonomy is more consistent with the point of view of the excluded than with the point of view of those who benefit from social hierarchies.101 People operating under more restrictive social expectations are more likely to be aware of the extent to which their choices and

Reconstructing Autonomy | 157

even their desires are limited by their social roles. When one’s own social role is understood to be the norm against which all other roles are measured—as it is for many white, heterosexual, middle-class men in modern America—then the constraints and meaning of that role are apt to disappear from view. When, on the other hand, one experiences oneself as outside that norm, the boundaries of one’s particular role are much clearer. The focus on the experience of those excluded can also be seen in the concern about the expressions of autonomy that are produced under conditions of oppression and that often go unrecognized by mainstream models of autonomy. The shift in attention away from choice and toward the story one tells is also consistent with a concern to better capture the experience of those excluded. And, perhaps most importantly, the focus on the pragmatic question about the capacities necessary for the exercise of autonomy and the social conditions under which those capacities are created and sustained represents direct attention to the concerns and needs of those traditionally excluded from social participation and denied social resources. At every step in the analysis, then, the narrative model of autonomy tries to take account of and express concern for those excluded. The narrative model also blends moral and epistemological concerns and categories. First, with respect to the particular category of knowledge about one’s self, such knowledge is not taken for granted in the narrative model of autonomy; rather, the process of acquiring such knowledge is a major part of the process of being autonomous. That is, the epistemological task of acquiring self-knowledge is a necessary aspect of autonomy. And, conversely, it is precisely through the practices and process of autonomy that one acquires such knowledge. Thus, one could not understand either this category of knowledge or autonomy except in relation to the other. With respect to knowledge about the world, the situation is the same. The fundamentally relational nature of autonomy means that it is impossible to engage in the process of telling one’s story without a certain amount of knowledge and understanding about the world one inhabits. For example, the agent must understand the conceptual categories available in order for her story to be intelligible to her interlocutors in the dialogue. Similarly, she must have a substantial amount of knowledge about how to engage in the practices of telling one’s story, a nonpropositional kind of knowledge often overlooked by traditional epistemology.

158 | Reconstructing Autonomy

The converse is also true: our identity—formed though the process of narrative autonomy—is deeply implicated in our knowledge of the world.102 Our stories shape our perceptions, and we see the world from within those perspectives. But we are able to recognize this shaping and the process of coming to understand it improves both our knowledge and our autonomy. The claim that epistemological issues are implicated at the deepest levels of our analysis of autonomy is not particularly shocking; even mainstream autonomy theories admit this. This complacency stands in marked contrast to the response to the parallel claim that moral and political issues are implicated in the foundations of our epistemology, which has been perceived as deeply threatening to the project of epistemology and traditionally resisted. The lack of parallelism here is due to the fact that moral and political theory is seen as a threat to the objectivity of epistemology, while epistemology poses no such threat to moral and political theory, which is already mired in subjectivity. The important point for this section is simply that a narrative model of autonomy takes these connections between epistemology and moral/political theory very seriously and sees them at every level of the analysis.

4. The Functions of Autonomy The central question for a narrative model seeking to fulfill the functions of autonomy is, how is some form of self-direction and/or self-ownership possible for a socially constructed self? All of the functions of autonomy described earlier require some notion of self-direction or self-ownership, but a thoroughly socially constructed self seems to leave little space for such a possibility. As Seyla Benhabib, puts it, “along with this dissolution of the subject into yet ‘another position in language’ disappear of course concepts of intentionality, accountability, self-reflexivity, and autonomy.”103 If people are so totally determined by their social context that their “retelling” of their stories consists of nothing more than a parroting back of the cultural elements they were given, it is difficult to see why this version of autonomy would be worth protecting at all. A social constructionist model must include some space for individual creativity in the retelling that allows autonomy to fulfill the functions described above. This section will first address this general issue concerning the possibility for creative retelling and then examine the particular functions of autonomy identified earlier in the chapter.

Reconstructing Autonomy | 159

a. the possibility and meaning of self-direction in a narrative model of autonomy I would like to divide this question into two parts: first, how can any relational, socially constructionist model of autonomy provide a basis for self-direction? And second, how might the particular model of narrative autonomy be especially helpful in explaining the process of self-direction? On the first issue, we need both a theoretical answer to explain how something more than mere “parroting” is possible and why it can provide the foundation for autonomy, and then a more pragmatic and contextual answer about the conditions under which that possibility is most likely to be realized. The theoretical answer grows out of the nature of the very symbolic systems through which the self is socially constructed. It is in the nature of such systems that meaning is created through them in ways that generate a permanent possibility of reinterpretation and transformation. Every use of a symbol is an interpretation of it: we never simply repeat, but always reinterpret. Our reinterpretations may be, and usually are, close enough to the previous uses that the meaning of the symbol in our social practice is not significantly changed, but change is always possible. Symbolic systems, while extremely powerful, are also very porous: they are created and recreated through a process that always contains the possibility for their transformation.104 Thus, “parroting” is never simply inevitable: the possibility for the exercise of creative reinterpretation always exists. As with the dichotomy between objectivity and subjectivity in Cartesianism, the dichotomy between free will and determinism in liberal autonomy theory is incapable of resolution and must be abandoned for a different understanding of the possibilities. The processes by which the past and present give rise to the future are more complicated and blended than this dichotomy can capture.105 We are always socially constructed, but that does not mean that we are never autonomous. Autonomy is not the opposite of social construction, but rather a particular form of it. The distinction between parroting back what we are given and using it in creative ways is meaningful and important in our lives, even though both of these possibilities are within the socially constructed universe we inhabit. Creativity does not take us out of social construction, but it does generate a particular sort of relationship between us and the social world, a relationship that has value for us.

160 | Reconstructing Autonomy

So, what is this relationship that, while socially constructed, is nonetheless worth calling creative? Why does this sort of creativity make the story I tell “mine” in a way that fulfills the need for self-direction and self-ownership that is at the heart of autonomy? And why is this relationship of creativity so valuable to us that it should count as autonomy? In our real lives, in which neither perfect freedom nor perfect determinism is ever our experience, what does it mean to exercise this kind of creative autonomy? This is an extremely complex question, in need of much study from the perspectives of a broad range of disciplines, so I offer here only a very preliminary—and probably very partial—answer. One of the important distinctions we experience between “parroting” and creativity involves the conscious application of our own standards to the reinterpretation of the cultural materials we are given. Creativity involves consciously reworking those materials in the light of our normative commitments, while parroting does not.106 While I believe the specifics of this experience require much greater attention and elucidation, this difference is sufficient to begin the process of explaining why creativity offers a meaningful foundation for autonomy. The conscious reworking of culturally given materials in light of our normative commitments is a process that makes the resulting choices and actions our own in a meaningful sense. In this process, we claim these materials and integrate them into our sense of self.107 We experience this process as active rather than passive: through this process, we transform these materials rather than simply absorb them. This process is the basis for self-direction because through it we actively shape our actions in the light of who we are and what we believe. Again, the creative process is not an escape from social construction: both our capacity and our inclination to engage in this process are, of course, socially constructed. The goal of a relational autonomy theory is not to get out of social construction but to find a meaning for autonomy within it. Thus, if one believes that creativity requires ultimate self-origination and that autonomy precludes determinism, then the description of agency offered here will look neither creative nor autonomous. I am suggesting that creativity might refer to a particular form of transformative agency (even if it is not ultimately self-originating) and autonomy might refer to a sense that one’s actions and choices grow out of a conscious process of reflection on one’s identity and commitments (even if the identity and commitments, along with the process itself, are not undeter-

Reconstructing Autonomy | 161

mined). Even if we are determined and not self-originating, it still matters whether or not we are creative and autonomous in this way. It matters because this experience of creativity and the autonomy it enables allow us to see ourselves and others as subjects rather than objects. The liberal fear of determinism arises from the important insight that we will lose something crucial if we see ourselves and others simply as objects.108 The liberal mistake lies, however, in assuming that if we are culturally constructed, we must be seen as objects. There is a basis for subjecthood that is not dependent on being unconstructed: it is a form of active, transformative agency that we exercise through conscious reflection on who we are and what we believe and value. This form of agency allows us to see ourselves and others as subjects regardless of whether or not there is some underlying causal determinism at work. Indeed, in everyday life, our perception of subjecthood is generally based on such agency rather than on some judgment that causal determinism has been ruled out. And seeing ourselves and others as subjects is, indeed, crucial because it means that we stand in a relationship of mutuality in which issues of respect, responsibility, and justice are central. Thus, when we experience this creative process in ourselves and expect and respond to it in others, we provide the foundation for just relationships between people. The theoretical answer to the question of how self-determination and self-ownership are possible in a social constructionist model, then, is that the porousness of symbolic systems always leaves open the possibility of creativity. Creativity, in turn, is a form of transformative agency that allows us to experience ourselves and others as self-determining and selfowning subjects, even as we are also subject to social construction. Thus, autonomy is possible. This theoretical answer, while necessary as a beginning, is not sufficient, however. We know that the degree to which this possibility for creativity is realized varies dramatically, within a single person’s life and also across persons and societies. Once creative transformation is established as theoretically possible, the interesting question is, what are the practical conditions under which such creativity is more likely to be exercised? Given that we can retell our stories in more or less creative ways, what allows us to be more creative? This new set of questions changes the inquiry in important ways. First, this inquiry is no longer simply a conceptual one, but has become a pragmatic effort to examine human experience. Instead of arguing

162 | Reconstructing Autonomy

about the relationship between causal determinism and the concept of free will, or even about the ways in which language is open to reinterpretation, we will be asking about our experiences of creativity. Under what circumstances do we have this experience? What mental or emotional processes do we go through on those occasions? The effort by many feminist writers to explore the capacities necessary for autonomy should be understood as an effort to move the discussion in this direction. This move is precisely parallel to the one suggested in chapter 5 with respect to truth: we must stop asking for a way to put together incompatible concepts and instead focus on how we build the social conditions necessary to ground the concepts that we need. In the analysis of truth, the challenge was to see that the slide into either relativism or conventionalism was retarded, not by the concept of truth itself, but by the practice of building communities of meaning that had sufficient resources to engage in both shared reality building and deep critique. Similarly, the way we assure autonomy in a social constructionist view is not by adopting a particular definition of it, but by understanding and constructing the social conditions that sustain it. Having a theoretical picture is a necessary first step; otherwise we can find ourselves paralyzed by the apparent impossibility of the task. But, once we have a conception that allows us to see autonomy as possible under social constructionism, the central task is to understand and create the conditions necessary to actualize that possibility. Moreover, we have no reason to expect that the answers to these practical questions will be singular or universal: the conditions conducive to creative retellings may vary from one culture to another and even across different contexts within the same culture. Many capacities and multiple and alternative strategies might contribute to such creativity. Thus, the pragmatic formulation of the question suggests that we should assess these conditions in particular contexts.109 Finally, the shift from the purely conceptual to the pragmatic question opens up an important normative dimension. When the self is understood as given in a presocial way, as in many forms of liberal theory, then the politics and morality of the construction of the self are hidden from view. We simply do not see the ways in which selfhood and autonomy for some people is constructed out of the exclusion and subordination of others.110 Once we recognize that the possibility of creative agency, of self-direction, is itself a product of particular social conditions, we can begin to ask about better and worse ways of constructing it. In other words, the cen-

Reconstructing Autonomy | 163

tral questions become not simply empirical, but also normative: which conditions lead to forms of self-direction that are good, and which lead to forms of self-direction that are morally or politically objectionable?111 What does a focus on narrative add to a relational approach to autonomy? As compared to a relational approach that focuses on rational reflection, a narrative model of autonomy offers several advantages. While the relational models tend to soften the rationalism of the liberal approach, the emphasis on reflection still shares the liberal focus on choice. All of the advantages of the narrative model flow from the shift in focus away from choice and toward stories. First, the focus on stories highlights creativity as the locus of autonomy rather than any conception—narrow or broad—of rationality. From a narrative perspective, rational analysis appears as only one of the many methods for performing the creative act of storytelling. When the focus is on choice, on the other hand, rational analysis seems to be the central activity through which autonomy is exercised: reason is the process through which we choose. The shift from choice to stories reduces the excessive rationalism. The focus on creativity is also helpful because it keeps our attention fixed on precisely the point at which the overly simplistic determinist chain is broken. The focus on rationality exacerbates our difficulties because it so easily degenerates into a replay of the liberal struggle over freedom and determinism. If we see choice in terms of rational analysis, then it appears that a choice will either be completely arbitrary or it will be completely determined: either reason compels assent or, if it does not, then our choice is arbitrary.112 When we see the central activity as creativity, however, then we are again directed to the pragmatic question about the social conditions that support that activity. Thus, a narrative model, with creativity at the center of its concern, directs our attention more systematically to precisely the pragmatic issues a relational model of autonomy generally suggests are central. Finally, the deemphasis on choice enables narrative autonomy to better represent the range of responses of persons experiencing oppressive social conditions. Choice may well be the most common locus of the experience of autonomy for those with power and opportunity, but a focus on choice obscures many of the subtler forms of resistance exercised by people suffering from various forms of social oppression.113 Some of the less privileged forms of resistance include refusing to stand out, pretending to accept the dominant values while not actually accepting them, and

164 | Reconstructing Autonomy

playing with irony the subordinate position assigned to one.114 When we focus on the actor’s own story about her life and identity and values, such devalued forms of resistance become recognizable as exercises of autonomy. This focus also highlights the subtle forms of collaboration to which oppressed people are subject. When a person’s choices are severely constrained by social oppression, both the person herself and others may be inclined to view her primarily as a victim of that oppression. It is crucial to take note, however, of the sometimes subtle ways in which victims collaborate in systems of oppression. The focus on narrative directs our attention to the ways in which the agent’s own story is shaped by those systems and the ways that she may be perpetuating them through her retelling of it, even if her choices are severely constrained. This narrative, relational model of autonomy serves all of the functions of autonomy identified earlier in this chapter. In many cases, however, the narrative model leads us to reconceive the correlative concepts— such as integrity, identity, and responsibility—in more relational terms. I will briefly explore the ways in which a narrative model of autonomy does the work we need it to do, while reorienting our understanding of certain basic moral concepts and relations.

b. the specific functions of autonomy 1. Self-Respect, Self-Trust, and Self-Esteem A sense of active agency and self-direction is necessary to ground the orientations of self-respect, self-esteem, and self-trust. One of the important functions of autonomy is to allow us access to this perspective on ourselves. Two questions then arise concerning the narrative model of autonomy. First, can this model provide a sufficient basis for a view of oneself as a self-directing agent to support these attitudes? And second, how, if at all, do self-trust, self-respect, and self-esteem change in such a model? Self-direction is possible within a relational model of autonomy because of the process of creative reinterpretation. As described in the previous subsection, we are never simply the passive recipients of forces beyond our control. But the narrative model also illuminates the extent to which self-direction is limited. First, self-direction is always partial: our stories are shaped by our own creativity, but also by all of the social and cultural materials we are given. We are the authors of our own stories, but we are never the only authors. Second, our capacity for self-direction is itself a product of social relations and conditions often beyond our indi-

Reconstructing Autonomy | 165

vidual control. The extent to which we enjoy the ability to direct ourselves is not itself determined only by us: we will have the necessary abilities (or not) because of a collection of social conditions many of which are controlled by people (like our parents, teachers, friends, and communities) other than ourselves. Thus, in a narrative model, our autonomy is always both partial and contingent. As a result, the nature of our self-trust, self-respect, and self-esteem must change. Take self-trust as an example. We cannot rest our trust on our control over our own lives, because we do not have the kind of control suggested by the liberal model. We can trust ourselves, not because we are the uncaused cases of our own lives (because we are not), but because our own creative agency is always part of what causes our lives. The self-direction on which our self-trust rests is meaningful, but substantially more limited than in the liberal model. This more limited and contingent version of self-trust offers, however, important advantages over the illusory security of the liberal version. The most important of these advantages is that, in a narrative model, our selftrust is in no way inconsistent with both humility and interdependence; indeed, it requires them. If we trust ourselves because of our role in the process of narrative autonomy, then we must realize both that this trust operates in spite of our limitations (rather than in the absence of them) and that we need others to help us develop and sustain that trust. We must relinquish the illusion of independence. The attention to the ways in which others contribute to our stories—either directly as coauthors or indirectly as supporting our own capacity for autonomy—should reduce the risk of pride attendant to the liberal version of autonomy and increase the likelihood of humility and compassion. The effects on self-respect and self-esteem are parallel. 2. Character and Integrity How can we understand the concepts of character and integrity in light of a narrative model of autonomy? To avoid the disintegration of identity experienced by Erik Erikson’s soldiers, the narrative model must provide a mechanism for claiming—and disclaiming—our internal landscape as our own. It must also offer a way to understand the process of changing our own characters and generate a useful meaning for integrity as a standard of moral evaluation. The narrative model of autonomy suggests that the process through which we claim our own characters—and discern those of other people— is fundamentally narrative: we tell stories about who we are, about the

166 | Reconstructing Autonomy

relationships we inhabit, and about the things we value.115 In the course of such stories, we may reinterpret older versions of the narrative or tell it in ways that represent a change in direction or identity. The concepts of character and of changing our character are coherent and important parts of a narrative model of autonomy. Nonetheless, character and the process of changing it are also given new and different meaning within this model. If one’s character is fundamentally narrative, then it is much more fluid than in the traditional view: it is never finished and it is always subject to interpretation. Moreover, this process of interpretation is essentially social: we construct and reconstruct our own characters and those of others through a dialogue with other people. This shift in understanding of character draws our attention to the fact that the social conditions under which character is constructed are not the same for everyone. “[W]e are not all in the same discursive positions any more than we are all in the same social ones. . . . There are moral problems with the social distribution of narrative resources and the credibility to use them.”116 The narrative model of autonomy allows us to see that the social construction of character and the possibility of integrity are moral and political issues with implications for the distribution of social resources. This shift in the meaning of character gives rise to new understandings of integrity as well. Margaret Urban Walker has proposed that integrity should be understood not as logical consistency, but as “reliable accountability.”117 Integrity requires not total integration or consistency, but “local dependability” and “responsiveness to the moral costs of error and change.”118 “Integrity is reliability not only at the outset, in the having of firm and coherent convictions and publicly expressing them, but also after the fact, in various reparative responses, sometimes including changes of moral course.”119 This conception of integrity does not fit very well with the older model of autonomy. In the liberal model, where autonomy is focused on choice, this sort of integrity seems to have little relation to it: there is a vast difference between living a life of one’s own choosing, on the one hand, and living a life one is forced to live with integrity, on the other hand. Indeed, Walker argues that on this understanding of autonomy, “[i]ntegrity in the reliability sense and autonomy in the social-political sense [i.e., freedom to choose] may co-vary inversely.”120

Reconstructing Autonomy | 167

But this reliability and responsiveness are central to a narrative model of autonomy. In the narrative model, relationships of trust and intelligibility are the foundation for autonomy: it is only within such relationships that the capacities for autonomy can be formed and sustained, only within them that the process of narrative interpretation that is central to autonomy can take place. Integrity, in Walker’s insightful new conception of reliability and responsiveness, is the foundation on which relationships of trust and intelligibility are built. As such, it is a crucial aspect of character from the perspective of narrative autonomy. 3. Responsibility In her wonderful book, Moral Responsibility and the Boundaries of Community: Power and Accountability from a Pragmatic Point of View, Marion Smiley traces the development of the concept of responsibility through classical and Christian versions to the modern, liberal (Kantian) model.121 She argues that the efforts of philosophers to adapt this model of responsibility to softer versions of free will that are more consistent with causal determinism are fundamentally unsuccessful.122 She suggests that we should, instead, reenvision responsibility as a complex collection of social practices. The role of free will, she argues, was to “provide us with the conceptual means of obscuring the social and political considerations that ground our judgments of causal responsibility and blameworthiness in practice.”123 If we scrutinize those social and political considerations directly, however, we discover that our practices of ascribing responsibility are essentially relational, interpretive, and pragmatic.124 She suggests that the form of moral agency required by these practices is not the contra-causal freedom over which philosophers struggle. Rather, we use a form of agency that “concedes a great deal more determinism than more traditional concepts of moral agency do, [but that] does not go as far as to dissolve the self in the way that postmodernist accounts of the individual do.”125 Such a relational model of responsibility suggests that our social practices of ascribing responsibility fulfill several functions: shaping future behavior, establishing and maintaining relationships of trust, and marking the boundaries of the relevant moral community. Moreover, while these practices rely upon a conception of autonomy, the relationship is reciprocal: autonomy also relies upon and is supported by the practices of responsibility. A narrative model of autonomy can support, and be supported by, this relational practice of responsibility.

168 | Reconstructing Autonomy

The first challenge is to explain how a narrative model of autonomy provides sufficient self-direction to ground the project of affecting future behavior. It was, after all, the apparent loss of contra-causal freedom in the face of causal determinism that led to the crisis over responsibility. First, of course, we must take note of the fact that even within a totally causally deterministic model, the social coercion represented by the practices of responsibility would have an effect on future behavior: such systems of reward and punishment would operate as part of the social forces determining individual behavior. But in this picture, the autonomy of the agent plays no role in the process through which the ascription of responsibility causes the future behavior. The question is whether a narrative model of autonomy can support a practice, such as ours, in which the causal efficacy of ascriptions of responsibility are understood to involve the autonomy of the agent, rather than flowing directly from the social practices themselves. Why is it that an ascription of responsibility is supposed to alter future behavior (aside from the behaviorist explanation)? What happens when a message of praise or blame is communicated to an autonomous agent? The narrative model of autonomy suggests that the result is twofold: first, the agent is presented with certain information that she must assess in terms of, and perhaps incorporate into, her stories about herself; and second, the agent’s relationships with her dialogic partners within the moral community are affected by those ascriptions of praise or blame, and she must respond to that as well. The process of revision and reinterpretation initiated by such an event is socially constructed both in its form and in its substance. But there is always the possibility of creative agency in the process of reinterpretation. The crucial question is what conditions lead us to a greater ability to exercise that creativity in productive ways? Here again, we are led away from the conceptual puzzle of determinism to the pragmatic consideration of the effect of social conditions on our capacities for the specific activities through which we experience autonomy. The important point is that the answer to the question of how we can be autonomous enough to direct our future behavior in response to ascriptions of responsibility is primarily practical rather than conceptual, and may vary with particular circumstances. A narrative model directs our attention to the pragmatic questions concerning the conditions that facilitate this process. For example, under conditions of social estrangement and isolation, people are less likely to be able to engage in this

Reconstructing Autonomy | 169

process of retelling their stories; as a result, they will be less likely to have ascriptions of praise or blame affect their behavior through autonomous means (although, of course, social coercion may still affect them in nonautonomous ways). Or, people whose cognitive or emotional skills are limited by lack of education, or by social training to restrictive roles, will be less likely to have the tools necessary to incorporate new information successfully into their stories; as a result, they will be less likely to have ascriptions of praise or blame affect their behavior through autonomous means. The ascription of praise or blame affects the agent’s future behavior by affecting the narrative process through which the agent exercises her autonomy, and the important questions about this process are the practical ones about the skills and conditions that facilitate or inhibit it. The narrative model helps to make sense of the second function of ascriptions of responsibility as well: the establishment and maintenance of relationships of trust. By drawing someone into the dialogic relationships in which narrative autonomy is produced, ascriptions of responsibility (if they work) generate the conditions necessary for trust. By allowing a victim to claim the attention and concern of a wrongdoer, and/or of the moral community they share, ascriptions of responsibility reestablish the norms of that community. This effect is created because the process of ascribing responsibility operates as part of the narrative through which the members of the community develop their own autonomous sense of self. It is because our sense of identity and autonomy is developed through narratives built in such a dialogue that the ascription of responsibility can create or maintain the social norms on which we base our trust. The same process explains why the practices of ascribing responsibility mark the boundaries of our moral community. The refusal to ascribe responsibility to someone is a refusal to stand in the reciprocal relationship with her through which the participants’ experience of autonomy is produced. The relationship on which “relational autonomy” rests is one of vulnerability. The vulnerability to other members of the community justifies the imposition of responsibility on them, and, in turn, the acceptance of responsibility by them makes the vulnerability bearable. Through the processes of narrative autonomy, then, we make each other the sorts of agents who can and should be held responsible. 4. Social Change and Democratic Politics A relational model of autonomy has a close connection to deliberative or discursive approaches to

170 | Reconstructing Autonomy

democracy. A shift to a narrative model of autonomy would, first, confirm the move in political theory toward deliberative conceptions of democracy; second, clarify the meaning and possibility of political movement in a social constructionist view; and, third, highlight the uses of narrative in our political life and the narrative nature of politics. A narrative model of autonomy, like relational models generally, suggests that democracy is better understood in discursive than in aggregative terms. Because our autonomous identities, and the interests, values, and concerns that flow from them, are created through a process of dialogue with others, it is a mistake to see them as endogenous to politics. Moreover, political interaction is one of the important realms in which we engage in building narrative autonomy. From a relational perspective, the goal of politics is not to preserve a preexisting sphere of individual autonomy, but to provide the types of interaction that constitute one part of our autonomy. Again, a deliberative model of democracy captures this dynamic, while an aggregative model tends to deny or overlook it. Democratic politics, in particular, is a form of engagement that seeks to ensure a given type or degree of equality among the participants. While the value of democracy in an aggregative model is primarily defensive— it protects us against incursions on our freedom by the government—the value from the point of view of relational autonomy is primarily constructive: democracy provides the opportunity for all citizens to participate in the political relationships that constitute one form of autonomy. The narrative view does not imply that the promotion of autonomy is the only goal of democratic politics. It is possible to understand political processes as serving a variety of goals, including generating positive policy outcomes126 and character building,127 along with facilitating the forms of self-direction and self-determination most directly relevant to autonomy. But the nature of each of these goals is affected by the underlying relational model of autonomy. First, the goals need to be seen in relational terms that recognize the interdependence of persons. For example, measures of welfare (such as gross domestic product) that ignore the impact of disparities in wealth on the relationships of citizens would be inconsistent with a relational understanding of autonomy. Second, the goals need to be seen in terms that incorporate the creative agency of the citizen and do not simply reduce her to a passive recipient of goods.128 Thus, the relational model of autonomy has relevance for our understanding of the other goals served by democratic politics beyond its role in individual autonomy.

Reconstructing Autonomy | 171

The narrative version of relational autonomy directs attention specifically to the use of stories in the political process. On one level, a narrative model asks us to understand the political process as a whole in a narrative way: to ask about the hermeneutics of our politics and the construction of identity through it. In particular, a hermeneutic approach to politics illuminates the nature of political movement. It suggests, for example, that political movements organized around “identity groups” need to be seen in a relational rather than an essentialist way. The “identity” does not precede the group, but is instead constituted by the members and their relations to each other. As a result, such groups do not require any particular degree of homogeneity.129 Nor should they be understood as preexisting social artifacts, but rather as created through the very process of political movement. Moreover, such collective identity creation should be seen as an important aspect of autonomy: one must participate in the process of giving meaning to the social identities one shares in order to fully tell one’s own story. Thus, from the perspective of narrative autonomy, political movement appears as inherently valuable, rather than simply instrumental, and actively creative, rather than simply a reflection of preexisting interests and groups. The narrative focus also asks us to look for stories at the microlevel of politics. As Iris Young has argued, narratives serve a variety of functions in a discursive model of democracy: helping people to give word and voice to harms that are presently without recognition in their society; helping to create and sustain subcommunities, such as those formed through the political movements described above; providing access to very different experiences and challenging stereotypes about those who are different; revealing the sources and lived experience of values and cultural meanings; and presenting knowledge from different social positions.130 One might summarize these functions by pointing to the fact that they all involve a combination of epistemological and moral concerns, they all involve creative and responsible agency, and they all facilitate the inclusion of differences within the polity. Thus, a narrative model of autonomy offers support to a particular vision of democracy, a vision in which democracy is a discursive practice within which we negotiate difference and create the interconnected identities that allow us to move forward together. The narrative model makes sense of our experience of political movement and highlights the creative and interdependent nature of our political lives. It grounds our

172 | Reconstructing Autonomy

commitment to democratic politics and offers a basis for criticism of present practices.

C. Conclusion The concept of autonomy serves important functions in our moral and political experience. A narrative model of autonomy can fulfill those functions, without the problematic individualism of the liberal model. Based on a fundamentally relational vision of the self and focused on the pragmatic and contextual questions about how we tell our own stories, the narrative model provides a basis for self-respect, self-trust, and selfesteem, grounds our sense of character and integrity, supports our practices of ascribing responsibility, and maintains the meaning and possibility of democratic political action. When seen through the lens of a narrative model, moreover, many of these moral and political concepts appear different in important and interesting ways, ways that increase our understanding of the relational and contextual nature of the moral and political space we inhabit.

part iv

Speech

This part of the book will move back towards the First Amendment by outlining the connections between the visions of truth and autonomy developed in the last part and the principle of free speech. The focus of this part will be the theory of free expression and the ways in which such theories could be based upon the new models of truth and autonomy, but along the way, I will offer some thoughts about the implications of these theories for particular legal doctrine. An important caveat is in order here: a relational truth theory or a narrative autonomy theory will no more generate determinate answers to particular cases than did the traditional truth and autonomy theories. The goal of a theoretical framework is not to provide such determinate answers, but to offer guidance about the issues and considerations that bear on the question, to suggest relative priorities among concerns, and to frame the possibilities in particular ways. The description of the free speech right that I offer in these chapters is intended in this spirit: not as specific lines of legal doctrine, but as a catalogue of issues and arguments out of which a variety of doctrinal approaches might be constructed. Chapter 7 will develop a theory based on relational truth, and chapter 8 will develop one based on narrative autonomy. In fact, however, these approaches are not clearly divisible, but are different facets of a single perspective brought to bear on the social life of which free speech practices are a part. In the conclusion, I will offer a more synthetic and comprehensive view from that perspective.

173

7 Free Speech and Truth

If we adopt a relational model of truth, is there a sense in which free speech is still necessary to the attainment of truth? Is there, in other words, an alternative truth theory of free speech? I think that there is indeed such a relational truth theory of free speech, but the ways in which speech contributes to the search for truth are systematically different from the ways it did so in the old truth theory. Speech is a useful, in fact a necessary, part of the process of truth making, where truth is understood to serve the purposes described in chapter 5. This chapter will outline the role of speech in constructing a shared reality and supporting deep critique, helping us to assess our social policies in terms of whether they “work,” and establishing a connection to a larger, nonhuman reality. The purpose is to demonstrate that a relational model of truth does give rise to a principle of free speech: speech deserves protection because it is a particularly important—indeed, a necessary—mechanism through which we construct relational truth. From the perspective of relational truth, however, the shape and nature of a free speech principle is very different from the shape and nature in the traditional truth theory. Speech, in a relational truth theory, is seen primarily as systems of relationships rather than as actions of individuals. As Robert Post has so persuasively argued, we must think of speech as a collection of social practices rather than as simply the communication of messages.1 I will try to identify some of the characteristics of these social practices and explain how those characteristics make speech particularly helpful in the process of knowledge making. “Truth-seeking is not merely a matter of sentences and propositions; it also involves habits of mind, priorities of reason, intersubjective orientations, and attitudes that, when taken together, make up what we recognize to be rational exchange or a collective search for knowledge.”2 We must add to this list: the process of constructing relational truth also involves the institutions, systems, and

175

176 | Free Speech and Truth

cultural practices that shape the relationships in which truth is made.3 If speech deserves protection because it helps us to construct truth, understood in terms of the functions addressed in chapter 5, then it is such systems and the relationships they create and sustain that should be the focus of legal protection rather than an individual speaker or a particular speech act. A legal right to free speech based on this relational truth theory would be a right to the protection of such systems and relationships. The functions of truth identified in chapter 5 included generating a shared reality, providing grounds for a deep critique of existing social practices and ideas, focusing our attention on whether our policies are “working,” and connecting us to a larger, nonhuman reality. The practices and systems of speech are some of the most important ways that we fulfill these functions; that is, speech is a primary mechanism for the construction of relational truth. From the perspective of relational truth, speech is important and deserving of protection.

A. Shared Reality and Deep Critique I suggested in chapter 5 that these two functions are closely related: they point to the need for a relational model of truth to avoid deteriorating into either simple relativism or conventionalism. In particular, a relational model, in which all knowledge is contextual, must not leave us helpless within a context that lacks the resources to generate both shared reality and deep critique. Cartesianism attempts to define away the risk that we will be unable to find the necessary resources simply by relying upon a particular concept of truth. This strategy has failed, but that does not mean that we are necessarily stuck in contexts that cannot support either shared reality or deep critique (i.e., that we are forced to choose between relativism and conventionalism). The sense that those are the only possible choices is itself a remnant of the Cartesian assumption that we must choose between subjectivity and objectivity, that there is either one truth or no truth. I argued in chapter 5 that the way for a relational model of truth to avoid this apparent dilemma was to look at the need for an appropriate context as a pragmatic problem rather than a conceptual one. The answer lies not in adopting a particular concept of truth, but in the practical conditions that support our ability to build a context with the resources we need. What keeps us from sliding into relativism or conventionalism is

Free Speech and Truth | 177

our active and creative agency in building contexts that allow for shared reality and deep critique. The central question for a relational model of truth, then, is, what are the conditions under which, the relationships within which, we find ourselves best able to engage in this activity? This question leads us directly to speech. Speech is, of course, the primary practice through which we manipulate symbolic and conceptual frameworks. Usually, the symbols involved are words, but not always. Use of nonverbal visual or aural symbols (a flag or a “raspberry” sound) is speech as well. Indeed, it is precisely when nonverbal activities involve such a symbolic or conceptual aspect that we are inclined to include them in the category of speech. The distinguishing characteristic of speech, understood in this broad sense, is that it is conceptual and symbolic.4 Why might such conceptual manipulation be important to the task of building and reinterpreting contexts so as to construct a shared reality and offer deep critique? The simple answer is that most, if not all, of our practices of building and reinterpreting contexts involve the conceptual manipulation that characterizes speech.5 There is, however, a deeper way to understand this connection: speech processes are a particularly important means of creating contexts in which shared reality is possible because they involve the types of relationships that reflect the moral foundations of the goal of a shared reality. The point in seeing shared reality as a goal is that we recognize our interdependence and the moral necessity for us to treat each other as joint participants in the project of building our collective life. The value of shared reality rests upon a particular moral commitment: to go forward together on the basis of trust and understanding. Speech, and the truth making processes that take place through it, are closely connected to that commitment because when we engage in the process of truth making through speech, we are led to exhibit this commitment to treat others as joint participants in the project. How does speech involve this moral commitment? The social practices we identify as speech involve an “intersubjective orientation” that we could describe as respect. In a dialogue, one must recognize the other participant as a moral agent deserving of respect.6 What exactly about an agent makes her deserving of this respect (e.g., her autonomy, her capacity for happiness, her reason) and what exactly such respect requires in terms of treatment of such agents are, of course, extremely complex and controversial questions. But the notion that dialogue entails some standard of respect is, at this point, almost a cliché in legal scholarship.7 As

178 | Free Speech and Truth

Seyla Benhabib puts it, “All argumentation entails respect for one’s conversation partners; such respect belongs to the idea of fair argumentation; to be a competent partner in such a conversation entails recognizing the principle of equal respect.”8 Speech is not, of course, the only way to guarantee such respect. A common alternative basis for respect is similarity or joint identification with some larger cultural identity (e.g., an ethnic or racial group or a political affiliation). In other words, we tend to respect people we think are like us. This is, of course, a highly problematic basis for respect since the usual (although perhaps not inevitable) corollary is that we do not respect people who are different. The dialogic basis for respect has the important epistemological and moral advantage of demanding respect across lines of difference.9 Thus speech is an extremely useful mechanism for placing people in the appropriate relationship and creating pressures and incentives for them to stay there, despite their differences. This connection between speech and respect goes beyond a merely theoretical relationship: the social practice of dialogue generates some very concrete experiences that provide a foundation for respect.10 The process of conversation with someone makes it more difficult to think of him as an object; it humanizes one’s opponent. It is much easier to demonize an opponent to whom one never actually has to speak. Seeing his efforts to persuade increases one’s own willingness to compromise and tends to soften the positions that one takes. Hearing him tell his side of the story makes it easier and more likely that one will make the leap of imagination necessary to see his point of view; in other words, it increases the possibility of empathy, which is conducive to respect. It is necessary to emphasize, however, that respect is not equivalent to polite manners or temperate styles of speaking or positive feelings towards one’s interlocutors. One can be respectful in a dialogue while being intemperate, angry, and even rude—it happens in families, classrooms, and political caucuses all the time. Respect here refers to the fundamental attitude of reciprocity necessary to a dialogue: the idea that one faces one’s interlocutor as an equal participant in a joint project. Failures of manners do not necessarily undermine the project of creating relational truth; failures of respect, in this foundational sense, do. Moreover, to conflate respect with politeness or a calm tone of voice is to systematically privilege speakers who already possess social privileges such as education and the relatively high rank in social hierarchy that allows one to feel comfortably calm in the face of controversy. By further

Free Speech and Truth | 179

marginalizing the voices heard less often, such rules of politesse can reduce our ability to produce relational truth, rather than promote it.11 The respect that is required is that one treat the dialogue as a joint project in which one’s interlocutors are equal participants. Speech, while far from being a guarantee, is an important mechanism for building and practicing such respect. Thus, the processes of truth creation through speech systematically promote the underlying moral relationship on which shared reality depends. Speech—both in theory and in practice—relies upon and generates respect. A relational model of truth, then, would suggest that we should protect speech because it is a method of fulfilling this function of truth that facilitates the respect that is necessary to shared reality. Speech is also a particularly useful mechanism for the truth function of offering deep critique of social conventions. In this context, it is useful to revive the concept of reason—in a non-Cartesian form, of course. In this form, giving reasons is a deeply context-dependent activity in which we try to make sense of our experience. One important part of the process of making sense of experience is the consideration of alternatives to our present perspective. In other words, what it means to be rational—or to give good reasons—is, at least in part, to consider the alternatives.12 The social constructionist model of truth makes clear why such consideration of alternatives is a requirement for the production of knowledge. If all knowledge is situated, and no perspective can claim the neutrality or objectivity available in a Cartesian framework, then the only mechanism available for insuring that one is not missing something quite important is the consideration of alternative perspectives. As many writers have pointed out, such consideration is a precondition for the only type of objectivity available in a social constructionist epistemology.13 Consideration of alternative perspectives is, moreover, part of the moral and political responsibility of the knower in this model. Part of what it means to be epistemologically responsible is to make yourself vulnerable to challenge by knowers with other perspectives. So, rationality, understood in this broad sense, requires the consideration of alternatives and, in so doing, it promotes the deep critique function of relational truth. Speech is connected to rationality because it is an important—perhaps even a unique—mechanism for the consideration of alternatives. Speech is the primary way in which we get access to other minds, allowing us to perceive alternatives we would otherwise miss. Such access is a necessary (although, of course, not a sufficient) basis for us to seriously consider the

180 | Free Speech and Truth

alternatives.14 Indeed, some have argued that communicative activities, uniquely, “contain within themselves the possibility of surveying the alternatives.”15 I would not go that far: there is no question that our eyes can be opened to a new way of seeing things by events other than speech. For example, the experience of a serious illness can give one a new perspective on the health care system. But we cannot, and would not wish to, create for ourselves and others all of the sorts of experiences—some quite painful—that would be necessary to generate such a shift in perspective. Moreover, some of the changes that would be required—like growing up with a different skin color—are simply impossible. Speech is the primary—perhaps the only—mechanism through which we can glimpse another perspective without actually changing the circumstances of our own lives. While all sorts of speech have this potential (and all would be potentially valuable in this model), it is worth noting that narrative seems to be a particularly effective form of speech for communicating an experience or perspective different from that of the listener. Narrative is more effective at cross-perspective communication than many other forms of speech, such as argumentation, for several reasons. First, narrative engages our empathy and imagination, encouraging us to try to see another experience from the inside rather than simply to judge it from within our existing perspectives.16 Second, narratives can help to overcome the prejudices that block our access to a different perspective. Often we bring to the dialogue assumptions that we are not even aware that we hold. These assumptions are difficult to challenge directly precisely because we are unaware of them. Narrative offers us a view from which such assumptions are both made visible and questioned.17 A model of truth that sees the ability to recognize and seriously entertain different perspectives as central should, therefore, value narrative highly. Speech, then, is a uniquely valuable mechanism for coming to understand other perspectives. As such, speech is an essential vehicle for reason, understood in a broad sense. Speech is therefore an indispensable mechanism for the fulfillment of the deep critique function of relational truth. The discussion so far has taken a view of speech relationships at the microlevel: the relationship between individual speakers and listeners. I have argued that this relationship is one that promotes both shared reality and deep critique. One of the advantages of a relational theory of truth is that it suggests that we must also consider speech relationships at the

Free Speech and Truth | 181

macrolevel: the systems of speech in which large numbers of people participate and through which they negotiate the conceptual and symbolic frameworks inherent in their truth making. Such systems would include the mass media, political campaigning, public education, and public fora. In order for a relational model of truth to function, these speech systems (through which people creatively reinterpret their frameworks) must be vibrant and healthy. More precisely, these systems of speech must facilitate both the creation of shared reality and the promotion of deep critique. The proper functioning of such systems is the precondition for the escape from the dilemma of relativism versus conventionalism. A free speech theory based on a relational model of truth would, therefore, see the protection of the function of such systems of speech as a central goal of the First Amendment. At this moment in the history of First Amendment scholarship, any theory that suggests that the First Amendment protects social systems must engage with the ground-breaking and influential work of Robert Post. Post has led the way in arguing that the First Amendment should be seen as protecting the values inherent in forms of social ordering, which he calls “domains,” and that “what are ordinarily called personal ‘rights’ to freedom of expression might perhaps most fundamentally be understood as mechanisms by which the law defines and establishes such domains.”18 The three primary domains he identifies are community, management, and democracy. He argues that very different values are inherent in these different domains, values that are in turn connected to different conceptions of the person.19 Moreover, while the domains are interdependent, they are also in competition: many social practices and institutions can be understood as part of more than one domain and which view one (or a court) chooses will have consequences for the types and degrees of constitutional protection for speech within that social practice.20 One important consequence of Post’s view is that there is no single rationale for the First Amendment’s protection of speech; rather, speech is protected for different reasons and in different ways in different domains.21 We have (at least) three First Amendment regimes, not one. The fundamental reorientation that Post has suggested for First Amendment theory is extremely important, and I believe that a relational model of truth helps to explain why. If truth is constructed and relational, if the central activity through which we construct it is the activity of reinterpretation that allows us to generate resources for shared reality and deep critique, then it is apparent that the value of speech for the pursuit

182 | Free Speech and Truth

of truth lies not only (or even primarily) in particular acts of speech but rather in the systems of speech that facilitate such practices of reinterpretation. Thus a relational model of truth is in agreement with Post’s revolutionary suggestion that First Amendment analysis must attend to systems of social practices and not only to the individual speech acts that currently dominate doctrinal analysis. Post goes on to articulate a complex and powerful theory based on the idea that there are three primary forms of social practice in which speech has value: democracy, community, and management. Each of these domains has its own inherent values, and the regulation of speech in each is guided by those values. In democracy, speech—in the form of “public discourse”—is the means through which individual autonomy is reconciled to collective self-determination. It is because one’s individual will helps to shape the collective will that guides the government that one is self-determining.22 Public discourse, then, cannot be regulated on the basis of community norms, such as civility. Such regulation would have at least two difficulties. First, it would foreclose the participation of some persons in the community (who wish to express a view considered uncivil) and thereby deny them the status of self-determining citizens. Second, it would set boundaries based on collective identity on the very conversation in which our collective identity is supposed to be formed.23 The principle governing speech within the domain of democracy, then, is the classic First Amendment commitment to bar any imposition of orthodoxy by government. Norms that “have the potential to constitute a specific form of community life” are “inconsistent with the neutrality essential to public discourse.”24 In the domain of community, on the other hand, speech is “the medium through which the values of a particular life are displayed and enacted.”25 Post argues that the privacy-based torts, such as defamation, invasion of privacy, and intentional infliction of emotional distress, serve the function of protecting socially and normatively defined spaces and practices that are understood as necessary to respect and personality.26 Enforcement of communal norms is crucial to the social order of community: civility rules, for example, are community defining.27 Thus, “[r]esponsive democracy posits persons with autonomous selves; community posits persons with socially embedded selves. Responsive democracy strives to open up the field of social choice; community to restrict it.”28 Post emphasizes that the domains are interdependent, and he acknowledges that democracy is dependent on community in several ways.

Free Speech and Truth | 183

We need communal institutions to inculcate the value of self-determination on which democratic commitment is based, as well as to instill a sufficient sense of identification with the culture so that people will experience the public discourse as being about a collective identity that includes them.29 And we also need a common culture that will maintain the civility and respect on which deliberative reason depends so that public discourse will not be experienced as irrational and coercive.30 Most importantly, Post admits that his model of democracy includes certain substantive commitments—to the value of self-determination and to the process of critical reflection31—so the boundaries of public discourse “cannot be fixed in a neutral fashion.”32 The first several forms of interdependence suggest only that democracy requires that a vibrant community continue to exist within the polity. They point to the practical dependence of democracy on community—the need for a culture outside of discourse to support it—that is theoretically consistent with the central commitment of the domain of democracy not to enforce communal norms within public discourse (although there may, of course, be practical difficulties with maintaining such a communal culture in the absence of government regulation of speech). But the last point above—the existence of substantive commitments underlying public discourse itself and the use of such commitments in determining the “boundaries” of discourse—goes considerably further. This point suggests that democracy is not just practically dependent on community, but actually interpenetrated by it, because the government does, in fact, regulate public discourse in order to protect certain norms. Post explains the Chaplinsky line of doctrine concerning fighting words as precisely such a minimalist but substantive regulation of public discourse in the interest of majority culture.33 Indeed, given that the boundaries of public discourse are themselves contested, such substantive regulation in accordance with communal norms is simply inescapable. The definition of the boundaries of public discourse cannot be neutral; it must represent the norms of some community concerning the meanings of rationality and critical thought. Thus, there is an inherent tension in democracy as Post describes it: public discourse requires government regulation (in the interest of community norms that define the boundaries of public discourse itself), but it also requires that government be disabled from regulating public discourse as a way of enforcing community norms. The commitments to self-determination and critical thinking are themselves the source of substantive limits on public discourse that are, simultaneously, necessary for

184 | Free Speech and Truth

and in tension with the goal of truly open dialogue about our collective identity.34 A relational model of truth helps to explain why this tension is inevitable, but it also indicates that the contrast between the domains of democracy and community is neither as stark nor as fundamental as Post’s approach suggests. The tension in democracy is simply a reflection of the more general tension between the need for both shared reality and deep critique. This tension is also present in community, although Post tends to emphasize shared reality in this domain and ignore the need for deep critique. When we attend to the issues highlighted by a relational theory of truth—-the need for both shared reality and deep critique—we find that the domains are neither as separate nor as different as Post’s analysis suggests. Post has set up democracy as the realm of deep critique, but he has found it impossible to banish the need for shared reality as decisively as his First Amendment principle suggests. His characterization of the domain of democracy as the area in which deep critique will be protected can be clearly seen in his criticisms of First Amendment theorists, like Cass Sunstein and Owen Fiss, who emphasize the need to improve the quality of public deliberations. Post argues that such theorists are inadequately radical in their view of what must be up for debate in public discourse. In his view, public discourse cannot assume any substantive norms about our collective identity because everything is open to dispute.35 Thus, public debate is the realm of deep critique. But this emphasis on deep critique must give way at some point because, as Post recognizes, it is not possible for everything to be open to dispute. The possibility of discourse itself requires some shared reality to ground it: some common fund of ideas and concepts, some common understanding of the nature of argument, some common commitment to noncoercive speech as a mechanism of defining collective identity. Thus, the ultimate commitment to deep critique that Post wishes to assert must be balanced against the need for shared reality, even within the domain of democracy. Community, on the other hand, he describes almost exclusively in terms of shared reality. He emphasizes the need for communities to enforce their norms, but he pays insufficient attention to the fact that within community, as within democracy, those norms are often contested. Communities must interpret their norms before they can enforce them, and

Free Speech and Truth | 185

often there are disagreements within a community over the meaning or importance of certain norms.36 Indeed, in modern cultural studies, cultures are seen as neither static nor homogeneous, but as “‘essentially contestable’—cultural meanings are ‘constantly up for grabs’ as individuals negotiate for the power to make and control them.”37 In other words, the dialectic between shared reality and deep critique plays itself out in both democracy and community. We cannot ascribe our need for deep critique to one domain and our need for shared reality to the other; instead, we must struggle with the balance between them in both domains. Unfortunately, then, assigning a speech practice to either community or democracy does not in fact determine whether we should favor dissent or the imposition of community norms. In both domains, both practices must be protected to some degree. Post, I believe, recognizes this in the domain of democracy, but it must be extended to community as well. When it is so extended, then the dynamic in the two domains is essentially the same. Nonetheless, Post is certainly right that the balance between shared reality and deep critique can and should vary with the context. In the editorials of a news magazine, for example, the law should protect a wide range of dissenting views, while in a workplace it might impose much stricter community norms of civility and respect. Perhaps the division of domains serves the purpose of identifying the places in our social life where one function of truth acquires presumptively greater weight in this balance. The difference between the domains of democracy and community would then be not a difference in fundamental nature, but simply a difference in emphasis. They are not separate realms based on incompatible models of the self, but instead a shorthand for the idea that the balance between deep critique and shared reality cannot, and should not, be struck exactly the same way in every area of social life. In this guise, I think that the domains are useful and that a relational model of truth helps to explain why. I would submit, however, that Post’s own highly nuanced analysis of particular issues in First Amendment law suggests that the contextual analysis that is necessary will generally be a more finely tuned one than simply assigning the speech practice at issue to one domain or another.38 In general, we will need to know a great deal about the particular contextual practice at issue in order to decide how to protect the abilities both to build shared reality and to generate deep critique. The

186 | Free Speech and Truth

identification of a speech practice as belonging to the domain of community or democracy might be the beginning of such analysis, but it would rarely, if ever, be conclusive. So, for example, a relational truth theory offers a particular and more limited understanding of what many have seen as the bedrock principle of the First Amendment: the idea that government may not silence speech simply because it disagrees with it.39 Such a basis for regulating speech is directly in conflict with the goals of relational truth, which include seeking deep critique of present ideas and practices. Thus, the simple claim that the government disagrees with the speech would always be an illegitimate basis for regulation because it is as if the government said that it was regulating the speech precisely to prevent critique. Such a government purpose is a violation of the First Amendment when free speech is justified by a relational theory of truth. Of course, the government almost never admits that it is regulating speech simply because it disagrees with it. Sometimes the circumstances of the regulation make it plain that, despite the government’s protestations to the contrary, this is the purpose and, in that situation, the Court rightly strikes down the legislation on that basis alone.40 More often, however, the government claims, quite plausibly, that it is regulating the speech for reasons that are logically distinct from disagreement with it, such as the harm it may cause to listeners or others. Here too, the general suspicion of government motives still has a place. It is plain that there is an inherent bias in government when it comes to the goals of shared reality and deep critique: while government can generally be trusted not to attempt to silence speech because it supports a shared reality, the case is quite different with deep critique. As Lee Bollinger explained in his important book on the First Amendment, there is a profound impulse to silence deep critique and the First Amendment should be seen as a response to this impulse.41 The response, however, is designed to promote not just tolerance generally, but the more specific goal of truth, understood in a relational way. As a result, the response to this impulse must always take account of the need for both shared reality and deep critique. Thus, where the government claims that it needs to silence speech in order to protect the capacity to generate a shared reality, this is not an example of the inherently illegitimate motive described above. We should regard such claims with some suspicion, because of the bias against critique, but we should not see it as the equivalent of the dislike and disagreement that is the gravamen of an unconstitutional purpose.

Free Speech and Truth | 187

The difference here turns on the distinction between protecting a particular shared reality that the government presently supports, on the one hand, and protecting the mechanisms through which our shared reality is reinterpreted and negotiated, on the other hand. The government has no legitimate interest in protecting a particular shared reality because the content of our shared reality does (and should) change constantly. Thus a government purpose to silence speech because it is critical of present policy or in violation of present social mores is unconstitutional. But the government does have a legitimate, indeed a compelling, interest in protecting the mechanisms through which shared reality is constructed and reinterpreted. This is what it means to say that the First Amendment protects speech because it contributes to relational truth. Thus, a government purpose to regulate speech in order to preserve the process of building shared reality and deep critique—for example, by preventing monopolies in the communications industries, or by regulating campaign contributions—is not presumptively illegitimate. The courts will, of course, still need to carefully examine the means chosen, but the purpose itself is consistent with a relational truth theory of the First Amendment. The difficulty is that protecting a particular shared reality is not, in the end, analytically distinct from protecting the process of constructing a shared reality. What is necessary for that process is itself a subject on which we have present views, which may themselves be contested and change over time. As a result, it will often be possible to characterize the government purpose at issue in either way. The distinction, then, is not a logical one, but a contextual one: which description of the government’s purpose is most plausible under the circumstances? Despite the indeterminacy of this analysis, it does make a difference if we frame the constitutional issue in this way. This approach suggests that when the government purpose is most plausibly understood, in context, as protecting the process, it should not be subject to the automatic invalidity that properly applies to government efforts to silence speech with which the state disagrees.42 The need to qualify this central First Amendment principle in this way arises because of the relationship between shared reality and deep critique. The traditional formulation of the principle sees only the tension between shared reality and deep critique: it casts the First Amendment in the role of protecting the latter as against a government determined to promote the former. While the tension is real, this view fails to recognize the equally important interdependence of shared reality and

188 | Free Speech and Truth

deep critique. Deep critique is not possible in the absence of a rich and meaningful shared reality: we must share a great deal in terms of our conceptual vocabulary and our practical understandings in order for criticisms of our cultural categories or values to be effective. Conversely (and more rarely recognized), shared reality depends upon the possibility and regular practice of deep critique. As John Stuart Mill so eloquently expressed it, the received opinion becomes “dead dogma” if it is not regularly challenged and defended.43 Thus, an understanding of the relationship between shared reality and deep critique leads us to a series of paradoxes: to refuse government the power to support the foundational aspects of shared reality may simultaneously protect critique and also undermine it. To refuse protection to deep criticism may help to preserve shared reality but also to weaken it. A relational understanding of truth suggests that the First Amendment’s protection for freedom of speech must be seen not simply as a weight on one side of this delicate equation, but as an effort to strike the sort of balance that facilitates both shared reality and deep critique. The free speech principle is not compromised by this sort of balancing; rather, the principle—when understood in terms of relational truth—has the balancing built into its foundation. A final word about the understanding of the free speech principle that emerges from consideration of the shared reality and deep critique functions of truth: this understanding raises powerfully the issue of moral remainders. Because the free speech principle calls for this balancing, it will often (perhaps always) involve some sacrifice of the very values it is intended to protect: either shared reality, or deep critique, or both. It will often (perhaps always) be the case that in whatever balance is struck, some of what we need, by way of shared reality or deep critique, is lost, even if the overall outcome is the best that we can manage under the circumstances. As a result, the compromises that we strike in First Amendment doctrine may leave moral remainders: moral demands that create pressures on the interpretation of the First Amendment in other contexts or on the interpretation of other constitutional provisions. The need for shared reality or deep critique that has been sacrificed here does not simply disappear; it must, instead, be acknowledged and accommodated elsewhere. Thus, I have argued elsewhere that if we interpret the First Amendment to protect hate speech in certain circumstances (for example, between adults in a public forum), then we must interpret the equal protection clause to require more active steps to redress the inequalities that make hate speech so damaging. The moral cost of hate speech to its tar-

Free Speech and Truth | 189

gets, generated by this interpretation of the First Amendment, creates a demand that must be satisfied through other aspects of constitutional theory.44 An understanding of the free speech principle based on relational truth builds tension, compromise, and the possibility of such moral debt into the idea of free speech itself.

B. “Working” The relational truth model’s emphasis on “working” requires that we open our own perspective to challenge by others who may be harmed by our choices. Caring about “working” means caring about the impact of our policies on other subjects who count. The way we facilitate the consideration of “working” is by opening up the possibility for those others to challenge us on moral and political grounds. Thus, central to the “working” function of truth claims is, first, the acceptance of a responsibility to make oneself vulnerable to such challenges45 and, second, the promotion of opportunities for challenge by those less likely to otherwise be heard. Speech, both as a relationship between individuals and as a collection of social systems, is central to both of these requirements. When seen at the microlevel, the social practices of speech—both for the speaker and for the listener—involve this sort of vulnerability. To engage in a dialogue is to make a commitment to a relationship in which one’s own identity and values are at stake. On a relational understanding of truth, it is through this process of interaction that our identities and values are formed and reconstituted: when we seek knowledge through dialogue with others, we open ourselves to their challenges. Obviously, people often find such challenge threatening and seek to avoid it. This avoidance can take the form of refusing to engage in the dialogue at all, but it can also take the form of particular dialogic maneuvers designed to derail the challenge (such as undermining the credibility of one’s challengers, talking over them, or whipping up one’s anger against them). Not all speech, then, is conducive to the production of relational truth. Only when we adopt a particular attitude of responsibility toward each other in the conversation does speech have this potential to make us vulnerable to challenge. Speech is, of course, not the only way in which we can make our identities and values vulnerable to others. There are many forms of vulnerability and many of them, other than speech, may have deep consequences

190 | Free Speech and Truth

for identity. An abused spouse, for example, may have her sense of self and her values profoundly affected by the experience of vulnerability to abuse.46 But speech is special. For most of these other sorts of vulnerability one must either establish an ongoing relationship with the person to be changed (as in the intimate association situation) or make a substantial change in that person’s life circumstances (for example, a total stranger can make a person vulnerable to changes in identity by raping her).47 Engagement in the social practice of speech can, however, make one vulnerable to changes in identity caused by people whom one has never met and who have altered nothing in the concrete circumstances of one’s life. Speech extends our vulnerability in a dramatic way, making us subject to challenge by people who would otherwise find us unreachable because of social and/or physical distance. By using examples such as spousal abuse and rape, I hope to highlight the fact that such vulnerability is dangerous, not entirely within our control, and not always a good thing. Relationships are the ground of our identity and our reality, but they are not thereby guaranteed to be positive: they can be destructive as well as productive; indeed, they can be both simultaneously.48 The demand for vulnerability is not based on some pollyanna-ish assumption that the vulnerability will always be harmless. Rather, the demand is based on the conviction that what safety is possible for us is to be had only within relationships—relationships marked by a certain form of responsibility—rather than in the retreat from them. Speech, of course, usually poses less danger than many other sorts of vulnerability, but even speech can be used in ways that are damaging and, where the necessary responsibility is missing, relational truth is not promoted by the speech. In particular, where the vulnerability on one side of a speech relationship is not matched by respect and reciprocal vulnerability on the other, there is a great likelihood that the relationship will be unable to serve the functions of truth. Speech promotes the functions of a relational model of truth only if it generates the types of relationships that allow for those functions to take place. Thus, the measure of protection for free speech in this model will be the relationships it generates. Speech, then, while far from harmless, represents the only real opportunity for vulnerability to all of the members of the community that must build truth together. To make that opportunity real requires that one engage in the speech process with the necessary attitude of responsibility. But the attitude alone would be insufficient without the opportunity for actual challenge presented by speech.

Free Speech and Truth | 191

The other requirement for facilitating the “working” function of truth claims is, therefore, to provide opportunities for challenge by those who wish to make such claims. The need to open up opportunities for challenges about “working” suggests that we must scrutinize our systems of speech at the macrolevel for their openness to a broad range of those affected by public policy, particularly those who might not otherwise be heard. From a relational truth perspective, Steven Shiffrin’s marvelous book, Dissent, Injustice, and the Meanings of America,49 offers a view of what the First Amendment might mean if we took seriously the need to protect the opportunity for people to make such challenges. Shiffrin argues that dissent plays a special role in our speech practices. He suggests that dissent is particularly important in our struggles over the meaning of America and in combating injustice.50 From the perspective of relational truth, the first of these functions appears to be a description of deep cultural critique, while the second function appears to concern the types of claims that I have called “working” claims: where people assert that our social policies or practices are taking inadequate account of the impact on subjects who matter. Working claims are, in other words, the category of dissent that focuses on injustice. Shiffrin argues that injustice is endemic in large-scale societies: it is simply a result of the self-perpetuating nature of the hierarchies that are inevitable in a complex society.51 Consequently, social institutions, both public and private, display a consistent inclination to stifle rather than encourage dissent. From the perspective of relational truth, injustice is built into the system of knowledge itself because our knowledge claims will reflect the perspectives of those with the power to impose their interpretations.52 Thus, injustice born of power differences may be simply ineradicable within societies, even democratic ones, but that conclusion suggests certain moral consequences. As Jane Mansbridge writes, “when we compromise with justice, we must design our lives and our institutions so that the justice that is compromised remains nagging in the margin somewhere, in a bracket that does not go away, to pique our souls and goad us into future action.”53 Shiffrin offers a detailed account of the forces—economic, cultural, and legal—that lead the mass media in our own society to stifle dissent.54 He concludes that such forces have undermined the opportunity for dissent to such a degree that it is necessary for government to undertake “an affirmative obligation to assure that the society encourages dissent against injustice whether in public or private hierarchies.”55 As this

192 | Free Speech and Truth

quotation suggests, Shiffrin largely rejects the public/private distinction on which Post insists. I think that this difference arises, in part, from Shiffrin’s attention to the centrality of “working” claims, while Post is focused more exclusively on the dialectic between shared reality and deep critique. The power differentials in which “working” claims have their genesis arise in private as well as public hierarchies. In order to effectively challenge them, to make the power holders vulnerable to these claims, the claims must sometimes be made within the institutions where the power is exercised. Critique could, of course, go forward in the realm of public discourse, and that process of public debate may often be helpful in redressing injustice, but the heart of a working claim is the direct challenge that demands vulnerability of the power holder. If such claims are important to relational truth, as I have argued that they are, then they must be accommodated, indeed, encouraged, within the broad range of institutions that generate such hierarchies. This does not mean, of course, that the First Amendment itself generates legal claims for freedom of speech against private institutions. The amendment is addressed to government and has consistently been interpreted to require “state action” as a basis for a constitutional claim.56 However, the fact that the same truth issues concerning “working” arise in private institutions as in public ones suggests that the First Amendment should be interpreted to allow government to regulate such institutions in the interest of promoting working claims. This is an instance where the amendment does not itself provide a legal remedy to a speaker, but where it supports rather than inhibits government action that does so. Another result of this analysis of the “working” function of truth is to undermine the separation of the last of Post’s three domains: management. Post describes management as an organization whose “boundaries can be recognized by the predominance of functionally defined organizational roles.”57 “[T]he distinction between management and governance turns on the priority accorded to proposed objectives” within the domain of management.58 He is clearly right that certain government institutions must be seen as engaged in an activity distinct from governance; such institutions are intended to serve a particular purpose and speech within them often can be regulated in the interest of that purpose.59 But the attention to working claims suggests that those holding power in such institutions must make themselves vulnerable to claims of injustice by other members and even by outsiders affected by the institution. Such institutions, in other words, are never only in the domain of management; they

Free Speech and Truth | 193

are always subject to claims of justice and not just to claims of instrumental rationality. The Supreme Court has often adopted a very deferential stance when reviewing regulations restricting speech within such “managerial” institutions.60 Such deference is highly problematic from a relational truth perspective. It is in the nature of power hierarchies to overestimate the need for their own control and underestimate the harms done to those subject to that control. For that reason, deference toward the people holding power in such institutions is generally less appropriate as the institutions are more hierarchical.61 From the perspective of relational truth, managerial institutions must be open to the types of claims that generate truth. Forms of institutional organization that preclude such claims entirely are inconsistent with the First Amendment commitment to truth. Thus, the First Amendment should have something to say about how closed to challenge institutions in our society may be. This view of the importance of working claims has consequences both for the issue of institutional control over the speech of members and for the issue of institutional control over resources sought by outsiders. With respect to control over the speech of their own members, such institutions must have sufficient control to carry out their functions, but they should generally tolerate the greatest amount of free speech by members consistent with that goal. In particular, they must tolerate the types of speech through which members make working claims concerning the exercise of power in the institution. So, for example, the school board may determine the textbooks to be used by a teacher, but may not fire a teacher for publishing an editorial in a local paper criticizing school board policy.62 Similarly, an attorney general may insist that his subordinates make certain arguments in court (for example, requesting the death penalty in certain cases), but he may not prevent them from circulating petitions challenging his transfer policy.63 For many people this argument will immediately bring to mind the military and its claim that unquestioning obedience to hierarchy is necessary to its functioning. A relational truth theory that sees the need for “working” claims suggests that institutional claims of this kind should be analyzed in the following way. Given the importance of “working” claims, a public institution should be allowed to be organized in ways that intentionally stifle such claims only if such organization is the least restrictive means of serving a compelling state interest. In other words, it must meet the highest level of constitutional scrutiny. Defense of the

194 | Free Speech and Truth

nation is undoubtedly such a compelling state interest; the interesting question here, of course, is whether or not the type of hierarchy presently characterizing the military is necessary to it. I suspect that this question could not be answered in a truly global way and would need to be assessed for particular contexts within military organizations, some of which might be consistent with greater amounts of dissent than others. For example, rules concerning “working” claims might be very different for speech during actual military operations (or even during training for them) than for speech during off-duty time or in recreational settings. In the “public forum” cases, a relational truth theory suggests that a publicly owned resource should be open to use by outsiders wishing to speak unless it would conflict with the primary purpose to which the resource is committed. Thus, public airports should be open to speech not inconsistent with their transportation uses,64 and public areas on military bases should be open to speech consistent with their particular uses.65 Decisions by managerial institutions to stifle speech should, in other words, generally be subject to an incompatibility analysis, as suggested in the Grayned66 case, rather than to the deferential approach presently taken by the Supreme Court in many such cases. As I suggested above, the goal of relational truth also suggests that government might seek to facilitate “working” claims within private institutions as well. Here again, readers may immediately wonder about a particular case: whether the state could insist that a religious institution open itself to internal dissent. This issue would turn on whether such a regulation interferes with another fundamental right. In some cases, government regulations requiring private groups to be open to dissent might well infringe upon the free exercise of religion.67 Intimate associational rights might also provide protection from government regulation for certain types of groups.68 But what are currently seen as “First Amendment” associational rights would not provide such protection where the purpose of the regulation is itself to promote the types of “working” claims that serve the functions of relational truth. This argument suggests that the Supreme Court was wrong in simply accepting the First Amendment associational claims of the Boy Scouts to revoke the membership of an openly gay adult scout.69 The civil rights laws represent a judgment that working claims by certain classes of persons have too often been suppressed or ignored. Expressive associations must therefore leave themselves open to such claims. If the goal of free speech is relational truth,

Free Speech and Truth | 195

this judgment promotes the values of the First Amendment rather than contradicting them and should not be unconstitutional. As a general matter, the First Amendment should be taken to endorse government efforts to protect “working” claims. In order to encourage dissent, Shiffrin suggests, a society must meet four conditions. First, it must have an educational system that promotes attitudes and teaches skills necessary to create citizens with the will and ability to challenge injustice. Second, it must open the channels of communication for expressing dissent. Third, it must keep legal barriers to dissent to a minimum. And fourth, it must arrange social and governmental institutions in such a way as to provide information to would-be dissenters.70 Shiffrin offers some concrete suggestions for the achievement of these goals, but he also recognizes that there are many possible ways to envision and pursue these conditions. The important point, for his argument and also for mine, is that the First Amendment should be seen as calling upon us to struggle with these possibilities, rather than as foreclosing them, as it does under some present doctrine.71 In light of the inevitable pressures toward injustice, we are not making ourselves vulnerable unless we are opening up channels for the formulation and expression of “working” claims in the way that Shiffrin describes. The strength of our systems of speech in facilitating these claims is essential to their ability to generate relational truth.

C. Connection to Nonhuman Reality This final function of truth claims is also often fulfilled by speech. While people connect to a larger, nonhuman reality through a variety of mechanisms, many of them involve speech that is intended either to describe or to induce that connection. Worship services, theological debate, and evangelism are all forms of speech that serve this function. One important point to be made here is about the overlap between the different functions of truth served by speech: much of the speech that serves this final function also serves the other functions of truth as well. Such speech may create a shared reality within a particular community (where people experience the inbreaking of the Holy Spirit); suggest a deep critique of some aspect of the majority culture (such as a critique of materialism from the perspective of an emphasis on spiritual life); or challenge social policies or practices on the claim that we are ignoring their

196 | Free Speech and Truth

impact on subjects who should matter to us (such as fetuses, or ecosystems). To recognize this final function of truth is, however, to insist that the free speech clause protects speech used to explore or express the connection to a larger reality even when that speech does not also serve one of the other functions of truth. So, for example, individual prayer is protected speech. Obviously, this function of truth is closely connected to the religion clauses of the First Amendment as well as the free speech clause. William Marshall has written a very interesting article suggesting that the search for truth is the underlying commitment that ties the speech and religion clauses together.72 The notion of truth he defends is closely related to the function I am discussing here: “the essential religious question of understanding one’s place in the universe is indivisible from the question of what is truth.”73 While Marshall insists that we need a transcendent model of truth to address these questions,74 he nonetheless seeks to justify this search for truth in terms of myth and narrative rather than in terms of a Cartesian correspondence theory.75 He argues that the validity of the search for truth as a justification “does not depend upon its grounding in a metaphysical reality. Instead, its validity depends upon its explanatory power and force.”76 And that power, in turn, rests on the fact that “the search for truth . . . responds to an essential human need.”77 While Marshall and I may disagree about the sense in which truth needs to be transcendent,78 I mean to be endorsing his insight that a particular aspect of truth can be seen at the root of both the free speech clause and the religion clauses. The search for understanding, for its own sake, is an important part of what we mean by truth and our lives would be poorer without it. The practices of both speech and religion serve as vehicles for such understanding, and that role is one source of their value. The connection between this function of truth and speech is, again, more than merely happenstance. People can and do experience this largescale understanding in nonconceptual ways, but the distinctive contribution of speech is that it provides the conceptual and symbolic frames for such experiences. Speech gives us the categories that enable us to recognize our experiences and the symbols that trigger them. When and if the search for this truth occurs through activities that do not involve speech, it should generally be protected under the religion clauses; but in the majority of cases, where speech is used, it should be protected under the free speech clause as well.

Free Speech and Truth | 197

As with the other functions of truth, a relational model asks us to consider the contexts and relationships within which this function takes place. When we look at the macrolevel, we find that this final function of truth is very often pursued through institutions and communities: religious, educational, and otherwise. One of the interesting implications of a relational approach to truth is that it suggests that such institutions and communities must be protected in order to facilitate this function of truth, in much the same way that access to mass communications must be protected in order to facilitate the critique and working functions. Thus, the Supreme Court’s understanding that associational rights can be based in First Amendment activities accurately reflects the need to protect the relationships through which the functions of truth are enacted.79

D. Conclusion Feminists and others who find the Cartesian tradition in epistemology problematic should, nonetheless, refuse to give up on a truth theory of free speech. We need truth for good and powerful reasons, and a relational model of truth can accommodate these important functions while retaining a close connection to the themes and commitments of feminism. Indeed, the relational model of truth can help us to see that there is a reason why the truth theory of free speech has such intuitive appeal: it captures our sense that the search for truth is crucial to our relationship to each other and to the world and that speech is the primary practice through which we seek truth. Our values are deeply implicated in our truth claims. We can only hope to pursue truth if we recognize and self-consciously address the moral and political assumptions that guide the process of knowledge creation. Speech is a uniquely valuable mechanism for such reflection because it both relies upon and (hopefully) actualizes the respect, responsibility, and reason that such reflection requires. Truth does and should matter to us; speech really is special when it comes to pursuing truth. Because truth fulfills such important needs, it can be a powerful engine for mobilizing social and political action. We should not allow those who would defend the existing hierarchies to hold a monopoly on truth. Too often, truth has been used to assuage the longing for certainty,80 a longing that sometimes blinds us to the demands of respect, responsibility, and

198 | Free Speech and Truth

reason. But “truth” is a word to conjure with,81 and its magic is not entirely a product of such illegitimate longings. We can, and should, reclaim truth as the symbol for our abiding commitment to live together—with other people and with the nonhuman world—in a way that meets our moral standards. Truth is the ultimate assertion that we are not alone and that our destiny, like our reality, must be created together. If we came to understand truth in this way, then the free speech clause of the First Amendment could become the repository of this commitment, rather than the fortress that protects us from our Cartesian anxieties.

8 Free Speech and Narrative Autonomy

In this chapter, I will begin the process of developing an alternative autonomy theory of free speech based on the narrative model of autonomy outlined in chapter 6. I will focus here on laying the theoretical foundations for this alternative theory, but I will mention implications for particular legal doctrine in passing. As in the previous chapter, the theory offered here would often allow for a variety of doctrinal formulae on a given issue. Like the traditional theories of free speech, it directs our attention to particular concerns and frames the issues, but it does not necessarily compel particular answers to our doctrinal questions. The purpose of this chapter is to explain how this theory might frame the issues in ways that highlight important aspects of speech that are cast into shadow by present doctrine and theory. If we adopted a narrative model of autonomy, why might that lead us to a principle calling for the protection of freedom of speech? Free speech is deserving of protection because it contributes to narrative autonomy in at least three ways. First, many speech acts would themselves be exercises of narrative autonomy: telling one’s story in the broadest sense. Second, some speech acts or systems of speech (such as the structure of the broadcasting market or the contents of school libraries) affect autonomy in the aggregate, for many people. They may have this effect on many people either by increasing or decreasing the opportunities for the exercise of narrative autonomy or by helping or hindering the development of the capacities needed for narrative autonomy. Finally, narrative autonomy connects to speech in a third way: narrative autonomy forms the normative foundation for social relationships and, in so doing, provides the basis for an argument about the relationship between speech and democracy. I will

199

200 | Free Speech and Narrative Autonomy

examine each of these three roles of speech in turn and explore their implications for the shape and purpose of the protection for free speech.

A. Individual Speech as an Exercise of Narrative Autonomy This first connection between speech and autonomy—in which an individual’s speech act is itself an exercise of narrative autonomy—is helpful in responding to one of the traditional criticisms of autonomy theories of free speech. The criticism is that autonomy theories do not seem to offer a principled means for distinguishing speech from other behaviors through which autonomy might be exercised and that do not receive the extraordinary protection accorded to speech. When autonomy is understood as choice, this criticism is quite powerful: choice can be exercised through any sort of activity, so why are exercises of autonomy through speech entitled to special protection not available to other autonomous choices?1 A narrative model of autonomy, however, offers a response to this criticism. In a narrative model, the central activity in autonomy is not choosing—which could pertain to any action—but telling one’s story, which is a matter of using a symbolic system (usually words) to create meaning. While it is certainly not the only way, speaking is undoubtedly the single most important way that we tell our stories. Thus, focusing protection on speech makes more sense when we see autonomy in narrative terms than when we see it in terms of choice. This argument immediately raises two issues. First, what about occasions when we tell our stories through means other than words? That is, what about symbolic speech cases? Second, what about exercises of autonomy that involve not telling our story, but living it? Why should protection for autonomy be limited to those activities through which we tell our stories rather than extended to the actions through which we live out the stories we have told? I will respond to these concerns in turn and then return to a more general consideration of the nature of a free speech right based on speech as an exercise of narrative autonomy.

1. Symbolic Speech If the protection for speech arises from its role as a mechanism for the exercise of narrative autonomy, then symbolic activities that play the same

Free Speech and Narrative Autonomy | 201

role should get the same protection. One way to see this is simply to recognize that speech (here understood as verbal behavior), while the most common method of telling one’s story, is by no means the only method. When we do use things other than words as symbols, for example, by burning a flag, then that also counts as telling our story and ought to be protected in the same way as more conventional speech because it is the same sort of exercise of narrative autonomy. The constitutional protection for “speech” should then be understood not as limited to verbal behavior, but rather as identifying a more general category of symbolic meaning-making activity (of which speech—conventionally understood, as verbal behavior—is the most common example). When the behavior in question is verbal, there is a strong presumption that it falls into this category.2 When the behavior is nonverbal, the presumption goes the other way, but it can be overcome by a showing that the behavior is in fact an example of “symbolic speech.” The Supreme Court’s present test for symbolic speech requires that the speaker intends to communicate a particular message and that the audience is likely to understand that message.3 The Spence test has been the subject of widespread criticism by scholars.4 One criticism argues that the test requires too much: it requires an intent to communicate a particularized message and thus would exclude from First Amendment protection large areas of art, which lack such a specific message but which most people think should be protected nonetheless.5 A second criticism argues that the test does not require enough: routine vandalism or traffic violations should not create an entitlement to First Amendment review whenever the perpetrator can claim that he or she was intending to express a meaning that the audience would have understood.6 If we examine the Spence test from a narrative autonomy perspective, however, we find that its basic approach is good, although it stands in need of some modification. The first criticism is correct in pointing out that this test is too focused on messages. This flaw arises from the test’s foundation in a Cartesian truth theory approach to speech: the test was designed to protect speech because—and only if—its message might lead us to the truth. In order to eliminate this problem, we must modify the test to require that the speaker intends to convey meaning and the audience would be likely to understand the act as an effort to convey meaning (even if, perhaps, they are uncertain or in disagreement about precisely what the meaning is).

202 | Free Speech and Narrative Autonomy

The Spence test’s focus on a speaker’s intent needs to be broadened rather than abandoned. An intent to convey meaning does not require that the meaning be capable of formulation in propositional form. It does not require that the symbolic speech take the form of an argument or that it appeal to reason rather than to other faculties, such as emotion or imagination. An intent to create meaning is, however, more than merely an intent to change the physical world in some direct way. Such an intent must include the purpose of contributing to the way people (oneself and perhaps others) understand something, where understanding includes emotional, imaginative, and other elements along with rational ones. This intent is what distinguishes destroying files at a nuclear weapons plant by pouring blood on them from destroying them secretly to hide evidence of embezzlement. Several aspects of a situation signal this intent, although no simple formula captures it and no single element is always necessary for it. Some relevant issues include the public nature of the act, the use of commonly understood symbols, and the connection of the act to the agent’s deep commitments. The central question is whether the act is valued, at least in part, because of its symbolic meaning (to the speaker and/or others) and not only for its physical effects in the world.7 It is this intent that must be present (although it need not be the only intention) in order for an activity to qualify as symbolic speech. Second, the requirement of audience understanding must be interpreted broadly in order to avoid wooden application to cases like the vandalism or speeding examples. The requirement of potential audience understanding should not be applied mechanically. It does not mean, for example, that there must be an actual audience; the Spence test should apply even to one whose speech is unheard by others. Thus, if O’Brien had burned his draft card alone in his basement, rather than on the steps of a public building, he should still have been able to claim that his act was speech within the meaning of a narrative autonomy theory.8 The audience understanding requirement instead captures the idea that our meanings must be made through dialogue—real or imaginary—with others.9 As a result, the issue of audience understanding should be interpreted to incorporate consideration of all of the contextual factors that affect whether others would understand our actions as meaningful. Robert Post has argued, for example, that the social practice of which the act is a part should be an important, perhaps controlling, factor in determining whether the act qualifies for First Amendment protection.10 I agree that such consideration is crucial, but I believe that the reference to

Free Speech and Narrative Autonomy | 203

audience understanding—when interpreted in light of a relational model of autonomy—can and should include such consideration. If the act is part of a practice that is normally understood to be expressive, then it is far more likely that the audience will understand the act at issue in that light. What difference is there, then, between Post’s approach and mine? The overlap between potential audience understanding that the activity is meaning creating, on the one hand, and the activity being part of a social practice of speech, on the other hand, is not perfect. A person might engage in a social practice generally understood as expressive, but do so in such a way that the likely audience would believe that he was not actually interested in the creation of meaning, but merely using this practice as a convenient way to achieve other ends. For example, I remember one of my professors in law school commenting that he stopped taking student protests seriously when he realized that they always seemed to occur on the first nice day in spring when no one wanted to be in class anyway. I expect that Post and I would agree that such suspicions about the motives of those who are using established social practices for speech should not take the speech (symbolic or otherwise) out of the protection of the First Amendment. While such suspicions are a necessary part of the analysis of symbolic speech where it is not part of an established social practice (e.g., the destruction of files), they should not be used to refuse protection to a particular instance of a symbolic activity that is generally recognized as expressive of meaning. This conclusion is a natural result of Post’s focus on the larger social practice of which the speech is a part. For me, the significant fact is that the audience in such a case understands the act as meaning creating, even if it believes that to be less important to the speaker than some other aspects of the activity. Either way, however, I believe we are in agreement about this case. The difference between these approaches, then, largely reduces to the possibility that someone might engage in an act so clearly expressive that the audience would see it as such even if it is part of a social practice not normally understood as expressive. Indeed, this is precisely the process through which social practices of symbolic expression are created and recreated (for example, people simply sitting in a place—which was not normally understood as a speech practice—became a “sit in”). We must protect such activities as speech in order to keep open the process of generating new speech practices. Even where the symbolic use of the activity is unusual, moreover, and gives rise to no such general social practice, if

204 | Free Speech and Narrative Autonomy

the meaning-making aspect is clear enough to be understood by a potential audience, then it should be protected as speech.

2. Living One’s Story The second issue I must address is why protection is limited to the process of telling one’s story rather than the more general process of living that story.11 Here I need to return to the discussion of the relationship between autonomy and freedom in chapter 6. Living one’s story is a good way of capturing the broad notion of freedom, of which autonomy is one part. In order to live one’s own story, one must be autonomous in the sense of being able to formulate a story that is properly called one’s own. But autonomy, although necessary, is not sufficient to freedom. To live that story, one also needs negative freedom (in the form of absence of coercion and constraint) and positive freedom (in the form of support for the opportunities and capacities that are required to live one’s story). Freedom, understood in this broad sense, is necessarily limited. No one ever operates under conditions of perfect liberty, negative or positive. Freedom, then, is always a matter of degree: we ask for an absence of constraint and a presence of support for a range of life options measured not only by their sheer number but also by their importance. We are more free when we have a greater ability to live out the more important parts of our stories. This broad freedom is protected through a variety of constitutional guarantees, of which the First Amendment protection for freedom of speech is only one. The Fourteenth Amendment’s protection of liberty through the doctrine of substantive due process, for example, can be understood as an effort to guarantee at least negative liberty for a particularly important collection of relationships through which we live out our stories. The religion clauses of the First Amendment can be understood in the same way. The distinctive aspect of the speech clause, when seen from this perspective, is that it is focused on protecting the autonomy part of this freedom: the part that involves formulating, expressing, and reinterpreting one’s story. This view of the relationship between autonomy, freedom, and the free speech guarantee has several implications. First, it suggests that many important freedoms will find their protection under parts of the Constitution other than the free speech clause. The fact that a particular activity does not involve the narrative autonomy that would qualify it for First

Free Speech and Narrative Autonomy | 205

Amendment protection does not mean that it will be left to the mercy of legislative majorities or executive policy. The free speech clause is only one part of a large and complex system of constitutional protection for freedom. Second, the relationship between autonomy and the other aspects of freedom is complicated. While it is true that one can be autonomous while suffering from a pronounced lack of both negative and positive freedom; it is also true that certain systematic denials of freedom may themselves limit one’s autonomy by interfering with the capacities or resources necessary to autonomy. So, for example, a denial of educational opportunities on the basis of race or gender is not only an interference with positive (and perhaps negative) freedom; it is also a limitation on autonomy because it can affect the very ability to formulate and reinterpret one’s story in the first place. While the First Amendment does not of itself protect activities such as going to school, it may, therefore, have implications for the way we assess restrictions on such activities under other provisions of the Constitution. Where a restriction on liberty (negative or positive) has the effect of substantially interfering with the capacities or resources necessary to autonomy, it should provide a powerful argument that the liberty is protected under some other clause of the Constitution. The First Amendment’s concern for autonomy ought to inform our understanding of other constitutional provisions, particularly the scope of the liberty interests protected under substantive due process and equal protection.12 One objection that might be made to this approach to autonomy is that it appears to require that we be able to distinguish between exercises of narrative autonomy and other aspects of freedom. In other words, if only narrative autonomy is protected, then wouldn’t we need to be able to tell whether someone is telling her story through a particular speech act or merely living it? The short answer is no. Speech activities often serve a great variety of purposes, but telling one’s story is almost always one of them. The symbolic aspects of speech guarantee that meaning making is almost always one of the functions of speech activities. Verbal activities can be assumed to be speech in this symbolic, meaning-making sense, regardless of the other functions they may also serve. Nonverbal activities may also have such symbolic aspects, and would then also qualify as speech. What we need to determine, in other words, is whether the activity at issue is speech; if it is speech (i.e., symbolic meaning making), then it can be assumed to be an exercise of narrative autonomy.

206 | Free Speech and Narrative Autonomy

This assumption holds true even if the content of the speech does not explicitly concern the speaker’s stories. The assumption rests on the connection between symbolic meaning making in general and narrative autonomy and highlights again the blending of epistemological and moral/political issues in a relational approach. Meaning refers “to a social realm that provides the basis for people to understand texts and other objects created by humans.”13 Given that, from a social constructionist perspective, all aspects of the world of our experience are, at least in part, created by humans, meaning in any context is social in this sense. Moreover, the process of making meaning is never separate from the process of making ourselves.14 Our identities are implicated in and transformed by the process of interpretation through which we make meaning.15 Thus, the search for relational truth and the process of narrative autonomy are carried out through the same activities. In a relational model, truth and autonomy are interdependent concepts: they are two different ways of understanding the process of making meaning, a process that is always both epistemological and normative. Thus, activities of symbolic meaning making—that is, speech activities—are prima facie exercises of narrative autonomy regardless of whether or not their content facially concerns the speaker’s story. The effort to understand ourselves and the effort to understand the world cannot be separated. It is the effort to understand that is crucial, to truth, to autonomy, and to speech.

3. The Nature of the Right From the perspective of individual acts of speech and with respect to this first argument—in which speech is protected because it is an exercise of autonomy—a narrative model will generate a free speech right that is not very different from the one generated by a choice-based, liberal model of autonomy. In both models, the emphasis in this argument is on the speaker and the use of speech as a vehicle for autonomy. The right in both cases could take positive or negative liberty forms, depending upon one’s understanding of the extent of government obligation to make the practice (of choice or of narrative autonomy, depending upon your theory) efficacious. The right in both cases could belong either to individuals or to groups engaged in the protected practice. The difference between seeing the speech as a product of choice and seeing it as a process of creative interpretive narration does, however, lead to a few significant differences in the shape of the legal right even from an

Free Speech and Narrative Autonomy | 207

individual perspective. For example, commercial speech would be protected under a narrative autonomy theory, while it is unprotected under the most prominent choice-based autonomy theory. More generally, and more significantly, however, a relational autonomy theory (like the narrative one) suggests that the individual perspective is not the only, or even necessarily the most important, one. In a relational model, in which the process of autonomy requires certain social relationships and certain social conditions, protecting speech as an exercise of autonomy requires us to protect those relations and conditions that make such an exercise possible. In other words, we must look at the speech systems that create the opportunities for narrative autonomy for people generally, and not only at the particular speech acts engaged in by individuals. I will briefly explore the differences between a narrative approach and a choice-based approach at the individual level in this section, and then consider the more systemic level of analysis opened up by a relational model of autonomy in the next section. At the level of individual speech acts, choice-based and narrative autonomy often look very much the same. Nonetheless, there are some differences in the ways the two models would structure the speech right. For example, there is a difference on the issue of commercial speech. Professor Baker has argued that commercial advertising should not be regarded as protected speech under his liberty theory because such speech is not the product of choice. He sees advertising as responding to market pressures and forces and not as an expression of the substantively valued commitments of the speaker.16 This argument relies upon a particular view of the inexorability of market pressures on economic enterprises and requires that we distinguish that realm from others in which, despite many social influences, individual choice remains meaningful.17 Thus, in advertising, it is the coercive effect of the market that makes the speech “unchosen” and therefore unprotected. From the perspective of narrative autonomy, however, it is not the breadth or narrowness of choice left open by social pressures (market driven or otherwise) that determines whether speech is protected. We all operate under such pressures at all times and they interact in ways that are complex and highly individual in determining the range of choices open to us. Some people obviously do engage in commercial speech that represents their value commitments, even if that speech is not the most economically profitable. Some people also engage in noncommercial forms of speech simply in response to social pressures without experiencing

208 | Free Speech and Narrative Autonomy

much choice in the matter. Thus, market pressure is not unique in the quality of coercion or the degree to which it forecloses choice of speech. And, more importantly and more generally, choice under conditions of pressure and influence is too much a matter of degree and of individual response to provide a coherent basis for determining whether or not speech is protected. From the perspective of a narrative model of autonomy, the hallmark of autonomous speech is not that it was chosen (in the sense that it was uncoerced by social forces), but that it is the product of a process of creative reflection, including normative reflection, on the identity and commitments of the speaker. Even a speaker operating under severely constrained choices can be narratively autonomous if she responds with creative agency. By this standard, it is possible for advertising to be autonomous, in the sense that it is a creative response to market pressures that seeks to express the values of the speaker. Indeed, it might well be the case that such autonomous advertising could be encouraged and supported by particular structures in the advertising industry and corporate world. As the next subsection will suggest, the First Amendment should welcome rather than inhibit regulation of the structure of such industries in the interest of facilitating more autonomous speech. It is also possible, of course, for commercial speech to be a relatively nonautonomous replication of the market forces shaping it. Particular instances of many other, noncommercial types of speech may also be relatively nonautonomous in this sense. The First Amendment does not require a judge to attempt to determine whether a particular instance of speech was the product of creative, normative reflection (whether it was autonomous) in order to find it protected speech. As I argued in chapter 6, it is the nature of symbolic meaning-making activity (i.e., speech) that it contains the potential for such creative, normative reworking of the materials one is given. The First Amendment’s protection for speech extends to all speech that possesses this potential, which includes commercial speech. The point in protecting speech is to protect and promote the exercise of this autonomy. This is best achieved by preserving those practices that invite the exercise of the autonomy as a whole, rather than by scrutinizing each instance of such a practice to determine if it was in fact autonomous. Such scrutiny would be both impractical and extremely chilling to the exercise of autonomy. Thus, a narrative model, by focusing on creativity and normative reflection rather than choice, finds com-

Free Speech and Narrative Autonomy | 209

mercial speech not different in kind from other forms of speech that are also subject to social pressures. In the case of all such forms of speech, a narrative autonomy model of free speech would protect the speech as a means of preserving the potential for the exercise of the creative, normative reflection that qualifies as autonomy.

B. Systems of Speech Protecting the potential for autonomy, however, requires more than simply allowing particular individual speech to go forward. A narrative model points us beyond individual speech acts as exercises of autonomy and toward the social systems that affect the potential for speech to be used in this way. In order to protect the potential to exercise autonomy through speech we must also examine, and where necessary modify, the social structures and practices that inhibit the fulfillment of this potential. Finding a particular speech act or category of speech protected, then, should not mean that the social structures that shape it are immunized from government regulation. Such structures can affect the exercise of autonomy in two ways. First, some social systems are themselves speech systems, which control the opportunities to use speech as an exercise of autonomy. Speech systems control whether or not one gets to speak, the conditions under which one speaks, and what one is allowed to say, and they thereby directly affect the opportunity for speech to function as an exercise of autonomy. Second, a much broader range of social systems can also affect the exercise of narrative autonomy through speech indirectly, by promoting or inhibiting the capacities necessary to narrative autonomy for large groups of people. I will address the direct role of speech systems first and then turn briefly to the indirect effect on capacities.

1. Speech Systems and the Opportunity for Exercising Autonomy A relational model of autonomy, with its pragmatic focus on the conditions necessary to support and encourage autonomy, requires that we look at speech as a matrix of social systems and not just as an individual act. In order to function as a realm for the exercise of autonomy, speech systems need to provide the conditions necessary for narrative autonomy.

210 | Free Speech and Narrative Autonomy

The protection for freedom of speech, then, could be understood as protection for the health and good functioning of such systems rather than only as protection for individual rights. This additional focus is one of the more important contributions of a relational model of autonomy to free speech jurisprudence. A relational model of autonomy facilitates the shift in focus from the individual speaker or listener to the social structures and practices that support speech on a large scale. This shift in focus allows us to seriously consider some of the kinds of problems current doctrine has difficulty even recognizing: the ways in which systems of speech restrict the opportunity for narrative autonomy for groups of people even if they do not violate the rights of any particular individual. While such a systemic focus is not uncommon in other forms of First Amendment theory, such as democracy theory,18 it is much rarer in autonomy theories, which traditionally take a more individualistic view.19 A relational model of autonomy allows an autonomy theory of free speech to take this broader, more systemic approach as well. In order to demonstrate how a narrative model of autonomy alters the analysis, I will briefly consider the issue of campaign finance reform. In this area, the scholarship has revolved around a conflict between autonomy theorists, who basically oppose regulations,20 and democracy theorists, who are more supportive of them.21 I will canvass the ways in which an autonomy theory of free speech has been understood to block campaign finance reform laws and demonstrate that a relational model of autonomy would reduce this resistance, both through its conception of the nature of individual autonomy and through its concern for systemic issues of autonomy. An autonomy-based theory of free speech has been understood to block campaign finance reform for at least three reasons. First, such laws are understood to violate the autonomy of the would-be speaker whose contributions or expenditures are restricted. Second, such laws serve a government purpose of restricting speech because of the fear that it will lead people (voters) to make bad choices and thus rest on a form of paternalism that the Court has found inconsistent with the autonomy of listeners. And third, such laws impose a particular normative vision on public discourse and therefore fail to respect the autonomy of citizens to use public discourse precisely for the purpose of debating such normative visions.

Free Speech and Narrative Autonomy | 211

The first argument focuses on the autonomy of the speaker restricted by such laws. The most famous example of this argument appears in the Supreme Court’s opinion in Buckley v. Valeo, in which the Court asserted that “the concept that the government may restrict the speech of some elements of our society in order to enhance the relative voice of others is wholly foreign to the First Amendment. . . .”22 The focus on the restriction on the speaker arises from the Court’s “preoccupation with individual autonomy.”23 The problem is that the government is sacrificing the autonomy of the speaker to achieve its goal of equality, when the very purpose of the First Amendment is to protect the speaker’s autonomy.24 From the perspective of a narrative, relational model of autonomy, two aspects of the situation are highlighted that reduce the impact of the speaker’s autonomy on the First Amendment analysis. First, the narrative model sees the autonomy interest of speakers here as more limited than would a choice-based model. Second, the narrative model sees autonomy interests at stake in other aspects of the situation as well. From the perspective of narrative autonomy, political campaign speech is a powerful and important exercise of autonomy. Such speech is a creative and normative effort to make meaning and affect understanding. Nonetheless, a relational model of autonomy highlights two ways in which the autonomy claims of speakers are limited. The first point applies only to contributions to candidates and not to expenditures. Expenditures for speech—either by a candidate or by an independent entity—are a direct exercise of narrative autonomy. Contributions to a campaign, on the other hand, have only a limited role as speech from the perspective of narrative autonomy. In order to qualify as speech under this theory, the activity must be engaged in for the purpose of creating meaning, rather than simply affecting the world directly, and it must be likely to be understood as an effort to create meaning.25 In many (probably most) cases, the nonverbal behavior of making a campaign contribution would include a symbolic meaning of support that is intended by the contributor and is likely to be recognized by the audience (other voters or the candidate). It should be protected to that extent. But much of the purpose of such contributions is practical rather than symbolic. A contribution is, in this sense, simply the equivalent of driving the candidate to the station so that he won’t miss a train: a practical contribution to the outcome of the election with no symbolic or inherent meaning. A narrative model of autonomy, with its focus on creative meaning making, makes it clear why

212 | Free Speech and Narrative Autonomy

this aspect of the contribution is not protected speech.26 It is certainly political action (like voting or running for office, which nonetheless can be carefully regulated),27 but it is not speech. The second way in which the speaker’s autonomy claims are limited applies, however, to expenditures as well as contributions. A choicebased model of autonomy sees the interference with campaign financing as a direct restriction on the choices of individuals. A relational model of autonomy, on the other hand, recognizes that the ability to tell one’s story is dependent on the provision of social goods, including opportunities, resources, and recognition by the interlocutors who are a necessary part of the process. The decision to restrict campaign spending and contributions is a decision to reallocate these social goods in a different way. The claim of the person whose narrative autonomy is reduced by such a decision must, therefore, be a claim to have the ordering remain in a particular form. I believe that the Constitution does, in fact, authorize such claims in certain circumstances; for example, it would generally be a violation of autonomy for the government to reallocate control over an adult person’s body to someone else.28 But it is much less clear that a speaker has an autonomy right to demand that the government allocate social goods in such a way as to assure him that he may spend unlimited money on political campaigns.29 In other words, from a narrative autonomy perspective, the speaker’s autonomy claim is not simply a claim to be left alone, as it might appear to be from a liberal autonomy perspective; instead, it is the much more difficult claim that the Constitution requires the government to allocate the social goods that facilitate autonomy in this particular way. Moreover, from the perspective of a narrative model of autonomy, there are other autonomy interests at stake besides those of the speaker. Because autonomy is dependent upon social relationships to sustain the process of creative reinterpretation, any system of speech can and should be examined to determine whether it supports or inhibits such social relationships. Vincent Blasi’s innovative approach to expenditure limits suggests that candidates for electoral office, for example, lose a great deal of their own narrative autonomy in the frantic and all-consuming rush to amass a “war chest” that will secure their election. The campaign is no longer an opportunity to tell their stories but simply a search for money.30 A relational model of autonomy, which recognizes the ways in which social conditions support or inhibit autonomy, allows us to see that this interest on the part of candidates can be understood as an autonomy interest.

Free Speech and Narrative Autonomy | 213

Similarly, the harms to citizens generally identified by proponents of campaign finance reform can also be seen as implicating narrative autonomy concerns. Unrestrained contributions and spending in political campaigns hurt the autonomy of those whose ability to participate in political dialogue and decision making is thereby unjustly reduced. Each part of this claim needs elucidation. First, such unlimited spending and contributions reduce the relative ability of less wealthy citizens to participate in political dialogue. This part of the claim is relatively straightforward, although hardly uncontroversial. It rests on the assumption that money buys a variety of political advantages, including the ability to set the agenda, the ability to persuade other voters to share one’s positions, the ability to give information to governmental decision makers, and—perhaps—the ability to influence the actions of candidates and elected officials.31 Second, this relative reduction in ability to participate is unjust. This claim depends upon the idea that differences in ability to participate in political dialogue and decision based on wealth are not justified. One could argue for this conclusion on a variety of bases, including political theories of the equal worth of citizens32 and instrumental arguments about the irrelevance of wealth to generating good political outcomes or improving democratic processes.33 One common response to such arguments is to point out that other irrelevant bases also generate inequality here: differences in time and talent, incumbency, and so on.34 The existence of such other failures of equality does not, however, undermine the claim that this particular failure is unjustified. Moreover, our inability to eliminate all failures is not a reason to ignore the failures we can address. Only if eliminating the impact of wealth exacerbated the influence of these other inequalities would their presence undermine the argument for campaign finance reform. This counterproductive effect with respect to incumbency is, in fact, a real risk of certain forms of regulations and should be taken quite seriously as a pragmatic issue in the drafting of regulations.35 But the possibility of such effects should be seen as an empirical and pragmatic issue in the design of such regulations rather than as an argument for why the effect of wealth is justified. Another common response is to suggest that the influence of wealth is a relatively harmless distortion either because it is not significant compared to the others,36 or because it does not generate a systematic impact on policies and political dialogue but rather is randomly distributed

214 | Free Speech and Narrative Autonomy

across the political landscape.37 Finally, some critics point out that restrictions on campaigns will do little to reduce the overall impact of wealth on politics because so much of the remaining political dialogue is dominated by those with money.38 There are, of course, arguments in response to these contentions, but they have been amply rehearsed in the existing literature. I would simply suggest that these issues go to the important, but secondary, question of whether a campaign finance regulation can be designed in such a way as to achieve its goals with minimal unintended, and undesired, consequences. This is a difficult question and clearly concerns the desirability of such regulation, but it does not explain why wealth is not an unjustified basis for differences in political influence. Finally, a narrative model of autonomy fills in the last piece: once one accepts that the impact of wealth on politics is unjustified, a narrative model demonstrates that the harm caused is a serious one from a First Amendment perspective because it is (among other things) a harm to autonomy. The argument is that the relative reduction in ability to participate in political dialogue and decision caused by the irrelevant factor of wealth is a harm to the autonomy of those disadvantaged by it. While the autonomy interests of other citizens are not unrecognized in the current dialogue,39 a narrative autonomy model provides a more powerful understanding of what they are and why they are at stake. Narrative autonomy concerns one’s ability to tell one’s own story, and that process takes place within social relations. The ability to participate in setting the agenda for political dialogue, in presenting one’s point of view to fellow citizens or to policy makers, and in determining who those policy makers will be is crucial to the processes of narrative autonomy. Such activities are fundamental because the story one tells must be told with others and must be recognized and reflected by others in order to be effective.40 As I will argue in the last section of this chapter, politics—while not the only social arena in which narrative autonomy operates—is a particularly important venue for narrative autonomy.41 The reduction of opportunities for political autonomy on the basis of the morally irrelevant characteristic of wealth is an injustice.42 In short, then, a narrative model highlights the limits of the speaker’s autonomy claims and the importance of potential autonomy concerns of other parties.43 The second autonomy-based objection to campaign finance reform concerns the government purpose served by such laws. The Court has accepted the prevention of corruption and the avoidance of the appearance of corruption as legitimate and important state goals.44 On the other

Free Speech and Narrative Autonomy | 215

hand, the Court has rejected the equalization of political influence as a legitimate goal.45 This rejection reflects the Court’s well-established doctrine that the government may not manipulate the information citizens receive in an effort to control their ideas, and presumably their behavior. The Court sees this sort of paternalism as so constitutionally suspect that it strikes down regulations based on it even when only lower-value categories of speech, such as commercial speech, are at issue.46 The difficulty with the paternalism argument is that it conflates two different sorts of government purposes, which are clearly distinct from the perspective of narrative autonomy. On the one hand, sometimes the government silences speech because it is afraid that the speech will cause listeners to behave in ways that the government—often legitimately— wishes to prevent. The government may, of course, prohibit the behavior itself in many cases, but the Court has held that it may not try to control or prevent that behavior by manipulating the speech to which people are exposed. The problem here is not properly characterized as paternalism. The behavior that the government wishes to prevent may be harmful to the listener herself (as in the case of liquor advertising), but it need not be so. Even if the behavior at issue is harmful only to others, the government generally may not attempt to control it by restricting speech. This is the lesson of the cases on politically subversive speech, which set an extremely high standard for allowing suppression of speech because of the response of listeners to the message.47 The harm to autonomy when the government attempts to control behavior by limiting speech is clear both in a liberal and a relational model. From the perspective of narrative autonomy, the government is seeking to control people’s behavior precisely by restricting the stories they may tell. Although its end—controlling the behavior—may be legitimate, the means chosen is one that robs people of resources for narrative autonomy, as well as opportunities for its exercise. Indeed, the problem lies precisely in the connection between the means and the end: the government is attacking narrative autonomy by attempting to preempt the process of creative reinterpretation that leads to action. As the Court has held, such a direct attack on autonomy should be allowed only under the most exigent and limited circumstances. On the other hand, it is neither paternalistic nor otherwise violative of autonomy for the government to seek to restructure the use and allocation of social resources so as to promote the opportunity for autonomy.

216 | Free Speech and Narrative Autonomy

As I will discuss below, the government is always in the business of allocating such resources and cannot avoid impacting autonomy in the process. To take account of that impact and attempt to make it positive rather than negative is not a violation of anyone’s autonomy. It is not paternalism, because it is a restriction imposed not for the good of the person restricted but rather for a larger social good. It is not a violation of anyone’s autonomy unless, as I discussed above, a person has a positive claim that the government must organize the particular resources at issue in a particular way. This will be the case on rare occasions, but such exceptions do not make suspect the entire category of government actions taken for this purpose. Campaign finance reform efforts serve a variety of purposes. Some of those purposes fall into the illegitimate category defined above, but many do not. If the purpose of the reform is to protect voters from their own foolish responses to campaign speech, then that is both paternalistic and a violation of autonomy. If the purpose is to protect voters from the foolish and corrupt responses of candidates to contributions, that is not paternalistic. Whether it violates the autonomy of the candidate depends upon whether the goal is to prevent the impact of the message or to prevent the impact of the money. The goal recently suggested by Professor Vincent Blasi—protecting candidates’ time—is also clearly neither paternalistic nor otherwise violative of autonomy.48 Most importantly, the purpose of reducing inequalities of participation is neither paternalistic nor inconsistent with autonomy. It does not seek to short-circuit the process of narrative autonomy, but to assure it to a greater number of people in a greater degree. Indeed, if the point in the protection for free speech is to promote narrative autonomy, then the government has a legitimate, even compelling, interest in assuring that systems of speech provide adequate opportunities for the exercise of that autonomy. In short, the goal of equalizing opportunities for political participation does not fall within either category of suspect government motives: it is neither paternalistic nor is it part of the larger category of illegitimate government interests in short-circuiting the process of autonomy. I made a parallel argument about the assessment of government motive under a relational truth theory in the previous chapter. As with truth, so with autonomy: the distinction for which I am arguing is important but inherently unstable. The difference between an illegitimate purpose to short-circuit autonomy and a legitimate purpose to improve the conditions for its exercise will often be contextual and a matter of descrip-

Free Speech and Narrative Autonomy | 217

tion.49 The question in such cases is, which description of the government’s purpose is most plausible under the circumstances? This, admittedly, indeterminate analysis calls for a judgment sensitive to the particular regulation at issue. When such analysis suggests that the government purpose is most plausibly understood, in context, as an effort to promote the opportunities for narrative autonomy, that purpose should not be subject to the automatic invalidity that properly applies to government efforts to short-circuit narrative autonomy. The third autonomy-based objection to campaign finance reforms is that the First Amendment is designed precisely to prevent government from intruding into such speech systems, even if its goal is to “improve” them. In response to calls for government regulation aimed at improving our public dialogue, for example, a number of scholars have argued that we lack standards to identify a fair speech system and that such arguments are “easily manipulated to justify censorship.”50 This objection rests on an autonomy claim at a higher level of abstraction: the claim is that the only way for a government to adequately respect the autonomy of its citizens is to refrain from imposing substantive standards on the public debate through which the political community constitutes itself.51 The standards for fairness are themselves the very stuff over which we struggle in our public discourse and could not possibly be set in advance by the government without constricting the range of that dialogue unjustly.52 As some of these critics recognize, however, there is no position of simple neutrality for government here.53 Even a “hands off” approach to speech leaves intact underlying rules of property and contract and tort— government regulations—that will shape the distribution of power and opportunity within speech systems.54 Indeed, from the perspective of narrative autonomy, the very idea of neutrality is problematic. If narrative autonomy is always exercised within social relationships, a society—either by what it has done or by what it has left undone—cannot help but shape the nature and distribution of opportunities for narrative autonomy. Any choice about the structure of speech systems will have implications for their role in supporting autonomy—there is no choice without such effects. The argument here is simply that such effects must be seen as relevant to the constitutionality of such systems if freedom of speech is justified because of its connection to narrative autonomy. Critics have attacked this argument, made most forcefully by Professor Cass Sunstein, on several grounds. First, some argue that this view

218 | Free Speech and Narrative Autonomy

improperly eliminates the state action requirement: the generally accepted idea that the restrictions of the Bill of Rights apply to government action only and not to private actors. Professor Lillian BeVier, for example, points out that the distinction between direct government distribution of a benefit, on the one hand, and government enforcement of a system in which private market decisions allocate benefits is “genuine and profound.”55 She argues that the latter cannot be seen as state action for the purposes of activating constitutional review without eliding this crucial difference.56 Two responses are in order. First, the argument here does not expand the category of actors subject to constitutional restraint beyond current doctrine.57 The state action doctrine concerns who is controlled by constitutional norms. To recognize the underlying state authorization of economic rights does not mean that private use of economic rights in ways that restrict the speech of others generally qualifies as state action for the purposes of First Amendment analysis. I have not argued that the action of the wealthy speakers violates the First Amendment. It is the government who is acting in campaign finance reforms. The argument here suggests that in applying constitutional norms to the government actor, we should recognize that when the government seeks to rearrange the underlying rules in ways that prevent abuses of private power, it is not, simply for that reason, violating the First Amendment. The point is not to treat the private action as state action, but to understand the constitutional limits on state action differently. Second, although the state action limit remains intact as a matter of doctrine, Professor BeVier is quite right to sense that the argument does call into question the importance of the distinction she wishes to rescue. The argument does not eliminate the state action doctrine, but it does suggest that the distinction between private action carried out pursuant to a grant of government power and direct government regulation is not as profound as Professor BeVier believes. Indeed, the importance of this distinction rests on an assumption that there is a fundamental moral distinction between the two categories, arising from the autonomy value implicit in private (and not in government) choices. A relational model of autonomy, however, makes it clear that autonomy and constraint may be at work on both sides of this line. The private choices are not “natural and inevitable,”58 but a direct product of government regulation and support, and the government’s choice to alter that regulation may support, rather than interfere with, the social conditions necessary for individual

Free Speech and Narrative Autonomy | 219

autonomy. The point, then, is not to eliminate the relevance of this distinction altogether—there are still significant reasons for preferring decentralized private decisions over centralized public ones—but to transform the choice between these two into a pragmatic issue in light of the goals of the First Amendment rather than a morally charged issue of principle. The second objection often raised against this view concerns the distinction between negative and positive rights. The deep structure of present constitutional law sees the First Amendment as a guarantee of a negative rather than a positive right. In other words, the First Amendment marks the limits of government regulation of speech systems, but does not usually generate positive rights on the part of citizens to demand that the government initiate regulations that it has not chosen to adopt. The distinction between negative and positive rights is different from but closely related to the state action requirement. While state action concerns who is bound by constitutional norms, the issue here is the nature of the obligation created by such norms: is it merely an obligation to refrain from acting or might it also include some obligation to undertake action? One could, theoretically, have a state action doctrine in a regime of positive rights (where the obligation to take action falls only on the government and not on private actors) or no state action doctrine in a regime of negative rights (where the obligation to refrain from interfering with the exercise of rights falls both on the government and on private actors). Despite this difference in focus, however, the two doctrines are closely related. Both arise from a vision in which the responsibility of government is limited to its own positive acts. Neither the state’s contribution to the actions of private persons nor the state’s passive inaction can, on this view, be the basis for constitutional responsibility. Thus, a citizen may demand that the state stop its own acts that interfere with her exercise of rights, but may not demand that the state reorder the social world (which, of course, includes other private actors) in ways that would benefit her exercise of rights. A relational view of autonomy, and the more general social constructionism on which it rests, undermines this view of the world by highlighting the ways in which private actors are dependent upon the provision of social resources and conditions by the state. Thus the government has always already acted, if only to set up the underlying property and contract regimes that generate speech opportunities, and private actors act in reliance on such systems.

220 | Free Speech and Narrative Autonomy

As I suggested above, the state action doctrine still makes sense within a social constructionist world. After all, even if the state has always acted, it is not the only actor. We might sensibly choose to ascribe certain results under certain circumstances to the activity of private parties even if that activity is facilitated by the state. The positive/negative rights distinction, however, is substantially undermined by the social constructionist challenge to the underlying vision of government responsibility. There is almost no situation in which one could not describe the obligation created by a right in either positive or negative terms. For example, the government is prohibited from excluding a speaker from a park or the government is required to provide a speaker with an equal opportunity to use the park. Indeed, the fundamental incoherence of the distinction between negative and positive rights was apparent even without the shift in perspective wrought by social constructionism; that shift merely makes the difficulties clearer. Thus, a relational model of autonomy might create some pressure for abandonment of this distinction. A narrative model of autonomy could provide the basis for generating a vision of particular positive rights. In the absence of such a large-scale revision of our legal culture, however, a relational model of autonomy would, at least, suggest that we should move further in the direction of recognizing positive rights in those situations in which the distinction is most untenable. For example, under current law a speaker can sometimes cast a challenge in terms that meet both the state action requirement and the negative rights restriction but that generate results that are effectively equivalent to a positive right. For example, a would-be speaker might argue that it is a violation of her First Amendment rights to allow a private property owner to use normal trespass laws to remove her from his property when she is seeking to use it to speak. If she wins such an argument, which is technically negative in form, it would result in what is effectively a positive right to call upon government to guarantee access to a certain resource that would otherwise be controlled by another private party. Present jurisprudence leaves open the possibility of such challenges, although it restricts them severely.59 The narrative model of autonomy would suggest loosening such restrictions, even if we did not move all the way to an abandonment of the negative/positive right distinction. A relational model of autonomy then, explains both why government regulation of speech systems is inevitable and why it implicates autonomy issues at the system level. This conclusion does not diminish the

Free Speech and Narrative Autonomy | 221

significance of the concern commonly voiced by critics about the risk of government censorship.60 The power to regulate speech systems (which is a power the government always exercises, either directly or indirectly, intentionally or without realizing it) has tremendous potential for abuse. Any view that calls upon the government to undertake this task self-consciously must offer some protection against such abuses.61 The inevitability of government influence over the preconditions for autonomy does, however, suggest a more modest understanding of the nature of abuse in this context. Government regulation that shapes the speech system in ways that affect the opportunity for narrative autonomy cannot automatically qualify as abuse because any government regulation (or failure to regulate) will do this. Moreover, ignorance is not a virtue. Given that any government choice may have such an effect, the Constitution should not require the government either to be actually blind or to adopt a pretense of blindness to such effects. Having an effect is not an abuse (nor is noticing or even intending the effect); only having a bad effect (or seeking that bad effect as a purpose) is an abuse. The great fear of government control over speech systems should not blind us to the fact that our present speech systems are also the product of government regulation. Once we recognize that we already live under conditions in which the government has structured our speech systems in ways that are often detrimental to our autonomy, perhaps the fear of allowing government to try to do better will abate a bit. Surely we cannot believe ourselves better off, our autonomy better protected, by refusing to allow government to do—with conscious consideration and according to constitutional standards—what it has already done unconsciously and with damaging consequences for the constitutional value of autonomy. Thus, a narrative autonomy theory removes or reduces several of the roadblocks that autonomy theory generally poses for campaign finance reform.

2. Speech Systems and the Development of the Capacities Necessary for Narrative Autonomy Speech may function not only as an actual exercise of narrative autonomy but also as a mechanism for developing the capacities necessary for the exercise of narrative autonomy. Because the capacities in a relational model of autonomy are very broad—including imagination, abilities to connect to others, and so on—the range of speech acts that might develop

222 | Free Speech and Narrative Autonomy

those capacities for any given individual is potentially very broad also. Moreover, nonspeech activities, such as getting an education, also develop these capacities. Indeed, for many of these capacities, nonspeech activities may be far more significant as mechanisms for the development of the capacity. But if the First Amendment were to protect all the activities that contributed to the development of these capacities for any individual, then the scope of First Amendment protection would be too broad to be practical. When we move from the individual level to the systemic level, we find, again, that speech systems are neither unique nor even the most important systems affecting the capacities for narrative autonomy. Many social structures and institutions affect the potential for people to develop the capacities necessary for narrative autonomy. Certainly, an economic system that generates and justifies great inequalities in wealth is at least as important in affecting the potential for many people to develop capacities for autonomy as any system of speech. Nonetheless, speech systems are one of the important categories of such background institutions. The educational system, the mass media, the internet, political campaigns, the system of government funding for artistic speech, the rules governing the use of public fora and other government property for speech purposes—these are all examples of speech systems. These systems constitute the background that broadly shapes the resources and opportunities available to large numbers of people to develop the capacities important to narrative autonomy. While other social systems also affect the capacities for narrative autonomy, it is unlikely that those capacities could be adequately promoted without protection for these systems of speech. Thus, well-functioning systems of speech, while not sufficient, are necessary to facilitate the development of the capacities for narrative autonomy for people generally. What about the other, nonspeech systems that also affect the development of autonomy for people generally, such as the economic system? The First Amendment itself does not address such systems directly, but the values underlying freedom of speech may have implications for the interpretation of other aspects of the Constitution as well. If we take narrative autonomy seriously, it should inform our understanding of the ways in which wealth is allowed to affect other opportunities important to the development of autonomy. So, for example, we might conclude that property-tax–based funding for public schools that results in massively un-

Free Speech and Narrative Autonomy | 223

equal funding for richer and poorer districts is a violation of the equal protection clause.62 With respect to speech systems and the regulation of speech within them, however, we should include the concern for the impact on the capacities necessary for autonomy in our First Amendment analysis. Of course, such systems serve many other purposes besides the promotion of the capacity for autonomy. And, of course, the effects of such systems on capacities such as reasoning, empathy, imagination, and so on is complex and difficult to determine in a reliable way. Thus, the limits imposed by a concern for this impact must leave a broad range open to the government in structuring speech systems. But, at a minimum, we could say that any government regulation of a speech system that has the purpose of reducing the capacity for autonomy is a violation of the First Amendment. As I have suggested, the Court’s present concern with paternalism is ill defined and often misapplied, but the kernel of truth at its core is captured by this systemic issue: it is illegitimate for the government to seek to inhibit the development of the capacities for autonomy.63 The important point here is that this concern applies whether or not any particular person’s opportunity to speak has been limited. When a speech system is designed in such a way as to interfere with the development of these capacities, it reduces the potential for autonomy of those who participate in it. The system thus undermines autonomy before it is even formed and there may be no particular person of whom it could be said that he or she would have developed the capacities for autonomy to a greater or deeper degree if the system had not interfered. The effect is cumulative and difficult to quantify. This sort of indirect effect can coexist with a direct impact on opportunities for autonomy, but it can also arise even in the absence of a direct effect on the opportunities of particular individuals. A concern about this effect of speech systems on the capacities for autonomy adds an important element to First Amendment jurisprudence. There is a category of cases that, under present doctrine, are understood to raise no serious First Amendment issue because no individual speaker’s rights are at stake, despite the fact that the government is acting to regulate a system of speech.64 In this category, the courts tend to see only the government, using its own money to run its own programs. But it is precisely here that a narrative autonomy theory, and in particular a focus on the development of the capacities for autonomy, suggests that the First

224 | Free Speech and Narrative Autonomy

Amendment places an outer limit. The government should not be allowed to regulate speech systems for the purpose of restricting people’s capacity for autonomy. This limit on government purpose should be understood to apply to the choice of which books to put into the school library as well as the choice about which to remove;65 to the limits placed on government funding for artistic speech;66 and to restrictions on the abortion-related speech of programs receiving federal family planning funds.67 These cases represent the government in its role as builder and facilitator of our speech culture. Even when no particular individual’s speech rights are violated by government action in this role, the First Amendment, if understood in terms of narrative autonomy, still places some restrictions on how the state may seek to shape this speech culture. Present doctrine— with its focus on whether the government has penalized speech or merely refused to subsidize it, and with its concern about whether the program at issue creates a public forum or not—draws lines within this large category that may be relevant for other reasons. But, regardless of where a regulation may fall in terms of those lines, if its purpose is to inhibit the development of the capacities necessary for autonomy, it violates the First Amendment.68

C. Autonomy, Speech, and Democracy A narrative model of autonomy also gives rise to a free speech principle by way of the connection between autonomy and democracy. Narrative autonomy, as I suggested in chapter 6, demonstrates the importance of a discursive model of democracy. The question then is how to understand the role of speech in a discursive democracy. Much work has already been done on this question, most notably by Professor Cass Sunstein.69 Professor Sunstein argues that the particular form of discursive democracy he is endorsing, which he calls Madisonian, requires a speech system that meets two minimum requirements: first, “it must reflect broad and deep attention to public issues,”70 and second, “there must be public exposure to an appropriate diversity of views.”71 I will explore briefly in this section some additional considerations that must be added to these two if one of the important goals of discursive democracy is narrative autonomy. First, from the perspective of narrative autonomy, it is not only the overall quality of public debate that matters; it is also the quality of par-

Free Speech and Narrative Autonomy | 225

ticipation in that debate by any individual. Autonomy is a concern for every person. If autonomy importantly includes participation in the process of political debate through which the parameters of one’s freedom are determined, then an individual denied such participation has suffered a diminution of autonomy even if the overall debate remains broad, deep, and diverse. One of the criticisms of more discursive models of democracy is that they tend to see social value only in the aggregate and to ignore the interests of individuals such as speakers.72 A discursive model that is based on narrative autonomy, however, will recognize both the systemic concerns and the vital interest of the individual speaker in her participation. This vital interest arises precisely from the way narrative autonomy depends upon political participation. In a liberal model of autonomy, the primary concern in politics is about outcomes and the ways in which they may expand or contract the realm of individual choice. A narrative model, on the other hand, sees politics as central to the construction of identity and the social relations in which it operates. It matters not only that we get the right result but also that I have the chance to participate in the process in ways that are constitutive of my ability to tell my own story. Because democratic politics is a necessary aspect of narrative autonomy, each individual has a powerful autonomy interest in participating. Moreover, the narrative model also helps to make clear why speech is the preferred method of political participation. Speech is symbolic meaning making, and the way in which politics is essential to narrative autonomy is as a mechanism for the social construction of meaning. The participation that is essential is participation in the collective project of determining these meanings. Speech is the primary means of such participation. Sunstein acknowledges and accepts the role of autonomy values in the justification of freedom of speech, arguing only that they are incomplete as an explanation and require the addition of democracy.73 In other words, he sees democracy and autonomy as alternative, and perhaps complementary, theories. The premise of this book is that democracy theories of free speech are ultimately applications of truth and autonomy theories, in the sense that the connections between speech and democracy are dependent upon the connections between speech and truth, or speech and autonomy. This does not mean that democracy theory is less useful than truth and autonomy theories, however. There are good reasons to

226 | Free Speech and Narrative Autonomy

believe that democracy theory may best capture the particular combination of truth and autonomy values that justifies freedom of speech and it could, therefore, be the most useful theoretical framework for the First Amendment. I make no attempt to answer that question here. My point here is simply that the meaning and function of democracy depends in part on the model of autonomy that lies behind it. Therefore, if one considers discursive democracy with an eye to the underlying value of narrative autonomy, one is led to greater recognition for the individual interest in political participation through speech. Second, a discursive model of democracy based on narrative autonomy would see greater continuity between politics and culture generally than Professor Sunstein’s approach suggests. Professor Sunstein defines political speech as speech that “is both intended and received as a contribution to public deliberation about some issue.”74 While this is a broader definition than some other advocates of democracy theory have adopted, it still seeks to clearly distinguish political speech from participation in culture more generally. Indeed, one of the advantages of a democracy theory, from Professor Sunstein’s point of view, is that it provides a basis for a “two-tier” First Amendment, in which political speech receives greater protection than other speech.75 From the perspective of narrative autonomy, however, there is no sharp division between political and nonpolitical speech. Politics is not inherently more important as a site for the negotiation of identity than many other cultural conversations. This is not to say, however, that there are no differences at all between political speech and other categories. Political participation might be seen as a particularly important form of cultural dialogue and one deserving of special concern and protection because of its consequences. Traditionally, political autonomy has been seen as the guarantor of other forms of autonomy and this argument still holds for narrative autonomy: one’s ability tell one’s own story in any aspect of life may be dependent on possession of the freedoms and resources whose distribution is determined by the political process. In other words, the political process is distinctive in terms of the importance of its consequences for autonomy generally. Given the practical, consequential significance of politics and also its lack of inherent significance, a democracy theory of free speech based on a narrative model of autonomy would be likely to generate a more flexible and contextual hierarchy of speech categories. Political speech would

Free Speech and Narrative Autonomy | 227

be presumptively significant because of its consequences for autonomy generally, but other categories of speech would not automatically fall into a lower tier. There are several other options: a continuum of constitutional protection (with political speech at one end) defined by the importance of speech categories to the social production of autonomy; a presumption in favor of strong protection for political speech, with other speech required to make a showing of its importance to autonomy to receive a similar level of protection; or a series of categories (not merely two) defined in advance by content and receiving varying levels of First Amendment protection. I believe that the choice between such options should probably be made on the basis of pragmatic and institutional concerns rather than on the basis of the theory of autonomy. My point here is simply that a narrative model of autonomy would suggest that a democracy theory might well include a more flexible and less clear division between political and nonpolitical speech than Professor Sunstein’s approach suggests. Finally, a narrative model of autonomy makes it clear that discursive democracy requires not only that public dialogue have a certain content (deep and broad attention to public issues and diverse views) but also that citizens approach that dialogue with certain attitudes and expectations. Those attitudes are perhaps best summed up by suggesting that discursive democracy can only serve the goal of narrative autonomy if citizens engage in it with a commitment to both respect and responsibility. In chapter 7, I outlined the ways in which a relational truth theory leads to similar conclusions. Here I will briefly explore the narrative autonomy path to the same end. Some of us are more vulnerable to the dangers of public dialogue than others. When one belongs to a social group that has been defined in derogatory ways, one is subject to greater threat through the discursive process. Initially, members of such groups may lack the credibility accorded to others when they speak. Issues of credibility—and the denial of credibility to women—have been central to feminist analyses of legal issues such as rape, sexual harassment, and pornography.76 Moreover, while we are all shaped by the group images negotiated through this process, the risks of the process fall disproportionately on those whose groups have historically been defined by damaging images. “In a society that is historically divided along harsh, unyielding axes of dominance and subordination, individuals cannot escape the influence of cultural

228 | Free Speech and Narrative Autonomy

stereotypes and other prescriptive representations of the groups they belong to, nor can they escape the influence of the social and economic advantages and disadvantages that institutions confer on these groups.”77 These inequalities suggest several requirements for the way in which public dialogue takes place. First, we must be self-conscious about the fact that identity and autonomy are at stake in our discursive politics. We cannot, either in reality or as a matter of pretense, remain oblivious to the dangers of such discussions for the people whose identities are implicated. In short, we must take responsibility for the moral implications of discursive politics. Taking responsibility means recognizing that the way we conduct our public dialogue has consequences for the autonomy and identity of many citizens. This recognition is largely missing from the symbolic politics that dominates our present political dialogue. Without this recognition, we blunder about, neither seeing nor counting the harm done to autonomy as part of the costs of social choices, including speech choices.78 In order for discursive democracy to serve the value of narrative autonomy, our broad, deep, and diverse conversation must also be a conversation in which we accept this responsibility. We must also recognize that the unequal impact on identity and autonomy infects the continuing conversation itself. Restrictions on autonomy for particular groups lead to particular conversational consequences: some people are less likely to speak, less likely to be heard, or less likely to be taken seriously or believed. As a result, the public dialogue through which our future stories are shaped is limited by the inequalities generated by the present dialogue. Not only must we come to the conversation with the knowledge that it carries consequences for autonomy, then, but we must also come with the understanding that our past dialogue will limit our present one in unequal and unfair ways unless we actively seek out and resist such inequalities. Promotion of respect for and attention to those whose participation has been affected by these inequalities is a necessary aspect of a well-functioning speech system.79 When we understand democracy as serving the value of narrative autonomy, then, the need for respect and responsibility in our public dialogue moves from the background into the foreground of our democracy theory of free speech.80

Free Speech and Narrative Autonomy | 229

D. Conclusion Our intuition that speech is connected to autonomy reflects a truth that is both profound and important. Speech is the primary mechanism through which we exercise narrative autonomy; speech systems are powerful determinants of both our opportunities for such exercise and of the capacities necessary for that exercise; and democracy is one of the most important arenas in which narrative autonomy is both exercised and protected. This chapter has offered a glimpse of just a few of the ways that a focus on narrative autonomy might alter our view of both the large issues of free speech theory and the specific analysis of certain doctrinal problems. The primary innovations of this autonomy theory are its focus on systemic concerns about autonomy, its emphasis on the social conditions required for the development and exercise of autonomy, and its blending of moral and epistemological concerns. A narrative model of autonomy preserves the important insight about our need to see ourselves as creative agents that lies at the heart of all autonomy theories, while recognizing that the process of creative agency is fundamentally social and interactive. A narrative autonomy theory of free speech sees in this process of creative agency the grounding for our First Amendment right.

conclusion

The Enlightenment Vision and the Interpretive Approach

The shift from the traditional models of truth and autonomy to the newer models suggested in this book is not merely a change in the form of these particular concepts. The shift advocated here is a fundamental reorientation: emotional as well as cognitive. In order to address the difficulties created by our present understandings of truth and autonomy, we need a new vision of the relationship between morality and epistemology, and of the ways we relate to ourselves, each other, and the world. The older models of truth and autonomy were part of a vision that developed out of the Enlightenment; that vision carried its own distinctive virtues and limitations and it represented a response to a particular emotional and intellectual agenda. The relational model of truth and the narrative model of autonomy are parts of a different vision that seeks to salvage some of the virtues of the Enlightenment view, while offering a different basic orientation, emotional and intellectual. This conclusion will outline these visions and their consequences.

A. The Enlightenment Vision The Cartesian model of truth and the liberal model of autonomy grow out of a single, unified vision. At the heart of this vision is both a dream and a nightmare: the dream is of liberation and the nightmare is of chaos. The genesis of the Enlightenment lies in this dream, and it is one that still resonates for us. But it is fear of the nightmare that provides the powerful emotional need for certainty and security that is the primary driving force behind the conceptions of truth and autonomy critiqued earlier. In the effort to evade the nightmare, the Enlightenment vision has, I believe,

231

232 | Conclusion

sacrificed the dream. If we are to move forward in search of newer dreams, we must understand this emotional/intellectual inheritance, the extent to which the Enlightenment project has failed, and why. The liberal model of autonomy and the Cartesian model of truth are deeply connected. The autonomous liberal is the Cartesian knower: they are essentially the same person. Reason is the primary faculty through which this person understands both himself and the world. Moreover, this person engages with the world—whether in the form of the natural world to be known or the social world to be inhabited—as an individual. His social background, his context, his cultural categories, are no essential part of him as either knower or chooser. What is most significant about him, both epistemologically and morally, is what is most abstract and, therefore, also what is most universal.1 In light of the extensive criticism offered earlier, it would be easy to lose sight of the promise of this vision. Both liberal autonomy and Cartesian truth promise us a freedom from external constraints that is still meaningful and important in the modern world. The Enlightenment project was to free us from the external and arbitrary constraints imposed by social institutions, religious and otherwise. We need only to contemplate the spectacle of Galileo, forced by the Church to recant his scientific discoveries, in order to be reminded of the value of such freedom. To free the human mind and spirit to go wherever reason dictates and ability allows was, and is, a noble ambition, and one that is far from realized. The danger of such freedom, however, is that it threatens to lead to chaos. Once the arbitrary limits imposed by society have been stripped of their legitimacy, what limits will remain? In order to see the terrifying nature of this danger, we need to examine in some detail the types of chaos that threaten to overwhelm us. The problem is not simply the practical difficulties attendant on a reduction in social order; it is much deeper than that. What exactly are liberals and Cartesians—and all of us who have been shaped by the world they made—afraid of? They are afraid, first, of the breakdown of personal identity. If the boundaries of one’s identity and the purposes of one’s life are not given in advance, by God or country or family, but chosen, then how is one to choose? The enormous weight of this choice threatens to crush the fragile sense of identity. Liberal theorists have struggled with the fear that, once the old structures lose their legitimacy, there might be little left to guide one’s choice. This leads to a dilemma: our choices appear to be either determined or arbitrary, neither of which is a sound basis for au-

Conclusion | 233

tonomous identity. In freeing us from determinism, the liberal fears he has merely made us prey to arbitrariness. It is worth noting that this arbitrariness can lead us in two different directions, both of which are problematic. On the one hand, we can decide that the arbitrariness of our identity boundary means that we are, and are responsible for, nothing. There would simply be no meaning to individual identities and no possibility for the concomitant ideas of autonomy, or individual responsibility, or moral personhood. Our idea of the self would shrink to the point where it could no longer sustain the moral life we understand. With no sense of ourselves as relevant moral units, we would fall into passivity and despair.2 On the other hand, we might decide that the arbitrariness of the boundary means that we are, and are responsible for, everything. Since the boundary is not set anywhere in particular, it is always possible that the appropriate unit for analysis in some situation is one in which we are included along with many other people and things. In this case, the relevant identities and the moral categories appropriate to them would be much larger, even infinitely expandable. But then our idea of the self could expand to the point where it would threaten to engulf all distinctions and make our moral lives unbearable and incomprehensible. How could we live feeling that our identity and responsibility are tied to everything? This is a prescription for mystics, but for the rest of us it is the path to madness. The threatened breakdown of the self is, indeed, terrifying. It is, in the final analysis, a loss of meaning for the self. The fear of this breakdown creates an overpowering need for some way to set the limits on identity that is neither socially determined nor arbitrary. The guidelines must be given in a way that is internal to the individual rather than socially imposed, but not totally subjective. The limits must be firm and stable—they cannot be constantly subject to the expansion and contraction that threatens disintegration—but they must not be arbitrary. The brilliant liberal solution is to look for the limits within the nature of the subject himself. It is the structure of the rational will, for Kant and those who follow him, that offers the reassurance that we will not find ourselves in this nightmare. The limits on our identity—which grow out of the very nature of liberal autonomy—guarantee that an individual’s identity will have a relatively stable structure. The point here is not so much to place limits on what people may or may not do, but to place limits on who may be held accountable for what; the issue is less social order

234 | Conclusion

and more moral order. Grounding the limits in the nature of our selves is a truly breathtaking solution: in a stroke, our weakness becomes our strength and our vulnerability is eliminated. The elegance and power of this deontological move need to be seen and appreciated. But, in order to protect us from disintegration, this solution must provide a high level of stability and security. It cannot be fluid or contextual; it must be fixed and universal. It cannot depend on any of the human capacities so obviously infected by social contingency—emotion, imagination, and so on—but only on a capacity conceived as abstracted from such contingencies: reason. It must be the process in which we place our faith, not the outcome, and that process must be neutral with respect to outcomes in a way that insulates it from claims of social determinism. The need for a high degree of certainty and stability arises, then, from the very nature of the fears that the liberal model of autonomy must address. The rationalism, the individualism, the abstraction, and all of the problematic aspects of the liberal model are direct responses to this need. In the end, it is fear that dictates the shape of the self in this model of autonomy. In addition to disintegration of the self, we—the inheritors of the liberal and Cartesian traditions—also fear the breakdown of reality. This is not really a separate problem from the first, of course, but simply a different way of seeing the same problem. Once the old authorities no longer have the power to determine reality, once we have been freed from the shackles of imposed belief, what then can we know? How can we determine what to believe in the absence of such authority? Where will we find the grounds for our beliefs? As with the loss of self, the prospect of the loss of such grounding for belief is deeply frightening. Without some foundation for knowledge claims, we would lose the ability to communicate, the ability to understand each other at all, the ability to recognize and respond to reality. In short, we would lose the sense of meaning in the world: it would be, perhaps, the buzzing, blooming confusion William James imagined.3 Such a loss would, again, plunge us into a kind of madness. The need for some grounding to take the place of the lost authorities is imperative. And, again, the grounding must somehow be insulated from the contingency of social conventions but stable and universal enough to support a reality we must inhabit together. The Cartesian model of truth looks to the very nature of the project of searching for truth to generate this foundation for knowledge. In the

Conclusion | 235

search for truth, the limits are provided by the requirements of objectivity and the methodologies of reason.4 These limits guarantee that the search will generate relatively stable, universal truths. They insulate our knowledge claims from the vicissitudes of context and individual desires. They assure us that the meaning in our vision of the world is not subject to challenge and disintegration simply because someone else has a different point of view. They look to the process of truth seeking—a process based on an abstract view of reason and designed to meet the standards of objectivity—as the source of security rather than any particular knowledge claim that may result from that process. Once again, the problematic aspects of the Cartesian model arise directly from the need to respond to this fear of chaos. Both the breakdown of identity and the breakdown of reality will lead, of course, to the breakdown of social order. The loss of social order is itself a substantial harm worth fearing, both in the war-torn Europe of the Enlightenment and in our own no-less-bloody time. But we are apt to miss the depth and power of the need for security if we focus only on the practical risks of social disorder. The deeper risks are to the understandings of the self and the world on which any hope of social order must rest. The greatest threat is to the possibility of meaning, which depends upon those understandings. The effort to secure social order is always a balance: the risk of disorder must be balanced against the gains in freedom or diversity. This is a pragmatic judgment that may vary with the context and that does not lend itself to abstract or final solutions. But the threat to meaning caused by the breakdown of truth and autonomy appears as an absolute in the liberal and Cartesian nightmare. It requires a form of security and certainty that is designed to shore up these concepts and eliminate our vulnerability, not merely to balance the risk. The need is intense and the solution proposed by the Enlightenment relies on reason, objectivity, individualism, and universality to respond to that need with the required degree of certainty. Unfortunately, the solution will not work. The security that it offers is an illusion, and one that has often been put to the service of illegitimate hierarchies based on gender or race. We cannot have the certainty and stability that the Enlightenment promised us. The fears are real, and we are right to be afraid, but we must find a way to face our fears without the kind of absolute reassurance that has seemed necessary and that liberals and Cartesians sought to achieve with their models of autonomy and truth. We are, in fact, vulnerable and we cannot hope to eliminate that

236 | Conclusion

vulnerability through the concepts of truth and autonomy. The effort to do so leads only to a loss of the very freedom we were seeking because it represents a vision of life based on fear rather than hope. We must, instead, find a new way to live with our vulnerability creatively.

B. The Interpretive Approach Just as the Cartesian model of truth and the liberal model of autonomy arise from a unified, Enlightenment vision, so too the relational model of truth and the narrative model of autonomy also share a foundation in a broader vision. I will call the view underlying these new models the interpretive approach. If the heart of the Enlightenment vision was liberation from external constraints on the mind and spirit, then the heart of the interpretive approach is a different sort of liberation. To illuminate the nature of this freedom, I must first describe this approach and then examine its relationship to the fears that overtook the Enlightenment project. The interpretive approach may be described in hermeneutic terms as involving both “hermeneutic universalism” and “hermeneutic contextualism.”5 Hermeneutic universalism asserts that all understanding is interpretive: “[t]here can be no appeal to experience, meaning, or evidence that is independent of interpretation or more basic than it.”6 Hermeneutic contextualism asserts that “interpretation always takes place within some context or background—such as webs of belief, a complex of social relations, tradition, or the practices of a form of life.”7 As a result, “all justification is circular . . . truth is relative to some interpretive circle or other.”8 Interpretation, understood in this way, is inherently pragmatic, that is, practical and concerned with application. It is inherently practical because it is tied to a particular context and to the purposes and practices that make up that context.9 This pragmatic focus does not mean that the interpretive approach is hostile to or uninterested in theory, however. Theory is “critical reflection on practice,” and, as such, “theories are instruments for transforming reality, rather than mirroring representations of its putative essential and invariable features.”10 The interpretive approach is also inherently moral and political. “[I]nterpretation is both a social and a political practice.”11 The forms of life that are the stuff of which interpretation is made are supported both by social consensus and by power relations, in varying degrees and com-

Conclusion | 237

binations. They are, in other words, the moral and political fabric of a society. Truth and autonomy, in this interpretive approach, are both names not for things but for interpretive processes. Truth is the interpretive process through which we construct reality; autonomy is the interpretive process through which we construct the stories of our lives. Interpretation, in these two forms, is the central activity through which we make meaning and come to understand ourselves and the world. To say that truth and autonomy are interpretive processes is to make a series of claims about their nature. First, their nature as processes means that they are necessarily incomplete, dynamic rather than static, and the product of human effort. Truth and autonomy are made, rather than found or given: they are a human achievement and it is always possible that we will fail in our efforts to create them. Even when we are succeeding, however, the interpretive approach suggests that our efforts must be continually ongoing in order for the process to continue to exist. And, in this continuing process, there is continual change. Second, the nature of the process as interpretive suggests that it possesses certain characteristics. Being interpretive, the processes of creating truth or autonomy are essentially social. Interpretation requires the use of conceptual categories and intellectual tools (reasoning, conventions of story telling, and so on), all of which are socially constructed and transmitted. Thus, the particular nature of our interpretations is deeply shaped by the social context in which we find ourselves. The social nature of our interpretive tools means that the process of interpretation is always contextual. The claim that interpretation is always contextual is not intended to rule out the possibility that the process of interpretation involves abstraction. Sometimes we interpret by abstracting from context; sometimes we interpret by filling in context. Indeed, the process of interpretation always requires some degree of both abstraction and contextualization, with the salient issue often being the balance between them. The point is not to limit interpretation to contextualization, as opposed to abstraction. The point is that regardless of which technique we may be using in a particular instance of interpretation, that process as a whole can only be understood in its context. In other words, the types of abstraction available to us and the ways in which they can be used are themselves a matter of our context. It is in this sense that the process of interpretation is always contextual.

238 | Conclusion

Moreover, the process of interpretation is social in a second sense: we engage in it through a dialogue, real or imaginary, with others who share the relevant culture. The process is learned and carried out in relations with others. Moreover, the process is one in which we share understanding in a way that allows us to share agency.12 Thus, the social nature of the process means also that it is relational. To say that the processes of creating truth and autonomy are interpretive, then, is to say that they are social, and therefore both contextual and relational. In addition, interpretive processes are also fundamentally evaluative. Interpretation is not a value-neutral activity. It is thoroughly permeated by moral considerations. Indeed, the cognitive aspects are inseparable from the moral ones: our values shape the categories through which we perceive the very facts to be interpreted, along with the interpretive frameworks we can bring to bear upon them. To say that the processes of creating truth and autonomy are interpretive, then, is also to say that they are evaluative in this moral sense. The interpretive nature of the processes also points to a particular understanding of the role of the agent. The person in the interpretive model is neither a passive recipient of truth or autonomy, nor is she an isolated individual generating reality and herself ex nihilo. She is, instead, a moral agent, acting within a web of social relations and cultural constructs, but acting in a creative way that requires the acceptance of personal, moral responsibility for her acts of interpretation.13 Interpretation is a creative act. It is accomplished always within existing cultural and social contexts, but it is accomplished through an act of creativity by an agent. This creativity is the locus of agency for the person. Interpretation is also a moral act. In our interpretations of both reality and ourselves—the search for truth and for autonomy—we exercise our creativity in and through relations to other persons that carry moral implications. In the interpretive approach, we are both active and morally responsible. The broader conception of reason that arose in both the relational model of truth and the narrative model of autonomy also grows out of this interpretive approach. The traditional view of rationality often associates it with logical inference and the process of manipulating the relations between existing bits of knowledge.14 This instrumental, abstract, and formal model of reason grows out of the Enlightenment view described above. But, as I suggested in previous chapters,15 in the interpretive process, the category of reason should cover all of the ways

Conclusion | 239

of making sense of our experience. This would include imaginative as well as inferential processes. Indeed, there is an interesting question— which I do not mean to answer here—whether there are any ways of interpreting (i.e., of creating meaning) that should not be understood as rational (i.e., as making sense). It might be useful to maintain the distinction between rational and other cognitive faculties, in order to be more precise about the different capacities involved in interpretive activity (i.e., to keep a contrast set to help secure the meaning of the term).16 If so, then we could do that within the interpretive approach, simply by specifying which subset of capacities for making meaning will be called rational and why. But the old borders of rationality were dependent on the Enlightenment vision and no longer make sense within an interpretive approach. Even if (perhaps) rationality does not apply to all methods of meaning making, my suggestion is that it will certainly apply to a much broader category than in the Enlightenment vision. Language, in the interpretive approach, is not representation, but it is also not simply a game with no connection to reality. Critics sometimes assume that, once representation has been abandoned, the only possibility remaining is that language is simply disconnected from reality. This view is encouraged by the playful attitude toward cultural forms typical of some types of postmodernism. But both the criticism and the playfulness are in fact based on Cartesian assumptions. The criticism is effective only if one assumes that representation is the only possible relationship between language and reality: if there are other alternatives, then the absence of representation would not necessarily mean that the connection to reality is completely severed. There are powerful arguments that the attitude of playfulness is also a remnant of Cartesianism. “[W]e can see our practices and history as items on hand for playful invention or unflinching conviction only if we first assume that we are, at a basic level, detached subjects before a world that has no necessary connection to our own identity.”17 The rejection of representation does not leave us with this form of detachment from the world. Instead, in an interpretive approach, language is itself a form of interaction with and creation of reality.18 Language is not a picture, but an act, with the same sort of real-world consequences as other actions. It is in this Wittgensteinian sense a “game”: a term that is intended to capture the interactive and regularized nature of the activity, not to connote frivolousness, arbitrariness, or unreality.19

240 | Conclusion

C. The Interpretive Response to the Enlightenment Fears What can the interpretive approach offer us as a response to the fears that drove the Enlightenment vision to seek security and certainty? The interpretive approach will never provide the kind of safety that the Enlightenment versions of truth and autonomy were intended to secure. We are, and will remain, vulnerable. But the interpretive approach offers liberation of a different sort and by a different path: it offers the possibility that we might live with our vulnerability without crippling fear. Indeed, as in the powerful Enlightenment insight, the interpretive approach seeks to make our vulnerability the source of our strength and our hope. First, unlike the Enlightenment vision, the interpretive approach does not attempt to eliminate our vulnerability through the nature of the concepts of truth and autonomy. If truth and autonomy are interpretive processes, then there is always the risk that we will fail to create the kind of truth or autonomy that we need in a given situation. We may find ourselves unable to make meaning about our reality or unable to effectively tell the story of our lives. We may suffer the consequences of being unintelligible to each other, or feeling that we do not understand our world, or experiencing the boundaries of our moral identity as unstable and threatened. The risks are real and the relational model of truth and the narrative model of autonomy do not seek to eliminate or hide them. But the interpretive approach does offer a way to live with these risks. First, the interpretive approach suggests that the risks of disintegration, while real and serious, are not global in the way that the Enlightenment nightmare imagines. We may fail, in any given instance, to secure the sort of truth or autonomy that we need, but we will not find ourselves without truth or autonomy in all situations and for all purposes. The global nature of the threat in the Enlightenment vision arises from the assumption that truth and autonomy are things that one might either have or not have. But if truth and autonomy are processes, and, indeed, if they are the processes of interpretation through which we create meaning, then it is not possible that we could lose them entirely. We will engage in these processes and we will experience them as successful or unsuccessful to varying degrees on different occasions, but they will never simply be lost to us.20 The dangers, from the perspective of the interpretive approach, then, are not conceptual, but practical. We are not in danger of losing meaning altogether, but of failing to make usable meaning in particular cases.

Conclusion | 241

Moreover, the risk of disintegration in any particular case is best understood in pragmatic terms rather than conceptual ones. Our ability to create sufficient truth or autonomy to fulfill the functions of those concepts depends primarily on moral, political, and practical considerations, rather than on the way we understand the concepts themselves. We cannot guarantee that we will have this ability simply by conceiving of truth or autonomy in a particular way: that was the Enlightenment mistake. But neither is our conception of truth or autonomy a bar to such ability. Our ability rests on other considerations, such as the openness of our truth-making processes to alternative perspectives, the equal distribution of resources for truth making and autonomy in our society, and so on. In other words, if we are worried about these risks in a particular case, we should be working on the conditions under which the processes of truth making and autonomy are taking place.

D. Speech And this, of course, is where speech comes into the picture. Speech is crucial to the processes of constructing truth and autonomy. The interpretive approach illuminates the ways in which speech is essential to these processes and suggests that certain understandings of the First Amendment better protect and promote this role of speech. With respect to truth, the interpretive approach offers a relational model of truth that can fulfill the functions for which we need the concept of truth: to establish shared reality, to construct deep critique, to make “working” claims, and to connect to a nonhuman reality. Speech is necessary to all of these endeavors, and all of these endeavors are, to varying degrees, implicated in all of the social institutions affected by free speech law. A relational model of truth suggests that First Amendment doctrine must pay attention to whether the law is serving to help or hinder the performance of these truth functions through speech. Such a focus would lead to modifications in a variety of First Amendment doctrines, including the limits on legitimate government purposes for regulating speech; the boundaries of the public forum doctrine; and the regulation of speech in “nonpublic” fora. With respect to autonomy, the interpretive approach suggests that we should understand our autonomy in terms of the ability to tell our own stories through creative reinterpretation of the materials we are given.

242 | Conclusion

When narrative autonomy becomes our central focus, we can see that speech serves this value in several ways. Speech is the primary mechanism through which we exercise narrative autonomy; speech systems are an essential component in providing opportunities for that exercise and in promoting the development of the capacities necessary for autonomy; and speech is the central democratic activity in the discursive model of democracy that grows out of a relational view of autonomy. Again, a First Amendment informed by this narrative model of autonomy would be focused on the practical questions about how the legal system supports, or fails to support, the role of speech in promoting narrative autonomy. Some of the areas of legal doctrine that would require modification include the understanding of symbolic speech; the role of commercial speech; campaign finance regulation; and the application of the First Amendment to some government-supported speech. To the extent possible, then, free speech law should focus our attention on the pragmatic questions the interpretive approach identifies as central. The purpose of the First Amendment is to protect our ability to use speech for the essential tasks of meaning making captured by our models of relational truth and narrative autonomy. Truth theories of free speech have always rested on the fundamental insight that speech is about understanding, and autonomy theories of free speech have always rested on the fundamental insight that speech is about freedom. An interpretive approach preserves these deep insights and demonstrates that, in a social constructionist world, the search for understanding and the search for freedom are connected at the deepest level. Indeed, it is in such a search that we find the only answers to the fears that truth and autonomy theory were designed to address.

E. Interdependence and Commitment The interpretive approach thus suggests that we can and must live with these risks and this vulnerability. Indeed, it was the desire to insulate ourselves from the risks that was the source of many of our problems. In attempting to construct concepts that would provide enough certainty and stability to eliminate these risks, we were forced to rely upon ideas like objectivity and individualism. Those ideas, in turn, construct the situation in a way that generates insoluble dilemmas (like arbitrariness versus determinism) and robs us of the very tools we need to live with those

Conclusion | 243

risks. As John Dewey advised, “intellectual progress usually occurs through sheer abandonment of questions together with both of the alternatives they assume—an abandonment that results from their decreasing vitality and a change of urgent interest. We do not solve them; we get over them.”21 In the flight from vulnerability, we cut ourselves off from each other and from the world; we purchase an illusion of security at the price of freedom; and we lose sight of the dream that was the very foundation of the Enlightenment vision. Ironically, perhaps, the only way to overcome our vulnerability is to stop trying to eliminate it. We cannot avoid it or run away from it, but we can live constructively with it. And the effort to do so will open up possibilities for liberation—liberation from fear—that will allow us to recapture the best of our Enlightenment heritage. How does one live constructively, creatively, with the vulnerability to these risks? The interpretive approach suggests that we must understand our situation as one in which we are fundamentally interdependent: we are dependent on other people and on the rest of the world, as they are on us. This interdependence is the source of our vulnerability, but it is also the source of our strength because it is in and through these relationships that truth and autonomy are made. We live constructively with our vulnerability when we embrace our interdependence, celebrate it, enrich it, and pay attention to the health of the relationships that make it up. In place of the isolated security of the Enlightenment solution, the interpretive approach offers interdependence. The interpretive approach also suggests that our fundamental situation is one of uncertainty in which we should never expect final answers or solutions. This uncertainty is a source of vulnerability, but it can also be a source of strength when understood in a particular way. Our answers are not given to us in some way that simply commands our assent. Instead, we build our answers through a process of commitment—moral and political as well as epistemological—and through the struggles over such commitments that is our social and political life. This process of commitment is at the heart of our ability to construct truth and autonomy and the shared life they allow. “The wish to find standards that stand on their own is unfulfillable. It is we who must stand on, and stand up for, them.”22 In place of certainty, the interpretive approach offers commitment. Interdependence and commitment, then, must take the place of certainty and security, and we must expand our understanding of free speech

244 | Conclusion

so that it captures this new vision. Freedom of speech is a worthy cultural icon; we are right to regard it as a central ideal. It represents the Enlightenment goals of understanding ourselves and the world, of seeking truth and autonomy, and these are goals we must not abandon. But we need to reinterpret the truth and autonomy we seek in order to escape the trap of illusory security and certainty. Freedom of speech should be seen as our central cultural practice of interdependence and commitment: it is the way we come together to construct, understand, and apply the commitments that constitute our world and our selves. When we see it in this light, free speech is a precious cultural inheritance and our greatest hope for the future.

Notes

notes to chapter 1 1. The other major contending approach to freedom of speech is sometimes characterized as “negative” theory. This approach looks not at why speech is particularly worth protecting but at why government regulation of speech is particularly dangerous. One of the primary exponents of negative theory is Frederick Schauer. See, e.g., Frederick Schauer, Free Speech: A Philosophical Inquiry (Cambridge: Cambridge Univ. Press 1982). 2. Frederick Schauer has characterized the truth theory as “throughout modern history the ruling theory in respect of the philosophical underpinnings of the principle of freedom of speech.” Id. at 16. 3. John Milton, Areopagitica in John Milton: Complete Poems and Major Prose 716, 746–47 (Merritt Y. Hughes, ed.) (Indianapolis: Odyssey Press 1957). 4. Id. at 739. 5. John Stuart Mill, On Liberty in The Utilitarians 473, 491 (New York: Anchor Books 1973). 6. Id. at 495. Mill also argues that silencing intellectual inquiry will lead to the deterioration of the character of the people—a form of autonomy argument. See id. at 507–9. 7. Id. at 509–15. 8. Id. at 522. 9. Id. at 530. 10. 12 Hening’s Laws of Virginia 84, 85–86 (January 1786). 11. James Madison, Memorial and Remonstrance in The Mind of the Founder: Sources of the Political Thought of James Madison 14–15 (Marvin Meyers, ed.) (Indianapolis: Bobs-Merrill 1973). 12. Abrams v. United States, 250 U.S. 616, 630 (1919) (Holmes, J., dissenting). For a nuanced analysis of this opinion, see Vincent Blasi, Reading Holmes through the Lens of Schauer: The Abrams Dissent, 72 Notre Dame Law Review 1343 (1997).

245

246 | Notes to Chapter 1 13. Dennis v. United States, 341 U.S. 494, 550 (1951) (Frankfurter, J., concurring). 14. Whitney v. California, 274 U.S. 357, 375 (1927) (Brandeis, J., concurring). 15. See New York Times v. Sullivan, 376 U.S. 254, 279 n. 19 (1964). 16. See Time v. Hill, 385 U.S. 374, 407–8 (1967). 17. Red Lion Broadcasting v. FCC, 395 U.S. 367, 390 (1969). 18. See Milkovitch v. Lorain Journal Co., 497 U.S. 1, 17–19 (1990). 19. For examples of this criticism, see Schauer, supra n.1 at 23–24. 20. See Paul J. Barber, Executing Two Tasks at Once in Acquisition and Performance of Cognitive Skills (Ann M. Colley & John R. Beech, eds.) (Chichester: Wiley 1989) (on the effect of multiple demands on our attention); Martha Augustinos & Iain Walker, Social Cognition: An Integrated Introduction 106–13 (London: Sage Publications 1995) (on how preexisting beliefs, perceived interests, and stereotypes affect perception). 21. See Paul H. Brietzke, How and Why the Marketplace of Ideas Fails, 31 Valparaiso University Law Review 951, 961–68 (1997). 22. Although the truth theory can withstand some skepticism about objectivity, as well as certainty, some minimal level of objectivity is necessary in order for it to be possible to determine that one position is more true than another. See Kent Greenawalt, Free Speech Justifications, 89 Columbia Law Review 119, 132 (1989). 23. For an argument that cuts back the truth theory to a narrow range of scientific knowledge in response to this skeptical attack, see Christopher Wonnell, Truth and the Marketplace of Ideas, 19 University of California at Davis Law Review 669 (1986). 24. Recognizing the questionableness of these assumptions does not necessarily make the truth theory simply untenable, but it does suggest the need for “[a] sensitive assessment [which] requires subcategorization among domains of truth and audiences, and recognition of degrees of possible interference.” Greenawalt, supra n.22 at 136. 25. See Martin Redish, The Value of Free Speech, 130 University of Pennsylvania Law Review 591, 603–4 (1982). 26. See Vincent Blasi, Free Speech and Good Character, 46 University of California at Los Angeles Law Review 1567, 1571–74 (1999); Thomas Scanlon, A Theory of Freedom of Expression, 1 Philosophy & Public Affairs 204, 216 (1972). 27. See C. Edwin Baker, Human Liberty 53–54 (New York: Oxford Univ. Press 1989). 28. See David A. J. Richards, Toleration and the Constitution 166–74 (New York: Oxford Univ. Press 1986). 29. Blasi, supra n.26 at 1569. In this wonderful speech, Professor Blasi pro-

Notes to Chapter 1 | 247 poses that the connection to character could be seen as a rationale independent of the three traditional free speech theories based on truth, autonomy, and democracy. See id. at 1568–69. He does not place the character argument within the autonomy theory because he defines the autonomy theory narrowly as holding that “the liberty to express one’s thoughts and to form them by unrestricted reading and listening is an essential attribute . . . of human autonomy—of what it means to be a self-directed person possessed of human dignity.” Id. Thus, he sees autonomy theory as focusing on a characteristic necessarily inherent in every human being, while his character approach focuses on characteristics that are goals far from universally achieved. His definition of autonomy theory corresponds to the second type of autonomy theory that I will discuss below. I have defined autonomy theory more broadly, however, to also include arguments that focus on the role of speech in the development of character, at least where those characteristics identified are ones traditionally associated with autonomy. I believe this description fits most of the character traits on Professor Blasi’s list, with one notable exception. In his expanded list of characteristics, Professor Blasi includes “intellectual and cultural empathy.” Id. at 1571. This is not a characteristic associated with the traditional view of autonomy that I will outline in chapter 2 and criticize in chapter 3. Its inclusion on the list indicates that Professor Blasi’s character argument may be based on a different conception of autonomy—perhaps more like the one I will suggest in chapter 6. 30. Cf. Richards, supra n.28 at 167 (“rationality itself flows from our capacities for speech and writing,” where rationality is defined in terms of a person’s development of her own conception of a life well lived). 31. Thomas Emerson, Toward a General Theory of the First Amendment, 72 Yale Law Journal 877, 879 (1963). 32. See John Rawls, Political Liberalism 32–33 (New York: Columbia Univ. Press 1993). See also David A. J. Richards, Constitutional Legitimacy, the Principle of Free Speech, and the Politics of Identity, 74 Chicago-Kent Law Review 779, 796 (1999) (“That interest is nothing less than the free exercise of the moral powers of their reason through which persons give enduring value to their lives and communities.”). 33. See Richards, supra n.28 at 168–69. For a similar argument based on the foundations of political legitimacy (and avoiding the overtly contractarian elements) see Baker, supra n.27 at 48–50. 34. See C. Edwin Baker, Scope of the First Amendment Freedom of Speech, 25 University of California at Los Angeles Law Review 964, 998 (1978). 35. John Locke, A Letter concerning Toleration (1689) in The Second Treatise of Government and a Letter concerning Toleration 125, 129 (J. W. Gough, ed.) (Oxford: Blackwell 1966, 3d ed.).

248 | Notes to Chapter 1 36. See id. at 154. 37. Mill, supra n.5 at 507–8. 38. See id. at 552–53. 39. See Leonard W. Levy, Jefferson and Civil Liberties: The Darker Side 42–69 (Cambridge: Harvard Univ. Press 1963). 40. James Madison, Memorial and Remonstrance against Religious Assessments in The Writings of James Madison Vol II 1783–1787 at 184 (Gaillard Hunt, ed.) (New York: Putnam’s 1901). 41. See Richards, supra n.32 at 798. 42. Stanley v. Georgia, 394 U.S. 557, 565 (1969). 43. Id. 44. FNB v. Bellotti, 435 U.S. 765, 804–5 (1978) (White, J., dissenting). 45. Virginia State Board of Pharmacy v. Virginia Citizens Consumer Council, 425 U.S. 748, 769 (1976). 46. 44 Liquormart, Inc. v. Rhode Island, 517 U.S. 484, 510 (1996). 47. Id. at 518. 48. See Virginia State Board of Pharmacy, supra n.45 at 763–64. 49. The Court has, however, been notoriously unwilling to accept other aspects of the theory, most notably the focus on protecting nontraditional forms of speech as self-expression. For example, Chief Justice Warren explicitly rejected “the view that an apparently limitless variety of conduct can be labeled ‘speech’ whenever the person engaging in the conduct intends thereby to express an idea.” U.S. v. O’Brien, 391 U.S. 367, 376 (1968). And the Court has consistently rejected the idea that obscenity might be protected speech because it serves the self-realization or self-fulfillment of its consumers. For a criticism of the Court’s approach based on an autonomy theory, see David A. J. Richards, Free Speech and Obscenity Law: Toward a Moral Theory of the First Amendment, 123 University of Pennsylvania Law Review 45, 72, 75 (1982). 50. See Greenawalt, supra n.22 at 143–44. 51. Schauer, supra n.1 at 56. 52. Id. at 57–58; see also Robert Bork, Neutral Principles and Some First Amendment Problems, 47 Indiana Law Journal 1, 25 (1971). This criticism does not apply to the work of Professor Baker, who is one of the few autonomy theorists to take on this challenge directly. Baker argues that speech is protected not just because it serves the values of self-realization and self-determination but because it serves those values in a particular way: “nonviolently and noncoercively.” See Baker, supra n.27 at 46. Any and all activities that serve these values in these ways should then receive First Amendment protection. See id. at 70. This argument provides precisely the type of answer needed by an autonomy theory: it specifies which activities will be protected and why. Moreover, the limits imposed by Baker are not external to his autonomy theory; rather, they are themselves required by the very model of autonomy that justifies the protection

Notes to Chapter 1 | 249 in the first place. It is respect for the autonomy of the listener that sets the limits on the activities that can qualify as protected. While one might raise objections to the particular answer Baker gives, he is not subject to the general objection raised in the text. I believe the more important objection to his theory concerns the model of autonomy that informs both the positive grant of liberty and the principle defining its limits. I will take up these issues in the next chapter. 53. For criticisms of Rawlsian social contract theory, see Richard Posner, The Problematics of Moral and Legal Theory 51 (Cambridge: Harvard Univ. Press 1999) (criticizing the risk aversion of the parties); Seyla Benhabib, Situating the Self: Gender, Community, and Postmodernism in Contemporary Ethics (New York: Routledge 1992) (criticizing the veil of ignorance). See generally Michael Sandel, Liberalism and the Limits of Justice (Cambridge: Cambridge Univ. Press 1982) (criticizing the conception of the self, of community, and of moral reasoning used by the parties). 54. A. John Simmons, Political Consent in The Social Contract Theorists 121, 134–35 (Christopher W. Morris, ed.) (Lanham, MD: Rowman and Littlefield 1999). 55. See Greenawalt, supra n.22 at 152–53. Baker’s argument may avoid this last difficulty. Baker appears to be suggesting that the agreement of autonomous persons to some restrictions on their autonomy is not the relevant standard: a government that fails to respect their autonomy simply does not generate an obligation to obey the law because it fails to treat them as the sort of autonomous subjects who could have such obligations. See Baker, supra n.27 at 48–50. This raises an interesting question: if it is true that such autonomous subjects might choose (unanimously) to forsake some of their autonomy in particular instances in order to achieve other ends, wouldn’t the government also be failing to respect their autonomy if it did not allow them to do so? I am not certain how Baker would resolve this conflict. Only if he would refuse to allow such choices to lead to restrictions on future autonomy can he avoid the objection in the text. 56. Alexander Meiklejohn, Political Freedom 109 (New York: Oxford Univ. Press 1960) (excerpt from testimony presented November 14, 1955, to the Senate Subcommittee on Constitutional Rights). 57. Id. at 118–19. 58. Alexander Meiklejohn, The First Amendment Is an Absolute, 1961 Supreme Court Review 245, 255. 59. Bork, supra n.52 at 26. For a middle path that defines as political all speech that is “both intended and received as a contribution to public deliberation about some issue,” see Cass R. Sunstein, Democracy and the Problem of Free Speech 130 (New York: Free Press 1993).

250 | Notes to Chapter 1 60. See Schauer, supra n.1 at 42. 61. Sunstein’s approach actually combines elements of both forms of democracy theory. For his discussion of the usefulness of speech to the listener in producing better-informed decisions, see Sunstein, supra n.59 at 243. 62. See Sunstein, supra n.59 at 18–19. 63. See id. at 244. 64. See id. at 244–45. 65. Id. at 245. 66. James Madison, Report of 1800 (January 7, 1800) in 17 Papers of James Madison 344, 341 (David Mattern, et al., eds.) (Charlottesville: Univ. Press of Virginia 1991). 67. New York Times v. Sullivan, 376 U.S. 255, 270 (1964). 68. See Harry Kalven, The New York Times Case: A Note on ‘The Central Meaning of the First Amendment,’ 1960 Supreme Court Review 1. Justice Brennan confirmed this reading of his opinion for the majority in William Brennan, The Supreme Court and the Meiklejohn Interpretation of the First Amendment, 79 Harvard Law Review 1 (1965). 69. Whitney v. California, 274 U.S. 357, 372 (1927) (Brandeis, J., concurring). 70. See Abrams v. U.S., 250 U.S. 616, 630 (1919). 71. On Holmes’s pragmatism, see Thomas C. Grey, Holmes and Legal Pragmatism, 41 Stanford Law Review 787 (1989). On his skepticism, see Yosal Rogat & James M. O’Fallon, Mr. Justice Holmes: A Dissenting Opinion—The Speech Case, 36 Stanford Law Review 1349 (1984). 72. See Stanley Ingber, The Marketplace of Ideas: A Legitimizing Myth, 1984 Duke Law Journal 1, 8 n.30. 73. Schauer, supra n.1 at 41. 74. See Meiklejohn, supra n.56 at 110. 75. See Baker, supra n.27 at 25, 28. 76. See Sunstein, supra n.59 at 243. 77. See Virginia Pharmacy, supra n.45 at 763–64 (applying this argument to economic life and commercial speech). 78. Schauer, supra n.1 at 42. 79. See David C. Williams, The Militia Movement and the Second Amendment: Conjuring with the People, 81 Cornell Law Review 879 (1996). 80. See, e.g., Martin H. Redish & Gary Lippman, Freedom of Expression and the Civic Republican Revival in Constitutional Theory: The Ominous Implications, 79 California Law Review 267 (1991). 81. See Redish, supra n.25 at 600–604. Redish defines the ultimate value as self-realization, and includes autonomy as only one aspect of it. I am defining autonomy much more broadly, so that his entire value of self-realization is included in the collection of concerns that I identify as autonomy.

Notes to Chapter 2 | 251

notes to chapter 2 1. See Mary Hawkesworth, From Objectivity to Objectivism: Feminist Objections in Rethinking Objectivity 151, 151 (Allan Megill, ed.) (Durham: Duke Univ. Press 1994). 2. See Alison M. Jagger & Susan R. Bordo, Introduction in Gender/ Body/Knowledge: Feminist Reconstructions of Being and Knowing 1, 3 (New York: Routledge 1989). See also Jane Flax, Postmodernism and Gender Relations in Feminist Theory in Feminism/Postmodernism 39, 41–42 (Linda J. Nicholson, ed.) (New York: Routledge 1990). 3. See Bertrand Russell, The Problems of Philosophy 126–30 (New York: Holt 1912). See generally Richard Rorty, Philosophy and the Mirror of Nature (Princeton: Princeton Univ. Press 1979) (describing this representational model and criticizing it). 4. Some feminists have seen this rationalist bias as tied to the division of mental and manual labor, and the preference for mental labor, that gives rise to a hierarchical structure in capitalist systems of production. See Hilary Rose, Beyond Masculinist Realities: A Feminist Epistemology for the Sciences in Feminist Approaches to Science 57, 69 (Ruth Bleier, ed.) (New York: Pergamon Press 1986). For discussion of how the rationalist bias affects law, see Lynne Henderson, Legality and Empathy, 85 Michigan Law Review 1574 (1987) and Robin West, Love, Rage, and Legal Theory, 1 Yale Journal of Law & Feminism 101 (1989). 5. Empiricism is perhaps the most common and influential version of Cartesianism. Empiricism maintains that all knowledge is derived from experience. See Bertrand Russell, supra n.3 at 73. See, e.g., John Locke, An Essay concerning Human Understanding (1694) (Clarendon Press ed. 1975). Rationalists, like Descartes himself, on the other hand, place more reliance on reason acting independent of sense perception as a mechanism for acquiring knowledge. See Rene Descartes, Discourses on the Method of Rightly Conducting the Reason and Seeking Truth in the Sciences, Part IV 62–63 (1637) (Anchor Books ed. 1974). Cartesianism, as I define it, is broad enough to encompass both varieties of theory. Cf. Dennis Patterson, Postmodernism/Feminism/Law, 77 Cornell Law Review 254, 266 (1992) (rationalism and empiricism are both foundationalist). 6. See Allan Megill, Introduction: Four Senses of Objectivity in Rethinking Objectivity, supra n.1 at 1. 7. Jagger & Bordo, supra n.2 at 3. 8. See Keith Lehrer, Theory of Knowledge 15 (Boulder, CO: Westview Press 1990). 9. See Syla Behabib, Epistemologies of Postmodernism: A Rejoinder to JeanFrancois Lyotard in Feminism/Postmodernism, supra n.2 at 107, 110–11.

252 | Notes to Chapter 2 10. See, e.g., Baker, supra chapter 1 n.27 at 12–14 (criticizing marketplace theory for relying on an objective model of truth reached by reason); David Cole, Agon at Agora: Creative Misreadings in the First Amendment Tradition, 95 Yale Law Journal 857, 876–77 (1986) (ascribing a belief in objective truth to Milton and Mill); Kent Greenawalt, supra chapter 1 n.22 at 131–32 (suggesting that the truth theory of free speech can survive much skepticism about the objectivity of truth, but not a total loss of that ideal); Stanley Ingber, supra chapter 1 n.72 at 15 (identifying the objective nature of truth as one of the assumptions of the “marketplace of ideas” model). 11. See Mill, supra chapter 1 n.5 at 503. 12. For a contrary interpretation of Mill’s use of truth, see Paul Chevigny, More Speech 6 (Philadelphia: Temple Univ. Press 1988) (“truth in any but a very special subjective sense matters not at all to most of Mill’s argument.”). 13. See Mill, supra chapter 1 n.5 at 494. 14. Id. at 518. 15. Chaplinsky v. New Hampshire, 315 U.S. 568, 572 (1942). 16. Whitney v. California, 274 U.S. 357, 375 (1927) (Brandeis, J., concurring), overruled by Brandenburg v. Ohio, 395 U.S. 444 (1969). 17. Milkovitch v. Lorain Journal Co., 497 U.S. 1, 19 (1990). 18. Id. at 21. 19. Abrams v. U.S., 250 U.S. 616, 630 (1919) (Holmes, J., dissenting). 20. See supra chapter 1 at 26. 21. See Vincent Blasi, Reading Holmes through the Lens of Schauer: The Abrams Dissent, 72 Notre Dame Law Review 1343, 1348–51 (1997). 22. Milkovitch, supra n.17 at 18 (quoting from Cianci v. New Times Publishing Co., 639 F.2d 54, 62 n. 10 (2d Cir. 1980)). 23. Id. 24. Int’l Brotherhood of Electrical Workers, Local 501 v. NLRB, 181 F.2d 34, 40 (2d Cir. 1950), aff’d 341 U.S. 695 (1951). 25. For a description of the development of the doctrinal confusion concerning symbolic speech, see Susan H. Williams, Content Discrimination and the First Amendment, 139 University of Pennsylvania Law Review 615, 650–54 (1991). The Court’s support for symbolic speech in the flag burning cases is not inconsistent with my claim that nonpropositional speech forms are disfavored. The flag burning cases are better understood as condemnation of the government purposes involved in the laws at issue than as sympathy for symbolic forms of speech. See Texas v. Johnson, 491 U.S. 397, 407–20 (1989); Elena Kagan, Private Speech, Public Purposes: The Role of Governmental Motive in First Amendment Doctrine, 63 University of Chicago Law Review 413, 504 (1996). 26. See, e.g., Texas v. Johnson, 491 U.S. at 431–32 (Rehnquist, C. J., dissenting) (describing flag burning as an inarticulate grunt).

Notes to Chapter 2 | 253 27. See, e.g., Steven D. Smith, Skepticism, Tolerance, and Truth in the Theory of Free Expression, 60 Southern California Law Review 651, 663 (1987). 28. See Joseph W. Singer, The Player and the Cards: Nihilism and Legal Theory, 94 Yale Law Journal 1, 4 n.8 (1984). 29. See Sandra Harding, Whose Science? Whose Knowledge? 152–53 (Ithaca, NY: Cornell Univ. Press 1991). 30. The skeptical interpretation of the Court’s opinions, while not inconsistent with their Cartesianism, certainly is inconsistent with my interpretation of their references to the truth theory. Because I believe that the Justices are neither dishonest nor confused when they refer to truth, but actually committed to the theory they say they are, I find the opinions to be better explained as based on faith in Cartesianism than as based on skepticism. 31. Joel Feinberg, Autonomy in The Inner Citadel 27, 28 (John Christman, ed.) (New York: Oxford Univ. Press 1989). 32. The literature on autonomy often distinguishes between three forms of self-determination in different realms: political autonomy, moral autonomy, and personal autonomy. Political autonomy will be relevant when we return to democracy theory, but for the remainder of this section, I will be discussing moral and personal autonomy. Moral autonomy concerns self-determination of one’s moral values, while personal autonomy concerns self-determination in life choices left open by moral constraints. See Marilyn Friedman, Autonomy and Social Relationships: Rethinking the Feminist Critique in Feminists Rethink the Self 40, 41 (Diana T. Meyers, ed.) (Boulder, CO: Westview Press 1997). 33. See Jean Grimshaw, Autonomy and Identity in Feminist Thinking in Feminist Perspectives in Philosophy 90, 91 (Morwenna Griffiths & Margaret Whitford, eds.) (Bloomington, IN: Indiana Univ. Press 1988). 34. Marilyn Friedman, Self-Rule in a Social Context: Autonomy from a Feminist Perspective in Freedom, Equality, and Social Change 158, 159 (Creighton Peden & James P. Sterba, eds.) (Lewiston/Lampeter/Queenston: Mellen 1989). 35. The popular image of autonomy may include more substantive elements as well. The exaggerated image of the “self-made man,” which enshrines a vision of autonomy as independence from all forms of commitment to relationships, represents an extreme form of atomistic individualism. The feminist critique has often been directed against this cultural image, see Friedman, supra n.32 at 42–47. The models of philosophers have, however, tended to eschew this substantive view in favor of a more process-oriented approach. One interesting question raised by feminist criticisms is the extent to which the commitment to a certain process of autonomous decision making may systematically lead to the adoption of the substantive values of independence and atomism. 36. For the classic identification approach, see Harry Frankfurt, Freedom of

254 | Notes to Chapter 2 the Will and the Concept of a Person, 68 Journal of Philosophy 5 (1971). For more recent writing including both the identification and historical approaches, see Gerald Dworkin, The Theory and Practice of Autonomy (New York: Cambridge Univ. Press 1988); Robert Young, Personal Autonomy: Beyond Negative and Positive Liberty (London: Croam Helm 1986). For an argument in favor of an historical approach and against the identification approach that focuses on “second-order desires,” see Alfred R. Mele, Autonomous Agents: From Self-Control to Autonomy 65–69, 156–76 (New York: Oxford Univ. Press 1995). 37. See Catriona Mackenzie & Natalie Stoljar, Introduction: Autonomy Refigured in Relational Autonomy: Feminist Perspectives on Autonomy, Agency, and the Social Self 1, 13 (Catriona Mackenzie & Natalie Stoljar, eds.) (New York: Oxford Univ. Press 2000). 38. See Mele, supra n.36 at 177–255 (arguing that there is a notion of autonomy that is consistent both with compatibilist visions (which argue that autonomy or free will can be understood even in the presence of physical determinism) and noncombatibilist or libertarian visions (which argue that free will is only possible in the absence of determinism)). 39. See Immanuel Kant, Foundations of the Metaphysics of Morals 29, 51 (Lewis White Beck, trans.) (Indianapolis: Bobs-Merrill 1959). 40. See, e.g., Feinberg, supra n.31 at 45. 41. See Dworkin, supra n.36 at 6. 42. See, e.g., Young, supra n.36 at 41. 43. See, e.g., Mele, supra n.36 at 228–29 (describing the process of becoming a free agent in terms of cognitive and motor skills with no reference to social interaction or social relations as necessary to it). 44. Paul Benson, Feminist Second Thoughts about Free Agency, 5 Hypatia 47, 49 (Fall 1990). 45. See Friedman, supra n.32 at 57 (“Mainstream theorists admit that people must be socialized for self-reflection and the other capacities required for the achievement of autonomy (could anyone really doubt this?), but mainstream theories do not generally define autonomy as intrinsically social. . . .”). 46. See Redish, supra chapter 1 n.25 at 606–7. 47. See id. at 610. 48. See Richards, supra chapter 1 n.28 at 166–74. 49. Id. at 171. 50. Id. at 71–72. 51. See id. 52. See id. at 80–81. 53. See id. at 81–82. Richards distinguishes his approach from “the Cartesian picture of the person as pure knower” on the grounds that such a picture does not take seriously enough “the practical roots of our rational and reason-

Notes to Chapter 2 | 255 able wills.” Id. at 78–79. This modification does represent a step away from the connected Cartesian/liberal vision and a move toward a more pragmatic approach, as evidenced by his citations to Dewey and Rorty. Id. at 79 n.54. Indeed, this move leads Richards to assert the interdependence of epistemology and morality in a distinctly non-Cartesian fashion. See id. Nonetheless, this rejection of Cartesianism grows more out of a desire to place the liberal model of autonomy at the center of theory than from an alternate model of autonomy. His real point is that the moral theory takes primacy over and provides the foundation for the epistemology, not that we need a different model of either truth or autonomy. See id. at 79 (quoting Santayana as saying, “What enables men to perceive the unity of nature is the unification of their own wills.”). Richards’s reliance on a fundamentally Kantian approach makes clear that, despite his criticisms of Cartesianism, he is firmly within the liberal tradition in moral and political thought. 54. See id. at 57–63. 55. See Baker, supra chapter 1 n.27 at 53–54 (describing creative uses of speech and how they generate new opportunities and capacities). 56. See id. at 53 (“any time a person engages in chosen, meaningful conduct . . . the conduct usually expresses and further defines the actor’s identity and contributes to his or her self-realization.”). 57. See id. at 49. 58. Id. at 52. 59. See id. at 59. 60. See id. at 56. 61. See id. 62. Id. 63. See id. at 12. 64. Id. at 278 fn. 65. See C. Edwin Baker, Injustice and the Normative Nature of Meaning, 60 Maryland Law Review 578, 583, 602 (1994). 66. Baker, supra chapter 1 n.27 at 278 fn. 67. See C. Edwin Baker, Of Course, More Than Words, 61 University of Chicago Law Review 1181, 1197 (1994) (respect for autonomy does not deny social construction, but it does insist that people have the right to participate in the processes of social construction through their expression). This combination raises the important question of how autonomy can continue to do this normative work once one has recognized the breadth and depth of social construction of identity. Part of the purpose of this book in general, and chapter 6 in particular, is to address that question. My answer, however, will require autonomy to be understood in terms different from the liberal focus on choice. 68. See Baker, supra chapter 1 n.27 at 52–54.

256 | Notes to Chapter 2 69. See 44 Liquormart Inc. v. Rhode Island, 517 U.S. 484, 518 (1996) (Thomas, J., concurring). 70. See id. at 510. 71. FCC v. Pacifica Foundation, 438 U.S. 726, 746 (1978) (quoting from Chaplinsky v. New Hampshire, 315 U.S. 568, 572 (1942)). 72. Texas v. Johnson, 491 U.S. 397, 432 (Rehnquist, C. J., dissenting). notes to chapter 3 1. See Deborah L. Rhode, Justice and Gender 5 (Cambridge: Harvard Univ. Press 1989); John Christman, Feminism and Autonomy in “Nagging” Questions: Feminist Ethics in Everyday Life 17, 18 (Dana E. Bushnell, ed.) (Lanham, MD: Rowman & Littlefield 1995). 2. See Susan H. Williams & David C. Williams, A Feminist Theory of Malebashing, 4 Michigan Journal of Gender & Law 35, 66–70 (1996). 3. See Christman, supra n.1 at 18. 4. Feminists disagree about many other things, however, and there are distinctive schools of feminist thought that give particular—and different—answers to some central questions. For example, feminist theory is often divided into categories, including liberal feminism, difference (or cultural) feminism, dominance feminism, radical feminism, Marxist and socialist feminism, and postmodernist feminism. For excellent summaries of these approaches (and others), see generally Rosemary Tong, Feminist Thought: A More Comprehensive Introduction (Boulder, CO: Westview 1998, 2d ed.). Like many, perhaps most, contemporary feminists, I do not propose to limit myself to one of these categories. I believe that all of the feminist theories have value and that different theoretical tools and frameworks are better for addressing different problems. 5. Peter L. Berger & Thomas Luckman, The Social Construction of Reality (New York: Doubleday 1966). 6. Ruth Hubbard, Some Thoughts about the Masculinity of the Natural Sciences in Feminist Thought and the Structure of Knowledge 1 (Mary McCanney Gergen, ed.) (New York: New York Univ. Press 1988). 7. See Naomi Schemann, Individualism and the Objects of Psychology in Discovering Reality: Feminist Perspectives on Epistemology and Metaphysics 225, 229 (Sandra Harding & Merrill Hintikka, eds.) (Boston: Kluwer 1983). 8. Kenneth J. Gergen, Feminist Critique of Science and the Challenge of Social Epistemology in Feminist Thought, supra n.6 at 29. 9. The Cartesian method attempts to correct for this particularity by insisting on the replicability of results. This process may correct for some of the more individualized factors that affect perception and recognition, but it does nothing to correct for very widely shared blindnesses caused by deeply held cultural as-

Notes to Chapter 3 | 257 sumptions or conceptual categories. See Sandra Harding, Whose Science? Whose Knowledge? 143–46 (Ithaca, NY: Cornell Univ. Press 1991). Replicability also does nothing to explain why we should regard these broadly shared cultural assumptions as a more reliable basis for knowledge than the more idiosyncratic ones. 10. For a thorough assessment of research on cognitive and psychological differences between the sexes, see generally Ruth Bleier, Science and Gender: A Critique of Biology and Its Theories on Women (New York: Pergamon Press 1984). 11. Sandra Harding, Introduction: Is There a Feminist Method? in Feminism and Methodology 1, 6 (Sandra Harding, ed.) (Bloomington, IN: Indiana Univ. Press 1987). 12. See Martha Chamallas, Feminist Constructions of Objectivity: Multiple Perspectives in Sexual and Racial Harassment Litigation, 1 Texas Journal of Women & Law 95, 113–17 (1992). 13. See Alison Jagger, Love and Knowledge: Emotion in Feminist Epistemology in Gender/Body/Knowledge, supra chapter 2 n.2 at 146. 14. See Dorothy E. Smith, Women’s Perspective as a Radical Critique of Sociology in Feminism and Methodology, supra n.11 at 93. 15. Jagger, supra n.13 at 164. 16. For an argument that science has internalized this critique but that legal theory still partakes of the positivism of an earlier scientific tradition, see Jeanne L. Schroeder, Abduction from the Seraglio: Feminist Methodologies and the Logic of Imagination, 70 Texas Law Review 109, 118, 128 n.44, 153 n.108 (1991). 17. Although even this claim is highly suspect in light of the very interesting literature on the nature of sexual difference as a continuum rather than as the two exclusive categories we usually assume. See, e.g., Marilyn Friedman, The Unholy Alliance of Sex and Gender, 27 Metaphilosophy 78 (Nos. 1 & 2, January/April 1996). 18. See Deborah L. Rhode, Introduction in Theoretical Perspectives on Sexual Difference 1, 4 (Deborah L. Rhode, ed.) (New Haven: Yale Univ. Press 1990) (“For example, we cannot understand sex-based differentials of height, weight, and physical strength without considering the influence of diet, dress, division of labor, and so forth.”). 19. Peggy Reeves Sanday, The Reproduction of Patriarchy in Feminist Anthropology in Feminist Thought and the Structure of Knowledge, supra n.6 at 49, 53. 20. Compare, for example, Jane Flax, Political Philosophy and the Patriarchal Unconscious: A Psychoanalytic Perspective on Epistemology and Metaphysics in Discovering Reality, supra n.7 at 258–61, with Hawkesworth, supra chapter 2 n.1 at 151.

258 | Notes to Chapter 3 21. See generally Myra Sadker & David Sadker, Failing at Fairness: How Our Schools Cheat Girls (New York: Scribner’s 1994). 22. See generally Janet Clark, Getting There: Women in Political Office, 515 Annals American Political Science Society 63 (Janet K. Boles, ed.) (1991). 23. See Catharine MacKinnon, Feminism Unmodified 32–45 (Cambridge: Harvard Univ. Press 1987). I do not mean to suggest, of course, that Cartesianism is the only, or even necessarily the primary, cause for these various phenomena, but it is a contributing factor. 24. Lorraine Code, What Can She Know? Feminist Theory and the Construction of Knowledge 77 (Ithaca: Cornell Univ. Press 1991). 25. Seyla Benhabib, Situating the Self: Gender, Community, and Postmodernism in Contemporary Ethics 156 (New York: Routledge 1992). 26. Jennifer Nedelsky, Reconceiving Autonomy: Sources, Thoughts, and Possibilities, 1 Yale Journal of Law & Feminism 7, 12 (1989). 27. Benhabib, supra n.25 at 157. 28. Carole Pateman, The Sexual Contract 1–2 (Stanford, CA: Stanford Univ. Press 1988). 29. See id. at 55–56. 30. See Friedman, supra chapter 2 n.32 at 52–53. It is worth noting that the critiques are often addressed to the older materials in the liberal philosophical tradition rather than to these contemporary accounts. Thus, while it is reasonable to point out that the critique does not fit the work of contemporary theorists very well, it is also important to remember that it never claimed to do so. It fits the work of the liberal classics—for which it was designed—much better. 31. See Trudy Govier, Self-Trust, Autonomy, and Self-Esteem, 8 Hypatia 99, 105–6 (Winter 1993). 32. See id. 33. See Maureen A. Mahoney & Barbara Yngvesson, The Construction of Subjectivity and the Paradox of Resistance: Reintegrating Feminist Anthropology and Psychology, 18 Signs 44, 53–55, 60 (1992). 34. See, e.g., Young, supra chapter 2 n.36 at 41; Feinberg, supra chapter 2 n.31 at 33. 35. See Young, supra chapter 2 n.36 at 41; Feinberg, supra chapter 2 n.31 at 33; Mele, supra chapter 2 n.36 at 228–29. 36. See Dworkin, supra chapter 2 n.36 at 162. Dworkin points out this neglect as a failing, demonstrating here, as elsewhere, that his approach may be more accurately characterized as a hybrid rather than as a straightforwardly liberal account. 37. Meir Dan-Cohen, Between Selves and Collectivities: Toward a Jurispru-

Notes to Chapter 3 | 259 dence of Identity, 61 University of Chicago Law Review 1213, 1243 (1994). 38. Nedelsky, supra n.26 at 11. 39. See supra chapter 2 at 42. 40. Feinberg, supra chapter 2 n.31 at 45. 41. Id. at 33–34. 42. Young, supra chapter 2 n.36 at 38. 43. Id. at 40. 44. See id. at 41–42. 45. See id. at 52–55. Young’s stated goal in this book is to defend a conception of autonomy that is distinct from individualism. See id. at 110. He therefore takes a large step away from the traditional liberal model of autonomy outlined in chapter 2. His rejection of individualism is, however, motivated by a political or policy goal: to show that paternalism may sometimes serve autonomy rather than always qualifying as an interference with it. See id. at 65, 74–76. As a result, his move away from individualism operates at this level of policy rather than at the deeper level of reconceptualizing autonomy altogether. In other words, he has moved away from some of the more superficially individualistic aspects of the liberal model—such as the assumption that all uncoerced choice is autonomous—but he has not abandoned the fundamental individualism inherent in the focus on the internal act of choice or will. The argument over the legitimacy of paternalism is one internal to liberal autonomy theory, much as the argument between empiricists and rationalists is internal to Cartesianism. 46. Michael Sandel, Liberalism and the Limits of Justice 181 (Cambridge: Cambridge Univ. Press 1982). 47. Susan Brison, Outliving Oneself: Trauma, Memory, and Personal Identity in Feminists Rethink the Self, supra chapter 2 n.32 at 29. 48. See Margaret Urban Walker, Moral Understandings: A Feminist Study in Ethics 117 (New York: Routledge 1998); Margaret Urban Walker, Picking Up Pieces: Lives, Stories, and Integrity in Feminists Rethink the Self, supra chapter 2 n.32 at 62–64. 49. See generally Carol Gilligan, In a Different Voice: Psychological Theory and Women’s Development (Cambridge: Harvard Univ. Press 1982). For a sample of the voluminous criticism and discussion of Gilligan’s research, see Linda Kerber, et al. On In a Different Voice: An Interdisciplinary Forum, 11 Signs 304 (1986). 50. See Carol Gilligan, Remapping the Moral Domain: New Images of Self in Relationship in Mapping the Moral Domain 8 (Carol Gilligan, et al., eds.) (Cambridge: Harvard Univ. Press 1988). 51. See Carol Gilligan, Moral Orientation and Moral Development in Women and Moral Theory 19, 23 (Eva Feder Kittay & Diana T. Meyers,

260 | Notes to Chapter 3 eds.) (Lanham, MD: Rowman & Littlefield 1987); Mary Field Belenky, et al., Women’s Ways of Knowing (New York: Basic Books 1986). 52. See Diana T. Meyers, The Socialized Individual and Individual Autonomy in Women and Moral Theory, supra n.51 at 143–45; Carrie Menkel Meadow, Portia Redux: Another Look at Gender, Feminism, and Legal Ethics, 2 Virginia Journal of Social Policy & the Law 75, 83 (1994). 53. See generally Nancy Chodorow, The Reproduction of Mothering (Berkeley: Univ. of California Press 1978). 54. See, e.g., Robin West, Jurisprudence and Gender, 55 University of Chicago Law Review 1, 20–27 (1988). 55. See Meyers, supra n.52 at 145. 56. See Gilligan, supra n.51 at 25–26. 57. See generally Susan Moller Okin, Women in Western Political Thought (Princeton: Princeton Univ. Press 1979). 58. See, e.g., Jagger, supra n.13 at 164. 59. See, e.g., Nell Noddings, Caring: A Feminine Approach to Ethics (Berkeley: Univ. of California Press 2003, 2d ed.); Sara Ruddick, Maternal Thinking: Toward a Politics of Peace (Boston: Beacon Press 1989); Nancy J. Hirschmann, Rethinking Obligation: A Feminist Method for Political Theory (Ithaca, NY: Cornell Univ. Press 1992). notes to chapter 4 1. Admittedly, the first situation is unlikely: in even the most repressive totalitarian society there are some contradictory cultural forces, if only among a minority. See Elizabeth Perry Hodges, Writing in a Different Voice, 66 Texas Law Review 629, 632 (1988). It would probably be more accurate to conceive of this as a continuum, running from the imaginary monolith at one end to the equally imaginary total chaos at the other end, with all actual societies falling somewhere in the middle. The argument in text simplifies this situation by treating it as two separate categories rather than a continuum, but it is valid nonetheless: at any point on the continuum (other than the end points) a society will suffer from both of the difficulties described in the text, but to differing degrees. 2. See Margaret J. Radin, The Pragmatist and the Feminist, 63 Southern California Law Review 1699, 1710, 1720 (1990). 3. See John Shotter & Josephine Logan, The Pervasiveness of Patriarchy: On Finding a Different Voice in Feminist Thought and the Structure of Knowledge, supra chapter 3 n.6 at 69–70. 4. See Audre Lorde, Sister Outsider 110 (Trumansburg, NY: Crossing Press 1984). 5. Catharine A. MacKinnon, Feminism, Marxism, Method, and the State: Toward a Feminist Jurisprudence, 8 Signs 635, 638 (1983).

Notes to Chapter 4 | 261 6. See Patterson, supra chapter 2 n.5 at 279. 7. MacKinnon, supra n.5 at 635. 8. See Jagger, supra chapter 3 n.13 at 160. 9. See id. See also Elizabeth Spelman, Anger and Insubordination in Women, Knowledge, and Reality: Explorations in Feminist Philosophy 270 (Ann Garry & Mary Pearsall, eds.) (Boston: Unwin Hyman 1989). 10. See Drucilla Cornell, Beyond Accommodation: Ethical Feminism, Deconstruction, and the Law 165–96 (New York: Routledge 1991); Drucilla Cornell, The Doubly Prized World: Myth, Allegory, and the Feminine, 75 Cornell Law Review 644, 696–98 (1990); J. Morawski, Impasse in Feminist Thought? in Feminist Thought and the Structure of Knowledge, supra chapter 3 n.6 at 191. 11. See Hilary Rose, Beyond Masculinist Realities: A Feminist Epistemology for the Sciences in Feminist Approaches to Science, supra chapter 2 n.4 at 58. 12. See Robin West, The Difference in Women’s Hedonic Lives: A Phenomenological Critique of Feminist Legal Theory, 3 Wisconsin Women’s Law Journal 81, 85 (1987) (when you have no social recognition for your pain, you may yourself experience it differently). 13. See Jagger, supra chapter 3 n.13 at 160. 14. See, e.g., MacKinnon, supra n.5 at 638. 15. See Smith, supra chapter 3 n.14 at 93; Rhoda K. Unger, Psychological, Feminist, and Personal Epistemologies: Transcending Contradiction in Feminist Thought and the Structure of Knowledge, supra chapter 3 n.6 at 136. 16. See Jean Grimshaw, Philosophy and Feminist Thinking 99–100 (Minneapolis: Univ. of Minnesota Press 1986). This is not to say that all value conflicts must be resolvable; some irreducible conflicts may remain in any moral system. But in order for feminism to claim any transformative potential, it must offer a way to assert that the point of view it provides is superior to (not just equally as valid as) the competing sexist viewpoint that is also enshrined in our cultural values. This is one value conflict on which feminism may not remain neutral. See Katharine Bartlett, Feminist Legal Methods, 103 Harvard Law Review 829, 879 (1990). 17. Seyla Benhabib, Epistemologies of Postmodernism: A Rejoinder to JeanFrancois Lyotard in Feminism/Postmodernism, supra chapter 2 n.2 at 129 n.35. 18. See Eloise A. Buker, Rhetoric in Postmodern Feminism: Put-Offs, PutOns, and Political Plays in The Interpretive Turn: Philosophy, Science, Culture 218, 224–25 (D. Hiley, J. Bohman, & R. Shusterman, eds.) (Ithaca, NY: Cornell Univ. Press 1991).

262 | Notes to Chapter 4 19. See Benhabib, supra n.17 at 122. We also need some standards for crosscultural judgment and criticism. Such judgments are epistemologically parallel to ones within a culture that is itself contradictory or composed of subcultures, but they may raise additional moral issues. 20. Donna Haraway, Situated Knowledges: The Science Question in Feminism and the Privilege of Partial Perspective, 14 Feminist Studies 1, 3 (1988). 21. Susan Bordo, Feminism, Postmodernism, and Gender-Skepticism in Feminism/Postmodernism, supra chapter 2 n.2 at 142. 22. Id. at 143. 23. See Harding, supra chapter 2 n.29 at 187. 24. See Patterson, supra chapter 2 n.5 at 276–77. 25. See James Boyle, Is Subjectivity Possible? The Postmodern Subject in Legal Theory, 62 University of Colorado Law Review 489, 497 (1991); see generally Pierre Schlag, The Problem of the Subject, 69 Texas Law Review 1627 (1991). But cf. Jane Flax, Thinking Fragments 216–17 (Berkeley: Univ. of California Press 1990) (arguing that the writing of some postmodernists reinstates a transcendent subject by incorporating a romantic/aesthetic form of heroic creativity). 26. Buker, supra n.18 at 227. 27. See Marilyn Friedman, Feminism in Ethics: Conceptions of Autonomy in The Cambridge Companion to Feminism in Philosophy 205, 220 (Miranda Fricker & Jennifer Hornsby, eds.) (Cambridge: Cambridge Univ. Press 2000). 28. Judith Butler, Gender Trouble, Feminist Theory, and Psychoanalytic Discourse in Feminism/Postmodernism, supra chapter 2 n.2 at 327. 29. See Patterson, supra chapter 2 n.5 at 260 n.25. 30. See Bartlett, supra n.16 at 879–80 n.220. 31. See Flax, supra n.25 at 207 (describing Foucault’s theory of power). 32. See Robin West, Feminism, Critical Social Theory, and Law, 1989 University of Chicago Legal Forum 59, 59–64. 33. See 1 William Blackstone, Commentaries ch. XV, sec. III (London: Cavendish 2001) (“[T]he husband and wife are one person in law: that is, the very being or legal existence of the woman is suspended during the marriage. . . .”). This denial of personhood is, of course, also one of the hallmarks of the oppression suffered by African American women and men. See Flax, supra n.25 at 219; see also Patricia Williams, The Alchemy of Race and Rights 153 (Cambridge: Harvard Univ. Press 1991). 34. West, supra n.32 at 89. 35. See Flax, supra n.25 at 220. 36. Christine DiStefano, Dilemmas of Difference: Feminism, Modernity, and Postmodernity in Feminism/Postmodernism, supra chapter 2 n.2 at 76.

Notes to Chapter 4 | 263 37. bell hooks, Sisterhood: Political Solidarity between Women in Feminism and Community 293, 295 (Penny A. Weiss & Marilyn Friedman, eds.) (Philadelphia: Temple Univ. Press 1995). 38. Patricia Hill Collins, Black Feminist Thought: Knowledge, Consciousness, and the Politics of Empowerment 95 (New York: Routledge 1991). 39. hooks, supra n.37 at 295. 40. Martha Minow, Surviving Victim Talk, 40 University of California at Los Angeles Law Review 1411, 1433 (1993). 41. See Elizabeth Schneider, Feminism and the False Dichotomy of Victimization and Agency, 38 New York Law School Law Review 387, 395 (1993). 42. See Kathryn Abrams, Sex Wars Redux: Agency and Coercion in Feminist Legal Theory, 95 Columbia Law Review 304, 348 (1995). 43. Adrienne Rich, If Not with Others, How? in Feminism and Community, supra n.37 at 399, 400. 44. Susan Wendell, Oppression and Victimization: Choice and Responsibility in “Nagging” Questions: Feminist Ethics in Everyday Life 41, 52–53 (Dana E. Bushnell, ed.) (Lanham, MD: Rowman & Littlefield 1995). 45. Id. at 54. 46. See id. at 57–60. Wendell describes yet a further perspective, the observer/philosopher perspective, from which people seek understanding on a large scale, including understanding of causes. What appears to be a choice from the perspective of the responsible agent may often appear to be caused by large-scale social forces from the perspective of the observer/philosopher. Wendell argues that neither of these perspectives is necessarily superior—both are necessary— but that we generally find ourselves alternating between them because they are difficult to sustain simultaneously for both intellectual and emotional reasons. See id. at 66–68. I focus on the responsible actor perspective because it is from that perspective that forward-looking choices are made and personal responsibility assumed and it is that perspective that is threatened by the total abandonment of autonomy. 47. Bordo, supra n.21 at 144. 48. Id. at 145. 49. See Iris Marion Young, The Ideal of Community and the Politics of Difference, in Feminism/Postmodernism, supra chapter 2 n.2 at 309–10. 50. See Mari Matsuda, Pragmatism Modified and the False Consciousness Problem, 63 Southern California Law Review 1763, 1768–71 (1990). 51. See Harding, supra chapter 2 n.29 at 304. But see Mary Jo Frug, A Postmodern Feminist Legal Manifesto (An Unfinished Draft), 105 Harvard Law Review 1045, 1047–48 (1992). 52. SeeNancy Hartsock, Money, Sex, and Power: Toward a Feminist

264 | Notes to Chapter 4 Historical Materialism 231–51 (Boston: Northeastern Univ. Press 1983); Jagger, supra chapter 3 n.13 at 162. 53. See Harding, supra chapter 2 n.29 at 124, 131–32; Nancy Hartsock, The Feminist Standpoint: Developing the Ground for a Specifically Feminist Historical Materialism in Feminism and Methodology, supra chapter 3 n.11 at 159–60; Uma Narayan, The Project of Feminist Epistemology: Perspectives from a Nonwestern Feminist in Gender/Body/Knowledge, supra chapter 2 n.2 at 265–66. 54. See Harding, supra chapter 2 n.29 at 126; Mari Matsuda, When the First Quail Calls: Multiple Consciousness as Jurisprudential Method, 11 Women’s Rights Law Reporter 7, 9 (1989). 55. See Narayan, supra n.53 at 266. 56. Jane Flax, Postmodernism and Gender Relations in Feminist Theory in Feminism/Postmodernism, supra chapter 2 n.2 at 56. 57. See DiStefano, supra n.36 at 72; Angela P. Harris, Race and Essentialism in Feminist Legal Theory, 42 Stanford Law Review 581, 608–9 (1990); Cherrie Moraga, La Guerra in This Bridge Called My Back: Writings by Radical Women of Color 27, 27–34 (Cherrie Moraga & Gloria Anzaldua, eds.) (New York: Women of Color Press 1983). 58. Narayan, supra n.53 at 266. 59. Harding, supra chapter 2 n.29 at 142. 60. Id. at 127. 61. Id. at 167. 62. Id. at 149 & n.17. 63. See id. at 151. 64. Id. at 169. 65. Id. 66. See id. at 152. 67. See Thomas C. Grey, Holmes and Legal Pragmatism, 41 Stanford Law Review 787, 798 (1989); Margaret J. Radin, The Pragmatist and the Feminist, 63 Southern California Law Review 1699, 1707–9 (1990). See generally Phylis Rooney, Feminist-Pragmatist Revisionings of Reason, Knowledge, and Philosophy, 8 Hypatia 15 (No. 2, Spring 1993). 68. See Margaret J. Radin & Frank Michelman, Pragmatist and Poststructuralist Critical Legal Practice, 139 University of Pennsylvania Law Review 1019, 1037 n.76 (1991). 69. Joan Williams, Rorty, Radicalism, Romanticism: The Politics of the Gaze, 1992 Wisconsin Law Review 131, 134. 70. See Radin, supra n.67 at 1726. 71. Williams, supra n.69 at 138. 72. See id. at 132. 73. Williams does not intend to address this issue: she is satisfied that cul-

Notes to Chapter 4 | 265 tural standards—the only ones available—are sufficient to make the arguments that concern her. See id. (“Words were tools even when we thought they were mirrors. The mere admission that they are no more than tools will not cause them suddenly to break.”) 74. See, e.g., Martha Minow & Elizabeth Spelman, In Context, 63 Southern California Law Review 1597, 1632–33, 1650 (1990); Radin, supra n.67 at 1711; Matsuda, supra n.50 at 1771–72. This principle is not, however, the only one that has been suggested. See Richard Rorty, Contingency, Irony, and Solidarity 53 (Cambridge: Cambridge Univ. Press 1989) (three aspirations are to reduce suffering and cruelty, maximize freedom, and create equal chances for all to pursue idiosyncratic fulfillment). 75. See generally Bartlett, supra n.16. 76. See id. at 880–81. 77. See id. at 881–82. 78. See id. at 881. 79. Id. at 884. Reality here serves the purpose identified by Sandra Harding of distinguishing what things are from what we wish things were, see Harding, supra chapter 2 n.29 at 160, rather than distinguishing something external to the person from something internal. 80. Bartlett, supra n.16 at 885. 81. See id. at 884. 82. Id. at 884. 83. See Katherine T. Bartlett, Tradition, Change, and the Idea of Progress in Feminist Legal Thought, 1995 Wisconsin Law Review 303. 84. See id. at 336. 85. See Bartlett, supra n.16 at 883. 86. Id. at 883. 87. See id. at 883–84 & n.235. 88. See id. at 884–85. 89. For a very interesting attempt to define such a standard from within an African American critical tradition, see generally Anthony Cook, Reflections on Postmodernism, 26 New England Law Review 751 (1992). 90. See, e.g., Seyla Benhabib, Critique, Norm, and Utopia 253–79 (New York: Columbia Univ. Press 1986). 91. See Jagger, supra chapter 3 n.13 at 161; see also Grimshaw, supra n.16 at 101 (suggesting a moral theory based on “interests” to inform epistemology). 92. See generally Joan Tronto, Women and Caring: What Can Feminists Learn about Morality from Caring? in Gender\Body\Knowledge, supra chapter 2 n.2 at 172. 93. See generally Drucilla Cornell, Toward a Modern/Postmodern Reconstruction of Ethics, 133 University of Pennsylvania Law Review 29 (1985).

266 | Notes to Chapter 4 94. See Martha C. Nussbaum, Non-Relative Virtues: An Aristotelian Approach, 13 Midwest Studies in Philosophy 32, 45 (1983). 95. See id. at 34–36. 96. Id. at 45. In addition, Nussbaum argues that certain ways of conceptualizing areas of life represent features of our common humanity, which cross cultural boundaries and define what it means to be human in so fundamental a way that to change them would be to change the kind of creatures we understand ourselves to be. As a result, these features may provide a foundation for standards of virtue that are valid across cultural lines. Such features might include our mortality, our dependence on the world outside our bodies for sustenance (physical and emotional), our cognitive and practical reason functions, and some notion of sociability. See id. at 48–49, 50. 97. See Martha C. Nussbaum, The Therapy of Desire: Theory and Practice in Hellenistic Ethics 23–24 (Princeton, NJ: Princeton Univ. Press 1994). 98. See, e.g., Martha C. Nussbaum, The Professor of Parody in The New Republic 37 (February 22, 1999) (sharply criticizing the work of Judith Butler). 99. See id. at 42; Martha C. Nussbaum, Sex and Social Justice 6–13 (Oxford: Oxford Univ. Press 1999). 100. Indeed, one principle or method, one truth, may always involve domination of some people by others. See Flax, supra n.56 at 48–49. 101. See generally Nussbaum, supra n.99; Nussbaum, supra n.97; Martha C. Nussbaum, Love’s Knowledge (New York: Oxford Univ. Press 1990); Martha C. Nussbaum, The Fragility of Goodness (New York: Cambridge Univ. Press 1986). notes to chapter 5 1. This approach is, perhaps, usefully analogized to the method of reflective equilibrium: beginning from our considered intuitions, subjecting them to (moral) analysis, and being willing to modify each in light of the other. See John Rawls, A Theory of Justice 20 (Cambridge: Harvard Univ. Press 1971). Such a method works best with less controversial issues and, although I offer some arguments for why the functions I identify are valuable, I expect that the functions of truth identified here will be relatively uncontroversial, both as a description of current practice and as normatively desirable. 2. This argument does not require a Cartesian idea of knowledge. A Wittgensteinian view of language as a practice still requires a form of intersubjective knowledge: we must all understand what to do sufficiently similarly or it is not a practice at all.

Notes to Chapter 5 | 267 3. Cf. Naomi Scheman, Who Wants to Know? The Epistemological Value of Values in (En)Gendering Knowledge 179, 185 (Joan Hartman & Ellen Messer-Davidow, eds.) (Knoxville, TN: Univ. of Tennessee Press 1991). 4. I am a little uncomfortable making this claim about how truth making works when I have just refused to actually define the boundaries of such truth making. I would like to suggest, however, that wherever that boundary may lie, it certainly includes the paradigm case of truth-making activities that help construct a shared reality: a discussion in which people try to reach agreement about what they believe and how they will proceed. The argument in text about the nature of the relationships engendered by truth making concerns this paradigm case, and this is, of course, the case that is most significant for the ultimate goal of this book: connecting the truth theory back to speech. If truth is limited to the making of shared reality through conceptual means, then perhaps all instances of truth processes fit this paradigm. If truth is not so limited, then they may not. I do not mean to answer this question. Whether or not every example of truth making fits this paradigm, truth making is valuable because much of it does. 5. See infra at 177–79 (arguing that speech leads us to this relationship and that it is speech that is the way we create a shared reality through truth claims). 6. See Claudia Card, Women, Evil, and Gray Zones, 31 Metaphilosophy 509 (No. 5, 2000) (examining the implications of the “Stockholm syndrome” phenomenon for feminism); Barbara A. Huddleston-Mattai & Rudy P. Mattai, The Sambo Mentality and the Stockholm Syndrome Revisited, 23 Journal of Black Studies 344 (No. 3, 1993) (applying this analysis to race issues). 7. The epistemological view—that shared reality is part of what truth is for— rests on a moral commitment: that we must regard each other in a particular way. This may sound like simply substituting a moral foundation for the metaphysical one I rejected in Cartesianism. There are several responses to this charge. First, the moral commitment here is not independent of the epistemology, but closely tied to it. It is because we come to recognize our epistemic interdependence (we make reality together) that the moral commitment becomes both salient and compelling. In other words, neither morality nor epistemology is “foundational”; they are instead mutually defining. Second, the sort of foundation provided by this moral/epistemological hybrid is not of the same nature as the foundationalism of Cartesian epistemologies. This is not an immovable rock upon which all other knowledge claims can rest. Instead, I offer this nonCartesian view of shared reality in the pragmatic spirit: I suggest that if we look at the world this way instead, we will find that it works better for us. I am not trying to justify the underlying moral claims in some objective way, any more than I am seeking to justify the epistemological claims objectively. Indeed, I am rejecting the idea that such a justification is either possible or meaningful. Instead, I offer a pragmatic justification: try it; you’ll like it.

268 | Notes to Chapter 5 8. As with extreme forms of skepticism generally, there is a serious question whether this position is either coherent or livable. See Phylis Rooney, FeministPragmatist Revisionings of Reason, Knowledge, and Philosophy, 8 Hypatia 15, 22 (No. 2, Spring 1993). If intersubjective knowledge is made completely impossible, then it is difficult to see, first, what could possibly be the epistemological status of that claim itself or, second, how one could live with other human beings at all. For an interesting argument on the livability of certain forms of skepticism, see generally Martha C. Nussbaum, Skepticism about Practical Reason in Literature and Law, 107 Harvard Law Review 714 (1994). 9. See Susan H. Williams, Feminist Legal Epistemology, 8 Berkeley Women’s Law Journal 63, 83–92 (1993). 10. See Martha C. Nussbaum, Valuing Values: A Case for Reasoned Commitment, 6 Yale Journal of Law & Humanities 197, 214 (1994). 11. See Richard Rorty, 1 Objectivity, Relativism, and Truth 29 (Cambridge: Cambridge Univ. Press 1991). 12. See, e.g., Joan Williams, Rorty, Radicalism, Romanticism: The Politics of the Gaze, 1992 Wisconsin Law Review 131, 133–34. 13. Perhaps it does not lose all of its value, however: we may still value the process because of the social bonds it creates or strengthens, because of the respect for participants that it expresses and instills, or because of the qualities of mind or character that it encourages. Cf. Vincent Blasi, The First Amendment and the Ideal of Civic Courage: The Brandeis Opinion in Whitney v. California, 29 William & Mary Law Review 653, 674–76 (1988) (arguing that freedom of speech—like democracy—can be justified on the basis of character creation). 14. See Robert Justin Lipkin, Indeterminacy, Justification, and Truth in Constitutional Theory, 60 Fordham Law Review 595, 627–28 (1992). 15. William James, Pragmatism 104 (Cambridge: Harvard Univ. Press 1975). 16. See Rorty, supra n.11 at 22. 17. See Radin, supra chapter 4 n.2 at 1710–11. 18. See Rorty, supra n.11 at 24. 19. Such attention to the concrete experiences of others can be made both less likely and more difficult—although perhaps also even more important—by the existence of social hierarchy. See Robin West, The Difference in Women’s Hedonic Lives: A Phenomenological Critique of Feminist Legal Theory, 3 Wisconsin Women’s Law Journal 81, 85 (1987). 20. Rorty, supra n.11 at 8. 21. Richard M. Thomas, Milton and Mass Culture: Toward a Postmodernist Theory of Tolerance, 62 University of Colorado Law Review 525, 575 (1991). 22. While the free exercise clause of the First Amendment also offers some

Notes to Chapter 5 | 269 protection to these beliefs, there are several reasons why we should not relegate this function of truth to the religion clauses alone. First, it is very difficult to define the boundary of this class of beliefs. In the religion clauses, such boundary disputes are crucial—because a belief or practice must qualify as religion to be entitled to the protection—whereas in the speech clause the definitional problem disappears. Second, the free exercise clause has been so weakened by recent Supreme Court cases that one should hesitate to abandon the expression of any belief to that minimal level of protection. See Employment Division v. Smith, 494 U.S. 872 (1990). Third, because we are concerned with the expression of belief, there are interests of listeners at stake as well: the listeners’ search for truth may be aided by hearing others’ opinions. These listeners’ interests are recognized by free speech doctrine but are alien to the highly individualistic focus of free exercise doctrine. 23. See Martha C. Nussbaum, “Secret Sewers of Vice”: Disgust, Bodies, and the Law in The Passions of Law 19, 21 (Susan A. Bandes, ed.) (New York: New York Univ. Press 1999). 24. See Chevigny, supra chapter 2 n.12 at 66. 25. I think the common complaints against social constructionists for internal inconsistency—that they assume that they know the truth while denying the possibility of truth—are based on this misunderstanding. At least some types of social constructionism, like the one I am proposing, do not require one to stop talking about truth and reality. They simply require that one understand the status of such claims differently: as claims about a complex construct that is always permeated by conceptual frameworks, including moral and political ones. We can still call it real, because that distinguishes it from the other categories of our experience that are not real: dreams, illusions, imaginings. What such critics seem to forget is that, for a social constructionist, being real or true does not mean being unconstructed. 26. See Deborah Rhode, Speaking of Sex: The Denial of Sexual Inequality 22–42 (Cambridge: Harvard Univ. Press 1997). 27. The distinction between physical and social constraints is particularly difficult to maintain in light of the impact of social institutions and practices on the physical environment we inhabit. The meaning of physical distance, for example, is very different if one lives in a technologically sophisticated society with trains and airplanes than if one does not. The nature of the constraint imposed by the physical reality is, thus, mediated by the social institutions. Conversely, as this example suggests, human institutions often have a physical aspect as well and generally have results in the physical world. What is real about them, however, is not just their physical aspect—the trains and tracks—but also their social aspect: the practice of carrying people from one place to another under certain specific circumstances (e.g., they have paid their money and it is the time when the train is scheduled to leave).

270 | Notes to Chapter 5 28. See supra at chapter 3, section C. 29. Thomas, supra n.21 at 550. See also Singer, supra chapter 2 n.28 at 66. 30. See Scheman, supra n.3 at 185. 31. See Robert Justin Lipkin, Kibbitzers, Fuzzies, and Apes without Tails: Pragmatism and the Art of Conversation in Legal Theory, 66 Tulane Law Review 69, 84–87 (1991). 32. See Bartlett, supra chapter 4 n.16 at 880–81. 33. See Lipkin, supra n.14 at 630 n.141. See also Rorty, supra n.11 at 13. 34. See Barbara Herrnstein Smith, The Unquiet Judge: Activism without Objectivism in Law and Politics in Rethinking Objectivity, supra chapter 2 n.1 at 302. 35. See Donna Haraway, Situated Knowledges: The Science Question in Feminism and the Privilege of Partial Perspectives, 14 Feminist Studies 575, 583 (1988). 36. See Sandra Harding, Who Knows? Identities and Feminist Epistemologies in (En)Gendering Knowledge, supra n.3 at 110. 37. See id. 38. See Abrams, supra chapter 4 n.42 at 326–29 (describing the work of Catharine MacKinnon). 39. See Drucilla Cornell, The Doubly-Prized World: Myth, Allegory, and the Feminine, 75 Cornell Law Review 644, 659 (1990) (recognizing this risk in deconstructive readings of gender). 40. See Bartlett, supra chapter 4 n.83 at 333. 41. See id. at 339. 42. See Kathryn Abrams, Hearing the Call of Stories, 79 California Law Review 971, 1020–24 (1991). 43. See, e.g., Steven Winter, Indeterminacy and Incommensurability in Constitutional Law, 78 California Law Review 1441, 1447–48 (1990) (using the work of French philosopher Maurice Merleau-Ponty to shed light on issues of constitutional interpretation). 44. See Martha C. Nussbaum, The Fragility of Goodness 1–8 (Cambridge: Cambridge Univ. Press 1986). 45. Is it possible to create a context so large that it includes all of humanity? For example, would it be possible to generate moral values to guide behavior that are appropriately applied to all human beings? See, e.g., Nussbaum, supra n.10 at 201 (arguing that truths that are “necessary” in the sense that they are “necessary for human thought and life” rather than immutable apart from history might still be possible). I do not think there is an a priori answer to this question. The extent to which a context can be expanded, while retaining its ability to provide a foundation for a shared reality, is a question that can only be answered in a pragmatic way. It will depend upon the issue that one is concerned about, the kind of shared reality that one is seeking to ground, and the concrete

Notes to Chapter 5 | 271 conditions of the people that one is attempting to include. Moreover, even these pragmatic questions may not be answerable in advance. In other words, the answer to the question, “Is a universal context possible?” will often be “Maybe; we won’t know until we try.” 46. This concern for the pain of others is not all that is involved in respect for personhood; indeed, it is not even the most important aspect of respect for personhood in many ethical theories. Deontological liberalism, for example, might find that recognition of one’s capacity for moral choice is far more central and that animals, for example, do not count as subjects, regardless of their abilities to feel pleasure and pain, because they lack this capacity for choice. I would suggest, however, that the capacity for happiness (which should, of course, be defined in ways that require rather more than merely pleasure and pain) is usually a necessary condition for personhood even in those theories for which it is not a sufficient one. See, e.g., John Rawls, A Theory of Justice 126–30 (Cambridge: Harvard Univ. Press 1971) (describing the circumstances of justice explicitly in terms of the conditions of human life). But see Immanuel Kant, Foundations of the Metaphysics of Morals 24 (Lewis White Beck, trans.) (Indianapolis, IN: Bobbs-Merrill 1959). 47. See Daniel A. Farber & Suzanna Sherry, Beyond All Reason: The Radical Assault on Truth in American Law (New York: Oxford Univ. Press 1997). 48. See Schenk v. U.S., 249 U.S. 47 (1919); Abrams v. U.S., 250 U.S. 616 (1919). Of course, the reason given for the repression in such cases rarely directly acknowledges that the speech is damaging or contrary to the political commitments of the society. The usual justification is that the speech poses some practical danger, like insurrection or revolution. But, as the social constructionist critique demonstrates, one’s estimate of the practical danger is deeply shaped by the political and moral commitments one brings to the exercise. This dynamic is amply demonstrated by many of the early First Amendment cases on politically subversive speech, in which the Court consistently overestimates the danger posed by “puny anonymities” and punishes them for ineffective speech. See id. at 629 (Holmes, dissenting). 49. See Ruth Bleier, Science and Gender (New York: Pergamon 1984); Anne Fausto-Sterling, Myths of Gender: Biological Theories about Women and Men (New York: Basic Books 1992); Ruth Bleier, Sex Differences Research: Science or Belief? in Feminist Approaches to Science 147 (Ruth Bleier, ed.) (New York: Pergamon 1986); Alison Wylie, Kathleen Okfuhlik, Leslie Thielen-Wilson, & Sandra Morton, Philosophical Feminism: A Bibliographic Guide to Critiques of Science, 19 Resources for Feminist Research 2–36 (1990). 50. Cf. Singer, supra chapter 2 n.28 at 55. 51. See, e.g., Mircea Eliade, The Sacred and the Profane: The

272 | Notes to Chapter 5 Nature of Religion 10 (New York: Harcourt Brace 1959) (“all that goes beyond man’s natural experience, language is reduced to suggesting by terms taken from that experience”). 52. It would be more accurate to describe the category of religious claims as an intersecting set rather than a subset. Some claims that are conventionally religious would fall into this category of describing the human role in or relationship to a nonhuman reality. Other religious claims might, however, fall into one of the other categories mentioned earlier—e.g., claims about a shared reality on which we should base joint decisions (“this is a Christian nation”) or claims about the effects of choices that we should care about (“secularism in the public schools destroys the faith of children”). And some religious claims may fall into the category of the forbidden Cartesian claims (“God does not hear the prayers of a Jew”). Because I have dealt with these other sorts of claims elsewhere in the text, I am here considering only those religious claims that fall within this last category of truth claims. 53. See Eliade, supra n.51 at 14–17. 54. See Note, Reinterpreting the Religion Clauses: Constitutional Construction and Conceptions of the Self, 97 Harvard Law Review 1468, 1475–86 (1984); Susan H. Williams & David C. Williams, Volitionalism and Religious Liberty, 76 Cornell Law Review 769, 850–906 (1991). 55. Compare, e.g., Kent Greenawalt, Religious Convictions and Political Choice (New York: Oxford Univ. Press 1988) with Mark Tushnet, Religion in Politics: Religious Convictions and Political Choice by Kent Greenawalt, 89 Columbia Law Review 1131 (1989). 56. See Susan H. Williams, Religion, Politics, and Feminist Epistemology: A Comment on the Uses and Abuses of Morality in Public Discourse, 77 Indiana Law Journal 267, 273 (2002). notes to chapter 6 1. See Diana T. Meyers, Self, Society, and Personal Choice 210–13 (New York: Columbia Univ. Press 1989). 2. See Robin S. Dillon, How to Lose Your Self-Respect, 29 American Philosophical Quarterly 125 (No. 2, April 1992). 3. See Trudy Govier, Self-Trust, Autonomy, and Self-Esteem, 8 Hypatia 99, 105–6 (Winter 1993). 4. See Dillon, supra n.2 at 138 n.24, 139 n.27. 5. See id. at 125. 6. See Julie Nelson-Kwa & Stephanie Riger, Women’s Agency in Psychological Contexts in Provoking Agents: Gender and Agency in Theory and Practice 169, 175 (Judith Kegan Gardiner, ed.) (Urbana: Univ. of Illinois Press 1995) (describing the “agentic” beliefs about oneself necessary to be au-

Notes to Chapter 6 | 273 tonomous); Keith Lehrer, Self-Trust 96 (Oxford: Oxford Univ. Press 1997). 7. See Thomas Hill, Autonomy and Self-Respect 4–18 (Cambridge: Cambridge Univ. Press 1991). 8. See Diana T. Meyers, The Politics of Self-Respect: A Feminist Perspective, 1 Hypatia 83, 84 (No. 1, Spring 1986). 9. See Dillon, supra n.2 at 133. 10. Id. 11. Id. at 134. 12. Id. 13. See id. at 135. 14. See Wendell, supra chapter 4 n.44 at 57–66. 15. See Richard Fallon, Two Senses of Autonomy, 46 Stanford Law Review 875, 893 (1994). 16. Owen Flanagan, Identity and Strong and Weak Evaluation in Identity, Character, and Morality: Essays in Moral Psychology 37, 48 (Owen Flanagan & Amelie Oksenberg Rorty, eds.) (Cambridge: MIT Press 1990). 17. See Dworkin, supra chapter 2 n.36 at 32. 18. Michele Moody-Adams, On the Old Saw That Character Is Destiny in Identity, Character, and Morality, supra n.16 at 111, 129. 19. See Young, supra chapter 2 n.36 at 16–17. 20. Compare, e.g., Dworkin, supra chapter 2 n.36 at 41 with Feinberg, supra chapter 2 n.31 at 39. 21. Diana Tietjens Meyers, Intersectional Identity and the Authentic Self? Opposites Attract in Relational Autonomy, supra chapter 2 n.37 at 153. 22. See, e.g., Morwenna Griffiths, Feminisms and the Self: The Web of Identity 183 (London: Routledge 1995) (suggesting we think in terms of authenticity rather than integrity); Natalie Stoljar, Autonomy and Feminist Intuition in Relational Autonomy, supra chapter 2 n.37 at 105; Meyers, supra n.21 at 170. 23. Moody-Adams, supra n.18 at 120. This connection may be a bit more complicated than the quotation in text suggests. First, there are situations in which the very fact that one is not responsible is a comment on one’s character (e.g., where a person is so weak or broken that we are inclined not to hold him responsible for his act, but we regard that weakness as itself a character flaw). Second, there are issues of collective guilt in which both responsibility and character, and their connection, are muddier. Nonetheless, in the majority of situations, the claim in text holds true. 24. See Marion Smiley, Moral Responsibility and the Boundaries of Community: Power and Accountability from a Pragmatic Point of View 246–47 (Chicago: Univ. of Chicago Press 1992); Genevieve

274 | Notes to Chapter 6 Lloyd, Individuals, Responsibility, and the Philosophical Imagination in Relational Autonomy, supra chapter 2 n.37 at 117; see also Sandra Lee Bartky, In Defense of Guilt in On Feminist Ethics and Politics 29 (Claudia Card, ed.) (Lawrence, KS: Univ. of Kansas Press 1999); George P. Fletcher, Liberals and Romantics at War: The Problem of Collective Guilt, 111 Yale Law Journal 1499 (2002). 25. See Smiley, supra n.24 at 49; Marcia Homiak, On the Malleability of Character in On Feminist Ethics and Politics, supra n.24 at 58; Bernard Williams, Making Sense of Humanity 14 (Cambridge: Cambridge Univ. Press 1995). But see Moody-Adams, supra n.18 at 120. 26. See Smiley, supra n.24 at 7, 14. 27. For a summary of this debate, see Smiley, supra n.24 at 5–7, 72–101. For some specific examples of this debate, see Dworkin, supra chapter 2 n.36; Frankfurt, supra chapter 2 n.36. 28. See Thomas E. Hill, Jr., The Kantian Conception of Autonomy in The Inner Citadel, supra chapter 2 n.31 at 93, 97. 29. See Frankfurt, supra chapter 2 n.36 at 69; Dworkin, supra chapter 2 n.36 at 16–20. 30. There is another debate over whether autonomy should be assessed in terms of each choice or act, or for an area of a life, or for a life as a whole. Compare, e.g., Dworkin, supra chapter 2 n.36 at 16 (autonomy “can only be assessed over extended portions of a person’s life. . . .”) with Diana T. Meyers, Personal Autonomy and the Paradox of Feminine Socialization, 84 Journal of Philosophy 619, 624–26 (Nov. 1987) (arguing for a concept of partial autonomy). Again, I am attempting to avoid this controversy until I have outlined an alternative conception of autonomy. 31. See Frankfurt, supra chapter 2 n.36 at 19–20; Young, supra chapter 2 n.36 at 13; Feinberg, supra chapter 2 n.31 at 42. 32. Marcia Homiak, On the Malleability of Character in On Feminist Ethics and Politics, supra n.24 at 57. 33. This view of responsibility as a collection of social practices is based on the approach taken by both Marion Smiley and Margaret Urban Walker. See Smiley, supra n.24; Margaret Urban Walker, Moral Understandings: A Feminist Study in Ethics (New York: Routledge 1998). My account borrows heavily from theirs. 34. Paul Benson, Feeling Crazy: Self-Worth and the Social Character of Responsibility in Relational Autonomy, supra chapter 2 n.37 at 79. 35. See id. at 81–82. 36. See Smiley, supra n.24 at 182; Walker, supra n.33 at 93. 37. Williams, supra n.25 at 16. 38. See Smiley, supra n.24 at 235–37.

Notes to Chapter 6 | 275 39. Walker, supra n.33 at 95. 40. Moody-Adams, supra n.18 at 125. 41. See Smiley, supra n.24 at 239. 42. Stephen L. White, Rationality, Responsibility, and Pathological Indifference in Identity, Character, and Morality, supra n.16 at 401, 420–21. 43. See Ronald Dworkin, Taking Rights Seriously 169 (Cambridge: Harvard Univ. Press 1977). 44. See generally David A. J. Richards, Toleration and the Constitution (New York: Oxford Univ. Press. 1986). 45. Moody-Adams, supra n.18 at 127. 46. Charles Taylor, The Ethics of Authenticity 35–38 (Cambridge: Harvard Univ. Press 1992). 47. See Lloyd, supra n.24 at 117. 48. See Walker, supra n.33 at 94–95; Smiley, supra n.24 at 48. 49. See Walker, supra n.33 at 94–95. 50. Cf. Taylor, supra n.46 at 33 (describing the same process for identity formation); Friedman, supra chapter 4 n.27 at 217–18 (describing the same process for autonomy itself). 51. See Smiley, supra n.24 at 249. 52. Meyers, supra n.21 at 152. 53. See Catharine A. MacKinnon, Toward a Feminist Theory of the State 83–96 (Cambridge: Harvard Univ. Press 1989); Elizabeth Schneider, The Dialectic of Rights and Politics: Perspectives from the Women’s Movement, 61 New York University Law Review 589, 602–3 (1986). 54. See Bartlett, supra chapter 4 n.16 at 864–66. 55. See Iris Marion Young, Inclusion and Democracy 18 (New York: Oxford Univ. Press 2000). 56. See id. 57. Id. at 19. 58. Id. at 22. See generally Sunstein, supra chapter 1 n.59. 59. See Carol C. Gould, Rethinking Democracy: Freedom and Social Cooperation in Politics, Economy, and Society 92 (Cambridge: Cambridge Univ. Press 1988). 60. Gould, supra n.59 at 93. 61. See, e.g., Phillip Pettit, Republicanism 8, 81–82 (Oxford: Oxford Univ. Press 1997) (democracy as a means to freedom understood as nondomination; autonomy is a further good not assured by the state). 62. See Isaiah Berlin, Two Concepts of Liberty in Four Essays on Liberty 122 (Oxford: Oxford Univ. Press 1969). 63. See Gould, supra n.59 at 35–36 (self-development as involving all three of these aspects).

276 | Notes to Chapter 6 64. See Young, supra n.55 at 20. 65. It is possible to hold a thoroughgoing utilitarian version of an aggregative theory, but it would lead to a very different sort of democracy from what we have or think we want. In such a model, the fact that an agent had a preference for something would be, at most, good evidence that it would make her happy, but the fact that it was her preference would have no independent weight in the argument. Such a thoroughly utilitarian theory is possible, but the utilitarian versions of democracy theory we normally consider as justifications for our existing or ideal systems are in fact hybrids that include a reliance on autonomy. 66. Gould, supra n.59 at 40. 67. Id. at 84–85. 68. See Young, supra n.55 at 17. 69. Id. at 32. 70. Id. at 33. 71. Hilde Lindemann Nelson, Resistance and Insubordination, 10 Hypatia 23, 27 (No. 2, Spring 1995). 72. Id. 73. Susan J. Brison, The Uses of Narrative in the Aftermath of Violence in On Feminist Ethics and Politics, supra n.24 at 214. 74. See Jennifer Nedelsky, Law, Boundaries, and the Bounded Self, 30 Representations 162, 168 (1990). 75. See Walker, supra n.33 at 118. 76. See Meyers, supra n.30 at 624–26. 77. See Meir Dan-Cohen, Between Selves and Collectivities: Toward a Jurisprudence of Identity, 61 University of Chicago Law Review 1213, 1243 (1994). 78. See Meyers, supra n.21 at 163. Meyers goes on to say that this benefit of intersectional identity supports the claim of standpoint epistemologies that subordinated groups have an epistemic advantage. See id. I have argued against a version of this position in chapter 4. Similar arguments apply in the autonomy context as well. First, because the position of intersectionality is the rule for people in our society rather than the exception, it does not suggest that any particular social group has an epistemic advantage. Second, while it is plain that intersectionality creates the opportunity for greater critical distance, it certainly does not guarantee that the person experiencing an intersectional identity will make use of that opportunity. As I discuss in chapter 4, many other, less useful responses are both possible and common (such as compartmentalization, or generalized alienation). With respect to narrative autonomy, as in the epistemological context, intersectionality should be understood as a resource rather than as the basis for assuming a privileged position. 79. See id. at 157. 80. See id. at 158.

Notes to Chapter 6 | 277 81. See Grimshaw, supra chapter 2 n.33 at 103. 82. Catriona MacKenzie, Imagining Ourselves Otherwise in Relational Autonomy, supra chapter 2 n.37 at 140. 83. See Taylor, supra n.46 at 101. 84. See id. at 102. 85. See Lorraine Code, The Perversion of Autonomy and the Subjection of Women: Discourses of Social Advocacy at Century’s End in Relational Autonomy, supra chapter 2 n.37 at 199. 86. See Homiak, supra n.32 at 52, 59. 87. See Walker, supra n.33 at 106–7. 88. Mackenzie, supra n.82 at 125. 89. See id. at 143. 90. See id. at 144. 91. See id. at 142. 92. These include memory, communication, volition, and self-respect, see Meyers, supra n.21 at 166. 93. Cf. id. 178 n.38 (making this point in terms of identity rather than stories). 94. Benson, supra chapter 2 n.44 at 54. 95. See Annette Baier, Postures of the Mind: Essays on Mind and Morals 83–90 (Minneapolis: Univ. of Minnesota Press 1985) (describing relationality as making us “second persons”). 96. See, e.g., Susan Brison, The Autonomy Defense of Free Speech, 108 Ethics 312, 337–38 (Jan. 1998); Barbara Rowland-Serdar & Peregrine Schwartz-Shea, Empowering Women: Self, Autonomy, and Responsibility, 44 Western Political Quarterly 605, 610–12 (No. 3, Sept. 1991). 97. See Walker, supra n.33 at 123–24. 98. See Flanagan, supra n.16 at 52. 99. See id. at 53–54; see also Kathryn Abrams, Redefining Women’s Agency: A Response to Professor Williams, 72 Indiana Law Journal 459, 461 (1997). 100. Uses of language or other media that do not tell a story but only create an aesthetic or emotional response are still speech and might well contribute to some of the other purposes of freedom of speech (e.g., the search for truth). In addition, such other uses of communication might well be the actions that result from a process of narrative autonomy: that is, through telling one’s own story, one might come to the realization that the creation of such acts of communication is central to acting on one’s own story. The limited—although hardly uncontroversial—claim that I am making here is that uses of media of communication that do not generate symbolic meaning are not themselves a part of the process of narrative autonomy. 101. See Stephanie M. Wildman & Adrienne D. Davis, Making Systems of

278 | Notes to Chapter 6 Privilege Visible in Stephanie M. Wildman, Privilege Revealed 7, 14–15 (New York: New York Univ. Press 1996). 102. Jagger, supra chapter 3 n.13 at 164. 103. Seyla Benhabib, Feminism and Postmodernism: An Uneasy Alliance in Seyla Benhabib, Judith Butler, Drucilla Cornell, & Nancy Fraser, Feminist Contentions: A Philosophical Exchange 17, 20 (introduction by Linda Nicholson) (New York: Routledge 1995). 104. See Drucilla Cornell, Transformations 4, 39 (New York: Routledge 1993); Judith Butler, Gender Trouble 39 (New York: Routledge 1999). 105. The development of chaos theory in the physical sciences is a useful metaphor here. Scientists have not rejected the idea of causal determinism in nature, but they have recognized that this process is so complex that tiny changes in input can have substantial (and unpredictable) changes in output. Moreover, one input that can have such effects is the very act of observation by the scientist. As a result, this porousness may actually characterize even the paradigm case of causal determinism: the physical sciences. The simple dichotomy between causal determinism and randomness, then, is breaking down even here. 106. I am describing moral creativity so I am limiting myself to moral standards. In artistic creativity, the standards might be aesthetic, moral, or some combination of the two. 107. Thus, the liberal autonomy theorists are not wrong in seeing reflection as central to autonomy (although they are mistaken in seeing it as focused on choice and rationality), but they are wrong in trying to find in this reflection an escape from social construction: this is a meaningful experience of autonomy even if we (and it) are socially constructed. 108. See Baker, supra chapter 2 n.67 at 1210. 109. See Linda Nicholson, Introduction in Feminist Contentions, supra n.103 at 1, 10. 110. See Judith Butler, Contingent Foundations: Feminism and the Question of “Postmodernism” in Feminist Contentions, supra n.103 at 35, 42, 47. 111. To the extent that certain poststructuralist views suggest that the constitution of a subject is always an act of violence that works to exclude others, such a position is unacceptable because it short-circuits this normative consideration. See Nancy Fraser, False Antitheses: A Response to Seyla Benhabib and Judith Butler in Feminist Contentions, supra n.103 at 59, 68. 112. See Kenneth Minogue, Ideal Communities and the Problem of Moral Identity, in Nomos XXXV: Democratic Community 41, 47 (John W. Chapman & Ian Shapiro, eds.) (New York: New York Univ. Press 1993). 113. See Jane Mansbridge, Feminism and Democratic Community in Nomos XXXV: Democratic Community, supra n.112 at 339, 365. 114. See Carole Anne Taylor, Positioning Subjects and Objects: Agency, Nar-

Notes to Chapter 7 | 279 ration, Relationality, 8 Hypatia 55, 60 (Winter 1993); Bessie Head, A Question of Power 39 (New York: Pantheon 1973); Patricia Hill Collins, Black Feminist Thought: Knowledge, Consciousness, and the Politics of Empowerment 111 (Boston: Unwin Hyman 1990). 115. See Walker, supra n.33 at 111. 116. See id. at 106–7. 117. Walker, Picking Up Pieces, supra chapter 3 n.48 at 62, 64. 118. See id. 119. See Walker, supra n.33 at 117. 120. See id. at 123. 121. See Smiley, supra n.24. 122. See id. at 7 n.8. 123. Id. at 18. 124. See id. at 91, 126. 125. Id. at 237. 126. I intend to leave open the variety of methods for measuring what counts as positive presently under debate in political theory. See generally, e.g., G. A. Cohen, Equality of What? On Welfare, Goods, and Capabilities in The Quality of Life 9 (Martha C. Nussbaum & Amartya Sen, eds.) (Oxford: Clarendon Press 1993) (discussing the advantages and disadvantages of using utility, primary goods, capabilities, and “midfare” as measures of well-being that should be equally distributed). 127. Cf. Blasi, supra chapter 1 n.26. 128. From the perspective of autonomy theory, the “capabilities” approach improves upon a “primary goods” approach because it looks to the subject’s use of goods rather than merely to the goods themselves as a measure of well-being. See Cohen, supra n.126 at 16; see also Amartya Sen, Capability and Well-Being in The Quality of Life, supra n.126 at 33–40. 129. See Young, supra n.55 at 89–90. 130. See id. at 72–76. notes to chapter 7 1. See Robert Post, Recuperating First Amendment Doctrine, 47 Stanford Law Review 1249, 1253 (1995). 2. Id. at 1272. 3. Professor Post himself is extremely sensitive to these institutional contexts and has envisioned a comprehensive and novel approach to the First Amendment that suggests that the constitutional analysis is and should be very different depending upon whether the speech at issue is part of the “domains” of democracy, community, or management. See Robert C. Post, Constitutional Domains 1–10 (Cambridge: Harvard Univ. Press 1995). I will discuss Post’s

280 | Notes to Chapter 7 approach at some length in this chapter and outline the areas of agreement and disagreement with a relational truth model. 4. This does not, of course, mean that it is either propositional or nonemotional (i.e., rational in Cartesian terms). Conceptual or symbolic representation can be highly emotional and can express complex reactions and attitudes that are not reducible to propositional form. For example, both burning a flag and erotic dancing could qualify as speech within this understanding, given the right circumstances, but lighting a bonfire to keep warm or simply to draw a crowd would not. See Susan H. Williams, Content Discrimination and the First Amendment, 139 University of Pennsylvania Law Review 615, 705–6 (1991) (discussing the distinction between expressive and facilitative aspects of speech). 5. In chapter 5, I left open the possibility that truth might itself be defined in conceptual terms. If one adopts that definition, then the connection between truth and speech is even closer: speech is the manipulation of conceptual and symbolic systems, and truth making is the conceptual or symbolic method of constructing shared reality. In this view, speech is how we make truth. Even if one adopts a broader definition of truth, however, in which nonconceptual forms of truth are possible, it would still be the case that much of our truth-making activity is conceptual, i.e., it takes place through practices we should (and generally do) call speech. 6. See Williams & Williams, supra chapter 3 n.2 at 81–97. 7. See, e.g., Lipkin, supra chapter 5 n.31 at 110. 8. Benhabib, supra chapter 3 n.25 at 31. Iris Young’s conception of the “reasonableness” required of participants in public dialogue is, I think, one attempt to spell out the nature of this respectful relationship in the context of democratic politics. See Young, supra chapter 6 n.55 at 24–25 (describing reasonableness as requiring “a disposition to listen to others, treat them with respect, make an effort to understand them by asking questions, and not judge them too quickly”). In this chapter, I am examining a truth justification not limited to democratic politics and it is possible that the precise requirements of respect might vary somewhat from one epistemological context to another (i.e., scientific laboratories, humanities classrooms, and democratic deliberations). But I think her description is a good example of the way one might fill in the notion of dialogic respect in a particular context. 9. See Iris Marion Young, The Ideal of Community and the Politics of Difference, in Feminism/Postmodernism, supra chapter 2 n.2 at 311. 10. The following arguments closely follow those presented in Lipkin, supra chapter 5 n.31 at 110. 11. See Young, supra chapter 6, n.55 at 39–40. 12. See Chevigny, supra chapter 2 n.12 at 66; Paul G. Stern, A Pluralistic Reading of the First Amendment and Its Relation to Public Discourse, 99 Yale Law Journal 925, 936 n.40 (1990).

Notes to Chapter 7 | 281 13. See Young, supra chapter 6 n.55 at 114–15; Harding, supra chapter 3 n.9 at 155–56; cf. Rorty, supra chapter 5 n.11 at 13 (reinterpret objectivity as intersubjectivity or solidarity); Singer, supra chapter 2 n.28 at 35 (“All objectivity means is agreement among people.”). 14. See Chevigny, supra chapter 2 n.12 at 69. 15. Stern, supra n.12 at 936. 16. See Kathryn Abrams, Hearing the Call of Stories, 79 California Law Review 971, 1002 (1991). 17. See Marc A. Fajer, Can Two Real Men Eat Quiche Together? Storytelling, Gender-Role Stereotypes, and Legal Protection for Lesbians and Gay Men, 46 University of Miami Law Review 511, 524–28 (1992). 18. Post, supra n.3 at 17. 19. See id. at 10. 20. See id. at 5–6. 21. See id. at 16–17. 22. Robert C. Post, Meiklejohn’s Mistake: Individual Autonomy and the Reform of Public Discourse, 64 Colorado Law Review 1109, 1115 (1993). 23. See id. at 1116. 24. Post, supra n.3 at 150. 25. Id. at 188. 26. See id. at 55, 62. 27. See id. at 151. 28. Id. at 188. 29. See id. at 192. 30. See id. at 193. 31. See id. at 142–43. 32. Id. at 177. 33. See id. at 105–8. 34. See id. at 146–47. Sometimes Post’s rhetoric seems to suggest that he thinks the domain of democracy really can be neutral and that the substantive commitments to self-determination and critical thinking represent a rejection of the claims of particular communities rather than being themselves the normative commitments of a particular community. See, e.g., id. at 138 (“The constitutional ‘shield’ established by Cantwell ensures that this competition [between communities] occurs on a level playing field, in which no particular community can obtain an unfair advantage and use the power of the state to prejudge the outcome of this competition by enforcing its own special norms or civility rules. This special neutrality . . .”); 144 (“At heart, therefore, the concept of critical interaction depends upon the continuous possibility of transcending what is taken for granted.”). At other times, he is much clearer about his recognition that these commitments are themselves norms on which particular communities disagree. See, e.g., id. at 146 (“we must understand rational reflection as itself a form of

282 | Notes to Chapter 7 social action that depends for its fulfillment upon a specific normative structure”). This ambiguity may arise from the fact that the book blends the projects of description and prescription; it is difficult to tell at certain points whether Post is simply describing the theoretical structure underlying the case law he so deftly explicates or whether he also means to incorporate that aspect of theory into his own approach. In light of his sophisticated and nuanced view of the self and his frequent recognition of the role of social construction, I have chosen to interpret the book as recognizing that these commitments are themselves the source of substantive limits on public discourse. 35. See Post, supra n.22 at 1117. 36. See generally Madhavi Sunder, Cultural Dissent, 54 Stanford Law Review 495 (2001). 37. Id. at 514. Post acknowledges the unstable and contested nature of communal norms, see Post, supra n.3 at 183, but I believe that his analysis of the difference between the domains of community and democracy does not adequately take account of this phenomenon. 38. See, e.g., Post, supra n.3 at 327–29 (analyzing hate speech on college campuses and suggesting that different models of education are possible, and might even apply to different areas on campus). 39. See, e.g., Jed Rubenfeld, The First Amendment’s Purpose, 53 Stanford Law Review 767, 778 (2001). 40. See Texas v. Johnson, 491 U.S. 397, 414 (1989). 41. See Lee Bollinger, The Tolerant Society 104–44 (New York: Oxford Univ. Press 1986). 42. A relational theory, thus, provides some help with one of the difficulties facing traditional truth theories: the need to explain why truth is so important that it should justify the protection of speech when other social values—such as national security, or the happiness of the general public—seem to require suppression of the speech. If truth is seen as one value among many, then we end up with a balancing approach that asks judges to consider the impact of speech on other values and decide when it is great enough to warrant a sacrifice of truth. If, on the other hand, we wish to establish an absolute right to freedom of expression, then we need to explain why truth is so much more important than these other values that may be harmed by speech. A relational truth theory highlights the fact that truth is not separate from our other values in the way imagined by this dilemma. Truth is the name of a process and many of our values are already part of this process because they do (or should) set the terms on which this activity takes place. In particular, the values of respect, reason, and responsibility provide the parameters within which a speech process must operate in order to produce relational truth. Thus, concerns about the ways in which those values are implicated in speech practices are in no sense external to the promotion of (relational) truth as a goal: speech practices

Notes to Chapter 7 | 283 that violate certain values will not create the types of relationships in which truth can be built and are, therefore, undeserving of protection under a truth theory. In a relational truth theory of speech, then, truth is not balanced against these other values; rather, these values define the scope of the protection for free speech: only relationships characterized by certain values generate the category of speech protected under a relational truth theory. Speech that does not meet these requirements should not receive protection under a relational truth theory (although it might, of course, serve values of autonomy or democracy, which I will discuss in the next chapter). 43. See Mill, supra chapter 1 n.5 at 509–15. 44. See Susan H. Williams, Feminist Jurisprudence and Free Speech Theory, 68 Tulane Law Review 1563, 1579–80 (1994). 45. See Young, supra chapter 6 n.55 at 25 (requirement of “publicity”—“a public in which people hold one another accountable”—as a form of such responsibility in democratic politics). 46. See generally Lenore Walker, The Battered Woman Syndrome (New York: Springer 1984); but cf. Martha Mahoney, Legal Images of Battered Women: Redefining the Issue of Separation, 90 Michigan Law Review 1 (1991). 47. See, for example, Susan Brison’s brave and profound work, Susan J. Brison, Aftermath (Princeton, NJ: Princeton Univ. Press 2002). 48. See Penny Weiss, Feminist Reflections on Community in Feminism and Community 3, 4–5 (Penny A. Weiss & Marilyn Friedman, eds.) (Philadelphia: Temple Univ. Press 1995) (arguing that feminists must examine “what it is in communities—what practices, structures, and values—that have made them or allowed them to be exclusionary, devouring, and violent, and what has made or allowed them to be sustaining, empowering, and respectful of individuality.”). 49. Steven H. Shiffrin, Dissent, Injustice, and the Meanings of America (Princeton, NJ: Princeton Univ. Press 1999). 50. See id. at xi. 51. See id. at 91–93. 52. Cf. Gerald Torres, Translation and Stories, 115 Harvard Law Review 1362, 1382 (2002). 53. Jane Mansbridge, Using Power/Fighting Power: The Polity in Democracy and Difference: Contesting the Boundaries of the Political 46, 59 (Seyla Benhabib, ed.) (Princeton, NJ: Princeton Univ. Press 1996). 54. See Shiffrin, supra n.49 at 97–112. See also C. Edwin Baker, Advertising and a Democratic Press (Princeton, NJ: Princeton Univ. Press 1994). 55. Shiffrin, supra n.49 at 20. He recognizes, however, that much of this obligation may fall on branches of government other than the judiciary. Id.

284 | Notes to Chapter 7 56. See, e.g., Brentwood Academy v. Tennessee Secondary School Athletic Association, 531 U.S. 288 (2001). 57. Post, supra n.3 at 252. 58. Id. at 245. 59. See id. at 245–50. 60. See Connick v. Myers, 461 U.S. 138 (1983); Greer v. Spock, 424 U.S. 828 (1976). 61. But see Post, supra n.3 at 261–63 (arguing that deference is appropriate when institutions are isolated, authority based, and total). 62. See Pickering v. Board of Education, 391 U.S. 563 (1968). 63. But see Connick v. Myers, 461 U.S. 138 (1983). 64. See Int’l Soc. for Krishna Consciousness, Inc. v. Lee, 505 U.S. 672 (1992). 65. But see Greer v. Spock, 424 U.S. 828 (1976). 66. Grayned v. City of Rockford, 408 U.S. 104, 116 (1972). 67. See Corp. of the Presiding Bishop of the Church of Jesus Christ of LatterDay Saints v. Amos, 483 U.S. 327, 334–46 (1987). 68. See Roberts v. United States Jaycees, 468 U.S. 609, 617–20 (1984). 69. See Boy Scouts of America and Monmouth Council, et al. v. Dale, 530 U.S. 640 (2000). 70. See Shiffrin, supra n.49 at 112–13. 71. For example, campaign finance laws, see Buckley v. Valeo, 424 U.S. 1 (1976) (striking down parts of the Federal Election Campaign Act of 1971 that restricted expenditures in support of a candidate) and access claims to the mass media, see Columbia Broadcasting System v. Democratic National Committee, 412 U.S. 94 (1973) (rejecting a claim of access to mass media). 72. See William P. Marshall, In Defense of the Search for Truth as a First Amendment Justification, 30 Georgia Law Review 1 (1995). 73. Id. at 16. 74. See id. at 8–10, 23–26. 75. See id. at 26–27. 76. Id. 77. Id. at 27. 78. See, e.g., id at n.6 (transcendent means “mind-independent, as opposed to subjective.”). 79. See, e.g., Roberts v. Jaycees, 468 U.S. 609, 622 (1984). 80. See Singer, supra chapter 2 n.28 at 29. 81. Cf. David C. Williams, The Militia Movement and Second Amendment Revolution: Conjuring with the People, 81 Cornell Law Review 879 (1996).

Notes to Chapter 8 | 285

notes to chapter 8 1. As I mentioned earlier, Baker’s liberty theory is not subject to this criticism because he offers an argument about how the noncoercive aspect of expressive activity distinguishes it from unprotected exercises of autonomy. See Baker, supra chapter 1 n.27 at 54–60. 2. We use verbal behavior (speech) for many purposes other than narrative autonomy, of course, but generally the speech also serves the purposes of narrative autonomy. I intend to leave open the possibility that, under some rare circumstances, verbal behavior might be so clearly not an instance of narrative autonomy that it would forfeit its status as speech. Perhaps Tourette syndrome, in which a person involuntarily shouts out words, would be an example of verbal behavior that is not an exercise of narrative autonomy. Susan A. Conners, Tourette Syndrome: An Inside Perspective, National Assoc. of School Psychologists Communique 8 (October 2002) (http://www.nasponline .org/publications/cq312tsconnoers.html). The issue in such a case is not whether the speech is coerced, but whether the speech represents an act of meaning making at all—or is simply a different sort of activity that happens to result in speaking a sound, like sneezing. Meaning making, even under conditions of coercion and pressure, is still speech and would still be deserving of protection under the First Amendment. See infra at chapter 8, section A3 (commercial speech discussion). As this example indicates, however, I also mean to suggest that cases in which verbal behavior does not serve the purpose of narrative autonomy should be a rare and anomalous occurrence. 3. See Spence v. Washington, 418 U.S. 405, 410–11 (1974). 4. See, e.g., Jed Rubenfeld, The First Amendment’s Purpose, 53 Stanford Law Review 767 (2001); Robert Post, Recuperating First Amendment Doctrine, 47 Stanford Law Review 1249 (1995). 5. See Rubenfeld, supra n.4 at 773. 6. See id. at 774–75; Post, supra n.4 at 1252. 7. Sometimes the physical effects will themselves be the product of changes in someone’s understanding that results from the speech. For example, hate speech can lead to violence by the listeners. Unlike Baker, I am not suggesting that harm that results from such changes in understanding is an illegitimate basis for regulating speech. See Baker, supra chapter 1 n.27 at 56. Rather, I am arguing that the intent to create an effect on someone’s understanding (even if it is only the speaker’s and whether or not that change in understanding leads to further effects in the world) is one of the requirements for finding an action to be symbolic speech. 8. See United States v. O’Brien, 391 U.S. 367 (1968). 9. See Baker, supra chapter 2 n.65 at 583.

286 | Notes to Chapter 8 10. See Post, supra n.4 at 1253. 11. I am grateful to Ed Baker for raising this issue during a presentation of an early version of this chapter at the Annual Meeting of the Law, Culture, and the Humanities Association. 12. Cf. Akhil Reed Amar, Architexture, 77 Indiana Law Journal 671, 691–96 (2002) (describing the process of interpreting one clause in the Constitution in light of others). 13. Baker, supra chapter 2 n.65 at 585–86. 14. See Jagger, supra chapter 3 n.13 at 164. 15. See Baker, supra chapter 2 n.65 at 591, 592. 16. See Baker, supra chapter 1 n.27 at 196, 202. 17. See id. at 206–7. 18. See generally Sunstein, supra chapter 1 n.59. 19. For an exception—an autonomy-based theory that includes social aspects of autonomy—see Steven Heyman, Righting the Balance: An Inquiry into the Foundations and Limits of Freedom of Expression, 78 Boston University Law Review 1275 (1998). 20. See, e.g., Post, supra chapter 7 n.3 at 268–89; Lillian R. BeVier, Campaign Finance Reform: Specious Arguments, Intractable Dilemmas, 94 Columbia Law Review 1258 (1994). Professor Baker is an interesting exception. He is an autonomy theorist who argues that campaign expenditures might be constitutionally regulated because elections are legally structured institutions designed to generate particular results: choosing good representatives through a process of equal participation. See C. Edwin Baker, Campaign Expenditures and Free Speech, 33 Harvard Civil Rights–Civil Liberties Law Review 1 (1998). 21. See, e.g., Sunstein, supra chapter 1 n.59 at 94–101; Burt Neuborne, Toward a Democracy-Centered Reading of the First Amendment, 93 Northwestern University Law Review 1055, 1070–73 (1999). 22. Buckley v. Valeo, 424 U.S. 1, 48–49 (1976). 23. Neuborne, supra n.21 at 1057. 24. See BeVier, supra n.20 at 1261. 25. It is not the presence of money that makes campaign contributions symbolic speech; obviously, many more traditional forms of speech also involve money (e.g., producing a newspaper). When regulations interfere with money that facilitates speech, the regulations are treated as affecting speech if they are targeted to such speech-related uses of the money (e.g., restrictions on money used for editorializing, FCC v. League of Women Voters, 468 U.S 364 (1984)), but not if the regulation of money is one that affects the speaker more generally, independent of whether or not she is speaking (e.g., general labor or environmental laws as applied to newspapers). The regulation might also trigger First Amendment scrutiny, even if it is aimed at general, nonspeech expenditures,

Notes to Chapter 8 | 287 where the burden falls disproportionately on speakers. See Minneapolis Star & Tribune v. Minnesota Commissioner of Revenue, 460 U.S. 575 (1983). In the case of regulations that restrict the resources for speech but are targeted neither at the facilitation of speech nor at speakers as a class, the Court has refused First Amendment review altogether. See Arcara v. Cloud Books, Inc., 478 U.S. 697 (1986). In campaign contribution regulations, the regulation affects the contribution of money, whether or not that money is used for speech purposes by the candidate. It is targeted neither to money that will be used for speech nor to speakers as opposed to other contributors or recipients. As a result, this is a regulation of speech only if the giving of money itself involves speech, independent of the uses to which it will be put by the recipient. In order for that to be the case, the act of giving the money must itself be symbolic speech. 26. The Supreme Court recognized this distinction itself in the Buckley case. See Buckley, 424 U.S. at 20–21. Liberal models of autonomy do, however, make this distinction harder to maintain. By focusing on choice rather than narrative, such models obscure the difference between interference with symbolic and facilitative aspects of the speech. See Kathleen M. Sullivan, Against Campaign Finance Reform, 1998 Utah Law Review 311, 315–16 (confusing this concern with the “expression v. action” distinction). Choice-based approaches then often fall back on the issue of government purpose as a substitute for a concern with the aspect of speech affected by the regulation. See id. A narrative model of autonomy highlights the importance of this distinction between the facilitative aspects of speech and the symbolic aspects. For expansion and explanation of this distinction, see Williams, supra chapter 7 n.4 at 659–62. 27. See Baker, supra n.20 at 28–33. 28. For example, by refusing to recognize rape within marriage. 29. See Frank I. Michelman, The Constitutional Question, 24 Harvard Journal of Law & Public Policy 17, 20 (2000). 30. See Vincent Blasi, Free Speech and the Widening Gyre of Fund-Raising: Why Campaign Spending Limits May Not Violate the First Amendment after All, 94 Columbia Law Review 1281 (1994). Professor Blasi uses these effects on candidates to argue for a government interest the Court has never considered: the interest in protecting candidates’ time so as to facilitate the fulfillment of their actual duties as representatives. See id. at 1282–83. 31. See Edward B. Foley, Equal-Dollars-Per-Voter: A Constitutional Principle of Campaign Finance, 94 Columbia Law Review 1204, 1226–27 (1994); Neuborne, supra n.21 at 1071. 32. See Foley, supra n.31 at 1220–30. 33. See Neuborne, supra n.21 at 1071–72. 34. See Sullivan, supra n.26 at 322–23; BeVier, supra n.20 at 1267–68. 35. See Cass R. Sunstein, Political Equality and Unintended Consequences, 94 Columbia Law Review 1390 (1994).

288 | Notes to Chapter 8 36. See BeVier, supra n.20 at 1267–68. 37. See id. at 1269. 38. See Sullivan, supra n.26 at 321–23. 39. See, e.g., Neuborne, supra n.21 at 1068. 40. See Meyers, supra chapter 6 n.21 at 158–68. 41. See Young, supra chapter 6 n.55 at 33. 42. This does not mean either that every citizen is guaranteed some minimum amount of political participation, nor that an arbitrary reduction is a constitutional violation. It means only that such an arbitrary reduction is a harm to autonomy created by a system of unlimited campaign financing, just as a restriction on political speech caused by expenditure limits imposes some harm on the exercise of autonomy of the speaker. 43. The analysis up to this point suggests an obvious concern: what happens when the individual does have an autonomy-based claim for the allocation of certain resources, but that claim conflicts with the allocation that the state believes would be better from a systemic point of view? First, the category of claims by individuals to a particular allocation of social resources should be quite narrow. For the most part the state is free to order resources in ways that it thinks best. Second, the only reason recognized here for reordering the resources in a speech system is to improve the opportunity for autonomy, not to promote some other social good. Thus, the conflict we must imagine is one in which the state believes that the general promotion of narrative autonomy through a particular speech system would require the violation of autonomy of an individual speaker. In such a conflict, I would suggest that a court might (1) look carefully at the context to see whether regulation might be possible in some times, places, or manners that would effectively preserve the individual autonomy interest at stake while serving the state’s goals; (2) recognize that if the conflict is inevitable, the protection of either side may leave a “moral remainder” that requires the government to seek out and act on ways of supporting the autonomy interest that was sacrificed in this instance. It may well be that hate speech poses such an intractable problem precisely because it is an example of such a conflict. In that case, we might look to contextual solutions, such as allowing the regulation of hate speech in settings where the harms to listener autonomy are most substantial (such as workplaces and schools). If we choose, however, to protect hate speech in the public forum (as present law does), then we have an obligation to acknowledge and accommodate the moral debt owed to listener autonomy elsewhere in the law. See supra 188–89. 44. See Buckley, 424 U.S. at 25–27. 45. See id. at 48–49, 54, 56–57. 46. See 44 Liquormart, Inc. v. Rhode Island, 517 U.S. 484 (1996). 47. See Brandenburg v. Ohio, 395 U.S. 444 (1969).

Notes to Chapter 8 | 289 48. See Blasi, supra n.30. 49. This distinction is parallel to the one that Professor Schauer draws between the substance and procedure of democratic decision making. See Frederick Schauer, Judicial Review of the Devices of Democracy, 94 Columbia Law Review 1326 (1994). Schauer argues that before we can determine the standards applicable to regulations of the electoral process, we must address the question of whether such decisions should be subject to judicial review or should be left to the democratic political process. See id. at 1330. My suggestion in text relies upon an analogous distinction between outcome and process: when the government is regulating in order to perfect the process of autonomy rather than to generate a particular substantive outcome, it is not acting in an inherently illegitimate way. On the jurisdictional issue that Schauer raises, I fall into the category that he finds underrepresented in the literature: I believe that the constitutionality of campaign finance reform is indeed for the courts to decide, but I think they should find at least some forms of regulation to be constitutional. See id. at 1344–45. 50. See Baker, supra chapter 2 n.67 at 1197; BeVier, supra n.20 at 1266. 51. See Post, supra chapter 7 n.3 at 278. 52. See id. at 1113–17. See also Sullivan, supra n.26 at 324. 53. See C. Edwin Baker, The Media That Citizens Need, 147 University of Pennsylvania Law Review 317, 338 n.53 (1998); Post, supra chapter 7 n.3 at 257. 54. See Sunstein, supra chapter 1 n.59 at 36–37. 55. See BeVier, supra n.20 at 1264. 56. See id. 57. Professor Sunstein, in the argument to which Professor BeVier is responding in the article cited, also does not appear to intend this expansion. See Sunstein, supra chapter 1 n.59 at 56. 58. BeVier, supra n.20 at 1265. 59. Compare Marsh v. Alabama, 326 U.S. 501 (1946) with Hudgens v. NLRB, 424 U.S. 507 (1976). 60. See, e.g., Sullivan, supra n.26 at 319–20. 61. Indeed, even an approach that does not ask the government to do this self-consciously should also offer some assurances against abuse because indirect and unintentional action can have profound effects on these systems and on the opportunity for autonomy. 62. See, e.g., McDuffy v. Sec. of Educ., 615 N.E. 2d 516 (Supreme Judicial Court, Mass. 1993). 63. This, admittedly relatively rare, illegitimate purpose may be the best explanation for the decisions in West Virginia State Board of Education v. Barnette, 319 U.S. 624 (1943), striking down a mandatory flag salute for school children, and in Board of Education v. PICO, 457 U.S. 853 (1982), striking

290 | Notes to Chapter 8 down the removal of books from a school library because of political objections to the ideas in them. 64. In the area of state-subsidized speech, this category is defined somewhat differently by different members of the Court. For Justices Scalia and Thomas, any government regulation that does not work a penalty but merely refuses to subsidize speech is largely immune to First Amendment challenge. See NEA v. Finley, 524 U.S. 569, 596 (Scalia, J., concurring). For Justice Souter, only those regulations that can plausibly be seen as controlling the government’s own speech (or perhaps the government’s purchase of speech for itself) are free of First Amendment review. See id. at 610–11 (Souter, J., dissenting). For most members of the Court, the line is less clear and appears to depend on balancing a variety of factors, including the purposes of the speech system, the competitiveness of it, the degree to which it resembles a public forum, the nature of the government restriction (subject matter v. viewpoint based), and so on. But, at some point, all member of the Court would agree, the government’s action in regulating a speech system enters a category in which the First Amendment is inapplicable. 65. See PICO, supra n.63. 66. See NEA, supra n.64. 67. See Rust v. Sullivan, 500 U.S. 173 (1991). 68. The more difficult question is whether a regulation in a speech system should be seen as unconstitutional if it has the substantial effect of inhibiting the capacities for autonomy, even if that is not its purpose. Because of the difficulties of such practical assessments, I would tentatively suggest that effect should not be an independent basis for invalidating a regulation, unless that effect is so substantial as to (1) call into question the plausibility of the apparent purpose or (2) raise a serious concern that the speech system at issue will become unable to support the capacities for autonomy in any meaningful way. These are intended to be extremely high standards that would rarely be met. 69. See generally Sunstein, supra chapter 1 n.59. 70. Id. at 20. 71. Id. at 21. 72. See, e.g., Kathryn Abrams, Law’s Republicanism, 97 Yale Law Journal 1591, 1599–1601 (1988); Don Herzog, Some Questions for Republicans, 14 Political Theory 473 (1986). 73. See Sunstein, supra chapter 1 n.59 at 137–44. 74. Id. at 130. 75. See id. at 124–37. 76. See Susan Estrich, Real Rape 42–44 (Cambridge: Harvard Univ. Press 1987); Martha Chamallas, Writing about Sexual Harassment: A Guide to the Literature, 4 University of California at Los Angeles Women’s Law Journal 37, 46–47 (1993); Catharine A. MacKinnon, Toward a

Notes to Conclusion | 291 Feminist Theory of the State 195–204 (Cambridge: Harvard Univ. Press 1989). 77. See Meyers, supra chapter 6 n.21 at 157. 78. I am thinking here of political advertisements like the notorious Willy Horton ad. 79. Sunstein clearly recognizes the impact of past inequalities on the nature of present preferences or desires, see Sunstein, supra chapter 1 n.59 at 19–20, but the point I am making here is somewhat different. First, I am suggesting that such inequalities affect not only the content of our political concerns but also our relative abilities to participate in the dialogue through which such concerns are shaped and recognized. Second, I am suggesting that, for that reason, resisting these effects of inequality must be part of the requirements for a well-functioning speech system. 80. Like Sunstein’s requirements, these additional concerns are not necessarily intended to provide a basis for positive First Amendment rights. Rather, they identify legitimate goals for government action in regulating speech systems. When the government seeks to structure a speech system in ways designed to promote these goals, its purpose should be understood as consistent with the demands of the First Amendment. notes to conclusion 1. See Susan J. Hekman, Moral Voices, Moral Selves: Carol Gilligan and Feminist Moral Theory 72 (University Park, PA: Pennsylvania State Univ. Press 1995). 2. See Charles Taylor, The Dialogic Self in The Interpretive Turn: Philosophy, Science, Culture 304, 305 (David R. Hiley, James F. Bohman, & Richard Schusterman, eds.) (Ithaca, NY: Cornell Univ. Press 1991) (“To escape all standards would not be a liberation, but a terrifying lapse into total disorientation. It would be to suffer the ultimate crisis of identity.”) 3. See William James, The Principles of Psychology (1890), reprinted in 53 Great Books of the Western World 318 (Robert Maynard Hutchins, ed.) (Chicago: W. Benton 1952). 4. This description, along with the one in chapter 2, might suggest that these models of reason and objectivity arose during the age of Descartes in the fullblown form we know today. Of course, this is not the case. Our ideas of objectivity, for example, have gone through a process of cultural development that continues in the present. For a very interesting exploration of the historical development of forms of objectivity, see Lorraine Daston, Objectivity and the Escape from Perspective, 22 Social Studies of Science 597 (1992). By attributing the modern version to the Enlightenment, I do not mean to deny this

292 | Notes to Conclusion process of development, but only to assert that it is properly seen as a process with its beginnings in the preoccupations of a particular era and that the later versions are in fact parts of a coherent line of development from the understandings of that time. 5. Introduction in The Interpretive Turn, supra n.2 at 7. The description in text closely follows these authors’ description of “the interpretive turn.” 6. Id. For an interesting challenge to hermeneutic universalism from within a hermeneutic approach, see Richard Schusterman, Beneath Interpretation in The Interpretive Turn, supra n.2 at 109 (describing “uninterpreted realities, experiences, and understandings as already perspectival, prejudiced, and corrigible; in short, as nonfoundationally given.”). Schusterman argues that we should recognize a category of nonfoundational but noninterpretive understanding for the purpose of maintaining the distinction between the conscious and deliberate activity of interpretation and the unconscious, unformulated understandings that form the background of our lives. He recognizes, but rejects, the alternative possibility of describing this background as unconscious interpretation. See id. at 126. Given that his model of this noninterpretive understanding is nonfoundational, selective, and active, see id. at 114–15, I am not sure that it matters to the interpretive approach as I have outlined it whether we call this category interpretation or not. I have chosen to adopt hermeneutic universalism here, however, in order to highlight the continuity rather than the difference between this sort of background and the conscious activity of interpretation. This is not to deny that there may be differences as well and that those differences may be important in certain circumstances. But I think those differences can be captured in other ways (as, for example, by describing different sorts of interpretive activity, along a continuum of consciousness), and for the purposes of my argument, the similarity is more significant. 7. Id. 8. Id. at 8. 9. See id. at 11. 10. Id. at 13. 11. Id. at 9. 12. See Taylor, supra n.2 at 311 (describing dialogic actions such as ballroom dancing and the way they involve a common agent). 13. See Paul A. Roth, Interpretation as Explanation in The Interpretive Turn, supra n.2 at 179, 191 (psychoanalysis involves telling a new story about the analysand in which she is seen as an agent rather than a victim). 14. See Richard Rorty, Inquiry as Recontextualization: An Anti-Dualist Account of Interpretation in The Interpretive Turn, supra n.2 at 62. 15. See supra at 106, 152–53. 16. So, for example, Charles Guignon suggests that rationality is “the ability to adopt multiple partially disengaged perspectives rather than a totally disen-

Notes to Conclusion | 293 gaged stance of pure ‘objectivity.’” See Charles B. Guignon, Pragmatism or Hermeneutics? Epistemology after Foundationalism in The Interpretive Turn, supra n.2 at 100. 17. Id. at 94 (criticizing Rorty). 18. See id. at 100. 19. One interesting issue raised by this approach is whether all meaningful experience is linguistic. In other words, assuming that all experience is interpretive—the hermeneutic universalism discussed above—does all interpretation take place through language? First, we would need to clarify what we mean by language here. Obviously, there are many forms of understanding that take place through nonverbal symbols, such as gestures, visual images, and nonword sounds. If we use language in its broadest sense, as all systems of symbols, then the question becomes whether there are forms of understanding that are totally nonsymbolic, nonconceptual? There are challenging examples of such understanding in the physical experience of learning a dance or a sport and coming to understand the motions in your body, and in the spiritual experiences many people have of losing a clear sense of their own boundaries. The possibility and nature of nonlinguistic interpretation is an interesting line of inquiry within an interpretive approach. 20. See Rorty, supra n.14 at 80 (we are going to find ourselves engaged in the process of recontextualization, that is, interpretation, “whatever happens”). 21. John Dewey, The Influence of Darwin on Philosophy in 4 The Middle Works of John Dewey 14 (Jo Ann Boydston, ed.) (Carbondale, IL: Southern Illinois Univ. Press 1977). 22. Walker, supra chapter 6 n.33 at 213.

Index

Abortion, restrictions on speech about, 224 Abstract values, Nussbaum on, 99 Abstraction: interpretation and, 237; need for certainty, stability, 234 Actions, narratives and, 66 Advertising, free speech and, 207–208 Aeropagitica (Milton), 13–14 Agency: capacity for, 133; creativity and, 160–161, 229; interpersonal recognition self-respect and, 133–134; prerequisites for experiencing it, 142; transformative agency and subjecthood, 161 Agentic recognition self-respect, one’s standing in a community, 133–134 Aggregative model of democracy: autonomy and, 145, 146; narrative autonomy theory of free speech, 170; tenets, 144; utilitarian version, 276n65 Animal rights, truth claims based on, 126 Antisubordination, in judging social practices, 87–88 Argumentation, narratives compared to, 180 Associational rights: of First Amendment, 194; Supreme Court’s understanding, 197

Audience understanding, in Spence test, 201–203 Autonomous man: Benhabib on, 59; Nedelsky on, 59 Autonomy: aggregative model of democracy and, 145, 146; being known and, 63–64; broadcasting and, 199; central focus, 149; character and, 135–137; choice and, 41–42, 46–47, 65–66, 163–164, 200; coercion and, 42, 46–47; collective identity creation, 171; connection to epistemology, 206; consciousness raising and, 144; contra-causal freedom and, 137–138; creativity and, 160–163, 229; cultural dependence of, 150–156; definition, 131, 145; democracy and, 30–31, 143–148, 224–228, 226; determinism and, 42, 158, 254n38; discursive/deliberative model of democracy, 145, 147, 224–228; feminism and, 79–80, 143; forms, 253n32; free speech and, 19–20; free will and, 42, 137–138; freedom and, 145–146, 204–205; functions of, 6, 9, 130–148, 158–172; gender bias and, 205; government influence on preconditions for, 221; happiness and, 146; Homiak on, 138;

295

296 | Index Autonomy (Continued) identity and, 228; individualism and, 259n45; inequality and, 154; integrity and, 64, 130, 135–137, 166; internal restraints and, 63; in interpretive approach, 237; jettisoning the concept, 71; justice and, 147; liberal model of (see Autonomy theory of free speech); living one’s story, 204–206; meanings, 41; narrative model of (see Narrative autonomy); narratives and, 65–66; nature of the self, 76–78; need to tell one’s own story, 155–156; oppression and, 154; parroting culture and, 158–161; politics and, 143–148, 225, 226, 228; practice of, 19–20; as a process, 150, 237–238, 240; promoting via government regulation of free speech, 215–221, 288n43, 289n49; race and, 205; reflection and, 278n107; in relational model of truth, 206; relational nature, 60–66, 69, 150–156, 157, 163; responsibility and, 130–131, 137–143, 167; rights and, 141; school libraries and, 199, 224; selfesteem and, 60–61, 130, 131–135; self-knowledge and, 151–152, 157; “self-made man,” 253n35; self-respect and, 130, 131–135; self-trust and, 60–61, 130, 131–135; selfworth and, 132; social change and, 131, 143–148; social constructionism and, 76–80, 159, 160–162; social nature of, 154, 237–238; social resource requirements, 205, 212, 215, 219, 241, 288n43; socialization and, 41, 42, 61–63, 254n45; speech and, 18, 200, 229, 241–242; speech systems and, 207,

209–224; trust and, 140; victimization and, 78–80; vulnerability and, 240; wealth and, 213; women and, 67–68 Autonomy theory of free speech, 18–24, 41–48, 58–70; Baker and, 46–47; Blasi on, 246n29, 248n52, 249n55; campaign finance reform and, 210–211, 212; challenges to, 51; commercial speech, 22; communitarianism and, 3; concern for the excluded, 69; conscience, 21; criticisms/critiques of, 3–4, 22–24, 30; definition, 246n29, 250n81; democracy theory of free speech and, 2, 27–31; determinate answers to particular cases and, 173; Enlightenment and, 32; epistemology and, 4–5; failure of, 9; feminism and, 3–5, 56–58, 71, 258n30; First Amendment and, 3; foundations, 3; free speech in, 206; free will/determinism dichotomy, 159; gender bias of, 4, 66–68, 70, 130; individualism, 43–44, 47, 63, 65–67, 69, 130, 253n35; instrumental version, 22–23; justification for free speech, 1–3; Kant and, 3, 41, 43; liberalism and, 3, 41–48, 63–66; liberty approach, 46–47; Locke and, 3, 41; Mill and, 21, 41, 245n6; morally suspect, 68, 130; obscenity, 48; paternalism and, 259n45; postmodernism and, 3; process-focused models, 42, 59–60; reason in, 152; Richards on, 45–46, 254n53; right to conscience, 45–46; self-realization and, 18–19, 22–23, 44–45, 250n81; self-reflection, 48; social constructionism and, 47, 255n67; social contract argument and, 18, 20, 23,

Index | 297 44, 59; speech acts, 207; Supreme Court decisions, 21–22, 47–48, 248n49; symbolic speech, 48; tenets, 2; variations, 41, 44–46; vulnerability and, 7–8, 9, 70. See also Narrative autonomy “Bad coherence,” 72 Baker, C. Edwin: on advertising, 207; campaign expenditures, 286n20; liberty approach to autonomy theory of free speech, 46–47, 248n52, 249n55; social constructionism and, 47 Bartlett, Katherine: positionality, 88–90; on tradition, 120 “Battered woman syndrome,” custody and, 79–80 Benhabib, Seyla: on autonomous man, 59; on dissolution of the subject, 158; on respect and speech, 178; on “self-made man,” 54 Benson, Paul, on responsibility, 139 BeVier, Lillian, campaign finance reform and the state action requirement, 218 Bill for Establishing Religious Freedom in Virginia (Jefferson), 15 Blame: narrative autonomy theory of free speech and, 168–169; responsibility and, 139–140 Blasi, Vincent: autonomy theory of free speech, 246n29, 248n52, 249n55; campaign finance reform, 216; on campaign financing, 212 Bordo, Susan: assumptions of Cartesianism, 33; on deconstructionism, 75–76 Boy Scouts, associational claims of, 194 Brandeis, Louis: self-reliance, 26; in Whitney v. California, 16, 26, 38

Brennan, William: commercial speech, 22, 47; New York Times v. Sullivan, 250n68 Broadcasting, autonomy and, 199 Buckley v. Valeo, 211 Campaign finance reform: autonomy theory of free speech and, 210–221, 212; failures of equality, 213; government regulation of free speech and, 215–221; narrative autonomy theory of free speech and, 7, 210–221; paternalism and, 215–217; positive and negative rights and, 219–220; state action requirement and, 218–219; symbolic speech and, 211, 286n25; wealth and campaign financing, 213–214 Cartesianism: alternative to (see Relational model of truth); assumptions, 33–36; conservative bias, 125; control and knowledge, 35–36; correspondence theory of truth, 33, 52–53, 115, 196; criticism and, 239; culture/nature dichotomy, 34–35; deep critique and, 176; dichotomies in, 34–35, 55, 90; empiricism and, 251n5; epistemological individualism, 33–34, 52–53, 232; in epistemology, 32–33; fact/value dichotomy, 33, 38, 105; feminist critique of, 3, 51–58, 68; free speech clause and, 198; free will/determinism dichotomy, 34–35; gender bias of, 51–52, 56–58, 124; knowledge and control, 35–36; knowledge’s foundation, 234–235; language in, 34; middle ground in, 90–92; mind/body dichotomy, 34–35; morally objectionable, 58;

298 | Index Cartesianism (Continued) Nussbaum on, 90–92; objective/subjective dichotomy, 34, 159, 176; playfulness and, 239; political correctness and, 123–126; rationalist bias, 34, 38, 39, 54, 251n4; reality in, 33, 34–35, 37–39, 55, 103; reason/emotion dichotomy, 34; reason in, 34; replicability of results, 256n9; repression and, 125; science and, 102; self (concept of) and, 76–78; separation of knower and known, 55, 56–57, 63–64; shared reality and, 176; social constructionism and, 3–4, 6, 51–58, 75–76, 90; Spence test and, 201; Supreme Court and, 37–40; tenets, 33, 36–37; thoughtcontrol and, 123–126; truth and, 32–37, 176; universal/particular dichotomy, 34; universalism of truth, 37, 38, 54–55, 76 Causal determinism. See Determinism Certainty: autonomy theory of free speech and, 7–8, 9; in Enlightenment, 231, 235; replacement for, 243–244; truth theory of free speech and, 7–8, 9, 17 Chaos theory, determinism and, 278n105 Chaplinsky v. New Hampshire, 38, 183 Character: autonomy and, 135–137; choice and, 142; narrative autonomy theory of free speech and, 165–167; responsibility and, 137, 273n23 Character development, democracy theory of free speech, 25, 29 Choice: autonomy and, 41–42, 46–47, 65–66, 163–164, 200; character and, 142; choosing be-

tween competing value judgments, 82–84, 89–92; creative interpretive narration compared to, 206; determinism and, 42, 254n38; facts and, 53; in liberal model of autonomy, 41–42; in narrative autonomy theory of free speech, 207–208; narratives and, 66, 149–150; oppression and, 164; responsible actor perspective, 263n46; social relationships and, 141–142; socialization and, 61–63 Civility: norms of, 182–183; reason and, 183 Code, Lorraine, on “self-made man,” 53–54 Coercion: autonomy and, 42, 46–47; responsibility and, 142; socialization and, 62–63 Commercial speech: autonomy theory of free speech, 22; First Amendment and, 208; narrative autonomy theory of free speech, 207–209; Supreme Court and, 47–48, 215 Commitment: free speech and, 244; in interpretive approach, 243; security’s replacement, 243–244 Common good, democracy theory of free speech, 29–30 Communitarianism, autonomy theory of free speech and, 3 Community: civility and, 182; deep critique and, 185; democracy and, 183, 185, 281n34; domain of, 181–185, 281n34; norms of, 184–185; shared reality and, 184–185 Conscience: autonomy theory of free speech, 21; right to, 45–46 Consciousness raising, in feminism, 143–144

Index | 299 Contextual frameworks, in relational theory of truth, 118–121 Contra-causal freedom: autonomy and, 137–138; responsibility and, 167 Control, knowledge and, 35–36 Conventionalism: not inevitable, 122; shared reality and, 176–177; social constructionism and, 99 Correspondence theory of truth: Cartesianism and, 33; objectivity and, 33; reality and, 33, 196; social construction of truth, 52–53, 115; tenets, 33 Creativity: experience of, 162; in interpretive approach, 238; narrative autonomy theory of free speech and, 160–163; self-determination and, 161; social constructionism and, 159–160; transformative agency and, 160–161 Critical theory, outlaw emotions and, 73 Cultural construction, relational model of truth, 120–122 Culture: cross-cultural value judgments, 261n19; cultural complexity, 74–75, 83–84; monolithic view of, 74; parroting culture, 158–161; straddling cultures, 83–84 Culture/nature dichotomy, in Cartesianism, 34–35 Decision making: consequences of, 100–103; goal achievement and, 100; shared reality and, 96 Decisional (moral) relativism, defined, 75 Deconstructionism, Bordo on, 75–76 Deep critique: Cartesianism and, 176; community and, 185; context and, 118–119, 185–186; democracy

and, 184, 185; First Amendment and, 186; free speech and, 176–189; government regulation of free speech and, 186–188; Post and, 184; realm of, 184; shared reality and, 118–119, 184–185, 187–188; speech and, 179–180; truth and, 99–100, 118–122 Defamation, Post on, 182 Democracy: aggregative model of, 144–145, 146, 170, 276n65; autonomy and, 30–31, 143–148, 224–228, 226; community and, 183, 185, 281n34; conceptions of, 144–145; consciousness raising and, 144; deep critique and, 184, 185; discursive/deliberative model of, 144–145, 147, 169–170, 171, 224–228; domain of, 181–185, 281n34; marketplace of ideas and, 26–27; Meiklejohn on, 27; narrative autonomy theory of free speech and, 169–170, 224–228; “representative democracy” model of science, 113–114; rhetoric and, 113, 228; shared reality and, 96, 99, 185; speech and, 199, 225–226; tension inherent in, 183–184; value of, 145; Young on (Iris), 144, 147, 171 Democracy theory of free speech, 24–31; applicability, 24–25; autonomy theory of free speech and, 2, 27–31; character development, 25, 29; common good, 29–30; criticisms of, 27–28, 30; derivative nature, 2, 27–31; free speech and, 2; Meiklejohn’s contribution, 24–25; Sunstein’s contribution, 25–26; tenets, 2, 24–26; truth theory of free speech, 2, 27–28, 29, 31

300 | Index Descartes, René, Enlightenment’s view of truth, 2–3 Descriptive relativism, tenets, 75 Desires, first- and second-order, 62–63, 156 Determinism: autonomy and, 158; chaos theory and, 278n105; choice and, 42, 254n38; fear of, 161; feminism and, 4; free will and, 159; interpretive agency and, 8; reason and, 43; responsibility and, 137; socialization and, 42 Dewey, John, on intellectual progress, 243 Dillon, Robin, self-respect, 133–134 Discursive/deliberative model of democracy: autonomy and, 145, 147; Madisonianism and, 224; narrative autonomy and, 169–170, 171; narrative autonomy theory of free speech, 224–228; respect and, 227–228; responsibility and, 227–228; tenets, 144 Dissent: encouragement for, 195; mass media and, 191–192; within military, 193–194; within religious organizations, 194; Shiffrin on, 191, 195 Dissent, Injustice, and the Meanings of America (Shiffrin), 191–192 Due process, liberty interests and, 205 Editorials, free speech and, 185 Educational opportunities, denial of, 205 Emerson, Thomas, free speech and autonomy, 19–20 Emotions: intentional infliction of emotional distress, 182; knowledge and, 54; outlaw emotions, 72–73; reason and, 106 Empiricism, Cartesianism and, 251n5

Enlightenment: absence of authority, 234; alternative to its vision (see Interpretive approach); autonomy theory of free speech, 32; breakdown of personal identity, 232–234; breakdown of reality, 234–235; breakdown of social order, 235; certainty to, 231, 235; chaos to, 231–235; Descartes and, 2–3; fear in shaping self, 234; global nature of threats in, 240; goals of, 244; liberation sought by, 231, 232; mistake made by, 241; rational will in, 233; security to, 231, 235, 240; truth theory of free speech, 2–3, 32; vision of, 231–236, 243 Epistemology: autonomy theory of free speech and, 4–5; connection to relational model of truth, 116–117; feminism and, 57–58, 69–70; mainstream tradition, 32; moral/political issues/theory and, 8, 113–115, 116–117, 123, 157–158, 206, 267n7; self-knowledge and, 157; social relationships and, 231; standpoint epistemologies, 82–86, 276n78; truth theory of free speech, 2–3, 4–5, 32; “working” truth claims and, 101, 108, 122–126. See also Cartesianism Equal protection clause: hate speech and, 188, 288n43; liberty interests and, 205; property tax–based funding of public schools, 222–223 Equality: campaign finance and failures of, 213; equalizing opportunities for political participation, 216 Erotic dancing, as speech, 280n4 Ethics: truth and, 91; “working” truth claims and, 101–102 Evaluative self-respect, Dillon on, 134

Index | 301 Fact/value dichotomy: in Cartesianism, 33, 38, 105; Supreme Court, 38 Facts, value judgments and, 53–54 False consciousness, 73 Feinberg, Joel: meanings of autonomy, 41; on socialization, 61–62 Feminism: autonomy and, 79–80, 143; autonomy theory of free speech and, 3–5, 58–68, 71, 258n30; categories of, 256n4; common themes, 51; concern for the excluded, 68–69, 113–116, 156–157; consciousness raising in, 143–144; critique of Cartesianism, 3, 51–58, 68; critique of liberal model of autonomy, 58–68; determinism and, 4; epistemology and, 57–58, 69–70; Habermas in, 90; hooks on, 78–79; individualism and, 69; interpretive agency and, 78; jettisoning truth, autonomy, 71; law and, 8–9, 89–90; relational model of truth and, 116, 197; relativism and, 4, 75; self (concept of) and, 76–78; social constructionism and, 3–4, 51, 56–58, 71–92; standpoint epistemologies and, 82–86; strategies for reconceiving social constructionism, 82–92; tradition and, 89, 120; truth and, 197; truth theory of free speech and, 3–5, 51–58, 68, 71; value conflicts and, 74–75, 261n16; victimization and, 78–80; women of color and, 78–79; women’s experiences, 72–74, 120, 143–144, 156 First Amendment: advertising and, 208; associational rights of, 194; autonomy theory of free speech and, 3; bedrock principle, 186; ca-

pacities contributing to narrative autonomy, 222–224; commercial speech, 208; as cultural icon, 1; deep critique and, 186; democracy theory and, 226; doctrinal development, 13; free exercise clause, 268n22; free speech clause, 196, 198, 204–205; hate speech and, 188–189, 288n43; interpretive approach and, 241–242; liberty interests and, 205; moral remainders from compromises, 188, 288n43; narrative autonomy theory of free speech and, 7, 199–228; nonspeech systems and, 222–223; politically subversive speech, 271n48; promotion of “working” truth claims within private institutions, 192–195; reasons for protecting free speech, 181, 202–203, 210; relational model of truth and, 187; relational theory of speech and, 7, 175–197; religion and, 128, 268n22; religion clauses, 196, 204; social practice systems and, 181–182; symbolic speech and, 203; theoretical frameworks for, 226; transcendent truths, 104; truth theory of free speech, 15–16, 18; “two-tier” version, 226 Fiss, Owen, Post and, 184 Flag burning, as speech, 201, 252n25, 280n4 Flanagan, Owen, soldiers suffering identity crisis, 135, 165 Fourteenth Amendment: doctrine of substantive due process, 204; equal protection clause, 188, 205, 223; narrative autonomy theory of free speech and, 204 Frankfurter, Felix, truth theory of free speech, 15

302 | Index Free speech, 175–198, 199–229; advertising and, 207–208; autonomy and, 19–20; autonomy theories of, 242; in autonomy theory of free speech, 206; capacities promoting, 45, 221–224; central practices of, 244; commitment and, 244; deep critique and, 176–189; editorials and, 185; foundations, 1–3, 13–31, 173 (see also Autonomy theory of free speech; Democracy theory of free speech; Narrative autonomy; Truth theory of free speech); functions of, 176, 179, 180, 181, 190, 195, 205; interdependence and, 244; Jefferson and, 21; Madison and, 21; marketplace of ideas, 15, 17, 26–27, 38–39, 115; narrative autonomy theory of free speech and, 199–229; “negative” theory of, 245n1; politically subversive speech, 215; protection of, 179, 180, 181, 199, 210, 216; relational truth theory of speech and, 175–198; self-government and, 26; shared reality and, 176–189; social constructionism and, 125; speakers and, 176; state action requirement, 218–220; systems and social relationships, 176, 180–181, 209–224; trespassing and, 220; truth theories of, 242; at the workplace, 185. See also Speech Free will: autonomy and, 42, 137–138; determinism and, 159; in liberal model of autonomy, 42; responsibility and, 167 Free will/determinism dichotomy, in Cartesianism, 159 Freedom: autonomy and, 145–146, 204–205; definition, 145–146; Gould on, 147; negative freedom,

204; positive freedom, 204; social resource requirements, 205 Freedom of speech. See Free speech Friedman, Marilyn, on criticisms of autonomy, 59 Gender: cultural frameworks and, 118; definition, 56 Gender bias: autonomy and, 205; in autonomy theory of free speech, 4, 66–68, 70, 130; in Cartesianism, 51–52, 56–58, 124; concept of the self and, 77; in truth theory of free speech, 4, 51–52, 70 Gender equality: commitment to, 70; contextual frames for, 119–120 Gilligan, Carol, moral reasoning in men and women, 67 Gould, Carol, on freedom, 147 Government: influence on preconditions for autonomy, 221; management domain and, 192–194 Government regulation of free speech: abortion-related speech, 224; abuse distinguished from having an effect, 221; campaign finance reform and, 215–221; deep critique and, 186–188; government’s disagreement with speech content, 186; politically subversive speech, 215; promoting autonomy via, 215–221, 288n43, 289n49; purpose in seeking, 187–188; reasons for disallowing, 2, 7, 186; reductions to capacities necessary for narrative autonomy, 223–224; shared reality and, 186–188; state action requirement, 218 Government subsidized speech, narrative autonomy theory of free speech and, 7, 224, 290n64 Grayned v. City of Rockford, 194

Index | 303 Habermas, Jurgen, feminists and, 90 Hand, Learned, on truth, 39 Happiness, autonomy and, 146 Harding, Sandra: objectivity, 84–86; standpoint epistemology, 84–86 Hate speech: consequences, 285n7; intractability as a problem, 288n43; protection for, 188–189 Hermeneutics: hermeneutic contextualism, 236; hermeneutic universalism, 236, 292n6, 293n19; politics and, 171 Holmes, Oliver Wendell: marketplace of ideas, 15, 26–27, 38–39; truth theory of free speech, 15–16 Homiak, Marcia, on autonomy and responsibility, 138 hooks, bell, on women of color and feminism, 78–79 Identity: agency and, 142; autonomy and, 228; breakdown of, 135, 165, 232–234; collective identity creation, 171; “intersectional” identities, 151, 276n78; knowledge and, 158; Meyers on, 151; politics and, 171, 225, 228; responsibility and, 137, 141; social relationships and, 190; soldiers suffering identity crisis, 135, 165; Taylor on, 141; vulnerability and, 189–190; Williams on, 86–87 Imagination, in narrative autonomy theory of free speech, 153 Independent moral standard, as strategy for reconceiving social constructionism, 88–92 Individualism: autonomy and, 259n45; in autonomy theory of free speech, 43–44, 47, 63, 65–67, 69, 130, 253n35; in Cartesianism, 33–34, 52–53, 232; dilemmas re-

sulting from, 242–243; in Enlightenment vision, 235; feminist view of, 69; need for certainty, stability, 234; “self-made man” as example, 253n35 Inequality: autonomy and, 154; public dialogue and, 227–228, 291n79; wealth and, 222 Injustice: knowledge and, 191; Mansbridge on, 191; pressure toward, 195; “working” truth claims and, 191–192 Integrity: autonomy and, 64, 130, 135–137, 166; definition, 136; narrative autonomy theory of free speech and, 165–167; reliability and, 166; responsibility and, 142; Walker on, 64, 166–167 Intellectual progress, Dewey on, 243 Intent: meaning and, 202; symbolic speech and speaker’s intent, 285n7 Interdependence: certainty’s replacement, 243–244; free speech and, 244; vulnerability and, 243 Interpersonal recognition self-respect, agency and, 133–134 Interpersonal skills, narrative autonomy theory of free speech and, 153–154 Interpretation: abstraction and, 237; choice compared to, 206–207; in critique of Cartesianism, 53; interpreting defined, 239; in interpretive approach, 237–238; in narrative autonomy, 149, 166; in relational truth, 105; value-neutrality and, 238 Interpretive agency: central concepts, 8; determinism and, 8; feminism and, 78; narrative autonomy, 8, 160–164; relational model of truth, 5, 8, 105, 116–117; relativism and, 8

304 | Index Interpretive approach, 231–244; autonomy in, 237; commitment in, 243; contextualism in, 237; creativity in, 238; First Amendment and, 241–242; interdependence and, 243; interpretation in, 237–238; language in, 239, 293n19; liberation sought by, 240, 243; morality in, 236–237, 238; narrative autonomy theory of free speech and, 236; personhood in, 238; politics in, 236–237; rationality in, 238–239; relational model of truth and, 236, 241; responsibility in, 238; security and, 240; theory in, 236; truth in, 237, 241; uncertainty and, 243; vulnerability and, 240–241, 242–243 Intersubjectivity, knowledge and, 98–99, 266n2, 268n8 Jagger, Alison: assumptions of Cartesianism, 33; outlaw emotions, 72–73 James, William, buzzing, blooming confusion, 234 Jefferson, Thomas: Bill for Establishing Religious Freedom in Virginia, 15; free speech, 21; truth theory of free speech, 15 Judgment, social construction and, 71–76 Justice, autonomy and, 147 Justification, social construction and, 71–76 Kant, Immanuel: autonomy theory of free speech, 3, 41, 43; reason in, 43, 233 King, Martin Luther, Jr., and Selma march, 121

Knowledge: consideration of alternatives and, 179–180; context sensitivity, 116, 118–122, 270n45; control and, 35–36; emotion in, 54; foundation for, 234–235; identity and, 158; injustice and, 191; interpretation in, 116–117; intersubjectivity and, 98–99, 266n2, 268n8; moral/political challenges to, 108–109; narratives and, 149; objectivity and, 235; personal responsibility for, 54, 116–117; a political process, 113–115, 122; positionality’s view, 88; power and, 35–36; purpose, 111; reason as basis for, 54; in relational model of truth, 122; relational nature, 116; separation of knower and known, 55, 56–57, 63–64; shared reality and, 97, 267n7; situatedness of, 116; social constructionism’s view, 51–58; social foundations of knowledge, 98–99; truth and, 97–99; value judgments and, 47, 53–54; weaving metaphor of, 109–113. See also Self-knowledge Language: in Cartesianism, 34; experience and, 293n19; in interpretive approach, 239, 293n19; narrative autonomy and, 155; in relational model of truth, 106–107; representation and, 239; shared reality and, 96; truth and, 36, 39; “victim talk,” 79 Law: determinate answers to particular cases, 173; feminism and, 8–9, 89–90; narratives in, 120; rights in, 141 Letter concerning Toleration (Locke), 20–21

Index | 305 Liberal model of autonomy, 41–48, 58–68; Baker and, 46–47; choice in, 41–42; feminist critique of, 58–68; free will in, 42; personhood in, 43, 47, 63–64; rationalism/rationality and, 47, 48; reason in, 43, 45; the self in, 43–44; social constructionism and, 61–63; socialization in, 42–43, 61–63; speech in, 44–47; substantive vs. procedural models, 58–59; Supreme Court Decisions, 47–48 Liberalism, autonomy theory of free speech and, 3, 41–48, 63–66 Liberation: Enlightenment’s concern, 231, 232; interpretive approach’s concern, 240, 243 Locke, John: autonomy theory of free speech, 3, 41; Letter concerning Toleration, 20–21 Lord, Audre, paradox of criticism, 72 Macrolevel speech: connection to nonhuman reality, 197; defined and applied, 181 Madison, James: free speech, 21; Memorial and Remonstrance against Religious Assessments, 15, 21; Richards on, 21; self-government and free speech, 26; truth theory of free speech, 15 Madisonianism, Sunstein’s, 224 Management, domain of, 181, 192–194 Mansbridge, Jane, on injustice, 191 Marketplace of ideas: democracy theory and, 26–27; Holmes and, 15; marketplace failures, 17; truth and, 15, 38–39, 115; universalism of truth, 38–39

Marshall, Thurgood: commercial speech, 47; in Stanley v. Georgia, 21–22 Marshall, William, on search for truth, 196 Marxism, social constructionism and, 56 Mass media: dissent and, 191–192; shared reality and, 181 Meaning: creation of (see Interpretation); losing it, 240; social constructionism and, 206; speaker’s intent and, 202; Spence test and, 201–202; threats to, 235 Meiklejohn, Alexander: on abandoning democracy, 27; democracy theory of free speech, 24–25 Memorial and Remonstrance against Religious Assessments (Madison), 15, 21 Metaphors for relationship between concepts and reality, 107–113 Meyers, Diana T., on intersectional identity, 151 Microlevel speech: defined, 180; vulnerability and, 189 Military, “working” truth claims in, 193–194 Milkovitch v. Lorain Journal Co., 16, 39 Mill, John Stuart: autonomy theory of free speech, 21, 41, 245n6; On Liberty, 14–15, 37; reality to, 37; on received opinion, 188; Supreme Court citing of, 16, 37; truth theory of free speech, 14–15, 37 Milton, John: Aeropagitica, 13–14; defense of free speech, 13–14 Mind/body dichotomy, in Cartesianism, 34–35 Minow, Martha, on “victim talk,” 79

306 | Index Moral (decisional) relativism, 75 Moral reasoning, in men and women, 67 Moral remainders, First Amendment compromises and, 188, 288n43 Moral responsibility. See Responsibility Moral Responsibility and the Boundaries of Community (Smiley), 167 Morality: epistemology and, 267n7; in interpretive approach, 236–237, 238; self and, 233; social relationships and, 231 Mystics, prescription for, 233 Narrative autonomy, 148–172, 199–229; aggregative model of democracy and, 170; autonomy in, 150–156, 165; blame and, 168–169; blending moral and epistemological concerns, 157–158; capacities contributing to narrative autonomy, 152–154, 157, 162, 167, 221–224; central activity, 200; central focus, 149; character and, 165–167; choice in, 149–150, 157, 163–164, 200; concern for the excluded, 156–157; creativity and, 160–163, 229; democracy and, 169–170, 224–228; determinate answers to particular cases and, 173, 199; discursive/deliberative model of democracy and, 169–170, 171; exercising rational capabilities, 152–153; feminist themes and, 156–158; First Amendment and, 7, 199–228; foundations, 6; free speech and, 199–229; government regulation of speech, 7; government subsidized speech, 7, 224, 290n64; hallmark of autonomous speech, 208; imagination in, 153;

integrity and, 165–167; interpersonal skills and, 153–154; interpretation in, 149, 166; interpretive agency and, 8, 160–164; interpretive approach and, 236; language and, 155; need to tell one’s own story, 155–156; parroting culture, 158–161; politics and, 171, 213; praise and, 168–169; rationalism and, 163; reason in, 152; responsibility and, 167–169; self-direction and, 159–165, 168; self-esteem and, 164–165; self in, 151; selfknowledge and, 151–152, 156, 157; self-respect and, 164–165; self-trust and, 164–165; social change and, 169–170; socially constructed selves, 158–161; symbolic speech and, 7, 200–204; tenets, 172; time in, 149; underlying view, 236; verbal behavior and, 285n2; vulnerability and, 169. See also Autonomy theory of free speech Narrative autonomy theory of free speech: advertising, 208; campaign finance reform, 7, 210–221; choice and social pressure, 207–208; commercial speech, 207–209; discursive/deliberative model of democracy, 224–228; distinction between negative and positive rights, 220; Fourteenth Amendment, 204; government subsidized speech, 7, 224, 290n4; intent and, 202; living one’s story, 204–206; neutrality and, 217; personhood in, 151–152; political compared to other speech, 226–227; political participation and, 225, 226–227; school libraries, 199, 224; social relationships and, 199; speech acts, 207; speech systems and, 207, 209–224;

Index | 307 Spence test, 201–203; symbolic speech, 200; wealth and, 222–223 Narrative model of autonomy. See Free speech; Narrative autonomy Narratives: analogies/metaphors in, 152; argumentation compared to, 180; choice and, 149–150; communication media for telling stories, 155, 277n100; consideration of alternatives and, 180; knowledge and, 149; in law, 120; living one’s story, 204–206; narrative process, 148–149; need to tell one’s own story, 155–156; politics and, 170–171; in relational truth theory of free speech, 180; Young on, 171 Nedelsky, Jennifer: on autonomous man, 59; on “self-made man,” 54 “Negative” theory of free speech, basis of, 245n1 Neutrality, narrative autonomy theory of free speech and, 217 New York Times v. Sullivan, 16, 26, 250n68 Nonhuman reality, connection to, 103–104; macrolevel speech and, 197; postmodernism and, 104; relational model of truth, 126–128, 195–197; relational truth theory of free speech, 195–197; social constructionism, 126–128; spirituality and, 103–104; truth and, 103–104, 126–128 Nonspeech systems, First Amendment and, 222–223 Nussbaum, Martha: on appeals to abstract values, 99; on Cartesianism, 90–92; common humanity, 266n96 Objective/subjective dichotomy: in Cartesianism, 34, 159, 176; truth and, 176

Objectivity: correspondence theory of truth and, 33; dilemmas resulting from, 242–243; in Enlightenment vision, 235; historical development of the concept, 291n4; knowledge and, 235; truth theory of free speech, 52–53, 246n1 Obscenity, autonomy theory of free speech, 48 On Liberty (Mill), 14–15, 37 Oppression: autonomy and, 154; choice and, 164; consciousness raising and, 144; diversity among the oppressed, 83–84; epistemic advantages of the oppressed, 83, 276n78; of women, 79, 82–83 Otherness, ineradicability of, 81 Outlaw emotions, defined, 72–73 Parroting culture, autonomy and, 158–161 Pateman, Carol, on social contract argument, 59 Paternalism: autonomy theory of free speech and, 259n45; Supreme Court, 215–216, 223 Patriarchy: bastions of, 70; monolithic view of culture, 74; “working” truth claims and, 101 Personal recognition self-respect, defined, 133–134 Personal responsibility. See Responsibility Personality, civility and, 182 Personhood: in interpretive approach, 238; liberal model of autonomy and, 43, 47, 63–64; in narrative autonomy theory of free speech, 151–152; respect for, 271n46; selfrespect and, 132; “working” truth claims and, 123 Perspective versus standpoint, 85

308 | Index Playfulness, 81; Cartesianism and, 239 Politeness, respect compared to, 178–179 Political correctness, relational model of truth and, 123–126 Politics: autonomy and, 143–148, 226, 228; concept of the self and, 77; consciousness raising and, 143–144; epistemology and, 113–115, 116–117; equalization of political influence as a goal, 215; equalizing opportunities for political participation, 216; hermeneutics and, 171; identity and, 171, 225, 228; in interpretive approach, 236–237; knowledge and, 113–115, 122; narrative and, 170–171; narrative autonomy and, 171, 213; political participation and narrative autonomy theory of free speech, 225, 226–227; political speech compared to other speech, 226–227; politically subversive speech, 215, 271n48; shared reality and, 96; Supreme Court and corruption in, 213; in “two-tier” First Amendment, 226; wealth and, 213–214 Pornography, credibility issue and, 227 Positionality, as strategy for reconceiving social constructionism, 88–90 Post, Robert: Chaplinsky v. New Hampshire to, 183; deep critique and, 184; on domains of social order, 181–185, 192–194, 279n3, 281n34; First Amendment protection, 202–203; on nature of speech, 175; on privacy-based torts, 182; relational truth theory

of free speech and, 182; Shiffrin and, 192; Sunstein and, 184 Postmodernism: autonomy theory of free speech and, 3; connection to nonhuman reality and, 104; social constructionism and, 56, 77, 98; truth theory of free speech, 3 Poststructuralism, social constructionism and, 56 Power: concept of the self and, 77; directionality of, 77; knowledge and, 35–36 Pragmatism: functionality standard, 87; social constructionism and, 56; as strategy for reconceiving social constructionism, 86–88; tenets, 86; truth theory of free speech, 3; “working” truth claims in, 101 Praise: narrative autonomy and, 168–169; responsibility and, 139–140 Prayer, protected speech, 196 Pride, Rich on, 80 Privacy-based torts, Post on, 182 Propositions, truth and, 37, 55 Public forum doctrine, in relational truth theory of free speech, 180 Public schools, property tax-based funding for, 222–223 Race, autonomy and, 205 Racism: inequality and, 154; social hierarchies and, 132–133 Radin, Margaret, “bad coherence,” 72 Rape, credibility issue and, 227 Rationalism/rationality: Cartesianism, 34, 38, 39, 54, 238–239, 251n4; consideration of alternatives and, 179–180; desire to be rational, 153; Enlightenment view, 238–239; exercising rational capa-

Index | 309 bilities, 152–153; in interpretive approach, 238–239; liberal model of autonomy and, 47, 48; meaning of, 179; means/end rationality, 152; narrative autonomy and, 163; need for certainty, stability, 234; rationality defined, 239, 292n16; relational model of truth and, 105–106. See also Reason Rationality, speech and, 179–180 Rawls, John, 20, 46 Reality: breakdown of, 234–235; in Cartesianism, 33, 34–35, 37–39, 55; Cartesianism and, 103; conceptual frameworks, 106–112; connection to nonhuman reality, 103–104, 126–128; constraints imposed by, 110–111, 269n27; correspondence theory of truth, 33, 52–53, 115, 196; experience and, 106–108; external reality, 112–113, 126–128; as finished cloth, 109; metaphors for relationship between concepts and reality, 107–113; in Mill, 37; physical versus social, 110–111; reality building, 267n4; in relational model of truth, 106–108, 112; shared reality, 118–122, 267n4, 267n7, 270n45, 272n52; unreality compared to, 103; “working” truth claims and, 102–103, 108, 122–126 Reason: in autonomy theory of free speech, 152; as basis for knowledge, 54; in Cartesianism, 34; civility and, 183; context sensitivity, 106; desire to be rational, 153; determinism and, 43; emotion and, 106; in Enlightenment vision, 235; faith in, 234; in Kant, 43, 233; liberal model of autonomy and, 43,

45; “making sense,” 106, 152, 239; means/end rationality, 152; in narrative autonomy theory of free speech, 152; reasoning by analogy, 152; relational model of truth and, 64–65, 180, 282n42; respect and, 183; social resource requirements, 153; truth and, 36, 37–38; understandings of, 64–65. See also Rationalism/rationality Reason/emotion dichotomy, 34; in Cartesianism, 34 Red Lion Broadcasting v. FCC, 16 Redish, Martin: relationship between democracy and autonomy, 30–31; self-realization, 45 Reflection, autonomy and, 278n107 Relational model of truth, 105–129, 175–198; alternative to (see Cartesianism); autonomy in, 206; concern for the excluded, 113–116; connection to epistemology, 116–117; connection to nonhuman reality, 126–128, 195–197; contextual frameworks in, 118–121; cultural constructionism and, 120–122; deep critique in, 118–122; determinate answers to particular cases and, 173; dichotomies and, 105; dilemma of conventionalism vs. relativism, 118–122; feminism and, 116, 197; First Amendment and, 187; foundations, 5–6; interpretive agency and, 5, 8, 105, 116–117; interpretive approach and, 236, 241; knowers in, 179; knowledge in, 122; language in, 106–107; metaphors for relationship between concepts and reality, 107–113; morality, 116–117; political correctness and, 123–126;

310 | Index Relational model of truth (Continued) rationalism and, 105–106; reality in, 106–108, 112; reason and, 64–65, 180, 282n42; relationship between knower and known, 116; responsibility and, 5–6, 139; science and, 113–114; shared reality in, 118–122; social constructionism and, 5, 105–113; tenets, 105, 129; thought-control and, 123–126; truth in, 206; truth’s importance, 282n42; underlying view, 236; universalism and, 105; weaving metaphor of knowledge production, 109–113; “working” truth claims and, 122–126, 189–195 Relational truth theory of free speech: connection to nonhuman reality, 195–197; government disagreement with speech, 186–188; macrolevel of speech, 181; managerial institutions, 192–194; microlevel of speech, 180; narrative in, 180; Post and, 182; public forum doctrine, 180; public vs. private institutions, 191–192, 194–195; religious speech, 195–197; respect and, 177–179, 182, 183, 280n8, 282n42; responsibility and, 189–190, 282n42; shared reality in, 176–177; vulnerability and, 179–180, 189–190; “working” truth claims and, 189–195 Relational truth theory of speech: First Amendment and, 7, 175–197; free speech and, 175–198 Relationships. See Social relationships Relativism: descriptive relativism, 75; feminism and, 4, 75; inevitability and, 122; interpretive agency and,

8; moral (decisional) relativism, 75; shared reality and, 99, 176–177; social constructionism and, 75, 87; types, 75 Reliability, integrity and, 166 Religion: dissent within, 194; First Amendment and, 128, 268n22; religion clauses of the First Amendment, 196; secularism and, 104; shared reality and, 272n72; social constructionism and, 127–128 Religious speech, in relational truth theory of free speech, 195–197 Representation, language and, 239 Repression, social constructionism and, 125–126 Respect: civility and, 182; discursive/deliberative model of democracy and, 227–228; politeness compared to, 178–179; reason and, 183; speech and, 177–179, 280n8, 282n42. See also Self-respect Responsibility: ascriptions of, 139–142, 167–169; autonomy and, 130–131, 137–143, 167; Benson on, 139; character and, 137, 273n23; coercion and, 142; determinism and, 137; discursive/deliberative model of democracy and, 227–228; epistemology and, 8; free will and, 167; Homiak on, 138; identity and, 137, 141; integrity and, 142; internalization of, 142; in interpretive approach, 238; for knowledge, 54, 116–117; membership in a community, 140; narrative autonomy and, 167–169; relational model of truth and, 5–6, 139; relational truth theory of free speech and, 189–190, 282n42; rights and, 141; self-esteem and,

Index | 311 142; self-trust and, 142; Smiley on, 167; trust and, 140, 169; White on, 141; “working” truth claims and, 189–190 Rich, Adrienne, on pride, 80 Richards, David A. J.: autonomy theory of free speech, 45–46, 254n53; on Madison, 21 Rights: autonomy and, 141; distinction between negative and positive rights, 219–220; responsibility and, 141 Rorty, Richard: need for transcendence, 104; social constructionism and, 99 Safety, social relationships and, 190 Sandel, Michael, self-knowledge, 63 Scalia, Antonin, government subsidized speech, 290n64 Schauer, Frederick: democratic decision making, 289n49; on restricting speech, 27; on truth theory of free speech, 245n2 School libraries, autonomy and, 199, 224 Science: Cartesianism and, 102; determinism in, 278n105; moral/political challenges to, 108; relational model of truth and, 113–114; “representative democracy” model, 113–114; value judgments in, 53, 54 Secularism, choosing, 104 Security: in Enlightenment, 231, 235, 240; interpretive approach and, 240; replacement for, 243–244 Self (concept of): autonomy and, 76–78; disintegration of, 98–99; fear in shaping it, 234; liberal model of autonomy, 43–44; morality and, 233; narrative autonomy

and, 151; social construction of, 158–161 Self-definition, self-worth and, 79 Self-determination: creativity and, 161; public discourse and, 183–184; social constructionism and, 161 Self-direction: narrative autonomy theory of free speech and, 159–165, 168; social constructionism and, 159–164 Self-esteem: autonomy and, 60–61, 130, 131–135; definition, 132; narrative autonomy and, 164–165; responsibility and, 142; self-respect and, 132; social conditions and, 132–133 Self-knowledge: autonomy and, 151–152, 157; epistemology and, 157; narrative autonomy and, 151–152, 156, 157; social process, 63–64 “Self-made man”: atomistic individualism of, 253n35; Benhabib on, 54; Code on, 53–54; Nedelsky on, 54 Self-realization, autonomy theory of free speech, 18–19, 22–23, 44–45, 250n81 Self-reflection, autonomy theory of free speech, 48 Self-respect: agentic recognition selfrespect, 133–134; autonomy and, 130, 131–135; definition, 132; entailments, 133; evaluative self-respect, 134; forms, 133–134; interpersonal recognition self-respect, 133–134; narrative autonomy and, 164–165; personal recognition selfrespect, 133–134; personhood and, 132; self-esteem and, 132 Self-trust: autonomy and, 60–61, 130, 131–135; definition, 132; narrative

312 | Index Self-trust (Continued) autonomy and, 164–165; responsibility and, 142; social conditions and, 132–133; vulnerability and, 132 Self-worth: autonomy and, 132; forms of, 132; self-definition and, 79 Selma to Montgomery march, television coverage of, 121 Sexism: inequality and, 154; monolithic view of culture, 74; social hierarchies and, 132–133 Sexual harassment, credibility issue and, 227 Shared reality: Cartesianism and, 176; community and, 184–185; context and, 185–186; conventionalism and, 176–177; decision making and, 96; deep critique and, 118–119, 184–185, 187–188; democracy and, 96, 99; extent of, 270n45; free speech and, 176–189; goal of, 177; government regulation of free speech and, 186–188; knowledge, 97; knowledge and, 267n7; language and, 96; mass media and, 181; politics, 96; in relational model of truth, 118–122; relational truth theory of free speech and, 176–177; relativism and, 99, 176–177; religion and, 272n52; social constructionism and, 99; social cooperation and, 96–97; speech and, 177; trust and, 97; truth and, 96–99, 181–182, 267n4, 280n5 Shiffrin, Steven: on dissent, 191, 195; Dissent, Injustice, and the Meanings of America, 191–192; Post and, 192

Sit-ins, as speech, 203 Situatedness: knowledge and, 116; social constructionism and, 86 Smiley, Marion, Moral Responsibility and the Boundaries of Community, 167 Social change: autonomy and, 131, 143–148; narrative autonomy and, 169–170 Social constructionism: autonomy and, 76–80, 159, 160–162; autonomy theory of free speech and, 47, 255n67; Baker and, 47; Cartesianism and, 3–4, 6, 51–58, 75–76, 90; concern for the excluded, 156–157; connection to nonhuman reality, 126–128; conventionalism and, 99; correspondence theory of truth and, 52–53, 115; creativity and, 159–160; experience in, 73–74; external reality and, 112–113; feminism and, 3–4, 51, 56–58, 71–92; feminist strategies for reconceiving it, 82–92; freedom of speech and, 125; implications, 53–54; internal consistency, 268n25; judgment and, 71–76; justification and, 71–76; knowledge in, 51–58; liberal model of autonomy and, 61–63; Marxism and, 56; meaning and, 206; monolithic view of culture, 74–75; otherness and, 81; playfulness, 81; postmodernism and, 56, 77, 98; poststructuralism and, 56; pragmatism and, 56; process-focused models of autonomy, 59–60; rate of cultural change, 121–122; relational model of truth, 5, 105–113; relativism and, 75, 87; religion and, 127–128; repression and, 125–126; schools of thought based

Index | 313 on, 56; self (concept of) and, 76–78; self-determination and, 161; self-direction and, 159–164; shared reality and, 99; situatedness and, 86; social foundations of knowledge, 98–99; standpoint epistemologies and, 82–86; state action requirement and, 220; symbolic systems and, 159; tenets, 58; transcendent truths, 126–128; truth and, 71–76; truth theory of free speech, 3–4; value judgments and, 126 Social contract argument: autonomy theory of free speech, 18, 20, 23, 44, 59; Pateman on, 59; patriarchal social order, 59 Social cooperation, shared reality and, 96–97 Social relationships: choice and, 141–142; epistemology and, 231; identity and, 190; morality and, 231; narrative autonomy theory of free speech and, 199; safety and, 190 Socialization: autonomy and, 41, 42, 254n45; choice and, 61–63; coercion and, 62–63; consciousness raising and, 144; determinism and, 42; Feinberg on, 61–62; liberal model of autonomy and, 42–43, 61–63 Souter, David, government subsidized speech, 290n64 Space, in autonomy theory of free speech, 149 Speech: autonomy and, 18, 200, 229, 241–242; as communication of messages, 175; conduct considered speech, 248n49; consideration of alternatives and, 179–180; cultural construction and, 121; deep cri-

tique and, 179–180; democracy and, 199, 225–226; distinguishing characteristic, 177; effort to understand, 206; government regulation of, 7, 23–24; institutional control over, 193–195; liberal model of autonomy, 44–47; macrolevel, 181, 197; microlevel, 180, 189; moral commitment in, 177–179; nonpropositional forms, 39; nonverbal forms, 177, 200–204, 205, 280n4, 293n19; political participation and, 225; political speech compared to other speech, 226–227; politically subversive speech, 271n48; Post on its nature, 175; rationality and, 179–180; reason and, 180; respect and, 177–179, 280n8; restricting, 27; shared reality and, 177; as social practices, 175; specialness, 2, 7, 200; speech systems and, 207, 209–224; suppression of, 14; theories of (see Autonomy theory of free speech; Narrative autonomy theory of free speech; Relational truth theory of speech); trust and, 177; truth and, 175, 177, 197, 241, 280n5; vulnerability and, 190. See also Commercial speech; Free speech; Hate speech; Symbolic speech Speech systems: examples of, 222; narrative autonomy theory of free speech and, 207, 209–224 Spence v. Washington, 201–203 Spirituality, connection to nonhuman reality, 103–104 Standpoint epistemologies: antisubordination and, 87–88; feminism and, 82–86; intersectional identity and, 276n78 Stanley v. Georgia, 21–22

314 | Index State action requirement: Bill of Rights and, 218–220; distinction between negative and positive rights and, 219 Stories. See Narratives Subjecthood, transformative agency and, 161 Sunstein, Cass: democracy theory of free speech, 25–26, 226; discursive/deliberative model of democracy, 224–227; Madisonianism, 224; Post and, 184; state action requirement, 217–218; “two-tier” first amendment, 226 Supreme Court: associational claims of the Boy Scouts, 194; autonomy theory of free speech, 21–22, 47–48, 248n49; Cartesianism and, 37–40; commercial speech, 47–48, 215; corruption in politics, 213; equalization of political influence as a goal, 215; fact/value dichotomy, 38; flag burning, 252n25, 280n4; free exercise clause of the First Amendment, 268n22; liberal model of autonomy, 47–48; Mill cited by, 21, 37; paternalism argument, 215–216, 223; skepticism, 40, 253n30; symbolic speech, 201, 252n25; truth theory of free speech, 15–16, 37–40; understanding of associational rights, 197 Symbolic speech: autonomy theory of free speech, 48; campaign contributions, 286n25; campaign contributions as, 211; facilitative aspects of speech and, 287n26; First Amendment and, 203; narrative autonomy, 7, 200–204; narrative autonomy theory of free speech and, 200; sit-ins, 203; speaker’s in-

tent and, 285n7; Supreme Court, 39, 252n25 Symbolic systems, porousness, 159, 161 Taylor, Charles, on identity, 141 Theory, in interpretive approach, 236 Thomas, Clarence, 22; government subsidized speech, 290n64 Thomas, Richard, longing for transcendence, 104 Thought-control, 123–126 Time: in narrative autonomy theory of free speech, 149 Time v. Hill, 16 Tradition: Bartlett on, 120; feminism and, 89, 120 Transcendence: First Amendment and, 104; need for, 104; transcendent truths, 126–128 Transformative agency: subjecthood and, 161 Trespassing, free speech and, 220 Trust: autonomy and, 140; responsibility and, 140, 169; shared reality and, 97; speech and, 177 Truth: ability to recognize, 17, 28, 38; accepting, 38; Cartesianism and, 32–37, 176; connection to nonhuman reality, 103–104, 126–128; consequences of decisions, 100–103; consideration of alternatives and, 241; context sensitivity, 106; correspondence theory, 33, 52–53, 115, 196; deep critique and, 99–100, 118–122; effort to understand, 206; ethics and, 91; feminism and, 197; functions of, 5, 9, 95–104, 118–128, 195; Hand on, 39; importance of, 282n42; interpretation in, 105; in interpretive approach, 237, 241; jettisoning the

Index | 315 concept, 71; knowledge and, 97–99; language and, 36, 39; marketplace of ideas and, 15, 38–39, 115; moral relationships and, 98; objective/subjective dichotomy and, 176; as a process, 237–238, 240; propositions and, 37, 55; reason and, 37–38; relational model of (see Relational model of truth); in relational model of truth, 206; search for understanding, 196; shared reality and, 96–99, 181–182, 267n4, 280n5; social constructionism and, 71–76; social nature of, 237–238; social resource requirements, 241; speech and, 175, 177, 197, 241, 280n5; universalism of, 34, 37, 38, 54–55, 76, 105; value of, 16–17; vulnerability and, 240; “working” truth claims (see “Working” truth claims) Truth theory of free speech, 13–18, 32–40, 51–58; ability to recognize truth, 17, 28, 38; accepting truth, 38; alternative to (see Relational truth theory of free speech); applicability, 17; autonomy theory of free speech, 7–8, 232; certainty and, 7–8, 9, 17; challenges to, 51; controversial issues, 16–18; critiques of, 3–4; democracy theory of free speech, 2, 27–28, 29, 31; determinate answers to particular cases and, 173; Enlightenment and, 2–3, 32; epistemology and, 2–3, 4–5, 32; failure of, 9; feminism and, 3–5, 51–58, 68, 71; First Amendment and, 15–16, 18; foundations, 2–3; gender bias of, 4, 51–52, 70; Holmes’s defense, 15; Jefferson’s use, 15; justification for

free speech, 1–3; Madison’s use, 15; Mill and, 14–15, 37; Milton’s contribution, 13–14; objectivity and, 52–53, 246n1; pedigree, 13; postmodernism and, 3; pragmatism and, 3; Schauer on, 245n2; social constructionism and, 3–4; Supreme Court decisions, 15–16; tenets, 1–2, 13; truth’s value, 16–17; vulnerability and, 7–8, 9, 70 Truths, transcendent truths, 126–128 Turette syndrome, verbal behavior and, 285n2 Uncertainty: interpretive approach and, 243; vulnerability and, 243 United States, cultural complexity, 74–75, 83–84 Universal/particular dichotomy, 34; in Cartesianism, 34 Universalism of truth: in Cartesianism, 34, 37, 38, 54–55; in Enlightenment vision, 235; hermeneutic universalism, 236, 292n6, 293n19; marketplace of ideas and, 38–39; in relational model of truth, 105 Value conflicts, feminism and, 74–75, 261n16 Value judgments: choosing between competing ones, 82–84, 89–92; context sensitivity of, 90; cross-cultural judgments, 261n19; facts and, 53–54; knowledge and, 47, 53–54; problem selection and, 53; science and, 53, 54; social constructionism and, 126 Verbal behavior, narrative autonomy theory of free speech and, 285n2 Victimization: autonomy and, 78–80; “battered woman syndrome,” 79–80; feminism and, 78–80;

316 | Index Victimization (Continued) powerlessness and, 80; “victim talk,” 79; Wendell on, 80 Vulnerability: autonomy and, 240; autonomy theory of free speech and, 7–8, 9, 70; elimination of, 234, 235–236; flight from, 243; forms of, 189–190; identity and, 189–190; interdependence and, 243; interpretive approach and, 240–241, 242–243; microlevel speech, 189; narrative autonomy and, 169; need to embrace, 70; relational truth theory of free speech and, 179–180, 189–190; self-trust and, 132; speech and, 190; truth and, 240; truth theory of free speech, 7–8, 9, 70; uncertainty and, 243; “working” truth claims and, 189 Walker, Margaret Urban, on integrity, 64, 166–167 Warren, Earl, conduct considered speech, 248n49 Wealth: campaign financing and, 213–214; inequality and, 222; narrative autonomy theory of free speech and, 222–223 Weaving metaphor of knowledge production: external reality and, 109–113 Welfare, measures of, 170 Wendell, Susan, on victimization, 80 White, Byron, commercial speech, 22, 47

White, Stephen, on responsibility, 141 Whitney v. California, 16, 26, 38 Williams, Joan, on identity, 86–87 Women: autonomy of, 67–68; oppression of, 79, 82–83; self-definition of, 79; victimization and, 78–80 Women of color: feminism and, 78–79; hooks on, 78–79 “Working” truth claims, 100–103, 122–126, 189–195; epistemology and, 101, 108, 122–126; ethics and, 101–102, 126; First Amendment and, 192–195; Grayned v. City of Rockford and, 194; injustice and, 191–192; institutional control of speech, 193–195; in military, 193–194; openness to challenges, 191; patriarchy and, 101; personhood and, 123; pragmatism and, 101; in private institutions, 192–195; reality and, 102–103, 108, 122–126; relational model of truth and, 122–126, 189–195; relational truth theory of free speech and, 189–195; responsibility and, 189–190; truth and, 100–103, 108, 122–126; vulnerability and, 189 Workplace, free speech at, 185 Young, Iris Marion: on democracy, 144, 147, 171; reasonableness in public dialogue, 280n5 Young, Robert, and socialization, 62–63

About the Author

Susan H. Williams is the Walter W. Foskett Professor of Law at Indiana University School of Law–Bloomington. She is the author of numerous articles on topics in First Amendment law, feminist jurisprudence, and legal theory. She lives in Bloomington, Indiana, with her husband, David, and their two children.

317